Speculative positioning does not detectably lead weekly USDA crop-condition reports — but it moves sharply around monthly WASDE releases

Summary

At the weekly grain of USDA's Crop Progress and Condition report (the subjective good/excellent ratings released every Monday during the growing season), we find no statistically detectable lead: managed-money net positioning in corn and soybean futures, measured on the last CFTC Commitments-of-Traders (COT) snapshot before each week's release, does not predict the direction or size of that week's condition change (corn: r = -0.06, p = 0.81, n = 19 weeks; soybeans: r = -0.17, p = 0.50, n = 18 weeks; 2025 season). The published literature and trade press instead document large speculative repositioning around a different, monthly report — USDA's WASDE (World Agricultural Supply and Demand Estimates) — where a 21-year study found corn and soybean price variance roughly 6.9–7.4x normal on release days, and where hedge funds have cut net-long positions by roughly a quarter in a single week ahead of a report. That is a real, sizeable, and separately documented effect; it is not the same report this question names, and the corpus queried directly here does not show the same effect for the weekly condition report.

What was tested and why

USDA's weekly Crop Progress/Condition report is a subjective survey (percent of the crop rated poor/fair/good/excellent), collected Sunday-to-Sunday and released the following Monday afternoon. If speculative funds have private information or better models (satellite imagery, proprietary weather forecasts) that anticipate what that Monday's report will say, their net position should already shift in the report's direction before release — i.e., the change in their net long/short position on the last COT survey (a Tuesday) before the report should correlate with the change the report then announces.

We tested this directly: for corn and soybeans in the 2025 season, we computed the week-over-week change in CFTC-reported managed-money net-long position (as % of open interest) as of the Tuesday COT report immediately preceding each week's Monday USDA release, and correlated it against the week-over-week change in an unweighted Corn-Belt average Good+Excellent% (Iowa, Illinois, Indiana, Ohio, Minnesota, South Dakota, Wisconsin, Missouri, Kansas, North Dakota — the states with usable weekly coverage in this corpus; see data limitation below). Both correlations were near zero and not significant.

Data limitation: no national weekly condition series in this corpus, and coverage is patchy across years

ag.nass_crop_production carries state-grain weekly condition only — NASS's own national aggregate is not loaded, so the Corn-Belt average used here is a proxy, not the published national Good+Excellent% figure. More importantly, state-level weekly condition coverage in this table is genuinely sparse and inconsistent across years: for corn in Iowa alone, only 2008, 2010–2013, 2015–2017, 2020 and 2025 have any weekly condition rows at all, and even those years are missing weeks. 2025 was the most complete season for the ten major Corn Belt states, which is why the analysis above is limited to that single season (n≈19 weekly observations per commodity) rather than a multi-year panel that could report a more stable, less noisy estimate. This is reported to the AskAmerica data-quality channel as a coverage gap, not treated as evidence of "no effect" beyond what one season can support.

What the literature says about a related but different report — WASDE

Independent reporting and research point to substantial speculative repositioning around USDA's monthly WASDE report, not the weekly condition report: a documented case shows managed money cutting its corn net long by roughly a quarter in one week ahead of a January 2023 USDA report, and a 21-year academic study found WASDE releases carrying NASS production estimates produce corn price variance about 7.38x normal and soybean variance about 6.87x normal on the release day (see sources). Trade coverage from 2026 describes managed money holding a combined net long of roughly 556,000 corn/soybean/wheat contracts (≈ 2.8 billion bushels equivalent) at a recent peak before unwinding sharply as crop conditions improved and energy prices fell — a real, large swing, but one that occurred over weeks to months around evolving fundamentals and the monthly WASDE cycle, not as a short-horizon lead on a single weekly condition release. The 2025 corn COT series pulled directly in this analysis shows the same order of magnitude: managed money's net-long position swung from +14.5% of open interest in January to -8.0% of open interest by early June — a real and large repositioning over the season, just not one that leads the weekly condition report on a week-by-week basis.

Bottom line

Answering the question as asked — the weekly USDA crop condition report specifically — the honest answer from data directly tested here is: measurably, not at all, at weekly grain, in the one season with usable data (2025). Speculative funds do reposition substantially over the growing season (swings of 15–20 percentage points of open interest are common), and they reposition sharply around the separate monthly WASDE report, but that repositioning does not show up as a statistically detectable lead on the following week's condition announcement in the data available. A reader should not generalize this null result confidently beyond the single season tested; a multi-year panel would need better NASS weekly-condition coverage than this corpus currently has.

