managing risk in farming: concepts, research, …managing risk in farming: concepts, research, and...

Managing Risk in Farming: Concepts, Research, and Analysis.By Joy Harwood, Richard Heifner, Keith Coble, Janet Perry, and AgapiSomwaru. Market and Trade Economics Division and Resource EconomicsDivision, Economic Research Service, U.S. Department of Agriculture.Agricultural Economic Report No. 774.

AAbbssttrraacctt

The risks confronted by grain and cotton farmers are of particular interest,given the changing role of the Government after passage of the 1996 FarmAct. With the shift toward less government intervention in the post-1996Farm Act environment, a more sophisticated understanding of risk andrisk management is important to help producers make better decisions inrisky situations and to assist policymakers in assessing the effectiveness ofdifferent types of risk protection tools. In response, this report provides arigorous, yet accessible, description of risk and risk management tools andstrategies at the farm level. It also provides never-before-published dataon farmers’ assessments of the risks they face, their use of alternative riskmanagement strategies, and the changes they would make if faced withfinancial difficulty. It also compares price risk across crops and time peri-ods, and provides detailed information on yield variability.

Keywords: Crop insurance, diversification, futures contracts, leasing,leveraging, liquidity, livestock insurance, marketing contracts, options con-tracts, production contracts, revenue insurance, risk, vertical integration.

AAcckknnoowwlleeddggmmeennttss

The authors gratefully acknowledge the reviews and comments of PeterBarry, University of Illinois; Joe Glauber, Office of the Chief Economist,USDA; Nelson Maurice, Risk Management Agency, USDA; George Wangand Fred Linse, Commodity Futures Trading Commission; and Don West,Cooperative State Research, Education, and Extension Service, USDA.Dave Banker, Bob Collender, and Nicole Ballenger, ERS, provided veryuseful assistance. Mae Dean Johnson and Erma McCray, ERS, developedthe charts and tables. Jack Harrison, Linda Hatcher, Deana Kussman,Mary Maher, Tom McDonald, Victor Phillips, and Cynthia Ray, ERS,provided editorial and production assistance.

Washington, DC 20036-5831 March 1999

Managing Risk in Farming: Concepts, Research, and Analysis Economic Research Service, USDAiiii

Man

agin

g R

isk

in F

arm

ing:

Co

nce

pts

,Res

earc

h,a

nd

An

alys

is

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv

What Is Risk? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Types of Risk Most Important to Producers . . . . . . . . . . . . . 4

Measuring Price and Yield Risk . . . . . . . . . . . . . . . . . . . . . . . . 8

Yield Randomness Varies Regionally . . . . . . . . . . . . . . . . . . . 8

Price Randomness Differs Among Commodities and Changes Over Time . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Yields and Prices Tend To Move in Opposite Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

How Farmers Can Manage Risk . . . . . . . . . . . . . . . . . . . . . . . 14

Enterprise Diversification. . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Vertical Integration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Production Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Marketing Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Hedging in Futures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Futures Options Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . 36

Maintaining Financial Reserves and Leveraging . . . . . . . . . 39

Liquidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

Leasing Inputs and Hiring Custom Work . . . . . . . . . . . . . . . 46

Insuring Crop Yields and Crop Revenues . . . . . . . . . . . . . . . 48

Off-Farm Employment and Other Types of Off-Farm Income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Other Ways of Managing Risk . . . . . . . . . . . . . . . . . . . . . . . 57

Farmers’ Reported Use of Risk Management Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

How Farmers Can Reduce Risk: Examples Using Hedging, Forward Contracting, Crop Insurance,and Revenue Insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Effects on Income Uncertainty Within the Year . . . . . . . . . . 65

Effects on Income Uncertainty Between Years . . . . . . . . . . . 68

Effects of Hedging, Forward Pricing, and Insurance on Average Returns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

TTaabbllee ooff CCoonntteennttss

Economic Research Service, USDA Managing Risk in Farming: Concepts, Research, and Analysis

Man

aging R

iskin

Farmin

g:Co

ncep

ts,Research

,and

An

alysis

iiiiii

Some Practical Aspects of Farm Risk Management . . . . . . 74

Farmers Often Are Willing To Accept Higher Risks To Obtain Higher Incomes. . . . . . . . . . . . . . . . . . . . 74

Crop Insurance and Forward Pricing Generally Can Reduce Income Uncertainty at Very Low Cost . . . . . . . . . 74

Risk Reduction From Forward Pricing Can Be Quite Small for Farms With High Yield Variability or Strongly Negative Yield-Price Correlations . . . . . . . . . . . 76

Elimination of Deficiency Payments Increases Risks for Many, But Not Necessarily All, Producers of Program Crops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Spreading Sales Before Harvest Tends To Reduce Risk, While Spreading Sales After Harvest Tends to Increase Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Forward Pricing May Help Reduce Price Uncertainty Not Only in the Current Year, but Also in Future Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Futures Prices Provide Information Useful in Making Production and Storage Decisions . . . . . . . . . . . . . . . . . . 80

Revenue Insurance Generally Does Not Fully Substitute for Forward Pricing . . . . . . . . . . . . . . . . . . . . . 82

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Appendix 1: Techniques for Measuring Risk . . . . . . . . . . . 107

Appendix 2: Analytical Tools for Assessing the Effectiveness of Risk Management Strategies . . . . . . . . 113

Appendix 3: Examples of Various Types of Crop and Revenue Insurance . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Managing Risk in Farming: Concepts, Research, and Analysis Economic Research Service, USDAiivv

Sum

mar

y

The risks confronted by grainand cotton farmers are of par-

ticular interest, given the changingrole of the Government after pas-sage of the 1996 Farm Act. Withthe shift toward less governmentintervention in the post-1996 FarmAct environment, a more sophisti-cated understanding of risk andrisk management is important tohelp producers make better deci-sions in risky situations and toassist policymakers in assessingthe effectiveness of different typesof risk protection tools. In response,this report provides a rigorous, yetaccessible, description of risk andrisk management tools and strate-gies at the farm level.

Risk is uncertainty that affects anindividual’s welfare, and is oftenassociated with adversity and loss.There are many sources of risk inagriculture, ranging from price andyield risk to the personal risksassociated with injury or poorhealth. In dealing with risky situa-tions, risk management involveschoosing among alternatives toreduce the effects of the varioustypes of risk. It typically requiresthe evaluation of tradeoffs betweenchanges in risk, changes in expect-ed returns, entrepreneurial free-dom, and other variables.

Several surveys have been conduct-ed asking about the types of riskmost important to farmers. Thesesurveys reach similar conclusions.A 1996 USDA survey, for example,indicates that producers are mostconcerned about changes in govern-ment laws and regulations (institu-tional risk), decreases in crop yieldsor livestock output (productionrisk), and uncertainty in commodi-ty prices (price risk). In general,

producers of major field crops tendto be more concerned about priceand yield risk, while livestock andspecialty crop growers are relative-ly more concerned about changes inlaws and regulations.

While concerns about risk varyacross types of producers, other fac-tors are also important in deter-mining the risk inherent in a pro-ducer’s situation. Yield risk, forexample, varies regionally, anddepends on soil type, climate, theuse of irrigation, and other vari-ables. Yield risk tends to be low inCalifornia, where irrigation is wide-spread, and higher in dryland pro-ducing areas in the Great Plains.In contrast to yield risk, price riskfor a given commodity tends not tovary geographically, and dependson such factors as commodity stocklevels and export demand.

Farmers have many options inmanaging agricultural risks. Theycan adjust the enterprise mix(diversify) or the financial struc-ture of the farm (the mix of debtand equity capital). In addition,farmers have access to manytools—such as insurance and hedg-ing—that can help reduce theirfarm-level risks. Off-farm earningsare a major source of income formany farmers that can help stabi-lize farm household income. Indeed,most producers combine the use ofmany different strategies and tools.

Because farmers vary in their atti-tudes toward risk, risk manage-ment cannot be viewed within a“one size fits all” approach. That is,it is not wise to say that “AllMidwestern corn farmers shouldhedge 50 percent of their crop infutures,” or that “No farmer should

SSuummmmaarryy

Economic Research Service, USDA Managing Risk in Farming: Concepts, Research, and Analysisvv

plan to obtain more than two-thirds of his or her income from asingle commodity.” Different farm-ers confront different situations,and their preferences toward riskand their risk-return tradeoffs havea major effect on decisionmaking ineach given situation. A large,industrialized operation, for exam-ple, may hire marketing expertiseto directly use hedging and options,while a smaller farmer may preferto forward contract with other par-ties better able to hedge directly.

Although farmers in similar situa-tions can differ greatly in theirresponse to risk, surveys providean overall view of producer choices.The results of a 1996 survey, con-ducted shortly after passage of the1996 Farm Act, indicate that opera-tors in the largest gross income cat-egories (more than $250,000 annu-ally) are more likely to use virtual-ly all risk management strategiesthan small-scale operators. Keepingcash on hand for emergencies andgood buys was the number onestrategy for every size farm, forevery commodity specialty, and inevery region.

Evaluating the effectiveness of dif-ferent strategies and tools requires

an understanding of the risk-returntradeoffs of individual producers.Several major points can be made,however, that generally apply torisk management. Most of the toolsdiscussed in this report tend toreduce intrayear income uncertain-ty, but may have only small or neg-ligible effects on multiyear uncer-tainties. In addition, some strate-gies—such as the combined use ofinsurance and forward pricing—tend to complement each other inreducing risks.

In short, understanding risk infarming is important for severalreasons. First, most producers areaverse to risk when faced withrisky outcomes. Someone who isrisk averse is willing to accept alower average return for loweruncertainty, with the tradeoffdepending on the person’s level ofrisk aversion. Thus, strategies can-not be evaluated solely in terms ofaverage or expected return, butalso must consider risk. Second,understanding risk helps farmersand others develop strategies formitigating the possibility ofadverse events, and aids in cir-cumventing extreme outcomes,such as bankruptcy.

Sum

mary

Farming is a financially riskyoccupation. On a daily basis,

farmers are confronted with anever-changing landscape of possi-ble price, yield, and other outcomesthat affect their financial returnsand overall welfare. The conse-quences of decisions or events areoften not known with certaintyuntil long after those decisions orevents occur, so outcomes may bebetter or worse than expected.When aggregate crop output orexport demand changes sharply,for example, farm prices can fluc-tuate substantially and farmersmay realize returns that differgreatly from their expectations.

The risks confronted by grain andcotton farmers are of particularinterest given the changing role ofthe Government after passage ofthe 1996 Farm Act. The Act elimi-nated deficiency payments which,between 1973 and 1995, providedprogram crop producers with priceand income support in years of lowprices. Now, participating crop pro-ducers instead receive contractpayments, which are fixedamounts scheduled to decline overtime between 1996 and 2002.Unlike deficiency payments, these

contract payments do not varyinversely with market prices. The1996 Farm Act also eliminatedannual supply management pro-grams, providing producers withthe flexibility to plant any crop(with certain restrictions for fruitsand vegetables) on any acre. TheAct also reduced government inter-vention in dairy markets.

This shift toward less governmentintervention in the post-1996 FarmAct environment creates a need fora more sophisticated understand-ing of risk and risk management.In response, this report provides arigorous, yet accessible, descriptionof risk and risk management toolsand strategies. It describes risk atthe farm level, examining situa-tions facing individual producers.It is designed for risk programmanagers, extension educators,farmers and other business people,and others interested in risk andrisk management issues.Understanding risk is a key ele-ment in helping producers makebetter decisions in risky situations,and also provides useful informa-tion to policymakers in assessingthe effectiveness of different typesof risk protection tools.


Man

aging R

iskin

Farmin

g:Co

ncep

ts,Research

,and

An

alysis

11

MMaannaaggiinngg RRiisskk iinn FFaarrmmiinngg

CCoonncceeppttss,, RReesseeaarrcchh,, aanndd AAnnaallyyssiiss

Joy Harwood, Richard Heifner, Keith Coble,Janet Perry, and Agapi Somwaru

RRiisskk mmaannaaggeemmeennttiinnvvoollvveess cchhoooossiinnggaammoonngg aalltteerrnnaattiivveesstthhaatt hhaavvee uunncceerrttaaiinnoouuttccoommeess aanndd vvaarryy--iinngg lleevveellss ooff eexxppeecctteeddrreettuurrnn..

Risk is uncertainty that affectsan individual’s welfare, and is

often associated with adversity andloss (Bodie and Merton). Risk isuncertainty that “matters,” andmay involve the probability of los-ing money, possible harm to humanhealth, repercussions that affectresources (irrigation, credit), andother types of events that affect aperson’s welfare. Uncertainty (a sit-uation in which a person does notknow for sure what will happen) isnecessary for risk to occur, butuncertainty need not lead to a riskysituation.

For an individual farmer, risk man-agement involves finding the pre-ferred combination of activitieswith uncertain outcomes and vary-ing levels of expected return. Onemight say that risk managementinvolves choosing among alterna-tives for reducing the effects of riskon a farm, and in so doing, affectingthe farm’s welfare position. Somerisk management strategies (suchas diversification) reduce risk with-in the farm’s operation, others(such as production contracting)transfer risk outside the farm, andstill others (such as maintainingliquid assets) build the farm’scapacity to bear risk. Risk manage-ment typically requires the evalua-tion of tradeoffs between changesin risk, expected returns, entrepre-neurial freedom, and other vari-

ables. The following examples illus-trate risk management in farmingand the types of tradeoffs faced byfarmers:

• Enterprise Diversification—Consider Farmer Smith, who isdebating the most appropriateenterprise mix on his operation.In particular, Smith is contem-plating switching 200 acres fromcorn to soybeans within his exist-ing operation of corn, hay, anddairy. By adding this new crop,Smith is less at risk that thefarm will generate low revenuebecause, in his location, incomefrom soybeans is less variablethan income from corn, andbecause individual commodityreturns do not move exactly intandem (they are less than per-fectly correlated). Smith mustconsider this risk reductionagainst the expected net returnsassociated with the new enter-prise, weighing any potentialdecline in net returns againstthe lower income variability thathe believes will be provided bysuch an additional crop.

• Crop Insurance—ConsiderFarmer Jones, who farms wherethe potential for drought is aconstant worry and yield vari-ability is high. Jones can pur-chase insurance to cover a largeportion of the potential loss, orcan self-insure and absorb any

Managing Risk in Farming: Concepts, Research, and Analysis Economic Research Service, USDA22

Wh

atis

Ris

k?

Risk is the possibility of adversity or loss, and refers to“uncertainty that matters.” Consequently, risk managementinvolves choosing among alternatives to reduce the effects ofrisk. It typically requires the evaluation of tradeoffs betweenchanges in risk, expected returns, entrepreneurial freedom,and other variables. Understanding risk is a starting pointto help producers make good management choices in situa-tions where adversity and loss are possibilities.

WWhhaatt IIss RRiisskk??

losses caused by low yields. Ininvestigating the purchase ofcrop insurance, he finds that theannual premium is quite highdue to the significant yield vari-ability in his area. As a result,Farmer Jones must consider therisk-return tradeoffs in decidingwhether or not to purchaseinsurance and, if he decides tobuy insurance, the level of cover-age that best suits his risk man-agement needs.

• Production Contracting—Consider Farmer Johnson, whois considering whether to enterinto a production contract with alarge broiler integrator. Theintegrator retains control overthe chicks as they are raised bythe producer, and prescribes spe-cific feeds, other inputs, and spe-cial management practicesthroughout the production cycle.In return for handing over man-agement decisions, the produc-er’s income risk is greatlyreduced, market access is guar-anteed, and access to capital isensured. Johnson must weighthese potential benefits againsthis reduced entrepreneurialfreedom and the risk of contracttermination on short notice.

As can be seen through these illus-trations, managing risk in agricul-ture does not necessarily involveavoiding risk, but instead, involvesfinding the best available combina-tion of risk and return given a per-son’s capacity to withstand a widerange of outcomes (Hardaker,Huirne, and Anderson). Effectiverisk management involves antici-pating outcomes and planning astrategy in advance given the like-lihood and consequences of events,not just reacting to those eventsafter they occur. That is, the fourmain aspects of risk managementinvolve (1) identifying potentiallyrisky events, (2) anticipating thelikelihood of possible outcomes andtheir consequences, (3) taking

actions to obtain a preferred com-bination of risk and expectedreturn, and (4) restoring (if neces-sary) the firm’s capacity to imple-ment future risk-planning strate-gies when distress conditions havepassed (Hardaker, Huirne, andAnderson; Patrick; Barry).

Because farmers vary in their atti-tudes toward risk and their abilityto address risky situations, riskmanagement cannot be viewedwithin a “one size fits all”approach. That is, it is not wise tosay that “All midwestern cornfarmers should hedge 50 percent oftheir crop in futures,” or that “Nofarmer should plan to obtain morethan two-thirds of his or herincome from a single commodity.”Different farmers confront differ-ent situations and structural char-acteristics, and as explained in thisreport, their preferences towardrisk and their risk-return tradeoffshave a major effect on decision-making in each given situation. Alarge, industrialized operation, forexample, may hire marketingexpertise to directly use hedgingand options, while a small familyfarm may prefer to forward con-tract with other parties better ableto hedge directly.

Understanding risk in farming isimportant for two reasons. First,most producers are averse to riskwhen faced with risky outcomes.Someone who is risk averse is will-ing to accept a lower averagereturn for lower uncertainty, withthe tradeoff depending on the per-son’s level of risk aversion. Thismeans that strategies cannot beevaluated solely in terms of aver-age or expected return, but thatrisk must also be considered.Second, identifying sources ofuncertainty helps farmers and oth-ers address the most importantstrategies for mitigating risk, andaids in circumventing extreme out-comes, such as bankruptcy.

Because farmers varyin their attitudestoward risk and theirability to address riskysituations, risk man-agement cannot beviewed within a “onesize fits all” approach


Wh

atis R

isk?

33

AAccccoorrddiinngg ttoo aa 11999966UUSSDDAA SSuurrvveeyy,, pprroodduucc--eerrss ooff mmaajjoorr ffiieellddccrrooppss aarree ggeenneerraallllyymmoorree ccoonncceerrnneeddaabboouutt pprriiccee aanndd yyiieellddvvaarriiaabbiilliittyy tthhaann aannyyootthheerr ccaatteeggoorryy ooffrriisskk..


Typ

es o

fR

isk

Mo

stIm

po

rtan

tto

Pro

du

cers

Several surveys have asked farm-ers about the most important

types of risk that they confront intheir farming operations. Thesetypes of questions are typically partof a larger survey that inquiresabout producers’ risk managementstrategies, and offers respondents alist of concerns that they can scorein terms of importance. Scores gen-erally are not ranked relative to oneanother, meaning that producersindependently analyze each concernon the list.

In 1996, USDA’s AgriculturalResource Management Study1

(ARMS), a nationwide survey offarm operators, questioned farmersas to their degree of concern aboutfactors affecting the operation oftheir farms. The ARMS is probabili-ty-based, and results can be expand-ed to reflect the U.S. farm sector.The concerns cited in the surveyvaried from “uncertainty in com-modity prices” to “ability to adoptnew technology.” Mean scores foreach concern were estimated byassigning a value to each measureof importance, with “not concerned”receiving a value of 1.00 and “veryconcerned” receiving a value of 4.00.

Wheat, corn, soybean, tobacco, cot-ton, and certain other producers

answering the survey were moreconcerned about yield and pricevariability than any of the othercategories (table 1). This may bepartly due to the 1996 Farm Act,which greatly reduced governmentintervention in markets for pro-gram crops (wheat, corn, cotton, andother selected field crops), and mayhave heightened producers’ wari-ness concerning price risk.Producers of other field crops, nurs-ery and greenhouse crops, beef cat-tle, and poultry were relativelymore concerned about changes inlaws and regulations, perhapsreflecting trepidation about changesin environmental and other policies.Across all farms, the ARMS resultsindicate that producers’ degree ofconcern was greatest regardingchanges in government laws andregulations (with a score of 3.02),decreases in crop yields or livestockproduction (with a score of 2.95),and uncertainty regarding commod-ity prices (with a score of 2.91).

Other surveys have also examinedproducers’ risk perceptions, mostoften focusing on crop production inspecific geographic areas. Theseother surveys, despite the limitedlocation and time period of theanalysis, generally support theARMS findings that price and yieldrisk are the most important con-cerns facing producers of major fieldcrops. One of the most comprehen-

A 1996 USDA survey indicates that producers are most con-cerned about changes in government laws and regulations(institutional risk), decreases in crop yields or livestock out-put (production risk), and uncertainty in commodity prices(price risk). In general, producers of major field crops tendto be more concerned about price and yield risk, while live-stock and specialty crop growers are relatively more con-cerned about changes in laws and regulations.

1The ARMS survey was formerly knownas the Farm Costs and Returns Survey(FCRS).

TTyyppeess ooff RRiisskk MMoosstt IImmppoorrttaanntt ttooPPrroodduucceerrss

sive studies of producers’ attitudestoward risk was conducted in 1983at a land-grant university (Patrickand others, 1985). This survey, cov-ering 12 States, was designed toelicit the most important types ofvariability faced by farmers and todetermine the importance of differ-ent types of variability across dif-ferent regions. Weather and outputprices were cited as the mostimportant sources of crop risk,regardless of location. Producersalso marked inflation, input costs,diseases and pests, world events,and safety and health as otherimportant sources of risk.

Interesting differences, however,appeared by farm-type grouping.For example, farmers in the South-east, where mixed (crop and live-stock) farming is important, andcorn, soybean, and hog producers inthe Midwest, gave less importanceto variability from commodity pro-grams than did cotton or smallgrain growers. Midwestern corn,soybean, and hog producers gavemuch greater importance to familyplans as a source of variability thandid the other farm-type groups.

Producers’ circumstances alsoaffected perceptions of risk in the

1983 Patrick survey. Using aslightly different sample thanabove, Patrick found that thegreater the debt-to-asset ratio, thegreater the importance given torisks associated with the cost ofcredit on crop farms. Risks associ-ated with hired labor increased inimportance as farm size increased.The producer’s level of educationappeared to be relatively unimpor-tant in influencing the importancegiven to different sources of vari-ability (Patrick).

More recently, participants inPurdue’s 1991 and 1993 TopFarmer Crop Workshops were ques-tioned about their attitudes towardfarm risks. They rated crop priceand crop yield variability as the topsources of risk in 1991, but rankedthem second and third in 1993(Patrick and Ullerich; Patrick andMusser). Concern about injury, ill-ness, or death of the operator wasthe highest rated source of risk in1993, significantly higher than in1991 (table 2). The importance ofchanges in government environ-mental regulations, land rents, andtechnology also increased signifi-cantly between 1991 and 1993.Respondents did not give muchimportance to livestock price or pro-

Farmers participat-ing in a 1993 PurdueWorkshop ratedinjury, illness, ordeath of the opera-tor as the highestrated source of risk,followed by cropprice and yield variability.


Typ

es of

Risk

Mo

stIm

po

rtant

to P

rod

ucers

55

Table 1—Farmers' degree of concern about factors affecting the continued operation of their farms

How concerned are you about eachfactor’s effect on the continued operation of your farm?

Oth

er c

ash

grai

ns

Whe

at

Cor

n

Soy

bean

s

Tob

acco

Cot

ton

Oth

er fi

eld

crop

s

Fru

it/nu

ts

Veg

etab

les

Nur

sery

/gre

enho

use

Bee

f

Hog

s

Pou

ltry

Dai

ry

Oth

er li

vest

ock

All

farm

s

Decrease in crop yields or livestock production 3.35 3.51 3.2 2.98 3.16 3.68 2.53 3.05 2.85 2.78 3.09 3.53 3.20 3.4 2.41 2.95

Uncertainty in commodity prices 3.41 3.83 3.4 2.93 3.15 3.75 2.48 2.88 2.82 2.63 2.96 3.31 3.09 3.54 2.47 2.91

Ability to adopt new technology 2.52 2.38 2.39 2.33 2.21 2.77 1.92 2.34 2.09 2.24 2.25 2.63 2.60 2.45 2.12 2.23

Lawsuits 2.43 2.47 2.03 2.46 1.89 2.78 2.07 2.39 2.66 2.06 2.36 2.70 2.32 2.36 2.00 2.26

Changes in consumer preferences 2.65 2.55 2.39 2.40 2.40 2.86 2.13 2.44 2.59 2.69 2.58 3.01 2.79 2.76 2.30 2.47for agricultural products

Changes in Government laws 3.31 3.36 3.15 2.79 2.77 3.54 2.88 2.97 2.75 3.09 3.03 3.23 3.34 3.31 2.88 3.02and regulations

11 = Not concerned, 2 = Slightly concerned, 3 = Somewhat concerned, 4 = Very concerned.

Source: Perry, Janet, editor, "Adaptive Management Decisions--Responding to the Risks of Farming," unpublished working paper, U.S. Dept. Agr., Econ. Res. Serv., December 1997.

Mean scores 1

In California, wherethe use of irrigationis common, outputrisks are secondaryto price risks amonggrowers.


Typ

es o

fR

isk

Mo

stIm

po

rtan

tto

Pro

du

cers

duction variability, likely reflectingthe limited importance of this enter-prise on their operations.

Other surveys of producers in theMidwest and Great Plains havefound similar results. Farmers andranchers in Nebraska indicated inthe mid-1990’s that output pricerisk and yield risk were the mostimportant sources of risk (Jose andValluru). On a 1-10 scale, therespondents rated output price fluc-tuations (6.07), input price fluctua-tions (5.98), and drought (5.73) asthe most important sources of risk.Although hail damage was ratedhigh in importance (6.58), the num-ber of farmers who selected hail asthe most important risk factor waslow. Survey research focusing onKansas lender-to-farming risks hasprovided similar findings (Mintert).

When California growers werequestioned, important regional vari-ations appeared. A 1992/93 surveyof 569 California growers, whichused a ranking scheme similar tothe ones in the Patrick studies,

reveals that output risks are sec-ondary to price risks among grow-ers in that State (Blank, Carter,and McDonald). These growersranked output price and input costsas first and second, respectively,among their risk concerns. Theseresults largely reflect the low yieldrisk faced in California in most sit-uations, due largely to the wide-spread use of irrigation.

Because of the apparent impor-tance of yield and output price riskto many producers, particularly inthe Midwest and Great Plains,these two risks are the focus of thefollowing section, which examinesthe measurement of risk. Disaggre-gate (farm- and county-level) dataare available to measure the priceand yield risk confronted by pro-ducers across the country. Thus, thefollowing section quantifies theprice and yield risks for producersin different locations, using corn asan example crop.

Table 2—Mean and standard deviation of importance ratings of sources of risk by Top Farmer Crop Workshop participants, 1991 and 1993 1

1991 1993(n = 80) (n = 61)

Standard StandardSources of risk Mean deviation Mean deviation

Changes in government commodity programs 3.83 1.08 3.66 1.03Changes in environmental regulations 3.81 1.03 4.13** .78Crop yield variability 4.21 .91 4.13 .78

Crop price variability 4.31 .87 4.16 .86Livestock production variability2 2.86 1.40 2.68 1.34Livestock price variability2 3.17 1.54 2.75 1.37

Changes in costs of current inputs 3.70 .89 3.89 .84Changes in land rents 3.18 1.16 3.56** .96Changes in costs of capital items 3.66 .94 3.77 .82

Changes in technology 3.54 1.03 3.84* .97Changes in interest rates 3.48 1.09 3.52 1.09Changes in credit availability 3.05 1.29 3.21 1.23

Injury, illness, or death of operator 3.86 1.30 4.39** .94Family health concerns -- -- 4.05 .91Changes in family relationships 3.36 1.42 3.73 1.29Changes in family labor force 2.96 1.28 3.11 1.25-- = Not applicable. n = Number. * The difference between years is statistically significant at the 10-percent confidence level. ** The difference between yearsis statistically significant at the 5-percent level.11 = Not important; 5 = Very important. 2In 1991, only 65 and 66 of 80 farmers responded to the livestock production and price variability questions. Had thenonrespondents been coded as a 1 (not important), the means would have been 2.50 and 2.79 for livestock production and price variability, respectively.

Source: Excerpted by ERS from Patrick, George F., and Wesley N. Musser, Sources of and Responses to Risk: Factor Analyses of Large-Scale CornbeltFarmers. Staff Paper No. 95-17, West Lafayette, IN: Purdue University, Department of Agricultural Economics, December 1995.

The sources of risk inagriculture rangefrom price and yieldrisk to financial andcontracting risk.


Typ

es of

Risk

Mo

stIm

po

rtant

to P

rod

ucers

77

Sources of Risk in FarmingSome risks are unique to agriculture, such as the risk of bad weathersignificantly reducing yields within a given year. Other risks, such asthe price or institutional risks discussed below, while common to allbusinesses, reflect an added economic cost to the producer. If thefarmer’s benefit-cost tradeoff favors mitigation, then he or she willattempt to lower the possibility of adverse effects. These risks includethe following (Hardaker, Huirne, and Anderson; Boehlje and Trede;Baquet, Hambleton, and Jose; Fleisher):

Production or yield risk occurs because agriculture is affected bymany uncontrollable events that are often related to weather, includ-ing excessive or insufficient rainfall, extreme temperatures, hail,insects, and diseases. Technology plays a key role in production risk infarming. The rapid introduction of new crop varieties and productiontechniques often offers the potential for improved efficiency, but mayat times yield poor results, particularly in the short term. In contrast,the threat of obsolescence exists with certain practices (for example,using machinery for which parts are no longer available), which cre-ates another, and different, kind of risk.

Price or market risk reflects risks associated with changes in theprice of output or of inputs that may occur after the commitment toproduction has begun. In agriculture, production generally is a lengthyprocess. Livestock production, for example, typically requires ongoinginvestments in feed and equipment that may not produce returns forseveral months or years. Because markets are generally complex andinvolve both domestic and international considerations, producerreturns may be dramatically affected by events in far-removed regionsof the world.

Institutional risk results from changes in policies and regulationsthat affect agriculture. This type of risk is generally manifested asunanticipated production constraints or price changes for inputs or foroutput. For example, changes in government rules regarding the use ofpesticides (for crops) or drugs (for livestock) may alter the cost of pro-duction or a foreign country’s decision to limit imports of a certain cropmay reduce that crop’s price. Other institutional risks may arise fromchanges in policies affecting the disposal of animal manure, restric-tions in conservation practices or land use, or changes in income taxpolicy or credit policy.

Farmers are also subject to the human or personal risks that arecommon to all business operators. Disruptive changes may result fromsuch events as death, divorce, injury, or the poor health of a principalin the firm. In addition, the changing objectives of individuals involvedin the farming enterprise may have significant effects on the longrunperformance of the operation. Asset risk is also common to all busi-nesses and involves theft, fire, or other loss or damage to equipment,buildings, and livestock. A type of risk that appears to be of growingimportance is contracting risk, which involves opportunistic behaviorand the reliability of contracting partners.

Financial risk differs from the business risks previously described inthat it results from the way the firm’s capital is obtained and financed.A farmer may be subject to fluctuations in interest rates on borrowedcapital, or face cash flow difficulties if there are insufficient funds torepay creditors. The use of borrowed funds means that a share of thereturns from the business must be allocated to meeting debt pay-ments. Even when a farm is 100-percent owner financed, the opera-tor’s capital is still exposed to the probability of losing equity or networth.

YYiieelldd vvaarriiaabbiilliittyy iisshhiigghheerr aatt tthhee ffaarrmmlleevveell tthhaann aatt tthheeSSttaattee oorr nnaattiioonnaalllleevveell..


Mea

suri

ng

Pri

ce a

nd

Yie

ld R

isk

YYiieelldd RRaannddoommnneessss VVaarriieessRReeggiioonnaallllyy

Yield variability for a given cropdiffers geographically and

depends on soil type and quality,climate, and the use of irrigation.Yield variability is often measuredby an indicator known as the “coef-ficient of variation,” which meas-ures randomness relative to themean (or average) value in theyield series. Using this measure,variability in corn yields, for exam-ple, ranges from about 0.2 to about0.4 across U.S. farms (fig. 1). Theseestimates were obtained by combin-ing 10 years of individual farm-level yield observations (obtainedfrom USDA’s Risk ManagementAgency (RMA) records) with longerseries of county yield observationsfrom USDA’s National AgriculturalStatistics Service (NASS).2

As can be seen from the map, yieldvariability tends to be lowest inirrigated areas and in the centralCorn Belt, where soils are deep andrainfall is dependable. Much cornproduction in Nebraska, for exam-ple, is irrigated, and yield variabili-ty is, as a result, quite low. Yieldvariability is also quite low in Iowa,Illinois, and other Corn Belt States,where the climate and soils providea nearly ideal location for corn pro-duction. In areas where cornacreage tends to be fairly low andin areas far removed from the cen-tral Corn Belt, yield variability isgenerally higher.

Yield variability can be measuredusing farm-, State-, or national-level data. Estimates tend to belower when variability is meas-ured at the higher State or nation-al levels of aggregation than at thefarm level of aggregation, asshown in the map. This is becauserandom deviations tend to offseteach other when averages aretaken across farms. Also, condi-tions across the region of aggrega-tion may vary widely. Farmers’risks can be seriously underesti-mated by using yield variabilitiesmeasured at the county level or athigher levels of aggregation.

MMeeaassuurriinngg PPrriiccee aanndd YYiieelldd RRiisskk

Price and yield risk, the most important types of risk facedby many producers, have interesting characteristics. Yieldrisk varies regionally and depends on soil type, climate, theuse of irrigation, and other variables. In contrast, price riskfor a given commodity depends on such factors as commodi-ty stock levels and export demand. As illustrated below, cropprices tend to be more volatile than livestock prices, reflect-ing the yield risk inherent in crop production. For a moredetailed understanding of risk measurement and how his-torical information can be used to estimate future risk, seeappendix 1.

2Yield variances were estimated by coun-ty for 1995 by regressing 1956-95 NASSyields on time using a generalized leastsquares estimator, which corrected for yieldheteroscedasticity. Variances of differencesbetween farm yields and NASS county yieldswere estimated for all farms in the RMArecords using 1985-94 observations for thetwo data sets. Farm yield variances by coun-ty were estimated as the sum of the estimat-ed county yield variance and the averagevariance of farm-county yield differences forfarms in the county. Covariances betweenfarm-county yield differences and countyyields were assumed to be zero, which istrue, on average, for all farms in a county.

PPrriiccee RRaannddoommnneessss DDiiffffeerrss AAmmoonngg CCoommmmooddiittiieess aannddCChhaannggeess OOvveerr TTiimmee

While yield expectations beforeplanting generally follow trends,price expectations often fluctuatesubstantially from year to yeardepending on commodity stock lev-els, export demand, and other fac-tors. Futures price quotes serve asuseful proxies for price expecta-tions for commodities traded onfutures exchanges. For example, aSeptember quote for the KansasCity wheat futures contract thatmatures the following July can beinterpreted as the market’s expec-tation in September of the value ofhard red winter wheat in that nextJuly.3

Price randomness can be estimatedby measuring futures price quotechanges from one trading date toanother. Thus, one measure of pricerisk for winter wheat at plantingtime is the standard deviation (orcoefficient of variation) of pricechanges from September to July inthe July wheat futures price. Thatis, the difference between theSeptember 1 quote and the nextJuly 1 quote on the July futurescontract can be obtained for severalyears, and the standard deviation(or coefficient of variation) calculat-ed on that annual series of price dif-ference observations.

Price variability or risk can bemeasured using ratios of successiveprices, Pt / Pt-1 , instead of differ-ences, Pt - Pt-1 , as used in theexample above. Ratios offer severaladvantages. First, the use of ratiosmay eliminate the need to makeadjustments for inflation, providedthat inflation rates are approxi-mately constant over the periodanalyzed. Second, ratios are unitfree, which facilitates comparisonsamong commodities. Third, measur-ing price variability using ratios

The level of futuresprices reflectstraders’ expectationsof prices at deliverytime.


Measu

ring P

rice and

Yield

Risk

99

3A futures contract is an agreemententered into on an exchange (such as theChicago Board of Trade or the ChicagoMercantile Exchange) between a seller whocommits to deliver and a buyer who com-mits to pay for a commodity. Exchangetrading and standardization foster competi-tion, and futures contract quotes are amongthe best current estimates of prices expect-ed at delivery time.

Coefficient of variation

0.20 to 0.240.25 to 0.290.30 to 0.40Greater than 0.40

Figure 1

Estimated farm-level corn yield coefficient of variation by county, 1995

Notes: Shaded areas include counties with at least 500 acres planted to corn. Lower corn yield variabilityindicates that farm yields fall within a narrower range. Based on farm-level data, 1985-94, and long-term county-level trend.Source: Constructed by ERS from USDA, NASS electronic county yield files, 1997, and USDA, RMAelectronic experience and yield record database.

allows the comparison of volatilitiesestimated over time intervals of dif-ferent lengths. For example, theprice volatility estimated with dailydata for a given month can be com-pared with the volatility estimatedfor a year using this procedure.

Futures quotes provide a vehiclefor observing price volatilitychanges over the growing season.To illustrate, volatilities in Decem-ber corn futures prices were esti-mated by month using a 10-yearaverage of the annualized standarddeviation of log (P t / P t-1) for “t”ranging over all trading days of themonth.4 In this example, volatilityin December corn prices tends tobe relatively low from the preced-ing December until just prior toplanting time in April (fig. 2).Volatility increases at plantingtime and is quite high during thecritical months of the growing sea-son as information (particularlyweather information) emerges andaffects prices. Volatility is loweragain in September and the follow-

ing months, when yields have beenlargely determined.

Price volatility differs among com-modities. To estimate volatility forthose commodities not traded onfutures markets, price expectationsmust be approximated in otherways. The preceding year’s price isone of the simplest proxies for theexpected price in a given year.Figure 3 reports estimates of pricevolatilities, based on annual obser-vations, for 20 commodities. Thevolatilities shown are the standarddeviations of the logarithms ofratios of the current year’s price tothe preceding year’s price for 1987-96. Price variability changes overtime, of course, as market condi-tions and government programschange, but relative price variabili-ties for the different commoditiestend to be similar between decades(Heifner and Kinoshita). In thisexample, crop prices were morevolatile than livestock prices,largely reflecting the importance ofyield risk in crop production.Those crops exhibiting the highestvolatilities (exceeding 20 percent)include dry edible beans, pears,lettuce, apples, rice, grapefruit,and grain sorghum. Volatilities forturkeys, milk, and beef cattle were

Futures prices tendto be most variableduring monthswhere weather hasthe greatest impacton yields.


Mea

suri

ng

Pri

ce a

nd

Yie

ld R

isk

Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov0

10

20

30

40Percent

Source: Estimated by ERS from Chicago Board of Trade data.

Figure 2

Volatility of December corn futures by month, averages for 1987-96

4For example, the January volatility esti-mate was constructed by averaging over the10 years the annualized standard devia-tions of logarithms of daily Decemberfutures settlement price relatives, log(Pt /Pt-1), over the trading days in each January.

less than 10 percent, while volatili-ties for the other commodities fellin the 10- to 20-percent range.

Price variability changes not onlywithin the year, but also betweenyears due to year-to-year differencesin crop prospects over the growingseason, changes in government pro-gram provisions, and changes inglobal supply and demand condi-tions. Figure 4 shows estimates ofcorn price volatility by decade, basedon annual observations and the log(P t / P t-1 ) procedure described ear-lier. Corn price variability was quitehigh during the 1920’s and 1930’s,largely due to the collapse of grainprices in the post-World War I peri-od and very low yields in 1934 and1936. Volatility was low during the1950’s and 1960’s, a period charac-terized by high government support,fairly stable yields, and consistentdemand. The 1970’s realized sizablepurchases by Russia early in thedecade, and poor crops in 1983 and1988 contributed to variability inthe 1980’s. Since 1990, variabilityhas been near long-term averagelevels. The same pattern applies tothe other grains as well.

Although price volatility (as wellas price levels) can vary substan-tially over time, prices are highlycorrelated geographically. Price dif-ferences between locations aremore or less held constant by thepotential for transporting com-modities from low-price areas tohigh-price areas, while price differ-ences between grades and classesare similarly constrained by thepossibility of substituting onegrade or class for another. How-ever, prices for grades or classesthat normally sell at a premiumon a more limited market, such ashigh protein spring wheat, may bemore variable than prices for thebulk of the commodity.

Hauling commodities is profitablewhenever the price differentialbetween two points exceeds haul-ing costs. These spatial price rela-tionships are re-established dailyfor those commodities traded onfutures exchanges as local buyersadjust the prices they offer tofarmers to maintain desired rela-tionships with the futures price.For example, consider a centralIllinois elevator operator in

Grain prices tend tobe more volatile thanlivestock prices, withsome fruits and veg-etables also exhibit-ing quite high year-to-year volatilities


Measu

ring P

rice and

Yield

Risk

1111

Dry

edi

ble

bean

sP

ears

for

fres

h us

eLe

ttuce

App

les

for fr

esh us

eR

ice

Gra

pefr

uit,

a ll u

ses

Sor

ghum

Pot

atoe

sO

nion

sA

ll w

heat

Cor

nS

oybe

ans

Ora

nges

, all

uses

Cal

ves

All

hogs

Upl

and

cot

ton

Fre

sh to

mat

oes

Lam

bsA

ll ha

yA

ll eg

gsLi

ve b

roile

rsT

urke

ysA

ll m

ilkC

attle

, all

beef

0

5

10

15

20

25

30

Figure 3

Price volatility, selected commodities, 1987-96

Percent

Source: Estimated by ERS from USDA, NASS, Agricultural Prices, various issues.

January, who sees the Marchfutures price for corn decline. Ifthe operator did not respond byreducing the price to producers,the elevator may end up payingmore to producers than the corncould be sold for in March.

In contrast to prices, yields aremuch less highly correlated geo-graphically. Yield differencesbetween locations vary from yearto year due to varying weatherconditions in different locations. In1988, for example, a major droughtgreatly reduced corn and soybeanyields in the Midwest. As a result,the yield differential between thecentral Corn Belt (Iowa andIllinois) and the Southeast wasmuch less than in years of wide-spread normal weather.

YYiieellddss aanndd PPrriicceess TTeenndd TToo MMoovvee iinn OOppppoossiittee DDiirreeccttiioonnss

Prices for agricultural commoditiesat the national or world level tendto be high when yields are low, andvice versa, because total demandfor food changes only moderatelyfrom year to year, while supply canfluctuate considerably due toweather in major producing coun-tries. Consumers bid up the pricefor crops in short supply, while

crops in abundant supply clear themarket only at low prices. Whentwo variables, such as price andyield, tend to move in oppositedirections, they are said to be neg-atively correlated.

The magnitude of price-yield corre-lation, which measures the strengthof the relationship between priceand yield, varies depending on thelevel of the comparison. Yield andprice on a farm, for example, neednot be related because the output ofone farm does not noticeably affectmarket prices. However, yieldsamong farms within a region tendto move together. As a result, indi-vidual farm yields in major produc-tion areas tend to be positively cor-related with national yields and,therefore, negatively correlated withprice. A negative yield-price correla-tion means that a farmer’s revenueis less variable from year to yearthan it would be otherwise. Themore negative the correlation, thegreater the “offsetting” relationship(or “natural hedge”) that works tostabilize revenues.

Estimates of the farm price-yieldcorrelation for corn in selectedcounties in the United States indi-cate that the correlation tends tobe more strongly negative in the

In major producingareas, the tendencyfor price to be highwhen yields are low(and vice versa) pro-vides farmers with a“natural hedge” thatmakes their incomesless variable thanotherwise.


Mea

suri

ng

Pri

ce a

nd

Yie

ld R

isk

1920's 1930's 1940's 1950's 1960's 1970's 1980's 1990-960

10

20

30

40

50

60Percent

Figure 4

Volatility of corn prices by decade

Source: Estimated by ERS from USDA, NASS, Agricultural Prices, various issues, and other historical price data published by USDA.

Corn Belt than in bordering areasof production (fig. 5). Thus, the nat-ural hedge is more effective in theCorn Belt, and natural movementsin price and yield work to inherent-ly stabilize incomes to a greaterextent in that area than elsewhere.In areas outside major producinglocations (such as in the Southeastor along the east coast), the natu-ral hedge is much weaker, meaningthat low yields and low prices (orconversely, high yields and highprices) are more likely to occursimultaneously. Wheat generallyexhibits lower yield-price correla-tions and weaker natural hedgesthan corn because production isless geographically concentrated.

