Monday, January 20, 2014

Surprise in crop data - a need for transparency

A review of selected USDA, WAOB, and NASS forecast and estimates in corn and soybeans developed by Scott Irwin of the University of Illinois is interesting reading for those who have gotten burned in grain markets around announcement dates. The author generally finds that forecasts of yield and production are unbiased, but for stocks to usage numbers there has been an increase in surprise volatility over the last few years.

What the author propose as a remedy for the dispersion of surprises around announcements is very simple. Tell the market how you calculate the estimates. Seems simple, yet there is a lot of confusion on how yields, production, acreage, and inventories are measured. You are not going to get good forecasts if you do not tell the market your methodology. Given the limited number of agriculture announcements  in grains, it seems like a simple fix to just provide transparency on how reports are generated. Of course, traders will have to read the how the methodology is employed.

No comments: