When you open an S-1 (the registration form a company files with the SEC before its shares go public), the headline numbers are easy to spot: how much money the company wants to raise and the price range it's targeting. Tucked in the back, in the section literally called "Underwriting," sits something most readers skip past entirely: exactly how the banks running the deal get paid and how that pay structure shapes the way shares get handed out on day one.
A single bank rarely runs an IPO alone. Instead, banks form a temporary group called an underwriting syndicate, and the way that group splits its fees and shares has a real effect on how a stock trades once it lists. This piece breaks down three things: how the fee gets divided, how the greenshoe (overallotment) option works, and how shares actually get allocated to investors.
The Gross Spread & How Underwriters Actually Get Paid
The gross spread is the difference between what underwriters pay the company for its shares and what they sell those shares for to the public. It's not pure profit. It's meant to cover compliance costs, the risk of the deal, and the bank's management fee all at once. J.P. Morgan's own investor education material describes gross spread the same way: as the difference between the purchase price and the public price (J.P. Morgan)
For decades, 7% was the widely cited baseline gross spread for mid-sized US IPOs, a figure that shows up repeatedly in academic research on IPO fees. But that percentage compresses hard as deal size grows, since banks compete harder for the biggest mandates.
Alibaba's massive 2014 IPO priced its gross spread at roughly 1.2%. Reuters reported the fee came to $300.4 million on a $25 billion deal, after the overallotment was exercised (Reuters via Yahoo Finance). That works out to $25B × 1.2%, or about $300M, confirming the reported figure.
Scanning a sample of recent S-1 underwriting sections directly, the gross spread percentage is usually one of the easiest numbers to locate on the page; it's almost always printed plainly on the prospectus cover or right at the top of the underwriting section, rather than buried deeper in the filing.
The 20/20/60 Split, and Why It's a Norm, Not a Rule
The gross spread doesn't get divided evenly among the syndicate. The most commonly cited framework splits it into three pieces: a management fee, an underwriting fee, and a selling concession.
Gross spread = Public offering price − Underwriter purchase price
Total underwriting fee = Gross spread × Shares sold
Fee pool after greenshoe = Gross spread × Total shares sold, including overallotment
Fee Component | Common Allocation | Who Gets It | What It Pays For |
Management Fee | 20% | lead-left bookrunner and joint bookrunners | Structuring the deal and managing the legal and regulatory process |
Underwriting Fee | 20% | The full syndicate, weighted by risk taken | Compensation for underwriting commitment, risk assumption, and deal execution. |
Selling Concession | 60% | Usually allocated to banks credited with distributing shares, subject to pot and syndicate-account rules. | Direct compensation for distributing stock to investors |
Sami Torstila's 2001 study of 4,186 US IPOs from the 1990s tested this split directly. The exact 20/20/60 split appeared in only 10% of deals in 1990. That share rose to 36% by 1999, still a minority overall (Torstila, Aalto University).
Here's the part worth being precise about. This 20/20/60 split is often described as an industry standard, and it is the most common reference point, butan academic study of 4,186 US IPOs from the 1990s found that only around 18% to 36% of deals actually used the exact 20/20/60 split, depending on how the rounding is measured.
The same research found a clear pattern underneath the variation: as deal size grows, the selling concession's share of the total fee tends to increase, while the management and underwriting fee shares shrink. In other words, the bigger the deal, the more of the fee gets weighted toward whoever actually sells the stock, not whoever structured it.
That's a meaningful nuance for anyone modeling underwriter economics. Treat 20/20/60 as the reference point everyone measures against, not a fixed rule every deal follows.
Why the Selling Concession Gets the Biggest Slice
The 60% weighting toward the selling concession is deliberate. It's designed to push the sales force to actually market the deal and place shares widely rather than sit back. For co-managers, banks that don't have a lead structuring role in the deal, this concession is essentially their entire economic reason for joining the syndicate at all.
Syndicate Hierarchy, Who Actually Runs the Deal
The syndicate isn't a flat group of equal partners. It's structured in tiers that reflect how much risk and reputational capital each bank is putting on the line.
The lead left bookrunner holds the most senior position. This bank originates the deal, runs the book-building process, and has final say on pricing and allocation. Because it carries the highest operational and reputational risk, it also claims the largest share of the management fee.
Joint bookrunners are peer banks brought in to share the workload. They actively participate in the investor roadshow and have a real seat at the table during pricing discussions, earning a meaningful share of the management fee, just less than the lead left.
Co-managers sit lower in the hierarchy. They don't structure the deal or run the book. Their role is mostly distribution and client relationships, which is why their economics run almost entirely through the selling concession rather than the management fee.
