· Analytic
Z4646 FGIS China-Share-of-US-Corn-Exports Velocity
The 4-week acceleration of China’s share of US weekly corn export inspections, winsorized at 2nd–98th percentile and z-scored. Captures demand-shock dynamics in Chinese purchasing of US corn , a rare destination that when-active produces visible announcement-fade dynamics in corn forward returns.
Current reading
+0.978
Extreme Pos
DAILY · SIGNAL
As of Oct 30, 2026. Updated nightly.
Current composite
+0.9865
On the plus-or-minus one scale; China share decelerating sharply (via polarity flip, extreme-pos signal = deep deceleration)
Signal state
Extreme POS
22 consecutive days in episode
All-time percentile
99.6th
Essentially maximum reading in 36-year history
moderate-pos corn fwd 60d
+7.26%
216 days (2.4%), HAC t +3.25 sig at 1% , small-N sharp bull tail
extreme-neg corn fwd 60d
-2.48%
1,635 days (18.2%), HAC t -3.83 sig at 1% , large-sample bear tail
Corn Extremes Contrarian V9 contribution today
-0.197
20 percent weight × -1 polarity × +0.978 value , LARGEST negative contribution in Corn Extremes Contrarian V9
OBJECTIVE
China is the world’s second-largest corn importer behind Mexico but historically a rare buyer of US-origin corn , Chinese purchases are primarily from Brazil (sanitary agreement reached 2022), Argentina, and Ukraine. US-China corn trade activates only during specific regimes: pandemic-era stockpiling (2020-2021), trade-deal purchases following Phase One (2020), origin-switch windows during Brazilian supply disruptions. This signal queries USDA FGIS Export by Destination weekly data, computes China’s share of total US corn exports as a 4-week rolling fraction, computes the 4-week change in that share (acceleration), winsorizes at the 2nd through 98th expanding percentiles (removes outliers from rare-single-week-surge events), and z-scores on a 252-day rolling window with 63-day minimum. The signal is polarity-inverted per the engine’s own 2026-07-10 rework comment: the author’s thesis was that China surge is bullish ("China rarely buys US corn , surge = bullish"), but empirical polarity testing found that high China share surge is SHORT-TERM BEARISH via "announcement fade" dynamics , the surge coincides with the market pricing-in the demand, after which the forward window captures mean-reversion. The signal is therefore polarity-inverted so that a surging China share maps to a NEGATIVE signal (contrarian-bearish for corn forward returns). The empirical pattern across 36 years shows a bimodal small-N structure. extreme-neg state (val below minus 0.5, deep China-share surge via inversion) delivers minus 2.48 percent corn fwd 60d (HAC Newey-West t minus 3.83, significant at 1 percent, 1,635 state-days , large sample). moderate-pos state (val between plus 0.2 and plus 0.5, moderate China-share deceleration) delivers plus 7.26 percent (HAC t plus 3.25, significant at 1 percent, 216 state-days , small-N but sharp). extreme-pos state (val above plus 0.5, deep deceleration) delivers plus 1.13 percent (HAC t plus 0.89, not statistically significant). The signal reads +0.978 as of October 23 2026 in extreme-pos territory for 22 consecutive trading days at the 99.6 percentile , essentially the deepest China-share deceleration in the modern sample. The signal contributes 20 percent weight to Corn Extremes Contrarian V9 Corn Extremes Contrarian with polarity sign MINUS one , the current reading contributes minus 0.197 to the Y composite (the LARGEST negative contribution among the 5 Corn Extremes Contrarian V9 members today). Read on for the mechanism, the signature historical cases, and the honest narrative-vs-code drift flagged by the persona code review.
LIVE CHART
Thirty-six years of China-share-of-US-corn-exports velocity, winsorized + z-scored + polarity-inverted.
COMPOSITE
1.0 / 100
STATE
Extreme Pos
PHASE
+1
FWD 60d
+2.96% (baseline +2.31%)
SAMPLE N
9,039 days
TODAY’S PLAYBOOK READING
STATE
Extreme Pos
STATE AND PERCENTILE
extreme-pos at the 99.6 percentile , essentially historic maximum reading.
