Will USDA Continue to Increase Soybean Yields in 2009?

October-13-09

In a corn article on our corn outlook web page, I looked at the strong tendency for USDA increases in corn yield estimates from September to October, as we have this year, to historically be related to even higher estimated yields in November and into the final January estimate. That tendency is not as strong for soybeans and there is much less predictive power this year to suggest soybean yields will continue to be revised higher.

Over the last 30 years from 1979 to 2008, 67% of the time a higher (lower) yield estimate in October was associated with a higher (lower) yield in November and 63% of the time a higher (lower) October yield estimate was associated with a higher (lower) final estimate in January.

While these percentages seem to point to a fairly strong relationship, the predictive power of the changes in the October yield to predict subsequent changes in yield is low. This was especially true when the October estimate was only slightly changed from September as it was this year (September 42.3 bpa and October 42.4 bpa).

As an example when the change in national soybean yield from September to October was less than .5 bushels per acre, the final yield changes ranged from a decline of 1.7 bushels to an increase of 1.7 bushels per acre. There was no ability for the change in yields from September to October to predict the direction of the final yield when the yield change from September to October was small as it was this year.

So, unlike corn, this historical analysis would suggest no ability to provide meaningful clues of the direction of the final soybean yield estimates for 2009.

Chris Hurt
October-13-09
Purdue University