英伟达市值持续暴涨，是否存在泡沫？

On Feb. 22, 2024, Nvidia’s stock performed one-day heroics probably never matched in the annals of capitalism. By 10 a.m., its shares had soared almost 15%, lifting its market capitalization to $1.91 trillion from $1.66 trillion at the previous close. The gain of $250 billion equals 40% the valuation for Tesla and 80% that of Oracle, and exceeds the worth of Intel by 20%. Nvidia is now the straw stirring the cocktail in the investment world. Its earnings beat, reported after the market close on Feb. 21, ignited a frenzy that sent the Nasdaq 100 soaring 3% by midafternoon on the 22nd.

Wall Street analysts cheered the results, with no fewer than 18 brokerages lifting their forecasts for gains beyond its current price of almost $800 by year-end. Goldman Sachs hiked to $875, while Bernstein raised to $1,000 and KeyBanc to $1,100.

The optimism is certainly understandable, given Nvidia’s sorcerous results, and the widespread view that the AI domain it dominates will prove the transformative technology of the millennium. For FY 2024 (ended Jan. 31), Nvidia grew GAAP revenues by 126% to $61 billion, and raised net income almost sixfold, to $29.8 billion. Nvidia is exhibiting operating leverage on a scale probably never before witnessed, meaning that it is booking big earnings increases as a share of every dollar of new revenue. For example, the gross margin rise of 57% to 73% shows that Nvidia this year pocketed 30% more on each comparable unit shipped to customers over last year, based on the selling price versus what it paid to make or purchase the products and materials. It gets better. Its R&D, overhead and other operating costs, less special items, rose a piddling 11%, one-tenth the rate of increase in its sales.

Despite those spectacular results, the bar just jumped enormously for future earnings

Analysts point out that Nvidia’s P/E multiple doesn’t look outrageous. Prior to the jump on February 22, it stood at 56 based on full FY 2024 earnings of $29.8 billion, and the Wall Street crowd used much lower numbers by citing forward estimates incorporating giant gains in the quarters to come. Overnight, the 15% jump in its share price raised Nvidia’s multiple to 64. In the investment community, that substantial figure appears to scare almost no one.

What should give pause is projections on what Nvidia will need to earn, say, a decade hence to be worth anything like its current cap of $1.91 trillion. In fact, it’s interesting to examine just how much the historic one day leap has raised the bar. Let’s say investors will want a 10% annual total return from Nvidia. After all, it’s a risky stock. The options action prior to the earnings release underscored that investors were extremely nervous that even mildly disappointing results would hammer the stock, simply because of the great expectations built into the price.

It’s especially interesting to compare the earnings Nvidia would have needed to generate when its market cap was $1.66 trillion in February 21 compared to the requirements today, after adding $240 billion.

So let’s start with what was required before the mega-bump, to reach returns of 10% a year. When Nvidia’s cap was $1.66 trillion, it had to notch a valuation of $4.33 trillion by 2034 to reach our goal. Assuming that by then it’s still growing pretty fast, but heading for maturity, we’ll assume it boasts a better-than-average PE of 25. In that scenario, the profit needed to grab the prize is $173 billion 10 years hence. Getting there equates to 18.9% yearly growth in profits, on average, over that span.

That’s one heck of a stretch. But on February 22, the stretch got a lot more elastic. At the current $1.91 trillion cap and our 25 assumption for a P/E, Nvidia must make not $173 billion, but $200 billion by 2034. That’s well over twice the sub-$100 billion numbers Apple and Microsoft are showing today. Adjusted for 2.5% yearly inflation over the next decade, it’s 50% more than Apple earned in the last four quarters. Hitting the $200 billion mark in 2034 requires annual increases of 21%, 2.1 points more than the 18.9% needed before the February 22 run-up.

That may not sound like a huge difference. But consider this: In 2034 over 2033, to add that annualized 21% in profits, Nvidia’s net income would need to wax, in a single year, by $35 billion. That compares with an increase in that 10th year of $27 billion just a day ago, when its valuation was 15% lower at $1.66 trillion.

Put simply, the problem isn’t that 21% earnings growth, for the tech superstar of superstars, looks so daunting. It’s that Nvidia already earns tons of money, and sells at a big multiple to those fat earnings. The cause for caution: The ultimate number Nvidia must achieve to give investors decent returns from here, and the multiple $15 or $20 and eventually $35 billion increases needed in the out years. Competitors will jump rush to capture a share of its sumptuous margins. Stiffer competition will cause prices and profitability to drop. The latest surge just made the adjustable mountain Nvidia must climb far steeper. No company has ever delivered such astounding results so quickly as Nvidia. And no company has ever had to scale such formidable heights to surmount the law of large numbers.