COT May 5 – Treasury Curve Split & Ag Crowding

Treasury positioning has fractured across the curve in a way I haven't seen in this dataset. The 30Y is at the 1st percentile—essentially the most net short hedge funds have been in two years. The 5Y is at the 98th percentile—among the most net short they've been, but in the opposite structural direction when you zoom out to ten years.

For those new to COT data: these percentiles measure where current speculative positioning sits relative to the past two years (primary window) and past ten years (structural context). When a market hits the 99th percentile, it means speculators are more net long than 99% of all prior weeks in that sample. When it hits the 1st percentile, they're more net short than 99% of weeks. It's a crowding gauge, not a price signal.

This analysis uses data from the CFTC Commitment of Traders Report, updated weekly. Full methodology is here. You can download this week's data (JSON) or view the interactive dashboard.

Top 10 Extreme Positions This Week

Here's where positioning crowding is most concentrated across the 43 futures markets I track:

Rank	Market	Net Position	2Y Percentile	10Y Percentile	Z-Score
1	30Y Treasury	-172,942	1st	8th	-2.93
2	Australian Dollar	+78,674	99th	99th	+2.22
3	Cotton	+100,682	99th	89th	+4.26
4	Wheat (HRW)	+14,371	99th	64th	+1.87
5	5Y Treasury	-1,421,299	98th	20th	+1.54
6	Brazilian Real	+66,797	98th	99th	+1.37
7	Soybean Oil	+168,887	98th	99th	+2.59
8	Soybean Meal	+134,714	97th	94th	+2.03
9	Soybeans	+232,198	97th	94th	+1.75
10	2Y Treasury	-1,673,329	4th	1st	-2.58

Treasury Curve: Structural Contradiction

The 30Y is net short 172,942 contracts—1st percentile over two years, 8th over ten. Non-commercial traders hold a net short position equal to 9.5% of open interest. That's not massive crowding by absolute standards, but it's at the low end of the range for this contract.

The 5Y is net short 1.42 million contracts—98th percentile over two years. But here's the contradiction: it's only at the 20th percentile over ten years. Speculators are heavily net short by recent standards, but this positioning level has plenty of precedent in earlier rate cycles. The net short represents 21.5% of open interest, which is concentrated.

The 2Y is at the 4th percentile over two years and the 1st over ten. Net short 1.67 million contracts. This one is structurally extreme across both windows.

What you're seeing is not a uniform Treasury view. It's a curve bet. Specs are most short at the front end, less short in the middle, least short at the long end. Whether that reflects rate path expectations or just crowded tactical positioning from earlier in the year, I can't tell you from this data alone.

Ag Complex: Broad Net Long Crowding

Fourteen markets hit the 95th or 5th percentile this week. Four of them are in the soy complex.

Soybeans: 97th percentile over two years, 94th over ten. Net long 232,198 contracts, representing 26.1% of open interest. Soybean meal: 97th and 94th. Net long 134,714, or 30.9% of OI. Soybean oil: 98th and 99th. Net long 168,887, or 51.4% of OI. That last one is the most concentrated—more than half of open interest is held net long by non-commercials.

Cotton is at the 99th percentile over two years (89th over ten), with a Z-score of +4.26—the most extreme reading in the dataset this week by that metric. Net long 100,682 contracts, or 30.3% of OI. Within this 104-week sample, readings beyond +4.0 have been uncommon.

Hard red winter wheat is at the 99th percentile over two years, though only the 64th over ten. Net long 14,371. This is a recent build, not a structural extreme.

The pattern is sustained accumulation. Compare the 4-week and 13-week deviations for soybeans: +229,934 vs. +244,606. That's consistent positioning growth over three months, not a single-week spike.

FX: Australian Dollar and Brazilian Real Both at Extremes

The Aussie dollar is at the 99th percentile in both the 2-year and 10-year windows. Net long 78,674 contracts, representing 28.1% of open interest. Commercials are net short 105,545 contracts (37.8% of OI), which is elevated hedging intensity. This is a structural extreme with conviction across both lookback periods.