The Greenshoe Option, A Quick Look at the Fee Side
Once the price is set, underwriters typically have the right to sell up to 15% more shares than originally planned, a mechanism called the greenshoe option (formally an overallotment option). What matters here is the fee impact: since the gross spread is calculated as a percentage of total proceeds, every extra share sold under the greenshoe grows the total fee pool for the whole syndicate.
SpaceX's June 2026 IPO is a clear real-world example of this at scale. Underwriters fully exercised the greenshoe, expanding the deal from an initial $75 billion raise to $85.7 billion. That extra $10.7 billion in proceeds didn't just grow SpaceX's capital raise; it expanded the fee pool the entire syndicate split. A fully exercised greenshoe often signals that the stock traded strongly enough for underwriters to cover the overallotment through the option rather than open-market purchases.
Reuters confirmed the deal grew from $75 billion to $85.7 billion after the greenshoe was exercised. If the gross spread were 1.0%, the fee pool would rise from $750 million to $857 million, a $107 million increase. If the spread were 1.2%, the pool would grow from $900 million to $1.028 billion, a $128.4 million increase. SpaceX's actual gross spread wasn't disclosed publicly at the time of writing.
A company sells 100 million shares at $20. Gross spread is $1.40 per share, or 7%. Total fee pool equals $140 million. Split 20/20/60; the management fee is $28 million. The underwriting fee is also $28 million. The selling concession comes to $84 million. If the greenshoe adds 15 million shares, that adds $21 million to the pool. Total fee pool rises to $161 million.
Allocation Mechanics & How Shares Actually Get Handed Out
Underwriters usually have broad discretion over institutional allocation, subject to law, issuer instructions, compliance rules, and syndicate procedures. The goal is to balance stable ownership, price support, issuer objectives, client relationships, and aftermarket liquidity.
Modern syndicates mostly use what's called the institutional pot system. Rather than each bank getting its own independent slice of shares to sell, all the shares get pooled into a single inventory controlled by the lead left bookrunner. Institutional investors submit Indications of Interest (IOIs, non-binding statements of how many shares they want and at what price) to various banks in the syndicate, but the lead left bank makes the final allocation call. The exact pot structure varies by deal and by market.
This gives the lead bank three practical advantages: it can reward institutional clients with a track record of holding stock long term, it can suppress day-one selling pressure by favoring funds unlikely to flip, and it gets a real-time read on price sensitivity by watching which investors push hardest for allocation in the final hours before pricing.
Older jump-ball arrangements are less favored in large deals, where the lead-left bookrunner wants centralized allocation control. That approach has mostly faded for large tech and mega-cap deals in favor of the centralized pot since it gives the lead left bank much tighter control over the deal's aftermarket behavior.
Where to Find Underwriter Economics in an S-1
Check the prospectus cover for price to the public, underwriting discount, and proceeds to the issuer. Check the Underwriting section for gross spread and selling restrictions. Check the overallotment clause, usually up to 15% of shares. Check the principal and selling shareholder tables. Check the exhibits for the full underwriting agreement and stabilization language.
Frequently Asked Questions
If the 20/20/60 split isn't followed on every deal, what actually determines how far a real deal deviates from it?
Deal size is the strongest documented factor. Research on the topic found the selling concession's share of the total fee rises steadily as the offering gets larger, while the combined management and underwriting fee share shrinks, likely reflecting the added distribution effort needed to place a bigger deal.
Do co-managers ever get access to the institutional pot or only the lead left bank?
Co-managers can source and submit investor orders into the pot, but they don't control final allocation. The lead left bookrunner makes that call, which is exactly why co-manager economics lean almost entirely on the selling concession rather than a share of pricing power.
Does a bigger syndicate, meaning more banks involved, always mean a lower gross spread for the issuing company?
Not necessarily. A larger syndicate can bring more distribution reach, which sometimes supports a lower percentage spread on very large deals, but it also means splitting the fee pool across more parties. Deal size and competitive tension among banks tend to matter more than syndicate size alone.
Can an issuer negotiate the gross spread percentage before the IPO prices?
Yes. The gross spread is a negotiated term between the company and its underwriters, disclosed in the underwriting agreement filed as an S-1 exhibit, not a fixed regulatory rate. Larger, more sought-after deals typically have more negotiating leverage to push the percentage down.
What happens to unsold shares if a syndicate can't place the full offering?
In a firm-commitment IPO, once the underwriting agreement is executed, the syndicate purchases shares from the issuer and resells them to investors. In practice, weak demand is usually addressed before final pricing through range cuts, downsizing, postponement, or withdrawal.
