Jim Simons on Medallion, Math, and the Art of Not Following the Pack
Jim Simons recounts a career that moved deliberately from pure mathematics to cryptography to academia to quantitative finance. His work on minimal surfaces and what became Chern-Simons theory was done without any investment thesis—only later did physicists like Witten find it essential. He regards the unpredictability of intellectual payoff as a defining feature of serious research, not a bug: "you do a piece of work and you don't really know where it's going to end up."
Renaissance Technologies and its Medallion Fund are described as the product of a deliberate hiring philosophy: no one from the investment industry. Simons recruited physicists, mathematicians, statisticians, and astronomers, operating on the belief that you can teach a scientist finance but not the reverse. The fund's edge came from accumulating many small market anomalies, running them through massive computing infrastructure, and testing every new signal rigorously in weekly all-hands science meetings before adoption—a collaborative culture he contrasts sharply with rival quant firms that silo teams and pay them on their individual system's performance.
On markets and the future of quant investing, Simons is measured. He doubts quant funds have made markets significantly more efficient, sees a durable role for value investing in the Buffett tradition, and thinks machine learning will prove useful without being transformative on its own. He notes Renaissance is still expanding computing capacity and that the anomalies, though minor, have persisted.
His principles for success are blunt and consistent: - Be guided by beauty in your work - Surround yourself with the best people and get out of their way - Don't run with the pack—if everyone is solving the same problem, find a different one - Don't give up easily, but don't confuse persistence with stubbornness - Hope for good luck—which he calls, only half-jokingly, the most important principle
Outside Renaissance, the Simons Foundation runs three philanthropic grant arms (math/physical science, life science, autism research) and the Flatiron Institute, an in-house computational research operation spanning astrophysics, quantum physics, neuroscience, and mathematics at roughly $100 million per year. He also founded *Quanta Magazine*, claiming his only contribution was the name. His memoirs-in-progress carry the title *Mathematics, Common Sense, and Good Luck*—a fair summary of the man's own account of his life.
Jim Simons traces his path from mathematician to founder of the most successful hedge fund in history, explaining how Renaissance Technologies built its edge by hiring scientists—not financiers—and hunting for market anomalies. He shares his core principles: pursue beauty, avoid the herd, hire slow thinkers if they're brilliant, and hope for good luck.
this fund has been the most successful hedge fund ever in terms of its returns annual returns
every day on average four papers are published rough in the physics archives four papers are published on average in invoking uh talking about churn simon
jesse douglas won the first fields medal i think in 2005 by showing that yes you could always do this with any frame and it would always be a smooth surface and it was called the plateau problem
up through dimension seven you could always ambient dimension seven you could always have if you had a boundary of dimension uh five you could always fill it fill it in with the minimal variety of dimension six
a guy named bombieri uh three italians showed that it was actually a counter example
i solved an outstanding problem in that curious field for which a special purpose computer was built in the national security agency and as far as i know that computer is still running to implement my little algorithm
you can teach a physicist finance but you can't teach a finance person physics
we didn't hire people from the investment industry
we hired statisticians physicists astronomers mathematicians
stock data and so on is open to the to the public i i don't think we have any uh we have a tremendous amount of it tremendous amount of it uh tick by tick even
tick data means every time a price changes which is every maybe 10 seconds or something like that or minutes
machine learning which is coming along and i believe that uh that will be uh useful
warren buffett kind of investing i think always will have a place
i went to stony brook at 30 years old to uh head their math department
about five years later i think it was whitten in princeton he was a physicist started using that math and then other physicists use that math
we had about 300 people working for us
Norway's sovereign wealth fund manages one and a quarter trillion dollars and is managed classically
there are quite a number of quant funds already out there besides renaissance and the medallion fund
A candidate rejected by interviewers for being too slow was hired after his Harvard thesis advisor said 'he solved the hardest problem I had' and went on to be extremely good at Renaissance
Simons attempted to hire Dennis Sullivan by admitting Sullivan's girlfriend as a graduate student in math, hoping it would anchor him, but Sullivan went to MIT instead
The Simons Foundation's in-house research operation at the Flatiron Institute costs about 100 million dollars a year
Simons Foundation in-house research units are: computational science led by Leslie Greengard (~30-35 people), astrophysics, quantum physics led by Antoine Georges, computational mathematics, and now computational neuroscience
Quanta Magazine was founded by the Simons Foundation, the name was Jim Simons's only contribution, and its readership keeps growing
quant funds may have made the markets somewhat more efficient but Simons is not sure, and he does not think markets are very much more efficient today
Renaissance could not operate without computers and is currently building more computing capacity
it was not a quant approach we didn't develop i didn't develop any at first it was just a high
it was just me and uh i've read the papers the newspapers i try to get a sense of where the major currencies might be going
i was not bad at it uh i was not bad at it so i hired a few other people
fundamental training it was not technical at all
i'm not the fastest runner i'm not the fastest thinker and i'm a pretty good thinker but i'm not so fast so do something original don't follow the pack
Renaissance hiring process included solving problems on the board, but Simons was not keen on that method
Other quant firms have groups with separate systems and are paid based on how their individual system does, keeping secrets from each other — Simons does not think that is a good way to work
be guided by beauty surround yourself by the smartest and best people you possibly can have and let them do that thing don't sit on top of them
don't run with the pack that's one of my if if everyone is trying to solve the same problem or a whole group of people if that's the latest and greatest thing to do don't do that
don't give up easily stick with it stick with it not forever but really give it a chance to uh to get where you're going
Adding a quant trading arm to a sovereign wealth fund would be 'another leg in the stool' rather than a replacement strategy
At Renaissance, once a week all scientists meet to share what they are working on, and new predictive signals are presented, discussed, and tested thoroughly before adoption
Simons Foundation philanthropy is organized into three grant-giving units — math and physical science, life science, and autism research — led by high-quality scientists with domain expertise
The way to make money in quant investing is to find anomalies — minor anomalies that have continued — and accumulate enough of them to build up returns
Shared with Earmark · earmark-ai.com