Agribusiness Management · Agri-sector specialisation · also ABM 524, Risk Management in Agribusiness (MANAGE), EXT 605 Risk & Climate Change Adaptation
The risks farmers and agribusinesses face and the tools to manage them: insurance, hedging and sound credit policy.
In the exams: agricultural master's entrance examinations, crop insurance (heavy); agricultural economics PhD entrance and fellowship examinations, risk and insurance (heavy); agricultural field officer recruitment in banking (heavy); agricultural development bank officer recruitment (present)
Agricultural risk management is the study of what can go wrong between the decision to sow, rear or buy and the moment money is finally in hand, and of the tools that farmers, agribusinesses, banks and governments use to live with that uncertainty. A farmer commits land, labour and borrowed money months before knowing the rainfall, the pest pressure or the harvest price. A processor commits to supply contracts before knowing what raw material will cost. A rural bank lends against a crop that may never be harvested. Each of them needs a way to think about risk clearly and a set of instruments to share or shed part of it.
This manual covers the whole ground examined in agribusiness and agricultural economics programmes: the types of risk and how they are measured; decision rules under risk and uncertainty; the strategies farmers use to diffuse risk on and off the farm; the economics of insurance, including moral hazard, adverse selection and basis risk; the history of crop insurance in North America, Asia, Africa and Latin America; indemnity, area-yield and weather-index insurance; the design of large publicly supported schemes such as those of the United States, India and China; livestock insurance; hedging with futures and options; and credit risk in agricultural lending.
Examiners approach the subject from different angles. Postgraduate entrance and fellowship papers in agricultural economics ask for definitions, scheme history, the threshold-yield and claim formula, and decision rules. Bank and development-finance recruitment papers ask about scheme enrolment, premium rates, crop-loan linkage and non-performing asset norms. Graduate programmes in the United States, Europe, Africa and Australia treat the same material through expected utility, index design and the political economy of subsidised insurance. Scheme rules change from season to season; wherever this manual gives a scheme parameter it says so, and a practitioner should confirm the current operational guidelines before acting on it.
Agriculture differs from most industries in three ways that make risk central. First, production is biological and exposed to weather, pests and disease that the manager cannot control. Second, there is a long lag between committing inputs and receiving output, so decisions are made on expectations rather than known outcomes. Third, many farms are small price-takers selling into markets where supply is inelastic in the short run, so a good harvest across a region can depress price at the same time as it raises yield. Risk in agriculture is therefore both large and often covariate, meaning that many farmers in the same area suffer the same shock at the same time. Covariate risk is what makes agricultural insurance harder to pool than, say, household fire insurance, where one house burning does not make the neighbour's house more likely to burn. Idiosyncratic risk, by contrast, falls on one farm or household, such as illness of the farmer or a fire in one barn, and can be pooled cheaply within a community or by an insurer. The practical question for every manager is which risks to carry, which to reduce and which to pass to someone else at a price.
Frank Knight drew the classic distinction in 1921. Risk describes situations in which the possible outcomes and their probabilities are known or can be estimated objectively, for example from long yield records. Uncertainty describes situations in which the probabilities cannot be known, such as the effect of a new pest or a policy change. Modern decision analysis, following Hardaker, Huirne, Anderson and Lien in Coping with Risk in Agriculture, softens this line by using subjective probabilities: a decision maker's degree of belief can be expressed as a probability and used in the same calculations. Examination questions still expect the Knightian definitions, and they expect students to match decision rules to each case: expected value and expected utility where probabilities are known, and rules such as maximin or minimax regret where they are not. A further distinction is between the probability of an event and its consequence: a frequent small loss and a rare ruinous loss may have the same expected value but demand quite different management.
Standard textbooks group agricultural risk into five sources. Production or yield risk comes from weather, pests, diseases, fire and the uncertain performance of new technology. Price or market risk comes from movements in output and input prices after production decisions are made, and from uncertainty over whether a buyer will actually take delivery. Institutional risk arises from changes in government policy and rules: minimum support prices, export bans, stock limits, subsidy changes, land and tenancy laws, environmental regulation and food-safety standards. Personal or human risk comes from death, illness, injury, divorce, disputes among heirs and the loss of key labour. Financial risk arises from the way the business is financed: interest-rate changes, the availability of credit, and the leverage that magnifies both gains and losses when borrowed capital is used. The first four together are often called business risk, the variability of returns the business would face even with no debt. Financial risk is added on top of business risk as the debt-to-equity ratio rises, which is why a highly leveraged farm can fail after a shock that an equity-financed neighbour survives. Some texts add a separate technological risk (the chance that a new variety or machine performs worse than promised) and some treat it inside production risk.
