What Is an IPO? The Two Prices, and Which One You Get
Two prices exist on the first day. Institutions get the first one. Everybody else gets the second one, and the gap between them is the whole story.
By Pavel Penev, MScFounder, TradeWize · 10+ years trading the marketsA private company decides to sell shares to the public for the first time. That sale is the IPO — the initial public offering.
Two prices exist on the day it happens. There's the price the shares are sold at, and there's the price they trade at once anybody can buy them. They are almost never the same number, and which one you get is decided before you ever see a headline.
The 10-second version
An IPO is a company selling new shares to the public for the first time, to raise money. Investment banks called underwriters set the offer price and hand most of the shares to their institutional clients. Across 1980-2025, 9,343 US IPOs closed their first day 19.0% above that offer price on average — a gap worth $250.1 billion against $1,190 billion actually raised. If you buy on the first day, you're paying the second price, and over the following three years those two starting points produced very different results.
What is actually being sold
A private company has shareholders already — founders, staff, early investors. What it doesn't have is a public market in its shares.
In an IPO the company creates new shares and sells them. The buyers hand over cash, the company keeps it, and the share count goes up. That's why an IPO raises money and an ordinary day of trading doesn't.
Existing shareholders can join the sale, and then they're called selling shareholders. Their proceeds go to them, not to the company — the prospectus says on its cover page how many shares are being sold by whom.
After that first sale the shares change hands between investors on an exchange, and none of that money reaches the company either. The IPO is the one moment the two things overlap.
What has to happen first
A company can't just announce a sale. It files a registration statement with the regulator — in the US, Form S-1 — and the bulk of that filing is the prospectus: the document describing the business, the terms, the finances and the risks.
Staff review it, comment on it, and eventually declare it effective. That word does a lot of work in headlines, so here is the regulator's own description of what it isn't: "the SEC's declaration of effectiveness does not represent an approval of the merits of the IPO or an indication that the information disclosed is complete or accurate".
Once the offering is priced, a final prospectus is filed — a 424B3 or a 424B4 — carrying the price that the earlier draft couldn't. From then on the company files quarterly and annual reports like every other listed company, Form 10-Q and Form 10-K.
All of it is public and free to read. The filings are the only track record a brand-new public company has.
Who sets the price, and who gets the shares
The underwriters are the investment banks running the sale. They market the offering, collect indications of interest from big investors, and use what comes back to recommend a price. The company signs off on the final number.
Notice what that process is not. Nobody holds an auction and lets the market clear. A room of professionals decides what the shares will be sold for, before anyone can trade them.
They also decide who gets them. In the regulator's words, underwriters and dealers will distribute most of the shares in the IPO to their institutional and high net-worth clients — mutual funds, hedge funds, pension funds, insurers and wealthy individuals.
So there are two ways in, and one of them isn't open to you
You can buy at the offer price if you're a client of an underwriter and get an allocation. Otherwise you buy the shares once they're trading, in the days after the offering, at whatever price they've reached by then. For almost every ordinary investor it's the second one.
The pop, and how big it usually is
Shares priced by a committee the night before tend to open higher than the price that committee picked. The jump is called the pop, and it is not a rare event.
Equally weighted mean first-day return, 1980 to 2025. The line is the 1980-2025 average of 19.0%.
Across 1980-2025, 9,343 US IPOs averaged 19.0% on their first day.
The average hides two things, and both matter. The first is that 1999 is in there: 476 IPOs at 71.2%, which is the dot-com bubble doing what bubbles do. The second is that the median across the whole period is 7.0% — far below the mean, because a handful of enormous first days pull the average up and most IPOs sit near the bottom of it.
The pop is also not a law. In two of the 46 years — 2008 and 2023 — the typical IPO finished its first day below the price it was sold at. 1984 averaged 3.7% across 171 deals.
One number to distrust: a big average on a tiny year
Those yearly averages count every IPO equally, so a quiet year of small deals can print a startling number on almost no money. 2022 is the case: 48.9% equally weighted, but only 14.2% once each deal is weighted by its size — across just 38 IPOs raising $6.99 billion all year.
Somebody pays for that pop
A share sold at one price that closes higher on the same day was sold too cheaply. The difference, multiplied by every share sold, is money the company could have raised and didn't. It has a name: money left on the table.
First-day close minus offer price, times shares offered, summed across each year's IPOs. 1980-2025 total: $250.1 billion, against $1,190 billion raised.
