COT Mar 31 – 11 Markets Hit Positioning Extremes
Mar 31, 2026
This week's CFTC positioning data shows hedge funds clustered at extremes across 11 markets, with 2-year Treasury shorts hitting a decade low and agricultural longs reaching their most crowded levels in two years.
Executive Summary
Last week's analysis highlighted positioning shifts across several markets. This week, the breadth of extremes widened sharply — 11 markets now sit at the 95th percentile or above (or 5th and below), the most simultaneous extremes observed in the 104-week dataset. 2Y Treasury shorts reached the 1st percentile in both 2-year and 10-year windows (-1.64M contracts, Z-score -2.99). Australian Dollar longs hit the 99th percentile across both timeframes (81.5K contracts, 30.9% of open interest). The agricultural complex shows coordinated crowding: soybean oil, soybean meal, cotton, and two wheat contracts all sit above the 98th percentile.
Methodology & Data Source
- Source: CFTC Commitment of Traders Report (Legacy Format)
- Analysis: Non-Commercial traders (large speculators — hedge funds, asset managers)
- Percentiles: Dual window — 2-year (104 weeks) primary + 10-year (520 weeks) structural context | Full Methodology →
- 📥 Download this week's data (JSON)
- 📊 View Interactive Dashboard →
Cross-Market Breadth: Eleven Extremes Across Five Asset Classes
The positioning landscape this week is defined not by a single dominant story, but by the sheer number of markets at statistical extremes. Eleven contracts reached the 95th percentile or higher (or 5th or lower) in the 2-year window. This breadth is notable: only twice in the past 104 weeks have more markets simultaneously clustered at these thresholds.
The extremes span five asset classes. Bonds, FX, grains, metals, and crypto each contribute at least one market to the list. This cross-sectional crowding suggests positioning decisions are being driven by macro themes rather than idiosyncratic market dynamics. Within the 104-week sample, periods with 10+ simultaneous extremes have historically coincided with elevated positioning volatility in subsequent weeks — though this is a structural observation, not a timing signal.
Open interest data for each market is static in this dataset (OI history is not tracked), so whether these extremes are forming on expanding or contracting liquidity cannot be assessed. This limits the ability to distinguish between conviction-driven builds and liquidity-driven squeezes.
Bonds: 2Y Treasury at Historic Short Extreme
2Y Treasury net shorts reached -1,637,324 contracts, the 1st percentile in both 2-year and 10-year windows. This is the lowest net positioning in the entire 520-week historical series. The Z-score of -2.99 places it nearly three standard deviations below the full-period mean.
The build has been sustained. Net shorts are -664,883 contracts below the 4-week average and -578,360 below the 13-week average. This is not a tactical spike — it is a multi-month accumulation of net short exposure. Non-commercial traders now hold a net short position equal to -34.7% of open interest (4.72M contracts). That crowding ratio is extreme. For context, the 90th percentile threshold for |NC%OI| in this market over the past two years is roughly 25%.
Commercials hold +1,572,073 net longs (33.3% of OI), the opposite position at similar concentration. This is the normal state — large speculators and commercials typically oppose each other. No alignment signal is present here.
The 5Y Treasury shows a different profile. Net shorts reached -1,586,840 contracts, but the percentile rank is only 81st over two years and 16th over ten. This divergence matters. The 2Y short extreme is structurally unprecedented. The 5Y short position is elevated in the recent window but mid-range in longer context — it has precedent in earlier cycles. The 10Y sits at the 48th percentile (2Y) and 12th percentile (10Y). The 30Y is at the 59th and 60th percentiles. The Treasury curve positioning is not uniform.
FX: Australian Dollar Crowding at 99th Percentile
Australian Dollar net longs hit 81,506 contracts, the 99th percentile in both 2-year and 10-year windows. The Z-score of 2.65 reflects persistent conviction. The build speed is sustained: +174,731 above the 4-week average, +144,428 above the 13-week average. Non-commercial crowding is at 30.9% of open interest (263,962 contracts). Commercials hold -109,041 net shorts (-41.3% of OI), indicating intense hedging pressure on the other side.
Within the 104-week sample, AUD net positioning has only been higher in two prior weekly reports. Both occurred during periods where price confirmed the positioning trend for 4–6 weeks before mean reversion began. The current reading is extreme, but extreme positioning across both timeframes has historically sometimes persisted for multiple weeks when price continues to validate the crowded trade.
