EXPERT REPORT OF FRANK SILLMAN€¦ · 18/10/2013 · EXPERT REPORT OF FRANK SILLMAN -2- ......

UNITED STATES BANKRUPTCY COURT SOUTHERN DISTRICT OF NEW YORK

In re: RESIDENTIAL CAPITAL, LLC, et al., Debtors.

) ) ) ) ) ) )

Case No. 12-12020 (MG) Chapter 11 Jointly Administered

EXPERT REPORT OF FRANK SILLMAN

-2-

I am an expert in the origination and securitization of mortgage loans. I have particular

experience with loan repurchases, loss reserves for mortgage loans, loan underwriting standards,

and the re-underwriting of loans underlying mortgage-backed securities.

I have been asked by the Debtors to offer expert testimony in connection with

confirmation of the proposed Chapter 11 plan for Residential Capital, LLC, et al. [Docket No.

4153].

SUMMARY OF OPINIONS

1. I have formed the following opinions concerning the proposed RMBS

Settlement:

Aggregate Lifetime Losses. The aggregate lifetime losses that will potentially be incurred by trusts included in the RMBS Settlement range from $42.1 billion to $42.9 billion.

Underwriting Defect Rates. My analysis of loans underwritten between 2004 and 2007, including a forensic re-underwriting of 1500 loans from that period, showed an average material underwriting defect rate of 43.5 percent. For loans issued before 2004, my analysis showed an average material underwriting defect rate of 35.1 percent.

Likely Damages After Legal Defenses and Litigation Costs. The likely amount of recoverable damages, after consideration of legal defenses and litigation costs, ranges from $7.38 billion to $8.6 billion.

2. I have formed the following opinions concerning the proposed Ambac

Settlement:

Aggregate Lifetime Losses. The aggregate lifetime shortfalls that will potentially be experienced by the Debtor-sponsored bonds insured by Ambac range from $225.6 million to $237.8 million.

Likely Damages. Ambac seeks rescissory damages or the equivalent thereof which, if awarded, would allow Ambac to

-3-

recover all its losses. Ambac could also recover additional amounts for interest, fees, and claims regarding the servicing of trusts not sponsored by the Debtors. The likely amount of recoverable damages with respect to the Debtor-sponsored trusts ranges from $180.5 million to $237.8 million.

3. I have formed the following opinions concerning the proposed Assured

Settlement:

Aggregate Lifetime Losses. The aggregate lifetime shortfalls that will potentially be experienced by the bonds insured by Assured range from $73.6 million to $77.6 million.

Likely Damages. Assured seeks rescissory damages or the equivalent thereof, which, if awarded, would allow Assured to recover all its losses. Assured could also recover additional amounts for interest and fees, and for certain servicing claims. The likely range of recoverable damages with respect to the Debtor-sponsored trusts (excluding certain servicing-related claims that I have not analyzed) ranges from $58.9 million to $77.6 million.

4. I have formed the following opinions concerning the proposed FGIC

Settlement:

Aggregate Lifetime Losses. The aggregate lifetime shortfalls that will potentially be experienced by the bonds initially insured by FGIC range from $1.64 billion to $1.71 billion.

Likely Damages. FGIC seeks rescissory damages or the equivalent thereof, which, if awarded, would allow FGIC to recover all its losses. FGIC could also recover additional amounts for interest and fees. The likely range of recoverable damages with respect to the Debtor-sponsored trusts ranges from $1.31 billion to $1.7 billon.

5. I have formed the following opinions concerning the proposed MBIA

Settlement:

Aggregate Lifetime Losses. The aggregate lifetime shortfalls that will potentially be experienced by the bonds insured by MBIA range from $2.19 billion to $2.22 billion.

-4-

Likely Damages. MBIA seeks rescissory damages or the equivalent thereof, which, if awarded, would allow MBIA to recover all its losses. MBIA could also recover additional amounts for interest, fees, and claims regarding trusts not sponsored by the Debtors. The likely amount of recoverable damages with respect to the Debtor-sponsored trusts ranges from $1.75 billion to $2.22 billion.

BACKGROUND AND QUALIFICATIONS

6. I serve as Managing Partner for Fortace, LLC, an advisory and consulting

firm to banks, mortgage companies, insurance companies, trustees and other investors. My

curriculum vitae is attached as Annex A.

7. A key area of my work with Fortace relates to reviewing and opining on

the reasonableness of demands for the repurchase of mortgage loans based on alleged breaches

of representations and warranties. I have performed repurchase demand work for insurers and

lenders who have issued repurchase demands to sellers based on alleged breaches of

representations and warranties. As part of this work, I have developed loan audit selection

criteria, reviewed contractual obligations, performed loan-level audits, made recommendations

as to whether or not a repurchase demand should be issued, and participated in negotiations with

sellers on whether to repurchase loans. I have also performed work for sellers who have

received repurchase demands from trustees, insurers, and lenders for alleged breaches of

representations and warranties. As part of this work, I have reviewed contractual obligations,

reviewed repurchase demands and related findings and supporting evidence, performed loan

level audits, made recommendations to sellers as to whether representations or warranties were

breached, and participated in negotiations on whether to repurchase loans.

8. I have approximately 25 years of experience in the mortgage banking

industry. I have held senior executive positions at a federally insured bank, at a Wall Street

-5-

investment bank, and at privately-held mortgage banking companies. I have managed residential

mortgage origination and loan operations, secondary marketing of mortgage loans, capital

market activities, treasury functions, and warehouse lending related to mortgage loan origination.

In particular, I have extensive experience in the residential mortgage market, including

origination, securitization, loss reserves, and repurchase-related activities related to Fannie Mae,

Freddie Mac, FHA, Prime Jumbo, Alt-A, subprime, home equity lines of credit, and closed-end

second lien residential mortgage loans.

DOCUMENTS, PUBLICATIONS, PRIOR TESTIMONY, EXHIBITS, AND COMPENSATION

9. Attached as Annex B is a list of the documents I considered in forming my

opinions. I have not authored any publications in the last ten years, and I have not testified as an

expert at trial or by deposition in the last four years other than in these proceedings. Attached as

Annexes C through I are the exhibits which summarize or support my opinions.

10. Fortace’s compensation is set forth in the Debtors’ Application for an

Order Authorizing Employment and Retention of Fortace LLC as Consultant to the Debtors

Nunc Pro Tunc to May 21, 2012, filed July 6, 2012 [Dkt. 704].

BASIS OF MY OPINIONS

I. BACKGROUND

11. The Plan seeks to resolve claims asserted by the trustees of RMBS trusts

sponsored and/or serviced by the Debtors, as well as claims asserted by four monoline insurance

companies that insured certain of the bonds. The RMBS Settlement, unlike the previous RMBS

settlement agreement, does not include bonds for which a monoline insurer provided insurance.

Insured bonds are the subject of separate settlement agreements. Accordingly, I have formed

-6-

separate opinions regarding the RMBS Settlement and monoline insurer settlements, which are

addressed in separate sections of this report.

A. Overview of the Debtors’ Mortgage Securitization Process.

12. The Debtors sponsored RMBS trusts covering a wide range of different

mortgage products. The Debtors also sold mortgage loans to third parties for inclusion in RMBS

sponsored by those third parties. In both situations, the loans are transferred to the RMBS trusts

through agreements in which the Debtors provided certain representations and warranties. The

representations and warranties may vary from RMBS trust to RMBS trust, but typically include,

among others, representations concerning accuracy of certain loan-level disclosures, that the

loans substantially complied with the Debtors’ underwriting guidelines, and that there was

proper documentation of the loans.

13. The RMBS trusts then issued bonds1 that are sold to investors or retained

by the Debtors. The structure of the trusts provides credit support that allows for payment to

some or all of the bonds even if some of the mortgage loans default or are liquidated with

collateral losses. There are a number of types of credit support, including overcollateralization

(placing more collateral in the trust than the face value of the bonds), excess spread (paying out

lower interest rates to bond holders than is received from the mortgagees), and subordination

(separating the bonds into tranches with different payment priorities so that subordinate bonds

absorb losses and provide support to the senior bonds).

14. As a way to further enhance the credit support of the bonds, a sponsor may

choose to obtain bond insurance for some of the bonds in a trust from a monoline bond insurance

company. The insurance, or “wrap,” guarantees full, complete and timely principal and interest 1 In this Report, the term bonds refers to all interests in the RMBS trusts, including any residual interests.

-7-

payments to the holders of wrapped bonds issued by the RMBS trust. For this guarantee, the

monoline charges a premium or fee. The presence of the bond wrap is an added third-party

guarantee to the wrapped bond holders in addition to the underlying credit structure of the

RMBS trust, which reduces the overall risk to the wrapped bond holders. The Debtors utilized

bond wraps on 142 of the 581 Debtor-sponsored trusts.

B. The Debtors’ Alleged Breaches of Representations and Warranties.

15. If a seller breaches its representations and warranties to the trust, typically

by providing loans that deviate from the representations, the seller may be required to repurchase

the defective loan(s).

16. Generally, the standard for analyzing a breach of representations and

warranties requires an assessment of: (a) whether the alleged loan defect or alleged breach is an

actual and material breach of representations and warranties, and (b) whether such breach was

material and adverse to the interests of the bond holders (cumulatively the “R&W Repurchase

Standard”). If the R&W Repurchase Standard is met, the seller is required to repurchase non-

liquidated loans at the purchase price, or to reimburse the RMBS trust for any losses incurred in

the liquidation of the loan, as defined in the applicable agreements. If the R&W Repurchase

Standard is not met, the seller does not have an obligation to repurchase the loan or reimburse the

RMBS trust for liquidated losses.

17. When a trust is subject to a bond wrap at issuance, sellers generally also

provide representations and warranties to the monoline issuing the wrap and provide that certain

of the representations and warranties made to the trusts inure to the benefit of the monoline.

-8-

18. The applicable agreements also provide remedies to the monoline that

include, if certain conditions are met, reimbursement from the seller for any payments made by

the monoline arising out of a breach of a representation and warranty.

II. OPINIONS REGARDING THE RMBS SETTLEMENT.

19. The RMBS Settlement covers all claims belonging to trusts sponsored by

the Debtors between 1999 and 2007.2 The 392 trusts sponsored between 2004 and 2007 were the

subject of a previous RMBS settlement agreement. The RMBS Settlement also releases all

claims belonging to 185 trusts sponsored by the Debtors between 1999 and 2004 (for a total of

581 trusts) that are the subject of this Report. 3

20. In addition, the RMBS Settlement covers all claims belonging to trusts

that were not sponsored by the Debtors but for which a Debtor either sold and/or serviced some

portion of the underlying mortgage loans. Of these, 160 trusts have representation and warranty

claims that are also included in my analysis.4

21. The RMBS Settlement does not provide for recovery to the trusts with

respect to any bonds insured by the monolines, so those wrapped bonds are not included in my

analysis of the RMBS Settlement.

2 The 741 trusts analyzed here are based on a list of trusts provided by the Debtors prior to the finalization of the Plan schedules filed on October 11, 2013. I am in the process of ascertaining and reconciling any discrepancies.

3 I understand that the trustees’ claims with respect to FGIC wrapped trusts were released as part of the FGIC settlement (other than claims with respect to any non-wrapped bonds), so the release of the trustees’ claims for the 47 FGIC wrapped trusts is not included in the valuation of the RMBS Settlement, leaving a total of 534 Debtor-sponsored trusts for the RMBS Settlement analysis.

4 I also understand that the RMBS Settlement provides for a $96 million payment in settlement of recognized cure claims based on allegations of improper loan servicing by the Debtors. I have not been asked to form any opinions regarding the settlement of those servicing claims, or certain allowed claim amounts in settlement of other servicing claims.

-9-

A. The Debtor-Sponsored Trusts’ Estimated Aggregate Lifetime Losses.

22. There are a variety of methods accepted in the financial services industry

to estimate RMBS lifetime losses. I am familiar and have experience with these methods. To

form the opinions stated in this report, I employed a trust-level methodology based on frequency

and severity rates. Frequency and severity rate-based loss forecasting and historically-based

assumption development are two of the accepted methods for deriving an estimate of potential

repurchase exposure. These two methodologies are regularly used by market participants,

financial institutions and experts to estimate repurchase exposures, including estimates provided

by financial institutions in their regulatory filings, and independent third-party expert reports.

Accordingly, the methodologies I used in this report are generally accepted in the industry as

sound means of estimating lifetime losses.

Model Process Overview

Remittance Data

Preliminary Assumptions Development

Process

AssumptionsValidation

Process

Illumination & Intex Loss

Models

Outputs

Industry RMBS

Remittance Data

Trusts Remittance

Data

User Input

Assumptions

Model Calculated

Assumptions

Lower Range

Forecasted Remaining Lifetime Losses

Settlement Trusts

Assumptions Validated

against actual Trusts

performance

Forecasted Remaining Lifetime

Loss Model Process Higher

Range Forecasted Remaining Lifetime Losses

-10-

23. Step 1 - The first step in developing estimated loss ranges for RMBS

trusts is to obtain the historical borrower loan payment remittance data (“Remit Data”) for both

(1) the trusts, and (2) other industry RMBS trusts which consist of loan products and

securitization structures similar to the trusts.5 This Remit Data contains hundreds of data fields

including loan level payment histories, prepayment data, default data and loan level losses. The

Remit Data may be available on either a loan level basis or at a trust level basis. For the 581

Debtor-sponsored trusts, we were able to obtain loan level data from Loan Performance6 (“LP”)

and Intex.7 We utilized Remit Data from September 2013.

24. Step 2a – I employed Illumination (formerly WestPat LLC) to run their

proprietary RMBS estimated loss and cash flow model (the “Illumination Model”) to determine

estimated lifetime loss ranges for the trusts and tranches of wrapped bonds within the trusts for

which loan level Remit Data was available. The Illumination Model requires loan level Remit

Data. The Illumination Model is a commercially available estimated loss and cash flow model

used by mortgage lenders, mortgage bond investors and money managers to estimate loan losses

and cash flows and to value RMBS mortgage bonds.

25. Step 2b – For the trusts for which only trust level Remit Data was

available, I utilized the Intex Model, as defined below, to determine estimated lifetime loss

ranges for both the trusts and the tranches of wrapped bonds within those trusts. The Intex

Model is a commercially available cash flow model used by mortgage lenders, mortgage bond

5 The Illumination Model groups the RMBS trusts into the following categories: Alt A/Sub Prime, Prime, HELOC & Fixed 2nds.

6 CoreLogic Loan Performance is a provider of RMBS loan remittance data.

7 Intex Solutions, Inc. is a provider of structured fixed income cash flow models and RMBS remittance data.

-11-

investors and money managers to estimate loan losses, cash flows and value RMBS mortgage

bonds (“Intex Model”).

26. Step 3 – Illumination and Intex Model assumption requirements and

discussion:

(a) Illumination Model assumptions:

(i) The Illumination Model independently develops its Validated Trust Assumptions for forecasting cash flows and estimated losses from actual historical performance of certain key data elements (“HIST PERF”) from the Remit Data for each of the trusts:

(a) Actual Trust Losses to date.

(b) Actual Severity Rates to date.

(c) Actual Constant Default Rates to date (“CDR”) aka Roll Rates or Frequency Rates.

(d) Actual Voluntary Constant Prepayment Rates (“VCPR”).

(e) Actual Loan Level Payment Histories to date (“PAY HIST”) aka Pay Strings.

(ii) Additionally, I provided a few macroeconomic assumptions to Illumination for use in the Illumination Model based on industry available data and my expert experience in developing these assumptions:

(a) Forward Yield Curve from September 30, 2013.

(b) The unemployment rate8 utilized was 8.1% from August 2013. The unemployment rate was held constant for the life of the loans.

(c) The current Combined Loan To Value (“CLTV”) was calculated using Case-Shiller9 home price data

8 U.S. Bureau of Labor Statistics.

9 S&P/Case-Shiller Home Price Index is a leading measure of the U.S. residential housing market.

-12-

as of July 2013. The model uses the zip code when available. If the zip code is not available, the model uses Metropolitan Statistical Area (“MSA”) level or state level data. Once the CLTV is updated, it is varied over time based on our Forward Home Price Index assumptions described below.

(d) FICO scores - The model does not update Borrowers’ FICO scores, the model utilizes the Borrowers’ origination FICO scores.

(e) LP and Intex Remit Data reflect the Trusts’ actual Losses to Date after applying any mortgage insurance claims paid to the Trusts. The LP and Intex Remit Data do not include any Monoline insurance claims paid to the Trustee for the benefit of the CertificateHolders.

(f) Forward Home Price Index (“HPI”) for distressed home sales.

(g) The Illumination Model varies time to foreclosure by state. The Illumination Model utilized time to foreclosure history through March 2012.

(b) Intex Model assumptions:

(i) The Intex Model requires the user to develop and input assumptions into the model. I provided assumptions for use in the Intex Model based on industry available data and my expert experience in developing these assumptions:

(a) Forward Yield Curve from September 30, 2013.

(b) VCPR – determined after reviewing each individual Settlement Trusts’ 6 month, 12 month and monthly time series trends.

(c) CDR - determined after reviewing each individual Settlement Trusts’ 6 month, 12 month and monthly time series trends.

(d) Severity Rates - determined after reviewing each individual Settlement Trusts’ monthly time series Severity trends.

-13-

27. Step 4a – The Illumination Model evaluates RMBS trust historical Remit

Data for loan products and securitization structures similar to the trusts from the available

industry Remit Data from LP or Intex (“Industry RMBS Remit Data”) to develop the Preliminary

Industry RMBS Loss Assumptions utilized to estimate the remaining lifetime losses for these

Industry RMBS Trusts.

28. Step 4b - The Illumination Model then performs a series of regression

analyses to validate the Preliminary Industry RMBS Assumptions against the actual performance

of these Industry RMBS Trusts to create the validated assumptions for the Industry RMBS Trusts

(“Validated Industry RMBS Trust Assumptions”).

Validating the Industry RMBS Assumptions

4a. The Illumination Model evaluates the Industry Remit Data to develop Preliminary Industry Loss Assumptions utilized to forecast remaining lifetime losses for Industry RMBS Trusts

4b. The Illumination Model performs a series of regression analyses to validate the Preliminary Industry Loss Assumptions against the actual performance of these Industry RMBS Trusts

4b. The Illumination Model reviews the results of these regression analyses, modifies its assumptions and reruns the regressions analyses over and over until the Preliminary Industry Loss Assumptions closely match the actual Industry RMBS losses to date.

4b. Once the Preliminary Industry Loss Assumptions accurately forecast the actual performance, they are deemed Preliminary Trusts Assumptions and ready to be applied against the Trusts Remit data

-14-

29. Step 4c - The Illumination Model then applies these Validated Industry

RMBS Trust Assumptions to the Trusts (“Preliminary Trusts Assumptions”). The Illumination

Model then performs a series of regression analyses to validate these Preliminary Trust

Assumptions against the actual performance of the trusts to obtain the validated Settlement Trust

assumptions (“Validated Trusts Assumptions”).

30. Step 4d - After this last regression analysis step, the Illumination Model

then utilizes the Validated Trust Assumptions for each of the trusts to forecast the remaining

lifetime losses for the trusts and for each tranche of bonds within each trust.

31. Step 5 - Determining the Forecasted Remaining Lifetime Losses for the

Non-Wrapped Trusts: I added the forecasted remaining lifetime losses of the non-wrapped trusts

Validating the Trusts Assumptions and Forecasted Remaining Lifetime Losses

4c. The Illumination Model then evaluates the Preliminary Trusts Assumptions for the useof forecasting the remaining lifetime losses for the Trusts.

4d. The Illumination Model performs a series of regression analyses to validate the Preliminary Trusts Assumptions against the actual performance of the Trusts.

4d. The Illumination Model reviews the results of these regression analyses , modifies its assumptions and reruns the regressions analyses over and over until the Preliminary Trusts Assumptions closely match the actual Trusts losses to dates.

4d. Once the Preliminary Settlement Trusts Assumptions closely match the actual performance, they are deemed Validated Trust Assumptions and they are applied to the Trusts to forecast the remaining lifetime losses.

-15-

from the both the Illumination and Intex Models, for both the lower and higher ranges. For the

429 non-wrapped Debtor-sponsored trusts, the aggregate results were:

Forecasted Remaining Lifetime Losses (in billions)

# of Trusts

Lower Range Higher Range

Non‐Wrapped Trusts 429 $7.4 $8.0

32. Step 6 - Determining the Actual Losses to Date for the Non-Wrapped

Trusts: I added the actual trust losses to date for the trusts (other than shortfalls to the wrapped

bonds) from both the LP and Intex Remit Data. For the 429 non-wrapped Debtor-sponsored

trusts, the aggregate actual trust losses to date are $29.6 billion.

33. Step 7 – Determining the Total Estimated Lifetime Loss ranges for the

Non-Wrapped Trusts: I added the total actual trust losses to date to the forecasted remaining

lifetime losses to determine the total estimated lifetime loss for both the lower and higher ranges.

The calculations are illustrated below:

Total Estimated Losses for the Non‐Wrapped Debtor‐sponsored Trusts (in billions)


Actual Losses to Date $29.6 $29.6

Forecasted Remaining Lifetime Losses $7.4 $8.0

Total Estimated Lifetime Losses $37.0 $37.6

34. The difference between the high and low ends of the range is explained by

the different assumptions supplied to the model. There are four assumptions that were varied

between the higher and lower estimated lifetime losses: constant default rate; severity; voluntary

constant prepayment rate; and forward home price index.

-16-

35. Step 8- Determining the Forecasted Remaining Lifetime Losses for the

Non-Wrapped Bonds Within the Wrapped Trusts: I added the forecasted remaining lifetime

losses for the non-wrapped bonds within the wrapped trusts other than those wrapped by FGIC,

from the both the Illumination and Intex Models, for both the lower and higher ranges. The

calculations for the forecasted remaining lifetime losses to the non-wrapped bonds within

wrapped trusts are illustrated below:

Forecasted Remaining Lifetime Losses

(in billions)

# of Trusts


105 $0.43 $0.46

36. Step 9 - Determining the Actual Losses to Date for the Non-Wrapped

Trusts: I added the actual trust losses to date for the non-wrapped bonds within the wrapped

trusts from both the LP and Intex Remit Data. For the non-wrapped bonds in the 105 wrapped

Debtor-sponsored trusts, the aggregate actual trust losses to date are $3.1 billion.

37. Step 10 – Determining the Total Estimated Lifetime Loss ranges for the

Non-Wrapped Bonds Within Wrapped Trusts: I added the total actual trust losses to date to the

forecasted remaining lifetime losses to determine the total estimated lifetime loss for the non-

wrapped bonds within wrapped trusts, for both the lower and higher ranges. The calculations are

illustrated below:

-17-

Total Estimated Losses for the Non‐Wrapped Bonds in Wrapped Debtor‐sponsored Trusts

(in billions)


Actual Losses to Date $3.1 $3.1

Forecasted Remaining Lifetime Losses $0.43 $0.46

Total Estimated Lifetime Losses $3.53 $3.56

38. Step 11 – Adding the Total Estimated Lifetime Loss ranges for the Non-

Wrapped Trusts and the Non-Wrapped Bonds Within Wrapped Trusts: for both the lower and

higher ranges, I added the total estimated lifetime loss for the non-wrapped trusts to the total

estimated lifetime loss for the non-wrapped bonds within wrapped trusts. The calculations are

illustrated below.

LOWER RANGE (in billions)

# of Trusts

Actual Losses to Date


Losses

Total Estimated Lifetime Losses

Non‐Wrapped Trusts 429 $29.6 $7.3 $36.9

Non‐Wrapped Bonds in Wrapped Trusts

105 $3.1 $0.43 $3.53

Total 534 $32.7 $7.73 $40.4

HIGHER RANGE (in billions)

# of Trusts

Actual Losses to Date


Losses

Total Estimated Lifetime Losses

Non‐Wrapped Trusts 429 $29.6 $8.0 $37.6

Non‐Wrapped Bonds in Wrapped Trusts

105 $3.1 $0.46 $3.57

Total 534 $32.7 $8.46 $41.13

39. The trust level data and calculations for the non-wrapped Debtor-

sponsored trusts are set forth in Annex C.

-18-

B. The Estimated Aggregate Lifetime Losses For Trusts Not Sponsored by the Debtors.

40. The RMBS Settlement also resolves claims belonging to 160 trusts that

were not sponsored by the Debtors. These non-Debtor-sponsored trusts will be allocated a

portion of the RMBS Settlement based on the percentage of the loans in each trust for which one

of the Debtors was the seller (typically meaning they originated the underlying loan). I have

conducted a preliminary analysis to estimate a range of potential lifetime losses for the non-

Debtor-sponsored trusts and my work to refine that analysis and narrow the estimated range is

ongoing.

41. Schedules 3g and 3r to the Plan identify the non-Debtor-sponsored trusts

with potential repurchase claims against the Debtors based on loans within the trusts sold by the

Debtors. Those schedules also set forth the percentage of loans within each trust that were sold

by the Debtors, based on analysis conducted by Duff & Phelps (“Duff”).

42. First, I reviewed the methodology that Duff used to determine the seller

percentage for the non-Debtor-sponsored trusts. Duff relied on the percentage of loans identified

in the bond offering documents as having been sold by the Debtors. Where that information was

not available in the offering documents, Duff used the percentage of loans being serviced by the

Debtors. Duff’s methodology is reasonable and is based on public historical information. I

independently confirmed the seller percentages set forth on Schedules 3g and 3r for a sample of

the trusts, based on the same methodology. Having found no errors, I relied on seller

-19-

percentages set forth in Schedules 3g and 3r for the remainder of the trusts.10 The average Seller

Percentage for the non-Debtor-sponsored trusts on the schedules is 7.5%.

43. Second, using the same data sources that I used for my analysis of the

Debtor-sponsored trusts, I independently collected the current outstanding balance and the total

collateral losses to date (as of September 2013 remittance) for each of the non-Debtor-sponsored

trusts identified on Schedules 3g and 3r. Actual losses to date make up the majority of the losses

for deals originated prior to 2007, as demonstrated by my analysis of the Debtor-sponsored

trusts. For the 160 non-Debtor-sponsored trusts, the aggregate actual trust losses to date are

$1.32 billion.

44. Third, I adjusted the actual losses to reflect the percentage of the collateral

that was sold by the Debtors, as reflected by the Seller Percentage described above, to determine

the portion of the trust collateral losses potentially attributable to the Debtors. I performed this

calculation at the trust level, applying each trust’s Seller Percentage to its respective actual trust

losses to date.

45. Fourth, I estimated that the non-Debtor-sponsored trusts would, on

average, experience future losses in the same proportion to outstanding collateral balance that

was forecast by my cash flow modeling of the Debtor-sponsored trusts. The non-Debtor-

sponsored trusts span a similar range of years (1999 to 2007) and product types, and the loans in

those trusts that were sold by Debtors are therefore likely to perform similarly, on average, to the

loans in the Debtor-sponsored deals from the same time period.

10 Schedules 3g and 3r were updated in a supplemental filing on October 11, 2013. My analysis relied on the prior version of the Schedules filed with the Plan on August 23, 2013, and I am in the process of updating my analysis to account for any discrepancies.

-20-

46. Fifth, I determined that for the 581 Debtor-sponsored trusts for which I

had generated forecasts using trust-level cash flow models, the lower and higher range estimates

of forecast lifetime potential losses were 14.6% to 15.9% of the outstanding collateral balance,

on average. The figures are:

Debtor‐sponsored Trusts

# of Trusts

Total Outsanding Principal Balance

(“OPB”)

Lower Range Forecasted Remaining

Lifetime Losses

Higher Range Forecasted Remaining

Lifetime Losses

billions % of OPB Billions % of OPB

581 $55.5 $8.1 14.6% $8.8 15.9%

47. Sixth, I used those percentages to estimate the total potential lifetime

losses to the portion of the collateral in the non-Debtor-sponsored trusts that was sold by the

Debtors, based on the current total outstanding balance for the non-Debtor-sponsored trusts. The

figures are:

Non‐Debtor‐sponsored Trusts

# of Trusts

Total Outsanding Principal Balance

(“OPB”)

Adjusted for Seller

Percentage

Lower Range Forecasted Remaining

Lifetime Losses

Higher Range Forecasted Remaining

Lifetime Losses

billions % of OPB billions % of OPB

160 $29.3 7.5% $0.3 14.6% $0.4 15.9%

48. Adding the Total Estimated Lifetime Loss ranges for the Debtor-

sponsored Trusts and the non-Debtor-sponsored Trusts: for both the lower and higher ranges, I

then added the total estimated lifetime loss for the Debtor-sponsored trusts to the total estimated

lifetime loss for the non-Debtor-sponsored trusts. The calculations are illustrated below

-21-

Total Non‐Wrapped Forecasted Lifetime Losses (in billions)


Debtor‐sponsored Trusts $40.5 $41.2

Non‐ Debtor‐sponsored Trusts $1.6 $1.7

Total $42.1 $42.9

49. The trust level data and calculations for the non-Debtor-sponsored trusts

are set forth in Annex D. I am in the process of updating and refining my analysis based on the

updated Plan schedules.

C. Underwriting Defect Rates For Debtor-Sponsored and Non-Sponsored Trusts.

50. The second step of my analysis involved estimating the number of loans

with material underwriting defects. This allowed me to calculate the losses that will be incurred

on loans sold in material breach of representations and warranties.

51. For my analysis of the prior RMBS settlement, I estimated the number of

loans with material underwriting defects using two distinct, and widely-accepted, methodologies.

The first methodology was based on the Debtors’ historical loan repurchase experience. After

performing that initial analysis, I had the opportunity to conduct a forensic re-underwriting of

approximately 1,500 randomly-selected loans securitized between 2004 and 2007. This allowed

me to determine, for a representative sample of loans within those trusts, the potential level of

material underwriting defects.11

52. The two methods produced similar results. For purposes of this Report, I

relied on the results from the second methodology—my forensic underwriting analysis—and

11 The expedited re-underwriting review did not consider the potential to cure missing documentation defects, or certain other arguments that might justify deviations from the general underwriting guidelines, so the actual level of material defects may be lower than the rates of potential underwriting defects identified here.

-22-

then multiplied the amount of estimated aggregate lifetime losses by the likely material defect

rate determined through that underwriting analysis in order to determine the amount of losses

that will be incurred on loans with likely material defects.

53. The approximately 1,500 loans reviewed for my underwriting analysis

were chosen from the 392 trusts involved in the prior settlement. Dr. Brad Cornell, a consultant

hired by the Official Committee of Unsecured Creditors, made a random selection of 1,500 loans

from across all 392 trusts involved in the earlier RMBS settlement. I assumed, for purposes of

my analysis, that his sample selection was properly randomized and representative of the entire

population of approximately 1.6 million loans underlying the 392 trusts involved in the earlier

settlement.

54. A team of forty-two loan underwriters, and three underwriting managers,

working under my direction, re-underwrote these loans using the Debtors’ own underwriting

guidelines. The underwriters have a minimum of three years of experience in one or more of the

following areas: residential mortgage underwriting, mortgage loan auditing, mortgage loan

quality control and mortgage re-underwriting experience. The underwriting managers have five

or more years of experience in one or more of these same areas.

55. The re-underwriting process involved a review of each loan’s “loan file.”

My team considered, and recalculated, the various credit and other metrics used in the

underwriting process, including debt-to-income ratios, loan-to-value ratios, and combined loan-

to-value ratios. My team also considered income documentation, employment documentation,

asset documentation; appraisals, credit scores and reports, insurance documentation, title

documentation and insurance, and other transaction details as necessary when available. The

-23-

team of underwriters also used various industry-standard third-party verification tools to help

validate origination information.

56. As a first step, my team of underwriters, underwriting managers and I

familiarized ourselves with the applicable underwriting guidelines and automated underwriting

system (AUS) loan approval formats to ensure that our re-underwriting analysis of the loan files

was based on those requirements. These included the GMAC RFC Client Guides, GMAC RFC

Client Guide Bulletins, other Lender underwriting guidelines, as applicable, and AUS Loan

Approval formats.

57. Next, my team conducted a review of the loan file (where available) for

each of the 1500 loans, and recalculated and recorded various credit and other metrics relevant to

the underwriting analysis. These metrics included, among others, the debt-to-income ratio, loan-

to-value ratio, and combined loan-to-value ratio. My team also otherwise reviewed the loans for

compliance with the AUS loan approval or applicable underwriting guidelines.

58. The re-underwriting process included, among other things, a review of the

following characteristics and a comparison of those characteristics to the applicable underwriting

guidelines or AUS, as applicable:

(a) Income: The income related documentation, if any, required to

satisfy the underwriting guidelines;

(b) Employment: The employment related documentation, if any,

required to satisfy the underwriting guidelines;

(c) Assets: The asset related documentation, if any, required to satisfy

the underwriting guidelines;

-24-

(d) Appraisal: The appraisal related documentation, if any, required to


(e) Credit: The credit related documentation, if any, required to


(f) Insurance: The insurance documentation, if any, required to


(g) Title: The title documentation, if any, required to satisfy the

underwriting guidelines; and

(h) Transaction details: The transaction related documentation, if any,

including the Purchase Contract if applicable, required to satisfy the underwriting guidelines.

59. During the course of the review, my team used various industry-accepted

third-party re-verification tools to help validate origination information.

(i) MERS: The MERS Link report allowed the underwriters to check

if there were any undisclosed mortgages at the time of origination of the subject loan that were

not disclosed by the borrower and not included in their debt-to-income ratios.

(j) Accurint: The Deep Skip Search report allowed the underwriters

to validate whether or not the borrower(s) were associated with the subject property during the

required period after the close of the subject loan for owner occupied transactions.

(k) The Work Number: The Current and Previous Employment report

available through The Work Number allowed the underwriters to validate income and

employment for borrowers whose employers provide employment data to The Work Number.

(l) Verbal Verification of Employment (VVOE): If the underwriters

were unable to obtain the employment information from The Work Number, then he or she

-25-

might elect to obtain the employment information directly from the employer and record the

results on a VVOE form.

60. The underwriters recorded the results of their review on a spreadsheet

template, entitled “Re-Underwriting Findings Summary Report,” that I designed for this purpose.

Among other things, the underwriters records kept track of the product type (such as subprime or

HELOC) of each loan reviewed.

61. Once the underwriters completed their work, the underwriting managers

performed quality control checks on a portion of the results to assess the accuracy and

completeness of the information presented. As part of these reviews, the underwriting managers

validated the accuracy of the underwriters’ findings against the data and documents in the

imaged loan files, checked some or all of the underwriters’ calculations, and referred to the

applicable guidelines or AUS as needed.

62. After the underwriting managers completed their review of the findings, I

then reviewed the Re-Underwriting Findings Summary Report for each and every one of the

loans, and made the ultimate determination as to whether a loan was “materially defective” based

on the information available. To the extent I deemed it necessary or desirable to refer to the

actual loan file or underwriting guidelines, the AUS or Loan Approval, or the governing

agreements, those materials were available to me and I made use of them. I also had

conversations with the underwriters and underwriting managers on my team during which I

posed questions and obtained clarification regarding various aspects of the loans reviewed if I

felt it was necessary.

-26-

63. If the loan was underwritten substantially in accordance with the

applicable guidelines or AUS, I designated the loan to be in substantial compliance (“In

Substantial Compliance”).

64. If the loan appeared based on the review to have defects that materially

increased the risk of the loan, I deemed the loan to be materially defective (“Materially

Defective”).12 Ultimately, there is a degree of subjectivity in the determination of whether a loan

substantially complies with the guidelines or whether any deviation is material, and I endeavored

to be conservative in those judgments so as not to understate the potential level of material

defects.

65. After reviewing the work performed by my team of underwriters, I

determined that 652 of the loans in the sample, or 43.5%, had material underwriting defects.

66. For trusts formed before 2004, I analyzed early payment defaults to

determine a defect rate for the loans prior to 2004. There is a known correlation between early

payment defaults and significant underwriting defects.13 Accordingly, I analyzed early

delinquency rates in order to estimate the defect rate for the pre-2004 trusts. Using aggregate

loan performance data for the Debtors’ RMBS shelves prior to and after 2004, as available from

the Vision website, I compared the rate of severely delinquent loans (i.e. loans that were 90 days

or more delinquent, or in foreclosure or REO) at 12 months after issuance of each vintage. From

that comparison, I created a ratio of the Debtors’ average severely delinquent percentage prior to

12 For purposes of this expedited review, I used a lack of substantial compliance with applicable guidelines as a proxy for breaches of contractual representations and warranties; consideration of additional information, including specific aspects of the securitization transaction documents, could impact the results.

13 See Early Payment Default – Links to Fraud and Impact on Mortgage Lenders and Investment Banks, BasePoint Analytics LLC (2007), attached hereto as Annex G.

-27-

2004, as compared to 2004 and after. To estimate the pre-2004 defect rate, I then multiplied that

ratio by the 43.5% defect rate of the 2004-2007 trusts.

67. On average, 4.23% of the 2004 to 2007 loans were severely delinquent

one year after issuance, whereas 3.41% of the pre-2004 loans were severely delinquent one year

after issuance. In other words, the severe delinquency rate prior to 2004 was 80.6% of the later

severely delinquent rate. Based on the correlation between severe delinquencies and

underwriting defects, I applied a defect rate for the pre-2004 loans that was 80.6% of the 2004 to

2007 defect rate, which works out to a 35.1% defect rate for the pre-2004 loans. The

delinquency ratio data and calculations are set forth in Annex E.

D. Likely Damages After Factoring In Legal Defenses And Litigation Expenses.

68. The third step in my analysis involved estimating the likely damages that

could be recovered by the trusts after consideration of legal defenses to repurchase demands and

litigation costs. I did this by applying a discount (the “Settlement Factor”) to the amount of

losses on loans with material defects. The discount was based on the Debtors’ own historical

repurchase experience, industry repurchase data, my own experience with loan repurchase

demands, and the opinions of other experts regarding legal defenses and litigation expenses. I

then multiplied the amount of losses that will be incurred on loans with material defects with the

Settlement Factor I calculated.

69. I determined it is appropriate to use a Settlement Factor of 41% to 47%.

70. The Debtors had, over the preceding five years, received and responded to

requests for the repurchase of thousands of loans based on alleged breaches of representations or

warranties. Most of the requests were made by Fannie Mae, Freddie Mac, and some came from

various mortgage loan insurers. The Debtors’ historical experience in accepting, or rejecting,

-28-

repurchase demands offers a robust pre-litigation source of verifiable and objective data

regarding the effectiveness of the Debtors’ legal and factual defenses.

71. Prior to March 2012, there was not much in terms of public disclosures

with any insight into Settlement Factors for alleged breaches of representations and warranties.

However, beginning in March of 2012, Fannie Mae, Freddie Mac and over a dozen Private Label

Sellers have filed Regulation AB 15-G repurchase demand data with the SEC, including their

rate of agreed repurchases.

72. The Debtors’ historical repurchase demand experience with Fannie Mae

and Freddie Mac is publicly reported in those entities’ SEC filings and has been summarized in a

report published by the Inside Mortgage Finance publication. I used data for repurchase

demands by Fannie Mae and Freddie Mac, rather than repurchase demand data for private label

securities (like those included in the RMBS settlement trusts), because the government-

sponsored enterprises’ data are much more robust and reliable.

73. The Debtors’ 2006-2008 GSE historical repurchase data, based on both

Fannie Mae and Freddie Mac’s Regulation AB 15-G SEC filings, as summarized in Inside

Mortgage Finance’s Special Report (“IMF Special Report”),14 are as follows:

Seller/Originators Repurchase Demands (millions)

Repurchased Pending Disputed

GMAC Mortgage / Ally (the Debtors)

$1,537.81 67.56% 2.60% .50%

All Seller / Originators

$65,836.91 49.54% 12.58% 4.15%

14 In this Special Report, the Debtor is referred to as “GMAC Mortgage / Ally.” An excerpt of this report is attached hereto as Annex H.

-29-

74. As reflected in the above chart, the average GSE Agree Rates for all

sellers was 49.54% and 67.56% for the Debtors. In our assumptions, we discount the GSE agree

rates based on the less stringent representations and warranties found in the Debtors’ trusts’

governing agreements when compared to the stronger representations and warranties found in the

Fannie Mae and Freddie Mac agreements. For example, in many of Debtors’ trusts’ governing

agreements there is little to no fraud representation or warranty language, and the requirements

to conform to the Underwriting Guidelines are often qualified with “generally” or “substantially”

in compliance with the Underwriting Guidelines, which are both lower standards than are found

in Fannie Mae or Freddie Mac agreements.

75. The GSEs have requested sellers to repurchase approximately $66 billion

in loans as noted in their recent SEC filings as summarized in IMF Special Report,15 while

industry estimates forecast that sellers of non-GSE securities, known as PLS, will repurchase

hundreds of billions in loans, resulting in seller losses of approximately $133 billion according to

Compass Point Research.16 As of the date of my review in connection with the prior RMBS

settlement, the data regarding repurchase demands concerning private label securities, suffered

from a large percentage of unresolved demands. For example, 82% of repurchase demands from

private-label trusts remain unresolved, while only 3.1% of Fannie Mae and Freddie Mac remain

unresolved. Rather than using an incomplete and unreliable data set, I relied on repurchase

demand data from Fannie Mae and Freddie Mac.

15 See Annex H.

16 See Annex I: Compass Point Research on Mortgage Repurchases Part II: Private Label RMBS Investors Take Aim, dated August 17, 2010.

-30-

76. According to data contained in SEC filings for Fannie Mae and Freddie

Mac, the Debtors accepted 67.5% of all repurchase demands asserted by Fannie Mae and Freddie

Mac. Industry data published by the Inside Mortgage Finance publication shows that all sellers

of mortgage-backed securities accepted 49.5% of all repurchase demands asserted by Fannie

Mae and Freddie Mac. Thus, the Debtors’ acceptance rate for repurchase demands is

substantially higher than that of the industry as a whole.

77. I recognize that, in the litigation context, plaintiffs would strongly dispute

the application of any discount based on the strength of representations and warranties at issue.

Nonetheless, in my analysis, I applied a substantial discount to the Debtors’ acceptance rate due

to the more stringent representations and warranties found in the governing agreements for loans

sold to Fannie Mae and Freddie Mac. In other words, I assumed that the Debtors would have

stronger repurchase defenses, and would prevail more often, in disputes with owners of the

private label mortgage-backed securities than the Debtors would in disputes with Fannie Mae

and Freddie Mac.

78. In preparation for my Original Declaration, I also reviewed the publicly-

available settlement information relating to the following settlements:

Seller/Originator Securitization Type Settlement Amount Date Bank of America PLS $8,500,000,000 June 201117

Lehman PLS $40,000,000 November 2011 Bank of America Fannie Mae $1,520,000,000 January 2012 Bank of America Freddie Mac $1,280,000,000 January 2012

79. Both the Bank of America (“BofA”) and Lehman PLS settlements and the

corresponding RMBS trusts are similar in terms of the securitization structure, issuance years,

mortgage loan product mix, governing agreements and R&W repurchase standards.

17 Bank of America settlement for 530 trusts is pending court approval.

-31-

80. Based on the above and in consideration of the costs, risks and

uncertainties if the parties do not mutually agree on the repurchase population and have to resort

to litigation to resolve their differences, we have discounted the Debtors’ GSE repurchase rates

and have assumed the trusts’ Settlement Factor ranges between a low of 41% and a high of 47%.

The following chart shows a comparison of this Settlement Factor to that used in the BofA

Expert Report and Lehman Expert Declaration:

Description Settlement Factor Assumptions Lower Range Higher Range

Trusts 41% 47% BofA Expert Report 40% 40%

Lehman Expert Declaration 30% 40%

81. The Settlement Factor assumptions for the lower range are similar in this

Report and the BofA Expert Report, while the Lehman Expert Declaration lower range

assumption is a more aggressive assumption than in my Report or the BofA Expert Report. The

Settlement Factor assumptions for the higher range utilized in this Report are higher than those

used in both the BofA Expert Report and the Lehman Expert Declaration. I concluded that

higher Settlement Factor assumptions in this Report are correlated to the Debtors’ substantially

higher actual rate of agreement with the GSEs when compared to the industry as a whole,

67.56% versus 49.54%. Given the above, these Settlement Factor assumptions are in my

professional opinion reasonable for the trusts.

82. Applying the total estimated lifetime loss, defect rate, and Settlement

Factor figures set forth above, the range of the Debtors potential liability to the trusts included in

the RMBS Settlement is $7.4 billion to $8.6 billion. The calculations are:

-32-

Estimated Lifetime Losses

x Defect Rate

=Losses on Defective

Loans x

Settlement Factor Range

= Final Range of

Likely Damages

Low High Low High Low High Low High

2004-2007 Trusts

$38.2 $38.9 43.5% $16.6 $16.9 41% 47% $6.8 $7.9

Pre-2004 Trusts

$3.89 $3.93 35.1% $1.36 $1.38 41% 47% $.56 $.65

Total $42.1 $42.9 $18.0 $18.3 $7.4 $8.6

III. OPINIONS REGARDING THE MONOLINE SETTLEMENTS.

83. The Plan seeks to resolve claims asserted by four monoline insurance

companies relating to RMBS trusts for which they provided insurance, promising to pay

investors to compensate for any shortfall in principal or interest payments to wrapped bonds.18

84. As a result, the first step in my analysis of the Debtors’ potential liability

to the monolines was to estimate the potential lifetime payments by each monoline to cover

shortfalls in principal and interest payments to the wrapped bonds.

A. Methodology for Estimating Aggregate Lifetime Shortfall Exposure.

85. I estimated the monolines’ aggregate lifetime shortfall exposure using the

same Illumination and Intex cash flow models discussed above, to simulate the cash flows to

each tranche of bonds. Those models allowed me to obtain forecasted cash flows not just at the

trust level, but also for specific tranches of bonds.

86. The shortfall to wrapped bonds is not the same as the collateral losses to

the trust issuing the bonds for a number of reasons. RMBS are premised on the assumption that

some mortgages will default, and one of the key features of RMBS securitization structures is to 18 I understand that certain monolines also asserted claims against Ally Financial, Inc., and that all four monoline settlements and the RMBS settlement form part of the Plan that includes a third party release of claims against AFI. I have been informed that the third party release is beneficial to the Debtors, the creditors and third parties but I have not performed any calculations to estimate the value of that third party release.

-33-

provide protection, or credit support, so that payments can be made to investors of certain more

senior bonds even if there are some losses on the underlying mortgages. The level of credit

support available to protect investors in any given bonds depends on the expected level of losses

at the time of the transaction, the credit support structure of the trust and the level of risk (and

resulting higher return) selected by the investor.

87. There are a number of types of credit support available to insulate

wrapped bonds from collateral losses, including overcollateralization, excess spread, and

subordination, as discussed above. Monoline insurers do not wrap the lowest, equity tranche of

bonds in the RMBS structure. In some transactions, there may also be additional tranches of

subordinate bonds that are not wrapped. Because of these and other credit enhancement features,

not every decline in the value of collateral will result in a failure to make required timely

payments of interest and principal to wrapped bonds.

88. The credit support available to wrapped bonds changes on a monthly

basis based on factors such as the accumulation of excess spread, so the timing of collateral

losses is important and one cannot determine the extent to which those losses will result in

shortfalls to any particular tranche of bonds without running a cash flow model that simulates the

payments coming in and out of the RMBS trust on a monthly basis.

89. Consistent with my analysis of the non-wrapped trusts, I used the

Illumination and Intex cash flow models to determine the extent to which future collateral losses

on the underlying mortgages would result in payment shortfalls to each tranche of wrapped

bonds and thus potential claims payments by the monolines.

90. In modeling the cash flows for the wrapped bonds, I used the same

variables and assumptions as I used to model lifetime collateral losses for the Debtors’ non-

-34-

wrapped trusts. However, the Illumination and Intex models have a feature to forecast payments

from the monoline insurer that reduce forecasted loss to the wrapped bonds, because the ultimate

loss to the bond holders, net of payments from the monoline, is an important consideration to

bond investors. For my purpose of estimating potential losses to the monolines, I set this feature

of the model to assume that there would be no payments from the monoline. As a result, the

model output of future losses to the wrapped bonds reflects the entire potential future shortfall of

principal and interest for which the monoline is potentially liable.19

B. Estimating Potential Monoline Damages and Accounting for Settlement

91. The second step of my analysis was to consider potential defenses and

other considerations in litigating or settling the monoline claims. The nature of the monolines’

claims is different than the trusts’ claims and therefore my evaluation of the Debtors’ potential

liability to the monolines, and of the litigation and settlement factors, is also different.

92. First, I understand that the monolines have asserted claims – such as

claims for fraudulent inducement and material breach of contract – under which, upon proving

material breaches of representations and warranties, they would be entitled to recover the entire

aggregate lifetime shortfalls on all the trusts for which they provided insurance. If the foregoing

is correct, then there is no need to consider the extent to which each monoline’s shortfall resulted

from material underwriting defects. Instead, the full amount of the shortfalls I calculated in step

one of my analysis could be recovered by each monoline without reduction.

93. Second, the monolines assert that they are also entitled to indemnification

of legal expenses, costs, and interest under the governing agreements. Therefore, in addition to

19 I did not evaluate the impact of any potentially offsetting premiums paid to the monolines, which I understand to be relatively small compared to the claims and, as noted below, would also have been factored in to the comparable settlements discussed herein.

-35-

defenses that might lead the monolines to recover less than the wrapped bond shortfalls, I also

must consider the possibility that the monolines would recover more than the wrapped bond

shortfalls.

94. I also understand that RMBS sponsors have argued that monolines may

only recover the shortfall in principal and interest payments caused by the Debtors’ failure to

repurchase defective loans.

95. As discussed above, for my analysis of loans in the Debtor-sponsored non-

wrapped trusts, I conducted a re-underwriting of approximately 1500 loans of various types sold

by the Debtors between 2004 and 2007. However, the sample of loans underwritten for that

analysis is not necessarily representative of the loans in any trust wrapped by a monoline. The

Debtors structured their RMBS issuances to securitize different types of loans in different trusts.

For example, the trusts with the designation “HE” in the name (such as GMACM 2005-HE3) are

backed by home equity loans, whereas the trusts designated RALI “QH” are backed by Alt-A

loans. Each trust therefore has distinct characteristics that may impact the defect rates.

Comparing the defect rates between different product types within my original sample revealed

wide range of likely defects, from approximately 20% to 60%, depending on the product type.

However, those subsets of my initial underwriting sample are not necessarily statistically

representative of any product type as whole from 2004 to 2007, nor within any particular vintage

or trust. Moreover, each monoline’s potential liability is unevenly distributed among the trusts it

wrapped, so variation in the defect rate for any particular vintage or trust could be material.

96. Given the time constraints, it was not possible to conduct additional

sampling and review the hundreds or thousands of additional loan files that would be necessary

to construct representative loan samples for each wrapped trust. Therefore, I did not apply a

-36-

defect rate to the wrapped trusts, and my work to evaluate the potential impact of varying defect

rates is ongoing.

97. I looked at recent monoline settlements to determine a likely range of

potential liability to the monolines based on the bond losses. The monoline settlement values, as

compared to the monolines’ potential bond loss exposure being settled, reflect the combined

influence of (i) any discount based on legal defenses or other litigation considerations, (ii) the

monolines’ claims for indemnification for fees, expenses and interest, and (iii) any potential

offset for premiums received by the monolines.

98. In preparation for this Report, I reviewed publicly-available settlement

information relating to the following settlements:

Monoline Seller/Originator Date Collateral Losses

Bond Losses

Settlement Payment

Notes

Assured Bank of America April 2011 $1.3b $0 $1.1b Plus 80% of future bond losses

Syncora Bank of America July 2012 - $402m $375m Plus transfer of securities

Assured UBS May 2013 - 308.2m20 358m Plus future loss sharing

MBIA Flagstar May 2013 $225m $165m $110m Plus other consideration

Assured Flagstar June 2013 - $89.1 $106.5 After trial awarding full losses, fees and

costs

99. The publicly available information did include not all material terms of the

settlements, including, for example, the value of certain non-cash consideration or expected

future claims. As a result, it is difficult to draw precise conclusions from these settlements. The

20 This figure was based the complaint and does not account for any additional bond losses incurred between that time and settlement.

