Momentum Selloff: Stay In Or Step Aside?
Goldman's "unconstrained momentum" basket yesterday logged its worst day since 2022, in what traders described as a sharper shock than the post-DeepSeek AI de-risking episode last year. The pain was concentrated in the very trades that had led year-to-date—high-beta, high-volatility winners—rather than a broad, indiscriminate market dump.
Stress first surfaced in high-volatility pockets like precious metals, then spread to software as markets priced in faster AI-driven disruption to legacy SaaS amid deteriorating credit conditions.
From there, it broadened into a cross-asset de-risking wave as tighter risk budgets and margin and VaR constraints pushed investors to raise cash by selling whatever was most liquid.
The pressure then moved into high-momentum AI-linked areas such as semiconductors and power. The semiconductor thesis still looks fundamentally solid, with capex guidance from several cloud hyperscalers that have reported so far coming in above expectations.
The problem is positioning. Crowding is simply too high, so when institutions’ first priority became raising cash, these crowded winners(AI beneficiaries, AI power, defense/security, Bitcoin miners, and other high-momentum pockets) were sold aggressively.
Sell-side commentary framed it as an extreme factor reversal—one of the biggest momentum drawdowns in more than three years—driven more by long selling than symmetric deleveraging.
So what now?
In the near term, the tone from both Goldman and Morgan Stanley was similar: momentum could still face downside pressure because crowded long positions haven’t fully reset, even if a short-term bounce shows up. But historically, sharp momentum drawdowns that aren’t accompanied by a full fundamental narrative collapse often create better medium-term entry points—once volatility starts to compress and positioning clears.
From an sector-logic standpoint, BofA's Vivek Arya argues it doesn’t make sense for software and chip stocks to sell off sharply at the same time.
Software weakness implies AI adoption is becoming so widespread that it could disrupt legacy SaaS business models, while chip weakness implies AI investment returns are deteriorating and future capex-driven demand is at risk. Those two narratives are hard to square, so he views the market’s reaction as an overreaction.
The practical playbook is less romantic and more survivability-oriented:
– Treat deleveraging as an active position, not a defeat. When correlation spikes, "diversified" high-beta books often behave like one trade.
– Hold more cash than usual. Cash isn't a view, it's flexibility. It keeps you from being forced to act in the worst tape and lets you buy when flows calm down.
– If you're looking to re-enter medium-term, do it with time and increments. The turn typically comes not when a headline changes, but when positioning stops bleeding and market structure flows stop dominating price.
Disclaimer: Moomoo Technologies Inc. is providing this content for information and educational use only.
Read more
Comment
Sign in to post a comment
Moomoo Insights
Moomoo Official Account
Decoding markets. Delivering alpha.
35K
Followers
10
Following
102K
Visitors
Follow
Eugh : Ai taking over job openings, gpt curse and billionaires taken advantage
okzai :
John TC : ee
一辈子好运 : Break even
75462370 : Cyclical fluctuations! Uncertainty on Wall Street! AI de-foaming, adjust your mindset and keep moving forward!
76273157 Eugh : Marie Joseph
75715823 : meet you there
75690250 : Due to the current spread of panic in the market, even the most valuable stocks are being suppressed. With increased caution, prioritizing risk aversion is wise. This is just my humble opinion.
74803952 : AI has replaced software, but the profit outlook for AI remains unclear, driven entirely by sentiment.
Slay2dudes : ok
View more comments...