The magnitude of the naturalhedge has implications for theeffectiveness of various risk-reduc-ing tools. A weaker natural hedge(with a slightly negative correla-tion between price and yield), forexample, implies that forward pric-ing by hedging in futures or byselling forward on the cash marketis more effective in reducingincome risk than when a strongnatural hedge exists, other factorsheld constant. In such situations,fixing a sales price for the crop

works to establish one componentof revenue, reducing the likelihoodof simultaneously low (or high)prices and yields. As a result,hedging corn can, at times, bemore effective in reducing risk inthose areas outside the major pro-ducing regions of the Corn Belt.

Because income risk depends onfactors other than the price-yieldcorrelation, however, the effective-ness of hedging in reducing risk ismore complicated. In particular,yield variability is an importantfactor. Corn yields are typicallymore variable outside the CornBelt, and hedging effectivenessdeclines as yield variability in-creases. Because yield variabilitytends to outweigh the impacts ofthe price-yield correlation, hedgingeffectiveness tends to be higher inthe Corn Belt than in less robustproducing areas. The interaction ofyield variability, price variability,and the price-yield correlation ininfluencing the effectiveness of riskmanagement tools are importantfactors affecting producers’ choiceof the risk management strategiesdiscussed in the next section.

The tendency forcorn yields andprices to be negative-ly related is strongerinside than outsidethe Corn Belt.


Measu

ring P

rice and

Yield

Risk

1133

Yield-price correlation

Less than to -0.40-0.39 to -0.30-0.29 to -0.15Greater than -0.15

Figure 5

Farm-level corn yield-price correlation by county, 1974-94

Note: Shaded areas include counties with at least 500 acres planted to corn.Source: Constructed by ERS from USDA, NASS electronic county yield files, 1997, and USDA, RMAelectronic experience and yield record database.

DDiivveerrssiiffiiccaattiioonn iiss aaffrreeqquueennttllyy uusseedd rriisskkmmaannaaggeemmeenntt ssttrraattee--ggyy tthhaatt iinnvvoollvveess ppaarr--ttiicciippaattiinngg iinn mmoorreetthhaann oonnee aaccttiivviittyy..

The price and yield risks dis-cussed earlier, along with a

farmer’s attitude toward risk, havea major impact on the choice ofrisk management strategies andtools. In analyzing the risk-returntradeoffs associated with differentapproaches, a producer must con-sider his or her expected return todifferent choices and the variancein returns. Economists have usedseveral approaches to capturethese tradeoffs, which vary in howthey describe a farmer’s “worldview” and how flexible they are inspecifying risk attitudes (seeappendix 2 for details).

EEnntteerrpprriissee DDiivveerrssiiffiiccaattiioonn

Diversification is a frequently usedrisk management strategy thatinvolves participating in morethan one activity. The motivationfor diversifying is based on theidea that returns from variousenterprises do not move up anddown in lockstep, so that when oneactivity has low returns, otheractivities likely would have higher

returns. A crop farm, for example,may have several productive enter-prises (several different crops orboth crops and livestock), or mayoperate disjoint parcels so thatlocalized weather disasters areless likely to reduce yields for allcrops simultaneously.

Many crop farms in the Corn Belt,for example, produce both corn andsoybeans. By producing both cropsinstead of only one, the farm isless at risk of having low revenuesbecause revenues from the twocrops are not perfectly positivelycorrelated. In some years, low cornrevenues may be counterbalancedby relatively high soybean rev-enues. Diversification in farminghas many similarities to the man-agement of financial instruments.Mutual fund managers, for exam-ple, tend to hold many stocks, thusdiversifying and limiting the lossesof a particular stock doing poorly.

The extent of farm diversificationin U.S. agriculture is a difficultconcept to measure. USDA analy-sis has measured diversificationusing an entropy index, whichaccounts for both the mix of com-modities and the relative impor-tance of each commodity (meas-ured by its estimated value) to


How

Farm

ers

Can

Man

age

Ris

k

HHooww FFaarrmmeerrss CCaann MMaannaaggee RRiisskk55

Farmers have many options in managing agricultural risks.They can adjust the enterprise mix (diversify) or the finan-cial structure of the farm (the mix of debt and equity capi-tal). In addition, farmers have access to various tools—suchas insurance and hedging—that can help reduce their farm-level risks. Indeed, most producers combine the use of manydifferent strategies and tools. Producers must decide on thescale of the operation, the degree of control over resources(including how much to borrow and the number of hours, ifany, worked off the farm), the allocation of resources amongenterprises, and how much to insure and price forward.

5The ordering of topics in this sectiondoes not reflect their relative importance.In addition, the length of discussion associ-ated with a particular topic is not meant toindicate importance, but reflects the com-plexity of the topic and diversity of the literature.

farm businesses (Jinkins). Theentropy index spans a continuousrange from 0 to 100. The value ofthe index for a completely special-ized farm producing one commodi-ty is 0. A completely diversifiedfarm with equal shares of eachcommodity has an entropy index of100. These entropy indexes forindividual producers can be aggre-gated to provide weighted averageentropy indexes by farm type, farmsize, and other classifications.

Based on the AgriculturalResource Management Study(ARMS), USDA’s entropy indexwork indicates that cotton farms(with an average index of 50) areamong the most diversified, pro-ducing substantial quantities ofcotton, cash grains, fruits and veg-etables, and in some cases, live-stock (table 3). Poultry farms,where 96 percent of the value ofproduction was from poultry in1990, were the least diversified,likely in part due to the impor-tance of production contracts. Suchcontracts can reduce producers’risk, reducing income variabilityand lessening producers’ interestin diversification as a risk man-agement tool (Dodson). In theNorthern Plains and Corn Belt,farms tended to be less diversifiedthan in other parts of the country,particularly when compared withfarms in the Southeast (Jinkins).

In addition, data from the ARMSsurvey indicate that a large por-tion of commercial farming opera-tions specialized in one or twoenterprises during the period1987-91.6 On average, one-third ofall commercial U.S. farms received

nearly all production from just twoenterprises during that period.Further, about one-third of aggre-gate farm production on commer-cial farms was from those engagedin only two enterprises (Dodson).

Many factors may contribute to afarmer’s decision to diversify. Theunderlying theory suggests thatfarmers are more likely to diversi-fy if they confront greater risks infarming, are relatively risk averse,and face small reductions inexpected returns in response todiversification. Other factors mayalso be important. Continuing thecorn and soybean example dis-cussed earlier, planting corn aftersoybeans may reduce the need forfertilizer because of the nitrogen-fixing properties of soybean plants,and planting both corn and soy-beans may spread out labor andmachine use over critical times inthe planting and harvesting sea-sons. In situations where livestockis part of the enterprise mix, theoperator may be kept busythroughout the year, and crop andanimal byproducts may be usedmore fully (Beneke).

Depending on the farm’s situation,however, the costs of diversifyingmay outweigh the benefits, andspecializing may be the preferredstrategy. Diversifying oftenrequires specialized equipment (forexample, different harvestingattachments), and may be limitedby managerial expertise and labor,the productive capacity of theland, and the market potential inthe surrounding area (Dodson).Diversifying requires a broaderrange of management expertisethan does producing only one com-modity, and does not typicallyallow for intensive management.As technologies become more com-plicated, such intensive manage-ment and greater farm specializa-tion may well be increasinglyimportant (Beneke).

Based on USDAresearch, cottonfarms are among themost diversified,while poultry farmsare among the mostspecialized.


How

Farmers C

an M

anage R

isk

1155

6Commercial farms in this specific paperwere defined as those that received at least$50,000 in annual sales, and where theoperator supplied at least 2,000 hours oflabor annually and designated farming ashis or her primary occupation. This defini-tion is more restrictive than is commonlyassumed.

As a result, farmers face tradeoffswhen examining diversification asa strategy versus specialization.Specializing can refine the expert-ise needed for a particular produc-tive activity, and may also lead tothe economies of scale that lowerper unit production costs, increas-ing the profitability of the opera-tion. Indeed, a producer’s decisionto specialize (or diversify) may bemotivated purely by expected prof-its, with no consideration given toreducing risk. Conversely, the ben-efits associated with diversifyingarise through the potential offset-ting revenue interactions amongenterprises, and the complemen-tarity of equipment and activitiesthat are used within the farmingoperation (Scherer).

Empirical analyses of diversifica-tion in farming have usuallyfocused on factors influencingenterprise choices. As an example,a study of over 1,000 crop farms inthe San Joaquin Valley ofCalifornia examined the relation-ship between diversification andsuch variables as farm size andwealth. The authors were interest-ed in the tradeoffs between riskreduction and potential sizeeconomies in a given activity. Theyfound that wealthier farmers aremore specialized, perhaps becausethey are less risk averse thanfarmers having lower net worths(Pope and Prescott). Similarly, they

found that corporate farms (withdiversified ownership and limitedliabilities) are more specialized, asare operators of smaller farms (asmeasured by cropped acreage) andyounger (or less experienced) oper-ators. Young farmers may startsmall and specialized operations,and perhaps become more diversi-fied as they expand their opera-tions. Farm size (measured byacres cropped) had a positive effecton diversification.

The effects of multiple enterpriseson reducing risk have also beenanalyzed. Schoney, Taylor, andHayward examined crop enterprisemixes for Saskatchewan farmers,and found that the gross incomesamong crops were highly correlat-ed. As a result, they concluded thatlittle risk reduction was gained bydiversifying beyond two or threecrops. In addition, they found that,although several crops typicallyhad a risk-reducing effect on theportfolio, these benefits were typi-cally outweighed by the lowergross incomes associated with suchlevels of diversification.

Several diversification studieshave also looked at combining live-stock and crop enterprises on anoperation, with the results depend-ing on the time period of theanalysis and other factors. Heldand Zink, for example, found thatadding livestock to a hypothetical

The benefits associ-ated with diversify-ing arise through thepotential offsettingrevenue interactionsamong enterprises,and the complemen-tarity of equipmentand activities thatare used within thefarming operation.


How

Farm

ers

Can

Man

age

Ris

k

Table 3—Diversification measurement using entropy indexes, 1990Extent of Value of Average enterprises

measure of production from in theCommodity diversification major commodity operation

Entropyindex Percent Number

Cotton 50 62 2.5Tobacco 42 76 2.2Cash grains 39 76 1.9

Vegetables, fruits, and nuts 39 80 1.9Nursery, greenhouse 32 57 1.2

Beef, hogs, and sheep 25 85 1.8Dairy 24 85 2.4Poultry 9 96 1.5Source: Excerpted by ERS from Jinkins, John, “Measuring Farm and Ranch Business Diversity,” AgriculturalIncome and Finance Situation and Outlook Report, AFO-45, U.S. Dept. Agr., Econ. Res. Serv., May 1992, pp. 28-30.

eastern Wyoming irrigated cropfarm could increase gross marginby 7 percent and reduce the coeffi-cient of variation from 0.63 to0.42. Woolery and Adams indicatedthat diversified land use, combinedwith livestock, could increase netincome and reduce relative incomevariability for South Dakota andWyoming farms. Other studieshave reached mixed results as tothe risk-return tradeoff (seePersaud and Mapp; Sonka andPatrick). Despite any benefits thatmay accrue to enterprise diversifi-cation, the opportunities are oftenlimited by resources, climatic con-ditions, market outlets, and otherfactors (Sonka and Patrick).

Geographical diversification(farming at several noncontiguouslocations) may also mitigate risksin crop production by reducing thechances that local weather events(such as hail storms) will have adisastrous effect on income.Nartea and Barry examined thisform of diversification usingIllinois corn and soybean data,and found that risk was notreduced significantly until landparcels were separated by at least30 miles. They accounted for thecosts associated with farmingacross widely dispersed plots (forexample, moving equipment andpeople and monitoring crop condi-tions), and concluded that widelydispersed tracts typically createunfavorable risk-return tradeoffsfor producers. When widely dis-persed parcels are observed, it islikely because of farmers’ desire toexpand their operations, and theirdifficulty in finding additionaltracts of farmland that are closeto their farming bases. Those mostlikely to gain from geographic dis-persion of parcels are institutionalinvestors with large acreages whodo not have to transport equip-ment and who use tenants to farmtheir holdings.

VVeerrttiiccaall IInntteeggrraattiioonn

Vertical integration is one of sever-al strategies that fall within theumbrella of “vertical coordination.”Vertical coordination includes allof the ways that output from onestage of production and distribu-tion is transferred to anotherstage. Farming has traditionallyoperated in an open productionsystem, where a commodity is pur-chased from a producer at a mar-ket price determined at the time ofpurchase. The use of open produc-tion has declined, however, andvertical coordination has increasedas consumers have becomeincreasingly sophisticated andimprovements in technology haveallowed greater product differenti-ation (Martinez and Reed; Allen).A vertically integrated firm, whichretains ownership control of a com-modity across two or more levels ofactivity, represents one type of ver-tical coordination (Mighell andJones).7

There are many examples of verti-cal integration in farming. Farmerswho raise corn and hay as feed fortheir dairy operations are vertical-ly integrated across both crop andlivestock production. Similarly, cat-tle producers who combine raisinga cow-calf herd, backgrounding theanimals to medium weights, andfeeding cattle to slaughter weights

Opportunities fordiversification canbe limited byresources, climaticconditions, marketoutlets, and otherfactors.


How

Farmers C

an M

anage R

isk

1177

7Other types of vertical coordinationreflect differing degrees to which a firm atone stage of production exerts control overthe quality or quantity of output at otherstages (Martinez and Reed). When produc-tion contracts are used, for example, thecontractor (or integrator) typically retainscontrol over the commodity and mostinputs, and the farmer usually receives anincentive-based fee for production services.In this case, the producer retains little con-trol over production decisions. When mar-keting contracts are used, in contrast, afirm commits to purchasing a commodityfrom a producer at a price formula estab-lished in advance of the purchase, and theproducer retains a large degree of decision-making control. Both production and mar-keting contracts are discussed in subse-quent sections in this report.

are vertically integrated. As theseexamples illustrate, vertical inte-gration can encompass changingthe form of the product (corn intolivestock), or combining stages inthe production process under own-ership by one entity (as in the cat-tle example).

From the farmer’s perspective, thedecision to integrate verticallydepends on many complex factors,including the change in profitsassociated with vertical integra-tion, the risks associated with thequantity and quality of the supplyof inputs (or outputs) before andafter integration, and other fac-tors. In particular, the relationshipbetween vertical integration andrisk involves an evaluation of theexpected returns and the varianceand covariance of the farmer’sreturn on investment for the cur-rent activity and the integrationalternative (Logan). If the correla-tion is positive and large acrossactivities, the gains in risk effi-ciency from vertical integrationmay be relatively low. In contrast,a negative correlation across activ-ities implies that integrating verti-cally may well reduce risk for thefarmer by internalizing processeswithin the operation.

In practice, vertical integration inagriculture often involves owner-

ship of both farm production andprocessing activities, particularlyin certain parts of the livestocksector (table 4). Vertical integra-tion is fairly common in the turkeyindustry, for example, where about30 percent of production takesplace on farms that perform multi-ple functions. On the largest oper-ations, the enterprise mix mayinclude a feed mill, a hatchery, agrow-out operation, a slaughterfacility, and a packing plant. Insuch cases, integration moves bothbackward into inputs (feed manu-facturing) and forward into the fin-ished, consumer-ready product.Similarly, egg producers with largeoperations may own their own feedmill, hatchery, laying operation,and freezing/drying plant for theprocessing of egg products(Manchester).

Vertical integration is also com-mon in certain specialty crops, par-ticularly for fresh vegetable andpotato operations (table 4). Inthese industries, vertical integra-tion often encompasses not onlyproduction of the crop, but alsosorting, assembling, and packagingproducts for retail sales. Large,vertically integrated vegetablegrowers, for example, often bothpack and sell their own vegetables,displaying their private brandnames on packages, and at times

Vertical integrationis much more com-mon in certain live-stock and specialtycrop industries thanin field crops.


How

Farm

ers

Can

Man

age

Ris

k

Table 4—Extent of farm production coordinated by vertical integrationCommodity 1970 1990

PercentLivestock:

Broilers 7 8Turkeys 12 28Hogs 1 6Sheep and lambs 12 28

Field crops:Food grains 1 1Feed grains 1 1

Specialty crops:Processed vegetables 10 9Fresh vegetables 30 40Potatoes 25 40Citrus 9 8Other fruits and nuts 20 25

Total farm output 5 8Source: Martinez, Steve W., and Al Reed, From Farmers to Consumers: Vertical Coordination in the FoodIndustry, AIB-720, U.S. Dept. Agr., Econ. Res. Serv., June 1996.

investing in research targeted atdeveloping new varieties.Incentives prompting an operationto adopt this type of verticallyintegrated strategy include theneed for extensive quality control(through control of cultural prac-tices and planting dates) and thedesire for brand-name identifica-tion of products, signaling known-quality produce to buyers(Powers).

While the above examples relate toindividual operations, farmers mayjoin together in a cooperativeorganization that is vertically inte-grated across functions.8 Examplesof farmer-owned, vertically inte-grated cooperatives include OceanSpray, which is owned by about 950cranberry and citrus growers inthe United States and Canada andmarkets fresh products and bottledjuices (Shee). Other vertically inte-grated cooperatives include LandO’Lakes (owned by dairy growers)and Sunkist (owned largely byCalifornia citrus growers).

There are also examples of grainfarmers who have cooperativelyintegrated into processing andother functions. Wheat growers inthe Fairmount, North Dakota, areajointly invested in the constructionof Dakota Valley Mills in late 1997,a farmer-owned mill supplied withwheat from local producers(Sosland Publishing Company,Sept. 1997). Similarly, a Kansas-based farmer cooperative, Twenty-first Century Grain Processing,secured an option in early 1997 tobuy a New Mexico flour mill.Producer members who participatein this venture deliver wheat undera marketing agreement to differentpoints in Kansas and Oklahoma fortransport to the New Mexicomilling site (Sosland PublishingCompany, Feb. 1997; Fee).

The Dakota Growers PastaCompany, which is run by produc-ers in a three-State area in theupper Midwest and includes a milland pasta plant, is similar in con-cept to the Twenty-first Centuryventure just discussed. EachDakota Growers farmer buys ashare of the company and entersinto a contract for delivery of apre-determined quality and quan-tity of wheat by a certain dateeach year (Martinez and Reed). Ifthe average open-market price fora given period exceeds the contractprice, a farmer’s payment isincreased above the initial contractamount. Conversely, if the averagemarket price is less than the con-tract price, the firm makes up thedifference. Premiums are paid forwheat of exceptional quality, andgrowers can purchase wheat fromcompany-held stocks in severeyield-loss years. By operating insuch a manner, the DakotaGrowers Pasta Company is notonly vertically integrated intomilling and pasta production, butalso relies on marketing contracts(see later discussion) among itsfarmer-members.

As noted earlier, the incentives forvertical integration can ariseeither from producers or from buy-ers further down the marketingchain who realize an opportunityto enhance their potential profitsor reduce their risk. A farmer (orgroup of farmers) may verticallyintegrate “downstream” (forwardin the marketing channel), forexample, to assure a market fortheir commodity and to capture agreater share of the value that isassociated with the productionprocess. By doing so, they mayenhance their profits by loweringtransactions costs and by usingmanagement and other resourcesmore efficiently. Risk can also bereduced by guaranteeing a marketoutlet and by avoiding the uncer-tainties of selling and purchasingintermediate commodities inimperfect markets. Conversely, a

The incentives forvertical integrationcan arise either fromproducers or frombuyers further downthe marketing chainwho realize anopportunity toenhance their poten-tial profits or reducetheir risk.


How

Farmers C

an M

anage R

isk

1199

8For more information on cooperatives,see Frederick. Although the discussion herefocuses on farmer cooperatives that alsoengage in processing, many other types ofcooperatives exist.

processor may vertically integrate“upstream” (backwards in the mar-keting channel) to exercise greatercontrol over the quality and timingof deliveries and the quality ofinputs used in the productionprocess. Again, reduced risk and/orgreater profits may result.

The risk-reducing benefits associ-ated with vertical integrationdepend to a great extent on thenature of the industry. Typically,the benefits associated with inte-gration increase as production andmarketing interrelationshipsbecome more complex and whenbreakdowns in marketplace com-petition are most likely (such asopportunistic behavior by contract-ing parties). For perfectly competi-tive industries, all firms are sub-ject to price fluctuations caused bysupply and demand shifts—whether or not they are verticallyintegrated—and integration can-not provide protection from suchrisks. In such industries, the bene-fits to integration may be small.When imperfect markets exist, incontrast, firms can benefit by somecombination of improved informa-tion access, internalized transac-tions costs, and efficiencies in mar-ket exchange (Perry, 1989). As aresult, firms tend to integratewhen the costs incurred in usingthe market price mechanismexceed the costs of organizingthose activities within the man-agement control of a single opera-tion (Scherer).

While vertical integration can leadto reduced risks and/or enhancedprofits for some firms or growers,others may find such a strategyunattractive. Depending on thesize of the firm and the extent ofthe proposed integration, the bene-fits associated with specializationand scale economies can be greatlyreduced or lost, particularly in per-fectly competitive markets. Forgrowers in such markets whochoose to vertically integrate, thegain may be primarily through

enterprise (or business) diversifi-cation (Perry, 1989). In addition,the size and scope of the operationcan have a major impact on inte-gration choice.

Empirical applications have exam-ined the linkage between verticalintegration and farm-level risk.One such study, focusing on cattleproduction in the Texas rollingplains, illustrates the importanceof size of firm and income growthon integration choice (Whitson,Barry, and Lacewell). This study,responding to concerns about priceuncertainty and the changingstructure of the livestock industry,evaluated the risk-return effects ofselling fed calves or holding themthrough subsequent stages in theproduction process. It included aweaned calf stage, a stocker phase(grazing on wheat pasture), a cus-tom feeding phase (bypassing thestocker phase and custom feeding),and other options.

The authors found that, at low-income levels, the preferredsequence involved production ofweaned calves with subsequentplacement in a feedlot, a resultconsistent with negative covari-ances. As growth in ranch incomeincreased, however, a wheat pas-ture activity was included in thevertical sequence to increaseincome and meet increased cashflow requirements over a 5-yearhorizon. The manager’s willingnessto accept risk and constraints tohis or her ability to borrow werecritical in determining the finalintegration choice.

PPrroodduuccttiioonn CCoonnttrraaccttss

Production contracts typically givethe contractor (the buyer of thecommodity) considerable controlover the production process (Perry,1997). These contracts usuallyspecify in detail the productioninputs supplied by the contractor,the quality and quantity of a par-ticular commodity that is to be

The benefits associ-ated with integra-tion typicallyincrease as produc-tion and marketingrelationshipsbecome more com-plex and when break-downs in market-place competitionare most likely.


How

Farm

ers

Can

Man

age

Ris

k

delivered, and the compensationthat is to be paid to the grower. Asan example, a broiler integrator(contractor) usually retains controlover the chicks as they are raisedby the producer, as well as pre-scribes specific inputs and specialmanagement practices throughoutthe production cycle. In return forrelinquishing control over decision-making, growers—particularly hogand broiler growers—are typicallycompensated with an incentive-based fee. According to USDA’sAgricultural Resource Manage-ment Study, commodities valued atapproximately $18 billion wereproduced under production con-tracts in 1997.

Firms commonly enter into pro-duction contracts with farmers toensure timeliness and quality ofcommodity deliveries, and to gaincontrol over the methods used inthe production process. Productioncontracting is particularly favoredwhen specialized inputs and com-plex production technologies areused, and the end product mustmeet rigid quality levels and pos-sess uniform characteristics.Production contracting is alsofavored when oversupply andundersupply have been problems,the risk-return tradeoffs areadvantageous to both the producerand the contracting firm, produc-tion technologies are specific, uni-form, and knowledge-based, cen-tralized management is feasible,

and the commodity is highly per-ishable (Kliebenstein andLawrence; Barry, Sonka, andLajili; Farrell; Harris). In addition,integrators may prefer to keepfixed capital assets (such as build-ings) off of their balance sheets forliquidity purposes (Barry).

Because the broiler industry pos-sesses many of these attributes,production contracting in thisindustry is particularly common.Indeed, about 99 percent of thevalue of broiler output was pro-duced under production contractsin 1997 (table 5). Such contractingis also commonly used in the eggand hog industries. For hogs, theuse of production contracts hasincreased rapidly in the past 5years, as the number of large, spe-cialized operations has acceleratedand size economies and new healthtechnologies have encouragedgreater concentration of animals.These contracts ensure that pack-ers receive a consistent supply ofhigh-quality hogs, allow processingfirms to operate at close to optimalcapacity, and allow the marketingsystem to be more responsive thanin the past to changes in consumerpreferences (Martinez, Smith, andZering).

Two basic types of production con-tracts are used, which differ in theamount of control, risk, and uncer-tainty the buyer and sellerassume: production management

Production contract-ing is most common-ly found in the broil-er, egg, and hogindustries.


How

Farmers C

an M

anage R

isk

2211

Table 5—Value of selected commodities produced under production contracts, 1997

Value of production underCommodity production contracts

Percent Million dollarsBroilers 99 6,664Cattle 14 4,280Eggs 37 773Hogs 33 3,097Vegetables 8 1,145

Total value of production under production contracts, all commodities1 12 18,215

1Includes $1,627 million in the crop category and $16,588 million in the livestock category. The total value ofagricultural production is $191,724 million.

Source: USDA, ERS, 1997 Agricultural Resource Management Study, special analysis.

contracts and resource-providingcontracts. The production manage-ment contract is commonly usedfor processing vegetables (sweetcorn, snap beans, and green peas).With these contracts, the buyergains additional control over deci-sions that would be made solely bythe grower in the absence of a con-tract, including planting schedulesand seed varieties (Powers). Byassuming this degree of control,the contractor increases the likeli-hood of receiving a commodity thathas specifically desired character-istics, and nearly all price risk isshifted to the contractor throughthe establishment of an agreed-upon price upon entry into thecontract. Some price risk remainswith the farmer, however, due toquality considerations, which mayresult in either a discount or a pre-mium relative to the establishedcontract price (Lucier). When cropsfail, growers receive no paymentunder these contracts and, hence,bear the production risk associatedwith crop shortfalls.

The second type of contract is theresource-providing contract, whichusually offers contractors a greaterdegree of control than do produc-tion management contracts. Thesecontracts are often used when spe-cialized inputs and managementare required to ensure final-prod-uct attributes, and are particularlycommon in the broiler industry.Under such an arrangement, thebroiler producer generally providesland, housing facilities, utilities,and labor, and covers operatingexpenses (repairs, maintenance,and manure disposal). The contrac-tor usually provides the chicks,feed, veterinary services, manage-ment, and transportation.Significant production decisions—such as the size and rotation offlocks, the flock’s genetic charac-teristics, and the capacity of chick-en houses—are made by the inte-grator (Perry, 1997). Thus, thegrower is essentially a custodian of

the production operation for theintegrator.

Much attention in recent years hasbeen focused on contracting in thebroiler industry and the implica-tions for producers’ risks andreturns. In this industry, paymentsare based on a grower’s perform-ance efficiency relative to all grow-ers in his or her group or “round,”and involve two components. Thefirst component is the “base pay-ment,” which is a fixed amount perpound of live meat produced. Thesecond component is the “incentivepayment,” which depends on thegrower’s efficiency in feed conver-sion, the poultry mortality rate forthat grower, and the weight of thatgrower’s finished birds relative toall growers in the round. Thesefactors are weighted in a calcula-tion that determines the grower’s“settlement cost.” If the settlementcost for all contractor flocks har-vested within a specified period inthe round is greater than the indi-vidual grower’s cost, he or shereceives a bonus. In contrast, apenalty is incurred if the grower’ssettlement cost is high relative toall other farmers in the round.

Several important risk-related fea-tures are associated with these“relative performance” contracts.Because grower payments dependlargely on production outcomes—and not feed or broiler prices—growers do not explicitly bear anyprice risk. Furthermore, the rela-tive nature of the contracts meansthat, in the presence of favorablegrowing conditions (like idealweather), the costs of all growersin the round are lower and, hence,no single grower receives a largerper pound payment. Althoughgrowers do not bear this type of“common” production risk faced byall growers in their round, they dobear the “idiosyncratic” risk specif-ic to their operation. For example,an operation that experienced anunusual disease outbreak and ahigher mortality rate would have a

Two basic types ofproduction contractsare used: productionmanagement con-tracts and resource-providing contracts.


How

Farm

ers

Can

Man

age

Ris

k

higher settlement cost—and apenalizing incentive payment—rel-ative to other growers in theround. Because of these factors,this type of production contractshifts price and common outputrisk from individual growers tointegrators, with growers retainingthe idiosyncratic risk specific tothe efficiency of their own opera-tion (Knoeber and Thurman).

The risk implications associatedwith production contracts arehighly conditional on the specificcontract terms. One recent studyof the broiler industry examinedthe risk-shifting associated withthe type of relative-performancecontract (a “contract with rounds”)discussed previously, comparingthe results to a “contracts withoutrounds” situation and to an “inde-pendent grower” situation(Knoeber and Thurman). Theydefined the payment in the con-tract-without-rounds case as afixed payment plus the amount bywhich the grower’s feed conversionperformance varies from a fixedstandard that did not change overtime. The “independent” caseassumed that the grower pur-chased inputs and sold broilers atmarket prices, and had not con-tracted with an integrator.

In that analysis, 89 percent of thegrowers realized risk reductionthat was significantly greater inthe relative performance contractsituation than in the contract-without-rounds case. This isbecause relative production con-tracts eliminate the risk commonto all growers in the round as wellas price risk, leaving only the idio-syncratic risk specific to produc-tion on a given operation. In con-trast, the without-rounds con-tracts—where payments are madeon a fixed standard representingthe average settlement cost for allgrowers for the entire sample—eliminate only price risk. Knoeberand Thurman also concluded thatrelative and without-rounds con-

tracts reduced risk by 97 and 94percent, respectively, comparedwith the independent grower case.In both situations, the reduction inprice risk accounted for the majorrisk-shifting component.

More recently, Martin analyzed therisk reduction associated with pro-duction contracts in the NorthCarolina pork industry. Martin’sresearch found that the risk-shift-ing capacity of relative productioncontracts was significantly greaterfor 36-70 percent of contract grow-ers compared with without-roundscontract growers. This is weakerevidence than realized by Knoeberand Thurman, and may partly beexplained by the greater homo-geneity associated with broilerproduction. Broilers usually areboth placed on farms and market-ed at uniform weights, while hogsmay be placed on farms at 30-60pounds and marketed at weightsvarying from 220-280 pounds.Because there is less output varia-tion common to all growers in thehog industry, relative contractshave less of an impact on growerrisk when compared with without-rounds contracts. Martin alsofound that without-rounds con-tracts shifted 90 percent of thegrower’s income risk to the inte-grator when compared with theindependent-grower situation,with 93.5 percent of income riskshifted in the case of relative pro-duction contracts.

In addition to risk-shifting capaci-ty, production contracts have otheradvantages for growers, as well asfor contractors. For contractors, theuse of production contracts canresult in sufficient input supplycontrol (without the need for verti-cal integration), as well asimproved response to consumerdemand. Evidence suggests thatfarmers enter production contractsto guarantee market access,improve efficiency, and ensureaccess to capital (Perry, 1997). Mostproduction contracts lower farmers’

One study of thebroiler industry con-cluded that “relative-performance’” pro-duction contractsreduced income riskby 97 percent, com-pared with the inde-pendent grower case.


How

Farmers C

an M

anage R

isk

2233

price risks when compared withrisks on the open market. The com-bination of lower market risks andless variable incomes was a majorreason cited by farmers for usingproduction contracting in at leastone survey (Rhodes and Grimes).This suggests that farmers are wellaware of the risk-shifting capacityjust illustrated. In addition,depending on contract terms, farm-ers can benefit from technicaladvice, managerial expertise, andaccess to technical advances (suchas high-quality breeding stock)that may not otherwise be readilyavailable (Perry, 1997).

Despite such advantages, however,production contracting has beencriticized. Some observers arguethat production contracting canlimit the entrepreneurial capacityof growers, and others cite therisks of contract termination onshort notice (Hamilton; Charlier;Harris). Contractors may requireupgrades to buildings and otherinfrastructure that are unexpectedby the grower, resulting in aninvestment risk. In addition, somegrowers under a relative perform-ance system believe that they areat an unfair disadvantage, arguingthat companies may not have theincentives to maintain strict accu-racy in the accounting and alloca-tion of inputs among growers, andthat absolute standards (as in the“contracts-without-rounds” case)may be most equitable and trans-parent (Jenner). Issues betweengrowers and integrators have led tolawsuits on various occasions, andIowa, Kansas, and Minnesota haveadopted some form of legislationregulating production contractingin agriculture (Johnson and Foster;Plain; Hamilton and Andrews).

MMaarrkkeettiinngg CCoonnttrraaccttss

Marketing contracts are either ver-bal or written agreements betweena buyer and a producer that set aprice and/or an outlet for a com-modity before harvest or before the

commodity is ready to be marketed(Perry, 1997). Since ownership ofthe commodity is generallyretained by the grower while thecommodity is produced, manage-ment decisions (such as varieties orbreeds, or input use and timing)typically remain with the producer.This latter characteristic—respon-sibility for management deci-sions—is critical in distinguishingmarketing contracts from produc-tion contracts (table 6).

Marketing contracts can takemany forms. They are at timesused by grain farmers to forwardprice a growing crop with a coun-try elevator, where they arereferred to as cash forward con-tracts. The contract terms varyacross contracts, but typicallyestablish a price (or contain provi-sions for setting a price at a laterdate) and provide for delivery of agiven quality (or grade) within aspecified time period. A “flat price”(or fixed price) forward contractmay, for example, state that afarmer will deliver 10,000 bushelsof No. 2 yellow corn to the localelevator at harvest for a price of$3.25 per bushel. Premiums anddiscounts may be established forgrain that does not meet specifiedquality standards. Flat price con-tracts are one of the most commontypes of forward contracts. Theprice typically is the elevator’s“bid price” for all farmer-deliveredgrain. This “bid price” is based ona current futures quote, less a“basis” adjustment that reflectsmarketing costs between the localelevator and the futures exchangelocation.9

One distinguishingdifference betweenmarketing contractsand production contracts is that pro-ducers using market-ing contracts takegreater responsibili-ty for managementdecisions.


How

Farm

ers

Can

Man

age

Ris

k

9Country elevators entering into suchmarketing contracts generally hedge theirpositions using futures markets. Hedgingprovides an offset to any price-levelchanges associated with the marketing con-tracts that elevators negotiate with produc-ers, and transfers price-level risk to basisrisk (uncertainty in the relationshipbetween futures and cash prices). See laterdiscussion of price-level and basis consider-ations in the "hedging" section.

The variety of marketing contractsavailable to grain farmers hasincreased over time, and nomen-clature and contract terms vary bylocation. Basis contracts, for exam-ple, are another type of marketingcontract, which provide for theprice to be determined by applyinga specified difference (basis) to aparticular futures contract price tobe observed later, usually whendesired by the farmer. This assuresthe farmer an outlet, for later own-ership transfer, while allowinggains or losses from changes in thefutures price. Other contractstransfer ownership immediatelywhile postponing payment, such asdelayed payment and delayedprice contracts. These contractsmay offer farmers tax advantages,while allowing the elevator to shipthe grain and open up storagespace. Among these contracts,delayed (deferred) payment con-tracts specify the price to be paid,while delayed price contracts(sometimes called “deferred price”

or “price later” contracts) providethat price will equal the elevator’sbid price, or the futures priceadjusted for basis at a time select-ed by the farmer.

In contrast, minimum-price con-tracts guarantee the producer aminimum price for harvest delivery,based on futures price quotes at thetime the contract is established,with the incorporation of a pricingformula that gives farmers theopportunity to sell at a higher priceif futures prices increase before thecontract expires (Catania). Whenhedge-to-arrive (HTA) contracts areused, the futures price is fixed inthe contract, but the basis is leftundetermined until a later time.HTA’s have effects similar tofutures hedges for the farmer,except that no commissions or mar-gin deposits are required and thefarmer deals with a local buyerinstead of a broker. (For a compre-hensive listing of the different

Marketing contractscan vary in theirtreatment of pricelevel risk and basisrisk.


How

Farmers C

an M

anage R

isk

2255

Table 6—Comparison of marketing and production contract characteristicsMarketing contracts Production contracts

Contractor: Contractor:

Buys a known quantity and quality of Arranges to have a specific quality and the commodity for a negotiated price quantity of commodity produced(or pricing arrangement)

Doesn’t own the commodity until it’s Usually owns the commodity being delivered produced

Has little influence over production Makes most of the production decisionsdecisions

Contractee (operator): Contractee (operator):

Has a buyer and a price (or pricing Provides a service and other fixed arrangement for commodities inputs (land, buildings, etc.) for a feebefore they are harvested)

Supplies and finances all or most Supplies a small part of the totalof the inputs needed to produce production inputs neededthe commodity

Owns the commodity while it’s being Usually does not own the commodityproduced

Makes all or most production decisions Makes few, if any, of the production decisions

Assumes all risks of production but reduced Bears few price or market uncertaintiesprice risk and limited production risks

Receives largest share of total value Receives a fee for production that doesof production not reflect the full market value of the

commoditySource: Excerpted by ERS from USDA, NASS, Costs and Returns Report Interviewer’s Manual, December 1996.

types of contracts and areas of riskexposure, see Kemp; Wisner.)

Most types of contracts do notcompletely eliminate price risk(table 7). The exception is flat-price contracts, which establish anexact price to be paid to the grow-er upon delivery and thus com-pletely eliminate price risk. In con-trast, nearly all other forwardmarketing contracts fix either thebasis (for example, 10 cents underthe Chicago Board of TradeNovember soybean contract) or thelevel of the futures price at thetime the contract is negotiated, butnot both. For those contracts thatestablish the basis, the risk ofprice-level variation is retained bythe producer until the time of cropdelivery and final sale. Conversely,fixing only the futures price in thecontract (as with HTA contracts)leaves the farmer with basis risk.When HTA’s are rolled over to suc-cessive months, the producer alsoincurs the risk associated withspreads across different futurescontract months.

Farmers who forward contract agrowing crop bear yield risk inaddition to price risk. As a result,farmers generally are advised toforward price substantially lessthan 100 percent of their expectedcrop until yields are well

assured.10 Difficulties associatedwith overcontracting arise if poorweather results in low yields, andproducers contracting a large pro-portion of their crop need to buy“replacement” bushels at an uncer-tain cash price to meet the termsof delivery on their forward con-tract. If the farmer’s shortfall iscaused by a severe drought, cashprices at the time such replace-ment bushels must be purchased,either as part of contract renegoti-ation or to meet delivery terms,may be quite high. When the cropsize is known, producers can safelyforward contract up to 100 percentof their crop. Deferred (or delayed)price contracts usually are notnegotiated until the grain is deliv-ered to the country elevator (whenyield risk no longer exists). Whensuch contracts are used, however,both futures price and basis riskstypically are faced by the producer(table 7).

To illustrate the risks associatedwith a simple forward contractingsituation, consider a farmer withirrigated corn acreage whoexpects, with considerable certain-ty, a crop of 50,000 bushels. This

Table 7—Most marketing contracts do not completely eliminate market risk 1

Market risks remainingType of marketing Guaranteed

contact Futures Basis Spread minimumFlat price2 ✔Basis ✔Deferred price ✔ ✔3

Minimum price ✔4 4 ✔

HTA (basic) ✔HTA (multiple crop) ✔ ✔

HTA= Hedge-to-arrive contract.1The extent to which yield risk is an issue depends on when the contract is entered. Deferred price contracts,for example, can be entered into before the crop is harvested (when yield risk is an issue) or after harvest(when crop size is known and yield risk is zero).2For definitions of the different types of contracts, see the glossary.3Some deferred price contracts specify the basis, which eliminates basis risk for farmers.4Downside risk is eliminated.

Source: Adopted by ERS from Kemp, Todd E., editor, Hybrid Cash Grain Contracts: Assessing, Managing, andControlling Risk, a White Paper by the Risk Evaluation Task Force on Hybrid Cash Contracts, National Grainand Feed Association, Washington, DC, 1996.

Farmers generallyare advised to for-ward price substan-tially less than 100percent of theirexpected crop untilyields are wellassured.


How

Farm

ers

Can

Man

age

Ris

k

10The minimum-risk amount to cash for-ward contract typically is slightly largerthan the minimum-risk futures hedgebecause basis risk is avoided. See subse-quent discussion of optimal hedging.

farmer is considering two optionsand plans to finalize his decisionin a few days (the end of April),just before planting time. Theoptions involve either (1) sellingthe crop in October, right afterharvest, for the cash market price,or (2) entering a flat price contractwith the local elevator for Octoberdelivery. The probability distribu-tion of cash prices in October,based on the farmer’s expectations,is shown in figure 6, and the eleva-tor at the time of the decision isoffering flat price contracts forharvest delivery. The fixed priceoffered under the contract isshown by the vertical line. The fig-ure illustrates both the downsideprotection and the upside potentialforgone if the farmer chooses thecontracting approach. Because thefarmer in this simple example hasno yield uncertainty, 100 percentforward contracting completelyeliminates revenue risk (since theprice for the entire crop is fixed). Ifthe producer forward contractsand cash prices turn out higher atharvest than the contract price, aprofit opportunity has been lost,but income is no less than expect-ed. (For other examples, see Jolly.)

Marketing contracts are used notonly for pricing field crops, butalso in the specialty crops sector(table 8). Among specialty crops,they appear most often in growersales of fruits, fresh pre-cut veg-etables, and processed vegetables(Powers). In the pre-cut vegetableindustry, for example, prepackagedshredded lettuce and cabbage,diced celery, and sliced carrots areused in large quantities by foodestablishments, institutions, andretailers. Purchasing pre-cut freshvegetables can reduce costs tothese buyers if sufficiently lesslabor is required in food prepara-tion. These buyers usually negoti-ate marketing contracts that speci-fy a tentative free-on-board price,quality, and delivery schedule for a6- to 12-month period. The tenta-tive contract price is based on theexpected price during the agree-ment’s duration, and may be rene-gotiated if warranted by marketconditions (Powers).

Although less common, marketingcontracts are also used in the live-stock sector. Fed cattle, for exam-ple, can be forward priced using acash forward contract—an agree-ment, in this case, by a cattle feed-

Marketing contractsare used not only forpricing field crops,but also in the spe-cialty crops and live-stock sectors.


How

Farmers C

an M

anage R

isk

2277

Probability density

Price

Fixed price

Downside potential

Source: Hypothetical example developed by ERS.

Forgone upside potential

Figure 6

Flat price contracts provide downside protection, but forgoupside potential

er to deliver a specified number ofcattle, in a designated futuremonth, to a packer. Two types ofcontracts are typically used.Similar to the situation for fieldcrops, “flat price” contracts specifythe price at the time the contractis negotiated between the two par-ties. In contrast, “basis” contractsspecify the basis level (the cashprice minus the futures price) atthe time the contract is signed,allowing the cattle feeder to waituntil a later time to fix the futuresprice, perhaps after the price levelhas increased. The final contractprice is calculated by adding thebasis specified in the contract tothe futures price on the day thecattle are priced (Elam).

Although the price risks associatedwith the different types of market-ing contracts vary, some types ofrisks are common to all contracts.All growers who enter into forwardcontracts face the risk of default bythe merchandiser offering the con-tract, who may be unable to meetthe financial obligation associated

with the contract. In addition, pro-ducers and merchandisers mayhave different understandings ofthe contract terms and the poten-tial financial impact (Kemp). As aresult, legal risks may be confront-ed by farmers, as was the case fornumerous individuals who hadentered into HTA contracts in 1995.

