reviewed by Rahul Malhotra
posted by Rahul Malhotra
The story of “black box” or “quantitative” trading is perhaps as old as the financial markets. More than two centuries ago, the Japanese rice markets saw Munehisa Homma become the most successful trader in history (in today's dollars) by employing a systematic approach of back‐testing and executing mean‐reversion strategies. While Homma was also recognized with the title of Samurai, today's quantitative fund managers settle for the more prosaic “2 and 20” i.e. 2% of assets managed and 20% of profits. The business remains highly lucrative, with hundreds of billions of dollars traded every day using systematic techniques. Quantitative methods can range from high frequency strategies, with holding periods measured in milliseconds, to long‐term strategies with monthly or yearly horizons. However, all these methods have enough common characteristics that quantitative trading has emerged as a distinct paradigm.
In this book, Rishi Narang does a great job of explaining the structure and thought processes used by quantitative funds. The different “modules” that form a full‐fledged quant system are well‐articulated – namely alpha generation, risk management, transaction cost modeling and finally trade execution. The book is remarkably equation‐free given that this field uses science (and scientists!) so heavily. I would therefore rate very highly its accessibility to the lay reader.
The world of quantitative strategy development is highly secretive, and most funds zealously protect their intellectual property by having their employees sign non‐compete agreements. Therefore, anyone looking for insights into the strategies themselves might feel a bit disappointed, as the author offers nary a glimpse of the crown jewels. That is expected, but for some reason even the chapters on risk management barely scratch the surface. Often, difficult problems (such as unstable correlations or over‐fitting in statistical models) are stated but no approaches are offered. “Machine learning” is mentioned in various parts of the book as a problem‐solving approach, but little explanation is provided as to what it is.
The book really shines when it comes to portfolio optimization, transaction cost modeling and trade execution. These tend to be less glamorous areas of quant trading but are as critical as alpha generation. The author himself estimates that quant managers “lose between 20‐50% of their alpha” to trading costs. The chapters on data: its sources, cleaning and importance are also good reading. The later section on the quant funds crisis of August 2007 touches on an important issue as this crisis did not coincide with a general market drop, and so is relatively unknown among the lay public. However, most analysts agree that it marked the beginning of the credit crunch that nearly devastated the financial markets one year later.
I would strongly recommend this book to students interested in a quant career and fundamental managers or fund‐of‐fund managers who would like a primer on quant methods. Any lay reader interested in the financial markets would also learn a lot. However, there is little in the book that practitioners in the field might find thought provoking. At best, quant readers may admire the author’s writing skill and feel validated in following similar processes.
The Econophysics Forum welcomes your comments