-37-

available information does demonstrate, however, that settlement of monoline claims at a small

discount, if any, to estimated bond losses, is within the range of recent settlements.

100. The Assured v. Bank of America (April 2011) settlement disclosed the

actual collateral losses to date ($1.3 billion) as well as settlement payment ($1.1 billion), which

reflects payment of approximately 85% of the collateral losses to date, even though none of the

collateral losses had yet translated into bond losses reimbursed by Assured. The settlement also

provided that Bank of America would pay 80% of any future losses on the wrapped bonds, up to

a cap of approximately five times the losses to date.

101. The more recent Syncora v. Bank of America settlement (July 2012)

disclosed the actual bond losses reimbursed by Syncora ($145 million) and the outstanding but

unpaid claims to Syncora ($257 million), as well as the up-front settlement amount ($375

million). The settlement reflected a payment of 93% of the claim amounts to Syncora to date.

No information was provided concerning expected future payments by Syncora. However, the

settlement did also provide for additional consideration to Syncora including the transfer of

certain securities, the value of which was not disclosed. Therefore, it is unclear whether the

settlement amount reflected any reduction from Syncora’s expected payment obligations.

102. In connection with the Assured v. Flagstar settlement (June 2013), I also

reviewed the opinion written by Judge Rakoff following the Assured v. Flagstar trial (February

2013). After trial, the court awarded Assured full reimbursement for its payments made under

the policy ($89.1 million), as well as legal expenses and other costs to be determined later. It did

so based on evidence that Flagstar’s obligation to repurchase defective loans would have been

sufficient to cover the entire loss to bonds wrapped by Assured. The parties subsequently settled

for $106.5 million, on top of which Flagstar will also reimburse Assured for all future claims on

-38-

its insurance policies. This settlement after trial thus reflects a roughly 15-20% premium over

the total shortfall to the wrapped bonds.

103. Based on these recent monoline settlements and the possibility that the

monolines could recover more than the wrapped bond losses, I determined that it was appropriate

to apply a Settlement Factor range from 80% of the total expected bond losses to 100% of the

total bond losses, while recognizing that a payment above the total bond losses to account for

potential indemnification expenses and interest payments, as occurred in Assured v. Flagstar, is

not outside the range of reasonableness.

104. The trust level data and calculations for the bonds wrapped by each of the

four monolines are set forth in Annex F.

C. Opinions Regarding the Ambac Settlement.

105. The Plan seeks to resolve claims asserted by Ambac relating to 64 Debtor-

sponsored RMBS trusts for which Ambac provided insurance, promising to pay investors to

compensate for any shortfall in principal or interest payments in the Wrapped bonds.21 My

analysis is limited to evaluating the claim for potential shortfall in principal and interest

payments on the Debtor-sponsored wrapped bonds.

106. As a result, the first step in my analysis of Ambac’s damages was to

estimate Ambac’s potential lifetime payments to cover shortfalls in principal and interest

payments to the wrapped bonds in the trusts wrapped by Ambac.

21 In addition, the Ambac settlement resolves claims with respect to non-Debtor-sponsored trusts wrapped by Ambac, for which Ambac has asserted claims against the Debtors based on the Debtors’ servicing. My analysis of the Ambac settlement does not address those additional claims. As noted above, my analysis of the non-Debtor-sponsored trusts is ongoing.

-39-

1. Ambac’s Estimated Aggregate Lifetime Shortfall Exposure.

107. I estimated Ambac’s aggregate lifetime shortfall exposure using the same

Illumination and Intex cash flow models discussed above to simulate the cash flows to each

tranche of bonds.

108. First, I collected the data concerning actual shortfalls to the wrapped

bonds in each trust insured by Ambac as of September 2013. Second, I forecasted the potential

lifetime shortfall for the specific tranches of bonds wrapped by Ambac, using the same low range

and high range assumptions that I used to calculate trust losses set forth above.

109. The actual shortfalls, and the low range and high range forecasted

remaining lifetime shortfalls to bonds wrapped by Ambac are:

Trust Actual Wrapped Bond Losses (in millions)

Forecasted Remaining Lifetime Wrapped Bond Shortfall

(in millions)


2004‐KS4 $2.2 $5.6 $6.5

2004‐RS1 $0.0 $0.0 $0.0

2004‐RS5 $1.3 $8.1 $9.3

2004‐RS9 $1.8 $2.3 $2.7

2006‐QH1 $49.8 $2.4 $3.9

GMACM 2001‐HLTV1 $0.0 $0.0 $0.0

GMACM 2001‐HLTV2 $0.0 $0.0 $0.0

GMACM 2002‐HLTV1 $0.0 $0.0 $0.0

GMACM 2002‐RP2 $0.0 $0.0 $0.0

GMACM 2005‐HE3 $98.0 $4.6 $9.9

RAMP 1999‐RS1 $0.0 $0.0 $0.0

RAMP 2001‐RS1 $0.0 $0.0 $0.0

RAMP 2001‐RS3 $0.0 $0.0 $0.0

RAMP 2002‐RP1 $0.0 $0.0 $0.0

RAMP 2002‐RS1 $0.0 $0.0 $0.0

RAMP 2002‐RS2 $0.0 $0.0 $0.0

RAMP 2002‐RS4 $0.0 $0.0 $0.0

RAMP 2002‐RS5 $0.0 $0.0 $0.0

RAMP 2002‐RS6 $0.0 $0.0 $0.0

RAMP 2002‐RS7 $0.0 $0.0 $0.0

-40-



(in millions)


RAMP 2002‐RZ4 $0.0 $0.0 $0.0

RAMP 2003‐RP1 $0.0 $0.0 $0.0

RAMP 2003‐RP2 $0.0 $0.0 $0.0

RAMP 2003‐RS1 $0.0 $0.0 $0.0

RAMP 2003‐RS2 $0.0 $0.0 $0.0

RAMP 2003‐RS3 $0.0 $0.0 $0.0

RAMP 2003‐RS4 $0.0 $0.0 $0.0

RAMP 2003‐RS5 $0.0 $0.0 $0.0

RAMP 2003‐RS6 $0.0 $0.0 $0.0

RAMP 2003‐RS7 $0.0 $0.0 $0.0

RAMP 2003‐RS8 $0.0 $0.0 $0.0

RAMP 2003‐RS9 $0.0 $0.0 $0.0

RAMP 2003‐RZ1 $0.0 $0.0 $0.0

RAMP 2003‐RZ2 $0.0 $0.0 $0.0

RAMP 2003‐RZ3 $0.0 $0.0 $0.0

RAMP 2003‐RZ4 $0.0 $0.0 $0.0

RAMP 2003‐RZ5 $0.0 $0.0 $0.0

RASC 2002‐KS1 $1.9 $8.6 $9.7

RASC 2002‐KS4 $12.0 $0.4 $0.6

RASC 2002‐KS6 $5.5 $1.2 $1.2

RASC 2002‐KS8 $8.3 $10.4 $11.1

RASC 2003‐KS4 $0.0 $0.0 $0.0

RASC 2003‐KS5 $0.0 $0.0 $0.0

RASC 2003‐KS9 $0.0 $1.2 $2.1

RFMSII 1999‐HI1 $0.0 $0.0 $0.0

RFMSII 1999‐HI4 $0.0 $0.0 $0.0

RFMSII 1999‐HI6 $0.0 $0.0 $0.0

RFMSII 1999‐HI8 $0.0 $0.0 $0.0

RFMSII 2000‐HI1 $0.0 $0.0 $0.0

RFMSII 2000‐HI2 $0.0 $0.0 $0.0

RFMSII 2000‐HI3 $0.0 $0.0 $0.0

RFMSII 2000‐HI4 $0.0 $0.0 $0.0

RFMSII 2000‐HI5 $0.0 $0.0 $0.0

RFMSII 2000‐HL1 $0.0 $0.0 $0.0

RFMSII 2001‐HI1 $0.0 $0.0 $0.0

RFMSII 2001‐HI2 $0.0 $0.0 $0.0

RFMSII 2001‐HI3 $0.0 $0.0 $0.0

RFMSII 2001‐HI4 $0.0 $0.0 $0.0

RFMSII 2001‐HS2 $0.0 $0.0 $0.0

RFMSII 2002‐HI1 $0.0 $0.0 $0.0

-41-



(in millions)


RFMSII 2002‐HI2 $0.0 $0.0 $0.0

RFMSII 2002‐HI3 $0.0 $0.0 $0.0

RFMSII 2003‐HI3 $0.0 $0.0 $0.0

RFMSII 2003‐HS4 $0.0 $0.0 $0.0

Total $180.8 $44.8 $57.0

110. Third, I added the actual bond shortfalls to the forecasted potential

shortfalls, to determine the total potential lifetime shortfalls to the wrapped bonds. The

calculations of the low range and the high range total potential lifetime shortfalls for the bonds

wrapped by Ambac are:

LOWER RANGE

Trust

Actual Wrapped

Bond Losses (in millions)

+

Forecasted Remaining Lifetime Wrapped Bond

Shortfall (in millions)

=

Total Potential Lifetime

Wrapped Bond Shortfall

2004‐KS4 $2.2

$5.6

$7.8

2004‐RS1 $0.0 $0.0 $0.0

2004‐RS5 $1.3 $8.1 $9.4

2004‐RS9 $1.8 $2.3 $4.0

2006‐QH1 $49.8 $2.4 $52.3

GMACM 2001‐HLTV1 $0.0 $0.0 $0.0

GMACM 2001‐HLTV2 $0.0 $0.0 $0.0

GMACM 2002‐HLTV1 $0.0 $0.0 $0.0

GMACM 2002‐RP2 $0.0 $0.0 $0.0

GMACM 2005‐HE3 $98.0 $4.6 $102.6

RAMP 1999‐RS1 $0.0 $0.0 $0.0

RAMP 2001‐RS1 $0.0 $0.0 $0.0

RAMP 2001‐RS3 $0.0 $0.0 $0.0

RAMP 2002‐RP1 $0.0 $0.0 $0.0

RAMP 2002‐RS1 $0.0 $0.0 $0.0

RAMP 2002‐RS2 $0.0 $0.0 $0.0

RAMP 2002‐RS4 $0.0 $0.0 $0.0

RAMP 2002‐RS5 $0.0 $0.0 $0.0

RAMP 2002‐RS6 $0.0 $0.0 $0.0

RAMP 2002‐RS7 $0.0 $0.0 $0.0

-42-

LOWER RANGE

Trust

Actual Wrapped


+



=



RAMP 2002‐RZ4 $0.0

$0.0

$0.0

RAMP 2003‐RP1 $0.0 $0.0 $0.0

RAMP 2003‐RP2 $0.0 $0.0 $0.0

RAMP 2003‐RS1 $0.0 $0.0 $0.0

RAMP 2003‐RS2 $0.0 $0.0 $0.0

RAMP 2003‐RS3 $0.0 $0.0 $0.0

RAMP 2003‐RS4 $0.0 $0.0 $0.0

RAMP 2003‐RS5 $0.0 $0.0 $0.0

RAMP 2003‐RS6 $0.0 $0.0 $0.0

RAMP 2003‐RS7 $0.0 $0.0 $0.0

RAMP 2003‐RS8 $0.0 $0.0 $0.0

RAMP 2003‐RS9 $0.0 $0.0 $0.0

RAMP 2003‐RZ1 $0.0 $0.0 $0.0

RAMP 2003‐RZ2 $0.0 $0.0 $0.0

RAMP 2003‐RZ3 $0.0 $0.0 $0.0

RAMP 2003‐RZ4 $0.0 $0.0 $0.0

RAMP 2003‐RZ5 $0.0 $0.0 $0.0

RASC 2002‐KS1 $1.9 $8.6 $10.5

RASC 2002‐KS4 $12.0 $0.4 $12.3

RASC 2002‐KS6 $5.5 $1.2 $6.7

RASC 2002‐KS8 $8.3 $10.4 $18.7

RASC 2003‐KS4 $0.0 $0.0 $0.0

RASC 2003‐KS5 $0.0 $0.0 $0.0

RASC 2003‐KS9 $0.0 $1.2 $1.2

RFMSII 1999‐HI1 $0.0 $0.0 $0.0

RFMSII 1999‐HI4 $0.0 $0.0 $0.0

RFMSII 1999‐HI6 $0.0 $0.0 $0.0

RFMSII 1999‐HI8 $0.0 $0.0 $0.0

RFMSII 2000‐HI1 $0.0 $0.0 $0.0

RFMSII 2000‐HI2 $0.0 $0.0 $0.0

RFMSII 2000‐HI3 $0.0 $0.0 $0.0

RFMSII 2000‐HI4 $0.0 $0.0 $0.0

RFMSII 2000‐HI5 $0.0 $0.0 $0.0

RFMSII 2000‐HL1 $0.0 $0.0 $0.0

RFMSII 2001‐HI1 $0.0 $0.0 $0.0

RFMSII 2001‐HI2 $0.0 $0.0 $0.0

RFMSII 2001‐HI3 $0.0 $0.0 $0.0

RFMSII 2001‐HI4 $0.0 $0.0 $0.0

-43-

LOWER RANGE

Trust

Actual Wrapped


+



=



RFMSII 2001‐HS2 $0.0

$0.0

$0.0

RFMSII 2002‐HI1 $0.0 $0.0 $0.0

RFMSII 2002‐HI2 $0.0 $0.0 $0.0

RFMSII 2002‐HI3 $0.0 $0.0 $0.0

RFMSII 2003‐HI3 $0.0 $0.0 $0.0

RFMSII 2003‐HS4 $0.0 $0.0 $0.0

Total $180.8 $44.8 $225.6

HIGHER RANGE

Trust

Actual Wrapped


+



=



2004‐KS4 $2.2

$6.5

$8.7

2004‐RS1 $0.0 $0.0 $0.0

2004‐RS5 $1.3 $9.3 $10.6

2004‐RS9 $1.8 $2.7 $4.4

2006‐QH1 $49.8 $3.9 $53.7

GMACM 2001‐HLTV1 $0.0 $0.0 $0.0

GMACM 2001‐HLTV2 $0.0 $0.0 $0.0

GMACM 2002‐HLTV1 $0.0 $0.0 $0.0

GMACM 2002‐RP2 $0.0 $0.0 $0.0

GMACM 2005‐HE3 $98.0 $9.9 $107.9

RAMP 1999‐RS1 $0.0 $0.0 $0.0

RAMP 2001‐RS1 $0.0 $0.0 $0.0

RAMP 2001‐RS3 $0.0 $0.0 $0.0

RAMP 2002‐RP1 $0.0 $0.0 $0.0

RAMP 2002‐RS1 $0.0 $0.0 $0.0

RAMP 2002‐RS2 $0.0 $0.0 $0.0

RAMP 2002‐RS4 $0.0 $0.0 $0.0

RAMP 2002‐RS5 $0.0 $0.0 $0.0

RAMP 2002‐RS6 $0.0 $0.0 $0.0

RAMP 2002‐RS7 $0.0 $0.0 $0.0

RAMP 2002‐RZ4 $0.0 $0.0 $0.0

RAMP 2003‐RP1 $0.0 $0.0 $0.0

RAMP 2003‐RP2 $0.0 $0.0 $0.0

RAMP 2003‐RS1 $0.0 $0.0 $0.0

-44-

HIGHER RANGE

Trust

Actual Wrapped


+



=



RAMP 2003‐RS2 $0.0

$0.0

$0.0

RAMP 2003‐RS3 $0.0 $0.0 $0.0

RAMP 2003‐RS4 $0.0 $0.0 $0.0

RAMP 2003‐RS5 $0.0 $0.0 $0.0

RAMP 2003‐RS6 $0.0 $0.0 $0.0

RAMP 2003‐RS7 $0.0 $0.0 $0.0

RAMP 2003‐RS8 $0.0 $0.0 $0.0

RAMP 2003‐RS9 $0.0 $0.0 $0.0

RAMP 2003‐RZ1 $0.0 $0.0 $0.0

RAMP 2003‐RZ2 $0.0 $0.0 $0.0

RAMP 2003‐RZ3 $0.0 $0.0 $0.0

RAMP 2003‐RZ4 $0.0 $0.0 $0.0

RAMP 2003‐RZ5 $0.0 $0.0 $0.0

RASC 2002‐KS1 $1.9 $9.7 $11.6

RASC 2002‐KS4 $12.0 $0.6 $12.6

RASC 2002‐KS6 $5.5 $1.2 $6.7

RASC 2002‐KS8 $8.3 $11.1 $19.4

RASC 2003‐KS4 $0.0 $0.0 $0.0

RASC 2003‐KS5 $0.0 $0.0 $0.0

RASC 2003‐KS9 $0.0 $2.1 $2.1

RFMSII 1999‐HI1 $0.0 $0.0 $0.0

RFMSII 1999‐HI4 $0.0 $0.0 $0.0

RFMSII 1999‐HI6 $0.0 $0.0 $0.0

RFMSII 1999‐HI8 $0.0 $0.0 $0.0

RFMSII 2000‐HI1 $0.0 $0.0 $0.0

RFMSII 2000‐HI2 $0.0 $0.0 $0.0

RFMSII 2000‐HI3 $0.0 $0.0 $0.0

RFMSII 2000‐HI4 $0.0 $0.0 $0.0

RFMSII 2000‐HI5 $0.0 $0.0 $0.0

RFMSII 2000‐HL1 $0.0 $0.0 $0.0

RFMSII 2001‐HI1 $0.0 $0.0 $0.0

RFMSII 2001‐HI2 $0.0 $0.0 $0.0

RFMSII 2001‐HI3 $0.0 $0.0 $0.0

RFMSII 2001‐HI4 $0.0 $0.0 $0.0

RFMSII 2001‐HS2 $0.0 $0.0 $0.0

RFMSII 2002‐HI1 $0.0 $0.0 $0.0

RFMSII 2002‐HI2 $0.0 $0.0 $0.0

RFMSII 2002‐HI3 $0.0 $0.0 $0.0

-45-

HIGHER RANGE

Trust

Actual Wrapped


+



=



RFMSII 2003‐HI3 $0.0

$0.0

$0.0

RFMSII 2003‐HS4 $0.0 $0.0 $0.0

Total $180.8 $57.0 $237.8

111. As reflected in the charts above, the aggregate total potential lifetime

shortfall to the bonds wrapped by Ambac ranges from $225.6 million to $237.8 million, which

does not include amounts for indemnification of fees, expenses and interest.

2. Estimating Potential Liability Accounting for Settlement

112. The second step in my analysis was to account for potential settlement.

As discussed above, based on a review of recent monoline settlements, I observed that monoline

claims had recently settled for between 80% and 120% of the wrapped bond shortfalls.

113. I understand that in addition to the shortfall in principal or interest

payments, Ambac asserts claims against the Debtors for interest and indemnification of

expenses. However, I have not reviewed the basis for those additional claimed amounts and my

analysis of Ambac’s claims is limited to evaluating potential liability with respect to the shortfall

to the wrapped bonds. I determined that settlement of the bond shortfall claims might range

from 80% to 100% of potential lifetime shortfalls. However, I recognize that monoline claims

have settled for amounts in excess of the bond shortfalls in certain circumstances.

114. To establish the high end of my range of potential liability with respect to

the bond shortfalls, I used 100% of the potential lifetime shortfall to the bonds based on my

higher range assumptions. For the lower end of my range, I started with 100% of the potential

-46-

lifetime shortfall to the bonds based on my lower range assumptions, and then discounted that

number to 80%. The calculations are:

Total Bond Shortfall

X

Settlement Factor

=

Range of Debtors’ Potential Liability to

Ambac

Low Range

High Range

Low Range

High Range

Lower Range Total Potential Shortfall

$225.6

80% ‐

$180.5 $237.8 Higher Range Total Potential Shortfall

$237.8 ‐ 100%

D. Opinions Regarding the Assured Settlement.

115. The Plan seeks to resolve claims asserted by Assured relating to three

Debtor-sponsored RMBS trusts for which Assured provided insurance, promising to pay

investors to compensate for any shortfall in principal or interest payments in the wrapped bonds.

My analysis is limited to evaluating the claim for potential shortfall in principal and interest

payments on the Debtor-sponsored wrapped bonds. 22

116. As a result, the first step in my analysis of Assured’s damages was to

estimate Assured’s potential lifetime payments to cover shortfalls in principal and interest

payments to the wrapped bonds in the trusts wrapped by Assured.

1. Assured’s Estimated Aggregate Lifetime Shortfall Exposure.

117. I estimated Assured’s aggregate lifetime shortfall exposure using the same


tranche of bonds.

22 I understand that Assured asserted servicing claims with respect to additional Debtor-sponsored trusts. I have not analyzed those claims.

-47-


bonds in each trust insured by Assured as of September 2013. Second, I forecasted the potential

lifetime shortfall for the specific tranches of bonds wrapped by Assured, using the same low

range and high range assumptions that I used to calculate trust losses set forth above.


remaining lifetime shortfalls to bonds wrapped by Assured are:



(in millions)


GMACM 2004‐HE3 $18.7 $2.7 $6.6

HEL 2006‐HSA3 $51.8 $0.4 $0.5

RFMSI 2005‐S5 $0.0 $0.0 $0.0

Total $70.5 $3.1 $7.1




wrapped by Assured are:

LOWER RANGE

Trust

Actual Wrapped


+



=



GMACM 2004‐HE3 $18.7 $2.7 $21.4

HEL 2006‐HSA3 $51.8 $0.4 $52.2

RFMSI 2005‐S5 $0.0 $0.0 $0.0

Total $70.5 $3.1 $73.6

-48-

HIGHER RANGE

Trust

Actual Wrapped


+



=



GMACM 2004‐HE3 $18.7 $6.6 $25.3

HEL 2006‐HSA3 $51.8 $0.5 $52.3

RFMSI 2005‐S5 $0.0 $0.0 $0.0

Total $70.5 $7.1 $77.6


shortfall to the bonds wrapped by Assured ranges from $73.6 million to $77.6 million, which

does not include amounts for indemnification of fees, expenses and interest.






payments, Assured asserts claims against the Debtors for interest and indemnification of

expenses. However, I have not reviewed the basis for those additional claimed amounts and my

analysis of Assured’s claims is limited to evaluating potential liability with respect to the

shortfall to the wrapped bonds. I determined that settlement of the bond shortfall claims might

range from 80% to 100% of potential lifetime shortfalls. However, I recognize that monoline

claims have settled for amounts in excess of the bond shortfalls in certain circumstances.




-49-



Total Bond Shortfall (millions)

X

Settlement Factor

=

Range of Debtors’ Potential Liability to Assured (millions)

Low Range

High Range

Low Range

High Range


$73.6 80% ‐ $58.9 $77.6

Higher Range Total Potential Shortfall

$77.6 ‐ 100%

E. Opinions Regarding the FGIC Settlement.

125. The Plan seeks to resolve claims asserted by FGIC relating to 47 Debtor-

sponsored RMBS trusts for which FGIC provided insurance, promising to pay investors to

compensate for any shortfall in principal or interest payments in the wrapped bonds. My

analysis evaluates the claim for potential shortfall in principal and interest payments on the

Debtor-sponsored wrapped bonds.

126. I understand that FGIC commuted its policies and therefore has no future

exposure with respect to the Debtor-sponsored trusts. My analysis estimates the potential

exposure to FGIC in the absence of a settlement and commutation.23 As a result, the first step in

my analysis was to estimate FGIC’s potential exposure in the absence of the FGIC settlement

and commutation, starting with FGIC’s potential lifetime payments to cover shortfalls in

principal and interest payments to the wrapped bonds in those two trusts.

23 I understand that, as part of the FGIC settlement agreement, the trustees already released certain claims with respect to the FGIC-wrapped trusts, and the value of those releases was considered in the context of evaluating the benefits of settlement. For purposes of this Report, in connection with evaluation of the FGIC allowed claims, I limit my analysis to the Debtors’ potential liability to FGIC based on the lifetime estimated shortfall to FGIC-wrapped bonds.

-50-

1. FGIC’s Estimated Aggregate Lifetime Shortfall Exposure.

127. I estimated FGIC’s aggregate lifetime shortfall exposure using the same


tranche of bonds.


bonds in each trust insured by FGIC as of September 2013. Second, I forecasted the potential

lifetime shortfall for the specific tranches of bonds wrapped by FGIC, using the same low range



remaining lifetime shortfalls to bonds wrapped by FGIC are:

Trust

Actual Wrapped Bond

Losses (in millions)


(in millions)


GMACM 2005‐HE2 $3.2 $2.0 $2.7

GMACM 2006‐HE1 $111.0 $45.2 $75.1

GMACM 2006‐HE2 $62.8 $8.7 $9.6

GMACM 2006‐HE3 $105.0 $2.0 $3.7

GMACM 2006‐HE5 $0.0 $0.0 $0.0

GMACM 2007‐HE2 $161.0 $7.0 $9.2

GMACM 2001‐HE3 $1.2 $0.0 $0.2

GMACM 2001‐HE2 $1.3 $0.0 $0.0

GMACM 2002‐HE1 $0.0 $0.3 $1.0

GMACM 2002‐HE4 $0.0 $0.0 $0.0

GMACM 2003‐HE1 $0.0 $0.0 $0.0

GMACM 2003‐HE2 $0.0 $0.0 $0.0

GMACM 2004‐HE1 $1.6 $0.0 $5.4

GMACM 2004‐HE5 $0.0 $0.0 $0.0

GMACM 2005‐HE1 $37.7 $3.3 $8.7

GMACM 2006‐HLTV1 $0.0 $0.0 $0.0

GMACM 2004‐HLTV1 $0.0 $0.0 $0.0

RFMSII 2005‐HS1 $13.8 $48.5 $49.5

RFMSII 2005‐HS2 $24.9 $58.6 $59.5

RFMSII 2005‐HSA1 $32.7 $38.9 $39.2

-51-

Trust Actual

Wrapped Bond Losses


(in millions)

RFMSII Trust 2006‐HSA2 $51.1 $71.2 $76.4

RFMSII 2006‐HI2 $0.0 $31.5 $32.4

RFMSII 2006‐HI3 $0.0 $12.3 $13.0

RFMSII 2006‐HI4 $7.3 $43.2 $44.6

RFMSII 2006‐HI5 $1.2 $47.9 $49.2

RFMSII 2007‐HI1 $0.0 $54.0 $55.6

RAMP 2004‐RS7 $8.6 $17.1 $18.5

RAMP 2004‐RZ2 $4.8 $7.1 $7.5

RAMP 2005‐EFC7 $38.8 $11.9 $14.6

RAMP 2005‐NC1 $62.9 $17.0 $19.6

RAMP 2005‐RS9 $110.5 $14.5 $16.5

RASC 2001‐KS1 $0.0 $2.8 $3.4

RASC 2004‐KS7 $4.2 $8.0 $8.8

RASC 2004‐KS9 $4.0 $3.0 $3.6

RASC 2005‐EMX5 $39.5 $4.9 $5.9

RASC 2007‐EMX1 $45.0 $55.7 $61.2

RFMSII 2003‐HS1 $0.0 $0.0 $0.0

RFMSII 2005‐HI1 $0.0 $0.0 $0.0

RFMSII 2002‐HS3 $0.0 $0.0 $0.0

RFMSII 2003‐HS2 $0.0 $0.0 $0.0

RFMSII 2004‐HI2 $0.0 $0.0 $0.0

RFMSII 2004‐HI3 $0.0 $0.0 $0.0

RFMSII 2004‐HS1 $0.0 $0.0 $0.0

RFMSII 2004‐HS3 $0.0 $0.0 $0.0

RFMSI 2005‐S2 $0.0 $0.0 $0.0

RFMSI 2005‐S7 $1.1 $3.0 $3.2

RFMSII 2006‐HSA1 $74.9 $6.1 $6.7

Total $1,009.9 $625.8 $704.6




wrapped by FGIC are:

-52-

LOWER RANGE

Trust

Actual Wrapped


+



=



GMACM 2005‐HE2 $3.2

$2.0

$5.2

GMACM 2006‐HE1 $111.0 $45.2 $156.2

GMACM 2006‐HE2 $62.8 $8.7 $71.4

GMACM 2006‐HE3 $105.0 $2.0 $107.1

GMACM 2006‐HE5 $0.0 $0.0 $0.0

GMACM 2007‐HE2 $161.0 $7.0 $168.1

GMACM 2001‐HE3 $1.2 $0.0 $1.2

GMACM 2001‐HE2 $1.3 $0.0 $1.3

GMACM 2002‐HE1 $0.0 $0.3 $0.4

GMACM 2002‐HE4 $0.0 $0.0 $0.0

GMACM 2003‐HE1 $0.0 $0.0 $0.0

GMACM 2003‐HE2 $0.0 $0.0 $0.0

GMACM 2004‐HE1 $1.6 $0.0 $1.6

GMACM 2004‐HE5 $0.0 $0.0 $0.0

GMACM 2005‐HE1 $37.7 $3.3 $41.0

GMACM 2006‐HLTV1 $0.0 $0.0 $0.0

GMACM 2004‐HLTV1 $0.0 $0.0 $0.0

RFMSII 2005‐HS1 $13.8 $48.5 $62.3

RFMSII 2005‐HS2 $24.9 $58.6 $83.4

RFMSII 2005‐HSA1 $32.7 $38.9 $71.6

RFMSII Trust 2006‐HSA2

$51.1 $71.2 $122.3

RFMSII 2006‐HI2 $0.0 $31.5 $31.5

RFMSII 2006‐HI3 $0.0 $12.3 $12.3

RFMSII 2006‐HI4 $7.3 $43.2 $50.5

RFMSII 2006‐HI5 $1.2 $47.9 $49.1

RFMSII 2007‐HI1 $0.0 $54.0 $54.0

RAMP 2004‐RS7 $8.6 $17.1 $25.7

RAMP 2004‐RZ2 $4.8 $7.1 $11.9

RAMP 2005‐EFC7 $38.8 $11.9 $50.6

RAMP 2005‐NC1 $62.9 $17.0 $79.8

RAMP 2005‐RS9 $110.5 $14.5 $125.0

RASC 2001‐KS1 $0.0 $2.8 $2.8

RASC 2004‐KS7 $4.2 $8.0 $12.2

RASC 2004‐KS9 $4.0 $3.0 $7.0

RASC 2005‐EMX5 $39.5 $4.9 $44.4

RASC 2007‐EMX1 $45.0 $55.7 $100.8

RFMSII 2003‐HS1 $0.0 $0.0 $0.0

RFMSII 2005‐HI1 $0.0 $0.0 $0.0

-53-

LOWER RANGE

Trust

Actual Wrapped


+



=



RFMSII 2002‐HS3 $0.0

$0.0

$0.0

RFMSII 2003‐HS2 $0.0 $0.0 $0.0

RFMSII 2004‐HI2 $0.0 $0.0 $0.0

RFMSII 2004‐HI3 $0.0 $0.0 $0.0

RFMSII 2004‐HS1 $0.0 $0.0 $0.0

RFMSII 2004‐HS3 $0.0 $0.0 $0.0

RFMSI 2005‐S2 $0.0 $0.0 $0.0

RFMSI 2005‐S7 $1.1 $3.0 $4.1

RFMSII 2006‐HSA1 $74.9 $6.1 $81.0

Total $1,009.9 $625.8 $1,635.7

HIGHER RANGE

Trust

Actual Wrapped


+



=



GMACM 2005‐HE2 $3.2

$2.7

$5.8

GMACM 2006‐HE1 $111.0 $75.1 $186.1

GMACM 2006‐HE2 $62.8 $9.6 $72.4

GMACM 2006‐HE3 $105.0 $3.7 $108.7

GMACM 2006‐HE5 $0.0 $0.0 $0.0

GMACM 2007‐HE2 $161.0 $9.2 $170.3

GMACM 2001‐HE3 $1.2 $0.2 $1.3

GMACM 2001‐HE2 $1.3 $0.0 $1.3

GMACM 2002‐HE1 $0.0 $1.0 $1.0

GMACM 2002‐HE4 $0.0 $0.0 $0.0

GMACM 2003‐HE1 $0.0 $0.0 $0.0

GMACM 2003‐HE2 $0.0 $0.0 $0.0

GMACM 2004‐HE1 $1.6 $5.4 $7.0

GMACM 2004‐HE5 $0.0 $0.0 $0.0

GMACM 2005‐HE1 $37.7 $8.7 $46.4

GMACM 2006‐HLTV1 $0.0 $0.0 $0.0

GMACM 2004‐HLTV1 $0.0 $0.0 $0.0

RFMSII 2005‐HS1 $13.8 $49.5 $63.3

RFMSII 2005‐HS2 $24.9 $59.5 $84.3

-54-

HIGHER RANGE

Trust

Actual Wrapped


+



=



RFMSII 2005‐HSA1 $32.7

$39.2

$71.8

RFMSII Trust 2006‐HSA2 $51.1 $76.4 $127.5

RFMSII 2006‐HI2 $0.0 $32.4 $32.4

RFMSII 2006‐HI3 $0.0 $13.0 $13.0

RFMSII 2006‐HI4 $7.3 $44.6 $51.9

RFMSII 2006‐HI5 $1.2 $49.2 $50.4

RFMSII 2007‐HI1 $0.0 $55.6 $55.6

RAMP 2004‐RS7 $8.6 $18.5 $27.1

RAMP 2004‐RZ2 $4.8 $7.5 $12.3

RAMP 2005‐EFC7 $38.8 $14.6 $53.4

RAMP 2005‐NC1 $62.9 $19.6 $82.5

RAMP 2005‐RS9 $110.5 $16.5 $127.0

RASC 2001‐KS1 $0.0 $3.4 $3.4

RASC 2004‐KS7 $4.2 $8.8 $13.1

RASC 2004‐KS9 $4.0 $3.6 $7.6

RASC 2005‐EMX5 $39.5 $5.9 $45.4

RASC 2007‐EMX1 $45.0 $61.2 $106.3

RFMSII 2003‐HS1 $0.0 $0.0 $0.0

RFMSII 2005‐HI1 $0.0 $0.0 $0.0

RFMSII 2002‐HS3 $0.0 $0.0 $0.0

RFMSII 2003‐HS2 $0.0 $0.0 $0.0

RFMSII 2004‐HI2 $0.0 $0.0 $0.0

RFMSII 2004‐HI3 $0.0 $0.0 $0.0

RFMSII 2004‐HS1 $0.0 $0.0 $0.0

RFMSII 2004‐HS3 $0.0 $0.0 $0.0

RFMSI 2005‐S2 $0.0 $0.0 $0.0

RFMSI 2005‐S7 $1.1 $3.2 $4.3

RFMSII 2006‐HSA1 $74.9 $6.7 $81.6

Total $1,009.9 $704.6 $1,714.5


shortfall to the bonds wrapped by FGIC ranges from $1.64 billion to $1.71 billion, which does

not include amounts for indemnification of fees, expenses and interest.

-55-






payments, FGIC asserts claims against the Debtors for interest and indemnification of expenses.

However, I have not reviewed the basis for those additional claimed amounts and my analysis of

FGIC’s claims is limited to evaluating potential liability with respect to the shortfall to the

wrapped bonds. I determined that settlement of the bond shortfall claims might range from 80%

to 100% of potential lifetime shortfalls. However, I recognize that monoline claims have settled

for amounts in excess of the bond shortfalls in certain circumstances.



higher range assumptions. For the lower end of my range, I started with the 100% of the

potential liability with respect to the Bond shortfalls is 100% of the potential lifetime shortfall to

the bonds based on my lower range assumptions, and then discounted that number to 80%. The

calculations are:


X

Settlement Factor

=

Range of Debtors’ Potential Liability to

FGIC (millions)

Low Range

High Range

Low Range

High Range


$1,635.7

80% ‐

$1,308.6 $1,714.5 Higher Range Total Potential Shortfall

$1,714.5 ‐ 100%

-56-

F. Opinions Regarding the MBIA Settlement.

135. The Plan seeks to resolve claims asserted by MBIA relating to 24 Debtor-

sponsored RMBS trusts for which MBIA provided insurance, promising to pay investors to

compensate for any shortfall in principal or interest payments in the wrapped bonds.24

136. As a result, the first step in my analysis of MBIA’s damages was to

estimate MBIA’s potential lifetime payments to cover shortfalls in principal and interest

payments to the holders of the wrapped bonds in those 24 trusts.

1. MBIA’s Estimated Aggregate Lifetime Shortfall Exposure.

137. I estimated MBIA’s aggregate lifetime shortfall exposure using the same

Illumination and Intex cash flow models discussed above to simulate the cash flows to the

holders of each tranche of bonds.


bonds in each trust insured by MBIA as of September 2013. Second, I forecasted the potential

lifetime shortfall for the specific tranches of bonds wrapped by MBIA, using the same low range



remaining lifetime shortfalls to bonds wrapped by MBIA are:



(in millions)


GMAC 2000‐HE2 $3.1 $0.2 $0.3

GMAC 2000‐HE4 $2.2 $0.6 $0.8

GMAC 2002‐HE3 $0.0 $0.0 $0.0

24 In addition, the MBIA settlement resolves claims with respect to non-Debtor- sponsored trusts wrapped by MBIA. My analysis of the MBIA settlement does not address those additional claims. As noted above, my analysis of the non-Debtor sponsored trusts is ongoing.

-57-



(in millions)


GMAC 2003‐GH1 $0.0 $0.0 $0.0

GMAC 2004 Variable $0.0 $0.0 $0.0

GMAC 2004‐HE4 $36.1 $3.4 $8.3

GMAC 2004‐J1 $0.0 $0.0 $0.0

GMAC 2004‐J2 $0.0 $0.0 $0.0

GMAC 2006‐HE4 $235.6 $13.9 $30.8

GMAC 2007‐HE1 $157.7 $6.4 $8.1

RALI 2002‐QS15 $0.0 $0.0 $0.0

RALI 2003‐QS1 $0.0 $0.0 $0.0

RALI 2003‐QS8 $0.0 $0.0 $0.0

RALI 2003‐S13 $0.0 $0.0 $0.0

RALI 2003‐S4 $0.0 $0.0 $0.0

RFC 2004‐HS2 $1.7 $0.7 $0.7

RFC 2004‐S4 $0.0 $0.0 $0.0

RFC 2006‐HSA4 $110.5 $3.0 $3.5

RFC 2006‐HSA5 $107.9 $5.9 $6.4

RFC 2007‐HSA1 $235.5 $5.6 $7.9

RFC 2007‐HSA2 $778.3 $24.7 $27.3

RFC 2007‐HSA3 $429.6 $25.9 $28.1

RFMSI 2004‐SR1 $0.0 $0.0 $0.0

RFMSII 2003‐HS3 $0.0 $0.0 $0.0

Total $2,098.2 $90.3 $122.0




wrapped by MBIA are:

LOWER RANGE

Trust

Actual Wrapped


+



=



GMAC 2000‐HE2 $3.1

$0.2

$3.3

GMAC 2000‐HE4 $2.2 $0.6 $2.8

GMAC 2002‐HE3 $0.0 $0.0 $0.0

-58-

LOWER RANGE

Trust

Actual Wrapped


+



=



GMAC 2003‐GH1 $0.0

$0.0

$0.0

GMAC 2004 Variable $0.0 $0.0 $0.0

GMAC 2004‐HE4 $36.1 $3.4 $39.5

GMAC 2004‐J1 $0.0 $0.0 $0.0

GMAC 2004‐J2 $0.0 $0.0 $0.0

GMAC 2006‐HE4 $235.6 $13.9 $249.5

GMAC 2007‐HE1 $157.7 $6.4 $164.1

RALI 2002‐QS15 $0.0 $0.0 $0.0

RALI 2003‐QS1 $0.0 $0.0 $0.0

RALI 2003‐QS8 $0.0 $0.0 $0.0

RALI 2003‐S13 $0.0 $0.0 $0.0

RALI 2003‐S4 $0.0 $0.0 $0.0

RFC 2004‐HS2 $1.7 $0.7 $2.4

RFC 2004‐S4 $0.0 $0.0 $0.0

RFC 2006‐HSA4 $110.5 $3.0 $113.5

RFC 2006‐HSA5 $107.9 $5.9 $113.8

RFC 2007‐HSA1 $235.5 $5.6 $241.2

RFC 2007‐HSA2 $778.3 $24.7 $802.9

RFC 2007‐HSA3 $429.6 $25.9 $455.5

RFMSI 2004‐SR1 $0.0 $0.0 $0.0

RFMSII 2003‐HS3 $0.0 $0.0 $0.0

Total $2,098.2 $90.3 $2,188.5

HIGHER RANGE

Trust

Actual Wrapped


+



=



GMAC 2000‐HE2 $3.1 $0.3

$3.4

GMAC 2000‐HE4 $2.2 $0.8 $2.9

GMAC 2002‐HE3 $0.0 $0.0 $0.0

GMAC 2003‐GH1 $0.0 $0.0 $0.0

GMAC 2004 Variable $0.0 $0.0 $0.0

GMAC 2004‐HE4 $36.1 $8.3 $44.4

GMAC 2004‐J1 $0.0 $0.0 $0.0

GMAC 2004‐J2 $0.0 $0.0 $0.0

-59-

HIGHER RANGE

Trust

Actual Wrapped


+



=



GMAC 2006‐HE4 $235.6 $30.8

$266.4

GMAC 2007‐HE1 $157.7 $8.1 $165.8

RALI 2002‐QS15 $0.0 $0.0 $0.0

RALI 2003‐QS1 $0.0 $0.0 $0.0

RALI 2003‐QS8 $0.0 $0.0 $0.0

RALI 2003‐S13 $0.0 $0.0 $0.0

RALI 2003‐S4 $0.0 $0.0 $0.0

RFC 2004‐HS2 $1.7 $0.7 $2.4

RFC 2004‐S4 $0.0 $0.0 $0.0

RFC 2006‐HSA4 $110.5 $3.5 $114.0

RFC 2006‐HSA5 $107.9 $6.4 $114.2

RFC 2007‐HSA1 $235.5 $7.9 $243.4

RFC 2007‐HSA2 $778.3 $27.3 $805.5

RFC 2007‐HSA3 $429.6 $28.1 $457.7

RFMSI 2004‐SR1 $0.0 $0.0 $0.0

RFMSII 2003‐HS3 $0.0 $0.0 $0.0

Total $2,098.2 $122.0 $2,220.2


shortfall to the bonds wrapped by MBIA ranges from $2.19 billion to $2.22 billion, which does

not include amounts for indemnification of fees, expenses and interest.




claims had recently settled for between 80% and 120% of the wrapped Bond shortfalls.


payments, MBIA asserts claims against the Debtors for interest and indemnification of expenses.

However, I have not reviewed the basis for those additional claimed amounts and my analysis of

MBIA’s claims is limited to evaluating potential liability with respect to the shortfall to the

-60-

wrapped bonds. I determined that settlement of the bond shortfall claims might range from 80%

to 100% of potential lifetime shortfalls. However, I recognize that monoline claims have settled

for amounts in excess of the bond shortfalls in certain circumstances.







X

Settlement Factor

=

Range of Debtors’ Potential Liability to MBIA (millions)

Low Range

High Range

Low Range

High Range


$2,188.5

80% ‐

1,750.8 $2,220.2 Higher Range Total Potential Shortfall

$2,220.2 ‐ 100%

-61- ny-1113838

CONCLUSION

145. In summary, for this Report I utilized a trust level cash flow model process

regularly used by market participants and financial institutions to estimate the Debtors’ potential

exposure to the Debtor-sponsored trusts based on collateral losses to the trusts. I used the same

model to estimate potential exposure to Ambac, Assured, FGIC, and MBIA for shortfalls on

wrapped bonds in Debtor-sponsored trusts. I also estimated the Debtors’ potential exposure for

loans sold by the Debtors and included non-Debtor-sponsored trusts, based on the performance

of the similar loans in the Debtor-sponsored trusts. Accordingly, the methodologies that I used

in this Report are generally accepted in the industry as a sound means of estimating repurchase

exposure. Based on my analysis described above, it is my opinion that:

(a) a reasonable range of potential exposure to the trusts included in

the RBMS Settlement is $7.38 billion to $8.6 billion;

(b) a reasonable range of potential exposure to Ambac is $180.5

million to $237.8 million;

(c) a reasonable range of potential exposure to Assured is $58.9

million to $77.6 million;

(d) a reasonable range of potential exposure to FGIC is $1.3 billion to

$1.7 billon; and

(e) a reasonable range of potential exposure to MBIA is $1.75 billion

to $2.22 billion.

Dated: October 18, 2013

________________

Frank Sillman

ANNEX A

Curriculum Vitae of Frank Sillman

Manhattan Beach, CA 90266 Office Phone (310) 545-4548 Email: [email protected]

Summary of qualifications

27 years of senior executive management experience in sales, marketing, operations,

secondary marketing, loan securitization, risk mitigation and consulting. This included managing four multi-billion dollar, nationwide lending divisions and building innovative, de novo operating divisions. My responsibilities also included managing Secondary Marketing, Treasury, IT systems design/implementation, Accounting and Forecasting.

Experience in building and managing scalable, profitable and low cost businesses.

Helped design and implement risk based management models, risk mitigation strategies and integrated imaging systems.

Hired and developed a strong team of senior sales, marketing, operational, finance and risk management executives.

Professional experience

Fortace LLC 2008 to Present Los Angeles, CA

Managing Partner, Re-Underwriting & Consulting practice Co-Founded and actively manages the leading Residential Mortgage Re-Underwriting & Consulting firm that assists National banks, leading law firms and nationwide Mortgage Banking firms in (1) managing risks associated with the origination and servicing of residential mortgage loans and (2) the development and implementation of cost management programs & processes. We developed a proprietary workflow system (FAS) that allows us to securely investigate, document and provide supporting documentation to help our clients reduce their RMBS and mortgage related risk exposure, mitigate potential losses and exceed their internal ROI objectives. Additionally we created a Cost Management and Accountability program that allowed our customers to dramatically cut costs and efficiently manage those costs on an ongoing basis.

IMB Bank 1998 to 2008 Pasadena, CA

Executive Vice President, Mortgage Banking Group Responsible for the Mortgage Banking businesses for a top ten nationwide residential mortgage originator, with peak mortgage production of $100 Billion in 2006. Built and managed our Retail, Wholesale, Correspondent and Warehouse Lending businesses. This included the recruiting and retention of our nationwide sales and sales management organization, development of a highly automated Marketing team that was leveraged over all four divisions and a nationwide, low cost, decentralized operational infrastructure including processing, underwriting, funding and post closing. We originated and sold via loan securitizations and whole loan sales Agency, FHA, Private Label, Heloc, and Jumbo residential loan production. I also helped design and implement an industry leading POS pricing, ratelock, automated underwriting and imaging platform that improved customer satisfaction, improved operational efficiencies, enhanced risk mitigation and lowered costs.

TCM/AHC Mortgage 1992-1998 Los Angeles, CA

Senior Vice President, Loan Production and Secondary Marketing Responsible for retail mortgage banking loan production and secondary marketing for a medium size mortgage banker. Grew the mortgage banking production volumes from scratch in 1992 to $100 million plus per month. I helped grow the business from an

Curriculum Vitae of Frank Sillman

Manhattan Beach, CA 90266 Office Phone (310) 545-4548 Email: [email protected]

unprofitable Mortgage Brokerage model to a profitable Mortgage Banking model while building the operation infrastructure to properly manage the compliance, underwriting, closing and sales of closed loans into the secondary market.

Shearson Lehman Mortgage 1986-1992 Newport Beach, Ca

Senior Vice President, Secondary Marketing, Treasury and Warehouse Lending Managed all aspects of our Secondary Marketing activities including hedging, loan sales. Securitization and pricing. I managed our Treasury department which including daily Treasury operations, bank and commercial paper financing arrangements and applicable accounting. I also oversaw our $500,000,000 Warehouse Lending business.

Education University of California, San Diego - Bachelor of Arts

Expert Witness Experience

Litigation Consulting & Expert Witness services in following areas: (1) Mortgage

Origination process and controls, (2) Client Underwriting Guidelines, (3) Quality Control processes, (4) Residential Loan Securitization including Origination and Servicing Securitization Representations & Warranties.

Fact Witness Deposition strategy and preparation

Contributory Loss Analysis for Securitization Representation and Warranty losses

Expert Report - Residential Capital, LLC Chapter 11 Bankruptcy – 9019 1st Expert Declaration on $8.6 billion RMBS Trust Settlement Agreement filed 6/11/12 - UNITED STATES BANKRUPTCY COURT SOUTHERN DISTRICT OF NEW YORK - Case No. 12-12020 (MG)

Expert Report - Residential Capital, LLC Chapter 11 Bankruptcy – 9019 Supplemental Expert Declaration on $8.6 billion RMBS Trust Settlement Agreement filed 9/28/12 - UNITED STATES BANKRUPTCY COURT SOUTHERN DISTRICT OF NEW YORK - Case No. 12-12020 (MG)

Expert Report - Residential Capital, LLC Chapter 11 Bankruptcy – 9019 Expert Reply Declaration on RMBS Trust Settlement Agreement filed 01/15/13 - UNITED STATES BANKRUPTCY COURT SOUTHERN DISTRICT OF NEW YORK - Case No. 12-12020 (MG)

Expert Report - Residential Capital, LLC Chapter 11 Bankruptcy – 9019 Supplemental Declaration on RMBS Trust Settlement Agreement filed 02/19/13 - UNITED STATES BANKRUPTCY COURT SOUTHERN DISTRICT OF NEW YORK - Case No. 12-12020 (MG)

Expert Deposition Testimony - Residential Capital, LLC Chapter 11 Bankruptcy – RMBS Trust Settlement Agreement Expert Deposition given 11/20/12 (7 hours 20 minutes) - UNITED STATES BANKRUPTCY COURT SOUTHERN DISTRICT OF NEW YORK - Case No. 12-12020 (MG)

ANNEX B

Annex B

Documents Relied On by Frank Sillman

Document Title, Bates Numbers Document Date

Filings in These Chapter 11 Proceedings

Debtor’s Application for an Order Authorizing Employment and July 6, 2012 Retention of Fortace LLC as Consultant to the Debtors Nunc Pro Tunc (Docket No. 704) Joint Chapter 11 Plan Proposed by Residential Capital, LLC, et al. July 3, 2013 (Docket No. 4153) Plan Schedules 3G and 4G, filed in In re Residential Capital, LLC, et al. August 23, 2013 (Docket No. 4819-2) Disclosure Statement for the Joint Chapter 11 Plan Proposed By August 23, 2013 Residential Capital, LLC, et al. and the Official Committee of Unsecured Creditors, Exhibit 9, “Methodology for Calculation of Recognized Claims,” (Docket No. 4819-3) Other Court Papers

Assured Guaranty Municipal Corp. v. UBS Real Estate Securities, Inc., Case No. 12-01579 (HB) (JCF), (S.D.N.Y.) Assured Guaranty Municipal Corp. v. Flagstar Bank, FSB, et al., Case No. 11-2375 (JSR), (S.D.N.Y.) Assured Guaranty Municipal Corp. v. Flagstar Bank, 11-cv-2375 (JSR), 2013 U.S. Dist. Lexis 16682 (S.D.N.Y.) MBIA Insurance Corp.v. Countrywide Home Loans, Inc., Bank of America Corp., Index No. 602825/2008 (Supreme Court of New York) MBIA Insurance Corp.v. Flagstar ABS, LLC et al., Index No. 13-0262 (S.D.N.Y) Syncora Guarantee Inc.v. Countrywide Home Loans, Inc., Bank of America Corp., Index No. 650042/2009 (Supreme Court of New York) In re Lehman Brothers Holdings Inc., Case No. 08-13555 (JMP), Docket January 12, 2012 No. 24255, Declaration of Zachary Trumpp in Support of Debtors’ Motion Pursuant to Section 8.4 of the Modified Third Amended Joint Chapter 11 Plan

Articles, Books, and Other Documents

BasePoint Analytics LLC, White Paper, “Early Payment Default – 2007 Links to Fraud and Impact on Mortgage Lenders and Investment Banks” Federal Housing Finance Agency Office of Inspector General, “Evaluation September 27, 2011 of the Federal Housing Finance Agency’s Oversight of Freddie Mac’s Repurchase Settlement with Bank of America,” Evaluation Report No. EVL-2011-006 Gamaitoni, Chris, CompassPoint Research & Trading, LLC, “Mortgage August 17, 2010 Repurchases Part II: Private Label RMBS Investors Take Aim – Quantifying the Risks” Inside Mortgage Finance Publications, Analyzing GSE Mortgage 2006-2008 Buyback Demands: Lender Impact Varies Significantly – GSE Buyback Demand Activity Lin, Brian, RRMS Advisors, Opinion Concerning Contemplated Settlement June 7, 2011 Amount for 530 Trusts Performance Summary Reports by Issue Year for RALI QA Series ARM, RALI QH Series ARM, RALI QO Series ARM, RALI QS Series 15-Yr FRM, RALI QS Series 30-Yr FRM, RAMP RS Series ARM, RAMP RS Series FRM, RAMP RZ Series ARM, RAMP RZ Series FRM, RASC KS Series ARM, RASC KS Series FRM, RFMSI S Series 15-Yr FRM, RFMSI S Series 30-Yr FRM, RFMSI SA Series ARM, RFMSII HI Series 125 CLTV, RFMSII HS Series CES, RFMSII HS Series HELOC, RFMSII HSA Series CES, RFMSII HSA Series HELOC

Data Sources

AUS Loan Approval Formats

CoreLogic Loan Performance

Deep Skin Search

GMAC RGC Client Guide Bulletins

GMAC RFC Client Guides

Illumination

Intex Solutions, Inc.

