on 16 June 2009
QRM is a technical book on risk management from a statistical point of view. It is definitely not a manual for practical implementation of QRM tools, so do not expect any how-tos. Rather, it is an excellent starting point for the risk manager who is keen on the technical aspects of risk measurement. Every chapter contains many references which point the reader to sources containing more detailed explanations. This book assumes a decent knowledge of statistics, particularly time series analysis. Also, the reader must be familiar with matrix algebra.
Good points: puts risk measurement into a formal, rigorous statistical framework; good overview of risk measurement topics; implementation in R of some tools is available for free as an R package.
Bad points: not very detailed in terms of how to implement many of the models; some chapters seem to be there more for completeness than for their practical value (I didn't find the chapter about copulas in particularly useful); too theoretical and very little emphasis on the practical side.
In summary, this book is for a risk manager who is very well trained in statistics and will be able (and willing) to implement the tools starting from the statistical concepts.