The expected value (EV) of a risky prospect is the probability-weighted average of its outcomes: EV = sum of p times x over all outcomes. Variance is the probability-weighted average of squared deviations from the expected value, and the standard deviation is its square root. Because the standard deviation is measured in the same units as the outcome, it is hard to compare across enterprises of different size, so analysts use the coefficient of variation (CV), the standard deviation divided by the mean, often expressed as a percentage. A lower CV means less risk per unit of expected return. When choosing between enterprises by the mean-variance (E-V) rule, a plan is efficient if no other plan offers a higher mean with the same or lower variance, or a lower variance with the same or higher mean. The set of such plans is the E-V efficient frontier; the decision maker's own risk attitude decides where on the frontier to be. Mathematical programming models such as quadratic programming and MOTAD (minimisation of total absolute deviations, due to Hazell) generate this frontier for whole-farm plans.
Expected value alone ignores the fact that most farmers dislike risk. The expected utility hypothesis, formalised by von Neumann and Morgenstern and applied to farming by Anderson, Dillon and Hardaker, says that a rational decision maker chooses the prospect with the highest expected utility, where utility is a function of wealth or income. A concave utility function represents risk aversion, a linear one risk neutrality and a convex one risk preference. The certainty equivalent (CE) of a risky prospect is the sure sum that gives the same utility as the prospect. The risk premium is EV minus CE: the amount of expected income a person would give up to be rid of the risk. For a risk-averse person the risk premium is positive, and this is exactly why a farmer may rationally buy insurance whose premium exceeds the expected indemnity. Stochastic dominance is a way to rank prospects without knowing the exact utility function: first-degree dominance holds for anyone who prefers more to less, and second-degree dominance holds for anyone who is also risk averse. The safety-first family of rules, beginning with Roy, instead asks the decision maker to keep the probability of income falling below a disaster level under a chosen limit.
When probabilities are unknown, textbooks teach five classical rules applied to a payoff matrix of actions against states of nature. The maximin rule (Wald) chooses the action whose worst outcome is best, a pessimistic rule suited to a poor farmer who cannot survive a disaster. The maximax rule chooses the action whose best outcome is best, an optimistic rule. The Hurwicz rule weights the best and worst outcome of each action by a coefficient of optimism alpha and chooses the highest weighted value; with alpha of zero it reduces to maximin and with alpha of one to maximax. The minimax regret rule (Savage) first builds a regret table, where regret is the difference between the best payoff available in a state and the payoff of the chosen action in that state, and then chooses the action whose maximum regret is smallest. The Laplace rule treats every state as equally likely and chooses the highest simple average. Students should practise building regret tables by hand, because examiners like to set a three-by-three matrix in which the rules point to different actions.
Farmers managed risk long before formal insurance existed, and most risk is still carried by these strategies. Diversification spreads the farm across enterprises whose returns are not perfectly correlated: crops with livestock, a rainy-season cereal with a pulse, or plots in different locations. Its benefit depends on correlation; two enterprises whose incomes move together give almost no protection. Mixed and intercropping reduce the chance that all components fail together. Technical measures reduce production risk directly: irrigation, drought-tolerant or pest-resistant varieties, staggered sowing, crop rotation, integrated pest management and good storage. Flexibility means keeping the ability to change course as information arrives: choosing machinery that suits several crops, renting rather than buying land, or keeping part of the land for a late-sown catch crop if the monsoon fails. Liquidity and credit reserves mean holding cash, grain stocks or unused borrowing capacity so that a bad year can be absorbed without selling productive assets. Selecting enterprises with a short production cycle reduces the time over which things can go wrong.
The second group of strategies moves part of the risk to someone else. Crop and livestock insurance transfer production risk to an insurer in return for a premium. Forward contracts fix a price with a buyer before harvest, and exchange-traded futures and options transfer price risk to speculators and other hedgers. Contract farming shares both price and, in some designs, production risk between farmer and firm. Share tenancy (sharecropping) spreads yield risk between landlord and tenant, which is one reason economists such as Cheung explained its persistence. Vertical integration moves the risk of one stage into a firm that controls the next. Cooperatives and producer organisations pool marketing, input purchase and sometimes price risk among members. Off-farm employment and remittances give the household income that is not correlated with the farm. Informal risk sharing through kin, village networks, rotating savings groups and gifts is important in low-income settings, though it fails when a shock hits the whole village at once.
