Momentum consensus trading experienced a "collapse-style" reversal on February 5th. Morgan Stanley believes that concentrated selling by crowded long positions, combined with passive selling from leveraged ETF rebalancing, rapidly amplified the losses. The retail investor buying that typically provides a cushion during declines was noticeably absent, leading to insufficient marginal demand.
This downturn exhibited a structural characteristic of "relative restraint in the indices, but intense internal market movement." 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 strongly, showing clear signs of rotation.
Looking ahead, a short-term technical rebound is possible, but Morgan Stanley favors a "sell-the-rally" approach. The reasoning is that the deleveraging/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 indicate that following a momentum-driven single-day sell-off of similar intensity, stock performance tends to be negative over the subsequent one to two months.
While index volatility appeared relatively contained, internal sector movements were severe. A notable feature of this decline was the "divergence between the index and its components." Goldman Sachs pointed out that at one point, nearly three-quarters of stocks were outperforming the S&P 500, even as the index itself trended lower, reflecting that the pain was concentrated in a few crowded momentum trades (which involve buying rising assets and selling falling ones, based on the belief that trends will continue).
Goldman's trading desk attributed this to a reversal driven by increased volatility and technical deviations, rather than a single fundamental "trigger." Morgan Stanley trader Bryson Williams also noted in his closing remarks that the day's action resembled a VaR shock triggering position rebalancing, with selling primarily coming from concentrated reductions in positions that had seen significant gains since the start of the year, rather than a symmetrical deleveraging of both long and short positions.
The momentum collapse saw crowded holdings being sold off heavily, with AI and high-beta themes under pressure. Morgan Stanley data showed that the long-short momentum index MSZZMOMO fell approximately 7.7% in a single day, an extreme move of about four standard deviations, primarily dragged down 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 characterized by "cutting longs, not shorts."
From a thematic perspective, selling pressure was highly concentrated in previously leading, crowded sectors: AI, national security, and Bitcoin mining stocks—all high-beta targets—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 continued 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 multiples when prices 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 over. Morgan Stanley notes that with volatility remaining elevated and stock leverage still near the upper 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 further amplified.
Additionally, although options market makers remain in a net positive gamma position (buying on dips and selling on rallies), this exposure has declined significantly. Combined with the negative gamma effect from leveraged ETFs (which buy on strength and sell on weakness), the overall market is nearing a "net negative gamma" state. This fund structure can exacerbate volatility and create a feedback loop with selling pressure.
The absence of retail buying has created a vacuum for marginal buyers and raises concerns about demand ahead of the tax season. Historically, retail investors' tendency to buy the dip has acted as a buffer during price declines, but this time their participation was markedly lacking. Morgan Stanley stated that net retail buying intensity was low (only 16% of days in the past year saw lower intensity), with most activity occurring during the afternoon market rebound.
Conversely, institutional net selling intensity was high (only 11% of days in the past year saw higher intensity), primarily occurring during the morning session. This left consensus long positions without crucial support at key moments.
Weak retail participation also directly suppressed the momentum factor. Morgan Stanley pointed out that the stocks favored by retail investors now show an unusually high overlap with those used in momentum long/short strategies. Therefore, when retail buying dries up, momentum strategies are pulled down together. Morgan Stanley also highlighted a seasonal pattern: stocks popular with retail investors often see January returns negatively correlated with February and March returns, historically coinciding with liquidity pressures from the tax season.
Regarding the outlook, a rebound may occur, but Morgan Stanley advises selling into it. Historical statistics from Morgan Stanley indicate that when the long-short momentum index MSZZMOMO experiences a single-day decline of 7% or more, accompanied by weakness on the long side (MSQQUMOL down 5% or more for the day), it typically serves as a negative signal for the following days. Median price performance tends to be negative over the next one to two months, turning positive only in the third month. The median peak-to-trough decline for MSZZMOMO is approximately 22%.
Within this pricing framework, a short-term technical rebound in the market is possible. Goldman Sachs' trading desk suggests that such significant momentum pullbacks may present buying opportunities in the medium term. However, Morgan Stanley maintains a more cautious stance, emphasizing that positioning has not fully unwound, systematic supply has yet to materialize completely, and the continued "absence" of retail investors as marginal buyers could make the next wave of selling pressure more difficult to absorb.