Man vs. Machine

Since the "flash crash" in May, there has been lots of talk about how computer trading algorithms  (i.e. algo trading) can crash the market by overwhelming the market with sell orders triggered after some event-based signal takes place.

Using computers to trade in the milisecond frequency makes it almost impossible for human beings to beat machines because there is no way to match their trading speed.

The FT last week had a very interesting story on how two (independent) Norwegian traders exploited a flaw in the algorithm of a U.S. firm to make money out of illiquid Norwegian stocks. So far so good, but the traders have been handed suspended prison sentences for market manipulation and a fine equal to their trading profits.

I find these charges a little odd. How come a computer that has a built in set of rules is not "manipulating" the markets when it places their orders? Two investors, doing a careful research job, find a flaw in the strategy and placed their bets, taking up the risk along with it. What's wrong with that? T

he U.S. firm is to blame for having a poor algorithm in the first place.Unfortunately this reminds me of that story saying that financial firms enjoy the profits but ask someone to bail them out in case of losses.

Behavioral FInance

Wow. Summer is busy even after classes are over! I thought I'd have more time to post but, as usual, we can't get what we want... Anyway, this is an interesting article that appeared in the FT about a hedge fund that uses behavioral finance techniques to invest.

Personally, I believe that behavioral finance brings essential insights on how financial models should strive to incorporate the idiosyncracies of human behavior rather than to take the easy way out and assume that investors are perfectly rational. However, up to now at least, I think that behavioral finance is still at a point in which is more like a collection of results challenging the current paradigm ("mainstream" finance) but still not able to come up with a good alternative theory. I'm not sure it ever will given how "strange" human beings can be whenever money is involved (or generally).

Altogether the article is very interesting and contains a useful introduction to behavioral finance. The trading strategy looks interesting too! Here is the beginning:

Before you even started reading this article, it had already been electronically scanned and its language examined by dozens of computers at hedge funds and investment banks.

At MarketPsy Capital in Santa Monica, California, remote servers will have rated how positive or negative it is on the economy and checked for emotional content on thousands of companies. Finding nothing useful, the computers will then move on.

Stress is Bad for your Trading Account

I read this a couple of weeks ago in the FT and forgot to blog about it. Basically Phillips and ABN Amro are teaming up to develop a bracelet to be used by traders that displays a warning if their stress levels are too high. The idea here is that investors might follow irrational strategies if they cannot think properly and should take a brake if they get to stressed out before taking any decisions.

This reminds me of a paper by Andre Lo in which he measures physiological characteristics of traders (like blood pressure) during live real trading sessions and finds significant correlation between changes in cardiovascular variables and market volatility.

Here is a video of how it should work. It looks like really good sci-fi stuff. I wonder if it might help in other situations, like preparing to an interview, approaching girls, or right before some World Cup penalty shoot-out...

Chance Favors the Prepared Mind

Just read on the FT that "Traders at Goldman made more than $100m in profits on 36 of the 65 days of the third quarter".

I guess that with less competition, those prepared for it and quick to adapt to the new times are making huge tons of money...