1 MN20211: Corporate Finance 2011/12: 1.Revision: Investment Appraisal (NPV). 2.Investment...

transcript

MN20211: Corporate Finance 2011/12:

1. Revision: Investment Appraisal (NPV).

2. Investment flexibility, Decision trees, Real Options.

3. Revision: Portfolio Theory => CAPM

4. Sources of Funding.

5. Capital Structure and Value of the Firm.

6. Optimal Capital Structure - Agency Costs, Signalling.

7. Dividend policy/repurchases.

8. Mergers and Acquisitions.

9. Venture Capital and Private Equity.

10. Introduction to Behavioural Finance.

11. Revision.

Corporate Finance: Three Major Decisions…

• Investment Appraisal (Capital Budgeting) – Which New Projects to invest in?

• Capital Structure (Financing Decision)- How to Finance the new projects – Debt or equity?

• Payout Policy – Dividends, Share Repurchases, Re-investment.

• => Objective: Maximisation of Shareholder Wealth.

First Topic: Investment Appraisal

• Brief revision of static NPV.

• => Flexibility

• => Decision trees

• => sensitivity analysis

• => Real Options

Investment Appraisal.

• Objective: Take projects that increase shareholder wealth (Value-adding projects).

• Investment Appraisal Techniques: NPV, IRR, Payback, ARR, Real Options….

• Which one is the Best rule for shareholder wealth maximisation?

Connections in Corporate Finance.

Investment Appraisal: Net Present Value with discount rate (cost of capital) given. Positive NPV increases value of the firm.

Cost of Capital (discount rate): How do companies derive the cost of capital? – CAPM/APT.

Capital Structure and effect on Firm Value and WACC.

• Debate over Correct Method• - Accounting Rate of Return.• - Payback.• - NPV.• - IRR.• - POSITIVE NPV Increases Shareholder

Wealth.• 2. Correct Method - NPV!• -Time Value of Money• - Discounts all future cashflows

Net Present Value

• .....)1()1(1 3

XINPV Perpetuities.

IRR =>

Take Project if NPV > 0, or if IRR > r.

.0.....)1()1(1 3

Example.Consider the following new project: -initial capital investment of £15m.-it will generate sales for 5 years.- Variable Costs equal 70% of sales value.- fixed cost of project £200k PA.- A feasibility study, cost £5000, has already been carried out.Discount Rate equals 12%.Should we take the project?

$000 2012 2013 2014 2015 2016 2017 SALES 14000 16000 18000 20000 22000 90000

VARIABLE COSTS -9800 -11200 -12600 -14000 -15400 -63000

OPERATING EXPENSES -200 -200 -200 -200 -200 -1000

EQUIPMENT COSTS -15000 -15000

CASHFLOWS -15000 4000 4600 5200 5800 6400 11000

DF @ 12% 1.00 0.893 0.797 0.712 0.636 0.567

NPV -15000 3571 3667 3701 3686 3632 3257

19.75 1.00 0.84 0.70 0.58 0.49 0.41

IRR = 19.75% -15000 3340 3208 3028 2820 2599 0

DO WE INVEST IN THIS NEW PROJECT?

NPV > 0.

COST OF CAPITAL (12%) < IRR (19.75%).

Note that if the NPV is positive, then the IRR exceeds the Cost of Capital.

NPV £m

Discount Rate %

19.7%0

CONFLICT BETWEEN APPRAISAL TECHNIQUES.YEAR A B C D DF: 10%

0 -1000 -1000 -1000 -1000 11 100 0 100 200 0.9092 900 0 200 300 0.8263 100 300 300 500 0.7514 -100 700 400 500 0.6835 -400 1300 1250 600 0.621

PAYBACK METHOD:

PROJECT A: 2 YEARS SELECT PROJECT APROJECT B: 4 YEARSPROJECT C: 4 YEARSPROJECT D: 3 YEARS

PROJECT A: -407PROJECT B: 511PROJECT C: 531 SELECT PROJECT CPROJECT D: 519

PROJECT A: -200%PROJECT B: 20.9%PROJECT C: 22.8%PROJECT D: 25.4% SELECT PROJECT D

Discount Rate10% 22.8%

PROJ C

PROJ D

COMPARING NPV AND IRR - 1

Select Project with higher NPV: Project C.

Discount Rate

COMPARING NPV AND IRR -2

Impossible to find IRR!!! NPV exists!

• COMPARING NPV AND IRR –3 Size Effect• Discount Rate: 10%• Project A : Date 0 Investment -£1000.• Date 1 Cashflow £1500.• NPV = £364.• IRR = 50%• Project B:- Date 0 Investment -£10• Date 1 Cashflow £18.• NPV = £6.36• IRR = 80%.• Which Project do we take?

• Mutually Exclusive project: firm can only take one (take project with highest positive NPV).Independent project: firm can take as many as it likes (take all positive NPV projects).Consider slide 10: Which project(s) would you take, and what would be the value-added, if projects are a) mutually exclusive, and b) independent?

Mutually Exclusive Versus Independent Projects.

Investment Flexibility/ Real options.

• Reminder of Corporation’s Objective : Take projects that increase shareholder wealth (Value-adding projects).

• Investment Appraisal Techniques: NPV, IRR, Payback, ARR

• Decision trees• Monte Carlo.• Real Options

Investment Flexibility, Decision Trees, and Real Options

Decision Trees and Sensitivity Analysis.•Example: From Ross, Westerfield and Jaffe: “Corporate Finance”.

•New Project: Test and Development Phase: Investment $100m.

•0.75 chance of success.

•If successful, Company can invest in full scale production, Investment $1500m.

•Production will occur over next 5 years with the following cashflows.

$000 Year 1 Year 2 - 6

Revenues 6000Variable Costs -3000Fixed Costs -1791Depreciation -300

Pretax Profit 909Tax (34%) -309

Net Profit 600Cashflow 900

Initial Investment -1500

Date 1 NPV = -1500 +

1 )15.1(

= 1517

Production Stage: Base Case

Decision Tree.

Do Not Test

Success

Failure

Invest

Do not Invest

Invest

NPV = 1517

NPV = 0

NPV = -3611

Date 0: -$100 Date 1: -1500

Solve backwards: If the tests are successful, SEC should invest, since 1517 > 0.

If tests are unsuccessful, SEC should not invest, since 0 > -3611.

P=0.75

P=0.25

Now move back to Stage 1.

Invest $100m now to get 75% chance of $1517m one year later?

Expected Payoff = 0.75 *1517 +0.25 *0 = 1138.

NPV of testing at date 0 = -100 + 15.1

1138 = $890

Therefore, the firm should test the project.

Sensitivity Analysis (What-if analysis or Bop analysis)

Examines sensitivity of NPV to changes in underlying assumptions (on revenue, costs and cashflows).

Sensitivity Analysis.

- NPV Calculation for all 3 possibilities of a single variable + expected forecast for all other variables.

NPV Expected Pessimistic or Best Optimistic

Market Size -1802 1517 8154Market Share -696 1517 5942Price 853 1517 2844Variable Cost 189 1517 2844Fixed Cost 1295 1517 1628Investment 1208 1517 1903

Limitation in just changing one variable at a time.

Scenario Analysis- Change several variables together.

Break - even analysis examines variability in forecasts.

It determines the number of sales required to break even.

Real Options

• A new investment appraisal method, analysed by academics in the 1980s

• Existing NPV method: static: one-off decision to take project or ‘throw it away’: once project is taken, committed to it.