Over 1980-2025 that came to $250.1 billion, against $1,190 billion the companies actually raised. For every dollar collected, roughly 21 cents went to whoever was allocated the shares instead.
1999 was the worst of it: $37.11 billion handed over in a single year, on $64.67 billion raised.
Which raises the obvious question, and it's a real debate rather than a scandal: a pop guarantees the deal gets sold, rewards the investors the bank needs next time, and buys a good first headline. Whether that is worth the discount is argued over every time a big offering prices.
Either way, the discount is real money and it moves in one direction. Away from the company, towards the people who got an allocation.
Learn it by doing
Reading about it is one thing — it clicks when you do it. Practise this hands-on in a free, interactive lesson (Stage 5: How Markets Work Globally).
Try the free lesson →What one offering looks like, priced out
Take an example offering: 20 million shares sold at $24, closing their first day at $33. Round numbers, not a real company.
Price the offering
$24.00 offered, $33.00 at the closeWhat the company raises
7% fee
$446,400,000
$480,000,000 comes in from the buyers of the offering. The underwriters keep $33,600,000, and the company is left with the rest.
Money left on the table
above the offer
$180,000,000
Ritter's definition: the first close minus the offer price, times the 20,000,000 shares sold. The company sold those shares for $24.00 on a day the market paid $33.00.
First-day return, if you were allocated
$24.00 → $33.00
37.5%
That is what a buyer of the offering made by the closing bell — and most of the shares go to the underwriters' institutional and high-net-worth clients, not to a queue anyone can join.
What 100 shares cost you
the later price is higher
$2,400.00 vs $3,300.00
$2,400.00 in the offering, $3,300.00 buying at the first close. Same company, same day, two prices — and the second one is the one open to anybody with a brokerage account.
The company sells at one price and the market trades at another, in the same afternoon. Drag the closing price under the offer price and the gap doesn’t vanish, it flips: the offering was sold above where the shares traded, and the people holding the loss are the ones who were allocated.
This is an example offering, not a real one — 20,000,000 shares at $24.00, closing at $33.00, on a 7% underwriting fee. Round numbers picked to be legible, not figures from any source. “Money left on the table” is Ritter’s measure — see the sources at the end. Nothing here is advice or a forecast.
At those settings the offering leaves $180 million on the table and hands an allocated investor 37.5% by the close. Meanwhile 100 shares cost $2,400 at the offer price and $3,300 at the first close. Same company, same day, $900 apart.
Drag the closing price below the offer price and the whole thing inverts. The deal was priced above what the market would pay, the allocation is a loss on day one, and nothing was left on the table at all. Both outcomes happen.
Two entry points, two different investments
Here is the part that makes the distinction worth caring about beyond the first afternoon.
Take every US IPO from 1980-2024 — 9,253 of them — and measure the next three years twice. Once from the offer price, and once from the first closing price. Same companies, same three years, two starting points.
Average three-year buy-and-hold returns on 9,253 US IPOs, 1980-2024, measured from each of the two prices. Market-adjusted subtracts the return on a broad US index over the same window.
| Measured from | 3-year return | Against the market | Against matched firms |
|---|---|---|---|
| The offer price | 36.3% | −3.3 points | 8.3 points |
| The first close | 19.1% | −20.5 points | −8.9 points |
The last two columns are differences in percentage points, not returns: the IPOs' average three-year gain minus the same three years in a broad index, and minus a matched company of similar size and valuation.
From the offer price, the average IPO returned 36.3% over three years and finished −3.3 points against the market. From the first close, 19.1% and −20.5 points.
That's a gap of 17.2 points over three years, and it's the first-day pop showing up again. The people who were handed shares at the offer price kept it. The people who bought once trading started paid it away.
So an average is doing a lot of work in both rows, and neither number is a forecast for any single company. What the two rows do settle is that "IPOs did well" and "buying IPOs did well" are different sentences with different answers.
The supply that arrives later
On day one, only a slice of the company's shares can actually be traded. That's deliberate, and it's temporary.
Insiders and early investors normally sign an agreement not to sell for a while. The regulator describes it like this: the existing shareholders have entered into a “lock-up agreement” in which they agree not to sell their shares for a certain period of time, typically 180 days.
The shares still exist. They just can't be sold yet. The market prices a small free-floating supply against whatever demand the offering has built up, which is one reason a hot IPO trades where it does.
Then the clock runs out. The regulator's own warning: when lock-up agreements expire, the share price may decline significantly if a large number of shares become available for sale all at once.
