home > archive > 2018 > this article


Is overconfidence to blame for the failure of financial forecasting?

By Jason Yu
web posted October 22, 2018

Economic forecasts are crucial for all economic activity because they influence business decisions and government policies. Inaccurate forecasts, whether they underestimate or overestimate an economy’s future success, incur great costs. Unfortunately, an economic recession, a decrease in a nation’s GDP for two consecutive quarters, can be difficult to forecast. This was very much the case leading up to the 2008 recession, known as the “Great Recession.? Many businesses were caught off guard by the recession; few even regarded the possibility of such a strong economic downturn. After collecting and analyzing 28000 predictions from 284 experts, a study by Philip Tetlock concluded that the average expert’s forecasts were only slightly more accurate than random guessing.

Richard Thaler, award-winning economist, shows that many predictions made by chief financial officers (CFO) are often very inaccurate. For example, in a survey conducted on Feb. 26, 2009, the CFO’s predicted a market return of just 2 percent. In fact, the market soared to 42 percent over the next year. What is causing these poor predictions? Thaler blames it on overconfidence, stating that “many of these executives apparently don’t realize that they lack forecasting ability.? Essentially, Thaler argues that the main problem facing economic forecasts is that these CFO’s are too confident in their abilities. Because these forecasters believe that they have the ability to accurately forecast the market, there is little improvement in economic predictions. This claim, however, falls flat because it is unrealistic to assume that forecasters are unaware of their shortcomings. Instead, the real reason why forecasts are often inaccurate is because forecasters base their predictions on mere intuition as opposed to objective data and models.

Overconfidence permeates our everyday lives. When I was younger, I remember watching “American Idol? with my mother. During the auditions, many contestants claimed that they were talented enough to win the show. My mother, who has perfect pitch, thought otherwise. She made good use of the fast-forward button on the remote control. In the same way, Thaler claims, many CFO’s do not recognize their apparent lack of foresight, leading to a continuous string of poor predictions. However, this is not a fair comparison. When denied from American Idol, many contestants continue to assert that they are still excellent singers and that “the judges just didn’t like my style.? Due to the subjective nature of musical preferences, the contestants remain overconfident. However, in market forecasts, the predictions are completely quantitative. CFO’s eventually see that their forecasts were off, and it is impossible to argue that the market somehow did not like their predictions. As such, any logical market forecaster will realize their mistakes and make an attempt to understand where he or she went wrong. It is unrealistic to claim, therefore, that these CFO’s are blind to their shortcomings.

Overconfidence is not to blame for inaccurate forecasts. Rather, the reason why forecasts are often so inaccurate is because many CFO’s apparently base their predictions on feelings. For example, Thaler mentions that predictions are more pessimistic during economic recessions and more optimistic during economic expansions. These predictions are not necessarily based on empirical data, but instead reflect the current feel of the times. On September 10, 2001, only 13 out of the 100 Blue Chip forecasters had answered in the affirmative to the question, “Has the United States slipped into a recession?? When asked again on September 19, 82 out of the 100 forecasters answered affirmatively. Revisions indicate that the economy was already in a recession by the second quarter, but it was only until after the September 11 attack that CFO’s acknowledged the recession. These findings are consistent with John Maynard Keynes, who famously suggested that business are prone to bouts of extreme overconfidence and underconfidence (overconfidence not in their own abilities, but in the economy). Clearly, these predictions are often poor because they are driven by human perceptions that are prone to bias and oversight.

Instead of relying on intuition and judgments based on the perturbations of recent events, it is more reasonable to make economic forecasts based on empirical data and quantitative forecasting models. Of course, economic systems are extremely hard to predict, even with highly complex models, but they are more likely to work than human feelings or intuition which are prone to blatant inaccuracies and oversights. Inaccurate predictions are to be expected, but the increased use of computer models and machine-learning algorithms over faulty human judgment is a step in the right direction. ESR

Jason Yu is a high school student currently studying AP Macroeconomics. (c) 2018 Jason Yu




Site Map

E-mail ESR



© 1996-2018, Enter Stage Right and/or its creators. All rights reserved.