Characterizing the hedging policies of commodity price‐sensitive corporations

• Date: 01 August 2020
• DOI: http://doi.org/10.1002/fut.22072
• Author: Stéphane Goutte,Ehud I. Ronn,Raphaël H. Boroumand
• Published date: 01 August 2020

J Futures Markets. 2020;40:1264–1281.wileyonlinelibrary.com/journal/fut1264

|

© 2019 Wiley Periodicals LLC

Received: 13 October 2019

|

Accepted: 22 October 2019

DOI: 10.1002/fut.22072

RESEARCH ARTICLE

Characterizing the hedging policies of commodity

price‐sensitive corporations

Raphaël H. Boroumand

1

|

Stéphane Goutte

2

|

Ehud I. Ronn

3

1

Department of Applied Economics, Paris

School of Business, Paris, France

2

Department of Economics and Business

Information, Université Paris 8,

Saint‐Denis, France

3

Department of Finance, McCombs

School of Business, University of Texas at

Austin, Austin, Texas

Correspondence

Ehud I. Ronn, Department of Finance,

McCombs School of Business, University

of Texas at Austin, 2100 Speedway Stop

B6600, Austin, TX 78712‐1276.

Email: eronn@mail.utexas.edu

Abstract

Many corporations face price and quantity uncertainty in commodities for

which existing futures and options contracts permit corporations to hedge their

risks. Finance theory has demonstrated frictions in capital markets are

equivalent to risk‐averse decision‐making: Taking prices and volatilities as

exogenous, decision‐makers make optimal hedge decisions as a trade‐off

between risk and return. In modeling risk aversion, we use mean‐variance and

mean‐value at risk‐utility functions. With options quantified as delta‐equivalent

futures, using data from the Commodity Futures Trading Commission and gold

companies, we document empirically corporations’hedge ratios appear to

respond to changing prices and volatilities in accordance with utility‐function

prescriptions.

KEYWORDS

optimal hedge ratios

JEL CLASSIFICATION

G11, G13, G31

1

|

INTRODUCTION

Many, if not most, corporations in the developed world face price‐and‐quantity uncertainty in commodities for

which there are traded assets—futures and options contracts—which permit these corporations to hedge away

some or all of the combined price and quantity risk to which they are exposed. Whether these corporations

are long the physical commodity, or short, they can take steps to minimize the shock of adverse price impact on

their operations.

The purpose of this paper is to determine whether corporations are following an optimal hedge strategy for this

combined price‐and‐quantity risk at the corporate (not trading‐desk) level. The inputs into the optimizing policy include

measures of price and quantity levels and risk, and the bias if any in forward prices relative to “expected prices”(i.e., a

non‐zero market price of commodity risk).

For a corporation based on gold extraction and marketing—and, therefore, long the physical gold output—Kolos and

Ronn (2008) demonstrated as corporate decision‐makers’risk‐aversion increases, the optimal hedge begins with long

position inputs, eventually reaching 100% of expected quantity. Their result obtains under the assumption of a mean‐

value at risk objective function, and is presented below

1

. For even greater risk aversion, the optimal hedge then

transitions to short futures contracts, eventually culminating in a 100% futures position.

1

See Section 2.2.

To date, the literature on hedging has addressed several issues. To begin, financial economics has been much

concerned with why firms hedge even when hedging activities are costly

2

. In the current work weassume the corporation

has made an affirmative decision to manage its risk, without requiring us to specify the firm’smotivation for doing so.

Second, the issue of the hedge implementation has been picked up by financial engineering, which deals with the

mechanics of using financial contracts (i.e., derivatives) to implement the hedge. In this context, an important

distinction has been made with respect to the use of futures versus options: Whereas futures are costless‐to‐enter, they

eliminate both downside risk as well as upside potential. In contrast, options are costly but they preserve upside

potential while eliminating downside risk.

