James Simons: former mathematician and now billionaire hedge fund manager | Hacker News Reader

James Simons: former mathematician and now billionaire hedge fund manager

seedmagazine.com

26 pointsgetp14 comments

His fund suffered loses recently, making it all suspicious that the math was just good old "sweeping the risk under the rug" kind:

You buy a decent assets that give you 7% return, but have a 2% chance of default. You borrow at 5% and leverage to the max. As long as your assets perform, you have returns that beat the market consistently with high probability. But with a low probability you lose a lot more, and you just hope that this doesn't happen while you rake all the fees and gains.

Medallion was up something like 70% after fees last year. Are you talking about REIF, which is benchmarked to the S&P, and which performed close to the S&P?

Edit: Now is as good a time as any to disclose that I work with Rentec, among other companies. I'm not sure if this makes me informed or just biased, though. Everything I've said about them is publicly available.

I'm talking about Renaissance. Medallion is closed and manages money only for Renaissance employees.

Renaissance is the fund management company. It manages several funds, which have different returns because they use different strategies and have different benchmarks. Which fund do you mean? What strategy are you talking about?

You're right, it was RIEF down for the year, but even Medallion suffered losses in August, here's his letter to investors explaining it: http://www.dealbreaker.com/2007/08/simons_warns_about_august...

"Medallion's 3.9 percent return during August, though that fund too was whipsawed by volatility, bolstered Simons's reputation as the silver-bearded wizard of quantitative investing."

http://www.bloomberg.com/apps/news?pid=20601109&refer=ho...

So you may mean that they suffered from volatility in August. But making 4% per month is not bad, even if there are 10% swings along the way.

Here's another one - from April this year:

http://www.bloomberg.com/apps/news?pid=20601103&sid=a6eS...

Oh no! A fund that correlates with the S&P is down about 12% from peaking in May to early April! Meanwhile, the S&P is down, um, 13.8% from peaking in May to early April! So the fund is slightly outperforming its benchmark, even though it's dropped. I don't think anyone invested in the fund hoping to lose 12%, but this is within the range Simons said to expect. One thing you claimed was that we should expect stable, high returns, and the occasional massive crash. And yet the massive crash you predict has not materialized! Instead, we have a slight decline, mirroring a slightly larger decline in the index one fund follows -- as well as a record-breaking year, in which Medallion appears to have earned over 100% before fees, on over $5 billion in assets.

It's suspicious to me that you point to their low-risk institutional fund's losses as proof that their high-risk strategy is dangerous, even as their high-risk fund is posting record gains.

I'm trying to make a point about their overall strategy:

It is easy to have low returns with low volatility - just buy bonds.

It is easy to have hight returns with high volatility - just buy penny stocks or emergent markets.

Their Medallion fund was doing something that defied that rule for 20 years, it had high returns with low volatility. Up until August last year.

That is incorrect. It is very hard to use simple rules like that to get high returns, because when enough people follow the rules, they become invalid. Renaissance is using a huge number of strategies, and they're constantly testing them. It's likely that any given strategy will be obsolete soon enough, but less likely that they'll all be wrong at once.

It might be useful to think of Renaissance as a market-maker, not a money manager -- given that you described two buy-and-hold strategies, and their average holding period is six minutes, it sounds like they might not be in the business you're talking about.

It's possible to get superior returns over time by having such a low cost or such a strong brand that your product is qualitatively different from someone else's: having a brand like Coca-Cola that is, say, 5% more recognizable than Pepsi can give you something that's worth much much more than Pepsi + 5%; being able to make trades 1% faster and 1% cheaper than anyone else (if Rentec can do that) means having a monopoly on taking advantage of one set of market discrepancies. They obviously have to spend a lot of money to maintain that advantage, but the advantage remains.

Ok, so you're saying that in the business of making money buying and selling stocks (short or long term doesn't matter) they have distinct competitive advantage that is not going away.

So what would be their advantage that holds for so long? Ability to properly calculate short-term risk/reward given all available public information?

Update: after reading your blog I found an answer - if stock market is a casino then quant funds are girls serving drinks to the audience.

Your concern is valid, but a bit misplaced. Rentec is actually one of the (very) few funds that seem not to bullshit their investors and have a fully rational investment strategy.

Majority of 'hedge' funds are simply volatility sellers in one form or another. They run disproportionate and unannounced risks of instant death. Again, it seems like Renaissance is not this type of a firm, even though it can have a few bad months.

Another profile:

http://www.bloomberg.com/apps/news?pid=20601109&refer=ho...

For folks interested in understanding his strategies, here is a thread by some quants:

http://www.nuclearphynance.com/Show%20Post.aspx?PostIDKey=48...

Can quants see into the future or is it just luck or market-custom algorithms?

Out of amusement I love it saying he's no longer a mathematician, because now he's a hedge fund manager! He must have given that up to play the market.

Reply on news.ycombinator.com