Benchmark · weekly · index (0–100)

Cross-asset speculative crowding

How crowded is speculative positioning across equity, FX, rates and commodity futures simultaneously?

The cross-asset speculative crowding index aggregates CFTC positioning benchmarks into one gauge of how extended speculative positioning is across futures markets at the same time. Each market's primary cohort is ranked against its own trailing five-year history; the distance from the median is averaged within each asset class and then equal-weighted across the four classes. A reading of 100 means the average constituent is at a historical extreme; 50 means the average is around its 25th or 75th percentile; near 0 means positioning is close to median across the board.

Readingsweekly

Overall crowding: 54.47; 4 August 2026
13-week change in index points: +0.33 pts; from 54.14
Most crowded asset class: FX; 65.58
Least crowded asset class: Equities; 43.86
Overall historical percentile: 54th; of 740 readings

Overall crowding

Equities

Rates

Commodities

Overall crowding median 53.70

Cross-asset speculative crowding: summary statistics (Max range)
Series	First	Latest	Min	Max
Overall crowding	57.56	54.47	37.52	70.86
Equities	30.60	43.86	7.19	94.83
FX	76.75	65.58	14.99	83.73
Rates	47.24	56.07	19.52	98.35
Commodities	75.65	52.39	20.40	90.75

Source: CFTC
Series: composite:overall composite:equities composite:fx composite:rates composite:commodities
Frequency: weekly
Data through: 4 August 2026
Refreshed: 11 Aug 2026

Copy as markdown

How this is calculated

Formula

2 × |percentile(primary Net/OI, trailing 260 weeks) − 50|, averaged per asset class, then equal-weighted across the four classes

Average positioning extremity across equity, FX, rates and commodity futures, with four asset-class subindices. Each contract's primary speculative cohort (leveraged funds for TFF, managed money for commodities) is ranked against its own trailing five-year history: the percentile of the latest Net/OI using only observations on or before each date, with a minimum of 104 prior observations. Crowding is twice the distance of that percentile from 50, so 0 marks a median reading, 100 a historical extreme, and roughly 50 marks the average constituent at the 25th or 75th percentile. Each asset-class subindex is the mean of its contracts' crowding scores, computed only when at least 60% of its intended contracts have valid scores. Overall is the equal-weighted mean of the four available class subindices. The eligible universe expands as contracts reach the 104-observation minimum, so early-sample readings reflect fewer constituents.

As of 4 August 2026, the latest reading is 54.47 index (0–100). That is down 2.4 index points over the past year and above its long-run median of 53.70 index (0–100).

Source series

CFTC

Plottedcomposite:overallOverall crowdingindex (0–100) · weekly · Not seasonally adjusted Plottedcomposite:equitiesEquitiesindex (0–100) · weekly · Not seasonally adjusted Plottedcomposite:fxFXindex (0–100) · weekly · Not seasonally adjusted Plottedcomposite:ratesRatesindex (0–100) · weekly · Not seasonally adjusted Plottedcomposite:commoditiesCommoditiesindex (0–100) · weekly · Not seasonally adjusted InputCFTC:COMPOSITE

How to read it

What a high or low reading means

A high reading means speculative positioning is historically extreme across many futures markets at once, typically during trending markets where consensus has built up on one side, marking periods of elevated reversal risk. A low reading means positioning is near median across markets and the index carries little contrarian signal. The scale runs 0 to 100, where 100 is a historical extreme and 50 is roughly the 25th or 75th percentile of the average constituent's own range.

How the index is built

Each constituent's primary cohort Net/OI is ranked against its own trailing five-year history using only past observations (no look-ahead), with a 104-week minimum before any score is emitted. Crowding is twice the distance of that percentile from 50, so extremes in either direction count equally. Each asset-class subindex averages its contracts' scores and requires at least 60% coverage; overall is the equal-weighted mean of the four available subindices, so FX's larger contract count does not dominate.

Why the eligible universe expands over time

A contract only contributes to the index once it has 104 weeks of history (the minimum for a stable percentile). Several contracts begin later than others (Russell 2000 in 2017, SOFR in 2018, Ultra-10Y in 2016), so early-sample readings reflect fewer constituents. The number of contributing contracts for each week is stored in the derived data and documents this expansion.

Limitations

The index averages net positioning and ignores gross exposure, spreads and option delta, so it understates crowding where speculators hold large offsetting positions. It covers the most-traded US futures markets but excludes many smaller markets where positioning can be more extreme. The rolling five-year window means the index's baseline shifts as old observations age out. Treat the chart as a broad crowding gauge, not a timing signal.

How this benchmark is used

Cross-asset risk regime assessment

Macro risk desks use the index as a regime label: high readings mark periods when a single narrative has pulled speculators into crowded positions across markets, raising the odds of a correlated unwind. It is one input to risk-budgeting and tail-hedging decisions, not a standalone trigger.

Asset-class crowding comparison

The four subindices reveal whether crowding is broad-based (many classes extended) or narrow (one class at an extreme while others are quiet). Comparing the subindices over time shows when positioning-driven risk is building or abating in specific parts of the futures complex.

Frequently asked questions

7 answers

What is the current cross-asset speculative crowding?

As of 4 August 2026, the latest reading is 54.47 index (0–100). That is down 2.4 index points over the past year and above its long-run median of 53.70 index (0–100).

How often is this benchmark updated?

This benchmark is built on weekly data. The page is refreshed when the source publishes new observations; the freshness block below the chart shows the exact data-through date and when PIER20 last fetched the file.

What data sources does this chart use?

The chart is built from Overall crowding, Equities, FX, Rates and Commodities, sourced from the CFTC Commitments of Traders report.

What does the cross-asset speculative crowding index measure?

How extended speculative positioning is across equity, FX, rates and commodity futures at the same time. Each market's primary cohort is ranked against its own trailing five-year history, the distance from the median is averaged within each asset class, and the four classes are equal-weighted into an overall index from 0 to 100.

What is a high reading?

The index runs 0 to 100. A reading of 100 means the average constituent is at a historical extreme; 50 means the average is around its 25th or 75th percentile; near 0 means positioning is close to median across the board. Readings in the high 50s and above mark episodes of broad cross-asset crowding.

How is this different from looking at any single market's positioning?

A single market only tells you about that market. The index aggregates across 24 positioning benchmarks, so it rises only when many markets are extended at once. A crowded yen position alone moves the index only a little; crowded yen, gold, bonds and equities together move it a lot.

Does a high reading mean a crash is coming?

Not directly. High readings mark elevated reversal risk because crowded positions unwind violently when the consensus breaks, but the trigger is usually external. The index is a risk-regime indicator, not a timing signal.

MarkdownMachine-readable version of this benchmark