Soybean Basis and Spreads
Implications for Marketing Decisions
April
To ND Soy Council Marketing Conference, Fargo
William W. Wilson
Dept. of Agribusiness and Applied Economics
NDSU
1
Basis Behavior –
◦ Concepts, Seasonality and Behavior
◦ Intermonth spreads
◦ Impacts of shipping costs
◦ Forecasting Basis
◦ Market year analogues for export basis
◦ Seasonal analogues
2
Typical: Stock/use vs price
2018/19: suggest prices 700-800c/b and take 4-6 years to return to normal
2019/20: Revised to .11 and now .104 prices about 875c
2020/21 Revised to .07 implies higher futures approaching 10$
Dept of Agribusiness & Applied
Economics,NDSU, Fargo - 58102 3
4
Taxonomy of Markets
Futures
Cash
B = C - F
Spot
Forward
-specific terms of
trade (location, time,
quantity)
Forward
-standardized terms
of trade
Arbitrage
Risk
Management
Rail Markets
Spot
Forward
Review of Basis
High Correlation Market B=C-F
Key Points:
1) correlation
2) risk: St deviation
97%
0.05$
Correlation =
Basis Standard Deviation =
$2.50
$2.70
$2.90
$3.10
$3.30
$3.50
$3.70
$3.90
$4.10
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Period
Pri
ce
Cash Futures
$(0.35)
$(0.30)
$(0.25)
$(0.20)
$(0.15)
$(0.10)
$(0.05)
$-
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
PeriodP
rice
Basis
Review of Basis
Low Correlation Market
39%
0.23$
Correlation =
Basis Standard Deviation =
$2.50
$2.60
$2.70
$2.80
$2.90
$3.00
$3.10
$3.20
$3.30
$3.40
$3.50
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Period
Pri
ce
Cash Futures
$(0.30)
$(0.20)
$(0.10)
$-
$0.10
$0.20
$0.30
$0.40
$0.50
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
PeriodP
rice
Basis
Factors impacting basis
Location: Shipping
Time: Storage cost
Quality: Premium/discounts
7
8
Delivery
B=DC
D = Storage
Cost
Harvest
B=C-F
Time
Overall Behavior of basis:
Reduced Cost of Storage and Delivery Month
Behavior
Recent basis behavior
MyDTN
Jamestown
PNW Track
Brazil
9
MyDTN Basis for soybean 4 14
10
MyDTN Basis for soybean 4 14
11
Jamestown Seasonal Soybean Basis Current: new crop -120 (65c/b under normal)
12
Kensal area Basis:Normal=-65 Now -110
13
Arthur (Ayr): 2019/20 about
50C/b<normal
Ayr: Extreme post-
harvest appreciation
Seasonal PNW Soybean BasisCurrent: Spot +80 (or less) vs. (vs average 130)
Most important factors impacting US Export Basis:◦ Brazil Export Basis
◦ China imports
◦ Etc.
Recent study: Bullock and Wilson, Factors impacting the US Export Soybean Basis (forthcoming)
16
PNW Basis update
Brazil Soybean Basis
18
Impacts of China Tariffs:
25% is about 190c/b
Impact◦ Delivered price China for US is increased 190c/b relative to all
competitors
◦ Value of Brazil, and all other potential exporters, increases toward this value, 190c/b
◦ Taken together, the world has a bifurcated market Those markets (China) with tariff on US imports
Those markets (all others) without tariff
Compounded by China Gov. Purchases◦ COF CO and Sinograin: Purchases made by these organizations,
and even from US are effectively tariff-free
◦ Commercial buyers in China do not have this advantage
◦ Impact: Purchase made through COFCO and Sinograin to provide good-faith to trade negotiations.
◦ Some may be cancelled, and/or switched to further months forward
19
Bzl would be at a $70 (190c/b) discount to the US.
bzl basis peaks in September/October and decreases rapidly there after.
