TL;DR. A US IPO is a choreographed six-to-twelve-month process. The company files a Form S-1 with the SEC, hires an underwriter syndicate that runs a two-week roadshow, prices the deal the night before trading opens, and then supports the stock in the aftermarket using a 15%-cap over-allotment option called the greenshoe. Underwriters typically take a 7% gross spread on mid-size deals. In 2025, 90 US operating companies went public and raised $39.0 billion, with an average first-day pop of 29.3%.
The moving parts of an IPO
Before the process, there are four groups in the room. The issuer is the company selling stock. The underwriter syndicate is a group of investment banks led by a “lead-left” bookrunner whose name appears in the top-left of the prospectus cover. The SEC reviews the registration statement for adequate disclosure (not merit) under the Securities Act of 1933. Finally, investors — mostly institutions in the initial allocation, then retail once the stock trades — buy the shares.
The exchange (NYSE or Nasdaq) does not price the IPO. Its job is to list the security and run the opening cross on day one.
Step 1: The S-1 registration statement
An IPO in the United States starts with a Form S-1 filing on the SEC’s EDGAR system. The S-1 is the registration statement required by Section 5 of the Securities Act of 1933 before a company can offer securities to the public. It contains the business description, risk factors, use of proceeds, dilution, management’s discussion and analysis (MD&A), executive compensation, related-party transactions, and audited financial statements — typically three years of income statements and two years of balance sheets for a domestic issuer.
Under the JOBS Act of 2012, an “emerging growth company” (annual revenue under a specified threshold, currently indexed for inflation) may submit the S-1 confidentially and only make it public 15 days before the roadshow. That is why you often see a company “file to go public” only weeks before it trades — the confidential draft has been going back and forth with SEC staff for months.
The SEC responds with comment letters. Rounds of amendments follow. The final version of the S-1, once cleared, becomes the prospectus. The version circulated to investors before the price is set is called a red herring, named for the red-ink legend on the cover warning that the information is subject to completion.
Step 2: The underwriters and the syndicate
An underwriter is a broker-dealer that agrees to purchase the shares from the issuer at the offer price and resell them to the public. In a firm-commitment underwriting — the norm for large US IPOs — the bank takes inventory risk between pricing and closing.
Modern IPOs are syndicated. In 2025, the average IPO had 7.9 managing underwriters (Ritter, Table 11); the lead-left bank runs the book and allocates shares, while co-managers add research coverage and distribution. Adding banks is also how issuers buy post-IPO research support: analysts from the syndicate are the most likely to initiate coverage.
For their trouble, underwriters take the gross spread — the difference between the price they pay the company and the price they sell to the public. On US IPOs raising between roughly $30 million and $160 million, the gross spread is almost always exactly 7%. Chen and Ritter documented this in their 2000 Journal of Finance paper “The Seven Percent Solution,” and it has held up: from 2001 through 2025, 93.3% of the 1,141 IPOs in that size band charged exactly 7%. Above roughly $160 million, spreads fall because scale creates real competition among banks. Visa (2008), General Motors (2010), Facebook (2012) and Uber (2019) all paid gross spreads under 3%.
Step 3: Bookbuilding and the roadshow
Once the SEC has cleared the S-1, management goes on the road. A roadshow is a two-week series of one-on-one meetings and group presentations with institutional investors — mutual funds, hedge funds, sovereign wealth funds, pensions — in New York, Boston, San Francisco, London and (sometimes) the Middle East and Asia.
During the roadshow, salespeople at the syndicate banks collect indications of interest: which funds want how many shares, and at what price. This is bookbuilding. The final price is set the night before trading, based on how oversubscribed the book is and how price-sensitive the demand looks. If demand is monstrous, the syndicate can price above the “file range” disclosed in the red herring; if demand is thin, they price below the range or pull the deal.
Bookbuilding is not a public auction. Allocations are discretionary — the lead bank decides who gets what — and there is a long-running academic literature on whether that discretion is used to reward favored clients. This is one reason direct listings (Spotify 2018, Slack 2019, Coinbase 2021) exist: they skip the syndicate and let the market discover the price via a normal opening auction. Ritter counts nine US direct listings in 2025.
Step 4: The greenshoe (over-allotment option)
The over-allotment option, universally called the greenshoe after the 1963 Green Shoe Manufacturing IPO where it first appeared, lets the underwriters sell up to 15% more shares than the base deal size within 30 days of pricing. FINRA Rule 5110(g)(9) prohibits any over-allotment option larger than 15% of the base offering.
