Stripe Could Have IPO'd Years Ago. It Built a Private Liquidity Machine Instead.

By every conventional script, Stripe should already be a public company. It processed $1.4 trillion in payments in 2024², pulled in roughly $5.1 billion in net revenue³, and crossed into the black with a $101.9 million pre-tax profit after a $1.2 billion loss the year before.¹ That is precisely the profile of a company that rings the bell at the New York Stock Exchange. Stripe did the opposite. It engineered a way to give its investors and employees their cash without ever filing an S-1 - and as of mid-2026, it still hasn't.¹⁰

The official story is that Stripe stays private because it is still in growth mode and an IPO would force it to think small. That is true, and it is also a polite cover for something sharper. Stripe is running a capital-markets arbitrage: it has turned the one thing an IPO is supposed to deliver - liquidity - into a repeatable private routine, while keeping everything an IPO would cost it.

First, the boring part: how the money actually moves

Stripe makes money the way a toll-taker does - a thin cut of every transaction it handles. The shorthand everyone quotes is '2.9% + $0.30,' but that's only the US domestic card rate on the pay-as-you-go tier. International cards run 3.1% + $0.30 with an extra 1.5% cross-border fee; bank transfers via ACH are just 0.8%, capped at $5; in-person payments through Stripe Terminal are 2.7% + $0.05.⁵ The point isn't the exact number - it's that the fee scales with volume, and the volume is enormous. Run that toll across $1.4 trillion of payments in a year² and the slivers compound into billions.

That last line is the whole growth story. A pure processor competes on price and watches its take rate erode toward zero. Stripe's answer is to sell software that rides on the payment rail it already owns. Stripe Billing - its recurring-revenue and subscriptions product - crossed a $500 million revenue run rate in 2024, manages nearly 200 million active subscriptions, and is used by more than 300,000 companies.⁹ A subscription business that processes through Stripe pays the swipe fee and the Billing fee, on the same transaction. The toll-taker becomes the software vendor, and the rate goes up.

The flywheel that replaces an IPO

Here is the part that gets misread. An IPO solves a specific problem: early investors and long-tenured employees are sitting on paper wealth they can't spend, and at some point they want out. The public market is the conventional exit ramp. Stripe built a private one. Since 2024 it has run tender offers roughly every six months - structured buybacks where the company and its backers purchase shares from employees and early holders at a set valuation.¹⁰ The February 2025 tender priced Stripe at $91.5 billion; by February 2026 a subsequent offer closed at $159 billion, a new all-time high backed by Thrive Capital, Coatue, and Andreessen Horowitz.⁷ Liquidity, delivered. No prospectus required.

What that buys Stripe is freedom from the one discipline an IPO imposes: the quarter. A public payments company answers to analysts every ninety days, and any heavy investment that dents next quarter's earnings gets punished by the stock. John Collison's framing is that public markets suit 'the extract stage rather than the expand stage,' and that Stripe is 'still growing very quickly and investing in new products.'⁶ Strip the diplomacy away and it's a wager: that the company can compound faster outside the earnings treadmill than inside it - and let private tender offers handle the liquidity the public market was supposed to provide.

The 2023 round is the cleanest proof of the strategy, even though it's usually filed under 'distress.' Stripe raised $6.5 billion at a $50 billion valuation - roughly half its 2021 peak.⁴ On the surface, a brutal down-round. Read what it was for: the cash went to buying back employee shares and covering the tax bills employees owed on equity that had been minted years earlier and never sold. Stripe said outright it didn't need the money to operate.⁴ This wasn't a company rescuing itself. It was a company paying its people the liquidity an IPO would have provided - and choosing to pay it privately instead.

Isn't this just a stalling tactic with better PR?

The fair objection is that 'staying private' often means hiding. Private companies disclose little, mark their own valuations, and can paper over weakness the public market would expose. The $95 billion peak in March 2021 and the $50 billion mark two years later weren't set by a market clearing millions of trades - they were set by investors in a room.⁴ A skeptic could argue Stripe simply prefers a scoreboard it controls.

But the numbers Stripe does disclose cut against the cynical read. A company hiding weakness doesn't volunteer that it swung from a $1.2 billion pre-tax loss to a $101.9 million profit¹, or that payment volume grew 38% to $1.4 trillion², or that a single newer product crossed a half-billion-dollar run rate.⁹ And the tender offers are the tell that this is choice, not avoidance: the company could exit anytime - the $159 billion 2026 valuation already exceeds its 2021 high⁷ - and it keeps declining. The honest version is that Stripe found a structure where it gets the upside of being public (liquidity, a rising mark, blue-chip backers) without the cost (the quarter), and it will keep using it until the arbitrage stops working.

Stripe makes its money the way every great toll operator does - a fraction of a percent on a flood of transactions it never originates, multiplied by volume until the fractions become billions, then thickened with software sold on top of the same rail. That part is conventional. What isn't conventional is the refusal to cash it in. The Collisons looked at the IPO - the bell, the prospectus, the ninety-day clock - and saw a transaction with a price tag attached. So they built a machine that delivers the proceeds without the price, and they keep running it every six months at a higher number than the last. The genius wasn't avoiding the public market. It was realizing they could borrow its best feature and leave the bill behind.