MERS Link Report

S&P/Case-Shiller Home Price Index

U.S. Bureau of Labor Statistics

All Annexes to this Report, to the extent not listed above

ANNEX C

ResCap Proposed Settlement

Trust Level Model

Debtor Sponsored Trusts

Non‐Monoline Trusts Only

September 2013 Distribution

LOWER RANGE HIGHER RANGE

Shelf Series Curr OB (in 000's) Cum Loss To DateLower Range ‐ Est. Loss

Going Forward

Lower Range ‐ Est.

Loss TotalCDR Severity CRR Cum Loss To Date

Higher Range ‐ Est.

Loss Going Forward


Loss TotalCDR Severity CRR

GMACM 2004‐AR1 71,130,200 4,677,779 2,509,952 7,187,731 1.64% 57.80% 19.33% 4,677,779 3,111,182 7,788,961 1.65% 68.16% 18.22%

GMACM 2004‐AR2 72,769,100 5,681,241 265,791 5,947,032 0.59% 11.09% 12.73% 5,681,241 425,711 6,106,952 0.69% 14.70% 12.10%

GMACM 2004‐GH1 43,058,300 4,284,730 1,497,660 5,782,390 1.08% 46.91% 5.74% 4,284,730 1,639,062 5,923,792 1.12% 48.51% 5.34%

RFMSII 2004‐HI1 17,692,700 27,074,938 1,132,846 28,207,784 3.56% 110.97% 10.23% 27,074,938 1,303,245 28,378,183 3.75% 110.95% 9.18%

GMACM 2004‐J1 41,449,700 1,069,794 762,025 1,831,818 0.44% 57.47% 7.41% 1,069,794 820,586 1,890,380 0.47% 56.95% 6.92%

GMACM 2004‐J2 51,868,300 1,750,506 823,787 2,574,292 0.51% 47.12% 8.92% 1,750,506 886,027 2,636,533 0.53% 47.19% 8.29%

GMACM 2004‐J3 47,476,800 776,404 1,330,583 2,106,987 0.73% 60.14% 9.19% 776,404 1,393,284 2,169,687 0.74% 60.03% 8.47%

GMACM 2004‐J4 80,562,000 2,828,046 1,385,929 4,213,975 0.56% 50.04% 9.84% 2,828,046 1,448,287 4,276,333 0.56% 50.05% 9.16%

GMACM 2004‐J5 74,924,500 3,033,075 2,551,886 5,584,960 1.92% 32.90% 11.62% 3,033,075 2,703,654 5,736,728 1.92% 33.34% 10.81%

GMACM 2004‐J6 54,200,200 1,487,621 787,023 2,274,644 0.63% 94.18% 8.79% 1,487,621 796,223 2,283,844 0.61% 95.04% 8.29%

RAMP 2004‐KR1 73,407,300 133,258,572 2,334,501 135,593,073 1.53% 61.65% 4.27% 133,258,572 2,654,123 135,912,695 1.57% 62.42% 3.90%

RAMP 2004‐KR2 57,072,500 100,636,556 5,243,125 105,879,681 5.11% 61.34% 3.36% 100,636,556 6,969,134 107,605,690 5.45% 63.50% 3.03%

RASC 2004‐KS1 64,134,100 40,554,112 6,491,589 47,045,701 1.66% 80.42% 4.61% 40,554,112 6,890,004 47,444,116 1.71% 81.74% 4.27%

RASC 2004‐KS10 75,024,900 73,104,584 16,012,354 89,116,938 3.62% 82.01% 4.90% 73,104,584 17,023,420 90,128,004 3.71% 83.52% 4.52%

RASC 2004‐KS11 45,108,200 54,655,152 11,340,463 65,995,615 3.71% 82.76% 3.21% 54,655,152 12,016,865 66,672,017 3.88% 83.54% 3.00%

RASC 2004‐KS12 40,221,700 46,410,304 8,065,094 54,475,398 3.00% 71.73% 1.70% 46,410,304 8,770,531 55,180,835 3.23% 72.97% 1.56%

RASC 2004‐KS2 72,261,000 46,719,720 10,627,714 57,347,434 2.69% 73.87% 4.30% 46,719,720 11,382,994 58,102,714 2.75% 75.90% 3.93%

RASC 2004‐KS3 53,876,500 35,315,496 9,713,270 45,028,766 3.26% 81.36% 5.04% 35,315,496 10,488,742 45,804,238 3.36% 83.58% 4.59%

RASC 2004‐KS5 90,354,000 59,102,836 11,617,347 70,720,183 2.26% 79.66% 5.36% 59,102,836 12,576,761 71,679,597 2.33% 81.28% 4.87%

RASC 2004‐KS6 73,056,700 61,229,444 8,480,531 69,709,975 1.96% 74.96% 3.87% 61,229,444 9,073,459 70,302,903 2.04% 76.25% 3.57%

RASC 2004‐KS8 57,444,700 39,351,372 5,509,589 44,860,961 1.72% 76.48% 5.58% 39,351,372 6,028,931 45,380,303 1.79% 78.24% 5.11%

RFMSI 2004‐PS1 5,980,287 30 6,966 6,996 0.10% 60.00% 22.00% 30 22,454 22,484 0.25% 68.00% 15.00%

RALI 2004‐QA1 16,543,200 3,229,475 787,234 4,016,709 1.14% 53.24% 5.43% 3,229,475 838,679 4,068,154 1.14% 55.31% 5.08%

RALI 2004‐QA2 43,350,100 9,869,127 2,371,420 12,240,547 1.54% 51.73% 7.58% 9,869,127 2,552,818 12,421,945 1.62% 51.89% 7.19%

RALI 2004‐QA3 34,917,300 5,779,931 2,705,734 8,485,665 1.95% 82.69% 13.85% 5,779,931 2,795,933 8,575,863 1.91% 82.40% 12.77%

RALI 2004‐QA4 36,353,500 6,987,875 875,181 7,863,055 0.61% 46.61% 4.90% 6,987,875 906,234 7,894,109 0.63% 46.15% 4.68%

RALI 2004‐QA5 51,874,200 9,747,886 2,754,024 12,501,910 1.86% 41.11% 7.70% 9,747,886 3,002,431 12,750,317 1.97% 41.50% 7.25%

RALI 2004‐QA6 116,803,000 33,420,062 10,864,515 44,284,577 3.04% 45.41% 6.90% 33,420,062 11,581,717 45,001,779 3.14% 45.94% 6.46%

RALI 2004‐QS1 53,749,000 5,166,856 2,526,323 7,693,179 1.12% 60.44% 7.21% 5,166,856 2,684,456 7,851,312 1.15% 60.90% 6.69%

RALI 2004‐QS10 42,536,000 5,152,431 1,843,499 6,995,930 1.18% 51.73% 7.05% 5,152,431 1,975,695 7,128,125 1.23% 51.66% 6.43%

RALI 2004‐QS11 41,108,800 4,882,028 1,265,738 6,147,765 0.84% 52.27% 7.41% 4,882,028 1,404,882 6,286,910 0.89% 52.57% 6.80%

RALI 2004‐QS12 90,791,900 8,263,536 5,034,804 13,298,340 1.40% 56.57% 7.25% 8,263,536 5,433,780 13,697,316 1.44% 57.66% 6.65%

RALI 2004‐QS13 19,586,100 897,439 220,604 1,118,042 0.60% 69.93% 3.85% 897,439 227,579 1,125,017 0.63% 69.05% 3.55%

RALI 2004‐QS14 47,721,600 5,112,041 2,003,578 7,115,619 1.03% 54.95% 6.67% 5,112,041 2,177,279 7,289,320 1.07% 55.85% 6.09%

Page 1 of 13


Trust Level Model






Going Forward




Loss Going Forward



RALI 2004‐QS15 44,823,500 7,241,985 2,157,660 9,399,645 1.08% 58.29% 6.13% 7,241,985 2,303,516 9,545,501 1.09% 60.00% 5.59%

RALI 2004‐QS16 122,331,000 14,065,906 8,089,458 22,155,364 1.85% 48.35% 5.95% 14,065,906 8,762,708 22,828,614 1.91% 48.96% 5.34%

RALI 2004‐QS2 55,629,300 5,925,063 2,173,061 8,098,124 1.41% 47.20% 10.21% 5,925,063 2,451,198 8,376,261 1.56% 47.33% 9.74%

RALI 2004‐QS3 25,989,500 1,043,520 150,964 1,194,483 0.30% 83.50% 7.21% 1,043,520 201,448 1,244,968 0.24% 84.91% 6.35%

RALI 2004‐QS4 60,224,800 5,043,315 2,514,519 7,557,833 1.49% 56.54% 6.95% 5,043,315 2,789,365 7,832,680 1.55% 57.11% 6.33%

RALI 2004‐QS5 56,830,500 5,772,119 3,088,620 8,860,739 1.60% 62.57% 6.18% 5,772,119 3,254,167 9,026,286 1.62% 63.78% 5.62%

RALI 2004‐QS6 21,433,700 897,816 140,477 1,038,293 0.33% 82.01% 6.28% 897,816 149,316 1,047,132 0.35% 80.74% 5.90%

RALI 2004‐QS7 95,533,100 8,740,955 4,159,185 12,900,140 1.25% 47.27% 6.10% 8,740,955 4,476,714 13,217,669 1.25% 49.44% 5.63%

RALI 2004‐QS8 60,188,900 6,018,703 2,170,631 8,189,334 1.27% 37.74% 5.81% 6,018,703 2,314,869 8,333,571 1.29% 38.65% 5.37%

RALI 2004‐QS9 14,266,900 1,092,000 122,849 1,214,849 0.37% 91.46% 4.89% 1,092,000 127,460 1,219,460 0.39% 90.07% 4.56%

RFSC 2004‐RP1 22,246,340 21,298,410 1,498,742 22,797,152 1.50% 70.00% 7.00% 21,298,410 1,756,222 23,054,632 2.50% 75.00% 5.00%

RAMP 2004‐RS10 118,295,000 132,690,520 11,697,699 144,388,219 1.84% 72.14% 5.75% 132,690,520 12,881,048 145,571,568 1.93% 73.54% 5.20%

RAMP 2004‐RS11 94,168,700 92,311,920 16,125,928 108,437,848 3.12% 67.86% 4.07% 92,311,920 17,591,061 109,902,981 3.30% 69.28% 3.77%

RAMP 2004‐RS12 111,300,000 103,672,080 11,176,891 114,848,971 1.77% 72.79% 5.16% 103,672,080 11,971,205 115,643,285 1.82% 73.70% 4.70%

RAMP 2004‐RS2 60,517,400 63,638,240 5,241,262 68,879,502 1.82% 69.25% 6.63% 63,638,240 5,619,555 69,257,795 1.86% 70.78% 6.16%

RAMP 2004‐RS3 79,427,200 29,713,598 9,383,593 39,097,191 2.48% 66.26% 5.28% 29,713,598 10,315,758 40,029,356 2.61% 68.01% 4.83%

RAMP 2004‐RS4 90,811,600 81,105,512 12,434,079 93,539,591 2.28% 77.15% 4.50% 81105512 13,731,051 94836562.55 2.45% 78.55% 4.20%

RAMP 2004‐RS6 84,914,900 73,307,024 5,580,781 78,887,805 1.60% 74.54% 10.80% 73,307,024 6,085,464 79,392,488 1.67% 75.15% 10.14%

RAMP 2004‐RS8 101,475,000 77,282,208 14,103,086 91,385,294 2.68% 87.90% 9.08% 77,282,208 14,917,840 92,200,048 2.70% 88.56% 8.29%

RAMP 2004‐RZ1 44,934,500 23,841,754 3,353,855 27,195,609 1.39% 71.59% 6.06% 23,841,754 3,706,498 27,548,252 1.44% 73.70% 5.42%

RAMP 2004‐RZ3 38,288,300 21,006,242 2,299,533 23,305,775 1.46% 69.78% 10.49% 21,006,242 2,513,640 23,519,882 1.50% 70.13% 9.40%

RAMP 2004‐RZ4 36,602,300 19,763,292 3,638,905 23,402,197 1.95% 67.57% 5.94% 19,763,292 3,974,611 23,737,903 2.02% 68.55% 5.31%

RFMSI 2004‐S1 29,979,300 387,804 247,108 634,912 0.35% 40.33% 10.51% 387,804 266,669 654,473 0.35% 41.40% 9.83%

RFMSI 2004‐S3 14,648,700 24,459 21,159 45,619 0.07% 85.11% 2.59% 24,459 21,365 45,825 0.07% 80.02% 2.30%

RFMSI 2004‐S5 52,723,400 636,206 911,463 1,547,669 0.73% 34.88% 7.79% 636,206 989,351 1,625,557 0.78% 34.65% 7.38%

RFMSI 2004‐S6 62,354,400 1,300,639 524,557 1,825,196 0.87% 16.34% 10.31% 1,300,639 655,066 1,955,705 0.87% 19.85% 9.81%

RFMSI 2004‐S7 8,834,810 11,889 219,359 231,248 0.85% 119.10% 5.97% 11,889 222,356 234,245 0.84% 114.89% 5.45%

RFMSI 2004‐S8 33,929,300 1,071,380 1,518,877 2,590,257 1.43% 54.28% 10.44% 1,071,380 1,602,432 2,673,812 1.45% 54.16% 9.76%

RFMSI 2004‐S9 92,269,400 4,352,376 1,506,477 5,858,852 0.73% 36.00% 10.02% 4,352,376 1,620,226 5,972,601 0.76% 35.57% 9.25%

RFMSI 2004‐SA1 33,397,000 1,551,026 1,436,724 2,987,750 1.14% 63.69% 10.02% 1,551,026 1,450,280 3,001,305 1.09% 64.89% 9.33%

RAMP 2004‐SL1 49,934,200 5,123,360 3,504,509 8,627,868 2.45% 73.77% 0.49% 5,123,360 4,286,744 9,410,103 2.49% 74.72% 0.44%

RAMP 2004‐SL2 39,977,000 6,650,167 40,041 6,690,208 0.58% 114.61% 99.97% 6,650,167 42,649 6,692,816 0.62% 114.84% 99.97%

RAMP 2004‐SL3 21,811,200 2,423,027 15,119 2,438,146 0.63% 68.15% 99.97% 2,423,027 16,123 2,439,150 0.44% 59.55% 99.97%

RAMP 2004‐SL4 23,294,100 1,522,422 368,148 1,890,570 7.84% 119.95% 99.97% 1,522,422 401,555 1,923,977 7.73% 119.90% 99.97%

Page 2 of 13


Trust Level Model






Going Forward




Loss Going Forward



RAAC 2004‐SP1 21,315,800 5,450,655 1,400,472 6,851,127 1.62% 85.74% 3.18% 5,450,655 1,456,468 6,907,123 1.66% 85.98% 2.87%

RAAC 2004‐SP2 12,489,300 306,458 333,147 639,605 1.79% 80.13% 10.34% 306,458 475,656 782,114 1.91% 81.34% 10.19%

RAAC 2004‐SP3 51,543,000 2,895,378 3,418,117 6,313,495 1.56% 72.44% 8.08% 2,895,378 3,700,319 6,595,697 1.63% 72.99% 7.47%

GMACM 2005‐AA1 43,397,324 15,694,805 2,509,595 18,204,400 0.99% 73.10% 6.19% 15,694,805 2,789,551 18,484,356 0.46% 72.96% 6.23%

GMACM 2005‐AF1 78,584,528 10,518,431 3,671,445 14,189,876 1.35% 60.50% 4.72% 10,518,431 3,988,255 14,506,685 1.44% 61.73% 4.41%

GMACM 2005‐AF2 92,898,500 37,280,396 10,983,417 48,263,814 2.37% 69.48% 7.47% 37,280,396 12,166,787 49,447,183 2.55% 69.70% 6.92%

RASC 2005‐AHL1 59,881,800 90,548,952 21,622,982 112,171,934 4.91% 78.78% 1.45% 90,548,952 23,021,946 113,570,898 5.17% 80.20% 1.31%

RASC 2005‐AHL2 77,305,800 89,717,216 20,313,986 110,031,202 3.75% 72.69% 1.96% 89,717,216 21,908,936 111,626,152 3.97% 74.39% 1.81%

RASC 2005‐AHL3 87,603,200 110,649,984 22,151,704 132,801,688 3.34% 74.53% 1.84% 110,649,984 24,010,163 134,660,147 3.54% 76.30% 1.66%

GMACM 2005‐AR1 79,879,700 8,390,226 764,048 9,154,274 0.38% 37.98% 10.83% 8,390,226 1,034,456 9,424,682 0.46% 42.36% 10.65%

GMACM 2005‐AR2 106,252,000 16,511,864 1,889,101 18,400,965 0.65% 38.90% 7.97% 16,511,864 2,158,636 18,670,500 0.72% 38.69% 7.33%

GMACM 2005‐AR3 121,242,000 17,472,642 1,429,311 18,901,953 2.01% 11.48% 13.32% 17,472,642 1,758,779 19,231,421 2.17% 12.89% 12.90%

GMACM 2005‐AR4 93,353,000 15,779,360 2,930,942 18,710,302 2.45% 76.40% 21.25% 15,779,360 3,814,577 19,593,937 2.44% 76.77% 20.03%

GMACM 2005‐AR5 165,753,000 24,410,584 7,298,751 31,709,335 0.95% 73.27% 9.68% 24,410,584 7,550,812 31,961,396 0.96% 72.32% 9.08%

GMACM 2005‐AR6 189,951,000 31,817,008 7,381,582 39,198,590 1.58% 45.95% 12.60% 31,817,008 7,980,478 39,797,486 1.61% 46.76% 11.80%

RAMP 2005‐EFC1 120,931,000 129,732,152 24,310,433 154,042,585 3.98% 62.35% 3.60% 129,732,152 26,416,180 156,148,332 4.21% 63.82% 3.34%

RAMP 2005‐EFC2 82,897,300 99,621,624 16,555,046 116,176,670 3.31% 67.16% 2.56% 99,621,624 18,112,156 117,733,780 3.52% 69.02% 2.35%

RAMP 2005‐EFC3 104,512,000 108,962,064 26,050,866 135,012,930 5.44% 66.51% 3.63% 108,962,064 31,755,460 140,717,524 5.76% 68.69% 3.33%

RAMP 2005‐EFC4 109,555,000 127,554,616 28,616,675 156,171,291 5.60% 65.20% 4.20% 127,554,616 30,775,436 158,330,052 5.86% 66.53% 3.85%

RAMP 2005‐EFC5 120,392,000 123,280,328 26,840,517 150,120,845 3.61% 68.38% 2.63% 123,280,328 29,187,042 152,467,370 3.83% 70.03% 2.40%

RAMP 2005‐EFC6 110,428,000 126,799,432 25,228,226 152,027,658 3.64% 67.04% 2.19% 126,799,432 27,469,947 154,269,379 3.90% 68.52% 2.01%

RASC 2005‐EMX1 76,515,500 50,803,252 17,304,653 68,107,905 3.76% 80.01% 3.93% 50,803,252 18,465,034 69,268,286 3.93% 80.82% 3.57%

RASC 2005‐EMX2 80,822,000 63,786,520 23,402,656 87,189,176 4.79% 84.12% 3.92% 63,786,520 24,701,910 88,488,430 4.92% 85.38% 3.59%

RASC 2005‐EMX3 96,797,700 111,250,344 34,534,681 145,785,025 5.75% 82.90% 2.83% 111,250,344 36,469,231 147,719,575 5.89% 84.18% 2.45%

RASC 2005‐EMX4 81,595,800 102,258,680 27,464,118 129,722,798 5.16% 83.20% 2.72% 102,258,680 29,166,638 131,425,318 5.36% 84.42% 2.40%

RFMSII 2005‐HI2 40,483,100 40,735,232 4,395,183 45,130,415 3.55% 109.68% 10.07% 40,735,232 5,719,056 46,454,288 3.67% 109.68% 9.12%

RFMSII 2005‐HI3 50,566,400 42,996,004 6,038,741 49,034,745 2.87% 109.39% 8.89% 42,996,004 6,422,393 49,418,397 2.98% 109.38% 8.07%

GMACM 2005‐J1 150,252,000 11,658,749 4,561,316 16,220,065 0.95% 48.89% 9.66% 11,658,749 4,947,801 16,606,550 0.99% 48.86% 8.83%

RASC 2005‐KS1 57,899,100 59,461,776 11,612,403 71,074,179 4.27% 71.71% 3.79% 59,461,776 14,641,987 74,103,763 4.44% 73.39% 3.46%

RASC 2005‐KS10 213,167,000 265,082,016 54,148,242 319,230,258 4.18% 72.07% 2.55% 265,082,016 61,175,371 326,257,387 4.41% 73.22% 2.36%

RASC 2005‐KS11 248,301,000 283,284,352 67,393,710 350,678,062 4.50% 68.84% 2.32% 283,284,352 72,778,888 356,063,240 4.80% 70.39% 2.17%

RASC 2005‐KS12 204,726,000 251,591,936 47,239,459 298,831,395 3.63% 68.77% 2.23% 251,591,936 51,146,247 302,738,183 3.86% 70.34% 2.07%

RASC 2005‐KS2 43,146,600 50,069,324 8,796,811 58,866,135 3.77% 79.80% 6.28% 50,069,324 9,302,536 59,371,860 3.92% 80.24% 5.97%

RASC 2005‐KS3 42,914,800 42,793,288 6,367,234 49,160,522 2.64% 65.70% 3.77% 42,793,288 7,003,734 49,797,022 2.81% 67.13% 3.46%

Page 3 of 13


Trust Level Model






Going Forward




Loss Going Forward



RASC 2005‐KS4 44,996,200 45,376,136 9,244,726 54,620,862 4.02% 67.74% 2.88% 45,376,136 10,981,489 56,357,625 4.17% 69.85% 2.70%

RASC 2005‐KS5 43,999,800 49,249,000 10,469,744 59,718,744 3.78% 72.91% 2.82% 49,249,000 11,256,082 60,505,082 3.96% 74.66% 2.62%

RASC 2005‐KS6 77,537,000 81,880,192 13,998,839 95,879,031 3.80% 64.12% 2.24% 81,880,192 18,472,647 100,352,839 4.03% 65.95% 2.09%

RASC 2005‐KS7 58,618,500 58,379,920 14,575,320 72,955,240 4.21% 72.11% 3.28% 58,379,920 15,700,912 74,080,832 4.41% 73.61% 2.99%

RASC 2005‐KS8 186,595,000 186,829,760 46,773,188 233,602,948 3.91% 72.32% 2.43% 186,829,760 50,438,008 237,267,768 4.11% 74.22% 2.23%

RASC 2005‐KS9 85,554,600 78,648,304 19,982,330 98,630,634 3.86% 70.33% 2.86% 78,648,304 21,822,866 100,471,170 4.11% 71.97% 2.60%

RALI 2005‐QA1 47,988,200 19,594,604 1,151,504 20,746,108 1.19% 26.07% 6.24% 19,594,604 1,273,240 20,867,844 1.28% 26.25% 5.81%

RALI 2005‐QA10 178,706,000 75,980,488 20,208,468 96,188,956 2.79% 50.03% 5.30% 75,980,488 22,591,620 98,572,108 2.99% 51.44% 4.87%

RALI 2005‐QA11 153,695,000 65,176,612 16,303,719 81,480,331 5.56% 61.05% 9.34% 65,176,612 29,664,180 94,840,792 5.63% 61.72% 8.81%

RALI 2005‐QA12 80,807,000 36,210,288 11,443,403 47,653,691 2.85% 69.62% 7.24% 36,210,288 12,402,314 48,612,602 2.97% 70.03% 6.62%

RALI 2005‐QA13 168,366,000 90,168,016 24,516,810 114,684,826 5.38% 44.35% 7.34% 90,168,016 28,101,785 118,269,801 5.69% 45.31% 6.72%

RALI 2005‐QA2 100,380,000 28,536,422 8,525,922 37,062,344 2.29% 43.51% 5.11% 28,536,422 9,790,815 38,327,237 2.48% 45.61% 4.76%

RALI 2005‐QA3 97,515,000 32,828,676 12,618,604 45,447,280 3.44% 59.73% 3.44% 32,828,676 16,901,112 49,729,788 3.58% 61.09% 3.13%

RALI 2005‐QA4 131,573,000 38,587,508 11,213,261 49,800,769 3.94% 39.63% 3.39% 38,587,508 14,210,644 52,798,152 4.15% 41.81% 3.11%

RALI 2005‐QA5 10,595,300 8,403,523 299,430 8,702,953 1.26% 31.28% 7.72% 8,403,523 340,515 8,744,038 1.34% 32.72% 7.28%

RALI 2005‐QA6 123,975,000 42,395,160 8,391,603 50,786,763 1.89% 48.91% 6.82% 42,395,160 9,280,099 51,675,259 2.01% 49.48% 6.28%

RALI 2005‐QA7 161,925,000 41,583,120 13,107,263 54,690,383 2.69% 43.71% 3.68% 41,583,120 17,691,067 59,274,187 2.91% 46.18% 3.39%

RALI 2005‐QA8 134,355,000 48,617,872 18,366,301 66,984,173 3.31% 51.13% 4.30% 48,617,872 20,294,497 68,912,369 3.43% 53.82% 4.01%

RALI 2005‐QA9 192,114,000 78,550,000 17,575,962 96,125,962 2.35% 45.48% 4.94% 78,550,000 19,280,061 97,830,061 2.47% 46.57% 4.52%

RALI 2005‐QO1 151,577,000 91,492,320 24,950,771 116,443,091 3.06% 64.85% 4.14% 91,492,320 26,633,609 118,125,929 3.23% 65.49% 3.97%

RALI 2005‐QO2 87,290,600 66,217,468 15,870,522 82,087,990 3.98% 58.22% 4.45% 66,217,468 16,770,550 82,988,018 4.19% 58.23% 4.20%

RALI 2005‐QO3 104,330,000 93,061,992 16,097,869 109,159,861 3.45% 49.80% 3.45% 93,061,992 17,469,788 110,531,780 3.72% 50.57% 3.34%

RALI 2005‐QO4 171,241,000 154,346,784 30,078,842 184,425,626 3.26% 63.07% 3.84% 154,346,784 32,090,752 186,437,536 3.43% 63.93% 3.68%

RALI 2005‐QO5 274,604,000 271,638,656 43,278,663 314,917,319 2.73% 61.94% 3.58% 271,638,656 46,553,973 318,192,629 2.91% 62.63% 3.45%

RALI 2005‐QS1 60,732,100 6,317,292 2,328,455 8,645,747 1.27% 50.37% 6.48% 6,317,292 3,070,848 9,388,140 1.34% 50.11% 5.88%

RALI 2005‐QS10 83,401,700 16,841,712 7,892,340 24,734,052 2.15% 67.18% 8.46% 16,841,712 8,435,999 25,277,711 2.16% 68.63% 7.70%

RALI 2005‐QS11 68,584,700 13,352,807 4,754,478 18,107,285 1.59% 58.94% 7.14% 13,352,807 5,258,712 18,611,519 1.64% 60.35% 6.39%

RALI 2005‐QS12 167,267,000 35,396,628 16,090,236 51,486,864 2.55% 51.41% 6.49% 35,396,628 17,477,194 52,873,822 2.60% 52.59% 5.82%

RALI 2005‐QS13 226,287,000 47,740,092 27,815,877 75,555,969 2.80% 55.19% 5.16% 47,740,092 30,202,165 77,942,257 2.88% 56.67% 4.63%

RALI 2005‐QS14 162,316,000 36,436,456 5,182,883 41,619,339 1.44% 71.23% 4.09% 36,436,456 5,306,232 41,742,688 1.46% 71.36% 3.75%

RALI 2005‐QS15 140,032,000 38,810,980 13,391,564 52,202,544 1.87% 67.35% 6.53% 38,810,980 14,881,733 53,692,713 1.94% 68.94% 5.77%

RALI 2005‐QS16 144,263,000 37,338,056 16,266,995 53,605,051 2.27% 61.22% 5.10% 37,338,056 17,517,711 54,855,767 2.33% 62.49% 4.59%

RALI 2005‐QS17 171,629,000 57,454,800 20,807,270 78,262,070 2.63% 58.53% 5.55% 57,454,800 22,582,171 80,036,971 2.74% 59.38% 5.05%

RALI 2005‐QS2 60,018,400 6,908,540 3,559,473 10,468,013 1.56% 50.78% 6.50% 6,908,540 3,813,463 10,722,002 1.60% 51.71% 5.97%

Page 4 of 13


Trust Level Model






Going Forward




Loss Going Forward



RALI 2005‐QS3 121,543,000 16,516,756 9,219,681 25,736,437 1.81% 61.99% 6.10% 16,516,756 11,148,005 27,664,761 1.89% 62.83% 5.52%

RALI 2005‐QS4 63,100,200 8,997,134 3,346,524 12,343,658 1.78% 51.51% 5.96% 8,997,134 4,506,358 13,503,492 1.84% 52.03% 5.34%

RALI 2005‐QS6 77,996,700 14,549,124 4,525,332 19,074,456 1.40% 59.18% 7.72% 14,549,124 4,902,785 19,451,909 1.44% 60.13% 7.03%

RALI 2005‐QS7 112,233,000 19,669,164 10,528,216 30,197,380 2.75% 47.09% 6.15% 19,669,164 11,601,839 31,271,003 2.87% 48.45% 5.62%

RALI 2005‐QS8 18,857,500 1,126,144 406,263 1,532,407 0.87% 86.15% 5.85% 1,126,144 426,599 1,552,742 0.90% 86.48% 5.40%

RALI 2005‐QS9 112,687,000 20,427,054 8,539,071 28,966,125 1.84% 58.75% 7.38% 20,427,054 9,326,729 29,753,783 1.89% 59.56% 6.57%

RAAC 2005‐RP1 49,275,460 40,328,883 2,600,866.78 42,929,749 0.75% 115.00% 8.00% 40,328,883 6,095,685.02 46,424,568 1.50% 125.00% 6.00%

RAAC 2005‐RP2 63,926,902 36,422,338 6,888,978.80 43,311,317 2.00% 80.00% 5.00% 36,422,338 11,171,980.87 47,594,319 3.00% 86.00% 4.00%

RAAC 2005‐RP3 61,442,327 48,812,005 8,208,847.30 57,020,853 2.00% 90.00% 5.00% 48,812,005 13,761,264.34 62,573,270 3.00% 100.00% 4.00%

RAMP 2005‐RS1 119,135,000 108,060,368 15,413,269.43 123,473,637 2.71% 67.41% 5.90% 108,060,368 16,665,866.19 124,726,234 2.77% 68.91% 5.29%

RAMP 2005‐RS2 76,812,500 86,405,144 12,503,454.19 98,908,598 2.59% 72.27% 3.64% 86,405,144 13,608,072.34 100,013,216 2.74% 73.45% 3.35%

RAMP 2005‐RS3 116,569,000 81,556,952 13,009,787.87 94,566,740 2.18% 68.04% 5.75% 81,556,952 14,351,578.66 95,908,531 2.32% 68.79% 5.26%

RAMP 2005‐RS4 77,809,900 68,499,384 14,347,234.78 82,846,619 3.47% 72.86% 5.45% 68,499,384 15,506,836.42 84,006,220 3.61% 74.10% 4.99%

RAMP 2005‐RS5 71,906,600 60,394,280 7,477,624.91 67,871,905 1.87% 73.35% 6.02% 60,394,280 8,175,356.28 68,569,636 1.99% 74.31% 5.64%

RAMP 2005‐RS6 158,462,000 154,750,976 28,063,271.17 182,814,247 2.98% 74.45% 4.88% 154,750,976 30,366,787.09 185,117,763 3.11% 75.29% 4.38%

RAMP 2005‐RS7 99,680,700 64,248,840 18,979,563.69 83,228,404 3.54% 67.29% 4.53% 64,248,840 20,789,695.43 85,038,535 3.74% 68.61% 4.09%

RAMP 2005‐RS8 133,600,000 110,883,920 25,141,049.02 136,024,969 3.40% 65.81% 3.81% 110,883,920 27,510,911.45 138,394,831 3.60% 67.26% 3.49%

RAMP 2005‐RZ1 29,751,100 18,001,520 2,468,632.55 20,470,153 1.57% 75.79% 7.65% 18,001,520 2,692,271.64 20,693,792 1.63% 76.39% 6.93%

RAMP 2005‐RZ2 52,815,700 49,061,136 7,084,538.39 56,145,674 4.29% 67.60% 8.40% 49,061,136 9,329,213.01 58,390,349 4.37% 68.83% 7.58%

RAMP 2005‐RZ3 69,017,300 66,842,724 11,594,913.44 78,437,637 2.70% 74.09% 4.25% 66,842,724 12,422,257.80 79,264,982 2.79% 75.12% 3.82%

RAMP 2005‐RZ4 86,481,900 87,630,912 13,068,332.18 100,699,244 2.96% 68.63% 6.05% 87,630,912 14,263,321.96 101,894,234 3.07% 69.91% 5.45%

RFMSI 2005‐S1 67,904,700 4,529,274 2,059,051.11 6,588,325 0.92% 52.34% 9.65% 4,529,274 2,125,972.68 6,655,247 0.92% 51.82% 8.98%

RFMSI 2005‐S3 15,126,800 117,860 113,407.78 231,268 0.43% 66.02% 7.49% 117,860 123,837.94 241,698 0.44% 69.15% 6.91%

RFMSI 2005‐S4 54,471,900 5,355,274 1,650,442.50 7,005,716 1.27% 42.56% 11.66% 5,355,274 1,754,288.27 7,109,562 1.29% 42.19% 10.71%

RFMSI 2005‐S6 94,730,700 5,918,915 1,701,824.69 7,620,740 1.11% 26.03% 10.23% 5,918,915 1,872,399.65 7,791,315 1.14% 26.30% 9.26%

RFMSI 2005‐S8 101,271,000 10,350,910 3,306,923.46 13,657,833 1.09% 44.60% 9.13% 10,350,910 3,888,095.44 14,239,005 1.20% 45.46% 8.26%

RFMSI 2005‐S9 94,515,200 16,396,569 2,643,061.75 19,039,631 1.36% 42.35% 8.97% 16,396,569 3,456,038.62 19,852,608 1.43% 42.89% 8.16%

RFMSI 2005‐SA1 42,935,500 5,297,655 1,048,756.19 6,346,411 1.56% 40.37% 6.83% 5,297,655 1,308,875.17 6,606,530 1.57% 42.67% 6.61%

RFMSI 2005‐SA2 125,388,000 17,543,368 4,238,928.37 21,782,296 1.30% 37.01% 7.80% 17,543,368 5,089,309.67 22,632,678 1.36% 41.15% 7.18%

RFMSI 2005‐SA3 162,767,000 27,648,460 3,213,149.19 30,861,609 0.74% 43.70% 10.72% 27,648,460 3,541,176.87 31,189,637 0.77% 44.31% 10.09%

RFMSI 2005‐SA4 260,666,000 42,539,240 5,614,998.38 48,154,238 0.82% 45.29% 11.39% 42,539,240 6,225,272.45 48,764,512 0.84% 46.56% 10.56%

RFMSI 2005‐SA5 110,919,000 26,268,070 6,601,646.30 32,869,716 2.70% 46.48% 9.49% 26,268,070 8,178,001.94 34,446,072 2.88% 47.33% 8.72%

RAMP 2005‐SL1 50,936,700 10,001,677 284.23 10,001,961 0.00% 120.00% 5.23% 10,001,677 284.23 10,001,961 0.00% 120.00% 4.97%

RAMP 2005‐SL2 29,122,000 6,526,876 ‐ 6,526,876 0.00% 0.00% 14.97% 6,526,876 ‐ 6,526,876 0.00% 0.00% 14.62%

Page 5 of 13


Trust Level Model






Going Forward




Loss Going Forward



RAAC 2005‐SP1 97,612,900 2,005,153 3,929,143.77 5,934,297 2.82% 54.66% 13.13% 2,005,153 4,369,170.70 6,374,323 2.79% 55.68% 12.31%

RAAC 2005‐SP2 86,620,200 27,690,902 11,225,821.28 38,916,723 2.41% 95.56% 2.27% 27,690,902 13,140,374.92 40,831,277 2.50% 96.26% 2.07%

RAAC 2005‐SP3 62,296,700 32,805,006 11,080,209.37 43,885,215 3.30% 86.73% 6.53% 32,805,006 11,795,375.71 44,600,382 3.41% 87.41% 6.02%

GMACM 2006‐AR1 167,861,000 41,549,824 9,025,553.63 50,575,378 1.83% 41.35% 8.27% 41,549,824 10,069,728.13 51,619,552 1.96% 41.98% 7.71%

GMACM 2006‐AR2 113,332,000 33,769,368 5,513,836.12 39,283,204 4.81% 53.56% 10.25% 33,769,368 7,160,254.89 40,929,623 5.04% 55.53% 9.71%

RAMP 2006‐EFC1 105,673,000 131,527,288 24,804,718.33 156,332,006 4.04% 65.08% 2.57% 131,527,288 26,856,994.84 158,384,283 4.28% 66.57% 2.36%

RAMP 2006‐EFC2 119,431,000 113,343,304 31,898,098.82 145,241,403 3.67% 73.76% 2.53% 113,343,304 34,677,640.35 148,020,944 3.91% 75.42% 2.33%

RASC 2006‐EMX1 76,411,700 103,261,992 26,022,508.12 129,284,500 4.21% 89.66% 1.58% 103,261,992 27,439,028.08 130,701,020 4.35% 90.71% 1.33%

RASC 2006‐EMX2 99,740,000 154,469,824 39,380,096.33 193,849,920 5.73% 86.28% 2.16% 154,469,824 41,426,749.41 195,896,573 5.93% 87.29% 1.89%

RASC 2006‐EMX3 148,997,000 248,248,192 64,452,974.86 312,701,167 6.39% 89.97% 2.64% 248,248,192 67,185,085.69 315,433,278 6.56% 90.75% 2.37%

RASC 2006‐EMX4 138,656,000 231,413,104 55,874,623.77 287,287,728 5.23% 89.15% 1.92% 231,413,104 58,601,959.69 290,015,064 5.39% 90.10% 1.67%

RASC 2006‐EMX5 129,272,000 212,265,456 47,994,865.64 260,260,322 4.54% 87.98% 1.90% 212,265,456 50,700,812.69 262,966,269 4.72% 89.15% 1.67%

RASC 2006‐EMX6 150,715,000 230,764,512 56,193,921.69 286,958,434 4.45% 89.76% 2.18% 230,764,512 59,740,004.05 290,504,516 4.65% 90.79% 1.91%

RASC 2006‐EMX7 122,508,000 191,037,456 54,838,415.24 245,875,871 6.35% 93.73% 3.65% 191,037,456 57,866,935.09 248,904,391 6.58% 94.54% 3.28%

RASC 2006‐EMX8 182,523,000 293,947,328 98,291,205.00 392,238,533 6.04% 98.39% 1.47% 293,947,328 102,053,195.26 396,000,523 6.18% 99.25% 1.25%

RASC 2006‐EMX9 202,529,000 314,238,976 108,753,051.28 422,992,027 6.61% 95.14% 1.68% 314,238,976 113,705,960.54 427,944,937 6.81% 96.31% 1.45%

RFMSII 2006‐HI1 57,295,600 51,822,240 7,793,874.56 59,616,115 3.15% 109.07% 8.61% 51,822,240 8,258,043.37 60,080,283 3.25% 109.09% 7.80%

GMACM 2006‐J1 154,537,000 21,031,638 10,032,389.02 31,064,027 2.13% 46.68% 8.77% 21,031,638 10,722,109.85 31,753,748 2.15% 47.35% 7.99%

RASC 2006‐KS1 161,535,000 181,602,512 41,608,461.66 223,210,974 4.05% 71.34% 2.45% 181,602,512 44,888,248.14 226,490,760 4.29% 72.80% 2.26%

RASC 2006‐KS2 186,041,000 225,987,888 46,915,521.85 272,903,410 3.96% 68.77% 2.09% 225,987,888 50,980,277.41 276,968,165 4.22% 70.49% 1.93%

RASC 2006‐KS3 207,859,000 295,889,728 54,348,661.45 350,238,389 4.41% 68.87% 2.96% 295,889,728 58,692,774.37 354,582,502 4.67% 70.16% 2.71%

RASC 2006‐KS4 154,981,000 179,118,544 30,386,815.29 209,505,359 2.99% 72.86% 3.58% 179,118,544 33,074,216.54 212,192,761 3.21% 73.62% 3.32%

RASC 2006‐KS5 179,152,000 195,692,576 44,314,070.74 240,006,647 4.01% 70.91% 3.22% 195,692,576 47,946,344.92 243,638,921 4.23% 72.68% 2.98%

RASC 2006‐KS6 148,413,000 154,937,872 35,658,316.29 190,596,188 3.29% 75.68% 2.58% 154,937,872 38,471,304.43 193,409,176 3.48% 77.22% 2.41%

RASC 2006‐KS7 150,402,000 157,809,344 35,248,182.76 193,057,527 3.63% 68.54% 2.68% 157,809,344 38,312,670.34 196,122,014 3.88% 69.95% 2.48%

RASC 2006‐KS8 159,988,000 170,108,208 46,022,787.43 216,130,995 4.34% 72.20% 2.64% 170,108,208 49,573,737.21 219,681,945 4.59% 73.74% 2.44%

RASC 2006‐KS9 351,699,000 437,484,384 110,516,764.79 548,001,149 4.43% 75.82% 2.57% 437,484,384 118,829,065.83 556,313,450 4.68% 77.37% 2.36%

RAMP 2006‐NC1 91,619,500 132,652,648 21,983,659.53 154,636,308 4.56% 65.80% 3.91% 132,652,648 23,892,055.99 156,544,704 4.85% 66.98% 3.59%

RAMP 2006‐NC2 160,407,000 195,440,320 40,624,129.70 236,064,450 3.65% 73.50% 2.69% 195,440,320 43,839,681.20 239,280,001 3.87% 74.57% 2.45%

RAMP 2006‐NC3 115,964,000 138,252,672 36,017,154.96 174,269,827 4.49% 77.65% 2.61% 138,252,672 38,349,631.20 176,602,303 4.68% 78.54% 2.31%

RALI 2006‐QA1 169,418,000 108,642,664 10,994,798.34 119,637,462 2.47% 33.81% 6.43% 108,642,664 12,867,329.36 121,509,993 2.75% 34.61% 5.79%

RALI 2006‐QA10 112,308,000 96,227,384 21,978,540.57 118,205,925 3.95% 56.11% 4.03% 96,227,384 23,696,527.27 119,923,911 4.17% 57.21% 3.78%

RALI 2006‐QA11 111,893,000 103,249,992 14,971,373.43 118,221,365 3.10% 46.34% 4.27% 103,249,992 16,643,961.17 119,893,953 3.34% 47.62% 4.01%

RALI 2006‐QA2 119,201,000 76,290,344 11,076,564.31 87,366,908 1.99% 41.93% 2.44% 76,290,344 12,367,008.47 88,657,352 2.16% 43.04% 2.26%

Page 6 of 13


Trust Level Model






Going Forward




Loss Going Forward



RALI 2006‐QA3 107,832,000 79,872,008 12,184,250.14 92,056,258 2.90% 42.58% 4.24% 79,872,008 13,572,986.23 93,444,994 3.07% 44.18% 3.91%

RALI 2006‐QA4 84,136,300 64,069,456 11,746,782.24 75,816,238 2.61% 57.22% 4.04% 64,069,456 12,753,170.22 76,822,626 2.75% 58.23% 3.75%

RALI 2006‐QA5 180,920,000 134,513,328 31,100,617.84 165,613,946 3.79% 55.83% 4.89% 134,513,328 33,661,413.81 168,174,742 3.97% 56.92% 4.52%

RALI 2006‐QA6 157,120,000 157,140,672 24,618,582.84 181,759,255 3.24% 50.73% 3.45% 157,140,672 26,876,632.82 184,017,305 3.44% 51.83% 3.16%

RALI 2006‐QA7 162,266,000 158,975,872 26,413,594.48 185,389,466 4.93% 44.14% 5.59% 158,975,872 29,144,596.11 188,120,468 5.21% 45.61% 5.17%

RALI 2006‐QA8 220,316,000 218,725,584 31,563,682.04 250,289,266 2.78% 52.56% 3.66% 218,725,584 34,711,601.54 253,437,186 2.96% 53.98% 3.42%

RALI 2006‐QA9 93,397,800 70,440,208 15,584,628.23 86,024,836 4.29% 47.62% 4.52% 70,440,208 17,255,411.08 87,695,619 4.56% 49.33% 4.20%

RALI 2006‐QO1 207,064,000 205,811,424 49,579,983.01 255,391,407 3.78% 82.38% 4.90% 205,811,424 51,609,108.80 257,420,533 3.81% 83.67% 4.61%

RALI 2006‐QO10 302,866,000 297,754,720 76,237,613.44 373,992,333 5.07% 58.29% 3.57% 297,754,720 81,540,579.54 379,295,300 5.32% 59.46% 3.35%

RALI 2006‐QO2 138,880,000 159,708,656 25,921,381.93 185,630,038 3.61% 61.83% 4.56% 159,708,656 27,763,223.49 187,471,879 3.80% 62.69% 4.33%

RALI 2006‐QO3 180,685,000 166,210,208 33,160,900.48 199,371,108 3.53% 56.79% 3.65% 166,210,208 35,697,972.44 201,908,180 3.78% 57.53% 3.52%

RALI 2006‐QO5 315,201,000 306,931,776 75,196,494.23 382,128,270 4.46% 62.52% 3.63% 306,931,776 79,146,117.44 386,077,893 4.66% 62.98% 3.44%

RALI 2006‐QO6 384,094,000 376,267,360 85,445,935.37 461,713,295 4.43% 55.85% 3.33% 376,267,360 92,088,304.22 468,355,664 4.70% 56.93% 3.13%

RALI 2006‐QO7 503,856,000 462,720,096 127,152,792.94 589,872,889 4.93% 63.37% 4.10% 462,720,096 135,153,224.66 597,873,321 5.17% 64.16% 3.86%

RALI 2006‐QO8 425,437,000 409,652,384 91,508,096.63 501,160,481 3.84% 61.42% 3.67% 409,652,384 96,702,852.11 506,355,236 4.03% 61.89% 3.50%

RALI 2006‐QO9 191,938,000 279,543,437 34,985,518.44 314,528,955 4.18% 43.54% 1.40% 279,543,437 37,134,282.28 316,677,719 4.33% 44.77% 1.29%

RALI 2006‐QS1 104,112,000 32,860,416 9,155,938.95 42,016,355 1.90% 52.85% 4.62% 32,860,416 9,999,910.46 42,860,326 2.04% 52.62% 4.18%

RALI 2006‐QS10 182,088,000 68,550,712 31,442,045.15 99,992,757 2.98% 68.85% 4.72% 68,550,712 33,387,546.32 101,938,258 3.06% 69.41% 4.26%

RALI 2006‐QS11 259,514,000 108,321,120 49,822,666.69 158,143,787 3.54% 65.83% 4.66% 108,321,120 53,600,409.15 161,921,529 3.66% 66.82% 4.18%

RALI 2006‐QS12 167,521,000 93,753,704 29,928,563.86 123,682,268 3.37% 72.25% 6.88% 93,753,704 32,089,736.17 125,843,440 3.40% 73.93% 6.28%

RALI 2006‐QS13 200,154,000 84,866,872 31,764,038.00 116,630,910 3.14% 61.01% 4.93% 84,866,872 34,692,652.86 119,559,525 3.31% 62.21% 4.56%

RALI 2006‐QS14 231,506,000 122,005,152 44,500,825.57 166,505,978 3.77% 66.52% 5.65% 122,005,152 47,977,858.93 169,983,011 3.89% 67.57% 5.13%

RALI 2006‐QS15 173,665,000 90,586,080 36,399,679.65 126,985,760 3.96% 66.91% 5.00% 90,586,080 39,254,489.71 129,840,570 4.07% 68.41% 4.50%

RALI 2006‐QS16 262,746,000 121,998,320 55,073,323.86 177,071,644 3.95% 67.09% 4.92% 121,998,320 59,401,932.92 181,400,253 4.09% 68.27% 4.44%

RALI 2006‐QS17 189,373,000 92,826,968 43,059,309.65 135,886,278 4.58% 67.17% 5.39% 92,826,968 45,827,887.62 138,654,856 4.66% 68.40% 4.89%

RALI 2006‐QS18 399,059,000 221,006,848 72,603,797.15 293,610,645 3.00% 71.37% 4.75% 221,006,848 77,733,977.42 298,740,825 3.11% 72.19% 4.33%

RALI 2006‐QS2 273,557,000 89,840,944 40,976,788.92 130,817,733 3.33% 58.84% 5.37% 89,840,944 44,381,312.53 134,222,257 3.42% 60.14% 4.83%

RALI 2006‐QS3 281,278,000 135,170,320 39,911,970.07 175,082,290 3.25% 61.94% 6.86% 135,170,320 43,417,228.58 178,587,549 3.37% 62.99% 6.24%

RALI 2006‐QS4 241,118,000 101,682,368 36,421,815.58 138,104,184 2.77% 65.78% 4.99% 101,682,368 39,717,172.26 141,399,540 2.88% 67.26% 4.48%

RALI 2006‐QS5 220,486,000 104,462,496 34,183,179.32 138,645,675 2.92% 63.97% 4.92% 104,462,496 36,660,690.79 141,123,187 2.99% 65.12% 4.46%

RALI 2006‐QS6 259,260,000 126,188,336 50,495,836.70 176,684,173 3.85% 64.69% 5.17% 126,188,336 54,298,068.14 180,486,404 3.98% 65.54% 4.65%