Mechanically, the syndicate deliberately sells short up to 115% of the deal to investors. If the stock trades up after listing, the underwriters exercise the greenshoe: they buy the extra shares from the issuer at the offer price (less the spread) and deliver them to close the short. The company gets more cash, everyone is happy.
If the stock trades down, the underwriters do not exercise the greenshoe. Instead they cover the short by buying stock in the open market at the depressed price, pocketing the difference. That buying is what supports weak IPOs in the days after listing — it is not charity, it is short-covering that also happens to be permitted stabilization under SEC Rule 104 of Regulation M.
A worked example: a $100M IPO
Suppose “NewCo” sells 10 million primary shares at $10, for base proceeds of $100 million. The lead-left files a 15% greenshoe, so the syndicate can sell up to 11.5 million shares. The gross spread is 7.0%, or $0.70 per share.
| Cash flow | Base deal only | Greenshoe fully exercised |
|---|---|---|
| Shares sold to the public | 10,000,000 | 11,500,000 |
| Offer price | $10.00 | $10.00 |
| Gross proceeds (to underwriters) | $100.0M | $115.0M |
| Underwriter gross spread (7%) | $7.0M | $8.05M |
| Net cash to NewCo | $93.0M | $106.95M |
| If stock closes at $13 day-one | “Money left on the table” = $30M | $34.5M |
Note the tension: the underwriters set the offer price, the underwriters’ institutional clients receive the allocation, and if the stock pops 30%, those clients — not the company — capture the pop. That is the meaning of “money left on the table.” From 1980 through 2025, US IPOs cumulatively left $250.1 billion on the table on $1.2 trillion in aggregate proceeds (Ritter, Table 1).
How often does the pop happen? Eleven years of data
The average IPO first-day return has been positive every year in the modern era, but the size varies wildly with the cycle. The 2020–2021 window produced some of the largest averages since the dot-com bubble; the 2022 shutdown left a tiny sample (N=38) skewed by a handful of biotech pops.
| Year | # of IPOs | Mean 1st-day return | Aggregate proceeds ($B) |
|---|---|---|---|
| 2015 | 118 | 19.2% | $22.00 |
| 2016 | 75 | 14.5% | $12.52 |
| 2017 | 106 | 12.9% | $22.98 |
| 2018 | 134 | 18.6% | $33.47 |
| 2019 | 113 | 23.5% | $39.28 |
| 2020 | 165 | 41.6% | $61.86 |
| 2021 | 311 | 32.1% | $119.36 |
| 2022 | 38 | 48.9% | $6.99 |
| 2023 | 54 | 11.9% | $11.92 |
| 2024 | 72 | 15.3% | $20.49 |
| 2025 | 90 | 29.3% | $38.97 |
Where beginners go wrong
A few things trip up first-time IPO watchers:
- “IPO price” is not the price you can buy at. Unless you are an allocated institutional client, you buy the stock in the open market after it starts trading — usually well above the offer price.
- The pop is not a mistake. A 15–25% first-day return is roughly the long-run average. It exists because underwriters need investors to willingly take unknown risk on an untested public company, and the discount is the compensation.
- The lock-up is real. Insiders and pre-IPO investors typically cannot sell for 180 days. Expiration days are catalysts — sometimes negative — because a large supply of shares suddenly becomes tradable.
- The prospectus is not marketing. The S-1’s risk factors section is written by lawyers to be comprehensive; read it, especially the pages that describe customer concentration, related-party transactions, and outstanding litigation.
Related concepts and what to learn next
An IPO is one of three main ways for a private company to become publicly traded in the US. A direct listing skips the underwriter syndicate and lets the market discover a reference price via a normal opening auction — useful for companies that do not need fresh cash. A SPAC merger (or “deSPAC”) lets a private company become public by merging with an already-listed shell; the mechanics of pricing and dilution are very different. A Regulation A+ offering is a smaller, retail-friendly registered offering — a fourth path used mostly by micro-caps.
Sources
- Jay R. Ritter, “Initial Public Offerings: Updated Statistics,” University of Florida, Warrington College of Business (July 7, 2026 update). All year-by-year counts, first-day returns, and money-left-on-the-table figures.
- Hsuan-Chi Chen and Jay R. Ritter, “The Seven Percent Solution,” Journal of Finance, June 2000. Original documentation of the persistent 7% gross spread.
- FINRA Rule 5110, Corporate Financing Rule — Underwriting Terms and Arrangements. Section (g)(9) sets the 15% over-allotment cap.
- SEC EDGAR, the official search interface for S-1 filings and prospectuses.
- SEC Rule 104 of Regulation M — permitted stabilization activities during a securities distribution.
Disclosure: This article is for informational purposes only and is not investment advice.