Just discovered something wild. A company with barely 3000 employees—Jane Street—made $20.5 billion in net trading revenue last year. That's more than Citigroup's entire trading division. More than Bank of America. And these are institutions with hundreds of thousands of employees.

I'm not exaggerating. Citigroup had $19.8B, Bank of America had $18.8B. Jane Street beat them both. The efficiency gap is almost absurd.

But here's the thing: most people have never heard of Jane Street. It was founded back in 1999 by three traders who left Susquehanna and a programmer from IBM. They started with something unglamorous—arbitraging ADRs in a windowless office. No grand narrative, no disruption ideology. Just an obsession with finding tiny price gaps and executing faster than anyone else.

Then they made a bet on ETFs when the market was still sleeping on them. And that decision changed everything. By the time institutions and retail investors woke up to ETFs, Jane Street had already built the infrastructure. Today, the company handles 24% of US ETF primary market share, 41% of bond ETF volume. Every time you buy an ETF, there's a solid chance Jane Street is on the other side of that trade.

What really gets me is their culture. No CEO. No hierarchy. About 40 senior employees run the show collectively, holding roughly $24 billion in equity. They don't use non-competes because they believe their advantage isn't in any single algorithm—it's the culture itself. Can't be copied.

And they code in OCaml. A functional programming language almost nobody in finance uses. Why? Because one line of buggy code can cost hundreds of millions. OCaml's type system catches errors at compile time. Side effect: Jane Street engineers are basically locked in. Their skills don't transfer. It's genius, actually—technology as a moat.

The hiring process is notorious. They don't want finance people. They want problem-solvers. Your interview? Probability puzzles. Game theory. Expected value calculations under pressure. They're testing your ability to think clearly in chaos, not your industry knowledge. And the pay? Starting interns get $300k. Experienced traders? Way more. SBF made $1M in his third year there, with projections of $75M annually if he'd stayed.

Speaking of SBF—here's where it gets interesting. In 2016, Jane Street commissioned him to build an election prediction system. Goal: know the results before CNN, then trade faster. On election night, when Trump's Florida data came in, the system flagged it immediately. Jane Street shorted the S&P 500 with positions worth several billion. They were up $300 million when SBF went to sleep.

Three hours later, the market reversed. Investors saw Trump as pro-business. Stocks rallied. Jane Street's shorts got squeezed. That $300M gain became a $300M loss overnight. A $600M swing. And Jane Street didn't punish SBF. They praised his forecasting system. The mistake wasn't math; it was market psychology.

But Jane Street's real story gets darker when you look at the legal stuff. In India, SEBI investigated them for market manipulation. The pattern they described is wild: on options expiration days, Jane Street's algorithm would buy massive amounts of index stocks and futures in the morning—sometimes over 20% of total volume. Then in the afternoon, they'd reverse, selling everything and pushing the index down. Their short option positions would print money. On one day SEBI examined, Jane Street lost $7.5M on spot and futures, but made $89M on options. Net: $81.5M.

From January 2023 to March 2025, SEBI says Jane Street made about $4 billion in India. Meanwhile, 93% of retail options traders in that same market lost money. SEBI suspended their trading privileges, froze accounts, demanded $560M in escrow. Later they were allowed back, but under investigation.

Then there's Terra/Luna. Todd Snyder, the liquidator, sued Jane Street in February 2026. The core accusation: insider trading. A Jane Street employee named Bryce Pratt used to work at Terraform and stayed connected to people there. In May 2022, hours before UST depegged, Terraform withdrew $150M from Curve. Ten minutes later, a Jane Street wallet withdrew $85M from the same pool. Together they drained the liquidity support. UST started falling. Jane Street allegedly made hundreds of millions covering exposure before the collapse hit the news. Then Pratt supposedly messaged Do Kwon offering to buy Luna at a discount.

Jane Street called the lawsuit "extortion." They're not wrong that Do Kwon committed fraud—he got 15 years. But the lawsuit raises a real question: if you have inside information and you exit before everyone else, is that just good trading or something darker?

Here's what I think: Jane Street is all of these things at once. The most profitable trading firm on Wall Street. A puzzle-solving machine that attracts elite talent. A company operating in the gray zones where information asymmetry, speed, and regulatory ambiguity intersect. They're not breaking laws in an obvious way. They're just better at finding edges that most people don't see—or can't see.

The math questions in interviews, the Terra mystery, the India investigation—they're all connected to the same thing: Jane Street's entire existence is built on information advantages. Finding what others miss. Moving faster. Operating at the boundary of what's allowed.

So who exactly is Jane Street? They're a collection of puzzle solvers, like they say. But they're also a reminder that in financial markets, the biggest profits often come from the deepest mysteries.