Anybody read this brilliant book by Mandelbrot? What did you think?
I would say that there is a lot of wisdom in it, but I don’t recall agreeing with everything and it has been some time since I read it.
For instance, I agree that volatility isn’t the same as risk. I think that the connection between risk and return is less tenuous than many academics. I agree that focusing on the tails and long-memory is important.
However, I don’t agree with using the stable paretian distribution since it often implies the mean is undefined and variance is infinite thus making optimization impossible. I’ve also never estimated multifractal models, so I really can’t determine whether they would result in improved performance.