Statistically Sound Indicators For Financial Market Prediction: Algorithms in C++ Paperback – October 22, 2019

One of those books that pretty much every trader should both have on his/her shelf and should read. Masters addresses techniques for ascertaining whether "here today, gone tomorrow, back the day after" results are an artifact of your system or market doldrums. If it is your system, some of his techniques could, possibly, cure the problem.Masters also uses several well known indicators such as RSI, MACD and On Balance Volume to demonstrate how at least one of the methods he provides does improve performance. This allows the reader to see how to use the math (C++ code provided) so that each of us can use the approach on our own inventions.Dr. Masters also provides sections on amalgamating two or more markets into systems, and the problems one can encounter as a result. I have not gone through these sections in detail in no small part because I am a day-trader in futures and have found no application for an amalgamation, at least not in the approaches I have been using. However, the math seems solid.Finally, the author provides a section on using selected portions of The Janus Theory which he personally uses and has found to be profitable.I have coded in several of his ideas. Some made no difference which was nice in that it implied statistical stability for that bit of code. One idea that I use idiosyncratically actually behave less well after the "standard stabilization fix" and a couple other were improved. I am not finished combining the author's ideas with my own, but, inasmuch as I posted a complaint against Amazon service, which Amazon quickly took down, under the aegis of a very negative score on this book, I thought it best to get my actual opinion of the book up quickly (now that I actually have it, only a little after the quick delivery I pay for -- he said grumpily) lest there be confusion as to my view of the importance (quite) and value (much) of this book.

The book is written in Master's semi-introductory style, which may take a bit of getting used to it. However, he delivers a depth of insight formed from years of testing and modifying stock indicator code. His insights are highly valuable and can rapidly save you years of hit and miss effort on your part.The most valuable part of the book is how to stabilize leading indicators for general statistical reliability. If you use Master's suggestions and apply them well your stock buy and sell positions should be significantly more profitable than those given by standard stock review sites.Be advised, Masters is not giving away his final code. You will have to apply his advice and do some rigorous testing to confirm his advice. This is not a book for amateurs. You will have to be able to program in C++ expertly to take advantage of his extensive experience or know someone who does.

At the outset, I wish to state that I have known the author for 25 years and will receive a very modest remuneration from the sale of the book. However, my share will go to two charities; Rotaplast International (San Francisco, CA) and Faces Foundation (Portland, OR), both of which are dedicated to providing free surgery and medical care to children in poor countries born with facial deformities - cleft lip and cleft palate. With the full disclosure out of the way on to the review.Tim Masters, hold a Phd in Statistics and has published numerous books on subjects including artificial intelligence ( neural networks, deep learning), Statistical Prediction & Classification and trading system development and testing.This book covers a large number of indicators with useful predictive information some of which are either thinly covered or not covered at all in other books on technical indicators. These include indicators such as; Detrended RSI, Deviation from linear, quadratic or cubic polynomial expectation, Reactivity ( a volume based indicator), normalized On Balance Volume, Indicators based on Information Theory, and the almost unknown Khalsa Follow-Through Index. In addition indicators derived from the behavior ofva multitude of markets, including Mahalanobis Distance and the Absorption Ratio, Shift and Coherence. Also included is a discussion of Gary Anderson's innovative approach to relative strength analysis, the Jauus Indicators.Beyond a discussion of the indicators there is an emphasis on improving predictive power by considering Entropy and Stationarity and how to induce stationarity, if it is not present. There is also a discussion of indicator Purification, a concept introduced by Norman Fosback.This is a book for serious students of technical technical indicators and their use trading system development.

The book is not bad but "statistically sound" in the title is misleading. There is no statistics associated to indicators presented. How accurate is a given indicator across markets, is it 52% or 50% (no better than coin flip essentially)? Or it could be 52% but only in certain market conditions. If no such statistics shown then one can come up with millions of indicators, and it's unclear what is the purpose.

Tim Masters's book is well-written and accessible despite the highly technical subject. Statistically Sound Indicators For Financial Market Prediction: Algorithms in C++ covers specific examples and usable code that takes the work of Gary Anderson and others to a point where it can be implemented and tested in practice. It is a terrific book for students of statistics that should open up a whole new perspective on trading systems development. While I can think of several technical trading hedge funds that would directly benefit from this work, I imagine many of them may not be aware of its existence. Hopefully Tim's excellent work becomes more widely known.

The book has some good information on many indicators according to the author. Sadly, I am unable to test any of them. The code that comes with the book is practically useless unless you're an expert in c++.