January is almost upon us and in the investment business, that means a wave of forecasts are headed our way. After a year of so much uncertainty on everything from trade to the elections, many will find these prognostications soothing. But unfortunately, both the nature of forecasting and our own behavioral biases suggest that we should not place too much confidence in most pundit’s words. Consider the following.
When the Roosevelt administration first proposed Social Security in 1935, they projected that total program expenditures would reach $1.3 billion in 1980 or 45 years into the future. It turns out they were a little off the mark; actual Social Security outlays in 1980 totaled $149 billion. The dismal history of oil price forecasting provides a more recent example of so-called “experts” getting it wrong. Back in 2007 with oil prices hovering around $100 per barrel, government researchers forecasted higher oil prices thanks to what they believed was peak global oil production and growing energy demand. While they were right for a short period of time (oil prices did spike to $140/barrel over the next six months), they failed to predict the rise of unconventional oil sources and new horizontal drilling technology. Less than one year later, oil prices had plunged to less than $40/barrel before recovering to their current $60 level.
These two examples got me thinking about forecasting in general. Why is it so difficult to make accurate predictions and why are we so captivated by them despite this? The answer on both fronts has to do with several common behavioral traits. Let’s start with the forecasters themselves. Experts are typically asked to make predictions about topics whose ultimate outcomes are based on multiple, complex inter-related issues; how fast will the U.S. economy grow next year? What about the direction of interest rates and what happens if the U.K. breaks out of the Euro zone – just to name a few today. But experts in many fields forecast not because they know but because they are asked. In most cases, these same experts have spent years studying the topic at hand. Interestingly, research has shown that more information – more data, more studies – often does not improve forecast results. What the additional information does consistently do, however, is increase the forecaster’s confidence level around their predictions.
If overconfident experts are partly to blame so are the individuals following them. Much of life requires us to make decisions in the face of uncertainty but as humans, we are hard-wired for “cognitive closure” and crave predictability. Whether warranted or not, “experts” meet this desire for predictability. Confirmation bias or the tendency to seek out information that reinforces our pre-existing views also factors in as individuals work to quell the anxiety created by uncertainty.
In the investing world, these behavioral biases can cause some particularly troublesome behavior including the tendency to underestimate risk, inadequately diversify and trade too much. Investors can employ a few easy techniques to combat these unhelpful behaviors. First, when making bets (in the markets or otherwise) be sure to keep detailed records of both your predictions and your confidence level around them. Intentionally, combat confirmation bias by seeking out information on opposing viewpoints. And finally, be willing to revise your beliefs as needed when new information comes your way. Good forecasting requires us to be aware of both what we know and what we don’t know. While rare, this sort of self-awareness can be helpful in both the stock market and in life.