Market Anomalies, Machine Learning, and the Limits of Quantitative Trading

There's something called the efficient market theory, which says that there's nothing in

the data, let's say price data, which will indicate anything about the future, because

the price is sort of always right.

The price is always right in some sense.

But that's just not true.

So there are anomalies in the data, even in the price history data.

For one thing, commodities especially used to trend, not dramatically trend, but trend.

So if you could get the trend right, you'd bet on the trend, and you'd make money more

often than you wouldn't, whether it was going down or going up.

That was an anomaly in the data.

But gradually we found more and more and more and more anomalies.

None of them is so overwhelming that you're going to clean up on a particular anomaly.

Because if they were, other people would have seen them.

So they have to be subtle things.

And you put together a collection of these subtle anomalies, and you begin to get something

that will predict pretty well.

How elaborate are these things?

Because in my head, I'm imagining some equation like Pythagoras equation, and you put a few

numbers in and something spits out.

But are these giant beasts of equations and algorithms, or are they simple things?

Well, the system as it is today is extraordinarily elaborate.

But it's not a whole lot of, you know, it's what's called machine learning.

So you find things that are predictive.

You might guess, oh, such and such should be predictive, might be predictive, and you test

it out in the computer, and maybe it is, and maybe it isn't.

You test it out on long-term historical data and price data and other things.

And then you add to the system this if it works, and if it doesn't, you throw it out.

So there aren't elaborate equations, at least not for the prediction part.

But the prediction part is not the only part.

You have to know what your costs are when you trade.

You're going to move the market when you trade.

Now, the average person will buy 200 shares of something, and he's not going to move the

market at all because it's too small.

But if you want to buy 200,000 shares, you're going to push the price.

How much are you going to push the price?

How are you going to, you know, are you going to push it so far that you can't make any

money because you've distorted things so much?

So you have to understand costs, and that's something that's important.

And then you have to understand how to minimize the volatility of the whole assembly of positions

that you have.

And so you have to do that.

That last part takes some fairly sophisticated applied mathematics, not earth-shattering,

but fairly sophisticated.

What discipline of mathematics or disciplines?

Is it multidisciplinary?

It's mostly statistics.

It's mostly statistics and some probability theory.

But I can't get into, you know, what things we do use and what things we don't use.

We reach for different things that might come, that might be effective.

So we're very universal.

We don't have any, but it's a big computer model.

For one thing, there is a capacity to the major model.

It can manage a certain amount of money, which is rather large, but it can't manage an enormous

amount of money because you're pushing, you're going to end up pushing the market around

too much.

So it was kind of a sweet spot as to how much it's reasonable to manage.

Therefore, it would never grow into some behemoth, which would, you know, take everybody out

and you'd be the only player.

I mean, well, of course, if you were the only player, there'd be no one to play against.

There are limitations, at least the way we see it.

But we keep improving it.

We have about 100 PhDs working for the firm.

That's what I mean.

I mean, how did you get to that point?

Did you start to think we need this, we need that?

We just hired smart people.

My algorithm has always been you get smart people together.