Canadian Dollar net shorts reached -32,684 contracts (91st percentile 2Y, 34th percentile 10Y). The 2-year extreme is notable, but the 10-year mid-range reading suggests this level has historical precedent. The crowding ratio is -15.1% of OI — meaningfully net short, though less concentrated than AUD's net long crowding.
The remaining eight FX markets tracked sit near median levels. Euro FX is at the 24th percentile (2Y), British Pound at the 11th, Japanese Yen at the 17th. No coordinated FX crowding is present beyond AUD and CAD.
Grains: Coordinated Long Crowding in the Soy Complex
Soybean Oil net longs reached 134,557 contracts (99th percentile 2Y, 98th percentile 10Y, Z-score 2.41). Soybean Meal hit 120,194 contracts (99th percentile 2Y, 90th percentile 10Y, Z-score 2.11). Cotton reached 48,387 contracts (98th percentile 2Y, 55th percentile 10Y, Z-score 3.24 — the highest Z-score in the dataset this week). Wheat HRW: 15,315 contracts (99th percentile 2Y, 65th percentile 10Y). Wheat SRW: -9,415 contracts (98th percentile 2Y — but note this is a net short position at the high end of the range, meaning shorts are crowded).
The soy complex crowding is particularly notable because it spans two related contracts. Both meal and oil are at the 99th percentile over two years, and open interest in each market is above 300,000 contracts. Non-commercial crowding ratios: soybean oil at 27.1% of OI, soybean meal at 19.8% of OI. These are concentrated trades.
Cotton's Z-score of 3.24 is the highest in the dataset, yet the 10-year percentile is only 55th. This suggests the current positioning level is unusual within the recent window but not unprecedented in historical context. The build has been sustained: +5,116 above the 4-week average, +40,749 above the 13-week average.
Cocoa sits at the 1st percentile (2Y) and 5th percentile (10Y), net short at -21,601 contracts. This is the inverse crowding scenario — speculative shorts are at an extreme.
Top 10 Most Extreme Positions
| Rank | Market | Net Position | 2Y Percentile | 10Y Percentile | Z-Score |
|---|---|---|---|---|---|
| 1 | 2Y Treasury | -1,637,324 | 1st | 1st | -2.99 |
| 2 | Australian Dollar | 81,506 | 99th | 99th | 2.65 |
| 3 | Bitcoin | 2,253 | 99th | 99th | 2.39 |
| 4 | Cocoa | -21,601 | 1st | 5th | -2.24 |
| 5 | Soybean Meal | 120,194 | 99th | 90th | 2.11 |
| 6 | Soybean Oil | 134,557 | 99th | 98th | 2.41 |
| 7 | Wheat (HRW) | 15,315 | 99th | 65th | 2.14 |
| 8 | Cotton | 48,387 | 98th | 55th | 3.24 |
| 9 | Wheat (SRW) | -9,415 | 98th | 56th | 1.62 |
| 10 | Silver | 23,904 | 5th | 29th | -2.00 |
Crypto and Metals: Bitcoin at 99th, Precious Metals Near Lows
Bitcoin net longs reached 2,253 contracts (99th percentile 2Y and 10Y, Z-score 2.39). The crowding ratio is 10.9% of open interest (20,626 contracts total). This is the highest net positioning in the 104-week dataset. Commercials hold -1,809 net shorts. The build has been sustained over multiple weeks.
Silver sits at the opposite extreme: 23,904 net longs, but at the 5th percentile (2Y) and 29th percentile (10Y). The Z-score is -2.00. This reflects speculative disengagement rather than aggressive short crowding. Gold is at the 8th percentile (2Y) and 32nd percentile (10Y), net long at 163,202 contracts — well below the full-period average of 231,282. Crowding ratios: silver at 20.8% of OI, gold at 45.2% of OI. Both metals show low speculative appetite relative to recent history.
Alignment Signal: Two Markets with BOTH LONG or BOTH SHORT
Two markets show the alignment signal this week — Large Speculators and Commercials both positioned on the same side, leaving Small Speculators holding the entire counterparty position.
