Nvidia: A Quick Example of Excessive Extrapolation

Last week we wrote a piece citing work done by some of the greats in behavioral finance – Benartzi, Tversky and Kahneman. Our work suggested that the recent wall of stock-based compensation could create unprecedented risk for employees.

Millions of individuals often refuse to diversify their stock comp due to well-understood behavioral errors. Two of the “culprits” we cited were representativeness and excessive extrapolation.

At the request of several FAs, today’s piece will walk through a simple example of soaring stock grants and excessive extrapolation. We’re going to use Nvidia because our ranking tools are neutral on the stock, and it has quickly become a household name.

Get our insights direct to your inbox: SUBSCRIBE

First, the stock-based compensation component. The chart below shows two things.

• The light blue bars are scaled to the left-hand axis and show the amount of stock-based compensation NVDA granted employees by the company’s Fiscal Year End

• The red line is the stock price of Nvidia over that same period adjusted for stock splits

When the stock price was low, the company’s employees received very small amounts of stock. Now that the company’s stock price has risen 44,083%, employees received ~$4bn in stock over just the last 12 months.

We are not suggesting impropriety. We picked Nvidia because our models are neutral on the stock. Yet the valuation, infused with AI hype, now suggests that recent profit growth will accelerate.

Before explaining why this might be a case of excessive extrapolation, we will first acknowledge our own biases. KCR has written a great deal about the expectations embedded in AI stocks.

We haven’t exactly been bullish:

• In AI, Oil & Gas: Misadventures in Capital Allocation, we showed that tech stocks were now as capital-intensive as energy stocks and that the earnings of tech stocks were just as cyclical as energy stocks

• In External Obsolescence: Tech Investors’ Newest Nightmare?, we demonstrated that the majority of the stock returns in the AI complex were driven by multiple expansion rather than improving fundamentals

• In Everything AI, we explained the brazen vendor financing underway in the tech complex and showed how it was an almost perfect analog to the dot.com bubble, which would lead to devastating losses

Our dim view of bloated valuations and high expectations has left egg on our face. Nvidia, the leader of the AI complex, has beaten earnings, and the stock is rising ever higher. We stand unbowed in our views.[1]