"Black Swan"

I picked up a copy of Nassim Taleb's 'Black Swan' at the airport on Friday. Interesting book with some basic and difficult to argue with premises:

• The 'normal' or 'Gaussian' probability distribution curve is only ever, at best, an approximation to reality.
• Although there are some aspects of reality for which it is a reasonably good approximation (human height and weight are both examples mentioned), there are many things out there for which it is a terrible approximation.
• Where it is a terrible approximation, it is normally underestimating the probabilities of "tail" events - unlikely disasters or triumphs.
• Where a terrible probability approximation is used to mechanistically calculate for outcomes with high impacts, 'bad things' ™ may happen.

Although the example he regularly uses for a variable which has a long and significant tail is human earnings or wealth, where he considers that some version of a power law distribution is in force, his money has been made from using this understanding to predict high risk events in the financial markets.

How can you make money from this? By taking it from others through insurance or hedging. If you know (or can place a damn good guess of) the real likelihood and impact of one of his 'Black Swan' events, and everybody else grossly underestimates them, you are not going to need to spend as much as you would if their actuaries were up to speed (this doesn't apply to high-ish probability events where the actuaries just look up the number of events and the assessed costs). You can then take the upside of the event happening or not and transfer the risk of the downside to somebody who understands it less well.

Overall - I would recommend this book for anybody interested in the financial markets and how people manage to get paid quite so much for doing quite so badly. I would note, as a mild warning, that it will be a bit too mathematical for many - although he does try to flag this and allow you to skip those bits. Personally, I will try his "Fooled by Randomness" when I see a copy although I would recommend Tim Harford's "Logic of Life" for an even more sideways look at 'homo economicus'.