Timberpost: Farecast for Stocks

Farecast (recently acquired by MIcrosoft) does just what you want when making an airplane ticket purchase: it predicts if the price is going to go up, down or stay level and advises when you should buy (now, wait). Timberpost is a small company founded by Peter Ross and Tim Taylor (Peter was a professor at Edinburgh University when I was there studying AI and Tim completed his PhD in the same department). Timberpost’s product – TRAITS – takes a crack at a real chestnut of an AI problem: predicting the stock market. The difference with this solution appears to be that it actually does a good job of it.

The graph below shows the performance of a hedge fund run (in simulation) by TRAITS compared against the FTSE EuroFund 300 Index.

A recent overview published by Timberpost says:

This portfolio is currently showing an annualised return of +23%, which would rank it 6th out of 200 peer funds according to the latest performance data on real European Long/Short Equity hedge funds published by EuroHedge magazine.

Timberpost describes their technology as follows:

Many machine learning techniques have been applied in finance, including neural nets, genetic algorithms, reinforcement methods and rule induction. We are developing a new approach that is inspired by ideas about how the human immune system functions. Like the immune system, our software can not only discover effective responses to new conditions (in our case, potential trading opportunities), it also adapts to remember past successes in order to be able to re-activate them quickly when conditions change.

In biological systems, recognition happens by molecular binding. In our software, recognition is based on elaborate mathematical expressions that describe features of the behaviour of stocks. The system is designed to be efficient; it can look at many thousands of elaborate expressions per second.

Comments

[Daniel Tunkelang]

I'd be less skeptical if you told us you were risking your life savings on its predictions. :)

[Prolific Programmer]

The large market makers use algorithms based on mathematics. If these algorithms can be reverse-engineered, predicting an index (and thereby outperforming it) is a simple task.

[Daniel Tunkelang]

Not sure whether you're being serious or sarcastic. But, as Niels Bohr said, prediction is hard--especially the future. But if you believe that a widely available strategy outperforms the market, then I've got a bridge I can sell you for a good price. :)

[Sandro Saitta]

Very interesting post. Botraiders is a french company doing roughly the same thing (predicting individual stocks for the CAC40). Here is their web address: http://www.botraiders.com/

[Geoff]

The graph of TRAITS shooting into the stratosphere certainly looks impressive. What is unclear is a breakdown of the total costs that would be incurred from the constant buying and selling. The costs incurred through pursuing such a strategy would present an insurmountable hurdle to beating the market average would it not?

These people are mistaken if they think they or their software can predict the future direction of the markets. This is not cynicism but a refusal to believe in magic. Any investment professional who speculates on the market’s future should be relegated to a fortune telling parlor.

[Geoff]

What is equally puzzling is that since the beginning of 2008, the world markets in general all took a dive: the FTSE 100, S&P 500, CAC-30 and so on.

My question is how could the TRAITS portfolio give such returns when all the markets in all sectors were collapsing, and that the TRAITS portfolio selection IS the market, albeit selected portions of it.

How can the market beat itself by such huge amounts especially in these inflationary times?

I find this VERY fishy.

[Timtaylor]

I just came across this old post while doing a Google search, and felt I had to respond to the last two comments by Geoff. I am one of the authors of this work. The first point to make, is that the performance included realistic trading costs, price slippage, volume constraints and other real-world factors. We were working closely with some real hedge funds, and using high quality data from Reuters. The second point, about how TRAITS did so well when the markets were collapsing: this was a long/short equity fund (as mentioned in the original post). I suggest Geoff does some research on short trading to understand how this works.

The commercial interest in this project took a knock during the financial crisis (to say the least). However, we have more recently applied this to currency trading, with some even more impressive results. A phoenix may yet rise from the ashes...