The OECD's holistic framework for agricultural risk divides risk into three layers. Normal risks are frequent, small losses that farmers should retain and manage through on-farm strategies and savings. Marketable risks are intermediate risks that private markets can price and share, through insurance, futures and contracts. Catastrophic risks are rare, large and highly covariate events that exceed the capacity of private markets; here governments step in through disaster relief, reinsurance, subsidised catastrophe cover and social safety nets. The framework warns that public support for normal risks crowds out farmers' own risk management and can create moral hazard, while failure to support catastrophic risk leaves a market gap. A second useful distinction is between ex ante risk management, taken before a shock (insurance, diversification, savings), and ex post coping, taken after it (selling livestock, withdrawing children from school, borrowing at high rates, cutting consumption). Poor households forced into ex post coping often lose productive assets and fall into a poverty trap, which is why development agencies promote ex ante tools.
Insurance pools many independent risks so that the average loss per member becomes predictable, which is the law of large numbers at work. An insurer charges a premium that covers the expected loss (the pure or actuarially fair premium), plus a loading for administration, the cost of capital held against bad years, and profit. For a risk to be commercially insurable, textbooks list several conditions: there must be many similar exposure units; losses must be accidental and outside the insured's control; the loss must be definite and measurable; probabilities must be estimable; losses should be independent across the insured; and the premium must be affordable. Agricultural risks fail several of these tests. Drought and flood are covariate, so losses are correlated across a whole district. Yield records at farm level are short or missing, so probabilities are hard to estimate for each farmer. And the farmer's own management affects the outcome. Insurers cope with covariance by buying reinsurance and by diversifying across regions and crops; governments often act as reinsurer of last resort.
Both problems arise from asymmetric information, where the insured knows more than the insurer. Adverse selection is a problem of hidden information before the contract: if the premium is set at the average risk, farmers who know their land is high risk find it a bargain and buy, while low-risk farmers find it expensive and stay out, so the pool becomes riskier than assumed and losses exceed premiums. Remedies include individual risk rating using farm yield histories, compulsory or credit-linked enrolment that keeps low-risk farmers in the pool, fixed enrolment deadlines before sowing or before the season's weather is known, and area-based products where the insurer and farmer have the same information. Moral hazard is a problem of hidden action after the contract: an insured farmer may take less care, apply less fertiliser or weed less, because the insurer bears part of the loss. Remedies include deductibles and coverage levels below 100 percent so the farmer always bears part of the loss, loss inspection, and index products where the payout does not depend on the farmer's own yield.
Basis risk in insurance is the mismatch between the payout an index product makes and the loss the farmer actually suffers. A farmer may lose her crop to a local hailstorm while the area yield or the rainfall at the reference station shows nothing wrong, so she receives nothing; or the index may trigger a payout to a farmer who suffered no loss. Three sources are usually distinguished. Spatial basis risk arises from distance between the farm and the measuring point, such as the weather station or the boundary of the insurance unit. Temporal basis risk arises when the index measures the wrong period, for instance counting rainfall in fixed calendar windows while the crop's sensitive stage shifted because of late sowing. Design or product basis risk arises when the index is a poor model of yield, for example using only rainfall when heat or pests caused the loss. The central trade-off of agricultural insurance design is now clear: indemnity products that measure each farmer's loss have low basis risk but high moral hazard, adverse selection and loss-assessment cost, while index products have low moral hazard and adverse selection and low cost but carry basis risk. Area-yield insurance sits in between.
Private hail insurance was sold in Europe from the eighteenth and nineteenth centuries, because hail damage is local and independent across farms and so meets the insurability tests. Multiple-peril crop insurance, covering drought and other covariate risks, was repeatedly tried by private firms and generally failed, which is why most national programmes are public or publicly supported. In the United States the Federal Crop Insurance Act of 1938 created the Federal Crop Insurance Corporation, first insuring wheat. Participation stayed low for decades. The Federal Crop Insurance Act of 1980 expanded the programme and brought in private insurance companies to sell and service policies, with the federal government subsidising premiums and reinsuring the companies under a Standard Reinsurance Agreement. The Federal Crop Insurance Reform Act of 1994 introduced catastrophic (CAT) coverage and linked some farm programme benefits to buying insurance; the Risk Management Agency was created in 1996 to administer the programme; and the Agricultural Risk Protection Act of 2000 raised premium subsidies substantially. Today crop insurance is the largest element of the US farm safety net.
