Managing Agricultural Commodity Risk Using Value-at-Risk

 

The emergence of agricultural insurance products and complex grain marketing contracts has complicated farm risk management decisions. With greater choices available, producers need to know how much risk can be reduced with each risk management product in order to choose a risk management portfolio that optimizes expected return relative to the risk they take. Without knowing the amount of risk reduction, or expected returns, producers may be duplicating risk reduction, and/or they may be choosing sub-optimal risk management portfolios.

 

So, how does a producer know how to optimize risk management given multiple contracts and changes in daily market prices? They can accomplish this by calculating their farm’s value-at-risk. What is farm value-at-risk? Value-at-risk (VaR) is a measure that quantifies the probability that a portfolio of assets will not decrease by more than a specified amount over a specified period. Value-at-risk is flexible in that it can incorporate basis, futures, and yield risks into a single revenue risk measure.

 

The concept is frequently used in the banking and business sectors to monitor if firms have adequate assets to cover potential losses-- ‘stress-testing’—worst case scenarios. This concept can be utilized in farm operations as well to monitor risk and evaluate risk reduction strategies relative to expected returns. It can also be used to determine the amount of risk an operation should tolerate given capital availability.

 

Producers’ should evaluate value-at-risk changes frequently as market risk and production risk changes occur, particularly throughout the production cycle. When production and market risk change, the value-at-risk measure will change, and action may be required through the marketing year to adjust the amount of risk being taken by producers.

 

How would value-at-risk be calculated for commodity production? The most direct approach is to couple historical yield and price data to build a copula. A copula is a model that best captures the historical relationship between variables that may be joint related. In this case, we are measuring if commodity price and yield at a local level that determine revenue are related. Copulas, also allow for uncertainty about the price and yield relationship given historical observances. For example, even though price and yield may have an inverse relationship historically that would represent a natural hedge, there may be varying levels of certainty to how often the relationship holds. With a copula model, simulations from the best fitting copula to historical prices and yields are transformed to real-time market and yield data to calculate real-time value-at-risk at any given point in time.