MW This man wiped $600 billion off Nvidia's valuation by marrying quant trading with AI.

By Joseph Adinolfi

DeepSeek was created by a top Chinese quantitative trading firm. It's not unusual for Wall Street quant luminaries to try their hand at tech.

Nvidia Corp. shares erased $593 billion off of their market capitalization on Monday, the biggest daily loss in the history of the U.S. market, according to Dow Jones Market Data.

One man was mostly to blame.

His name is Liang Wenfeng, and he is the co-founder of the China-based quantitative hedge fund High-Flyer Quant and its artificial-intelligence lab business, DeepSeek. High-Flyer is one of the largest quantitative hedge funds operating in Chinese markets, the South China Morning Post reported, using algorithms, mathematical models and computers to bet on moves in financial markets.

DeepSeek shot to international prominence over the weekend after introducing its latest model. U.S.-based users can access it via smartphone app.

As of late Sunday, DeepSeek's app had climbed to the top ranking among the most-downloaded apps in the Apple Inc. $(AAPL)$ app store. Meanwhile, social-media platforms like X exploded with testimonials about how DeepSeek's latest model rivaled publicly-available Western models like those produced by OpenAI and Anthropic, after DeepSeek released its latest model on Jan. 20.

According to some metrics, DeepSeek's model has performed at least as well as certain mass-market products from by OpenAI, which produced the ChatGPT family of large-language models, and Anthropic, the creator of Claude. It has also managed to accomplish this on a shoestring budget, and - the company claims - without the aid of the most-advanced Nvidia $(NVDA)$ chips.

DeepSeek's sudden success sent U.S. stocks reeling on Monday as investors wondered about the possibility of the U.S. forfeiting its status as the global leader in AI and whether Nvidia's pricey chips would always be required to fuel AI innovation.

The model's performance also raised uncomfortable questions about whether the massive amounts of money being spent by companies like Microsoft Corp. $(MSFT)$, Alphabet Inc. $(GOOGL)$ $(GOOG)$ and Amazon.com Inc. $(AMZN)$ on infrastructure to bolster their AI capabilities would ultimately pay off.

Shares of all three companies sank on Monday, although the carnage was nowhere near as intense as the selloff that afflicted Nvidia. The pioneering AI chipmaker saw its shares finish down 16.9% on Monday, translating to a loss in market capitalization of $588.9 billion, according to Dow Jones Market Data.

Read more: Does DeepSeek spell doomsday for Nvidia and other AI stocks? Here's what to know.

A team of strategists at Goldman Sachs Group $(GS)$ has said it expects leading U.S. tech giants and other companies will spend about $1 trillion on AI over the next year. Sequoia's David Cahn forecast in a blog post published in July that the gap between AI spending and revenues would reach $600 billion by the fourth quarter of 2024.

Who is Liang?

Liang, 40, grew up in China's southeastern province of Guangdong, and attended the prestigious Zhejiang University in Hangzhou, where he studied electronics information and computer vision, according to a report in the SCMP.

He founded High-Flyer with two college friends in 2015. Earlier this month, he appeared at a symposium hosted by Chinese Premier Li Qiang in Beijing, according to the SCMP. His attendance was interpreted as a sign of DeepSeek's growing importance to China's nascent AI industry.

High-Flyer has said it began building a cluster of Nvidia chips a few years ago. DeepSeek has said its models were trained on fewer than 10,000 Nvidia A100 GPUs, a claim that some in the U.S., including Tesla Inc. $(TSLA)$ CEO and GrokAI founder Elon Musk, have pushed back against.

Liang's hedge fund reportedly manages $8 billion, but he is hardly the first quantitative trader to build on skills he refined in the quantitative trading arena and use them to build technologies with broader applications. Years before Liang arrived in Hangzhou to study at a prestigious university there, Jeff Bezos, who would go on to launch Amazon.com, was working at D.E. Shaw & Co., a pioneering quantitative trading firm on Wall Street, before leaving to launch his own company. Bezos built on the skills he learned at D.E. Shaw to create an online bookstore that, over time, would sell just about everything to retail consumers. Amazon would also build a large cloud-computing platform called Amazon Web Services.

Bezos is hardly the only investing luminary to have made the jump from tech to finance, or gone at it from the other direction.

David Siegel, a co-founder of the big Wall Street quant shop Two Sigma Investments, had previously co-founded a tech startup, Blink.com Inc. before heading into quantitative trading. Both Siegel and his Two Sigma co-founder, John Overdeck, worked at D.E. Shaw & Co. before founding their own firm. Overdeck also served a brief stint as a vice president in the early days of Amazon.com.

Even David E. Shaw, the founder of D.E. Shaw & Co., handed off management of his eponymous hedge-fund firm to launch D.E. Shaw Research, which uses supercomputers to aid in drug research.

Robert Mercer and Peter Brown, of legendary Long Island-based quant fund Renaissance Technology, had previously worked at an IBM Research Lab.

Mark Gorton, who helped found early algorithmic trading firm Tower Research Capital, went on to create the peer-to-peer file-sharing service LimeWire.

When quant and tech overlap

According to Gareth Shepherd, co-head of Voya Machine Learning Intelligence, there's a significant amount of overlap in the skill sets required to excel in quant investing and AI research.

"Even the tech stack that's applied in quant finance, it's the same set of tools that Big Tech are using. It's reinforcement learning, and deep learning, and natural language processing and Bayesian networks," he said. "You get across all of those different AI techniques and approaches, they're applicable in both."

If anything, Liang's decision to expand his focus from finance to AI underscores just how competitive - and how potentially lucrative - AI is expected to be.

For years, Wall Street had an edge in competing for top tech talent, said Evan Feagans, a co-portfolio manager for the TCW Artificial Intelligence ETF AIFD, during an interview with MarketWatch.

"For a long time, the quant hedge funds were aggressively hiring out of tech," Feagans said. "There's not a lot of people with this kind of expertise in data science and machine learning, and going to a quant hedge fund was one of the most lucrative things you could do."

But the advent of the AI revolution has helped give some giant tech companies the upper hand. Given the immense cost of developing these models, companies like OpenAI are willing to ensure that they can pay what is necessary to bring top talent in-house.

"Now, we're starting to see a bit of that tech and AI pendulum swing back. The talent is starting to flow toward tech again," Feagans said.

-Joseph Adinolfi

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

January 27, 2025 17:58 ET (22:58 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.