Asian programmes show the same path from costly individual cover to area-based and subsidised schemes. Japan built a farmer-mutual crop insurance system under its 1947 agricultural loss compensation law, with heavy state reinsurance, and the Philippines created a public crop insurer in 1978. India's first experiment, an individual-farm scheme for cotton in 1972-73, showed that individual assessment was expensive and exposed to moral hazard; the economist V. M. Dandekar argued for a homogeneous area approach, in which all farmers in an area receive the same indemnity based on area yield, and a pilot in 1979-84 tested it. India then ran credit-linked area-yield schemes from 1985, extended them to non-borrowing farmers in 1999, created a specialised public insurer in 2002 and piloted weather-index cover from 2007. China began central premium subsidies in 2007, delivered through commercial insurers, and within a decade became one of the largest agricultural insurance markets in the world by premium. Across the region the lessons were similar: area units reduce moral hazard and cost, linkage with seasonal credit raises enrolment, and subsidy is needed to reach smallholders.
India consolidated its schemes in 2016 into an area-yield scheme (PMFBY) and a restructured weather-index scheme (RWBCIS); Chapter 8 uses these, with the US and Chinese programmes, to show how public schemes are designed. Internationally, area-yield insurance was proposed by Halcrow in 1949 and later analysed by Miranda (1991), and the US offered a Group Risk Plan based on county yields from the 1990s. Weather-index insurance for smallholders spread from the early 2000s with World Bank support: a groundnut rainfall-index pilot linked to credit in Malawi in 2005 is one of the most cited early cases, and pilots followed in Ethiopia, Kenya, Mexico and elsewhere. Mongolia's index-based livestock insurance, using district-level livestock mortality rates, started in the mid-2000s. In East Africa, the International Livestock Research Institute (ILRI) and partners launched index-based livestock insurance for pastoralists in northern Kenya in 2010, using a satellite vegetation index. Sovereign risk pools such as the African Union's African Risk Capacity insure governments rather than farmers against drought. The long history teaches three lessons: purely private multi-peril crop insurance rarely survives; uptake among smallholders is low without subsidy or credit linkage; and product design must balance basis risk against moral hazard and cost.
Indemnity-based crop insurance pays each farmer for the loss measured on his or her own farm. Named-peril policies cover specific hazards such as hail or fire; multiple-peril crop insurance (MPCI) covers yield shortfall from most natural causes. In the standard yield-protection design used in the United States, the insured's yield guarantee is the farm's average historical yield, the Actual Production History (APH), multiplied by a chosen coverage level (commonly offered from 50 to 85 percent). If harvested yield falls below the guarantee, the indemnity is the shortfall multiplied by a price election. Revenue protection extends this by guaranteeing revenue: the guarantee is APH times coverage level times the higher of the projected (pre-planting) and harvest futures-based price, and the indemnity is the amount by which actual revenue (actual yield times harvest price) falls short. Revenue insurance therefore covers price and yield together and relies on the natural hedge between them. Individual products are accurate but need farm-level yield records, field loss adjustment and strong controls against moral hazard and fraud, which make them costly for very small farms.
Area-yield index insurance, often called yield-index insurance, pays every insured farmer in a defined insurance unit the same percentage indemnity when the average yield of that unit falls below a threshold. The area yield is measured by crop cutting experiments (CCEs) or, increasingly, by remote sensing and models. Because no single farmer can influence the area average much, moral hazard is weak, and because the insurer and the farmer have the same information about the area's history, adverse selection is reduced. In schemes such as India's the logic is: threshold yield = average yield of a reference period multiplied by the indemnity level; if actual area yield is below threshold, claim = (threshold yield minus actual yield) divided by threshold yield, multiplied by the sum insured. India's PMFBY, for example, takes the reference yield as the average of the best five of the past seven years. For example, if the best-five average is 30 quintals per hectare and the indemnity level is 80 percent, the threshold is 24 quintals; if the measured area yield is 18 quintals, the shortfall is 6 of 24, or 25 percent, so a farmer with a sum insured of 40,000 currency units receives 10,000 currency units. Students commonly err by dividing the shortfall by the average yield instead of the threshold yield.
A crop cutting experiment harvests a small plot of fixed size chosen by random sampling within the insurance unit, weighs the produce, dries it where required, and converts it to yield per hectare. Its statistical design comes from mid-twentieth-century sample-survey work, notably P. V. Sukhatme's at FAO and in South Asia. The accuracy of an area-yield product depends on the number of CCEs, the honesty of their conduct and the homogeneity of the unit. Large units carry more spatial basis risk because a farmer's yield is less correlated with a large area's average; small units need many more CCEs, which strains state agriculture and revenue staff. This is why several schemes, India's among them, have moved the unit for major crops down to the village, and why technology such as smartphone recording of CCEs, satellite imagery and crop-simulation models is being used to improve and verify yield estimates. Area-yield products also cannot capture localised losses such as a hailstorm on one field; some schemes therefore add individual assessment for named localised calamities (see Chapter 8). Area-revenue products, which multiply area yield by a market price, extend the idea to price risk.
