John Huston

John Huston
Contributor since: 2012
Tom, This is a nice exploration of this topic. Two extensions occur to me.
1. To add to your theory section, interests rates should also directly affect equity prices. A more attractive risk free interest rate should reduce demand for stocks. That it would add a force moving us towards equilibrium.
2. Another way of tackling the mean reverting question would be a unit root test. As you point out, the tricky part is the time horizon. Over Shiller-like time horizons, (1891-2012) they may be stationary. Only extremely healthy investors could make us of that knowledge!
Thanks for the article. I enjoy this sort of analysis.
I share your skepticism about some research on stationarity. If there have been regime changes then by definition the series hasn't been stationary. Or at best, if we conclude something is mean reverting within a regime one is left wondering if the current regime is stable.
This series is a bit different since it is (p/e)*i Even with the addition of the treasury rate whether or not it is stationary would depend on the time frame. We chose Shiller's long time horizon but had we looked at various shorter periods, e.g. 1975 to 2000, it would not be stationary.
The many comments on the unusually low treasury interest rates are largely on target. In fact, there are parallels with some of the other time periods mentioned in the article. During WWI and WWII there were pressures on the Fed to keep interest rates artificially low so as to reduce the costs of government borrowing.