The Brazilian real is at the 98th percentile over two years and 99th over ten. Net long 66,797 contracts, or 50.9% of OI. That's more concentrated than the Aussie on a relative basis, even though the absolute net position is smaller.

Both builds have been sustained. The Aussie's net position is +137,330 above the four-week average and +113,092 above the 13-week. The real is running +66,797 net vs. a full-period average of +51,347.

Equity Index: Nasdaq and Dow Both Net Short, Alignment Signal Active

The Nasdaq 100 is at the 4th percentile, net short 1,832 contracts. That's not a huge absolute number, but here's what makes it notable: both large speculators and commercials are net short. That means small speculators—retail public traders—hold the entire net long position as the counterparty to both groups.

Futures markets are zero-sum. When two groups align on one side, the third group holds 100% of the other side by mechanical necessity. In this case, small specs are long while both hedge funds and commercials are short.

The Dow shows the same structure—both large specs and commercials net short—but the Dow is only at the 34th percentile, so the positioning level itself isn't extreme. The Nasdaq's 4th percentile reading amplifies the signal.

Historically, when large speculators and commercials have aligned on the same side while positioning reaches an extreme percentile, small speculator positions have sometimes coincided with periods of elevated positioning volatility. This is a configuration flag, not a timing tool. It tells you the structure is unusual, not what happens next.

What Different Scenarios Might Mean

For 30Y Treasury (1st percentile, sustained net short build):

• If the net short position extends further while long-end yields continue rising, then this is consistent with a momentum trade that has worked and may attract additional tactical shorts—though the 2-year extreme suggests limited room before positioning becomes unusually concentrated even by long-term standards.
• If yields stabilize or fall while positioning remains at the 1st percentile, then this may indicate speculative shorts are offsides and vulnerable to a technical unwind—particularly given that the net short represents 9.5% of open interest, meaning unwinding flows could move price if they occur rapidly.
• If positioning drifts higher (less net short) without a major yield move, then this may reflect routine monthly repositioning or expiry-driven adjustments rather than a directional macro bet—speed of the unwind would distinguish between the two.

For Soybean Oil (98th percentile both windows, highest NC crowding ratio in the ag complex):

• If the net long position persists into June while price continues to confirm, then this is consistent with a structural fund positioning tied to biodiesel policy expectations or weather risk—the 13-week sustained build (+197,294 above the 13-week average) suggests this is not a tactical punt.
• If price stalls or reverses while positioning remains at 98th percentile and 51.4% of OI, then this may indicate a crowded long trade vulnerable to cascading exits—particularly given that commercials hold a net short of 64.8% of OI, which is unusually concentrated hedging pressure on the opposite side.
• If net longs begin unwinding quickly (a sharp drop in the next 2-3 reports) without major price disruption, then this may reflect expiry rollover mechanics or fund rebalancing rather than a change in the fundamental view—watch whether OI contracts or just net positioning shifts.

For Nasdaq 100 (4th percentile, alignment signal active):

• If both large specs and commercials extend their net short positions while small specs hold net longs and price declines, then this is consistent with a scenario where retail-held longs are stopping out into a down move—the alignment structure means small specs have no natural counterparty buffer if they liquidate.
• If the alignment persists into next week's report while price stabilizes or rallies, then this may be consistent with a structural positioning condition that has historically sometimes coincided with small speculator liquidation or capitulation—though this is a configuration flag, not a directional indicator, and the timing cannot be determined from positioning data alone.
• If large specs or commercials flip to net long in the next 1-2 reports, then the alignment breaks and the configuration is no longer mechanically notable—this would return the market to its normal three-way tug-of-war and reduce the structural significance of the current setup.

This analysis is for educational purposes only and does not constitute financial advice.

Next week's report covers the period ending May 12, 2026.

Explore More: Interactive Dashboard | Methodology | Download Data | Previous Analysis