Weather-index insurance pays according to a measured weather variable at a reference weather station rather than according to measured crop loss. The variable is chosen because it explains a large part of yield variation for a crop in a place: cumulative rainfall in defined phases, number of consecutive dry days, excess rainfall, maximum or minimum temperature, relative humidity, wind speed, or frost. A contract is written as a term sheet. For a deficit-rainfall cover, the term sheet specifies the period and phase, the trigger (the rainfall level below which payment starts), the exit (the level at or below which the maximum payout is made, usually treated as crop failure), the payout rate per unit of shortfall and the maximum payout. Between trigger and exit the payout typically rises linearly: payout = (trigger minus actual) divided by (trigger minus exit) times the maximum payout, or equivalently shortfall in millimetres times a rate per millimetre. Excess-rainfall, heat and humidity covers are written the same way with the inequality reversed. Payouts can be settled quickly after the cover period because no field loss assessment is needed.
Weather-index insurance has clear strengths: payouts are objective and transparent, loss adjustment costs are low, moral hazard is essentially removed because the farmer cannot influence rainfall, adverse selection is reduced if sales close before the season's weather becomes known, and the product can be reinsured on international markets because the risk is measured independently. Its main weakness is basis risk. The density of weather stations is decisive: rainfall in monsoon climates can vary sharply over a few kilometres, so a farm far from the station may suffer drought while the station records adequate rain. A second weakness is the need for long, clean historical weather data, often thirty years or more, to price the contract; gaps can be filled with satellite rainfall estimates or reanalysis data, but these add their own error. Third, the index captures only the weather peril it measures; pest outbreaks, disease and management failures are not covered. Farmers who receive nothing after a real loss lose trust, so poor design damages uptake for years.
Where weather stations are sparse, satellite data provide an alternative index. The Normalised Difference Vegetation Index (NDVI) measures greenness from the difference between near-infrared and red reflectance; it is used as a proxy for pasture availability and crop condition. The index-based livestock insurance programme in northern Kenya and southern Ethiopia, developed by ILRI with partners, pays pastoralists when the NDVI-based forage index for their area falls low enough to predict high livestock mortality; later versions pay earlier, as asset protection, so that households can buy feed or water before animals die. Kenya's government later bought such cover for vulnerable pastoralists under a public livestock insurance programme. In Ethiopia, the R4 Rural Resilience Initiative, which grew from the HARITA pilot of Oxfam America and partners with the World Food Programme, allowed cash-poor farmers to pay premiums with labour on community risk-reduction projects, a model called insurance-for-work. At the sovereign level, the African Risk Capacity insures participating governments against drought using a rainfall-based model, so that relief can be funded early. Evidence from these pilots is that uptake at unsubsidised prices is low, bundling with credit, inputs or savings helps, and trust built by prompt payouts matters as much as the design.
Large public crop insurance schemes share common objectives: to compensate farmers for losses from events they cannot control, to stabilise income so that they stay in farming and keep investing, and to keep credit flowing to agriculture. They differ in delivery and enrolment. The US federal programme is voluntary, sold and serviced by private insurers that the government subsidises and reinsures. India's Pradhan Mantri Fasal Bima Yojana (PMFBY), from 2016, covers food crops, oilseeds and annual commercial and horticultural crops notified by state governments, with insurers chosen by bidding for clusters of districts; it was compulsory for borrowers until 2020 and is now voluntary, with borrowers enrolled by their bank unless they opt out before a cut-off date. China's subsidised policies are sold by commercial insurers, often with village collectives organising enrolment. Tenants and sharecroppers are eligible in some schemes and not in others, which matters because they bear the crop risk.
Almost every large scheme subsidises premiums, because correlated risk and high delivery costs make unsubsidised premiums too high for most farmers. Subsidy rules differ. India's PMFBY caps the farmer's share at 2 percent of the sum insured for main-season food and oilseed crops, 1.5 percent for second-season food and oilseed crops and 5 percent for annual commercial and horticultural crops, with the rest of the actuarial premium paid as subsidy shared between the national and state governments; since 2020 the national share applies only up to set premium rates, so a state wanting cover at higher rates bears the excess. The US programme instead pays a share of the premium that falls as the farmer buys more coverage (for basic and optional units, 59 percent at 70 percent coverage but 38 percent at 85 percent), which keeps catastrophic cover cheap while making farmers pay more for shallow losses. In credit-linked schemes the sum insured per hectare is usually tied to the crop-loan limit per hectare (the scale of finance), so a claim can repay the loan. Governments and insurers may also share risk directly: in a cup-and-cap design, first used in one Indian district, the insurer's liability is capped at a multiple of premium and part of the premium is returned when claims are low. Because subsidy rules and caps are revised from time to time, current guidelines should always be checked.
Public schemes typically cover several stages of the season, and India's PMFBY illustrates the full set. Prevented sowing: if most insured farmers in an area cannot sow because of rainfall deficit or adverse conditions, an indemnity of up to 25 percent of the sum insured is paid and cover for that crop ends. Standing crop: yield losses from non-preventable risks such as drought, dry spell, flood, widespread pest and disease attack, landslide, natural fire, storm and cyclone are paid on the area approach. Mid-season adversity: an on-account payment can be made when expected area yield is likely to be very low, to get cash to farmers before final estimates are complete. Post-harvest: crops left to dry in the field are covered for a short period (14 days in PMFBY) against hail, cyclone and unseasonal rain, assessed farm by farm. Localised calamities: hailstorm, landslide, inundation, cloudburst and lightning fire affecting isolated farms are assessed individually if reported promptly. Losses from war, nuclear risk, malicious damage and other preventable causes are excluded. The standing-crop claim uses (threshold yield minus actual yield) divided by threshold yield times sum insured, with indemnity levels of 70, 80 or 90 percent for high-, moderate- and low-risk areas, and actual yield measured by crop cutting experiments supplemented by remote sensing. The US programme offers a parallel menu at farm level: yield protection, revenue protection (using the higher of the projected and harvest price) and area-based plans.
Several governments run weather-index products alongside area-yield schemes. They use weather parameters as a proxy for yield: deficit and excess rainfall, temperature, humidity and wind recorded at reference weather stations. For each crop and area a term sheet specifies the phases, triggers, exits and payout rates. India's restructured weather-based scheme (RWBCIS), launched with PMFBY in 2016, has the same farmer premium shares and is used mainly for horticultural and plantation crops where weather-station networks are dense; Mexico's catastrophic agricultural insurance uses weather and area-yield indices bought by state governments on behalf of smallholders. Claims are calculated automatically from weather data, without crop cutting experiments, which makes settlement fast but leaves the basis risk described in Chapter 7. Examiners often ask students to contrast the two families: area-yield index schemes measure actual yields and add individual assessment for localised and post-harvest losses; weather-index schemes are parametric, with no field assessment.
Livestock insurance protects the owner against the death of an animal, and in some policies against permanent total disability. The conventional policy for cattle and buffalo, common to general insurers in India and similar in principle elsewhere, covers death from accident (including fire, lightning, flood, cyclone and similar perils), disease contracted during the policy period, surgical operations, and riots or strikes. Typical exclusions are theft and clandestine sale, intentional slaughter except on a veterinarian's advice, malicious injury or neglect, war and nuclear perils, pre-existing disease and injury, and transport beyond limits stated in the policy. Disease cover usually starts after a short waiting period, commonly 15 days from the start of the policy, which guards against insuring an animal already ill; the exact terms vary by insurer and country. The sum insured is normally the market value of the animal, assessed by a veterinarian on age, breed, milk yield and body condition. Premium rates vary by species, region and whether the scheme is subsidised or bank-linked. Similar policies exist for sheep and goats, pigs, poultry (often on a flock basis), horses and aquaculture stock.
Livestock insurance faces moral hazard and fraud in an especially sharp form, because one animal looks much like another. The central control is identification: the insured animal carries a unique ear tag (or, increasingly, a microchip or muzzle-print image) whose number is recorded on the policy, with a photograph of the animal and the owner. To claim, the owner must notify the insurer promptly after death and produce the ear tag, a veterinarian's death certificate, a post-mortem report where required, and photographs. Missing or replaced tags are the most common source of rejected claims, so owners must report a lost tag and have the animal re-tagged. Premiums for bank-financed animals are often paid through the loan, which gives the bank a stake in timely claims. Many governments subsidise livestock premiums: India's livestock insurance, now part of its National Livestock Mission, and China's subsidised insurance for sows and dairy cows are large examples. Subsidy shares, the number of animals per household and eligible species have changed with the scheme's revisions and differ between general and disadvantaged categories, so the current guidelines must be checked.
