Mean Reversion Trading Systems
This book is an extension of Quantitative Trading Systems
Subtitles might be:
- Practical Methods for Swing Trading
- Applying Statistics to Trading
- Transformations and Indicators
- Looking for the Signal in the Noise
Based on the method used to identify the entry, a system can be placed in one of the categories:
- Trend following
- Mean reverting
To state the obvious, every profitable trade is a trend following trade for the period it is held. The sell price must be higher than the buy price, no matter which order these occur.
Based on holding period, a system can be placed in one of a different set of categories (although your definitions may vary):
- Buy and hold — forever (investing)
- Long term — months to years (trading)
- Short term — weeks to months
- Swing — hours to days
- Scalping — minutes to hours
The focus of this book is systems that enter when price is overextended — far from the recent mean — anticipating that price will revert to the mean, with holding periods of a few days.
By most definitions, systems such as these are defined to be mean reversion and swing trading.
Readers of my companion book, Modeling Trading System Performance, will recall my analysis of trading frequency and holding period, with the conclusion that the sweet spot for holding period is one to two days, and it is desirable to trade frequently. This book is discussing systems that closely fit those sweet spots.
Equity prices typically have two cycles per month. Trading each leg of each cycle leads to about 50 trades per year. The analysis described in this book centers around systems that identify those cycles.
Entries are based primarily on indicators and patterns. Most of the systems use only the daily Open, High, Low, and / or Close, (and rarely Volume) for the issue being traded. Some use auxiliary series and some use multiple time frames. Both entries and exits are typically Market on Close or at a limit price. Often, the price at which an entry or exit signal will be generated can be computed in advance, so limit or market orders can be placed as day orders, removing the need to watch the market throughout the trading day.
All systems are fully disclosed, enabling readers to replicate the results described and to use the methods as bases for further development.
Mean reversion systems are often criticized as being excessively risky. That topic is addressed thoroughly, including analysis of risk and practical techniques for dealing with it.
The book includes over 80 program listings, downloadable and ready to run using AmiBroker, that illustrate the topics being discussed.
These links each open a pdf file related to the book.
- Table of Contents
- Introduction (Chapter 1)
- Development and Analysis (Chapter 2)
- Naive system (Chapter 3)
The code shown in this book can be downloaded. The link to the code is a separate page (to avoid complications with historical and legacy documents)
You can learn more about the book, read additional pages, read reviews, and purchase a copy — all at Amazon.