Do Big Corn Crops Get Bigger? Implications for 2009

October-13-09

The old market adage is that big crops keep getting bigger as USDA makes monthly estimates starting in August and extending to their final estimate in January. So is this true? What does it mean for this year?

I will use the last 30 years of U.S. corn and soybean yields to examine the first question. This includes monthly estimates of yields for August, September, October, November and January (USDA does not make an estimate of yields in December). The 30 years are from 1979 to 2008.

If big crops tend to get bigger, then small crops should also tend to get smaller as USDA makes their monthly yield estimates from August to January. From these last 30 years, how often does a larger (smaller) yield estimates in the early months relate to an even larger (smaller) estimate in subsequent months. I looked at the change in yield from September to October and how that related to a change in yield from October to the estimate in November and the final in January.

For corn the odds were 73% that an increase (decrease) in the October yield estimate was related to an increase (decrease) in the November estimate. For the final yield in January 76% of the time higher (lower) yields in the October report related to higher (lower) yields in the final January report.

When the October report had at least a 2 bushel higher, or at least a 2 bushel lower, yield than the September report, then the odds were 92% that the final yield estimate in January would go in the direction of the October estimate.

The recently released corn estimate on October 9 was 164.2 bushels per acre compared to 161.9 bushels per acre in September, an increase of 2.3 bushels. Large yield adjustments in the October report tend to be related to further large yield adjustments in the same direction for subsequent reports, but with some variation.

For corn these relationships have held well enough in the last 30 years to at least consider the implications for 2009. As mentioned, the October estimate raised yields by 2.3 bushels per acre. Using the historical relationships this would be associated with an increase of 1.4 bushels per acre in the national yields for the November report to 165.6 and an additional .2 bushel in the January estimate to 165.8 bushels per acre.

Keeping harvested acres the same, this would raise the size of the corn crop by 127 million bushels. It is interesting to compare this historically derived estimated increase in production to my 130 million bushel yield loss from the recent frosts/freeze damage during October 9 to October 12. This suggests that the recent weather related production loss may be offset by higher production revisions.

This would mean that the USDA ending stocks number of 1.672 billion bushels would not have to be revised lower as many in the market seem to expect. That would also tend to suggests that corn prices immediately after the weather damage may not be sustainable into the peak of harvest.

Chris Hurt
October-13-09
Purdue University