The AI boom era has led Citigroup to set a $12 trillion market-capitalization projection for SpaceX. Here's what investors need to know about investment-bank research.

After pocketing $500 million in underwriting fees from SpaceX's record IPO, Wall Street banks are now eyeing multitrillion-dollar valuations for the company.

Wall Street celebrated Christmas in June as SpaceX prepared to list on the Nasdaq in a record-breaking public offering.

Lead underwriter Goldman Sachs erected models just a few weeks ago of SpaceX rockets in the lobby of its New York headquarters. Joint lead bookrunner and stabilization agent Morgan Stanley adorned its lobbies, elevator banks and turnstiles with SpaceX logos and pictures of Mars. Bankers wore green Nike sneakers in a nod to SpaceX's "greenshoe" option, which allowed the underwriters to sell more shares than originally planned in the event of heavy demand. JPMorgan hosted hundreds of SpaceX employees at its headquarters in Manhattan to celebrate the stock's listing, capping off the event with a light show featuring rocket launches.

SpaceX's stock hit a high of $225 in June and then sank as low as $104.83 in July. It recently changed hands for $141.50, giving it a market capitalization of $1.9 trillion.

Twenty-two Wall Street banks collected $500 million in fees for underwriting, marketing and allocating shares in the SpaceX IPO, which priced the stock at $135 and raised $75 billion, a regulatory filing shows. Twenty-five days later, on July 7, banks published a flurry of price targets as they initiated coverage of SpaceX as a regulatory quiet period on such research expired. Many of the banks did so with enthusiasm, describing the far-off potential of the company. Deutsche Bank called SpaceX the "apex of civilizational ambition" and put a $255 price target on the stock. JPMorgan analysts said SpaceX's potential impact on humanity was bigger than anything they had ever seen before and projected the stock would hit $225 by December 2027.

Raymond James analyst Brian Gesuale weighed in with a price target of $800 a share, or close to 500% above SpaceX's initial-public-offering price. Gesuale projected that, by 2035, SpaceX would generate $5.2 trillion in annual revenue and $2.63 trillion in free cash flow. Discounting those cash flows back to present value yields a valuation of $896 per share. Apply a 10% haircut for "risk adjustment," and you arrive at $800. Citigroup said SpaceX's stock would hit $200 by the end of the year, "a milestone along the path to $900+," implying a market capitalization of about $12.2 trillion.

SpaceX's stock hit a high of $225 in June and then sank as low as $104.83 in July. It recently changed hands for $141.50, giving it a market capitalization of $1.9 trillion. Most investors who have bought SpaceX shares in the stock market have lost money, at least on paper.

Citigroup, Raymond James, Deutsche Bank and JPMorgan have all maintained their price targets on SpaceX. Raymond James did not immediately respond to a request for comment. Deutsche, Citigroup and JPMorgan declined to comment. Overall, 75% of analysts covering the company are bullish, according to FactSet data.

After a multiyear drought in tech IPOs, Wall Street's equity capital-markets desks are raking in fees again thanks to the artificial-intelligence boom. SpaceX's half-billion-dollar fee pool is expected to be a mere appetizer as Anthropic and OpenAI prepare to launch their own mammoth IPOs in the coming months.

Wall Street's age-old structural paradox is back in the spotlight: juggling conflicts of interest between sell-side research analysts and their employers' dealmaking businesses.

As a result, America's financial institutions, with the exception of a few major skeptics, have found themselves as AI's biggest cheerleaders in a capacity that extends beyond IPO road shows. Wall Street's age-old structural paradox is back in the spotlight: juggling conflicts of interest between sell-side research analysts and their employers' dealmaking businesses.

'There's really very little incentive to be bearish on a stock'

Gwynne Shotwell, president of SpaceX, rang the opening ball as the SpaceX IPO launched on the Nasdaq exchange.

Morgan Stanley expects SpaceX to spend $64 billion on AI this year, much more than earlier forecasts. SpaceX says not to worry about that spending, since it believes it can pay it off in less than a year and is on track to record $100 billion in annualized revenue by the end of 2026.