Soybean Basis: Gulf and PNW
Basis values at U.S. Gulf and PNW are highly correlated
PNW is normally a premium to US Gulf reflecting the greater shipping costs to that port; and the lower ocean shipping costs from that port
Current: FOB Basis
22
Spatial Competition for Northern
Soybean Jan 15 2020, and 4 9 2020
23
Intermonth spreads
24
25
Price of Storage Discuss
◦ Inverted and normal markets
◦ Theory of the price of storage
Illustrate
◦ Normal and inverted
◦ Futures, basis, logistics: Inverse/normal could occur in any of these markets\
POINT: Intermonth spread and total spread represent the ‘market determined price of
storage’ Ps, vs storage cost S
◦ If Ps >S store (store and sell futures e.g., storage hedge)
◦ Ps <S store (don’t store; sell cash)
Futures
Month
Normal Market Inverted Market
Price
level
Intermonth
Spread
Total
Spread
Price
level
Intermonth
Spread
Total
Spread
March 800 800
May 820 +20 775 -25
July 835 +15 +35 745 -30 -65
Supply and Price of Storage
Elements of the S or Storage:
◦ Downward sloping convenience yield (ct)
◦ Upward sloping risk premium (rt)
Large crops/stocks results in
◦ large intermonth storage costs; or,
◦ discounts for nearby relative to deferred delivery periods
Congestion◦ As shipping demand (for all/any commodities)
increases, congestion accrues, resulting in increases in shipping costs
Congestion costs can:
◦ Increase to very large levels
◦ Increase sufficient to divert shipments into storage, or to other routes/modes
◦ Common in Ports
Port congestion
Rail capacity
Crucial in international competition
Convenience Yield
Physical Cost
Risk Premium
Supply
Stocks
Price Of Storage
3.Supply and Demand for Storage
Downward sloping demand
curve
◦ Inventory (K) will result in an
inverted market (Q)
◦ Inventory (L) will result in a
carry market (R)
Current Soybean Market:
April 13, 2020
Simple:
◦ Narrower carry
◦ Store
◦ Sell nearby
28
Soybean Prices: Apr 2020
2020Futures Diff Cumu
lative
Months Per
month
Return to
Storage
Cost to Store 5
May 854
July 862 8 8 2 4 -2
August 865 3 11 3 4 -4
Sept 866 0.75 11.75 4 3 -8.25
Nov 870 4.5 16.25 6 3 -13.75
Jan-20 873 3 19.25 8 2 -20.75
FUTURES SPREADS ONLY! Basis declines Jan-July
Price of Storage
29
Forward Curve
Definition:
◦ Curve representing the prices at which
the market is willing to transact future
business today.
◦ Reflects normal vs inverted markets
30
2020 Soybean:
Forward Curves CME Soybeans Normal Market
31
DTN: Cost of storage calculator
32
Logistics
33
Rail Rates and Car Markets
Elements of rail shipping costs
◦ Tariff Rates (below)
◦ Car markets: Primary and secondary
(Below)
◦ FSC---now cancelled
34
Rail Car Markets
Primary car auctions are at nil premium (avg $54/car)
Secondary car values (DCV)
◦ More risky
◦ Historical: 4 peaks 2006, 2008, 2010, 2013
◦ Sharp increase in 2013/14, and, sharp decrease in recent weeks
◦ Large premiums allocate demand and storage (i.e. don’t ship!)
◦ Now: converge to 0 or <0 premium!
Shippers
◦ Have to option to take coverage
◦ Not all shippers are covered; some benefit while others are out of the market
35
Railroad Pricing Mechanisms
36
Rail Car Market: Taxonomy
Primary market: Initial rail car auction from RR
◦ Weekly
◦ 12-24 months forward
◦ Other terms
Secondary market:
◦ Brokerage
◦ Offers from buyers (e.g., CHS, etc.) as part of procurement
GrainCO: has 3 ways to make trades of trains in secondary transactions◦ sold on a trip-by-trip basis to other
companies;
◦ used by company owned elevators; and
◦ forward contracted to other companies for a set quantity, delivery period and duration.
GrainCO: in trades to affiliated and/or target supplier
◦ typically sells their shuttles to elevators for $25-$50/car/trip over the premium that they paid the carrier.
◦ assumes all risk and liability in regard to the cars being placed on time.
