How Active Investor Positioning Can Explain (and Predict) Mega-Cap Moves
Tracks Active Long Capital: The Long-to-Short (LTS) Ratio highlights where institutional investors are putting real money to work, measured by Active Long Interest value versus Short Interest value.
Reveals Crowded Winners: When normalized by stock-specific ownership volatility, LTS surfaces names that are unusually heavily owned.
Rotates Toward Leadership: Within the Mag-7, the stock with the highest LTS z-score (above +1.5 std. dev.) triggers a signal — flagging where positioning and momentum are most aligned. We call this the Mag-7 Crowded Momentum Model.
Delivers Outperformance: Since 2020, the strategy has triggered 790 times, delivering an average 20-day return of +4.1%, compared to a +3.0% baseline across all Mag-7 trading days. NVDA is the current signal, having triggered 17 times year-to-date with an average forward return of +4.8% and a 65% win-rate, versus +4.1% and 58%, respectively, for all 20-day periods YTD.
Flexible Framework: While a Mag-7 rotation model is one use case, LTS works across sectors and styles, with the impact of Active Long Interest and Short Interest illuminating the most dynamic exposures in the market.
The Long-to-Short Ratio (LTS), calculated as Active Long Interest market value divided by Short Interest notional value, offers a direct lens into institutional positioning. When normalized by stock-specific volatility using a 6-month trailing z-score, LTS becomes a powerful tool to identify emergent conviction expressed through the combined lens of active long and short positioning.
This note focuses on one application: Crowded Momentum, a strategy that flags days when a Mag-7 stock is heavily owned (LTS z-score > +1.5 standard deviations) relative to its own historical levels. The signal is designed to capture continuation in institutional favorites, with the results amongst the Mag-7 suggesting not to fade crowding but to lean into it when the tape supports it. In the event multiple Mag-7 constituents generate signals on the same day, the stock with the largest deviation from its average LTS is selected.
Since 2020, the strategy has triggered 790 times, providing a large and statistically useful sample. Across those signals, the average 20-day forward return was +4.1%, compared to +3.0% across all Mag-7 trading days. Stocks like NVDA, MSFT, AMZN, and AAPL delivered especially strong results, while GOOGL remains a notable exception with negative returns for post-signal periods. This dispersion highlights how LTS-based signals offer not only direction but differentiation — flagging which names are most likely to lead and which may be lagging despite similar index membership.
The most recent signal is NVDA, which has returned an average of 4.8% over the 20 days following a Mag-7 Crowded Momentum signal in 2025 year-to-date (17 occurrences). This compares to a 4.1% average return across all 20-day periods. Following the signal, returns were positive 65% of the time, versus 58% across all 20-day windows.
Want to know more? Access this data in real time using S3’s BLACK APP & BLACK MAP
The information herein (some of which has been obtained from third party sources without verification) is believed by S3 Partners, LLC (“S3 Partners”) to be reliable and accurate. Neither S3 Partners nor any of its affiliates makes any representation as to the accuracy or completeness of the information herein or accepts liability arising from its use. Prior to making any decisions based on the information herein, you should determine, without reliance upon S3 Partners, the economic risks, and merits, as well as the legal, tax, accounting, and investment consequences, of such decisions.