10 ms·

I’ve found RenTech to be fascinating over the years and highly recommend the book on it “Man who solved the market” Two big takeaways for his success - 1) He

by probe 6y ago

I’ve found RenTech to be fascinating over the years and highly recommend the book on it “Man who solved the market” Two big takeaways for his success - 1) He was pretty early, and quite contrarian, in betting on computer and quant strategies and thus took the “low hanging fruit” early on (def wasn’t low hanging back then when no one knew or believed in computer trades strategies) 2) From the book, Rentech’s main strategy was based on “reversion to the mean” - I.e “We make money from the reactions people have to price moves”. Trading on how you think OTHERS will trade and systemizing it (ex vol and momentum) is powerful but clearly doesn’t scale when you become the market yourself And a bonus one - despite being a math genius, he basically was failing till he brought on others. He hired the right people (ie those interested in math not finance), created the right environment, took care of logistics, and pushed on a key insight (model to trade). He couldn’t have done it by himself.

hogFeast 6y ago
I actually found the last part was my main takeaway. Even in the early 1990s, Simons had basically checked out of the fund and was mainly doing venture stuff. He clearly made some good hires pre-1990s (I can't remember but the data guy clearly seemed to give them a huge edge over the competition, they clearly had data that no-one had) but it was that sequence of hires after this point that really elevated things: Peter Brown, Nick Patterson, Robert Mercer, etc. Very humbling. Of course, everyone will continue to think the strategies are the secret sauce. Also, I think it highlights that quant investing starts out being very scalable but stops scaling quite quickly (and most similar firms hire people that, on paper, are very smart and get nowhere...so RenTech is the best example of scalability). At the top end, fundamental investing is still more scalable (which is what common sense would indicate). As an aside, the article is totally pointless. Finance professors are engaged in an argument with themselves. They know they believe things that make no sense, and so spend all their time grappling with facts to fit them into their model. Humans do not reason perfectly, when you put a trade on you move the market, effects can last for ages (you have pure arbs that take years to close)...the whole discussion is just non-sensical, and any academic examination of finance should start from reality, not what theories are fun to teach. It is kind of tragic to see intelligent people do this to themselves...but some people just prefer Haskell to Python.
- nightski 6y ago
  Not sure what Haskell has to do with it. I've written algorithms that take in 2D images and produce depth maps with live visualization in OpenGL using Haskell. It's incredibly practical once you learn it. But I also disagree from the point that if physics did what you suggested we'd be no where at all. If they had to start with reality before producing useful models then we would of skipped pretty much all of modern physics today. All models are wrong, but some are useful as they say.
  - hogFeast 6y ago
    "once you learn it"...yes, everything is incredibly practical once you remove all the disadvantages. The point is: some people prefer complexity for complexity's sake, and this doesn't work well in a team environment (where the "once you learn it" part becomes quite relevant, one person who prefers complexity for complexity's sake will take down the whole group, not understanding when something should be simple is an indication of ignorance...finance professors rarely have any understanding of actual finance, their ignorance on this is total). These models aren't useful. Also, the saying is wrong. The reason why is that close to 100% of finance professors will quote that saying (srs, I think I have heard this 20-30 times now) because they use models that are wrong and not useful but this model seems to give them an intellectual reason for doing so: any "wrong" model could actually be good, according to this idea. But wrongness is neither nor there because wrongness for a model is utility, they are identical. The only point is utility. And the reason why these models aren't useful, as I have said already, is that they aren't used outside of academia. Their only utility is giving finance professors something fun to teach. And again, the solution is to build models from the way the world actually is (and btw, these are numerous...almost every successful investor, fundamental or quant, has a systematic process...but these models aren't fun to teach).
