CFTC S&P 500 Positioning vs SPY

Cash SPY tells you where price is. CFTC managed-money net positioning in the E-mini S&P 500 contract tells you how the leveraged side of the buyer base is paying for that price. The combined report (futures plus options, reduced to E-mini-equivalent contracts) covers the standard, E-mini and Micro E-mini complex and is published every Friday for the prior Tuesday's open interest, with a 3-day lag that the market has worked into the read since the Disaggregated COT format was introduced in September 2009. Goldman's Prime Brokerage cross-asset desk and Bank of America's Hartnett-led Flow Show have both used this series as the equity leg in their cross-asset positioning composites since at least 2014. The institutional thesis the pair tests is whether speculator conviction confirms or contradicts the cash trend, with the recurring breakpoint being the January 26, 2018 reading: asset managers ran their longest E-mini exposure on record (roughly 1.05 million contracts) into the Volmageddon week of February 5, 2018, where the VIX spiked from 17.31 to 37.32 in a single session and SPY fell 4.1 percent. That single episode is the textbook case for why a record-positive positioning read into a 9.14 VIX low (the November 3, 2017 close, an all-time low at the time) is itself a setup, not a confirmation. Convex tracks net positioning normalised by total open interest rather than absolute contracts, because the E-mini complex has roughly tripled in open interest since 2010 and absolute contract levels are not comparable across the full series.

Three positioning extremes anchor the regime atlas. The January 26, 2018 record long preceded the Volmageddon drawdown that took SPY from 286.58 (January 26, 2018 close) to 252.40 (February 8, 2018 close), a 12 percent drawdown in nine sessions and the largest one-day move since 2011 on February 5. The March 17, 2020 COT unwind, captured against the COVID liquidation that drove SPY from 339.08 (February 19, 2020 close) to 222.95 (March 23, 2020 close, a 34 percent drawdown), saw asset managers cut net long exposure by nearly 250,000 contracts in three weeks while leveraged funds flipped to net short, the fastest documented positioning unwind in the series. The October 4, 2022 reading produced the largest commercial-hedger short surge on record at the index level, coinciding with the cycle low of 348.11 (October 12, 2022 SPY close, down from 477.55 January 4, 2022). Each episode shares one feature: positioning hit a percentile extreme (above the 90th or below the 10th over a rolling five-year window) before price did. The January 2020 net long peak of 1.101 million contracts sat above the 99th percentile of the 2010-2020 window and the COVID drawdown began three weeks later. Treating the COT report as a leading indicator works at extremes; treating it as a leading indicator at typical readings produces noise, because mid-percentile positioning has roughly the same forward-return distribution as the unconditional sample.

The transmission channel from CFTC positioning to SPY runs through dealer gamma and through the basis between the front-month E-mini and cash. When asset managers and leveraged funds carry a large directional position, the dealer block warehousing the other side hedges in cash equity, primarily through the SPDR creation/redemption mechanism for SPY and through systematic delta-hedging of index options. The April 22, 2026 reading of approximately 145,000 contracts net long across managed money is mid-percentile (52nd percentile over the 2015-2026 window), which is why the spread between positioning and price has been stable through the first quarter of 2026: the dealer book has neither a stretched short-gamma profile nor a meaningfully constrained inventory. The cleanest historical example of the mechanism breaking is March 2020, when liquidity in the cash SPY/E-mini basis decoupled by as much as 80 basis points intraday on March 18, 2020, because primary dealers hit balance-sheet capacity before redemptions could clear. Bloomberg's Prophet desk later reconstructed the basis dislocation: the COT had already shown the unwind starting in the Tuesday March 17 release, but cash SPY did not bottom until six trading days later. The 6-day lag during that window is the largest in the post-2010 sample and remains the worst-case calibration point for the mechanism. The 2018 episode produced a smaller but related dislocation: the SPDR creation-unit basket traded at a 12 bp premium to cash SPY on February 5, 2018 as authorised participants struggled to source the underlying basket during the volatility spike.

