|
| 
 enlarge | Manufacturer: Random House
Category: EBooks
List Price: $17.95
Buy New: $9.99
You Save: $7.96 (44%)
Avg. Customer Rating:  352 reviews
Sales Rank: 71
Format: Kindle Book
Media: Kindle Edition
Edition: 1
Number Of Items: 1
Pages: 400
Dewey Decimal Number: 003.54
ASIN: B000PDZFCK
Publication Date: April 17, 2007
Availability: Usually ships in 24 hours
|
| Customer Reviews:
 Narcissistic, not worth buying September 22, 2008
3 out of 5 found this review helpful
If you want to read this book, check it out of the library. The basic premise of the book tires out quickly. The writing held a narcissistic flavor in my opinion.
 Brilliant and Refreshing September 9, 2008
1 out of 2 found this review helpful
This book reminds us that skepticism and empiricism are virtues that we ignore at great peril. It paints broad strokes so don't expect excessive details since it's meant as general advice and as a warning. It's rich in sarcasm and humor, so if you like your books dry and mirthless you might not like it's style. It's content however, is concise and effective in making its point that something is wrong (decision making madness on a colossal scale) and tips to avoid making those mistakes in your daily life. I think that people who like to make up their own minds (or are often called upon to make them for others) will find this book invaluable for it's timeliness and clarity. Personally, I also loved the fact he has Eco and Mandelbrot in this book.
Very clever book September 9, 2008
1 out of 2 found this review helpful
The book is full of ideas and observations of reality, very clever ones, I enjoy reading it. More when I finish it.
 good ideas but unfair criticism too September 8, 2008
23 out of 26 found this review helpful
As a mathematical statistician I am a little taken aback by Taleb's lack of scholarship in understanding and appreciating what professional statisticians do. He puts down economists, Nobel laureates, philosopher and statisticians among others. There is a degree of unseemly arrogance on his part and I am sure that some of the other 300 or so reviews on amazon take him to task on that.
It is a shame too because for many of us, it spoils the really good main point of the book which is that Black Swans exist, make predictions difficult if not impossible but can be handle in the stock trading business at least by using his barbell approach.
I found the first 1/3 rd of the book very philosophical extremely redundant yet provocative. The rest of the book was much more interesting to me particular the last few chapter which had the most technical discussion and many points to agree with and also to quibble with.
A Black Swan is an extreme event that is very rare but so significant that it creates instability in averages and can ruin predictions and be either castastrophic (the negative Black Swan) or bring great fortune (the positive Black Swan). These Black Swans are real and Taleb cites many examples. Taleb is also right with his point that some economists are blind to the Black Swan or at least the unpredictability of them. I have often seen major declines in the stock market explained after the fact with seemingly logical but very suspicious and dubious rationalizations. Taleb deserves credit for recognizing this and realizing that in the world he calls extremistan where the Black Swans exist they must be accounted for but no should attempt the futile business of predicting them!
He also recognizes that there is another world where the Gaussian distribution and other light-tailed parametric distributions prevail and he calls this the world of mediocrastan. Here, the usual parametric statistics is useful but in Taleb's view it is not very common in practice to be in a mediocrastan world. This is the world of parametric statistics and is the place where most elementary courses in statistics reside. But here is also where I think Taleb makes a big mistake. He assume that this is the world where all statisticians and econometricians live and play and so these teachings are irrelevant to the practical world. Well, in many of the areas he discusses the parametric statistical models do not work. But probabilist, statisticians and econometricians have realized this for at least the past 60 years. In the 1930s and 1940s the field of nonparametric statistics developed through the work of Pitman, Mann and Whitney and Wilcoxon to name a few. Also the theory of extreme value distributions goes back to Fisher and Tippett in 1929 and was rigorously developed by Gnedenko in the 1940s. Nonparametric statistics deals with general distributions that do not have a simple parametric form and includes the heavy-tailed distributions that Taleb cares about. Also the asymptotic theory of extreme values that Fisher and Tippett, Gumbel and Gnedenko discovered showed that the extreme events had systematic behavior based on the three extreme-value types of distributions. So the extremes can be treated using asymptotic statistical theory just as well as the averages can be characterized asymptotically through the central limit theorem and the stable laws (in the case of a heavy-tailed population distribution). So in some ways Taleb is off and out of gas because he doesn't address or perhaps is even ignorant of this theory.
In the area of finance as well as in other areas, time series models have been useful in developing forecasts. In the world of mediocrastan the Box-Jenkins ARIMA models are very useful for problems in forecast and stochastic control. This was well established with the very popular book by Box and Jenkins that was first published in 1970. However financial data often falls into the world of extremistan and the stationary distributions when they exist are non-Gaussian and heavy-tailed. It is in this context that ARIMA models fail but the statisticians and econometricians have developed other models including the GARCH models which handle this type of data and allow for better predictions. Taleb mentions the GARCH models but only to make fun of them in a very superficial way that does not discuss any of the mathematics associated with these models. Again, I am not sure if Taleb is ignorant about this body of literature or just dismisses it because he see other models that cannot be used to predict as more appropriate.
Taleb is enamored with Mandelbrot and his theory of fractal geometry and the apparent natural properties of fractals. Well at least fractals look like coastlines on the world globe as well as other common items in our natural environment. But is this enough to say that fractals are the only models relevant to extremistan? I am not yet convinced.
This August I went to the Joint Statistical Meetings in Denver. There was a session on the Black Swan and to his credit Taleb was brave enough to accept the invitation of the statistical community to come to discuss the issues in his book. Unfortunately, I was not able to attend that session. But it got mew curious enough to want to read the book and see what Taleb's premise was all about. I do not yet know much about what came out of that session. I hope that at least Mr. Taleb came out of it with a better appreciation of the intelligence of statisticians and the more sophisticated models that he appears to be ignorant of based on the lack of discussion of them in his book.
Another branch of nonparametric statistics developed in the 1970s that is now called resampling methods. One of the more successful of these methods is the bootstrap. I have done some research into bootstrap methods as well as having authored a text on the topic. I believe that the bootstrap approach to time series analysis is another way that these time series with non-Gaussian innovation distributions or the stationary distributions of the time series model can be handled. I am not yet convinced that in the world of extremistan the hope of some form of forecasting must be abandoned as is Taleb's thesis.
Fantastic perspectives that increase your own horizons September 5, 2008
1 out of 3 found this review helpful
Loved it. Great thinking, fantastic perspectives..like this gem referring to Umberto Eco's library "He is the owner of a large personal library (containing thirty thousand books), and separates visitors into two categories: those who react with "Wow! Signore professore dottore Eco, what a library you have! How many of these books have you read?" and the others -- a very small minority -- who get the point that a private library is not an ego-boosting appendage but a research tool. Read books are far less valuable than unread ones. The library should contain as much of what you do not know as your financial means, mortgage rates, and the currently tight real-estate market will allow you to put there. You will accumulate more knowledge and more books as you grow older, and the growing number of unread books on the shelves will look at you menacingly. Indeed, the more you know, the larger the rows of unread books."..
And that is just one example. See the wealth of other reviews here for more. If you even think you might be interested, don't hesitate!
|
|
|
Powered by Associate-O-Matic
| |
