Santoli: AI bull market almost surely needs tech stocks to recover to keep advancing

Get Mike Santoli's Market Memo in your inbox first. SUBSCRIBE HERE

The road to bull-market glory does not run through Walmart and Berkshire Hathaway and Eli Lilly.

There's been a virtual buying panic in traditionally defensive, high-quality stocks over the past few weeks, a near-equal-and-opposite offset to the selling pressure in semiconductors and AI-build-out spenders. The three stocks above — the only members of the $1 trillion market-cap club not in Big Tech – have collectively added $550 billion in market value in the past month, almost making up for what Alphabet has lost.

The Nasdaq-100 and the S&P 500 Low Volatility ETF (SPLV) are rushing toward one another in a satisfying picture of symmetry that nonetheless reflects an almost comical level of tactical rotation executed by an investment community accustomed to playing one factor against another.

While this to-and-fro has so far protected the S&P 500 as a whole from an appreciable setback for months, it has left it stymied around levels that date back to May — before we got two earnings-reporting seasons showing 30% annual profit growth.

A week ago here, I suggested the AI theme was nearing a make-or-break moment to stage a recovery or risk surrendering its leadership status and destabilizing the broader market. Thus far, this trade has failed to respond to moderate oversold conditions or a run of cosmetically strong hyperscaler quarterly results, further loosening its grip on the benefit of the doubt.

In stressed markets undergoing soul-searching about the long-term bullish thesis, prices often need to travel further to find buyers with conviction while the market action must disappoint dip-buyers to extinguish complacency. Hard to know if we're there, but presumably we're now closer.

The Nasdaq-100 closed Monday right at its 200-day moving average, off almost 11% from its record high. After an enervating failed Monday-morning bounce attempt, the Philadelphia Semiconductor Index finished down another 3.8% and settled under the late-July low that was set as the Situational Awareness hedge fund was finished liquidating.

"Markets did not get the memo that bearish seasonal pressure typically subsides in mid-October and instead ended the month with some of the most acute selling pressure that has been realized during the late-summer/early-fall period," notes JPMorgan technical strategist Jason Hunter. He's advising clients to remain on the defensive "until we see clear evidence of a bottoming process on the charts or until there are clear signs of capitulation."

This is a time-tested playbook, though many a sharp market pullback has ended without what would qualify as unqualified capitulation.

The market action is winning some plaudits for the way non-tech stocks have held up better, many groups benefiting from the "quality" rotation while abandoned consumer cyclicals get a reprieve from easing oil prices and the slide in the 10-year Treasury yield back below 5%.

This gives active managers plenty to do in playing against a highly concentrated and tech-dependent S&P 500. But I'll continue to argue that the bull market built on the AI story will almost surely need the AI story to get back on track to keep advancing. A prolonged period when the equal-weight S&P 500 outperforms the market-cap-weighted version is otherwise known – usually – as a bear market.

The AI-related stocks' weighting in the S&P 500 is off the highs but remains near 45%, according to Morgan Stanley.

As the chart shows, AI has wobbled relative to the rest of the market during this uptrend before reasserting itself.  But there's a trickier element to the current gut check: In prior phases of the buildout, the key limiting bottleneck gave investors something to buy in order to exploit it.

The initial scarcity of GPUs meant to own Nvidia above all else from 2023 through 2024. When cloud capacity, power and cooling were found to be the crucial scarce resources, hyperscaler stocks worked and the AI-adjacent industrials were drastically revalued. Then memory emerged as the chokepoint and Micron and Sandisk went stratospheric.

Now, as Nvidia's Jensen Huang has said, capital is the constraint and that not only leaves no asset to buy, but it raises the cost of capital broadly as the mega-caps go free-cash-flow-negative.

Overlapping feedback loops and scrambled correlations across the markets and the real economy have further sapped investor conviction. Treasury yields no longer move with oil but largely with AI-capex sentiment, with both in retreat lately. AI-stock setbacks also mute inflation momentum and the equity wealth effect, meaning fewer Federal Reserve rate hikes are anticipated now versus a month ago.

This all probably sounds like an elaborate trap, and it has become one for the previously crowded trades. Could this reflect the end game for the AI-exuberance market chapter, which as I keep pointing out peaked around the SpaceX IPO in June? Impossible to say for sure, but if it were, I'm not sure things would look much different to this point.

But I don't want to dismiss the possibility that a lot of the known challenges are largely reflected in prices.

I argued in August that we should assume the valuation peak for this cycle was in. We've now retrenched to a forward P/E of 18.5 on the S&P 500 from 23 a year ago and the Nasdaq-100 at 20x is at the slimmest premium to the S&P 500 valuation in nine years. I'm on board with the idea that the market is telling us companies are now "over-earning," but if nothing else there is less valuation risk now than in recent months.

Sentiment toward AI stocks is in the dumps. Funds' positioning in tech has retreated below neutral. Important trading lows this year have happened near the turn of the month on March 30 and July 29, and the Nasdaq-100's relative-strength reading is exactly as oversold as it was at those turns.

So, stay alert for possible relief bounces that could evolve into something more consequential if indeed the midterm elections pass in a way that allows traders to unclench a bit into the year's home stretch.

Market Temperature Gauge

 John Kolovos of Macro Risk Advisors on the latest reading: "Similar to valuation, sentiment is not a timing tool, but it is getting harder and harder to find any bulls, as a near-perfect storm of tight financial conditions is being met with tech earnings that have gone from bad to worse in the blink of an eye." 

Around the Street

-Another good big-picture conversation on the practical business and investment implications of the next phase of AI proliferation on Patrick O'Shaughnessy's Invest Like the Best podcast, this one with investor Cliff Sosin. His comparison of AI infrastructure economics to internet connectivity providers and his argument for how most of the productivity boost will go toward "consumer surplus" rather than corporate profits are particularly interesting.

-For my birthday Monday, my colleague and friend Sara Eisen wondered how I would write a birthday message to myself. I immediately regretted suggesting she ask ChatGPT, which generated this credible and somewhat embarrassing result:

"Another year older, though the longer-term trend remains constructive. Some modest deterioration in the underlying internals, perhaps, but nothing that suggests a decisive breakdown.

"The candles on the cake are approaching historically elevated levels, though valuation concerns have persisted for years without derailing the advance. At this stage, the key question isn't whether I've gotten older — that's largely priced in — but whether the wisdom gains can continue to offset the inevitable compression in energy multiples.

"For now, I'm maintaining a cautiously optimistic outlook, with upside optionality, downside support from family and friends, and no immediate need for Fed intervention. Happy birthday to me. The secular bull market remains intact."

Market on Close

The stampede into perceived defensive stocks has been a crucial driver of Apple shares' massive outperformance over the Nasdaq-100 since the AI-capex agita flared up months ago. Apple heading into Monday's earnings report had outperformed the Nasdaq-100 by nearly 20 percentage points over the prior six months.

Its singular refusal to join the aggressive spending binge to build AI capacity preserved its free cash flow and made it the perfect counterweight to the capex trade, while inflating Apple's valuation to a 15-year high.

But its advantage in free-cash-flow yield has now narrowed, thanks to its stock-price gains and the collapse of peer shares.

SUBSCRIBE TO MIKE SANTOLI'S MARKET MEMO HERE