Wednesday, June 12, 2013

Davenport & Kim, Keeping Up with the Quants

It may be an overstatement to describe data analysis as “the sexiest job of the 21st century” (CNBC headline, 6/5/13), but data analysts are definitely in demand. An overwhelming amount of data has been and continues to be collected. These mountains of raw data often seem akin to all that stuff hoarders can’t bear to part with, taking up space (though of course not as much as physical junk does) and having no apparent function. Amazon is the ultimate data hoarder, claiming never to throw anything away. After all, you never know when with a new kind of pick and shovel or a better map you just might find gold in them thar hills.

Keeping Up with the Quants: Your Guide to Understanding and Using Analytics by Thomas H. Davenport and Jinho Kim (Harvard Business Review Press, 2013) is a terrific book for those who aspire to be data-savvy consumers or managers, even for those who might become quants one day themselves. In clear, non-technical prose the authors explain the ingredients common to all types of quantitative analysis: framing the problem (problem recognition, review of previous findings), solving the problem (modeling and variable selection, data collection, and data analysis), and communicating and acting on results. They draw examples from a range of fields, including finance, and in the process expand the reader’s horizons.

Take Florence Nightingale, for instance. Did you know that she was an early adopter of quantitative methods? She was appalled at the conditions and high mortality rates in a makeshift British military hospital in Turkey during the Crimean War. She began collecting data. “Nightingale’s greatest innovation, however, was in presentation of the results.” She developed innovative diagrams, “a kind of pie chart with displays in the shape of wedge cuts. Nightingale printed them in several colors to clearly show how the mortality from each cause changed from month to month.” Eventually, death rates in the hospital fell dramatically and, when she returned to England, Nightingale “found herself a celebrity and praised as a heroine. Nightingale became a Fellow of the Royal Statistical Society in 1859—the first woman to become a member—and an honorary member of the American Statistical Association in 1874.” (pp. 103-104)

Keeping Up with the Quants is at its core a how-to book in practical problem solving using scientific (analytic) methods coupled with sound business practice. But that description makes it seem boring. And the book is anything but boring. It takes a page from Xiao-Li Meng’s Harvard undergraduate course called Real-Life Statistics: Your Chance for Happiness (or Misery) which includes modules on such topics as romance, wine and chocolate, finance, medicine, and the stock market. The course aims, in Meng’s words, to make statistics “not just palatable, but delicious.” (p. 96) Well, I wouldn’t call Keeping Up with the Quants a delicious book, mainly because I have problems analogizing from food to the written text, but it’s a book I plan to read a second time. (And, by contrast, my delicious dinner from last night is long gone.)

Those readers with a narrow financial focus will learn how the Australian authorities solved the Simon Hannes insider trading case and why being able to adjust models frequently has helped make James Simons’s flagship Medallion fund so phenomenally successful. Anyone contemplating cooking the books should heed the first-digit (or Benford’s) law. “By just looking at the first digit of each data entry and comparing the actual frequency of occurrence with the predicted frequency, one can easily finger concocted data. In general, faked or fraudulent data appear to have far fewer numbers starting with 1, and many more starting with 6, than do true data.” (p. 146)

I would recommend this book to anyone who is using data, big or small, to make decisions—actually, that pretty well encompasses everyone. Whether you’re picking stocks or deciding if you should get a pet, you will learn something from Davenport and Kim.

No comments:

Post a Comment