However, SpaceX's capacity to fund that capital spending largely relies on compute deals with firms like Anthropic and Google that can be reneged on with a few months' notice, according to regulatory filings. Meanwhile, it's so far unclear whether SpaceX's new AI models will be able to snag enterprise customers away from Anthropic and OpenAI, which dominate the U.S. business market. And plans for SpaceX AI data centers in space are likely at least a few years away.

'There's really very little incentive to be bearish on a stock, especially a high-profile stock.'Craig Coben, formerly of Bank of America

"Skeptics will push back that planned capex is higher ... but we believe this all reflects the red-hot AI data capacity demand signal - you've got to spend money to make money, and SpaceX is accelerating its capacity investments accordingly," Stifel's Jonathan Siegmann said in a client note this month, brushing off concerns.

Such bullish attitudes are common right now on Wall Street. There are laws and regulations that are supposed to ensure sell-side analysts can express their sincere views about a company, independent of the business incentives of their firms' fee-making divisions. The banks themselves also have mechanisms to prevent conflicts of interest. But experts say sell-side analysts are still unlikely to be publicly skeptical of prominent companies, specifically ones that are paying their institutions millions of dollars in fees.

"There's really very little incentive to be bearish on a stock, especially a high-profile stock," said Craig Coben, former global head of equity capital markets at Bank of America.

"You're likely to upset a certain constituent. You may upset the company, which means that you can't provide corporate access. You may upset investors, some of whom may be your clients," he told MarketWatch. "You may upset the investment bankers."

Taken to the extreme, conflicted sell-side research fed into the dot-com bubble. Jack Grubman became a prominent telecommunications analyst in the late 1990s and earned large pay packages while issuing buy ratings on stocks like WorldCom and Global Crossing, both of which would later become mired in accounting scandals and file for bankruptcy.

Regulators have sought to eliminate analysts' conflicts of interest, including by tying their compensation to the quality of their research, rather than how much business they've driven for investment-banking colleagues. Research analysts are barred from joining investment bankers on the road shows that precede stock offerings. Banks must disclose active client relationships, and underwriting banks must honor quiet periods following an IPO before issuing research reports.

However, the incentive structure has remained "relatively intact," Coben said. The dynamic of research departments being bullish on companies that their banks underwrite is widely understood to persist in a less formalized, relationships-based manner, said Jay Ritter, emeritus professor of finance and director of the IPO Initiative at the University of Florida.

Investment banks have ethical screens that prevent research analysts from talking to bankers working on an IPO deal, "but there's a complete understanding that if the analyst isn't supportive of the company being taken public by that investment bank, the investment bankers are going to have a problem with that," Ritter told MarketWatch.

Beyond the major Wall Street banks that priced the SpaceX IPO and took the largest share of the fees, minor broker-dealer players also helped with placing the shares and received the smallest slice of fees from the SpaceX IPO, a regulatory filing shows.

In Ritter's view, the junior co-managers were brought in less for deal execution and more for promotional reach. He contended that these banks were "basically getting a little bit of the gross spread revenue ... in return for an understanding that their analyst would cover the company typically with a buy recommendation."

'If I'm an analyst and I have a buy recommendation the company's CFO is willing to talk to me, and if I've got a sell recommendation the CFO won't return my phone calls.'Jay Ritter, IPO Initiative, University of Florida

Sell-side analysts face continual pressure to remain in management teams' good graces, as their research is partially informed by one-on-one meetings with corporate leaders. This leads to a problem: "If I'm an analyst and I have a buy recommendation the company's CFO is willing to talk to me, and if I've got a sell recommendation the CFO won't return my phone calls," Ritter said.

Analysts may also issue "overly optimistic" forecasts to secure private meetings between clients and corporate executives called nondeal roadshows, researchers said in a 2022 study. Brokers of those meetings are "rewarded" with greater commissions by institutional investors, they found.

For equity research departments, such meetings are among the few ways they can claim to drive revenue for their companies, said Matt McClintock, a former sell-side analyst covering retail stocks who now runs his own research firm. Analysts' compensation is also often tied to corporate access, he said - another reason to stay on good terms with executives.

"If you do more corporate access you get paid more," McClintock added. "And if a company is reading your report, and you say something the companies don't like, then they're not going to give you corporate access."