July, 2011
The task of this essay is to examine price discovery, volatility spillovers and impacts of speculation in the dairy sector.
I have developed a method that allows for the calculation of implied cheese futures prices for a period before cheese
futures actually started trading. Evidence for periods when cheese futures did trade suggests that utilized approximation
methods perform very well. Examining the time series properties of cheese cash and implied futures price I find that the
unit root hypothesis is strongly rejected for cash prices, while unit roots cannot be rejected for nearby futures prices
in the framework that carefully controls for rollovers. To explain this result, I built a model that illustrates the time
series properties of the nearby futures price series for a futures contract written on a second-order stationary cash series
and identified the mean-reverting nonlinear dynamics that will occur at rollovers. Given the time series properties of the
cash and futures series I propose an error-correction-type model using spreads between cash and the second nearby futures
instead of the cointegration vector. To account for volatility dynamics I employ the GARCH-BEKK structure. I find that the
flow of information in the mean model is predominantly from futures to cash, while volatility spillovers are bidirectional.
It is possible that cash prices that include unfilled bid/offers react differently to increases in volatility in futures
prices than sales cash prices, though this result may not be robust and further research is needed to identify if liquidity
in the cash market is reduced with increase in conditional volatility of the futures price. I propose an extension of the
BEKK variance model that I refer to as GARCH-MEX. That model does not restrict the sign of the additional regressors on the
conditional variances, and can easily insure positive-definiteness of the conditional covariance matrix. Utilizing that model
to evaluate the impact of speculation I find strong evidence against the hypothesis that excessive speculation is increasing
the conditional variance of futures prices. If anything, speculation may in fact be inadequate, and further research with
daily speculative positions and high-frequency futures prices is needed to identify the effect of increased speculation on
realized volatility of futures prices, bid-ask spread and magnitude of slippage.
July, 2011
Options on agricultural futures are popular financial instruments used for agricultural price
risk management and to speculate on future price movements. Poor performance of Black's
classical option pricing model has stimulated many researchers to introduce pricing models
that are more consistent with observed option premiums. However, most models are motivated
solely from the standpoint of the time series properties of futures prices and need for
improvements in forecasting and hedging performance. In this paper I propose a novel arbitrage
pricing model motivated from the economic theory of optimal storage, and consistent with
implications of plant physiology on the importance of weather stress. I introduce a pricing
model for options on futures based on a Generalized Lambda Distribution (GLD) that allows
greater flexibility in higher moments of the expected terminal distribution of futures price.
I use times and sales data for corn futures and options for the period 1995-2009 to estimate
the implied skewness parameter separately for each trading day. An economic explanation is then
presented for inter-year variations in implied skewness based on the theory of storage. After
controlling for changes in planned acreage, I find a statistically significant negative
relationship between ending stocks-to-use and implied skewness, as predicted by the theory
of storage. Furthermore, intra-year dynamics of implied skewness reflect the fact that
resolution of uncertainty in corn supply is resolved between late June and middle of October,
i.e. during corn growth phases that encompass corn silking through grain maturity. Impacts
of storage and weather on the distribution of terminal futures price jointly explain upward
sloping implied volatility curves.
This paper begins with an account of the evolution of dairy futures markets in the
U.S. A partially overlapping time series (POTS) model is then estimated to examine price
behavior in simultaneously traded Class III milk futures contracts. POTS is a latent factor model
that measures price changes in futures as a linear combination of a common factor, i.e.
information affecting all traded contracts, and an idiosyncratic term specific to each contract.
This paper contributes to the literature by showing that the importance of a common factor in
price volatility determination for dairy is related to capital production factors, i.e. the dairy herd.
Finally, it is shown that Class III milk futures volatility decreases as contracts approach maturity.
This "Inverse Samuelson Effect" comes from the fact that Class III milk futures are cash-settled
contracts that settle against a formula-based price. The results suggest that the importance of the
common factor declines as one approaches maturity, implying that individual contract months
are poor substitutes in hedging a specific month's cash price risk. Thus, despite low liquidity in
the market, it is useful to have 12 contract delivery months per year.
July, 2011
Download PDF
The econometric analysis contained in this study adopts the modeling principle currently used by USDA for
policy analysis whereby milk production is modeled as a product of dairy herd size and yield per cow,
stochastic elements that are modeled separately (USDA 2007). We account for herd size dynamics along
the lines of models proposed by Chavas and Klemme (1986) and Schmitz (1997). We contribute to the
literature by introducing a mixed frequency approach that allows us to model yield using quarterly data,
while dairy herd and replacement heifer pool is modeled using annual frequency. That allows us to model
the yield as a trend-stationary process with the economic environment inducing both short-term shocks and
impacting the speed of reversion to trend yields. In addition, we design and implement dynamic residual-based
bootstrap technique that can be used in testing for changes in non-marginal simulated long-run supply
responsiveness. We obtain strong in-sample predictive power and very high significance of key economic and
herd structure variables. Several conclusions emerge from our study. First, given the large difference between
short-run and long-run responses of production to price changes, policy makers should consider more than short
run responses to future policy changes. What may in the short-run seem like a minor impact that does not
disturb market equilibrium can indeed lead to large production surpluses after more time has passed and dairy
herd size has had adequate time to adjust to the new policy environment. Second, despite dramatic yield improving
technological change, improved genetics and the increasing importance of large farms, all of which we would
expect to increase the milk supply price responsiveness, we find a declining trend in long-run supply responsiveness
from 1975 through 2005. If such decline were to persist or continue that would be a major cause for worry, as
ever larger price swings would be needed to quickly equilibrate the market in face of demand shocks. However,
we find that milk supply is getting more responsive in recent several years both to milk and feed prices. We
recommend extending this analysis using micro-level data to examine farm-level behavior in face of price-swings.
Increasing responsiveness to feed prices further justifies focusing the next generation of dairy policy toolbox
on managing dairy profit margins rather than just revenue streams. Further research needs to be done to develop
a partial equilibrium model of the U.S. dairy sector based on insights on structural characteristics of the
production presented in this article. We believe that when combined with a model of the demand for dairy products
our work has a potential to improve reliability of long-run projections of the U.S. milk production.
July, 2011
Download PDF