What's priced in?

What the trade expects before each big USDA report, and how the last one landed against it.

Data: USDA reports and the published trade estimates before each one.

Last report, the October WASDE (Oct 9): 3 of 7 graded figures landed in line. Biggest miss, 2026/27 corn ending stocks: 1.849 bil bu vs 1.677 bil bu expected (+10.3%, bearish). Dec '26 corn settled 20½¢ lower that day, $5.00¼ to $4.79¾ (CBOT daily settles as Yahoo Finance reports them). Every figure
November WASDE: Nov 10, 11:00 a.m. Central. If you are holding unpriced bushels, know the number to beat before that morning. Your own numbers: break-even · safe pre-sell.

Next report: what the trade expects from the November WASDE

November WASDE2026-11-10 · 11:00 a.m. CT · 14 days out
The number to beatUSDA’s October WASDE (Oct 9), standing until the November WASDE
NumberUSDA now
2026/27 corn yield181.2 bu/acre
2026/27 corn production16.034 bil bu
2026/27 corn ending stocks1.849 bil bu
2026/27 soybean yield53.1 bu/acre
2026/27 soybean production4562 mil bu
2026/27 soybean ending stocks315 mil bu
2026/27 wheat production1534 mil bu
2026/27 wheat ending stocks740 mil bu
USDA’s own figures (WASDE-676). A survey average, high and low go beside them once a pre-report survey posts, with the source and time. USDA WASDE
No trade range yet: No pre-report trade survey is on file for this report yet. The survey usually publishes one to two days ahead of the release; the range is typed in from it and is never estimated here.
No market odds: No prediction market is quoting this report. When one is, its odds appear here with the venue named.
No surprise thresholds: A surprise is a print past the trade average by more than its band: 0.5% for a yield, 1% for production, 2% for stocks. The thresholds arrive with the trade average.
Fund positioning: Managed money: Corn net long 164,423 contracts, down 82,002 on the week; Soybeans net long 221,325 contracts, down 28,482 on the week; Wheat net short 52,005 contracts, down 19,450 on the week. CFTC Commitments of Traders, positions as of October 20, 2026.

Track record

How past reports landed against what the trade expected, and which way price moved.

How the expectation lined up against the actual print, and which way price moved. The July WASDE went unscored: expectations were not filed that window; the gap stays on the record.