A third issue to which the analysis of this paper refers pertains to the issue of causality: That is, are futures and

option prices the consequence of so‐called “hedging pressure”? The articles in this category include de Roon, Nijman,

and Veld (2000), Bessembinder (1992), and Hirshleifer (1990) and those of the financialization literature such as

Singleton (2012), Cheng and Xiong (2013), and Goldstein and Yang (2017). We hold the countervailing view. In our

view, the role of financial players in the commodity markets has been inappropriately exaggerated: We propose the

hypothesis that futures and option prices—and their associated risk premiums—are determined by economy‐wide

factors reflecting the geopolitical and economic realities of the respective markets, and the compensation for bearing

risk in the economy. Accordingly, the tests we propose examine the empirical response of commercial agents’hedge

ratios to changes in the prices of futures and options.

In light of this critical assumption, a fourth issue addresses the quantitative objective function to be utilized in

optimizing the amount and instrument to be used in hedging. Much of the literature here has focused on variance‐

minimization or mean‐variance efficiency, but that is a particularly undesirable objective function in the presence of

options’nonlinear payoffs. In fact it was shown by Lapan, Moschini, and Hanson (1991) that under mean‐variance

efficiency the optimal hedging strategy does not involve options. In contrast, in this paper we consider objective functions

that explicitly utilize nonlinear instruments to capture the notion of downside risk aversion together with upside capture.

Finally, the issue of the biasedness/unbiasedness of futures contracts is inescapable in this context. The analyst, and

indeed the decision‐maker, must take a position on whether they believe futures prices are biased or unbiased

predictors of expected prices in the real (not the risk‐neutral) world: While our model in no way infringes on the fair‐

valuation of future contracts, that does not imply that the risk premium on the futures contracts is necessarily zero.

A large body of literature related to hedging of both quantity and price exists in relation to risk management in

commodity markets: See the review by Tomek and Peterson (2001). The general principle of portfolio theory is well‐

known per Huang and Litzenberger (1988) and Artzner, Delbaen, and Heath (1999): namely, the decision‐maker selects

the composition of the firm’s portfolio to maximize expected utility. The early literature on using futures markets

considered simple portfolios consisting of a commodity inventory and a short position in futures contracts Johnson

(1960). An optimal hedge was derived assuming that the quantity to be hedged was given exogenously, that only output

price risk was important, and hence that the decision was about the optimal size of the futures position. The objective

function maximized gross returns, subject to a risk constraint based on the variance. The resulting optimal futures

position is identical to the position which minimizes the variance of returns if the futures price is an unbiased forecast

of the terminal price at the completion of the hedge (or if the hedger is extremely risk‐averse). If the problem is specified

as a joint decision about the quantity and the size of the futures position, the optimum still reduces to a ratio of futures

to cash positions which minimizes the variance of returns and is obtained as the ratio of a covariance of futures and

cash prices to a variance of futures prices Kahl (1983). Other models have been developed to consider optimal positions

in futures that take account of price and yield risk jointly Newbery (1983). In multiperiod settings, the important

questions are how to hedge when only limited funding is available to support the hedging program or when mark‐to‐

market gives rise to additional risks. These questions are addressed in Emmer, Kuppelberg and Korn (2001), Lagcher

and Leobacher (2003), Lien and Li (2003), and Fehle and Tsyplakov (2004).

It was a natural extension to consider positions in option contracts as part of the portfolio. If the model incorporates

options markets in a mean‐variance framework, and if the options premiums and futures prices are unbiased, then

options turn out to be redundant hedging instruments; the optimal hedging strategy involves only futures Lapan,

Moschini and Hanson (1991). This result was obtained assuming normally distributed prices (which allow for negative

prices). When the distribution of returns is skewed to the right, options contracts enter optimal portfolios Vercammen

2

The reasons frequently cited include: Decreasing a firm’s expected tax payments; Reducing the costs of financial distress; Allowing firms to better plan for their future capital needs and reduce their

need to gain access to outside capital markets; Improving the design of management compensation contracts and allowing firms to evaluate their top executives more accurately; and Improving the

quality of the decisions made. See Grinblatt and Titman (2001), as well as Smith and Stulz (1985) and Froot, Scharfstein, and Stein (1992).

BOROUMAND ET AL.

|

1265