WHAT THE RAW DATA IS DOING
China’s share of US corn exports has decelerated sharply , current China-origin buying is near-zero while total US corn exports are elevated. The sharp deceleration via polarity flip maps to a near-maximum positive signal.
POSTURE FOR THE WEEKS AHEAD
extreme-pos historical mean +1.13 percent (not sig). The information-rich tail in this signal is extreme-neg (large sample, HAC t -3.83 sig at 1 percent) which represents the opposite regime (China share surging) , not the current reading. The current extreme-pos reading at historic extreme is NEW TERRITORY empirically.
WHAT THE CHART SHOWS
The composite sits at +0.98, saturated in the China Share Accelerating band. The FGIS 4-week acceleration of China’s share of US weekly corn export inspections is firing deep into the Accelerating extreme. The parallel Z4642 CNY Weak signal confirms the upstream currency mechanism driving the surge. CBOT front-month corn (orange, right axis) sits near $4.48/bu.
TOP INPUTS FIRING
- Primary composite +0.98 State China Share Accelerating, saturated extreme band. The 4-week acceleration of China’s share of US corn export inspections is reading deep extreme.
- FGIS weekly corn export inspections, China-share 4-week acceleration USDA Federal Grain Inspection Service weekly export inspection reports. China-destination share, 4-week acceleration, winsorized at 2nd-98th percentile and z-scored.
- IC vs CBOT corn fwd 60d: Spearman +0.165, p=1.5e-24 (N=3,796) Significantly positive IC on full sample. Current rolling 252d IC -0.182 at p=0.012 suggests recent regime has been slightly mean-reverting after acceleration peaks; interpret current reading with awareness that the full-sample mechanism is positive but near-term dynamics can be contrarian.
THE TAKEAWAY
What this means for you today
THE CHINA DEMAND VELOCITY CHANNEL
China’s corn-import share from US origin has historically been volatile. In 2020-2021 Phase One trade-deal windows, China share reached 20-30 percent of weekly US corn exports. Post-2022 Brazilian sanitary-agreement opening, China share has returned to near-zero in most weeks. This signal captures the ACCELERATION (change in share over 4-week windows) rather than the absolute level, so it fires on changes-of-regime rather than steady-state levels. The 2026-10-25 reading of 99.6 percentile indicates the sharpest 4-week deceleration in 36 years of history.
THE ANNOUNCEMENT-FADE POLARITY
The engine’s 2026-07-10 rework polarity-inverted the signal based on empirical direction-testing. The author’s original thesis was that China surge is BULLISH for corn ("China rarely buys US corn , surge = bullish"). The empirical history showed the opposite: when China share surges, the surge arrives AFTER corn has already rallied in anticipation (announcement fade), and the forward 60-day window captures mean-reversion. The polarity inversion makes a China-share SURGE map to NEGATIVE signal (contrarian-bearish for corn), and a China-share DECELERATION map to POSITIVE signal. The current near-maximum positive reading reflects historic deceleration , the opposite regime from surge.
THE HONEST CAVEAT , ENGINE MECHANISM DIFFERS FROM CONFIG METADATA
The engine’s current `_signal()` function (code reworked 2026-07-10 per in-code comment) computes **China-share-of-US-corn-exports × velocity**. However, the engine’s CONFIG metadata , the `description`, `website_metadata.display_name`, `website_metadata.short_description`, `website_metadata.what_it_measures`, `website_metadata.why_it_matters`, and `website_metadata.historical_context` fields , all still describe the LEGACY mechanism: "USDA FGIS weekly corn export inspections YoY." The CONFIG metadata was NOT updated when the engine was reworked. **Readers consuming the CONFIG metadata would be badly misled** about what this signal actually computes. This page describes the ACTUAL engine mechanism, not the stale CONFIG. See §11 Methodology Limitations for the full persona code review finding.
