Tomorrow = Today + 1

As we head into the final quarter of this year, corporate executives and analysts have to shore up their forecasts for next year’s earnings growth. In a world where corporate guidance on earnings is common place, most analysts are taking their cue from the people who should know best what the prospects for next year are: CEOs and CFOs.

Unfortunately, forecasting is hard particularly the future and corporate executives are no exception. It’s not that they consciously mislead the market. A new paper shows that there is no systematic exaggeration of earnings forecasts by managers. Instead, managers fall for simple extrapolation bias. They look at earnings this year and then simply assume that this year’s growth is a good indication for next year’s growth. But because strong earnings growth tends to weaken over time (and weak earnings growth tends to strengthen), the managers that extrapolate the most and have the highest current earnings growth tend to make the biggest error in their forecasts. This effect is so big that in the long run, companies with the highest projected earnings growth underperform the companies with the lowest projected earnings growth.

This simple extrapolation of current earnings growth becomes particularly fraught with error after a year like 2021. After all, the growth rates many companies saw this year are extremely high and heavily distorted by the low base from the pandemic. For investors, it is worth knowing that this year corporate executives likely overestimate next year’s earnings growth by a wider margin than normal. And it is useful to know which company executives are more likely to extrapolate current earnings into the future and become too optimistic about 2022.

As I said, some fraudsters and con artists aside, corporate executives make an honest attempt at forecasting next year’s earnings accurately. And the ones that do worst are the ones that work at companies that have earnings that are particularly hard to forecast.

Managers in companies with very erratic earnings are more prone to extrapolate this year’s earnings growth into next year. After all, if your earnings are all over the place, it is difficult to make accurate forecasts and you are likely to use current earnings as a stronger anchor for your expectations.

The second group of companies that tend to suffer from extrapolation bias are companies with very salient earnings. A company that has long been loss-making but this year became profitable for the first time ever (or at least in many years) is more likely to extrapolate the success of this year into the future. On the other hand, companies that did poorly this year will be more cautious in their earnings forecasts, even if this year’s poor results were clearly a one-off and are highly unlikely to repeat next year. These are the companies that are more likely to surprise to the upside next year and outperform.