The S&P 500 Hit a Record. The Earnings Headline Is Misleading - And the Consumer Is Running on Empty.

The S&P 500 closed at a record 7,737 on Tuesday, ending a two-month summer swoon and putting the Dow past 54,000 for the first time. The catalysts are familiar: strong earnings, a tech rebound, and Middle East peace hopes driving oil lower. The consensus line, headlined across financial media, is that earnings strength will carry the market higher from here.

I have to push back. The earnings number the market is celebrating is materially misleading, and the engine driving those earnings - the American consumer - is running on historic lows of savings. Record highs don't tell you whether the setup is asymmetric or just expensive.

The 47.4% headline is a mirage

FactSet's blended year-over-year earnings growth rate for the S&P 500 in Q2 sits at 47.4%, the highest since Q2 2021. On the surface, that justifies the rally. Strip out Alphabet and Amazon, though, and it falls to 28.8%. That's still solid - the seventh consecutive quarter of double-digit growth - but it's nowhere near the explosion the headline implies.

The distortion comes from two one-off events. Alphabet's Q2 GAAP EPS included a $98 billion gain. Amazon's included $53.4 billion in pre-tax other income, primarily from its investment in Anthropic. These are non-recurring windfalls that inflated the aggregate earnings surprise to 31.4% - the highest since FactSet began tracking in 2008. Exclude those two names and the surprise drops to 9.2%. That's above the 5-year average of 7%, sure, but it's far from record-breaking.

The market is celebrating an earnings bonanza that is materially propped up by two mega-cap accounting events that won't repeat next quarter. That's not a thesis - that's a headline trick.

The consumer is the real worry

Corporate earnings are only as durable as the economy that feeds them, and consumer spending accounts for roughly two-thirds of U.S. GDP. The personal savings rate... fell to 2.7% in June. Compared to the roughly 8% average across the past several decades, 2.7% is extraordinarily thin.

Households are spending faster than they earn. Personal spending... held up so far, and consumer confidence surveys look worse than actual behavior - people say one thing and spend another. But the savings buffer that allowed consumption to ride through past slowdowns is nearly gone. When that cushion is this thin, earnings growth that depends on consumer spending becomes fragile. The market's arguably baked in continued consumer resilience without pricing the risk that the tank is almost empty.

Beating consensus isn't enough anymore

There's another layer the headline doesn't capture. The bar for what counts as "good" earnings has moved so high that beating Wall Street's published estimate is no longer sufficient to keep a stock elevated. AMD fell more than 8% after beating expectations. SpaceX dropped 10% after reporting narrower losses and stronger-than-expected revenue. Both companies won the consensus game and lost the market's.

The issue is whisper numbers - informal, elevated expectations that circulate among traders and institutional investors and sit well above the official consensus. When a stock trades at a premium, the whisper number becomes the real benchmark, and missing it triggers selling even when the published beat looks good. Future quarters will need to deliver not just growth, but growth that exceeds the increasingly inflated informal bar. That's a harder game to win, and it means the earnings-driven rally has a much shorter fuse than the headline suggests.

What the valuation picture actually says

Here's the piece of good news in the setup. The S&P 500's forward P/E... sits at 19.6, below the index's own five-year average of 19.9. Rapid earnings growth has actually compressed valuations over the past year from peaks near 22. The numbers are growing fast enough that the multiple has cooled.

But "below the five-year average" is not the same as dirt cheap. At 19.6, the market is pricing in continued double-digit earnings growth. If that growth decelerates - whether from fading AI capex tailwinds, consumer softening, or simply the difficulty of sustaining 28%+ growth on a massive earnings base - the multiple could contract rather than expand. That's the meltup unwinding, and it doesn't require earnings to collapse, just to normalize.

Breadth has improved - for now

One genuinely positive signal: roughly 71% of S&P 500 stocks now trade above their 200-day moving averages, up from 55% when the index last peaked in June. That's broader participation than the concentrated chip-fueled rally of earlier this year. About 61% of S&P 500 companies have reported Q2 results, with an 86% beat rate on EPS - above the 5-year average of 78%.

The SPY ETF is sitting at $771.72, up 13.2% year-to-date and 5.8% over the past five trading days. The RSI-14 - a momentum oscillator where readings above 70 signal overbought conditions - is at 66, approaching that zone. The index remains above both its 50-day moving average ($746.40) and its 200-day average ($701.95), which is the technically constructive setup bulls point to.

Hold, but don't chase

This isn't a case of the market panicking and throwing good stocks out with the bathwater. The earnings data, even stripped of Alphabet and Amazon's one-offs, is solid. Breadth has improved. The secular AI capex cycle remains a genuine tailwind. But the risk/reward at an all-time high isn't as one-sided as the consensus headline implies.

I don't think investors should flee the market here. The thesis for moderate participation remains intact. But this isn't the kind of setup where you chase records with conviction. The better risk/reward for new capital sits in individual names where growth, valuation, and competitive moat still show a genuine disconnect - not in broad index exposure that's already priced in a lot of optimism.

For those already positioned, the move is to hold and watch. Reassess if the consumer savings rate continues toward the low end of history while spending softens - that's the trigger for back-half earnings downgrades. Reassess if the whisper-number phenomenon spreads beyond a few high-multiple names - that means valuations are baked to levels that leave little room for error.

The market isn't wrong about earnings strength. It's wrong about what the headline number actually means, and it's ignoring the consumer cushion that's about to disappear.