My latest book, "Statistically Sound Indicators for Financial Market Prediction" is now available from Amazon and all major online booksellers.
WHEW!!! I just received notice this morning that this book won the Book of the Year award from The Technical Analyst!
To examine this book on Amazon, click here.
In my decades of professional experience as a statistical consultant in the field of financial market trading, the single most important lesson that I’ve learned about trading is this: the quality of the indicators is vastly more important than the quality of the trading algorithm or predictive model. If you are sloppy about your indicator computation, no high-tech model or algorithm is going to bail you out. Garbage in, garbage out still rules.
This book presents numerous traditional and modern indicators that have been shown to carry significant predictive information. But it will do far more than just that. In addition to a wealth of useful indicators, you will see the following issues discussed:
-->There are simple tests that let you measure the potential information-carrying capacity of an indicator. If your proposed indicator fails this information-capacity test, you should consider revising it. This book describes simple transformations that raise the information-carrying capacity of your indicators and make them more useful for algorithmic trading.
-->You will learn how to locate the regions in your indicator’s domain where maximum predictive power occurs so that you can focus on these important values.
--> You will learn how to compute statistically sound probabilities to help you decide whether the performance of an indicator is legitimate or just the product of random good luck.
--> Most traditional indicators examine one market at a time. But you will learn how examining pairs of markets, or even large collections of markets simultaneously, can provide valuable indicators that quantify complex inter-market relationships.
--> Govinda Khalsa devised a powerful indicator called the Follow-Through Index which reveals how likely it is that an existing trend will continue. This indicator is extremely useful to trend-following traders, but due to its complexity it is not widely employed. This book presents its essential theory and implementation in C++.
--> Gary Anderson developed a detailed and profound theory of market behavior that he calls The JANUS Factor. This theory enables computation of several powerful indicators that tell us, among other things, when trading opportunities are most likely to be profitable and when we should stay out of the market. This book provides the fundamental theory behind The JANUS Factor along with extensive C++ code.
Whether you compute a few indicators and trade by watching their plots on a computer screen, or do simple automated algorithmic trading, or employ sophisticated predictive models, this book provides tools that help you take your trading to a higher, more profitable level.
To download a zip file containing the SINGLE, PAIRED, MULT, and ROC programs, click here.
To download a zip file containing complete source code for these programs, click here. When unzipped, you will get five more zip files, one for each program, and one for miscellaneous things. You MUST unzip these zip files into separate directories, because they contain source files that are different but have the same name.
ERRATA... In late March 2020 I released a Second Edition that contains a few minor corrections. No errors in the code have been reported. With only one exception, these errors are misspellings or slightly unclear wording, so if you have already purchased the First Edition there is no reason to buy the Second. The only notable error is on Page 206, where I speak of total volume when I should have said mean volume. Thus, Equation 7.25 should be modified. Both the numerator and denominator should have the sum divided by the number of terms in the sum, converting the total to the mean. This is done correctly in the code on the next page. Also, on Page 255, CLOSE MINUS MA should have read CLOSE MINUS MOVING AVERAGE, which is made clear in the example.
These programs are intended to run in a Windows console (command prompt). The source is complete, containing all source code modules. No external libraries are required when compiled in a Windows environment, and they compile cleanly under Microsoft Visual C++. If you compile in a different environment, you may need to substitute some basic routines such as for keyboard input and memory allocation.