October WASDE4 surprises2026-10-09
2026/27 corn yieldbearish
177.8 bu/acre → 181.2 bu/acre
+1.9% vs trade · inside the range, near the top
Dec '26 corn settled 20½¢ lower that day, $5.00¼ to $4.79¾ (CBOT daily settles as Yahoo Finance reports them).
2026/27 corn ending stocksbearish
1.677 bil bu → 1.849 bil bu
+10.3% vs trade · inside the range, near the top
Dec '26 corn settled 20½¢ lower that day, $5.00¼ to $4.79¾ (CBOT daily settles as Yahoo Finance reports them).
2026/27 wheat ending stocksbearish
722 mil bu → 740 mil bu
+2.5% vs trade · inside the range, near the top
Dec KC wheat finished 17¢ lower at $7.19¼ and Chicago spot wheat 12¢ lower that day, as DTN's report review printed them on Oct 9, 2026. AGSIST keeps no wheat settles to check these against. Source
2026/27 corn productionbearish
15.716 bil bu → 16.034 bil bu
+2.0% vs trade · inside the range, near the top
Dec '26 corn settled 20½¢ lower that day, $5.00¼ to $4.79¾ (CBOT daily settles as Yahoo Finance reports them).
2026/27 soybean yieldin line
52.9 bu/acre → 53.1 bu/acre
+0.4% vs trade · inside the range
Nov '26 soybeans settled 4½¢ higher that day, $12.87½ to $12.92 (CBOT daily settles as Yahoo Finance reports them). DTN's report review printed $12.94, up 4½¢. The move agrees; the level is 2¢ apart and no CME settlement copy is on file to say which is right.
2026/27 soybean ending stocksin line
311 mil bu → 315 mil bu
+1.3% vs trade · inside the range
Nov '26 soybeans settled 4½¢ higher that day, $12.87½ to $12.92 (CBOT daily settles as Yahoo Finance reports them). DTN's report review printed $12.94, up 4½¢. The move agrees; the level is 2¢ apart and no CME settlement copy is on file to say which is right.
2026/27 soybean productionin line
4541 mil bu → 4562 mil bu
+0.5% vs trade · inside the range
Nov '26 soybeans settled 4½¢ higher that day, $12.87½ to $12.92 (CBOT daily settles as Yahoo Finance reports them). DTN's report review printed $12.94, up 4½¢. The move agrees; the level is 2¢ apart and no CME settlement copy is on file to say which is right.
2026/27 wheat productionno trade estimate
n/a → 1534 mil bu
Dec KC wheat finished 17¢ lower at $7.19¼ and Chicago spot wheat 12¢ lower that day, as DTN's report review printed them on Oct 9, 2026. AGSIST keeps no wheat settles to check these against. Source
September Grain Stocks2 surprises2026-09-30
Corn stocks, all positions, Sept 1bearish
1.918 B bu → 2.095 B bu
+9.2% vs trade · outside the trade range
Corn stocks printed 2.095 billion bushels, above the entire pre-report trade range (1.843–2.005B), not just the 1.918B average. A real miss: more old-crop corn was sitting in bins and elevators on September 1 than any surveyed analyst had guessed.
Soybean stocks, all positions, Sept 1bullish
0.324 B bu → 0.315 B bu
-2.8% vs trade · inside the range, near the bottom
Soybean stocks printed 0.315 billion bushels, 2.8% under the 0.324B trade average, outside the 2% band, so graded bullish. It was still inside the survey's 0.304 to 0.349B range, near the bottom.
Wheat stocks, all positions, Sept 1no trade estimate
n/a → 1.846 B bu
Wheat stocks printed 1.846 billion bushels. No trade estimate was on file to grade this against, the only survey found had a wheat figure that didn't reconcile against a cited WASDE number, so it was withheld rather than published. The print stands on the record; it just isn't scored.
September WASDE3 surprises2026-09-11
2026/27 soybean yieldbearish
52.5 bu/acre → 52.8 bu/acre
+0.6% vs trade · inside the range, near the top
No price move on file for Sep 11, 2026. The settles kept here cover Dec corn and Nov soybeans in the February and October price windows only, and no move is typed in without a dated source.
2026/27 corn ending stocksbearish
1.533 bil bu → 1.567 bil bu
+2.2% vs trade
USDA put 2026/27 corn carryout at 1.567 billion bushels, 34M above the 1.533B trade average (Dow Jones survey, via DTN). No survey low or high is on file for this figure, so it is graded on the 2% band.
2026/27 soybean ending stocksbearish
290 mil bu → 310 mil bu
+6.9% vs trade
USDA put 2026/27 soybean carryout at 310M, 20M above the 290M trade average (Dow Jones survey, via DTN). Reuters' survey averaged 298M; the print was above that too. No survey low or high is on file for this figure, so it is graded on the 2% band.
2026/27 corn yieldin line
178.1 bu/acre → 178.5 bu/acre
+0.2% vs trade · inside the range
No price move on file for Sep 11, 2026. The settles kept here cover Dec corn and Nov soybeans in the February and October price windows only, and no move is typed in without a dated source.
August WASDE1 surprise2026-08-12
2026/27 corn yieldbullish
182.0 bu/acre → 180.7 bu/acre
-0.7% vs trade
USDA's first survey-based corn yield came in at 180.7, 2.3 under July and 1.3 under the trade's 182.0. More acres (planted up 1.4M to 96.7M) kept production near-record at 16.013B bu, but the market traded the yield cut and the 1.653B carryout (down from 1.79B): December corn settled up 15 cents at $4.78.
2026/27 soybean yieldin line
52.9 bu/acre → 52.7 bu/acre
-0.4% vs trade
Soybean yield printed 52.7, a couple tenths under both July and the trade, effectively in line. The acreage add (planted up 1.4M to 86.8M) still makes 4.519B bu the largest soybean crop on record, and carryout ticked up to 320M. November beans settled up 12 cents at $11.84, riding corn's coattails more than their own numbers.
June WASDE1 surprise2026-06-11
2026/27 wheat ending stocksbullish
765 mil bu → 744 mil bu
-2.7% vs trade · inside the range, near the bottom
New-crop wheat carryout came in at 744M, 21M (2.7%) under the 765M trade average and 18M under May, so graded bullish; it was still inside the survey's 740 to 804M range, near the bottom. The cut came from a smaller Hard Red Winter crop (all-wheat production trimmed to 1,543M, yield 47.0); the season-average farm price was cut $0.50 to $6.00.
2025/26 corn ending stocksin line
2.138 bil bu → 2.145 bil bu
+0.3% vs trade · inside the range, near the bottom
Old-crop carryout landed right on the trade guess, a non-event for corn.
2026/27 corn ending stocksin line
1.947 bil bu → 1.96 bil bu
+0.7% vs trade · inside the range
New-crop carryout nudged up fractionally on old-crop demand tweaks; essentially unchanged at $4.40.
2025/26 soybean ending stocksin line
338 mil bu → 340 mil bu
+0.6% vs trade · inside the range
Old-crop beans unchanged from May and on top of the trade, no reaction.
2026/27 soybean ending stocksin line
312 mil bu → 310 mil bu
-0.6% vs trade · inside the range
New-crop beans held at 310, a hair under the trade, quietly friendly but no fireworks.
2025/26 wheat ending stocksin line
942 mil bu → 935 mil bu
-0.7% vs trade · inside the range, near the bottom
Old-crop wheat held at 935, just under the trade guess.
May WASDEnot scored2026-05-12
2026/27 winter wheat productionnot graded
1208 mil bu → 1048 mil bu
-13.2% vs trade
Not graded: the trade survey's date is not on file, so nothing shows it was published before the report.
Smallest winter wheat crop in six decades on Southern Plains drought - July KC wheat ran limit-up and futures hit 2-yr highs.
2026/27 soybean ending stocksnot graded
355 mil bu → 310 mil bu
-12.7% vs trade
Not graded: the trade survey's date is not on file, so nothing shows it was published before the report.
Trade looked for a higher carryout; the tighter print pushed Nov beans to a 2-yr high near $12.
2026/27 corn ending stocksnot graded
1.922 bil bu → 1.957 bil bu
+1.8% vs trade
Not graded: the trade survey's date is not on file, so nothing shows it was published before the report.
New-crop carryout landed near the trade guess - ample supply into 2027 capped the upside.
2025/26 soybean ending stocksnot graded
345 mil bu → 340 mil bu
-1.4% vs trade
Not graded: the trade survey's date is not on file, so nothing shows it was published before the report.
Old-crop carryout close to the trade estimate - little reaction.
2025/26 corn ending stocksnot graded
2.131 bil bu → 2.14 bil bu
+0.4% vs trade
Not graded: the trade survey's date is not on file, so nothing shows it was published before the report.
Old-crop carryout in line - the week's move came from the new-crop sheets.

