Forecasting Corn Futures Volatility in the Presence of Long Memory, Seasonality and Structural Change

Price volatility in the corn market has changed considerably globalization and stronger linkages to the energy complex. Using data from January 1989 through December 2009, we estimate and forecast the volatility in the corn market using futures daily prices. Estimates in a Fractional Integrated GARCH framework identify the importance of long memory, seasonality, and structural change. Recursively generated forecasts for up to 40-day horizons starting in January 2005 highlight the importance of seasonality, and long memory specifications which perform well at more distant horizons particularly with rising volatility. The forecast benefits of allowing for structural change in an adaptive framework are more difficult to identify except at more distant horizons after a large downturn in volatility.