Goldman vs. Morgan Stanley fight to own lead-left IPO power for OpenAI, Anthropic

Goldman Sachs already grabbed one of the biggest prizes in the IPO pipeline, and now it is running straight back into a new knife fight: the “lead left” book runner spot for the OpenAI and Anthropic listings. In the SpaceX deal scheduled for June 12, Goldman clinched the “lead left” position. That matters because “lead left” is the designation that determines how many shares go to which funds, while other underwriters help on fees and coverage but do not participate in the share allocation.

Fortune reports that both OpenAI and Anthropic have chosen Goldman and Morgan Stanley as their two top book runners, but the issuers have not determined which bank will get the all-important lead left designation. Translation: the deal is not about who is invited, it is about who gets to pull the allocation lever. And the market tells the story in real time through the behavior of hedge funds, mutual funds, insurers, and endowments, which are “super-hungry” for bigger IPO allocations in these mega-offerings.

So why is this “lead left” job treated like a position of hidden kingmaking? Jay Ritter, a University of Florida professor and the “world's top academic expert on IPOs,” explains that the lead left banker controls the share split among funds. Other underwriters are expected to provide analyst coverage and share in fees, but they do not participate in allocation. The consequence is that funds do not just evaluate the bank’s relationship, they evaluate how it will translate into actual shares.

That is where “soft dollars” come in, and yes, the name sounds like marketing. In Ritter’s description, soft dollars are defined as the amount that the commissions paid on trades exceed the bank’s cost to execute transactions. Rule of thumb, as Ritter frames it: the more soft dollars a fund sends the lead left underwriter, the larger the portion of IPO shares they get in return. And if a fund does not know who will win lead left, Ritter says they hedge by sending soft dollars to both firms so they are well placed either way.

The economic logic is brutal. SpaceX is the cleanest example of why the “lead left” spot can create multiples beyond standard underwriting fees. The fees for raising an expected $86 billion are roughly $600 million, but the real Wall Street money flows from the “pop,” the jump over the offer price on the first day of trading. The underwriting process typically underprices the stock to generate that bump, with the pretext that a quick investor gain supports loyalty among long-term shareholders. Ritter’s data shows that around three-quarters of all offerings get a pop, with an average increase of 19%.

In a SpaceX scenario where the stock rises about 20% on day one, investors could gain more than $17 billion. Ritter reckons that at least 30% of that windfall would cycle back to the banks in soft dollars. That estimate lands at roughly $5 billion, which is more than eight times the amount collected in fees, and the lead left underwriter would capture by far the biggest portion. So for Goldman, the SpaceX outcome is not just a headline win. It is proof it can secure allocation power, then monetize it.

Now, the stakes get even bigger in the AI IPO race. Fortune notes that SpaceX represents less than half the total potential take for the “Big Three” AI offerings of 2026, with OpenAI and Anthropic expected to raise at least $60 billion each, totaling $120 billion-plus. At the SpaceX fee rate of 0.75%, the gross spread would come to $900 million for those two deals. The math gets the same additional twist as always: first-day pops can turn soft dollars into a fee-swamping flood.

If both OpenAI and Anthropic ramp 20% on day one, Fortune says funds and a smaller group of retail clients would gain $24 billion across those offerings. If the usual soft dollars boomerang back to the banks, Ritter estimates they'd pocket over $7 billion, again with most of it going to the lead left. Meanwhile, market-implied probabilities show how intensely everyone is already positioning: Kalshi gives Goldman a 73% chance of prevailing for lead left in OpenAI, while PolyMarket puts odds at even for Anthropic. Even though both OpenAI and Anthropic have filed confidential draft registration statements and the timing of debuts is not known, the “soft dollars” behavior is apparently arriving early, based on both banks’ perceived chances.

Financial performance reinforces that expectation. Fortune reports that Goldman grew 18% in sales and trading in 2025 to a record $41.5 billion, while Morgan Stanley expanded 17% to $33.1 billion. In other words, if you are a fund allocating commissions, you are not just paying for access to a hot IPO, you are paying for momentum in the banks most likely to win lead left. Ritter expects more great results from both firms in sales and trading in coming quarters based on that trend.

For executives and boards watching from the sidelines, the strategic takeaway is simple but easy to miss: the “lead left” decision is an allocation lever disguised as an underwriter designation. When IPOs are this massive, the distribution of shares can reshape who gets access to the first-day liquidity event, which then feeds back into future trading incentives and underwriting outcomes. If you are advising a fund complex, underwriting the process, or overseeing capital markets strategy at a peer bank, the question is no longer “who is on the books.” It is “who controls the left side of the prospectus,” because that is where the allocation power, and potentially billions in soft-dollar economics, concentrates.