• Real Options: recognises flexibility in decision-making.

Real Options.

A digression: Financial Options

A call option gives the holder the right (but not the obligation) to buy shares at some time in the future at an exercise price agreed now.

A put option gives the holder the right (but not the obligation) to sell shares at some time in the future at an exercise price agreed now.

European Option – Exercised only at maturity date.

American Option – Can be exercised at any time up to maturity.

For simplicity, we focus on European Options.

Example:

• Today, you buy a call option on Marks and Spencer’s shares. The call option gives you the right (but not the obligation) to buy MS shares at exercise date (say 31/12/10) at an exercise price given now (say £10).

• At 31/12/10: MS share price becomes £12. Buy at £10: immediately sell at £12: profit £2.

• Or: MS shares become £8 at 31/12/10: rip option up!

Factors Affecting Price of European Option (=c).

-Underlying Stock Price S.

-Exercise Price X.

-Variance of of the returns of the underlying asset ,

-Time to maturity, T.

.0,0,0,02

The riskier the underlying returns, the greater the probability that the stock price will exceed the exercise price.

The longer to maturity, the greater the probability that the stock price will exceed the exercise price.

Options: Payoff Profiles.

Buying a Call Option.

WSelling a put option.

Selling a Call Option. Buying a Put Option.

Pricing Call Options – Binomial Approach.

1- q dS=13.40

uS=24.00

S = £20. q=0.5. u=1.2. d=.67. X = £21.

1 + rf = 1.1.

Risk free hedge Portfolio: Buy One Share of Stock and write m call options.

uS - mCu = dS – mCd => 24 – 3m = 13.40.

M = 3.53.

By holding one share of stock, and selling 3.53 call options, your payoffs are the same in both states of nature (13.40): Risk free.

Cu = 3

Since hedge portfolio is riskless:

.))(1( uf mcuSmcSr

1.1 ( 20 – 3.53C) = 13.40.

Therefore, C = 2.21.

This is the current price per call option. The total present value of investment = £12 .19, and the rate of return on investment is

13.40 / 12.19 = 1.1.

Alternative option-pricing method

• Black-Scholes

• Continuous Distribution of share returns (not binomial)

• Continuous time (rather than discrete time).

Real Options

• Just as financial options give the investor the right (but not obligation) to future share investment (flexibility)

• Researchers recognised that investing in projects can be considered as ‘options’ (flexibility).

• “Real Options”: Option to delay, option to expand, option to abandon.

• Real options: dynamic approach (in contrast to static NPV).

Real Options

• Based on the insights, methods and valuation of financial options which give you the right to invest in shares at a later date

• RO: development of NPV to recognise corporation’s flexibility in investing in PROJECTS.

Real Options.

• Real Options recognise flexibility in investment appraisal decision.

• Standard NPV: static; “now or never”.

• Real Option Approach: “Now or Later”.

• -Option to delay, option to expand, option to abandon.

• Analogy with financial options.

Types of Real Option

• Option to Delay (Timing Option).

• Option to Expand (eg R and D).

• Option to Abandon.

Option to Delay (= call option)

Project value

Value-creation

Investment in waiting:

(sunk)

Option to expand (= call option)

Project value

Value creation

Investment in initial project: eg R and D (sunk)

Option to Abandon ( = put option)

Project value

Project goes badly: abandon for liquidation value.

Valuation of Real Options

• Binomial Pricing Model

• Black-Scholes formula

Value of a Real Option

• A Project’s Value-added = Standard NPV plus the Real Option Value.

• For given cashflows, standard NPV decreases with risk (why?).

• But Real Option Value increases with risk.

• R and D very risky: => Real Option element may be high.

Simplified Examples

• Option to Expand (page 241 of RWJ)

Build First Ice

If Successful

Expand

If unsuccessful

Do not Expand

• NPV of single ice hotel

• NPV = - 12,000,000 + 2,000,000/0.20 =-2m

• Reject?

• Optimistic forecast: NPV = - 12M + 3M/0.2

• = 3M.

• Pessimistic: NPV = -12M + 1M/0.2 = - 7m

• Still reject?

Option to Expand (Continued)

Option to expand (continued)

• Given success, the E will expand to 10 hotels

• =>

• NPV = 50% x 10 x 3m + 50% x (-7m) = 11.5 m.

• Therefore, invest.

Option to abandon.

• NPV(opt) = - 12m + 6m/0.2 = 18m.

• NPV (pess) = -12m – 2m/0.2 = -22m.

• => NPV = - 2m. Reject?

• But abandon if failure =>

• NPV = 50% x 18m + 50% x -12m/1.20

• = 2.17m

• Accept.

Option to delay and Competition (Smit and Ankum).

•-Smit and Ankum present a binomial real option model:

•Option to delay increases value (wait to observe market demand)

•But delay invites product market competition: reduces value (lost monopoly advantage).

•cost: Lost cash flows

•Trade-off: when to exercise real option (ie when to delay and when to invest in project).

•Protecting Economic Rent: Innovation, barriers to entry, product differentiation, patents.

•Firm needs too identify extent of competitive advantage.

Option to delay versus competition: Game-theoretic approach

Firm 1\Firm 2 Invest early Delay

Invest early NPV = 500,NPV = 500 NPV = 700, NPV = 300

Delay NPV = 300, NPV = 700 NPV = 600,NPV = 600

Option to delay versus competition: effects of legal system

Firm 1\ Firm 2 Invest early Delay

Invest early NPV = 500,NPV = 500 NPV = 700- 300, NPV = 300+300

Delay NPV = 300+300, NPV = 700-300

NPV = 600,NPV = 600

Monte Carlo methods

• BBQ grills example in RWJ.

• Application to Qinetiq (article by Tony Bishop).

Use of Real Options in Practice

• . •

SECTION 2: Risk and Return/Portfolio Decision/ Cost Of Capital.

The cost of capital = investors’ required return on their investment in a company. It provides the appropriate discount rate in NPV.

Investors are risk averse.

Future share prices (and returns) are risky (volatile).

The higher the risk, the higher the required return.

An investor’s actual return is the percentage change in price:

Risk = Variability or Volatility of Returns, Var (R).

We assume that Returns follow a Normal Distribution.

Var(R).

./)....()( 21 TRRRAverageRE T

• • Risk Aversion.

Investors prefer more certain returns to less certain returns.

Wealth

150100 200

Risk Averse Investor prefers £150 for sure than a 50/50 gamble giving £100 or £200.

• Portfolio Analysis.

Two Assets: Investor has proportion a of Asset X and (1-a) of Asset Y.

).()1()(.)( YXp REaREaRE

).,().1(2)(.)1()(.)( 22 yxCovaaYVaraXVaraRVar p

Combining the two assets in differing proportions.

• Portfolio of Many assets + Risk Free Asset.

M.Efficiency Frontier.

All rational investors have the same market portfolio M of risky assets, and combine it with the risk free asset.

A portfolio like X is inefficient, because diversification can give higher expected return for the same risk, or the same expected return for lower risk.

• The Effect of Diversification on Portfolio Variance.

Number of Assets.

An asset’s risk = Undiversifiable Risk + Diversifiable Risk

= Market Risk + Specific Risk.

Market portfolio consists of Undiversifiable or Market Risk only.

Relationship between Investor Portfolio

Decision and Firm’s Cost of Capital

• Investors can diversify away all specific risk; therefore, should only be rewarded for holding each firm’s market risk => CAPM.