THE REGIME MAP
The three China-share-of-US-corn-exports velocity regime states, forward returns and sample share
Behavior across the 4-week acceleration of China’s share of US weekly corn export inspections (FGIS data), winsorized at 2nd-98th percentile and z-scored. Captures demand-shock dynamics in Chinese purchasing of US corn. CBOT corn benchmark.
| State | Phase | Share | Fwd 60d | HAC t | Edge vs baseline | Last firings |
|---|---|---|---|---|---|---|
| China Share Decelerating | – | 22.7% (2057 d) | +0.44% | +0.15pp | Jun 2026, Oct 2025, Jan 2025 | |
| China Share Neutral | 0 | 66.1% (5976 d) | +5.33% | +5.04pp | Apr 2026, Dec 2025, Sep 2025 | |
| China Share Accelerating TODAY | + | 11.2% (1011 d) | +7.42% | +7.13pp | Oct 2026, Jan 2026, Sep 2024 |
The historical pattern
Why it matters
THE TRACK RECORD
State changes
| State | Start | End | Duration | Fwd 60d SPY |
|---|---|---|---|---|
| Extreme Neg | 10/23/2024 | 11/27/2024 | 26d | +4.92% |
| Normal | 11/29/2024 | 01/29/2025 | 41d | -1.39% |
| Extreme Pos | 01/30/2025 | 09/26/2025 | 166d | -8.96% |
| Normal | 09/29/2025 | 10/06/2025 | 6d | +3.66% |
| Extreme Pos | 10/07/2025 | 12/05/2025 | 43d | +2.10% |
| Normal | 12/08/2025 | 01/20/2026 | 29d | -1.65% |
| Extreme Neg | 01/21/2026 | 04/13/2026 | 57d | +3.61% |
| Normal | 04/14/2026 | 06/17/2026 | 46d | +8.71% |
| Extreme Pos | 06/18/2026 | 09/30/2026 | 72d | +1.43% |
| Extreme Neg | 10/01/2026 | 10/23/2026 | 17d |
Scroll inside the table for the full history. Forward 60-day SPY is measured from each run’s start date; a reader cannot trade off a start-date they only identify in hindsight, so this is a transparency log, not an alpha claim. Asterisked (*) returns are partial. The run is less than 60 trading days old. The full 60-day comparable fills in as more data arrives.
What this measures
The engine queries `fundamentals_data.USDA_FGIS_EXPORT_BY_DESTINATION_WEEKLY` with `commodity=’CORN’`, groups by `week_ending × country`. Filters to China rows (country name contains “CHINA”). Computes `china_share = china_mt / total_mt` per week. Applies 4-week rolling smooth. Computes 4-week change (acceleration). Applies expanding 2nd-to-98th percentile winsorization (requires minimum 52 weeks of history). Forward-fills daily. Z-scores on 252-day rolling window with 63-day minimum. **Polarity-inverts** (`zsc = -_z(…)`) per the empirical rework finding that high China-share surge is short-term bearish via announcement fade. Tanh-bounds. Five-state machine on the final the primary composite.
GO DEEPER
Want the full depth?
The full research brief adds methodology detail on the FGIS Grain Export Inspections report structure, how China-share is computed, why 4-week acceleration isolates demand-shock dynamics from secular trend, and the winsorization choice. Includes historical case studies anchored to actual Chinese import episodes (2015-2016 devaluation surge, 2020-2021 COVID stockpile, 2019-2020 trade war), the three-regime state-conditional table, and the methodology for pairing with CNY Strength, Rail Index, Ending Stocks, and the Corn Cycle DNA playbook.
Alert drafts
Not investment advice · The analysis on this page is informational and educational. Signal readings, historical case figures, and empirical statistics are derived from publicly available data and are provided without warranty. Hypothetical performance results have inherent limitations, including the benefit of hindsight. Readers who intend to trade should consult a licensed investment professional and see the research brief for the full disclaimer.
Last refreshed Nov 02, 2026 12:33 PM ET