It seems ungracious to find fault with the 2026 stock market, but that isn’t stopping some observers from expressing unease with some anomalous tape action. From a distance, the half-year now ending has been rewarding from point to point. The S & P 500 is up 8.7%, which equates to an annualized total return approaching 20% with dividends. The Dow Industrials and equal-weighted S & P are at or within a fraction of a record. The advance over six months has nicely alternated leadership between a broad majority of stocks (January, February and the past few weeks) and the narrow, high-momentum AI-hardware theme (the rest of the time). The Nasdaq 100 , weighed down as it’s been since peaking on June 3 by a prolonged malaise in the Magnificent Seven cohort, is holding its uptrend, merely settling back to its 50-day moving average, from which it bounced both Friday and Monday. Upward earnings-forecast revisions are on the gallop, fiscal support remains with a Federal deficit of 6% or GDP and capital-markets activity is humming. Still, the behavioral oddities and internally erratic behavior keep inviting questions about the energy reserves and structural integrity of the rally. .SPX YTD mountain S & P 500, YTD There’s no doubt the overall market has decelerated and settled into a churning, unresolved range since mid-May, a moment when the aggression of AI-related buying went so far as to bid the Cerebras IPO to a stark day-one peak while calling the old wounded soldiers of the last generation’s tech wars into service ( Cisco Systems and HP Enterprise shares getting exaggerated pops). Internal dysfunction Then there’s the incoherent message of a high-dispersion market. Last week, the S & P 500 was down five straight days while more of its components were up than down each day. Hadn’t happened since 2000. A related point highlighted by Goldman Sachs: The rolling one-year correlation between the S & P 500 and its equal-weighted version – these are the same 500 stocks, mind you – has fallen to 0.79, the lowest ever. For 25 years it’s averaged 0.96. Evercore ISI strategist Julian Emanuel goes deeper to feature “negative-beta stocks” – those that move opposite to the S & P on a daily basis. There are now more than 80 such names, “more than there have been at any time since the Tech Bubble unwound in 2000-01.” He sees these stocks as valuable diversifiers for investors anxious about the AI-centric benchmark but unwilling to pay up for outright downside protection. The Philadelphia Semiconductor Index in June has averaged a 4.1% daily move. The only periods more hyperactive than this month were either in the desperate crash phase of a bear market or in the furious rebound phase after a nasty correction. MU YTD mountain Micron, YTD Much discussed but still worthy of mention: Micron Technology last week reported a spreadsheet-melting upside surprise to revenue, earnings and multi-year forward profit and margin guidance. And the stock remains below where it closed two days before those gaudy numbers hit. Too much reliance on AI? While all of the above quirks and superlatives assert that this is an uncommonly kinetic and limit-pushing market backdrop, they are all slightly different ways to portray a single underlying reality: The market as gauged by the S & P 500 benchmark has grown unbalanced and acutely reliant on the AI-capex theme. This is broadly understood at a high level, generally with a sense of discomfort, which is why the consensus always applauds periods of “broadening,” such as the past few weeks when healthcare and financials and the like have gathered momentum. Can a bull market truly escape the core life force that animated it in the first place, though? I’m somewhat dubious that broadening phases are more than restorative interludes but am open-minded. BCA Research in a new report runs the numbers to determine how many investment “factors” – fundamental or technical characteristics – have effectively been overtaken by AI risk. Without getting into the math of it all, the firm concludes that the following factors are now dominated simply by AI: beta, volatility, asset growth, free cash flow, gross-profit-to-assets, earnings volatility and earnings momentum. This at a time when the market is in the throes of a rethink over the costs, benefits and duration of AI investment, where the value will accrue and what is already priced in. The notion that memory will remain in a multi-year shortage with supply rationed through surge pricing for years to come has animated that subsector. But now hardware makers are pushing back and asking for approval to buy Chinese chips and the bull case on Cerebras is memory efficiency. Investors are alert to the dynamic that has seen Nvidia shares stall out for eight months, again, as stellar fundamentals continue but seem to have been recognized and front-loaded by the market and as the hyperscalers footing the bill have been sold down and cheapened in the process. Pause needed? As dicey as these fits and starts might be, it’s probably a plus that the market is working identify these nuances and draw distinctions between the ripe opportunities and the played-out theses. I was talking last week about how it wouldn’t be surprising to see the hyperscalers get rediscovered before too long, as traders observe that they’ve become unloved, under-owned, relatively inexpensive and potentially defensive in a risk-averse tape. Whether Monday represented the start of such a pendulum swing, who knows, but Meta Platforms now trades cheaper than eBay on a forward P/E basis. Would the market need more than a seven-week trading range to work of excessive optimism and offsides positioning in leveraged semi plays? Plausibly. Observers are right to point out an aggressive rise in margin debt, in excess of underlying equity-market appreciation, as a possible hazard – but at least folks are pointing such things out. The sell-side might be starting to over-promise, too: FactSet computes a “bottom-up” S & P 500 target, using the consensus stock-price target for each index component among industry analysts. This hypothetical target now sits at 8918 – up more than 20% from here and far in excess of any current strategists’ forecast. FactSet senior earnings analyst John Butters tells me the average bottom-up implied upside target among this vocationally bullish crowd over the past five, ten, 15 and 20 years is 13-15%. Approaching four years into this bull market, Wall Street pros are not shy about letting the good times roll.
Mike Santoli: A great first half, but some erratic behavior raises doubts about the rally’s integrity