• CAPM provides the firm’s cost of equity.

].)([)( fmfi rrErrE

)( mrE

Security Market Line.

Capital Asset Pricing Model

• Estimating Cost of Equity Using Regression Analysis.

We regress the firm’s past share price returns against the market.

• • Weighted Average Cost of Capital (WACC).

When we have estimated Cost of Debt, and Cost of Equity-

if we have market values of debt and equity, we can calculate WACC – discount rate in NPV of new investments.

ed KequityKdebtWACC *%*%

Lecture 5 and 6: Capital Structure and Dividends.

Positive NPV project immediately increases current equity value (share price immediately goes up!)

oo EBV Pre-project announcement

New project: .IVNPV n INew capital (all equity)

Value of Debt oBIVE n 0

New Firm Value

Original equity holders

New equity

Example:

oo EBV =500+500=1000.

IVNPV n 60 -20 = 40.

oB = 500.

IVE n 0 = 500+40 = 540

I = 20

nVV =1000+60=1060.

Value of Debt

Original Equity

New Equity

Total Firm Value

Positive NPV: Effect on share price.

Assume all equity.

Market No of Price per Market No of Price per£K Value Shares Share Value Shares Share

Current 1000 1000 1 1040 1000 1.04

New Project 20 19 1.04

Project Income 60 1060 1019 1.04

Required Investment 20

NPV 40

Value of the Firm and Capital Structure

Value of the Firm = Value of Debt + Value of Equity = discounted value of future cashflows available to the providers of capital.

(where values refer to market values).

Capital Structure is the amount of debt and equity: It is the way a firm finances its investments.

Unlevered firm = all-equity.

Levered firm = Debt plus equity.

Miller-Modigliani said that it does not matter how you split the cake between debt and equity, the value of the firm is unchanged (Irrelevance Theorem).

Value of the Firm = discounted value of future cashflows available to the providers of capital.

-Assume Incomes are perpetuities.

Miller- Modigliani Theorem:

TNCFBTVV

Irrelevance Theorem: Without Tax, Firm Value is independent of the Capital Structure.

Note that ed KequitytKdebtWACC *%)1(*%

D/E D/E

Without Taxes With Taxes

Examples

• Firm X

• Henderson Case study

MM main assumptions:

- Symmetric information.

-Managers unselfish- maximise shareholders wealth.

-Risk Free Debt.

MM assumed that investment and financing decisions were separate. Firm first chooses its investment projects (NPV rule), then decides on its capital structure.

Pie Model of the Firm:

MM irrelevance theorem- firm can use any mix of debt and equity – this is unsatisfactory as a policy tool.

Searching for the Optimal Capital Structure.

-Tax benefits of debt.

-Asymmetric information- Signalling.

-Agency Costs (selfish managers).

-Debt Capacity and Risky Debt.

Optimal Capital Structure maximises firm value.

Combining Tax Relief and Debt Capacity (Traditional View).

D/E D/E

Section 4: Optimal Capital Structure, Agency Costs, and Signalling.

Agency costs - manager’s self interested actions. Signalling - related to managerial type.

Debt and Equity can affect Firm Value because:

- Debt increases managers’ share of equity.

-Debt has threat of bankruptcy if manager shirks.

- Debt can reduce free cashflow.

But- Debt - excessive risk taking.

AGENCY COST MODELS.

Jensen and Meckling (1976).

- self-interested manager - monetary rewards V private benefits.

- issues debt and equity.

Issuing equity => lower share of firm’s profits for manager => he takes more perks => firm value

Issuing debt => he owns more equity => he takes less perks => firm value

Jensen and Meckling (1976)

If manager owns all of the equity, equilibrium point A.

Slope = -1

If manager owns half of the equity, he will got to point B if he can.

Slope = -1

Slope = -1/2

If manager owns half of the equity, he will got to point B if he can.

Final equilibrium, point C: value V2, and private benefits B1.

Slope = -1

Slope = -1/2

Jensen and Meckling - Numerical Example.PROJECT PROJECTA B

EXPECTED INCOME 500 1000

MANAGER'S SHARE:100% 500 1000

VALUE OF PRIVATE 800 500BENEFITS

TOTAL WEALTH 1300 1500

MANAGER'S SHARE:50% 250 500

VALUE OF PRIVATE 800 500BENEFITS

TOTAL WEALTH 1050 1000

Manager issues 100% Debt.

Chooses Project B.

Manager issues some Debt and Equity.

Chooses Project A.

Optimal Solution: Issue Debt?

Issuing debt increases the manager’s fractional ownership => Firm value rises.

Debt and risk-shifting.

State 1 100 0 0.5

State 2 100 170 0.5

100 85

Values: Debt 50 25

Equity 50 60

OPTIMAL CAPITAL STRUCTURE.

Trade-off: Increasing equity => excess perks.

Increasing debt => potential risk shifting.

Optimal Capital Structure => max firm value.

Other Agency Cost Reasons for Optimal Capital structure.

Debt - bankruptcy threat - manager increases effort level. (eg Hart, Dewatripont and Tirole).

Debt reduces free cashflow problem (eg Jensen 1986).

Agency Cost Models – continued.

Effort Level, Debt and bankruptcy (simple example).

Debtholders are hard- if not paid, firm becomes bankrupt, manager loses job- manager does not like this.

Equity holders are soft.

Effort Level

High Low Required

Income 500 100 200

What is Optimal Capital Structure (Value Maximising)?

Firm needs to raise 200, using debt and equity.

Manager only cares about keeping his job. He has a fixed income, not affected by firm value.

a) If debt < 100, low effort. V = 100. Manager keeps job.

b) If debt > 100: low effort, V < D => bankruptcy. Manager loses job.

So, high effort level => V = 500 > D. No bankruptcy => Manager keeps job.

High level of debt => high firm value.

However: trade-off: may be costs of having high debt levels.

Free Cashflow Problem (Jensen 1986).

-Managers have (negative NPV) pet projects.

-Empire Building.

=> Firm Value reducing.

Free Cashflow- Cashflow in excess of that required to fund all NPV projects.

Jensen- benefit of debt in reducing free cashflow.

Jensen’s evidence from the oil industry.

After 1973, oil industry generated large free cashflows.

Management wasted money on unnecessary R and D.

also started diversification programs outside the industry.

Evidence- McConnell and Muscerella (1986) – increases in R and D caused decreases in stock price.

Retrenchment- cancellation or delay of ongoing projects.

Empire building Management resists retrenchment.

Takeovers or threat => increase in debt => reduction in free cashflow => increased share price.

Jensen predicts:

young firms with lots of good (positive NPV) investment opportunities should have low debt, high free cashflow.

Old stagnant firms with only negative NPV projects should have high debt levels, low free cashflow.

Stultz (1990)- optimal level of debt => enough free cashflow for good projects, but not too much free cashflow for bad projects.

Income Rights and Control Rights.

Some researchers (Hart (1982) and (2001), Dewatripont and Tirole (1985)) recognised that securities allocate income rights and control rights.

Debtholders have a fixed first claim on the firm’s income, and have liquidation rights.

Equityholders are residual claimants, and have voting rights.

Class discussion paper: Hart (2001)- What is the optimal allocation of control and income rights between a single investor and a manager?

How effective are control rights when there are different types of investors?

Why do we observe different types of outside investors- what is the optimal contract?