    - nightski 6y ago
      Are you implying software isn't complex? Or that imperative languages have low complexity? Haskell takes the complexity of software and provides useful constructs to generalize and abstract some of these complexities. Does it take time to learn? Absolutely. Is it easy for newbies to understand? Definitely not, because it's hard to appreciate their value until you have encountered these issues time and again in software. But it's most definitely not complexity for complexity sake. It can vastly simplify software in practice by restricting the domain in which you are working with a very powerful type system. That is the entire point of it all after all. I'm not sure which models you are talking about - but models such as Modern Portfolio Theory, or Black Scholes, while inherently flawed have been massively useful in the real world. Claiming they aren't useful is simply not true. But again, you don't mention any specific models so it's hard to even know what you are talking about.
WalterBright 6y ago
> but clearly doesn’t scale when you become the market yourself Any market-beating strategy will no longer work when the market adopts it. I.e. if you have such a strategy, keep it to yourself as long as practical.
- fractionalhare 6y ago
  That's not necessarily true, it depends. For example risk parity is common knowledge but it still beats the market. You don't really need any secret sauce to use it effectively. You could do it, personally, and you would probably do well. However if your strategies are well known people typically won't pay you much (if anything) to manage their money, because a bunch of shops will be offering comparable results with the same thing. Continuing with risk parity: there are walkthroughs of how this works with code and math available online: https://cryptm.org/posts/2020/08/01/parity.html https://cryptm.org/posts/2020/08/01/parity.html Note the alpha, beta, volatility and Sharpe measures comparing a straightforward risk parity strategy to SPY. It's not controversial to anyone in the actual industry that you can beat the market on a risk-adjusted basis. Very often the techniques for doing that are well known and can be levered up to safely beat SPY on a total basis with less overall risk. What's truly difficult (and secret) is beating the market by several standard deviations.
  - WalterBright 6y ago
    The link doesn't show it doing better than the S&P 500.
    - fractionalhare 6y ago
      Fair, it's not explicit. I misrecalled the control strategy. But the point still stands for the example in that article: over longer timespans SPY tends to return 7 - 10% or so. It has a beta of 1 (basically by definition). Levering up SPY will give you a better return, but at the cost of exposing you more to market volatility. In comparison the given risk parity strategy has a beta of about 0.5, and a natural return of about 10% (i.e. before leverage). You can safely lever the risk parity strategy to a higher total return than the historical market return without getting your beta beyond 1.
xkjkls 6y ago
> He was pretty early, and quite contrarian, in betting on computer and quant strategies and thus took the “low hanging fruit” early on (def wasn’t low hanging back then when no one knew or believed in computer trades strategies) Also the data back then was much harder to acquire. Bloomberg didn't even exist at the time.
- gabereiser 6y ago
  You could subscribe to a service and get pricing data and news. My father did it (with reverse effect) and I remember him using his fancy 9600baud modem to get his portfolio prices for the day (last 24h summaries, 5m and 1m candles). There was no Bloomberg but there was compuserve and AOL and usenet and various other forms of financial forums.
  - icedchai 6y ago
    Yep, my dad was getting live stock quotes back in the 80's.
MrMember 6y ago
His hiring strategy was fascinating. He specifically avoided people from traditional finance backgrounds. He'd target people with doctorate degrees in math, or electrical engineering, or other non-traditional backgrounds and assume (clearly correctly) that they could pick up any necessary knowledge on finance as needed.

