Nvidia: Too big to excite?
The chip designer has transformed from a high-beta leader to a low-beta laggard within the AI theme it created.
Editor’s note: An abridged version of this post appeared in the May 18 edition of EntryPoint, a markets newsletter from Sherwood featuring data on how Robinhood traders are navigating the daily twists and turns and the key trends that define the price action. Sign up here.
On the surface, it’s a ridiculous thing to say about the world’s most valuable company, a stock that’s up 20% this year and 36% since March 30, but...
Like Rodney Dangerfield said, Nvidia don’t get no respect. At least, not that much compared to its peers these days.
Its GPUs are the OG neurons behind ChatGPT, and remain the brains of the AI boom. Annual revenue growth is expected to keep reaccelerating when the chip designer posts its Q1 results this Wednesday.
That’s a remarkable feat considering how fast its top line has been growing for so long, with annual sales up 700% from 2022 through 2025.
But, this is a boom, and when you’re the First Big Thing, it’s hard to also be the Next Big Thing.
There’s too big to fail, and in the case of Nvidia, there may be such a thing as too big to excite. In its past three quarterly reports, the chip designer delivered better-than-expected revenues, earnings per share, and sales guidance — and dropped the following session.
While any chip company would be hard-pressed to match Nvidia’s size or importance to this theme, traders seem more focused on finding firms that will match, or exceed, its growth going forward.
Price action reflects this quest for what’s next: traders’ desire to chase bottlenecks in memory, networking, and CPUs has led to Nvidia both trailing its peers in the VanEck Semiconductor ETF and becoming a stock that’s less volatile than the overall fund.
For what it’s worth, Nvidia is somewhat of a victim of its own success in terms of attracting a disproportionate amount of inflows. Because it’s such a big weight in the index, a portfolio manager wanting to take a meaningfully overweight position would effectively need to run some substantial underweights elsewhere for that exposure to move the needle, possibly running into concentration limits in the process.
But there’s also one way I think Nvidia shot itself in the foot: its June 2024 stock split. Stock splits are generally considered as a way to make your share price a little more accessible and get a better multiple through expanding the potential pool of buyers. But, by making it easier to buy the underlying, via a cheaper stock price, Jensen and co. seem to have cannibalized demand for leveraged exposure via options.
It’s said the stock market can be a Keynesian beauty contest — in which we’re all making decisions based on our guesses about what other people will do, rather than some objective standard of goodness — and in this world, Nvidia’s maturity seems to be a strike against its beauty.
Like a millennial replaced by a Gen Zer, Nvidia’s morphed from a high-beta leader to a low-beta laggard within the AI theme. The same can be said for another group of mature companies that also happen to be its biggest customers: the Magnificent 7 hyperscalers.
By and large, this group isn’t trading like clear AI winners, despite their hundreds of billions spent to reorient their future around this theme.
This continues to tell us a lot about what any “AI bubble” is and isn’t: at this juncture, it’s an attempt to find out the beneficiaries of the hundreds of billions in capex (in some cases, extrapolating demand years down the road), and not a desire to bet on a high ROI from that spending.