Few empirical analyses haveexamined the reduction in pricerisk associated with the use of dif-ferent marketing contracts, partlybecause time-series data capturingcontract prices are generally notavailable to researchers. However,futures price data for field cropsand livestock are readily availablefrom exchanges. Because of thestructure of most forward con-tracts for field crops and livestock,the use of futures prices in esti-mating optimal hedging amountsand risk reduction provides a closeapproximation to the analogousestimates that would exist in ananalysis of many forward contract-ing situations.

One key issue forproducers and contractors is thatboth understand theterms of the contract.


How

Farm

ers

Can

Man

age

Ris

k

Table 8—Value of selected commodities produced under marketing contracts, 1997

Value of production underCommodity marketing contracts

Percent Million dollars

Barley 19 167Canola 46 44Cattle 9 2,735Corn 8 1,720Cotton 33 1,923

Dry edible beans 3 82Eggs 6 131Fruits 59 7,199Oats 3 8

Peanuts 41 372Peas 9 4Potatoes 43 720Rice 31 573Sorghum for grain 6 86

Soybeans 9 1,672Sugar beets 82 973Sunflowers 8 51Vegetables 24 3,374

Total value of production under marketing contracts, all commodities1 22 41,6101Includes $21,323 million in the crop category and $20,287 million in the livestock category. The total value ofagricultural production is $191,724 million.


HHeeddggiinngg iinn FFuuttuurreess

Futures contracts provide farmers(as well as processors, merchandis-ers, and others) with a method forreducing their risks. Futures con-tracts were almost exclusivelytraded on commodity prices in thepast, although innovations inrecent decades also have intro-duced contracts on interest rates,foreign exchange rates, priceindexes, and crop yields. A primaryuse of futures involves shiftingrisk from a firm that desires lessrisk (the hedger) to a party who iswilling to accept the risk in ex-change for an expected profit (thespeculator). Also, hedgers withopposite positions in the markettrade with each other, and specula-tors with opposing views of themarket may also trade.

A futures contract is an agreementpriced and entered on an exchangeto trade at a specified future time acommodity or other asset with spec-ified attributes (or in the case ofcash settlement, an equivalentamount of money). The U.S.exchanges that trade agriculturalfutures contracts are the ChicagoBoard of Trade; the ChicagoMercantile Exchange; the KansasCity Board of Trade; theMinneapolis Grain Exchange; theNew York Coffee, Sugar, and CocoaExchange; and the New YorkCotton Exchange. Trading is con-ducted either through “open outcry”on the floor of the exchange or elec-tronically. The December corn con-tract traded on the Chicago Boardof Trade, for example, specifies lotsof 5,000 bushels for No. 2 yellowcorn and a December delivery peri-od. Contracts for major field crops(including corn, wheat, soybeans,cotton), four types of livestock andanimal products (live cattle, feedercattle, live hogs, and pork bellies),and sugar and frozen concentratedorange juice have been traded foryears. More recently, futures con-tracts for rice, boneless beef, anddairy products have been intro-

duced. Because contracts are stan-dardized, the only issue to be nego-tiated at trading time is price.Enforcing contract terms is a keyfunction of the exchanges wheretrading occurs, and guaranteeingcontracts is a key function of theexchange clearinghouse.

Most futures contracts are offset byopposite trades before delivery time,with each party to the transactionselling (or buying) a futures con-tract that was initially bought (orsold). For example, if a farmer(through his or her brokerage houseand its trader on the Chicago Boardof Trade) sells a corn contract inMay for December delivery, his orher position may be offset by buyinga December corn contract at anytime before the end of the deliveryperiod, which is about December 20.Such an offset usually occursbecause the major motive in tradingfutures is to hold a temporary posi-tion, and then trade for money, andnot to physically deliver or acquirea commodity (Hieronymus). Mosthedgers offset because making ortaking delivery on futures would bemore costly than delivering throughnormal channels, while speculatorsgenerally do not want to own theactual commodity.

Because futures contracts are com-mitments to trade in the future,actual delivery and payment arenot required until the contractmatures. However, both buyers andsellers are required to make mar-gin deposits with their brokers toguarantee their respective commit-ments. Because the margin depositis small (typically 5-10 percent ofthe underlying value of the con-tract), speculators (who provide liq-uidity) are attracted to the market.The exchanges set minimum mar-gins by contract, which can beraised by brokers to provide theprotection they deem necessary.Using the December corn contractas an example, and assuming a$2.00 per bushel price quote, a cat-tle feeder who buys one contract

Most futures con-tracts are offset byopposite tradesbefore delivery time,with each party tothe transaction sell-ing (or buying) afutures contract thatwas initially bought(or sold).


How

Farmers C

an M

anage R

isk

2299

(5,000 bushels) makes a $10,000commitment. With a 10-percentmargin, the feeder must post$1,000 with his or her broker. A“margin call” occurs when the priceof the contract moves against thetrader, say to $1.90 in this exam-ple. When a margin call occurs, theproducer must post additional mar-gin with his or her broker to coverthe loss and restore the deposit.Similarly, when the price movesfavorably for the trader by a speci-fied amount, money can be with-drawn from the margin deposit.

Because futures prices reflect val-ues of commodities at futuresdelivery points, the local cashprices confronted by farmers usu-ally vary from futures contractquotes at a given point in time.The differences between futuresand cash prices are termed “basis,”and reflect differences in priceacross space (due to transportationcosts), time (which are associatedwith storage costs), or quality(such as differences in protein pre-miums for wheat). The basis is cal-culated as the difference betweenthe cash price (at a given locationand at a given point in time) andthe futures price (associated witha specified exchange and contractmonth).11 The basis varies overtime, and reflects only transporta-tion costs and quality differencesas the contract reaches maturity.As seen below, hedging largelyeliminates price level uncertainty,but not basis uncertainty, whichgenerally has a smaller variance.

Two categories of hedging exist:“long” hedging (where a futurescontract is purchased) and “short”hedging (where a futures contract

is sold). Either type of hedgeinvolves holding a futures positionin anticipation of a later transac-tion in the cash market, and inboth cases, the futures position isopposite to the cash position.Because futures and cash pricestend to move up and down togeth-er, losses and gains in the twomarkets tend to offset each other,leaving the hedger with a returnnear what was expected (the ini-tial futures price plus the end-of-period basis—see later discussion).Thus, hedging helps protect thebusiness from changes in price lev-els. Farmers may choose to hedgein many different situations,including the following:

• Storage hedging—Farmers ormerchants who own a commodi-ty can protect themselves fromdeclines in the commodity’sprice by short hedging. Thisinvolves selling futures con-tracts as the commodity is har-vested or acquired, holding theresulting short futures positionduring the storage period, andbuying it out when the cashcommodity is sold. Losses(gains) in the value of the cashcommodity due to unexpectedprice changes will be largely off-set by gains (losses) in the valueof the futures position leavingthe owner of the commoditywith approximately the expectedreturn from storage.

• Production hedging—Crop andlivestock producers can protectthemselves from declines inprices of expected outputs byshort hedging. This generallyinvolves selling futures con-tracts at the beginning of orduring the growing or feedingperiod, holding the resultingshort futures position until theproduct is ready to sell, and buy-ing the futures as the output issold. Losses (gains) in the valueof the output due to unexpectedprice changes tend to be offsetby gains (losses) in the value of

Two categories ofhedging exist:

• “long” hedging—where a futurescontract is pur-chased.

• “short” hedging—where a futurescontract is sold.


How

Farm

ers

Can

Man

age

Ris

k

11Basis is sometimes calculated as cashprice minus futures price and sometimes asfutures price minus cash price, as mutuallyunderstood by the parties involved. Cashprices may be quoted relative to thefutures, such as "10 cents over" or "20 centsunder" the futures price. In this report,basis is calculated as cash minus futures,so that an "under" basis has a negative signand an "over" basis has a positive sign.

the futures position. However,yield variability reduces therisk-reducing effectiveness ofhedging for crop growers andgenerally makes it inadvisableto sell futures equal to morethan one-half to two-thirds ofthe expected crop.

• Hedging expected purchases—Livestock feeders anticipatingthe purchase of corn or feedercattle can protect themselvesfrom price increases by longhedging. This involves buyingcorn or feeder cattle futures con-tracts to match anticipatedrequirements and selling theresulting long futures positionsas these inputs are purchasedon the cash market. Increases(declines) in the cost of feedersor feed due to unexpected pricechanges will be partly offset bygains (losses) in the value of thefutures position leaving thefeeder with approximately theexpected costs of inputs.Feeders’ overall price risks maybe further reduced by sellingfutures on prospective outputs,as discussed above.

To better understand the impor-tance of basis risk in hedging, con-sider the example of a corn produc-er with irrigated acreage who isconsidering the pricing of hisgrowing crop. Because the produc-er irrigates and faces few othernatural perils, he knows the size ofhis crop with a great deal of cer-tainty and is concerned only withprice risk. If the farmer does nothedge, his risk is solely associatedwith the harvest cash price (P2),which can also be calculated as theharvest futures price (F2) plus theharvest basis (B2). Thus, thefarmer’s net return in a cash-sale-at-harvest situation (Ru) can becalculated as the cash price (F2 +B2) at harvest multiplied by actualproduction (Y2 ), minus productioncosts (C):

Ru = [(F2 + B2) * Y2 ] - C.

Suppose now that the producerplaces a short hedge (for example,sells a futures contract) to reducethe risk of a price decline and alower sales price for his growingcrop. The expected final net returnat harvest (Rh) is based on thecash price at harvest (F2 + B2),and the profit or loss associatedwith the farmer’s futures marketposition (F1 - F2). The farmer’sactual level of production is desig-nated as Y2 in the following equa-tion, and the quantity hedged is h * Y1, where h is the hedge ratioand Y1 is expected production:

Rh = [(F2 + B2) * Y2 ] +[(F1 - F2) * (h * Y1)] - C.

Assuming that output is knownwith certainty at the time thehedge is placed, and that actualproduction equals the quantityhedged (for example, Y2 = h * Y1),gives the following:12

Rh = [(F2 + B2) * Y2] + [(F1 - F2) * Y2] - C,

or:

Rh = [Y2 * (F1 + B2 )] - C.

This last equation indicates thatthe price component of the farmer’snet return depends on the futuresprice at the time the hedge isplaced plus the harvest basis.Because the futures price is knownwith certainty at planting, and out-put is known with certainty in thisexample, the only risk faced by thefarmer is the risk associated withthe harvest basis. Thus, price levelrisk is eliminated by this anticipa-tory hedge, and the only risk facedby the grower is basis risk (theuncertain nature of B2).

The existence of basis risk is a keyfactor distinguishing the risk asso-ciated with futures hedging andthe use of many types of cash for-

When output isknown with certain-ty, the hedger elimi-nates price level riskand only confrontsbasis risk.


How

Farmers C

an M

anage R

isk

3311

12This example also applies to a storagehedge, where output is known with certainty.

ward contracts (see previous sec-tion). When a producer enters intoa “flat price” forward contract withhis or her local elevator, for exam-ple, the basis risk he or she facesis zero. In addition, forward con-tracts are generally less standard-ized than futures contracts, andspecific terms may vary across ele-vators. Physical delivery to thelocal elevator at harvest is gener-ally required, and no margin callsexist when cash forward contractsare used (table 9).

Using a numerical example toillustrate hedging, suppose thecorn producer discussed earlierwishes to reduce his income uncer-tainty by selling a futures contractat planting time. Because thefarmer irrigates his corn crop, heis not concerned about yield risk,and the hedge quantity is assumedto equal actual output. The farmerin this example observes a $2.75per bushel futures price at plant-ing time. He expects a harvestbasis of -$0.25, giving an expectedcash price of $2.75 plus -$0.25, or

$2.50 per bushel (table 10). Twooutcomes are shown in the table, a$0.25-price decrease betweenplanting and harvest, and a $0.25-price rise. In both cases, the real-ized harvest basis is -$0.25, asexpected.13 With hedging, thereturn per bushel is $2.50 in bothcases. This return can be calculat-ed as (1) the futures price at plant-ing time ($2.75) plus the harvestbasis (-$0.25 in both scenarios), or(2) the cash price at harvest ($2.25or $2.75, depending on the sce-nario) plus the gain or loss fromthe futures market position(+$0.25 or -$0.25).

Reality differs from the exampleillustrated above in that neitherbasis nor yields can be anticipatedwith certainty, even for an irrigat-ed farm. Table 11 illustrates suchyield and price risk for a 500-acrecorn farm. The corn farmer’s

Forward contractsare generally lessstandardized thanfutures contracts,and specific termsmay vary across elevators.


How

Farm

ers

Can

Man

age

Ris

k

Table 9—A comparison of “flat price” cash forward contracts and futures hedging

“Flat price”Characteristic cash forward contracts Futures hedging

Competitiveness of price Depends on margintaken by elevator Yes

Basis risk No Yes

Default risk Some No

Ease of recontractingor offset Depends Yes

Physical delivery Yes Seldom

Margin calls No YesSource: For more information, see Heifner, Richard G., Bruce H. Wright, and Gerald E. Plato, Using Cash,Futures, and Options Contracts in the Farm Business, AIB-665. U.S. Dept. Agr., Econ. Res. Serv., April 1993.

Table 10—Effects of hedging on a hypothetical corn grower’s return per bushelPrice decrease Price increase

Cash/futures price scenario scenario

Dollars per bushel

Cash price expected at harvest 2.50 2.50

Cash price realized at harvest 2.25 2.75

Futures price at planting 2.75 2.75

Futures price at harvest 2.50 3.00

Futures return to the producer +0.25 -0.25

Net price realized with hedging 2.50 2.50Source: Hypothetical example developed by ERS.

13The level of the basis varies with prox-imity to major markets, and in major corngrowing areas, tends to be between $0.10and $0.30 per bushel below the futuresprice at harvest.

return from a cash sale at harvest(with no short hedge) ranges from-$55,500 to +$70,000 across the sixscenarios. Now suppose that thefarmer weighs his alternatives anddecides to hedge in the spring. Heanticipates that his output will notlikely fall below 30,000 bushels inany given year, and thus considershis optimal hedge level to be sixcontracts (at 5,000 bushels percontract). At harvest time, he liftshis short hedge by buying back hisfutures contract, and sells his cashcrop in the marketplace. Hereceives the proceeds from thecash sale of the crop (less produc-tion costs of $150,000) plus thegains or losses associated with thefutures transaction (less commis-sion charges of $420). Over the sixscenarios, the producer’s netreturn, including the return to thehedge, varies from -$48,000 to+$68,080. The variability inreturns indicates that the returnto hedging is less variable (asmeasured by either the range ofoutcomes or the standard devia-tion) than is associated with acash-sale only strategy. Theexpected return over time, howev-er, is approximately identical inboth cases.

As can be seen by this example,the estimation of hedging amountsand risk reduction is much more

complicated in the presence ofyield risk. Generally, the effective-ness of hedging in reducing riskdiminishes as yield variabilityincreases and the correlationbetween prices and yields becomesmore negative. Although hedgingcan reduce income uncertainty formany farmers, it never completelyeliminates such uncertainty.

In addition to the considerationsdiscussed previously in this sec-tion, hedging involves possiblecosts for interest forgone on mar-gin deposits and for bias in futuresprices. These costs generally aresmall relative to the value of thepositions taken, but they partlyoffset the risk-reducing benefitsfrom hedging (see box on “TheCost of Forward Pricing”). The pos-sibilities that expected incomescan actually be increased by hedg-ing are discussed in a later sectionof this report.

An extensive literature addressesfarmer hedging. Much of this liter-ature analyzes the risk reductionassociated with hedging and thecalculation of optimal (generallyrisk-minimizing) hedge ratios,which specify the proportion of thecommodity that would be hedgedto minimize risk. When output iscertain, risk-minimizing hedgeratios are generally close to, but

The estimation ofhedging amountsand risk reduction ismuch more compli-cated in the presenceof yield risk.


How

Farmers C

an M

anage R

isk

3333

Table 11—Returns to a cash sale at harvest and a futures hedge for a hypothetical corn producerNet return

Harvest Spring Harvest Revenue from cash sale Net return DifferenceScenario Production cash price futures futures from crop at harvest with hedging in returns

Bushels Dollars per bushel Dollars

1 80,000 2.60 3.20 2.85 208,000 58,000 68,080 10,0802 80,000 2.75 2.90 3.00 220,000 70,000 67,000 -3,000

3 50,000 3.15 2.70 3.40 157,500 7,500 -13,500 -21,0004 50,000 2.80 3.15 3.05 140,000 -10,000 -7,000 3,000

5 30,000 3.15 3.65 3.40 94,500 -55,500 -48,000 7,5006 30,000 3.30 3.65 3.55 99,000 -51,000 -48,000 3,000

Averagenet return -- -- -- -- -- 19,000 18,580 -420

-- = Not applicable.Note: This illustration assumes that a producer has a 500-acre farm. The average yield for those acres varies across scenarios, from 160 bushels per acre inscenarios 1 and 2 to 100 bushels per acre in scenarios 3 and 4 to 60 bushels per acre in scenarios 5 and 6. The producer hedges six contracts at 5,000bushels per contract. The commission on the purchase and sale of the contract is $35. For a total of six round-turn transactions, the cost is 6 * $70, or $420.The farmer’s production costs total $150,000.


slightly less than, 1.0 (100 percentof the commodity would behedged).14 In contrast, when out-put is uncertain (such as would bethe case for a growing seasonhedge when a producer does notirrigate), the hedge quantity isusually substantially less than 100percent of expected output. Thissituation exists due to the correla-tion between random productionand random price. Because thesevariables are negatively correlatedin most cases, a “natural hedge”stabilizing revenue is inherent inthe system and the optimal strate-gy is to hedge a quantity lowerthan the producer’s expected out-put (McKinnon).

Many optimal hedge models exist,and vary in their assumptionsabout futures and cash price deter-mination, the unbiasedness offutures prices, risk aversion, andother factors (Berck; Miller andKahl; Rolfo; Plato; Ward andFletcher). In one study, Grant, forinstance, examined the risk-mini-mizing optimal hedge for corn andsoybean farmers in Iowa,Nebraska, and North Carolina inthe presence of price and yieldrisk. His findings indicated thatselling futures equal to 50-80 per-cent of expected production mini-mizes revenue risks for most cornand soybean producers in thoseStates. These hedges would, onaverage, eliminate less than 50percent of producers’ revenue vari-ance over the growing season dueto basis and yield risk. Becauseselling futures involves costs forcommissions and interest forgoneon margin deposits and a 20-per-centage-point reduction in theoptimal hedge has a relativelysmall effect on risk reduction,Grant concluded that the beststrategy may be for farmers tohedge between 30 and 50 percentof their expected crop.

Recently, studies have exploredother aspects of hedging. Lapanand Moschini examined the hedg-ing problem, and assumed thatprice, output, and basis are ran-dom variables. Their findings indi-cate that, unlike earlier resultsusing different approaches, theoptimal hedge depends on the levelof a producer’s risk aversion.Further, they conclude that theeffects of unhedgeable basis riskare exacerbated by yield risk.Other studies have, among othertopics, examined the effectivenessof futures in providing revenueprotection over a period of severalyears (Plato; Gardner, 1989).

Futures and options contractshave been traded on crop yields aswell as prices. The Chicago Boardof Trade introduced Iowa cornyield futures and options tradingin 1995. Subsequently, trading incorn yields for four other Corn BeltStates and the United States wasadded (Chicago Board of Trade,1998; Grenchik and Campbell).Yield futures contracts remainingopen at maturity are settled bycash payments based on USDA’sNational Agricultural StatisticsService (NASS) yield estimates.Although yield futures and optionscontracts provide potential hedg-ing vehicles for crop insurers aswell as farmers, trading volumehas been low. Relatively low corre-lations between individual farmyields and State yields limit theeffectiveness of such contracts inshifting farmers’ yield risks. Inaddition, insurers appear to findthe reinsurance provided by USDAthrough the Standard ReinsuranceAgreement under the Federal cropinsurance program satisfactory forshifting risk (Maurice; Lehman).

Recent research that examineshedging in both price and yieldfutures found that a risk-minimiz-ing firm can reduce its incomevariance by simultaneously hedg-ing in both price and yield futures(Vukina, Li, and Holthausen;

Futures and optionscontracts have beentraded on crop yieldsas well as prices.


How

Farm

ers

Can

Man

age

Ris

k

14When output is known, basis risk gen-erally causes the optimal hedge to be lessthan 1.0.

McNew; Heifner and Coble, 1996).Vukina, Li, and Holthausen foundthat such hedging in both types ofcontracts can be more effectivethan using price futures alone,

with the effectiveness of the two-instrument hedge depending onthe volatility of the yield contract.As the variance of the underlyingyield increases, the effectiveness of

Hedging involvescosts that appearmodest comparedwith the riskreduction for mostfarmers.


How

Farmers C

an M

anage R

isk

3355

The Cost of Forward Pricing

When hedging with futures, farmers must pay commissions and forgointerest or higher earning potential on money placed in margindeposits. Those who use cash forward contracts may incur such costsindirectly, to the degree that local buyers lower prices paid to covertheir hedging costs. Moreover, the prices obtained by hedgers may dif-fer from the price expected at delivery by the amount that speculatorsrequire as compensation for standing by to take hedgers’ trades and/orfor bearing risks.

Commissions for futures trading vary by brokerage house and by sizeof trade, but typically the commission to both buy and sell is less thanone-half percent of the value of the contract. Margin requirements alsovary. For example, if the margin deposit averages 10 percent for 6months, the interest forgone on the money at a 6 percent interest ratewould be 0.1 x 6/12 x 0.06 = 0.3 percent. Traders who meet marginrequirements by depositing government securities with their brokersavoid such interest costs, although they may sacrifice some income byholding securities that yield lower returns than could be earned withother investments.

Farmers avoid directly paying commissions or making margin depositsby forward contracting with local buyers. However, elevators and otherfirms who buy from farmers generally incur hedging costs that mustbe covered. Such firms usually avoid interest costs on margin depositsby depositing government securities. Moreover, forward contractingassures them a flow of commodities through their facilities, which mayincrease their returns. Thus, they may or may not seek to recover theircommission costs by paying slightly lower prices to farmers.

Futures markets facilitate forward pricing because short-term specu-lators (“scalpers”) are present to take the opposite sides of sell or buyorders as they arrive from traders outside the exchange. Scalpers thenquickly trade out of their positions with the expectation of a smallprofit, say one-quarter to one-half cent per bushel. This profit, whichcompensates scalpers for helping make the market liquid, constitutesa modest cost for hedgers.

Futures prices also may be less than expected cash prices becausespeculators require a risk premium for carrying hedgers’ risks. Keynessuggested that such risk premiums could be expected in marketswhere short hedging exceeds long hedging. Dusak later pointed outthat any such bias should be small because low transactions costsallow such risks to be spread very efficiently. Empirical studies indi-cate that futures price biases are small or nonexistent for most com-modities, particularly for the grains, where active long hedging helpsbalance short hedging (Zulauf and Irwin; Bessembinder; Kolb).However, the issue is not fully resolved.

In summary, forward pricing involves costs that appear modest com-pared with the risk reduction obtained for most farmers, but that maymake forward pricing more attractive to farmers who are very riskaverse.

hedging in both price and yieldfutures declines relative to hedg-ing in price futures alone. Theyalso concluded that hedging effec-tiveness depends critically on theprice and yield bases.

FFuuttuurreess OOppttiioonnss CCoonnttrraaccttss

A commodity option gives the hold-er the right, but not the obligation,to take a futures position at a speci-fied price before a specified date.The value of an option reflects theexpected return from exercisingthis right before it expires and dis-posing of the futures positionobtained. If the futures pricechanges in favor of the option hold-er, a profit may be realized eitherby exercising the option or sellingthe option at a price higher thanpaid. If prices move so that exercis-ing the option is unfavorable, thenthe option may be allowed toexpire. Options provide protectionagainst adverse price movements,while allowing the option holder togain from favorable movements inthe cash price. In this sense,options provide protection againstunfavorable events similar to thatprovided by insurance policies. Togain this protection, a hedger in an

options contract must pay a premi-um, as one would pay for insurance.

Options markets are closely tied tounderlying futures markets.Options that give the right to sell afutures contract are known as“put” options, while options thatgive the right to buy a futures con-tract are known as “call” options.The price at which the futures con-tract underlying the option may bebought (for a call option) or sold(for a put option) is called the“exercise” or “strike” price. As anexample, suppose a wheat producerpurchases a put option having astrike price of $3.00 per bushel. Iffutures prices move to $2.80, theoption may be exercised for a netprofit of $0.20 ($3.00-$2.80), minusthe premium paid for the option. Ifthe harvest cash price is $2.70 perbushel, the farmer’s return is $2.90per bushel ($2.70 plus $0.20),minus the premium.

The effects on realized returnsfrom hedging with futures and putoptions are compared for a rangeof possible futures price outcomesin figure 7. In this example, corn isstored in November and sold inMay, output risk is absent, and the

Options provide pro-tection against unfa-vorable events simi-lar to that providedby insurance policies.


How

Farm

ers

Can

Man

age

Ris

k

2.20 2.40 2.60 2.80 3.00 3.20 3.402.00

2.20

2.40

2.60

2.80

3.00

3.20

Realized return, $/bu.

Futures price at end of storage period, $/bu.

Put option hedge


Cash sale at end of storage period

Figure 7

Effects of futures and options hedging on exposure to price variation at marketing time, a storage example

Futures hedge

hedge ratio is 1.0. The May futuresprice is $2.80 per bushel at thebeginning of the storage periodand the expected May basis is -$0.20. By hedging with futures,the farmer obtains an expectedreturn for the corn in storage of$2.80 - 0.20, or $2.60. Alternative-ly, the farmer can buy an at-the-money put option with a $2.80strike price for a $0.20 premium.The put guarantees a price equalto the strike price, minus the pre-mium, minus the basis, or $2.80 - 0.20 - 0.20 = $2.40, while allow-ing the farmer to gain if thefutures price rises above $3.00 inMay. By not hedging, the farmergets the futures price minus thebasis. The figure shows that therange of possible prices is greatestwith the cash sale and least withthe futures hedge. Unlike futureshedging, the put does not limit thepotential profits associated withincreasing prices, but the pricemust rise more than the premiumcost before a profit is realized.

The premium paid for an optiontypically consists of “intrinsic”value and “time” value. The intrin-sic value reflects the differencebetween the underlying futuresprice and the strike price. If theprice of the underlying futurescontract is $2.90 per bushel, forexample, and the strike price is$2.70, then the holder of a calloption could gain $0.20 by exercis-ing the option immediately.Consequently, the premium in thiscase must be at least $0.20 perbushel, and the option is “in themoney.” If the strike price is abovethe futures price, the intrinsicvalue of the call option is zero andthe put is said to be “out of themoney.” When the strike priceequals the futures price, the optionis “at the money.”

The time value of an option, incontrast, depends on several fac-tors, including the volatility of theunderlying futures contract, thetime until the option expires, the

interest rate, the strike price, andthe underlying futures price. Timevalue refers to the money thatbuyers are willing to pay for thepossibility that the intrinsic valueof an option will increase overtime. An option on a futures con-tract with very low volatility, forexample, will have a small timevalue because traders do notexpect the intrinsic value tochange to a great extent over time.If the futures price is volatile, incontrast, the probability is highthat the option’s intrinsic valuewould increase and traders wouldbe willing to pay more for thechance of such a gain (Sarris). Inaddition, intrinsic value dependson the time until the option’s expi-ration. The greater the time hori-zon, the greater the intrinsic valuebecause price uncertainty isgreater. Observed options pricescan be used to provide informationabout anticipated price variability(see box, p. 40).

Table 12 illustrates the situationfor a central Illinois producer onMarch 15 who plans to produce 500acres of corn and hedge with putoptions. The December futuresprice is $3.00 per bushel at plant-ing time and the premium for at-the-money puts is $0.20 per bushel.His expected yield is 150 bushelsper acre, and his production costsare estimated at $150,000. Becausethe farmer expects his productionto fall no lower than 50,000bushels, he buys 10 put contracts(5,000 bushels per contract * 10contracts = 50,000 bushels), andselects a strike price of $3.00. Thecost associated with this purchaseis $10,000 in premiums (at anassumed cost of $0.20 per bushel),and $350 in commissions (10 con-tracts at $35 per contract).

As a simplifying assumption, sup-pose that the producer makes hisdecision on October 20 as to thesale of the option. If the futuresprice that day is $3.00 (equal tothe strike price), the option has no

Unlike hedging infutures, put options(or call options) donot limit the poten-tial profits associat-ed with increasing(decreasing) prices.


How

Farmers C

an M

anage R

isk

3377

intrinsic value (since the option isat the money), and an assumedtime value of $0.07 per bushel(reflecting the probability that thefutures price will decline beforethe option expires, raising theoption’s intrinsic value). Using thisestimate, the producer’s return tothe purchase of the option is thetime value on October 20 at $3,500($0.07 * 50,000 bushels), less thepremium cost of $10,000, and thecommission cost of $700 (10 con-tracts * $35 per trade * 2 trades),or -$7,200. If the producer had soldhis crop at the harvest cash price,his return would have been$37,500, instead of the $30,300($37,500 - $7,200) earned in thishypothetical put option situation.

The farmer’s return to buyingoptions depends largely on thefutures price at harvest. With ahigh futures price in October (say,$3.75), the producer’s loss associ-ated with the option is even high-er, while a low futures price (say,$2.50) would result in a highergain than in the cash-sale-at-har-vest-only case. In an efficient mar-ket, the producer’s return frombuying put options over a series ofmany years is expected to equal

the return to either hedging withfutures or simply selling the cropat harvest, except for commis-sions. Although returns areapproximately the same in allthree cases, hedging with eitherput options or futures reducesuncertainty about return.

Several studies have explored therisk-return properties of options asthey affect the farm firm. Many ofthese studies have found options tobe a potentially useful method forstabilizing returns (Heifner andPlato; Curtis, Kahl, andMcKinnell). Various surveys havealso found that options are used toat least the same extent asfutures, including a study of theIowa cattle sector (Sapp). In addi-tion, large-scale Corn Belt produc-ers participating in Top FarmerCrop Workshops in 1994 and 1995indicated that they used options asfrequently, or more frequently,than hedging and to price a signifi-cant portion of their crops (Patrick,Musser, and Eckman). In a 1988survey of Iowa producers, about 11percent of the grain producersresponding indicated that theyused hedging and about 11 percentindicated that they used options

Many studies havefound options to be apotentially usefulmethod for stabiliz-ing returns.


How

Farm

ers

Can

Man

age

Ris

k

Table 12—An illustration of net returns to a corn farmer who uses put optionsto protect against price riskItem Revenues, costs, and net returns Dollars

1 Revenue from crop sale on October 20:(75,000 bushels * $2.50/bushel) 187,500

2 Total production costs 150,000

3 Net return from crop sale 37,500

4 Premium for put option paid on March 15:(10 put options * 5,000 bushels * $0.20 premium/bushel) 10,000

5 Return from put sale received on October 20:(10 put options * 5,000 bushels * $0.07/bushel) 3,500

6 Commissions on put purchase and sale:(2 * 10 put options * $35 commission) 700

7 Net return from put hedge(5 - 4 - 6) -7,200

8 Net return from cash sale and put hedge(3 + 7) 30,300

Note: The put option has a $3.00 strike price and a $0.07 time value on October 20. Although net returns fromthe hedge in this example are negative, the example could as easily have been constructed to show a positivenet return. (See discussions in the text regarding the expected return to forward pricing.)


(Edelman, Schmiesing, and Olsen).Hog and fed cattle producers, how-ever, were more likely to hedgethan to use options.

The conclusions of the literature,however, are not definitive as tothe effectiveness of options con-tracts in reducing risk, based ondifferent underlying assumptions.One study, for example, analyzesproduction, hedging, and specula-tive decisions in futures andoptions markets given the pres-ence of basis risk (Lapan,Moschini, and Hanson). Theseresearchers, assuming no produc-tion risk, found that options are aredundant hedging tool whenfutures and options markets areunbiased and when cash prices area linear function of futures prices.They indicate that the optimalhedging strategy involves usingonly futures contracts (the returnsof which are linear in futuresprices) because they dominateoptions contracts (the returns ofwhich are nonlinear in futuresprices). If futures prices or optionspremiums are biased, however, theresults indicate that options, usedalong with futures, provide theoptimal strategy for insuringagainst price risk. They concludethat options are more appealing asa speculative tool to exploit privateinformation about price distribu-tions than as a hedging tool.

Intrigued by a comparison of sur-vey findings with the Lapan,Moschini, and Hanson research,Sakong, Hayes, and Hallam ques-tioned the conditions under whichproducers find options useful forhedging. Introducing both outputand price uncertainty, theseauthors found that it is almostalways optimal for farmers to buyput options and to underhedge onthe futures market. Their resultslend support to the practice ofhedging the minimum expectedyield on the futures market, whilehedging the remainder of expectedoutput against downside price risk

using put options. Theseresearchers also found that theirresults are strengthened if the pro-ducer expects local production toinfluence national prices and ifrisk aversion is higher at lowincome levels.

MMaaiinnttaaiinniinngg FFiinnaanncciiaall RReesseerrvveessaanndd LLeevveerraaggiinngg

Leveraging refers to the producer’suse of debt to finance the opera-tion. Increasing the degree ofleverage increases the likelihoodthat in a year of low farm returnsthe producer will be unable tomeet his or her financial obliga-tions, and heightens the potentialfor bankruptcy. Thus, in general,highly leveraged producers operatein an environment of greaterfinancial risk than do producerswho choose a less highly leveragedfarm structure.15

A producer’s choice of debt (rela-tive to equity) depends on manyfactors, including the farmer’srisk aversion, the size and type ofoperation, the farmer’s marketrelationships with input suppliersand output purchasers, lenders’willingness to provide loans, andthe availability of governmentprograms for managing risk.Increasing the farm’s leverage(that is, borrowing) increases thecapital available for production,allowing expansion of the busi-ness, but also entails incurring arepayment obligation and createsthe risk of loan default because ofthe risks inherent in the farmingoperation. Because of these manyfactors, a farmer’s use of debt tofinance the operation interactswith both the production andmarketing risks faced by the pro-ducer (Barry and Baker; Gabrieland Baker).

Highly leveraged pro-ducers operate in anenvironment ofgreater financial riskthan do producerswho choose a lesshighly leveragedfarm structure.


How

Farmers C

an M

anage R

isk

3399

15An increase in leveraging means that thefarmer is at increased risk. In contrast,farmers who increase their use of mostother risk management tools covered in theaccompanying sections—such as hedgingand insurance—reduce their risk.

Options traders cananticipate pricevolatility throughoutthe growing season,but not very wellprior to plantingtime.


How

Farm

ers

Can

Man

age

Ris

k

Commodity Options Quotes Provide Estimates ofAnticipated Price Randomness

Information about anticipated price variability within a given yearcan be obtained from commodity options quotes. The value of a com-modity option depends on the volatility of the underlying futuresprice, the futures price level, the strike price, the interest rate, andthe time to maturity (Black). Holding the last four variables con-stant, a higher volatility implies a higher price for both puts andcalls. Volatility cannot be observed until after the fact. However, ifan options price is observed along with the last four variables, an“implied volatility” can be calculated. Such implied volatilitiesembody the current judgments of traders—who have money on theline—as to the actual volatility likely to be realized.

To examine traders’ ability to anticipate volatility in corn and soybeanprices, the actual volatility was regressed on implicit volatility for theyears 1987-95 using the December corn contract and the Novembersoybean contract. The actual volatility (the dependent variable) wascalculated using the log (P t / P t-1) procedure applied to futures pricesfrom the end of each month to the last trading day preceding thefutures delivery month. For the February estimate of corn pricevolatility, for example, actual volatility was captured by the standarddeviation of the daily log of relative futures prices from March 1 toNovember 30. Implicit volatility (the independent variable) was calcu-lated by applying the Black formula to at-the-money puts and averag-ing over the trading days in the month (for February, using the previ-ous example). The R2 from this equation is illustrated in the accompa-nying figure for each month prior to expiration of the December corncontract and the November soybean contract. The chart indicates thatoptions traders can anticipate price volatility from May and Junethrough the growing season, but not very well prior to planting timewhen such information would be most valuable.

Farmers and marketers can potentially use implied volatilities inmaking planting and storage decisions. Implied futures price volatili-ties, together with futures quotes, the producer’s expected yields, andthe producer’s expectation of yield variability, may indicate thatplanting corn, for example, would result in higher and less volatilereturns than planting soybeans (or vice versa). Farmers or marketerswho are storing a crop may be able to make more use of implied pricevolatility information than those who are making planting decisionsbecause output risk can be disregarded when storing.

Jan Feb Mar Apr May Jun Jul Aug Sep Oct0

0.2

0.4

0.6

0.8R squared

December corn

November soybeans

Source: Estimated by ERS from Chicago Board of Trade prices.

Proportion of corn price volatility to harvest anticipated by options traders by month

The risk management decisionconfronting a farmer who mustchoose the degree of leverage canbe illustrated using the portfolioapproach. Table 13 illustrates theeffect of borrowing on the variabil-ity of returns to owned equity,where the expected rate of returnto farming (Ra) is 12.5 percent andthe interest rate for both borrow-ing and saving (id) is assumed tobe 7.5 percent. The standard devi-ation of farming returns (σa) is 5percent, and the standard devia-tion of the risk-free asset (σd) iszero. The higher rate of return tofarming is consistent with theassumption that returns must behigher than the risk-free rate ofreturn or risk-averse individualswould not invest in farming.

The first column of the tablereports various levels of debt-to-equity ratios. A negative debt-to-equity ratio reflects a farm thathas invested a portion of its equityin risk-free savings at a 7.5-per-cent return. In contrast, a positivedebt-to-equity ratio indicates thatthe operator has borrowed toexpand the operation. The expect-ed return to equity capital (Re) is:

Re = (Ra * Pa) - (id * Pd)

In the equation, Pa and Pd are pro-portions of the two assets relativeto equity, with the holdings of therisky asset (Pa) plus holdings of

the risk-free asset (Pd) equaling1.0 in the total portfolio. Leverag-ing implies negative holdings ofthe risk-free asset, resulting in aminus sign in the equation. Thisinformation is used in the secondand third columns, which illus-trate the tradeoff between expect-ed returns and the variability ofreturns, with the standard devia-tion of the return to equity calcu-lated as the weighted standarddeviation of the risky asset:

σe = σa Pa

As an example, consider a debt-to-equity ratio of 0.25. In this case,holdings of the risky asset are 1.25and holdings of the risk-free assetare -0.25. Thus, the expectedreturn is calculated as (1.25 *0.125) - (0.25 * 0.075), or 13.75percent. The standard deviation ofthe return to equity is (0.05 *1.25), or 6.25 percent.

As shown in the table, the -0.5debt-to-equity ratio results in thelowest expected return and thelowest level of risk. In this situa-tion, the producer holds thegreatest proportion of his or herassets in the low-return risk-freeinvestment, and the smallest pro-portion in the higher return riskyasset. As more capital is investedin the higher risk, higher returnfarming operation, expectedreturns increase, as does the

In this example, themore highly lever-aged the farmbecomes, the greaterthe risk and thegreater the expectedreturn—which maynot always be thecase for a farmer inreal life.


How

Farmers C

an M

anage R

isk

4411

Table 13—An example of the effect of borrowing on the variability of returns to owned capital

Standard deviation of Debt/equity ratio Expected return to equity (Re) expected return (σe)

Percent-0.50 10.00 2.50-0.25 11.25 3.750.00 12.50 5.000.25 13.75 6.25

0.50 15.00 7.500.75 16.25 8.751.00 17.50 10.001.25 18.75 11.25

Note: The expected return to farming equals 12.5 percent, with a standard deviation of 5 percent. The expect-ed interest rate is 7.5 percent, with a standard deviation of zero.

Source: Adapted by ERS from Barry, Peter J., and C. B. Baker, “Financial Responses to Risk in Agriculture,”Risk Management in Agriculture, ed. Peter J. Barry, Ames: Iowa State University Press, 1984, pp. 183-199.

standard deviation of returns toequity. Thus, the more aggressive-ly the farmer borrows, the morehighly leveraged the farmbecomes and, in this example, thegreater the risk and the greaterthe expected return.

Various research studies haveexamined producers’ use of lever-aging. For example, a recent studyexamined the sources of capitalused by farm operators across theUnited States, averaging datafrom the 1991-93 AgriculturalResource Management Study(ARMS). Over the 3 years, lenderswere found to supply 10 percent ofthe $638 million in total capitalmanaged by commercial farms.Most commercial farm capital washeld either as owned equity (55percent) or was leased (35 per-cent), generally from landlords(table 14). As anticipated, relianceon debt and leased capital financ-ing declined as wealth and ageincreased (Koenig and Dodson).

In another study, use of debtrepayment capacity, measured asthe ratio of actual debt relative tothe maximum amount of debt sup-ported by net cash income avail-able for loan payments, was ana-

lyzed for various sales class sizes(Ryan). Use of debt repaymentcapacity was found to increaseacross all commercial farm sizeclasses between 1991 and 1994,especially in the smallest category(defined as $40,000 to $99,999). Inthis category, use rose from lessthan 50 percent in 1991 to over 70percent in 1994.

For those producers who are high-ly leveraged, understanding andmanaging price and yield risk canassume heightened importance.This is because highly leveragedfarmers must be concerned aboutmeeting their financial obligations,and high yield and price risk insuch situations may increase thelikelihood of insolvency and bank-ruptcy. Thus, farmers’ decisionsabout leveraging (and hence, thefinancial risk they confront) mustbe considered in the context of thebusiness risks they confront ontheir operations.

Several studies have examinedthis interaction between price andyield (business) risk and producerbehavior with regard to financialrisk. In particular, one line ofresearch has addressed the topicof “risk balancing,” and considered

Research indicatesthat farm policiesthat create a lowerrisk business envi-ronment mightinduce financialchoices that increasetotal farm risk


How

Farm

ers

Can

Man

age

Ris

k

Table 14—Sources of capital for various groups of commercial farm operators, 1991-93 average

Young commercial1 Older commercial2

Low- Low- AllItem resource Aspiring Wealthy resource Traditional Wealthy commercial

Percent

Share of total managedcapital that is:

Leased capital 69 52 27 60 34 21 35Debt capital 17 12 10 15 10 8 10Owner capital 14 36 62 25 56 71 55Total 100 100 100 100 100 100 100

Farms

Number of farms 40,260 66,845 33,062 180,194 240,105 67,576 562,866

Dollars

Average total managedcapital per farm 543,361 805,375 1,937,478 587,534 952,741 2,827,468 1,133,906

1Young farmers are under 40 years of age. The category definitions are low resource (less than $150,000 net worth), aspiring ($150,000-$500,000 net worth),and wealthy (greater than $500,000 net worth).2Older farmers are over 40 years of age. The category definitions are low resource (less than $250,000 net worth), traditional ($250,000-$1 million net worth),and wealthy (greater than $1 million net worth).

Source: Excerpted by ERS from Koenig, Steven R., and Charles B. Dodson, “Sources of Capital for Commercial Farm Operators,” Regulatory, Efficiency, andManagement Issues Affecting Rural Financial Markets, Proceedings of the NC-207 Regional Committee, Staff paper SP0196, ed. Bruce L. Ahrendsen,Fayetteville: University of Arkansas, Department of Agricultural Economics and Rural Sociology, January 1996, pp. 70-83.

the producer’s financial leveragingstrategy in the presence of govern-ment farm programs that helpstabilize prices and/or yields. Inthis context, “risk-balancing”refers to adjustments in businessand financial risk that result froman exogenous shock (such as a sta-bilizing policy) that affects theexisting balance. The seminalwork in this area was conductedby Gabriel and Baker, who devel-oped a conceptual framework thatlinked production, investment,and financing decisions via a riskconstraint. Their model indicatedthat, in the aggregate, farmersmake financial adjustments lead-ing to decreased (or increased)financial risk in response to a rise(or fall) in business risk. Thus,farm policies that create a lowerrisk business environment mightinduce financial choices thatincrease total farm risk.