Indemnity livestock insurance works poorly for extensive pastoral herds, where animals graze over large, remote areas and individual deaths cannot be verified cheaply. Index products avoid verification. Mongolia's index-based livestock insurance pays herders when the officially measured mortality rate of a species in their district exceeds a trigger, so the individual herder's loss need not be assessed; small losses are retained by herders, intermediate losses are covered by commercial insurance, and catastrophic losses are backed by the government, a practical application of the risk-layering idea. The satellite-based programmes in Kenya and Ethiopia described in Chapter 7 predict mortality or forage scarcity from NDVI. Both kinds of index carry basis risk: a herder may lose many animals in a district where the average mortality is below the trigger. For dairy and intensive units, other products cover disease outbreaks, business interruption and loss of milk income, and in the United States federal programmes include livestock price and gross-margin insurance that protect against falls in livestock prices or in the margin between livestock and feed prices.
A forward contract is a private agreement to deliver a commodity at a future date at a price fixed today; it removes price risk but carries counterparty risk and is hard to exit. A futures contract is a standardised forward traded on an exchange, with a clearing house that stands between buyer and seller, daily marking to market and margin deposits that control default risk. A hedger takes a futures position opposite to the cash-market position. A farmer who will sell grain at harvest is naturally long the crop and sells futures: a short hedge. A processor or exporter who must buy later is naturally short and buys futures: a long hedge. Hedging replaces price risk with basis risk. The basis is the local cash price minus the futures price. At the time the hedge is lifted, the net price received by a short hedger equals the futures price at which the hedge was placed plus the basis at the time it is lifted. If the basis turns out stronger (higher) than expected the short hedger gains; if weaker, the hedger loses. Basis reflects transport cost, local supply and demand, quality differences and storage cost, and it is usually far more predictable than the price itself. Futures markets are supervised by national regulators, such as the Commodity Futures Trading Commission in the United States, the China Securities Regulatory Commission and India's securities regulator, which absorbed its commodity regulator in 2015.
An option gives its buyer the right, but not the obligation, to buy (call) or sell (put) a futures contract at a fixed strike price, in return for a premium paid to the seller (writer). A farmer who buys a put option establishes a price floor while keeping the chance to benefit from higher prices. The expected minimum net price is the strike price minus the premium plus the expected basis. If prices fall below the strike, the put gains value and offsets the lower cash price; if prices rise, the farmer lets the put expire, loses only the premium and sells at the higher cash price. A buyer of grain can similarly buy a call to set a price ceiling. Options are therefore closer to insurance than futures are: the premium is the cost, and the upside is kept. Option sellers bear open-ended risk and must post margin. Combinations such as a fence (buying a put and selling a higher-strike call) reduce the net premium by giving up upside above the call strike.
A hedge ratio is the size of the futures position relative to the cash exposure. A naive one-to-one hedge is rarely optimal when the cash and futures prices are not perfectly correlated or when the quantity to be sold is itself uncertain. The minimum-variance hedge ratio is h* = rho times sigma of the cash price change divided by sigma of the futures price change, which is also the slope from regressing cash price changes on futures price changes; the number of contracts is h* times the cash quantity divided by the contract size. Because yield is uncertain when the hedge is placed, a farmer who hedges the full expected crop may be over-hedged after a drought, when prices are high and output is short, and must buy back futures at a loss with no crop to offset them. For this reason farmers commonly hedge only part of expected output before harvest, or combine revenue insurance with hedging so that the insurance covers the yield shortfall. Hedging also needs cash for margin calls, which makes it a liquidity decision as well as a price decision; small farmers usually gain access through aggregators such as farmer producer organisations.
Credit risk is the risk that a borrower does not repay principal or interest in full and on time. Agricultural lending carries extra credit risk for reasons already seen: income depends on weather and prices; losses are covariate, so many borrowers in a district default together after a drought; land titles and tenancy records may be incomplete, weakening collateral; and loan waivers and political interventions can damage the repayment culture, a form of moral hazard at the level of the credit system. Banks manage these risks through credit policy, appraisal, pricing, security, portfolio diversification, and linkage with insurance. Crop insurance improves the creditworthiness of farmers, which is why several countries, India among them, link crop insurance with seasonal crop loans; the claim can repay the loan after a crop failure instead of leaving an overdue account. Lenders also face concentration risk when a large share of a branch's portfolio is in one crop or one district, and liquidity risk when many borrowers seek rescheduling at once.
