Many on Wall Street like to style themselves as contrarians, but few are swimming against the current as hard as Jay Goldberg, the lone bear on high-flying Nvidia.
Sitting in the cluttered home office on the ground floor of his three-story Victorian near San Francisco’s Haight-Ashbury neighborhood, the good-humored bespectacled 54-year-old senior analyst at Seaport Global Securities hardly looks or sounds like a stock-market iconoclast. With his collection of ancient technological devices, including a brick-shaped cell phone from the 1980s that looks like it could’ve been used by Gordon Gekko in the movie “Wall Street,” the married father of three comes across as a low-key gadget geek.
His sell rating on the chipmaker at the epicenter of all things artificial intelligence, on the other hand, has turned him into Wall Street’s version of a punk rocker.
“There’s a lot more that can go wrong with Nvidia than can go right,” Goldberg said.
Eighty analysts cover the chipmaker, 73 of whom have the equivalent of a buy rating on the stock. Six rate it a hold. And then there’s Goldberg.
The mass bullishness makes sense when you consider that anyone who has tried to stand in the way of Nvidia’s rise in the past few years has gotten trampled amid the insatiable demand for Nvidia’s chips, called graphics processing units or GPUs, that are essential for AI computing. Its stock is up more than 3,000% since the start of 2020, making it by far the best performer in the S&P 500 Index over that time, almost doubling the return of its next closest competitor, Super Micro Computer.
With the AI boom driving the U.S. economy and propelling the stock market to record after record, few people are eager to bet against the key company behind the rally.
“AI is a multiyear generational cycle, and one that we’re in the early innings of, not even the middle innings,” said Jim Awad, senior managing director at Clearstead Advisors, which owns Nvidia shares. “Nvidia is a significant player within that, involved in every aspect of the AI trade. It is powering the economy and the stock market. I say let the horse ride.”
Against the grain
Goldberg, however, has no qualms about going against the grain.
“I’m probably a little cantankerous by nature, so I’m skeptical of all of the hype around AI right now,” he said. “This is not my first bubble.”
Goldberg covers 12 companies, and Nvidia is his only sell rating. He has buys on Apple, Netgear and chipmakers Broadcom and Arm Holdings, both of which are heavily exposed to the AI trade.
In his view, the investing thesis around AI comes down to spending by six companies: Microsoft, Alphabet, Amazon, Meta Platforms, Oracle and OpenAI.
The sextet are racing to build out infrastructure for the technology, and their demand for Nvidia chips has made it the biggest company in the world, at $4.5 trillion. The five public companies in that group are expected to devote nearly $400 billion to capital expenditures this year, up more than 67% from last year. OpenAI, the enormously influential closely held owner of ChatGPT, has committed to spending over $1 trillion.
But with four of those companies reporting earnings next week, investors are beginning to focus on how little these businesses have to show for all that spending so far.
Goldberg likens this to the telecommunications-infrastructure buildout during the dot-com bubble in the late 1990s and early 2000s, a period when he was getting his MBA at the University of Chicago before landing jobs at Lazard and Deutsche Bank and later at chipmakers Peregrine Semiconductor and Qualcomm, followed by a software startup and venture capital firm. When the anticipated internet traffic didn’t materialize right away, companies like Cisco Systems, whose shares were turbocharged by the spending, got hammered. More than two decades later, Cisco’s stock price still hasn’t touched its 2000 peak.
“That feels very strongly like the pattern we’re seeing now,” Goldberg said. “We’re going to build up all this AI stuff for what are largely psychological reasons. At some point the spending will stop, and the whole thing will tumble down and we’ll reset.”
His view hasn’t been borne out so far.
Goldberg initiated coverage of Nvidia on April 30, and the stock has soared more than 70% since then, beating the Philadelphia Stock Exchange Semiconductor Index and some of the companies he has buy ratings on, including Arm Holdings. And the bullish calls keep coming even as chatter among investors about an AI bubble gets louder. The average analyst price target has steadily risen and now stands at $220, implying upside of 18% from Nvidia’s Friday closing price of $186.26.
The bull case
“It’s still early in the investment phase,” said Moon Surana, portfolio manager at Harding Loevner, which holds Nvidia shares. “There are no signs of overcapacity yet. There are no GPUs sitting idle.”
Nvidia’s biggest bull is Frank Lee of HSBC, the analyst who last week raised his rating to buy from hold with a price target of $320. He sees the market for AI accelerators growing significantly as demand moves beyond Nvidia’s biggest customers, a view that has been supported by recent deals in the private market.
However, the level of demand that Lee sees would mean that Nvidia is basically sold out of its AI processing chips. That has led Goldberg to question where any additional upside could come from. His price target stands at $100, by far the lowest on the Street.
That said, Goldberg is quick to express his admiration for Nvidia and the leadership of CEO Jensen Huang. In his view, the sell rating means the stock will underperform its peers like Broadcom, Qualcomm and Advanced Micro Devices, not that investors should short it.
Electric concerns
“There aren’t many swing factors on the upside,” he said.
For one, it’s not yet clear where all of the incremental electricity required to power new data centers is going to come from, according to Goldberg. In addition, there’s a lot of leverage building up around their development.
“Once you trace down where all these GPUs are going, you get into the weeds of the neoclouds and all these electricity and property deals that are taking place,” he said. “It’s easy to see how some obscure company fails and that cascades down the rest of the supply chain.”
As a reminder of how quickly the industry evolves, Goldberg’s office features a stack of shelves filled with tech relics: Dozens of cell phones trace the evolution of the device from the novelty of the 1980s to the ubiquity of the current decade. There are also a camcorder and an old server that was developed by Qualcomm but never made it to market.
While there isn’t a spot on the shelf reserved for one of Nvidia’s Blackwell chips that are powering the current generation of AI innovation, it’s reasonable to assume that one day it too will be considered a piece of history.
Even on bullish Wall Street, there is mounting concern that AI enthusiasm has reached a breaking point, with the circular nature of deal financing coming under particular scrutiny.
Goldman Sachs CEO David Solomon recently compared it to the dot-com euphoria. Some money managers are positioning for a shift away from big tech. And a record share of global fund managers see a bubble in AI stocks, according to the latest survey from Bank of America. Even OpenAI CEO Sam Altman said “yes” when asked if there was an AI bubble.
All of which helps give Goldberg the courage to stick to his convictions.
Bloomberg’s Ian King contributed.