Analyst estimates, graded, corn & soybean yield forecasts

Who went on record, and how it landed, public forecasters scored against the actual print on how far off they were, how often they beat the trade consensus, and which way they lean. Yield calls and ending-stocks calls are scored in separate groups, and no order is printed until a record is long enough to support one.

Scored records

A finishing order needs at least two forecasters with 10 scored calls each in the same group. One forecaster has 3 or more so far, so this board shows records, not places.
AnalystCallsAvg errorBeat tradeBias
Yield calls · 12 scored calls. Errors are averaged inside this group only: a 2% miss on a yield and a 2% miss on ending stocks are not the same miss, and this page does not add them together.
AGSIST Yield Nowcastoursagsist.com (model)1× closest filed, wider than trade61.73%calls ran 0.77–2.46%17% · 1 of 6+1.73% high
Still building. Listed above the line once a forecaster has 3 scored calls in this group. Their accuracy so far is real and is shown; their position is not, because 2 calls is not a record.
Arlan SudermanStoneX1× closest filed, wider than trade2 of 31.42%calls ran 0.57–2.27%0% · 0 of 2+1.42% high
Robert McClureMcClure Trades1× closest, beat trade2 of 30.98%calls ran 0.06–1.90%50% · 1 of 2-0.92% low
Jerry GulkeGulke Group1 of 30.03%midpoint of a stated range100% · 1 of 1≈ even
Pro FarmerPro Farmer1 of 32.97%0% · 0 of 1-2.97% low

Read the Beat trade column before the Avg error column. A forecast that landed further from the USDA print than the free trade consensus did not beat anything, however close to the print it looks on its own; the count of calls that did beat it is printed beside every record on this board.

Avg error = average % off the USDA actual, averaged within one group only: a yield miss is never added to an ending-stocks miss, because the trade does not read them at the same scale (why). Calls ran = the best and worst single call behind that average. Beat trade = share of calls closer to the print than the trade consensus, with the raw count beside it. Bias = average signed % error: high means systematically over, low means under. Calls = scored metrics in that group; a forecaster needs 3+ in one group to be listed above the line, and this page prints no places at all until two or more forecasters have 10+ each: on three or four calls the gap between two averages is inside their own spread. ours marks our own model, which is scored by the same rules as everyone else and is not exempt from them.

Filed for the next report

Every forecast already filed for an upcoming report, timestamped now, graded automatically once USDA prints. Lean is versus the trade consensus, or USDA's standing projection where no survey exists yet.

Tracked forecasters

Every public forecaster this page follows.

Scored, by report

Each report, each forecaster on record, and how far the trade consensus itself landed.

Forecasters who went on record, scored against the print. Every metric line shows how far the trade consensus itself landed from the print, so you can see what each forecast was up against. ★ marks a call that was both the closest filed and closer than the consensus. Closest-of-those-who-filed is not a win on its own and does not get a star.

How forecasters are scored Average error, beat-trade and bias, and why groups are never pooled.