37
BNSF Primary Auction for COTs
38
BNSF Primary Auction for COTs 11
2018
39
2nd Car Market
40
Tradewest 2nd Market August 2017--typical
41
2nd Market Feb 11 2019
Nearby values
inflated
Likely due to
◦ Service problems
◦ Anticipated/real
nearby exports
soybean and HRS
Wide B/A spread
Inverted
New crop nil offers
42
Recent 2nd market Values 4 13 2020
43
44
45
USDA Grain Transportation
http://www.ams.usda.gov/AMSv1.0/ge
tfile?dDocName=STELPRDC5069058
BIG DEAL: Improve transparency
45
USDA Oct 2018
46
Company offers in 2nd transactions
47
Data Behavior
For illustration
As used, in most
cases below in
empirical models
48
Expected Quantity per Shuttle—
Primary Market Shipments per month defined as:
◦ V=Velocity per month
E.g., 2.5 trips per month
V is random
◦ Qshuttle= # shuttles bought in
primary market
E.g. 10*V=25 trains (110 cars)
per month
◦ Cars: 100-110 rail option
Role of 2nd market
◦ If V*Q>shipping demand: sell on
2nd market
◦ If V*Q<shipping demand: buy on
2nd market
49
Velocity (See data below)
Actual cars derived as:
Ca=V*To*110
Ca = actual cars
V=Velocity
To= Trains ordered
110= cars/train (actual = [100,110] rail option
Conversion: 3750b/c
Velocity
Trips/month 1 2 3 1 2 3
2.7 2.7 5.4 8.1 1,113,750 2,227,500 3,341,250
2.8 2.8 5.6 8.4 1,155,000 2,310,000 3,465,000
2.9 2.9 5.8 8.7 1,196,250 2,392,500 3,588,750
3 3 6 9 1,237,500 2,475,000 3,712,500
3.1 3.1 6.2 9.3 1,278,750 2,557,500 3,836,250
3.2 3.2 6.4 9.6 1,320,000 2,640,000 3,960,000
3.3 3.3 6.6 9.9 1,361,250 2,722,500 4,083,750
Trains bought
Car/month
Trains bought
Bushels/month
50
Velocity
51
52
Correlations—Very Interesting!
Correlation Basis PNW
Basis
Jamestown ND
PNW-
Jamestown
Spread
Nearby Soybean
Futures DCV ($/car)
Basis PNW 1.000
Basis
Jamestown ND 0.311 1.000
PNW-
Jamestown
Spread 0.777 -0.219 1.000
Nearby
Soybean
Futures 0.514 0.090 0.490 1.000
DCV ($/car) 0.500 -0.199 0.589 0.251 1.000
53
Rail Car Market: Summary
Feature Primary Secondary
Allocation
mechanism
Bid to buy from Railroad Acquire from brokers or
other shippers
Quantity
allocated
Bid on 12, 24, 36 months of
trains by month
Offers by other shippers
Tradable Yes Yes
Quantity Receive X*N where X is rail
cycle (e.g., 2.9 trips/month
random) and N is number of
trains
Amount purchased
Car placement By period (10 day) s.t. car cycles Assured, by period (10
day)
Guarantee None Yes, by month
Cancellation Possible s.t. penalty None (re-sell in 2nd
market)
Price auction: Average=54$/c s.t. min Market: Average=$225;
no min/max
54
Grain Pricing I:
Basis derivation Implications
Define: B=basis, c=cash, f=futures, T=tariff rail rate;M=margin; d1,d2 are 2 different destiatins, FSC is fuel service
charge; Car is car premium
1 Bo=Co-FConventional: Basis is constant; highly
predictable;
2 Bo=Bd-M-T-F Little more complex
3 Bo=MAX[(Bd1-T1), [(Bd2-T2), [(Bd3-
T3)] -M-F
Multi-markets: makes basis at origin more
dependent on basis and shipping costs to
multiple markets. More volatility in basis!
