Firstly I have to say that I loved this book. It has its downfalls as a polemic, but it is one of the few books on economics, but I read it with a smile on my face most of the time with occasional gasps of astonishment at the man's maverick genius.

For the uninitiated, Benoit Mandelbrot is essentially a mathematician, but one who has applied his original thinking to a large number of disciplines from economics, geometry, hydrology and wind turbulence. Most people (myself included) would have first encountered him during the early nineties when James Gleick's book on Chaos Theory first popularised his "fractal" view of the world. Fractals (the greatest example of which is the eponymous Mandelbrot Set) are essentially geometrical patterns that are repeated on ever smaller scales to produce irregular shapes and surfaces that cannot be reproduced by classical Euclidian geometry - and Mandelbrot has used these techniques to give us a new way of looking at different subjects that have (to date) been explained by simpler mathematical techniques (in this case the financial markets).

The first part of the book is nothing more than a rant - but an amusing one. He opines on a subject dear to my heart, and one that the industry as a whole should be addressing but are not, that the behaviour of market prices cannot be adequately explained (or measured) using the Gaussian normal distribution. He plots the history of modern portfolio theory, and whilst not belittling the achievements of Markowitz, Sharpe, Samuelson, Black, Scholes and Merton, systematically tears apart the theories of EPM, CAPM, and Black-Scholes - predicated as they all are on the thin-tails of the normal curve.

Nothing new here I guess, but the rest of the book goes on to explain his fractal view of the markets. He proposes power law distributions as the alternative the random walk model and uses the Old Testament examples of Noah ("...the flood came and went - catastrophic but transient...") and Joseph ("...one might call him the first international arbritrageur..." for stockpiling grain) to explain the variability effects in the market. More pertinent and much more scarily, Mandelbrot does show that the exponents needed to model these power law distributions for different markets or instruments are so diverse and intractable as to make general market models meaningless.

As the sleeve cover says "This volume, will not make you richer ... but it may prevent you from getting poorer."