Agricultural finance texts, especially in South Asia, teach three sets of principles. The five Cs of credit are character (willingness to repay, judged from past record), capacity (ability to repay from income), capital (the borrower's own stake), collateral (security offered) and conditions (the economic and loan conditions). The three Rs of credit are returns (the additional income from the investment financed), repayment capacity (the ability to repay out of those returns after meeting family and other needs) and risk-bearing ability (the capacity to withstand losses). The seven Ps of farm credit are principle of productive purpose, personality, productivity, phased disbursement, proper utilisation, payment and protection. Appraisal of a crop loan uses the scale of finance per hectare multiplied by the area under each crop, with additions for post-harvest, household and maintenance needs; revolving farmer credit cards operate on the same logic. Term loans are appraised on the project's cash flow, using measures such as the debt service coverage ratio (cash available for debt service divided by the debt service due), net present value and internal rate of return, and the repayment schedule is matched to the income cycle of the enterprise, with a moratorium or grace period where income starts late.
A bank's expected loss on a loan is the product of the probability of default (PD), the loss given default (LGD, the share of exposure not recovered) and the exposure at default (EAD). The Basel framework uses these components to set capital held against unexpected loss. Loans are classified by repayment status. Under prudential norms that follow Basel guidance, a loan generally becomes non-performing (an NPA) when interest or principal is overdue for more than 90 days; some regulators adapt this for agricultural loans to the crop cycle (India's central bank, for example, rules that a loan for a short-duration crop becomes an NPA when overdue for two crop seasons and a loan for a long-duration crop when overdue for one crop season). NPAs are then classed as substandard, doubtful or loss according to how long they have been non-performing and whether security is impaired, and banks must provision against them. When natural calamities strike, regulators in many countries allow crop loans to be converted or rescheduled and fresh loans to be given, subject to conditions, so that affected borrowers are not pushed into default. Interest subvention on short-term crop loans (a concessional rate with an additional incentive for prompt repayment) and credit guarantee funds for some categories further reduce the effective cost and risk of lending; their current rates and limits should be checked each year.
A student should be able to classify any farm problem by risk source, compute expected value, coefficient of variation and certainty equivalent, apply the classical decision rules to a payoff matrix, and explain in one paragraph why index insurance reduces moral hazard but brings basis risk. The history of crop insurance should be known in order, together with the threshold-yield formula, typical premium caps and subsidy rules, and the stages of cover; the principles are the same whether the scheme in front of you is the US federal programme, India's PMFBY, China's subsidised insurance or another national scheme.
A field officer in a bank or insurer with five years' experience enrols loanee and non-loanee farmers before cut-off dates, checks that the sum insured matches the scale of finance and land records, explains the opt-out rule, and helps farmers report localised losses within the notice period with photographs and survey numbers. On the lending side the officer appraises revolving crop credit limits, watches repayment against the crop calendar so accounts do not slip into NPA, and uses claim settlements to regularise loans. An agribusiness manager at this level places simple short hedges and reads the local basis before contracting with farmers.
A senior manager with twenty-five years in a bank, insurer, state agriculture department or agribusiness firm decides portfolio questions: which crops and districts to cover under area-yield or weather-index schemes, what indemnity level fits each area, how to bid for clusters and structure reinsurance or cup-and-cap arrangements, and how much concentration in one district's crop loans the bank can bear. Such a manager designs relief responses after a drought with the state, builds CCE and remote-sensing verification systems that reduce disputes, and sets the firm's hedging policy, including limits, hedge ratios and margin funding.
A researcher with a doctorate and long field experience studies why uptake of insurance is low and what design would raise welfare: estimating basis risk from yield and weather panels, testing satellite and crop-model indices, running randomised trials of bundled insurance and credit, measuring the effect of insurance on input use and investment, and evaluating the fiscal cost and distribution of premium subsidies. Such a researcher advises governments on risk layering, sovereign risk pools and the interaction of insurance with price support and trade policy, and publishes the evidence that shapes the next revision of the schemes.
This manual gives principles and the stable rules of the large public schemes used as examples, not their season-by-season notifications: actuarial rates, notified crops, cut-off dates, regional risk-sharing models and subsidy shares are left out because they change often. Livestock insurance subsidy shares are not stated for the same reason. Canada, the European Union and Brazil run large programmes that are not covered in detail. Quantitative evidence on uptake, loss ratios and welfare effects is not reported. Tables, charts and worked term sheets will be added when the platform supports them.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions. Scheme parameters (PMFBY premium caps, subsidy sharing, interest subvention) are those in force at the last review and should be checked against current guidelines.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.
Last updated · How Laisha writes and checks its pages · Report a mistake
US$10 a month · one price worldwide · cancel any time
An AI operating system for agriculture: free guides, manuals and tools written for every country.