Pairs trading attempts to profit from the relative mispricing of a pair of stocks by buying one stock in the pair and shorting the other, and holding these positions until the prices converge. This book explains two approaches to pairs trading: statistical arbitrage (identify the mispricing by analyzing historical data) and risk arbitrage (when a corporate event such as a merger, is expected). A minimal mathematical background is required; the math is not hard, but it is essential.
The first part of the book is a gentle introduction to timeseries analysis and arbitrage pricing theory – topics of importance in quantitative finance – that many readers will find accessible. In other books, the treatment of these subjects is often academic and dry.
For the arbitrage pairs, the author uses a factor model and cointegration tests to determine the pairs. He proposes a way to estimate the holding time for pairs and details an approach to calculate the trading bands, however don’t expect a ‘ready to trade’ recipe. The method requires tools which are not readily available to individual traders (factor models, cointegration). In addition, choosing crucial model and trading parameters (the time window to identify the pairs for example, or the choice of time-based stops) is left as an exercise for the reader.
The third and last part explores pairs trading in the context of an announced merger, using a Kalman filter to separate in the spread dynamics, the deal risk component from the noise.
Despite the shortcomings, this book is an interesting read. It provides an insight on how the pros approach the topic. Contrast this with the easier but ad-hoc approaches, such as trading correlated pairs.