30Y Treasury (BOTH SHORT): Large speculators hold -31,633 net shorts (-1.8% of OI). Commercials hold -69,728 net shorts (-3.9% of OI). Small speculators therefore hold the entire net long position. The COT percentile for 30Y is 59th (2Y) and 60th (10Y) — near mid-range, so this alignment is not coinciding with an extreme. Futures markets are zero-sum. When two groups align net short, the third group must hold 100% of the net long. Historically, similar configurations have sometimes coincided with periods of elevated positioning volatility for the retail-held side — though this is a structural observation, not a directional indicator.
S&P 500 (BOTH SHORT): Large speculators hold -38,787 net shorts (-2.0% of OI). Commercials hold -31,371 net shorts (-1.6% of OI). Small speculators hold the entire net long. The COT percentile is 70th (2Y), slightly elevated but not extreme. This alignment check is a positioning concentration flag, not a standalone directional indicator.
"So What?" — Positioning Scenarios
2Y Treasury (Net Short Extreme: 1st Percentile)
If 2Y yields continue to rise while net shorts remain below the 5th percentile for multiple weeks, then this is consistent with a sustained directional conviction trade — historically, when builds of this magnitude (sustained over 13+ weeks) have persisted at extreme percentiles, positioning has sometimes remained at extremes for 6–10 weeks before mean reversion began.
If yields stall or reverse while the non-commercial crowding ratio remains above 30% of open interest, then this may indicate elevated unwind risk — concentrated net short positions (above -30% of OI) have historically sometimes been followed by rapid repositioning when price stops confirming the trade, though timing cannot be determined from COT data alone.
If net shorts begin to cover quickly (4-week deviation narrows by >200k contracts) without a corresponding yield move, then this may reflect tactical repositioning or month-end book squaring rather than structural conviction change — speed of repositioning relative to price action is a key distinction between noise and trend shifts.
Australian Dollar (Net Long Extreme: 99th Percentile)
If AUD/USD price continues to make new highs while net longs persist above the 95th percentile, then this is consistent with trend confirmation — the sustained build over 13 weeks (vs 4-week and 13-week averages both showing large positive deviations) suggests structural positioning rather than tactical entry, and historically, such builds have sometimes continued for multiple weeks when price validates the trade.
If price stalls while the non-commercial crowding ratio remains above 30% of open interest, then this may indicate exhaustion risk — concentrated net long positions at the 99th percentile in both 2-year and 10-year windows have historically sometimes preceded positioning unwinds when price momentum fades, though the configuration can persist.
If commercials continue to add net shorts (currently -41.3% of OI) while speculators hold steady, then this may reflect intensifying hedging pressure — in past cycles, when commercial hedging exceeded 40% of OI while speculators were at extremes, this has sometimes been consistent with a structural imbalance that eventually required positioning adjustment, though the timing has varied widely.
Cotton (Net Long Extreme: 98th Percentile, Highest Z-Score)
If price continues to trend higher while the Z-score remains above +3.0, then this is consistent with a high-conviction macro trade — the Z-score of 3.24 is the highest in the dataset this week, yet the 10-year percentile is only 55th, suggesting this positioning level has historical precedent and may not be as unusual as the recent window implies.
If the crowding ratio (currently 13.9% of OI) increases further while price fails to confirm, then this may indicate late-cycle crowding — when Z-scores exceed +3.0 and crowding ratios push above 15% of OI, historically this has sometimes coincided with positioning vulnerability, though it is not predictive on its own.
If the build speed shifts (4-week deviation currently +5,116, well below the 13-week deviation of +40,749), then this may suggest the positioning trend is decelerating — a narrowing 4-week deviation relative to the 13-week average has historically sometimes preceded mean reversion in other markets at similar extremes, though the pattern is not consistent across all cycles.
Key Questions for Next Week
Will 2Y Treasury net shorts extend below -1.7M contracts, or does the 1st percentile mark a positioning floor? The current reading is the lowest in the 520-week historical series — further extension would be structurally unprecedented.
Do soybean oil and soybean meal net longs remain coordinated at the 99th percentile, or does one contract begin to diverge? Coordinated extremes in related contracts have historically sometimes unwound in sequence rather than simultaneously.
Does the alignment signal in 30Y Treasury and S&P 500 persist into next week's report, and do the COT percentiles shift toward extremes? The signal is amplified when it coincides with 90th+ or 10th- percentile readings.
This analysis is for educational purposes only and does not constitute financial advice.
Check back next Friday for the latest COT report analysis covering the week ending April 7, 2026.
Explore More: Interactive Dashboard | Complete Methodology | Download Data | Previous Analysis