RESEARCH
·
No. 139
·
The IPO Playbook
Our approach with new IPOs, and what the last 25 years of data say about the first two years.
Jacob Radke
3 min read
First twenty-four months of performance for 56 large U.S. IPOs since 2000, indexed to each stock’s day-one open.
Source · Yahoo Finance and Fjell Capital working cohort
Median return at month 24
−12%
Across the large U.S. IPO cohort since 2000.
25th percentile at month 24
−54%
Downside range across the IPO cohort.
Above day-one open
44%
Share of the cohort above its day-one open at month 24.
Category
Charts
Share
SpaceX, Anthropic, OpenAI, and Stripe all have IPO rumors. The combined private-market valuations run into the trillions, and some of them will eventually be public.
Before any of them do, this is what the last 25 years say about the first 24 months for a large IPO.
The cohort is 56 U.S. companies that went public since 2000 at $5 billion or more in market cap. SPACs are excluded. Each name is indexed from the day-one open, the price you can actually transact at, not the offering price most investors do not get allocated to.
The median new IPO is down 12% two years in. The 25th percentile is down 54%, and only 44% of the cohort is above its day-one open at month 24, which means a majority is below it.
There is a wide spread above the median too. Google, Peloton, and Reddit each tripled. Meta worked through a rough debut and compounded. For every name that worked, there is a Coinbase, a Rivian, or an Affirm down 80% or more two years on.
Our approach with any new IPO is to wait. We do not buy in the first six months. After that window there is actual trading history, at least one earnings report, and usually the first lockup expiry, enough to assess whether the price reflects the business.
History does not tell us which side of the band the next batch lands on. It tells us the dispersion is real, and the buyer with six months of data has a better dataset than the buyer on day one.
Until next week,
Jacob
Sources & method
The cohort is a hand-curated working list of 56 U.S. IPOs since 2000 with offering market capitalizations of at least $5 billion. SPACs are excluded and direct listings are included.
Five delisted names could not be pulled from Yahoo Finance: TWTR, ZNGA, LNKD, YNDX, and KFT. Including them would likely pull the bands and median slightly lower. The y-axis is capped at +150%; GOOGL, PTON, and RDDT exceeded that level.
Day-one open is the index baseline rather than the offering price because it represents the price a retail investor can actually transact at. Source: Yahoo Finance and Fjell Capital working cohort.
Written by
Jacob Radke
Partner, Wealth Advisor
Jacob writes the weekly charts. Jacob leads the firm’s platform, including trading, research, portfolio systems and technology. He also sits on the firm’s board and investment committee, and advises client families.
Contact
KEEP READING
More from the archive
M&A and IPOs lead the banks higher
The six largest U.S. banks have returned roughly twice the market since the start of 2025. The deal machine is why.
READ
The AI Trade That Cost $45 Billion
The index finished July almost exactly where it started, yet two sides of the biggest trade of the year moved about fifty points apart.
READ
–$5.9 Billion
Alphabet’s free cash flow went negative for the first time in its public history. Not because the business is shrinking, but because it’s spending.
READ
tmrw, delivered straight to your inbox.
One letter a week, and the charts on Tuesdays. No sales material, and no list sharing. Unsubscribe in one click.
Like what you are reading?
Let’s talk about it.
If something we published applies to your own situation, that is the right reason to call.
See if you are a fit
Educational content only. Nothing published here is personalized investment, legal or tax advice, a recommendation of any security, or an offer to buy or sell. Views are as of the date of publication and are not updated. Past performance does not guarantee future results.
© 2026 Fjell Capital, LLC