Futures hedging on both procurement risk and sales risk under correlated prices and demand

Abstract

The profitability of a manufacturer could be largely affected by underlying uncertainties embedded in the fast-changing business environment. Random factors, such as input material price at the procurement end or output product price and demand at the sales end, might produce significant risks. Effective financial hedging therefore needs to be taken to mitigate these risk exposures. Although it is common to use commodity futures to control the risks at either end separately, little has been done on the hedging of these risk exposures in an integrated manner. Therefore, this study aims to develop a planning approach that performs financial hedging on both the procurement risk and the sales risk in a joint manner.

This planning approach is based on a framework that has a risk-averse commodity processor that procures input commodity and sells output commodity in the spot market, while hedging the procurement risk and sales risk through trading futures contracts in the commodity markets. Both the input and output commodities futures are used for the hedging. A both-end-hedging model is developed to quantitatively evaluate the approach. The evaluation is based on an objective function that considers both profit maximisation and risk mitigation. Decisions on spot procurement, input futures hedging position, and output futures hedging position are optimised simultaneously. As the input commodity is the main production material for the output commodity, positive correlation between the input material price and the output product price is considered. The customer demand is considered negatively correlated with the output product price.

An ethanol plant using corn as the main input material is employed as an example to implement the proposed model. The model is represented as a stochastic program, and the Gibson-Schwartz two-factor model is employed to describe the stochastic commodity prices. Historical commodity price data are used to estimate the parameters for the two-factor model with state-space form and Kalman filter. By generating various scenarios representing evolving prices and the random customer demand, the stochastic program could be solved using linear programming algorithms under its deterministic equivalent.

Numerical experiments are carried out to demonstrate the benefit that could be gained from applying the both-end-hedging approach proposed in this study. Comparing with traditional no-hedging model or single-end-hedging models, the improvement obtained from the proposed model is found to be significant. The effectiveness of the model is further tested in various price trend and price correlation, demand elasticity and volatility, and risk attitude of the decision maker. It is found that the proposed approach is robust in these various circumstances, and the approach is especially effective when the price trend is uncertain and when the decision maker has a strong risk-averse attitude.