The mystery about how Rentec makes a ton of money is very real - nobody seem to have a clue. It's most probably indeed due to some excellent math / engineering work, but could very well be a fraud too - wouldn't be a first time that something is 'very secretive' for a reason.
Btw, another interesting trading company is XTX - founded by a Russian-born mathematician Alexander Gerko - they are huge (in terms of trading volumes and profits, not in terms of headcount), and apparently using a lot of modern machine learning.
I really doubt RenTec is a scam. I talked to someone who is friends with a partner there, it seems unlikely it's a scam. What's more likely is that they have an entrenched market position: agreements with counterparties that only they can get, knowledge about tech that is only useful with the rest of the setup, and generally good management.
The thing about this space is everyone thinks it's magical, that you somehow by being smart can build a money machine. My view is it's like any other business, there's a mix of factors that contribute.
XTX is eating a lot of the FX space both by being smart about tech and by focusing on certain things within the ecosystem, for instance not just about being fast but having good distribution as well.
I don't doubt that you may be true, but HN was full of comments explaining how FTX/Alameda is not a scam, but how they use deep knowledge, domain information, how they are ruthless and efficient, how they are scientific and disciplined, how they are "real sharks", etc.
Alameda was legit making money; their strategy was to be long everything with leverage - all the time. Mathematical geniuses in a raising market, lost 100bn in a declining market.
I seriously doubt it's a scam as the flagship medallion fund is only available for insiders (people who actually work at rentec and know how it works).
I can recommend "The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution" by Gregory Zuckerman, but of course don't expect to learn any market-beating strategies from it.
Quanta Magazine, a source for articles often linked here, is funded by his foundation.
Actually the book mentions several market beating strategies, you just need to have read some other books to tie them together. Naturally it's not all there for you as a nice GitHub library, but I recognised the skeleton of a stat arb strategy in there.
I agree that the book indicates the type of data and maybe setup that rentech is trading, but, imo, it still doesn't fully explain the mystery.
I've worked many years in a top systematic fund (~30% yoy, high sharpe return for the internal fund). And this fund was also trading everything that I could see mentioned when people talk about Medallion. Despite this, they were only reaching 30% return on USD 1B (and had to limit the fund's capacity). Medallion is something like 40% on 5-10B.
It's the size of the returns together with the size of the fund that I find baffling re RenTech.
I agree that last part is what requires explanation, how can they have a bit more return on a lot more money.
Answer is probably a combination of getting there first, thus leaving fewer opportunities for others in the data, and slight variation in how they approach the problem leading to discovery of extra capacity. Once you find something you have opportunity cost in changing your system. Island hopping an archipelago, once you find an island with a bunch of fruit, how likely are you to find another even better island?
Could it be that they simply got there first? With such a huge fund that pushes the market around but a strategy that starts to lose above that, couldn't it make sense that the market has "settled around" them, so to speak?
That plus having decades of experience in automating signal discovery.
since this story is a bust, while we're here, has Simons every said anything interesting somebody could point me to? I've listened to various interviews but he doesn't really say much or spark my imagination. With Musk, you can like him or not, but he talks about a lot of interesting stuff, let's you know what he thinks. What Simon's accomplished is phenomenal, but it's just a black box really.
It's because James Simons is a serious and actually intelligent person with depth of knowledge, and thus, he is one that doesn't need to opine on random things. He also makes a lot of money in a very secretive way. I'm not sure why anyone expects him to expound in detail on how Renaissance Technologies does it. Simons also spends more money than some governments on very serious and advanced scientific research. I suppose he's rather busy with all his investments and philanthropy. Musk is busy selling himself and filling his unfillable ego.
You can see in Simons' TED interview with Chris Anderson how he gets rather annoyed with Chris at times, particularly when Chris mentions Simons' "amazing mind" at around 8:30. Simons seems, to me, to be a rather pragmatic and humble person who doesn't particularly enjoy flattery. Chris even throws out a quote by Musk at the very end of the interview, fishing for some similar platitude from Simons, but Simons deftly avoids giving out such an empty quote.
Most interestingly, you can tell Simons' like of science and mathematics and alludes to his interest when he mentions his investment into the origins of life towards the end.