Conflict Benefits of Debt Costs of Debt

Breaking MM Tax Relief Fin’l Distress/ Debt Capacity

Agency Models

JM (1976) Managerial Perks

Increase Mgr’s Ownership

Risk Shifting

Jensen (1986) Empire Building Reduce Freecash Unspecified.

Stultz Empire Building Reduce Freecash Underinvestment.

Dewatripont and Tirole, Hart.

Low Effort level Bankruptcy threat =>increased effort

DT- Inefficient liquidations.

Signalling Models of Capital Structure

Assymetric info: Akerlof’s (1970) Lemons Market.

Akerlof showed that, under assymetric info, only bad things may be traded.

His model- two car dealers: one good, one bad.

Market does not know which is which: 50/50 probability.

Good car (peach) is worth £2000. Bad car (lemon) is worth £1000.

Buyers only prepared to pay average price £1500.

But: Good seller not prepared to sell. Only bad car remains.

Price falls to £1000.

Myers-Majuf (1984) – “securities may be lemons too.”

Asymmetric information and Signalling Models.

- managers have inside info, capital structure has signalling properties.

Ross (1977)

-manager’s compensation at the end of the period is

DVCVVrM

DVVVrM

if )1(

D* = debt level where bad firm goes bankrupt.

Result: Good firm D > D*, Bad Firm D < D*.

Debt level D signals to investors whether the firm is good or bad.

Myers-Majluf (1984).

-managers know the true future cashflow.

They act in the interest of initial shareholders.P = 0.5 Do

Nothing:

Good Bad

IssueEquity

Good BadAssetsin Place

250 130 350 230

NPV ofnewproject

0 0 20 10

Value ofFirm

250 130 370 240

Expected Value 190 305

New investors 0 100

Old Investors 190 205

Consider old shareholders wealth:

Good News + Do nothing = 250.

Good News + Issue Equity =

Bad News and do nothing = 130.

.69.248)370(305

Bad News and Issue equity = .31.161)240(305

Donothing

Issueandinvest

GoodNews

250 * 248.69

BadNews

130 161.31*

Old Shareholders’ payoffs EquilibriumDonothing

Issueandinvest

GoodNews

250 * 248.69

BadNews

130 140 *

Issuing equity signals that the bad state will occur.

The market knows this - firm value falls.

Pecking Order Theory for Capital Structure => firms prefer to raise funds in this order:

Retained Earnings/ Debt/ Equity.

Evidence on Capital structure and firm value.

Debt Issued - Value Increases.

Equity Issued- Value falls.

However, difficult to analyse, as these capital structure changes may be accompanied by new investment.

More promising - Exchange offers or swaps.

Class discussion paper: Masulis (1980)- Highly significant Announcement effects:

+7.6% for leverage increasing exchange offers.

-5.4% for leverage decreasing exchange offers.

Practical Methods employed by Companies (See Damodaran; Campbell and Harvey).

-Trade off models: PV of debt and equity.

-Pecking order.

-Benchmarking.

-Life Cycle.

Increasing Debt?

Trade-off Versus Pecking Order.

• Empirical Tests.• Multiple Regression analysis (firm size/growth

opportunities/tangibility of assets/profitability…..• => Relationship between profitability and leverage

(debt): positive => trade-off.• Or negative => Pecking order:• Why? • China: Reverse Pecking order

Capital Structure and Product Market Competition.

• Research has recognised that firms’ financial decisions and product market decisions not made in isolation.

• How does competition in the product market affect firms’ debt/equity decisions?

• Limited liability models: Debt softens competition: higher comp => higher debt.

• Predation models: higher competition leads to lower debt. (Why?)

Capital Structure and Takeovers

• Garvey and Hanka:

• Waves of takeovers in US in 1980’s/1990’s.

• Increase in hostile takeovers => increase in debt as a defensive mechanism.

• Decrease in hostile takeovers => decrease in debt as a defensive mechanism.

Garvey and Hanka (contiuned)

V Trade-off: Tax shields/effort levels/FCF/ efficiency/signalling Vs financial distress

Practical Capital Structure: case study

Lecture 6: Dividend Policy

• Miller-Modigliani Irrelevance.

• Gordon Growth (trade-off).

• Signalling Models.

• Agency Models.

• Lintner Smoothing.

• Dividends versus share repurchases.

• Empirical examples

Early Approach.

• Three Schools of Thought-• Dividends are irrelevant (MM).• Dividends => increase in stock prices

(signalling/agency problems).• Dividends => decrease in Stock Prices (negative

signal: non +ve NPV projects left?).• 2 major hypotheses: Free-cash flow versus

signalling

Important terminology

• Cum Div: Share price just before dividend is paid.

• Ex div: share price after dividend is paid < Cum div.

Example

• A firm is expecting to provide dividends every year-end forever of £10. The cost of equity is 10%.

• We are at year-end, and div is about to be paid. Current market value of equity = 10/0.1 + 10 = £110

• Div is paid. Now, current market value is• V = 10/0.1 = £100.• So on…

CD = 110

ED = 100

Common Stock Valuation Model

• You are considering buying a share at price Po, and expect to hold it one year before selling it ex-dividend at price P1: cost of equity = r.

)1()1(11

What would the buyer be prepared to pay to you?

)1()1(22

0 )1()1(1 r

Therefore:

Continuing this process, and re-substituting in (try it!), we obtain:

0 )1(t tr

Price today is discounted value of all future dividends to infinity (fundamental value = market value).

Dividend Irrelevance (Miller-Modigliani)

• MM consider conditions under which dividends are irrelevant.

• Investors care about both dividends and capital gains.

• Perfect capital markets:-• No distorting taxes• No transactions costs.• No agency costs or assymetric info.

Dividend Irrelevance (MM): continued

• Intuition: Investors care about total return (dividends plus capital gains).

• Homemade leverage argument

• Source and application of funds argument => MM assumed an optimal investment schedule over time (ie firm invests in all +ve NPV projects each year).

Deriving MM’s dividend irrelevance

• Total market value of our all-equity firm is •

10 )1(

Sources = Uses

1)1( ttttt FrIDFCF

Re-arranging:

1)1( ttttt FrIFCFD

FrIFCFFrIFCFS

10000 ...)1(

)1()1(

010 FCF

Substitute into first equation:

At t =0,

FrIFCFIFS

000 ...)1(

Successive substitutions

00 )1(

•Current value of all-equity firm is present value of operating cashflows less re-investment for all the years (residual cashflow available to shareholders) Dividends do not appear!

•Assn: firms make optimal investments each period (firm invests in all +ve NPV projects).

•Firms ‘balance’ divs and equity each period: divs higher than residual cashflow => issue shares.

•Divs lower than free cashflow: repurchase shares.

Irrelevance of MM irrelevance (Deangelo and Deangelo)

• MM irrelevance based on the idea that all cash will be paid as dividend in the end (at time T).

• Deangelo argues that even under PCM, MM irrelevance can break down if firm never pays dividend!

Irrelevance of MM irrelevance (continued)

• Consider an all-equity firm that is expected to produce residual cashflows of £10 per year for 5 years.

• Cost of equity 10%.

• First scenario: firm pays no dividends for the first 4 years. Pays all of the cashflows as dividends in year 5.

• Now it is expected to pay none of the cashflows in any year: Vo = 0 !

?)1.1(

10 t tV

“Breaking” MM’s Irrelevance

• MM dividend irrelevance theorem based on:

• PCM

• No taxes

• No transaction costs

• No agency or asymmetric information problems.