How forecasters are scored. Public pre-report forecasts measured against the USDA actual: avg error is the mean absolute % miss, beat-trade is the share landing closer than the consensus, bias is the mean signed % error. Errors are pooled only inside a metric group (yield with yield, ending stocks and production with each other) because a 2% miss on a 180 bu yield and a 2% miss on a 310 million bushel carryout are not the same miss; the site’s own in-line bands (0.5% for yields, 2% for stocks) have graded every report since Oct 10, 2026. A forecaster needs 3+ scored calls in one group to be listed above the line, and this page prints no finishing order until at least two of them have 10+ each. It tracks the most public forecasters, not necessarily the best. It is a record of who called what, not advice or an endorsement.

Methodology

What “expected” means, what gets measured, how a report is scored and where the actual comes from.

What "expected" means. The estimate is the pre-report trade survey: the average of the figures grain analysts and trade firms submit in the days before a USDA report, the same numbers the wire services aggregate and the market trades into ahead of the release. The range is the spread of those guesses. It is what the market has already assumed, before USDA says a word.

What gets measured. The headline supply figures that move price: U.S. ending stocks (carryout) for corn, soybeans, and wheat, plus production and yield in the months USDA reports them, plus the quarterly Grain Stocks report, a different measure: a survey estimate of the bushels held on farms and in elevators and processing plants on the first of the quarter, built from NASS's surveys of farm operators and off-farm storage, rather than a balance-sheet projection. These are the numbers traders watch first, because a surprise there ripples through the whole balance sheet.

How a report is scored. Each figure is graded on how far it landed from the trade average. Within 0.5% for a yield, 1% for production and 2% for stocks is in line: two per cent of a 183 bu/acre yield is 3.7 bushels, which would score almost any print as unremarkable, while two per cent of a 2.1 billion bushel carryout is about 42 million and is a fair band. Past the band, a supply figure below the trade average is bullish (less supply than expected) and above it is bearish (more supply than expected). The survey's low and high are printed beside the grade as context ("inside the range, near the top", "outside the trade range") and do not change it. A figure with no trade estimate on record is not scored at all and says so. This rule took over on Oct 10, 2026, and every report on this page was regraded with it. From Oct 3 to Oct 9 any print inside the trade range was graded in line however far from the average; October’s corn ending stocks landed 10.3% over the average and read in line on a day December corn fell 20½¢. A trade estimate only counts if its source is dated before the report. One whose date could not be established is kept and marked "source date unknown". The bullish and bearish surprise thresholds on the report card are the trade average minus and plus that same band: a print past either one is graded a surprise, whether or not it sits inside the analysts’ range. The page does not yet hold enough graded reports to say how price responds when one is.

Where the actual comes from, and when. The corn and soybean yields this board grades are NASS's, published in the Crop Production report at the same hour as the WASDE. A watcher asks NASS for the release month's forecast by name every five minutes inside the release window and fills the figure in the moment it exists. Nothing is written before the report is public, and a figure the source does not publish is left blank with the reason rather than estimated. Ending stocks and the rest of the balance sheet are printed only in the WASDE itself and are still filled in by hand. Grain Stocks prints on its own separate release day and is filled in by hand from that report, not from the WASDE.

Sources. Report dates from the USDA WASDE and NASS release schedules; trade estimates from the published pre-report surveys, linked on each figure; actuals from USDA NASS (yields and Grain Stocks) and the WASDE itself (ending stocks, typed in by hand). Fund positioning from the CFTC Commitments of Traders, dated on the card. This is a map of expectations, not a trade recommendation or financial advice.

Grading method and sources above. · Reviewed by Sigurd Lindquist, CCA

Common questions

What does "priced in" mean?

Markets move on surprises. If the trade already expects a bearish number, that expectation is built into today's price: "priced in." Price reacts to the gap between the report and what was expected, which is why the expectation matters as much as the report.

Where do the expectations come from?

The estimate range is the pre-report trade survey, typed in from the published survey and left blank with the reason when none has been filed. Positioning is the weekly CFTC Commitments of Traders, dated on the card, because a position is only a fact about the Tuesday it was taken. Implied odds appear when a prediction market is quoting the report and the card says so when none is. Together they show what the market has already assumed.

What's a "surprise threshold"?

The trade average plus or minus its band. Since Oct 10, 2026 every figure is graded on its distance from the pre-report trade average: within 0.5% for a yield, 1% for production and 2% for stocks it is in line, and past that it is a surprise. For a supply number like stocks, production or yield, below the band is bullish and above it is bearish. So a print can sit inside the analysts' range and still be a surprise: the October corn yield, 181.2 against a 177.8 average, was inside the 173.2 to 182.1 range and graded bearish. The range is printed beside the grade as context. This page does not yet hold enough graded reports to say how price responds when a print lands past a threshold.

Is this a trade recommendation?

No. It's a map of expectations, not advice. It tells you where the bar is set so you can judge a report's reaction for yourself.

Built and maintained by a CCA, available at no charge for non-commercial use. Questions: sig@agsist.com · 715-797-2428. Nothing here is marketing or financial advice.