I spend most of the time focusing on forward earnings estimates, whether it be for individual stocks or the S&P 500 index as a whole. The reason for this is when you buy stock, you are buying into a share of that company's future earnings. The market doesn’t care about the past, it's a “what have you done for me lately” business. It’s true that companies with a history of stability and profitability increase the odds of success going forward, but stocks trade on a company's future growth potential and how those results come in compared to those estimates.
I get a lot of questions as to why I pay so much attention to forward-looking estimates when the future is uncertain. First, we have to understand and accept the fact that nothing is going to be 100% foolproof. There is no indicator or data point that is going to predict the outcome, especially the shorter time frame.
But as it turns out, analyst earnings estimates are pretty reliable on balance. For this example, I have taken the average forward earnings estimates for the S&P 500 every week over the last 15 years and compared them to what got reported 12 months later.
The average over that 15-year time frame is -0.15%, so basically even. This means that on average, the forward earnings estimates have pretty much equaled what got reported 12 months later. In other words, analysts do a pretty good job.
70% of the time, actual earnings are reported between +5% and -5% of what analysts had predicted they would be 12 months earlier.
And this even includes the COVID shutdown (something no one could have ever predicted) and recovery. Analysts usually don’t include economic recession probabilities in their forward earnings estimates. So when we do get those inevitable recessions, all bets are off. In 2020, earnings results came in about 20% below what was first estimated beforehand. But, the recovery pushed earnings results over 20% above estimates. So in the end it was basically a wash.
No one method of analysis is perfect. You should never rely on one data point alone. This blog focuses on forward earnings, but also many economic indicators as well. The point of this is to estimate the probability of a recession. If recession probabilities are low, earnings estimates have better odds of being reliable.
In the end, forward earnings estimates are a good starting point. They are pretty reliable for the most part. And they balance out in the long term.