Watch Out for Economic Forecasts and Negative Extrapolation

by: Scott Rothbort

That was truly a horrible jobs report that was delivered on Friday before the market opened. It will no doubt engender a series of large downward estimates. However, be careful of negative extrapolation. It is my opinion that most economic data points’ estimates are nothing more than guess work.

When I was at The Wharton School of Business in the early 1980s (which had a similar but not exactly same economic environment to that of today) I had the pleasure of taking an advanced macroeconomic class taught by Professor Robert Shiller . Prof. Shiller is now at Yale University and is widely known as co-creator of the Case-Shiller Index. For the term project we were asked to develop an econometric forecast of many data points such as unemployment, U.S. Treasury interest rates and GDP.

It was a fun project for which I devoted plenty of time. I even visited the Federal Reserve to try to get more data and insight. Recall that this was before the days of personal computers and the internet. We had to perform time consuming research with crude tools by today’s standards. I submitted my forecasts along with a written paper (which was typed on an old fashioned electric typewriter) explaining my research, assumptions and conclusions. Prof. Shiller aggregated the entire class’ forecasts and compared them to the actual results. My predictions were at the bottom of the class. However, I received an “A” on the paper / project. The reason according to Prof. Shiller was that my methods and reasons were the best in the class.

So be careful as you start to see economists and market strategists begin to forecast data points in the future. I don’t put much weight into such forecasts. Be even wearier of the less rigorous analysts who just use a ruler and extrapolate in a negative fashion.