About Those Forecasts for 2017

The guys at Bespoke recently posted a chart looking at how well analysts have done in predicting the S&P 500. As you might guess, it’s not a great track record.

I was curious to see if there’s a method to the analysts’ madness, and indeed there is. Their forecasts for next year are largely dependent on what the market did this year. I was surprised to see how strong the relationship is.

The analysts are paid tons of money to tell us that the market is subject to…are you ready for this…simple mean reversion. Good years will be followed by bad years, and bad years by good years. The correlation works out to -0.627.

Here’s the scatterplot for the last 16 years.

The horizontal axis is the market’s return for this year. The vertical is the gain expected by Wall Street analysts. Note the fairly strong negative relationship. You’ll also notice the tight range of analysts’ forecasts compared with the real world. (Bear in mind that the horizontal axis is really four times wider than the vertical. I had to zoom in to make it more readable.)

The standard deviation in returns for the analysts is only 4%. It’s nearly 18% for reality.

I’ll simplify what the equation means. Assuming the market will return 10% next year, divide this year’s gain by six, and subtract it from 10%…and presto, you’re now a Wall Street analyst.

So what’s the real world explanatory of this year’s returns for next year? Here you go:

The R^2 is less than 1%.

Posted by on December 30th, 2016 at 12:02 pm

The information in this blog post represents my own opinions and does not contain a recommendation for any particular security or investment. I or my affiliates may hold positions or other interests in securities mentioned in the Blog, please see my Disclaimer page for my full disclaimer.