on June 29, 2007
1) Highly recommand this book to serious Quants.
2) Graduate lever math is required for serious reader.
3) Good reference book and good for self-study
4) Well written, easy read.
5) worth the money.
The field of quantitative techniques have developed so much in the last 10 years, but almost no book cover enough serious topics about these new directions. I had already learn a bit of the robust techniques while working, including robust estimates, robust portfolio construction, error control, bayesian estimates and others. But those are all picked up in pieces, at different times, and with much research efforts. So you can imagine my delight to see a book that covers a lot of the pieces concisely.
This book itself is very well written, occasionally misspelled math labels are easily corrected by more math inclined reader, and will not interfare with casual reading. Like many of Fabozzi books, overall organization is slightly loose, so that you can start any chapter in the book and still get pretty much decent view about that subject. But better written for quants than some of Fabozzi's early books (which are mainly used as reference books)
what is missing in this book?
just one: sometimes, reference papers or books are given even though a little more details would save serious reader a lot more time. Yes, I know, those are advance topics, still would like to see them as a serious reader. Maybe as appendix for relevent chapers.
Over all, worth every penny of it.
on October 1, 2007
This book is fabulous. It covers the latest optimization and statistical methods in Finance as well as the modern portfolio theory and some risk management and can be used for audience with Finance background, or optimization or statistics background. It is definitely one of the best books serving as an interface between Finance and Operations Research (O.R.). The other excellent book is "Optimization Methods in Finance," by Cornuejols and Tutuncu (2007), which discusses O.R. techniques with financial applications. This one is almost the opposite - it focuses on modern portfolio theory and discusses how O.R. and statistical methods can be applied. Both books discuss some latest optimization techniques - however, this one has much more details on a modern method for handling uncertainty called Robust Optimization, in which Dr. Pachamanova (third author) is an expert, as well as some relatively advanced stat methods such as robust statistics and Bayesian estimation. Despite the advanced methods, I found this book Fabozzi (2007) clearly written and quite easy to follow, just like Cornuejols and Tutuncu. The only comment I have is that I wish there was a list of references at the back, just like most other books.
Another text I would recommend is Ruppert's "Statistics and Finance: An Introduction" which serves as an interface between statistics and finance, as the title indicated.
on October 1, 2007
This book is similar to most other Fabozzi books, sharing similar strengths and weaknesses. The good part is that it provides a relatively complete and very up-to-date reference on current research in portfolio optimization. It will save you a lot of time in doing literature review.
I do not really like two features about the book. First, there are gaps and repetitions in the book (I guess this is not surprising, given it's written by many authors, but someone should try to put everything together in a coherent fashion). Second, the examples in the book tend to be overly simplistic and the authors did not include enough details for readers to replicate them (I can not really blame the authors, because most books are just like that).
on August 18, 2010
This is the best quant finance book I've yet read. The symbols on the cover may look daunting, but the text actually keeps notation simple. Many topics are covered quickly and accessibly; this is a math book you can actually skim, or skip around in. I think that's due to good writing.
Also: I stand firmly in the Robust camp. After my class with Karen Kafadar, I'm confident that Robust models are easier to explain and more reliable. Her typical example was to mis-type just one of the data by repeating a digit or moving the decimal place -- and how likely is that! -- and see how much the output changed. Ideally your real-world recommendation shouldn't change too much based on just one data point. (If that's unavoidable, you should withdraw any recommendation.)
So many mathematical questions or ideas yield up a flowering of possible tweaks and adjustments that can be made to a model, with no recommendation of which parameter value to use. A good answer is: whatever is most stable across different potential scenarios.
There is a wide variance among the Frank J. Fabozzi series (Advanced Stochastic Optimization, for example, is way worse than this). If you only have time to read one, read this one.