Nerds on Wall Street: Math, Machines, and Wired Markets

10.2. Never Met a Data Vendor I Didn't Like

The preceding discussion about collective social investing is an outsider's view. I was never a believer in the "Kumbaya" portfolio. As the manager of $6 billion in quant equities in the 1990s, I did believe in data, and bought a great deal of it, millions of dollars' worth every year. Quality was sketchy, and since computers were ingesting the stuff instead of people, I tried to buy everything at least three times for cross-checking and error correction. Some compact single-value macro and market statistics came in over modems, while larger volumes of data on thousands of individual stocks would often arrive on tape via FedEx packages. When data vendors asked, "What's FTP?"^[] we'd go show them. By the late 1990s, along with other data hog quant managers, I'd begun to notice that the Web was changing the pace of the game (see Bob Butman's before-and-after earnings surprise charts in Chapter 4).

A particularly eye-opening moment about the shifting world of financial information came in November 1998. The Department of Labor (DoL) was scheduled to release its usual monthly job statistics data load on the 9th of the month. Then my three premium-priced macro-economic data vendors would collect it, clean it, type it in, and combine it with the rest of the day's statistics so we could retrieve it by modem overnight. But on November 5 an overeager civil servant accidentally posted the data on the DoL web site. There was a near-immediate market ...