Can't Beat the Heat

Ahem.

Poor Weather Pushes Prices Up for Corn and Soybeans:

Prices for corn and soybeans jumped in commodities markets on Thursday after the Department of Agriculture said that the onslaught of bad weather, from heat and drought in some areas to heavy rains and flooding in others, would reduce yields for those critical crops.

Okay, I'll try to not let this go to my head.  But the heat really kills.  Maybe not as much when it's wet.  But it still kills.

The growing season isn't over yet.  USDA was too optimistic last year at this time.  But I think that was mainly because the heat wave came after surveying for the August crop report was finished (about August 1).  If the weather stays good from now until harvest I'd guess the forecast will be about right.

I should also note that at least some of the price bump today could have been a demand shock.  Aggregate demand is highly uncertain right now and closely tied to the stock market, which spiked about as much as commodities.

Supply shifts in, demand shifts out, prices go boom!

Update: An interesting tidbit (maybe) for ag. commodity wonks out there:  About a week before each USDA forecast comes out, a couple private-market forecasts can be purchased.  I don't have links handy, but I suspect all the major commodity players do in fact buy these private market forecasts.  Some analysis by Scott Irwin and Daryl Good at UofI shows that private market forecasts are as accurate as USDA forecasts.  Still, USDA forecasts move the market quite substantially when they are announced.  Why?

Well, it turns out that combining the private-market forecast and USDA forecasts gives a forecast that significantly beats both of them.  That means the USDA forecast still contains a lot of information.  It also means that the market "surprise" is something far smaller than what was announced, and smaller even than the difference between private-market and USDA forecasts.  The implication of all this is that prices are even more sensitive to quantities than they may appear, and about twice as sensitive (I think) as Irwin and Good estimated.