Is Long-Term Forecasting Valuable?

A Financial Times article highlights and challenges the common practice of long-term forecasting. Opening with a George Eliot quote – “Among all forms of mistake, prophesy is the most gratuitous” – the piece makes no bones about challenging the value of long-term forecasts. It argues: “Either they are so short term they are often liable to be as wrong as any other prediction; or they are so long-run that they are practically useless and only show the long-term historical returns.” Nonetheless, recent forecasts of GMO and JPMorgan Asset Management’s “long-term capital markets assumptions” (LTCMA). As depicted in the images below, GMO’s 7-year predictions show significant real returns in emerging markets and modest losses in the U.S. The LTCMA, which predicts returns over the next 10 to 15 years, suggests somewhat higher returns in most asset classes, albeit in “an environment of steady inflation and subdued long-term growth.” The LTCMA notes that the firm’s 2004 predictions were only off by 0.2% (7.1 predicted versus 6.9 actual) despite the financial crisis. This leads to the question: “If forecasts are not even budged by the greatest global financial crisis in generations, are they useful as anything other than a general yardstick to remind investors that markets tend to rise over time?”

gmo forecast.PNG

jpm forecast