John Maynard Keynes had famously said that in the long run we were all dead. The point of the current note is not to argue with the great man, but to add that, apparently, in the long run, economic policy is, in some ways, irrelevant. Our evidence for this is the observation in I. Rivin’s recent paper.

What is looked at is the (dividend-reinvested) return of the Standard and Poor’s 500 stock index for the last 150 years not adjusted for inflation. Here is the graph:

Natural logarithm of the size of a $1 investment in the S&P 500, with dividends reinvested.

The graph above uses monthly data, as provided by Robert Shiller on his very helpful site.

Now, the graph above looks an awful lot like a straight line. To see how much like a straight line it is, there is a regression analysis in Rivin’s paper, and the trend accounts for 99% of the variance! (for the curious, the annualized slope of the line is 0.1035, in other words the expected gain (again, not adjusted for inflation) is 10.9% a year.

The main question this leads to is the obvious: Why is the trend so strong?

The main takeaways are:

  • Go long when the curve goes below trend, go short when below.
  • Since (for the third time) the graph is not inflation adjusted, inflation is the real return killer.

To make the last point clearer, let’s now plot the “deflated” returns of the S&P 500 (also in the referenced paper):

Log of an inflation-adjusted $1 invested in the S&P 500 (with dividends reinvested).

The graph is much less straight now (correlation is now “only” 90%, vs 99.5% before), and one can see long flat sections.

Leave a Reply