The desk has run a consumer-caution thesis since 1 July 2026, built on a string of earnings misses across autos, apparel and packaged goods that the S&P 500 and Nasdaq Composite largely ignored. On 3 July 2026 the desk noted that a weak Non-Farm Payrolls print, which met its own stated falsifier for that thesis, still left the S&P 500 closing flat with futures higher, and concluded the caution case rested on earnings-level dispersion rather than a macro trigger. The question now is whether the AI-capex names flip that dynamic in the other direction: strong supply-chain data pulling the index up even as consumer names stay soft.
The supply-chain evidence is genuinely strong on its own terms. Hon Hai's report of stronger-than-expected sales is framed explicitly around sustained AI demand, and Foxconn's second-quarter revenue jump corroborates that from the assembly side of the same ecosystem, even with the company's own caution on geopolitics attached. A separate report on 5 July 2026 described the semiconductor ETF rebounding after a volatile week, with weak jobs data and Broadcom-related Fed jitters described as overshadowed by the move. Three data points, all pointing the same direction, all landing the same day.
Set that against the tape. The Nasdaq Composite, the index most exposed to the names this story is about, closed 2 July 2026 down 0.8% on the day, down 4.66% over the trailing month, and sitting well off its 20-session high of 26830.96. The S&P 500 closed the same session flat on the day but up 1.71% over five sessions, a milder version of the same pattern. If the AI-capex trade were genuinely reasserting itself over consumer caution, the Nasdaq Composite would be the first place that showed up, not the last.
Strong AI-supply-chain earnings and a semiconductor rebound are necessary conditions for the capex trade to retake leadership from consumer caution, but the Nasdaq Composite's own monthly decline of 4.66% shows they are not yet sufficient.
The split worth naming is between the earnings evidence and the index confirmation, and they are not the same claim. Hon Hai and Foxconon's numbers are real and specific to AI-linked demand; the semiconductor ETF's rebound is real and specific to a single volatile week. Neither of those requires the broader Nasdaq Composite to have turned yet, and the data shows it has not. Reading strong component-level prints as an index-level reassertion is the mistake the last week's consumer-caution episodes should have already inoculated this desk against: earnings evidence and index price action have been diverging since 1 July 2026, not converging.
The weakest link in the reassertion case is timing: Hon Hai and Foxconn's numbers are single-source reports, and the semiconductor ETF's rebound is described only over one volatile week, not a sustained run. What the desk is watching next is whether semiconductor and AI-linked names actually outperform the S&P 500 over the coming sessions despite these prints; if they instead lag the broader index, the AI-reassertion thesis fails and the consumer-caution narrative the desk has carried since 1 July 2026 stays dominant. The ISM Services PMI due 6 July 2026 (forecast 54.2, against a previous 54.5) and the FOMC Meeting Minutes due 8 July 2026 are the next tests of whether the broader macro backdrop gives either trade a reason to move.
