Investors and investment analysts start with certain underlying assumptions assumptions that are more often an article of faith than necessarily based on empirical evidence. Very few investors actually ever bother to check whether their underlying assumptions actually hold with any degree of statistical certainty. This book is a ground-breaking work which attempts to test those assumptions about stock market behaviour. The author is chief researcher at the American journal Technical Analysis of Stocks and Commodities and he has also done a lot of work on neural nets.
Broadly, market investment thought can be divided into three schools. There are the academics epitomised by Malkiel who wrote the influential Random Walk Down Wall Street. There are the fundamental investors who idolise Graham, Buffet and Lynch and look to either discover unrealised value, or target unusual growth patterns in stocks by using various earnings/growth-related models. And there are technical analysts who study price-volume movements looking for trends.
The academics believe that the more efficient the market mechanism and the information dissemination system, the more random the price movements and the closer to impossible it is for any investment policy to consistently earn exceptional returns. In other words, a monkey randomly placing buy/sell orders will make about as much in the long run as a Warren Buffet.
The fundamentalists believe prices vary directly with earning expectations, but only in the long run. They ignore timing considerations and the psychological factors that create long periods of over and under estimation which make for long bull and bear markets. Most fundamental analysts also believe prices vary independently and randomly in the short run.
Technical analysts believe that all available information translates into supply/demand trends which can be decoded by studying only price movements. They also argue that mob psychology which drives all markets may be complex but it hasnt really changed over the centuries. In addition, they argue price changes are neither random nor independent of previous prices because investors base their trading decisions on previous prices.
Over a 25-year period, Sherry set up rigorous tests to examine stock movements for evidence about their randomness, stationarity and independence. From a technical analysts angle the answers to certain questions are crucial yet most of them dont bother to even think about them.
Stationarity, independence and randomness are all three well-defined statistical concepts and standard statistical tests can be carried out on time series such as stock price movements, to discover their presence or absence. As Sherry explains, the toss of a fair coin is a simple example of a stationary, random and independent process. The rules governing it do not change, the result is random, and it is independent of any previous results.
But, while stationarity is desirable to make any form of rational analysis worthwhile randomness and independence are undesirable for a stock market investor. If the rules governing changes themselves change rapidly, any pattern is likely to be spurious. But, if movements are random, it doesnt matter whether the rules are stationary or not an investors profits will be governed by chance. If the movement is independent of previous moves, there is no point in studying price history at all.
Assuming that a stock does show signs of stationarity, non-randomness and some form of dependence, it is worthwhile for a technical analyst to then look for trend persistences and search for patterns. Yet, most analysts skip the first three points and start from here. A perusal of Sherry will give you the tools and methodology to start with the basics.
The book is replete with copious worked examples of techniques to test for all the above. Applying the tests to Indian stock market movements is easy and also leads to fascinating conclusions. It is possible to use the pattern tests here to optimise and refine your trading techniques and trading time-frames for either specific stocks or indices representing large chunks of the market.
It is a great book that could provide the intellectual stimulus to turn moderately competent traders into big winners. It is both brilliantly argued and very simply written. Few assumptions are made about the extent of previous knowledge. It is desirable but not essential for the reader to be statistically competent the worked examples are detailed enough to walk a mathematical virgin through the labyrinth of numbers. At the offered price its also excellent value for money.
The Mathematics of Technical Analysis Clifford J Sherry Vision Books Rs 280/328 pages
