A Decision Model to Determine Class III Milk Hedging Opportunities
Travis Holt, Spencer, Wis., defended his thesis, “A Decision Model to Determine Class III Milk Hedging Opportunities,” on Monday, August 6, 2007. Holt is an Investment Officer for Citizens State Bank in Loyal, Wis. He graduated from Kansas State University in December with a Master’s in Agribusiness (MAB).
Milk has become one of the largest agricultural commodities, as measured by gross sales, produced in the United States. Since the federal government began to loosen its control over dairy prices in the early 1980’s, farm level milk prices have seen dramatic increases in volatility. More dairy farmers are carrying debt to off-set further shrinking profit margins. Because of the greater leverage in the industry and reduced government support, many producers desire to find mechanisms by which to reduce price risk. In his thesis, Holt analyzed strategies to reduce risk and developed a model-based hedging strategy for Class III milk.
“The decision models developed and tested in this thesis not only reduced price volatility, they also increased the mean Class III price obtained as compared to a ‘cash-only’ strategy. While the decision models were successful in-sample, unfortunately their out-of-sample testing proved to be considerably less successful as all of the model-based strategies underperformed the cash market,” Holt said.
Because of the newness of dairy-related futures contracts, there has not a lot of studies completed on trading strategies, which makes Holt’s findings important. While the decision models were not successful strategies to reduce risk, that finding itself is a key finding, said Dr. Kevin Dhuyvetter, Agricultural Economics professor and Holt’s thesis advisor.
“Using in-samples and out-of-samples, you learn you can’t out-guess the market. There is no magic model,” Dhuyvetter said. “We learned using a statistical model does not always mean it is the truth and that is important to remember.”
The full thesis publication can be found online on Kansas State University’s Research Exchange at http://hdl.handle.net/2097/391.
