17

Primary submission: 07.07.2012 | Final acceptance: 17.04.2013

Price Risk and Risk Management

in Agriculture
Udo Broll1, Peter Welzel2, Kit Pong Wong3

ABSTRACT This note studies the risk-management decisions of arisk-averse farmer. The farmer faces multiple
sources of price uncertainty. He sells commodities to two markets at two prices, but only one of
these markets has afutures market. We show that the farmers optimal commodity futures market
position, i.e., a cross-hedge strategy, is actually an over-hedge, a full-hedge, or an under-hedge
strategy, depending on whether the two prices are strongly positively correlated, uncorrelated, or
negatively correlated, respectively.

KEY WORDS: agricultural price risk; risk management; commodity futures; correlation; cross-hedge

JEL Classification: D73, D78, D83

1
Technische Universitt Dresden, Germany; 2University of Augsburg, Germany; 3University of Hong Kong, Hong Kong

Introduction cereal prices more than doubled in real terms before

In todays economy, farmers and agribusiness firms falling sharply during the second half of 2008. From
face ahigh degree of risk because of certain new fac- 2010 through 2012, prices again increased rapidly and
tors, such as the increased price volatility of inputs and generated fears that the global market was entering
outputs, climate change, international trade restric- a phase of sustained commodity price volatility. Al-
tions, new and more stringent food safety standards, though short-term coping strategies are important
and greater public concern about the environment, to particularly in economies that are negatively affected
name only a few. These developments coincide with by rising food prices (Prakash, 2011; Prakash, & Gil-
major changes in the fundamentals of the agriculture bert, 2011) an important part of the economic policy
market (UNCTAD, 2011). A critical issue in agricul- debate addresses the question of whether these recent
ture is adjusting supply and price risk. The role of price fluctuations are the result of speculation in the
economic risk is particularly important for pricing in futures markets or whether they simply reflect the un-
commodity futures markets. In developed economies, derlying economic fundamentals.
farmers have access to commodity futures markets in However, futures markets in the real world are far
which they can hedge the price risk. from complete. For example, less-developed countries
Since 2006, there have been significant price fluc- and economies in transition places in which risk-
tuations in major food commodities on international sharing markets are embryonic and markets are heav-
markets. Between early 2006 and mid-2008, grain and ily controlled are unlikely to have futures markets for
agricultural goods. Farmers that are exposed to com-
Correspondence concerning this article should be addressed to: modity price uncertainty must therefore rely on com-
Udo Broll, Technische Universitt Dresden, 01062 Dresden, Ger- modity futures contracts on related goods to indirectly
many, e-mail: [email protected] hedge against their price-risk exposure (Allen and Lu-

www.ce.vizja.pl Vizja Press&IT

18 Vol.7 Issue 2 2013 17-20 Udo Broll, Peter Welzel, Kit Pong Wong

eck, 2003; UNCTAD, 2011). Such a risk-management hedging instruments for the random price of the sec-
technique is referred to as cross hedging (see, e.g., ond good, ~p2 .
Anderson and Danthine, 1981; Broll, Wong, & Zilcha, The farmers profits are given by
1999; Chang, & Wong, 2003; Frechette, 2000; Haigh, &
~
Holt, 2000; Hudson, 2007). An important part of the = ~
p1 x1 + ~
p2 x2 + ( p1f ~
p1 )h, (1)
policy debate in agriculture addresses the question of
whether commodity price volatility results from specu- where h is the number of futures contracts sold (pur-
lation in the futures market or whether it simply reflects chased, if negative). The farmer is risk averse and possess-
the underlying economic fundamentals (e.g., Gilbert, es avon Neumann-Morgenstern utility function, U ( ) ,
2010; Food and Agriculture Organization [FAO], 2011; which is defined over its profits, , with U ( ) > 0 and
Ziegler, 2012). In this context, there is growing concern U ( ) < 0 . For agiven production, the farmers decision
that financial investment into commodity derivatives problem is to choose his futures market position, h , to
may have become an independent source of price be- maximize the expected utility of its profits
havior and is causing the recent volatility.
~
The purpose of this note is to provide theoreti- max E[U ( )] (2)
h
cal insights into optimal cross-hedging strategies for
farmers contracts. To that end, we consider a risk- where E() is the expectation operator. The first-order
averse farmer who sells his output to two markets; condition for program (2) is given by
however, only one of these markets has afutures mar-
~
ket to which the farmer has access. We show that the E[U ( *) ( p1f ~
p1 )] = 0, (3)
farmers optimal futures position hinges on the bivari-
ate dependence of the random commodity prices in where an asterisk (* ) indicates an optimal level. Given
the separate markets. To derive concrete results, we the assumed properties of U ( ), the second-order
propose the concepts of strong correlation. We show condition for program (2) is satisfied.
that the farmer can find an optimal solution through
over-hedging, full-hedging, or under-hedging strate- Optimal Risk Management Policy
gies, depending on whether the two random prices are To examine the farmers optimal futures position, h*,
strongly positively correlated, uncorrelated, or nega- we write equation (3) as
tively uncorrelated, respectively.
~ ~
The rest of this note is organized as follows. The E[U ( * )][ p1f E( ~
p1 )] cov[U ( * ), e~1 ] = 0, (4)
next section develops the model of a farmer facing
price risk and cross-hedging opportunities, and the where cov(,) is the covariance operator. (Please note
section following that describes the farmers optimal that for any two random variables, ~ x and ~ y , we have
commodity futures hedging strategy. The final section ~ ~ ~ ~ ~ ~
cov( x , y ) = E ( x y ) E ( x ) E ( y ) .) Evaluating the left-
concludes. hand side of equation (4) at h* = x1 yields