In further investigating “risk bal-ancing,” other research has exam-ined the impacts of income-sup-porting farm policies on leverage.Collins, for example, developed astructural model of a risk-averseproducer’s overall debt-equity deci-sion, supporting Gabriel andBaker’s conclusions that risk-reducing farm policies may wellincrease financial risk-takingbehavior on the part of producers.Collins also concluded that, forrisk-averse producers, greater risk-taking behavior also may be asso-ciated with policies intended toraise expected farm income. Otherwork has shown that risk-reducingand income-enhancing policiesmay, due to increased leveraging,increase the likelihood of farmerslosing part of their equity or goingbankrupt (Featherstone, Moss,Baker, and Preckel). Further,research that includes more com-plex specifications, such as taxlaws, credit subsidies, and otherfactors, reach similar conclusions(Moss, Ford, and Boggess;Ahrendsen, Collender, and Dixon).

In addition to approaches thathave examined the links betweenfinancial risk and business risk,other research has examined opti-mal farm decisionmaking, includ-ing links that span financial, mar-keting, and production considera-tions. The underlying tenet of thisline of research is that certainmarketing strategies often work tostabilize business risk, and also toreduce the risks associated withdebt repayment by ensuring morepredictable incomes. Thus, afarmer may choose either a for-ward contracting, hedging, or otherbusiness risk strategy accommo-dating lenders’ preferences forgreater liquidity (see next section)and loan repayment certainty(Barry and Baker). These modelsoften are based on risk program-ming and stochastic simulations,and typically assume risk aversionon the part of the producer.

Several of these studies havefocused on hedging (a financialstrategy for which data are readilyavailable) and its relationship to aproducer’s use of leverage. Thesestudies generally conclude thathedging tends to increase as thefarm’s debt level rises. Using anoptimal hedging model that explic-itly accounts for the financialstructure of the farm, one suchtheoretical article concludes thathedging is positively related todebt because hedging reduces busi-ness risk, offsetting to some extentthe increased financial risk associ-ated with leveraging (Turvey andBaker, 1989).

Empirical research tends to sup-port these findings, including stud-ies focusing on a hypothetical cornand soybean farm in Indiana(Turvey and Baker, 1990) andFlorida orange growers (Moss andvan Blokland; Moss, Ford, andCastejon). A survey of Indianafarmers also indicates that highlyleveraged farmers are more likelyto hedge than other producersbecause they perceive that hedging

Studies indicate thathighly leveragedfarmers are morelikely to hedge thanother producers.


How

Farmers C

an M

anage R

isk

4433

could increase their net returnsand/or reduce their risk (Shapiroand Brorsen). Farmer’s use ofleveraging (and the resulting debtpayment obligations) is closelyrelated to liquidity management,the topic of the next section.

LLiiqquuiiddiittyy

Another aspect of financial riskmanagement is liquidity, whichinvolves the farmer’s ability to gen-erate cash quickly and efficientlyin order to meet his or her finan-cial obligations (Barry and Baker).The liquidity issue relates to cashflow and addresses the question:“When adverse events occur, does afarmer have assets (or other mone-tary sources) that can easily beconverted to cash to meet his orher financial demands?”

Asset liquidity depends on therelationship between the firm’sassets and the expected cash pro-ceeds from the sale of each of thoseassets (Barry, Baker, and Sanint).An asset is perfectly liquid if itssale generates cash equal to, orgreater than, the reduction in thevalue of the firm due to the sale.Illiquid assets, in contrast, cannotbe quickly sold without a produc-er’s accepting a discount, reducingthe value accruing to the firm bymore than the expected sale price.Examples of liquid assets includegrain in storage, cash, and compa-ny stock holdings, while illiquidassets include land, machinery,and other fixed assets. Factors thatinfluence liquidity include mar-ketability of the asset, the lengthof time allowed for liquidationbefore the cash is needed, transac-tions costs, and the asset’s income-generating role in the firm (Barry,Baker, and Sanint; Pierce).

Liquidity management is interre-lated with risk responses in pro-duction and marketing, and alsowith the farm’s degree of leverage.The more highly leveraged thefarm, everything else being equal,

the greater the need for carefulliquidity management in order tomake timely payments on loansand other farm financial obliga-tions. Some of the methods thatfarmers use to manage liquidity,and hence their financial risk,include the following:

• Selling Assets—A producer’swillingness to sell assets is animportant financial response torisk, particularly in crisis situa-tions (Barry and Baker). If afarmer faces a low net income ina given year, selling liquidassets (such as stored grain ornonfarm assets, such as stocks)is a first step in meeting expens-es for the year. Holding liquidassets, however, may be costlybecause they typically earnlower returns than when used inthe production process (assum-ing the economic viability of theoperation). If the use of liquidassets is not adequate to meetfinancial demands, additionalsteps—such as the sale of lessliquid assets—may be necessary.Because many farmers are heav-ily invested in illiquid assets,such as land, livestock, andmachinery, maintaining liquidityto meet shortfalls in returnsmay at times be difficult.

• Managing the Pacing ofInvestments and Withdrawals—Maintaining flexibility in thetiming of farm investments andwithdrawals is also a response tofinancial risk. In low incomeperiods, for example, a producermay postpone the purchase ofnew machines and other equip-ment. This is an approachfavored by many producers dur-ing times of adversity. It avoidslarge financial outlays duringsuch periods, builds equity,reduces indebtedness, and allowsthe strengthening of profitabilityin a rapidly expanding farmoperation (Barry and Baker).

The more highlyleveraged the farm,the greater the needfor careful liquiditymanagement inorder to make timely payments on loans and otherobligations.


How

Farm

ers

Can

Man

age

Ris

k

• Holding Liquid CreditReserves—Producers commonlymaintain liquid credit reservesto manage their financial risk.Credit reserves reflect unusedborrowing capacity, and general-ly reflect additional capital fromlenders in the form of an openline of credit. Maintaining creditreserves avoids the costs associ-ated with liquidating assets tomeet cash demands, and the pos-sible later reacquisition of thoseassets when the adversity haspassed (Barry and Baker). Inaddition, drawing from creditreserves when needed does notdisrupt the farm’s asset struc-ture, the transactions costs aretypically low, and institutionalsources of funds are generallyavailable to producers in ruralareas (Barry, Baker, and Sanint).Several implicit costs are, how-ever, associated with suchreserves. For example, they rep-resent an opportunity cost fromforgone leveraging. Further,interest must be paid on newloans, and noninterest charges(such as loan fees) are at timesused by lenders to compensatefor establishing lines of credit(Barry, Baker, and Sanint).

Farmers’ reliance on the last strate-gy listed above—accessing creditreserves to obtain liquidity duringtimes of adversity—introduces riskin terms of lenders’ responses.Lenders’ decisions regarding theavailability of credit are directly

affected by a farm’s capital struc-ture (the degree to which the farmis leveraged), conditions in the agri-cultural sector (such as the level ofmarket prices), and financial mar-ket conditions (such as interestrates) (Barry, Baker, and Sanint).

Partly due to significant loan loss-es in the 1980’s, agriculturallenders increasingly have empha-sized credit quality and manage-ment of credit risk in their loanportfolios. Both price responses(risk-adjusted interest rates) andnonprice responses (differentialloan limits, security requirements,or loan supervision requirements)may be employed to address creditrisk (Miller, Ellinger, Barry, andLajili). In a 1992 survey of morethan 1,000 banks in Illinois,Indiana, and Iowa, respondentsindicated that they were quite ableto distinguish between high- andlow-risk borrowers, and to monitortheir performance. Of the respon-dents, 70 percent indicated thatthey differentially priced loans tofinance farm production, and 59percent differentially priced loanssecured by farm real estate. Asmaller percentage of respondentsindicated the use of risk-adjustedinterest rates on loans (table 15).16

Agricultural lendersincreasingly haveemphasized creditquality and manage-ment of credit risk intheir loan portfolios.


How

Farmers C

an M

anage R

isk

4455

Table 15—Selected price and nonprice responses of respondent banks inIllinois, Indiana, and Iowa, 1992Price and nonprice response Indication of use

PercentDifferential loan pricing on loans to finance

agricultural production 70

Differential loan pricing on loans secured byfarm real estate 59

Risk-adjusted loan pricing on loans to financeagricultural production 57

Risk-adjusted loan pricing on loans secured byfarm real estate 40

Charging of fees on agricultural loans 43Source: Excerpted by ERS from Miller, Lynn H., Paul N. Ellinger, Peter J. Barry, and Kaouthar Lajili, “Price andNonprice Management of Agricultural Credit Risk,” Agricultural Finance Review 53 (1993): 28-41

16Commercial banks may use differen-tial rates for other reasons than credit risk.Complexity of pricing, a bank's capital posi-tion, bank size, and risk-distinguishingability generally are associated with theuse of differential and risk-adjusted pricing(Miller, Ellinger, Barry, and Lajili).

In addition, lenders may requirethat producers use one or morerisk management strategies toincrease the likelihood of timelypayments on financial obligations.Indeed, lenders’ recommendationscan have an important influenceon producers’ risk managementdecisions. A survey of Texaslenders and producers in the late1980’s, for example, indicated thatthe use of risk management prac-tices—including hedging, forwardcontracting, crop insurance, farmprogram participation, and diversi-fication—resulted in lenders view-ing loan requests more favorably(Knight, Lovell, Rister, and Coble).Using a logit model, this researchalso found that lenders can greatlyincrease the probability of theirborrowers adopting certain riskmanagement practices if the use ofthose practices is recommended bythe lender.

Regardless of lender recommenda-tions, empirical research providesevidence of the effectiveness ofsuch risk management strategies.As discussed earlier, studies showthat the use of hedging or optionsreduces financial risk andimproves cash flow, potentiallylowering a farmer’s credit risk(Turvey and Baker, 1989). Becauseof this risk reduction, high-debtproducers with low credit reserveswould be expected to hedge morethan low-debt producers with largecredit reserves (Turvey and Baker,1990). Turvey and Baker’s resultssupport the notion that lenderswill benefit from producers’ hedg-ing (and presumably, their use ofother risk management strategies)because it decreases portfolio riski-ness (Heifner, 1972a).

LLeeaassiinngg IInnppuuttss aanndd HHiirriinnggCCuussttoomm WWoorrkk

Producers can also manage theirfarming risks by either leasinginputs (including land) or hiringworkers during harvest or other

peak months. Leasing refers to acapital transfer agreement thatprovides the renter (the actualoperator) with control over assetsowned by someone else for a givenperiod, using a mutually agreed-upon rental arrangement (Perry,1997). Farmers can lease land,machinery, equipment, or livestock.

Leasing has similarities withleveraging (a topic discussed previ-ously in this section), in that bothare methods used to expand con-trol over resources. In addition,both commit the farmer to regularpayments. Leasing appears, how-ever, to have some advantages.One advantage is that control canbe gained over long-life inputs(such as land and machinery),without making long-term pay-ment commitments. In addition,leasing provides producers withflexibility in allocating their assetportfolios—a producer can be ineither the farming business or theland ownership business, withoutbeing in both.

Leasing has potential advantagesto those who are renting. Leasingimproves the renter’s flexibility torespond to changing market condi-tions. In addition, leasing reducesthe long-term fixed payments onborrowed capital that may strainliquidity in years of reduced out-put, and can reduce both financialand production risk for the renter(Sommer and others, 1998). Inessence, leasing limits fixed costs,providing greater flexibility for therenter to adapt. It also offers away to enter farming or to managethe size of the operation withoutrequiring large investments of cap-ital. One disadvantage, however, isthat renting may limit the short-term borrowing capacity of anoperation because of the absence ofcollateral to back a loan (Sommerand others, 1998).

Advantages may further accruefrom the perspective of the owner.Leasing allows the owner of the

Leasing has similari-ties to leveraging, inthat both are meth-ods used to expandcontrol overresources.


How

Farm

ers

Can

Man

age

Ris

k

asset to receive a return on his orher investment, and may reducerisk when a share rental arrange-ment is used if the owner is confi-dent of the renter’s managementability. In the case of a cashlease, the owner’s income risk isreduced substantially.

Leasing of land is common in U.S.agriculture. According to USDA’sARMS data, about 9 percent of U.S.farm operators leased all of theirland in 1995, 36 percent operatedon at least some rental land, and55 percent owned all of the landthat they operated (Sommer andothers, 1998). The ARMS resultsalso indicated that full-ownerfarms (those that rented neitherland nor other production assets)accounted for proportionally small-er shares of acreage, income, andsales than part-owner farms thatrented land and other assets (fig.8). Farms that rented land andother productive assets operatedmore than twice the U.S. averageacreage, and had income and sales3.5 to 4 times the national average.Although apparently increasing inrecent years, leasing of nonrealestate assets is at a lower level

than of farmland (Barry; also seeKoenig and Dodson).

Land rental arrangements can falleither in the category of “sharerenting” or “cash renting.” Withshare renting, the landlord andtenant share in the operation’sreturns and each provides a prede-termined set of inputs. The twoparties usually share input costs inthe same proportions as outputsand share the risk of yield variabil-ity. They typically have equal sayin management decisions, althoughthe tenant usually carries out mostof the production decisions. Often,the owner provides land, while therenter provides machinery andlabor. In practice, the renter (aswell as the owner) may have sever-al such arrangements.

The risk benefit of this type ofarrangement is derived from thefinancial sharing of potential loss-es between the partners. If netreturns are negative in a particu-lar year, for example, losses arespread across the participants inthe share rental arrangement. Inessence, share leasing is a highlyrisk efficient form of financing, inthat the operator’s rental obliga-tion moves in a perfectly correlat-


How

Farmers C

an M

anage R

isk

4477

Full owner Part owner Tenant0

10

20

30

40

50

60

70

80Farms

Acres operated

Gross cash income

Gross value of sales

Percent

Source: Sommer, Judith E., Robert A. Hoppe, Robert C. Green, Penelope J. Korb, Structural and Financial Characteristics of U.S. Farms, 1995: 20th Annual Family Farm Report to Congress, AIB-746, U.S. Dept. Agr., Econ. Res. Serv., November 1998.

Figure 8

Distribution of farms, acres operated, gross cash income, and gross value of sales by tenure class, 1995

Land rental arrange-ments can be classi-fied as either “sharerenting” or “cash renting.”

ed way with receipts from theoperation, thus stabilizing theafter-rent income position relativeto a fixed-payment cash lease.

Share-rental arrangements can bedifficult to manage, however, andthe trend has been away fromshare leasing to cash leasing.Some of the impetus for this trendis on the part of landowners, par-ticularly if the owner is absenteeand questions arise regarding therenter’s practices and skills. Theowner may decide that his or herincome risk is too great and thatmonitoring the management skillsof the renter is too time consum-ing, and may instead opt for a cashrental arrangement. Some of theimpetus for this trend is also fromoperators. It is easier to bid foradditional tracts of land usingcash bids than share bids, andcash leasing avoids the sharing ofmanagement responsibilities withseveral landlords.

With cash renting, the tenant rentsthe land for a pre-specified, fixedamount per acre. Cash rentingaffords the renter flexibility, as in ashare-rent agreement. All of theyield and price risk are absorbedby the renter in a cash rentingarrangement, and none remainswith the owner, who receives onlythe agreed-upon cash rent payment(Perry, 1997). In addition, therenter typically provides inputsother than the land (including themachinery), reducing the fixedcosts committed by the landowner.To better match rental arrange-ments with the needs of landlordsand tenants, “hybrid” contracts arenow being used. These “flexible”cash rents incorporate the risk-sharing advantages of share leases,without the sharing of responsibili-ties (Barry).

Research suggests that accountingrates of return may vary systemat-ically with a farm’s tenure posi-tion, but that these differences donot necessarily have implications

for performance in terms of eco-nomic rates of return. Accountingrates of return for owned farmlandhave been low historically, withempirical research indicating that,as tenancy increases, accountingrates of return to assets and lever-age positions tend to increase(Ellinger and Barry). Differencesacross tenure classes largelyreflect the nondepreciability offarmland and its inherently lowrate of return and low debt-carry-ing capacity because part of thereturns to land ownership occur ascapital gains rather than as cur-rent income (Barry and Robison).Low accounting rates of returnmay mask underlying economicrates of return, and provide pro-ducers with liquidity problemsthat worsen with the degree offinancial leverage.

Owners who hire custom help(who provide skilled labor andtheir own equipment) can lowerthe costs associated with commit-ting capital to fixed inputs.Producers may, at times, find thathiring workers full-time for theentire year may be costly whenthose workers are only essentialduring harvest or other peakmonths. With the use of customworkers (or hired or contractlabor), the owner has a great dealof flexibility, potentially lowers hisor her costs, and obtains special-ized labor (Perry, 1997). The use ofsuch arrangements, however, mayincrease the owner’s risk becausehe or she would have less controlover resources than if equipmentwere owned outright or workershired full-time.

IInnssuurriinngg CCrroopp YYiieellddss aanndd CCrroopp RReevveennuueess

Insurance is often used by cropproducers to mitigate yield (andhence, revenue) risk, and is obvi-ously prevalent outside of agricul-ture. Property, health, automobile,and liability insurance are all

Cash renting affordsthe renter flexibility,although the renterabsorbs all of theyield and price risk.


How

Farm

ers

Can

Man

age

Ris

k

forms of insurance regularly pur-chased by individuals to mitigaterisk. For an individual, the use ofinsurance involves the exchange ofa fixed, relatively small payment(the premium) for protection fromuncertain, but potentially large,losses. When losses occur, virtuallyall types of insurance policiesrequire a deductible, meaning thatthe individual must assume a por-tion of the value of the loss.Indemnities compensate individu-als for losses up to the level of theinsurance guarantee, which isbased on the deductible chosen bythe insured (within ranges set bypolicy terms).

A key characteristic of an insur-ance market involves the conceptof risk pooling. Risk poolinginvolves combining the risks facedby a large number of individualswho contribute through premiumsto a common fund, which is used topay the losses due any individualin the pool (Ray). More specifically,when an insurance company sellspolicies to many different individu-als who have less than perfectlycorrelated risks, the total portfoliowill be less risky than the averageof the individual policies. This isbecause, at any point in time, theodds of all insureds in the poolhaving a claim are extremely low.Thus, the insurer diversifies non-systemic (uncorrelated) risksacross the insurance pool(Goodwin and Smith; Miranda).

In part because of several “marketfailure” arguments, theGovernment operates the multi-peril crop insurance (MPCI) pro-gram. One market failure argu-ment is based on the idea thatmany of the natural disaster risksassociated with crop production(such as drought, flooding, and dis-ease) are correlated across wide-spread geographical areas. As aresult, it has been argued thatpooling risks on a scale that is fea-sible for most private insurers isdifficult (Miranda and Glauber;

Ray).17 Others argue that privatemulti-peril insurance fails becauseother types of producer responsesto risk—such as diversificationand smoothing of consumptionover time through savings and bor-rowing—greatly reduce the addi-tional effect of insurance insmoothing consumption, and makeinsurance unattractive to farmerswhen offered at competitive mar-ket prices (Wright and Hewitt).

In addition, research has shownthat moral hazard and adverseselection are problems that signifi-cantly affect the viability of multi-ple peril crop insurance (Ahsan,Ali, and Kurian; Chambers;Goodwin and Smith). Moral hazardis present when an insured individ-ual can increase his or her expect-ed indemnity by actions takenafter buying insurance. Adverseselection occurs when a farmer hasmore information about the risk ofloss than the insurer does, and isbetter able to determine the fair-ness of premium rates. Both moralhazard and adverse selection affectthe actuarial soundness of insur-ance, and pose a particularly diffi-cult dilemma in multi-peril cropinsurance. A lack of extensive pro-ducer-specific yield-risk informa-tion, which is needed to controladverse selection, has been a prob-lem historically, and monitoringfarmers’ protection against losses,which is the basis for controllingmoral hazard, is also difficult.Empirical research has used vari-ous data sets and approaches, andprovides evidence of moral hazardin multi-peril crop insurance (Justand Calvin, 1993a; Coble, Knight,Pope, and Williams), as well as

In part because ofseveral “market fail-ure” arguments, theGovernment oper-ates the multi-perilcrop insurance program.


How

Farmers C

an M

anage R

isk

4499

17This argument has been countered bythose who argue that a wide array of rein-surance options are available in interna-tional markets that allow systematic risksto be diversified. Goodwin and Smith, forexample, state that, "Such markets aremore than able to permit a sufficientdegree of diversification to permit risksthat appear to be systematic to individualmarkets to be spread across a wider rangeof activities and markets."

adverse selection (Just and Calvin,1993b; Goodwin; Luo, Skees, andMarchant; Quiggin, Karagiannis,and Stanton).

In contrast to multi-peril cropinsurance, certain other agricul-tural risks—such as the risks asso-ciated with hail damage or thedeath of livestock—are insured byprivate companies with no govern-ment subsidization or reinsurance(see box for information on live-stock insurance, p. 53). Unlikemultiple peril crop insurance,these markets are generally char-acterized by risks that are nonsys-temic across producers, similar tothe risks underlying liability, auto-mobile, life, and other types of pri-vate-market insurance.

The Federal multi-peril crop insur-ance program has been the focus ofinterest in recent years, and theFederal Crop Insurance ReformAct of 1994 increased the level ofthe premium subsidy provided toproducers, as well as grower par-ticipation. With passage of the1994 Act, Congress introduced cat-astrophic (CAT) coverage, forwhich growers do not pay a premi-um. Rather, producers who chooseto obtain CAT must pay an admin-istrative fee.18 CAT policies pay forlosses below 50 percent of a pro-ducer’s average yield (based on a4- to 10-year “actual productionhistory,” or “APH,” yield series forthe grower). When losses qualify,indemnity payments are made at arate of 55 percent of the maximumprice set by USDA’s RiskManagement Agency (RMA).19

Growers can select among a widevariety of coverage levels underthe program. More specifically, agrower can obtain multi-peril crop

insurance at levels between 50 and75 percent of his or her APH yield,using 5-percent increments.20

Growers can also select a pricecoverage level of up to 100 percentof the established price set byRMA. Coverage above the CATlevel, up to a maximum of 75/100(the first number refers to theyield coverage and the secondnumber to the price coveragelevel), is termed “buy-up” coverage.Producers receive indemnitiesunder the program according tothe following equation:

Indemnity = Max [(GuaranteedYield - Actual Yield), 0] *

Price Guarantee.

Within this equation, the guaran-teed yield is calculated by multi-plying the producer’s APH yield bythe coverage level that he or sheselects. To illustrate, assume thata soybean producer has an APHyield of 40 bushels per acre, andselects a coverage of 75 percent.The guaranteed yield is then 30bushels per acre (0.75 * 40). If theactual yield is 20 bushels in agiven year, an indemnity would bepaid on the 10 bushels (30 - 20) ofshortfall from the yield guarantee.If the actual yield is above theguarantee, the farmer receives noindemnity. The price guaranteeplaces a dollar value on the loss. Ifthe farmer chooses a $5.50 priceelection, for example, his or herindemnity would total $5.50 * 10bushels, or $55 per acre.

Except at the CAT level, producersmust pay a premium for coverageunder the multi-peril crop insur-ance program. The key to insur-ance rate setting is the accurateestimation of expected indemnities.

The Federal CropInsurance ReformAct of 1994 increasedthe level of the pre-mium subsidy pro-vided to producers,as well as growerparticipation.


How

Farm

ers

Can

Man

age

Ris

k

18Before 1999, producers paid an admin-istrative fee of $50 for CAT coverage.Beginning in 1999, the fee is $60.

19Starting in 1999, CAT coveragedeclined from 50-percent yield and 60-per-cent price coverage to 50-percent yield and55-percent price coverage.

20Starting in 1999, APH coverage isavailable at the 85-percent yield coveragelevel in selected areas and for selectedcrops. Various revenue insurance products(see upcoming discussion, p. 52) will alsoallow 85-percent coverage in selected loca-tions and for selected crops. Higher levelcoverage is also available under the GroupRisk Plan (see upcoming discussion, p. 52).

In effect, insurers must set actuari-ally sound premium rates so thatthe premiums collected are in bal-ance with total expected indemni-ties. Under an actuarially soundprogram with no subsidization, theaverage insured individual would,in the long run, expect to receivethe same amount in indemnities asis paid in premiums.

While actuarially fair rates pro-vide a starting point, insurancepremiums generally must coveradditional costs. Private insurancecompanies must price their prod-ucts in order to recover overhead,operating costs, and a desiredreturn on equity. When these costsare added into the premium, thecost of insurance over time exceedsindemnities that will be paid out.Individuals are willing to acceptsuch contracts for automobile,medical, and other private insur-ance products (as well as hail andlivestock insurance) due to riskaversion (see appendix 2). In short,private insurance is priced accord-ing to the following formula:

Premium = (Actuarially FairPremium + Administrative Costs)

> Expected Indemnity.

In contrast, Federal multi-perilcrop insurance attempts to encour-age participation by providing fourprimary types of subsidies. Thesecategories include the following:21

• Premium subsidy—The premiumpaid by producers has been sub-sidized since 1980, with the sub-sidy depending on the level ofcoverage. Currently, the maxi-mum subsidy for multi-peril crop

insurance (other than for CAT,which is subsidized at 100 per-cent), is 41.7 percent of the totalpremium, and is offered at the65/100 coverage level. The sub-sidy varies with other levels ofcoverage and by type of product.

• Delivery expense reimburse-ment—The private companiesdelivering policies to farmers are,as of 1998, reimbursed for theirsales and service expenses at 11percent of (implicit) total premi-um for CAT coverage and 24.5percent of total premium at buy-up levels. In the absence of gov-ernment involvement, privatecompanies would include thisexpense in the premium paid bythe producer.

• Reinsurance—The Governmentreinsures private companies thatsell policies (that is, the Govern-ment shares in the risk of loss) tohelp reduce financial losses inyears of widespread disasters.Companies can also earn under-writing gains when certain condi-tions are met, as determined inthe Standard ReinsuranceAgreement signed betweenUSDA and the companies.

• Excess losses—Indemnities arepaid to qualifying farmersregardless of the level of premi-um income. Such “excess losses”are paid by the Government inyears when indemnity paymentsexceed total premiums. TheFederal Crop Insurance ReformAct of 1994 legislates that oper-ation of the program (includingthe setting of premiums) is to beconducted in a manner so thatthe loss ratio (total indemnitiesdivided by total premium) is notto exceed an expected maximumof 1.075 over the long run.

Thus, premiums charged thefarmer under multi-peril cropinsurance are priced according tothe following equation:

Federal multi-perilcrop insuranceattempts to encour-age participation byproviding four pri-mary types ofsubsidies.


How

Farmers C

an M

anage R

isk

5511

21In addition, the administrative costsassociated with funding USDA's RiskManagement Agency are appropriatedannually by Congress. This funding sup-ports RMA's development of rates and poli-cy terms for the multi-peril crop insuranceprogram, the research and developmentcosts associated with new RMA productintroductions, compliance functions associ-ated with private companies deliveringpolicies, and other activities.

Premium = (Actuarially FairPremium - Premium Subsidy) <

Expected Indemnity.

As a result, farmers have twoincentives for obtaining multi-perilcrop insurance. Because the pro-gram is subsidized, participantsare expected to receive indemnitiesin excess of their premium cost,resulting in a positive net return.In addition, research has confirmedthe risk-reducing effectiveness ofcrop insurance, particularly in situ-ations of high yield variability.

Risk protection is greatest whencrop-yield insurance (which pro-vides yield risk protection) is com-bined with forward pricing orhedging (which provide price riskprotection). Using an example,research indicates that a corn pro-ducer in North Carolina—a fairlyhigh-risk corn-producing area—would expect that his or her rev-enue would fall below 70 percentof expected revenue about 23 per-cent of the time. With the purchaseof 75/100 crop insurance, the per-centage falls to 17 percent, andwith the use of both crop insur-ance and an optimal hedge, thepercentage falls to 7 percent.Generally, revenue insurance pro-vides protection similar to thecombination of crop insurance andan optimal hedge (Harwood,Heifner, Coble, and Perry).

Since 1990, Congress and theAdministration have becomeincreasingly interested in encour-aging the development of newtypes of policies. Group Risk Plan(GRP) insurance, which is basedon county (rather than individual)yields, was first introduced on apilot basis in 1993, and has sincebeen expanded to nearly all majorfield crops in the late 1990’s(Skees, Black, and Barnett).Because it is based on area (notindividual) yields, producers withsignificant yield losses may findthemselves unprotected becausethe county yield does not warrant

an indemnity payment. Variousstudies have shown that GRP ismost effective at protecting indi-vidual yield risk when a strongcorrelation exists between individ-ual and county-level yields(Miranda; Skees; Glauber,Harwood, and Skees).

In addition, both producers andpolicymakers have expressed con-siderable interest since the early1980’s in the concept of revenue(and cost of production) insurance.In the 1981 Farm Act, for example,Congress mandated a study on thefeasibility of revenue insurance. Inthe 1994 Federal Crop InsuranceReform Act, Congress mandated acost of production insurance planthat was to compensate producersfor reductions in yield and/or priceresulting from an insured cause.And, in the 1996 Federal Agricul-ture Improvement and Reform(FAIR) Act, Congress clearly sig-naled the need for introducing pilotrevenue insurance programs.

As of 1998, three revenue insur-ance products were available toproducers of major field crops inselected areas: Crop RevenueCoverage, Income Protection, andRevenue Assurance (see appendix3). These products complementmany strategies, such as the useof diversification, and provide amore comprehensive alternative tothe use of multi-peril crop insur-ance. In designing these alterna-tives, policymakers, program ana-lysts, and insurance companieshave benefited from witnessingCanada’s experience with theGross Revenue InsuranceProgram (GRIP) in the early1990’s. Canada’s GRIP was expen-sive (in terms of both governmentand farmer costs) and interferedwith market signals and plantingdecisions, largely because it usedlong-term average prices in estab-lishing the guarantee (Sands;Turvey and Chen). Based on thisexperience, U.S. revenue insuranceproducts use an intrayear futures

Generally, revenueinsurance providesprotection similar tothe combination ofcrop insurance andan optimal hedge.


How

Farm

ers

Can

Man

age

Ris

k

price guarantee, rather than aguarantee based on long-termaverage prices.

Revenue insurance was first intro-duced in the United States in 1996.The Income Protection product wasdeveloped by USDA’s Risk Manage-ment Agency in 1996. Crop Reve-nue Coverage, which was designedby American Agrisurance, Inc., aprivate company, was also intro-duced in 1996. These programswere expanded from limited cover-age in 1996 to new geographicareas in 1997 and 1998. A morerecent product, Revenue Assur-ance, was offered in 1997 for cornand soybeans in Iowa, and wasdeveloped by the Iowa Farm

Bureau. Research has indicatedthat the effectiveness of revenueinsurance in reducing farm-levelincome risk can be substantial, andis similar to the effectiveness ofcombining the purchase of cropinsurance with hedging (Harwood,Heifner, Coble, and Perry).

As implemented for 1998 crops,these three plans have many simi-lar features, but also differ in manyways. Each of the products com-bines price and yield risk protec-tion in one program. Indemnitiesunder each plan equal the amount,if any, by which guaranteed rev-enue exceeds the revenue realizedat harvest. All calculate guaran-teed and realized revenues from

Hail insurance andseveral types of live-stock insurance areavailable through theprivate sector.


How

Farmers C

an M

anage R

isk

5533

Several Types of Livestock Insurance AreAvailable Through the Private Sector

Several types of insurance covering livestock are widely available inthe United States through the private sector. Like private hail insur-ance, these products are not subsidized by the Federal Government.Livestock and hail insurance are quite different from multi-peril crop-yield and crop-revenue insurance in that coverage is typically limitedto those losses that are independent geographically, such as hail (inthe case of hail insurance) or fire, lightning, hail, collision, and othersuch perils (in the case of livestock).

One of the most popular private products used by livestock producers isa blanket farm personal property policy. Under this type of policy, live-stock coverage is included as part of the farm’s business property, and issubject to the same terms, conditions, and limitations faced by otherproperty associated with the farm business. The insured selects anamount of protection and pays premium on that amount, with each pol-icy limiting the actual cash value or market value per animal. Typicalper animal values are $2,000 per head of cattle and $500 per head ofswine (Anderson). At least one insurer offers an endorsement to theirblanket policy, for an added premium, that includes coverage for freez-ing or smothering in blizzards or snowstorms (Skees and Pyles).

An alternative to a blanket policy is a “stated value” policy. Under thistype of policy, the insured provides the insuring company with a list ofthe value of individual animals to be insured. Coverage is for animaldeath caused by named perils for animals on the farm or in transportto another location. Perils typically covered by such policies includefire, lightning, aircraft or falling objects, collision with a vehicle,smoke, vandalism, and theft (Anderson).

A third type of insurance is livestock mortality coverage, which is all-risk term life insurance. This coverage is typically used to insure high-value show or performance animals and covers loss due to death ortheft. Livestock insured under such policies must pass a veterinarian’sinspection and their values must be substantiated at the time of poli-cy issuance (Anderson).

farm yields and from futures pricesat signup and at harvest time.They all use policy terms associat-ed with basic coverage under themulti-peril crop insurance pro-gram. In addition, each productrequires that producers pay a pre-mium for coverage, which is subsi-dized by the Federal Governmentin a manner similar to multi-perilcrop insurance. The FederalGovernment also reinsures privatecompanies against a portion of thelosses associated with each of theproducts, and provides reimburse-ment for delivery expenses. Theuniqueness of each product, interms of the specification of theguarantee and other variablesestablishing the producer’s cover-age, is explained in the appendix.In addition, each product is uniquein its rating methodology and theproducer’s ability to subdivideacreage into individual parcels forloss adjustment purposes.

Among the revenue insuranceproducts currently available, CropRevenue Coverage (CRC), which isavailable over the widest geo-

graphic areas and has been themost widely publicized, has thehighest enrollment. Despite CRC’shigher premium rates, sales werestrong relative to multi-peril(APH) crop insurance in 1996 and1997 in many areas (see figs. 9 and10 for 1997 data). Indeed, littlecorrelation appears to existbetween premium rates (relativeto APH) and the proportion ofacreage covered by CRC.

CRC sales were particularly strongfor corn and soybeans in Iowa andNebraska in 1997. Two factors like-ly explain this result. First, Iowaand Nebraska were the only Stateshaving prior experience with CRCin 1996, and producers were likelymore familiar with the programthan in locations in which 1997was the first year of CRC coverageavailability. Second, Nebraska hasa large American Agrisurance, Inc.(the CRC-developing company)sales force, and agent enthusi-asm—a key to the successful mar-keting of insurance policies—waslikely strong. Third, AmericanAgrisurance, Inc., invested consid-

Among the revenueinsurance productsavailable, CropRevenue Coveragehas the highestenrollment.


How

Farm

ers

Can

Man

age

Ris

k

PercentLess than 3031 to 5051 to 75Greater than 75

Figure 9

Proportion of corn Crop Revenue Coverage (CRC) acres to allbuy-up insured corn acres, 1997

Note: Shaded areas include counties with at least 500 acres planted to corn.Source: Estimated by ERS from USDA, Risk Management Agency, electronic experience and yieldrecord database, 1996 and 1997.

erable time and effort in promo-tional activities (such as agent andcommodity group meetings) inIowa and Nebraska (Cleaveland).

OOffff--FFaarrmm EEmmppllooyymmeenntt aanndd OOtthheerrTTyyppeess ooff OOffff--FFaarrmm IInnccoommee

Earning off-farm income is anoth-er strategy that farmers may useto mitigate the effects of agricul-tural risk on farm family house-hold income. Not only can off-farmincome supplement householdincome, it may also provide a morereliable stream of income thanfarm returns. In essence, off-farmincome can offer a form of diversi-fication. The incentives for diversi-fying income sources depend onthe level and variability of returnswhen considering a risk-averseproducer. If farm households arerisk averse, then they will be will-ing to supply relatively more laborto stable off-farm occupations thanthey would otherwise (Mishra andGoodwin, 1997). Or, they may seekout other types of off-farm income(such as interest and dividends) tocounter negative fluctuations infarm income.

According to USDA’s ARMS data, alarge percentage of farm familiesearn off-farm income, and the lev-els of off-farm income relative tofarm income can be significant.ARMS data for 1996, for example,indicate that 82 percent of all farmhouseholds had off-farm incomethat exceeded their farm income(Hoppe). For each farm type catego-ry (including very large farms), atleast 28 percent of the householdswithin the category had off-farmincome exceeding farm income.

Farm household income can be cat-egorized as earned off-farm income(wages and salaries), unearned off-farm income (social security, pen-sions, and investments), and farmnet cash income (fig. 11). As illus-trated in the figure, reliance on off-farm income is related to farmsize. About 10 percent of farmhouseholds were classified as pri-marily engaged in farming andhaving sales between $100,000and $249,999 in 1996. These farmsrelied on off-farm sources for about57 percent of their total householdincome. In contrast, householdsoperating very large farms (those

Off-farm income can offer a form of diversification.


How

Farmers C

an M

anage R

isk

5555

Percent

Figure 10

Proportion of soybean Crop Revenue Coverage (CRC) acres to all buy-up insured soybean acres, 1997

Less than 3031-5051-75Greater than 75

Note: Shaded areas include counties with at least 500 acres planted to soybeans.Source: Estimated by ERS from USDA, Risk Management Agency, electronic experience and yieldrecord database, 1996 and 1997.

with sales of $500,000 or more)accounted for 3 percent of allfarms and relied on off-farmsources for a relatively small percentage of their average income(Hoppe).22

Several studies have modeled fac-tors, such as off-farm work, thataffect inequality in the distribu-tion of income among farmers.Gardner (1969) found that off-farmwork reduced both shortrun andlongrun income inequality, andpostulated that off-farm work mayenable poor farmers to add to theirown capital stock. A study focusingon New York farmers reached sim-ilar conclusions, finding that ifincomes are improved by increas-ing income from nonfarm sources,inequality among farm familieswould likely be reduced (Boisvertand Ranney). Using 1991 ARMSdata, another study found that thedistribution of income in the NorthCentral region was most equalamong U.S. regions, and mostunequal in the West (El-Osta,Bernat, and Ahearn). In addition,

results indicated that farm opera-tor households that did not partici-pate in off-farm employment expe-rienced higher income inequalityas a group than did their partici-pating counterparts.

Research has also addressed thedecision to engage in off-farm workand the hours of off-farm worksupplied by farmers. One study,focusing on off-farm labor supplyin Illinois, found that off-farmwork was quite sensitive to eco-nomic incentives, and that a 10-percent increase in the off-farmwage entailed an 11-percentincrease in hours of off-farm work,holding other factors constant(Sumner). A study focusing onMassachusetts farmers in 1986/87concluded that the hours of worksupplied by the farm operatordepended on the participationdecision of the spouse. In addition,family and farm characteristicswere important to both the partici-pation decision and hours workedby the farm operator (Lass,Findeis, and Hallberg).

Various empirical studies haveexamined the relationship between

Evidence suggeststhat the riskiness offarm income is posi-tively related toworking off thefarm.


How

Farm

ers

Can

Man

age

Ris

k

<$100,000 $100,000-$249,999 $250,000-499,999-0.2

0.0

0.2

0.4

0.6

0.8

1.0Farm income

Earned off-farm

Unearned off-farm

Percent


Sales class

Figure 11

Farm household income by sales class, 19961

1Note: For sales classes less than $250,000, the operator's principal occupation is farming. Farm operator households are associated with farms organized as individual operations, partnerships, or family corporations, and are generally closely held by the operator household. Household income includes income from farming activities and earnings from nonfarm sources by all household members in the reporting year.

22For more information on historical off-farm earnings, see Hoppe and others;Hamrick; Kassel and Gibbs.

off-farm employment and farmincome variability. In one study, atimes series analysis of aggregatedata indicated that the fraction oftotal farm family income earnedfrom off-farm sources was higherin the 1980’s than in the early1970’s, and suggested that theriskiness of farm income is posi-tively related to working off thefarm (Kyle). A study focusing onproducer responses to a survey inDodge County, Georgia, in the1980’s indicated that risk and lowincomes were major disadvantagesassociated with full-time farming(Bartlett). In another study, farmhousehold total income was foundto be significantly less variable ifproducers and their spousesworked off the farm (Sander).

More recent research has moreexplicitly linked the decision towork off the farm with farmincome variability and other fac-tors (Mishra and Goodwin, 1997).Mishra and Goodwin’s analysis,using a simultaneous-equationTobit model, confirmed that theoff-farm labor supply of farmers ispositively correlated with the risk-iness of farm income amongKansas farmers. Their results alsoindicated that off-farm work (forboth the farmer and spouse) ispositively correlated with off-farmexperience and with the degree ofleverage associated with the farm.Further, operators of larger farmsand those receiving governmentsupports were less likely to workoff the farm. In a followup study,Mishra and Goodwin (1998) alsofound a positive and significantcorrelation between farm incomevariability and the decision byfarm operators in North Carolinato work off the farm.

Although the focus of this sectionhas been on off-farm employment,off-farm income may be derivedfrom other sources as well (such asinterest and dividends). Indeed,several studies have concludedthat the low correlation between

financial assets (stocks, bonds, cer-tificates of deposit) and farmassets suggests that diversifyinginto financial assets may yieldimportant gains in risk efficiencyfor farm households. A quadraticprogramming analysis of a repre-sentative Illinois grain farm, forexample, indicated that variouslevels of diversification couldreduce the relative variability ofthe farm’s rates of return on assetsby 15-25 percent compared withholding farm assets alone (Youngand Barry). Conversely, otherresearch has focused on nonfarmequity investment in agriculture,generally concluding that investorscan gain from inclusion of farmassets in their investment portfo-lios (Crisostomo and Featherstone;Moss, Featherstone, and Baker).

OOtthheerr WWaayyss ooff MMaannaaggiinngg RRiisskk

The strategies and tools just dis-cussed in detail are by no meansall inclusive. Many other diversestrategies for farm risk manage-ment are commonly used by pro-ducers on their operations. Some ofthese additional strategies includethe following:

• Adjusting inputs and outputs—Producers can respond to risk byaltering output levels, input use,or some combination of the two.Research indicates that greateroutput price risk results inlower levels of both input useand final output. Given thatpreferences toward risk and cir-cumstances can vary greatlyacross producers, the final inputand output levels chosen by pro-ducers can, accordingly, varyconsiderably for individuals insimilar situations. (See Sandmo;Hawawini; Ishii; Robison andBarry; and Just and Pope formore detail.)

• Cultural practices—Culturalpractices can be used to reduceyield and, hence, income risk.One such practice involvesplanting short-season varieties

Farm householdtotal income hasbeen found to be sig-nificantly less vari-able if producers andtheir spouses workoff the farm.


How

Farmers C

an M

anage R

isk

5577

that mature earlier in the sea-son, protecting against the riskof early frost and yield loss.Supplemental irrigation due toabnormal weather is anothermeans to protect against yieldloss.

• Excess machine capacity—Afarmer may have enoughmachine capacity so that plant-ing and harvesting crops canoccur more rapidly than neededunder normal weather condi-tions. By having such resources,the farmer can avoid delays ateither planting or harvest thatmay reduce yield losses.

Other methods of risk manage-ment in farming are also impor-tant, and focus on other types of

issues than those specific to pro-duction, marketing, and finance.Legal risks and issues associatedwith farm liability, for example,have become increasingly impor-tant. In addition, tax concerns area key issue in managing theincome risks associated with year-to-year income flows, as well asestate transfers from generation togeneration (Keller; Keller andRigby-Adcock; Baquet, Hambleton,and Jose). Government pay-ments—such as contract paymentsunder the 1996 Farm Act—canalso be used to provide liquidity,for example, or to pay the premi-um for an options contract or a“buy up” crop insurance policy.


How

Farm

ers

Can

Man

age

Ris

k

AA 11999966 UUSSDDAA SSuurrvveeyyffoouunndd tthhaatt kkeeeeppiinnggccaasshh oonn hhaanndd wwaasstthhee nnuummbbeerr oonnee rriisskkmmaannaaggeemmeenntt ssttrraattee--ggyy ffoorr eevveerryy ssiizzeeffaarrmm,, ffoorr eevveerryy ccoomm--mmooddiittyy ssppeecciiaallttyy,, aannddiinn eevveerryy rreeggiioonn..