Gordon Growth Model.

• MM assumed firms made optimal investments out of current cashflows each year

• Pay any divs it likes/ balanced with new equity/repurchases.

• What if information problems etc prevent firms easliy going back to capital markets:

• Now, real trade-off between investment and dividends?

Gordon Growth Model.

Where does growth come from?- retaining cashflow to re-invest.

.)1)(1(001

KKrNCFg

Constant fraction, K, of earnings retained for reinvestment.

Rest paid out as dividend.

Average rate of return on equity = r.

Growth rate in cashflows (and dividends) is g = Kr.

Example of Gordon Growth Model.£K 19x5 19x6 19x7 19x8 19x9 Average Profits After Tax (NCF) 2500 2760 2635 2900 3100Retained Profit (NCF.K) 1550 1775 1600 1800 1900

Dividend (NCF(1-K)) 950 985 1035 1100 1200

Share Capital + retentionsB/F 30000 31550 33325 34925 36725C/F (= BF + Retained Profit) 31550 33325 34925 36725 38625

Retention Rate K 0.62 0.64 0.61 0.62 0.61 0.62r on opening capital 0.083 0.087 0.079 0.083 0.084 0.083

g = Kr = 0.05.

How do we use this past data for valuation?

Gordon Growth Model (Infinite Constant Growth Model).

Let %12

05.012.0

1260)05.1(1200)1( 100

= 18000

Finite Supernormal Growth.

-Rate of return on Investment > market required return for T years.

-After that, Rate of Return on Investment = Market required return.

)(.. 1

rTNCFK

If T = 0, V = Value of assets in place (re-investment at zero NPV).

Same if r = .

Examples of Finite Supernormal Growth.

T = 10 years. K = 0.1.

A. Rate of return, r = 12% for 10 years,then 10% thereafter.

1018)1.01(1.0

)1.012.0(10).100.(1.0

B. Rate of return, r = 5% for 10 years,then 10% thereafter.

955)1.01(1.0

)1.005.0(10).100.(1.0

Dividend Smoothing V optimal re-investment (Fairchild 2003)

• Method:-• GG Model: derive optimal retention/payout

ratio• => deterministic time path for dividends,

Net income, firm values.• => Stochastic time path for net income:

how can we smooth dividends (see Lintner smoothing later….)

Deterministic Dividend Policy.

• Recall

• Solving

• We obtain optimal retention ratio

.)1)(1(01

.)1)((

Analysis of

• If

• If with

• Constant r over time => Constant K* over time.

r .0* K

r ],1,0[*K .0*

Deterministic Case (Continued).

• Recursive solution:

• => signalling equilibria.

• Shorter horizon => higher dividends.

tt rKKND )*1*)(1(0

When r is constant over time, K* is constant. Net Income, Dividends, and firm value evolve deterministically.

Stochastic dividend policy.

• Future returns on equity normally and independently distributed, mean r.

• Each period, K* is as given previously.

• Dividends volatile.

• But signalling concerns: smooth dividends.

• => “buffer” from retained earnings.

Agency problems

• Conflicts between shareholders and debtholders: risk-shifting: high versus low dividends => high divs => credit rating of debt

• Conflicts between managers and shareholders: Jensen’s FCF, Easterbrook.

Are Dividends Irrelevant?

- Evidence: higher dividends => higher value.

- Dividend irrelevance : freely available capital for reinvestment. - If too much dividend, firm issued new shares.

- If capital not freely available, dividend policy may matter.

C. Dividend Signalling - Miller and Rock (1985).

NCF + NS = I + DIV: Source = Uses.

DIV - NS = NCF - I.

Right hand side = retained earnings. Left hand side - higher dividends can be covered by new shares.

Div - NS - E (Div - NS) = NCF - I - E (NCF - I)

= NCF - E ( NCF).

Unexpected dividend increase - favourable signal of NCF.

Prob 0.5 0.5

Firm A Firm B E(V)

NCF 400 1400 900

New Investment 600 600 600

Dividend 0 800 400New shares 200 0 100

E(Div - NS) = E(NCF - I) = 300.

Date 1 Realisation: Firm B: Div - NS - E (Div - NS) = 500 = NCF - E ( NCF).

Firm A : Div - NS - E (Div - NS) = -500 = NCF - E ( NCF).

Dividend Signalling Models.

• Bhattacharya (1979)

• John and Williams (1985)

• Miller and Rock (1985)

• Ofer and Thakor (1987)

• Fuller and Thakor (2002).

• Fairchild (2009/10).

• Divs credible costly signals: Taxes or borrowing costs.

Competing Hypotheses.

• Dividend Signalling hypothesis Versus Free Cashflow hypothesis.

• Fuller and Thakor (2002; 2008): Consider asymmetric info model of 3 firms (good, medium, bad) that have negative NPV project available

• Divs used as a) a positive signal of income, and b) a commitment not to take –ve NPV project (Jensen’s FCF argument).

• Both signals in the same direction (both +ve)

Signalling, FCF, and Dividends.Fuller and Thakor (2002)

• Signalling Versus FCF hypotheses.

• Both say high dividends => high firm value

• FT derive a non-monotonic relationship between firm quality and dividends.

Firm Quality

Fairchild (2009, 2010)

• Signalling Versus FCF hypotheses.

• But, in contrast to Fuller and Thakor, I consider +ve NPV project.

• Real conflict between high divs to signal current income, and low divs to take new project.

• Communication to market/reputation.

Cohen and Yagil

• New agency cost: firms refusing to cut dividends to invest in +ve NPV projects.

• Wooldridge and Ghosh

• 6 roundtable discussions of CF.

Agency Models.

• Jensen’s Free Cash Flow (1986).

• Stultz’s Free Cash Flow Model (1990).

• Easterbrook.

• Fairchild (2009/10): Signalling + moral hazard.

Behavioural Explanation for dividends

• Self-control.

• Investors more disciplined with dividend income than capital gains.

• Mental accounting.

• Case study from Shefrin.

• Boyesen case study.

D. Lintner Model.

Managers do not like big changes in dividend (signalling).

They smooth them - slow adjustment towards target payout rate.

)..( 11 DivepstTKDivDiv ttt K is the adjustment rate. T is the target payout rate.

Dividend Policy -Lintner Model

1 2 3 4 5 6 7 8

FIRM A B CK 0.5 0 1

YEAR EPS DIV DIV DIV

1 30.00 13.25 11.50 15.002 34.00 15.13 11.50 17.003 28.00 14.56 11.50 14.004 25.00 13.53 11.50 12.505 29.00 14.02 11.50 14.506 33.00 15.26 11.50 16.507 36.00 16.63 11.50 18.008 40.00 18.31 11.50 20.00

Using Dividend Data to analyse Lintner Model.

In Excel, run the following regression;

ttt cEpsbDivaDiv 1

...)1( 1 epstTKDivKDiv tt

The parameters give us the following information,

a = 0, K = 1 – b, T = c/ (1 – b).

Dividends and earnings.

• Relationship between dividends, past, current and future earnings.

• Regression analysis/categorical analysis.

Dividends V Share Repurchases.

• Both are payout methods.

• If both provide similar signals, mkt reaction should be same.

• => mgrs should be indifferent between dividends and repurchases.

Dividend/share repurchase irrelevance

• Misconception (among practitioners) that share repurchasing can ‘create’ value by spreading earnings over fewer shares (Kennon).

