Recently read: A Non-Random Walk Down Wall Street
A Non-Random Walk Down Wall Street, by Andrew W. Lo & A. Craig MacKinlay
Consider the "last trade" published price for various stocks: Those trades didn't all happen at exactly the same time. This is the sort of book that can take that observation and discuss it for 19 pages: Heavy statistics about what sort of anomalous effect (autocorrelation) you'd see if you ignored the effect, and how this explanation is inadequate to explain the much larger observed anomalous autocorrelation.
In short, a very careful book, full of statistics, concluding that the random walk model should be rejected, without particularly revealing how the market actually works. (There's always room for further investigation.)
The sections I found most interesting were the discussion of futures-index arbitrage, and the last chapter about unusual events observed during the 1987 crash. I'm not sure to what extent the insights are actually useful to me (statistics can be like that), and I have some doubts regarding the contrarian-strategy section, but I'm still glad I read the book. The outlook is useful, and over time I may find more applications.
Conditionally recommended: It does a good job of what it sets out to do, but I expect that most people would lack the necessary math and/or interest in the subject matter.
Consider the "last trade" published price for various stocks: Those trades didn't all happen at exactly the same time. This is the sort of book that can take that observation and discuss it for 19 pages: Heavy statistics about what sort of anomalous effect (autocorrelation) you'd see if you ignored the effect, and how this explanation is inadequate to explain the much larger observed anomalous autocorrelation.
In short, a very careful book, full of statistics, concluding that the random walk model should be rejected, without particularly revealing how the market actually works. (There's always room for further investigation.)
The sections I found most interesting were the discussion of futures-index arbitrage, and the last chapter about unusual events observed during the 1987 crash. I'm not sure to what extent the insights are actually useful to me (statistics can be like that), and I have some doubts regarding the contrarian-strategy section, but I'm still glad I read the book. The outlook is useful, and over time I may find more applications.
Conditionally recommended: It does a good job of what it sets out to do, but I expect that most people would lack the necessary math and/or interest in the subject matter.