4 ms·

I have a feeling this could be one of those stories on HN where it gets a bunch of upvotes but not very much discussion. If you enjoy pop economics/trading boo

I have a feeling this could be one of those stories on HN where it gets a bunch of upvotes but not very much discussion. If you enjoy pop economics/trading books ala Michael Lewis then I'd recommend Fortunes Formula. Great book focusing mostly on Ed Thorpe who is one of the fathers of modern day quantitative trading. https://www.amazon.ca/Fortunes-Formula-Scientific-Betting-Casinos/dp/0809045990 https://www.amazon.ca/Fortunes-Formula-Scientific-Betting-Ca... I don't want to speak for an industry, but I don't see it used very much in practice in investing anymore, mostly because firms nowadays seem to make much smaller bets much more frequently( think HFT firms) so even pension funds that don't do investments under say $20-$30 million are usually doing sizes that amount to 1/50 of Kelly. It may have use for very small individual investors, think someone with under $100,000 to invest that wants to pick their own holdings and not jut buy the market via an ETF. Having said that, if you want to work on the buy side soon er or later someone will ask you about it during an interview so its in your best interest to understand it.

chillydawg 10y ago
Kelly has almost no use for your random small investor. Kelly is all about when you have a very good idea of your edge over the market (first assumption is that you have an edge, very few do). If you're a bedroom trader, it's unlikely you have enough data and a big enough trading history to accurately gauge your market edge. You might well be up 30% in the last year, but that does not mean you've got a 30% edge. When you do have a good estimate of edge, then kelly can come into play. But even then, the theoretical basis of kelly is that money in infinitely indivisible and you can trade at tiny fractions of your total bankroll. So, you normally need to bet less than the kelly stake, but how much less? It's a black art in markets where edges vary with time, along with your skill, perceived skill and measurable skill.
gwern 10y ago
I second the recommendation. I expected it to be mediocre at best, but _Fortune's Formula_ is actually both good on the the technical & philosophical aspects of Kelly, and on the history & people involved.
- hudibras 10y ago
  William Poundstone is one of my Guilty Pleasures. I've been reading (and re-reading...) his books for years, starting with his Big Secrets series, and he is always interesting and thought-provoking, while not taking himself too seriously. Think a better Malcolm Gladwell, but with a worse PR team.
pvitz 10y ago
Also, if you know a little bit about finance and are interested in Ed Thorp, read his papers. For example, https://books.google.de/books?id=uOPa9XtzoJ4C&lpg=PA181&ots=Ro1P_SOeXB&dq=Risk%20Arbitrage%20in%20the%20Nikkei%20Put%20Warrant%20Market%20of%201989-1990&hl=de&pg=PA181#v=onepage&q=Risk%20Arbitrage%20in%20the%20Nikkei%20Put%20Warrant%20Market%20of%201989-1990&f=true https://books.google.de/books?id=uOPa9XtzoJ4C&lpg=PA181&ots=... should give you a feeling of how his hedge fund was working.

chillydawg 10y ago

Kelly has almost no use for your random small investor. Kelly is all about when you have a very good idea of your edge over the market (first assumption is that you have an edge, very few do). If you're a bedroom trader, it's unlikely you have enough data and a big enough trading history to accurately gauge your market edge. You might well be up 30% in the last year, but that does not mean you've got a 30% edge. When you do have a good estimate of edge, then kelly can come into play. But even then, the theoretical basis of kelly is that money in infinitely indivisible and you can trade at tiny fractions of your total bankroll. So, you normally need to bet less than the kelly stake, but how much less? It's a black art in markets where edges vary with time, along with your skill, perceived skill and measurable skill.

gwern 10y ago

I second the recommendation. I expected it to be mediocre at best, but _Fortune's Formula_ is actually both good on the the technical & philosophical aspects of Kelly, and on the history & people involved.

hudibras 10y ago
William Poundstone is one of my Guilty Pleasures. I've been reading (and re-reading...) his books for years, starting with his Big Secrets series, and he is always interesting and thought-provoking, while not taking himself too seriously. Think a better Malcolm Gladwell, but with a worse PR team.

pvitz 10y ago

Also, if you know a little bit about finance and are interested in Ed Thorp, read his papers. For example, https://books.google.de/books?id=uOPa9XtzoJ4C&lpg=PA181&ots=Ro1P_SOeXB&dq=Risk%20Arbitrage%20in%20the%20Nikkei%20Put%20Warrant%20Market%20of%201989-1990&hl=de&pg=PA181#v=onepage&q=Risk%20Arbitrage%20in%20the%20Nikkei%20Put%20Warrant%20Market%20of%201989-1990&f=true https://books.google.de/books?id=uOPa9XtzoJ4C&lpg=PA181&ots=... should give you a feeling of how his hedge fund was working.