Previous sections in this reporthave focused on addressing the

myriad strategies that producerscan use to manage their farm-levelrisks and their effectiveness. Thissection, in contrast, addresses thequestions: “How have producersused these tools and strategies ontheir farming operations?” “Whatfactors are associated with farm-ers’ use of different strategies?”Several surveys of farmers’ use ofvarious risk management strate-gies have been conducted over thepast 10-15 years. These surveystypically focus on asking producerswhether or not they use hedging,crop insurance, and forward con-tracts, as well as whether theymanage risk through diversifica-tion, keeping cash on hand, andother strategies.

Two difficulties are present inassessing and interpreting theresults of these surveys, whichmust be kept in mind while read-ing the results presented in thenext paragraphs. First, many ofthe surveys are focused on specificStates or areas. Because differentquestions are asked of differentgroups of farmers at differenttimes, it is difficult to compareresponses on a “one-for-one” basisacross studies or across time.

Second, farmers are typically ques-tioned as to their use of a strategyto manage risk. Some producersmay indicate that they use a givenstrategy (such as diversification orhedging), even though profit maxi-mization (and not risk reduction)may be their primary motivation.

The most comprehensive survey offarmers’ use of selected risk man-agement strategies is USDA’sAgricultural Resource Manage-ment Study (ARMS). Results ofthe 1996 ARMS survey, conductedshortly after passage of the 1996Farm Act, indicate that operatorsin the largest gross income cate-gories (more than $250,000 annu-ally) are most likely to use hedg-ing, forward contracting, and virtu-ally all other risk managementstrategies. In contrast, operatorswith less than $50,000 in saleswere less likely to use forward con-tracting or hedging, and signifi-cantly fewer reported diversifica-tion as a method for reducing risk(fig. 12). Keeping cash on hand foremergencies and good buys wasthe number one strategy for everysize farm, for every commodityspecialty, and in every region.

The 1996 ARMS survey also askedproducers about the impact of the


Farmers’R

epo

rted U

se of

Risk

Man

agemen

tStrategies

5599

FFaarrmmeerrss’’ RReeppoorrtteedd UUssee ooff RRiisskkMMaannaaggeemmeenntt SSttrraatteeggiieess

Several surveys of farmers’ use of risk management strate-gies have been conducted over the past 10-15 years. Theresults vary. Results of the 1996 Agricultural ResourceManagement Study (ARMS), for example, conducted shortlyafter passage of the 1996 Farm Act, indicate that operatorsin the largest gross income categories (more than $250,000annually) are most likely to use virtually all risk manage-ment strategies. In contrast, operators with less than$50,000 in sales were less likely to use forward contractingor hedging. Keeping cash on hand for emergencies and goodbuys was the number one strategy for every size farm, forevery commodity specialty, and in every region.

1996 Farm Act in influencingwhether or not they were consider-ing the use of new strategies. Atthe U.S. level, about one-third ofthe producers responding to thesurvey reported receiving directgovernment commodity payments.Of those receiving governmentpayments, between 5 and 8 per-cent indicated that they increasedtheir use of at least one risk man-agement strategy or tool (forwardcontracting, futures hedging, use ofoptions, use of insurance, or otherstrategy) in 1996 in response tothe 1996 Farm Act.23 Responseswere fairly consistent across allU.S. regions. With less governmentintervention in farming andgreater trade liberalization, farm-ers appear to be increasingly rely-ing on forward contracting andother risk management tools toreduce their farm-level risks.

A recent Farm Futures survey alsoquestioned its readers nationally asto their use of various risk manage-ment strategies. The 690 respon-dents reflect a nonrandom pool of

the magazine’s readers. These read-ers are generally in the top 10 per-cent of all U.S. farmers, with about75 percent located in the Corn Beltand almost all have sales exceeding$100,000 annually. Commonly usedstrategies reported by a high pro-portion of these respondents includ-ed using government farm pro-grams, diversifying into both cropsand livestock, planting varietieswith different maturity dates, con-tracting inputs to lock in a favor-able price, buying crop insurance,and using crop-share rentalarrangements (table 16).

Several surveys of producers’ useof risk management strategieshave been conducted by universityextension specialists. In a surveyconducted in the mid-1990’s,Nebraska producers were ques-tioned about their use of alterna-tive marketing tools, includingcash forward contracts, hedgingwith futures, hedging with options,hedge-to-arrive contracts, basiscontracts, and minimum price con-tracts. They were also asked thepercentage of their crops for whichthese tools, if any, were used, withuse of a tool considered to be

Of producers receiv-ing government pay-ments, 5-8 percentindicated that theyincreased their useof at least one riskmanagement tool orstrategy in 1996 inresponse to the 1996Farm Act.


Farm

ers’

Rep

ort

ed U

se o

fR

isk

Man

agem

ent

Stra

tegi

es

< $50,000 All U.S.0

20

40

60

80

100

Forward contracting

Diversifying

Cash on hand

Hedge in futures


Sales class

Note: For all sales classes, the principal occupation of the operator may or may not be farming.

Percent

Figure 12

Farmers’ use of alternative risk management strategies by sales class and all United States, 1996

$500,000+$50,000-$249,999

$250,000-$499,999

23This is approximately 1-3 percent ofall U.S. farmers.

important if it was used to marketmore than 50 percent of the pro-ducer’s crop (Jose and Valluru).The results indicate that cash for-ward and basis contracts were themost commonly used marketingtools for any percentage of theNebraska producers’ crops. Ofthose using cash forward con-tracts, about 47 percent indicatedthe use of this tool to price 75-100percent of their crop. Similarly, ofthose using basis contracts, 49 per-cent indicated use of this tool toprice 75-100 percent of their crop.

Among the responding producersparticipating in Top Farmer CropWorkshops held at PurdueUniversity in 1993, 1994, and1995, about two-thirds indicatedthat they used cash forward con-tracts. Producers participating inthe workshops indicated that thesecontracts were, on average, used toprice 20-30 percent of their cornand soybean crops. Hedging wasused by approximately 10-20 per-cent of the participants, dependingon the specific crop and year(Patrick, Musser, and Eckman;Musser, Patrick, and Eckman).

Evidence also exists from theGreat Plains. A 1992 survey ofKansas producers indicated thatover 30 percent of the respondentsused forward contracting to price aportion of their wheat, corn, andsoybean crops during the 1990-92period. Corn was hedged in futuresmost frequently (reported by 11percent of the respondents), fol-lowed by cattle (8 percent of therespondents). Nearly 15 percent ofthe wheat producers and about 10percent of the cattle and corn pro-ducers had used options comparedwith less than 5 percent of soy-bean, grain sorghum, and hog pro-ducers (Goodwin and Schroeder).

Several surveys provide informa-tion historically on the use of for-ward contracting and futures hedg-ing, and suggest that the use ofthese strategies may haveincreased over time. In a 1986Wisconsin study, for example,about 20 percent of the respon-dents had used cash forward con-tracts at least once in the mostrecent 5 years, and 8 percent hadused futures within that period(Campbell and Shiha). The surveyalso indicated that large-scale pro-ducers were more frequent users of

Regional surveysoften point to thecommon use of cashforward contracts byproducers.


Farmers’R

epo

rted U

se of

Risk

Man

agemen

tStrategies

6611

Table 16—Results of a Farm Futures magazine questionnaire on farmers’ use ofvarious risk management strategies, 1997

Percentage of respondents indicatingTool or strategy use of tool or strategy

PercentUsed government farm program 69Diversified operation by raising crops and livestock 39Planted seed varieties with different maturity dates 39Contracted inputs to lock in a good price 35

Bought crop insurance 30Used crop-share land rents 25Kept a credit line open to take advantage

of attractive input prices 20Used multiyear leases 16

Irrigated 13Shared expenses with landlord 10Refinanced loans to take advantage of lower interest rates 8Hired custom operator to reduce machinery expenses 6

Hired custom operator to improve timeliness of crop operations 6Diversified by growing crops not normally grown in the area 3Leased equipment rather than bought 3Rented equipment rather than bought 2Source: Excerpted by ERS from Knorr, Bryce A., editor of Farm Futures magazine, Testimony before theSubcommittee on Risk Management and Specialty Crops, U.S. House of Representatives, April 10, 1997.

both forward contracting andfutures than were small-scale pro-ducers.24 In another study, only 7percent of Kansas grain producersreported hedging in 1983, and 18percent had forward contracted atany time in prior years (Tierney;Mintert). Based on a different sam-ple, the Kansas Crop and LivestockReporting Service reported thatless than 5 percent of farmershedged any of their grain in each ofthe years 1980-85.

Data from the 1996 ARMS surveyindicate that more farmers inmany areas may be using variousrisk managements strategies—such as forward contracting, diver-sifying, hedging, or keeping cashon hand—than reported in region-al and State-level studies in theearly- and mid-1980’s (fig. 13). Forexample, about 40 percent of pro-ducers in the Corn Belt andNorthern Plains regions used for-ward contracting and approxi-mately 25 percent used futures in1996. Producers in the SouthernPlains were somewhat less likelyto use many of the risk manage-ment strategies listed than thosein the Corn Belt or NorthernPlains, as were producers in theNortheast and Appalachia.

Empirical studies have at timesextended survey data and exam-ined the relationship between theuse of various strategies and pro-ducer characteristics. A study of 41selected farmers in Indiana in1985, for example, found that theuse of hedging was positivelyrelated to the farmer’s perceptionof the income-stabilizing potential

Data from the 1996ARMS survey indi-cate that more farm-ers in many areasmay be using variousrisk managementstrategies than inthe 1980’s.


Farm

ers’

Rep

ort

ed U

se o

fR

isk

Man

agem

ent

Stra

tegi

es

Corn Belt N. Plains Southeast S. Plains All U.S.0

20

40

60

80

100Forward contracting

Diversifying

Cash on hand

Hedge or futures

Percent

Note: For all regions, the principal occupation of the operator may or may not be farming.

Figure 13

Farmers' use of alternative risk management strategies by selected regions and United States, 1996


24ARMS survey results from 1983 pro-vide information on the value of sales mar-keted by various methods regionally, andimplicitly support the idea that large-scaleproducers are more likely to use forwardcontracts. The ARMS data indicate thatbetween 50 and 60 percent of corn salesdelivered at harvest in 1983 in Illinois,Iowa, Minnesota, and Ohio were priced byforward contract, while less than 30 per-cent of corn sales in Kansas, Michigan, andMissouri were priced using this method(Harwood, Hoffman, and Leath). Similarly,more than 50 percent of soybean salesdelivered off-farm at harvest were forwardcontracted in Illinois and Minnesota in1982 and 1983 compared with fewer than25 percent of sales in Kansas and manySoutheastern States (Leath). In contrast,less than 15 percent of wheat sales at har-vest were forward contracted in 1983 inmost major wheat-producing States, includ-ing Kansas and North Dakota (Hoffman,Harwood, and Leath).


Farmers’R

epo

rted U

se of

Risk

Man

agemen

tStrategies

6633

Risk Management Education Can Use Many Avenues

Risk management education has been an important initiative, as witnessed by a fiscal year 1998 effort jointly spon-sored by USDA’s Risk Management Agency, USDA’s Cooperative State Research, Education, and Extension Service,and the Commodity Futures Trading Commission. These efforts have focused on the use of a wide variety of work-shops, education programs, information events, and research to better help educate producers and understand theneeds of farmers in the learning process. These efforts complement longstanding work undertaken in the coopera-tive extension community (Anderson and Mapp; Schroeder, Parcell, Kastens, and Dhuyvetter).

In theory, a producer’s decision to obtain the human capital necessary to adopt a new technology—whether involv-ing a new conservation technique or a new forward pricing strategy—is based on factors related to the expectedreturns and costs associated with adoption and the producer’s risk attitude. Producers evaluate the discountedvalue of their expected returns from education (net of investment costs) to evaluate whether or not they shouldparticipate. If discounted expected net returns are positive, a producer would tend to participate (Ben-Porath).Using an example, discounted expected returns to investment in education fall as the time horizon decreases.Thus, the expected returns to education are expected to decline with the age of the producer, meaning that olderfarmers are less likely to participate in educational programs than younger farmers.

This theoretical basis was used to evaluate Kansas producers’ participation in risk management and marketingeducation programs in 1992 (Goodwin and Schroeder). This research found, as expected, that more experienced(older) farmers are less likely to participate in educational programs. The percentage of crop acres on the farm,total farm acres, the degree of farm leverage, the educational level of the operator, and a preference for risk wereall positively related to participation. Similarly, preference for farm-related education, measured by hours perweek spent reading farm publications, also had a significant positive effect on seminar attendance. Importantly,the authors found that participation in marketing and risk management education seminars and programs sig-nificantly increased farmers’ adoption of forward pricing techniques.

Farmers use many educational sources other than seminar attendance. Ford and Babb, for example, found thatfarm magazines, other farmers, and family and friends were among the most important information sources for asample of producers in Indiana, Illinois, Iowa, and Georgia in the 1980’s. In another study conducted in the 1980’s,researchers conducting a random survey of Ohio cash grain producers found that older farmers and operators ofsmall farms often cited radio and television broadcasts as the most useful source of marketing information, whileoperators of larger farms and those with at least some college education tended to cite marketing professionals asmost useful (Batte, Schnitkey, and Jones).

Further, a nonrandom sample of large, commercial farm operations in the Corn Belt in 1991 found that producersspent an average of $2,578 per year on information sources, and that consultants accounted for 60 percent of totalexpenditures (Ortmann, Patrick, Musser, and Doster). Despite the importance of consultants, the use of “own farmrecords/budgets” were the highest-rated information source for production, marketing, and financial decisions forthese producers. These results support recommendations by the extension service and others encouraging pro-ducers to keep and use farm records and to prepare farm budgets for planning purposes.

A series of questions included on USDA’s 1996 Agricultural Resource Management Study provides information onpriority educational needs. In this section of the survey, producers were questioned as to changes that they wouldmake in their farming operation under adverse circumstances (“what would you do differently if faced with finan-cial difficulty?”). The respondents were provided a listing of production, marketing, and financial activities fromwhich to choose. Producers in the $50,000 and higher sales classes indicated consistently that they would adjusttheir costs, improve their marketing skills, restructure their debt, and spend more time on management (seetable). These responses indicate the wide-ranging—yet interrelated—risk management education needs of pro-ducers, and can be effectively provided by both private and public sector interests.

Changes that producers would make in their operations if faced with financial difficulty, 1996Sales Class

Item Less than $50,000 $50,000-$249,999 $250,000-$499,999 $500,000+ All

PercentRestructure debt 24.3 47.7 45.8 48.7 30.3Sell assets 31.1 27.8 31.2 28.5 30.4Use more custom services 7.4 17.5 17.4 19.9 10.1

Scale back 25.6 23.1 20.0 23.7 24.8Diversify 11.8 22.9 20.9 20.6 14.5Spend more time on management 18.7 37.7 47.3 44.4 24.3

Use advisory services 18.8 22.1 28.0 26.3 20.1Adjust costs 33.9 54.0 58.8 57.2 39.5Improve marketing skills 29.5 47.2 53.1 53.4 34.6Source: USDA, ERS, 1996 Agricultural Resource Management Study, special analysis.

of hedging, debt position, and farmsize (Shapiro and Brorsen).Contrary to expectations, educa-tion was found to be inverselyrelated to hedging, a result thatmay be peculiar to the sample ofproducers analyzed in the study. Inaddition, risk attitudes were notsignificantly related to the use offorward pricing methods.

Other studies of this type havebeen based on larger participantsamples. A survey of 677 Iowagrain, swine, and fed cattle pro-ducers in 1988, for example, indi-cated that use of hedging forgrains was positively and signifi-cantly related to gross farm salesand the use of other forward pric-ing tools over the prior 2 years(Edelman, Schmiesing, andOlsen). These same variables alsohad the greatest significance inexplaining use of futures to hedgeswine and fed cattle.

In a study of 595 producers partic-ipating in USDA’s Futures andOptions Marketing Programbetween 1986 and 1988, modelresults indicate that prior use offorward contracts, possession of abachelor’s degree or above, mem-bership in a marketing club, andgross sales had the greatest posi-tive impact on the probability of

using futures and options (Makus,Lin, Carlson, and Krebill-Prather).A survey of 1,963 Kansas farms in1992 found that the use of forwardpricing techniques is positivelyand significantly related to yearsof formal education, croplandacreage, total farm acres, leverage,risk preference, input intensity,marketing seminar participation,and the use of crop insurance(Goodwin and Schroeder).

These studies, by providing infor-mation on producer characteristicsand the use of forward pricingtechniques, suggest strategies forproducer education (see box).Operators of larger farms, thosethat are most highly leveraged,and those with prior experienceusing forward contracts would bemost likely to be interested inusing futures or options. In con-trast, education on cash forwardcontracts would likely be moreeffective for the general farm pop-ulation than education on futuresand options. At least one study hasfound that the use of marketingclubs (which often emphasize alearning-by-doing approach)appears to be quite effective inintroducing producers to futuresand options (Makus, Lin, Carlson,and Krebill-Prather).

Many different char-acteristics are associ-ated with farmers’use of hedging,including farm sizeand prior use of for-ward contracts.


Farm

ers’

Rep

ort

ed U

se o

fR

isk

Man

agem

ent

Stra

tegi

es

TThhee eeffffeeccttiivveenneessss ooffffoorrwwaarrdd pprriicciinngg iinnrreedduucciinngg rriisskkddeeppeennddss oonn yyiieellddvvaarriiaabbiilliittyy,, yyiieelldd--pprriicceeccoorrrreellaattiioonn,, aannddwwhheetthheerr oorr nnoott tthheeccrroopp iiss iinnssuurreedd..

We have seen in the “HowFarmers Can Manage Risk”

section of this report that riskmanagement strategies canreduce farmers’ risk of incomeloss, at times substantially. Thissection provides a detailed analy-sis of the effectiveness of severalrisk management tools on incomeuncertainty within the year(intrayear income risk), incomeuncertainty between years (multi-year income risk), and farmers’average returns.

The analysis focuses on selectedtools including futures hedges,cash forward contracts, crop insur-ance, and revenue insurance,rather than the many tools dis-cussed earlier. There are severalreasons. First, forward pricing andinsurance are widely available tofarmers and among the more effec-tive risk-reducing tools. They arefairly easy to use and involve nocommitment beyond the currentcrop year. Second, their optimaluse is somewhat independent of

differences between farms inwealth, debt, rental arrangements,off-farm earning opportunities, andthe use of other risk managementtools. Third, focusing on this limit-ed set of tools allows results for afew representative farms to havewide applicability.

EEffffeeccttss oonn IInnccoommee UUnncceerrttaaiinnttyyWWiitthhiinn tthhee YYeeaarr

The many different options avail-able for managing income risk leadto questions about their effective-ness across different producingregions and about how they canbest be combined to reduce produc-ers’ risks. A producer’s choiceamong strategies is particularlycomplicated when both price andyield (output) risk is present—thecase for a farmer with a growingcrop in the field. In this situation,the degree to which strategies,such as forward contracting orhedging, reduce income riskdepends on yield variability, thecorrelation between price and


How

Farmers C

an R

edu

ce Risk

6655

HHooww FFaarrmmeerrss CCaann RReedduuccee RRiisskk:: EExxaammpplleessUUssiinngg HHeeddggiinngg,, FFoorrwwaarrdd CCoonnttrraaccttiinngg,,CCrroopp IInnssuurraannccee,, aanndd RReevveennuuee IInnssuurraannccee

To what extent can hedging, forward contracting, and cropand revenue insurance reduce uncertainty within the year(intrayear) and over longer periods (multiyear), and changefarmers’ average returns? All of these tools tend to reduceintrayear income uncertainty, but have only small or negli-gible effects on multiyear uncertainties. Some strategies—such as the combined use of insurance and forward pric-ing—tend to complement each other in reducing risks. Theirrisk-reducing effectiveness varies by crop and location,depending on yield variability and the degree to which farmyields and prices move together. In addition, crop and rev-enue insurance likely increase average returns slightly formost farmers because they are subsidized, but forward con-tracting and hedging, lacking subsidies, have little directimpact on average or expected returns.

yield, and whether or not the cropis insured.25

Recent research at the EconomicResearch Service (ERS) examinedthe effectiveness of several strate-gies—use of hedging,26 crop insur-ance, and revenue insurance—inreducing farmers’ income risksover the growing season in variouscorn-growing locations. Thesestrategies were compared with theuse of a “no risk-reducing strategy,”which assumes that producers selltheir crops at harvest for the localcash market price and do notinsure. Four counties with differingyield variabilities and yield-pricecorrelations were selected for theanalysis, including the following:

• Iroquois County—Located ineast central Illinois, this countyhas relatively low yield variabil-ity and a strongly negativeyield-price correlation.

• Anderson County—In east cen-tral Kansas, this county repre-sents an area with relativelyhigh yield variability and a highyield-price correlation.

• Lincoln County—In west centralNebraska, Lincoln County repre-sents an irrigated area whereboth yield variability and yield-price correlation are low.

• Pitt County—In east centralNorth Carolina, this county rep-resents an area of relativelyhigh yield variability and lowyield-price correlation.

For each county, a “hypothetical”corn farm was specified, and riskreduction was estimated. The riskmeasure used is the probability ofrevenues falling below 70 percentof their average or expectation.This measure is unit free (as it isexpressed in percentage terms),and facilitates comparisons acrossfarms having different averageyields.

The results indicate that a repre-sentative corn farm in AndersonCounty, Kansas, or Pitt County,North Carolina, has a much higherlikelihood of very low revenueswhen no strategy is used (a cashsale at harvest and no crop insur-ance) than a corn farm in IroquoisCounty, Illinois, or Lincoln County,Nebraska (fig. 14). The probabilitiesare 21 percent in Anderson County,25 percent in Pitt County, 9 percentin Iroquois County, and 8 percent inLincoln County. The risk of cata-strophically low returns is consider-ably higher in the Kansas andNorth Carolina counties becauseyields vary more in those countiesthan in counties where crops areirrigated (as in Nebraska) or whereweather risk is inherently low (asin central Illinois).

In addition, the “natural hedge”(the price-yield correlation) is a fac-tor in explaining risk outcomes. Inmajor producing areas in the CornBelt, such as in Iroquois County,widespread low yields can signifi-cantly increase prices. Conversely,low prices are often associated withbumper-crop years. This negativerelationship between prices andyields tends to stabilize farmer rev-enues in these areas, and con-tributes to the low risk of loss inIroquois County. Pitt County, incontrast, is more likely to have lowcorn prices and low yields (or highprices and high yields) at the sametime, making corn revenues inher-ently more variable. This isbecause such areas have lessimpact than the central Corn Belton national output and prices.

Revenue risk in corngrowing tends to behigher outside thaninside the Corn Belt.


How

Farm

ers

Can

Red

uce

Ris

k

25As discussed earlier, the absence ofyield risk greatly simplifies decisionmak-ing. In the extreme case, the absence ofboth yield risk and basis risk means thateither forward contracting or hedging caneliminate revenue uncertainty completely.For example, the owner of a harvested cropcan lock in a return to storage by contract-ing for a fixed price at the end of the stor-age period. When output is known for cer-tain, but basis risk is present, hedging canreduce (but not completely eliminate)income risk.

26Although hedging is used in thisexample, the results associated with for-ward contracting are similar.

Hedging an optimal level of ex-pected output (shown by the sec-ond set of bars) modestly reducesrevenue risk compared with theno-strategy case, although theimpact varies greatly across loca-tions.27 The greatest impact is inLincoln County, Nebraska, wherethe probability of income below 70percent of expected income isreduced from 8 to 2 percent. Theimpact is most pronounced in thiscounty because it has low yieldvariability due to irrigation and aweak price-yield correlation. Insuch locations, establishing anexpected price (less harvest basis)greatly reduces revenue risk.Strong yield-price correlations oryield variabilities that exceed pricevariabilities prevent hedging fromgreatly reducing risk.

Crop insurance is generally moreeffective than hedging in reducingthe risks of very low revenuesacross the four counties. Whencrop insurance alone is used by aproducer (the third bar associatedwith each county),28 the probabili-ty of very low revenues is reducedgreatly in all counties exceptLincoln County, Nebraska. In thiscounty, irrigation is widely used,which protects against yield short-falls and essentially substitutes forinsurance. In the other locations,crop insurance has an advantageover forward pricing because farm-level yields generally are relativelymore variable than prices.

The fourth set of bars for eachcounty represents risk reductionwhen 75-percent crop insurancecoverage is combined with opti-mal hedging. As shown in the fig-ure, risk is reduced substantiallyin each of the locations, and theuse of crop insurance and hedgingin concert is much more effectiveat reducing risk than either toolused alone. By protecting both

Crop insurancereduces risk morethan forward pricingin most areas, butthe greatest riskreductions areobtained by combin-ing insurance andforward pricing.


How

Farmers C

an R

edu

ce Risk

6677

27Optimal hedging, as defined for thecalculations, involves selling Decemberfutures in March, with the hedge magni-tude set at the level resulting in the great-est risk reduction. More specifically, theoptimal hedge ratio–defined as the propor-tion of expected output that is hedged tominimize risk–ranges from 40 percent inAnderson County, Kansas, to 60 percent inLincoln County, Nebraska, and Pitt County,North Carolina.

Iroquois, IL Lincoln, NE Anderson, KS Pitt, NC0

0.1

0.2

0.3No strategy

Forward price through hedging

Crop insurance

Forward price and crop insurance

Revenue insurance

Source: Estimated by ERS.

Probability of revenue less than 70 percent of expectation

Figure 14

Effect of forward pricing (hedging) and insurance on probability of revenue less than 70 percent of expectation

28We assumed 75-percent yield coverage.

yield and price, the combined useof both tools is very effective, par-ticularly in areas with a weakprice-yield correlation and highyield variability.

The combined effects of crop insur-ance and hedging on risk reduc-tion are similar to the situationwhere producers use revenueinsurance (see the last set of barsassociated with each county). (Therevenue insurance plan assumedhere is an intraseasonal guaranteebased on individual farm yieldsand a futures price projection, anddoes not include a replacementcoverage component—only a basicrevenue guarantee. Thus, it ismore similar to the IncomeProtection product than to CropRevenue Coverage.) Such coveragereduces the probability of revenuesless than 70 percent of expecta-tions to near zero, except for therisks associated with differencesbetween local prices and futuresprices at harvest (e.g., basis risk).

EEffffeeccttss oonn IInnccoommee UUnncceerrttaaiinnttyyBBeettwweeeenn YYeeaarrss

Farmers, like everyone else, faceuncertainty about future incomesas well as current income. In par-ticular, the payments required onthe substantial debt needed tofinance investments in land,machines, and equipment make aregular cash flow particularlyimportant in farming. The impor-tance of future income comparedwith current income partly dependson the farmer’s ability to borrow ordraw from savings to cover tempo-rary income shortfalls. Producerswho have low savings and littleborrowing capacity must focus oncovering their current expenses andloan obligations. In contrast, thosewith liquid savings or short-termborrowing capacity sufficient tocover temporary income shortfallsmay be more concerned with pro-tecting future income flows orwealth. The question posed in this

section is “Can the use of hedgingand crop insurance help protectagainst income variability beyondthe current year?”

The risks in farming would be sub-stantially less if outputs could beinsured and priced forward overperiods more nearly matching theexpected life of the specializedmachines and equipment requiredfor production. However, activetrading in contracts that maturemore than 18 months in the futurehas not evolved for agriculturalcommodities, and crop insurance isoffered only on the current year’syields. Thus, hedging (or forwardcontracting) and crop insurancecannot directly assure farmers’incomes beyond 9 to 18 months.

At the same time, the use of cropinsurance may indirectly help farm-ers stabilize their incomes aroundlonger term trends. For example, aproducer can anticipate expectedyields in future years becauseinsured yields change only gradual-ly as each new yield is added to thefarm’s yield history. Thus, knowingthat insurance will be available infuture years can reduce uncertaintyabout those future years’ incomes,even though the farmer cannot yetobtain such insurance.

The multiyear variability issuessurrounding hedging (or forwardcontracting) are more complex.Unlike crop insurance, where yieldguarantees change only graduallyover time, the preplanting futuresprice quotes at which hedges canbe made often vary markedlyacross years, depending on old cropstocks and anticipated demand.Tomek and Gray concluded thatforward prices before planting aremore stable from year to year thanharvest prices for commoditiesthat cannot be stored between cropyears, such as potatoes. They alsoconcluded that little stability wasto be gained by forward pricingstorable crops, such as grains.

Combinations ofcrop insurance andforward pricing areclose substitutes forrevenue insurance inreducing risk.


How

Farm

ers

Can

Red

uce

Ris

k

Figures 15 and 16 illustrate howfutures prices for harvest deliveryvaried over the season for corn andsoybeans for the years 1977-96.The charts suggest that the pricesat which those crops can behedged vary almost as much fromyear to year as harvest prices.

There is, however, a weak tenden-cy for harvest futures prices forcorn and soybeans to be less vari-able in January than at contractmaturity (tables 17 and 18),although this does not seem to bethe case for July wheat in the pre-ceding August (table 19).

Forward selling haslittle effect on theyear-to-year variabili-ty in prices receivedby corn and soybeanproducers.


How

Farmers C

an R

edu

ce Risk

6699

400

500

600

700

800

900

1,000

1,100

1978 80 82 84 86 88 90 92 94 96

Cents per bushel

Source: Constructed by ERS from Chicago Board of Trade prices.

Figure 16

First of month price for November soybean futures, 1977-96 contracts

150

200

250

300

350

400

450

1978 80 82 84 86 88 90 92 94 96

Cents per bushel

Source: Constructed by ERS from Chicago Board of Trade prices.

Figure 15

First of month price for December corn futures, 1977-96 contracts

Lack of futures contracts withmore distant maturities has led tothe consideration of using futurescontracts that mature in the cur-rent year to hedge subsequentyear’s production. This could beaccomplished by selling futurescontracts to cover more than 1year’s crop production, and thensequentially rolling over thefutures positions to later years asthe later maturing contractsbecame available for trading.Gardner concluded in 1989 thatsuch rollover hedging would not bevery effective in reducing risk.

Indeed, the difficulties arisingwhen this year’s futures contractsare used to hedge next year’scrops was demonstrated in 1996.Some corn and soybean producerswho had entered hedge-to-arrivecontracts on their 1995 crops, andlost out on the subsequent pricerise, hoped to roll over the con-tracts to allow delivery of 1996crops. How-ever, the low prices offutures on 1996 crops comparedwith 1995 crops made suchrollovers unprofitable. Moreover,the elevators involved were anx-ious to settle the contracts, close

out their futures hedges, andrecover the large margin depositsthat had been required.

EEffffeeccttss ooff HHeeddggiinngg,, FFoorrwwaarrddPPrriicciinngg,, aanndd IInnssuurraannccee oonnAAvveerraaggee RReettuurrnnss

While many risk managementtools involve trading off consider-able expected return to reducerisk, forward pricing and insur-ance often can reduce risk with lit-tle or no sacrifice in averagereturns. Indeed, the use of cropinsurance may increase averagereturns over time for many farm-ers due to the Government’s subsi-dization of many crop and revenueinsurance products. The costs offorward pricing, while not zero,generally are small. Thus, thefarmer’s optimal forward pricingand insurance strategy often canbe closely approximated by mini-mizing risk.

The subsidization of crop and rev-enue insurance policies creates aninteresting situation regarding theexpected returns to producers.Crop (and revenue) insurance pre-miums are set so that expected

Farmers’ costs forhedging or forwardpricing are small


How

Farm

ers

Can

Red

uce

Ris

k

Table 17—Mean and standard deviation of first-of-month December corn futuresprices, 1977-96 Statistic January July December

Dollars per bushelMean 2.62 2.73 2.54

Standard deviation 0.41 0.52 0.52

Source: Calculated by ERS from Chicago Board of Trade data.

Table 18—Mean and standard deviation of first-of-month November soybeanfutures prices, 1977-96 Statistic January July November




Table 19—Mean and standard deviation of first-of-month July wheat futuresprices, 1978-97 Statistic August January July




indemnities are approximately inbalance with total premium. Totalpremium is shared between thefarmer and the Government, withthe Government paying 41.7 per-cent of the total premium cost atthe 65/100 coverage level. In addi-tion, the Government pays for anyexcess of indemnities over totalpremium (excess losses) in theevent of disasters. Thus, subsidiza-tion of these programs causesindemnities to exceed producer-paid premiums, resulting inincreased average returns to farm-ers as a group over the long run.

Based on these concepts, an indi-cator of a farmer’s expected returnto obtaining crop-yield insuranceor crop-revenue insurance is theamount of indemnity per dollar ofproducer-paid premium. For cropyears 1994-97, this return measure(in aggregate, across all crops andregions) ranged from $0.87 to$2.14 per dollar of farmer-paidpremium (table 20). Over time,rates charged for insurance havebeen increased to bring expectedindemnities closer in line withtotal premiums, reducing over thelong run the likelihood of substan-tial excess losses to the Govern-ment. When total indemnities aredivided by total premiums (includ-ing the subsidy), the resulting datain recent years are well below 1.0.

As can be seen from the subsidystructure, those producers whoconfront the highest total premium

receive the largest premium sub-sidy because the dollar value ofthe premium subsidy is calculatedas a percentage of total premium.Total per acre premium may behigh due to the yield risk associat-ed with production of that crop inthe area, the value of the crop, andother factors. Figure 17 maps theper acre premium subsidy forwheat, in dollar value terms,across major growing areas in theUnited States. As examples, farm-ers in Montana received a premi-um subsidy of $2.85 per acre for65/100 coverage, while wheatgrowers in North Dakota receivedan average subsidy at the 65/100level of about $2.66 per acre.

In contrast to crop and revenueinsurance, where government sub-sidies result in increased incomeson average for most participants,hedging and forward contractingmay lower average incomes due tocommissions or other costs, or fromslightly lower prices received (seeearlier box, “The Cost of ForwardPricing,” p. 35). On the other hand,forward pricing may raise farmers’income if any of the following hold:(1) the farmer can time sales to hithigher than average prices; (2)reduced risks obtained throughforward pricing allow the farmerto borrow at lower interest ratesand/or expand operations; or (3)price information provided byfutures quotes enables the farmerto make better decisions aboutproduction or storage.

Because of subsidiza-tion, crop-yield insur-ance and crop-rev-enue insuranceresult in increasedaverage returns tofarmers as a groupover the long run.


How

Farmers C

an R

edu

ce Risk

7711

Table 20—Indemnities, premiums, and loss ratios, 1994-97 1

Item 1994 1995 1996 1997

Million dollarsIndemnities:

Total 601 1,400 1,342 947

Premiums:Total 949 1,087 1,409 1,425Producer-paid 694 654 856 872

Ratio

Loss ratios (no units):Based on total premiums 0.63 1.29 0.95 0.66Based on producer-paid premiums 0.87 2.14 1.57 1.09

1Data for 1995-97 are based on buyup policies only.

Source: Calculated by ERS from USDA, Risk Management Agency, Crop Insurance Actuarial InformationPublic Access Server, FTP download page, http://www.act.fcic.usda.gov/actuarial/ftpdwnld.html

If the farmer can forecast futuresprice changes, hedges can be timedand adjusted to take advantage ofthe forecast. For example, if thefutures price is expected to rise, asmaller short hedge or no hedgemay be indicated. Alternatively, ifthe futures price is expected tofall, a larger short hedge may beindicated. The possibilities for tim-ing trades are unlimited, rangingfrom hedging as soon as thefutures or options contract isopened for trading on the ex-change to simply selling at deliv-ery. Futures contracts for corn,soybeans, and wheat are now list-ed for trading up to 3 years beforedelivery, although trading volumegenerally is light for contractsmaturing more than 18 months inthe future. Moreover, a farmermight change his/her futures posi-tion several times during the year,particularly if he/she were veryconfident in predicting prices.

To profitably time hedges requiresthe same price forecasting skills aspure speculation. Forecasts can bederived in many ways, rangingfrom simply looking for repeating

patterns in price behavior (techni-cal analysis) to analyzing supply/demand conditions (fundamentalanalysis). Grossman and Stiglitznoted that when information iscostly, markets cannot reflect allpossible information, which leavesroom for speculative profits forthose with superior access to infor-mation or analytical ability.Indications are that some specula-tors do profit, but the majorityapparently lose (Zulauf and Irwin).

The difficulties in forecastingfutures price changes can be visu-alized by examining historicalfutures price behavior, whichreflects traders’ expectations forthe market price at contract matu-rity. As new information becomesavailable in the market, expecta-tions change. Seasonal movementsin the December corn andNovember soybean futures con-tracts for 20 years are shown infigures 15 and 16, with each linein the figures connecting succes-sive beginning-of-month futuresprices for one contract over the 12months preceding the contract’smaturity date. Gaps in the lines

Farmers and otherhedgers may profitby adjusting posi-tions to take advan-tage of anticipatedprice changes—butonly if they havesuperior price fore-casting ability.


How

Farm

ers

Can

Red

uce

Ris

k

Figure 17

Illustration of the per acre premium subsidy, for wheat implicit in 65/100 crop insurance coverage, selected counties and States, 1996

Subsidy per acreLess than $1.00$1.01 to $2.00$2.01 to $3.00Greater than $3.00

Note: Shaded areas include counties with at least 500 acres planted to wheat.Source: Estimated by ERS from USDA, Risk Management Agency, electronic experience and yielddatabase, 1997.

reflect the transition to each suc-cessive year’s harvest contracts.

The figures show that futuresprice movements over the seasondiffer markedly from year to year.In 13 of the 20 years, for example,prices fell from January (the firstmonth of trading) to December forthe corn contract, and fromJanuary to November for the soy-bean contract. The averagedeclines were $0.08 for corn and$0.08 for soybeans (see tables 17and 18). In comparison, the Julywheat futures prices increasedfrom the preceding August to Julyin 11 of 20 years by an average of$0.05 (table 19).

The figures suggest that corn andsoybean futures prices tend to peakin midsummer. For example, theaverage price of the December cornfuture on July 1 during the 1977-96 period was $2.73 per bushel,while the average price on Decem-ber 1 was $2.54. Producers whoroutinely sold forward in early Julythus averaged a $0.19 largerreturn per bushel than those whoroutinely sold immediately afterharvest in December. Novembersoybean prices exhibited a similarpattern during 1987-96, but notduring the 1977-86 period.

If such seasonal patterns contin-ued, pure speculators as well asproducers could profit easily byroutinely selling in July and buy-ing in November or December. Asmore traders followed this prac-tice, however, July prices would bedriven down and harvest priceswould be driven up, diminishingthe potential for trading profits.Indeed, such profit potentials canbe expected to virtually disappear

under the intense competition offutures trading. The result wouldbe an “efficient market,” where thecurrent price captures all availableinformation about the price to beexpected at contract maturity.

There is much difference of opin-ion among those who advise farm-ers about timing sales in forwardmarkets. Numerous studies havefound possibilities for profits fromparticular strategies. Other stud-ies show that futures marketsappear to be quite efficient, leav-ing little room for profiting fromtiming trades. In reviewing thevarious studies, Zulauf and Irwinconclude that, for most producers,such strategies have limited abili-ty to enhance income.

Can farmers convert lower risksobtained through hedging intohigher average incomes? Thisdepends on how much risks arelowered, on the farmer’s financialsituation, and on whether his orher lender is willing to increaseloans when the farmer hedges orprices forward.

Do futures quotes provide informa-tion that farmers can use toimprove production and storagedecisions? For example, can farm-ers gain by storing if and only ifthe difference between the pricefor the future that matures at theend of the storage period exceedsthe current futures price by morethan the marginal cost of storage?Heifner (1966) found some evi-dence that this would work forstorage, as did Tomek. However,others have found little evidence tosupport this possibility for produc-ers (Irwin, Zulauf, and Jackson).

Forward pricing doesnot directly raiseaverage pricesreceived for mostfarmers, but theresulting lower risksmay allow returns tobe raised indirectly.


How

Farmers C

an R

edu

ce Risk

7733

FFaarrmmeerrss oofftteenn mmuusstt cchhoooosseebbeettwweeeenn hhiigghheerr aavveerraaggee rreettuurrnnssaanndd lloowweerr rriisskkss..

Risk management implies differ-ent things for different people,

depending on their attitudestoward risk, their financial situa-tions, and the opportunities avail-able to them. In some cases, man-aging risk involves minimizing riskfor a given level of expected outputor revenue. In other cases, itinvolves keeping risk withinbounds while seeking higherexpected returns. More generally,the goal of risk management is toobtain the best available combina-tion of expected income and incomecertainty, given the individual’sresources and risk preferences.

FFaarrmmeerrss OOfftteenn AArree WWiilllliinngg TTooAAcccceepptt HHiigghheerr RRiisskkss TToo OObbttaaiinnHHiigghheerr IInnccoommeess

Farming, like any business enter-prise, involves taking risks toobtain a higher income or highersatisfaction than might beobtained otherwise. Some farmersappear to virtually disregard risk.But for most, the amount of riskthat can be accepted is limited.Thus, risk management is not amatter of minimizing risk, but ofdetermining how much risk totake, given the farmer’s alterna-tives and preference tradeoffsbetween risk and expected return.

To use an example, consider a pro-ducer who has just harvested10,000 bushels of corn and is exam-

ining three alternatives: (1) sellingthe crop and placing the income ina certificate of deposit (CD); (2)storing the corn until March, or (3)selling the crop and using thereturns to custom feed cattle.Figure 18 shows expected outcomesfor the three strategies in terms ofexpected profit on the horizontalaxis and the probability of returnless than $25,000 on the verticalaxis. The CD provides zero proba-bility of loss and the lowest expect-ed profit, the cattle feeding alterna-tive offers the highest risk andhighest expected return, and thestorage alternative is in the middle.

For farmers having similar wealthand farming situations, the mostrisk averse would likely choose thecertificate of deposit. Those whoare less risk averse would be morelikely to choose storage. The leastrisk-averse farmers would tend tochoose feeding cattle, the riskiestchoice among the alternatives, butalso the strategy with the highestexpected return. In short, optimalchoices under risk for producers insimilar situations can differ widelyamong individuals.

CCrroopp IInnssuurraannccee aanndd FFoorrwwaarrddPPrriicciinngg GGeenneerraallllyy CCaann RReedduucceeIInnccoommee UUnncceerrttaaiinnttyy aatt VVeerryy LLooww CCoosstt

Reducing risk generally involvessome cost or reduction in expected


Som

e P

ract

ical

Asp

ects

of

Farm

Ris

kM

anag

emen

t

SSoommee PPrraaccttiiccaall AAssppeeccttss ooff FFaarrmm RRiisskk MMaannaaggeemmeenntt

To get desirable combinations of risk and return, farmersmust selectively and carefully use the various risk manage-ment tools available to them. Only tools that have goodprospects for reducing income uncertainty or increasingexpected income are candidates. This section considers cer-tain practical aspects of risk management, which are easilyoverlooked or contrary to intuition.

income. Consider Farmer Smith,for example, who is contemplatingdiversification, but knows that hisexpected net returns are maxi-mized by planting continuous cot-ton. By diversifying into othercrops, all of which have fairly sta-ble (but relatively low) yields,Farmer Smith estimates that hereduces his average net return byabout 15 percent. He calculatesthat the standard deviation in hisincome, however, is likely to beabout 20 percent lower becausethe net returns to the various dif-ferent crops he is considering areless than perfectly correlated. Inthis example, undertaking a risk-reducing strategy results in sub-stantially lower net returns toFarmer Smith, which he mustweigh relative to the benefits oflower income risk.

In contrast, strategies, such ashedging in futures, buying options,or forward contracting with a localelevator, tend to lower risk withlittle change in expected netreturns. The low cost of forwardpricing occurs because futuresprices exhibit little bias, meaningthat the price for each trade close-

ly approximates the price thenexpected to prevail when the con-tract matures. Most studies havefound little or no bias in futuresprices for commodities, such asgrains, with active trading andsubstantial long as well as shorthedging, but not all analysts agree(see Zulauf and Irwin).

