Monte Carlo Methods in Finance - Peter Jaeckel

I got this book on loan, and to cut things short, I think that's the way I'd like to read this - it's worth the read, but I'd feel ripped of paying 40 quid for it (let alone the 60 pounds "list price"). It's not terribly coherent, but rather the 14 chapters (in 200 scant pages) feel like papers on "what I learnt about X" for various aspects of Monte Carlo simulation (both more and less directly relevant). It's all about the author's personal experience and feels rather biased because of that in places. Having said that, it did expose me to various aspects I'd not really come across before, and provided a good companion to the training I've received at work.

As for the details, I must admit I did get lost in a few bits - the excuse being I was rather skimming the book for concepts rather than trying to learn everything inside out. I suspect that this book works better providing ideas to a seasoned Monte Carlo user who can take them with a pinch of salt and compare them against recently published papers (and their own experiments), rather than as an introduction. Having said that, the book's quite good at providing references to papers for areas it doesn't go into in so much depth.

I think the thing that let the book down the most for me was the end of the final chapter, where he gives a few words on multi-processor implementations of Monte Carlo algorithms, where, quite frankly, it gets a bit waffly and vague with technical terms, trying to explain findings he apparently didn't fully understand. Given this is the area of the book I am strongest at, I found this rather disconcerting, and it rather shook my faith in the author on the areas I felt I understood rather less well.

Perhaps I'll buy the book one day, if I find a cheap copy.

Posted 2005-11-16.