RALI 2006‐QS7 159,391,000 85,710,288 30,575,528.10 116,285,816 3.66% 62.09% 4.00% 85,710,288 32,855,523.02 118,565,811 3.77% 63.61% 3.63%

RALI 2006‐QS8 306,202,000 151,491,712 63,733,463.97 215,225,176 3.93% 69.26% 5.29% 151,491,712 67,940,100.93 219,431,813 4.02% 70.22% 4.79%

RALI 2006‐QS9 161,707,000 89,479,344 37,088,955.96 126,568,300 5.54% 58.33% 5.21% 89,479,344 39,510,970.72 128,990,315 5.64% 59.69% 4.75%

Page 7 of 13


Trust Level Model






Going Forward




Loss Going Forward



RAAC 2006‐RP1 73,894,256 54,513,153 3,737,774.28 58,250,927 0.75% 95.00% 7.00% 54,513,153 8,987,458.93 63,500,612 1.50% 105.00% 5.00%

RAAC 2006‐RP2 84,118,192 95,656,452 16,434,700.15 112,091,152 3.00% 90.00% 5.00% 95,656,452 23,756,250.75 119,412,703 4.00% 100.00% 4.50%

RAAC 2006‐RP3 76,299,251 103,009,997 23,762,741.97 126,772,739 4.00% 100.00% 4.00% 103,009,997 31,802,289.68 134,812,287 5.00% 110.00% 3.50%

RAAC 2006‐RP4 117,579,211 99,996,947 24,307,553.83 124,304,501 3.00% 100.00% 6.00% 99,996,947 34,230,514.15 134,227,461 4.00% 105.00% 5.00%

RAMP 2006‐RS1 240,398,000 268,328,864 62,394,721.83 330,723,586 4.52% 65.18% 2.82% 268,328,864 67,185,690.41 335,514,554 4.75% 66.57% 2.56%

RAMP 2006‐RS2 179,109,000 186,614,528 47,423,907.17 234,038,435 4.57% 68.03% 3.25% 186,614,528 51,164,116.56 237,778,645 4.82% 69.41% 2.97%

RAMP 2006‐RS3 173,811,000 163,307,936 50,007,682.72 213,315,619 4.81% 68.78% 2.89% 163,307,936 53,507,872.31 216,815,808 5.02% 69.98% 2.58%

RAMP 2006‐RS4 218,022,000 275,999,616 69,375,644.80 345,375,261 4.91% 76.15% 3.07% 275,999,616 74,318,840.48 350,318,456 5.14% 77.44% 2.77%

RAMP 2006‐RS5 107,452,000 100,960,232 30,760,843.65 131,721,076 5.75% 68.05% 4.25% 100,960,232 32,826,864.68 133,787,097 5.96% 69.06% 3.83%

RAMP 2006‐RS6 94,082,200 116,071,472 23,260,826.30 139,332,298 4.01% 71.70% 3.78% 116,071,472 25,092,692.84 141,164,165 4.20% 73.06% 3.45%

RAMP 2006‐RZ1 111,985,000 111,517,128 17,451,496.49 128,968,624 2.74% 72.99% 5.48% 111,517,128 18,818,217.20 130,335,345 2.83% 73.97% 4.96%

RAMP 2006‐RZ2 82,348,200 107,809,872 18,297,242.91 126,107,115 4.23% 69.76% 5.06% 107,809,872 19,712,858.32 127,522,730 4.40% 70.89% 4.59%

RAMP 2006‐RZ3 189,473,000 234,627,360 57,705,632.56 292,332,993 4.69% 76.71% 3.24% 234,627,360 69,803,609.63 304,430,970 4.90% 78.20% 2.98%

RAMP 2006‐RZ4 250,781,000 291,810,560 68,179,730.11 359,990,290 3.79% 76.64% 2.77% 291,810,560 73,216,840.14 365,027,400 4.02% 77.66% 2.56%

RAMP 2006‐RZ5 140,548,000 167,821,296 39,942,042.79 207,763,339 4.23% 75.01% 3.53% 167,821,296 43,113,325.12 210,934,621 4.48% 75.94% 3.21%

RFMSI 2006‐S1 104,563,000 18,580,054 3,513,946.15 22,094,000 1.35% 39.23% 10.08% 18,580,054 3,957,418.96 22,537,473 1.42% 40.35% 9.32%

RFMSI 2006‐S10 235,770,000 48,625,508 14,579,519.15 63,205,027 2.17% 42.14% 8.84% 48,625,508 15,728,869.05 64,354,377 2.20% 42.97% 8.14%

RFMSI 2006‐S11 166,975,000 34,687,496 10,495,719.59 45,183,216 2.95% 34.31% 9.33% 34,687,496 11,425,153.61 46,112,650 2.99% 35.21% 8.56%

RFMSI 2006‐S12 317,167,000 55,954,076 9,731,112.38 65,685,188 0.41% 77.91% 16.11% 55,954,076 11,144,205.05 67,098,281 0.45% 80.05% 15.32%

RFMSI 2006‐S2 72,130,500 14,480,171 3,477,914.87 17,958,086 1.76% 43.11% 9.73% 14,480,171 3,910,139.57 18,390,311 1.86% 43.89% 8.89%

RFMSI 2006‐S3 94,131,000 22,820,258 6,954,659.35 29,774,917 2.52% 46.98% 9.55% 22,820,258 7,682,237.25 30,502,495 2.61% 48.15% 8.80%

RFMSI 2006‐S4 94,980,300 15,215,365 4,556,751.01 19,772,116 2.16% 32.29% 8.33% 15,215,365 5,055,995.15 20,271,360 2.19% 34.05% 7.69%

RFMSI 2006‐S5 192,066,000 40,615,088 14,224,863.11 54,839,951 2.56% 43.61% 8.72% 40,615,088 15,678,032.56 56,293,121 2.63% 44.87% 7.94%

RFMSI 2006‐S6 151,955,000 37,926,424 8,092,714.37 46,019,138 1.95% 43.97% 10.23% 37,926,424 8,984,662.98 46,911,087 2.04% 44.70% 9.45%

RFMSI 2006‐S7 126,798,000 28,830,190 8,981,241.52 37,811,432 2.13% 50.47% 9.12% 28,830,190 9,924,508.21 38,754,698 2.20% 51.75% 8.40%

RFMSI 2006‐S8 111,925,000 25,765,740 10,281,756.98 36,047,497 2.97% 50.38% 9.81% 25,765,740 10,836,751.60 36,602,492 2.97% 51.05% 9.13%

RFMSI 2006‐S9 102,249,000 24,611,726 5,694,669.03 30,306,395 1.55% 54.39% 9.72% 24,611,726 6,313,981.11 30,925,707 1.60% 55.79% 8.94%

RFMSI 2006‐SA1 93,343,700 26,678,300 3,423,521.26 30,101,821 1.91% 29.88% 9.97% 26,678,300 4,035,427.32 30,713,727 2.07% 31.38% 9.28%

RFMSI 2006‐SA2 260,918,000 83,832,176 20,547,779.61 104,379,956 4.48% 32.66% 8.38% 83,832,176 24,922,709.72 108,754,886 4.70% 32.35% 7.86%

RFMSI 2006‐SA3 94,405,800 28,849,486 5,157,551.35 34,007,037 1.91% 55.08% 13.80% 28,849,486 5,356,601.00 34,206,087 1.92% 54.19% 12.87%

RFMSI 2006‐SA4 79,874,400 31,590,606 1,281,942.59 32,872,549 0.56% 33.61% 6.99% 31,590,606 1,484,907.46 33,075,513 0.63% 33.96% 6.72%

RAAC 2006‐SP1 56,160,400 67,229,632 10,071,483.48 77,301,115 2.56% 66.94% 1.49% 67,229,632 12,087,403.06 79,317,035 2.82% 67.04% 1.39%

RAAC 2006‐SP2 76,004,700 73,882,032 16,076,430.43 89,958,462 3.49% 72.44% 3.28% 73,882,032 17,306,740.96 91,188,773 3.66% 73.60% 2.95%

RAAC 2006‐SP3 75,144,000 62,703,300 15,458,661.19 78,161,961 4.31% 66.74% 4.72% 62,703,300 16,529,658.79 79,232,959 4.52% 67.43% 4.40%

Page 8 of 13


Trust Level Model






Going Forward




Loss Going Forward



RAAC 2006‐SP4 81,881,600 53,271,208 11,468,942.70 64,740,151 2.34% 79.54% 3.90% 53,271,208 12,206,139.70 65,477,348 2.44% 80.16% 3.65%

GMACM 2007‐HE3 114,190,447 107,431,287 8,866,701.50 116,297,989 3.43% 103.96% 8.71% 107,431,287 9,400,679.18 116,831,967 3.49% 104.18% 7.72%

RASC 2007‐KS1 139,154,000 137,901,744 33,271,442.75 171,173,187 3.57% 66.70% 2.47% 137,901,744 36,304,514.25 174,206,258 3.81% 68.32% 2.25%

RASC 2007‐KS2 340,770,000 366,682,272 99,783,176.83 466,465,449 3.80% 74.11% 1.80% 366,682,272 107,470,237.27 474,152,509 4.05% 75.47% 1.66%

RASC 2007‐KS3 486,350,000 466,503,616 147,374,458.46 613,878,074 4.56% 71.74% 2.52% 466,503,616 157,673,643.01 624,177,259 4.80% 73.18% 2.33%

RASC 2007‐KS4 87,434,500 95,489,152 29,333,235.22 124,822,387 4.82% 75.25% 2.08% 95,489,152 31,142,218.03 126,631,370 5.04% 76.77% 1.93%

RALI 2007‐QA1 131,898,000 106,556,560 21,912,241.12 128,468,801 3.26% 51.79% 3.33% 106,556,560 23,891,198.63 130,447,759 3.45% 53.18% 3.11%

RALI 2007‐QA2 104,185,000 100,350,624 15,854,384.39 116,205,008 3.40% 46.20% 3.39% 100,350,624 17,195,223.48 117,545,847 3.59% 47.41% 3.20%

RALI 2007‐QA3 280,502,000 270,874,944 43,346,422.33 314,221,366 2.87% 52.59% 3.30% 270,874,944 47,365,205.02 318,240,149 3.10% 53.26% 3.09%

RALI 2007‐QA4 64,657,100 71,406,680 13,082,303.77 84,488,984 3.26% 62.37% 3.26% 71,406,680 14,062,919.33 85,469,599 3.43% 63.22% 3.02%

RALI 2007‐QA5 160,637,000 132,693,904 22,236,256.92 154,930,161 3.66% 41.87% 4.03% 132,693,904 24,183,418.08 156,877,322 3.84% 42.98% 3.72%

RALI 2007‐QH1 205,060,000 173,080,880 34,784,408.47 207,865,288 3.00% 57.99% 3.29% 173,080,880 37,026,554.92 210,107,435 3.18% 58.34% 3.12%

RALI 2007‐QH2 126,633,000 118,749,672 33,939,086.50 152,688,758 6.65% 52.81% 3.42% 118,749,672 36,139,043.31 154,888,715 7.03% 53.91% 3.26%

RALI 2007‐QH3 129,829,000 124,914,440 24,581,446.95 149,495,887 3.09% 59.43% 2.79% 124,914,440 26,441,687.54 151,356,128 3.31% 60.19% 2.67%

RALI 2007‐QH4 170,646,000 127,040,496 42,935,026.44 169,975,522 6.12% 53.30% 3.78% 127,040,496 45,954,731.13 172,995,227 6.40% 54.77% 3.57%

RALI 2007‐QH5 214,651,000 167,367,632 40,277,435.92 207,645,068 3.15% 57.23% 2.77% 167,367,632 42,616,208.67 209,983,841 3.29% 58.04% 2.65%

RALI 2007‐QH6 260,873,000 198,236,480 54,346,655.03 252,583,135 4.05% 53.91% 3.10% 198,236,480 58,266,218.67 256,502,699 4.33% 54.59% 2.95%

RALI 2007‐QH7 159,235,000 108,551,888 35,612,510.47 144,164,398 6.07% 45.92% 3.42% 108,551,888 37,896,656.69 146,448,545 6.38% 46.87% 3.24%

RALI 2007‐QH8 241,780,000 190,100,416 54,200,233.17 244,300,649 4.48% 52.80% 2.72% 190,100,416 57,993,811.53 248,094,228 4.76% 53.77% 2.60%

RALI 2007‐QH9 264,160,000 189,462,720 65,035,399.56 254,498,120 4.37% 56.25% 1.99% 189,462,720 69,084,524.07 258,547,244 4.65% 57.03% 1.91%

RALI 2007‐QO1 227,154,000 203,088,016 61,246,053.84 264,334,070 4.96% 57.40% 2.31% 203,088,016 65,365,494.17 268,453,510 5.23% 58.80% 2.20%

RALI 2007‐QO2 189,409,000 173,127,472 50,263,824.10 223,391,296 4.92% 62.04% 3.39% 173,127,472 53,268,419.32 226,395,891 5.14% 63.08% 3.22%

RALI 2007‐QO3 114,697,000 95,646,920 27,085,364.58 122,732,285 4.43% 60.31% 3.95% 95,646,920 28,722,433.47 124,369,353 4.62% 61.34% 3.76%

RALI 2007‐QO4 199,702,000 158,369,120 46,732,124.37 205,101,244 3.92% 64.18% 3.77% 158,369,120 49,783,051.83 208,152,172 4.12% 65.00% 3.56%

RALI 2007‐QO5 101,488,000 71,781,592 28,244,090.97 100,025,683 6.81% 57.16% 4.63% 71,781,592 29,830,337.00 101,611,929 7.11% 57.98% 4.39%

RALI 2007‐QS1 514,166,000 205,924,944 65,799,314.55 271,724,259 2.64% 55.75% 5.00% 205,924,944 71,520,337.50 277,445,282 2.75% 56.96% 4.57%

RALI 2007‐QS10 187,295,000 88,095,432 31,183,417.81 119,278,850 2.99% 60.85% 4.23% 88,095,432 34,052,702.42 122,148,134 3.11% 62.62% 3.84%

RALI 2007‐QS11 115,299,000 64,526,632 20,848,480.66 85,375,113 3.77% 58.23% 4.94% 64,526,632 22,381,103.65 86,907,736 3.92% 59.00% 4.51%

RALI 2007‐QS2 218,323,000 86,062,256 46,286,092.33 132,348,348 4.22% 62.83% 4.62% 86,062,256 49,862,528.25 135,924,784 4.30% 64.61% 4.14%

RALI 2007‐QS3 400,111,000 178,849,408 84,874,864.84 263,724,273 3.86% 65.43% 4.39% 178,849,408 91,459,690.85 270,309,099 3.99% 66.93% 3.98%

RALI 2007‐QS4 281,095,000 132,726,448 60,795,619.89 193,522,068 5.63% 61.00% 7.32% 132,726,448 64,839,033.20 197,565,481 5.79% 61.83% 6.79%

RALI 2007‐QS5 185,069,000 79,613,944 30,436,299.45 110,050,243 3.12% 63.78% 5.24% 79,613,944 32,820,042.86 112,433,987 3.24% 64.60% 4.77%

RALI 2007‐QS6 318,949,000 162,357,904 49,608,340.29 211,966,244 3.63% 55.49% 5.85% 162,357,904 53,606,396.37 215,964,300 3.77% 56.42% 5.38%

RALI 2007‐QS7 353,706,000 137,847,936 50,281,147.26 188,129,083 2.81% 53.23% 3.73% 137,847,936 54,666,675.19 192,514,611 2.94% 54.51% 3.41%

Page 9 of 13


Trust Level Model






Going Forward




Loss Going Forward



RALI 2007‐QS8 299,528,000 110,000,536 56,705,801.27 166,706,337 3.37% 62.60% 4.03% 110,000,536 61,022,599.03 171,023,135 3.48% 63.90% 3.63%

RALI 2007‐QS9 314,816,000 136,486,560 50,961,218.29 187,447,778 3.23% 57.11% 4.57% 136,486,560 55,686,732.88 192,173,293 3.38% 58.43% 4.16%

RAAC 2007‐RP1 125,009,784 97,583,582 16,000,283.63 113,583,865 1.50% 95.00% 4.00% 97,583,582 28,793,868.39 126,377,450 2.50% 102.00% 3.00%

RAAC 2007‐RP2 94,633,637 77,819,856 25,695,858.68 103,515,714 3.00% 102.00% 3.00% 77,819,856 35,870,900.12 113,690,756 4.00% 110.00% 2.50%

RAAC 2007‐RP3 124,499,411 142,543,564 43,630,746.41 186,174,311 5.00% 85.00% 3.50% 142,543,564 56,848,785.18 199,392,349 6.00% 92.00% 2.50%

RAAC 2007‐RP4 103,992,215 105,847,953 43,357,279.56 149,205,232 4.50% 100.00% 3.00% 105,847,953 58,467,075.88 164,315,029 5.50% 110.00% 2.00%

RAMP 2007‐RS1 155,948,000 125,306,384 50,965,975.08 176,272,359 5.13% 80.81% 4.35% 125,306,384 54,120,616.48 179,427,000 5.31% 81.25% 3.88%

RAMP 2007‐RS2 134,519,000 136,505,712 41,203,760.94 177,709,473 4.99% 77.26% 4.39% 136,505,712 44,057,083.31 180,562,795 5.18% 78.11% 3.93%

RAMP 2007‐RZ1 105,726,000 114,385,728 27,964,050.39 142,349,778 3.90% 71.86% 3.04% 114,385,728 30,325,335.86 144,711,064 4.11% 73.25% 2.73%

RFMSI 2007‐S1 163,319,000 32,508,594 10,603,296.26 43,111,890 2.63% 40.62% 10.29% 32,508,594 12,021,568.50 44,530,162 2.72% 42.25% 9.33%

RFMSI 2007‐S2 159,531,000 27,262,366 13,717,619.70 40,979,986 3.11% 44.99% 9.84% 27,262,366 15,049,178.15 42,311,544 3.18% 46.06% 8.99%

RFMSI 2007‐S3 189,809,000 39,978,232 9,956,023.21 49,934,255 1.73% 42.37% 8.34% 39,978,232 10,974,340.56 50,952,573 1.79% 43.07% 7.57%

RFMSI 2007‐S4 113,085,000 22,693,986 6,233,224.84 28,927,211 1.36% 52.38% 7.79% 22,693,986 6,831,775.79 29,525,762 1.40% 53.28% 7.05%

RFMSI 2007‐S5 192,919,000 34,898,348 14,906,868.60 49,805,217 2.66% 42.64% 8.38% 34,898,348 16,235,808.64 51,134,157 2.72% 43.56% 7.69%

RFMSI 2007‐S6 254,708,000 59,653,252 16,106,551.41 75,759,803 2.30% 35.02% 6.72% 59,653,252 17,746,265.24 77,399,517 2.36% 36.46% 6.20%

RFMSI 2007‐S7 167,964,000 30,685,534 9,969,674.36 40,655,208 1.97% 44.34% 9.29% 30,685,534 11,176,028.45 41,861,562 2.03% 45.90% 8.41%

RFMSI 2007‐S8 156,350,000 38,799,972 12,133,632.87 50,933,605 2.80% 42.39% 9.03% 38,799,972 13,310,509.79 52,110,482 2.87% 43.36% 8.26%

RFMSI 2007‐S9 52,375,600 10,835,202 3,071,243.89 13,906,446 1.21% 73.13% 10.01% 10,835,202 3,301,525.80 14,136,728 1.23% 72.73% 9.08%

RFMSI 2007‐SA1 92,558,200 37,352,004 1,318,561.88 38,670,566 0.75% 23.09% 7.53% 37,352,004 1,420,268.57 38,772,273 0.77% 23.78% 7.30%

RFMSI 2007‐SA2 124,351,000 46,953,572 12,908,179.50 59,861,752 4.09% 54.60% 9.43% 46,953,572 17,372,667.98 64,326,240 4.28% 54.39% 9.10%

RFMSI 2007‐SA3 128,708,000 48,091,736 1,497,854.76 49,589,591 0.51% 34.56% 11.24% 48,091,736 1,567,928.94 49,659,665 0.59% 28.68% 10.65%

RFMSI 2007‐SA4 138,022,000 50,466,544 1,475,677.42 51,942,221 0.53% 38.54% 15.33% 50,466,544 1,891,581.91 52,358,126 0.61% 40.93% 14.47%

RAAC 2007‐SP1 115,973,000 49,744,780 21,829,237.53 71,574,018 3.15% 78.87% 5.12% 49,744,780 23,489,426.64 73,234,207 3.28% 79.81% 4.66%

RAAC 2007‐SP2 112,366,000 84,580,024 32,138,663.32 116,718,687 3.83% 78.43% 3.05% 84,580,024 34,690,498.02 119,270,522 4.01% 79.91% 2.72%

RAAC 2007‐SP3 134,150,000 92,288,840 31,957,279.80 124,246,120 2.80% 80.42% 2.82% 92,288,840 35,171,475.06 127,460,315 3.01% 81.80% 2.55%

GMACM 2003‐J10 10,537,500 1,342 268,947 270,289 1.19% 98.19% 4.63% 1,342 271,771 273,113 1.21% 97.04% 4.25%

GMACM 2003‐AR1 45,197,800 2,108,188 74,231 2,182,418 0.12% 30.36% 16.81% 2,108,188 70,916 2,179,103 0.12% 29.58% 16.05%

GMACM 2003‐AR2 57,176,500 1,127,589 82,643 1,210,232 0.33% 5.08% 6.62% 1,127,589 159,142 1,286,731 0.36% 9.06% 6.44%

GMACM 2003‐GH2 40,839,400 5,142,159 1,926,488 7,068,648 1.35% 54.22% 5.66% 5,142,159 2,048,878 7,191,037 1.38% 55.32% 5.32%

GMACM 2003‐J5 11,212,100 172,479 16,661 189,140 0.07% 89.22% 3.94% 172,479 17,406 189,885 0.08% 89.08% 3.57%

GMACM 2003‐J6 39,249,900 422,020 365,014 787,035 0.29% 50.99% 8.96% 422,020 363,541 785,561 0.29% 49.67% 8.39%

GMACM 2003‐J7 49,888,000 620,441 975,966 1,596,408 0.72% 41.42% 8.08% 620,441 1,027,331 1,647,772 0.74% 41.12% 7.57%

GMACM 2003‐J8 57,316,100 851,528 757,588 1,609,116 0.34% 57.69% 7.95% 851,528 779,728 1,631,256 0.34% 57.51% 7.43%

GMACM 2003‐J9 50,256,400 999,199 1,441,001 2,440,200 1.05% 44.38% 9.17% 999,199 1,508,484 2,507,683 1.04% 45.38% 8.62%

Page 10 of 13


Trust Level Model






Going Forward




Loss Going Forward



GMACM 2010‐1 25,581,091 8,944,455 1,516,202 10,460,656 1.00% 105.00% 12.00% 8,944,455 3,836,248 12,780,703 2.00% 110.00% 8.00%

GMACM 2010‐2 214,522,150 16,647,190 4,412,633 21,059,823 1.20% 41.50% 1.59% 16,647,190 4,906,889 21,554,079 1.25% 42.60% 1.51%

RALI 1999‐QS4 180,155 27,670 ‐ 27,670 0.00% 0.00% 0.00% 27,670 ‐ 27,670 0.00% 0.00% 0.00%

RALI 2001‐QS13 1,796,650 301,852 7,218 309,070 0.34% 90.74% 3.99% 301,852 7,455 309,306 0.35% 90.24% 3.44%

RALI 2001‐QS16 6,619,120 1,949,395 260,165 2,209,561 0.89% 67.25% 6.72% 1,949,395 279,747 2,229,142 0.93% 68.04% 6.40%

RALI 2001‐QS17 5,432,610 1,885,256 515,502 2,400,758 1.65% 89.72% 6.60% 1,885,256 503,536 2,388,792 1.63% 87.70% 6.39%

RALI 2001‐QS18 9,718,020 2,588,240 825,244 3,413,483 2.22% 58.50% 6.09% 2,588,240 871,701 3,459,940 2.29% 59.28% 5.83%

RALI 2001‐QS19 2,597,030 312,485 24,174 336,660 0.56% 111.86% 3.07% 312,485 24,566 337,051 0.58% 110.19% 2.65%

RALI 2002‐QS1 11,179,300 1,904,898 268,854 2,173,752 0.71% 52.75% 7.75% 1,904,898 298,839 2,203,737 0.76% 54.01% 7.36%

RALI 2002‐QS10 2,496,550 571,188 349 571,537 0.02% 12.64% 9.02% 571,188 554 571,742 0.03% 13.37% 8.68%

RALI 2002‐QS11 14,429,100 2,954,325 543,175 3,497,500 1.25% 48.86% 8.61% 2,954,325 599,219 3,553,544 1.30% 50.93% 8.15%

RALI 2002‐QS12 18,731,000 3,179,002 1,626,344 4,805,346 2.08% 66.02% 6.83% 3,179,002 1,724,525 4,903,527 2.09% 68.16% 6.43%

RALI 2002‐QS13 6,296,570 654,041 114,829 768,870 1.19% 86.39% 5.74% 654,041 114,280 768,321 1.17% 87.10% 5.38%

RALI 2002‐QS14 14,914,300 1,819,878 1,166,803 2,986,681 4.75% 29.16% 7.32% 1,819,878 1,246,765 3,066,643 4.66% 31.08% 6.95%

RALI 2002‐QS16 5,138,820 245,594 80,333 325,928 0.80% 108.05% 4.69% 245,594 81,985 327,580 0.81% 107.85% 4.04%

RALI 2002‐QS17 28,919,100 4,004,214 1,599,039 5,603,253 1.34% 60.27% 6.31% 4,004,214 1,709,851 5,714,065 1.37% 61.40% 5.85%

RALI 2002‐QS18 11,831,600 602,532 186,089 788,621 0.83% 98.69% 4.98% 602,532 186,586 789,118 0.84% 97.14% 4.63%

RALI 2002‐QS19 48,032,600 5,399,650 2,188,235 7,587,885 0.98% 66.24% 6.33% 5,399,650 2,311,628 7,711,278 1.01% 66.59% 5.99%

RALI 2002‐QS2 8,713,780 1,369,443 446,086 1,815,530 1.03% 67.66% 5.23% 1,369,443 466,110 1,835,553 1.03% 69.90% 4.88%

RALI 2002‐QS3 16,069,600 3,444,894 1,197,211 4,642,104 1.78% 71.75% 8.75% 3,444,894 1,245,236 4,690,130 1.82% 71.16% 8.23%

RALI 2002‐QS4 4,338,830 412,889 34,905 447,794 0.49% 102.46% 2.84% 412,889 36,602 449,491 0.51% 102.64% 2.51%

RALI 2002‐QS5 14,504,800 3,697,314 762,892 4,460,206 1.73% 46.27% 6.39% 3,697,314 811,449 4,508,763 1.78% 47.35% 6.14%

RALI 2002‐QS6 20,885,800 3,656,902 1,149,907 4,806,809 1.00% 80.33% 6.40% 3,656,902 1,220,861 4,877,764 1.04% 80.41% 6.05%

RALI 2002‐QS7 15,217,800 2,401,107 352,100 2,753,207 0.88% 41.04% 7.77% 2,401,107 389,896 2,791,003 0.91% 43.12% 7.41%

RALI 2002‐QS8 3,388,100 349,587 35,728 385,315 0.69% 93.42% 4.73% 349,587 35,858 385,445 0.69% 93.66% 4.44%

RALI 2002‐QS9 14,107,500 2,780,649 362,529 3,143,178 0.65% 70.27% 10.63% 2,780,649 373,361 3,154,010 0.66% 69.32% 10.06%

RALI 2003‐QA1 12,675,700 2,666,857 853,816 3,520,672 2.11% 46.94% 6.97% 2,666,857 951,133 3,617,990 2.26% 48.51% 6.75%

RALI 2003‐QS10 84,633,100 5,837,929 2,443,816 8,281,745 0.61% 65.06% 6.17% 5,837,929 2,629,886 8,467,815 0.63% 65.53% 5.73%

RALI 2003‐QS11 86,692,000 6,531,428 2,890,632 9,422,060 0.76% 60.37% 6.19% 6,531,428 3,012,661 9,544,089 0.76% 61.21% 5.75%

RALI 2003‐QS12 22,515,500 581,491 258,234 839,725 0.63% 86.02% 5.25% 581,491 262,716 844,207 0.64% 86.23% 4.79%

RALI 2003‐QS13 96,865,400 5,656,400 4,932,346 10,588,746 1.14% 63.53% 6.47% 5,656,400 5,257,091 10,913,490 1.16% 65.00% 6.04%

RALI 2003‐QS14 20,261,500 592,329 132,615 724,944 0.32% 89.26% 2.78% 592,329 138,948 731,278 0.33% 89.93% 2.50%

RALI 2003‐QS15 88,816,100 6,431,615 4,301,167 10,732,782 1.16% 55.13% 5.20% 6,431,615 4,587,646 11,019,261 1.19% 56.03% 4.82%

RALI 2003‐QS16 17,796,900 739,870 256,858 996,728 0.64% 101.18% 4.72% 739,870 253,986 993,856 0.63% 101.79% 4.42%

Page 11 of 13


Trust Level Model






Going Forward




Loss Going Forward



RALI 2003‐QS17 111,242,000 7,073,351 4,883,486 11,956,836 0.97% 60.24% 5.66% 7,073,351 5,069,494 12,142,844 0.98% 60.47% 5.29%

RALI 2003‐QS18 16,428,400 258,305 100,053 358,358 0.28% 96.28% 3.75% 258,305 100,233 358,538 0.28% 95.03% 3.47%

RALI 2003‐QS19 73,498,300 5,503,560 1,121,994 6,625,554 0.62% 33.47% 6.48% 5,503,560 1,201,314 6,704,874 0.61% 35.59% 6.05%

RALI 2003‐QS2 38,176,100 3,894,436 1,160,090 5,054,526 0.96% 46.97% 7.13% 3,894,436 1,290,829 5,185,265 1.02% 48.15% 6.71%

RALI 2003‐QS20 21,343,400 690,376 239,320 929,696 0.48% 96.52% 1.47% 690,376 261,949 952,325 0.53% 96.07% 1.44%

RALI 2003‐QS21 46,079,200 5,354,450 2,103,864 7,458,314 0.89% 67.33% 5.66% 5,354,450 2,255,661 7,610,110 0.91% 68.41% 5.25%

RALI 2003‐QS22 57,256,200 4,282,535 2,897,840 7,180,375 1.31% 52.71% 5.91% 4,282,535 3,112,867 7,395,402 1.35% 53.60% 5.47%

RALI 2003‐QS23 17,029,000 596,536 340,621 937,157 1.02% 86.64% 5.79% 596,536 353,642 950,178 1.04% 88.10% 5.38%

RALI 2003‐QS3 12,428,500 631,877 28,027 659,904 0.13% 83.06% 4.20% 631,877 28,037 659,914 0.13% 84.86% 3.86%

RALI 2003‐QS4 43,827,400 4,683,848 1,843,400 6,527,248 0.97% 64.54% 7.12% 4,683,848 1,928,630 6,612,477 0.99% 64.23% 6.67%

RALI 2003‐QS5 15,902,100 797,833 126,996 924,830 0.36% 109.37% 4.92% 797,833 124,015 921,848 0.35% 108.68% 4.59%

RALI 2003‐QS6 34,531,800 3,477,561 851,191 4,328,752 0.94% 40.93% 7.99% 3,477,561 935,802 4,413,363 0.99% 41.57% 7.46%

RALI 2003‐QS7 43,519,500 2,879,596 806,514 3,686,110 0.38% 62.95% 5.49% 2,879,596 831,804 3,711,400 0.38% 64.40% 5.09%

RALI 2003‐QS9 15,217,500 568,672 214,311 782,983 0.69% 97.30% 4.48% 568,672 215,306 783,978 0.69% 97.30% 4.13%

RAMP 2001‐RS2 4,399,014 11,231,765 492,746 11,724,511 2.00% 100.00% 6.00% 11,231,765 794,929 12,026,694 3.00% 102.00% 4.00%

RAMP 2002‐RS3 18,844,200 22,155,498 2,054,789 24,210,287 2.16% 70.25% 2.99% 22,155,498 2,214,804 24,370,302 2.26% 72.04% 2.82%

RAMP 2002‐RZ2 10,061,900 11,376,191 1,606,771 12,982,962 3.67% 78.55% 8.24% 11,376,191 1,742,006 13,118,197 3.78% 79.72% 7.42%

RAMP 2002‐RZ3 21,991,200 21,050,438 1,855,922 22,906,360 1.69% 69.61% 5.51% 21,050,438 2,030,353 23,080,791 1.76% 71.37% 5.01%

RAMP 2002‐SL1 10,692,946 293,200 188,894 482,094 0.50% 95.00% 12.00% 293,200 249,997 543,197 0.60% 98.00% 10.00%

RAMP 2003‐RS10 80,556,100 83,885,856 5,238,297 89,124,153 1.35% 71.29% 6.81% 83,885,856 5,783,544 89,669,400 1.41% 72.97% 6.23%

RAMP 2003‐RS11 140,888,000 89,645,064 19,292,976 108,938,040 2.74% 73.95% 6.04% 89,645,064 20,923,221 110,568,285 2.87% 74.79% 5.50%

RAMP 2003‐SL1 28,974,595 1,392,562 545,124 1,937,686 0.75% 50.00% 10.00% 1,392,562 865,305 2,257,867 1.00% 55.00% 8.00%

RASC 2001‐KS2 51,174,600 101,113,080 7,108,827 108,221,907 2.91% 79.93% 3.98% 101,113,080 7,659,123 108,772,203 3.06% 81.39% 3.73%

RASC 2001‐KS3 60,639,700 121,518,872 10,828,074 132,346,946 3.41% 81.67% 2.25% 121,518,872 11,510,204 133,029,076 3.53% 83.61% 2.09%

RASC 2002‐KS2 31,971,100 79,880,448 3,733,160 83,613,608 2.26% 77.26% 3.61% 79,880,448 3,992,289 83,872,737 2.34% 78.96% 3.35%

RASC 2003‐KS10 97,438,000 45,396,460 10,701,596 56,098,056 2.15% 73.17% 5.46% 45,396,460 11,505,830 56,902,290 2.21% 74.96% 5.07%

RASC 2003‐KS11 84,476,200 73,522,408 8,224,917 81,747,325 1.78% 71.02% 4.16% 73,522,408 8,841,744 82,364,152 1.83% 72.97% 3.82%

RASC 2003‐KS12 83,572,900 62,672,284 11,640,485 74,312,769 2.64% 74.19% 3.88% 62,672,284 12,599,844 75,272,128 2.76% 76.06% 3.62%

RASC 2003‐KS3 11,554,900 19,063,186 3,038,630 22,101,816 5.14% 72.64% 3.89% 19,063,186 3,266,051 22,329,237 5.39% 74.84% 3.76%

RASC 2003‐KS6 13,520,000 20,004,462 2,212,624 22,217,086 2.11% 88.93% 3.51% 20,004,462 2,367,126 22,371,588 2.21% 90.36% 3.34%

RASC 2003‐KS7 95,194,400 40,266,224 10,401,121 50,667,345 2.26% 73.57% 5.74% 40,266,224 11,381,290 51,647,514 2.39% 75.37% 5.40%

RASC 2003‐KS8 60,280,000 28,246,624 7,311,108 35,557,732 2.22% 74.59% 4.37% 28,246,624 7,884,594 36,131,218 2.31% 76.44% 4.06%

RFMSI 2003‐S10 29,743,000 518,521 47,408 565,929 0.14% 20.15% 10.56% 518,521 45,720 564,241 0.12% 20.52% 9.93%

RFMSI 2003‐S11 16,473,200 355,856 213,333 569,189 0.63% 96.85% 6.47% 355,856 209,115 564,971 0.62% 96.02% 6.05%

Page 12 of 13


Trust Level Model






Going Forward




Loss Going Forward



RFMSI 2003‐S12 32,946,300 1,576,319 18,927 1,595,245 0.07% 18.83% 15.26% 1,576,319 16,158 1,592,477 0.06% 18.02% 14.63%

RFMSI 2003‐S14 11,756,000 1,114 94,340 95,454 0.84% 44.23% 3.99% 1,114 98,776 99,890 0.86% 44.98% 3.66%

RFMSI 2003‐S15 12,163,100 36,076 13,471 49,547 0.15% 32.83% 6.39% 36,076 16,322 52,398 0.15% 40.13% 5.88%

RFMSI 2003‐S16 15,849,400 79,567 2,306 81,872 0.01% 68.43% 2.08% 79,567 2,672 82,239 0.01% 69.63% 1.81%

RFMSI 2003‐S17 53,865,700 454,644 1,030,393 1,485,037 0.70% 46.64% 10.07% 454,644 1,120,032 1,574,676 0.72% 47.77% 9.49%

RFMSI 2003‐S18 17,318,600 32,224 108,785 141,010 0.27% 97.99% 2.40% 32,224 110,864 143,088 0.28% 98.06% 2.01%

RFMSI 2003‐S19 38,975,600 717,160 528,027 1,245,187 0.37% 61.81% 10.34% 717,160 546,829 1,263,989 0.37% 61.45% 9.66%

RFMSI 2003‐S6 8,103,380 22,752 17 22,770 0.00% 120.00% 10.93% 22,752 0 22,753 0.00% 120.00% 10.28%

RFMSI 2003‐S7 62,850,000 734,738 551,224 1,285,961 0.34% 46.79% 11.24% 734,738 575,111 1,309,848 0.34% 47.29% 10.54%

RFMSI 2003‐S9 5,632,760 96,178 2,551 98,729 0.06% 17.34% 14.64% 96,178 3,170 99,348 0.05% 22.20% 14.02%

RFMS2 1998‐HI2 ‐ 36,874,298 ‐ 36,874,298 0.00% 0.00% 0.00% 36,874,298 ‐ 36,874,298 0.00% 0.00% 0.00%

RFMS2 2002‐HI4 8,345,310 29,426,514 1,019,481 30,445,995 4.27% 112.18% 10.87% 29,426,514 1,085,504 30,512,018 4.46% 112.20% 9.98%

RFMS2 2002‐HI5 8,906,660 23,183,006 1,042,973 24,225,979 4.17% 111.87% 11.99% 23,183,006 1,117,854 24,300,860 4.36% 111.81% 10.86%

RFMS2 2002‐HS1 1,744,870 3,813,717 34,168 3,847,885 0.95% 111.65% 11.83% 3,813,717 35,572 3,849,289 0.98% 111.89% 11.36%

RFMS2 2002‐HS2 2,358,680 3,841,103 295,469 4,136,572 5.18% 110.70% 8.67% 3,841,103 304,789 4,145,892 5.31% 110.76% 8.20%

RFMS2 2003‐HI1 10,162,800 21,516,042 956,846 22,472,888 2.84% 111.78% 9.23% 21,516,042 1,022,489 22,538,531 2.97% 111.82% 8.41%

RFMS2 2003‐HI2 13,542,700 26,120,772 1,209,162 27,329,934 2.62% 109.41% 9.64% 26,120,772 1,275,156 27,395,928 2.70% 109.46% 8.83%

RFMS2 2003‐HI4 23,149,400 33,091,528 2,269,820 35,361,348 2.92% 110.23% 10.90% 33,091,528 2,451,552 35,543,080 3.07% 110.26% 9.83%