Farmers who hedge directly infutures incur costs for commis-sions and interest forgone on mar-gin deposits, but these generallysum to less than 2 percent of thevalue of the product. When optionsare used, a premium must be paidbut, on average, the option holdergets the premium back as gainsfrom exercising or selling theoptions. Farmers’ costs also typi-cally are low when crops are for-ward priced through contractswith local buyers. No commissionsor margins are required from thefarmer, although the buyer typical-ly incurs such costs to hedge his orher position. Many country eleva-tors appear willing to bear thesecosts in order to assure a timelyflow of commodities into theirfacilities. Some may pass alongpart of their hedging costs to farm-

Crop insurance andforward pricing areways to lower riskwith little sacrifice inexpected returns.


Som

e Practical A

spects o

fFarm

Risk

Man

agemen

t

7755

0.00

0.01

0.02

0.03

0.04

0.05

24,500 25,000 25,500 26,000 26,500 27,000

Expected return, dollars

Probability of return < $25,000

Figure 18

Risk and return for alternative uses of a 10,000-bushel corn crop


Sell cropand buy CD

Store cornuntil March

Sell cropand customfeed cattle

ers through a wider basis and thusa slightly lower price for forwardcontracts than expected to prevailat harvest time.

Unlike most risk managementtools, crop-yield insurance and thenew crop-revenue insurance prod-ucts—which are subsidized by theGovernment—provide a specialcase where income risk is reducedand expected returns areenhanced. Because the privatecompanies delivering policies tofarmers are reimbursed directlyfor their selling expenses (whichinclude agent sales commissionsand data processing costs), suchexpenses are not incorporated inthe total premium. Moreover, thetotal premium is also subsidizedby the Government, meaning thattotal indemnities exceed farmer-paid premiums over time. As aresult, buying crop-yield or crop-revenue insurance raises averagereturns as well as reduces risk formost participating farmers. Inshort, the relationship betweenrisk and returns depends on thegiven tool or strategy, and theunique situations confronted byindividual producers.

RRiisskk RReedduuccttiioonn FFrroomm FFoorrwwaarrddPPrriicciinngg CCaann BBee QQuuiittee SSmmaallll ffoorrFFaarrmmss WWiitthh HHiigghh YYiieelldd VVaarriiaabbiilliittyyoorr SSttrroonnggllyy NNeeggaattiivvee YYiieelldd--PPrriicceeCCoorrrreellaattiioonnss

Farmers can reduce their priceuncertainty through several mech-anisms, including hedging infutures or options or entering intocash forward contracts. The effec-tiveness of these forward pricingtools, however, can vary greatly,depending on the yield risks facedby the given farmer, the interac-tions between price and yield, andthe other risk management toolsthat are used on the operation.Table 21 illustrates the effective-ness of hedging and crop insuranceon farms that have different price-yield correlations and yield vari-abilities. Although futures hedgingis used as a proxy in the table forall types of forward pricing strate-gies, results for hedging with com-modity options, or forward con-tracting, would be similar.

More specifically, the table illus-trates how crop insurance andfutures hedging work together toreduce risks for farmers in differ-

Crop insurance andhedging typically canreduce probabilitiesof revenues less than75 percent of averageby about half,depending on yieldvariability and price-yield correlation.


Som

e P

ract

ical

Asp

ects

of

Farm

Ris

kM

anag

emen

t

Table 21—Effect of futures hedges and crop insurance on the probability ofreturns less than 75 percent of expectations

Yield coefficient of variation (standard deviation/mean)Price-yield Riskcorrelation strategy 0.1 0.2 0.3 0.4 0.5

0 None 0.14 0.19 0.21 0.24 0.25MPCI .12 .15 .18 .18 .20

MPCI+hedge .06 .08 .12 .14 .18

-0.1 None .14 .17 .20 .22 .25MPCI .12 .14 .15 .17 .18

MPCI+hedge .06 .06 .11 .15 .17

-0.2 None .13 .17 .20 .21 .24MPCI .11 .13 .14 .16 .17

MPCI+hedge .06 .06 .10 .14 .16

-0.3 None .13 .15 .19 .20 .23MPC .11 .12 .13 .14 .16

MPCI+hedge .06 .06 .10 .13 .14

-0.4 None .12 .15 .18 .19 .22MPCI .10 .11 .10 .13 .14

MPCI+hedge .06 .07 .10 .11 .13

-0.5 None .11 .14 .17 .19 .20MPCI .10 .10 .09 .12 .13

MPCI+hedge .06 .07 .08 .10 .11Source: Estimated by ERS.

ent risk situations. Yield riskincreases from left to right acrossthe table, while negative yield-price correlations increase downthe left column. The entries in thecells reflect the probabilities ofrevenues falling below 75 percentof expectations. Price volatility isassumed to be 20 percent, regard-less of yield variability or theprice-yield correlation.

Because the table is constructedusing a wide range of parameters,it applies to many different farm-ing situations. Corn producers inthe Corn Belt, who confront fairlylow yield variability and a stronglynegative price-yield correlation,tend to lie near the lower left cor-ner of the table. Dryland wheatgrowers and corn growers in areasdistant from the Corn Belt tend tolie near the upper right corner.Producers who irrigate tend to lienear the upper left corner, as theyexperience low yield variabilityand tend to be outside major pro-ducing areas. The lower right cor-ner, in contrast, is of minor inter-est because yield-price correlationsare generally not strong whereyield variability is highest.

Three risk management strategiesare shown: no insurance or hedg-ing (denoted by “none” in thetable), crop insurance at the 75-percent yield coverage level(“MPCI”), and crop insurance com-bined with a minimum-riskfutures hedge (“MPCI + hedge”).Risk-reducing effectiveness can begauged by comparing the probabil-ities of low revenues across the dif-ferent strategies. As expected, therisk-reducing effectiveness of cropinsurance increases as yieldsbecome more variable. In contrast,the added risk reduction obtainedby hedging an insured crop dimin-ishes as the yield coefficient ofvariation increases. Thus, theeffects of changes in yield variabil-ity or price-yield correlation on thetotal risk reduction obtained frominsurance and hedging can differ

substantially for farmers in differ-ent situations.

EElliimmiinnaattiioonn ooff DDeeffiicciieennccyyPPaayymmeennttss IInnccrreeaasseess RRiisskkss ffoorrMMaannyy,, BBuutt NNoott NNeecceessssaarriillyy AAllll,,PPrroodduucceerrss ooff PPrrooggrraamm CCrrooppss

Replacing commodity programs(deficiency payments and supplymanagement programs) with fixedAgricultural Market TransitionAct (AMTA or contract) paymentsand planting flexibility in the 1996Farm Act dramatically altereddecades of significant governmentintervention in the markets forprogram crops.29 Deficiency pay-ments were in effect between 1973and 1995, and provided compensa-tion in years of low prices by pay-ing farmers the difference, if posi-tive, between a pre-establishedtarget price, and the higher of theaverage market price for the cropover a specified period or the loanrate. These payments averagedover $5 billion annually between1990 and 1995, and accounted formore than one-half of total farmprogram outlays over that period(USDA, 1998). Their eliminationhas raised concerns about greaterrisk in farming, with someobservers arguing that the elimi-nation of deficiency payments—and their replacement with “con-tract payments”—removes thesafety net in low-price years(Conrad).

One recent study, however, indi-cates that the effectiveness of defi-ciency payments in stabilizingincome risk varied, depending onthe correlation between individualand aggregate yields and the rela-tionship between aggregate yieldsand prices (Glauber and Miranda).The study suggests that deficiencypayments were least effective instabilizing farmers’ incomes in

Negative price-yieldcorrelations reducethe effectiveness ofdeficiency paymentsin stabilizing farm-ers’ incomes.


Som

e Practical A

spects o

fFarm

Risk

Man

agemen

t

7777

29Program crops eligible for deficiency pay-ments between 1973 (the start of the pro-gram) and 1995 included corn, sorghum,barley, oats, wheat, cotton, and rice.

areas where farm-level yields andprices are strongly negatively cor-related. In the study, about 29 per-cent of corn acreage and 26 percentof wheat acreage was found to belocated in counties where revenueswere destabilized by deficiencypayments. In several situations—such as that for Illinois corn andNorth Dakota wheat—the propor-tion of output in counties whereincome was destabilized wasgreater than 50 percent (table 22).

The relationship between localyields and prices is significantlyrelated to the effectiveness of defi-ciency payments. In major produc-ing areas, high prices tend to offsetlow yields (which can be stronglycorrelated with national yields),and vice versa. In the absence ofdeficiency payments, this relation-ship tends to stabilize revenuesand is termed the “natural hedge.”In major producing areas wherethe natural hedge is strong, defi-ciency payments may actuallyincrease income variability by pro-viding producers higher-than-aver-

age incomes in high-yield (low-price) years, while having only asmall effect in low-yield (high-price) years. Outside major grow-ing areas, the natural hedge isweaker, and deficiency paymentstend to stabilize incomes comparedwith situations where producersdepend only on the market. Thus,in markets where incomes areinherently most variable, deficien-cy payments, by stabilizing price,work to reduce revenue risk.

Demand considerations are alsoimportant in judging the risk-reducing effectiveness of deficiencypayments. Such payments mayprovide even producers in majorgrowing areas some protectionagainst prolonged slumps indemand, such as might accompanya worldwide downturn in economicconditions. Thus, deficiency pay-ments would likely be relativelymore effective in protecting pro-ducer incomes in years like 1998,when large worldwide suppliesand low prices resulted in weakdemand for several U.S. crops.

Research has foundthat deficiency pay-ments reduced farmrevenue variabilityon less than one-third of U.S. corn andwheat farms.


Som

e P

ract

ical

Asp

ects

of

Farm

Ris

kM

anag

emen

t

Table 22—Effects of deficiency payments on farm revenue variability

Crop State Average percentage of production destabilized

Percent

Corn Iowa 33.1Illinois 50.7

Nebraska 25.1Minnesota 35.9

Indiana 23.1

Ohio 4.0South Dakota 52.3

Wisconsin 19.4Missouri 53.8Michigan 0.0

United States 28.9

Wheat Kansas 37.4North Dakota 74.6

Montana 1.4Oklahoma 3.6

South Dakota 46.9

Texas 15.7Colorado 6.5

Minnesota 74.9Washington 0.0Nebraska 9.4

United States 25.5Source: Excerpted by ERS from Glauber, Joseph W., and Mario J. Miranda, “Price Stabilization, RevenueStabilization, and the Natural Hedge,” Working Paper, U.S. Dept. Agr., Office of the Chief Economist, and OhioState University, October 30, 1996.

SSpprreeaaddiinngg SSaalleess BBeeffoorree HHaarrvveessttTTeennddss TToo RReedduuccee RRiisskk,, WWhhiilleeSSpprreeaaddiinngg SSaalleess AAfftteerr HHaarrvveessttTTeennddss TToo IInnccrreeaassee RRiisskk

Spreading sales over time appears,on the surface, to be a form of diver-sification, which is a sound meansfor reducing risk. Indeed, spreadingpre-harvest sales of a crop over timeto increase the amount forwardpriced as yield or output becomesmore certain can reduce risk forproducers in many cases. The risk-reducing effectiveness of such for-ward pricing actions depends on thedegree of price-yield correlation andthe yield variability confronted bythe farmer. For most farmers, theminimum-risk forward sale at thetime of planting is no more than 70percent of the expected crop, and itmay approach zero for farmers witha strong “natural hedge” or veryhigh yield variability (Grant; Millerand Kahl; Lapan and Moschini;Coble and Heifner). It is higher forfarmers with low yield variability orwho carry crop insurance.

In contrast, once production isknown with certainty, risk is mini-mized by fixing the price regard-less of the time of delivery, if acompetitive forward market isavailable. This is because forwardprices tend to follow “randomwalks,” meaning that successivechanges are determined by chanceand independent of one another.When producers postpone estab-lishing the price for such activitiesas grain storage or livestock feed-ing, where output is known, riskactually is increased because thefinal price realized equals the cur-rent price, plus a series ofunknown random price changes.Postponing the pricing of all orpart of an assured output makessense only for producers who confi-dently expect that prices will risein the future. In such cases, thefarmer takes greater risk in thehope of obtaining a higher expect-ed return.

FFoorrwwaarrdd PPrriicciinngg MMaayy HHeellppRReedduuccee PPrriiccee UUnncceerrttaaiinnttyy NNoottOOnnllyy iinn tthhee CCuurrrreenntt YYeeaarr,, bbuutt AAllssoo iinn FFuuttuurree YYeeaarrss

Forward pricing with futures,options, or cash forward contractsreduces or eliminates price uncer-tainty between the time the for-ward sale is made and deliverytime. It serves farmers mainly as atool for reducing uncertainty aboutprices for commodities to be sold,or bought, within the year. How-ever, forward pricing offers someopportunities to reduce priceuncertainty over a longer horizon.Futures for corn and soybeans arenow traded up to 2 or more yearsahead of maturity, allowing farm-ers to forward price more than 1year’s crop. However, low tradingvolume in the later maturing con-tracts means that hedgers must bemore concerned about liquidityand possible price bias.

Another possibility is to hedgefuture years’ anticipated produc-tion in contracts that mature thisyear or next, and then roll over thepositions to contracts that maturein successive years as they becomeavailable for trading. For example,a farmer might sell contracts inthis year’s harvest time futures tocover parts of several future years’expected crops, and then succes-sively roll over the contracts tolater maturing contracts thatwould be bought back as eachfuture year’s crop is marketed.Although superficially appealing,this strategy holds little promiseeither for increasing averagereturns or decreasing risk. It isineffective in projecting a highprice for the current year’s cropinto future years because rollingover the contract generally wouldinvolve buying the old crop futureat a high price and selling the newcrop future at a lower price.Moreover, such interyear rolloverstrategies hold little promise forreducing risk due to the variability

“Rolling over” futurespositions to coverparts of severalfutures years’expected crops holds little promiseeither for increasingaverage returns orreducing risk.


Som

e Practical A

spects o

fFarm

Risk

Man

agemen

t

7799

of interyear spreads (price differ-ences between contracts maturingin different years) (see Gardner,1989). In addition, trading costswould be substantial.

Simply forward pricing each year’sexpected output before plantingreduces uncertainty about returnsin future years, to the extent thatplanting time forward pricesdiverge less from longrun equilib-rium prices than do harvest timeprices. Tomek and Gray showedthat such forward selling wasmore effective in stabilizingreturns for nonstorable commodi-ties, such as potatoes, than forstorable commodities, such asgrains and oilseeds. For storables,a large or small crop tends toaffect prices for more than 1 yearbecause stocks are carried fromone year to the next. The currentyear’s price is affected to the great-est extent by a very large or verysmall crop, but the impacts canresonate over a period of years.Table 23 shows that corn and soy-bean futures prices for harvestdelivery have been slightly lessvariable from year to year inMarch than at harvest.

The effectiveness of forward pric-ing in reducing uncertainty aboutreturns in future years depends onyield variability and the yield-price correlation, as previouslyshown for current-year risks.Finally, forward pricing cannotprotect against longer term varia-tions in demand, such as mightarise from business cycles.

FFuuttuurreess PPrriicceess PPrroovviiddeeIInnffoorrmmaattiioonn UUsseeffuull iinn MMaakkiinnggPPrroodduuccttiioonn aanndd SSttoorraaggeeDDeecciissiioonnss

Futures prices, which representthe best estimates of well-informedtraders at a given point in time,reflect the foreseeable effects ofpotential production adjustments.Thus, no forward pricing rulebased on price levels, or price lev-els relative to costs, is likely to beconsistently profitable for eitherhedgers or speculators (see Zulaufand Irwin). In other words, thefarmer who bases forward pricingdecisions on future price levelsgenerally takes on more priceuncertainty than necessary withlittle assurance of a higher aver-age return.

Although the level of futuresprices relative to costs provides lit-tle guidance about whether toprice forward, it does provideinformation useful in decidingwhether to produce or store. Theappropriate rule is “produce (orstore) when variable costs can becovered,” not “price forward onlywhen costs can be covered.”Variable production costs are thosecosts, such as for seed, fertilizer,custom work, and rent for land orstorage space, that vary with thelevel of output. This contrasts withfixed costs, such as interest anddepreciation on buildings andequipment, which must be metregardless of the level of output. Ifthe price covers both fixed andvariable costs, production likelywill be profitable. If it covers vari-able costs, but not fixed costs, lossis minimized by producing. If it

Forward pricing ismore effective in sta-bilizing returns fromnonstorable thanfrom storable crops.


Som

e P

ract

ical

Asp

ects

of

Farm

Ris

kM

anag

emen

t

Table 23—Standard deviations of first-of-month prices for harvest-time futuresin March and at harvest-time, 1977-96 1

Month December corn contract November soybean contract

Dollars per bushelMarch 0.41 0.77

Last full month of trading .51 1.111Calculations are based on annual observations.


does not cover variable cost, loss isminimized by not producing.

Forward markets allow price set-ting and delivery at differenttimes. Being able to price forwardgives farmers opportunities to lockin returns when storage is prof-itable. Indeed, farmers can usemarket signals, together with for-ward pricing, to both increasetheir profits and reduce the risksassociated with crop storage. Forexample, farmers can profitablystore crops after harvest if theirown storage costs are lower thanthe “market price of storage.” Theprice of storage offered by the mar-ket is indicated by spreadsbetween futures prices for succes-sive months.

To illustrate, suppose that it isNovember and a farmer with newcrop corn in a local elevator isdeciding whether to store the cornuntil March. The local price forcorn is $2.40 a bushel, the Marchfutures contract is trading for$2.75, and the expected basis inMarch is $0.20 under. The elevatorstorage charge is $0.025 perbushel per month, all of which isvariable cost since it can be avoid-ed by not storing. Should thefarmer store or sell the corn? Theexpected return from storage is$2.75 - $0.20 - $2.40, or $0.15 perbushel. The cost of bin space for 4months is 4 * $0.025, or $0.10 perbushel. In addition, the farmermust cover the cost of interest onthe grain. If the farmer’s interestrate is 6 percent, the interest costper bushel is 0.06 * 1/3 year *$2.40 = $0.05 per bushel. Thus,storage is a break-even operationat current prices. This is becausethe cost of bin space plus interest($0.10 + $0.05 = $0.15) equals theexpected return from storage. Ifthe local cash price should fall rel-ative to the March future, thenstorage would be profitable; other-wise, it is not profitable. The pro-ducer can minimize the risk ofstorage by entering a fixed price

forward sale for delivery in March,or by selling March futures con-tracts in an amount equal to thequantity in storage.

The importance of variable costscan also be illustrated by consider-ing a second producer, who is hold-ing corn in his own bins, which oth-erwise would be empty. The costs forthe bin, including interest, deprecia-tion, and insurance, are fixed andcannot be avoided by not storing.This farmer’s variable (or added)cost for storing an additional 4months is $0.02 per bushel forinsurance on the grain and insectcontrol plus $0.05 per bushel forinterest, equaling $0.07 per bushel.The expected return above variablecost for storage is $0.15 - $0.07, or$0.08 per bushel. If the $0.08 morethan covers the total cost of the bin,the farmer would make a profit. Ifthe total cost associated with thebin exceeds $0.08 per bushel, theproducer would be sustaining a lossover the long run. The farmer is,however, still better off to store thanto leave the bin empty because hecan cover his variable costs, plus aportion of his fixed costs that wouldbe incurred anyway.

The return per month from cornstorage declines after harvest asmonth-to-month storage chargesaccumulate (see table 24 for anexample). In the example, a pro-ducer with storage costs of 2-1/2cents per bushel per month, forexample, might expect to store hiscrop until May. At this point intime, the per month expectedreturn to storage is $0.05 3/4divided by 2 months, or $0.0288,while the variable cost of storageis nearly equal, at $0.025. A pru-dent policy for this farmer involvesselling the May or July futurescontract when the corn was put instorage in October, and holding ituntil the expected return fromstorage no longer covers storagecosts. At that time, the producerwould sell the corn crop in thecash market and buy back the

Production orstorage is advisableonly when the for-ward price coversvariable costs.


Som

e Practical A

spects o

fFarm

Risk

Man

agemen

t

8811

Yield insurance tendsto raise, and revenueinsurance tends tolower, optimal hedgeratios, but theseeffects are small atinsurance levels upto 75 percent.


Som

e P

ract

ical

Asp

ects

of

Farm

Ris

kM

anag

emen

t

futures contract. On average, thosefarmers with the lowest costs ofstorage would want to store for alonger period of time. Unlike deci-sions about crop production or live-stock feeding, decisions regardingstorage can be reversed at anytime when the forward price nolonger covers costs.

RReevveennuuee IInnssuurraannccee GGeenneerraallllyyDDooeess NNoott FFuullllyy SSuubbssttiittuuttee ffoorrFFoorrwwaarrdd PPrriicciinngg

By protecting against both pricedeclines and low yields, revenueinsurance partially substitutes forboth forward pricing and cropinsurance. It does not in all cases,however, completely replace hedg-ing or forward contracting in pro-tecting against price declines. Thisis because revenue insurance guar-

antees no more than 75 percent ofexpected revenue (85 percent forsome commodities and locations),whereas forward pricing can guar-antee as much as 100 percent of theexpected market price. Thus, forexample, a farmer with irrigatedland and low yield risk (or a farmerwith crop insurance) might reducerisk to a greater degree by guaran-teeing 100 percent of the expectedprice with an at-the-money putoption (or a short futures hedge)than with the purchase of 75-per-cent revenue insurance.

Recent research shows how rev-enue insurance as well as cropinsurance affects risk-minimizinghedge ratios for corn producers(Coble and Heifner). The effect ofdifferent levels of yield and rev-enue insurance on optimal hedge

Table 24—Expected return from storing corn between futures delivery months,1997 crop 1

Futures Futures price, Difference from Storage Expected returncontract month 10/31/97 previous delivery month interval to storage

$ per bushel Months $ per bushel per month

December 1997 2.79 ¾ -- -- --March 1998 2.89 ¼ 0.09 ½ 3 0.0317May 1998 2.95 0.05 ¾ 2 0.0288July 1998 2.99 ¼ 0.04 ¼ 2 0.0212September 1998 2.91 -0.08 ¼ 2 -0.0412-- = Not applicable.1Estimates based on October 31, 1997, futures prices.Source: Calculated by ERS from Chicago Board of Trade data.

ratios for Iroquois County, Illinois,is illustrated in figure 19, wherethe minimum-risk hedge withoutinsurance is estimated to be 25percent (see Heifner and Coble,1998). The figure shows that with50-percent yield or revenue insur-ance, the optimal hedge is essen-tially the same as with no insur-ance. With 75-percent yield insur-

ance, the optimal hedge ratio risesto 40 percent, while it remains atnear 25 percent with 75-percentrevenue insurance. In other words,75-percent revenue insurance haslittle impact on the optimalamount to hedge. The figure showsthat higher levels of revenueinsurance, if available, wouldreduce optimal hedge ratios.


Som

e Practical A

spects o

fFarm

Risk

Man

agemen

t

8833

0.500 0.625 0.750 0.875 1.0000

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Insurance level

Optimal hedge ratio

Source: Estimated by ERS.

Yield insurance

Revenue insurance

Note: Assumes expected utility maximization for a farmer with 500 acres of corn, a $300,000 net worth, and average risk aversion.

Figure 19

Effect of insurance level on optimal hedge ratio, Iroquois County, Illinois

Ahrendsen, Bruce L., Robert N.Collender, and Bruce L. Dixon.“An Empirical Analysis ofOptimal Farm Capital StructureDecisions,” Agricultural FinanceReview 54 (1994): 108-119.

Ahsan Syed M., Ali A. G. Ali, andN. John Kurian. “Toward aTheory of AgriculturalInsurance,” American Journal ofAgricultural Economics 64(August 1982): 520-529.

Allen, Kristen. “A View ofAgriculture’s Future Through aWide-Angle Lens,” Choices(Second Quarter 1993): 34-37.

Anderson, Jock R., John L. Dillon,and J. Brian Hardaker.Agricultural Decision Analysis.Ames: Iowa State UniversityPress, 1977.

Anderson, Kim B., and Harry P.Mapp. “Risk ManagementPrograms in Extension,”Journal of Agricultural andResource Economics 21 (July1996): 31-38.

Anderson, Richard, insurance spe-cialist. Personal communication.Kansas City, MO: U.S. Dept.Agr., Risk Management Agency,October 1998.

Antle, John M. “EconometricEstimation of Producers’ RiskAttitudes,” American Journal ofAgricultural Economics 69(August 1997): 509-522.

Baquet, Alan, Ruth Hambleton,and Doug Jose. UnderstandingAgricultural Risks: Production,Marketing, Financial, Legal,and Human Resources. U.S.

Dept. Agr., Risk ManagementAgency, December 1997.

Bard, Sharon K., and Peter J.Barry. “Developing a Scale forAssessing Farmers’ RiskAttitudes,” paper presented atU.S. Dept. Agr., Econ. Res. Serv.,Washington, DC, November 4,1998.

Barry, Peter J., professor. Personalcommunication. Urbana, IL:University of Illinois,Department of AgriculturalEconomics, June 1998.

____________, and C. B. Baker.“Financial Responses to Risk inAgriculture,” Risk Managementin Agriculture. Ed. Peter J.Barry. Ames: Iowa StateUniversity Press, 1984, pp. 183-199.

____________, C. B. Baker, andLuis R. Sanint. “Farmers’ CreditRisks and Liquidity Manage-ment,” American Journal ofAgricultural Economics 63 (May1981): 216-227.

____________, and Lindon J.Robison. “Economic VersusAccounting Rates of Return forFarm Land,” Land Economics62 (November 1986): 388-401.

____________, Steven T. Sonka, andKaouthar Lajili. “VerticalCoordination, FinancialStructure, and the ChangingTheory of the Firm,” AmericanJournal of AgriculturalEconomics 74 (December 1992):1219-1225.

____________, and David R.Willman. “A Risk Programming


Ref

eren

ces

RReeffeerreenncceess

Analysis of Forward Contractingwith Credit Constraints,”American Journal ofAgricultural Economics 58(February 1976): 62- 70.

Bartlett, Peggy F. “Motivations ofPart-time Farmers,” MultipleJob Holding Among FarmFamilies. Ed. M. C. Hallberg, JillL. Findeis, and Daniel A. Lass.Ames: Iowa State UniversityPress, 1991.

Batte, Marvin T., Gary D.Schnitkey, and Eugene Jones.“Sources, Uses, and Adequacy ofMarketing Information forCommercial Midwestern CashGrain Farmers,” North CentralJournal of AgriculturalEconomics 12 (July 1990): 188-196.

Beneke, Raymond R. “ProgressComes Dressed in StrangeClothing,” Choices (FirstQuarter 1998): 42-44.

Ben-Porath, Yoram. “TheProduction of Human Capitaland the Life Cycle of Earnings,”Journal of Political Economy 75(August 1967): 352-365.

Berck, Peter. “Portfolio Theory andthe Demand for Futures: TheCase of California Cotton,”American Journal ofAgricultural Economics 63(August 1981): 466-474.

Besant, Lloyd, editor. CommodityTrading Manual. Chicago:Chicago Board of Trade, 1985.

Bessembinder, Hendrick. “AnEmpirical Analysis of RiskPremia in Futures Markets,”Journal of Futures Markets 13(September 1993): 611-630.

Binswanger, Hans P. “AttitudesToward Risk: ExperimentalMeasures in Rural India,”American Journal of

Agricultural Economics 62(August 1980): 395-407.

Black, Fischer. “The Pricing ofCommodity Contracts,” Journalof Financial Economics 3(September 1976): 167-179.

Blank, Steven C. “The New RiskEnvironment in CaliforniaAgriculture,” Agribusiness: AnInternational Journal 11(March/April 1995): 155-168.

____________, Colin A. Carter, andJeffrey McDonald. “Is theMarket Failing AgriculturalProducers Who Wish to ManageRisks?” Contemporary EconomicPolicy 15 (July 1997): 103- 112.

____________, and JeffreyMcDonald. “Preferences for CropInsurance When Farmers AreDiversified,” Agribusiness: AnInternational Journal 12(November/December 1996):583-592.

Bodie, Zvi, and Robert C. Merton.Finance. Upper Saddle River,NJ: Prentice Hall, 1998.

Boehlje, Michael D., and Larry D.Trede. “Risk Management inAgriculture,” Journal of theAmerican Society of FarmManagers and Rural Appraisers41 (1977): 20-29.

Boisvert, Richard N., andChristine Ranney. “Accountingfor the Importance of NonfarmIncome on Farm Family IncomeInequality in New York,”Northeastern Journal ofAgricultural and ResourceEconomics 19 (April 1990): 1-11.

Campbell, Gerald R., and AmrShiha. Wisconsin Corn andSoybean Producers’ Knowledgeand Use of Options and RelatedMarketing Instruments. WorkingPaper No. 276. Madison:University of Wisconsin,


Referen

ces

8855

Department of AgriculturalEconomics, December 1987.

Catania, Patrick J., editor. Grains:Production, Processing,Marketing. Chicago: ChicagoBoard of Trade, 1992.

Chambers, Robert G. “Insurabilityand Moral Hazard inAgricultural InsuranceMarkets,” American Journal ofAgricultural Economics 71(August 1989): 604-616.

Charlier, Marj. “The BroilerBusiness Consolidates, and Thatis Bad News to Farmers,” WallStreet Journal (January 4,1990): 1.

Chicago Board of Trade. Contractlistings and quotes.http://www.cbot.com/index2.htmlAccessed in September 1998.

____________. Daily futures pricequotes. Compiled by TechnicalTools, Inc., financial data down-loading and charting system. TTChartbook 95. Boca Raton, FL.Accessed on various dates in1997 and 1998.

Cleaveland, Barry, senior vicepresident. Personal communica-tion. Council Bluffs, IA:American Agrisurance, Inc.,October 1998.

Coble, Keith, and Joy Harwood.“The Effects of the 1996 FarmAct on Agricultural RiskManagement.” Organized sym-posia presentation at theAmerican AgriculturalEconomics Association annualmeeting, Toronto, Canada, 1997.

____________, and Richard Heifner.“The Effect of Crop or RevenueInsurance on Optimal Hedging,”Research SymposiumProceedings, NCR-134Conference on AppliedCommodity Price Analysis,Forecasting, and Market Risk

Management, Chicago, April 20-21, 1998.

____________, Thomas O. Knight,Rulon D. Pope, and Jeffrey R.Williams. “An ExpectedIndemnity Approach to theMeasurement of Moral Hazardin Crop Insurance,” AmericanJournal of AgriculturalEconomics 79 (February 1997):216-226.

Collins, Robert A. “ExpectedUtility, Debt-Equity Structure,and Risk Balancing,” AmericanJournal of AgriculturalEconomics 67 (August 1985):627-629.

____________, and Edward E. Gbur.“Risk Analysis for Proprietorswith Limited Liability: A Mean-Variance Safety-FirstSynthesis,” Western Journal ofAgricultural Economics 16 (July1991): 156-162.

Commodity Futures TradingCommission. The CFTCGlossary: A Layman’s Guide tothe Language of the FuturesIndustry. CFTC P-105. RevisedJanuary 1997. http://www.cftc.gov/glossary.html

Conrad, Senator Kent.Congressional Record:Proceedings and Debates of the104th Congress, Second Session.Vol. 142, No. 17, S1046,February 7, 1996.

Crisostomo, Mario F., and Allen M.Featherstone. “A PortfolioAnalysis of Returns to FarmEquity and Assets,” NorthCentral Journal of AgriculturalEconomics 12 (January 1990): 9-21.

Curtis, Charles Jr., Kandice H.Kahl, and Cathy S. McKinnell.Optimal Soybean MarketingStrategies: The South CarolinaCase. Working Paper. Clemson,SC: Clemson University, 1989.


Ref

eren

ces

Dillon, John L., and Pasquale L.Scandizzo. “Risk Attitudes ofSubsistence Farmers inNortheast Brazil,” AmericanJournal of AgriculturalEconomics 60 (August 1978):425-435.

Dodson, Charles B. “FarmDiversification—Still aProblem?” Agricultural Incomeand Finance Situation andOutlook Report. AFO-50. U.S.Dept. of Agr., Econ. Res. Serv.,September 1993, pp. 44-46.

Dusak, Katherine. “FuturesTrading and Investor Returns:An Investigation of CommodityMarket Risk Premiums,”Journal of Political Economy 81(November/December 1973):1387- 1406.

Edelman, Mark A., Brian H.Schmiesing, and Douglas R.Olsen. “Use of SelectedMarketing Alternatives by IowaFarmers,” Agribusiness: AnInternational Journal 6 (March1990): 121-132.

Elam, Emmett. “Cash ForwardContracting versus Hedging ofFed Cattle, and the Impact ofCash Contracting on CashPrices,” Journal of Agriculturaland Resource Economics 17(July 1992): 205-217.

Ellinger, Paul N., and Peter J.Barry. “The Effects of TenurePosition on Farm Profitabilityand Solvency: An Application toIllinois Farms,” AgriculturalFinance Review 47 (1987): 106-118.

El-Osta, Hisham S., G. AndrewBernat, Jr., and Mary C.Ahearn. “Regional Differences inthe Contribution of Off-FarmWork to Income Inequality,”Agricultural and ResourceEconomics Review 24 (April1995): 1-14.

Farrell, Kenneth R. VerticalCoordination in FoodProduction, Processing, andDistribution. NationalMarketing Service Workshop,Asheville, NC, U.S. Dept. Agr.,Econ. Res. Serv., 1969.

Featherstone, Allen M., andCharles B. Moss. “QuantifyingGains to Risk DiversificationUsing Certainty Equivalence ina Mean-Variance Model: AnApplication to Florida Citrus,”Southern Journal ofAgricultural Economics 22(December 1990): 191-197.

___________, Charles B. Moss,Timothy G. Baker, and Paul V.Preckel. “The Theoretical Effectsof Farm Policy on OptimalLeverage and the Probability ofEquity Losses,” AmericanJournal of AgriculturalEconomics 70 (August 1988):572-579.

Fee, Rod. “Business: Livestock,”Successful Farming, 96 (Mid-February 1998): 21.

Fleisher, Beverly. Agricultural RiskManagement. Boulder andLondon: Lynne RiennerPublishers, Inc., 1990.

Ford, Stephen A., and Emerson M.Babb. “Farmer Sources andUses of Information,”Agribusiness: An InternationalJournal 5 (September 1989):465-476.

Frederick, Donald A. Co-ops 101:An Introduction to Cooperatives.Cooperative Information Report55. U.S. Dept. Agr., RuralBusiness Cooperatives Service,June 1997.

Freund, Rudolph J. “TheIntroduction of Risk Into aProgramming Model,”Econometrica 24 (July 1956):253-263.


Referen

ces

8877

Gabriel, Stephen C., and C. B.Baker. “Concepts of Businessand Financial Risk,” AmericanJournal of AgriculturalEconomics 62 (August 1980):560-564.

Gardner, Bruce L. “Crop Insurancein U.S. Farm Policy,” Economicsof Agricultural Crop Insurance:Theory and Evidence. Ed.Darrell L. Hueth and WilliamH. Furtan. Norwell, MA: KluwerAcademic Publishers, 1994,pp.17-44.

____________. “Determinants ofFarm Family IncomeInequality.” American Journal ofAgricultural Economics 51(November 1969): 753-769.

____________. “Rollover Hedgingand Missing Long-Term FuturesMarkets,” American Journal ofAgricultural Economics 71 (May1989): 311-318.

____________, and R. A. Kramer.“Experience with CropInsurance Programs in theUnited States,” Crop Insurancefor Agricultural Development,Issues and Experience. Ed. PeterHazell, Carlos Pomareda, andAlberto Valdes. Baltimore: TheJohns Hopkins UniversityPress, 1986, pp. 195-222.

Glauber, Joseph W., Joy L.Harwood, and Jerry R. Skees.An Alternative for ReducingFederal Crop InsuranceProgram Losses. AER-668. U.S.Dept. Agr., Econ. Res. Serv., May1993.

____________, and Mario J.Miranda. “Price Stabilization,Revenue Stabilization, and theNatural Hedge.” Working Paper.U.S. Dept. Agr., Office of theChief Economist, and OhioState University, October 30,1996.

Good, Darrel L., Scott H. Irwin,Thomas E. Jackson, andGregory K. Price. 1995 PricingPerformance of Market AdvisoryServices for Corn and Soybeans.AgMAS Project Research Report1997-01. Urbana, IL: TheAgMAS Project, University ofIllinois at Urbana-Champaign,March 1997.

Goodwin, Barry K. “Premium RateDetermination in the FederalCrop Insurance Program: WhatDo Averages Have to Say AboutRisk?” Journal of Agricultureand Resource Economics 19(December 1994): 382-395.

____________, and Ted C.Schroeder. “Human Capital,Producer Education Programs,and the Adoption of Forward-Pricing Methods,” AmericanJournal of AgriculturalEconomics 76 (November 1994):936-947.

____________, and Vincent H.Smith. The Economics of CropInsurance and Disaster Aid.Washington, DC: AmericanEnterprise Institute Press,1995.

Grant, Dwight. Optimal FuturesPositions for Corn and SoybeanGrowers Facing Price and YieldRisk. TB-1751. U.S. Dept. Agr.,Econ. Res. Serv., March 1989.

Grenchik, Mark, and JeffCampbell. “Chicago Board ofTrade Crop Yield InsuranceFutures and Options Contracts:New Tools for Managing CropYield Risk.” Presentation at the1995 Meeting of the FederalCrop Insurance Corporation/Extension Service AdvisoryCouncil, Kansas City, MO,March 15, 1995.

Grossman, Sanford J., and JosephE. Stiglitz. “On the Impossibilityof Informationally Efficient


Ref

eren

ces

Markets,” American EconomicReview 70 (June 1980): 393-408.

Hamilton, Neil D. AgricultureWithout Farmers? IsIndustrialization RestructuringAmerican Food Production andThreatening the Future ofSustainable Agriculture? Whitepaper 94-1. Des Moines, IA:Drake University Law School,Agricultural Law Center.February 1994.

____________, and Greg Andrews.State Regulation of ContractFeeding and Packer Integrationin the Swine Industry. Whitepaper 92-4. Des Moines, IA:Drake University Law School,Agricultural Law Center.November 1992.

Hamrick, Karen, issue editor.“Appendix Tables,” RuralConditions and Trends 8(September 1997): 96-108. U.S.Dept. Agr., Econ. Res. Serv.

Hardaker, J. Brian, Ruud B. M.Huirne, and Jock R. Anderson.Coping With Risk inAgriculture. New York: CABInternational, 1997.

Harris, Marshall. “VerticalCoordination Under ProductionContracts,” Entrepreneurship inAgriculture. Monograph No. 12.Iowa City: University of Iowa,College of Law, AgriculturalLaw Center. undated.

Harwood, Joy, Dick Heifner, KeithCoble, and Janet Perry.“Strategies for a New RiskEnvironment,” AgriculturalOutlook. AO-234. U.S. Dept.Agr., Econ. Res. Serv., October1996, pp. 24-30.

____________, Linwood A. Hoffman,and Mack N. Leath. “Marketingand Pricing Methods Used byMidwestern Corn Producers,”Feed Situation and OutlookReport. Fds-303. U.S. Dept. Agr.,

Econ. Res. Serv., September1987, pp. 33-39.

Hawawini, Gabriel A. “A Mean-Standard Deviation Expositionof the Theory of the Firm UnderUncertainty: A PedagogicalNote,” American EconomicReview 68 (March 1978): 194-202.

Hazell, P. B. R. “A LinearAlternative to Quadratic andSemivariance Programming forFarm Planning UnderUncertainty,” American Journalof Agricultural Economics 53(February 1971): 53-62.

Heifner, Richard G. “Implicationsof Hedging for the AgriculturalLender,” Agricultural FinanceReview 33 (July 1972a): 8-14.

____________. “Optimal HedgingLevels and HedgingEffectiveness in Cattle Feeding,”Agricultural EconomicsResearch 24 (April 1972b): 25-36.

____________. “The Gains fromBasing Grain Storage Decisionson Cash-Futures Spreads,”Journal of Farm Economics 48(December 1966): 1490-1495.

____________, and Keith Coble.“The Effects of Combining Priceand Yield Hedges and CropInsurance on Corn Growers’Risks.” Paper presented at theChicago Board of Trade 1996Annual Fall Research Seminar,Chicago, IL, 1996.

____________, and Keith Coble.“The Risk-ReducingPerformance of AlternativeTypes of Crop and RevenueInsurance When Combined withForward Pricing.” Report to theRisk Management Agency. U.S.Dept. Agr., Econ. Res. Serv.,November 1998.


Referen

ces

8899

____________, Joseph W. Glauber,Mario J. Miranda, Gerald E.Plato, and Bruce H. Wright.Futures, Options, and FarmPrograms: Report to Congress ona Study Mandated by the FoodSecurity Act of 1985. StaffReport No. AGES9003. U.S.Dept. Agr., Econ. Res. Serv.,January 1990.

____________, and RandalKinoshita. “Differences AmongCommodities in Real PriceVariability and Drift,”Agricultural EconomicsResearch 45 (Fall 1994): 10-20.

____________, and Gerald Plato.“The Efficiency of OptionsCompared to Fixed PriceContracts for Shifting RevenueRisk in Crop Production.” Paperpresented at the AmericanAgricultural EconomicsAssociation meetings, Reno, NV,July 27-30, 1986.

____________, Bruce H. Wright,and Gerald E. Plato. UsingCash, Futures, and OptionsContracts in the Farm Business.AIB-665. U.S. Dept. Agr., Econ.Res. Serv., April 1993.

Held, Larry J., and Richard A.Zink. “Farm Enterprise Choice:Risk-Return Tradeoffs for Cash-Crop Versus Crop-LivestockSystems,” North CentralJournal of AgriculturalEconomics 4 (July 1982): 11-19.

Hieronymus, Thomas A. Economicsof Futures Trading. New York:Commodity Research Bureau,Inc., 1978.

Hoffman, Linwood A., Joy L.Harwood, and Mack N. Leath.“Marketing and PricingMethods Used by Selected U.S.Wheat Producers,” WheatSituation and Outlook Report.WS-281. U.S. Dept. Agr., Econ.Res. Serv., May 1988, pp. 19-24.

Hoppe, Robert A., agriculturaleconomist. Personal communica-tion. Washington, DC: U.S. Dept.Agr., Econ. Res. Serv., October1998.

____________, Robert Green, DavidBanker, Judith Z. Kalbacher,and Susan E. Bentley.Structural and FinancialCharacteristics of U.S. Farms,1993: 18th Annual Family FarmReport to Congress. AIB-728.U.S. Dept. Agr., Econ. Res. Serv.,October 1996.

Irwin, Scott H., Carl R. Zulauf, andThomas E. Jackson. “MonteCarlo Analysis of MeanReversion in CommodityFutures Prices,” AmericanJournal of AgriculturalEconomics 78 (May 1996): 387-399.

Ishii, Yasunori. “On the Theory ofthe Competitive Firm UnderPrice Uncertainty: Note,”American Economic Review 67(September 1977): 768-769.

Jagger, Craig, and Joy Harwood.“Module 6: Crop Insurance andRevenue Insurance,” USDA’sCrop/Commodity ProgramsAfter the 1996 Farm Act. USDAGraduate School Course, winterquarter, 1999.

____________, and Joy Harwood.“Ms. Farmer’s Wheat ProgramAdventures—Before and Afterthe 1996 Farm Bill: A CaseStudy for the Classroom,”Review of AgriculturalEconomics 20 (Spring/Summer1998): 259-272.

Jenner, Mark W. “ImprovedGrower/Integrator RelationsWould Enhance PoultryIndustry,” Feedstuffs 70 (March16, 1998): 1.

Jinkins, John. “Measuring Farmand Ranch Business Diversity,”Agricultural Income and


Ref

eren

ces

Finance Situation and OutlookReport. AFO-45. U.S. Dept. Agr.,Econ. Res. Serv., May 1992, pp.28-30.

Johnson, C. Scott, and Kenneth A.Foster. “Risk Preferences andContracting in the U.S. HogIndustry,” Journal ofAgricultural and AppliedEconomics 26 (December 1994):393-405.

Jolly, Robert. “Gains fromImproved Risk Management,”Managing Change-ManagingRisk: A Primer for Agriculture.Ames: Iowa State University,Department of Economics,January 1997, pp. 45-48.

