Predicting where the stock market is heading has been the holy grail of financial professionals for over a century. Every year investment companies and forecasting firms dedicate enormous sums pitching their prognostications to investors, and it’s easy to understand why: the payoff can be huge. For brokerage houses and their sales representatives, it is a proven strategy for attracting uninformed investors looking to earn superior returns with minimal risk.
Unsurprisingly, the most ardent defenders of forecasting stock prices are those who make their living from it, including investment firms and financial news outlets. Most academic economists, on the other hand, are skeptical that any forecasting model—even the most sophisticated—can predict turning points in the markets. An overwhelming amount of evidence gathered over the last 80 years supports their position. A groundbreaking report titled Can Stock Market Forecasters Forecast? published in 1933 by Alfred Cowles III, an economist and founder of the Cowles Commission at Yale University, concluded, after thorough analysis, that it was “doubtful.” Since its publication, I am not aware of one credible report contradicting Cowles’ findings. Rather, several provide corroboration.
A 2012 study conducted by the Vanguard Group applied over a dozen different metrics (e.g. price to earnings ratios, dividend yield, etc.) to past stock returns for the period between 1935 and 2012. The goal was to identify reliable signals that might provide guidance for future performance. The researchers concluded that forecasting stock returns is essentially impossible in the short term. In other words, we simply have no way of predicting which way or by how much stocks will move over the next few days, weeks or months. Furthermore, the researchers determined that even over longer time horizons of 10 years or more, many metrics assumed to have predictive value, in fact, had little to none.
In a 2015 article, Robert Shiller, Yale professor, and 2013 Nobel Laureate, wrote that after searching the news archives about the country’s major recessions beginning in 1920, he “found virtually no warning from economists of a severe crisis. Instead, the newspapers emphasized the views of business executives or politicians, who tended to be optimistic.” The almost-universally-unanticipated 2008 collapse of our country’s financial system offers a prime example. No model, including those developed at the world’s most elite institutions including the United States Federal Reserve Bank, foresaw a collapse until it occurred. Most were projecting continued GDP growth and rising asset values into 2009.
If we can guide rockets into outer space and back to earth with precision, why can’t some of the world’s smartest people using the most sophisticated technology develop reliable market forecasts? An explanation that has gained widespread acceptance is that current models are doomed to fail because they are vast simplifications of reality and cannot capture the interaction of forces that drive the markets. Instead, they rely on the conventional economic theory that assumes that the financial markets are made up of players who make decisions to buy and sell in a logical and predictable fashion. In 2000 and then again in 2008, investors worldwide witnessed firsthand the gap between theory and reality: in the real world, things get chaotic. When this happens, investors—swept up in the euphoria or panic of the moment—abandon logic and behave unpredictably.
In my next post, we’ll talk about a fundamentally new way of thinking about the financial markets.