"A word of warning before the USDA crop report on Wednesday. The trade is looking for somewhat higher yields for both corn and beans. The analysts' range is tight. Just like last year they are clustering around the last USDA number, and we can all remember what kind of cluster became of that!"

"Frankly I have no idea what kind of data they are looking at, and I suspect that they can't see beyond the "better than expected" comments. After such a difficult growing season, I am sure that many farmers had low expectations. Personally I don't calculate yields on expectations. What I do is gather as many actual yields, anecdotal and test plots and compare them to a vast data base going back 11 years. With an early harvest last year it was easy to call the corn yield below 155 by mid September, when the USDA was at 165, and analysts between 165 and 170. This year the corn harvest is much slower, but it has been clear for at least two weeks that the corn yield is going down. The anecdotal yields have been consistently lower from day 1. Last year's anecdotal yields averaged 172.7. This year the anecdotal yields have been between 160 and 163. With around 60 % of expected yields in, they currently stand at 160.1 or down 12.6 bushels from a year ago. This compares to the USDA's September projection down 4.7 bushels. The test plots tell a similar story. With thousands of test plots reporting, perhaps 20 percent of what we had last year, the projected yield is far below the USDA. Iowa is meant to be 2 bushels above a year ago, but is averaging 4 below. Illinois is meant to be 4 above, but is 4 below. Indiana is more than 20 bushels below the USDA. Kansas is more than 10 below, and Nebraska is more than 5 below. The test plots are pointing to a national yield of between 141 and 142. However about 75 % of the yields reported are from plots planted before May 10 .Nationally only 40 % of the crop was planted before May 10. My data shows a very significant drop off of in yield in the Eastern cornbelt for corn planted after May 10. This would be the corn that pollinated in the heatwave that lasted from July 17 until August 5 .According to a well respected agronomist whom I spoke to last week, the biggest detriment to yield in those circumstances would be shallow kernel depth which is not something that the USDA would have picked up on in September.

"As for soybeans, the trade has been touting bigger and bigger yields as the days go by. Once more, I cannot find any evidence to support this. The anecdotal yields have always suggested lower yields, and today are running at a whopping 7 bushels, or 11 % below a year ago. The test plots have been flooding in the last few days, mainly from the Western belt. Iowa, which is projected to be unchanged from a year ago is currently almost 3 bushels below. Minnesota is averaging 10 bushels below , with the USDA at 4 below, S Dakota is 8 below, with the USDA at unchanged, and Wisconsin is in line at 5 below. The anecdotal yields for Illinois and Indiana are running more than 10 bushels below a year ago.

"The USDA moves in mysterious ways. Just take a look at the last several stocks reports. 300 million bushels has now become just a fudge factor. Every USDA report is dangerous territory. I have no idea when they will see the writing on the wall for these crops. Perhaps they will wake up in November, but the trade is leaning horribly wrong , looking for bigger crops."

Eoin Treacy's view Thank you for this educative email which I have no doubt will be well received by the Collective. Grains and beans were among the best performers in the commodity sector from mid 2010 as they rallied from comparatively depressed levels on the back of strong fundamentals and heightened speculative interest. They were not spared in the unwinding of leverage since August and fell precipitously along with a number of other commodities.

Corn rested its June peak just below 800¢ in early September and pulled back to test the 600¢ area where it has at least paused. A similar bounce to that posted in early July would increase confidence that a medium-term low is being formed, while a sustained move below 575¢ would complete a medium-term Type-3 top formation.

Soybeans hit a medium-term peak near 1400¢ in February and ranged with a mild downward bias until late August when it hit a new high. However, it was unable to sustain the upward break and quickly fell to the lower side of the range and extended the decline below the 200-day MA. Prices fell for 19 of 23 sessions and have become oversold as they pause in the region of 1150¢. Potential exists for at least a relief rally and the short-term upside can probably be given the benefit of the doubt in the absence of a sustained move below 1150¢. A move above the 200-day MA, currently near 1310¢, held for more than a few days would be required to suggest a return to medium-term demand dominance.

Wheat has been forming a large volatile first step above the 2009 & 2010 base since late last year. It is currently testing the lower side near 600¢ and a clear upward dynamic is required to reconfirm the return of demand in this area. Oats rallied more than wheat but has also returned to the region of the lower side of a relatively lengthy range.