Jose, H. Douglas, and Ram S. K.Valluru. “Insights from the CropInsurance Reform Act of 1994,”Agribusiness: An InternationalJournal 13 (December 1997):587-598.

Just, Richard E., and LindaCalvin. “Moral Hazard in U.S.Crop Insurance: An EmpiricalInvestigation.” Unpublishedmanuscript. College Park:University of Maryland, Collegeof Agriculture and NaturalResources, October 1993a.

____________, and Linda Calvin.“Adverse Selection in U.S. CropInsurance: The Relationship ofFarm Characteristics toPremiums.” Unpublished manu-script. College Park: Universityof Maryland, College ofAgriculture and NaturalResources, April 1993b.

____________, and Rulon D. Pope.“Stochastic Specification ofProduction Functions andEconomic Implications,” Journalof Econometrics 7 (February1978): 67-86.

Kansas Crop and LivestockReporting Service. KansasGrain Marketing and Transpor-

tation: Data for 1984 Crop andHistorical Data for 1979-83.Topeka, KS, January 1986.

____________. Kansas GrainMarketing: Data for 1985 Crops.Topeka, KS, no date.

Kassel, Kathleen, and Robert M.Gibbs. “Rural NonfarmEarnings Edge Up,” RuralConditions and Trends 7(December 1996): 22-24, U.S.Dept. Agr., Econ. Res. Serv.

Keller, Des. “Estate Tax Peace ofMind,” Progressive Farmer,March 15, 1998, pp. 18-20.

____________, and Leigh Rigby-Adcock. “Finding Shelter Underthe Liability Umbrella,”Progressive Farmer, March 15,1998, pp. 40-41.

Kemp, Todd E., editor. HybridCash Grain Contracts:Assessing, Managing, andControlling Risk. A White Paperby the Risk Evaluation TaskForce on Hybrid CashContracts, National Grain andFeed Association, Washington,DC, 1996.

Kenyon, David, and Craig Cooper.“Selected Fundamental andTechnical Pricing Strategies forCorn,” North Central Journal ofAgricultural Economics 2 (July1980): 137-144.

Keynes, J. M. A Treatise on Money.Vol. 2. London: MacMillan andCo., 1930.

King, Robert P., and George E.Oamek. “Risk Management byColorado Dryland WheatFarmers and the Elimination ofthe Disaster AssistanceProgram,” American Journal ofAgricultural Economics 65 (May1983): 247-255.

____________, and Lindon J.Robison. “Risk Efficiency


Referen

ces

9911

Models,” Risk Management inAgriculture. Ed. Peter J. Barry.Ames: Iowa State UniversityPress, 1984, pp. 68-81.

Kinsey, Jean. “Measuring the Well-Being of Farm Households:Farm, Off-Farm, and In-KindSources of Income: Discussion.”American Journal of Agricul-tural Economics 67 (December1985): 1105-1107.

Kliebenstein, James B., and JohnD. Lawrence. “Contracting andVertical Coordination in theUnited States Pork Industry,”American Journal of Agricul-tural Economics 77 (December1995): 1213-1218.

Knight, F. H. Risk, Uncertainty,and Profit. New York: Harperand Row Publishers, 1921.(Reprinted in 1965.)

Knight, Thomas O., Ashley C.Lovell, M. Edward Rister, andKeith H. Coble. “An Analysis ofLenders’ Influence onAgricultural Producers’ RiskManagement Decisions,”Southern Journal of Agricul-tural Economics 21 (December1989): 21-33.

Knoeber, Charles R., and Walter N.Thurman. “Don’t Count YourChickens...: Risk and RiskShifting in the Broiler Industry,”American Journal of Agricul-tural Economics 77 (August1995): 486-496.

Knorr, Bryce A., editor of FarmFutures magazine. Testimonybefore the Subcommittee onRisk Management and SpecialtyCrops, U.S. House ofRepresentatives, April 10, 1997.

Koenig, Steven R., and Charles B.Dodson. “Sources of Capital forCommercial Farm Operators,”Regulatory, Efficiency, andManagement Issues AffectingRural Financial Markets.

Proceedings of the NC-207Regional Committee. Staffpaper SP0196. Ed. Bruce L.Ahrendsen. Fayetteville:University of Arkansas,Department of AgriculturalEconomics and Rural Sociology,January 1996, pp. 70-83.

Kolb, Robert W. “Is NormalBackward Normal?” Journal ofFutures Markets 12 (February1992): 75-91.

Kroll, Yoram, Haim Levy, andHarry M. Markowitz. “Mean-Variance Versus Direct UtilityMaximization,” Journal ofFinance 39 (March 1984): 47-61.

Kyle, Steven C. “The RelationBetween Farm Production andOff-Farm Income,” Agriculturaland Resource Economics Review22 (October 1993): 179-188.

Laffont, Jean-Jacques. TheEconomics of Uncertainty andInformation. Boston: MIT Press,1989.

Lapan, Harvey, and GiancarloMoschini. “Futures HedgingUnder Price, Basis, andProduction Risk,” AmericanJournal of AgriculturalEconomics 76 (August 1994):465-477.

____________, Giancarlo Moschini,and Steven D. Hanson.“Production, Hedging, andSpeculative Decisions withOptions and Futures Markets,”American Journal ofAgricultural Economics 73(February 1991): 66-74.

Larson, Donald K., and Thomas A.Carlin. “Income and EconomicStatus of People with FarmEarnings,” Southern Journal ofAgricultural Economics 6(December 1974): 73-79.

Lass, Daniel A., Jill L. Findeis, andM. C. Hallberg. “Off-Farm


Ref

eren

ces

Employment Decisions byMassachusetts FarmHouseholds,” NortheasternJournal of Agricultural andResource Economics 18 (October1989): 149-159.

Leath, Mack. “Pricing StrategiesUsed by Soybean Producers.”Selected paper presented at theSouthern AgriculturalEconomics Association annualmeeting, Orlando, FL, February2-5, 1986.

Lehman, Dave, group manager ofcommodity products. Personalcommunication. Chicago, IL:Chicago Board of Trade, October1998.

Levy, H., and H. M. Markowitz.“Approximating ExpectedUtility by a Function of Meanand Variance,” AmericanEconomic Review 69 (June1979): 308-317.

Lin, William, G. W. Dean, and C. V.Moore. “An Empirical Test ofUtility vs. Profit Maximizationin Agricultural Production,”American Journal ofAgricultural Economics 56(August 1974): 497-508.

Logan, Samuel H. “A ConceptualFramework for AnalyzingEconomies of VerticalIntegration,” American Journalof Agricultural Economics 51(November 1969): 834-848.

Love, H. Alan, and Steven T.Buccola. “Joint Risk Preference-Technology Estimation with aPrimal System,” AmericanJournal of AgriculturalEconomics 73 (August 1991):765-774.

Lucier, Gary, senior specialty cropseconomist. Personal communica-tion. Washington, DC: U.S. Dept.Agr., Econ. Res. Serv., March1998.

Luo, Haiping, Jerry R. Skees, andMary A. Marchant. “WeatherInformation and the Potentialfor Inter-temporal AdverseSelection in Crop Insurance,”Review of AgriculturalEconomics 16 (September 1994):441-451.

MacMinn, Richard D., andAlphonse G. Holtman.“Technological Uncertainty andthe Theory of the Firm,”Southern Economic Journal 50(July 1983): 120-136.

Makus, Larry D., Biing-Hwan Lin,John Carlson, and Rose Krebill-Prather. “Factors InfluencingFarm Level Use of Futures andOptions in CommodityMarketing,” Agribusiness: AnInternational Journal 6(November 1990): 621-631.

Manchester, Alden C. “TheIndustrialization of U.S.Agriculture: The Role ofContracting and VerticalIntegration.” Unpublished work-ing paper. U.S. Dept. Agr., Econ.Res. Serv., March 1998.

Mapp, Harry P., Michael L.Hardin, Odell L. Walker, andTillak Persaud. “Analysis ofRisk Management Strategies forAgricultural Producers,”American Journal of Agricul-tural Economics 61 (December1979): 1071-1077.

___________, and Glenn A.Helmers. “Methods of RiskAnalysis for Farm Firms,” RiskManagement in Agriculture. Ed.Peter J. Barry. Ames: Iowa StateUniversity Press, 1984, pp. 116-128.

Markowitz, Harry M. PortfolioSelection: Efficient Diversifi-cation of Investments. NewHaven, CT: Cowles Foundationfor Research in Economics, YaleUniversity, 1959.


Referen

ces

9933

Martin, Laura. “ProductionContracts, Risk Shifting, andRelative Performance Paymentsin the Pork Industry,” Journal ofAgricultural and AppliedEconomics 29 (December 1997):267-278.

Martinez, Steve W., and Al Reed.From Farmers to Consumers:Vertical Coordination in theFood Industry. AIB-720. U.S.Dept. Agr., Econ. Res. Serv.,June 1996.

____________, Kevin Smith, andKelly Zering. VerticalCoordination and ConsumerWelfare: The Case of the PorkIndustry. AER-753. U.S. Dept.Agr., Econ. Res. Serv., August1997.

____________, and Kelly D. Zering.“Optimal Dynamic HedgingDecisions for Grain Producers,”American Journal ofAgricultural Economics 74(November 1992): 879-888.

Maurice, Nelson, senior under-writer. Personal communication.Kansas City, MO: U.S. Dept.Agr., Risk Management Agency,October 1998.

McKinnon, R. I. “Future Markets,Buffer Stocks, and IncomeStability for PrimaryProducers,” Journal of PoliticalEconomy 75 (December 1967):844-61.

McNew, Kevin. “Testing theSuccess of the Iowa Corn YieldFutures Contract.” Paper pre-sented at the Chicago Board ofTrade Spring Research Seminar.Chicago, IL, 1996.

Mighell, Ronald L., and LawrenceA. Jones. Vertical Coordinationin Agriculture. AER-19. U.S.Dept. Agr., Econ. Res. Serv.,February 1963.

Miller, Lynn H., Paul N. Ellinger,Peter J. Barry, and KaoutharLajili. “Price and NonpriceManagement of AgriculturalCredit Risk,” AgriculturalFinance Review 53 (1993): 28-41.

Miller, Stephen E., and Kandice H.Kahl. “Performance ofEstimated Hedging RatiosUnder Yield Uncertainty,”Journal of Futures Markets 9(August 1989): 307-319.

Mintert, James. “Farmers CurrentMarketing Practices andAttitudes.” Paper presented atthe American AgriculturalEconomics AssociationExtension Workshop, EastLansing, MI, July 31-August 1,1987.

Miranda, Mario J. “Area-YieldCrop Insurance Reconsidered,”American Journal of Agricul-tural Economics 73 (May 1991):233-242.

____________, and Joseph W.Glauber. “Systemic Risk,Reinsurance, and the Failure ofCrop Insurance Markets,”American Journal ofAgricultural Economics 79(February 1997): 206-215.

Mishra, Ashok K., and Barry K.Goodwin. “Farm IncomeVariability and the Supply ofOff-Farm Labor,” AmericanJournal of AgriculturalEconomics 79 (August 1997):880-887.

____________, and Barry K.Goodwin. “Income Risk andAllocation of Labor Time: AnEmpirical Investigation.”Unpublished working paper.U.S. Dept. Agr., Econ. Res. Serv.,and North Carolina StateUniversity, January 1998.

Moscardi, Edgardo, and Alain deJanvry. “Attitudes Toward Risk


Ref

eren

ces

Among Peasants: AnEconometric Approach,”American Journal ofAgricultural Economics 59(November 1977): 710-716.

Moss, Charles B., Allen M.Featherstone, and Timothy G.Baker. “Agricultural Assets inan Efficient MultiperiodInvestment Portfolio,”Agricultural Finance Review 47(1987): 82-94.

____________, Stephen A. Ford, andWilliam G. Boggess. “CapitalGains, Optimal Leverage, andthe Probability of Equity Loss: ATheoretical Model,” AgriculturalFinance Review 49 (1989): 127-134.

____________, Stephen A. Ford, andMario Castejon. “Effect of DebtPosition on the Choice ofMarketing Strategies forFlorida Orange Growers: A RiskEfficiency Approach,” SouthernJournal of AgriculturalEconomics 23 (December 1991):103-111.

____________, and P. J. vanBlokland. Debt Choice andMarketing Instruments: TheCase of Frozen ConcentratedOrange Juice. Staff Paper 402.Gainesville: University ofFlorida, Department of Foodand Resource Economics,January 1990.

Musser, Wesley N., Harry P. Mapp,Jr., and Peter J. Barry.“Applications I: RiskProgramming,” Risk Manage-ment in Agriculture. Ed. Peter J.Barry. Ames: Iowa StateUniversity Press, 1984, pp. 129-147.

____________, J. Ohannesian, andF. J. Benson. “A Safety FirstModel of Risk Management forUse in Extension Programs,”North Central Journal of

Agricultural Economics 3(January 1981): 41-46.

____________, George F. Patrick,and David F. Eckman. “Risk andGrain Marketing Behavior ofLarge-Scale Farmers,” Journalof Agricultural Economics 18(January 1996): 65-77.

____________, and Kostas G.Stamoulis. “Evaluating the Foodand Agricultural Act of 1977with Firm Quadratic RiskProgramming,” AmericanJournal of AgriculturalEconomics 63 (August 1981):447-456.

Nartea, Gilbert V., and Peter J.Barry. “Risk Efficiency and CostEffects of GeographicalDiversification,” Review ofAgricultural Economics 16(September 1994): 341-351.

Nelson, Carl H., and Paul V.Preckel. “The Conditional BetaDistribution as a StochasticProduction Function,” AmericanJournal of AgriculturalEconomics 71 (May 1989): 370-378.

Newbery, David M. G., and JosephE. Stiglitz. The Theory ofCommodity Price Stabilization.New York: Oxford UniversityPress, 1985.

“On Our Own,” Farm Futures, 25(January 1997), special reporton safety nets and risk manage-ment.

Ortmann, Gerald F., George F.Patrick, Wesley N. Musser, andD. Howard Doster. “Use ofPrivate Consultants and OtherSources of Information by LargeCornbelt Farmers,”Agribusiness: An InternationalJournal 9 (July 1993): 391-402.

Patrick, George F. “Producers’Attitudes, Perceptions, andManagement Responses to


Referen

ces

9955

Variability,” Risk Analysis forAgricultural Production Firms:Concepts, Information Require-ments, and Policy Issues.Proceedings of a SeminarSponsored by Southern RegionalProject S-180. AE-4574. Urbana:University of Illinois,Department of AgriculturalEconomics, July 1984.

____________, and Wesley N.Musser. Sources of andResponses to Risk: FactorAnalyses of Large-ScaleCornbelt Farmers. Staff PaperNo. 95-17. West Lafayette, IN:Purdue University, Departmentof Agricultural Economics,December 1995.

____________, Wesley N. Musser,and David T. Eckman. “ForwardMarketing Practices andAttitudes of Large-ScaleMidwestern Grain Producers,”Review of AgriculturalEconomics 20 (Spring/Summer1998): 38-53.

____________, and StantonUllerich. “Information Sourcesand Risk Attitudes of Large-Scale Farmers, Farm Managers,and Agricultural Bankers,”Agribusiness: An InternationalJournal 12 (September/October1996): 461-471.

____________, Paul N. Wilson,Peter J. Barry, William G.Boggess, and Douglas L. Young.“Risk Perceptions andManagement Responses:Producer-Generated Hypothesesfor Risk Modeling,” SouthernJournal of AgriculturalEconomics 17 (December 1985):231-238.

Perry, Janet, editor. “AdaptiveManagement Decisions—Responding to the Risks ofFarming.” Unpublished workingpaper. U.S. Dept. Agr., Econ. Res.Serv., December 1997.

Perry, Martin K. “VerticalIntegration: Determinants andEffects,” Handbook of IndustrialOrganization, Volume I. Ed.Richard Schmalensee andRobert D. Willig. Oxford,England: Elsevier SciencePublishers Ltd., 1989, pp. 183-255.

Persaud, Tillak, and Harry P.Mapp. “Analysis of AlternativeProduction and MarketingStrategies in SouthwesternOklahoma: A MOTADApproach,” Risk Analysis inAgriculture: Research andDevelopments. Staff Paper No.AE-4492. Urbana: University ofIllinois, Department ofAgricultural Economics, 1980.

Pierce, J. L. “Commercial BankLiquidity,” Federal ReserveBulletin 52 (1966): 1093-1102.

Plain, Ronald, livestock extensionspecialist. Personal communica-tion. Columbia, MO: Universityof Missouri, Department ofAgricultural Economics, April1998.

Plato, Gerald E. “Effectiveness ofFutures and Options inReducing Farmer RevenueRisk,” Options, Futures, andAgricultural CommodityPrograms: SymposiumProceedings. Ed. Bruce Wright.Staff Report No. AGES870911.U.S. Dept. Agr., Econ. Res. Serv.,February 1988, pp. 84-93.

Pope, Rulon D., and RichardPrescott. “Diversification inRelation to Farm Size andOther SocioeconomicCharacteristics,” AmericanJournal of AgriculturalEconomics 62 (August 1980):554-559.

Powers, Nicholas J. MarketingPractices for Vegetables. AIB-702. U.S. Dept. Agr., Econ. Res.Serv., August 1994.


Ref

eren

ces

Querin, Scott F., and William G.Tomek. Hedging StrategiesUtilizing Technical Analysis: AnApplication to Corn in WesternNew York. A.E. Res. 83-12.Ithaca, NY: Cornell University,Department of AgriculturalEconomics, 1983.

Quiggin, J., G. Karagiannis, and J.Stanton. “Crop Insurance andCrop Production: An EmpiricalStudy of Moral Hazard andAdverse Selection,” Economicsof Agricultural Crop Insurance:Theory and Evidence. Ed.Darrell L. Hueth and WilliamH. Furtan. Norwell, MA: KluwerAcademic Publishers, 1994, pp.253-272.

Rawlins, Richard B., and Daniel J.Bernardo. “IncorporatingUncertainty in the Analysis ofOptimal Beef-Forage ProductionSystems,” Southern Journal ofAgricultural Economics 23 (July1991): 213-225.

Ray, Parimalkumar K.Agricultural Insurance: Theoryand Practice and Application toDeveloping Countries. New York:Pergamon Press, second edition,1980.

Rhodes, V. J., and G. Grimes. U.S.Contract Production of Hogs: A1992 Survey. AgriculturalEconomics Report No. 1992-2.Columbia: University ofMissouri, Department ofAgricultural Economics, 1992.

Robison, Lindon J., and Peter J.Barry. The Competitive Firm’sResponse to Risk. New York:MacMillan Publishing Company,1987.

____________, Peter J. Barry, andWilliam G. Burghardt.“Borrowing Behavior UnderFinancial Stress by theProprietary Firm: A TheoreticalAnalysis,” Western Journal of

Agricultural Economics 12(December 1997): 144-151.

____________, Peter J. Barry, JamesB. Kliebenstein, and George F.Patrick. “Risk Attitudes:Concepts and MeasurementApproaches,” Risk Managementin Agriculture. Ed. Peter J. Barry.Ames: Iowa State UniversityPress, 1984, pp. 11-30.

Rolfo, Jacques. “Optimal HedgingUnder Price and QuantityUncertainty: The Case of aCocoa Producer,” Journal ofPolitical Economy 88 (February1980): 100-116.

Roy, A. D. “Safety-First and theHolding of Assets,”Econometrica 20 (July 1952):431-449.

Ryan, James T. “Utilization of DebtRepayment Capacity byCommercial Farm Operators.”Selected paper presented at theAmerican AgriculturalEconomics Association AnnualMeeting, Indianapolis, IN,August 6-9, 1995.

Saha, Atanu, C. Richard Shumway,and Hovav Talpaz. “JointEstimation of Risk PreferenceStructure and Technology UsingExpo-Power Utility,” AmericanJournal of AgriculturalEconomics 76 (May 1994): 173-184.

Sakong, Yong, Dermot J. Hayes,and Arne Hallam. “HedgingProduction Risk with Options,”American Journal of Agricul-tural Economics 75 (May 1993):408-415.

Sander, William. “Off-FarmEmployment and Incomes ofFarmers,” The Oxford AgrarianStudies 12 (1983): 34-47.

Sandmo, Agnar. “On the Theory ofthe Competitive Firm UnderPrice Uncertainty,” American


Referen

ces

9977

Economic Review 61 (March1971): 65-73.

Sands, Laura. “Losing Their GRIP:Canadians Want to Let Go ofRevenue Insurance,” TopProducer (Mid-February 1994):20.

Sapp, S. Iowa Cattle Production: ASurvey of Iowa Cattle Producersin 1990. Ames: Iowa StateUniversity, 1990.

Sarris, Alexander H. “PriceVariability in Cereal Marketsand Risk ManagementStrategies for MarketParticipants.” Discussion paperAGR/CA/APM/CFS(97)8.Organisation for Economic Co-operation and Development,Group on Cereals, AnimalFeeds, and Sugar of the WorkingParty on Agricultural Policiesand Markets, October 1997.

Scherer, F. M. Industrial MarketStructure and EconomicPerformance. Chicago: RandMcNally College PublishingCompany, second edition, 1980.

Schoney, R. A., J. S. Taylor, and K.Hayward. “Risk Reduction fromDiversification and CropInsurance in Saskatchewan,”Economics of Agricultural CropInsurance: Theory and Evidence.Ed. Darrell L. Hueth andWilliam H. Furtan. Norwell,MA: Kluwer AcademicPublishers, 1994, pp. 293-305.

Schroeder, Ted C., Joseph L.Parcell, Terry L. Kastens, andKevin C. Dhuyvetter.“Perceptions of MarketingStrategies: Producers versusExtension Economists,” Journalof Agricultural and ResourceEconomics 23 (July 1998): 279-293.

Shapiro, B. I., and B. WadeBrorsen. “Factors AffectingFarmers’ Hedging Decisions,”

North Central Journal ofAgricultural Economics 10 (July1988): 145-153.

Shee, Jody. “Ocean Spray BusinessSets Goal: $3 Billion,” ThePacker, March 16, 1998, p. 1.

Skees, Jerry R. “The PoliticalEconomy of a Crop InsuranceExperiment.” W. I. MyersLecture Series Paper. Ithaca,NY: Cornell University,Department of Agricultural,Resource, and ManagerialEconomics, October 14, 1994.

___________, J. Roy Black, andBarry J. Barnett. “Designingand Rating an Area Yield CropInsurance Contract,” AmericanJournal of AgriculturalEconomics 79 (May 1997): 430-438.

___________, Joy Harwood, AgapiSomwaru, and Janet Perry. “ThePotential for Revenue InsurancePolicies in the South,” Journalof Agricultural and AppliedEconomics 30 (July 1998): 47-61.

___________, and Sandra Pyles.“Availability and Design ofLivestock Insurance in theUnited States.” Unpublishedworking paper. Lexington, KY:University of Kentucky,Department of AgriculturalEconomics, October 1998.

Sommer, Judith E., David E.Banker, Robert C. Green, JudithZ. Kalbacher, Neal R. Peterson,and Theresa Y. Sun. Structuraland Financial Characteristics ofU.S. Farms, 1994: 19th AnnualFamily Farm Report to theCongress. AIB-735. U.S. Dept.Agr., Econ. Res. Serv., July 1997.

____________, Robert A. Hoppe,Robert C. Green, and PenelopeJ. Korb. Structural andFinancial Characteristics of U.S.Farms, 1995: 20th Annual


Ref

eren

ces

Family Farm Report toCongress. AIB-746. U.S. Dept.Agr., Econ. Res. Serv., November1998.

Sonka, Steven T., and George F.Patrick. “Risk Management andDecision Making in theAgricultural Firm,” RiskManagement in Agriculture. Ed.Peter J. Barry. Ames: Iowa StateUniversity, 1984, pp. 95-115.

Sosland Publishing Company.“Farmer coop. in bid to buy NewMexico mill,” February 26, 1997.http://www.sosland.com/old-news/articles/022697_2.htm

____________. “Work begins on newflour mill; production alreadycontracted,” September 25,1997. http://www.sosland.com/oldnews2/articles/092597p.htm

Sumner, Daniel A. “The Off-FarmLabor Supply of Farmers,”American Journal ofAgricultural Economics 64(August 1982): 499-509.

Takayama, Akira. AnalyticalMethods in Economics. AnnArbor: University of MichiganPress, 1993.

Tauer, Loren W. “Target MOTAD,”American Journal ofAgricultural Economics 65(August 1983): 606-610.

Telser, L. “Safety-First andHedging,” Review of EconomicStudies 23 (1955-56): 1-16.

Tierney, William I. Unpublishedsurveys. Manhattan: KansasState University, various years.

Tomek, William G. Effects ofFutures and Options Trading onFarm Incomes. Staff paper No.87-9. Ithaca, NY: CornellUniversity, Department ofAgricultural Economics, May1987.

____________, and Roger W. Gray.“Temporal Relationships AmongPrices on Commodity FuturesMarkets: Their Allocation andStabilizing Roles,” AmericanJournal of AgriculturalEconomics 52 (August 1970):372-80.

Tronstad, Russell, and C. RobertTaylor. “Dynamically OptimalAfter-Tax Grain Storage, CashGrain Sale, and HedgingStrategies,” American Journal ofAgricultural Economics 73(February 1991): 75-88.

Turvey, Calum G., and Timothy G.Baker. “A Farm-Level FinancialAnalysis of Farmers’ Use ofFutures and Options UnderAlternative Farm Programs,”American Journal ofAgricultural Economics 72(November 1990): 946-957.

____________, and Timothy G.Baker. “Optimal Hedging UnderAlternative Capital Structuresand Risk Aversion,” CanadianJournal of AgriculturalEconomics 37 (March 1989):135-143.

____________, and Kevin Chen.“Canadian Safety Net Programsfor Agriculture.” Paper present-ed at the American AgriculturalEconomics Association meet-ings, San Diego, CA, August 8-10, 1994.

____________, H. C. Driver, andTimothy G. Baker. “Systematicand Unsystematic Risk in FarmPortfolio Selection,” AmericanJournal Agricultural Economics70 (November 1988): 831-36.

U.S. Department of Agriculture,Economic Research Service.Agricultural Outlook. AO-234.October 1998, p. 57.

____________, Economic ResearchService. 1997 Agricultural


Referen

ces

9999

Resource Management Study.Special analyses.

____________, Economic ResearchService. 1996 AgriculturalResource Management Study.Special analyses.

____________, National Agricul-tural Statistics Service.Agricultural Prices. Variousissues.

____________, NationalAgricultural Statistics Service.Costs and Returns ReportInterviewer’s Manual, December1996.

____________, National Agricul-tural Statistics Service.Electronic county yield files,1997.

____________, National Agricul-tural Statistics Service.Electronic county price files,1997.

____________, Risk ManagementAgency. Crop InsuranceActuarial Information PublicAccess Server, FTP downloadpage. http://www.act.fcic.usda.gov/actuarial/ftpdwnld.html

____________, Risk ManagementAgency. Electronic experienceand yield record database (vari-ous record types and aggrega-tion levels), 1996 and 1997.

Van Kooten, Cornelis G., DouglasL. Young, and Jeffrey A.Krautkraemer. “A Safety-FirstApproach to Dynamic CroppingDecisions,” European Review ofAgricultural Economics 24(1997): 47-63.

Vukina, Tomislav, Dong-feng Li,and Duncan M. Holthausen.“Hedging with Crop YieldFutures: A Mean-VarianceAnalysis,” American Journal of

Agricultural Economics 78(November 1996): 1015-1025.

Walker, Odell L., and Glenn A.Helmers. “Applications II: RiskProgramming,” Risk Manage-ment in Agriculture. Ed. Peter J.Barry. Ames: Iowa StateUniversity Press, 1984, pp. 148-165.

Ward, Ronald W., and Lehman B.Fletcher. “From Hedging to PureSpeculation: A Micro Model ofOptimal Futures and CashMarket Positions,” AmericanJournal of AgriculturalEconomics 53 (February 1971):71-78.

Whitson, Robert E., Peter J. Barry,and Ronald D. Lacewell.“Vertical Integration for RiskManagement: An Application toa Cattle Ranch,” SouthernJournal of AgriculturalEconomics 8 (December 1976):45-50.

Wisner, Robert. “Using GrainContracts for RiskManagement,” ManagingChange-Managing Risk: APrimer for Agriculture. Ames:Iowa State University, January1997, pp. 11-14.

____________, E. Neal Blue, and E.Dean Baldwin. “Can Pre-harvestMarketing Strategies IncreaseNet Returns for Corn andSoybean Growers?” ResearchSymposium Proceedings, NCR-134 Conference on AppliedCommodity Price Analysis,Forecasting, and Market RiskManagement, Chicago, April 21-22, 1997, pp. 26-41.

Woolery, B. A., and R. M. Adams.The Trade-off Between Returnand Risk for Selected Big HornBasin Crop and Cattle FeedingSystems. Report No. RJ-149.University of WyomingAgricultural ExperimentStation, February 1980.


Ref

eren

ces

Wright, B. D., and J. A. Hewitt. “AllRisk Crop Insurance: Lessonsfrom Theory and Experience.”Economics of Agricultural CropInsurance: Theory and Evidence.Ed. Darrell L. Hueth andWilliam H. Furtan. Norwell,MA: Kluwer AcademicPublishers, 1994, pp. 73-112.

Wright, Bruce H., Joy L Harwood,Linwood A. Hoffman, andRichard G. Heifner. ForwardContracting in the Corn Beltand Spring Wheat Areas, 1988.Staff Report AGES881102. U.S.Dept. Agr., Econ. Res. Serv.,December 1988.

____________, Linwood Hoffman,Gerald Plato, and Ted Covey.“HTA Contracts: Risks andLessons,” Agricultural Outlook.AO-234. U.S. Dept. Agr., Econ.Res. Serv., October 1996, pp. 31-34.

Young, Renna P., and Peter J.Barry. “Holding Financial Assetsas a Risk Response: A PortfolioAnalysis of Illinois GrainFarms,” North Central Journalof Agricultural Economics 9(January 1987): 77-84.

Zulauf, Carl R., and Scott H. Irwin.“Market Efficiency andMarketing to Enhance Incomeof Crop Producers,” ResearchSymposium Proceedings, NCR-134 Conference on AppliedCommodity Price Analysis,Forecasting, and Market RiskManagement, Chicago, April 21-22, 1997, pp. 1-25.

Zwingli, Michael E., William E.Hardy, Jr., and John L. Adrian,Jr. “Reduced Risk Rotations forFresh Vegetable Crops: AnAnalysis for the Sand Mountainand Tennessee Valley Regions ofAlabama,” Southern Journal ofAgricultural Economics 21(December 1989): 155-165.


Referen

ces

110011

“Actual Production History”(APH) yield—The basis for deter-mining the producer’s guaranteeunder either the multi-peril cropinsurance program or most feder-ally subsidized revenue insurancepolicies. The producer’s APH yieldis calculated as a 4- to 10-yearsimple average of the producer’sactual yield on the insured parcelof land. If a producer does nothave actual yields, the series (upto 4 years) is filled in with a “tran-sition yield,” based on either coun-ty or program yields.

Actuarial soundness—An insur-ance term describing the situationwhere indemnities paid out, onaverage, equal total premiumsand the insurance program“breaks even.”

Adverse selection—A situation inwhich an insured has more infor-mation about his or her risk of lossthan does the insurance provider,and is better able to determine thesoundness of premium rates.

Agricultural ResourceManagement Study (ARMS)—Aprobability-based annual survey offarmers and ranchers in the con-tiguous 48 States conducted by theEconomic Research Service andthe National AgriculturalStatistics Service of USDA. Thesample data can be expanded byusing appropriate weights to rep-resent all farms. The ARMS waspreviously known as the FarmCosts and Returns Survey (FCRS).

Basis—The difference between aspecific futures price and a specificcash price for the same or relatedcommodity.

Basis contract—A forward con-tract that calls for the price to bedetermined by applying a specifieddifference (the basis) to a particu-lar futures contract price to beobserved at a future time, asselected by one of the trading par-ties. Both contracting parties areleft with price level risk until thefinal price is established.

Basis risk—The risk associatedwith an unexpected widening ornarrowing of the basis between thetime a hedging position is estab-lished and the time that it is lifted.

Call option—An option contractthat entitles the holder the right,without obligation, to buy afutures contract at a specifiedprice during a specified time peri-od. The buyer pays a premium tothe seller for this right. A calloption is purchased to protectagainst, or with the expectation of,a potential rise in the futuresprice.

Cash renting—A type of landrental agreement between a land-lord and tenant. Typically, the ten-ant rents the land for a fixedamount per acre that is pre-speci-fied in the agreement.

Certainty equivalent—The cer-tain return that would provide anindividual the same level of satis-faction as a specified uncertainprospective return. The largest cer-tainty equivalent outcome is pre-ferred when comparing alternativerisky choices.


Glo

ssar

y

GGlloossssaarryy3300

30For additional definitions on futuresand options terms, see Heifner, Glauber,Miranda, Plato, and Wright; CommodityFutures Trading Commission.

Coefficient of variation (c.v.)—A measure of variability in a dataset. It is calculated as the stan-dard deviation of the data, dividedby the mean (or average). Thelower the c.v., the smaller the rela-tive variability of that data set.

Commodity Futures TradingCommission—The Federal agencyestablished in 1974 that regulatesfutures and commodity optionstrading under the CommodityExchange Act.

Contract payments—Direct pay-ments to producers of programcrops authorized by the 1996 FarmAct. These payments are in effectfor fiscal years 1996-2002, totaling$35 billion over that time horizon.Contract payments do not varydepending on market prices or pro-duction levels. They are alsoknown as Agricultural MarketTransition Act (AMTA) payments.

Cooperative State Research,Education, and ExtensionService—A U.S. Department ofAgriculture agency that helps fundand administer research and edu-cational programs done jointlywith land grant universities andother institutions.

Deficiency payments—A govern-ment payment initiated under the1973 Farm Act and eliminatedwith the 1996 Farm Act. Deficiencypayments were made to farmerswho participated in wheat, feedgrains, rice, or cotton programs.The payment rate (expressed inper bushel or per pound terms)was based on the differencebetween the price level establishedby law (the target price) and thehigher of the market price duringa period specified by law or theprice per unit at which theGovernment provided loans tofarmers to enable them to holdtheir crops for sale at a later date(the loan rate). The paymentequalled the payment rate, multi-plied by the acreage planted for

harvest, up to the maximum pay-ment acres, and then multiplied bythe program yield established forthe particular farm.

Deferred price contract—(1)Any forward contract where priceis left to be determined later. (2) Aforward contract that transfersownership before price is deter-mined.

Delayed price contract—A for-ward contract that transfers own-ership of a commodity from afarmer to a buyer while providingfor price to be set and payment tobe made at a later date. May alsobe called a “deferred price con-tract” or a “price later” contract.

Delayed payment contract—-Aforward contract that establishesprice and transfers ownership of acommodity from the farmer to thebuyer while providing for paymentto be made at a later date. Used byfarmers for shifting incomebetween years for tax purposes.May also be called a “deferred pay-ment contract.”

Efficient market—A market inwhich price fully reflects all rele-vant information.

Empirical probability distribu-tion—A table or figure describingthe likelihood of events that isbased entirely on observed relativefrequencies without making anyassumptions about the shape ofthe distribution.

Exercise (or strike) price—Theprice specified in an option con-tract at which the option holdercan buy (call) or sell (put) afutures contract.

Farm Service Agency—The U.S.Department of Agriculture agencythat administers various farm pro-grams, including contract pay-ments and farm ownership andoperating loans.


Glo

ssary

110033

Flat (or fixed) price contract—A forward contract that establish-es the specific price to be paid bythe buyer to the seller.

Forward contract—An agree-ment between two parties callingfor delivery of, and payment for, aspecified quality and quantity of acommodity at a specified futuredate. The price may be agreedupon in advance, or determined byformula at the time of delivery orother point in time.

Forward pricing—Agreeing onprice for later delivery. “Forwardpricing” is used broadly in thisreport to refer both to hedging infutures or options, or forward con-tracting.

Futures contract—An agreementpriced and entered on an exchangeto trade at a specified future timea commodity, or other asset, withspecified attributes (or in the caseof cash settlement, an equivalentamount of money).

Hedge-to-arrive contract—Anagreement between a farmer andbuyer that calls for the farmer todeliver and the buyer to pay for acommodity on a future date at thecurrent futures price plus a differ-ential (basis) to be determined atdelivery time. Some hedge-to-arrive contracts allow for rollingover to later maturing futures con-tracts. From a farmer’s standpoint,a hedge-to-arrive contract is simi-lar to a short hedge in that thefutures price is fixed but the basisis left to be determined later.

Hedging—Taking a position in afutures or options market whichtends to reduce risk of financialloss from an adverse price change.

Idiosyncratic risk—Risk that isspecific to an operation and that isnot common to all producers in thearea. A broken water pipe, forexample, reflects idiosyncraticrisk.

In-the-money option—An optioncontract that would yield a posi-tive return to the holder if exer-cised. An option is in-the-money ifthe strike price exceeds the mar-ket price for a put, or is less thanthe market price for a call. Themagnitude of this difference is theintrinsic value of the option.

Indemnity—The compensationreceived by an individual for quali-fying losses paid under an insur-ance program. The indemnity com-pensates for losses up to the levelof the insurance guarantee.

Intrinsic value—The value of anoption if immediately exercised.The amount by which the currentprice for the underlying commodityor futures contract is above thestrike price of a call option, orbelow the strike price of a putoption.

Leverage—Use of borrowed fundsto finance a business, such asfarming, or an investment.

Liquidity—The extent to whichassets can be quickly converted tocash without accepting a discountin their value. An asset is perfectlyliquid if its sale generates cashequal to, or greater than, thereduction in the value of the firmdue to the sale. Illiquid assets, incontrast, cannot be quickly soldwithout a producer accepting adiscount, reducing the value accru-ing to the firm by more than theexpected sale price.

Long hedging—Purchasing afutures contract or a call option tooffset the risk of a price increasein the cash market.

Margin—The money or collateralguaranteeing the customer’sfutures or options trades, deposit-ed by a customer with his or herbroker for the purpose of insuringthe broker against a loss on anopen futures contract. The initialmargin is the amount required to


Glo

ssar

y

enter a futures position. The main-tenance margin is the amountrequired to continue a futuresposition without receiving a mar-gin call.

Margin call—A request from abroker to a customer for additionalmargin to cover the customer’sfutures position after a pricechange unfavorable to the cus-tomer. The broker may close outthe customer’s position if the mar-gin call is not met.

Marketing contract—A verbal orwritten agreement between aprocessor or handler and a growerestablishing an outlet and a price,or a formula for determining theprice, for a commodity before har-vest or before the commodity isready to be marketed.

Minimum price contract—Acontract providing the farmer withprotection against a decline inprice below a minimum level,while leaving the final pricinguntil a later date.

Moral hazard—The ability of aninsured to increase his or herexpected indemnity by actionstaken after buying the insurance.

Natural hedge—A tendency foryield and price deviations fromexpectations to offset each other intheir effects on crop revenue. Thistendency is strongest in majorgrowing areas.

Normal distribution—A sym-metric, bell-shaped mathematicaldistribution that is widely used instatistical analysis because itclosely approximates manyobserved distributions. Normal dis-tributions are fully described bytheir means and variances.

Off-exchange option—An agree-ment between two parties enteredwithout the services of an organ-ized exchange that gives one partythe right to buy or sell an asset at

a specified price over a specifiedtime interval.

Optimal hedge—The size of thefutures or options position thatminimizes a hedger’s risk. Oftenexpressed as a ratio of the futuresor options position to the cashposition. Yield risk and/or basisrisk generally cause the optimalhedge ratio to be less than 1.0.

Option contract—A contract thatgives the holder the right, withoutobligation, to buy or sell a futurescontract at a specific price within aspecified period of time, regardlessof the market price of the futures.

Out-of-the-money option—Anoption contract that cannot beprofitably exercised at the currentmarket price. An option is out-of-the-money if the market priceexceeds the strike price for a putor is less than the strike price fora call.

Premium—An amount of moneypaid to secure risk protection.Option buyers pay a premium tooption sellers for an options con-tract. Similarly, the purchaser ofan insurance policy pays a premi-um in order to obtain coverage.

Price-yield correlation—A sta-tistical measure of the closeness ofthe relationship between pricesand yields.

Production contract—A verbalor written agreement between aprocessor (integrator) and a grow-er that usually specifies in detailthe production inputs supplied bythe processor, the quality andquantity of a particular commoditythat is to be delivered, and thecompensation that is to be paid tothe grower. In return for relin-quishing complete control overdecisionmaking, the producer isoften compensated with a pricepremium or lower market risk.


Glo

ssary

110055

Put option—An option contractthat gives the holder the right,without obligation, to sell a futurescontract at a specific price (the“strike price”) within a specifiedperiod of time, regardless of themarket price of the futures. A putoption normally is purchased toprotect against a cash price decline(hedger) or with the expectation ofa futures price decline (speculator).

Reinsurance—A method ofspreading insurance companies’risk over time and space. Forapproved agricultural insuranceprograms (Federal crop insuranceand approved revenue insuranceproducts), the Risk ManagementAgency shares the risk of loss witheach private insurance companydelivering policies to producers.Private reinsurance is also avail-able, where one insurance compa-ny transfers part of the risk toanother company by agreements,which vary as to the terms appli-cable to the risk transferred.

Revenue insurance—An insur-ance program offered to farmersthat pays indemnities based onrevenue shortfalls. As of 1998,three revenue insurance programswere offered to producers in select-ed locations. These three programs(Crop Revenue Coverage, IncomeProtection, and RevenueAssurance) are subsidized andreinsured by the RiskManagement Agency.

Risk—Uncertainty in outcomesthat are not equally desirable tothe decisionmaker, and that mayinvolve, among other outcomes,the probability of making (or los-ing) money, harm to humanhealth, repercussions that affectresources (such as credit), or othertypes of events that affect a per-son’s welfare. Risk is uncertaintythat “matters.”

Risk aversion coefficient—Ameasurement of an individual’spreference for a certain outcome

over an uncertain outcome withequal expected value.

Risk Management Agency—The U.S. Department of Agricul-ture agency that provides over-sight, subsidization, and reinsur-ance for approved risk manage-ment programs, such as theFederal multi-peril crop insuranceprogram and various revenueinsurance programs.

Share renting—Renting landunder a contract that calls for thecrop to be divided between the ten-ant and the landlord in fixed per-centages. May also provide forsharing certain inputs.

Short hedging—Selling a futurescontract to offset the risk of a pricedecline in the cash market.

Standard deviation—One of themost widely used measures of dis-persion, calculated as the squareroot of the variance.

Time value—The portion of anoption’s premium that exceeds theintrinsic value. It reflects the prob-ability that the option will movein-the-money, or deeper in-the-money. The longer the timeremaining until expiration of theoption, the greater its time value.

Unbiasedness—Characterizes anestimate, forecast, or forward(including futures) price that isneither systematically high or low,but correct on average.

Uncertainty—Lack of sure knowl-edge or predictability because ofrandomness.

Utility function—A mathemati-cal expression that can be used torepresent a decisionmaker’s riskpreferences.

Variance—One of the most widelyused measures of dispersion, calcu-lated as the average squared devia-tion from the mean or expectation.


Glo

ssar

y

Risk must be quantified in orderto evaluate whether various

risk management tools and strate-gies are effective in achieving pro-ducers’ risk reduction goals. Thisprocess involves measuring uncer-tainty and quantifying the rela-tionship between uncertainty andan individual’s well being. Thissection discusses how risks can bequantified and provides represen-tative estimates for selected loca-tions—focusing on price variabili-ty, yield variability, and the corre-lation between prices and yields(the extent to which prices andyields move together).

MMeeaassuurriinngg UUnncceerrttaaiinnttyy

The measurement of uncertaintyinvolves estimating the probabili-ties of future outcomes. Estimatesmay be made, for example, of theprobability of yield less than 100bushels per acre, the probability ofprice falling below $2.25 perbushel, or the probability of rev-enue less than $200 per acre. Moregenerally, one would like to esti-mate the joint probability distribu-tion of yield, price, and revenue sothat one might, for example, speci-fy the probability of revenuefalling below any specified level. Toestimate such probabilities, wegenerally start by observing his-

torical outcomes and separatingrandom variability from systemat-ic variability.