• Impossible in perfect world:

• Fairchild (JAF).

Dividend/share repurchase irrelevance (continued)

• Fairchild: JAF (2006):

• => popular practitioner’s website argues share repurchases can create value for non-tendering shareholders.

• Basic argument: existing cashflows/assets spread over fewer shares => P !!!

• Financial Alchemy !!!

The Example:….

• Kennon (2005): Eggshell Candies Inc

• Mkt value of equity = $5,000,000.

• 100, 000 shares outstanding

• => Price per share = $50.

• Profit this year = £1,000,000.

• Mgt upset: same amount of candy sold this year as last: growth rate 0% !!!

Eggshell example (continued)

• Executives want to do something to make shareholders money after the disappointing operating performance:

• => One suggests a share buyback.

• The others immediately agree !

• Company will use this year’s £1,000,000 profit to but stock in itself.

Eggshell example (continued)

• $1m dollars used to buy 20,000 shares (at $50 per share). Shares destroyed.

• => 80,000 shares remain.• Kennon argues that, instead of each share

being 0.001% (1/100,000) of the firm, it is now .00125% of the company (1/80)

• You wake up to find that P from $50 to $62.50. Magic!

Kennon quote

• “When a company reduces the amount of shares outstanding, each of your shares becomes more valuable and represents a greater % of equity in the company … It is possible that someday there may be only 5 shares of the company, each worth one million dollars.”

• Fallacy! CF: no such thing as a free lunch!

MM Irrelevance applied to Eggshell example

At beginning of date 0:

At end of date 0, with N0 just achieved, but still in the business (not yet paid out as dividends or repurchases:

)1(100

Eggshell figures

000,000,5000,000,1

000,000,1)1(1

Cost of equity will not change: only way to increase value per share is to improve company’s operating performance, or invest in new positive NPV project. Repurchasing shares is a zero NPV proposition (in a PCM).

Eggshell has to use the $1,000,000 profit to but the shares.

Eggshell irrelevance (continued)

• Assume company has a new one-year zero NPV project available at the end of date 0.

• 1. Use the profit to Invest in the project.

• 2. Use the profit to pay dividends, or:

• 3. Use the profit to repurchase shares.

Eggshell (continued)

50$000,000,525.0

000,000,1000,000,10 PV

40$000,000,425.0

000,000,10 PV Ex div

50$000,000,425.0

000,000,10 PV

Each year –end: cum div = $50, ex div = $40

Long-term effects of repurchase

• See tables in paper:• Share value pre-repurchase = $5,000,000 each

year.• Share value-post repurchase each year =

$4,000,000• Since number of shares reducing, P .by 25%, but

this equals cost of equity.• And is same as investing in zero NPV project.

Conclusion of analysis

• In PCM, share repurchasing cannot increase share price (above a zero NPV investment) by merely spreading cashflows over smaller number of shares.

• Further, if passing up positive NPV to repurchase, not optimal!

• Asymmetric info: repurchases => positive signals.• Agency problems: FCF.• Market timing.• Capital structure motives.

Dividend/share repurchase irrelevance

• See Fairchild (JAF 2005)

• Kennon’s website

Evidence.

• Mgrs think divs reveal more info than repurchases (see Graham and Harvey “Payout policy”.

• Mgrs smooth dividends/repurchases are volatile.

• Dividends paid out of permanent cashflow/repurchases out of temporary cashflow.

Motives for repurchases (Wansley et al, FM: 1989).

• Dividend substitution hypothesis.

• Tax motives.

• Capital structure motives.

• Free cash flow hypothesis.

• Signalling/price support.

• Timing.

• Catering.

Repurchase signalling.

• Price Support hypothesis: Repurchases signal undervaluation (as in dividends).

• But do repurchases provide the same signals as dividends?

Repurchase signalling: (Chowdhury and Nanda Model: RFS 1994)

• Free-cash flow => distribution as commitment.

• Dividends have tax disadvantage.

• Repurchases lead to large price increase.

• So, firms use repurchases only when sufficient undervaluation.

Open market Stock Repurchase Signalling:

McNally, 1999

• Signalling Model of OM repurchases.

• Effect on insiders’ utility.

• If do not repurchase, RA insiders exposed to more risk.

• => Repurchase signals:

• a) Higher earnings and higher risk,

• b) Higher equity stake => higher earnings.

Repurchase Signalling :Isagawa FR 2000

• Asymmetric information over mgr’s private benefits.

• Repurchase announcement reveals this info when project is –ve NPV.

• Repurchase announcement is a credible signal, even though not a commitment.

Costless Versus Costly Signalling:Bhattacharya and Dittmar 2003

• Repurchase announcement is not commitment.

• Costly signal: Actual repurchase: separation of good and bad firm.

• Costless (cheap-talk): Announcement without repurchasing. Draws analysts’ attention.

• Only good firm will want this

Repurchase timing

• Evidence: repurchase timing (buying shares cheaply.

• But market must be inefficient, or investors irrational.

• Isagawa.

• Fairchild and Zhang.

Repurchases and irrational investors.

Isagawa 2002• Timing (wealth-transfer) model.• Unable to time market in efficient market

with rational investors.• Assumes irrational investors => market

does not fully react. • Incentive to time market.• Predicts long-run abnormal returns post-

announcement.

Repurchase Catering.

• Baker and Wurgler: dividend catering

• Fairchild and Zhang: dividend/repurchase catering, or re-investment in positive NPV project.

Competing Frictions Model:From Lease et al:

Asymmetric Information

Agency Costs

Payout

High Payout

Low Payout

High PayoutLow Payout

Dividend Cuts bad news?

• Fairchild’s 2009/10 article.• Wooldridge and Ghosh:=>• ITT/ Gould• Right way and wrong way to cut dividends.• Other cases from Fairchild’s article.• Signalling/FCF hypothesis.• FCF: agency cost: cutting div to take –ve NPV project.• New agency cost: Project foregone to pay high dividends.• Communication/reputation important!!

Lecture 9: Venture Capital/private equity

• Venture capitalists typically supply start-up finance for new entrepreneurs.

• VC’s objective; help to develop the venture over 5 – 7 years, take the firm to IPO, and make large capital gains on their investment.

• In contrast, private equity firms invest in later stage public companies to take them private….

Private Equity.

• PE firms generally buy poorly performing publically listed firms.

• Take them private• Improve them (turn them around).• Hope to float them again for large gains• Our main focus in this course is venture capital,

But will look briefly at PE later.• “Theory of private equity turnarounds” plus PE

leverage article, plus economics of PE articles.

Venture capitalists

• Venture capitalists provide finance to start-up entrepreneurs

• New, innovative, risky, no track-record…• Hence, these Es have difficulty obtaining

finance from banks or stock market• VCs more than just investors• Provide ‘value-adding’ services/effort• Double-sided moral hazard

Venture capital process

• Investment appraisal stage: seeking out good entrepreneurs/business plans: VC overconfidence?

• Financial contracting stage: negotiate over cashflow rights and control rights.

• Performance stage: both E and VC exert value-adding effort: double-sided moral hazard.

• Ex post hold-up/renegotiation stage? Double sided moral hazard

• => exit: IPO/trade sale => capital gains (IRR)

VC process (continued)

• VCs invest for 5-7 years.

• VCs invest in a portfolio of companies: anticipate that some will be highly successful, some will not

• => attention model of Gifford.

