The Buyback That Went Nowhere

Michael Burry, the "Big Short" investor famous for calling the housing crisis, just dropped a reality check on Nvidia Corp. (NVDA). His latest target? The chip giant's absolutely massive $112.5 billion stock buyback program since 2018, which Burry says delivered exactly zero additional value to shareholders.

That's not a typo. Despite spending more than a hundred billion dollars buying back its own shares, Burry argues the whole exercise was essentially running on a treadmill. Fast, expensive, and you end up in the same place you started.

The Math Behind the Critique

Here's where it gets interesting. In a detailed X post, Burry walked through Nvidia's financial data and spotted something curious: the disconnect between aggressive share repurchases and a rising share count. Between 2018 and now, Nvidia generated impressive numbers—$205 billion in net income and $188 billion in free cash flow. The company also spent $20.5 billion on stock-based compensation.

So what happened with all those buybacks? According to Burry, they basically served as an expensive defensive mechanism against dilution from stock-based compensation rather than actually reducing shares outstanding.

"But it bought back $112.5B worth of stock and there are 47 million MORE shares outstanding," Burry tweeted. He added, "The true cost of that SBC dilution was $112.5B, reducing owner's earnings by 50%."

In other words, every dollar spent buying back shares was immediately offset by new shares issued to employees. The net result? A whole lot of money spent with nothing to show for it in terms of actual share count reduction. For long-term investors looking at "owner's earnings"—the real economic value flowing to shareholders—this paints a considerably less rosy picture.

Meanwhile, in AI Land

The timing of Burry's critique is particularly notable given Nvidia's absolute dominance in the AI chip market. The company just delivered blockbuster third-quarter results with record revenue of $57 billion, a 62% jump year-over-year. CEO Jensen Huang confidently declared that "AI is going everywhere," and based on the numbers, he's not wrong.

Nvidia's earnings call showcased overwhelming demand for its Blackwell and Rubin GPU architectures, with the company projecting $500 billion in revenue from these platforms by the end of calendar year 2026. That's the kind of growth that makes investors forget to ask uncomfortable questions about capital allocation.

CFO Colette Kress addressed the buyback strategy during the recent earnings call, emphasizing the importance of maintaining a strong balance sheet to fund growth initiatives and secure critical supply chains. Huang also outlined plans for continued stock repurchases alongside strategic ecosystem investments with companies like OpenAI and Anthropic, designed to expand the reach of Nvidia's CUDA platform.

The Stock Performance Paradox

Despite Burry's concerns about capital allocation efficiency, Nvidia shareholders have had little to complain about lately. The stock has crushed the broader market this year, climbing 34.86% year-to-date compared to gains of 17.03% for the Nasdaq Composite and 17.47% for the Nasdaq 100.

On Wednesday, Nvidia (NVDA) shares finished the regular session up 2.85% at $186.52, then surged another 5.08% in after-hours trading. Over the past year, the stock has gained 27.85%.

The stock maintained stronger price trends over medium and long-term periods, though it showed weakness in the short term with a poor value ranking according to market data.

So who's right here? Nvidia is printing money, dominating the most important technology trend of the decade, and its stock keeps climbing. But Burry's pointing out that beneath all that success, there's a capital allocation strategy that looks more like financial theater than genuine value creation for shareholders. Sometimes both things can be true at once.