To illustrate, appendix figure 1(and appendix table 1) show cornyields for Iroquois County, Illinois(in the east central part of theState) for the years 1956-95. Thejagged line links the actual yields,averaged across the county, foreach year, while the straight linerepresents the systematic upwardtrend in yields. This upward trendmay be considered to be a “known”source of variation that will repeatitself in the future. It has beencaused by several factors, includ-ing the development of higheryielding varieties and the intro-duction of improved chemicals andfertilizers. In contrast, the yielddeviations from trend—mainlycaused by weather—constitute therandom variability.31

Quantifying yield randomness gen-erally involves summarizing what


Ap

pen

dix

1:Techn

iqu

es forM

easurin

g Risk

110077

AAppppeennddiixx 11:: TTeecchhnniiqquueess ffoorr MMeeaassuurriinngg RRiisskk

Understanding how risk is measured is a starting point forhelping farmers make choices about the most appropriatestrategies for their individual situations. This appendixprovides information on the different approaches that canbe used to quantify risk, and illustrates how probability dis-tributions can be used to characterize the outcomes associ-ated with risky choices. In order to make the best decisionsfor their individual operations, farmers and other decision-makers often use historical and current information aboutprices, yields, weather conditions, and other variables toestimate future risk.

31Whether the yields shown are ade-quately represented by a linear trend canbe questioned. Linear yield trends often areused for forecasting, but there is no strongreason why yield trends should be linear, orfollow any other specific mathematicalform. Trend projections inevitably involve adegree of subjectivity, not only in choosingthe mathematical function to use, but alsoin selecting the years to be included in cal-culating the trend.

is known about deviations fromexpected yields, as measured bytrend. Randomness can bedescribed by converting such devi-ations into a frequency distribu-tion, or histogram, as depicted inappendix figure 2.32 Each bar onthe figure shows the number oftimes that yield deviations fromtrend in Iroquois County fell with-in a particular 10-bushel-per-acrerange. For example, the barlabeled “-5 to +5” illustrates thatyields fell between -5 bushels and+5 bushels from trend in 7 of the40 years between 1956 and 1995,and the bar labeled “5 to 15” illus-trates that yields fell between 5bushels and 15 bushels abovetrend in 9 of the years. Frequen-cies are greatest near the middleand the least at the lower andupper ends, which is typical ofyields, prices, and revenues. This isbecause extreme weather events—such as the 1988 drought—are lesslikely than weather events havinga more modest effect.

The degree of randomness isreflected in the width of the distri-bution and in the number of obser-vations that are distant from themean. Note that appendix figure 2is not symmetrical (like the tradi-tional bell curve), but that thelower tail is longer than the uppertail. This so-called negative skew-ness is typical of yield distributions.This shape occurs because devas-tating weather can cause very sig-nificant yield declines (as low aszero), while very good weather islikely to only moderately boostyields above trend due to the physi-ological limitations of the plant.

For many purposes, a single num-ber is a more convenient measureof randomness (or dispersion) thanis an entire distribution. The mostwidely used measures of random-ness are the variance and itssquare root, the standard deviation.Variance is the average squareddeviation from the mean, or trend.By using the variance of deviationsfrom trend, a large part of the sys-tematic variation is removed.

One problem with the varianceand standard deviation is that


Ap

pen

dix

1:Te

chn

iqu

es f

orM

easu

rin

g R

isk

60 64 68 72 76 80 84 88 9240

60

80

100

120

140

160

180

1956

Trend-adjusted yield

Actual yield

Bushels per acre

Note: Actual yield is county average.

Appendix figure 1

Actual and trend-adjusted corn yields, Iroquois County, Illinois, 1956-95

Source: Constructed by ERS from USDA, NASS electronic county yield files, 1997.

32To construct appendix figure 2, ran-domness was assumed to have remainedunchanged, although appendix figure 1suggests that it may be increasing overtime.

they are difficult to interpret with-out knowing the level or magni-tude of the underlying variable. Avariance of 10 bushels, for exam-ple, has quite different implica-tions for the tightness of the distri-bution when the mean yield(adjusted for trend) is 50 bushelsper acre than when it is 160bushels. As a result, proportionalvariability—or variability relativeto the mean—is often measured tofacilitate comparisons. The mostcommonly used measure of rela-tive variability is the coefficient of

variation, which equals the stan-dard deviation divided by themean.

The variance (or alternatively, thestandard deviation or coefficient ofvariation) is a good measure ofvariability for approximately sym-metric, bell-shaped distributions.It fully describes the variability ina normal distribution, which is aparticular bell-shaped mathemati-cal distribution that closelyapproximates many observed dis-tributions. Most yield distribu-


Ap

pen

dix

1:Techn

iqu

es forM

easurin

g Risk

110099

Appendix table 1—Calculation of yield variability for Iroquois County, IllinoisYear Actual yield Trend-adjusted Deviation

yield (actual minus trend)

Bushels per acre

1956 73.4 75.7 -2.21957 65.8 77.0 -11.11958 66.1 78.3 -12.21959 60.2 79.6 -19.4

1960 71.8 80.9 -9.11961 76.6 82.3 -5.71962 90.4 83.6 6.91963 92.0 84.9 7.11964 85.3 86.2 -0.9

1965 100.6 87.5 13.01966 87.8 88.9 -1.01967 100.4 90.2 10.21968 87.6 91.5 -3.91969 108.0 92.8 15.2

1970 85.8 94.1 -8.31971 118.6 95.5 23.11972 113.8 96.8 17.01973 99.1 98.1 1.01974 83.5 99.4 -15.9

1975 115.0 100.7 14.21976 113.3 102.1 11.21977 90.7 103.4 -12.71978 110.1 104.7 5.41979 123.6 106.0 17.5

1980 74.6 107.3 -32.81981 118.9 108.7 10.21982 136.7 110.0 26.71983 85.3 111.3 -26.01984 111.4 112.6 -1.3

1985 143.2 113.9 29.31986 117.9 115.3 2.61987 134.7 116.6 18.11988 59.6 117.9 -58.31989 137.1 119.2 17.8

1990 127.6 120.5 7.01991 66.9 121.9 -54.91992 148.6 123.2 25.51993 113.7 124.5 -10.81994 161.2 125.8 35.41995 99.0 127.1 -28.1

Note: The equation estimated from these data for detrending yields is: E(Yt) = 1.72 + 1.32(T), where T is theyear, minus 1900.Source: Calculations made by ERS from USDA, NASS, electronic county yield files, 1997.

tions, however, appear to be non-normal with long lower tails asshown in appendix figure 2.Moreover, some tools used to man-age farmers’ risks, particularlycrop insurance and commodityoptions, impose non-normality bysetting bounds or limits on thelower tails of the yield or price dis-tributions realized by the farmer.Producers generally prefer yieldand price distributions that arebounded from below because itlimits their losses. However, thestandard deviation may not pro-vide a satisfactory measure of riskunder such distributions, whichclearly are non-normal.

Other measures of variability ordispersion may be useful for distri-butions that are clearly non-nor-mal. One such measure is theprobability of outcomes belowsome critical level. The probabilityof yield less than 70 percent of itsexpectation, for example, might bea useful measure of risk for somefarmers. If the trend yield is 127bushels per acre, the 70 percentpoint would equal 0.70 * 127, or 89bushels per acre. This is 38bushels below trend. In appendixfigure 2, the probability of such a

yield (or lower) is two occurrencesin 40 years, or a probability of 2/40= .05. Individual farmers mightchoose higher or lower cutoffpoints, depending on their differ-ing financial circumstances anddegrees of risk aversion.

EEssttiimmaattiinngg PPrroobbaabbiilliittiieess ooffFFuuttuurree EEvveennttss

Farmers, like other decisionmakers,are fundamentally concerned withrandomness in future events, notthe distribution of past outcomes asillustrated in the previous section.They are concerned about the prob-abilities of outcomes to be observedin the future and the effects ofthese outcomes on their economicwelfare. The probability associatedwith any given outcome indicatesthe strength of one’s belief thatsuch an outcome will occur, rangingfrom zero (which represents no pos-sibility) to 1 (representing absolutecertainty).

Two sources of information aboutsuch probabilities are available:logic and experience. In puregames of chance, logic rules. Forexample, in flipping a coin, twoequally likely outcomes are possi-


Ap

pen

dix

1:Te

chn

iqu

es f

orM

easu

rin

g R

isk

-65

to -

55

-55

to -

45

-45

to -

35

-35

to -

25

-25

to -

15

-15

to -

5

-5 to

+5

5 to

15

15 to

25

25 to

35

35 to

45

0

2

4

6

8

10

10-bushel ranges below and above trend

Appendix figure 2

Frequency distribution of corn yield deviations from trend, Iroquois County, Illinois, 1956-95


Frequency (number of occurrences)

ble—heads or tails—and thus aprobability of 0.5 can be assignedto each. In business decisions,however, historical observationsoften must be relied upon to esti-mate probabilities. Each of the fre-quencies illustrated in appendixfigure 2, for instance, could bedivided by the total number ofyears, 40, to obtain estimates ofthe probabilities of yields withineach of the intervals. The resultingdistribution is often referred to asan “empirical” probability distribu-tion because it is estimatedthrough the use of a specific set ofhistorical observations. Supposethat the projected mean yield is130 bushels per acre. Referring toappendix figure 2, the estimatedprobability of the yield fallingbetween 115 and 145 bushels (thatis, between -15 and +15 bushelsfrom the trend expectation) is23/40 = 0.575.

An alternative way to describe dis-persion graphically is to plot prob-abilities of outcomes falling at orbelow specific values. This is calleda cumulative distribution.Appendix figure 3 is a cumulativedistribution of the Iroquois Countyyield deviations. Cumulative dis-

tributions are particularly usefulfor representing continuous vari-ables because probabilities can beread directly from the vertical axisinstead of by summing areasunder a curve. Cumulative distri-butions are useful in safety-firstanalysis and stochastic dominanceanalysis, which are discussed inthe next section of this report.

Relative frequencies derived fromhistorical observations are not nec-essarily the best estimates offuture probabilities. Sometimes,the decisionmaker has additionalinformation—such as regardingrecent rainfall or temperature con-ditions—which needs to be takeninto account. Moreover, most his-torical series include events thathave small probability of recur-ring, or fail to catch events, thatthough uncommon, have a non-zero likelihood. To reduce theimpacts of such sampling errors,forecasters often impose smooth-ness and a degree of symmetry byfitting mathematical distributionsto historical observations.

The normal distribution, which issymmetrical and bell-shaped, isfrequently used as an approxima-


Ap

pen

dix

1:Techn

iqu

es forM

easurin

g Risk

111111

-65

to -

55

-55

to -

45

-45

to -

35

-35

to -

25

-25

to -

15

-15

to -

5

-5 to

+5

5 to

15

15 to

25

25 to

35

35 to

45

0

0.2

0.4

0.6

0.8

1

10-bushel ranges below and above trend

Cumulative probability

Appendix figure 3

Cumulative probability of corn yield deviations from trend, Iroquois County, Illinois, 1956-95


tion. Although yield distributionsare typically negatively skewed, asdiscussed earlier, the normal dis-tribution is computationally con-venient because it is fullydescribed by its mean and vari-ance. In addition, yield deviationsfrom normality may not be great.The mean and variance in appen-dix table 1, for example, can beused as parameters of a normaldistribution of yield deviationsfrom trend. Appendix figure 4

illustrates realized corn yields forIroquois County over the 1956-95period and a projected probabilitydensity function for the 1997 yield.The projected distribution reflectsthe belief that the true probabilityfunction is continuous and recog-nizes that observed historicalobservations between 1956 and1995 are only a sample of the pos-sible outcomes.


Ap

pen

dix

1:Te

chn

iqu

es f

orM

easu

rin

g R

isk

60 64 68 72 76 80 84 88 92 9640

60

80

100

120

140

160

180

1956

Bushels per acre


Projected yield

Projected yield distribution

Appendix figure 4

Projected 1997 corn yield distribution for Iroquois County, Illinois, based on 1956-95 observations

The following sections examineselected approaches that are

commonly used by economists inanalyzing decisionmaking underrisk. The approaches discussedhere range from one of the sim-plest (the “safety-first” approach)to one of the more complex (theuse of “expected utility”).“Certainty equivalence” also is dis-cussed, which involves measuringrisk in terms of differences inexpected income. The use of thesedifferent approaches allowsresearchers to rank alternativestrategies, and to help producersmake optimal choices in differentsituations.

TThhee ““SSaaffeettyy--FFiirrsstt”” AApppprrooaacchh

The “safety-first” approach to riskmanagement applies if a decision-maker first satisfies a preferencefor safety (such as minimizing theprobability of bankruptcy) whenmaking choices as to the firm’sactivities. Only when the safetyfirst goal is met at a thresholdlevel can other goals (such as max-imizing expected returns) beaddressed. Thus, attaining thehighest-priority goal serves as aconstraint on goals that have suc-cessively lower priorities (Robison,Barry, Kliebenstein, and Patrick).

Safety-first methods are particu-larly applicable where survival ofan individual or business is theparamount concern. However, inmost business risk managementsituations, the use of safety-firstmethods is somewhat arbitrarybecause no single goal is clearlydominant.

The safety-first criteria can bespecified in various ways in empir-ical applications. One of the firstuses of this approach was devel-oped in 1952, and involves choos-ing the set of activities with thesmallest probability of yielding anexpected return (Y) below a speci-fied disaster level of return (Y-min)(Roy). To aid in understanding thevarious safety-first criteria, appen-dix table 2 shows the expectedincome and the probability ofincome less than the disaster level,which is assumed to be $50,000,for three hypothetical strategies,A, B, and C. For strategy B, forexample, the expected return is$500,000, and the probability thatreturns under this strategy willfall below $50,000 is 4 percent.Strategy A has the highest expect-ed return and the highest riskamong the strategies illustrated,while strategy C has the lowestexpected return and lowest risk.


Ap

pen

dix

2:An

alytical Tools for

Assessin

g the Effectiven

ess of

Risk

Man

agemen

tStrategies

111133

AAppppeennddiixx 22:: AAnnaallyyttiiccaall TToooollss ffoorr AAsssseessssiinnggtthhee EEffffeeccttiivveenneessss ooff RRiisskk MMaannaaggeemmeennttSSttrraatteeggiieess

Different modeling approaches are used by economists tocapture decisionmaking in risky situations. Theseapproaches are based on the idea that each risky strategyoffers farmers a different probability distribution of income,and that determining the best strategy involves describingthe different distributions and developing rules to chooseamong them. These approaches differ, however, in the waysin which they incorporate risk attitudes, and in the degreeof flexibility allowed in specifying risk-return trade-offs.

Under Roy’s safety-first criteria,the optimal activity choice occurswhere the probability of expectedreturn falling below the $50,000threshold is minimized. StrategyC, which has the lowest probabilityof disaster, best meets this criteria.When returns are normally dis-tributed, the solution occurs wherethe disaster level (Y-min) is thegreatest number of standard devi-ations away from the expectedincome. Roy’s criteria can beexpressed mathematically as:

Minimize Prob (Y < Y-min)

A second type of approach, intro-duced by Telser in 1955, assumesthat the decisionmaker maximizesexpected returns, E(Y), subject tothe constraint that the probabilityof a return less than or equal to aspecified minimum disaster level(Y-min) does not exceed a givenprobability (P). Mathematically,Telser’s approach is expressed as:

Maximize E(Y)subject to: Prob (Y < Y-min) ≤ P

To apply the Telser criterion to theexample in appendix table 2, sup-pose that the critical probability is4 percent. Then, the Telser criteri-on would choose strategy B, whichmaximizes expected income amongthose strategies for which P is notgreater than 4 percent. Alter-natively, if the critical probabilitywere 3 percent, then strategy Cwould be selected, while if it were5 percent, strategy A would beselected. This example illustratesthat safety-first results can bequite sensitive to initial assump-

tions about what constitutes a crit-ical loss.

The topics that have beenaddressed by these various typesof safety-first criteria vary widely.They include: optimal hedging(Telser), dynamic cropping deci-sions in southeastern Washington(Van Kooten, Young, andKrautkraemer), farm extensionprograms (Musser, Ohannesian,and Benson), and attitudes towardrisk regarding fertilizer applica-tions among peasants in Mexico(Moscardi and de Janvry).

The safety-first approach has bothadvantages and disadvantages. Itdoes not require the specificationof a farmer’s risk aversion coeffi-cient (see accompanying box onrisk aversion, p. 119), and it is notlimited to specific distributionalassumptions, other than that utili-ty increases with returns (subjectto varying constraints dependingon the specification) (appendixtable 3). As a result, it is straight-forward to use. On the downside,however, it is limited in its abilityto address producers’ varying lev-els of aversion to risk (althoughthe threshold probability can serveas a proxy for risk aversion), anddifficulties can also arise in choos-ing the critical cutoff level for dis-aster returns. In addition, any out-come below the cutoff is treated asequivalent to any other. In reality,observations far below the cutoffdisaster level are more adverse tothe farmer than those that arenearer the cutoff point.


Ap

pen

dix

2:A

nal

ytic

al T

ool

s fo

rA

sses

sin

g th

e Ef

fect

iven

ess

of

Ris

kM

anag

emen

tSt

rate

gies

Appendix table 2—Comparison of Roy’s safety-first approach with Telser’s safety-first approach

Strategy Expected Minimum disaster Probability of fallingincome return below minimum

E(Y) (Y-min) disaster return (P)

Dollars Percent

A 1,000,000 50,000 5B 500,000 50,000 4C 200,000 50,000 3


TThhee ““EE--VV”” AApppprrooaacchh aanndd QQuuaaddrraattiicc PPrrooggrraammmmiinngg

A classic problem in risk analysisinvolves determining an optimalallocation of resources across anarray of risky alternatives. Theproblem was first solved byMarkowitz in the context of select-ing optimal stock portfolios. Hissolution was to find the set of allo-cations that maximize expectedtotal return for different levels ofvariance of total return. This iscalled the “expected value-variance(or E-V) efficient” set or frontier.The heavy line in appendix figure

5 represents an E-V efficient fron-tier. On this frontier, expectedreturn can be increased only byaccepting a larger variance ofreturn. The optimal portfolio ispresumed to come from this fron-tier and depends on the decision-maker’s preference tradeoffsbetween expected return and vari-ance of return.33


Ap

pen

dix

2:An

alytical Tools for

Assessin

g the Effectiven

ess of

Risk

Man

agemen

tStrategies

111155

Appendix table 3—Methods for ranking probability distributions, and assumptions and parameters required for each

Method Assumptions required Parameters requiredabout utility to use method

Roy’s safety-first Increases with prob (Y > L) Prob (Y < L)

Telser’s safety-first Increases linearly with Y ifprob (Y < L) < P, is zero otherwise E(Y) and prob (Y < L)

E-V efficiency Increases with Y at a decreasingrate plus normality or quadratic utility E(Y) and Var(Y)

1st degree stochastic dominance Increases with Y Complete distribution of Y

2nd degree Increases with Y at a stochastic dominance decreasing rate Complete distribution of Y

Expected utility and Complete distribution of Ycertainty equivalents Known function of Y and utility functionNote: Y = income; E(Y) = expected income; L = critical level of income; P = critical probability.

Source: Compiled by ERS.

33The E-V efficient set will include theportfolio that maximizes expected utility ifthe decisionmaker's utility function isquadratic or returns on all activities arenormally distributed. See subsequent sec-tion on "The Expected Utility Approach."


Indifference curves

Optimum

E-V efficient frontier

Expected income

Variance of income

Appendix figure 5

Example E-V efficient frontier and indifference curves

The E-V efficiency criterion can beused in allocating a farm’sresources among alternative riskyenterprises. A risk-averse farmerdesires high expected return andlow variance of return, whichinvolves moving upward and/or tothe left in the figure. The optimalcombination of activities for thefarmer occurs at the point on theE-V frontier that provides the pre-ferred combination of expectedreturn and variance of return. Toillustrate the farmer’s preferences,three indifference curves areshown as dashed lines. Each con-nects combinations of risk andexpected return that are equallydesirable to the producer. The opti-mal point on the E-V frontier isthe point that touches the highestattainable indifference curve.

This approach has on many occa-sions been applied to farming deci-sions, particularly to decisionsabout enterprise choice and diver-sification. E-V efficient combina-tions of crop and livestock enter-prises can be identified and thecombination that offers the pre-ferred mix of expected return andvariability of returns can be cho-sen. Determining E-V efficientcombinations requires estimates ofthe variances in returns and thecorrelations of returns for thoseenterprises under consideration, aswell as estimates of expectedreturns for those enterprises.

The attractiveness of E-V analysisis that it leads to relatively conven-ient solutions using quadratic pro-gramming. The exact formulationof the problem can vary. Oneapproach is to maximize a quadrat-ic function of activity levels subjectto linear constraints as follows:

subject to:

where:xi = the level of the ithactivity;

ui = the expected returnper unit of the ith activity;

λ= the risk-return trade-off;

σij = the covariance ofreturn on activities i and j;

aji = coefficients in m lin-ear constraints on theactivity levels.

bj = levels of the linearconstraints.

To trace out the E-V frontier, thequadratic programming problemmust be solved parametrically asthe risk aversion coefficient, λ,varies from 0 to ∞. This is a rathercomplicated problem, but computeralgorithms are available. If thefarmer is risk neutral (λ=0), theproblem collapses to a an incomemaximization problem, which canbe solved with ordinary linear pro-gramming. As risk becomesincreasingly important, the riskaversion coefficient increases andthe E-V portfolio becomes increas-ingly diversified (Anderson, Dillon,and Hardaker). Other factors, suchas limitations imposed by resourceconstraints, may also lead to adiversified portfolio.

The quadratic programmingapproach has been applied to farmenterprise selection by manyresearchers. The first application,for example, involved the evalua-tion of four production activitiesand several resource constraintson a representative farm in east-ern North Carolina (Freund).Quadratic risk programming hassince been applied to many otherevaluations of optimal farm enter-


Ap

pen

dix

2:A

nal

ytic

al T

ool

s fo

rA

sses

sin

g th

e Ef

fect

iven

ess

of

Ris

kM

anag

emen

tSt

rate

gies

a x b j m

x i n

ji i ji

n

i

≤ =

≥ ==∑ 1 2

0 1 21

, ,

, , ,

K

K

Maximize x u x xii

n

i i jj

n

i

n

= ==∑ ∑∑−

1 11

λ σ

prise choice, including studies byBarry and Willman, and Musserand Stamoulis.34

The use of the E-V approach hasboth advantages and disadvan-tages. As in the case of “safety-first” analysis, E-V analyses mayor may not include an explicitmeasure of the producer’s riskaversion (as illustrated in theexample above). E-V analysis islimited, however, in that itassumes that the producer has anoutcome distribution that is nor-mal35 or, alternatively, a utilityfunction (which expresses riskpreferences) that is quadratic. Inaddition, the farmer is assumed toalways prefer more (rather thanless) of the variable in question(such as income), and is assumeduniversally not risk preferringwith respect to that variable(Hardaker, Huirne, and Anderson).As with various approaches to riskanalysis, estimation of the vari-ance-covariance matrix can pres-ent methodological pitfalls (Mappand Helmers).

TThhee SSttoocchhaassttiicc DDoommiinnaanncceeAApppprrooaacchh

Unlike E-V analysis, which isbased on the mean and variance of

a distribution, stochastic domi-nance involves comparing pointson two or more entire distribu-tions. That is, when stochasticdominance is used, alternativesare compared in terms of the fulldistributions of outcomes. Becausecomparisons must be made at eachspecified point along each distribu-tion in a pairwise fashion, the con-ceptual complexity and computa-tional task associated with thisapproach are greater than when E-V analysis is used (Hardaker,Huirne, and Anderson).

The first concept of stochastic effi-ciency was formalized in the early1960’s, and is known as “first-degree” stochastic dominance. Thisapproach rests on the notion thatdecisionmakers prefer more of agiven variable (such as income) toless. Using an example, supposethere are three plans, A, B, and C,each having a probability distribu-tion of income outcomes, “x.” Thecumulative density functions(CDFs) associated with the plansare FA (x), FB (x), and FC (x),respectively, as shown in appendixfigure 6. The CDFs reflect the“accumulated” area under theprobability density function (PDF)at each level of income for eachplan. At the extreme right side ofthe chart, the entire PDF issummed, and the probability ofrealizing an income for any of theplans that is equal to or less thanthe amount designated on the axisis 1.00.

For one plan to dominate anotherin the first-degree sense, the CDFfor the first plan must nowhere behigher than the CDF for the sec-ond plan, and it must lie below theCDF for the second plan at somepoint. Mathematically, first degreestochastic dominance (FSD) can beexpressed, for two representativeplans A and B, as:

FA (x) ≤ FB (x), for all levels of x


Ap

pen

dix

2:An

alytical Tools for

Assessin

g the Effectiven

ess of

Risk

Man

agemen

tStrategies

111177

34Other risk programming methods areavailable. MOTAD programming, for exam-ple, minimizes the mean absolute deviationin net income, which simplifies the problemto one of linear programming (Hazell).Target MOTAD, developed in 1983, mini-mizes deviations from a target level ofincome (Tauer). Discrete stochastic pro-gramming can also be used to determineefficient enterprise choices for farmers. Fora discussion of the advantages and disad-vantages of these approaches, as well asexamples, see Hardaker, Huirne, andAnderson; Musser, Mapp, and Barry; Mapp,Hardin, Walker, and Persaud; and Walkerand Helmers.

35The normality assumption may bereasonable, particularly if the number ofrisky prospects is not too small and therisky prospects are diverse (Anderson,Dillon, and Hardaker). In addition, severalstudies have concluded that the E-Vapproach is quite robust to violations of thenormality assumption (Levy andMarkowitz; Kroll, Levy, and Markowitz).

In appendix figure 6, FC (x) domi-nates FA (x) in the first-degreesense, meaning that the probabili-ty of exceeding any given level ofincome is greater under plan Cthan plan A, and that plan A can-not be a member of the first-degree stochastic dominant (FSD)set. Because the CDF for plan Bcrosses both plans A and C, plan Bdoes not dominate, and is not dom-inated by, either A or C in the firstdegree sense (Hardaker, Huirne,and Anderson).

Second-degree stochastic domi-nance (SSD) is applicable if thedecisionmaker is risk averse andprefers higher incomes to lowerincomes. In contrast to first-degreestochastic dominance, SSDinvolves the comparison of areasunder the CDFs for various plans,and, in general, has more discrimi-natory power than does FSD (Kingand Robison). For a representativeplan, A, to be SSD over anotherplan, B, requires that:

In appendix figure 6, for example,note that the accumulated areaunder FB(x) is less than underFA(x) at all levels of income. Thus,B exhibits SSD over A. C alsoexhibits SSD over A. However, Cdoes not exhibit SSD over Bbecause there is a range in thelower tail where the accumulatedarea under C exceeds that underB. This example illustrates thestronger discriminatory power ofSSD compared to FSD while show-ing that SSD does not discriminateamong all distributions.

As with other approaches to riskanalysis, the use of stochastic dom-inance methods has both pros andcons. Although it provides a rigor-ous assessment, the number ofefficient sets may remain undulylarge. SSD is more discriminatingthan FSD, but nearly one-half ofrandomly generated farm plans inone study, for example, were foundto be within the SSD set (King andRobison). The assumption of riskaversion required by SSD may notalways hold, and the pairwise com-parisons that are necessary indetermining the efficient set canbe computationally burdensome.


Ap

pen

dix

2:A

nal

ytic

al T

ool

s fo

rA

sses

sin

g th

e Ef

fect

iven

ess

of

Ris

kM

anag

emen

tSt

rate

gies


0

0.2

0.4

0.6

0.8

1

CA B

Income

Cumulative probability

Appendix figure 6

Cumulative income distributions under three alternative strategies

∫ ≤ ∫∞ ∞

0 0F F for all

values of x.

A B( ) ( ),x x


Ap

pen

dix

2:An

alytical Tools for

Assessin

g the Effectiven

ess of

Risk

Man

agemen

tStrategies

111199

Different Approaches Can Be Used to Estimate Risk Aversion

To determine a farmer’s best risk management strategy, information isneeded about his or her risk preferences among the different incomedistributions generated by those alternative strategies. Individualswho accept a lower average return to reduce the variability of returnsare said to be risk averse. Many individuals are believed to be riskaverse, as evidenced by the widespread demand for automobile, prop-erty, and health insurance. Premium costs for these products general-ly exceed expected indemnities due to administrative costs, but buyersoften find the price acceptable to mitigate potentially disastrous out-comes.

While risk aversion is acknowledged as widespread, the degree of riskaversion varies among individuals and is difficult to ascertain. Twogeneral approaches typically have been used in empirical analyses.The first approach measures risk aversion directly by confronting thedecisionmaker with a choice (either actual or hypothetical) among sev-eral alternatives, at least some of which involve risk (Newbery andStiglitz). Such approaches have been used to determine risk aversionamong farmers in northeast Brazil (Dillon and Scandizzo) and ruralIndia (Binswanger), among others.

Measuring risk preferences directly can, however, lead to unstableresults. Interview methods, in particular, are faced with the inevitableproblem that individuals may not be able to reveal their attitudestoward decisions they have never taken or seriously contemplated(Binswanger). In addition, such studies have typically focused on asmall-scale basis. Recently, work has been undertaken to measurefarmers’ risk attitudes on a large-scale basis, using rating scales of riskmanagement questions to ascertain farmers’ risk preferences (Bardand Barry).

The second approach does not involve interviews with decisionmakersor experimental determination of attitudes toward risk. Rather, thiscategory involves: 1) focusing on testing hypotheses econometricallyregarding risk preference structure; or 2) directly estimating utilityfunctional forms or risk aversion coefficients using data on actual firmchoices (Saha, Shumway, and Talpaz; Antle; Love and Buccola).Several studies have used this second category, with estimates of rela-tive risk aversion ranging widely.

More technically, measures of either “absolute” or “relative” risk aver-sion can be used to quantify an individual’s attitude toward risk. Bothmeasure the curvature of the utility function, and represent the degreeto which the satisfaction obtained from an additional unit of incomedeclines as income increases. The units of measurement must be con-sidered in interpreting estimates of absolute risk aversion, whereasrelative risk aversion is unit free.

Varying estimates of relative and absolute risk aversion result fromdifferent approaches and data sets. Researchers generally agree that areasonable relative risk aversion coefficient, for example, is in theneighborhood of 2.0, or “rather risk averse.” Relatively risk-aversefarmers would be likely to maintain substantial financial reserves asprotection against income shortfalls, while those who are less riskaverse would be inclined to borrow to near their limit in order toincrease their expected incomes (Hardaker, Huirne, and Anderson).

Further, once the efficient set ofplans is determined, identificationof the optimal choice within thisset depends on knowing moreabout a decisionmaker’s preferencethan merely that an unquantifiedaversion to risk exists (Anderson,Dillon, and Hardaker).

TThhee EExxppeecctteedd UUttiilliittyy AApppprrooaacchh

If a decisionmaker’s risk prefer-ences can be described mathemati-cally and the probability distribu-tions associated with each riskyalternative are known, his or herchoice among the risky alternativescan be optimized directly. Expectedutility provides a convenient way torepresent risk preferences. Thebasic idea is that decisionmakersmaximize expected utility, whereutility is an indicator of satisfactionmeasured in arbitrary units. Utilityincreases less than proportionatelywith income for decisionmakerswho are risk averse. In other words,utility is an increasing, but down-ward bending, function of incomefor such persons (Anderson, Dillon,and Hardaker; Robison and Barry;Laffont; Takayama).

Many different specifications canbe used to capture the curvature ofthe utility function, and each rep-resents the degree to which thesatisfaction obtained from an addi-tional unit of income changes asincome increases. One such utilityfunction specification can beexpressed as:

U = 100 - (1,000,000/Y)

As can be seen from this equation,an increase in Y, say from $20,000to $30,000 (50 percent), results inan increase in utility from 50 to 67(32 percent). This utility functionexhibits constant relative riskaversion, which means that thedegree of risk aversion decreaseswith income. The coefficient of rel-ative risk aversion in this exampleis 2, which is considered by manyeconomists to be about average.

To illustrate the use of expectedutility, suppose that Farmer Smith,whose utility function is specifiedas above, is choosing between twostrategies: (1) continuing to farm,and (2) taking a job in town. Underthe first strategy, Smith has an 80-percent chance of a net income of$70,000 and a 20-percent chance ofa net income of $20,000. Workingin town provides a sure income of$55,000. The expected income andstandard deviation in income foreach strategy are shown in appen-dix table 4. Neither strategy domi-nates the other from the stand-point of E-V efficiency because thefirst strategy has the highestexpected income, while the secondhas the lowest standard deviationin income (zero in this case).

Using Smith’s utility function, theresulting utilities for each level ofincome shown in the table are:

Income Utility

$20,000 50.0

$55,000 81.8

$70,000 85.7


Ap

pen

dix

2:A

nal

ytic

al T

ool

s fo

rA

sses

sin

g th

e Ef

fect

iven

ess

of

Ris

kM

anag

emen

tSt

rate

gies

Appendix table 4—Expected income and standard deviation of income undertwo strategies available to Farmer Smith

StandardProbabilities of expected incomes Expected deviation of

Strategy $20,000 $55,000 $70,000 income income

Percent Dollars

1 0.2 0.0 0.8 60,000 20,0002 0.0 1.0 0.0 55,000 0



Ap

pen

dix

2:An

alytical Tools for

Assessin

g the Effectiven

ess of

Risk

Man

agemen

tStrategies

112211

Certainty Equivalence Allows Estimation of Risk inDollars of Expected Income

Decisionmakers often would like the losses from risk or the gainsfrom risk reduction to be measured in terms of dollars of expectedincome. A sure outcome that an individual finds equally desirable toa given risky prospect is called a certainty equivalent outcome(Anderson, Dillon, and Hardaker; Laffont). Knowing certainty equiva-lent outcomes allows one not only to rank risky alternatives, but alsoto estimate the cost of risk, or the premium that the individual wouldpay to avoid the risk. Certainty equivalence simultaneously accountsfor the probabilities of the risky prospects and the preferences for theconsequences (Anderson, Dillon, and Hardaker). Because decision-makers seldom have similar attitudes toward risk, certainty equiva-lents vary among individuals, even for the same risky prospect(Hardaker, Huirne, and Anderson).

Certainty equivalents can be calculated when E-V analysis is used orwhen the utility function is known and expected utility analysis isused. In the latter case, a certainty equivalent can be calculated byfirst calculating expected utility and then finding the sure outcomethat would provide equal utility. This involves applying the inverse ofthe utility function to expected utility. For the utility function abovethis gives:

CE(Y) = 1,000,000 / [100 - E(U)]

Applying this inverse function to the expected utilities calculated inthe accompanying table gives the estimates shown in the accompany-ing table. These results, shown in the last column, indicate that thefirst strategy yields a certainty equivalent income of $46,667, com-pared to an expected income of $60,000. In other words, the cost ofuncertainty in farming for Smith is $60,000 - $46,667 = $13,333.Thus, Smith prefers the second strategy, which gives a certaintyequivalent income of $55,000, and is $8,333 higher than obtained ifhe had chosen the first strategy.

Certainty equivalent estimates must be used with care, primarilybecause they tend to convey an unwarranted sense of precision. Theymust be taken as rather rough approximations, and depend heavilyon how accurately the underlying utility functions and probabilitydistributions are estimated. Utility functions, in particular, differmarkedly among individuals, are not directly observable, and arelikely estimated with substantial errors in most cases.

Certainty equivalent income under two strategies for farmer SmithExpected Certainty equivalent

Strategy utility income

Dollars

1 78.6 46,667

2 81.8 55,000Source: Hypothetical example developed by ERS.

Expected utility under each strat-egy is calculated by weightingeach utility level by its probability.The results for the two strategies,shown in appendix table 5, indi-cate that the second strategyyields a slightly higher expectedutility and is therefore preferred.Expected utility results can alsobe used to estimate certaintyequivalents (see accompanyingbox, p. 121).

The use of expected utility hasboth advantages and disadvan-tages. One advantage is that this

approach is quite generalizable,allowing a wide choice of utilityfunctions and probability distribu-tions. Unlike stochastic dominanceand E-V efficiency criteria, whichtypically leave some alternativesunranked, expected utility general-ly ranks all alternatives. Themajor drawback of the approach isthat utility functions are difficultto estimate and known onlyapproximately, at best. Moreover, itassumes that decisionmakersexhibit a high level of rationality,which does not always seem to bethe case.


Ap

pen

dix

2:A

nal

ytic

al T

ool

s fo

rA

sses

sin

g th

e Ef

fect

iven

ess

of

Ris

kM

anag

emen

tSt

rate

gies

Appendix table 5—Example calculation of expected utility under two strategiesfor farmer Smith

Probabilities of expected utilities ExpectedStrategy 50.0 81.8 85.7 utility

Probability

1 0.2 0.0 0.8 78.62 0 1.0 0 81.8


The accompanying tables illus-trate the mechanics of several

types of federally-subsidized cropand revenue insurance. Thesetypes are:

Catastrophic (CAT) coverage—A type of Federal crop insurancethat guarantees 50 percent of theproducer’s actual production histo-ry (APH) yield at 55 percent of theprice election under crop insurancereform. Producers pay a processingfee of $60 per crop. The processingfee is tied to persons who have aninterest in the land, not theacreage itself.

“Additional” (or Buy Up) cover-age—A type of Federal crop insur-ance that refers to coverage thatequals or exceeds a 50-percentyield guarantee at 100 percent ofthe price election. Farmers mustpay a processing fee and a premi-um in order to receive added cov-erage.

Income Protection (IP)—A typeof revenue insurance that protectsproducers against reductions ingross income when a crop’s priceor yields decline from early-seasonexpectations. Indemnities are paidif the producer’s gross income (asmeasured by the product of the

producer’s realized yield and theharvest futures price) falls below apredetermined guarantee. Thres-hold trigger levels are based on aproducer’s APH yield and a plant-ing-time price for the harvestfutures contract. Coverage is basedon enterprise units. Coverageoptions in most areas range from50 to 75 percent in 5-percent incre-ments.

Crop Revenue Coverage(CRC)—A type of revenue insur-ance that provides revenue insur-ance plus replacement-cost protec-tion to producers. Indemnities arepaid if the producer’s gross income(as measured by the product of theproducer’s realized yield and theactual harvest price) falls below apredetermined guarantee (asmeasured by the product of theproducer’s APH yield and the high-er of the early-season price projec-tion or the actual harvest price).Since CRC uses the higher of theplanting-time price for the harvestfutures contract or the actualfutures contract quote at harvest,the producer’s guarantee mayincrease over the season, allowingthe producer to purchase “replace-ment” bushels if yields are low andprices increase during the season.


Ap

pen

dix

3:Examp

les ofV

ariou

s Typ

es of

Cro

p an

d R

evenu

e Insu

rance

112233

AAppppeennddiixx 33:: EExxaammpplleess ooff VVaarriioouuss TTyyppeess ooffCCrroopp aanndd RReevveennuuee IInnssuurraannccee


Ap

pen

dix

3:Ex

amp

les

ofV

ario

us

Typ

es o

fC

rop

an

d R

even

ue

Insu

ran

ce

Appendix table 6—Example of a Catastrophic Crop Insurance Policy for YIELD Loss 50 percent yield coverage and 55 percent of price electionThe producer pays no premium

A B CUnit Yield #1 Yield #2 Yield #3

Step Item (30) (13) (0)

1 Net indemnity to producer $/acre 0 15.40 44.00

2 = Indemnity paid to producer $/acre 0 15.40 44.00

3 = Calculated yield loss bu/acre 0 7 20

4 = Yield guarantee bu/acre 20 20 205 = percent of yield coverage percent 50.0 50.0 50.06 * APH yield guarantee bu/acre 40 40 40

7 - Actual harvested yield bu/acre 30 13 0

8 * Producer price election $/bu 2.20 2.20 2.209 = percent of price election percent 55.0 55.0 55.010 * FCIC price election $/bu 4.00 4.00 4.00

11 - Producer premium $/acre 0 0 0

12 = Total premium per acre $/acre 4.50 4.50 4.50

13 - Premium subsidy per acre $/acre 4.50 4.50 4.5014 = Subsidy percent percent 100 100 10015 * Total premium per acre $/acre 4.50 4.50 4.50Source: Excerpted from Jagger, Craig, and Joy Harwood. “Module 6: Crop Insurance and Revenue Insurance,”USDA’s Crop/Commodity Programs After the 1996 Farm Act. USDA Graduate School Course, winter quarter,1999.

Appendix table 7—Example of a Multi-Peril Crop Insurance Policy for YIELD Loss The producer chooses: percent yield coverage and percent of price election The producer pays part of the premium


Step Item (30) (13) (0)

1 Net indemnity to producer $/acre -3.50 48.50 100.50

2 = Indemnity paid to producer $/acre 0 52.00 104.00

3 = Calculated yield loss bu/acre 0 13 26

4 = Yield guarantee bu/acre 26 26 265 = percent of yield coverage percent 65.0 65.0 65.06 * APH yield guarantee bu/acre 40 40 40

7 - Actual harvested yield bu/acre 30 13 0

8 * Producer price election $/bu 4.00 4.00 4.009 = percent of price election percent 100 100 10010 * FCIC price election $/bu 4.00 4.00 4.00

11 - Producer premium $/acre 3.50 3.50 3.50


13 - Premium subsidy per acre $/acre 2.50 2.50 2.5014 = Subsidy percent percent 41.7 41.7 41.715 * Total premium per acre $/acre 6.00 6.00 6.00Source: Excerpted from Jagger, Craig, and Joy Harwood. “Module 6: Crop Insurance and Revenue Insurance,”USDA’s Crop/Commodity Programs After the 1996 Farm Act. USDA Graduate School Course, winter quarter,1999.


Ap

pen

dix

3:Examp

les ofV

ariou

s Typ

es of

Cro

p an

d R

evenu

e Insu

rance

112255

Appendix table 8--Example of an Income Protection Insurance Policy for REVENUE LossThe producer chooses: percent of revenue coverageThe producer pays part of the premium


Step Item Price #1 Price #2 Price #3

1 Net indemnity to producer $/acre 10.68 -3.32 -3.32

2 = Indemnity paid to producer $/acre 14.00 0 0

3 = Revenue guarantee = planting guarantee $/acre 104.00 104.00 104.00

4 = Harvest price as projected at planting time $/bu 4.00 4.00 4.00

5 * APH yield bu/acre 40 40 406 * Revenue coverage level percent 65.0 65.0 65.0

7 - Realized revenue $/acre 90.00 120.00 150.00

8 = Realized harvest price $/bu 3.00 4.00 5.009 * Actual yield bu/acre 30 30 30




Appendix table 9--Example of a Crop Revenue Coverage Insurance Policy forREVENUE Loss The producer chooses: percent of revenue coverage The producer pays part of the premium


Step Item Price #1 Price #2 Price #3

1 Net indemnity to producer $/acre 8.35 -5.66 -5.66

2 = Indemnity paid to producer $/acre 14.00 0 0

3 = Revenue guarantee = higher of (a) or (b) $/acre 104.00 104.00 104.00

4 (a) Harvest guarantee $/acre 78.00 104.00 130.005 = Realized harvest price $/bu 3.00 4.00 5.006 * APH yield bu/acre 40 40 407 * Revenue coverage level percent 65.0 65.0 65.0

8 (b) Planting guarantee $/acre 104.00 104.00 104.009 = Harvest price as projected

at planting time $/bu 4.00 4.00 4.0010 * APH yield bu/acre 40 40 4011 * Revenue coverage level percent 65.0 65.0 65.0

12 - Realized revenue $/acre 90.00 120.00 150.0013 = Realized harvest price $/bu 3.00 4.00 5.0014 * Actual yield bu/acre 30 30 30




managing risk in farming: concepts, research, …managing risk in farming: concepts, research, and...

Documents