The “Dean of Valuation” Says Nvidia Stock Could Plunge by 31%. Here’s My Contrarian Take on Why DeepSeek Could Fuel It to New Highs, Instead.


Aswath Damodaran is an accomplished professor at New York University’s Stern School of Business. In particular, Damodaran specializes in valuation — having written several books on the topic, and often publishing his models and forecasts to the public. Over the years, Damodaran has become known as the “Dean of Valuation” among financial journalists and media personalities.

Last week, Damodaran published a new forecast around Nvidia (NASDAQ: NVDA) — calling for a 37% drop in share price from current levels (as of Feb. 5).

Below, I’m going to detail Damodaran’s logic to help explain why he’s calling for such a drop. From there, I’ll give my take on why I’m not fully aligned with his bearish forecast.

By now, you’re probably familiar with AI’s newest talking point — namely, a Chinese start-up called DeepSeek. DeepSeek is the latest company to emerge in the AI realm, claiming it’s developed game-changing applications for a fraction of the cost used to build mainstream models from OpenAI or Anthropic.

In Damodaran’s analysis, he states that DeepSeek has “changed the AI story” that will “create a bifurcated AI market, with a segment of low-grade AI products that is commoditized and highly competitive and a segment of premium products.”

On the surface, I understand what Damodaran is getting at. If (key word “If”) DeepSeek has built a platform on par with or superior to existing AI models and did so with less costly infrastructure, Nvidia’s position as the king of the chip realm would appear jeopardized.

To me, the above contention is still more of a theory than anything. It seems that each hour, more stories are publishing about DeepSeek — many of which are now alleging the start-up was funded with much more than the initial $6 million it claimed. If that’s the case, then Nvidia has less to worry about.

But in a world where DeepSeek was built for far less than funding compared to what was plowed into OpenAI and its cohorts, I still don’t see such a notion as a bad thing for Nvidia. The reason actually lines up with Damodaran’s point of chipware becoming commoditized.

Right now, it’s well known that many of Nvidia’s largest customers include cloud hyperscalers such as Microsoft, Alphabet, and Amazon. Moreover, big tech giants such as Meta Platforms and Tesla are also some of Nvidia’s biggest adopters. What is also known is that many of these companies are investing heavily into internal chipware and working with lower-cost providers, such as Advanced Micro Devices.


Leave a Comment