AI whack-a-mole as markets continue repricing the cost of disruption

US equities endured a rough session, with the Dow down 600 points, the S&P 500 posting a third consecutive loss, and the Nasdaq sliding 2%, as the market’s relationship with artificial intelligence and adjacent proxies continues to pivot from infatuation to interrogation. What was once a rising escalator of multiple expansions now feels more like a corridor of trapdoors. The dominant feature of this phase is not indiscriminate liquidation, but a disciplined exit through sectors perceived to sit directly in the blast radius of AI disruption.

Almost anything previously branded under the AI risk umbrella, whether the threat is tangible or merely narrative-driven, is now under sustained pressure.

The tape has turned into a game of whack-a-mole. Each session a new industry surfaces and capital swings the hammer. Software and SaaS absorbed the initial shock. Private credit followed. Insurance brokers came next. Wealth managers and diversified financials were pressured. Real estate services rolled over. Most recently logistics names were hit hard as investors recalibrated what AI enabled automation means for freight matching, labor intensity, and margin durability. The magnitude of moves in select transport stocks reflects not simply volatility, but a rapid repricing of business model resilience under an automation lens.

Technically, the Nasdaq 100 has broken below its 100-day moving average while the S&P 500 slipped beneath its 50-day. Momentum has deteriorated, and we are approaching the mechanical CTA selling zone, not a favourable development to end the week.

The macro backdrop offered little support. Existing home sales registered one of the sharpest monthly declines on record, reinforcing concerns around rate sensitivity and nudging flows into traditional defensives such as utilities, staples, and healthcare.

Treasury yields declined, with the 10Y probing the 4.10% area, as duration reasserted its role as ballast during equity stress. Notably, markets largely discounted the prior strong payrolls report, signalling that forward earnings durability now outweighs backward labour strength in investor calculus.

Cross asset price action reinforced the deleveraging tone. Commodities and crypto, including gold and bitcoin, initially showed relative stability but then weakened as traders raised liquidity to offset equity losses. This pattern suggests portfolio mechanics rather than a structural shift in their underlying narratives. When hedges are monetized to fund drawdowns, the signal is a balance-sheet adjustment, not a thematic collapse.

Perhaps the most telling development has been the absence of reflexive dip buying. In prior episodes, short-dated options flows frequently fueled late-session reversals. This time, the closing minutes lacked momentum. The defining characteristic has not been overwhelming supply but a clear unwillingness to intervene in a snowballing thematic unwind.

Strategically, the AI trade has entered a more demanding phase. The market is no longer rewarding the scale of ambition. Investors are now auditing execution. The central question is shifting from total addressable market to sustainable margin capture. Automation promises efficiency, but it also risks commoditization.

The secular AI theme remains intact. However, the easy multiple expansion regime appears to have transitioned into a more analytical and selective environment. Capital is becoming more discriminating. Liquidity is more conditional. Valuations must now be supported by cash-flow visibility and competitive durability.

Until there is clearer evidence on where durable earnings power ultimately resides within the AI ecosystem, the whack-a-mole dynamic is likely to persist. In this regime, rallies will be tested, defensives will attract relative flows, and duration will continue to serve as a hedge against growth uncertainty.