A consensus momentum trade experienced a "collapse-style" reversal on February 5th. Morgan Stanley believes that concentrated long position unwinding, combined with forced selling from leveraged ETF rebalancing, rapidly amplified the market decline. The retail investor buying that typically provides a cushion during downturns was notably absent, leading to insufficient marginal demand. This sell-off displayed a structural characteristic of a "relatively restrained index paired with intense internal market movements": selling pressure was highly concentrated in the Nasdaq, technology, and semiconductor sectors—previously leading high-beta themes with crowded positioning (such as AI, national security, and Bitcoin mining stocks). In contrast, cyclical sectors, chemicals, and banking performed more robustly, indicating a significant rotation.

For the near-term outlook, a technical rebound is possible, but Morgan Stanley favors a "sell-the-rally" approach. The reasoning is that the deleveraging and rebalancing chain may not be over, and retail capital is likely to remain weak ahead of the tax season, making the next wave of selling pressure harder to absorb. Historical statistics also show that following a momentum-driven sell-off of similar intensity, stock performance tends to be negative over the subsequent one to two months.

While the index volatility appeared relatively contained, internal sector movements were severe. Goldman Sachs pointed out that at one point during the session, nearly three-quarters of stocks were outperforming the S&P 500, even as the index itself declined. This reflects that the pain was concentrated in a small number of crowded momentum trades (which involve buying rising assets and selling falling ones based on the expectation that trends will continue). Goldman Sachs' trading desk attributed this to a reversal triggered by rising volatility and technical deviations, rather than a single fundamental catalyst. Morgan Stanley trader Bryson Williams also noted in his closing remarks that the day's action resembled a VaR shock-induced position rebalancing, with selling primarily stemming from concentrated unwinding of stocks that had rallied sharply early in the year, rather than a symmetrical deleveraging of both long and short positions.

The momentum collapse involved concentrated unwinding of crowded positions, putting pressure on AI and other high-beta themes. Data from Morgan Stanley showed its long/short momentum index (MSZZMOMO) fell approximately 7.7% in a single day, an extreme move of about four standard deviations, primarily driven by the long side. The momentum long index (MSQQUMOL) fell about 5.7%, while the momentum short index (MSQQUMOS) rose only about 1.9%, indicating a one-sided adjustment focused on cutting long positions. Thematically, selling pressure was heavily concentrated in previously leading, crowded sectors: AI, national security, and Bitcoin mining stocks were at the core of the decline. Meanwhile, cyclical stocks, chemicals, and banking sectors relatively outperformed, forming an "anti-AI" style rotation.

Leveraged ETF rebalancing amplified the volatility and poses a risk of ongoing selling pressure. According to calculations by Morgan Stanley's quantitative team, rebalancing by leveraged ETFs, which generated approximately $18 billion in selling pressure on US stocks, was a key driver of the day's decline (as leveraged ETFs must sell stocks to maintain their target leverage ratio when the underlying assets fall). This supply was concentrated in the Nasdaq, technology, and semiconductor sectors, impacting several popular individual stocks with single-day selling exceeding $100 million. More importantly, this type of selling pressure is not necessarily over. Morgan Stanley indicated that with volatility remaining elevated and stock leverage still near the high end of its historical range, the risk of further selling from leveraged ETF rebalancing persists over the coming week, with an estimated scale of around $10 billion. If deleveraging continues, volatility in highly leveraged stocks is likely to be amplified further. Additionally, although option market makers still hold a positive gamma position (buying on dips, selling on rallies), this exposure has declined significantly. Combined with the negative gamma effect from leveraged ETFs (buying high, selling low), the market overall is nearing a "net negative gamma" state—a capital structure that can exacerbate volatility and create a feedback loop for selling pressure.

The absence of retail buying created a vacuum in marginal demand, raising concerns ahead of the tax season. Historically, retail investors' dip-buying has often acted as a buffer during price declines, but their participation was clearly missing this time. Morgan Stanley reported that net buying intensity from retail investors was low (ranking in the bottom 16% of days over the past year), occurring mainly during the afternoon market rebound. Concurrently, institutional net selling intensity was high (ranking in the top 11% of days over the past year), concentrated primarily in the morning session. This left consensus long positions without crucial support at a key moment. The weakness in retail activity also directly suppressed the momentum factor. Morgan Stanley noted a remarkably high overlap currently between the stocks retail investors favor and those used in momentum long/short strategies; consequently, when retail buying dries up, momentum strategies are dragged down alongside. Morgan Stanley also highlighted a seasonal pattern: returns on stocks popular with retail investors in January often show a negative correlation with returns in February and March, historically coinciding with liquidity pressures related to the tax season.

Looking ahead, while a rebound may materialize, Morgan Stanley advises selling into any strength. Historical analysis from Morgan Stanley shows that when the MSZZMOMO index experiences a single-day decline of 7% or more, accompanied by weakness on the long side (MSQQUMOL down 5% or more), it typically signals negative performance in the following days. The median performance over the subsequent one to two months is usually negative, turning positive only by the third month. The median peak-to-trough decline for MSZZMOMO is around 22%. Within this pricing framework, while a short-term technical bounce is possible—and Goldman Sachs' trading desk suggests such significant momentum pullbacks can offer medium-term buying opportunities—Morgan Stanley's assessment remains more cautious. The core reasons are that positioning has not fully unwound, systematic supply has yet to fully materialize, and the continued "absence" of retail investors as marginal buyers could make the next wave of selling pressure more difficult to absorb.