on 5 April 2007
Jaeckel's book is an outstanding one and I will be using it for computational finance teaching at King's College London. Its many strengths include a proper appreciation of the importance of quantile functions (inverse CDFs) and discussions of different approaches to multivariate simulation including copula techniques (introductory) and low-discrepancy sequence methods (more detailed). Jaeckel carefully explains why a long-standing standard approaches to Normal simulation (Box-Muller) is problematic, and includes a helpful chapter on managing "bad" correlation matrices, which is vital for multivariate simulation. Variance reduction is discussed in detail. Applications focus more on the interest rate side and BGM models, and there is little on modern Monte Carlo methods for American options. So while this book does not cover everything, what it does do is done brilliantly.