The Convex Risk Appetite Index (CRAI) consumes the CFTC E-mini net long figure as one of seven equity-positioning inputs, alongside ICI mutual-fund flows, Bank of America Bull-Bear, AAII sentiment net bull-bear, NAAIM exposure, dealer gamma estimates, and the put-call ratio. The CFTC leg carries the highest weight (28 percent) inside the equity-positioning bucket because it has the longest unbroken weekly history (back to 1992 for the standard contract, 2009 for the disaggregated breakdown) and it is the only series where positions are matched against required regulatory disclosure rather than survey response. The April 22, 2026 reading is consistent with CRAI's neutral classification of the equity positioning regime: not extended (no contrarian short setup), not capitulated (no contrarian long setup), and consistent with SPY's mid-cycle behaviour through 2026. The most informative episodes for the pair are the disagreement events, where positioning and price-action signals point opposite directions. The clearest example was the August 5, 2024 unwind: SPY fell 3.0 percent intraday on the Yen-carry liquidation, but the CFTC report two business days later showed managed-money long exposure largely intact, predicting (correctly, in retrospect) that the move would not extend beyond the initial volatility shock. The August 5, 2024 VIX intraday peak hit 65.73 before retracing to close at 38.57, a single-session move comparable to February 2018, but with positioning resilient and SPY recovering within nine sessions, the CRAI cross-check filtered what would otherwise have looked like a regime break.

The April 22, 2026 read is mid-range and consistent with SPY's $568.40 close. The actionable framework is to act on the pair only when both legs signal in alignment: positioning above the 90th percentile coupled with SPY within 3 percent of a multi-year high is the contrarian short setup; positioning below the 10th percentile coupled with SPY in a defined downtrend (50-day moving average below the 200-day) is the contrarian long setup. Between those extremes the pair carries weak signal and the position size should reflect that. Historical base rate at the contrarian-short setup, defined over the 2010-2025 sample, is a 9.4 percent average drawdown over the subsequent 90 days against an unconditional 90-day path of plus 1.8 percent, with the worst single instance being the January-February 2018 episode (negative 12.0 percent drawdown). Historical base rate at the contrarian-long setup is a plus 14.2 percent rally over the subsequent 90 days against the same unconditional baseline, with March 2020 producing the strongest signal (plus 24.7 percent in 60 days from the COT trough). The horizon for the signal is one to three months; using it on a daily basis produces noise because the COT release cadence is itself weekly. The asymmetry is itself notable: contrarian-long setups have produced higher absolute returns than contrarian-short setups have produced absolute drawdowns, consistent with the broader equity-market drift, and the position-sizing implication is to weight the long side more aggressively when it triggers.

Three indicators trip the pair from its current neutral classification into an actionable regime. First, a managed-money net long reading above 700,000 contracts in any single COT release would push positioning above the 90th percentile and combine with SPY's current proximity to all-time highs to trigger a contrarian-short setup. Second, a leveraged-funds net short reading exceeding 200,000 contracts (the configuration last seen in October 2022 at the cycle low) would put the pair in contrarian-long territory if it occurred during a defined SPY downtrend. Third, the spread between asset managers (typically structurally long) and leveraged funds (typically more nimble) is itself a regime indicator; when the two cohorts disagree by more than 1 standard deviation versus their pair-correlation history, the next four-week price action has historically been more volatile (realised volatility 22 to 28 percent) than when they agree (realised volatility 12 to 16 percent). Convex updates the spread reading every Friday at 3:30 pm ET when the COT release lands, and surfaces the percentile rank against the post-2010 distribution rather than the absolute contract count, which is the only consistent way to compare across the structural growth in open interest. Total open interest in the E-mini complex grew from approximately 3.0 million contracts in 2010 to over 9 million in 2024, making absolute-contract comparisons across the full 1992-2026 series structurally misleading.