RFSC 2002‐RM1 4,174,950 261,814 190 262,003 0.01% 27.04% 14.43% 261,814 118 261,931 0.01% 34.00% 14.00%

RFSC 2003‐RM1 19,661,200 267,819 94,108 361,927 0.23% 36.38% 10.91% 267,819 103,322 371,141 0.23% 37.78% 10.27%

RFSC 2003‐RM2 29,451,200 480,656 295,912 776,568 0.38% 41.87% 8.96% 480,656 294,608 775,264 0.37% 41.71% 8.45%

Page 13 of 13

ANNEX D


Non Debtor Sponsored Trust Forecasted Loss Analysis

Name Current Collat Bal Collateral Losses to DateSeller

Percentage

Debtor Collateral

Losses To Date

Forecasted Losses for

Debtors Loans

LOWER RANGE

Forecasted Losses

for Debtors Loans

HIGHER RANGE

BAFC 2005‐3 B1 46,122,697.72 1,269,318.97 4.30% 54,581$ 319,307$ 349,652$

BAFC 2005‐4 B1 70,633,534.00 1,132,495.00 6.30% 71,347$ 716,436$ 784,520$

BAFC 2005‐5 B1 116,254,211.04 6,106,889.63 16.22% 990,537$ 3,035,886$ 3,324,389$

BAFC 2005‐6 B1 116,126,157.51 11,559,009.03 7.32% 845,542$ 1,367,635$ 1,497,603$

BAFC 2005‐7 XB1 226,602,473.84 15,132,022.11 2.60% 393,433$ 948,558$ 1,038,700$

BAFC 2005‐8 B1 165,117,351.03 12,904,064.20 9.14% 1,179,431$ 2,429,768$ 2,660,671$

BAFC 2006‐1 B1 133,175,040.69 12,828,598.85 3.96% 508,013$ 849,071$ 929,759$

BAFC 2006‐4 B1 107,617,528.57 15,920,476.16 15.90% 2,531,356$ 2,754,901$ 3,016,702$

BAFC 2007‐4 SB1 374,631,864.82 110,490,119.40 2.42% 2,673,861$ 1,459,641$ 1,598,352$

CARR 2006‐RFC1 M1 188,518,789.66 177,453,492.52 100.00% 177,453,493$ 30,351,525$ 33,235,863$

CARR 2007‐RFC1 M1 352,439,398.34 235,854,273.44 100.00% 235,854,273$ 56,742,743$ 62,135,066$

GSR 2004‐10F B1 119,729,567.90 1,214,279.34 17.47% 212,135$ 3,367,598$ 3,687,624$

GSR 2005‐5F B1 267,976,137.60 8,039,575.23 4.61% 370,624$ 1,988,946$ 2,177,957$

GSR 2006‐2F B1 242,722,593.48 30,647,030.70 1.20% 367,764$ 468,940$ 513,504$

GSR 2006‐4F B1 211,115,424.41 38,180,162.27 18.88% 7,208,415$ 6,417,233$ 7,027,070$

GSR 2007‐AR1 B1 551,503,329.33 153,290,998.52 5.00% 7,664,550$ 4,439,602$ 4,861,502$

GSR 2007‐OA2 B1 175,584,650.31 139,784,331.56 5.00% 6,989,217$ 1,413,456$ 1,547,779$

GSR 2007‐4F B1 250,572,320.52 40,678,893.35 2.73% 1,110,534$ 1,101,341$ 1,206,002$

HVMLT 2003‐1 B1 13,325,184.80 310,198.79 53.17% 164,933$ 1,140,685$ 1,249,086$

HVMLT 2003‐2 B1 52,419,897.69 2,075,474.82 0.16% 3,321$ 13,503$ 14,787$

HVMLT 2004‐4 B1 18,997,613.22 4,037,526.93 5.32% 214,796$ 162,718$ 178,182$

HVMLT 2007‐2 B1 438,147,007.80 473,299,029.09 10.28% 48,655,140$ 7,251,683$ 7,940,819$

HVMLT 2007‐7 B1 687,432,372.74 403,255,328.40 12.06% 48,632,593$ 13,347,599$ 14,616,036$

LXS 2006‐12N M1 541,595,693.40 452,457,102.57 16.77% 75,877,056$ 14,622,921$ 16,012,553$

LXS 2007‐12N 1M1 481,227,054.99 414,854,693.06 2.73% 11,325,533$ 2,115,137$ 2,316,141$

LXS 2007‐15N M1i 1,194,475,460.76 839,423,396.47 15.50% 130,110,626$ 29,808,135$ 32,640,834$

LXS 2007‐2N M1 442,711,580.23 423,197,520.70 35.47% 150,108,161$ 25,281,797$ 27,684,353$

LXS 2007‐4N M1 730,049,817.03 656,792,828.16 14.63% 96,088,791$ 17,195,812$ 18,829,949$

MLMI 2003‐A2 1M1 41,833,644.42 575,621.62 5.11% 29,414$ 344,170$ 376,876$

MLMI 2003‐A4 M1 15,627,901.33 1,598,267.12 11.96% 191,153$ 300,925$ 329,522$

SARM 2007‐6 M1 256,667,820.06 188,543,866.46 0.75% 1,414,079$ 309,926$ 339,379$

SASC 2002‐4H B1 7,260,032.40 1,150,499.26 19.86% 228,489$ 232,137$ 254,197$

ARMT 2005‐10 CB1 291,823,806.58 109,131,627.32 4.50% 4,910,923$ 2,114,263$ 2,315,184$

ARMT 2005‐11 CB1 277,803,109.02 171,982,573.12 4.50% 7,739,216$ 2,012,684$ 2,203,951$

ARMT 2005‐9 CB1 226,421,855.07 94,578,444.99 9.00% 8,512,060$ 3,280,853$ 3,592,636$

BAFC 2006‐2 B1 228,956,363.31 27,324,741.29 0.99% 270,515$ 364,934$ 399,614$

BAFC 2007‐3 3B1 301,910,869.22 144,575,443.24 1.84% 2,660,188$ 894,381$ 979,375$

BAFC 2007‐7 XB1 262,133,367.49 97,001,333.69 0.71% 688,709$ 299,645$ 328,120$

BSABS 2004‐AC1 M1 25,982,686.10 4,917,772.69 1.36% 66,882$ 56,892$ 62,298$

BSABS 2004‐AC7 M1 58,623,547.58 10,417,928.68 2.40% 250,030$ 226,521$ 248,048$

BALTA 2005‐4 1B1 474,967,929.93 136,195,982.91 0.32% 435,827$ 244,703$ 267,958$

CSFB 2005‐10 CB1 335,350,443.65 79,958,062.24 4.58% 3,662,079$ 2,472,807$ 2,707,801$

CSFB 2005‐11 CB1 204,781,429.16 48,428,055.60 3.02% 1,462,527$ 995,688$ 1,090,310$

CSFB 2005‐12 B1 284,518,724.58 129,779,877.73 3.35% 4,347,626$ 1,534,552$ 1,680,382$

CSFB 2005‐3 CB1 205,340,258.27 17,841,546.85 9.00% 1,605,739$ 2,975,380$ 3,258,134$

CSFB 2005‐4 CB1 186,248,487.75 13,478,507.07 9.00% 1,213,066$ 2,698,741$ 2,955,205$

CSFB 2005‐5 CB1 169,739,598.31 11,867,711.14 2.54% 301,440$ 694,133$ 760,097$

CSFB 2005‐6 CB1 177,800,723.35 37,228,999.25 7.63% 2,840,573$ 2,184,157$ 2,391,720$

CSFB 2005‐8 B1 346,630,511.64 81,085,500.62 3.39% 2,748,798$ 1,891,875$ 2,071,662$

CSFB 2005‐9 CB1 268,676,244.50 56,479,462.48 2.77% 1,564,481$ 1,198,215$ 1,312,083$

CSMC 2006‐1 CB1 319,049,007.94 63,126,907.00 0.19% 119,941$ 97,597$ 106,872$

CSMC 2006‐8 CB1 100,139,703.08 52,851,576.29 2.50% 1,321,289$ 403,062$ 441,366$

CSMC 2006‐9 CB1 384,769,281.97 52,402,966.45 0.09% 47,163$ 55,753$ 61,051$

CSMC 2007‐6 CB1 230,873,399.93 87,071,115.46 0.49% 426,648$ 182,136$ 199,445$

CSMC 2007‐7 CB1 105,579,478.51 53,159,377.14 0.21% 111,635$ 35,696$ 39,089$

GPMF 2006‐HE1 AX 178,750,225.91 916,315,108.20 0.44% 4,031,786$ 126,627$ 138,660$

GSMPS 2006‐RP2 B1 150,485,404.38 4,395,009.33 3.55% 156,023$ 860,099$ 941,835$

GSR 2003‐2F B1 6,324,259.93 168,313.20 32.89% 55,358$ 334,888$ 366,713$

GSR 2005‐6F B1 209,893,038.63 16,656,342.52 2.68% 446,390$ 905,646$ 991,711$

GSR 2005‐8F B1 257,586,313.65 27,882,335.28 9.00% 2,509,410$ 3,732,426$ 4,087,122$

GSR 2005‐9F 1B1 253,077,549.16 40,471,050.23 0.42% 169,978$ 171,131$ 187,394$

GSR 2005‐AR3 1B1 263,202,444.06 51,928,750.67 7.89% 4,097,178$ 3,343,434$ 3,661,164$

GSR 2005‐AR7 1B1 541,742,306.18 55,078,809.77 2.82% 1,553,222$ 2,459,618$ 2,693,359$

Page 1 of 3




Percentage

Debtor Collateral

Losses To Date


Debtors Loans

LOWER RANGE

Forecasted Losses

for Debtors Loans

HIGHER RANGE

GSR 2006‐3F B1 206,316,647.64 32,062,166.40 1.45% 464,901$ 481,646$ 527,418$

GSR 2006‐AR1 B1 445,081,525.21 92,797,904.97 5.00% 4,639,895$ 3,582,906$ 3,923,394$

GSR 2007‐HEL1 A 27,476,093.65 59,353,463.35 2.50% 1,483,837$ 110,591$ 121,101$

HALO 2007‐AR2 B1 165,083,926.22 59,793,616.24 0.33% 197,319$ 87,709$ 96,044$

HVMLT 2006‐13 B1 58,327,757.06 28,413,406.99 2.18% 619,412$ 204,719$ 224,173$

LXS 2006‐10N 1M1 382,187,689.94 315,091,243.17 0.46% 1,449,420$ 283,048$ 309,947$

MASTR 2003‐2 30B1 24,350,423.66 403,626.41 9.00% 36,326$ 352,838$ 386,368$

MASTR 2003‐4 B1 55,308,119.33 58,048.65 0.38% 221$ 33,838$ 37,053$

MLMI 2005‐A6 B1 238,997,124.73 101,722,445.69 5.00% 5,086,122$ 1,923,927$ 2,106,760$

RBSGC 2007‐B 1B1 230,329,486.75 59,256,348.76 0.11% 65,182$ 40,791$ 44,668$

RBSGC 2005‐A B1 127,023,989.81 17,487,597.03 4.50% 786,942$ 920,289$ 1,007,745$

SAIL 2006‐2 B1 209,909,319.89 358,738,849.52 0.78% 2,798,163$ 263,604$ 288,655$

SAMI 2003‐AR1 B1 37,877,040.02 4,732,583.08 100.00% 4,732,583$ 6,098,203$ 6,677,722$

SARM 2007‐3 M1 410,080,988.69 186,136,312.24 2.95% 5,491,021$ 1,947,680$ 2,132,770$

SASC 2002‐9 M1 36,747,563.22 2,181,496.67 0.80% 17,452$ 47,331$ 51,829$

SASC 2006‐BC2 M1 298,566,729.28 376,223,415.61 0.90% 3,386,011$ 432,623$ 473,736$

BAFC 2006‐5 B1 186,498,486.14 25,870,286.61 2.50% 646,757$ 750,656$ 821,992$

CSFB 2002‐34 CB1 16,539,377.44 6,243,406.45 9.00% 561,907$ 239,656$ 262,430$

CSFB 2002‐AR33 CB1 7,345,786.47 1,832,263.04 9.00% 164,904$ 106,440$ 116,556$

FMRMT 2003‐A A 371,000.00 0.00 100.00% ‐$ 59,731$ 65,407$

GSMPS 2004‐4 B1 207,260,266.89 8,345,613.22 9.00% 751,105$ 3,003,201$ 3,288,599$

GSMPS 2005‐LT1 M1 17,666,468.76 13,584,197.07 3.44% 467,296$ 97,844$ 107,142$

GSMPS 2005‐RP2 B1 231,817,456.40 7,846,283.19 2.36% 185,172$ 880,814$ 964,518$

GSMPS 2005‐RP3 B1 243,809,547.73 7,421,963.37 2.23% 165,510$ 875,349$ 958,535$

GSMPS 2006‐RP1 B1 204,967,165.37 2,802,580.36 5.00% 140,129$ 1,649,986$ 1,806,786$

GSR 2005‐7F B1 89,474,278.65 5,226,284.39 9.00% 470,366$ 1,296,482$ 1,419,688$

LMT 2005‐1 B1 314,801,584.26 44,031,767.40 1.37% 603,235$ 694,358$ 760,343$

SASC 05‐RF2 B1 46,337,074.48 998,298.00 9.00% 89,847$ 671,424$ 735,230$

SASC 05‐RF4 B1 65,857,439.58 2,535,253.21 9.00% 228,173$ 954,274$ 1,044,960$

SASC 05‐RF6 B1 31,434,360.49 1,490,600.68 9.00% 134,154$ 455,484$ 498,769$

SASC 2001‐8A B1i 3,630,674.96 281,888.42 9.00% 25,370$ 52,608$ 57,608$

SASC 2002‐12 A6 97,226,745.85 1,069,268.39 9.00% 96,234$ 1,408,816$ 1,542,697$

SASC 2005‐RF1 B1 52,953,753.51 1,542,184.32 9.00% 138,797$ 767,300$ 840,217$

SASC 2005‐S7 M1 68,433,230.85 156,172,007.09 9.00% 14,055,481$ 991,598$ 1,085,830$

SASI 1993‐6 B1 1,751,608.03 3,969,649.23 4.50% 178,634$ 12,690$ 13,896$

TMTS 2005‐11 1B1 66,492,102.17 236,911,559.96 4.50% 10,661,020$ 481,735$ 527,515$

TMTS 2005‐13SL B1 41,194,133.87 207,622,651.03 4.50% 9,343,019$ 298,451$ 326,814$

GSAA 2005‐9 B1 187,166,225.22 68,198,725.33 19.48% 13,285,112$ 5,870,057$ 6,427,895$

LUM 2006‐3 1M1 171,265,032.41 156,450,712.16 28.35% 44,353,777$ 7,817,136$ 8,560,006$

LUM 2006‐4 B1 93,250,428.28 110,100,804.37 11.87% 13,068,965$ 1,782,081$ 1,951,434$

LUM 2006‐5 B1 135,120,732.80 119,719,056.93 51.86% 62,086,303$ 11,281,852$ 12,353,978$

SARM 2004‐4 B1 211,455,292.52 14,334,860.08 0.06% 8,601$ 20,427$ 22,368$

SEMT 2005‐2 B1 41,285,244.15 1,700,864.02 14.65% 249,177$ 973,774$ 1,066,313$

ARMT 2004‐5 CB1 139,809,733.62 38,841,928.74 4.50% 1,747,887$ 1,012,922$ 1,109,181$

ARMT 2005‐1 CB1 131,026,190.77 47,799,123.49 4.50% 2,150,961$ 949,285$ 1,039,496$

BSABS 2007‐SD2 1B1 126,945,319.11 66,297,117.34 0.01% 6,630$ 2,044$ 2,238$

BSABS 2007‐SD3 M1 163,591,733.89 101,154,998.56 0.71% 718,200$ 187,002$ 204,773$

BALTA 2006‐4 1B1 1,188,978,495.84 1,268,938,617.75 0.19% 2,410,983$ 363,709$ 398,272$

BALTA 2006‐5 1B1 301,710,595.75 335,916,414.84 0.20% 671,833$ 97,151$ 106,383$

BALTA 2006‐8 1B1 372,575,295.91 334,951,789.44 0.52% 1,741,749$ 311,920$ 341,562$

BSSLT 2007‐SV1A M1 252,944,594.77 443,417,142.35 2.50% 11,085,429$ 1,018,102$ 1,114,853$

GSMPS 2005‐RP1 B1 237,765,150.17 6,370,646.80 1.35% 86,004$ 516,783$ 565,893$

GSR 2006‐AR2 1B1 312,256,066.25 66,254,679.05 5.00% 3,312,734$ 2,513,661$ 2,752,537$

MALT 2004‐12 B1 102,976,582.62 7,773,956.81 2.50% 194,349$ 414,481$ 453,869$

MALT 2004‐4 B1 102,935,784.12 5,411,522.17 2.50% 135,288$ 414,317$ 453,689$

MALT 2004‐6 B1 123,236,768.29 7,502,240.72 4.50% 337,601$ 892,850$ 977,699$

MALT 2004‐7 B1 133,077,768.93 7,196,677.51 4.50% 323,850$ 964,148$ 1,055,772$

MALT 2004‐8 B1 97,366,165.39 7,672,622.47 4.50% 345,268$ 705,418$ 772,454$

MALT 2005‐3 B1 125,812,710.72 10,502,859.71 2.50% 262,571$ 506,396$ 554,520$

MALT 2005‐4 B1 69,024,119.83 12,647,354.06 4.50% 569,131$ 500,080$ 547,603$

MALT 2005‐5 B1 116,712,304.30 18,236,215.02 2.50% 455,905$ 469,767$ 514,409$

MALT 2006‐1 B1 101,688,520.15 22,995,242.18 0.72% 165,566$ 117,877$ 129,079$

MALT 2007‐HF1 B1 83,629,002.26 48,146,776.73 4.80% 2,311,045$ 646,285$ 707,702$

Page 2 of 3




Percentage

Debtor Collateral

Losses To Date


Debtors Loans

LOWER RANGE

Forecasted Losses

for Debtors Loans

HIGHER RANGE

MARP 2005‐1 B1 93,190,654.67 2,290,006.22 9.00% 206,101$ 1,350,333$ 1,478,656$

MARP 2005‐2 B1 131,510,321.99 3,688,057.92 0.89% 32,824$ 188,441$ 206,349$

MARP 2006‐1 B1 98,051,734.13 2,738,868.61 0.17% 4,656$ 26,837$ 29,387$

MARP 2006‐2 B1 78,129,508.40 2,614,919.56 4.42% 115,579$ 555,985$ 608,821$

MASTR 2002‐7 B1 14,805,774.06 591,786.94 5.81% 34,383$ 138,495$ 151,656$

MASTR 2004‐1 B1 64,402,369.38 282,841.71 9.00% 25,456$ 933,190$ 1,021,872$

MASTR 2004‐10 B1 44,067,883.52 1,050,096.55 9.00% 94,509$ 638,544$ 699,225$

MASTR 2004‐11 B1 56,791,450.69 1,560,433.21 8.04% 125,459$ 735,131$ 804,992$

MASTR 2004‐3 B1 31,533,417.85 126,042.44 9.00% 11,344$ 456,919$ 500,341$

MASTR 2004‐4 B1 23,982,657.04 548,503.73 2.65% 14,535$ 102,322$ 112,046$

MASTR 2004‐5 B1 25,717,911.09 695,297.66 3.29% 22,875$ 136,225$ 149,171$

MASTR 2004‐6 B1 72,833,505.09 406,535.99 2.80% 11,383$ 328,333$ 359,535$

MASTR 2004‐9 30B1 70,024,324.32 1,316,768.33 5.95% 78,348$ 670,798$ 734,545$

MSSTR 2005‐2 B1 30,763,034.12 974,046.97 1.37% 13,344$ 67,854$ 74,302$

SACO 2007‐1 B1 32,159,533.01 165,850,750.85 5.00% 8,292,538$ 258,884$ 283,486$

SEMT 2004‐10 B1 74,785,297.02 5,930,963.12 4.50% 266,893$ 541,819$ 593,309$

SEMT 2004‐11 B1 92,260,808.38 3,452,979.83 2.90% 100,136$ 430,766$ 471,702$

SEMT 2004‐12 B1 73,390,264.41 4,125,968.69 3.10% 127,905$ 366,291$ 401,100$

SEMT 2004‐4 B1 51,691,737.27 3,614,964.97 1.99% 71,938$ 165,615$ 181,354$

SEMT 2004‐5 B1 62,109,278.03 2,981,797.87 4.50% 134,181$ 449,982$ 492,744$

SEMT 2004‐6 B1 68,571,799.65 4,515,964.90 4.19% 189,219$ 462,579$ 506,538$

SEMT 2004‐7 B1 96,288,490.72 3,726,487.99 4.41% 164,338$ 683,658$ 748,627$

SEMT 2004‐8 B1 83,748,130.83 4,425,869.66 3.88% 171,724$ 523,158$ 572,874$

SEMT 2004‐9 B1 88,217,333.73 4,194,499.50 4.50% 188,752$ 639,135$ 699,872$

SEMT 2005‐1 B1 54,122,801.65 3,113,942.36 4.50% 140,127$ 392,120$ 429,383$

SEMT 2005‐3 B1 56,109,354.53 3,010,209.45 4.50% 135,459$ 406,512$ 445,144$

SEMT 2005‐4 1B1 63,201,958.30 990,812.32 2.35% 23,284$ 239,125$ 261,849$

SEMT 2007‐1 B1 266,670,302.01 54,148,315.15 1.66% 898,862$ 712,703$ 780,432$

SEMT 2007‐2 1B1 471,810,637.03 36,561,138.98 2.50% 914,028$ 1,899,038$ 2,079,505$

SEMT 2007‐3 1B1 318,436,503.08 36,315,480.38 2.50% 907,887$ 1,281,707$ 1,403,509$

SEMT 2007‐4 B1 79,653,463.36 10,154,058.02 2.50% 253,851$ 320,605$ 351,073$

FNR 2002‐66 A1 72,889,552.08 0.00 4.50% ‐$ 528,085$ 578,269$

LMT 2006‐7 B1 295,490,128.27 128,870,358.42 2.45% 3,157,324$ 1,165,561$ 1,276,325$

SASC 2008‐RF1 B1 61,908,177.98 3,718,569.58 5.00% 185,928$ 498,361$ 545,721$

Page 3 of 3

ANNEX E


Pre 2004 Defect Rate Analysis

Shelf / Series Year VintageSince

Issue Mth

Deal

Count Original Pool Balance Current Pool Balance

90+ Day & FC &

REO Delq

% Curr Bal

RFMSI S ‐ 15 Yr FRM 2003 Pre‐2004 12 11 2,790,880,068$ 1,995,086,546$ 0.00%

RFMSI S ‐30 Yr FRM 2003 Pre‐2004 12 8 4,057,362,334$ 3,057,866,905$ 0.00%

RFMSI SA ‐ ARM 2001 Pre‐2004 12 1 199,421,312$ 76,342,036$ 0.00%

RFMSI SA ‐ ARM 2002 Pre‐2004 12 2 780,001,352$ 221,155,588$ 0.00%

RFMSI S ‐ 15 Yr FRM 2000 Pre‐2004 12 4 692,917,501$ 477,921,310$ 0.13%

RALI QS ‐ 15 Yr FRM 2003 Pre‐2004 12 9 1,823,919,708$ 1,479,403,124$ 0.10%

RFMSI S ‐30 Yr FRM 2002 Pre‐2004 12 8 3,665,732,589$ 1,793,076,823$ 0.11%

RALI QS ‐ 15 Yr FRM 2000 Pre‐2004 12 2 214,317,765$ 166,142,450$ 0.13%

RALI QS ‐ 15 Yr FRM 2002 Pre‐2004 12 5 1,014,182,335$ 685,091,255$ 0.11%

RFMSI S ‐ 15 Yr FRM 2002 Pre‐2004 12 12 3,126,622,483$ 1,373,978,543$ 0.05%

RFMSI S ‐ 15 Yr FRM 2001 Pre‐2004 12 8 1,809,165,990$ 1,099,656,041$ 0.08%

RFMSI S ‐30 Yr FRM 2001 Pre‐2004 12 21 8,628,974,031$ 5,499,682,637$ 0.12%

RFMSI S ‐30 Yr FRM 2000 Pre‐2004 12 12 4,102,984,959$ 2,688,786,040$ 0.39%

RALI QS ‐ 15 Yr FRM 2001 Pre‐2004 12 3 514,006,874$ 377,481,588$ 0.24%

RALI QS ‐ 30 Yr FRM 2003 Pre‐2004 12 14 6,661,907,841$ 5,102,859,658$ 0.28%

RALI QS ‐ 30 Yr FRM 2002 Pre‐2004 12 14 5,701,209,991$ 3,610,204,790$ 0.82%

RALI QS ‐ 30 Yr FRM 2001 Pre‐2004 12 16 4,346,273,626$ 3,029,452,305$ 0.86%

RAMP RZ ‐ ARM 2003 Pre‐2004 12 1 125,000,140$ 97,778,067$ 1.10%

RALI QS ‐ 30 Yr FRM 2000 Pre‐2004 12 12 3,061,833,143$ 2,125,993,789$ 1.60%

RAMP RZ ‐ FRM 2003 Pre‐2004 12 5 3,305,001,828$ 2,653,618,153$ 1.41%

RAMP RZ ‐ FRM 2000 Pre‐2004 12 1 175,000,110$ 115,550,205$ 2.51%

RASC KS ‐ FRM 2003 Pre‐2004 12 8 4,375,000,775$ 3,513,908,518$ 2.28%

RAMP RZ ‐ FRM 2002 Pre‐2004 12 4 1,890,005,063$ 1,466,742,553$ 2.48%

RAMP RS ‐ FRM 2003 Pre‐2004 12 11 6,030,006,620$ 4,620,294,185$ 2.62%

RAMP RZ ‐ FRM 2001 Pre‐2004 12 4 890,004,267$ 642,168,114$ 2.72%

RASC KS ‐ ARM 2003 Pre‐2004 12 11 9,925,011,602$ 7,468,368,229$ 4.50%

RASC KS ‐ FRM 2002 Pre‐2004 12 5 3,450,001,931$ 2,750,984,965$ 5.30%

RAMP RS ‐ ARM 2003 Pre‐2004 12 11 7,035,010,545$ 5,569,704,088$ 5.55%

RAMP RS ‐ FRM 2002 Pre‐2004 12 7 2,862,343,599$ 1,933,227,693$ 5.61%

RASC KS ‐ ARM 2002 Pre‐2004 12 7 8,050,009,344$ 6,486,375,932$ 5.93%

RASC KS ‐ FRM 2000 Pre‐2004 12 5 2,450,005,810$ 2,004,796,337$ 8.24%

RAMP RS ‐ FRM 2001 Pre‐2004 12 3 892,170,872$ 593,954,111$ 7.20%

RAMP RS ‐ FRM 2000 Pre‐2004 12 4 475,133,800$ 341,382,277$ 10.44%

RASC KS ‐ FRM 2001 Pre‐2004 12 3 2,475,000,871$ 1,969,344,779$ 7.66%

RASC KS ‐ ARM 2000 Pre‐2004 12 5 3,850,008,581$ 3,217,912,914$ 9.16%

RAMP RS ‐ ARM 2002 Pre‐2004 12 6 893,147,154$ 687,667,220$ 8.67%

RASC KS ‐ ARM 2001 Pre‐2004 12 4 3,750,002,373$ 3,003,774,055$ 8.78%

RAMP RS ‐ ARM 2000 Pre‐2004 12 4 600,060,614$ 391,394,531$ 12.49%

RALI QA ‐ ARM 2003 Pre‐2004 12 5 $ 2,422,605,271 $ 2,144,218,614 17.45%

RAMP RS ‐ ARM 2001 Pre‐2004 12 3 449,775,137$ 320,782,831$ 15.94%

RFMSII HS ‐ CES 2003 Pre‐2004 12 3 1,225,000,337$ 811,465,513$ 0.11%

RFMSII HS ‐ CES 2002 Pre‐2004 12 3 960,001,568$ 490,852,687$ 0.24%

RFMSII HS ‐ HELOC 2003 Pre‐2004 12 4 940,000,558$ 572,476,529$ 0.26%

RFMSII HS ‐ CES 2001 Pre‐2004 12 3 985,001,268$ 625,420,196$ 0.28%



RFMSII HI ‐ 125 CLTV 2003 Pre‐2004 12 4 1,042,007,690$ 859,022,067$ 0.76%

RFMSII HI ‐ 125 CLTV 2000 Pre‐2004 12 5 2,636,318,548$ 2,340,431,633$ 0.81%



RFMSI SA ‐ ARM 2003 Pre‐2004 No Sheet

TOTAL Pre‐2004 95,278,323,692$ 3.41%

Page 1 of 3




Issue Mth

Deal


90+ Day & FC &

REO Delq

% Curr Bal

RFMSI S ‐ 15 Yr FRM 2004 2004‐2007 12 6 868,881,829$ 744,384,875$ 0.00%

RFMSI SA ‐ ARM 2004 2004‐2007 12 1 250,052,758$ 204,860,794$ 0.00%

RFMSI S ‐ 15 Yr FRM 2005 2004‐2007 12 2 386,415,265$ 340,051,174$ 0.00%

RFMSI S ‐ 15 Yr FRM 2006 2004‐2007 12 2 419,584,468$ 341,067,256$ 0.00%

RALI QS ‐ 15 Yr FRM 2007 2004‐2007 12 1 97,351,539$ 83,271,255$ 0.50%

RFMSI S ‐ 15 Yr FRM 2007 2004‐2007 12 3 146,222,744$ 123,139,820$ 0.38%

RFMSI S ‐30 Yr FRM 2004 2004‐2007 12 7 2,502,195,211$ 2,067,961,424$ 0.06%

RFMSI S ‐30 Yr FRM 2005 2004‐2007 12 8 2,441,753,199$ 2,243,349,920$ 0.12%

RFMSI SA ‐ ARM 2005 2004‐2007 12 5 2,676,963,213$ 2,283,344,294$ 0.13%

RFMSI S ‐30 Yr FRM 2006 2004‐2007 12 12 6,380,807,461$ 5,534,485,759$ 0.24%

RALI QS ‐ 15 Yr FRM 2004 2004‐2007 12 5 702,581,251$ 590,128,321$ 0.13%

RALI QS ‐ 15 Yr FRM 2005 2004‐2007 12 3 338,991,135$ 291,832,118$ 0.30%

RALI QS ‐ 15 Yr FRM 2006 2004‐2007 12 4 446,277,599$ 378,390,618$ 0.42%

RALI QS ‐ 30 Yr FRM 2004 2004‐2007 12 12 3,662,371,110$ 2,822,432,158$ 0.42%

RFMSI SA ‐ ARM 2006 2004‐2007 12 4 1,699,516,781$ 1,425,857,835$ 1.04%

RALI QS ‐ 30 Yr FRM 2005 2004‐2007 12 16 5,762,944,371$ 4,817,911,156$ 0.66%

RALI QA ‐ ARM 2004 2004‐2007 12 6 $ 2,222,050,021 $ 1,585,675,851 0.51%

RALI QO ‐ ARM 2005 2004‐2007 12 5 3,708,889,469$ 3,028,072,258$ 0.42%

RALI QA ‐ ARM 2005 2004‐2007 12 13 6,380,157,493$ 4,919,610,442$ 0.77%

RFMSI S ‐30 Yr FRM 2007 2004‐2007 12 9 4,051,078,712$ 3,787,295,984$ 1.17%

RASC KS ‐ FRM 2004 2004‐2007 12 12 2,290,003,927$ 1,857,193,279$ 1.57%

RAMP RS ‐ FRM 2005 2004‐2007 12 9 1,570,722,582$ 1,280,104,412$ 1.88%

RALI QH ‐ ARM 2006 2004‐2007 12 1 340,487,638$ 317,106,521$ 3.28%

RAMP RZ ‐ FRM 2004 2004‐2007 12 4 915,001,304$ 716,628,397$ 1.45%

RAMP RS ‐ FRM 2004 2004‐2007 12 12 4,175,005,102$ 3,211,323,198$ 1.69%

RASC KS ‐ FRM 2005 2004‐2007 12 12 1,440,489,275$ 1,193,916,665$ 2.69%

RALI QO ‐ ARM 2006 2004‐2007 12 10 10,101,365,345$ 8,929,712,386$ 2.70%

RAMP RZ ‐ FRM 2005 2004‐2007 12 4 308,114,946$ 262,183,722$ 2.44%

RAMP RZ ‐ ARM 2004 2004‐2007 12 4 685,017,067$ 515,459,652$ 1.97%

RFMSI SA ‐ ARM 2007 2004‐2007 12 4 1,474,619,663$ 1,317,036,207$ 3.07%

RALI QS ‐ 30 Yr FRM 2006 2004‐2007 12 18 12,312,652,224$ 10,122,196,185$ 3.49%

RASC KS ‐ ARM 2004 2004‐2007 12 12 7,800,004,501$ 5,619,699,191$ 3.13%

RAMP RZ ‐ FRM 2006 2004‐2007 12 5 689,737,543$ 619,996,187$ 3.94%

RAMP RZ ‐ ARM 2005 2004‐2007 12 4 1,007,295,593$ 824,836,869$ 3.55%

RASC KS ‐ FRM 2006 2004‐2007 12 9 1,527,435,360$ 1,290,436,374$ 5.41%

RALI QA ‐ ARM 2006 2004‐2007 12 11 5,554,418,370$ 4,504,963,952$ 6.28%

RAMP RS ‐ ARM 2005 2004‐2007 12 9 5,464,282,280$ 4,103,626,011$ 4.98%

RAMP RS ‐ ARM 2004 2004‐2007 12 12 8,040,021,691$ 6,197,378,136$ 5.27%

RAMP RS ‐ FRM 2006 2004‐2007 12 6 1,687,209,733$ 1,362,035,662$ 5.24%

RAMP RZ ‐ FRM 2007 2004‐2007 12 1 99,886,968$ 90,061,248$ 6.32%

RASC KS ‐ ARM 2005 2004‐2007 12 12 7,620,783,002$ 5,680,409,729$ 6.10%

RALI QO ‐ ARM 2007 2004‐2007 12 5 2,211,048,783$ 2,084,083,430$ 8.00%

RALI QS ‐ 30 Yr FRM 2007 2004‐2007 12 11 7,599,672,338$ 6,867,864,734$ 7.07%

RALI QH ‐ ARM 2007 2004‐2007 12 9 4,232,718,503$ 4,087,541,897$ 7.47%

RASC KS ‐ FRM 2007 2004‐2007 12 4 804,057,336$ 731,093,762$ 9.00%

RAMP RZ ‐ ARM 2006 2004‐2007 12 5 2,297,037,625$ 1,926,433,207$ 10.06%

RAMP RS ‐ ARM 2006 2004‐2007 12 6 2,753,023,961$ 2,154,863,598$ 10.55%

RASC KS ‐ ARM 2006 2004‐2007 12 9 5,762,450,629$ 4,582,615,995$ 11.35%

RAMP RZ ‐ ARM 2007 2004‐2007 12 1 240,449,694$ 194,726,645$ 14.55%

RAMP RS ‐ FRM 2007 2004‐2007 12 2 542,044,050$ 476,377,192$ 15.45%

RALI QA ‐ ARM 2007 2004‐2007 12 5 2,422,605,271$ 2,144,218,614$ 17.45%

RAMP RS ‐ ARM 2007 2004‐2007 12 2 341,002,371$ 303,018,201$ 18.63%

RASC KS ‐ ARM 2007 2004‐2007 12 4 2,204,033,433$ 1,932,292,683$ 20.40%

Page 2 of 3




Issue Mth

Deal


90+ Day & FC &

REO Delq

% Curr Bal

RFMSII HSA ‐ HELOC 2005 2004‐2007 12 1 100,973,458$ 59,445,355$ 1.56%

RFMSII HS ‐ HELOC 2005 2004‐2007 12 2 525,000,326$ 296,993,085$ 0.72%

RFMSII HS ‐ HELOC 2004 2004‐2007 12 3 720,000,109$ 430,847,271$ 0.26%

RFMSII HSA ‐ HELOC 2006 2004‐2007 12 4 1,044,310,329$ 673,917,056$ 2.22%

RFMSII HSA ‐ HELOC 2007 2004‐2007 12 2 784,105,449$ 581,319,792$ 5.51%

RFMSII HS ‐ CES 2004 2004‐2007 12 2 635,000,276$ 438,990,801$ 0.31%

RFMSII HI ‐ 125 CLTV 2004 2004‐2007 12 3 730,015,540$ 585,009,719$ 0.79%

RFMSII HS ‐ CES 2005 2004‐2007 12 2 900,000,035$ 694,318,106$ 0.25%

RFMSII HSA ‐ CES 2005 2004‐2007 12 1 178,529,931$ 129,571,020$ 1.11%

RFMSII HI ‐ 125 CLTV 2005 2004‐2007 12 3 705,002,822$ 576,387,924$ 0.78%

RFMSII HSA ‐ CES 2006 2004‐2007 12 2 763,769,849$ 576,499,330$ 0.71%

RFMSII HI ‐ 125 CLTV 2006 2004‐2007 12 5 1,204,469,744$ 1,059,274,961$ 1.08%

RFMSII HSA ‐ CES 2007 2004‐2007 12 2 1,891,463,190$ 1,501,889,357$ 7.31%

RFMSII HI ‐ 125 CLTV 2007 2004‐2007 12 1 257,532,198$ 236,395,082$ 1.45%

TOTAL 2004‐2007 132,955,416,113$ 4.23%

TOTAL COMBINED 228,233,739,805$ 3.89%

Ratio: 80.7%

2004‐2007: 43.5%

Pre‐2004: 35.1%

Page 3 of 3

ANNEX F


Trust Level Model

Monoline Estimated Loss


Shelf Series Class Curr OB

Collateral Cum Loss To

Date

Monoline Payment

To Date Bonds Loss To Date

Lower Range ‐ Bonds

Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR

Higher Range ‐ Bonds

Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher

Range ‐ CRRMonoline CE %

GMACM 2004‐HE1 A1 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HE1 A2 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HE1 A3 97,747,569 ‐ 920,856 ‐$ 3,287,780$ FGICGMACM 2004‐HE1 VPRN 67,380,121 ‐ 634,773 ‐$ 2,135,951$ FGICGMACM 2004‐HE1 Total 167,920,024 72,368,422 ‐ 1,555,630 ‐$ 12,880,947 85,249,369$ 2.00% 99.00% 8.00% 5,423,731$ 20,399,000 92,767,422 3.00% 100.00% 6.00% FGIC 1.66%GMACM 2004‐HE3 Total 168,811,827 80,546,315 18,691,890 ‐ 2,742,524$ 7,696,057 88,242,372$ 2.00% 90.00% 10.00% 6,652,133$ 11,207,542 91,753,857 3.00% 95.00% 7.00% FSA 0.00%GMACM 2004‐HE3 A1 ‐ ‐ ‐ ‐$ ‐$ FSAGMACM 2004‐HE3 A2 ‐ ‐ ‐ ‐$ ‐$ FSAGMACM 2004‐HE3 A3 98,334,904 11,098,742 ‐ 1,638,809$ 3,980,302$ FSAGMACM 2004‐HE3 A1VN ‐ ‐ ‐ ‐$ ‐$ FSAGMACM 2004‐HE3 A2VN 67,275,321 7,593,148 ‐ 1,103,715$ 2,671,831$ FSAGMACM 2004‐HE3 A3VN ‐ ‐ ‐ ‐$ ‐$ FSAGMACM 2004‐HE4 Total 179,534,395 103,103,306 36,140,625 ‐ 3,380,461$ 11,459,858 114,563,165$ 2.00% 93.00% 10.00% 8,256,622$ 19,048,742 122,152,048 3.00% 99.00% 7.00% MBIA 0.00%GMACM 2004‐HE4 A1 ‐ ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐HE4 A2 ‐ ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐HE4 A3 94,228,042 18,968,289 ‐ 1,792,040$ 4,388,692$ MBIAGMACM 2004‐HE4 A1VN ‐ ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐HE4 A2VN 85,306,353 17,172,336 ‐ 1,588,421$ 3,867,930$ MBIAGMACM 2004‐HE4 A3VN ‐ ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐HE5 Total 44,349,329 17,112,486 ‐ ‐ ‐$ 3,164,245 20,276,730$ 1.59% 104.70% 8.66% ‐$ 3,377,805 20,490,291 1.66% 104.94% 8.11% FGIC 6.40%GMACM 2004‐HE5 A1 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HE5 A2 ‐ ‐ ‐ ‐$ ‐$ FGIC

GMACM 2004‐HE5 A3 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HE5 A4 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HE5 A5 35,634,099 ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HE5 A6 5,877,239 ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HE5 AIO ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI2 Total 25,382,300 37,300,736 ‐ ‐ ‐$ 2,127,370 39,428,106$ 2.26% 109.74% 9.19% ‐$ 2,240,111 39,540,847 2.33% 109.74% 8.42% FGIC 12.40%RFMSII 2004‐HI2 A1 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI2 A2 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI2 A3 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI2 A4 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI2 A5 35,634,099 ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI2 A6 5,877,239 ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI2 AIO ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI3 Total 24,294,581 32,141,006 ‐ ‐ ‐$ 1,563,701 33,704,707$ 1.73% 109.83% 11.18% ‐$ 1,672,028 33,813,034 1.79% 109.83% 10.14% FGIC 13.64%RFMSII 2004‐HI3 A1 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI3 A2 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI3 A3 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI3 A4 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HI3 A5 20,980,285 ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HLTV1 Total 16,536,830 20,046,152 ‐ ‐ ‐$ 956,082 21,002,233$ 1.23% 102.57% 10.96% ‐$ 1,069,815 21,115,967 1.31% 102.76% 9.77% FGIC 17.29%

GMACM 2004‐HLTV1 A1 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HLTV1 A2 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HLTV1 A3 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2004‐HLTV1 A4 13,677,385 ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HS1 Total 20,680,200 13,037,183 ‐ ‐ ‐$ 667,857 13,705,040$ 1.03% 108.97% 13.76% ‐$ 725,331 13,762,514 1.09% 108.93% 12.87% FGIC 11.00%/12.40%RFMSII 2004‐HS1 AIIO ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2004‐HS1 AI6 2,186,542 ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2004‐HS1 AI1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2004‐HS1 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2004‐HS1 AI3 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HS1 AI4 2,848,384 ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HS1 AI5 7,300,000 ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HS1 AII 5,064,106 ‐ ‐ ‐$ ‐$ FGICRFMSII 2004‐HS2 Total 32,502,285 19,130,618 1,675,855 ‐ 707,751$ 1,729,401 20,860,019$ 1.53% 105.79% 12.13% 745,470$ 1,815,032 20,945,650 1.56% 105.90% 11.30% MBIA 0%/1.39%RFMSII 2004‐HS2 AI1 ‐ ‐ ‐ ‐$ ‐$ MBIARFMSII 2004‐HS2 AI2 ‐ ‐ ‐ ‐$ ‐$ MBIARFMSII 2004‐HS2 AI3 ‐ ‐ ‐ ‐$ ‐$ MBIARFMSII 2004‐HS2 AI4 ‐ ‐ ‐ ‐$ ‐$ MBIARFMSII 2004‐HS2 AI5 17,980,425 927,091 ‐ 682,232$ 717,976$ MBIARFMSII 2004‐HS2 AI6 1,636,052 84,357 ‐ 25,519$ 27,494$ MBIARFMSII 2004‐HS2 AII 11,636,809 664,407 ‐ ‐$ ‐$ MBIARFMSII 2004‐HS3 Total 13,987,300 10,353,150 ‐ ‐ ‐$ 741,183 11,094,333$ 1.22% 102.53% 11.05% ‐$ 794,774 11,147,924 1.28% 102.62% 10.30% FGIC 1.35%RFMSII 2004‐HS3 A 12,600,282 ‐ ‐ ‐$ ‐$ FGICRASC 2004‐KS4 AI6 8,092,111 43,618 ‐ 44,335$ 61,030$ AmbacRASC 2004‐KS4 AI1 ‐ ‐ ‐$ ‐$ AmbacRASC 2004‐KS4 AI2 ‐ ‐ ‐$ ‐$ AmbacRASC 2004‐KS4 AI3 ‐ ‐ ‐$ ‐$ AmbacRASC 2004‐KS4 AI4 9,595,465 51,721 ‐ 93$ 835$ Ambac

Page 1 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RASC 2004‐KS4 AI5 29,700,000 160,088 ‐ 1,356,045$ 1,698,438$ AmbacRASC 2004‐KS4 A2A 14,373,704 77,477 830,783 429,707$ 520,665$ AmbacRASC 2004‐KS4 A2B1 ‐ ‐ ‐$ ‐$ AmbacRASC 2004‐KS4 A2B2 ‐ ‐ ‐$ ‐$ AmbacRASC 2004‐KS4 A2B3 17,774,603 95,808 981,150 3,742,439$ 4,224,809$ AmbacRASC 2004‐KS4 Total 76,934,910 52,487,248 428,712 1,811,933 5,572,619$ 7,846,870 60,334,118$ 1.99% 66.72% 4.45% 6,505,777$ 8,586,723 61,039,983 2.10% 68.36% 4.13% Ambac 3.15%/0%RASC 2004‐KS7 Total 69,362,406 44,466,644 ‐ 4,248,427.00 7,990,972$ 10,232,873 54,699,517$ 2.79% 70.54% 4.20% 8,824,581$ 11,025,110 55,453,990 2.92% 71.93% 3.90% FGIC 0%/0%RASC 2004‐KS7 AI1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RASC 2004‐KS7 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RASC 2004‐KS7 AI3 ‐ ‐ ‐ ‐$ ‐$ FGICRASC 2004‐KS7 AI4 11,634,081 ‐ 50,958 122,605$ 193,290$ FGICRASC 2004‐KS7 AI5 21,114,184 ‐ 85,816 3,393,952$ 3,821,586$ FGIC

RASC 2004‐KS7 AI6 6,029,235 ‐

28,107

134,658$ 174,639$ FGIC

RASC 2004‐KS7 A2A 16,086,883 ‐

1,984,380

2,123,503$ 2,257,616$ FGICRASC 2004‐KS7 A2B1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RASC 2004‐KS7 A2B2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RASC 2004‐KS7 A2B3 16,834,791 ‐

2,099,166

2,216,254$ 2,377,450$ FGICRASC 2004‐KS9 Total 53,596,000 33,635,636 ‐ 3,988,990 3,037,370$ 8,401,285 42,036,921$ 2.62% 74.66% 3.05% 3,632,007$ 8,991,522 42,627,158 2.73% 76.17% 2.80% FGIC 0%/0%RASC 2004‐KS9 AI1 ‐ ‐ ‐ ‐$ ‐$ FGICRASC 2004‐KS9 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RASC 2004‐KS9 AI3 ‐ ‐ ‐ ‐$ ‐$ FGICRASC 2004‐KS9 AI4 5,174,464 ‐ 133,019 ‐$ 260$ FGIC

RASC 2004‐KS9 AI5 17,875,973 ‐

424,027

2,038,987$ 2,282,852$ FGICRASC 2004‐KS9 AI6 5,947,705 ‐ 161,054 68,261$ 95,077$ FGICRASC 2004‐KS9 AII1 ‐ ‐ ‐$ ‐$ FGIC

RASC 2004‐KS9 AII2 ‐ ‐ ‐$ ‐$ FGICRASC 2004‐KS9 AII3 11,158,449 ‐ 1,453,729 430,766$ 574,561$ FGICRASC 2004‐KS9 AII4 13,948,062 ‐ 1,817,161 499,356$ 679,258$ FGIC

RAMP 2004‐RS1 AI6B 21,746,752 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 MII1 19,225,889 ‐ ‐ ‐$ ‐$ Ambac

Page 2 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RAMP 2004‐RS1 AI6A 16,886,053 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 MII2 8,797,661 ‐ ‐ 1,001,258$ 1,164,109$ AmbacRAMP 2004‐RS1 MI1 8,099,314 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 AI7 7,123,967 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 MI2 4,198,266 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 MI3 944,383 ‐ 2,260,304 ‐$ ‐$ Ambac

RAMP 2004‐RS1 MII3 587,730 ‐ 1,448,912 587,730$ 587,730$ Ambac

RAMP 2004‐RS1 AI1 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 AI2 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 AI3 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 AI4 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 AI5 ‐ ‐ ‐ ‐$ ‐$ AmbacRAMP 2004‐RS1 MI4 ‐ ‐ 5,432,345 ‐$ ‐$ Ambac

RAMP 2004‐RS1 AIIA ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 AIIB ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS1 MII4 ‐ ‐ 2,789,984 ‐$ ‐$ Ambac

RAMP 2004‐RS1 MII5 ‐ ‐ 2,966,501 ‐$ ‐$ Ambac

RAMP 2004‐RS1 MII6 ‐ ‐ 6,194,908 ‐$ ‐$ Ambac

RAMP 2004‐RS1 Total 87,343,300 81,504,648 ‐ 21,092,955 1,588,988$ 7,229,517 88,734,165$ 1.71% 68.02% 5.92% 1,751,839$ 8,034,466 89,539,114 1.83% 69.45% 5.53% Ambac 25.13%

RAMP 2004‐RS5 AI6 9,280,347 45,652 171,506 360,470$ 450,937$ Ambac

RAMP 2004‐RS5 AI1 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS5 AI2 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS5 AI3 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS5 AI4 ‐ ‐ ‐$ ‐$ Ambac

Page 3 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RAMP 2004‐RS5 AI5 55,684,418 273,921 851,683 7,722,679$ 8,834,106$ Ambac

RAMP 2004‐RS5 MII1 25,949,117 ‐ 623,596$ 692,587$ Ambac

RAMP 2004‐RS5 MII2 8,584,993 ‐ 4,281,816$ 4,537,161$ AmbacRAMP 2004‐RS5 MII3 2,101,300 ‐ 2,101,300$ 2,101,300$ Ambac

RAMP 2004‐RS5 MII4 1,897,431

1,113,107

1,897,431$ 1,897,431$ Ambac

RAMP 2004‐RS5 A2A ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS5 A2B1 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS5 A2B2 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS5 A2B3 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS5 MII5 ‐ 2,941,033 ‐$ ‐$ Ambac

RAMP 2004‐RS5 Total 100,487,000 72,405,816 319,573 5,077,329 16,987,291$ 14,771,337 87,177,153$ 2.76% 67.36% 4.81% 18,513,521$ 16,523,008 88,928,824 2.95% 69.86% 4.49% Ambac 0.00%RAMP 2004‐RS7 AI1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RAMP 2004‐RS7 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RAMP 2004‐RS7 AI3 ‐ ‐ ‐ ‐$ ‐$ FGIC

RAMP 2004‐RS7 AI4 15,648,379 ‐ 863,731 495,563$ 560,586$ FGIC

RAMP 2004‐RS7 AI5 53,278,528 ‐ 2,051,472 10,819,022$ 11,637,422$ FGIC

RAMP 2004‐RS7 AI6 8,695,475 ‐ 525,380 562,938$ 622,147$ FGIC

RAMP 2004‐RS7 A2A 17,693,238 ‐ 2,016,555 2,515,792$ 2,663,273$ FGIC

RAMP 2004‐RS7 A2B1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RAMP 2004‐RS7 A2B2 12,537,764 ‐ 2,534,376 1,555,811$ 1,651,965$ FGIC

RAMP 2004‐RS7 A3 23,482,583 ‐ 565,015 1,171,417$ 1,366,559$ FGICRAMP 2004‐RS7 Total 128,740,313 81,547,166 ‐ 8,556,529 17,120,543$ 20,764,468 102,311,634$ 2.90% 71.09% 4.79% 18,501,951$ 22,285,082 103,777,658 3.04% 71.84% 4.46% FGIC 0.00%

RAMP 2004‐RS9 MII1 29,376,197 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 AI1 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 AI2 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 AI3 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 AI4 5,341,529 38,518 173,576 ‐$ ‐$ Ambac

RAMP 2004‐RS9 AI5 36,420,808 262,630 927,128 2,225,305$ 2,599,331$ Ambac

RAMP 2004‐RS9 AI6 9,022,853 65,064 285,293 59,796$ 79,293$ Ambac

RAMP 2004‐RS9 MII2 10,431,546 ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 MII3 5,107,686 ‐ ‐ 3,646,915$ 4,356,373$ AmbacRAMP 2004‐RS9 MII4 1,652,275 ‐ 3,555,945 1,652,275$ 1,652,275$ Ambac

RAMP 2004‐RS9 AII1 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 AII2 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 AII3 ‐ ‐ ‐ ‐$ ‐$ Ambac

RAMP 2004‐RS9 MII5 ‐ ‐ 4,506,631 ‐$ ‐$ Ambac

RAMP 2004‐RS9 Total 95,231,600 89,536,504 366,211 9,448,573 7,584,291$ 9,090,644 98,627,148$ 2.09% 68.75% 7.49% 8,687,272$ 10,173,097 99,709,601 2.18% 70.77% 6.69% Ambac 0.00%RAMP 2004‐RZ2 Total 54,656,300 29,608,290 ‐ 4,797,309 7,085,430$ 5,110,705 34,718,995$ 2.23% 66.27% 8.24% 7,506,209$ 5,640,104 35,248,394 2.31% 67.97% 7.50% FGIC 0%/0.52%RAMP 2004‐RZ2 AIO ‐ ‐ ‐ ‐$ ‐$ FGICRAMP 2004‐RZ2 AI6 8,947,673 ‐ 1,271,248 190,081$ 227,189$ FGICRAMP 2004‐RZ2 AI1 ‐ ‐ ‐ ‐$ ‐$ FGICRAMP 2004‐RZ2 AI2 ‐ ‐ ‐ ‐$ ‐$ FGICRAMP 2004‐RZ2 AI3 ‐ ‐ ‐ ‐$ ‐$ FGICRAMP 2004‐RZ2 AI4 12,406,553 ‐ 1,309,854 166,742$ 207,430$ FGICRAMP 2004‐RZ2 AI5 26,683,793 ‐ 2,216,207 6,728,606$ 7,071,590$ FGICRAMP 2004‐RZ2 AII 7,873,710 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2004‐S4 1AV 33,647,388 ‐ ‐ ‐$ ‐$ MBIA

Page 4 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFMSI 2004‐S4 2AV 13,122,082 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1A4 11,560,765 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2A7 4,278,748 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2A6 3,308,683 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2A8 2,476,127 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2A3 1,706,734 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1M1 1,696,570 ‐ ‐ 115$ 123$ MBIA

RFMSI 2004‐S4 2A2 1,014,237 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1A8 692,364 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1M2 589,516 ‐ ‐ 271,697$ 277,132$ MBIA

RFMSI 2004‐S4 1A3 18,909,000 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2M1 151,415 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1M3 135,078 ‐ 164,815 130,629$ 130,711$ MBIA

RFMSI 2004‐S4 2M2 60,554 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2M3 45,391 ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2B3 30,299 ‐ 351 19,900$ 19,746$ MBIA

RFMSI 2004‐S4 2B1 30,267 ‐ ‐ 1,352$ 1,098$ MBIA

RFMSI 2004‐S4 2B2 15,124 ‐ ‐ 4,605$ 4,371$ MBIA

RFMSI 2004‐S4 1A1 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1A2 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1A5 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1A6 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1A7 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2A1 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2A4 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 2A5 ‐ ‐ ‐ ‐$ ‐$ MBIA

RFMSI 2004‐S4 1B1 ‐ ‐ 314,252 ‐$ ‐$ MBIA

RFMSI 2004‐S4 1B2 ‐ ‐ 219,708 ‐$ ‐$ MBIA

RFMSI 2004‐S4 1B3 ‐ ‐ 256,701 ‐$ ‐$ MBIA

RFMSI 2004‐S4 Total 47,036,800 950,675 ‐ 955,827 428,297$ 272,246 1,222,921$ 0.15% 57.92% 8.00% 433,181$ 273,730 1,224,405 0.15% 55.96% 7.49% MBIA 7.20%

GMACM 2004‐VFT Total 101,305,729 31,316,190 ‐ ‐ ‐$ 8,561,489 39,877,678$ 2.00% 95.00% 10.00% ‐$ 14,404,161 45,511,751 3.00% 98.00% 7.00% MBIA 22.26%

GMACM 2004‐VFT A1 78,757,785 ‐ ‐ ‐$ ‐$ MBIARAMP 2005‐EFC7 AI1 ‐ FGICRAMP 2005‐EFC7 AI2 ‐ FGICRAMP 2005‐EFC7 AI3 37,671,785 3,587,505 9,179,882 FGICRAMP 2005‐EFC7 AI4 58,068,756 5,529,920 9,085,080 9,124,934$ 11,219,860$ FGICRAMP 2005‐EFC7 A2 32,832,436 3,126,651 8,267,560 2,730,108$ 3,361,785$ FGICRAMP 2005‐EFC7 Total 128,609,888 143,956,196 12,244,076 26,532,522 11,855,042$ 32,526,320 176,482,516$ 3.99% 65.62% 1.59% 14,581,645$ 34,970,114 178,918,978 4.24% 67.04% 1.48% FGIC 0.00%RASC 2005‐EMX5 A1 ‐ ‐ ‐ ‐$ ‐$ FGICRASC 2005‐EMX5 A2 29,501,252 8,311,029 12,503,475 ‐$ 12,991$ FGICRASC 2005‐EMX5 A3 34,155,212 9,622,132 9,048,338 4,926,903$ 5,869,726$ FGICRASC 2005‐EMX5 Total 65,022,000 101,494,616 17,933,161 21,551,813 4,926,903$ 20,289,661 121,784,277$ 4.35% 87.44% 2.86% 5,882,717$ 21,362,556 122,857,172 4.46% 88.65% 2.57% FGIC 0.00%GMACM 2005‐HE1 A1 ‐ ‐ ‐ ‐$ ‐$ FGIC

GMACM 2005‐HE1 A2 106,178,190 6,174,990 13,019,543 1,681,193$ 4,416,170$ FGIC

GMACM 2005‐HE1 A3 90,924,911 5,287,908 11,149,190 1,439,677$ 3,781,755$ FGIC

GMACM 2005‐HE1 A1VN 11,387,036 662,234 1,396,276 180,299$ 473,610$ FGIC

GMACM 2005‐HE1 A2VN ‐ ‐ ‐ ‐$ ‐$ FGIC

GMACM 2005‐HE1 A3VN ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2005‐HE1 Total 209,096,697 128,660,543 12,125,132 25,565,008 3,301,168$ 15,185,192 143,845,735$ 2.00% 90.00% 10.00% 8,671,535$ 25,968,314 154,362,166 3.00% 95.00% 7.00% FGIC 0.00%GMACM 2005‐HE2 Total 114,102,000 44,595,335 ‐ 3,190,404 1,987,286$ 7,810,001 52,405,336$ 1.58% 105.44% 11.08% 2,656,066$ 8,655,252 53,250,588 1.69% 105.62% 10.11% FGIC 0.00%

GMACM 2005‐HE2 AIO ‐ ‐ ‐ ‐$ ‐$ FGIC

GMACM 2005‐HE2 A1 ‐ ‐ ‐ ‐$ ‐$ FGICGMACM 2005‐HE2 A2 ‐ ‐ ‐ ‐$ ‐$ FGIC

GMACM 2005‐HE2 A3 ‐ ‐ ‐ ‐$ ‐$ FGIC

GMACM 2005‐HE2 A4 66,216,355 ‐ 1,850,137 1,152,437$ 1,540,266$ FGIC

GMACM 2005‐HE2 A5 28,833,569 ‐ 805,621 501,823$ 670,701$ FGICGMACM 2005‐HE2 A6 19,134,910 ‐ 534,646 333,026$ 445,099$ FGIC

GMACM 2005‐HE3 A1 ‐ ‐ ‐ ‐$ ‐$ AmbacGMACM 2005‐HE3 A2 128,539,943 28,961,668 13,263,332 1,988,028$ 4,273,533$ Ambac

GMACM 2005‐HE3 A3 105,606,820 23,794,547 10,896,989 1,633,339$ 3,511,082$ Ambac

Page 5 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


GMACM 2005‐HE3 A1VN 64,230,439 14,471,927 6,627,587 993,403$ 2,135,452$ Ambac

GMACM 2005‐HE3 A2VN ‐ ‐ 0 ‐$ ‐$ AmbacGMACM 2005‐HE3 A3VN ‐ ‐ 0 ‐$ ‐$ Ambac

GMACM 2005‐HE3 Total 298,377,202 189,242,815 67,228,142 30,787,908 4,614,770$ 6,991,130 196,233,945$ 2.00% 90.00% 9.00% 9,920,068$ 19,968,962 208,498,200 3.00% 92.00% 8.00% Ambac 0.00%RFMSII 2005‐HI1 Total 30,778,800 37,017,148 ‐ ‐ ‐$ 3,016,282 40,033,430$ 2.55% 109.60% 9.84% ‐$ 3,252,984 40,270,132 2.68% 109.62% 8.95% FGIC 15.64%

RFMSII 2005‐HI1 A1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HI1 A2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HI1 A3 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HI1 A4 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSII 2005‐HI1 A5 25,965,961 ‐ ‐ ‐$ ‐$ FGICRFMSII 2005‐HS1 AI1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HS1 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HS1 AI3 38,311,579 3,513,115 ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HS1 AI4 49,232,000 4,514,501 ‐ 33,282,431$ 33,849,125$ FGIC

RFMSII 2005‐HS1 AI5 20,089,167 1,842,147 ‐ 547,018$ 568,223$ FGIC

RFMSII 2005‐HS1 AII 42,979,477 3,941,154 ‐ 14,665,016$ 15,062,433$ FGICRFMSII 2005‐HS1 Total 119,885,000 101,797,208 13,810,917 ‐ 48,494,466$ 9,991,903 111,789,111$ 2.12% 105.54% 14.29% 49,479,781$ 10,684,664 112,481,872 2.18% 105.59% 13.01% FGIC 0.00%RFMSII 2005‐HS2 AI1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HS2 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HS2 AI3 37,748,782 7,954,042 ‐ 12,087,824$ 12,698,300$ FGIC

RFMSII 2005‐HS2 AI4 26,450,000 5,573,277 ‐ 26,450,000$ 26,450,000$ FGICRFMSII 2005‐HS2 AI5 17,922,200 3,776,385 ‐ 1,499,931$ 1,486,243$ FGICRFMSII 2005‐HS2 AII 35,828,933 7,549,511 ‐ 18,544,567$ 18,861,279$ FGICRFMSII 2005‐HS2 Total 80,255,954 100,440,256 24,853,214 ‐ 58,582,322$ 9,075,794 109,516,050$ 2.69% 105.46% 12.25% 59,495,822$ 9,834,832 110,275,088 2.81% 105.49% 11.17% FGIC 0.00%RFMSII 2005‐HSA1 Total 30,371,900 73,263,216 32,655,876 ‐ 38,904,154$ 5,460,400 78,723,616$ 4.25% 107.30% 10.55% 39,188,950$ 5,690,737 78,953,953 4.35% 107.32% 9.90% FGIC 0.00%RFMSII 2005‐HSA1 AI1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HSA1 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2005‐HSA1 AI3 15,463,683 9,204,374 ‐ 11,459,185$ 11,601,896$ FGIC

RFMSII 2005‐HSA1 AI4 13,717,000 8,164,704 ‐ 13,717,000$ 13,717,000$ FGIC

RFMSII 2005‐HSA1 AI5 11,121,944 6,620,062 ‐ 3,727,074$ 3,768,899$ FGICRFMSII 2005‐HSA1 AII 14,560,431 8,666,736 ‐ 10,000,895$ 10,101,154$ FGICRAMP 2005‐NC1 AI1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RAMP 2005‐NC1 AI2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RAMP 2005‐NC1 AI3 41,596,569 6,837,034 9,664,682 197,206$ 289,821$ FGIC

RAMP 2005‐NC1 AI4 51,227,557 8,420,035 7,384,124 8,772,312$ 10,058,405$ FGIC

RAMP 2005‐NC1 AII 84,088,096 13,821,169 16,742,018 8,001,058$ 9,241,506$ FGICRAMP 2005‐NC1 Total 176,912,223 195,351,552 29,078,238 33,790,825 16,970,576$ 39,954,246 235,305,798$ 3.38% 69.95% 2.68% 19,589,732$ 44,156,473 239,508,025 3.64% 71.46% 2.37% FGIC 0.00%RAMP 2005‐RS9 AI3 19,770,898 5,424,434 8,248,907.30 ‐$ ‐$ FGIC

