Michael Burry is not your typical hedge fund manager. He did not go to business school, nor did he complete a traditional analyst program at an investment bank. Rather, Burry's background is in medicine -- completing an MD at Vanderbilt University and previously working as a neurology assistant at Stanford.
Burry eventually took to the capital markets and started his own investment firm. He is primarily known for being one of the few investors who predicted the subprime mortgage crisis back in 2008.
Fast forward to today, and Burry has a new short -- artificial intelligence (AI). According to his fund's latest 13F filing, Burry bought put options on semiconductor powerhouse and AI king Nvidia (NVDA +0.79%) last quarter.
Let's delve into Burry's beef with Nvidia and assess whether the acclaimed investor has just uncovered a bombshell revelation.
NASDAQ: NVDA
Key Data Points
Why is Michael Burry shorting Nvidia?
Over the last three years, the S&P 500 and Nasdaq Composite have repeatedly notched new highs. Much of these gains can be attributed to an overwhelmingly bullish AI trade -- with valuations among big tech soaring in lockstep with the broader market.
When it comes to Nvidia, however, Burry's concerns revolve less around valuation and more around accounting -- specifically, a concept known as depreciation.
Let's say a company spends $1 billion on AI hardware, including GPUs, server racks, and networking equipment. If management expects the useful life of these assets to last five years, then the actual expense will be recognized ratably -- depreciated -- over that timeline. This means that on a company's income statement, investors will see a depreciation charge for $200 million as opposed to the full $1 billion that was outlaid.
Nvidia is innovating at a record pace -- releasing new GPU architectures approximately every 18 months. However, hyperscalers like Microsoft, Alphabet, Amazon, Oracle, and Meta Platforms -- each of which are major Nvidia customers -- are depreciating their AI infrastructure over a timeline that exceeds the actual useful life of these assets.
Burry is contesting that big tech is under-accounting their expense profiles and artificially inflating earnings. Nvidia is Burry's primary target in the investigation, given the company's efforts to quickly release new chips, which shorten the product life cycles of existing GPU architectures and render them obsolete after a couple of years at most.
In Burry's eyes, Nvidia is giving way to a coordinated accounting fraud exercise among leading AI developers.
Image source: Getty Images.
Burry's accounting concerns are legitimate, but is big tech really committing fraud?
Burry is correct in that big tech is depreciating its AI infrastructure over a longer horizon than these products actually last. However, labeling these practices as fraudulent is a bit hyperbolic.
First, public companies report their financial statements in compliance with generally accepted accounting principles (GAAP). While Burry's claim that GAAP profits are overstated checks out, it's important for investors to understand that metrics like earnings per share (EPS) do not fully reflect a company's profitability.
Moreover, GAAP figures can be misleading as they are susceptible to noncash charges, such as depreciation.
Sometimes, it can be more helpful to analyze non-GAAP (adjusted) figures, such as free cash flow, to get an idea of a company's ability to generate and sustain excess cash.
Moreover, many of the hyperscalers are audited by the Big Four accounting firms: PwC, EY, Deloitte, and KPMG. If tech companies were truly engaging in fraudulent accounting, a Big Four auditor would not sign off on the documentation that goes to the Securities and Exchange Commission (SEC).
Image source: Getty Images.
What's next for Nvidia?
While I do think Burry has raised an interesting point, I do not see the accounting issue as a red flag.
Burry is subtly implying that big tech's capital expenditures (capex) may not pay off in the long run. Should this prediction come to fruition, then he's forecasting profit growth to decelerate, which could lead to a collapse in the AI trade.
In reality, big tech is already recognizing a return on its AI infrastructure investments. Among hyperscalers, revenue is consistently accelerating, gross margins are widening, and free cash flow is accumulating.
Against this backdrop, I think Burry is being a bit pedantic and pessimistic about the broader AI opportunity. As a long-term investor, I am far less concerned about profits next quarter than I am about a company's profile decades from now -- once AI has evolved into a more mature concept.
For these reasons, I am not worried about Nvidia's growth prospects. The company's backlog for new chips remains robust, and with new use cases emerging, I see Nvidia remaining an influential force in the AI realm for years to come.
As an investor in Nvidia, I remain optimistic about the company's ability to generate strong revenue and profits well into the future and view the semiconductor leader as a core buy-and-hold position in my portfolio.
