RANGE-BASED PRICE FORECASTS AND A TRADING STRATEGY FOR CORN AND SOYBEANS FUTURES

A TRADING STRATEGY FOR CORN AND SOYBEANS FUTURES

Abstract

The high volatility of food prices over the past decade has made price
forecasting increasingly important to policy makers and market participants
alike. Food price forecasts are undertaken on a regular basis by various
government agencies, and there is appreciable evidence that these forecasts
have implications for government food policies. It is noted that most existing
studies on food price forecasts are based on periodic averages or close-toclose price data. On the other hand, considerable literature has
accumulated over the past few years regarding the use of range-based
forecasting methods. One such method is based on the observation that
movements in the daily high and low prices are tied up in the long run by a
condition closely approximated by the daily price range. This paper applies
range-based method to forecasting the daily high and low prices of corn and
soybeans futures. It is found that this approach offers significant advantages
over the traditional ARIMA and random walk methods in terms
of out-ofsample forecast accuracy. Another attraction of this method is that it is very
easy to implement. While there are many avenues in which the high and low
price forecasts can be put to use, as one application this study develops a

trading strategy of corn and soybeans futures that makes use of these price
forecasts. This strategy generally yields very reasonable profits, and its
success depends in part on the accuracy of the price forecasts produced by
the underlying model.