RAMP 2005‐RS9 AI4 122,495,752 33,608,495 18,302,640.14 8,735,990.480$ 9,921,154$ FGIC

RAMP 2005‐RS9 AII 93,216,656 25,575,348 19,301,373.20 5,806,808.950$ 6,590,800$ FGIC

RAMP 2005‐RS9 AI1 ‐ ‐ ‐$ ‐$ FGIC

RAMP 2005‐RS9 AI2 ‐ ‐ ‐$ ‐$ FGICRAMP 2005‐RS9 Total 235,483,305 249,793,280 64,608,278 45,852,920.64 14,542,799$ 42,845,550 292,638,830$ 3.53% 59.15% 3.60% 16,511,954$ 46,124,953 295,918,233 3.74% 59.97% 3.36% FGIC 0.00%RFMSI 2005‐S2 AV 58,044,888 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S2 A2 15,255,841 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S2 A6 14,226,337 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S2 M1 3,447,589 ‐ 137,194 2,762,817$ 3,043,532$ FGIC

RFMSI 2005‐S2 A3 1,203,720 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S2 A1 23,350,000 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S2 A4 721,727 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S2 AP 561,402 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S2 M2 ‐ ‐ 1,468,533 ‐$ ‐$ FGIC

RFMSI 2005‐S2 M3 ‐ ‐ 740,276 ‐$ ‐$ FGICRFMSI 2005‐S2 B1 ‐ ‐ 498,104 ‐$ ‐$ FGIC

RFMSI 2005‐S2 B3 ‐ ‐ 401,973 ‐$ ‐$ FGIC

RFMSI 2005‐S2 B2 ‐ ‐ 374,380 ‐$ ‐$ FGIC

RFMSI 2005‐S2 A5 ‐ ‐ ‐ ‐$ ‐$ FGICRFMSI 2005‐S2 Total 58,183,700 3,922,714 ‐ 3,620,460 2,762,817$ 1,392,773 5,315,488$ 1.18% 30.21% 8.50% 3,043,532$ 1,605,788 5,528,502 1.25% 31.63% 7.81% FGIC 5.94%

RFMSI 2005‐S5 AV 57,203,292 ‐ ‐ ‐$ ‐$ FSARFMSI 2005‐S5 A8 25,326,000 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 A6 15,926,729 ‐ ‐ ‐$ ‐$ FSA

Page 6 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFMSI 2005‐S5 A1 5,368,657 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 A5 5,288,120 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 M1 3,418,414 ‐ 58,184 2,745,130$ 3,000,452$ FSA

RFMSI 2005‐S5 A7 1,122,364 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 A4 388,296 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 A2 25,326,000 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 A3 281,432 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 AP 83,281 ‐ ‐ ‐$ ‐$ FSA

RFMSI 2005‐S5 M2 ‐ ‐ 1,339,005 ‐$ ‐$ FSA

RFMSI 2005‐S5 M3 ‐ ‐ 730,845 ‐$ ‐$ FSA

RFMSI 2005‐S5 B1 ‐ ‐ 488,173 ‐$ ‐$ FSA

RFMSI 2005‐S5 B2 ‐ ‐ 368,566 ‐$ ‐$ FSA

RFMSI 2005‐S5 B3 ‐ ‐ 369,051 ‐$ ‐$ FSA

RFMSI 2005‐S5 Total 57,368,200 3,721,383 ‐ 3,353,825 2,745,130$ 1,298,614 5,019,997$ 1.11% 31.46% 9.23% 3,000,452$ 1,509,153 5,230,536 1.19% 32.73% 8.53% FSA 5.98%RFMSI 2005‐S7 AV 89,583,610 ‐ ‐ ‐$ ‐$ FGIC

RFMSI 2005‐S7 A6 23,155,340 ‐ ‐ 1,192,621$ 1,258,680$ FGIC

RFMSI 2005‐S7 A8 17,135,513 ‐ 148,372 2,139,783$ 2,297,889$ FGICRFMSI 2005‐S7 A5 10,937,630 ‐ 94,706 1,365,827$ 1,466,747$ FGICRFMSI 2005‐S7 A1 6,936,533 ‐ ‐ 632,434$ 698,548$ FGICRFMSI 2005‐S7 A9 2,461,743 ‐ 221,811 1,701,600$ 1,814,730$ FGICRFMSI 2005‐S7 A4 2,252,540 ‐ ‐ 205,397$ 226,934$ FGICRFMSI 2005‐S7 A3 1,836,512 ‐ 15,902 229,333$ 246,278$ FGIC

RFMSI 2005‐S7 A2 24,048,761 873,493 208,239 3,003,069$ 3,224,962$ FGIC

RFMSI 2005‐S7 AP 472,597 ‐ ‐ 25,455$ 28,407$ FGIC

RFMSI 2005‐S7 A7 345,577 ‐ 82,558 337,509$ 338,152$ FGIC

RFMSI 2005‐S7 M1 ‐ ‐ 5,868,203 ‐$ ‐$ FGIC

RFMSI 2005‐S7 M2 ‐ ‐ 2,098,547 ‐$ ‐$ FGIC

RFMSI 2005‐S7 M3 ‐ ‐ 1,055,104 ‐$ ‐$ FGIC

RFMSI 2005‐S7 B1 ‐ ‐ 908,180 ‐$ ‐$ FGIC

RFMSI 2005‐S7 B2 ‐ ‐ 757,311 ‐$ ‐$ FGICRFMSI 2005‐S7 B3 ‐ ‐ 623,646 ‐$ ‐$ FGICRFMSI 2005‐S7 Total 89,940,900 12,553,896 873,493 12,082,579 10,833,027$ 5,077,417 17,631,313$ 2.04% 40.96% 8.30% 11,601,325$ 5,682,238 18,236,134 2.15% 41.75% 7.56% FGIC 0.00%GMACM 2006‐HE1 A 451,180,891 37,548,715 73,430,059 45,248,782$ 75,107,895$ FGICGMACM 2006‐HE1 Total 453,248,309 308,889,156 37,548,715 73,430,059 45,248,782$ 59,478,479 368,367,636$ 3.00% 87.00% 10.00% 75,107,895$ 92,492,636 401,262,337 4.00% 92.00% 7.00% FGIC 0.00%GMACM 2006‐HE2 Total 121,696,000 96,742,991 12,062,112 50,694,790 8,678,432$ 9,831,967 106,574,958$ 1.94% 104.27% 12.88% 9,619,734$ 10,818,443 107,561,434 2.06% 104.51% 11.90% FGIC 0.00%

GMACM 2006‐HE2 A1 ‐ 256,787 ‐$ ‐$ FGIC

GMACM 2006‐HE2 A2 ‐ 1,968,259 ‐$ ‐$ FGIC

GMACM 2006‐HE2 A3 56,519,785 5,595,565 29,047,166 1,220,136$ 1,400,113$ FGICGMACM 2006‐HE2 A4 65,317,422 6,466,547 19,422,578 7,458,296$ 8,219,621$ FGICGMACM 2006‐HE3 A1 ‐ ‐ ‐$ ‐$ FGIC

GMACM 2006‐HE3 A2 44,143,671 5,870,631 16,302,965 428,356$ 777,262$ FGIC

GMACM 2006‐HE3 A3 78,420,493 10,429,077 28,961,946 760,967$ 1,380,792$ FGIC

GMACM 2006‐HE3 A4 39,041,841 5,192,142 14,418,778 378,849$ 687,431$ FGIC

GMACM 2006‐HE3 A5 47,487,863 6,315,372 17,538,029 460,807$ 836,145$ FGICGMACM 2006‐HE3 Total 209,094,000 168,566,510 27,807,222 77,221,718 2,028,979$ 17,191,845 185,758,355$ 1.94% 106.42% 15.04% 3,681,630$ 19,697,368 188,263,878 2.10% 106.56% 13.60% FGIC 0.00%

GMACM 2006‐HE4 Total 429,159,045 323,464,071 235,612,379 ‐ 13,862,207$ 41,038,016 364,502,087$ 3.00% 90.00% 10.00% 30,807,104$ 50,161,741 373,625,812 4.00% 93.00% 8.00% MBIA 0.00%

GMACM 2006‐HE4 A1 192,162,837 107,201,764 ‐ 6,617,235$ 14,400,405$ MBIA

GMACM 2006‐HE4 A2 127,248,796 70,988,208 ‐ 4,381,884$ 9,535,841$ MBIAGMACM 2006‐HE4 A3 102,931,633 57,422,407 ‐ 2,863,088$ 6,870,858$ MBIA

GMACM 2006‐HE4 A1VP ‐ ‐ ‐ ‐$ ‐$ MBIA

GMACM 2006‐HE4 A2VP ‐ ‐ ‐ ‐$ ‐$ MBIA

GMACM 2006‐HE4 A3VP ‐ ‐ ‐ ‐$ ‐$ MBIAGMACM 2006‐HE5 1A1 141,631,690 ‐ ‐ ‐$ ‐$ FGIC

GMACM 2006‐HE5 2A1 ‐ ‐ ‐ ‐$ ‐$ FGIC

GMACM 2006‐HE5 2A2 96,957,048 ‐ ‐ ‐$ ‐$ FGICGMACM 2006‐HE5 Total 261,728,831 213,825,691 ‐ ‐ ‐$ 10,565,077 224,390,768$ 1.35% 103.26% 10.82% ‐$ 11,402,551 225,228,242 1.40% 103.46% 9.81% FGIC 8.84%RFMSII 2006‐HI2 Total 70,497,500 60,756,772 ‐ ‐ 31,451,643$ 8,501,547 69,258,319$ 2.59% 108.63% 8.25% 32,444,176$ 9,159,965 69,916,737 2.72% 108.63% 7.43% FGIC 0.00%

RFMSII 2006‐HI2 A1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2006‐HI2 A2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2006‐HI2 A3 13,493,433 ‐ ‐ ‐$ ‐$ FGIC

Page 7 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFMSII 2006‐HI2 A4 72,434,000 ‐ ‐ 31,451,643$ 32,444,176$ FGICRFMSII 2006‐HI3 Total 72,195,200 57,675,528 ‐ ‐ 12,294,425$ 9,312,343 66,987,871$ 2.54% 108.37% 8.65% 13,041,656$ 10,036,981 67,712,509 2.65% 108.39% 7.75% FGIC 0.00%

RFMSII 2006‐HI3 A1 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2006‐HI3 A2 ‐ ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2006‐HI3 A3 14,735,150 ‐ ‐ ‐$ ‐$ FGICRFMSII 2006‐HI3 A4 65,142,000 ‐ ‐ 12,294,425$ 13,041,657$ FGICRFMSII 2006‐HI4 A1 ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2006‐HI4 A2 ‐ ‐ ‐$ ‐$ FGIC

RFMSII 2006‐HI4 A3 23,872,278 1,734,052 ‐ ‐$ ‐$ FGICRFMSII 2006‐HI4 A4 76,826,000 5,580,544 ‐ 43,188,451$ 44,607,234$ FGICRFMSII 2006‐HI4 Total 78,225,500 86,402,568 7,314,597 ‐ 43,188,451$ 14,340,252 100,742,820$ 3.74% 107.93% 8.86% 44,607,234$ 15,474,795 101,877,363 3.91% 107.96% 7.95% FGIC 0.00%RFMSII 2006‐HI5 A1 ‐ ‐ ‐ ‐ ‐ FGIC

RFMSII 2006‐HI5 A2 ‐ ‐ ‐ ‐ ‐ FGIC

RFMSII 2006‐HI5 A3 29,191,116 328,322 ‐ ‐ ‐ FGICRFMSII 2006‐HI5 A4 77,476,000 871,399 ‐ 47,931,156 49,174,333 FGICRFMSII 2006‐HI5 Total 83,707,200 77,030,264 1,199,721 ‐ 47,931,156$ 15,678,836 92,709,100$ 4.03% 108.29% 8.36% 49,174,333$ 16,662,923 93,693,187 4.15% 108.29% 7.45% FGIC 0.00%GMACM 2006‐HLTV1 Total 55,971,600 50,035,564 ‐ ‐ ‐$ 4,445,759 54,481,323$ 1.81% 105.84% 14.81% ‐$ 5,111,520 55,147,084 1.98% 105.92% 13.52% FGIC 0.00%

GMACM 2006‐HLTV1 A1 ‐ ‐ ‐ ‐ ‐ FGIC



GMACM 2006‐HLTV1 A4 27,264,010 ‐ ‐ ‐ ‐ FGICGMACM 2006‐HLTV1 A5 28,901,170 ‐ ‐ ‐ ‐ FGICRFMSII 2006‐HSA1 Total 72,413,600 103,180,136 42,509,719 32,424,895 6,107,118$ 12,058,088 115,238,224$ 4.00% 106.29% 11.81% 6,693,047$ 12,902,569 116,082,705 4.13% 106.30% 10.62% FGIC 0.00%

RFMSII 2006‐HSA1 A1 ‐ ‐ ‐ ‐ ‐ FGIC

RFMSII 2006‐HSA1 A2 ‐ ‐ 3,686,164 ‐ ‐ FGIC

RFMSII 2006‐HSA1 A3 23,728,118 13,929,367 10,388,601 907,667 1,041,109 FGIC

RFMSII 2006‐HSA1 A4 33,068,374 19,412,475 9,848,627 4,679,880 5,060,168 FGICRFMSII 2006‐HSA1 A5 15,617,112 9,167,877 8,501,504 519,571 591,769 FGICRFMSII 2006‐HSA2 AI1 ‐ ‐ ‐ ‐ ‐ FGIC

RFMSII 2006‐HSA2 AI2 ‐ ‐ ‐ ‐ ‐ FGICRFMSII 2006‐HSA2 AI3 34,995,879 17,828,177 ‐ 25,287,643 27,606,925 FGICRFMSII 2006‐HSA2 AI4 20,949,000 10,672,184 ‐ 20,949,000 20,949,000 FGIC

RFMSII 2006‐HSA2 AI5 18,155,559 9,249,104 ‐ 5,480,865 6,793,859 FGIC

RFMSII 2006‐HSA2 AII 26,143,079 13,318,238 ‐ 19,516,175 21,070,132 FGICRFMSII 2006‐HSA2 Total 63,558,218 124,693,080 51,067,703 ‐ 71,233,683$ 14,376,925 139,070,005$ 5.00% 92.00% 8.00% 76,419,916$ 18,861,621 143,554,701 6.00% 96.00% 6.00% FGIC 0.00%

RFMSII 2006‐HSA3 Total 24,513,461 57,970,296 51,769,902 ‐ 413,828$ 1,456,752 59,427,048$ 1.96% 103.88% 8.22% 493,819$ 1,823,012 62,703,696 2.03% 104.01% 7.58% FSA 0.73%

RFMSII 2006‐HSA3 A 24,333,449 51,769,902 ‐ 413,828 493,819 FSARFMSII 2006‐HSA4 Total 59,299,200 126,816,360 110,487,924 ‐ 3,037,613$ 10,265,806 137,082,166$ 2.86% 104.81% 8.46% 3,476,040$ 11,159,837 137,976,197 3.01% 104.81% 7.64% MBIA 0.00%

RFMSII 2006‐HSA4 A 55,665,741 110,487,924 ‐ 3,037,613 3,476,040 MBIARFMSII 2006‐HSA5 Total 44,802,800 114,097,480 107,897,626 ‐ 5,891,787$ 9,995,420 124,092,900$ 4.51% 104.43% 10.99% 6,351,865$ 10,635,391 124,732,871 4.60% 104.52% 10.00% MBIA 0.00%

RFMSII 2006‐HSA5 A 42,812,413 107,897,626 ‐ 5,891,787 6,351,865 MBIA

RALI 2006‐QH1 A1 92,334,354 5,971,617 ‐ 84 107 Ambac

RALI 2006‐QH1 A2 17,419,698 1,126,599 21,670,936 2,438,860 3,856,099 Ambac

RALI 2006‐QH1 A3 ‐ ‐ 21,042,963 ‐ ‐ Ambac

RALI 2006‐QH1 Total 120,038,000 119,826,680 7,098,216 60,588,899.25 2,438,945$ 21,288,500 141,115,180$ 3.45% 52.16% 2.88% 3,856,206$ 22,993,386 142,820,066 3.71% 52.89% 2.76% Ambac 0.00%

RALI 2006‐QH1 M1 ‐ ‐ 8,342,000 ‐ ‐ Ambac

RALI 2006‐QH1 M2 ‐ ‐ 4,426,000 ‐ ‐ Ambac

RALI 2006‐QH1 M3 ‐ ‐ 1,703,000 ‐ ‐ Ambac

RALI 2006‐QH1 M4 ‐ ‐ 1,702,000 ‐ ‐ Ambac

RALI 2006‐QH1 M5 ‐ ‐ 1,702,000 ‐ ‐ AmbacRASC 2007‐EMX1 A11 ‐ ‐ 1,407,086 ‐ ‐ FGIC

RASC 2007‐EMX1 A12 ‐ ‐ 2,413,677 ‐ ‐ FGIC

RASC 2007‐EMX1 A13 90,989,548 390,659 13,795,939 10,118,628 11,651,215 FGIC

RASC 2007‐EMX1 A14 40,294,097 173,001 6,054,458 18,761,097 20,074,872 FGIC

RASC 2007‐EMX1 A2 117,792,077 505,734 20,303,877 26,870,133 29,520,716 FGICRASC 2007‐EMX1 Total 250,913,000 234,697,392 1,069,394 43,975,036.65 55,749,857$ 99,058,996 333,756,388$ 5.33% 77.94% 1.71% 61,246,803$ 105,276,599 339,973,991 5.58% 79.63% 1.55% FGIC 0.00%GMACM 2007‐HE1 Total 282,517,000 225,916,482 157,697,528 ‐ 6,364,839$ 15,014,891 240,931,373$ 1.33% 103.11% 12.13% 8,086,741$ 17,497,491 243,413,973 1.42% 105.22% 10.11% MBIA 0.00%

GMACM 2007‐HE1 A1 ‐ ‐ ‐ ‐ ‐ MBIA

GMACM 2007‐HE1 A2 ‐ ‐ ‐ ‐ ‐ MBIA

GMACM 2007‐HE1 A3 71,215,595 39,751,726 ‐ 1,555,463 1,983,271 MBIA

GMACM 2007‐HE1 A4 123,871,000 69,143,368 ‐ 2,838,135 3,599,015 MBIA

Page 8 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


GMACM 2007‐HE1 A5 87,430,022 48,802,433 ‐ 1,971,241 2,504,455 MBIAGMACM 2007‐HE2 A1 18,378,965 1,232,841 7,162,708 372,589 488,581 FGIC

GMACM 2007‐HE2 A2 75,893,365 5,090,843 29,577,400 1,515,200 1,986,855 FGIC

GMACM 2007‐HE2 A3 97,534,024 6,542,475 38,011,265 1,947,253 2,553,398 FGIC

GMACM 2007‐HE2 A4 77,188,926 5,177,748 30,082,310 1,541,066 2,020,772 FGIC

GMACM 2007‐HE2 A5 28,378,373 1,903,590 11,059,708 566,570 742,933 FGIC

GMACM 2007‐HE2 A6 55,131,520 3,698,161 21,486,028 1,100,693 1,443,319 FGICGMACM 2007‐HE2 Total 358,932,000 277,281,942 23,645,658 137,379,419 7,043,371$ 23,375,343 300,657,285$ 1.55% 104.36% 14.29% 9,235,858$ 26,637,723 303,919,665 1.66% 104.31% 12.77% FGIC 0.00%RFMSII 2007‐HI1 A1 ‐ ‐ ‐ ‐ ‐ FGIC

RFMSII 2007‐HI1 A2 ‐ ‐ ‐ ‐ ‐ FGIC

RFMSII 2007‐HI1 A3 41,412,224 ‐ ‐ ‐ ‐ FGIC

RFMSII 2007‐HI1 A4 78,740,000 ‐ ‐ 53,956,576 55,611,957 FGICRFMSII 2007‐HI1 Total 92,469,800 82,770,256 ‐ ‐ 53,956,576$ 14,418,733 97,188,989$ 3.39% 108.10% 9.49% 55,611,957$ 15,734,643 98,504,899 3.53% 108.17% 8.27% FGIC 0.00%

RFMSII 2007‐HSA1 Total 98,175,000 244,985,296 235,545,026 ‐ 5,646,108$ 19,050,819 264,036,115$ 2.99% 104.56% 7.14% 7,902,774$ 23,235,607 268,220,903 3.70% 104.41% 6.88% MBIA 0.00%

RFMSII 2007‐HSA1 A 91,726,187 235,545,026 ‐ 5,646,108 7,902,774 MBIARFMSII 2007‐HSA2 Total 219,385,040 726,288,892 778,266,280 ‐ 24,666,496$ 44,560,878 770,849,769$ 3.22% 106.37% 4.30% 27,263,247$ 47,711,738 774,000,630 3.39% 106.39% 3.71% MBIA 0.00%

RFMSII 2007‐HSA2 A1V ‐ ‐ ‐ ‐ ‐ MBIA

RFMSII 2007‐HSA2 A1F ‐ ‐ ‐ ‐ ‐ MBIA

RFMSII 2007‐HSA2 A2 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSII 2007‐HSA2 A3 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSII 2007‐HSA2 A4 44,637,239 158,350,167 ‐ 4,382,736 4,957,372 MBIA

RFMSII 2007‐HSA2 A5 71,956,000 255,263,205 ‐ 13,746,386 14,957,181 MBIA

RFMSII 2007‐HSA2 A6 102,791,801 364,652,907 ‐ 6,537,375 7,348,694 MBIARFMSII 2007‐HSA3 Total 160,736,314 424,692,955 429,603,755 ‐ 25,938,071$ 26,997,073 451,690,028$ 3.75% 106.71% 3.25% 28,064,051$ 28,570,606 453,263,560 3.91% 106.73% 2.72% MBIA 0.00%

RFMSII 2007‐HSA3 AI1 ‐ ‐ ‐ ‐ ‐ MBIA



RFMSII 2007‐HSA3 AI4 31,468,912 84,107,706 ‐ 5,262,628 5,688,159 MBIA



RFMSII 2007‐HSA3 AII 50,216,633 134,215,184 ‐ 7,498,473 8,258,703 MBIA

GMACM 2000‐HE2 Total 3,442,378 19,597,626 3,085,230 ‐ 221,417$ 198,391 19,796,018$ 2.00% 102.00% 6.00% 319,458$ 307,920 19,905,546 3.00% 105.00% 5.00% MBIA 0.00%

GMACM 2000‐HE2 A1 3,317,985 2,967,294 ‐ 216,592 312,569 MBIA

GMACM 2000‐HE2 A2 131,874 117,936 ‐ 4,825 6,889 MBIA

GMACM 2000‐HE4 Total 4,057,665 10,995,712 2,189,373 ‐ 616,045$ 506,158 11,501,870$ 4.00% 103.00% 5.00% 756,772$ 660,404 11,656,116 5.00% 107.00% 4.00% MBIA 0.00%

GMACM 2000‐HE4 A1 3,627,559 1,957,302.72 ‐ 571,564 701,834 MBIA

GMACM 2000‐HE4 A2 430,106 232,070 ‐ 44,480 54,938 MBIAGMACM 2002‐HE4 Total 10,341,976 6,610,637 ‐ ‐ ‐$ 418,575 7,029,212$ 1.00% 95.00% 8.00% ‐$ 869,475 7,480,112 2.00% 98.00% 7.00% FGIC 29.71%

GMACM 2002‐HE4 A1 ‐ ‐ ‐ ‐ ‐ FGIC

GMACM 2002‐HE4 A2 7,269,423 ‐ ‐ ‐ ‐ FGICGMACM 2002‐HE4 AIO ‐ ‐ ‐ ‐ ‐ FGICGMACM 2002‐HE1 Total 14,162,836 10,407,694 ‐ 44,711 316,246$ 1,271,734 11,679,428$ 3.00% 95.00% 15.00% 989,269$ 1,812,762 12,220,456 4.00% 98.00% 13.00% FGIC 0.13%

GMACM 2002‐HE1 A1 ‐ ‐ 11,178 79,062 247,317 FGICGMACM 2002‐HE1 A2 7,269,423 ‐ 33,534 237,185 741,952 FGIC

GMACM 2002‐HE3 Total 34,602,128 14,803,800 ‐ ‐ ‐$ 869,209 15,673,009$ 1.00% 93.00% 18.00% ‐$ 1,848,311 16,652,111 2.00% 95.00% 16.00% MBIA 8.88%

GMACM 2002‐HE3 A1 ‐ ‐ ‐ ‐ ‐ MBIA

GMACM 2002‐HE3 A2 ‐ ‐ ‐ ‐ ‐ MBIA

GMACM 2002‐HE3 A3 ‐ ‐ ‐ ‐ ‐ MBIAGMACM 2003‐HE1 Total 36,788,966 17,735,703 ‐ ‐ ‐$ 820,238 18,555,941$ 1.00% 90.00% 20.00% ‐$ 1,760,687 19,496,390 2.00% 93.00% 18.00% FGIC 2.91%

GMACM 2003‐HE1 A1 ‐ ‐ ‐ ‐ ‐ FGIC

GMACM 2003‐HE1 A2 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2003‐HE1 A3 35,719,160 ‐ ‐ ‐ ‐ FGIC

GMACM 2001‐HLT1 AI1 ‐ ‐ ‐ ‐ ‐ Ambac


GMACM 2001‐HLT1 AI3 ‐ ‐ ‐ ‐ ‐ AmbacGMACM 2001‐HLT1 AI4 ‐ ‐ ‐ ‐ ‐ Ambac



GMACM 2001‐HLT1 AI7 2,906,599 ‐ ‐ ‐ ‐ Ambac

Page 9 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


GMACM 2001‐HLT1 AII ‐ ‐ ‐ ‐ ‐ AmbacGMACM 2001‐HLT1 Total 4,406,599 29,469,015 ‐ ‐ ‐$ 265,940 29,734,955$ 2.00% 93.00% 15.00% ‐$ 440,519 29,909,535 3.00% 95.00% 12.00% Ambac 34.00%GMACM 2001‐HE2 Total 10,045,635 14,582,601 ‐ 1,330,249 ‐$ 1,203,491 15,786,092$ 4.00% 95.00% 3.00% 2,915$ 1,542,508 16,125,109 5.00% 97.00% 2.00% FGIC 2.41%GMACM 2001‐HE2 IA1 3,950,455 ‐ 315,088 ‐ 2,915 FGIC

GMACM 2001‐HE2 IA2 1,975,789 ‐ 1,015,161 ‐ ‐ FGIC

GMACM 2001‐HE2 IIA1 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2001‐HE2 IIA2 ‐ ‐ ‐ ‐ ‐ FGIC

GMACM 2001‐HE2 IIA3 ‐ ‐ ‐ ‐ ‐ FGIC

GMACM 2001‐HE2 IIA4 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2001‐HE2 IIA5 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2001‐HE2 IIA6 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2001‐HE2 IIA7 2,879,579 ‐ ‐ ‐ ‐ FGICGMACM 2001‐HE3 Total 4,620,410 5,359,493 ‐ 1,151,551 ‐$ 421,066 5,780,559$ 2.00% 95.00% 6.00% 194,373$ 648,551 6,008,044 3.00% 97.00% 5.00% FGIC 0.18%GMACM 2001‐HE3 A1 2,680,196 ‐ 753,313 ‐ 120,223 FGICGMACM 2001‐HE3 A2 1,895,866 ‐ 398,237 ‐ 74,150 FGICGMACM 2001‐HLT2 AI 3,827,057 ‐ ‐ ‐ ‐ AmbacGMACM 2001‐HLT2 AII ‐ ‐ ‐ ‐ ‐ AmbacGMACM 2001‐HLT2 Total 4,773,332 16,984,409 ‐ ‐ ‐$ 700,609 17,685,018$ 4.00% 90.00% 8.00% ‐$ 924,766 17,909,175 5.00% 91.00% 6.00% Ambac 19.82%

GMACM 2002‐HLT1 A1 4,979,166 ‐ ‐ ‐ ‐ Ambac

GMACM 2002‐HLT1 A2 ‐ ‐ ‐ ‐ ‐ Ambac

GMACM 2002‐HLT1 Total 6,034,166 19,599,349 ‐ ‐ ‐$ 910,995 20,510,344$ 4.00% 95.00% 8.00% ‐$ 1,212,539 20,811,888 5.00% 97.00% 6.00% Ambac 17.48%RFSC 2002‐RP2 A1 6,619,593 ‐ ‐ ‐ ‐ AmbacRFSC 2002‐RP2 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2002‐RP2 Total 7,486,933 17,801,727 ‐ ‐ ‐$ 1,207,536 19,009,263$ 3.00% 100.00% 6.00% ‐$ 1,783,218 19,584,945 4.00% 105.00% 4.00% Ambac 11.58%GMACM 2003‐GH1 M1 1,450,052 ‐ ‐ ‐ ‐ MBIAGMACM 2003‐GH1 M2 1,318,229 ‐ ‐ ‐ ‐ MBIAGMACM 2003‐GH1 B 970,440 ‐ ‐ ‐ 83,158 MBIAGMACM 2003‐GH1 A1 ‐ ‐ ‐ ‐ ‐ MBIA

GMACM 2003‐GH1 A2 ‐ ‐ ‐ ‐ ‐ MBIAGMACM 2003‐GH1 A3 ‐ ‐ ‐ ‐ ‐ MBIAGMACM 2003‐GH1 A4 ‐ ‐ ‐ ‐ ‐ MBIAGMACM 2003‐GH1 A5 15,323,486 ‐ ‐ ‐ ‐ MBIA

GMACM 2003‐GH1 Total 20,181,700 4,560,366 ‐ ‐ ‐$ 1,471,000 6,031,366$ 1.55% 76.27% 4.90% 83,158$ 1,552,659 6,113,025 1.58% 77.23% 4.51% MBIA 24.07%GMACM 2003‐HE2 Total 15,250,458 6,614,661 ‐ ‐ ‐$ 481,915 7,096,576$ 1.00% 90.00% 12.00% ‐$ 1,062,319 7,676,980 2.00% 95.00% 10.00% FGIC 20.81%GMACM 2003‐HE2 A1 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2003‐HE2 A2 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2003‐HE2 A3 ‐ ‐ ‐ ‐ ‐ FGICGMACM 2003‐HE2 A4 6,548,778 ‐ ‐ ‐ ‐ FGICGMACM 2003‐HE2 A5 5,528,450 ‐ ‐ ‐ ‐ FGICGMACM 2003‐HE2 AIO ‐ ‐ ‐ ‐ ‐ FGICRALI 2002‐QS15 CB 15,145,480 ‐ ‐ ‐ ‐ MBIARALI 2002‐QS15 NB3 5,379,000 ‐ ‐ ‐ ‐ MBIARALI 2002‐QS15 M1 5,331,927 ‐ ‐ ‐ ‐ MBIARALI 2002‐QS15 M2 2,088,081 ‐ ‐ 249,118 325,163 MBIA

RALI 2002‐QS15 M3 1,579,864 ‐ ‐ 1,147,396 1,190,212 MBIARALI 2002‐QS15 B1 194,595 ‐ 466,547 177,194 176,130 MBIARALI 2002‐QS15 AP 21,007 ‐ ‐ ‐ ‐ MBIARALI 2002‐QS15 NB2 5,379,000 1,573 ‐ ‐ ‐ MBIARALI 2002‐QS15 AV ‐ ‐ ‐ ‐ ‐ MBIARALI 2002‐QS15 NB1 ‐ ‐ ‐ ‐ ‐ MBIARALI 2002‐QS15 B2 ‐ ‐ 1,368,036 ‐ ‐ MBIARALI 2002‐QS15 B3 ‐ ‐ 1,198,455 ‐ ‐ MBIA

RALI 2002‐QS15 Total 29,882,400 3,114,670 1,573 3,033,038 1,573,708$ 1,316,093 4,430,763$ 1.15% 53.53% 5.94% 1,691,506$ 1,434,093 4,548,763 1.21% 54.39% 5.59% MBIA 30.92%

RALI 2003‐QS1 M1 5,644,504 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A13 2,483,466 ‐ ‐ 15 20 MBIA

RALI 2003‐QS1 A14 2,483,466 ‐ ‐ ‐ ‐ MBIARALI 2003‐QS1 M2 2,270,525 ‐ ‐ 447,553 525,410 MBIA

RALI 2003‐QS1 A6 1,703,536 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 M3 1,698,988 ‐ 361,401 1,459,322 1,490,062 MBIA

RALI 2003‐QS1 A2 1,571,968 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A4 929,201 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A5 929,201 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A1 743,308 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A3 445,985 ‐ ‐ ‐ ‐ MBIARALI 2003‐QS1 A9 364,897 ‐ ‐ ‐ ‐ MBIA

Page 10 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RALI 2003‐QS1 A10 364,897 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 AP 201,609 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A8 19,075,000 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A7 ‐ ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A11 ‐ ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 A12 ‐ ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS1 AV ‐ ‐ ‐ ‐ ‐ MBIARALI 2003‐QS1 B1 ‐ ‐ 1,288,762 ‐ ‐ MBIA

RALI 2003‐QS1 B2 ‐ ‐ 1,355,631 ‐ ‐ MBIA

RALI 2003‐QS1 B3 ‐ ‐ 1,390,444 ‐ ‐ MBIA

RALI 2003‐QS1 Total 37,305,100 4,360,272 ‐ 4,396,237 1,906,890$ 1,293,469 5,653,741$ 1.07% 50.05% 7.93% 2,015,492$ 1,390,645 5,750,917 1.09% 51.20% 7.42% MBIA 25.89%

RALI 2003‐QS8 A6 12,814,656 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 A4 4,533,549 ‐ ‐ ‐ ‐ MBIARALI 2003‐QS8 A5 4,533,549 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 M1 3,564,665 ‐ ‐ 631,176 727,713 MBIA

RALI 2003‐QS8 A1 3,147,651 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 A2 1,573,825 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 M2 1,489,375 ‐ ‐ 1,359,943 1,380,381 MBIARALI 2003‐QS8 A7 1,476,889 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 M3 206,738 ‐ 1,247,050 181,426 180,595 MBIA

RALI 2003‐QS8 AP 198,753 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 A3 5,921,000 ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 AV ‐ ‐ ‐ ‐ ‐ MBIA

RALI 2003‐QS8 B1 ‐ ‐ 872,964 ‐ ‐ MBIA

RALI 2003‐QS8 B2 ‐ ‐ 1,020,774 ‐ ‐ MBIA

RALI 2003‐QS8 B3 ‐ ‐ 899,815 ‐ ‐ MBIA

RALI 2003‐QS8 Total 35,098,500 4,084,189 ‐ 4,040,603 2,172,545$ 1,537,855 5,622,044$ 1.04% 60.49% 6.63% 2,288,689$ 1,635,812 5,720,001 1.07% 61.44% 6.26% MBIA 15.06%RAMP 2001‐RS1 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS1 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2001‐RS1 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS1 AI4 5,838,303 ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS1 AII 3,092,229 ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS1 Total 9,547,120 23,733,322 ‐ ‐ ‐$ 1,296,015 25,029,337$ 4.53% 68.60% 6.60% ‐$ 1,342,778 25,076,100 4.47% 71.20% 6.37% Ambac 0%/15.44%

RAMP 2001‐RS1 AIIO ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS3 AI5 8,704,343 ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS3 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS3 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS3 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS3 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2001‐RS3 AII 4,352,387 ‐ ‐ ‐ ‐ AmbacRAMP 2001‐RS3 Total 14,795,300 29,441,794 ‐ ‐ ‐$ 2,359,843 31,801,637$ 3.83% 77.91% 3.60% ‐$ 2,490,105 31,931,899 3.90% 80.02% 3.38% Ambac 8.33%/14.55%

RAMP 2001‐RS3 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2002‐RP1 A1 5,056,112 ‐ ‐ ‐ ‐ Ambac

RFSC 2002‐RP1 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2002‐RP1 Total 6,084,370 17,156,209 ‐ ‐ ‐$ 547,502 17,703,711$ 3.08% 64.78% 8.99% ‐$ 584,988 17,741,197 3.18% 66.35% 8.69% Ambac 16.90%

RAMP 2002‐RS1 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS1 AI5 7,768,232 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS1 MII2 1,936,284 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS1 MI1 1,811,974 ‐ 343,940 1,344,901 1,425,430 Ambac

RAMP 2002‐RS1 MII3 421,059 ‐ 2,211,798 177,928 192,459 Ambac

RAMP 2002‐RS1 AIIO ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2002‐RS1 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS1 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS1 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS1 MI2 ‐ ‐ 1,813,310 ‐ ‐ Ambac

RAMP 2002‐RS1 MI3 ‐ ‐ 2,080,041 ‐ ‐ Ambac

RAMP 2002‐RS1 AII ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS1 MII1 ‐ ‐ ‐ ‐ ‐ Ambac

Page 11 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher



RAMP 2002‐RS2 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS2 AI5 9,269,436 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS2 MII2 1,953,827 ‐ ‐ ‐ ‐ AmbacRAMP 2002‐RS2 MI1 1,452,115 ‐ ‐ 349,768 411,823 Ambac

RAMP 2002‐RS2 MI2 740,115 ‐ 104,750 524,426 523,689 Ambac

RAMP 2002‐RS2 MII3 151,028 ‐ 1,573,316 ‐ ‐ Ambac


RAMP 2002‐RS2 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS2 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS2 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS2 MI3 ‐ ‐ 2,622,288 ‐ ‐ Ambac

RAMP 2002‐RS2 AII ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2002‐RS2 MII1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS2 Total 13,484,700 19,218,928 ‐ 4,300,354 874,194$ 1,504,700 20,723,628$ 2.31% 87.04% 6.22% 935,512$ 1,579,022 20,797,950 2.35% 88.33% 5.86% Ambac 17.15%

RAMP 2002‐RS4 AI6 11,223,398 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS4 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS4 AI2 ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2002‐RS4 AI5 1,587,973 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS4 AII 1,481,977 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS4 Total 16,541,300 23,580,908 ‐ ‐ ‐$ 1,414,540 24,995,448$ 1.91% 73.65% 6.16% ‐$ 1,544,386 25,125,294 2.03% 74.85% 5.82% Ambac 10.12%RAMP 2002‐RS4 AIIO ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2002‐RS5 AI6 9,251,316 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS5 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS5 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS5 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS5 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2002‐RS5 AI5 2,969,083 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS5 AII 5,739,940 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS5 Total 21,044,800 16,883,254 ‐ ‐ ‐$ 3,040,182 19,923,436$ 3.42% 82.57% 8.02% ‐$ 3,171,378 20,054,632 3.44% 83.86% 7.57% Ambac 14.38%

RAMP 2002‐RS6 AI7 13,262,395 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS6 AI6 4,359,844 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS6 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS6 AI2 ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2002‐RS6 AI5 5,818,160 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS6 AII 2,392,439 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RS6 Total 31,454,600 26,800,322 ‐ ‐ ‐$ 3,720,403 30,520,725$ 2.87% 63.07% 4.68% ‐$ 3,922,162 30,722,484 2.74% 67.76% 4.30% Ambac 9.86%


RAMP 2002‐RS7 A1 17,884,963 ‐ ‐ ‐ ‐ AmbacRAMP 2002‐RS7 Total 19,047,700 6,999,832 ‐ ‐ ‐$ 2,007,614 9,007,446$ 2.11% 76.50% 5.77% ‐$ 2,142,407 9,142,239 2.21% 77.32% 5.55% Ambac 5.37%

RAMP 2002‐RS7 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RZ4 A 12,522,420 ‐ ‐ ‐ ‐ Ambac

RAMP 2002‐RZ4 Total 21,756,500 18,108,402 ‐ ‐ ‐$ 1,924,548 20,032,950$ 2.03% 69.92% 7.42% ‐$ 2,113,995 20,222,397 2.10% 72.08% 6.83% Ambac 42.21%

RAMP 2002‐RZ4 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2003‐RP1 A1 ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2003‐RP1 M2 10,436,266 ‐ ‐ ‐ ‐ AmbacRFSC 2003‐RP1 M3 2,679,664 ‐ 6,976,336 528,910 1,579,751 Ambac

RFSC 2003‐RP1 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2003‐RP1 M1 ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2003‐RP1 Total 13,174,896 26,215,997 ‐ 6,976,336 528,910$ 1,546,380 27,762,377$ 2.00% 95.00% 6.00% 1,579,751$ 2,598,288 28,814,284 3.00% 100.00% 4.00% Ambac n/a

RFSC 2003‐RP2 A1 682,621 ‐ ‐ ‐ ‐ Ambac

RFSC 2003‐RP2 M2 9,209,556 ‐ ‐ ‐ ‐ Ambac

RFSC 2003‐RP2 M1 4,118,000 ‐ ‐ ‐ ‐ Ambac

Page 12 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFSC 2003‐RP2 Total 19,870,454 15,818,761 ‐ ‐ ‐$ 1,093,210 16,911,971$ 1.00% 105.00% 10.00% ‐$ 2,416,143 18,234,904 2.00% 110.00% 8.00% Ambac 96.56%

RFSC 2003‐RP2 M3 ‐ ‐ ‐ ‐ ‐ Ambac

RFSC 2003‐RP2 M4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS1 AI6 9,911,838 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS1 AII 7,373,117 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS1 AI5 6,583,451 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS1 MI1 1,679,505 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS1 Total 27,442,800 32,232,516 ‐ ‐ ‐$ 2,033,237 34,265,753$ 1.60% 62.42% 3.56% ‐$ 2,190,028 34,422,544 1.65% 64.62% 3.41% Ambac 7.58%

RAMP 2003‐RS1 MI2 940,337 ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS1 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS1 AI2 ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS1 MI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS2 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS2 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS2 AI3 ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS2 AI6 15,548,402 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS2 AII 11,303,649 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS2 Total 54,562,500 58,608,528 ‐ ‐ ‐$ 7,027,819 65,636,347$ 2.73% 68.43% 4.28% ‐$ 7,659,151 66,267,679 2.89% 69.66% 3.97% Ambac 4.66%/20.27%

RAMP 2003‐RS3 AI5 4,823,972 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS3 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS3 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS3 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS3 AI4 16,682,154 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS3 AII 9,388,971 ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS4 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS4 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS4 AI3 ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS4 AI6 6,462,850 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS4 AIIA 5,293,965 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS4 AIIB 5,569,046 ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS5 AI6 5,890,993 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS5 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS5 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS5 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS5 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS5 AI5 43,448,371 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS5 AIIA 7,191,734 ‐ ‐ ‐ ‐ Ambac



RAMP 2003‐RS6 AI6 8,358,902 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS6 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS6 AI2 ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS6 AI5 50,013,814 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS6 AIIA 3,878,493 ‐ ‐ ‐ ‐ Ambac



RAMP 2003‐RS7 AI5 57,822,916 ‐ ‐ ‐ ‐ Ambac

Page 13 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RAMP 2003‐RS7 MII1 11,342,104 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS7 AI6 8,477,589 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS7 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS7 MI2 6,146,451 ‐ ‐ 268,395 535,163 Ambac

RAMP 2003‐RS7 MI1 5,209,821 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS7 MII3 4,160,477 ‐ 2,440,821 3,833,041 3,980,933 Ambac

RAMP 2003‐RS7 MII2 3,960,718 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS7 MI3 1,075,774 ‐ 6,948,027 1,075,774 1,075,774 Ambac

RAMP 2003‐RS7 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS7 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS7 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS7 AIIA ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS7 AIIB ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS7 Total 98,904,400 59,388,992 ‐ 9,388,848 5,177,210$ 5,240,000 64,628,992$ 1.05% 78.87% 7.62% 5,591,871$ 5,691,300 65,080,292 1.10% 79.21% 7.06% Ambac Class PIF

RAMP 2003‐RS8 AI6A 35,148,315 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 AI8 13,722,032 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS8 MII1 12,932,405 ‐ ‐ 1,233,088 1,372,132 Ambac

RAMP 2003‐RS8 AI7 7,784,254 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 MI1 6,814,278 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 MI2 6,736,116 ‐ ‐ 1,569,421 2,238,459 Ambac

RAMP 2003‐RS8 MII2 5,504,525 ‐ ‐ 3,740,573 3,753,494 AmbacRAMP 2003‐RS8 AI6B 18,679,353 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 MI3 2,952,737 ‐ ‐ 2,937,312 2,928,002 Ambac

RAMP 2003‐RS8 MII4 1,866,440 ‐ ‐ 1,866,440 1,866,440 Ambac

RAMP 2003‐RS8 MII3 1,758,916 ‐ 4,343,746 1,758,916 1,758,916 Ambac

RAMP 2003‐RS8 MII5 186,290 ‐ 1,702,363 186,290 186,290 AmbacRAMP 2003‐RS8 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 AI5 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS8 AIIA ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS8 AIIB ‐ ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RS9 AI6A 23,714,528 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 MII1 15,476,850 ‐ ‐ 1,105,987 1,378,372 Ambac

RAMP 2003‐RS9 MI1 7,811,995 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS9 AI7 7,757,021 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 MII2 5,766,544 ‐ ‐ 4,701,196 4,816,438 Ambac

RAMP 2003‐RS9 MI2 5,697,978 ‐ ‐ 3,762,947 4,668,304 Ambac

RAMP 2003‐RS9 AI6B 15,863,325 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 MII3 2,297,660 ‐ ‐ 2,297,660 2,297,660 Ambac

RAMP 2003‐RS9 MII4 2,232,834 ‐ 168,612 2,232,834 2,232,834 AmbacRAMP 2003‐RS9 MI3 1,560,818 ‐ 4,331,902 1,465,097 1,452,000 Ambac

RAMP 2003‐RS9 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 AI5 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 AIIA ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RS9 AIIB ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RS9 MII5 ‐ ‐ 2,582,543 ‐ ‐ Ambac



RAMP 2003‐RZ1 AI7 5,245,891 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ1 AI6 6,121,499 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ1 AI1A ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ1 AI1B ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ1 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

Page 14 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RAMP 2003‐RZ1 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ1 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ1 AI5 11,523,771 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ1 AII 12,803,927 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ1 Total 42,504,000 27,276,392 ‐ ‐ ‐$ 3,419,205 30,695,597$ 2.04% 66.73% 8.40% ‐$ 3,712,718 30,989,110 2.11% 68.00% 7.73% Ambac 15.71%

RAMP 2003‐RZ2 M1 2,525,334 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ2 A1 12,679,088 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ2 M2 2,044,927 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ2 Total 18,651,600 10,509,932 ‐ ‐ ‐$ 1,153,649 11,663,581$ 1.85% 65.67% 11.97% ‐$ 1,287,489 11,797,421 1.97% 66.53% 11.18% Ambac 31.80%RAMP 2003‐RZ2 M3 81,880 ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RZ3 A6 14,261,959 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ3 A5A 9,831,106 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ3 M1 4,431,961 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ3 M2 3,472,635 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ3 A5B 10,923,451 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ3 Total 45,513,000 20,395,800 ‐ ‐ ‐$ 2,763,577 23,159,377$ 1.49% 66.18% 8.83% ‐$ 3,094,942 23,490,742 1.54% 68.52% 7.98% Ambac 22.82%

RAMP 2003‐RZ3 M3 150,218 ‐ ‐ ‐ ‐ Ambac


RAMP 2003‐RZ3 A1 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ3 A2 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ3 A3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ3 A4 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ4 M1 14,842,099 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ4 A7 12,517,792 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ4 M2 11,943,189 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ4 M3 3,671,656 ‐ ‐ 2,886,755 3,556,944 Ambac

RAMP 2003‐RZ4 A6 64,197,530 ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ4 A1 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ4 A2 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ4 A3 ‐ ‐ ‐ ‐ ‐ Ambac

RAMP 2003‐RZ4 A4 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ4 A5 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ4 Total 107,536,000 35,008,260 ‐ ‐ 2,886,755$ 6,573,823 41,582,083$ 1.57% 62.96% 8.75% 3,556,944$ 7,422,318 42,430,578 1.67% 64.38% 7.92% Ambac 28.42%RAMP 2003‐RZ5 A6A 21,808,429 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 A7 15,265,480 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 M2 10,861,779 ‐ ‐ 2,405,887 2,967,396 AmbacRAMP 2003‐RZ5 M1 8,089,380 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 A6B 28,577,794 ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 M3 548,504 ‐ 6,076,302 507,333 501,324 AmbacRAMP 2003‐RZ5 A1 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 A2 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 A3 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 A4 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 A5 ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 AV ‐ ‐ ‐ ‐ ‐ AmbacRAMP 2003‐RZ5 Total 85,439,000 34,950,384 ‐ 6,076,302 2,913,221$ 6,755,985 41,706,369$ 2.04% 63.68% 8.88% 3,468,720$ 7,477,893 42,428,277 2.12% 65.21% 8.06% Ambac 22.90%RASC 2001‐KS1 AI6 10,275,061 ‐ ‐ 15,631 30,574 FGICRASC 2001‐KS1 AI1 ‐ ‐ ‐ ‐ ‐ FGICRASC 2001‐KS1 AI2 ‐ ‐ ‐ ‐ ‐ FGICRASC 2001‐KS1 AI3 ‐ ‐ ‐ ‐ ‐ FGICRASC 2001‐KS1 AI4 ‐ ‐ ‐ ‐ ‐ FGICRASC 2001‐KS1 AI5 18,970,444 ‐ ‐ 2,734,814 3,360,343 FGICRASC 2001‐KS1 AII 622,768 ‐ ‐ ‐ ‐ FGICRASC 2001‐KS1 Total 43,762,800 128,338,992 ‐ ‐ 2,750,444$ 9,843,256 138,182,248$ 4.27% 86.28% 2.12% 3,390,917$ 10,474,788 138,813,780 4.45% 87.97% 1.93% FGIC 0.00%RASC 2002‐KS1 AI6 17,285,122 76,169 420,668 690,462 824,468 AmbacRASC 2002‐KS1 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS1 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS1 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS1 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS1 AI5 42,594,940 187,700 924,349 6,346,802 7,207,389 Ambac

RASC 2002‐KS1 AIIA 10,141,322 44,689 106,687 852,858 886,261 AmbacRASC 2002‐KS1 AIIB 8,727,616 38,459 97,148 739,467 768,134 AmbacRASC 2002‐KS1 Total 77,861,800 148,825,456 347,017 1,548,852 8,629,589$ 14,335,728 163,161,184$ 3.60% 76.22% 1.94% 9,686,251$ 15,358,738 164,184,194 3.77% 78.04% 1.80% Ambac 0.00%

Page 15 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RASC 2002‐KS4 AI6 9,114,477 915,048 1,470,443 5,720 13,634 AmbacRASC 2002‐KS4 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS4 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS4 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS4 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS4 AI5 28,481,554 2,859,404 2,712,287 291,508 567,207 AmbacRASC 2002‐KS4 AIIA 14,536,652 1,459,407 718,465 32,956 34,688 AmbacRASC 2002‐KS4 AIIB 12,390,712 1,243,965 589,794 28,451 29,850 AmbacRASC 2002‐KS4 Total 64,979,500 115,200,320 6,477,823 5,490,990 358,635$ 10,931,170 126,131,490$ 3.14% 85.61% 4.42% 645,379$ 11,626,724 126,827,044 3.23% 87.23% 4.04% Ambac 0.76%RASC 2002‐KS4 AIIO ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS6 AI6 8,527,643 425,085 722,111 ‐ ‐ AmbacRASC 2002‐KS6 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS6 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS6 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2002‐KS6 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS6 AI5 25,920,883 1,292,100 1,207,831 ‐ ‐ Ambac

