Two Sigma, Jane Street, and the End of the Floor

The trading floor died the way most things die in finance: not all at once, but in a series of migrations that each seemed reasonable at the time and collectively amounted to extinction. The open outcry pits at the Chicago Mercantile Exchange, the NYSE specialist posts, the commodity rings in London; they didn’t close because someone decided floors were obsolete. They closed because the people making money had already left for rooms full of servers.

The firms that inherited the quant revolution after Renaissance Technologies proved it could be done are not imitators. They are the second generation of a species that evolved from Simons’ original proof of concept, and they built something that looks nothing like what came before. Two Sigma and Jane Street are the clearest examples of what quant finance became once it stopped being an experiment and started being the architecture of the market itself.

Two Sigma Built the Factory

Two Sigma was founded in 2001 by David Siegel and John Overdeck, both of whom had passed through D.E. Shaw, the fund that Jeff Bezos left to start Amazon. The Shaw lineage matters because Shaw was the first major fund to treat quantitative finance as a technology problem rather than a math problem. Renaissance had the math. Shaw had the engineering. Two Sigma took the engineering and scaled it into something that looks more like Google than Goldman Sachs.

The fund manages roughly $60 billion in assets. It employs over 1,500 people, and a substantial portion of them are software engineers, data scientists, and machine learning researchers who would be equally at home at a big tech company. The trading floor aesthetic; the shouting, the hand signals, the guys in colored jackets; is not just absent from Two Sigma. It is conceptually impossible there. The firm’s entire thesis is that markets generate data, data contains patterns, and patterns can be extracted by systems that operate at scales and speeds no human trader could match.

What Two Sigma represents in the evolution of quant finance is the industrialization of the insight. Simons proved that statistical arbitrage works. Two Sigma proved that you could build an institution around that proof. Not a cult of personality, not a secretive research lab, but a machine; a technology company that happens to produce returns instead of software products. The firm publishes research. It sponsors academic conferences. It maintains open-source software projects. It recruits from the same talent pool as Facebook and Google, and it competes for that talent by offering the same things: interesting problems, massive data sets, and the resources to explore them.

This is a fundamentally different organism from Renaissance. Medallion operated like a sealed laboratory; no outside money, no public presence, no explanations. Two Sigma operates like a platform. The secrecy isn’t gone, exactly; nobody is publishing their actual trading strategies; but the culture is outward-facing in a way that Simons would have found alien. Two Sigma wants to be seen as a technology leader. Renaissance wanted to be invisible.

The difference matters because it reflects two possible futures for quant finance: the closed kingdom and the open factory. Simons chose the kingdom. Two Sigma chose the factory. Both print money. But only one of them is building something that could survive the founders.

Jane Street Is the New Black Box

Jane Street is harder to talk about than Two Sigma, which is itself a signal worth paying attention to.

The firm was founded in 2000. It operates primarily as a market maker, which means it provides liquidity; it stands in the middle of transactions, buying from sellers and selling to buyers, collecting the spread between those prices. Market making is one of the oldest functions in finance, older than stock exchanges themselves. What Jane Street did to market making is roughly what Amazon did to retail: they didn’t invent the function; they rebuilt it from first principles using technology, and in the process they became so efficient at it that the older version of the function essentially ceased to exist.

Jane Street handles a staggering percentage of global ETF trading volume. The exact numbers are difficult to pin down because the firm publishes almost nothing about its operations, but industry estimates consistently place it as one of the largest ETF liquidity providers in the world. When someone buys or sells a share of SPY or QQQ or any of the thousands of exchange-traded funds that have become the default investment vehicle for most of the planet, there is a meaningful probability that Jane Street is on the other side of that trade.

The firm reportedly generated over $10 billion in net trading revenue in a single recent year. For context, that figure would make Jane Street more profitable than most of the investment banks that the general public thinks of as the apex predators of finance. Goldman Sachs’ entire trading division generates revenue in the same neighborhood. Jane Street does it with roughly 2,500 employees. Goldman does it with tens of thousands.

The efficiency gap is the whole story. Jane Street can provide tighter spreads, faster execution, and more reliable liquidity than traditional market makers because its systems are better. Not marginally better. Categorically better. The firm hires from math olympiad winners, competitive programming champions, and the top PhD programs in quantitative fields. It pays them accordingly; starting compensation for new graduates reportedly exceeds $300,000, and experienced traders and engineers earn multiples of that. The firm doesn’t need a large headcount because the systems do the work. The humans design, maintain, and improve the systems. The systems execute.

What makes Jane Street genuinely strange in the landscape of modern finance is the combination of its scale and its opacity. This is a firm that touches a meaningful fraction of global equity market activity every single day, and most people outside the industry have never heard of it. It doesn’t advertise. It doesn’t do media interviews. It doesn’t have a Wikipedia page that tells you much of anything useful. It publishes some technical blog posts and sponsors some open-source work in the OCaml programming language, which it uses as its primary development platform; a choice that is itself revealing, because OCaml is a functional programming language that almost nobody else in finance uses, which means Jane Street’s entire codebase exists in an ecosystem that makes it structurally difficult for competitors to poach their systems.

