Whenever Perry Kaufman writes a new book, and he’s not exactly prolific, system traders take note. After all, his New Trading Systems and Methods (now in its fourth edition) is considered to be a classic. Alpha Trading: Profitable Strategies That Remove Directional Risk (Wiley, 2011) is his latest effort. In it, Kaufman introduces the reader to stat-arb setups, primarily pairs trading in stocks and futures.

What separates Kaufman’s work from that of many other writers on trading strategies is that he goes into the nuts and bolts of designing and testing system candidates. Let’s say that a trader has an itch to go long Ford when it’s relatively oversold and pair that trade with a short in GM when it’s relatively overbought assuming that the two stocks will come back in line with one another. Does this trade make sense? Does it have a shot at being profitable once slippage and commissions are factored in?

Kaufman would probably send this trader to the back of the class while he worked with those students who were willing to learn the fundamentals of pairs trading. For starters, are the stocks correlated? What is the volatility of each stock? How should the position be sized? (We clearly can’t just buy 100 shares each of Ford and GM.) How can we identify movements away from equilibrium?

Kaufman explains, usually with the appropriate Excel formulas, the process of developing a pairs trading strategy. Volatility is an essential ingredient. First, without sufficient volatility trading profits will be eaten up by the cost of doing business, commissions, and slippage, so a volatility filter might be appropriate. Second, in identifying adaptive buy and sell levels volatility must be normalized. Third, position sizing should be volatility adjusted. Fourth, when combining pairs into a portfolio the trader has to take into account target volatility.

The book begins with examples from the world of stocks. But for the trader who is not part of the high-frequency elite, pairs trading with stocks can be frustrating: “the returns for many stocks are small, and the demands on good execution are high.” (p. 93) So Kaufman moves on to pairs trading using futures. Although the basic concepts remain the same, he provides information specific to futures trading and offers examples of pairs that were relatively consistent in his 2007-2009 test (crude-EURUSD), poor performers (crude-gold), and inconsistent (EURUSD-gold).

Classic pairs trading assumes that prices are mean reverting. It therefore focuses on relatively short time frames where noise dominates price. (By the way, there’s a very good chapter on the importance of price noise with efficiency ratios for 44 markets. Not surprisingly, the equity markets are the noisiest, with the Russell and the S&P leading the pack.) Trends begin to surface over longer periods of time. Therefore, risk-adjusted spreads are an option for trend traders. Kaufman illustrates these spreads using LME data. He also studies cross-market trading, using such pairs as soybean-ADM and crude-XOM.

I should note in closing that the website that accompanies the book has spreadsheets that make the calculations necessary for setting up trades ever so much easier.

Alpha Trading is somewhat repetitive because Kaufman uses only a few ingredients in his strategies. Anyone who is looking for 99 ways to win in the pairs or spread trades arena should turn elsewhere. What Kaufman offers is different—essentially a do-it-yourself manual for the trader who wants to develop strategies that remove directional risk and unearth opportunities for creating alpha.