Tuesday, 30 September 2008

Fooled by Randomness

Just finished book by Nassim Taleb - "Fooled by Randomness" - despite the intermittent intellectual snobbery (which leads to a little verbosity) I think this is an absolute must read for any person with an interest in high finance, economics and math.

Taleb's site: http://www.fooledbyrandomness.com

This book is about the influence of luck on life, and how as humans we often over ascribe luck to our skill and under ascribe bad luck to our poor performance.

A quote in the book, "Machiavelli ascribed 50% of the success in his life as luck" describes Taleb's precept aptly. His view that most people encounter this statistic, but don't ascribe that higher degree of luck to their successes. I know I am guilty of this, albeit unintentionally.

Another quote regarding "legendary traders": "It is very likely that a large enough sample of traders that one of them will have an inordinately lengthy lucky streak".

In his vernacular this luck is known as "randomness" and he suggests we need to be much more cognisant as it could equally apply to bad luck, and therefore the downfall of people's lives and well being. None more so than in the trading world of high finance. He defines this "outlying" events as "Blackswans" (his subsequent book is titled as such) - very rare events that are game changing and almost always not accounted for.

Taleb's thesis (for me at least) is that despite the belief that we can be rational 100% of the time (he has a considerable go at the neo-classical economists - which I agree with), we are not genetically designed to do so, hence we cant. We should therefore we should plan accordingly. Our emotions will always try and play a part in our decisions, is his assertion. His solution is to develop tricks to bypass the emotional component as much as possible. He lays these tricks out in the book.

There is good reason for us being innately programmed to use emotion because it is what "energises us to act" he says. This makes a lot of sense if you think about it too...

He details how human's suffer from biases or heuristics that are based on limited proof. One example is "survivorship bias" - exaggeration of performance. It makes a lot of sense for us to have this ability because it quickly allows us to interoperate with our environments, e.g., "tiger looks dangerous, lets run" - whereas to validate that rule one would probably die in the process.

The last interesting thing he puts forward (for me) is that human brains are not designed for non-linearity, ie, more than two variables impacting an outcome (aka multivariate analysis). Yet in reality, this is exactly how stuff works. As such, our minds tend to "smooth" out these mutli variability, framing it in x and y axis model.

A nice way of explaining this is the learning of a sport. One doesnt usually become constant incrementally good at something. It is more staccato - you will be bad for a while, then all of a sudden you "click" and can start doing "that" move or swing or play. So improvement is punctuated by short sharp upward successes rather than a constant linear progression... yet our minds perceive it is a constant trend.

I wonder how Taleb would view the latest "blackswan" (financial meltdown)?

No comments:

Real Time Web Analytics