The objective of the study is to identify risk efficient marketing strategies for a representative Texas cotton farm. To accomplish this, we rank the net revenue distributions of alternative marketing strategies to discover which is dominant. A Monte-Carlo simulation model is used to make the comparison among 1) forward pricing with put options, 2) selling the crop at harvest using spot price cash sale, 3) using the naïve strategy (selling a percentage of the crop each month for a year), and 4) putting the crop in the CCC loan program for deferral of sale. The loan program will include a comparison of the current loan program and a flexible loan program that will work by using an Olympic average.
Historical data of cotton futures settlement prices, historical daily put option premiums, historical local prices from county data from one specific county in West Texas, and historical adjusted world price (AWP) will be used in the simulation model. Each strategy will use these price data as the bases for the forecasted prices used in the model. The county yield data will be taken from multiple farms within one county from Crop Insurance records. A cross section of the data among many farms over several years will provide a significant amount of data points to build a representative model. The data will be separated into dryland and irrigated farms to essentially have two models to represent the varied risk associate with this variable.
A multivariate empirical distribution is placed on the variables to account for their correlation. Each strategy is ranked based in terms of certainty equivalents as the measurement of risk aversion over a defined range. This approach, stochastic efficiency with respect to a function (SERF), reveals the strategy that is most preferred for the representative farm over the historical data.
Audio File