C. Venture Capital Financing

• Active Value-adding Investors.

• Double-sided Moral Hazard problem.

• Asymmetric Information.

• Negotiations over Cashflows and Control Rights.

• Staged Financing

• Remarkable variation in contracts.

Features of VC financing.

• Bargain with mgrs over financial contract (cash flow rights and control rights)

• VC’s active investors: provide value-added services.

• Reputation (VCs are repeat players).

• Double-sided moral hazard.

• Double-sided adverse selection.

Kaplan and Stromberg

• Empirical analysis, related to financial contract theories.

Financial Contracts.

• Debt and equity.

• Extensive use of Convertibles.

• Staged Financing.

• Control rights (eg board control/voting rights).

• Exit strategies well-defined.

Fairchild (2004)

• Analyses effects of bargaining power, reputation, exit strategies and value-adding on financial contract and performance.

• 1 mgr and 2 types of VC.

• Success Probability depends on effort:

VCiM eeP

},1,0{iwhere => VC’s value-adding.

Fairchild’s (2004) Timeline

• Date 0: Bidding Game: VC’s bid to supply finance.

• Date 1: Bargaining game: VC/E bargain over financial contract (equity stakes).

• Date 2: Investment/effort level stage.• Date 3: Renegotiation stage: hold-up

problems• Date 4: Payoffs occur.

Bargaining stage

• Ex ante Project Value

• Payoffs:

.0).1( PRPPRV

Optimal effort levels for given equity stake:

Optimal equity proposals.

• Found by substituting optimal efforts into payoffs and maximising.

• Depends on relative bargaining power, VC’s value-adding ability, and reputation effect.

• Eg; E may take all of the equity.

• VC may take half of the equity.

Equity Stake

Payoffs

E’s choice of VC or angel-financing

• Explain Angels.

• Complementary efforts

• Ex post hold-up/stealing threat

• Fairchild’s model

To come

• Legal effects: (Fairchild and Yiyuan)

• => Allen and Song

• => Botazzi et al

• Negative reciprocity/retaliation.

Ex post hold-up threat

• VC power increases with time.

• Exit threat (moral hazard).

• Weakens entrepreneur incentives.

• Contractual commitment not to exit early.

• => put options.

Other Papers

• Casamatta: Joint effort: VC supplies investment and value-adding effort.

• Repullo and Suarez: Joint efforts: staged financing.

• Bascha: Joint efforts: use of convertibles: increased managerial incentives.

Complementary efforts (Repullo and

Suarez).

• Lecture slides to follow…

Control Rights.

• Gebhardt.

• Lecture slides to follow

Asymmetric Information

• Houben.

• PCP paper.

• Tykvova (lock-in at IPO to signal quality).

E’s choice of financier

• VC or bank finance (Ueda, Bettignies and Brander).

• VC or Angel (Chemmanur and Chen, Fairchild).

Fairness Norms and Self-interest in VC/E Contracting: A Behavioral Game-theoretic

Approach• Existing VC/E Financial Contracting Models

assume narrow self-interest.• Double-sided Agency problems (both E and VC

exert Value-adding Effort) (Casamatta JF 2003, Repullo and Suarez 2004, Fairchild JFR 2004).

• Procedural Justice Theory: Fairness and Trust important.

• No existing behavioral Game theoretic models of VC/E contracting.

My Model:

• VC/E Financial Contracting, combining double-sided Moral Hazard (VC and E shirking incentives) and fairness norms.

• 2 stages: VC and E negotiate financial contract.

• Then both exert value-adding efforts.

How to model fairness? Fairness Norms.

• Fair VCs and Es in society.

• self-interested VCs and Es in society.

• Matching process: one E emerges with a business plan. Approaches one VC at random for finance.

• Players cannot observe each other’s type.

Timeline

• Date 0: VC makes ultimatum offer of equity stake to E;

• Date 1: VC and E exert value-adding effort in running the business

• Date 2 Success Probability• => income R.• Failure probability • =>income zero

1],1,0[

VCEEE eeP

• Expected Value of Project

• Represents VCs relative ability (to E).

ReePRV VCEEE )(

Fairness Norms

• Fair VC makes fair (payoff equalising) equity offer

• Self-interested VC makes self-interested ultimatum offer

• E observes equity offer. Fair E compares equity offer to social norm. Self-interested E does not, then exerts effort.

Expected Payoffs

• PRrePR UFEUE )(2

2])1)[(1(])1[( VCFUSUVC eRPrRPr

If VC is fair, by definition, FU

Solve by backward induction:

• If VC is fair;

• Since

• for both E types.

• =>

EFE ePR FS PP

2)1( VCFVC ePR

VC is fair; continued.

• Given FU

Optimal Effort Levels:

e EFVC

Fair VC’s equity proposal (equity norm):

VC is self-interested:

• From Equation (1), fair E’s optimal effort;

FSFU PP

e EUFUE

Self-interested VC’s optimal Equity proposal

• Substitute players’ optimal efforts into V= PR, and then into (1) and (2). Then, optimal equity proposal maximises VC’s indirect payoff =>

Examples;

• VC has no value-adding ability (dumb money) =>

• =>

• r =0 =>

• r => 1 ,

Example 2

• VC has equal ability to E; =>

• r =0 =>• r => 1 ,

• We show thatas r => 1

],1,0[ FU

Table 1.

Table of venture performance

Graph of Venture Performance.

Future Research.

• Dynamic Fairness Game:ex post opportunism (Utset 2002).

• Complementary Efforts.

• Trust Games.

• Experiments.

• Control Rights.

Private Equity

• JCF paper: slides to follow…

• PE and leverage: slides to follow….

Lecture 10: Introduction to Behavioural Corporate Finance.

•Standard Finance - agents are rational and self-interested.•Behavioural finance: agents irrational (Psychological Biases).•Irrational Investors – Overvaluing assets- internet bubble? Market Sentiment?•Irrational Managers- effects on investment appraisal?•Effects on capital structure?•Herding.

Development of Behavioral Finance I.

• Standard Research in Finance: Assumption: Agents are rational self-interested utility maximisers.

• 1955: Herbert Simon: Bounded Rationality: Humans are not computer-like infinite information processors. Heuristics.

• Economics experiments: Humans are not totally self-interested.

Development of Behavioral Finance II.

• Anomalies: Efficient Capital Markets.• Excessive volatility.• Excessive trading.• Over and under-reaction to news.• 1980’s: Werner DeBondt: coined the term

Behavioral Finance.• Prospect Theory: Kahnemann and Tversky

1980s.

Development III

• BF takes findings from psychology.

• Incorporates human biases into finance.

• Which psychological biases? Potentially infinite.

• Bounded rationality/bounded selfishness/bounded willpower.

• Bounded rationality/emotions/social factors.

Potential biases.

• Overconfidence/optimism • Regret.• Prospect Theory/loss aversion.• Representativeness.• Anchoring.• Gambler’s fallacy.• Availability bias.• Salience….. Etc, etc.

Focus in Literature

• Overconfidence/optimism

• Prospect Theory/loss aversion.

• Regret.

Prospect Theory.

Eg: Disposition Effect:

Sell winners too quickly.

Hold losers too long.

Risk-averse in gains

Risk-seeking in losses

Overconfidence.

• Too much trading in capital markets.

• OC leads to losses?

• But : Kyle => OC traders out survive and outperform well-calibrated traders.

Behavioral Corporate Finance.

• Much behavioral research in Financial Markets.