RASC 2002‐KS6 AII 15,484,958 771,892 1,108,580 1,208,063 1,208,063 Ambac

RASC 2002‐KS6 Total 50,236,600 73,390,432 2,489,076 3,038,522 1,208,063$ 5,674,229 79,064,661$ 2.23% 73.50% 3.44% 1,208,063$ 6,064,703 79,455,135 2.27% 76.25% 3.23% Ambac 3.39%

RASC 2002‐KS6 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS8 A6 15,637,678 712,310 851,529 1,031,150 1,123,548 Ambac

RASC 2002‐KS8 A1 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS8 A2 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS8 A3 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS8 A4 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2002‐KS8 A5 50,099,005 2,282,053 4,412,030 9,364,073 10,016,676 Ambac

RASC 2002‐KS8 Total 66,313,300 46,179,160 2,994,362 5,263,559 10,395,223$ 11,640,736 57,819,896$ 3.07% 84.18% 3.96% 11,140,224$ 12,411,615 58,590,775 3.18% 85.73% 3.66% Ambac 0.00%

RASC 2002‐KS8 AIO ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 AI5 32,057,300 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 AI6 16,627,780 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 MI1 8,294,550 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 MI2 5,924,678 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 MI3 4,882,106 ‐ ‐ 2,295,624 3,556,092 Ambac

RASC 2003‐KS4 AIIA 4,043,332 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 AIIB 5,214,003 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 AIII 3,121,101 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 AIIO ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS4 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2003‐KS4 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2003‐KS4 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRASC 2003‐KS4 Total 83,145,900 52,333,468 ‐ ‐ 2,295,624$ 10,366,115 62,699,583$ 2.32% 74.77% 4.26% 3,556,092$ 11,135,049 63,468,517 2.42% 76.09% 3.99% Ambac 29.09%/8.14%

RASC 2003‐KS5 AI6 4,408,601 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 AI5 24,217,361 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 AIIA 6,791,278 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 AIIB 4,534,845 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS5 Total 44,589,000 29,806,584 ‐ ‐ ‐$ 6,289,252 36,095,836$ 2.50% 78.52% 4.11% ‐$ 6,694,593 36,501,177 2.55% 80.60% 3.82% Ambac 8.11%/13.38%

RASC 2003‐KS5 AIIO ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS9 AI6 10,052,842 ‐ ‐ ‐ 2,532 Ambac

RASC 2003‐KS9 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS9 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS9 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS9 AI4 4,418,891 ‐ ‐ ‐ ‐ Ambac

RASC 2003‐KS9 AI5 49,853,000 ‐ ‐ ‐ 807,069 Ambac

RASC 2003‐KS9 A2A 10,824,007 ‐ ‐ 690,810 746,207 Ambac

RASC 2003‐KS9 A2B 11,716,122 ‐ ‐ 487,626 551,683 Ambac

RASC 2003‐KS9 Total 90,067,900 48,063,936 ‐ ‐ 1,178,436$ 9,305,239 57,369,175$ 1.98% 76.02% 5.82% 2,107,492$ 10,006,237 58,070,173 2.05% 77.34% 5.37% Ambac 4.82%/0.00%

Page 16 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFMSI 2003‐S13 AV 29,747,847 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 A2 18,260,000 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 A3 9,027,916 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 M1 1,517,369 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 M2 465,435 ‐ ‐ 310,639 332,401 MBIA

RFMSI 2003‐S13 M3 354,978 ‐ ‐ 344,505 344,915 MBIA

RFMSI 2003‐S13 AP 122,049 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 B1 100 ‐ 312,335 ‐ ‐ MBIA

RFMSI 2003‐S13 A1 18,260,000 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 A4 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 A5 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 A6 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 A7 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S13 B2 ‐ ‐ 284,853 ‐ ‐ MBIA

RFMSI 2003‐S13 B3 ‐ ‐ 513,538 ‐ ‐ MBIA

RFMSI 2003‐S13 Total 29,843,200 1,152,820 ‐ 1,110,725 655,144$ 92,155 1,244,975$ 0.34% 15.00% 9.82% 677,316$ 94,758 1,247,578 0.33% 15.36% 9.29% MBIA 7.86%

RFMSI 2003‐S4 AV 20,050,469 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A4 11,146,417 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 M1 1,329,421 ‐ ‐ 7,809 9,211 MBIA

RFMSI 2003‐S4 M2 460,832 ‐ ‐ 134,981 142,250 MBIA

RFMSI 2003‐S4 M3 362,732 ‐ ‐ 307,131 312,218 MBIA

RFMSI 2003‐S4 AP 155,861 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 B1 67,596 ‐ 49,766 60,368 60,396 MBIA

RFMSI 2003‐S4 A3 6,527,000 ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A1 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A2 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A5 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A6 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A7 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A8 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A9 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A10 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A11 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A12 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 A13 ‐ ‐ ‐ ‐ ‐ MBIA

RFMSI 2003‐S4 B2 ‐ ‐ 222,214 ‐ ‐ MBIA

RFMSI 2003‐S4 B3 ‐ ‐ 353,689 ‐ ‐ MBIA

RFMSI 2003‐S4 Total 20,132,600 639,159 ‐ 625,669 510,289$ 372,188 1,011,347$ 0.69% 48.98% 11.21% 524,074$ 379,757 1,018,916 0.68% 49.46% 10.64% MBIA 11.08%

RFMS2 1999‐HI1 A1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI1 A2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI1 A3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI1 A4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI1 A5 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI1 A6 1,203,928 ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI1 Total 3,402,371 40,895,988 ‐ ‐ ‐$ 420,343 41,316,331$ 3.71% 111.80% 11.24% ‐$ 445,284 41,341,272 3.86% 111.86% 10.47% Ambac 64.62%

RFMS2 1999‐HI4 A1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI4 A2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI4 A3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI4 A4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI4 A5 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI4 A6 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI4 A7 661,684 ‐ ‐ ‐ ‐ Ambac


RFMS2 1999‐HI6 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI6 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI6 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI6 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

Page 17 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFMS2 1999‐HI6 AI5 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI6 AI6 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI6 AI7 2,236,286 ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI6 AI8 887,847 ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI6 AII ‐ ‐ ‐ ‐ ‐ Ambac


RFMS2 1999‐HI8 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI8 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI8 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI8 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI8 AI5 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI8 AI6 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI8 AI7 1,762,877 ‐ ‐ ‐ ‐ Ambac

RFMS2 1999‐HI8 AI8 287,862 ‐ ‐ ‐ ‐ Ambac



RFMS2 2000‐HI1 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI1 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI1 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI1 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI1 AI5 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI1 AI6 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI1 AI7 2,046,593 ‐ ‐ ‐ ‐ Ambac



RFMS2 2000‐HI2 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI2 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI2 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI2 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI2 AI5 2,238,412 ‐ ‐ ‐ ‐ Ambac



RFMS2 2000‐HI3 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI3 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI3 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI3 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI3 AI5 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI3 AI6 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI3 AI7 3,341,017 ‐ ‐ ‐ ‐ Ambac



RFMS2 2000‐HI4 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI4 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI4 AI3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI4 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI4 AI5 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI4 AI6 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI4 AI7 4,683,582 ‐ ‐ ‐ ‐ Ambac



RFMS2 2000‐HI5 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI5 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HI5 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HI5 AI4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2000‐HI5 AI5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HI5 AI6 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HI5 AI7 6,106,260 ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HI5 AII ‐ ‐ ‐ ‐ ‐ Ambac

Page 18 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFMS2 2000‐HI5 Total 8,981,260 57,645,180 ‐ ‐ ‐$ 843,807 58,488,987$ 2.60% 111.59% 8.68% ‐$ 908,289 58,553,469 2.75% 111.62% 7.96% Ambac 32.01%RFMS2 2001‐HI1 A 3,103,105 ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI1 Total 4,333,590 25,786,334 ‐ ‐ ‐$ 337,795 26,124,129$ 2.09% 111.12% 10.01% ‐$ 368,287 26,154,621 2.24% 111.15% 9.35% Ambac 28.39%RFMS2 2001‐HI2 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 AI5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 AI6 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 AI7 1,916,227 ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 AII ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI2 Total 2,984,370 20,035,914 ‐ ‐ ‐$ 164,827 20,200,741$ 1.66% 111.15% 11.98% ‐$ 179,818 20,215,732 1.76% 111.23% 11.06% Ambac 35.79%RFMS2 2001‐HI2 AIO ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI3 AI1 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2001‐HI3 AI2 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2001‐HI3 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI3 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI3 AI5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI3 AI6 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI3 AI7 7,136,117 ‐ ‐ ‐ ‐ Ambac

RFMS2 2001‐HI3 AII ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI3 Total 9,313,230 42,530,960 ‐ ‐ ‐$ 898,174 43,429,134$ 2.52% 110.99% 9.81% ‐$ 969,536 43,500,496 2.66% 110.98% 8.98% Ambac 23.38%RFMS2 2001‐HI3 AIO ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI4 A1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI4 A2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI4 A3 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2001‐HI4 A4 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI4 A5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI4 A6 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2001‐HI4 A7 7,081,215 ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HI4 Total 9,206,220 41,715,440 ‐ ‐ ‐$ 883,576 42,599,016$ 2.74% 111.09% 11.60% ‐$ 945,401 42,660,841 2.85% 111.10% 10.68% Ambac 23.08%RFMS2 2001‐HI4 AIO ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HS2 A1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HS2 A2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HS2 A3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HS2 A4 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HS2 A5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2001‐HS2 Total 1,398,380 4,097,670 ‐ ‐ ‐$ 6,045 4,103,715$ 0.23% 112.76% 8.48% ‐$ 6,671 4,104,340 0.25% 112.53% 7.89% Ambac Classes PIFRFMS2 2001‐HS2 AIO ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 A1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 A2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 A3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 A4 ‐ ‐ ‐ ‐ ‐ Ambac

RFMS2 2002‐HI1 A5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 A6 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 A7 6,980,699 ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 AIO ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI1 Total 9,230,700 37,815,172 ‐ ‐ ‐$ 819,438 38,634,610$ 2.95% 112.02% 11.34% ‐$ 866,646 38,681,818 3.06% 112.06% 10.51% Ambac 24.38%

RFMS2 2002‐HI2 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AI5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AI6 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AI7 3,795,681 ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AII 1,788,996 ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 AIO ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI2 Total 7,209,690 27,284,040 ‐ ‐ ‐$ 569,200 27,853,240$ 2.63% 111.31% 9.99% ‐$ 590,962 27,875,002 2.65% 111.32% 8.90% Ambac 22.54%RFMS2 2002‐HI3 A1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 A2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 A3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 A4 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 A5 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 A6 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 A7 8,375,182 ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 AIO ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2002‐HI3 Total 10,250,200 32,152,084 ‐ ‐ ‐$ 1,045,575 33,197,659$ 3.27% 112.36% 7.90% ‐$ 1,112,837 33,264,921 3.42% 112.32% 7.16% Ambac 18.29%RFMS2 2002‐HS3 Total 3,594,250 3,914,926 ‐ ‐ ‐$ 231,035 4,145,961$ 2.36% 111.88% 11.64% ‐$ 243,082 4,158,009 2.44% 111.86% 10.61% FGIC 71.46%RFMS2 2002‐HS3 1A1 ‐ ‐ ‐ ‐ ‐ FGIC

Page 19 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RFMS2 2002‐HS3 1A2 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2002‐HS3 1A3 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2002‐HS3 1A4 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2002‐HS3 1A5 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2002‐HS3 1A6 449,294 ‐ ‐ ‐ ‐ FGICRFMS2 2002‐HS3 2A 634,765 ‐ ‐ ‐ ‐ FGICRFMS2 2002‐HS3 AIO ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HI3 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HI3 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HI3 AI3 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HI3 AI4 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HI3 AI5 6,767,415 ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HI3 AII 6,672,672 ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HI3 Total 16,390,400 26,662,514 ‐ ‐ ‐$ 2,015,106 28,677,620$ 3.95% 110.79% 10.99% ‐$ 2,128,118 28,790,632 4.08% 110.80% 10.10% Ambac 16.47%/19.50%RFMS2 2003‐HS1 Total 10,071,400 7,310,844 ‐ ‐ ‐$ 456,802 7,767,645$ 1.75% 107.73% 14.79% ‐$ 463,788 7,774,631 1.75% 107.97% 14.18% FGIC 28.94%RFMS2 2003‐HS1 AIIO ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS1 AI6 824,095 ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS1 AI1 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS1 AI2 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS1 AI3 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS1 AI4 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS1 AI5 3,934,641 ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS1 AII 1,633,532 ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 AI4 7,474,135 ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 Total 17,222,500 10,341,978 ‐ ‐ ‐$ 388,053 10,730,031$ 0.79% 108.87% 14.58% ‐$ 416,473 10,758,451 0.84% 108.84% 13.82% FGIC 23.50%RFMS2 2003‐HS2 AIIA 1,620,875 ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 AIIB 2,467,479 ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 MI1 345,976 ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 AIIO ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 AIA ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 AIB ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 AI2 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 AI3 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 MI2 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS2 MI3 ‐ ‐ ‐ ‐ ‐ FGICRFMS2 2003‐HS3 Total 23,314,700 10,098,988 ‐ ‐ ‐$ 1,339,457 11,438,445$ 1.98% 108.87% 16.27% ‐$ 1,460,304 11,559,292 2.11% 108.87% 15.18% MBIA 0.00%RFMS2 2003‐HS3 AI1 ‐ ‐ ‐ ‐ ‐ MBIARFMS2 2003‐HS3 AI2 ‐ ‐ ‐ ‐ ‐ MBIARFMS2 2003‐HS3 AI3 ‐ ‐ ‐ ‐ ‐ MBIARFMS2 2003‐HS3 AI4 13,651,768 ‐ ‐ ‐ ‐ MBIARFMS2 2003‐HS3 A2A 2,474,433 ‐ ‐ ‐ ‐ MBIARFMS2 2003‐HS3 A2B 2,053,044 ‐ ‐ ‐ ‐ MBIARFMS2 2003‐HS3 AIIO ‐ ‐ ‐ ‐ ‐ MBIARFMS2 2003‐HS4 AIA 3,436,178 ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HS4 AIB 2,378,740 ‐ ‐ ‐ ‐ AmbacRFMS2 2003‐HS4 Total 8,567,730 5,533,338 ‐ ‐ ‐$ 424,538 5,957,876$ 1.20% 102.72% 8.96% ‐$ 462,365 5,995,703 1.29% 102.57% 8.49% Ambac 14.59%RASC 1999‐RS1 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRASC 1999‐RS1 AI2 ‐ ‐ ‐ ‐ ‐ AmbacRASC 1999‐RS1 AI3 1,792,364 ‐ ‐ ‐ ‐ AmbacRASC 1999‐RS1 AII 1,120,192 ‐ ‐ ‐ ‐ AmbacRASC 1999‐RS1 Total 3,652,234 3,793,159 ‐ ‐ ‐$ 10,992 3,804,151$ 0.23% 33.13% 10.11% ‐$ 13,777 3,666,011 0.27% 34.20% 9.13% Ambac 17.49%RFMS2 2000‐HL1 AI1 ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HL1 AI2 361,718 ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HL1 AII ‐ ‐ ‐ ‐ ‐ AmbacRFMS2 2000‐HL1 Total 1,031,029 8,105,735 ‐ ‐ ‐$ 21,183 8,126,918$ 1.22% 45.31% 7.69% ‐$ 25,197 8,130,932 1.41% 46.32% 7.35% Ambac 64.92%GMACM 2004‐JI IO 37,035,243.92 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A3 16,774,000.00 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A20 16,660,094.88 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI M1 2,361,777.14 ‐ ‐ 77,530$ 121,027$ MBIAGMACM 2004‐JI M2 953,508.19 ‐ ‐ 587,998$ 624,219$ MBIAGMACM 2004‐JI A21 876,916.16 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI M3 633,882.45 ‐ ‐ 633,882$ 633,882$ MBIAGMACM 2004‐JI B1 179,507.84 ‐ 229,904 165,534$ 164,818$ MBIAGMACM 2004‐JI PO 163,997.89 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A4 ‐$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A5 ‐$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A6 ‐$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A7 ‐$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A8 781,707$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A9 ‐$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A10 ‐$ ‐$ ‐$ ‐$ ‐$ MBIA

Page 20 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


GMACM 2004‐JI A11 ‐$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A12 ‐$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A13 1,693,031$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐JI A1 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A2 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A14 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A15 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A16 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A17 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A18 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI Total 41,078,423$ 1,079,779$ ‐$ 1,113,980$ 1,464,944$ 1,774,653 2,854,432$ 0.43% 57.02% 6.40% 1,543,947$ 1,852,225 2,932,004 0.47% 56.92% 6.29% MBIA 10.05%GMACM 2004‐JI A19 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI A22 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐JI B2 0 ‐ 404,083 ‐$ ‐$ MBIAGMACM 2004‐JI B3 0 ‐ 479,992 ‐$ ‐$ MBIAGMACM 2004‐J2 WAC_IO 30,679,090.47 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A10 15,303,291.14 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A7 7,203,716.72 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A4 3,027,366.41 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 M1 3,014,988.16 ‐ ‐ 93,243$ 159,730$ MBIAGMACM 2004‐J2 A9 1,445,278.95 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 M2 1,283,429.16 ‐ ‐ 1,074,921$ 1,116,901$ MBIAGMACM 2004‐J2 A11 830,234.16 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 PO 715,390.36 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 M3 690,504.76 ‐ 60,560 667,972$ 666,957$ MBIAGMACM 2004‐J2 A12 471,157.88 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A8 17,566,000$ ‐$ ‐$ ‐$ ‐$ MBIAGMACM 2004‐J2 Total 51,551,358$ 1,776,708$ ‐$ 1,683,480$ 1,836,135$ 2,236,917 4,013,625$ 0.49% 47.01% 7.70% 1,943,588$ 2,349,927 4,126,635 0.53% 48.59% 7.54% MBIA 9.68%GMACM 2004‐J2 A1 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A2 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A3 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A5 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 A6 0 ‐ ‐ ‐$ ‐$ MBIAGMACM 2004‐J2 B1 0 ‐ 551,762 ‐$ ‐$ MBIAGMACM 2004‐J2 B2 0 ‐ 513,428 ‐$ ‐$ MBIAGMACM 2004‐J2 B3 0 ‐ 557,730 ‐$ ‐$ MBIAGSR 2007‐HEL1 Total 26,894,539$ 56,359,724$ 65,349,240$ ‐$ 4,469,033$ 5,134,028 61,493,752$ 3.42% 105.81% 11.11% 4,992,354$ 5,772,318 62,132,042 3.65% 105.84% 9.82% MBIA 0.00%GSR 2007‐HEL1 A 26,089,153$ 63,388,763$ ‐$ 4,335,203$ 4,842,853$ MBIAGSR 2007‐HEL1 S 805,386$ 1,960,477$ ‐$ 133,830$ 149,501$ MBIARAST 2003‐A10 AX 47,538,659.35 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A10 A4 18,887,098.24 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A10 A1 14,384,320.49 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A10 A3 9,150,000.00 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A10 B1 6,155,771.76 ‐ ‐ 1,914,790$ 2,138,816$ MBIARAST 2003‐A10 A5 3,665,430.26 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A10 B2 1,455,136.94 ‐ ‐ 1,356,057$ 1,370,208$ MBIARAST 2003‐A10 B3 1,145,750.05 ‐ ‐ 1,111,669$ 1,109,363$ MBIARAST 2003‐A10 PO 470,749.07 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A10 B4 7,900.37 ‐ 633,769 ‐$ ‐$ MBIARAST 2003‐A10 A2 13,725,000$ ‐$ ‐$ ‐$ ‐$ MBIARAST 2003‐A10 Total 69,047,157$ 2,041,563$ ‐$ 2,453,310$ 4,382,516$ 2,982,547 5,024,110$ 1.08% 51.80% 5.29% 4,618,388$ 3,198,258 5,239,821 1.16% 51.67% 5.14% MBIA 12.69%RAST 2003‐A10 B5 0 ‐ 362,650 ‐$ ‐$ MBIARAST 2003‐A10 B6 0 ‐ 1,456,891 ‐$ ‐$ MBIARAST 2003‐A7 A3 20,754,852.88 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 A4 20,754,852.88 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 A9 15,081,000.00 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 B1 10,745,024.60 ‐ ‐ 131,687$ 209,910$ MBIARAST 2003‐A7 A12 3,538,858.26 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 B2 2,710,243.41 ‐ ‐ 1,441,889$ 1,594,424$ MBIARAST 2003‐A7 B3 2,237,768.32 ‐ ‐ 2,235,149$ 2,236,162$ MBIARAST 2003‐A7 A5 2,000,000.00 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 A7 1,984,866.76 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 B4 938,532.19 ‐ ‐ 938,532$ 938,532$ MBIARAST 2003‐A7 A10 937,820.85 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 B6 623,315.07 ‐ 1,580,565 623,315$ 623,315$ MBIARAST 2003‐A7 B5 377,621.67 ‐ ‐ 377,622$ 377,622$ MBIARAST 2003‐A7 A11 255,769.32 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 PO 41,208.48 ‐ ‐ 1,746$ 1,774$ MBIARAST 2003‐A7 A8 15,081,000$ 18,182$ ‐$ ‐$ ‐$ MBIARAST 2003‐A7 Total 62,345,039$ 1,175,993$ 18,182$ 1,580,565$ 5,749,940$ 3,873,567 5,049,560$ 1.43% 58.25% 5.81% 5,981,739$ 4,078,193 5,254,186 1.50% 58.29% 5.66% MBIA 28.47%RAST 2003‐A7 A1 0 ‐ ‐ ‐$ ‐$ MBIA

Page 21 of 22


Trust Level Model





Date

Monoline Payment



Expected Loss Going

Forward

Lower Range ‐

Collateral Expected

Loss Going Forward

Lower Range ‐

Collateral Expected

Cum Loss Total

Lower

Range ‐

CDR

Lower Ranger ‐

Severity

Lower Range ‐

CRR


Expected Loss Going

Forward

Higher Range ‐

Collateral Expected

Loss Going Forward

Higher Range ‐

Collateral Expected

Cum Loss Total

Higher

Range ‐

CDR

Higher Range ‐

Sev

Higher


RAST 2003‐A7 A2 0 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 A6 0 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 A13 0 ‐ ‐ ‐$ ‐$ MBIARAST 2003‐A7 A14 0 ‐ ‐ ‐$ ‐$ MBIARAST 2006‐A1 AX 234,874,680.81 ‐ ‐ ‐$ ‐$ MBIARAST 2006‐A1 1A3 69,396,696.16 ‐ 3,811,327 13,307,868$ 14,319,282$ MBIARAST 2006‐A1 1A6 67,920,441.78 ‐ 3,730,250 14,515,374$ 15,512,720$ MBIARAST 2006‐A1 1A7 67,920,441.78 ‐ ‐ ‐$ ‐$ MBIARAST 2006‐A1 1A1 43,891,814.40 ‐ ‐ 8,574,945$ 9,234,153$ MBIARAST 2006‐A1 3A2 32,665,486.00 ‐ 266,794 12,404,079$ 12,980,809$ MBIARAST 2006‐A1 1A8 28,684,982.25 ‐ 1,575,404 5,500,780$ 5,918,846$ MBIARAST 2006‐A1 3A1 28,580,504.08 ‐ 233,430 10,852,887$ 11,357,494$ MBIARAST 2006‐A1 2A1 24,132,474.25 ‐ 502,653 4,810,076$ 5,149,205$ MBIARAST 2006‐A1 3A3 8,535,538.07 ‐ 69,714 3,241,203$ 3,391,904$ MBIARAST 2006‐A1 3A4 8,489,121.21 ‐ 971,663 3,223,578$ 3,373,458$ MBIARAST 2006‐A1 1A4 3,917,545.31 ‐ 215,155 837,224$ 894,750$ MBIARAST 2006‐A1 1A5 3,917,545.31 ‐ ‐ ‐$ ‐$ MBIARAST 2006‐A1 PO 1,609,522.26 ‐ 383,419 179,750$ 200,139$ MBIARAST 2006‐A1 1A2 1,032,608.90 ‐ 2,428,762 39,994$ 35,532$ MBIARAST 2006‐A1 WRAP_1A8 312,371,618$ 5,754,498$ ‐$ 3,487,603$ 3,752,781$ MBIARAST 2006‐A1 2A2 0 ‐ 1,508,529 ‐$ ‐$ MBIARAST 2006‐A1 3A5 0 ‐ 7,418,722 ‐$ ‐$ MBIARAST 2006‐A1 IB1 0 ‐ 12,856,714 ‐$ ‐$ MBIARAST 2006‐A1 IB2 0 ‐ 13,702,221 ‐$ ‐$ MBIARAST 2006‐A1 IB3 0 ‐ 3,359,416 ‐$ ‐$ MBIARAST 2006‐A1 IIB1 0 ‐ 5,081,912 ‐$ ‐$ MBIARAST 2006‐A1 IIB2 0 ‐ 2,130,089 ‐$ ‐$ MBIARAST 2006‐A1 IIB3 0 ‐ 1,423,707 ‐$ ‐$ MBIARAST 2006‐A1 P 0 ‐ ‐ ‐$ ‐$ MBIARAST 2006‐A1 IB4 0 ‐ 3,082,041 ‐$ ‐$ MBIARAST 2006‐A1 IB5 0 ‐ 2,526,398 ‐$ ‐$ MBIARAST 2006‐A1 IB6 0 ‐ 1,691,211 ‐$ ‐$ MBIARAST 2006‐A1 IIB4 0 ‐ 1,071,324 ‐$ ‐$ MBIARAST 2006‐A1 IIB5 0 ‐ 834,296 ‐$ ‐$ MBIARAST 2006‐A1 IIB6 0 ‐ 716,936 ‐$ ‐$ MBIARAST 2006‐A1 Total 312,371,618$ 70,130,992$ 5,754,498$ 71,592,088$ 80,975,360$ 35,730,838 105,861,830$ 2.61% 56.84% 5.89% 86,121,071$ 39,731,986 109,862,978 2.77% 58.30% 5.39% MBIA 6.65%/4.75%RFMSI 2004‐SR1 Total 14,503,630$ 5,046,456$ ‐$ ‐$ ‐$ ‐ 5,046,456$ 0.13% 27.11% 24.31% ‐$ ‐ 5,046,456 0.19% 28.31% 22.69% MBIA 3.34%RFMSI 2004‐SR1 A5 484,300.30 ‐ ‐ ‐$ ‐$ MBIARFMSI 2004‐SR1 A4 14,019,000$ ‐$ ‐$ ‐$ ‐$ MBIA

RFMSI 2004‐SR1 A1 0 ‐ ‐ ‐$ ‐$ MBIARFMSI 2004‐SR1 A2 0 ‐ ‐ ‐$ ‐$ MBIARFMSI 2004‐SR1 A3 0 ‐ ‐ ‐$ ‐$ MBIA

x x x x x x x x x x x x x x x x x x x x x

Page 22 of 22

ANNEX G

© 2007 BasePoint Analytics LLC. All Rights Reserved.

White Paper EARLY PAYMENT DEFAULT – LINKS TO FRAUD AND IMPACT ON MORTGAGE LENDERS AND INVESTMENT BANKS

Executive Summary

BasePoint wanted to investigate the link between fraud and its impact on early payment performance

BasePoint Analytics worked with several lenders over the past year to investigate and understand the increasing trends in Early Payment Default (EPD). Early Payment Defaults are typically classified as loans which become delinquent by more than sixty days in their first year. However, this definition varies by lender. Overall, lenders are experiencing an increase in serious delinquency within the first several months of the loan life. In particular, non-prime lenders are being hit by this trend.

One of the goals of the BasePoint study was to investigate the link between fraud and payment trends during the early life of a loan. In general, the earlier the serious delinquency occurs, the more likely it is linked to a significant misrepresentation on the original loan application.

Millions of historic loans were examined during the study

The fraud scientists at BasePoint Analytics, a leading fraud analytics and consulting firm, analyzed over 3 million loans that were originated between 1997 and 2006. The 3 million loans reviewed included over 16,000 examples of early payment defaults and loans that went into foreclosure.

BasePoint found that the percentage of EPD loans that can be attributed to a fraudulent misrepresentation on the original application varies from between 30% and 70%. The ratio varies based on a number of factors, including Portfolio type, Loan program, Underwriting policies, and EDP/FPD definitions and measurement periods.

The main drivers of EPD identified in the study were:

1. Loan program risk – e.g. Purchase, 80/20, Stated Income 2. Broker risk - Prior and current risk of broker 3. Borrower credit profile 4. Borrower income levels and employment characteristics 5. First Time Home Buyers 6. Seller Concessions 7. Geographic risk related to property and prior fraud experience

in that area

© 2007 BasePoint Analytics LLC All Rights Reserved. BasePoint Confidential 2

A small percentage of brokers accounted for most of the risk

The study also demonstrated that a small percentage of brokers contribute a disproportionate amount of loans that result in EPD. At one lender studied, less than 10% of brokers accounted for 100% of EPD, fraud and repurchase requests. This outcome was not unique. This study revealed that nearly all the EPD loans came through a small number of brokers.

Traditional credit risk tools don’t predict default risk well when the underlying information in the application is fraud

And finally, the study found that when egregious misrepresentations are made on a mortgage application, the value of the credit score to predict EPD deteriorates. Traditionally mortgage lenders and investment banks have relied on credit scores to assess the risk of a borrower. Credit scores effectively predict risk when the facts on the application are true. However, when a borrower or broker misrepresents fundamental characteristics such as: income, employment, debt or the value of the property, the credit score risk assessment isn’t as effective.

BasePoint also found that predictive analytics can be employed to effectively identify 40% or more of Early Payment Default risk prior to loan funding. This provides lenders and investment banks with a valuable tool in controlling their EDP losses.


Understanding the Relationship between Early Payment Default (EPD) and Mortgage Fraud

Through its work with leading mortgage lenders, BasePoint has had the opportunity to review data from several lenders to analyze linkages between EPD and material misrepresentations on the original loan application, classified as fraud.

BasePoint analyzed over 3 million loans that were originated between 1997 and 2006. The 3 million loans reviewed included over 16,000 loans that were confirmed to contain egregious misrepresentations in the loan file that later led to a default. These misrepresentations included fraud such as: income inflated by as much as 500%, appraisals that overvalued the property by 50% or more, fictitious employers and falsified tax returns.

The study concluded that misrepresentations can grossly affect the risk of a loan, finding that loans containing egregious misrepresentations were up to 5 times more likely to default in the first six months than loans that did not.

BasePoint found that the percentage of EPD loans that can be linked to a fraudulent misrepresentation on the original application varies from between 30% and 70%. The ratio varies based on a number of factors including Portfolio type, Loan program, Underwriting policies, and EDP/FPD definitions and measurement periods.

The main drivers of fraud and EPD identified in the study are:

Loan program risk (for example, stated income, low doc, 100% CLTV, and low credit profile)

Broker risk

Borrower credit profile (either low credit profile or credit profile er)

els and employment characteristics

Seller Concessions

Geographic risk related to property and prior fraud experience in that area

not matching loan program – an indication of a straw buy

Borrower income lev

First Time Home Buyers


The BasePoint study uncovered that the most common reasons for EPD were comprised of these layered risk factors (several factors combined to create high risk applications), including:

Credit risk not matching with stated income (for example, low credit scores and high income)

Age not matching with income (young borrower, very high income)

Key misrepresentation types that demonstrated to be indicators of Earl P ult include:

Stated income

Stated income loans were higher risk overall in terms of fraud levels. The stu floans m hers:

ir total book of business were higher risk.

performance and submitting stated income loans were more likely to

Eighth Periodic Mortgage Fraud Case Report to the Mortgage Bankers Association in April 2006 that one of their customers reviewed a sample of 100 stated income loans and found that when compared to IRS figures, “almost 60% of the stated amounts were exaggerated by more than 50%.” Clearly these types of income exaggerations significantly increase the risk of default.

First Time Homebuyer with 100% CLTV loans

Self Employed Borrowers combined with other risk factors

y ayment Defa

Occupancy

Property valuation (for example, appraisals not matching the area or neighborhood)

dy ound that there are specific drivers that made certain stated income ore risky than ot

Degree to which the stated income exceeded normal income for a given ZIP code.

Broker’s risk level - brokers that submitted more stated income loans as a percentage of the

Broker’s historical performance in terms of the number of prior loans that resulted in fraud or EPD. Brokers that had higher risk of bad

be involved in fraud.

Stated income compared to the Debt to Income ratio (DTI). A higher DTI and a higher income indicated a higher risk of income fabrication.

The Mortgage Asset Research Institute, Inc. (MARI) reported in their


The BasePoint study also determined that a small percentage of brokers contribute a disproportionate amount of loans that result in EPD. At one non-prime lender studied, less than 10% of brokers accounted for 100% of EPD, fraud and repurchase requests. This was not unique to this lender. In general, the study revealed that EPD is concentrated within certain brokers, with nearly all the bad loans linked to a relatively small amount of brokers.

Early Payment Default risk is concentrated to a very few brokers

BasePoint discovered that most risk of early payment default for lenders came from approximately 6% to 8% of brokers. The overwhelming majority of brokers submitted no early payment defaults, even those brokers that submitted extremely high volumes.

Brokers that submit loans with homogenous characteristics, such as common employers, and loans with more combined risk factors, such as CLTV, Loan Type, Credit Score, and Employment Type, will be more likely to have loans involved in early pay defaults, discount sales and repurchase requests.

Mortgage brokers originate the majority of non-prime mortgages and the current broker incentive structure is completely focused on closing as many loans as possible at the most favorable terms for the lender. Brokers are not accountable for the quality or performance of the loan once it closes. When a loan closes, in addition to their broker fee, they may also receive a yield-spread premium from the lender for closing a higher interest rate loan. The brokers have a financial incentive to place borrowers into loans with higher interest rates than they might otherwise qualify to obtain. This incentive structure, along with stated income and low documentation loan programs, provide ample opportunity for the small number of unscrupulous brokers to take advantage of the system by perpetrating fraud and by closing loans with an increased propensity to default.

The study also revealed that geography impacts EPD rates. At one lender included in the study, the top 20 risky areas accounted for about 14% of volume, but accounted for 45% of EPD, 25% of confirmed frauds, and 20% of repurchased loans.


And finally, the study found that when egregious misrepresentations are made on a mortgage application, the value of the credit score to predict EPD deteriorates. Traditionally mortgage lenders and investment banks have relied on credit scores to assess the risk of a borrower. Credit scores effectively predict risk when the facts on the application are true. However, when a borrower or broker misrepresents fundamental characteristics such as: income, employment, debt or the value of the property, the credit score risk assessment isn’t as effective.

Findings on the Impact of EDP

The increase in early payment defaults is placing a profitability strain on originators. The majority of US mortgage loans are packaged and sold on the secondary market to investment banks. The investment banks have the right to return purchased loans if they experience EPD generally in the first 90 days after purchase (however the time period can vary by agreement).

Correlated to the increase in EPD being seen throughout the industry, an increase in associated repurchase requests is putting financial pressure on many lenders, especially those with large non-prime portfolios. This has been the cause of several prominent lenders going out of business or selling their non-prime wholesale lending units during late 2006 and early 2007. Other lenders have been forced to dramatically increase their loss reserves, causing quarterly losses and significant impacts on company valuations.

“Many lenders are facing increases in repurchase requests and early payment defaults. In an effort to help lenders deal with these challenges, BasePoint has rigorously studied the issue and found a direct correlation between EPD and mortgage fraud,” said Tim Grace, president and CEO of BasePoint. “We can demonstrate for lenders and investment banks how they can substantially reduce their EPD losses, often within a short period of time. The cost of mortgage fraud is borne by every person or family who buys or sells a home. That’s why BasePoint continues to focus on developing advanced software solutions to put a stop to mortgage fraud before it happens.”

One of the study’s primary findings was the tremendous impact of EPD on lender profitability. Lenders are experiencing as much as $100 million per month in First Payment Defaults alone. The study found that EPD rates vary considerably by lender, based on factors such as loan program mix, and underwriting policy. What the lenders share is a common concern regarding an EPD escalation trend.


Lender impact examples:

EPD Rate (percentage of origination volume)

Measurement Period

Lender A 2.5% First month

Lender B 4.0% First month

Lender C 8.0% First 8 months

As a result of the increases in EPD and associated loan repurchase requests lenders are taking drastic action to curtail losses:

Making sweeping changes to underwriting policy

Eliminating underperforming loan programs

sentially forces lenders to reduce their origination volumes (through major underwriting policy changes or

e results in higher loss ratios and less revenue to offset losses which will continue to roll through from loans

core cutoffs. For non-prime lenders, this can actually make

Selling portfolios

Modifying profitability projections

In addition, investment banks are no longer willing to purchase pools of loans from loan programs that have demonstrated higher incidents of EPD. This gives the lender few options other than to curtail those programs entirely. This es

loan program eliminations).

While many of these changes may appear prudent given the current environment, BasePoint believes that there are more targeted, data-driven solutions to retain the good origination volume that will now be lost, while substantially reducing both fraud and EPD risk. Now is the time that lenders need to preserve as many new loans which would perform as well as possible, while strategically eliminating fraud as well as loans with a high risk of EPD. Eliminating good loan volume through these dramatic and overarching policy changes can in fact exacerbate their profit squeeze in the short-term. Increasing defaults and shrinking origination volum

originated in 2005 and 2006.

It is also important to note that as this study has revealed, EPD is not solely a credit-quality issue. Many lenders, especially in the non-prime market, have reacted to this increase in their EPD losses by making sweeping policy changes, including somewhat arbitrary increases in credit s


both the fraud and EPD problem worse, as adverse selection takes place.

Here is an example of how this adverse selection can happen. In an effort to fend off increasing delinquencies, a lender may choose to increase their minimum credit score cut-offs. In doings so, many performing loans will now been eliminated with the credit score cut-off increase. While it may lower the overall number of EPD loans, the rate may actually get worse. Now the loans they are booking are higher credit score, non-prime loans. These higher credit score borrowers should be able to qualify for a prime product. BasePoint has found this to be a risky population in sub-prime portfolios. They may actually carry more risk than the segment that has been eliminated. As a

manage their delinquency risk. However, this study found that while average credit scores have been

creasing in many loan pools, fraud rates, default rates and loan rates are also on the rise.

nt study was also designed to evaluate the effectiveness of predictive analytics in identifying EPD risk during the

the loans to be rejected pre-funding. In fact, predictive models such as BasePoint’s FraudMark™ for Origination

D risk. The following chart demonstrates that at a 15% review rate, FraudMark for Origination detects over 35% of EPD dollars re-funding, which is 260% better than the credit score in predicting

EPD risk.

result of the increased credit score requirement, the originator may adversely select a more risky population.

Mortgage originators are not alone in their concern for this increase in EPD. Investment banks that purchase pools of mortgage loans are also impacted by this trend. Traditionally, investment banking firms relied on credit and compliance tools to

inrepurchase

Tools to Mitigate EPD Losses

In addition to investigating the relationship between fraud and EPD, and understanding the impact of EPD on lenders and investment banks, the BasePoi

origination process for lenders, and the due diligence process for investment banks.

The study concluded that predictive models can be deployed early in the loan process to help lenders predict which loans are likely to default early, enabling

correctly identify 40% of a lender’s loans pre-funding that, if booked, would result in EPD.

As a part of this study, a sample of loans was evaluated to determine the effectiveness of the credit score and the FraudMark score to identify EP

p


R eviewRate

C redit ScoreEPD $

De n tectioRat e

FraudMark™ for O ion riginat

EPD $ Detect ate ion R

FraudMark Lift over

C redit ScoreDetecti n o

5% 5% 14% 180%

10% 7% 27% 286%

15% 10% 36% 260%

This illustrates that a predictive model built specifically to target fraud can also accurately identify EPD risk. This is not surprising now that

ed

other credit risk factors and socio-economic factors that

ed in various sectors of the financial services industry for many

we understand the inherent relationships between EPD and fraud.

FraudMark uses sophisticated analytic scoring technology to identify suspicious mortgage loan applications, enabling the lender to quickly and effectively stop them before funding. These models accurately predict the likelihood of a loan containing fraud that will result in financial loss to the lender. FraudMark enables lenders to decrease fraud losses by automatically identifying loans that have the highest associated fraud risk. Over the past year FraudMark has helplenders prevent nearly $1 billion in suspicious loans from funding.

But as demonstrated, the overlap between EPD and fraud is not complete. There are credit risk factors and other significant events such as natural disaster, medical crisis, and loss of employment that impact EPD and are unrelated to fraud. BasePoint is in the process of designing a predictive analytic solution to more specifically target EPD, and will include both the indicators of fraud that are also linked to EPD, as well on thepredict EPD.

The concept of mining historical loan applications for data patterns that provide clues to when an application is fraudulent or has a high propensity for EPD is a technique called “pattern recognition”. This technique to detect fraud in its earliest stages has been successfully employyears.

One of the most important aspects of building a successful predictive model is selecting the model target. When a single model is used to attack multiple business problems, thus multiple targets, often the


performance of the model deteriorates. That is why BasePoint recommends a multiple-model approach to successfully attack the unique but related business problems of fraud and Early Payment

Due to the impact of broker risk on EPD losses, broker monitoring is

hnology produces a score that indicates the level of risk on any given broker, and is

monitoring solutions can help lenders and investment banks quickly and precisely reduce their EPD

sses and improve their profitability.

itigate their EPD exposure. These tools have been successfully used in other segments of the financial industry for

Lenders and investment banks alike are taking action to adopt analytic solutions to help control EPD.

Default.

another way to help lenders control their risk.

BasePoint had found that brokers who submit loans with homogenous characteristics, such as common employers, and loans with more combined risk factors, such as CLTV, riskier loan types, credit score, and employment type, are more likely to have loans involved in early pay defaults, discount sales and repurchase requests. Because of this, BasePoint has developed patent-pending Dynamic TRAITS™ technology, which tracks the history of a broker to catch anomalies and escalation of risk over time. The TRAITS tec

refreshed daily to augment the FraudMark score.

These predictive analytic and broker

lo

Lenders are Taking Action to Mitigate EPD Risk

EPD is a significant and growing financial issue for lenders. While there are thousands of unique scenarios that can lead to EPD, the patterns and trends can be detected with advanced analytic solutions. In addition to established underwriting and due diligence practices, advanced analytics and pattern recognition can be leveraged to help lenders m

decades.


About BasePoint Analytics

BasePoint Analytics is a leading fraud analytics and consulting company serving the mortgage and banking industries. Using science to analyze historical transactions, BasePoint develops advanced fraud and EPD scores to fit each organization's unique needs. Lenders benefit from sophisticated predictive analytics that quickly identify fraudulent and other risky activity, minimizing losses while accelerating the processing of performing applications and transactions. Leveraging a client’s existing technology, BasePoint provides clients with immediate results and quick return on investment.

Leading Scientists and Top Industry Consultants BasePoint's team of renowned scientists is dedicated to building state-of-the-art predictive models using the latest advanced techniques. Our fraud and EPD specialists have innovated transactional, application and account-based models that are detecting fraud and EPD in some of the largest organizations across the globe. BasePoint is committed to investing heavily in research and development to provide you with the industry's most effective defense against fraud.

Having successfully managed fraud and risk operations for dozens of the world's highest profile organizations, BasePoint's elite team of consultants has deep domain expertise. Clients can have confidence these professionals will successfully guide you to significant fraud and risk reduction through the integration of analytic models, tools, strategy alignment, and operational best-in-class processes.

A Global Focus Fraud and EPD risk do not have geographic boundaries and neither does BasePoint. Our experts have spent years understanding the global nature of fraud and risk migration, and more than a decade researching fraud, risk trends and management throughout the world. Whether your organization is local, national, or spans many continents, we have the expertise and solutions to help.


ANNEX H

Analyzing GSE Mortgage Buyback Demands: Lender lmpact Varies Significantly

GSE Buyback Demand Activity: 2006-2008(Dollars in Miilions)

Repiiichasd Demandt::ir,' Ll---- jt i iii bispqsiiion,at oCmands

Rank Seller/Oriqinator Volume Pct Assets Repurchased Wthdrawn Disputed Pending

IIttIIIIllIIlIllIIIIIrl

1

2

3

4(6

7

q

10

11

12

13

14

'16

17

'18

19

20

21

22

23

24

25

26

27

28

29

30

31

33

34

35

COUNTRYWDE

WELLS FARGO

CHASE HOME FINANCE

BANK OF AMERICA

CITIMORTGAGE

SUNTRUST MORTGAGE INC.

GMAC MORTGAGE/ALLY

TAYLOR, BEAN & W-IITAKER MORTGAGE

FLAGSTAR BANK, FSB

U.S. BANK N.A.

AMSOUTH BANK

WASHINGTON MUTUAL

NATIONAL CITY BANK

INDYMAC BANK, FSB

WACHOVIA MORTGAGE, FSB

LEHMAN BROTHERS

MORGAN STANLEY

HSBC MORTGAGE CORPORATION (USA)

FIRST HORIZON HOME LOAN

ABN AMRO MORTGAGE GROUP, INC.

EMC MORTGAGE CORPORATION

FIFTH THIRD BANK

GREENPOINT MORTGAGE FUNDING, INC.

OHIO SAVINGS BANK

DB STRUCTURED PRODUCTS, INC.

PHH MORTGAGE/CENDANT

FREEDOM MORTGAGE CORPORATION

BRANCH BANKING & TRUST

GOLDMAN SACHS MORTGAGE COMPANY

HOMEBANC MORTGAGE CORPORATION

PULTE MORTGAGE LLC

REGIONS BANK

DLJ MORTGAGE CAPITAL INC,

BANKUNITED, FEDERAL SAVINGS BANK

MORTGAGE ACCESS/WEICHERT FINANCIAL

s16,2'16.06

$7,073.59

$6,766.26

$5,373.05

$3,966.04

$3,026.1 1

$1,537.81

$1,464.46

$1,224.15

s1,094.07

$996.93

$979.84

$744.1 E

$736.87

$714.08

$71 1.96

$638.70

$580.65

$521.36

$493.62

$491.62

$490.1 2

$403.94

$326.03

$283.0 1

$279.84

$278.35

$21 2.39

$1 97.98

$110.12

$95.95

$90.30

$80.05

977.94

$69.79

$69.08

966.81

$65.60

$53.36

$44.31

$42.76

$40.22

$36.91

$32.23

$30.82

$29.41

$27.54

$26.1 0

$25.75

$23.36

$22.55

$21.1 6

$19.7s

$18.04

$17.00

$1 6.94

5.47% 15.45%

6.14% 7.75%

2.860/0 8.25o/o

8.460/o 11.25%

1.49% 16.?204

1.81% 12.37%

0.50% z.ovlo

0.63% 0.26%

1.360/o 19.'l9o/o

5.290k 4.93%

2.37o/o 10.35%

3.77o/o 12.93o/o

1.11Vo 19.67%

1.360/0 2.34%

3.72o/o 9.79%

O.O7o/o 67 .89o/o

1.74o/o 4.6o0/o

O.74o/o 9.20ok

2.43Vo 18.24%

2.88% 12.37%

4.81Vo 9.96%

0.80o/o 2.99o/o

1.18% 25.12%

1.72Vr 7 .O7ok

6.69% 4.270k

1.46% 16.O10/o

1.1'lo/o 18.05%

2.960/o 3.84%

5.64"/o 7.8O%

3.12o/o 2.354/o

6.810/o 2'1.98o/o

2.40o/o 4.7 4o/o

9.25o/o 3.28%

0.00% o.48%

3.51o/o 15.68%

4.49o/o 2.78o/o

1.91Vo 7 .77%

2.97o/o 16.06%

9.19% 2.500/o

o.770/o 12.51%

0.85% 0.00%

'16.310/o 2.260/0

6.85% 1o.83o/o

1.17o/o 6.310/o

0.00% 0.00%

7.98o/o LO4Vo

1s.15% 22.01%

0.00% 1.24%

0.00% 54.35o/o

1.24o/o 10.05%

1.660/o 16.53%

3.87"/" 2.860/o

0.00% '14.650/o

4.31o/o 2.78o/o

15-72o/o 4.77o/o

0.55% 12.14o/o

3.13%

1.AsVo

3.24%

3.44%

2.430h

3.08%

1.49%

3.12Vo

2.460/o

1.9AYo

2.79o/"

0.96%

1.85Vo

1.30o/o

1.63%

2.600/o

6.25%

2.60%

1.90%

1.61%

7.O4Vo

2.15o/o

9.90%

1.57%

8.48%

0.79%

4.76Vo

1.O2%

3.04Vo

3.59o/o

1.28o/o

1.14%

4.14%

6.15%

1.93o/o

1.98%

O.47Vo

0.84Vo

5.88%

1.31%

20.1'lo/o

2.91o/o

1.55%

0.87%

41.97o/o

o.72%

2.12o/o

17.O8%

2.75%

1.05o/o

3.11o/o

1.48o/o

O.7Eo/o

1.43o/o

1.650/o

2.2Oo/o

47.71o/o 32.40%

49.76% 43.11o/o

52.060/0 4O.72d/o

39.46% 42.18%

53.50% 29.72o/o

57.39% 32.77o/o

67.56"/o 35.62%

24.12Vo 23.2Oo/o

4D.92o/o 39.02o/o

49.04% 45.45%

51.47o/o 36.10%

90.33% 36j70h

64.460/o 16.21%

82.91o/o 13.14Vo

65.92% 2O.85o/o

3.7Oo/o 9.05%

36.65% 70.310/o

69.43% 21.960/o

46.170/o 34.91%

50.10% 40.79%

53.53% 66.50%

56.89% 43.02%

39.33% 38.670/o

53.25o/o 38.21Vo

45.74% 56.'13%

34.54o/o 50.84%

55.89% 25.35%

61.05% 33.39%

35.58% 61.68Yo

45.17o/o 52.76Vo

15.58o/o 55.620/0

72.08o/o 25.39%

40.35% 74.22%

7.27o/o 92.03%

23.72% 58.98%

31.600/o 67 .49%

53.97% 36.340/o

55.55% 28.33o/o

22.35% 68.29%

56.99% 29.73%

57.35ok 62.9'10/o

41.31% 53.80%

21 .83Vo 61.27oh

63.61% 33.60%

0.00% 100.0070

21.43% 63.47%

29.78% 33.05%

55.33% 79.41%

35.28% 'l9.37o/o

62.06% 27.17%

5O.28o/o 31.52o/o

59.15% 36.140/o

67.880/o 17.47Vo

1.45o/o 41.35o/o

23.A6% 68.72d/o

34.46% 52.85o/o

36 PROVIDENT FUNDING ASSOCIATES

37 USAA FEDERAL SAVINGS BANK

38 SOVEREIGN BANK

39 E"TRADE BANK

40 IRWN MORTGAGE CORPORATION

41 FIRST NATIONAL BANK OF NEVADA

42 CENTEX/HAR\AIf,ODSTREETFUNDING

43 CHEVY CHASE BANK FSB

44 PNC MORTGAGE

45

46

47

48

49

50

51

53

54

55

56

GOLDEN FIRST MORTGAGE CORPORATION

M&T MORTGAGE CORPORATION

CTX MORTGAGE COMPANY LLC

NOMURA CREDIT & CAPITAL, INC.

