How Jim Simons Built the World’s Greatest Quant Fund

By John Banks

Jim Simons, the legendary mathematician, hedge fund manager, and quant, retired a few weeks ago from board chairman duties for Renaissance Technologies (RenTech), the firm he started with partners more than 40 years ago. He is stepping back at 82 years old.

Simons and his firm are best known for their Medallion Fund, which first began trading in 1988.

The Medallion Fund is the most profitable quant hedge fund of all time. Simons, as chief architect of the Medallion Fund, thus has one of the greatest investor track records of all time. As of 2021, his personal wealth had compounded to more than $25 billion.

The Medallion Fund — named after mathematical awards that Simons and an early partner had won — is closed to new investors, and mostly runs money for employees of the firm. The fund stopped accepting outside money in 2005.

The size of the Medallion Fund is capped around $10 billion in assets due to size constraints. For more than three decades, the fund has averaged about 40% returns per year, net of all management fees.

If you take out the fees, the Medallion fund has average annual returns in the range of 60 to 70%.

On top of that, the Medallion fund has very few losing weeks, let alone losing months or quarters. Their consistency is remarkable; the fact they’ve kept it up for decades is even more remarkable still.

Normally, when a quant hedge fund finds a series of exploitable edges, other quant funds eventually discover the same edges and profitability erodes. Not with Renaissance. After 30-plus years, they seem uncatchable.

Nor does volatility hurt them. The fund likes “action,” as Simons has called it. The Medallion Fund reportedly made 76% — or more than 116% before its hefty fees — in 2020.

The “secret sauce” of the Medallion fund cannot be replicated, though many have tried.

But it could be eye-opening, and possibly instructive, to get a sense of how they do it.

Simons, a brilliant mathematician, is a recipient of the Oswald Veblen Prize, a kind of lesser-known Nobel Prize for geometry. He is also the co-discoverer of a breakthrough mathematical theory, known as Chern-Simons theory, that is actively used in a range of fields to this day.

In his early twenties, Simons was a “code cracker” — a specialist in detecting and decoding encrypted messages for the Pentagon.

After World War II, code crackers were in a kind of arms race with the senders and receivers of code, who tried to make their transmissions invisible.

If a group of soldiers on the ground had to communicate with air support, or a submarine had to communicate with central command, the idea was to make the coded transmission as hard to detect as possible — ideally to make it seem like static or white noise, and not a transmission at all. 

Simons and other code crackers would attempt to identify these coded signals — which meant spotting faintly detectable patterns in the midst of random noise — and then to figure out the encryption method.

Early in his career, Simons worked for the Pentagon’s Institute for Defense Analyses, a highly secretive outfit that shielded its work from the outside world.

At the Institute for Defense Analyses, Simons learned about more than code-cracking techniques. He also learned how to build a research team, with an atmosphere that was open and collaborative on the inside but closed and secret to the outside world.

Simons was eventually fired from the Institute for Defense Analyses for making public statements in opposition to the Vietnam War. But this gave Simons the chance to recruit other brilliant researchers, from the Institute and elsewhere, to help him start Renaissance Technologies (RenTech).

Through much of the 1980s, RenTech focused on traditional trend-following strategies in commodity markets. But as the years passed, the strategies started to feel crowded.

In the mid-1980s, one of the brilliant researchers on Simons’ team started looking for “ghost patterns” in market prices. This work was grounded in the same type of code cracking that Simons and others had done for the Pentagon: It was all about finding faintly detectable patterns that others could not see. 

In 1988, the Medallion Fund was launched as a combination trend-following fund and quant fund. By 1990, the fund had relaunched with a total focus on the quant side and earned 56% in its first year. That was the beginning of an incredible run that continues to this day. 

The Medallion Fund is unique in many ways.

For example, Simons only hires scientists and mathematicians whose minds are far from Wall Street. There are no economists or fundamental analysts or traditional Wall Street investors or traders. The more removed they are from traditional investing, the better.

The Medallion Fund maintains a library of at least 8,000 signals, based on the short-term patterns detected by the hundreds of scientists who work for Renaissance Technologies.

The Medallion Fund then uses these signals to trade in and out of markets, thousands if not tens of thousands of times per day, with a trading reach that covers exchanges around the globe.

The win/loss ratio for Medallion’s trades is said to be just 2% — a difference of 51% versus 49%.

But because they can apply this edge thousands of times per day, the fund’s performance has the consistency of a giant casino. For a single trade, a 51% edge is almost the same as pure randomness. But with thousands of trades per day, day in and day out, the iron law of averages means that a tiny2% edge, 51% versus 49%, can deliver profitability on a near-constant basis.

The reason the Medallion Fund is size-capped at $10 billion is almost certainly due to leverage. Because profits are so reliable, the fund can use a lot of leverage, and could easily be levered as much as 10 to 1, meaning $9 of borrowed funds for every $1 of capital.

That would mean a $10 billion max size actually means a $100 billion trading footprint, and would further mean the fund’s 60-70% return on $10 billion is, in reality, more like 6-7% on $100 billion (with $90 billion of it borrowed).

One of the reasons the Medallion Fund is so hard to copy — and so hard to beat — is surely because of the infrastructure required to run such a high-precision operation.

In 2016, Simons’ firm quietly filed a 16-page technical document with the U.S. Patent and Trademark Office for “executing synchronized trades in multiple exchanges.” The concept involved using atomic clocks, the most high-precision time instruments on earth, to synchronize order transmissions down to a few billionths of a second.

To replicate the success of the Medallion Fund, a competitor would not only need a close-knit team of some of the most brilliant scientists in the world — with deep training in code cracking and machine-language translation — they would need the ability to not only execute thousands of orders per day, but also to absorb, interpret, and incorporate trillions of data points per day.

In order to keep its signal library updated and refreshed, while constantly rotating into signals that are working and out of ones that have stopped working, Renaissance Technologies has to process insane amounts of data, constantly, and incorporate it into minor system adjustments in real time.

Not only do the fund’s researchers look at all the price data one could imagine — including all of the bid and ask orders that don’t get filled — they consider any form of clean data to be grist for their signal extraction mill, even down to the level of snowfall in New York’s Central Park. (Back when trading floors were physical, weather could have an impact on floor traders coming into the exchange.)

If you want to know more about Simons, and the fascinating journey he took in building the Medallion Fund into a success, an excellent read is The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution by Gregory Zuckerman.

It should be noted, though, that even the Medallion Fund has limits to its success. They likely keep the fund capped around $10 billion because, after the leverage of borrowed money is applied, their signals simply can’t handle greater size.

Then, too, Renaissance Technologies has launched other funds, making them available to institutional investors — but the other funds aren’t nearly as successful, and have even had painful losing years. Why? Because the other funds can’t make use of the same “ghost signals” the Medallion Fund does — so they have to do more conventional things, and thus return to earth in their results.

Looking beyond the spectacular profits, one of the most notable things Jim Simons did — which wound up being a cornerstone of his success — was figuring out how to build a collaborative, tight-knit research team that functions like a family.

The doors to the outside world are closed, but on the inside all doors are open. From the beginning it was a spirit of teamwork, and close partnership among like minds, that made the Medallion Fund a success (and still do to this day).