E{U [ p1f x1 + ~
p2 x2)]}[ p1f E( ~
p1 )] + cov{U [ p1f x1 + ~
p2 x2 ], ~
p1}
The Model
+~
{U [ p1f x1 two
We consider afarmer who Eproduces p2 final p1f E( ~
x2)]}[outputs, p1 )] + cov{U [ p1f x1 + ~
p2 x2 ], ~
p1} . (5)
indexed by i = 1 and 2. Let xi and pi be the amount
of outputs and the per-unit selling commodity price If the above expression is positive, zero, or negative,
~
in market i , where i = 1 and 2. Profit risk comes from equation (4) and the strict concavity of E[U ( )] im-
two sources, ~p1 and ~p2, which denote the pricing of ply that h* is greater than, equal to, or less than x1 ,
the random goods in markets 1 and 2, respectively (the respectively.
tilde ( ) denotes arandom variable). Cross hedging is It is impossible to determine the sign of expression
modeled by allowing the farmer to trade infinitely di- (5) without imposing concepts of bivariate depen-
visible futures contracts in the market for the first good dence upon ~p1 and ~p2 . Thus, we offer the following
at the forward rate, denoted by p1f . There are no direct definition.

CONTEMPORARY ECONOMICS DOI: 10.5709/ce.1897-9254.79

 Price Risk and Risk Management in Agriculture 19

Definition: The random variable, ~ x , is said to be If ~p1 and ~p2 are strongly positively (negatively)
strongly positively correlated, uncorrelated, or negatively correlated, we have cov( ~p1 , ~p2 ) > (<) 0 . According
correlated to the random variable, ~y , if, and only if, to equation (7), the farmer finds it optimal to set
cov[~ x, f (~y )] is positive, zero, or negative, respectively, h > (<) x1 to reduce the variability of its profits. When
for all strictly increasing functions, f () . p1f > (<) E( ~ p1 ) , a speculative motivation may induce
This definition is motivated by similarly ordered the farmer to sell (purchase) forward contracts. Thus,
random variables in Hardy, Littlewood, and Plya the over-hedging (under-hedging) incentive for risk
(1934) and Ingersoll (1987). An example of strongly minimization is reinforced by speculative motivations
correlated random variables is the following linear when p1f () E( ~p1 ) .
specification: ~ p2 = + ~p1 + ~ , where and are If ~p1 and ~p2 are strongly uncorrelated, we have
scalars, and ~ is a zero-mean random variable inde- cov( ~
p1 , ~
p2 ) = 0. Thus, equation (7) implies that h = x1
pendent of ~p1. This linear specification is widely used minimizes the variability of the farmers domestic
in the hedging literature. profits. The farmer deviates from this full hedge only
Result 1: Given that the farmer is allowed to trade when p1f E( ~p1 ). If p1f > (<) E( ~p1 ) , speculative mo-
commodity futures contracts, if ~p1 and ~p2 are strongly tivations induce the farmer to sell (purchase) forward
uncorrelated, then the farmers optimal future posi- contracts, thereby making the over-hedging (under-
tion, h*, is greater than, equal to, or less than x1 , de- hedging) strategy optimal.
pending on whether p1f is greater than, equal to, or
less than E( ~ p1 ) , respectively. If ~ p1 and ~
p2 are strongly Conclusion
positively (negatively) correlated, then the farmers op- In todays economy, farmers and agribusiness firms
timal future position, h*, is greater (less) than x1 , when face ahigh degree of risk because of certain new fac-
p1f () E( ~ p1 ) . tors, such as greater price volatility for inputs and out-
puts, climate change, international trade restrictions,
Proof: If ~p1 and ~p2 are strongly uncorrelated, then and new and more stringent food safety standards.
the covariance term of expression (5) vanishes. Thus, These developments coincide with major changes in
expression (5) is positive, zero, or negative, depending the fundamentals of the agriculture market. Acritical
on whether p1f is greater than, equal to, or less than issue in agriculture is price risk.
E( ~
p1 ) , respectively, which implies that h* is greater An important topic in the policy debate in agri-
than, equal to, or less than x1 , respectively. culture concerns the question of whether the volatil-
If ~p1 and ~p2 are strongly positively (negatively) ity in commodity prices results from speculation in
correlated, then the covariance term of expression (5) the futures market or whether this volatility simply
is positive (negative). Thus, expression (5) is positive reflects the underlying economic fundamentals. In
(negative) when p1f () E( ~p1 ) , such that h* > (<) x1 . this context, there is growing concern that financial
The logic of Result 1 may be shown as follows. investment into commodity derivatives based on the
Considering the variance on both sides of equation replication of futures indices has become an indepen-
(1), we have dent source of price behavior and is causing recent
patterns of price volatility in these markets.
~ ~ ~ ~ ~
var( ) = var( p1 ) (x1 h) + var( p2 ) x2 + 2 cov( p1 , p2 ) (x1 h) x2 In this note, we have examined the optimal risk-
2 2

