The title of the “Bond King” is somewhat in contention. Many would give it to Bill Gross, though over the years, a younger Jeffrey Gundlach has at least shared the title. In any case, when Gundlach posts something like this, it is worth . . . peeling back, shall we say?
Gundlach took aim at Wall Street’s newest financial engineering trend. His beef? Nvidia’s (NVDA) agreement with major private credit and asset management firms such as Apollo, BlackRock, Blackstone, KKR, and Goldman Sachs.
The consortium's goal is to fund massive artificial intelligence (AI) data center infrastructure and chip purchases by issuing long-term debt backed by the chips themselves as collateral. Gundlach calls out a glaring duration and depreciation mismatch.
And, while time will tell if he is right, I repeat to my audience weekly my own concerns that AI compute spending has reached the point where lofty expectations (via stock prices) cannot possibly be met. But hey, that’s just the outcome, not the current excitement. There’s a big difference, as we learned in both the dot-com bubble and more recently in the aftermath of the cryptocurrency runup.
Breaking Down Gundlach's Argument
At the heart of Gundlach's critique is that this is the type of collateral that spoils like bananas. That is, Nvidia’s GPU product refresh cycles run on a fast 18-month to 24-month round trip.
Thus, issuing long-dated debt secured by hardware that may be technologically obsolete or vastly diminished in value in a few years creates a huge mismatch. One which, in the short attention span zone we all seem to live in nowadays, is going to be a big problem for tech investors — even if it’s not right now. That is why Gundlach sarcastically compared securitizing GPUs to issuing a 30-year bond backed by "warehouses of newly engineered bananas."
Gundlach noted in subsequent comments that when market tops form, they rarely ring a bell. Instead, they showcase financial innovation relying on dubious ratings to turn speculative tech capital expenditures into a pseudo-safe asset class.
Mark Cuban responded to that post in agreement, noting that “chips as an asset class will be the new crypto.” To me, it smacks of both 2008 and 1987, whereby financial magic turned into, well, rotten bananas.
The ETF Effect
If this ends up as a pie in the face of NVDA shareholders, that will only be the start of it. As the ETF guy around here, I feel the need to point out that 475 U.S.-listed exchange-traded funds (ETFs) own NVDA. That does not include the many more that get to the stock via indirect means, such as swap contracts and the like.
If debt markets re-price the risk on private credit platforms funding these mega-purchases, hyperscaler access to easy leverage shrinks. Without continuous cheap debt fueling $500 billion-plus hardware buying sprees, hardware sales growth naturally decelerates toward realistic end-user demand. And that bleeds through to hundreds of NVDA-infused ETFs.
As we see above, NVDA is a central part of the ETF landscape. More than 10% of all ETFs, including those not related to equities, large-cap, or tech stocks, include a position in the stock. Given its high market capitalization, when an ETF owns NVDA, it is often a pivotal part of its overall strategy.
I cut this list off for brevity, but we see just how many funds have double-digit percentage positions. If you run it on this section of the sidebar on most Barchart pages, you’ll see it for any stock in an instant. It is a very useful tool.
That chart above does not even include the very biggest ETFs, in which NVDA is around 7% to 10% of assets.
NVDA itself looks somewhat neutral, chart-wise. But it is only up about 5% since late last year, so its rapid acceleration pace has slowed. Gundlach’s comments are more intermediate-term, and that’s where I think NVDA likely leads the broader market down. Because it is the Jenga piece that holds this market together — for now.
The Bottom Line
Gundlach isn't attacking Nvidia’s current technology. Instead, he is warning against Wall Street credit overreach. When debt duration outlives the realistically useful life of the collateral backing it, credit markets eventually force a re-rating — and because NVDA anchors nearly 500 ETFs, that re-rating spreads far beyond a single ticker.
That doesn’t matter in a short-attention-span phase of the stock market cycle, but I think it will eventually. And that’s when years of stock price gains can evaporate.
Rob Isbitts is a semi-retired CIO, former fiduciary investment advisor, and Barchart columnist. Check out his other work at ETFYourself.com (featuring the Fresh Charts weekly trading post), and ROAR.PiTrade.com, helping investors to better-manage their own portfolios.
On the date of publication, Rob Isbitts did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. For more information please view the Barchart Disclosure Policy here.