Thursday, March 27, 2014

A Regime Shift in Interest Rates and Stocks

I thought I would do an update on a pattern I initially mentioned last summer.

10 Year Treasury Rate (green, left scale)
real S&P 500 Level (blue, right scale)
There seem to be two regimes, regarding the relationship between equity returns and interest rates.  (I am using 10 year treasury rates.)

From about 1968 to 1995, inflation remained above 3%.  During this time, interest rates were negatively correlated with equity prices.  When interest rates went up, equity prices usually went down.

10 Year Treasury Rate (green, left scale)
real S&P 500 Level (blue, right scale)
From 1997 to the present, when inflation has generally been below 2.5%, interest rates and equity prices have been positively correlated.  Rates and equities have moved up & down together.

I wonder if this is because during the earlier period, the Fed was erring on the side of being too loose, so higher rates reflected a risk of suboptimally high inflation; but during the recent period, the Fed has been too tight, so that low rates reflect a risk of deflation and low real rates associated with economic decline.

Simple regressions for both periods produce an r-squared value of about .2 (data is monthly).  In the earlier period, a 1% YOY increase in 10 year treasury yields was associated with a 4% YOY loss in the real S&P500 level.  In the current period, a 1% interest rate increase is associated with an 11% increase in the real level of the S&P500.

This relates, I think, to my recent posts on asset allocation.  It is only in the current, low inflation regime where bonds provide strong asset class diversification.  In this regime, equities have declined when interest rates have fallen (long-duration bonds gain when rates fall).  But, this negative beta may only be in effect when interest rates are very low, when there is a dual drag on bonds - both from limited income and limited potential for capital gains.

One issue to watch if we move back to a high inflation regime would be TIPS.  It might be possible that inflation protected bonds will provide very low or negative beta with equities in both regimes.

PS.  I forgot about this post from last year on this topic.  I guess it took less than a year to start repeating myself.

No comments:

Post a Comment