Nvidia's Earnings Show Why Momentum Can Still Work

Here is the case against momentum, and it sounds reasonable. By early September only about 37% of S&P 500 stocks were trading above their 50-day average, against a historical norm closer to 57%. The strategy that carried 2026 rolled over back in June. And the trade's poster children — the AI chip names — just absorbed the steepest rate-driven correction in years, with the 30-year Treasury near 5.26%, its highest since 2007, and the semiconductor index in its biggest drawdown since the 2022 bear market. If momentum is a crowded trade that is finally breaking, stepping aside looks like the mature move.

The problem is that "momentum" is doing the work for two very different ideas. One is chart velocity — buying whatever is going up, on the theory that going up begets going up. That version earned its reputation this year: Societe Generale finds the iShares MSCI USA Momentum ETF is up more than 20% year to date against roughly 8% for the S&P 500, an unusual gap between winners and losers. The other version is earnings momentum — buying the companies whose financials are accelerating, and letting the market slowly pay up for it. They are not the same thing, even when they move together. The first is crowd flow; the second is a lagging read on fundamentals.

Nvidia is the cleanest place to tell them apart, because its latest report made the two visibly separate. For the quarter ended July 26, the company reported revenue of $96.2 billion, up 106% from a year earlier — an acceleration from the roughly 62% and 73% growth it printed in the two quarters before that. Data center, the business that matters, grew 117%. Gross margin held at 75%. That is not a momentum story that stopped; it is one that got faster.

Here is the part most people miss. The stock is up roughly 20% year to date while its earnings more than doubled. Price rose, and the stock got cheaper anyway. Each dollar of Nvidia now buys meaningfully more earnings than it did a year ago — trailing earnings trade around 28 times, deep in the same zone as an S&P 500 that is growing at a single-digit rate. When a high-grower's price underperforms its own earnings, the market is compressing the multiple, not inflating a bubble. Momentum "working" has made this trade less expensive, not more.

That, and not any chart pattern, is why the strategy can still work here. The concern the selloff raised — that the market got ahead of the fundamentals — is contradicted by the actual numbers: the re-rating was downward while the earnings were upward. What the late-summer yield shock did was reprice the multiple because rates repriced everyone's discount rate, not because Nvidia's cycle turned. Far from ending the trade, it handed fresh money a chance to buy the strongest earnings stream in the index at a single-digit multiple premium to the market.

The honest caveat is that this is a premium price, not a bargain. Roughly 28 times trailing earnings on a company that must keep outgrowing the market is a decision to pay up for a supercycle, not a discount on solidity. The new guidance sits at about $108 billion for the current quarter — which assumes no data center compute revenue from China and still points to growth staying near or above 90% — but the whole case leans on the acceleration continuing into next year. Momentum stops working the moment the growth does. The invalidation is specific: if Nvidia's year-over-year revenue growth decelerates toward the market's rate while the multiple stays where it is, the compression was a one-time repricing and holding a rich multiple for market-rate growth makes no sense.

So how does a reader act on this rather than cheer it? Don't chase the vertical. The setup is constructive, but after a near-doubling in earnings the better risk/reward has always shown up on the macro-driven pullbacks the strategy is famous for swallowing — and the late-summer episode was exactly that. The broader lesson is the one that survives the trade's ups and downs: momentum that mirrors accelerating earnings is a durable way to own quality; momentum that is only crowd flow is a rumor. The selloff just sorted the two, and it did it at a discount.