Get ready. USDA releases what may be its last market-moving World Agricultural Supply And Demand Estimates of 2023 next week. The report, featuring updated corn and soybeans production forecasts, is likely to set the tone for trading into the end of the year, and perhaps beyond.

To be sure, the government also publishes a December WASDE, but typically doesn’t change much in its estimates of U.S. supply or demand. So, without big weather changes in South America, or yet another surprise from volatile world markets and events, traders won’t have that much to chew on as activity winds down for the holidays.

Moreover, history suggests the November WASDE could feature more surprises on U.S. crop estimates. Corn and possibly soybeans could be in line for another round of yield reductions if the old adage about small crops getting smaller holds true.

In any event, when it comes to production guesses, the WASDE is the only game in town. With harvest half over, USDA already suspended Crop Progress ratings for corn and soybeans. Satellite vegetation maps also don’t provide fresh insight with a growing percentage of fields bare. Usefulness of Vegetation Health Index began fading weeks, if not months ago.

Accuracy of weekly Crop Progress condition reports, of course, have attracted plenty of criticism from both farmers and traders. Though this “beauty contest,” like the VHI, does have a statistically significant correlation with yields, the gold standard for measuring bushels per acre remains USDA’s monthly surveys of farmers and their fields first reported in August. The next, and final survey, won’t be released until Jan. 12, though USDA sometimes makes changes nine months later in its end of September grain stocks inventory after the end of the marketing year in August.

USDA officials reiterated the importance of the surveys as their primary tool for making changes to WASDE production at the agency’s annual fall data users meeting earlier in October. While they look at satellite maps, weather and crop progress ratings, the surveys trump those by far.

So here’s what those surveys suggest we could see from the Nov. 9 WASDE.

How much corn?

Trying to predict WASDE November yield estimates based on October results, should be easy, because there’s a very strong positive correlation between the two months, implying a 93% chance the two will track.

Based on this, November should come at 173.15 bushels per acre, up slightly compared to the 173 printed Oct. 12. On average over the past 30 years, November yields were up about that same amount, though the agency issued no October WASDE in 2013 due to a government shutdown.

Still, the annual estimate put out in January could be down to 172.4 bpa based on this same logic, because on the average for January is lower than November or October, with later revisions after the September Grain Stocks and Ag Census knocking a little more off the yield, though by then 2023 production likely will be just a footnote.

Another pattern teases a lower November figure right from the get-go, without waiting until the New Year. Eight times since 1993, yields fell in both the September and October reports, following the first monthly estimates in August. In five of those years yields also fell in November, losing an average of 1.4%, which would cut the yield to 170.6 bpa. And in six of the years, January yields also fell, which would take the average yield down to 169.5 bpa if the pattern holds.

Of course, last year yields fell in September and October, then rose in November and January, so this is far from a sure thing, and not enough to be statistically significant. So, those ideas about small crops may not be a real trend at all, but just a fluke.

But, if the trend does hold, production could be down 200 million bushels in November, with another 100 million reduction in January.

Numbers like those could finally take projected ending stocks for the marketing year down below 2 billion bushels, assuming usage doesn’t also tumble. That would likely confirm fall lows, but not be enough to spark any type of major rally without damage to crops in South America or another black swan event that resets the table.

Minor soybean changes

Soybean stocks are tighter relatively than corn, so cuts that would take 2023 crop carryout lower could jumpstart futures if South American production doesn’t keep growing. But history suggests any potential impacts on soybean yields from prior WASDE estimates would be somewhat limited.

The correlation between October and November soybean yields is even stronger than the connection for corn. Average yields over the past 30 yields showed a similar pattern, increasing from August through October, before slipping a little in November and gaining slightly again in January. Projecting the October to November tend from all 30 years, like corn, shows a little higher yield, but up less than a tenth of a bushel per acre, or 10 million bushels total.

Soybean yield estimates for 2023 fell from August to September and October, which happened in seven of the 30 years since 1993, one time less than corn. Still, this pattern projects a cut of a little more than .2 of a bushel, taking the 49.6 bpa printed in June down to 49.4 bpa. Based on this logic, January could also be down, but basically just enough to be a rounding error. That would make soybeans even more dependent on South American unless China begins buying much more aggressively, say as part of an effort to deflect criticism of its record on technology, human rights, foreign policy, military expansion, etc.

Of course, surprises happen, especially when there’s plenty of error built into these guesses and even error statistics have outliers as well.

But it is a caution against hoping for too much help from the November soybean report, at the least. The good news, for those who trade USDA numbers? At least the agency has funding to put the report out. Whether that’s true after the current deal in Congress expires Nov. 17 remains very much an open question.