The book has numerous, unforgivable typos and mistakes in equations. It appears that no one other than the author has ever reviewed the materials before it was published. It is truly disappointing to see that, as both the publisher and the author are highly regarded. This can be an extremely dangerous book if an average reader just takes the results and equations at face value and apply them. Unless you work through all the derivations yourself, you never know which equation has mistake, and which does not, except for those mistakes that are too obvious for anyone to miss. As a result, this book is most useful for one type of readers: those with math skill nearly as good as the author's but don't have the time to solve these problems themselves. Only these readers can apply the results with peace in mind, because if they are willing to spend the time to work through the math, most of them should be able to spot those many mistakes. But how many readers have this level of skill?
Suggestion to the author: IMMEDIATELY, do the readers and the profession a favor, go publish a second edition with correction; better to use a new publisher which is willing to hire a reviewer, if it doesn't violate your contract. You have done a very good job indeed. Too bad you ruined it with all these careless mistakes and typos. Two stars were taken.
The pro is thus that there are lots of information, mostly of practical relevance, mostly well explained, much of it previously unknown even as regards its existence.
The fifth star is taken away by the cons, of which in my judgment there are three: (1) a LARGE number of typos, mostly innocuous, some not, always annoying. For a $90 book this is bad slippage from the publisher and should be remedied in a second edition. (2) The material is somewhat disconnected; especially in large topics we have more a smorgasbord of things than a smooth discourse; (3) some details are not worked out in a totally kosher fashion. Two examples: (i) on risk budgeting, attention should be paid to singular covariance matrices (since passive managers will have near-zero tracking error volatility) which means violating the equality of marginal return/risk ratios. This crops up in the examples but is not explained in the text. (ii) Black-Litterman is not explained in full generality. See Litterman's new book or Lee for that.
Overall, however, you should buy and read this book and end up the wiser.