SpaceX lowballed its bankers on fees. Goldman Sachs has another way to win big

The real money comes after the opening bell.

It’s been widely reported that SpaceX will provide its underwriters a “gross spread” of just 0.75% or less for shepherding the biggest initial public offering of all time. According to Jay Ritter, a professor at the University of Florida who’s the nation’s top academic expert in the field, that number ties the lowest percentage on record for a conventional IPO. Typically, in debuts raising tens of billions, the banks get a substantially larger cut of between 1% and 3%. For context, the figures when Facebook (2012) and Uber (2019) entered the public markets were between 1.1-1.3%. Still, the amount their institutional and retail clients are paying in total for the rocket and AI giant’s shares on offer is so stupendous that Wall Street will still pocket the largest dollar fee bonanza of all time.

Ritter recounts the scenario the only other time a U.S. IPO got done at a spread this small. In late 2010, the U.S. government—after bailing General Motors from bankruptcy following the great Financial Crisis—took the auto giant public. “Goldman was competing with W.R. Hambrecht & Co., a firm that proposed auctioning GM’s shares instead of using an underwriting. Its founder Bill Hambrecht told me he offered the Treasury a super-low fee of 0.75%.” As it turned out Goldman Sachs agreed to match that fee to win the deal.

Now, Goldman as leader on SpaceX is orchestrating a trophy offering at the same record-thin percentage that’s a good deal for Wall Street simply because the deal’s so immense. Though SpaceX will generate a never-before-seen fount of fees, they’ll still likely furnish a small portion of the most sumptuous total prize Wall Street’s ever seen. And Goldman’s looking to collect the lion’s share.

Let’s start with the reward that will be explicitly stated on the prospectus, and gets most of the media attention, that gross spread. SpaceX has already announced in its latest, amended S-1 that it’s selling 555.6 million shares at $135 in the underwriting, for $75 billion—and is aiming for a $1.75 trillion valuation, biggest on record. The banks are also getting an extra allocation called an “over-allotment” of 15% in case their investors want more once SpaceX starts trading. Since that additional demand is a virtual certainty, the full raise should come to 639 million shares and just over $86 billion. That’s over triple the $25 billion Alibaba amassed in the 2014 IPO that until now ranked as the largest ever mounted on a U.S. exchange. Hence, the firms distributing the shares will split approximately $646 million in fees. That’s more than twice the roughly $300 million for Alibaba, and dwarfs the less than $100 million Uber paid and the around $200 million cost to Meta.

The roster of underwriters ranges from such marquee names as Goldman Sachs and Morgan Stanley to boutique investment banks such as Allen & Co. and William Blair to giant foreign universal banks Societe Generale, Santander and Mizuho. All of these participants get shares to distribute. Ritter reckons that many of the smaller banks will get batches only for their retail clients. They’ll still do well. The fees get split according to the percentage of the shares each bank gets to distribute; if your allocation is 5%, you’d receive 5% of the roughly $646 million. Hence, membership in the SpaceX IPO club will bring a nice share of the fees even to the banks that distribute a few points of the total.

The King’s Ransom will flow from not from fees but “soft dollars,” and Goldman looks to be the big winner

Though the gross spread’s rich, the really big money goes to the player who determines which ultra-hungry funds and brokers get the shares. The banks will send a far bigger part of their portions to the retail side than in most IPOs, an estimated 30% versus the usual 5% or less. Among the recipients are Charles Schwab, Morgan Stanley’s E*Trade, and Robinhood. But while all the 23 firms do the distributing, only one bank decides to what clients the vast bulk of the offering will go. And that’s what’s known as the “lead left underwriter.” SpaceX displays Goldman Sachs in that position, at the top left on the front page of the IPO prospectus. Just after Goldman on the same top row come the four other “joint book-running managers,” Morgan Stanley, B ofA Securities, Citigroup, and J.P. Morgan.

Ritter explains that the joint book-runners will get more than the average percentage of shares to distribute since they dominate the institutional part of the offering, and hence they’ll receive an outsized part of the fees. But they’ll have little say on which hedge and mutual funds, insurers and endowments those shares go to. That’s mainly Goldman’s purview. “Goldman Sachs will allocate the vast majority of the shares on its own authority as lead-left underwriter,” he says.

The big question is how much the stock will jump on Day 1. For SpaceX, a good “pop,” though by no means assured, is likely. As Ritter points out, Wall Street makes a practice of underpricing IPOs. Result: Around three-quarters of the offerings end day one above the underwriting price, and the average increase is 19%. The second amended S-1 disclosed that SpaceX has reserved 5% of the shares to be purchased at the offer price of $135 by employees, friends and family of the executives, and people SpaceX does business with. “That suggests Elon Musk would want a ‘pop,'” says Ritter. “If employees immediately lose money in an IPO, they’re not happy campers.” By the way, that privileged group can sell their stock any time, including hours soon after the opening bell rings on the Nasdaq, since they’re exempted from the lock-up provisions that apply to pre-IPO shareholders including Musk and others in the C-suite.

In exchange for getting underpriced shares, and immediate risk-free gains for their portfolios, the hedge funds and other money managers return part of the bounty to the lead underwriter in what’s called “soft dollars.” That’s officially the amount that the “commissions” exceed the actual cost of executing the trades. “The soft dollar amount paid can be multiple the pennies spent on execution,” observes Ritter. Typically, he says, about 30% of the first day profits boomerang back to the bankers in soft dollars, most of the bounty going to the lead left underwriter.

Let’s say that SpaceX shares finish their first day $27 or 20% higher, at $162. In a single session, the owners anointed by the underwriters, and chiefly by Goldman, would book profits of $17.3 billion. That would set a record for money “left on the table.” (The previous record was the $8 billion from Alibaba’s 2014 offering.) If the nearly one-third of the first day bump that’s the norm flows back to the underwriters in soft dollars, the 20% pop in SpaceX could hand Wall Street a windfall of over $5 billion (30% of the $17.3 billion pop). “And the biggest beneficiary by far would be Goldman as the lead left underwriter that decided who got the shares,” says Ritter. Big as the fees are, even at that low-sounding percentage, the soft dollar fount is around eight-times larger. “I’m expecting two tremendous quarters in sales and trading for Goldman and maybe Morgan Stanley and others helped by SpaceX, plus the profits from the coming OpenAI and Anthropic IPOs,” says Ritter.

AI is disruptive and transformative. But the IPO process doesn’t change, and Wall Street wouldn’t want it any other way.

This story was originally featured on Fortune.com