There are two interesting talks or interviews I know of with Simons:
He bought those companies, he didn’t found Them. Anybody can buy a car company or a rocket company if they are rich enough. The only thing he invented was PayPal, and that was not original. And it wasn’t even his idea alone.
And we can like Elon or not (I'm definitely in the latter camp lately) but if it was as easy as you say, there would be tons of other companies making such breakthroughs.
Well, not sure how else to say it but there's 0 benefit in Simons saying _anything_ that might give insight into how Rentec operates. He's been secretive about it his entire life.
Imo, there is no "secret." There is just a bunch of smart people in a room, constantly turning out and testing ideas.
They were some of the first to apply linear regression and what we know as "machine learning" to equities. Edward Thorpe is another dude to look into.
Their trading really took off after hiring some IBM engineers (accounting for their ramp up time). Those engineers were working on speech detection. I think it really is just a matter of having an insanely smart set of people, working on a problem in a room together... with the capital to back it up. IMO, Jim Simons alone couldn't do it _but_ he is/was arguably one of the best mathematicians of our time. That certainly had to have helped.
I got a bit off track, but regarding your comment here:
>has Simons every said anything interesting
What are you looking for? Do you mean in regard to rentec or just in general? I've done a lot of research on them. I could take a stab at answering them. (for reference, I operate a small algotrading shop but have always been interested in the history of algo trading)
I meant in general. I'm a smart educated person and I like listening to smart educated people, doesn't matter what they're interested in, doesn't matter if I like them or agree with them. I find Bill Gates a huge disappointment, not claiming he's dumb, but he just doesn't say interesting things. His philanthropic work is a bore: it's all just too big and obvious, I might as well just look at a list of things the United Nations is up to. Not saying there's anything wrong with it, but him talking about it doesn't interest me.
Same with Simons. He's gotta be smart, but I guess he just doesn't talk.
I don’t know if we’re listening to the same people, but both Bill Gates and Jim Simons say very insightful things. You do need to know what they’re talking about or hinting at. What I’d do to chat with either of them.
For one, the Numberphile interview with Simons another commenter linked to. You can tell a lot about what he thinks is important if you listen carefully.
Yes, it will be hard to get much from Simons. I think he is a man that knows his strengths and weaknesses and I feel he keeps to his domain. He doesn't even talk much about that though, haha.
I collected some info on Rentec in my blog[0] because I wanted to find out more after reading Zuckerman's "The Man Who Solved the Market." I hope you find it useful.
Maybe comes from all his time working in coding breaking and top secret shit for DOD. Yea, I've looked and read, but not a lot. lol. Also very little turn over or public info from people who worked at Renaissance
The markets guys jealousy guard their edge in the market. The less you say the better, less chance of slipping and letting everyone know the recipe to the secret sauce.
Musk seeks attention to help market his businesses.
If I'm reading this correctly, the top stock holding in Euclidean Capital only accounts for 0.1% of Simons' net worth. The fact that the top holdings are biotech doesn't mean much when they're still such a small percentage.
The tldr is there’s some stocks his fund holds that are the result of private VC investments going public. That’s it. I was waiting for some kind of punchline but it didn’t arrive
> Euclidean itself is a cipher — with no ready information on its assets, strategy, or how closely Simons is involved.
> An email to Simons seeking comment was not returned. A spokesman for Renaissance declined to comment on the record.
> The fund has disclosed ownership of ten stocks — mostly biotechnology companies and many lacking revenues.
> The ten positions as of October 30 totaled just $107.7 million, though that is likely only a part of the fund. That’s small beer for Simons, whose net worth is estimated by Bloomberg Billionaires at $25 billion.
The mystery about how Rentec makes a ton of money is very real - nobody seem to have a clue. It's most probably indeed due to some excellent math / engineering work, but could very well be a fraud too - wouldn't be a first time that something is 'very secretive' for a reason.
Btw, another interesting trading company is XTX - founded by a Russian-born mathematician Alexander Gerko - they are huge (in terms of trading volumes and profits, not in terms of headcount), and apparently using a lot of modern machine learning.