Blog

The Armbruster Capital Management News & Education section of our website incorporates articles, vidcasts, and newsletters specifically geared towards issues that our clients are facing today.

14 Nov

Break free from the traps of the investment world

Mark Armbruster
VIEWPOINT

Question: What do you call an economist with a forecast? Answer: Wrong. We are constantly inundated with forecasts from “the experts.” This is true not only from economists but also from political pundits, Fed watchers, stock market strategists and other prognosticators. It is fun to watch and consider the soothsayers’ predictions, but is there any actual value to their prognostications?

Take the example of Nouriel Roubini, a professor of economics at the New York University Stern School of Business. He is a well-respected author, educator, and economist who correctly called the collapse of the housing market in 2008, as well as the subsequent global recession. Clearly, given his experience and credentials, he is well positioned to make prescient forecasts.

However, in the years following the 2008 crisis, Roubini continued to warn of further economic calamity, all while the economy rebounded and the stock market rose to record highs. Back when I worked on Wall Street, I saw this first hand. Stock strategists at various firms would rise to prominence based on a correct market call, but would then fall out of favor when they missed the next big event. I started to wonder if this wasn’t akin to a stopped clock being right twice per day.

It turns out my observation wasn’t a new one, nor was it unique to the field of investing. Philip Tetlock, currently a professor at the University of Pennsylvania, had been studying this phenomenon for some time. He studied predictions from 284 different experts across several disciplines. He tracked their forecasts for more than 20 years and had a database of 82,361 predictions to study. The result: The experts are not right very often. In fact, they were correct in their forecasts less than you would expect from random chance…