Hacker News Reader

Comment by vecter | Hacker News Reader

I don't know where you're getting your data from, but I know of at least one high frequency algorithmic trading firm that make ~$1B a year using mathematical models. The models aren't simple, but they're entirely automated and they behave exactly the opposite of how you describe them: you turn them on and they print unbelievable gobs of money.

View on HN

High frequency trading is a different game, though: quoting from an excellent article [1]:

"Most HFTs run a market making strategy. What this means is they play both sides of the table - they take no position on whether a stock will go up or down. Instead, they try to offer securities both to buy and sell. If you want to buy, they will sell to you at $20.10. If you want to sell, they’ll buy from you at $20. As long as their buys and sells match don’t get too out of whack, the HFT will collect $0.10 = $20.10 - 20.00."

[1] http://www.chrisstucchio.com/blog/2012/hft_apology.html

Anybody else think this is like, inherently bad? I mean making money from nothing, producing nothing, doing no service to anybody. The only way you could possible get that billion without doing nothing is to take it from other people, essentially stealing it. Why is this legal?

RenTech doesn't do "nothing". They do a ton of research and write a lot of code. Then, they purchase securities from people who want to sell them, and sell them to people who want to buy them. Why is this more palatable if there is a trader in between?

They create a more efficient market. Ensuring, for example, that the future price of a commodity matches the spot price when the future expires.

They provide an anonymous financial service.

To a financial-impaired mind like mine those looks like great explanations, thanks you all that responded me, now it does seems a little more fair.

Is your insurance company stealing from you when your house doesn't burn down?

When someone buys or sells a security, they take on risk in exchange for money. When someone is on the other end of the transaction, they give up money to reduce their risk. That seems fair to me. Just because it's a zero-sum game with respect to money doesn't mean it's a zero-sum game in totality. It's just that you can buy less iPods with risk than you can buy with cash, so you mentally overvalue cash and undervalue the lack of risk.

They aren't producing anything, but they are providing a service. Algorithmic futures traders are typically uncorrelated with for example the S&P so pensions funds use their services to smooth out volatility and generate better returns.

Not knowing the exact firm I would still debate that its purly mathmatical. _Every_ professional I have talked to has pretty much said that no matter how good your math there is still additional information required other then the raw data about the security.

Also I am talking about trading not market making which can be done automatically, market making only works when you can beat the other guys at making deals and yes you can do that with pure numbers, but again that's not really what were discussing here.

Algorithmic hedge funds trade on statistical models successfully. I work for one at the moment, and over the long term they expect to make 15-20% per year. We have no traders and there's no intervention - it's purely the statistical model that determines how and when to trade.

So it can be done, but you need a brain the size of a planet. (I don't have a brain the size of a planet, so I don't build the models)

I can confidently predict that no, you won't make 15-20 percent as a long term average, though you may well manage a number of good years in a row. Save while the going is good.

May I ask where your confidence in this prediction comes from?

You comment about needing the brain the size of a planet is what leads off my doubts. Its like people who say that credit swaps are too complex to understand so don't even pay attention to them. Your basically admitting you have no idea whats going on and trying to cover it up by saying its just way to complex to understand

That was a somewhat flippant comment... The guys who build the models have phds in physics plus advanced degrees in statistics and stochastic calculus etc. I understand the basics of how the system works, but due to my lack of that advanced maths education and a certain amount of secrecy I don't have every detail.

So in summary: it is too complex to understand, unless you have a very advanced education in the statistical techniques they use to build the models.

Agree, that always bothers me, hand waving under the guise that it's too complicated to be understood. It may take some time but it can always at minimum be on a high level understood. If it can't then someone is lying or hiding something.

(For example I don't have a PHD in physics but a few years of reading and following up on linear algebra, and I can hold my own in a conversation on sting theory with with a PHD Physicist).

One example, there are many:

These guys certainly knew their stuff, and they also had a system: http://en.wikipedia.org/wiki/Long-Term_Capital_Management

Reply on news.ycombinator.com