Basketball-Street closures, Secret Service-level security expected as Trump to attend NBA Finals

The selloff looks less like a fundamental reset and more like a positioning air pocket in a name that had become a consensus beneficiary of AI capex. With only modest direct sensitivity to the article’s subject matter, the market’s reaction suggests investors used the headline as a convenient excuse to de-risk after a crowded move; that usually creates a short-duration dislocation rather than a durable change in thesis. The key second-order effect is that the real price action may be in semis broadly: any sign of slowing server procurement or digestion at hyperscalers tends to hit multiple cycles ahead of actual revenue revisions.

The contrarian read is that the market may be underestimating how quickly DRAM pricing can reaccelerate if AI infrastructure spend remains intact. Memory is one of the cleanest operating-leverage stories in semis, so even a small improvement in industry supply discipline can translate into outsized EPS upside over the next 2-3 quarters. If the tape is punishing the stock on sentiment rather than on channel evidence, the setup favors buying after forced selling, not before it.

Tail risk is that this is not just sentiment: if hyperscaler order cadence slips for even one quarter, multiple compression can persist for 6-12 weeks and drag peers lower. The cleaner tell will be management commentary and any change in capex cadence from the largest customers; absent that, the downside is likely technical while the upside remains fundamental. In that scenario, the stock can retrace a meaningful portion of the move quickly as short interest and systematic de-risking unwind.