Dealing with downside risk in a multi‐commodity setting: A case for a “Texas hedge”?

Abstract

The authors thank an anonymous reviewer and the editor, Robert I. Webb, for their helpful comments, which improved the quality of the manuscript. The authors are grateful to David Anderson and Steven Klose for their help with locating data and for useful discussion of the feedlot operation model. The authors acknowledge financial support from Hatch projects RI-9262 and TEX09258. GABRIEL J. POWER* DMITRY VEDENOVThis study analyzes the problem of multi-commodity hedging from the downside risk perspective. The lower partial moments (LPM 2 )-minimizing hedge ratios for the stylized hedging problem of a typical Texas panhandle feedlot operator are calculated and compared with hedge ratios implied by the conventional minimumvariance (MV) criterion. A kernel copula is used to model the joint distributions of cash and futures prices for commodities included in the model. The results are consistent with the findings in the single-commodity case in that the MV approach leads to over-hedging relative to the LPM 2 -based hedge. An interesting and somewhat unexpected result is that minimization of a downside risk criterion in a multicommodity setting may lead to a "Texas hedge" (i.e. speculation) being an optimal strategy for at least one commodity.

Optimal hedge ratio estimation and hedge effectiveness with multivariate skew distributions

“…This conclusion has two major implications. First, there are more essential information than just the first two orders of moments (Power and Vedenov, 2010). It is advisable to consider the first four orders of moments together to estimate MVHR, while the traditional methods aim only at estimating mean and variance.…”

Forecasting Hourly Peak Call Volume for a Rural Electric Cooperative Call Center

“…Copulas have a dependence function that is separate from the marginal distributions (Cherubini et al , ). Past research has used copulas in a variety of situations such as optimal hedge ratios (Power and Vedenov, ), revenue insurance where price and yield are correlated and modeled separately (Tejada and Goodwin, ) and correlated crop yields (Vedenov, ). A centralized call center is expected to generate efficiencies in managing peak call volume because severe weather events that generate the majority of power outage calls do not occur simultaneously across a wide geographical region.…”