management decisions of a risk-averse farmer who

) (x1 h) 2 + var( ~
p2 ) x22 + 2 cov( ~ p1 , ~
p2 ) (x1 h) x2 , (6) is facing multiple sources of commodity price uncer-
tainty. The farmer sells commodities to two markets,
where var () is the variance operator. Partially differ- but only one of these has afutures market. We have
entiating equation (6) with respect to h and evaluating shown that the farmers optimal forward position is
the resulting derivative at h = x1 yields an over-hedge, afull-hedge, or an under-hedge strat-
egy, depending on whether the two random com-
~ ~ ~
var( ) |h = x = 2 cov( p1 , p2 ) x2 . (7) modity prices are strongly positively correlated, un-
h 1
correlated, or negatively correlated, respectively.

www.ce.vizja.pl Vizja Press&IT

20 Vol.7 Issue 2 2013 17-20 Udo Broll, Peter Welzel, Kit Pong Wong

References Ziegler, J. (2012). Wir lassen Sie verhungern. Muenchen:

Allen, D. W., Lueck, D. (2003). The Nature of the Farm: Bertelsmann.
Contracts, Risk and Organizsation in Agriculture.
Cambridge, MA: MIT Press.
Anderson, R. W., Danthine, J.-P. (1981). Cross Hedging.
Journal of Political Economy, 89(6), 1182-1196.
Broll, U., Wong, K. P., & Zilcha, I. (1999). Multiple Cur-
rencies and Hedging. Economica, 66(264), 421-432.
Chang, E. C., Wong, K. P. (2003). Cross-Hedging with
Currency Options and Futures. Journal of Finan-
cial and Quantitative Analysis, 38(3), 555-574.
Food and Agriculture Organization (2011, June 2).
Price Volatility in Food and Agricultural Markets:
Policy Responses. Retrieved from: http://www.
globalbioenergy.org/uploads/media/1106_FAO_
et_al._-_Price_volatility_in_food_and_agricul-
tural_markets_policy_responses_.pdf
Frechette, D. L. (2000). The Demand for Hedging and
the Value of Hedging Opportunities. American
Journal of Agricultural Economics, 82(4), 897-907.
Gilbert, C. L. (2010). How to Understand High Food
Prices. Journal of Agricultural Economics, 61(2),
398-425.
Haigh M. S. Holt, M. T. (2000). Hedging Multiple
Price Uncertainty in International Grain Trade.
American Agricultural Journal of Economics,
82(4), 881-896.
Hardy, G. H., Littlewood, J. E., Plya, G. (1934). In-
equalities. London, UK: Cambridge University
Press.
Hudson, D. (2007). Agricultural Markets and Prices.
Malden, MA: Blackwell.
Ingersoll, J. E. Jr. (1987). Theory of Financial Decision
Making. Savage, MD: Rowman and Littlefield.
Prakash, A. (2011). Why Volatility Matters. In A.
Prakash (Ed.), Safeguarding Food Security in Vola-
tile Global Markets (pp. 1-24). Rome: FAO.
Prakash, A., Gilbert, C. L. (2011). Rising Vulnerability
in the Global Food System: Beyond Market Fun-
damentals. In A. Prakash (Ed.), Safeguarding Food
Security in Volatile Global Markets (pp. 45-66),
Rome: FAO.
UNCTAD. (2011). Price Information in Financial-
ized Commodity Markets: The Role of Informa-
tion. New York, NY: United Nations Publica-
tion. Retrieved from http://unctad.org/en/Docs/
gds20111_en.pdf

CONTEMPORARY ECONOMICS DOI: 10.5709/ce.1897-9254.79