That last detail is worth sitting with. When a firm chooses an obscure programming language for its core infrastructure, it is not making a technical decision. It is making a moat decision. Every engineer trained in Jane Street’s systems becomes more valuable to Jane Street and less portable to competitors. The knowledge is embodied in a language that the rest of the industry doesn’t speak. This is the kind of strategic thinking that doesn’t show up in annual reports because the firm doesn’t produce annual reports.

The Floor Trader Had No Chance

The open outcry trader; the guy in the pit with the hand signals and the colored jacket; was already a nostalgia act by the time Jane Street hit its stride. But it is worth understanding precisely why, because the reasons illuminate something about what markets have become.

A floor trader’s advantage was information asymmetry generated by physical presence. Standing in the pit, you could see who was buying, who was selling, how urgently, and in what size. You could read the body language. You could hear the voice. You could feel the energy of a crowded pit and know, before the price moved, which direction it was about to go. This was real edge. It made real money. And it was entirely dependent on the market being a physical place where humans met to trade.

Electronic trading destroyed that edge not by replicating it but by making it irrelevant. When the market moved to screens, the information asymmetry shifted. Now the advantage went to whoever could process the data coming off the electronic order book fastest; whoever could see the pattern in the stream of bids and offers and execute on it in microseconds rather than the seconds or minutes it took a human to shout across a pit.

The quant firms didn’t just replace floor traders. They replaced the entire epistemology of trading. The floor trader knew things through embodied experience; years of standing in the pit, reading crowds, developing intuition. The quant firm knows things through statistical analysis of data at a scale that no human could process. Both are forms of knowledge. One of them scales. One doesn’t.

The floor at the NYSE still exists in a ceremonial capacity. The opening bell still rings. People in suits still stand around looking important for the cameras. But the actual price discovery; the mechanism by which markets determine what things are worth; happens in data centers in New Jersey and Illinois, in server racks owned by firms like Jane Street and Citadel and Two Sigma and Virtu. The floor is a set. The data center is the theater.

What They Actually Built

The world that Two Sigma and Jane Street and their peers created is not the world anyone designed. It is an emergent system; the product of thousands of firms each optimizing locally, each trying to extract a few more basis points of edge, each building systems slightly faster and slightly smarter than the last iteration.

The result is a market that is, by most measurable standards, more efficient than any market in human history. Bid-ask spreads are narrower. Transaction costs are lower. Liquidity is deeper. Price discovery is faster. An individual investor in 2026 gets a better execution on a stock trade than a professional floor trader got in 1996. That is a real, material improvement, and the quant firms deserve credit for it.

But the same architecture that produces efficiency also produces fragility. When the liquidity providers are algorithms, and the algorithms all respond to the same classes of signals, the liquidity can vanish simultaneously. It happened in 2010. It happened again in 2015. It has happened in smaller, less publicized ways dozens of times since. The quant firms didn’t set out to create a market that could lose a trillion dollars of value in ten minutes. They set out to make markets more efficient. They succeeded. The fragility came free.

Jane Street and Two Sigma don’t cause flash crashes. But they are load-bearing elements of the structure that makes flash crashes possible. The market they operate in is a market they substantially created; not through conspiracy or coordination, but through the simple accumulation of rational decisions by intelligent systems that all happen to work the same way.

The floor trader couldn’t have caused a flash crash. He didn’t have the power. But he also couldn’t have made the market this efficient, this liquid, this cheap to participate in. The trade-off between efficiency and fragility is not a bug in the quant-built market. It is the market. It is the thing they made.

The Question Nobody Asks

The real inheritance of the quant revolution is not Two Sigma’s assets under management or Jane Street’s trading revenue. It is the fact that markets now run on an infrastructure that almost nobody understands. Not the regulators. Not the politicians who write the rules. Not the retail investors whose retirement savings flow through these systems every day. Not even, in some meaningful sense, the firms themselves; because when you build a system complex enough, the system develops behaviors that the builders didn’t predict and can’t always explain.

The floor trader was legible. You could walk onto the floor and see what was happening. You could understand, at a human level, how prices were being set. The quant market is not legible. It is a network of algorithms responding to each other at speeds that are literally faster than human thought, making decisions based on patterns in data that no human selected and no human reviews. It works beautifully almost all of the time. When it doesn’t work, nobody can explain why in real time, because the explanation requires understanding the interaction of thousands of systems that were not designed to interact.

Two Sigma and Jane Street are the apex of this architecture. They are the firms that most fully embody what quant finance has become: brilliant, profitable, efficient, opaque, and operating at a scale that makes the old world of floor trading look like a village market. The floor trader is dead. They built the thing that killed him. Whether what they built is better depends on what you mean by better; and who you think markets are for.