4 Bo=Bd-[Rt-FSC-Car]-M-F Numerous rail mechanisms makes greater
uncertainty in some elements of shipping costs;
for those shippers not covered, values have
greater volatility (risk)
5 Bo,t+n=Bd,t+n -
[Rt+n-FSCt+n-Cart+n]*-(M+RP)-Ft+n
Forward transaction results in volatility in basis
values resulting from un-covered shippers
having to infer expectations of relevant values;
including an implicit risk premium in margins
55
Summary
US Grain Shipping has evolved radically in past 2 decades
◦ Greater flexibility
◦ Greater efficiency in shipping
◦ Declining pressure on shipping rates
3 components to rail pricing
◦ Tariff
◦ FSC (if applied)
◦ Shuttle premium
Implications for traders
◦ Knowledge of the market
◦ Integrate commodity and freight positions: Long commodity/short freight Evaluate risk of freight positions
◦ Forward purchases reduce risks for suppliers and facilitates better offers
◦ A portion of the seasonal price (futures and basis) is partially offset by shuttle premiums and barge rates
56
56
Break: QUestions
57
Basis Analytics
SOurces
MyDtn
GeoGrain
DTN-ProphetX
Tools
Mapping
Projections
◦ Seasonal
◦ MACD
Spatial competition
58
Grain Hedge: Basis Soybeans
59
MyDTN Basis for soybean 4 14
60
Projections: Short term and
Seasonal
Shorter term:
◦ Simple comparisons of spot vs.. moving
averages
◦ Turning points! MACD
◦ Projections based on MACD (shorter term)
Seasonal indexes and current period
basis values
Seasonal Projections
61
Price Behavior Using Moving Averages
62
Kensal area Basis:Normal=-65 Now -110
63
Ayr: Extreme post-
harvest appreciation
65
Seasonal Analysis and Simple
Projections
• Seasonal Motivations
• Concept of seasonally produced crops
• Seasonal factors affecting basis behavior
• Seasonal Indexes
• Derivation of Seasonal Indexes2
2 A moving average seasonal index was calculated using monthly data from 1985-1996. The seasonal index value
is calculated by estimating the center average of 12 months of prices, then the current price is divided by this
average price. Index scores are then averaged across years for each month to get the average seasonal index
score for a month.
Jamestown Average Soybean
Basis Seasonal basis behavior
shows that on average◦ Basis is at a bottom in July-
August
◦ Basis is at a peak in December
Seasonal changes◦ May-August: -20c
◦ August-Dec: +40c
Basis variability (risk) is shown by the st. deviations◦ Basis is least risk at and
after harvest time
◦ Basis risk escalate in months following January, no doubt due to uncertainties of the impact of Brazilian crop and basis competition
66
67
Seasonal Jamestown Soybean Basis Index
Average of 2010/11-2018/19 (Sept-August)
67
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
Seaso
nal
Ind
ex
-1 Std. Dev.
Mean
+1 Std. Dev.
-1 Std. Dev. 0.77 0.67 0.93 1.09 1.06 1.01 0.76 0.54 0.83 0.86 0.35 0.27
Mean 0.98 1.05 1.15 1.20 1.16 1.11 1.01 0.89 0.98 0.98 0.67 0.61
+1 Std. Dev. 1.18 1.42 1.37 1.30 1.26 1.22 1.25 1.23 1.14 1.09 0.98 0.95
Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug
Seasonal Analog Derivation Using
Cluster Analysis - PNW
Factors Influencing PNW and Gulf
Export Basis
The marketing year average basis level for both Gulf and PNW is highly correlated and is
primarily influenced by competitive pressures – both international and domestic.
Most important impacting basis level:
◦ Brazil Basis
◦ Chinese imports
No single dominant seasonal pattern for either Gulf or PNW basis – the seasonality
characterization varies substantially from year-to-year and is not correlated between the two
markets.
Seasonality of the basis is primarily influenced by
◦ the level of export activity
◦ internal logistical factors such as railcar bottlenecks / delays,
◦ pace of farmer deliveries into the market,
◦ and transportation costs – primarily barge and ocean freight.
Break: QUestions
71
Spatial Competition for Northern
Soybean
72
Stl: Level; and seasonal
appreciation: -40 to +30
73
PNW Track: Level and seasonal
appr: 25 to +90
74
Current Soybean: StL vs PNW Spot and New
Crop
Net Returns 4 14
Competition:
◦ Stl vs PNW
◦ Current result
Spot: PNW Dominates
◦ Basis at PNW is stronger
◦ Net returns favors shipping to PNW
New Crop
◦ Basis in both markets extremely depressed
◦ Net return slightly favor PNW, but only by 6/b (other factors could influence shipping decisions (e.g. barge, congestion, velocity)
75
St. Louis
Barge
Transfer
PNW Track StLouis
Barge
transfer
PNW Track
Central ND
Track Values 30 115 12 90
Freight to Track: Excl car premiums 100 153 100 153
Net Jamestown -70 -38 -88 -63
Jamestown bid -95 -95 -125 -125
Gross Return 25 57 37 62
C/b
New Crop 2020Spot 2019
C/b
Forward basis curve 11 30
Critical: Basis FOB for months forward
Competitive during nearby
Brazil basis falls sharply (70c/b) into March 2020
Like pressure will be on US values, port and interior!