by probe 6y ago

hogFeast 6y ago

I actually found the last part was my main takeaway. Even in the early 1990s, Simons had basically checked out of the fund and was mainly doing venture stuff. He clearly made some good hires pre-1990s (I can't remember but the data guy clearly seemed to give them a huge edge over the competition, they clearly had data that no-one had) but it was that sequence of hires after this point that really elevated things: Peter Brown, Nick Patterson, Robert Mercer, etc. Very humbling. Of course, everyone will continue to think the strategies are the secret sauce. Also, I think it highlights that quant investing starts out being very scalable but stops scaling quite quickly (and most similar firms hire people that, on paper, are very smart and get nowhere...so RenTech is the best example of scalability). At the top end, fundamental investing is still more scalable (which is what common sense would indicate). As an aside, the article is totally pointless. Finance professors are engaged in an argument with themselves. They know they believe things that make no sense, and so spend all their time grappling with facts to fit them into their model. Humans do not reason perfectly, when you put a trade on you move the market, effects can last for ages (you have pure arbs that take years to close)...the whole discussion is just non-sensical, and any academic examination of finance should start from reality, not what theories are fun to teach. It is kind of tragic to see intelligent people do this to themselves...but some people just prefer Haskell to Python.

nightski 6y ago
Not sure what Haskell has to do with it. I've written algorithms that take in 2D images and produce depth maps with live visualization in OpenGL using Haskell. It's incredibly practical once you learn it. But I also disagree from the point that if physics did what you suggested we'd be no where at all. If they had to start with reality before producing useful models then we would of skipped pretty much all of modern physics today. All models are wrong, but some are useful as they say.
- hogFeast 6y ago
  "once you learn it"...yes, everything is incredibly practical once you remove all the disadvantages. The point is: some people prefer complexity for complexity's sake, and this doesn't work well in a team environment (where the "once you learn it" part becomes quite relevant, one person who prefers complexity for complexity's sake will take down the whole group, not understanding when something should be simple is an indication of ignorance...finance professors rarely have any understanding of actual finance, their ignorance on this is total). These models aren't useful. Also, the saying is wrong. The reason why is that close to 100% of finance professors will quote that saying (srs, I think I have heard this 20-30 times now) because they use models that are wrong and not useful but this model seems to give them an intellectual reason for doing so: any "wrong" model could actually be good, according to this idea. But wrongness is neither nor there because wrongness for a model is utility, they are identical. The only point is utility. And the reason why these models aren't useful, as I have said already, is that they aren't used outside of academia. Their only utility is giving finance professors something fun to teach. And again, the solution is to build models from the way the world actually is (and btw, these are numerous...almost every successful investor, fundamental or quant, has a systematic process...but these models aren't fun to teach).
  - nightski 6y ago
    Are you implying software isn't complex? Or that imperative languages have low complexity? Haskell takes the complexity of software and provides useful constructs to generalize and abstract some of these complexities. Does it take time to learn? Absolutely. Is it easy for newbies to understand? Definitely not, because it's hard to appreciate their value until you have encountered these issues time and again in software. But it's most definitely not complexity for complexity sake. It can vastly simplify software in practice by restricting the domain in which you are working with a very powerful type system. That is the entire point of it all after all. I'm not sure which models you are talking about - but models such as Modern Portfolio Theory, or Black Scholes, while inherently flawed have been massively useful in the real world. Claiming they aren't useful is simply not true. But again, you don't mention any specific models so it's hard to even know what you are talking about.

WalterBright 6y ago

> but clearly doesn’t scale when you become the market yourself Any market-beating strategy will no longer work when the market adopts it. I.e. if you have such a strategy, keep it to yourself as long as practical.

fractionalhare 6y ago
That's not necessarily true, it depends. For example risk parity is common knowledge but it still beats the market. You don't really need any secret sauce to use it effectively. You could do it, personally, and you would probably do well. However if your strategies are well known people typically won't pay you much (if anything) to manage their money, because a bunch of shops will be offering comparable results with the same thing. Continuing with risk parity: there are walkthroughs of how this works with code and math available online: https://cryptm.org/posts/2020/08/01/parity.html https://cryptm.org/posts/2020/08/01/parity.html Note the alpha, beta, volatility and Sharpe measures comparing a straightforward risk parity strategy to SPY. It's not controversial to anyone in the actual industry that you can beat the market on a risk-adjusted basis. Very often the techniques for doing that are well known and can be levered up to safely beat SPY on a total basis with less overall risk. What's truly difficult (and secret) is beating the market by several standard deviations.
- WalterBright 6y ago
  The link doesn't show it doing better than the S&P 500.
  - fractionalhare 6y ago
    Fair, it's not explicit. I misrecalled the control strategy. But the point still stands for the example in that article: over longer timespans SPY tends to return 7 - 10% or so. It has a beta of 1 (basically by definition). Levering up SPY will give you a better return, but at the cost of exposing you more to market volatility. In comparison the given risk parity strategy has a beta of about 0.5, and a natural return of about 10% (i.e. before leverage). You can safely lever the risk parity strategy to a higher total return than the historical market return without getting your beta beyond 1.

xkjkls 6y ago

> He was pretty early, and quite contrarian, in betting on computer and quant strategies and thus took the “low hanging fruit” early on (def wasn’t low hanging back then when no one knew or believed in computer trades strategies) Also the data back then was much harder to acquire. Bloomberg didn't even exist at the time.

gabereiser 6y ago
You could subscribe to a service and get pricing data and news. My father did it (with reverse effect) and I remember him using his fancy 9600baud modem to get his portfolio prices for the day (last 24h summaries, 5m and 1m candles). There was no Bloomberg but there was compuserve and AOL and usenet and various other forms of financial forums.
- icedchai 6y ago
  Yep, my dad was getting live stock quotes back in the 80's.

MrMember 6y ago

His hiring strategy was fascinating. He specifically avoided people from traditional finance backgrounds. He'd target people with doctorate degrees in math, or electrical engineering, or other non-traditional backgrounds and assume (clearly correctly) that they could pick up any necessary knowledge on finance as needed.