So the 180-day mark is a date worth knowing about before you buy, not after. It's in the prospectus.
It isn't a US quirk
Underpricing turns up in essentially every market that has been studied. What differs is how much of it there is.
| Market | Average first-day return | IPOs | Period |
|---|---|---|---|
| China | 159.3% | 5,410 | 1990-2025 |
| Saudi Arabia | 102.3% | 242 | 2003-2024 |
| India | 80.4% | 3,456 | 1990-2024 |
| Japan | 48.7% | 4,306 | 1970-2025 |
| Germany | 21.3% | 868 | 1978-2025 |
| Australia | 20.1% | 2,578 | 1976-2025 |
| United States | 17.7% | 14,079 | 1960-2025 |
| United Kingdom | 15.5% | 5,457 | 1959-2025 |
| France | 9.3% | 929 | 1983-2025 |
| Canada | 6.8% | 816 | 1971-2025 |
Equally weighted average initial returns, from a table covering 55 countries. Each row is that market's own sample and period, and they differ — a bigger number is not a like-for-like comparison with a smaller one.
The range runs from 6.8% in Canada to 159.3% in China, with the United States at 17.7% across 14,079 offerings. Rules on how shares are priced and allocated differ market to market, and the gap between those rows is largely a gap between rulebooks.
If you're buying an IPO where you live, the local rules are the ones that decide whether you can get an allocation at all.
What to read before deciding anything
This article takes no view on any particular offering. But if you're going to look at one, the document exists, it's free, and here is what's in it.
- Risk Factors — what management itself says could go wrong with the business or the shares.
- Use of Proceeds — what the company plans to do with the money it raises.
- Dilution — the gap between what you are paying and what earlier shareholders paid.
- Management's Discussion and Analysis — management explaining, in words, why the numbers moved.
A new listing has no history of public results, no years of quarterly reports to compare against, and no long record of what management does when things go wrong. The prospectus is what stands in for all of that.
And the first day's price is a price, not a verdict. It's what a small, deliberately restricted supply of shares fetched from whoever wanted them most, on one afternoon.
What is an IPO?
An initial public offering is the first time a company sells its shares to the general public. The company files a registration statement — Form S-1 in the US — containing a prospectus that describes the business, the terms and the risks, and investment banks called underwriters price and sell the offering. The money from the new shares goes to the company, which is what makes an IPO a way of raising capital rather than just a listing.
Can an ordinary investor buy at the IPO price?
Usually not. There are two ways in: an allocation at the offer price, which requires being a client of an underwriter, or buying in the market once the shares are trading. The regulator's own bulletin says underwriters and dealers will distribute most of the shares in the IPO to their institutional and high net-worth clients. Almost everybody else is buying at the second price.
How much do IPOs usually rise on the first day?
Across 1980-2025, 9,343 US IPOs averaged 19.0% on their first day — but the median was 7.0%, so most were far below the average. In two years — 2008 and 2023 — the typical IPO ended its first day below the offer price.
What does "money left on the table" mean?
It's the first-day closing price minus the offer price, multiplied by the shares sold: what the company would have raised had it priced at where the shares actually traded. US IPOs left $250.1 billion on the table across 1980-2025, against $1,190 billion raised. The worst year was 1999, at $37.11 billion.
Are IPOs a good investment?
That depends entirely on which price you pay, and the record separates the two. Over 1980-2024, 9,253 US IPOs returned 36.3% over three years measured from the offer price — −3.3 points against a broad index — and 19.1% measured from the first closing price, which is −20.5 points against the same index. Those are averages across thousands of companies, not a prediction about any one of them.
What is a lock-up period?
An agreement by existing shareholders not to sell their shares for a set time after the IPO. 180 days is the usual length. It keeps the tradeable supply small at first, which is part of why a sought-after IPO trades where it does. When it expires, when lock-up agreements expire, the share price may decline significantly if a large number of shares become available for sale all at once.
Does the SEC approve an IPO?
No. It reviews the disclosure and can declare the registration statement effective, but in its own words, "the SEC's declaration of effectiveness does not represent an approval of the merits of the IPO or an indication that the information disclosed is complete or accurate". Effectiveness means the paperwork clears the bar, not that the investment is sound.
See where new shares come from, in the free course
TradeWize's free track runs 20 stages, from what a share is to a finished investor playbook, with a zero-risk simulator wired into the lessons. Stage 5 is how markets work globally: who issues shares, who trades them afterwards, and where your money actually goes. No card, and the investing curriculum stays free.