• Not so much in Behavioral CF.

• Relatively new: Behavioral CF and Investment Appraisal/Capital Budgeting/Dividend decisions.

Forms of Irrationality.

a) Bounded Rationality (eg Mattson and Weibull 2002, Stein 1996).

- Limited information: Information processing has a cost of effort.

- Investors => internet bubble.

b) Behavioural effects of emotions:

-Prospect Theory (Kahneman and Tversky 1997).

- Regret Theory.

- Irrational Commitment to Bad Projects.

- Overconfidence.

C) Catering – investors like types of firms (eg high dividend).

Bounded rationality (Mattson and Weibull 2002).

-Manager cannot guarantee good outcome with probability of 1.

-Fully rational => can solve a maximisation problem.

-Bounded rationality => implementation mistakes.

-Cost of reducing mistakes.

-Optimal for manager to make some mistakes!

-CEO, does not carefully prepare meetings, motivate and monitor staff => sub-optimal actions by firm.

Regret theory and prospect theory (Harbaugh 2002).

-Risky decision involving skill and chance.

-manager’s reputation.

Prospect theory: People tend to favour low success probability projects than high success probability projects.

-Low chance of success: failure is common but little reputational damage.

-High chance of success: failure is rare, but more embarrassing.

Regret theory: Failure to take as gamble that wins is as embarrassing as taking a gamble that fails.

=> Prospect + regret theory => attraction for low probability gambles.

Irrational Commitment to bad project.

-Standard economic theory – sunk costs should be ignored.

-Therefore- failing project – abandon.

-But: mgrs tend to keep project going- in hope that it will improve.

-Especially if manager controlled initial investment decision.

-More likely to abandon if someone else took initial decision.

Real Options and behavioral aspects of ability to revise (Joyce 2002).

-Real Options: Flexible project more valuable than an inflexible one.

-However, managers with an opportunity to revise were less satisfied than those with standard fixed NPV.

Overconfidence and the Capital Structure (Heaton 2002).

-Optimistic manager overestimates good state probability.

-Combines Jensen’s free cashflow with Myers-Majluf Assymetric information.

-Jensen- free cashflow costly – mgrs take –ve NPV projects.

-Myers-Majluf- Free cashflow good – enables mgs to take +ve NPV projects.

-Heaton- Underinvestment-overinvestment trade-off without agency costs or asymmetric info.

Heaton (continued).

-Mgr optimism – believes that market undervalues equity = Myers-Majluf problem of not taking +ve NPV projects => free cash flow good.

-But : mgr optimism => mgr overvalues the firms investment opportunities => mistakenly taking –ve NPV project => free cash flow bad.

-Prediction: shareholders prefer:

-Cashflow retention when firm has both high optimism and good investments.

- cash flow payouts when firm has high optimism and bad investments.

Rational capital budgeting in an irrational world. (Stein 1996).

-Manager rational, investors over-optimistic.

- share price solely determined by investors.

-How to set hurdle rates for capital budgeting decisions?

- adaptation of CAPM, depending on managerial aims.

- manager may want to maximise time 0 stock price (short-term).

-May want to maximise PV of firm’s future cash flows (long term rational view).

Effect of Managerial overconfidence, asymmetric Info, and moral hazard on Capital Structure Decisions.

Rational Corporate Finance.

-Capital Structure: moral hazard + asymmetric info.

-Debt reduces Moral Hazard Problems

-Debt signals quality.

Behavioral Corporate Finance.

-managerial biases: effects on investment and financing decisions

-Framing, regret theory, loss aversion, bounded rationality.

-OVERCONFIDENCE/OPTIMISM.

Overconfidence/optimism

• Optimism: upward bias in probability of good state.

• Overconfidence: underestimation of asset risk.

• My model =>

• Overconfidence: overestimation of ability.

Overconfidence: good or bad?

• Hackbarth (2002): debt decision: OC good.

• Goel and Thakor (2000): OC good: offsets mgr risk aversion.

• Gervais et al (2002), Heaton: investment appraisal, OC bad => negative NPV projects.

• Zacharakis: VC OC bad: wrong firms.

Overconfidence and Debt

• My model: OC => higher mgr’s effort (good).

• But OC bad, leads to excessive debt (see Shefrin), higher financial distress.

• Trade-off.

Behavioral model of overconfidence.

Both Managers issue debt:

.ˆ,ˆ qqpp

.)ˆ1(ˆ2

ˆ bpqp

IpRpM g

.)ˆ1(ˆ2

ˆ bqqp

IqRqM b

Good mgr issues Debt, bad mgr issues equity.

.)ˆ1(ˆ

ˆ bpIp

pRpM g

qRqM b

Both mgrs issue equity.

ˆ Iqp

pRpM g

ˆ Iqp

qRqM b

Proposition 1.

,)ˆ1()ˆ1()(

)(ˆbpbqI

}.{ DSS bg

,)ˆ1()(

)(ˆ)ˆ1( bpI

}.,{ ESDS bg

c) ,)(

)(ˆ)ˆ1()ˆ1( I

qpqbpbq

}.{ ESS bg

Overconfidence leads to more debt issuance.

Overconfidence and Moral Hazard

• Firm’s project: 2 possible outcomes.

• Good: income R. Bad: Income 0.

• Good state Prob:

• True:

• Overconfidence:

• True success prob:

].1,0()( eP .0.0

Manager’s Perceived Payoffs

.)ˆ1()(ˆˆ 2 IPDebPDRPM D

.)1(ˆˆ 2 IPReRPM E

Optimal effort levels

Effect of Overconfidence and security on mgr’s effort

• Mgr’s effort is increasing in OC.

• Debt forces higher effort due to FD.

Manager’s perceived Indirect Payoffs

bIDbDRbDR

)()(ˆ22

IDDRDR

)()(ˆ22

))(2()(ˆ22

bbDbDRb

True Firm Value

))()(()( b

bRbDRbbRPV DD

RPV EE

Effect of OC on Security Choice

))(2()0(ˆ

bbDbIRb

.0)(ˆ CDM

],,0[ C

Manager issues Equity.

Manager issues Debt.

Effect of OC on firm Values

))()(()( b

bRbDRV CD

bDRRbDbbR

)()2)(( 22

Results

• For given security: firm value increasing in OC.• If• Firm value increasing for all OC: OC good.• Optimal OC: • If • Medium OC is bad. High OC is good.• Or low good, high bad.

,0)( CDV

,0)( CDV .* max

Results (continued).

• If

• 2 cases: Optimal OC:

• Or Optimal OC:

,0)( CDV

.* max

Effect of Overconfidence on Firm Value

0 0.1 0.2 0.3 0.4 0.5

Overconfidence

1 2 3 4 5 6 7 8 9 10

Overconfidence

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

OverconfidenceV

Conclusion.

• Overconfidence leads to higher effort level.

• Critical OC leads to debt: FD costs.

• Debt leads to higher effort level.

• Optimal OC depends on trade-off between higher effort and expected FD costs.

Future Research

• Optimal level of OC.

• Include Investment appraisal decision

• Other biases: eg Refusal to abandon.

• Regret.

• Emotions

• Hyperbolic discounting

• Is OC exogenous? Learning.

Herding

Hyperbolic Discounting

Emotional Finance

• Fairchild’s Concorde case study.

1 MN20211: Corporate Finance 2011/12: 1.Revision: Investment Appraisal (NPV). 2.Investment...

Documents