UNIVERSAL MORTGAGE CORPORATION

COLONIAL SAVINGS FA

OPTEUM FINANCIAL SERVICES, LLC

R&G MORTGAGE CORPORATION

DO\\NEY SAVINGS AND LOAN ASSOCIATION

AMERICAN HOME MORTGAGE CORPORATION

MORTGAGE LENDERS NETWORK USA, INC

METLIFE HOME LOANS

Copyright @ lnside Mortgage Finance Publications (301) 951-1240 Page 11

Analyzing GSE Mortgage Buyback Demands: Lender lmpact Varies Significantly

Reaurclase Demands Disposition of DemandsRank Seller/Orioinator Volume Pct Assets Reourchased Wthdrawn DisDuted Pendino

646 THE FARMERS AND MECHANICS BANK

648 MVB MORTGAGE CORPORATION

649 HERITAGE FEDERAL CREDIT UNION

650 ALTRA FEDERAL CREDIT UNION

653 COMMUNITY NATIONAL BANK

652 MINSTER BANK

651 PEOPLES COMMUNITY BANK

655 MAUCH CHUNKTRUST CO,

654 THE CITIZENS SAVINGS BANK

656 FINANCIAL PLUS FEDERAL CREDIT UNION

657 AMERICAN BANK & TRUST

658 VANDYK MORTGAGE CORPORATION

659 FARMERS CITIZENS BANK

660 CHRISTIAN COMMUNITY CREDIT UNION

661 MARKLEBANK

662 DAKOTALAND FEDERAL CREDIT UNION

663 DHCU COMMUNIry CREDIT UNION

664 CARLSBAD NATIONAL BANK

665 DELTA COUNTY CREDIT UNION

666 COMMUNITY TRUST BANK, INC

668 GOLDEN MORTGAGE BANKERS

667 THE NATIONAL BANK

669 HEART\ /ELL MORTGAGE CORPORATION

670 FIRST FARMERS BANK & TRUST

671 FIRST NATIONAL BANK OF GRANT PARK

672 TO\AA AND COUNTRY BANC IiIORTGAGE SERVICES

673 MID-IVIISSOURI MORTGAGECOMPANY

674 NEWREPUBLIC SAVINGS BANK

675 WEST END BANK, S.B.

676 INDIANA UNIVERSIry CREDIT UNION

677 AMERICANTRUST FEDERAL SAVINGS BANK

678 THE STATE BANK AND TRUST COMPANY

579 HERGETBANK, NATIONALASSOCIATION

680 CHEVIOT SAVINGS BANK

681 FIRST FEDERAL SAVINGS BANK OF IOWA

682 CFCU COMMUNITY CREDIT UNION

683 BAYBANK

684 PULASKI BANK, A SAVINGS BANK

685 IDAHO CENTRAL CREDIT UNION

686 SOY CAPITAL BANK AND TRUST COMPANY

687 MACKINAC SAVINGS BANK

688 NORTHERN MICHIGAN BANK & TRUST

Grand Total

$0.08

$0.08

$0.07

$0.07

$0.07

$0.07

$0.07

$0.07

$0.07

$0.07

$0.07

$0.07

$0.07

$0.07

$0.06

$0.06

$0.06

$0.06

$0.06

$0.06

$0.06

$0.06

$0.06

$0.06

s0 06

0.28o/o

5.28o/o

0.14o/o

0.06%

O.19Yo

0.52o/o

1.O8o/o

o.37%

o.45%

0.86%

o.1404

$0.07 11.26Vo

o.92%

0.27%

0.37%

1.48Vo

0.20%

O.'l9o/o

0.31o/o

0.05%

0.55%

1j8%0.78%

0.08%

0_06%

$0.06 12.10%

o.oo%

o.oo%

o.oo%

1 00.00%

0.00%

0.00%

0.00%

1 00.00%

0.00%

0.00%

1 00.00%

100.00%

100.00%

1 00.00%

0.00%

100.00%

1 00.00%

1 00.00%

100.00%

100.00%

O.0oo/o

0.00%

0.00%

0.00%

0.00%

100.00%

0.007o

100.00%

0.00%

1 00.00%

O.0oo/o

0.00%

1 00.00%

0.00%

1 00.00%

1 00.00%'t 00.00%

1 00.00%

0.00%

0.00%

0.00%

100.00%

45.54%

1 00.00%

100.000/o

0.00%

0.00%

0.00%

O.OOo/o

0.00%

0.00%

0.00%

0.00%

O.OOo/o

O.OOVI

0.00%

0.00%

o.oo%

0.00%

o.o0%

o.o0%

0.00%

0.00%

0.00%

O.0oo/o

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

o.oo%

0.00%

o.oo%

O.00Vo

0.00%

0.00%

0.00%

0.00%

O.0Oo/o

0.00%

O.jOYo

0.00%

O.OOo/o

0.00%

0.00o/o

0.00%

0.00%

0.00%

0.00%

12.58%

0.00% 100.00%

0.00% 100 00%

$0.06

$0.05

$0.05

$0.05

$0.0s

$0.05

$0.05

$0.04

$0.04

$0.03

$0.03

$0.03

90.03

$0.02

$0.02

$65.836.91

0.41%

0.28Vo

0.41o/o

0.39%

0.11vo

O.260/o

o.25Vo

0.09%

0.17%

0.24%

0.190h

0.18%

0.19%

7.75%

0.03%

2.40%

1 00.00%

0.00%

o.oovo

1 00.00%

0.00%

0.00%

425.OOV|

0.00%

'100.00%

0.00%

0.00%

0.00%

0.00%

0.00%

39.34Vo

1 00.00%

1 00.00%

1 00.00%

0.00%

O.OOVI

0.00%

0.00%

1 00.00%

0.00%

'100.00%

1 00.00%

0.00%

100.00%

0.00%

0.00%

0.00%

0.00%

1 00.00%

1 00.00%

1 00.00%

0.00%

35.75%

0.000/o

0.00%

0.000/o

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

O.0Oo/o

0.00%

0.00%

o.00vo

0.00%

o.00%

0.00%

0.00%

1 00.00%

0.007o

0.00%

0.00%

0.00%

0.00%

0.00%

o.o0%

0.00%

0.00%

O.OOo/o

0.00%

0.00%

0.00%

0.00%

0.00%

0.o0%

0.00%

4.15%

0.00% 100.00%

Note: Data cover rcpurchase demands on moftgages secuitized by Fannie Mae and Freddie Mac frcm 2006 through 2008. Sellerhiginator data are for sellers of

Source: /nside Moftgage Finance analysis of Fannie Mae and Freddie Mac SEC dlsclosures

Copyright @ lnside Mortgage Finance Publications (301) 951-1240 Page22

ANNEX I

Mortgage Finance Mortgage Repurchases Part II: Private Label RMBS Investors Take Aim - Quantifying the Risks

See Important Disclosures on the Last Page of this Report

Summary During the course of mortgage loan sales, selling lenders make certain representations and warranties to buyers such as the GSEs and bond investors that hold the securitized loans. Breaches of these representations and warranties cause the selling lender to have to repurchase the loan or indemnify the buyer against future losses. As analyzed in our March 15, 2010 report “GSE Mortgage Repurchase Risk Poses Future Headwinds: Quantifying the Losses”, we estimated the potential unrecognized liability related to GSE repurchase requests. Due to increasing litigation activity by private label RMBS investors, we believe that liability may also lurk for originators/underwriters of the initial securitizations and could approach 5% to 15% of tangible book value. As such, based upon information contained in pending lawsuits, we have analyzed securitization data in an attempt to frame the potential liability that could exist. See the table below for a summary of estimated losses.

Key Points

FHLB lawsuits. Since late 2009, several FHLBs have filed suit against multiple underwriters of Alt-A and subprime MBS deals citing inaccurate claims in the initial prospectus such as the percentage of high LTV loans, amount of investor properties, or number of underwriting exceptions. Utilizing sales information from foreclosed properties within the deal, the suits have compiled convincing data to show that the loan underwriting was materially worse than stated in the initial prospectus. Combined, the lawsuits (FHLBs of Pittsburgh, Seattle, San Francisco) are requesting rescission on about $25.6B in MBS purchases.

Investor syndicate with substantial clout gearing to pursue loan buybacks. An investor group representing $500B in MBS securities has sent letters to Trustees of mortgage backed securitizations requesting that they enforce servicing breaches related to improperly originated loans. According to a July 21 Reuters article, the group has topped the required 25% ownership threshold needed to enforce Trustees to compel the servicers to hand over documentation (i.e. loan files), or be removed from the deal.

FHFA subpoenas. On July 12, the Federal Housing Finance Agency (FHFA), issued 64 subpoenas seeking documents for MBS securities that Freddie and Fannie had invested in. Previously, the GSE’s had been requesting documentation (i.e. loan files) to determine potential reps and warranty breaches; however, due to a lack of success, the FHFA was forced to use their subpoena power to compel the documentation.

Potential liability. With the majority of the subprime/Alt-A originators out of business, most of the litigation is targeted at the underwriters of the initial securitizations. The suits generally claim, among other items, that the underwriters of the securitizations misrepresented the profile of loan standards within the initial prospectus.

Chris Gamaitoni 202-534-1387 [email protected]

Jason Stewart 202-540-7306 [email protected]

Mike Turner 202-534-1380 [email protected]

August 17, 2010

Source: Compass Point Research & Trading LLC, Bloomberg, Inside MBS & ABS, Asset Backed Alert

Total Alt‐A & Subprime RMBS Repurchase Request Loss Estimates

Company Ticker Rating Loss ($M) Per Share* % of TBV Loss ($M) Per Share* % of TBV Loss ($M) Per Share* % of TBV

Bank of America BAC NR 44,977 $2.69 22% 35,204 $2.11 17% 16,728 $1.00 8%

JP Morgan JPM NR 32,922 $4.93 19% 23,941 $3.59 13% 9,006 $1.35 5%

Deutsche Bank DB NR 20,892 $18.65 31% 14,070 $12.56 21% 4,463 $3.98 7%

Goldman Sachs GS NR 15,103 $16.77 15% 11,194 $12.43 11% 4,197 $4.66 4%

RBS Greenwich RBS NR 15,282 $0.16 19% 9,417 $0.10 12% 1,919 $0.02 2%

Credit Suisse CS NR 12,151 $6.15 30% 8,898 $4.50 22% 3,743 $1.89 9%

UBS UBS.N NR 12,262 $1.94 22% 8,350 $1.32 15% 2,830 $0.45 5%

Morgan Stanley MS NR 8,312 $3.56 15% 7,855 $3.37 14% 4,498 $1.93 8%

Citigroup CS NR 9,964 $0.21 5% 7,819 $0.16 4% 3,729 $0.08 2%

Barclays BCS NR 3,789 $0.19 4% 3,583 $0.18 3% 2,068 $0.10 2%

HSBC HBC NR 3,555 $0.12 2% 3,515 $0.12 2% 2,071 $0.07 1%

Total 179,210 133,846 55,253

* after‐tax (assume 40%)

Worst Case Base Case Best Case

Compass Point Research & Trading, LLC Mortgage Repurchases Part II: Private Label RMBS Investors Take Aim—Quantifying the Risks 2

Chris Gamaitoni | 202-540-7387 | [email protected] Jason Stewart | 202-540-7306 | [email protected] Mike Turner | 202-534-1380 | [email protected]

Litigation Background: a Brief History

Since September 2008, there have been a number of lawsuits aimed at originators of subprime and Alt-A mortgages by either investors in private label (non-government guaranteed) RMBS securities, or the companies that insured them. In 2008 and 2009, bond insurers MBIA, Syncora, and FGIC all filed separate lawsuits against Countrywide (later amended to include Bank of America). Generally, these lawsuits claim that a significant portion of the loans underlying the securitizations that they guaranteed failed to comply with the underwriting guidelines or other reps and warranties. In December 2008, Greenwich Financial, on behalf of a bondholder group, filed suit against Countrywide charging that they violated securitization agreements in modifying loans as part of their $8B settlement with Attorney Generals from multiple states. Since late 2009/early 2010, lawsuits have been filed on the behalf of the Federal Home Loan Banks of Pittsburgh, Seattle and San Francisco. Similar to some of the mortgage insurer lawsuits, the lawsuits all claim that, among other things, a significant portion of the loans underlying the securitizations did not comply with the standards that were cited within the securitization prospectus. However, unlike the lawsuits by the mortgage insurers which are directed at the originator, the FHLB suits are against the underwriters of the securitizations. Accordingly, the suits believe the underwriters should be held liable since they misrepresented the information contained in the prospectus. They are seeking rescission on approximately $25.6 billion in RMBS purchases. In July 2010, an investor syndicate purportedly representing $500B in MBS sent letters to numerous trustees of mortgage backed securitizations requesting that they enforce servicing breaches related to improperly originated loans. The group was formed in order to assemble enough representation to exceed the required 25% or 50% thresholds needed to compel the trustee to take action against the servicer. For reference, the trustee technically manages the securitization trust, and has the duty to ensure the servicer complies with all requirements in the securitization documents. Statements from the syndicate’s attorneys have stated that they have 25% voting rights for over 2,300 deals, 50% in over 900 deals, and 66% in more than 450 deals. The group is represented by Talcott Franklin, a Dallas-based firm that was founded by an attorney who previously worked on a bondholder lobbying effort that was related to the Greenwich Financial litigation. The firm appears to have been established specifically for taking on this effort.

Date Action Amount

Sep‐08 a MBIA sues Countrywide and BAC $1.4B

Status: In April 2010, the Judge denied motion to dismiss (some

counts). All parties have appealed the Judge's ruling, and such

appeals are pending. Discovery has commenced.

Dec‐08 b Greenwich Financial sues Countrywide Decl. Jdg.

Status: Awaiting ruling from NY State Supreme regarding

Countrywide's motion to dismiss

Jan‐09 c Syncora sues Countrywide and BAC $0.4B

Status: In April 2010, the Judge granted Defendant's motion to

dismiss (some counts). Appeals are pending. Judge has ordered

Countrywide to produce all loan files regarding 3 securitizations.

Defendants' have filed counterclaims against Syncora for breach

of contract. Syncora has agreed to stay proceedings against BAC.

Claims against Countrywide continue.

Sep‐09 d FHLB Pittsburgh lawsuits ‐ multiple defendants $2.6B

Status: After being removed from state court to federal court,

the cases were remanded back to Court of Common Pleas in Dec.

2009. Defendants in each lawsuit have filed motions to dismiss

with the Court, and a hearing on the motions is scheduled for

August 25, 2010.

Dec‐09 e FGIC sues Countrywide (now BAC) $1B

Status: Judge granted Countrywide's motion to dismiss only as to

the claims of negligent misrepresentation and breach of implied

covenant of good faith and fair dealing. The Judge denied the

motion to dismiss as to the claims of fraud. Both parties have

filed appeals, which are pending.

Dec‐09 f FHLB Seattle lawsuits ‐ multiple defendants $4B

Status: Cases moved to Federal court. On July 29, 2010, Plantiffs

argued motion to remand all cases to State court. Awaiting

decision.

Mar‐10 g FHLB San Francisco lawsuits ‐ mutiple defendants $19B

Status: Cases filed in state court and removed to federal court by

defendants. FHLB has filed motion to remand to state court.

Motion hearing set for 9/17/10.

Jul‐10 h FHFA issues 64 subpoenas for loan files N/A

Status: unknown, private

Jul‐10 i Investor group announces intentions to file suit $500B

Status: nothing publicly filed yet

Aug‐10 j NY Federal Reserve engages in actions to enforce $70B

repurchases on faulty mortgages acquired through Bear

Stearns and AIG

Status: unknown, private

Sources

a. http://www.mbia .com/investor/lega l_proceedings .html

b. Greenwich Financia l Services , et a l . v. Countrywide Finacnia l Corp., et al .; SCROLL

c. Syncora Guarantee Inc. v. Countrywide Home Loans , Inc., et a l .; SCROLL

d. FHLB of Pittsburgh's Form 10‐Q for the Quarter Ended June 30, 2010; PACER

e. Financia l Guaranty Insurance Company v. Countrywide Home Loans , Inc.; SCROLL

f. FHLB of Seattle 's Form 10‐Q for the Quarter Ended June 30, 2010; PACER

g. FHLB of San Francisco 's Form 10‐Q for the Quarter Ended June 30, 2010; PACER

h. July 12, 2010 Federal Hous ing Finance Agency news release

i . July 21, 2010 Reuters article "Mortgage bond holders get lega l esge: buybacks seen"

j. Aug 4, 2010 Bloomerg article "N.Y. Fed May Require Banks to Buy Back faul ty Mortgages , Assets"


All the lawsuits generally make similar claims—that a significant portion of the underlying loans failed to comply with the underwriting guidelines or other reps and warranties and thus misrepresentations and material omissions were made in connection with the sale of private label RMBS. As background, during the securitization of loans, the underwriter (or originator, in the case of the mortgage insurer) makes certain representations and warranties that the underlying loans conform with the standards set forth in the securitization prospectus. Some of the most common misrepresentations cited in the lawsuits that have been filed are:

Stated loan-to-value ratios were lower than actual LTVs Failure to disclose additional liens on properties Property values were based on overstated valuations Overstating the number of mortgages on primary residences Originators of mortgage loans securing collateral pools departed from underwriting standards

In order to have conclusive proof that a significant portion of the underlying loans did not conform to the initial underwriting guidelines, the best source of information is loan file documentation. This point is made clear via statements in the FHFA subpoenas; “… the Conservator is seeking the contents of loan files, which include documents used in the underwriting process, such as loan applications and property appraisals.” (July 12, 2010 FHFS news release) While the GSEs, via the FHFA, have the power to subpoena the servicers of the securitization to turn over the documentation, other RMBS investors, such as the FHLB, do not have direct access to the files and must litigate in an attempt to gain access to the loan files. Based on the information provided, there appear to be two routes currently implemented by investors: File suit against the securitization underwriter. Utilizing statistical analyses of trust performance, the FHLB suits have

attempted to prove that the only way for the underlying loan performance to have performed as poorly as they did was if the underwriting was materially different than stated. If a judge does not dismiss the case, the plaintiffs are likely to gain access to the loan files via the discovery phase of the litigation (there has been no decision in the FHLB cases yet). To date, among the various lawsuits listed above, only in Syncora v. Countrywide/BAC have the defendants been ordered to produce loan files.

or Garner the required 25% or 50% voting rights from securitization investors in order to compel the trustee to force the servicer

to provide the required documentation (or be removed as acting trustee). This is the route the $500B investor group is initially taking. Thus, the group conceivably should have a greater chance of accessing loan files as the deciding factor may not hinge on a judge’s decision.

As previously noted, the FHLB suits are requesting rescission of about $25.6B in RMBS purchases. However, we believe these suits, the investor syndicate, the GSE’s and the Fed, ultimately are looking to have the underwriter, or the originator (if they are not bankrupt), repurchase only the underlying loans that did not abide by the underwriting standards stated in the prospectus.

Also in July 2010, the FHFA, acting on behalf of Fannie and Freddie, issued 64 subpoenas seeking documents related to private-label mortgage backed securities in which they invested. The FHFA intends to utilize the information to determine whether the issuers (underwriters) and others may be liable for certain losses suffered. The ultimate goal is “to determine whether misrepresentations, breaches of warranties, or other acts of omissions occurred that would require them to repurchase loans underlying the securitizations.” (July 12, 2010 Federal Housing Finance Agency news release) Most recently, the New York Federal Reserve stated in August that they are engaged in actions to enforce repurchases on faulty mortgages acquired through Bear Stearns and AIG. (August 4, 2010 Bloomberg article)

Litigation Background: The Real Issue—Access to the Loan Files

At first glance, many of the lawsuits sound like a Hail Mary by investors that have lost money on soured RMBS purchases. Our skepticism increases substantially when you consider that the claims of “faulty” mortgages are being made by entities such as the GSEs, FHLBs or mortgage insurers that have deep access to mortgage data and are deemed experts. However, a closer look at the FHLB lawsuits provide fairly convincing evidence that the loans were significantly worse than stated and the cases could have merit. Recall, as stated above, one of the primary goals of the lawsuit is to gain access to the loan files, as they will likely provide more convincing proof of their claims. Thus, the initial lawsuit only needs to provide enough evidence to convince the judge to deny motions to dismiss and enter the discovery phase which will potentially provide the plaintiffs access to the loan files.

Litigation Background: Do the Lawsuits Stand a Chance?



Accordingly, below are two examples that were cited in the San Francisco FHLB’s lawsuit of underwriting misrepresentations allegedly made in connection with the sale of Adjustable Rate Mortgage-Backed Pass-Through Certificates, Series 2007-1. “Untrue or misleading statements about the LTVs of the mortgage loans.” Utilizing an Automated Valuation Model (AVM), the FHLB estimated the actual average loan-to-values for underlying mortgages and compared them to statements made in the prospectus. Their analysis of 2,578 loans (58% of the entire pool), found that 414 loans, or 16%, had LTVs in excess of 100%, versus the statement in the prospectus that zero loans had LTVs in excess of 100%. Below is the results of their analysis taken from the lawsuit:

Source: Schedule 1 to First Amended Complaint, FHLB San Francisco v. Credit Suisse Securities (USA) LLC, et al. (emphasis added)

“Untrue or misleading statements about owner-occupancy of the properties that secured the mortgage loans” Based on their analysis, the FHLB estimated that among the 4,345 loans in this securitization, misstatements were made regarding 521 loans. Below is the info included in the lawsuit:

Source: Schedule 1 to First Amended Complaint, FHLB San Francisco v. Credit Suisse Securities (USA) LLC, et al.

In summary, the lawsuit claims that the defendants made untrue or misleading statements on 50.6% of the loans securitized in Adjustable Rate Mortgage-Backed Pass-Through Certificates, Series 2007-1 (p. 3, First Amended Complaint, FHLB San Francisco v. Credit Suisse Securities (USA) LLC, et al.) And, that is just one of the 116 securitizations that the San Francisco FHLB alleges were misrepresented. Where do the FHLB lawsuits stand? None of them have entered discovery. The Pittsburgh cases were moved from state court to federal court, then back to state court and are awaiting a ruling regarding the defendants’ motions to dismiss. The Seattle and San Francisco suits have been moved to federal court, but the FHLB has pending motions to remand those proceedings to state court. While the FHLB lawsuits are in limbo, the lawsuit filed by MBIA has had more progress that could have negative implications for the defendants of the other suits. In April 2010, Judge Bransten partially denied Bank of America’s motion to dismiss, and held that BAC is the successor-in-interest to Countrywide and thus vicariously liable for the conduct of

Item 62. Details of the results of the AVM analysis:

Number of loans 4,345

Number of properties on which there was enough information for the

model to determine a true market value 2,578

Number of loans on which the stated value was 105% or more of the

true market value as reported by the model 1,741

Aggregate amount by which the stated value of those properties

exceeded their true market values as reported by the model $159,299,961

Number of loans on which the stated value was 95% or less of the

trust market value as reported by the model 289

Aggregate amount by which the true market values of those

properties exceed their stated values $18,366,289

Number of loans with LTVs over 100% as stated by Defendants ‐

Number of loans with LTVs over 100% , as determined by the model 414

Weighted‐average LTV, as staed by Defendants (group 3) 72.2%

Weighted‐average LTV, as determined by the model (group 3) 86.6%

Items 96. Details of properties that were stated to be owner‐occupied, but were not:

(a) Number of loans on which the owner of the property instructed tax authorities to

send the property tax billed to him or her at a different address: 243

(b) Number of loans on which the owner of the property could have, but did not,

designate the property as his or her homestead: 325

(c) Number of loans on which the owner of the property owned three or more

properties: 30

(d) Eliminating duplicates, number of loans about which one or more of statements (a)

through (c) is true: 521



Countrywide if Countrywide is ultimately found liable (p. 15, April 29, 2010 Order of Judge Bransten, MBIA Insurance Corp. v. Coutnrywide Home Loans, Inc., et al.). The case was ordered to move forward on the fraud and breach of implied covenant of good faith and fair dealing causes of action. Since the Judge’s decision in April, both Bank of America and the FHLB have appealed the ruling. The same Judge is also sitting for the Syncora and FGIC lawsuits which are similar to the MBIA case. Importantly, in Syncora’s case against Countrywide, in May of this year Judge Bransten ordered Countrywide to produce to Syncora the loan origination files for all of the loans in three separate securitizations originated by Countrywide and insured by Syncora (May 7, 2010 Order of Judge Bransten, Syncora Guarantee Inc. v. Countrywide Home Loans, Inc., et al.). This ruling may set a precedent for the MBIA and FGIC lawsuits should Countrywide and BAC resist producing the loan origination files in those cases. While these lawsuits could be extremely slow to progress, we believe the FHFA subpoenas, Fed requests, and the actions being taken on behalf of the investor syndicate may proceed at a faster pace, given they are likely to gain access to the coveted loan files much sooner. With access to loan files potentially a matter of when, not if, the next question we consider is whether access to loan files will really be the smoking gun many expect. To gain some perspective on how pervasive the problem of defective mortgages was, we refer investors to the April 7, 2010 testimony of Richard Bowen, III, before the Financial Crisis Inquiry Commission. Mr. Bowen was the Business Chief Underwriter for Correspondent Lending in the Consumer Lending Group at Citigroup in charge of over $90B in residential mortgage production. Below are excerpts of his testimony: “In mid-2006, I discovered that over 60% of these mortgages purchased and sold were defective. Because Citi had given reps and warrants to the investors that the mortgages were not defective, the investors could force Citi to repurchase many billions of dollars of these defective assets. This situation represented a large potential risk to the shareholders of Citigroup. I started issuing warnings in June of 2006 and attempted to get management to address these critical risk issues. These warnings continued through 2007 and went to all levels of the Consumer Lending Group. We continued to purchase and sell to investors even larger volumes of mortgages through 2007. And defective mortgages increased during 2007 to over 80% of production.” Source: http://subprimeshakeout.blogspot.com/2010/06/sec-demands-more-disclosure-from-jp.html We defer investors to legal experts to opine on the potential outcomes of the outstanding lawsuits; however, given the potential evidence that the loan files could uncover, it would not be surprising to us to see settlements develop once data from the loan files access has been attained.

Who is Exposed to Alt-A Underwriting Risk?

With the majority of the top Alt-A and subprime mortgage originators out of business, the litigation has largely been centered on the underwriters of the securitizations. Should investor suits ultimately be successful in recovering damages from the underwriters, we would expect the underwriters to turn to the originators of the loans (so long as they are not affiliated with the underwriter or bankrupt) and attempt to recover those damages. Since this process is likely to take some time and we have quantifiable data points with regard to underwriter exposure, we have focused this report only on framing the potential liability of Alt-A and subprime RMBS underwriters. We believe that there is a material risk related to the past underwriting of Alt-A loans in the banking sector due to representation and warranties underwriters made to the buyers of Alt-A RMBS. Based on data compiled from Inside MBS & ABS, our analysis of the FHLBs suits, and actual performance data of the ‘05 to ‘07 Alt-A RMBS vintages, we estimate that the total liability for rescission requests on Alt-A RMBS to be $67.9 billion. Our worst and best case estimates for industry wide losses is $99.1 billion and $13.4 billion, respectively. JP Morgan (JPM—NR) tops the list with $13.1 billion of estimated losses largely due to the company’s acquisition of Bear Stearns, who topped the underwriting league tables with $132.9 billion of Alt-A RMBS underwritten during that time (according to Inside MBS & ABS). Deutsche Bank sits at the number two spot with $10.3 billion of estimated losses and Bank of America comes in third with $10.2 billion of estimated losses largely due to their acquisition of Countrywide, which underwrote $85.4 billion of Alt-A RMBS, or 86% of Bank of America’s total exposure, during the time period (according to Inside MBS & ABS). See the following table for complete details on company specific exposure.



Using data from Inside Mortgage Finance, we start with the league tables recording the top lead underwriters of Alt-A RMBS from 2005 through 2007. Since the majority of the rescission requests in the FHLBs suits were focused on loans underwritten in the years 2005 through 2007, we confined our initial data set to Alt-A RMBS underwritten and issued during those years. Ultimate losses will be dependent on three main factors; rescission percentage, default rate, and severity of loss on repurchased loans. Since these factors will vary based on vintage (or year underwritten), we use average statistics by vintage to estimate the liability. While these factors may also vary by issuer, we have not been able to identify any meaningful public statistic that correlates to the FHLBs suits rescission request percentage. Therefore, while we acknowledge there may be slight rescission rate differences between issuers, we believe using a vintage average is a suitable data point for framing the analysis. Worst Case Alt-A Loss Estimate In the worst case scenario, we assume that the rescission requests identified in the FHLB suits are indicative of the total potential pool of loans that could be rescinded industry-wide. While we cannot opine on whether or not the suit’s rescission percentage will ultimately be proven accurate, we believe that the data set forth in each particular suit is substantial enough to establish a worst case scenario. We then apply a success ratio, assuming that not all rescission requests will be honored or result in a loss. Finally, we apply a loss severity estimate to produce a net loss for loans repurchased. The mathematical equation used to estimate worst case losses is set forth below:

(weighted average rescission request by year) x (success ratio) x (severity of loss) = loss estimate

Methodology for Quantifying Risk

Source: Compass Point Research & Trading LLC, Bloomberg, Inside MBS & ABS

Worst Case Alt‐A Net Repurchase Loss Estimates'05 ‐ '07 % of orig. 2007 2006 2005

Bear Stearns 21,080 15.9% 6,686 8,965 5,429

Lehman Brothers 20,264 16.6% 8,143 7,545 4,576

Deutsche Bank 16,763 16.9% 7,268 5,941 3,553

Countrywide Securities 13,300 15.6% 3,798 5,852 3,650

Bank of America 3,085 22.1% 2,407 678 0

Total Bank of America 16,386 16.5% 6,205 6,530 3,650

RBS Greenwich Capital 15,282 15.5% 4,415 6,485 4,382

Goldman Sachs 9,625 16.9% 3,361 4,821 1,444

UBS 8,989 15.8% 3,052 3,467 2,469

Credit Suisse 6,801 21.1% 4,629 2,172 0

Citigroup 4,164 22.5% 3,442 722 0

Total 119,354 47,202 46,648 25,504



Alt‐A Worst Case Scenario Assumptions2007 2006 2005

FHLB Rescission Rate 54.5% 49.1% 43.2%

Success Ratio 75.0% 60.0% 50.0%

Severity of Loss 60.0% 55.0% 50.0%

Alt‐A RMBS Repurchase Request Loss Estimates


JP Morgan JPM NR 21,080 $3.16 12% 13,110 $1.96 7% 2,718 $0.41 2%





UBS UBS.N NR 8,989 $1.42 16% 5,472 $0.87 10% 1,148 $0.18 2%


Citigroup C NR 4,164 $0.09 2% 2,527 $0.05 1% 683 $0.01 0%

Total 99,090 67,920 13,371





Base Case Alt-A Loss Estimate In the base case scenario, we assume that rescission requests are limited to all seriously delinquent and defaulted loans that have occurred up to and including July 2010. We then apply a success ratio, assuming that not all rescission requests will be honored or result in a loss. Finally, we apply a loss severity estimate to produce a net loss for loans repurchased. The mathematical equation used to estimate worst case losses is set forth below: (total 60+ day delinquent loan balance & cumulative gross defaults through July 2010) x (success ratio) x (severity) = loss estimate

As a point of reference, First Horizon (FHN—NR) noted in the company’s latest 10-Q filing that they have witnessed average rescission rates of between 40% and 50% of the repurchase and make-whole requests (similar to our “success ratio”) and observed loss severities (measured as a percentage of the unpaid principal balance) ranging between 50% and 55% of the repurchased loans. This would result in an approximate loss severity of between 20% and 28%. The majority of FHN’s loan repurchase requests made to date have occurred on prime loans, which should bear a lower ultimate severity than Alt-A loans. We believe this benchmark compares favorably to our base case scenario for Alt-A loan repurchase risk. Best Case Alt-A Loss Estimate In the best case scenario, we assume that rescission requests are limited to all seriously delinquent and defaulted loans that occurred up to eighteen months after issuance. We then apply a success ratio, assuming that not all rescission requests will be honored or result in a loss. Finally, we apply a loss severity estimate to produce a net loss for loans repurchased. The mathematical equation used to estimate worst case losses is set forth below:

(total 60+ day delinquent loan balance & cumulative gross defaults @ 18 months after issuance) x (success ratio) x (severity) = loss estimate

Base Case Alt‐A Net Repurchase Loss Estimates'05 ‐ '07 % of orig. 2007 2006 2005

Bear Stearns 13,110 9.9% 3,765 7,303 2,042

Lehman Brothers 12,453 10.2% 4,586 6,146 1,721

Deutsche Bank 10,269 10.4% 4,093 4,840 1,336

Countrywide Securities 8,279 9.7% 2,139 4,767 1,373

Bank of America 1,908 13.6% 1,356 552 0


RBS Greenwich Capital 9,417 9.5% 2,486 5,283 1,648

Goldman Sachs 6,363 11.2% 1,893 3,927 543

UBS 5,472 9.6% 1,719 2,825 928

Credit Suisse 4,376 13.6% 2,607 1,769 0

Citigroup 2,527 13.7% 1,938 588 0

Total 74,174 26,583 38,001 9,590

Best Case Alt‐A Net Repurchase Loss Estimates'05 ‐ '07 % of orig. 2007 2006 2005

Bear Stearns 2,718 2.0% 1,120 1,319 279

Lehman Brothers 2,709 2.2% 1,364 1,110 235

Deutsche Bank 2,274 2.3% 1,217 874 183

Countrywide Securities 1,685 2.0% 636 861 188

Bank of America 503 3.6% 403 100 0

Total Bank of America 2,188 2.2% 1,039 961 188

RBS Greenwich Capital 1,919 1.9% 739 954 225

Goldman Sachs 1,346 2.4% 563 709 74

UBS 1,148 2.0% 511 510 127

Credit Suisse 1,095 3.4% 775 319 0

Citigroup 683 3.7% 576 106 0

Total 16,080 7,905 6,863 1,312





Alt‐A Base Case Estimate Assumptions2007 2006 2005

Balance 71.6% 52.1% 38.0%

Net Losses 3.8% 5.2% 1.0%

Severity 60.0% 55.0% 45.0%

Gross Losses 6.3% 9.4% 2.2%

REO 2.0% 2.4% 1.2%

Foreclosure 9.8% 13.6% 6.7%

Bankrupt 2.2% 3.0% 1.8%

Delinquent Loans 17.3% 20.6% 11.2%

Gross SDQ 37.7% 48.9% 23.1%

Success Ratio 80.0% 80.0% 80.0%

Alt‐A Best Case Estimate Assumptions2007 2006 2005

Balance 88.1% 79.4% 71.6%

Net Losses 0.3% 0.1% 0.0%

Severity 60.0% 55.0% 50.0%


REO 1.5% 1.4% 0.3%


Bankrupt 0.6% 0.5% 0.2%


Gross SDQ 6.9% 4.0% 0.9%

Success Ratio 60.0% 60.0% 60.0%



We believe that there is material risk related to the past underwriting of subprime loans in the banking sector due to the representation and warranties underwriters made to the buyers of subprime RMBS. While we have yet to see a lawsuit, we believe the consortium of investors represented by the law firm Talcott Franklin P.C. intends to pursue a strategy that ultimately results in the rescission of loans that they believe breach the underwriters representation and warranties. Should investors be successful in recovering damages from the underwriters, we would expect the underwriters to turn to the originators of the loans (so long as they are not affiliated with the underwriter or bankrupt) and attempt to recover those damages. Since this process is likely to take some time and we now have quantifiable data points with regard to underwriter exposure, we have focused this report only on framing the potential liability for underwriters and not originators.

Based on data compiled from Asset Backed Alert, our analysis of the FHLBs suits, and actual performance data of the ‘05 to ‘07 subprime RMBS vintages, we estimate that the total liability for rescission requests on subprime RMBS to be $80.3 billion. Our worst and best case estimates for industry wide losses is $89.3 billion and $46.6 billion, respectively. Bank of America (BAC—NR) tops the list with $25.0 billion of estimated losses largely due to their acquisition of Countrywide and Merrill Lynch, who underwrote $86.0 billion and $45.7 billion of subprime RMBS, respectively, during the time period. JP Morgan (JPM—NR) sits at the number two spot with estimated losses of $10.8 billion based on subprime underwriting exposure of $60.2 billion based in part on the company’s acquisition of Bear Stearns, who underwrote $37.4 billion of subprime RMBS during that time. See the at the top of the following page for complete details on company specific loss exposure.

Who is Exposed to Subprime Underwriting Risk?

Subprime Issuance by Year ($Mil.)Rank* Issuer Total '05‐'07 Mkt Share 2007 2006 2005 2004 2003 2002 2001 2000

1 Countrywide 85,993 15.8% 19,509 26,345 40,140 42,650 9,671 4,591 3,381 1,631

2 Lehman Brothers 49,597 9.1% 18,652 17,635 13,310 13,773 8,774 10,213 10,702 8,942

3 RBS Greenwich 47,721 8.8% 19,520 11,207 16,993 21,461 10,634 8,211 8,408 4,361

4 Merri l l Lynch 45,667 8.4% 21,936 12,019 11,712 7,318 2,899 200 649 176

5 Morgan Stanley 37,572 6.9% 23,656 6,373 7,543 8,523 6,433 6,393 1,634 1,343

6 Bear Stearns 37,382 6.9% 13,360 11,169 12,854 13,095 10,783 9,336 6,748 10,097

7 Credit Suisse 31,436 5.8% 7,161 9,732 14,543 11,930 3,727 7,121 9,573 2,122

8 Goldman Sachs 31,274 5.8% 6,802 13,166 11,307 9,506 2,538 4,314 0 346

9 Citigroup 28,588 5.3% 14,026 5,888 8,674 4,368 12,077 0 0 0

10 Bank of America 24,487 4.5% 10,179 3,956 10,352 14,128 6,368 4,508 4,792 2,417

11 J.P. Morgan 22,833 4.2% 11,360 7,001 4,472 8,453 13,690 3,717 5,773 0

12 Deutsche Bank 20,066 3.7% 10,169 4,313 5,584 9,681 7,785 5,567 3,120 0

13 UBS 18,068 3.3% 5,366 5,830 6,873 5,050 3,580 3,038 0 237

14 Barclays 17,723 3.3% 9,578 4,738 3,406 1,717 0 0 0 0

15 HSBC 16,890 3.1% 6,708 9,678 504 0 0 0 0 0

16 WaMu Capital 11,284 2.1% 3,488 2,142 5,655 3,903 0 0 0 0

17 GMAC RFC 5,402 1.0% 987 2,335 2,080 497 242 0 0 0

18 Friedman Bill ings Ramsey 4,002 0.7% 0 324 3,678 660 0 0 0 0

19 Terwin Capital 3,375 0.6% 166 2,307 902 1,082 96 0 0 0

20 Wachovia 2,225 0.4% 1,062 648 515 0 0 1,651 451 0

21 Societe Generale 991 0.2% 177 814 0 0 0 0 0 0

22 RBC Capital 899 0.2% 386 513 0 0 0 246 0 0

23 BMO Capital 196 0.0% 106 90 0 0 0 0 0 0

24 SunTrust 185 0.0% 185 0 0 0 0 0 0 0

25 Banc One Capital 0 0.0% 0 0 0 0 892 100 0 0

Total 543,855 204,540 158,222 181,093 177,795 100,190 69,205 55,229 31,673

Source: Compass Point Research & Trading LLC, Bloomberg, Asset Backed Alert



Using data from Asset Backed Alert, we start with the league tables recording the top lead underwriters of subprime RMBS from 2005 through 2007. Since the majority of the rescission requests in the FHLB suits were focused on loans underwritten in the years 2005 through 2007, we confined our initial data set to subprime RMBS underwritten and issued during those years. Ultimate losses will be dependent on three main factors; rescission percentage, default rate, and severity of loss on repurchased loans. Since these factors will vary based on vintage (or year underwritten), we use average statistics by vintage to estimate the liability. While these factors may also vary by issuer, we have not been able to identify any meaningful public statistic that correlates to the FHLB suits rescission request percentage. Therefore, while we acknowledge there may be slight rescission rate differences between issuers, we believe using a vintage average is a suitable data point for framing the analysis. Worst Case Subprime Loss Estimate In the worst case scenario, we assume that the rescission requests identified in the FHLB suits are a good proxy for the total potential pool of loans that could be rescinded industry-wide. While we cannot opine on whether or not the suit’s rescission percentage will ultimately be proven accurate, we believe that the data set forth in each particular suit is substantial enough to establish a worst case scenario. We then apply a success ratio, assuming that not all rescission requests will be honored or result in a loss. Finally, we apply a loss severity estimate to produce a net loss for loans repurchased. The mathematical equation used to estimate worst case losses is set forth below:

(weighted average rescission request by year) x (success ratio) x (severity of loss) = loss estimate

Methodology for Quantifying Risk


Subprime Worst Case Scenario Assumptions

2007 2006 2005

FHLB Rescission Rate 54.5% 49.1% 43.2%

Success Ratio 80.0% 80.0% 80.0%

Severity of Loss 65.0% 55.0% 50.0%

Worst Case Subprime Net Repurchase Loss Estimates'05 ‐ '07 % of orig. 2007 2006 2005

Countrywide 14,609 17.0% 5,188 4,657 4,763

Merrill Lynch 9,348 20.5% 5,834 2,125 1,390

Bank of America 4,635 18.9% 2,707 699 1,228


Bear Stearns 7,052 18.9% 3,553 1,974 1,525

J.P. Morgan 4,789 21.0% 3,021 1,238 531

Total J.P. Morgan 11,842 19.7% 6,574 3,212 2,056

RBS Greenwich 9,189 19.3% 5,191 1,981 2,017

Morgan Stanley 8,312 22.1% 6,291 1,127 895

Credit Suisse 5,350 17.0% 1,904 1,721 1,726

Goldman Sachs 5,478 17.5% 1,809 2,327 1,342

Citigroup 5,800 20.3% 3,730 1,041 1,029

Deutsche Bank 4,129 20.6% 2,704 763 663

UBS 3,273 18.1% 1,427 1,031 816

Barclays 3,789 21.4% 2,547 838 404

HSBC 3,555 21.0% 1,784 1,711 60

Total 89,309 47,689 23,232 18,388

Subprime RMBS Repurchase Request Loss Estimates



JP Morgan JPM NR 11,842 $1.77 7% 10,831 $1.62 6% 6,288 $0.94 4%


Morgan Stanley MS NR 8,312 $3.56 15% 7,855 $3.37 14% 4,498 $1.93 8%

Citigroup CS NR 5,800 $0.12 3% 5,292 $0.11 3% 3,047 $0.06 2%




Barclays BCS NR 3,789 $0.19 4% 3,583 $0.18 3% 2,068 $0.10 2%

HSBC HBC NR 3,555 $0.12 2% 3,515 $0.12 2% 2,071 $0.07 1%

UBS UBS.N NR 3,273 $0.52 6% 2,878 $0.46 5% 1,681 $0.27 3%

Total 89,309 80,329 46,626







Base Case Subprime Loss Estimate In the base case scenario, we assume that rescission requests are limited to all seriously delinquent and defaulted loans that have occurred up to and including July 2010. We then apply a success ratio, assuming that not all rescission requests will be honored or result in a loss. Finally, we apply a loss severity estimate to produce a net loss for loans repurchased. The mathematical equation used to estimate worst case losses is set forth below: (total 60+ day delinquent loan balance & cumulative gross defaults through July 2010) x (success ratio) x (severity) = loss estimate

Best Case Subprime Loss Estimate In the best case scenario, we assume that rescission requests are limited to all seriously delinquent and defaulted loans that occurred up to eighteen months after issuance. We then apply a success ratio, assuming that not all rescission requests will be honored or result in a loss. Finally, we apply a loss severity estimate to produce a net loss for loans repurchased. The mathematical equation used to estimate worst case losses is set forth below:

(total 60+ day delinquent loan balance & cumulative gross defaults @ 18 months after issuance) x (success ratio) x (severity) = loss estimate





Subprime Base Case Estimate Assumptions

2007 2006 2005

Balance 60.2% 29.2% 16.5%

Net Losses 19.0% 16.3% 5.6%

Severity 65.0% 60.0% 55.0%

Gross Losses 29.3% 27.1% 10.1%

REO 4.1% 4.4% 3.3%

Foreclosure 16.4% 15.9% 11.5%

Bankrupt 3.1% 3.6% 4.0%


Gross SDQ 65.2% 60.3% 35.0%

Success Ratio 80.0% 80.0% 80.0%

Subprime Best Case Estimate Assumptions2007 2006 2005

Balance 82.1% 78.7% 55.5%

Net Losses 4.3% 2.0% 0.4%

Severity 60.0% 40.0% 40.0%


REO 6.0% 5.4% 2.1%


Bankrupt 1.8% 1.7% 1.4%


Gross SDQ 6.9% 4.0% 0.9%

Success Ratio 60.0% 60.0% 60.0%

Base Case Subprime Net Repurchase Loss Estimates'05 ‐ '07 % of orig. 2007 2006 2005

Countrywide 12,321 14.3% 5,161 4,653 2,508

Merrill Lynch 8,657 19.0% 5,803 2,123 732

Bank of America 4,038 16.5% 2,693 699 647


Bear Stearns 6,310 16.9% 3,534 1,973 803

J.P. Morgan 4,521 19.8% 3,005 1,236 279

Total J.P. Morgan 10,831 18.0% 6,539 3,209 1,082

RBS Greenwich 8,205 17.2% 5,164 1,979 1,062

Morgan Stanley 7,855 20.9% 6,258 1,126 471

Credit Suisse 4,522 14.4% 1,894 1,719 908

Goldman Sachs 4,831 15.4% 1,799 2,325 706

Citigroup 5,292 18.5% 3,710 1,040 542

Deutsche Bank 3,801 18.9% 2,690 762 349

UBS 2,878 15.9% 1,419 1,030 429

Barclays 3,583 20.2% 2,534 837 213

HSBC 3,515 20.8% 1,775 1,709 31

Total 80,329 47,440 23,210 9,680

Best Case Subprime Net Repurchase Loss Estimates

'05 ‐ '07 % of orig. 2007 2006 2005

Countrywide 7,215 8.4% 2,916 2,859 1,440

Merrill Lynch 5,004 11.0% 3,279 1,304 420

Bank of America 2,322 9.5% 1,522 429 371


Bear Stearns 3,670 9.8% 1,997 1,212 461

J.P. Morgan 2,618 11.5% 1,698 760 160

Total J.P. Morgan 6,288 10.4% 3,695 1,972 621

RBS Greenwich 4,744 9.9% 2,918 1,216 609

Morgan Stanley 4,498 12.0% 3,536 692 271

Credit Suisse 2,648 8.4% 1,070 1,056 522

Goldman Sachs 2,851 9.1% 1,017 1,429 405

Citigroup 3,047 10.7% 2,097 639 311

Deutsche Bank 2,188 10.9% 1,520 468 200

UBS 1,681 9.3% 802 633 246

Barclays 2,068 11.7% 1,432 514 122

HSBC 2,071 12.3% 1,003 1,050 18

Total 46,626 26,808 14,260 5,557



What Reserves have been Recorded?

Based upon our review of quarterly filings, JPM appears to be the only underwriter that has potentially reserved for repurchases as it relates to private label litigation. In 1Q10, JPM recorded a $2.3B charge in litigation reserves for “mortgage-related” matters. When asked a question on their earnings call regarding the charge, management responded “to think about that as we have repurchase reserves that we’ve talked about related to the GSEs as an ongoing expense we’ve been reserving for. This (charge) relates to the broader question of all other ideas for claims against us from private investors”. A review of the litigation section of JPM’s 2009 10-K and their 1Q10 10-Q shows that the only change is the mention of the FHLB San Francisco lawsuit (the Seattle and Pittsburgh lawsuits were mentioned in the 10-K). Interestingly, the charge was also recorded in the quarter immediately following a request from the SEC for more information regarding their repurchase reserves. Two weeks following the release of their 4Q09 earnings, JPM received a letter on January 29, 2010 from the SEC requesting disclosures on how the company establishes repurchase reserves for various reps and warranties, including GSE’s, monoline insurers and any private loan repurchase requests (http://www.sec.gov—JPM March 2, 2010 Correspondence). Our review of quarterly filings found that BAC had a $3.9B reserve for all mortgage repurchase requests (on $11.1B in requests made), JPM had a $2.3B reserve for mortgage repurchases (which is separate from their $2.3B litigation reserve charge in 1Q10), and Citigroup had a $727MM reserve for mortgage repurchases. Importantly, BAC’s 2Q10 quarterly filing noted that they have only received $33MM in private label MBS repurchase requests thus far. Below is a table of the applicable reserves.

Source: Company filings, Compass Point Research

Unpaid principal bal. ‐ in millions BAC JPM C

Unresolved mortgage repurchase requests 11,100 2,880 4,478

GSEs 5,600 1,400 4,166

Monolines 4,000 1,700 98

Other investors 1,400 na 214

Private label MBS investors 33 na na

Reserve for repurchases 3,900 2,332 727

Litigation reserve (estimate) na 2,300 na

Subtotal 3,900 4,632 727


Compass Point Research & Trading, LLC 12

Important Disclosures Analyst Certification I, Chris Gamaitoni, hereby certify that the views expressed in this research report accurately reflect our personal views about the subject securities or issues. We further certify that we have not received direct or indirect compensation in exchange for expressing specific recommendations or views in this report.

Ownership and Material Conflicts of Interest As of the end of the month immediately preceding the date of publication of this research report (or of the second most recent month if the publication date is less than 10 calendar days after the end of the most recent month), neither Compass Point Research & Trading, LLC, nor any of its affiliates own any of the subject company(ies)'s equity securities. The research analyst named in the certification above holds a financial interest in the common stock of Citigroup (NYSE: C), which is the subject of this report. There are no material conflicts of interest of Compass Point Research & Trading, LLC or of the research analyst named in the certification above of which the research analyst knows or has reason to know at the time of publication of this report. Neither the research analyst named in the certification above, any member of that analysts' household, nor any person that depends upon him for financial support, is an officer, director or advisory board member of the subject company(ies) mentioned in the research report. The research analyst named in the certification above does not receive any compensation from Compass Point Research & Trading, LLC that is in any way related to Compass Point Research & Trading, LLC's investment banking revenues. Compass Point Research & Trading, LLC does not compensate its research analysts for investment banking services, but rather provides research analysts with a salary and bonus based upon the research analyst’s individual performance and quality of research, the correlation between the analyst’s recommendations and the stock price performance, and overall ratings received from clients, sales employees, and other employees independent of Compass Point Research & Trading, LLC’s investment banking department. The research analyst named in the certification above has not received any compensation from any company that is the subject of this research report. Compass Point Research & Trading, LLC has never managed or co-managed a public offering of securities for any company that is the subject of this research report and has never had any investment banking relationship with any company that is the subject of this research report, and therefore has not received any compensation for investment banking services from any such companies in the past 12 months of publication of this report and does not expect or intend to receive compensation for investment banking services from the subject companies with the next three months from the publication of this report. Compass Point Research & Trading, LLC has received no compensation from any company that is the subject of this research report for any products or services rendered to such companies, and neither the research analyst named in the certification above nor any Compass Point employee with ability to influence the substance of this research report has any knowledge of such compensation to Compass Point or any affiliate. No (none) of the company(ies) that are the subject of this research report have ever been clients of Compass Point Research & Trading, LLC. Compass Point Research & Trading, LLC has never, and as of the publication of this research report does not, act as a market maker in the securities of any of the companies that are the subject of this report.

Compass Point Research & Trading, LLC 13

Important Disclosures, cont’d Global Disclaimer This report is based upon public information that Compass Point Research & Trading, LLC and the research analyst named in the attestation above assume to be correct. Assumptions, opinions, forecasts, and estimates constitute the research analyst’s judgment as of the date of this material and are subject to change without notice. The research analyst’s judgments may be wrong. Neither Compass Point Research & Trading, LLC nor its affiliates, nor the research analyst, are responsible for any errors, omissions, or results obtained from the use of this information. Past performance is not necessarily indicative of future results. The securities and/or financial instruments mentioned in this research report, and the trading strategies related thereto, may not be suitable for all investors. You must consider your specific investment goals and objectives prior to transacting in any security or financial instrument. Consult with your financial advisor before making any transactions or investments. This research report is not intended as an offer or solicitation for the purchase or sale of any security or other financial instrument.

Date post:	04-Jul-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times