A massive structural shift is on the way in financial markets, where artificial intelligence hardware could increasingly become collateral for securitized credit. During the August 11 episode of CNBC’s Mad Money, host Jim Cramer detailed how major investment institutions are preparing to issue “compute bonds” backed by data center equipment, drawing direct parallels to established securitized credit markets.
Cramer opened his commentary by pointing out how Wall Street leadership, including Goldman Sachs CEO David Solomon, aligned with NVIDIA Corporation (NASDAQ:NVDA) CEO Jensen Huang to champion data center assets as long-lasting collateral:
Today, we’re learning about the possibility of institutions offering compute bonds. This time, trading compute, that’s a data center asset class, not unlike those securitizations in auto loans. There’s Jensen Huang, CEO of NVIDIA, talking about the viability of the data center, long-lasting value of the chips in these warehouses full of servers. Then you had a series of titans of finance talk about how logical the whole thing is. Then David Solomon, the CEO of Goldman Sachs, piped up and pushed them too.
Securitization Mechanics and Asset Durability
To digest the mechanics of these proposed instruments, Cramer highlighted that compute bonds operate under the exact same structural framework as mortgage-backed or auto loan securities:
What you didn’t hear, or at least I didn’t hear until I digested it, was that these would be securitizations, just like securities backed by home loans or auto loans. Hence the strong endorsement of Goldman’s David Solomon who can see the logic… Now, you may be wondering, will these pieces of paper be backed by the full faith and credit of Jensen Huang and NVIDIA? That’s very funny, but no. But neither are the securities based on auto loans or home loans. There’s a construct at work here, though. There’s a belief that the assets don’t depreciate quickly, if at all.
Comparing Microchip Resilience to Automobile Depreciation
Expressing strong personal confidence in the asset class, Cramer argued that graphics processing units produced by NVIDIA Corporation (NASDAQ:NVDA) retain value far better than traditional vehicles. He said:
I can tell you I’m a believer. The NVIDIA chips that are at the core of the data center haven’t shown much depreciation at all… Older models have held up better than cars which are worth less than the moment they leave the lot… At the beginning, it wouldn’t shock me if NVIDIA itself took down bonds by the tens of millions, augmenting their cash program while demonstrating their fidelity. I know that’s what I’d do.
A New Asset Class Catering to Yield-Hungry Investors
Concluding his analysis, Cramer called the invention of compute bonds a brilliant financial strategy, comparing its disruptive potential to gold, real estate, and residential mortgages. He remarked:
If there are losses, NVIDIA would absorb them. But then again, I don’t think there will be. Instead, I think Jensen Huang and his confederates have invented a whole new asset class like gold, like real estate, like home loans, like auto loans. Oh, and let me just say one thing: I think it’s brilliant. I believe these notes will entice people who want a little extra yield. And there are a lot of money managers who love to chase yield, whether you think that’s a good idea or not.
Depreciation Realities and Technology Obsolescence Risks
While bullish commentators view compute bonds as a groundbreaking yield generator, skeptics warn that securitizing hardware carries some structural risks. Unlike real estate or physical gold, semiconductor technology inherently faces rapid technological obsolescence. If next-generation architecture delivers significant efficiency leaps, existing server clusters housing older chips from NVIDIA Corporation (NASDAQ:NVDA) could experience sudden valuation write-downs, which could impair the underlying collateral backing these debt notes.
Furthermore, risk analysts warn against comparing physical data center hardware to consumer loans. If end-user demand for AI inference or model training cools down, secondary market rental rates for compute power could decline significantly.
Smart Money Backing
While fixed-income skeptics warn of hardware depreciation risks, institutional smart money continues increasing its footprint in the semiconductor king. NVIDIA Corporation (NASDAQ:NVDA) was held by 275 hedge funds in Q1 2026, up from 264 funds in the prior quarter, showing rising high-conviction backing across active money managers.
At the same time, the short percentage of float for NVIDIA Corporation (NASDAQ:NVDA) sits at a minimal 1.26%. Despite lingering macroeconomic debates, broad market bears means few investors are betting against the stock.
NVIDIA Corporation (NASDAQ:NVDA) remains the central pillar of the artificial intelligence trade. While financial engineering around compute bonds introduces new fixed-income variables, steady hedge fund accumulation and modest short seller presence provide a durable backdrop. If data center hardware maintains its structural value, it could keep validating Cramer’s bullish vision for the chip titan.
While we acknowledge the risk and potential of NVDA as an investment, our conviction lies in the belief that some AI stocks hold greater promise for delivering higher returns and doing so within a shorter time frame. If you are looking for an AI stock that is more promising than NVDA and that has 10,000% upside potential, check out our report about this cheapest AI stock.
READ NEXT: Jim Cramer Discusses Retailers He Likes: Williams-Sonoma (WSM) and Ralph Lauren (RL) and Jim Cramer Endorses Lockheed Martin as “Sensational”.
Disclosure: None. Follow Insider Monkey on Google News.
