The nonpartisan Congressional Budget Office regularly revises its forecast of economic growth, the deficit, and other variables it studies. The economists at the International Monetary Fund likewise periodically revise their forecasts, at one point claiming that “downward revisions to growth forecasts . . . highlight continued fragilities”—which, translated, means: “Our forecasts were wrong because we didn’t foresee weaknesses in various economies.”
Recently released minutes of meetings of Federal Reserve Board policymakers show that they were musing about the dangers of inflation just as the failure of Lehman Brothers was about to shake the world financial system to its core and tip the world into a deep recession. And, more recently, Janet Yellen, newly ensconced in the chairman’s seat at the Fed, told the Senate that she can’t yet tell whether some new economic data reflect slowing or snowing, but that she would soon huddle with her staff and policymakers to try to “get a handle” on it.
There’s more, but you get the idea: Economic forecasting involves peering through the windshield in an effort to see where the road is taking us, remaining alert to be able to change course at the emergence of the latest bit of economic data. Don’t blame economists: Those who are in the forecasting business are responding to the market’s demand for some guidance as to whether to buy or sell, build or pile up cash, hire or fire, or, in the case of policymakers, whether to raise or lower interest rates, step up government spending at the expense of higher deficits, or retire government debt. These poor souls must not only peer through heavy fog as they appraise the road ahead, but they must do so even though the rearview mirror is equally fogged, with frequent revisions to data purporting to tell something about past performance.
“Economists,” reports the Economist, “are notoriously bad at predicting sudden turning-points in global growth.” And, the writer might have added, much else. The good news for economists is that their continued failures only whet corporate appetites for more. As the Wall Street Journal reports, “With more data available than ever before and markets increasingly unpredictable, U.S. companies—from manufacturers to banks and pharmaceutical companies—are expanding their corporate economist staffs.” But before concluding that, for economists, nothing succeeds like failure, take a look at two books that, between them, serve as a history of economic forecasting: Walter A. Friedman’s Fortune Tellers is true to its title, while Emanuel Derman’s Models. Behaving. Badly. paints a picture, warts and all, of the current state of play in the forecasting business.
Friedman traces the careers of Roger Babson, Irving Fisher, John Moody, C. J. Bullock, and Warren Persons, men who thrived in the boom years following World War I, thanks to what Friedman calls “the inherent wish of human beings to find certainty in life by knowing the future,” but who failed to predict the Great Depression. (Oddly, or not, Babson, Fisher, and Persons all contracted tuberculosis: In the days preceding the discovery of antibiotics, the responsiveness of the disease to recommended cures such as lots of fresh air was disconcertingly unpredictable.)
Readers interested in the details of these prognosticators’ lives, careers, and methods are in for a treat, as Friedman writes with a combination of accuracy and dramatic flair. Others might benefit from his comparison of those early days of forecasting to the situation today. Friedman concedes that today’s globalized economy is more complex, the amount of data available to today’s economists infinitely greater, and the volume of transactions a multiple of what it was in the days of his early “fortune tellers.” But the desire for certainty remains. And when forecasts go wrong, the cast of characters and the blame game that has gone on since the 2008 financial upheaval seem amusingly similar. Then, it was Yale’s Irving Fisher; now, it is Yale’s Robert Schiller—the difference being that Fisher did not see the Great Depression coming, while Schiller did predict the Great Recession. In 1930, the press “blamed the Harvard Economic Society for promoting false optimism”; in 2008, the Financial Times ran a piece titled “Blame It on Harvard.”
[I]n the recent crisis the pioneer methods of prediction—using historical patterns, mathematical models, expectations, and empirical analogies—each continue to have purveyors and believers. . . . The events leading up to 2008 revealed a similar account of the mistaken hubris of investors in the 1920s. . . . Both episodes, then and now, emphasize the important role that skepticism should play in evaluating rosy economic scenarios and the promises of market gurus.
This caution is repeated by Derman, who comes to his views the hard way, by experience, rather than by studying history, which is Friedman’s vocation at the Harvard Business School. Derman was chief “quant” at Goldman Sachs and remains in the game as a principal in KKR Prisma, which “manages hedge fund portfolios . . . utilizing [a] . . . quantitative approach to risk management,” while directing Columbia University’s program in financial engineering. He argues, “The great financial crisis has been marked by the failure of models both qualitative and quantitative. . . . Models are always inaccurate, and financial models especially so. . . . After more than 20 years of hubris, models collapsed [in 2007].”
By concentrating on these models, Derman brings Friedman’s work to the present day, although he spends less time on the biographies of the bright young things sent by business schools and math departments straight to Wall Street. The current crowd of forecasters, notes Friedman, began with an “optimism that was equal to that of the pioneer forecasters . . . [but] like the pioneering generation, would eventually confront harsh economic realities not foreseen by the models, in this case the stagflation of the 1970s. But by that time they, again like the pioneers, had come to enjoy great influence on economic institutions and ideas.”
Fortunately, Derman does not spend a great deal of time merely trying to debunk the use of models. After all, to use a layman’s description, they do provide the user with a way of putting what-if questions to the modeler. What if interest rates fall rather than rise, as you are predicting? What if the dollar strengthens rather than weakens? That sort of thing. Instead, Derman walks a middle line between those who would do away with models completely and those “naïve idealists [who] pin their faith on the belief that somewhere just offstage there is a model that will capture the nuances of markets, a model that will do away with the need for common sense.” He concludes from his training and experience that models can be useful if the user would only “begin boldly but expect little” and the modelers would “remember as they write their equations . . . the humans behind the equations” and be “humble in applying mathematics to markets.”
In short, Derman writes, “You must start with models but then overlay them with common sense and experience.” Investors have learned this the hard way in recent years, and they seem prepared to admit it. Policymakers not so much. It is comforting that Janet Yellen describes herself as a “sensible central banker,” less comforting that so many policy-making forecasters insist that it is reality that gets it wrong, rather than their forecasts.
Irwin M. Stelzer, a contributing editor to The Weekly Standard, is a columnist for the Sunday Times (London).