I kept hearing recommendations for this book. But I'm not really a finance guy, just a recreational math guy. I'd already heard this book's gist: financiers keep getting into trouble because their formulas assume that market returns follow a normal distribution. (Here, "normal" in the statistical particular-bell-curve-shape sense.) But market returns don't follow that distribution; it's more like a power curve. The book goes into details on this, but for me, not-a-finance-guy, those details pretty much went in one ear and out the other. Oh, there's some formula called CAPM? Maybe I heard of that? Black-Scholes is another formula? OK, I'm sure I heard of that one, probably because it has a more memorable name than CAPM. And I was able to follow a discussion about it… but now it's a few days later and all that has faded. So I guess I recommend this book if you're a finance guy. For me, who already knew the gist—a few days later, I can still say that I know the gist, but I don't think I absorbed anything new from this book.
Oh right, another part of the gist that I already knew: Fractals are neat-o. (This book is by Mandelbrot.) But he shows graphs of simulated investment prices generated based on a normal distribution and based on a fractal and says "Which of these looks more realistic?" As not-a-finance-guy, all I could do was shrug. Maybe if I were a trader I would have got more out of that part, too? Especially if I didn't already know about fractals?