Comparable FOB Basis: Current 4 13
77
Export fob basis 4 13 2020
78
International Arbitrage 4 13
Current market impacted by
◦ Cheap PNW and Bzl Basis,
◦ Cheap PNW ocean shipping vs USG
C&F Values (China) favor
Bzl gains advantage due to cheaper basis in April-August
[Exclude perceived/real quality differentials vs Bzl]– 10-20c/b
79
International USG PNW Bzl
FOB Port (M&P) c/b 75 115 82
Ocean shipping (US ports to China)$/mt 36 17 23
Total Basis + Ocean (delivered China) 172 160 145
Total Cost: Futures (893)+ Basis + Ocean $/mt 391 387 381
USG PNW Bzl Bzl Adv
April 374 369 364 6
May 380 375 370 6
June 380 375 370 6
July 381 377 373 3
Aug 381 377 374 2
Sept 382 377 378 -1
Oct 382 377 380 -3
Nov 385 380 383 -3
Derivation delivered China 4 13 2020
Delivered China $/mt
2020: Chinese Purchases and Coverage
80
Potential slots:
◦ June-Aug;
◦ New crop (US)
Gut-month for
competition:
◦ Bzl vs US
◦ Bzl seasonal low
basis to May!
◦ Bzl origins
currently favored
for these positions
China Imports: Mar 31 2020
Month
Purchase
To date
Yet to
buy
%
Covera
ge
%
Low High
Sep-19 7.3 7.3 7.3 0 100
Oct 7.3 7.3 7.3 0 100
Nov 9.6 9.6 9.6 0 100
Dec 5 5 5 0 100
Jan-20 3.5 3.5 3.5 0 100
Feb 6.7 6.7 6.7 0 100
Mar 7.5 7.5 7.5 0 100
Apr 8.5 8.5 8.5 0 100
May 8.5 9 8.5 0.25 97
June 9.5 9.5 6.7 2.8 71
July 7.25 7.25 4.2 3.05 58
Aug 7 7 1.2 5.8 17
2019/20 Total 87.65 88 76 12 86
Feb-21 Estimate 7 1.50 5 22
21-Mar 8 2.00 6 27
21-Apr 9 0.80 8 9
21-May 9 0.60 8 7
21-Jun 10 0.50 9 5
21-Jul 7 0.40 7 6
2020/21 Est total 48 5.80 43 12
mmt
Demand Range
81
Implications for Growers
Basis provides direction for decisions related to:
Making effective hedging decisions
Basis impacts
Timing or shipping (not purchasing) decisions
Storage decisions
Quality decisions
Shipping strategies (more later)
Understanding basis is critical in negotiations with suppliers/handlers for
both flat price, basis contracts (discussed later) and premiums/discounts for
quality
Major results
Basis had been increasing, now declining in to new crop
Outlook is for lows in October, increasing into Dec/Jan and then declining
Intermonth spreads: All are inverted.
Sell and don’t store
Don’t‘ store to new crop
Sell new crop for as nearby as possible
Basis Fundamentals: 2019
Old crop: Issues China trade-deal related
purchases
US Logistics◦ USGulf flooding
◦ Rail velocity
Bzl harvest, and lower basis
Bzl logistical problems: potential shifting some sales to PNW
Normal seasonal declines Feb-July
New Crop China trade resumption
◦ Elimination of import tariffs
◦ Govt vs private buying
Bzl logistics in Aug-Oct period
Spreads in US futures
PNW dominance; vs US Gulf
US is competitive Sept forward until Bzl new crop
Intense competition in Aug-Sept◦ US marketing 2 crops
◦ Weakening of basis: -20 to 50cb to August
◦ Narrowing intermonth spreads discourage storage post harvest. Only Basis
To Revert to historically High
Values
Requirements
◦ Increase in Brazil basis
Increased shipping costs
◦ Substantial increase in China purchases
◦ Rail rate reductions (substantial)
83
Grain Sellers: Fundamental Market Strategies (Wilson
Interpretation)
Expectation Likelihood Best Cons. Worst
Basis/Spread Futures
High Sell Futures Buy Puts (ITM) Long cash
Low Buy Puts (OTM) Sell Futures Long Cash
High Long Cash Buy Puts
Low Long Cash Out-of-Money
Puts
Sell Forward
Cash Contracts
Buy puts
Basis contract
NPE or DPC
HTA or sell
futures
Buy puts
As confidence about the underlying increase in futures
*Increases: Shift to more direct futures of ITM Options (higher delta)
*Lessens: Shift to more options and lower OTM options (lower delta)
84