What This Report Does Not Answer

• Use a multi-year panel for robustness: ag.nass_crop_production's state-level weekly condition series has genuine, patchy year-to-year coverage gaps (confirmed live); only the 2025 season had complete-enough Corn Belt state coverage to build a full weekly series, so the empirical test is limited to n≈19 weekly observations in one season rather than several years pooled

Every query behind this report

16 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT DISTINCT statisticcat_desc FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND agg_level_desc='NATIONAL'

SELECT DISTINCT agg_level_desc FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND statisticcat_desc ILIKE '%COND%'

SELECT DISTINCT short_desc FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND statisticcat_desc ILIKE '%COND%' LIMIT 20

SELECT contract_market_name, cftc_contract_market_code FROM cftc.cot_disaggregated_futures WHERE contract_market_name ILIKE '%CORN%' GROUP BY contract_market_name, cftc_contract_market_code

SELECT DISTINCT state_alpha FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND statisticcat_desc ILIKE '%COND%' ORDER BY state_alpha

SELECT year, state_alpha, reference_period_desc, short_desc, value, load_time FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND short_desc='CORN - CONDITION, MEASURED IN PCT GOOD' AND state_alpha='IA' AND year=2023 ORDER BY reference_period_desc FETCH FIRST 30 ROWS ONLY

SELECT "year", state_alpha, reference_period_desc, short_desc, "value" FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND short_desc='CORN - CONDITION, MEASURED IN PCT GOOD' AND "year"=2023 FETCH FIRST 30 ROWS ONLY

SELECT "year", state_alpha, reference_period_desc, "value" FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND short_desc='CORN - CONDITION, MEASURED IN PCT GOOD' AND "year"=2023 AND state_alpha='IA' ORDER BY reference_period_desc

SELECT "year", COUNT(*) FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND short_desc='CORN - CONDITION, MEASURED IN PCT GOOD' AND state_alpha='IA' GROUP BY "year" ORDER BY "year"

SELECT state_alpha, COUNT(*) as n FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND short_desc='CORN - CONDITION, MEASURED IN PCT GOOD' AND "year"=2025 GROUP BY state_alpha ORDER BY n DESC

SELECT "year", state_alpha, COUNT(*) as n FROM ag.nass_crop_production WHERE commodity_desc='CORN' AND short_desc='CORN - CONDITION, MEASURED IN PCT GOOD' AND "year" IN (2016,2017,2020,2025) AND state_alpha IN ('IA','IL','IN','OH','MN','SD','WI','MO','KS','ND','NE') GROUP BY "year", state_alpha ORDER BY "year", state_alpha

SELECT report_date, open_interest_all, money_manager_long_all, money_manager_short_all
FROM cftc.cot_disaggregated_futures
WHERE cftc_contract_market_code='002602' AND "year"=2025
ORDER BY report_date

SELECT reference_period_desc,
  AVG(CASE WHEN statisticcat_desc='CONDITION' AND short_desc='CORN - CONDITION, MEASURED IN PCT GOOD' THEN "value" END) as good,
  AVG(CASE WHEN statisticcat_desc='CONDITION' AND short_desc='CORN - CONDITION, MEASURED IN PCT EXCELLENT' THEN "value" END) as excellent
FROM ag.nass_crop_production
WHERE commodity_desc='CORN' AND "year"=2025
  AND state_alpha IN ('IA','IL','IN','OH','MN','SD','WI','MO','KS','ND')
  AND short_desc IN ('CORN - CONDITION, MEASURED IN PCT GOOD','CORN - CONDITION, MEASURED IN PCT EXCELLENT')
GROUP BY reference_period_desc
ORDER BY reference_period_desc

SELECT contract_market_name, cftc_contract_market_code FROM cftc.cot_disaggregated_futures WHERE contract_market_name ILIKE '%SOYBEAN%' GROUP BY contract_market_name, cftc_contract_market_code

SELECT reference_period_desc,
  AVG(CASE WHEN short_desc='SOYBEANS - CONDITION, MEASURED IN PCT GOOD' THEN "value" END) as good,
  AVG(CASE WHEN short_desc='SOYBEANS - CONDITION, MEASURED IN PCT EXCELLENT' THEN "value" END) as excellent
FROM ag.nass_crop_production
WHERE commodity_desc='SOYBEANS' AND "year"=2025
  AND state_alpha IN ('IA','IL','IN','OH','MN','SD','WI','MO','KS','ND')
  AND short_desc IN ('SOYBEANS - CONDITION, MEASURED IN PCT GOOD','SOYBEANS - CONDITION, MEASURED IN PCT EXCELLENT')
GROUP BY reference_period_desc
ORDER BY reference_period_desc

SELECT report_date, open_interest_all, money_manager_long_all, money_manager_short_all
FROM cftc.cot_disaggregated_futures
WHERE cftc_contract_market_code='005602' AND "year"=2025
ORDER BY report_date