Thursday, July 4, 2013

A Just-So story about the recent divergence of stock prices and interest rates (Pt. 2)

I ran across this reaction to the June Fed announcement from JP Morgan:

JP Morgan June 21, 2013 Market Bulletin

They agree with me that the main change from the Fed announcement was the decline in uncertainty about the Fed's balance sheet & future inflation, not a change in the average expectation of Fed actions.  The place where they diverge from my previous post is that they think the drop in equities was an overreaction.

In addition to the idea of equities being an inflation hedge, which could explain a small sell-off on this news. It has occurred to me that an additional factor could be at play:

Here is a graph of the 60 day daily beta between a long term TIPS ETF (LTPZ) and the S&P 500:

Inflation protected bonds have been highly negatively correlated to stocks as a result of tight Fed policy.  As the economy normalizes, and there is less danger of erroneous Fed discretion, this relationship should not be as strong.  As shown here, the correlation began breaking down in late May and the reversal strengthened at the Fed announcement.  This can become a self-fulfilling outcome, because, per modern portfolio theory, the high negative correlation could have been driving leveraged long positions in both equities and TIPS bonds.  A sudden shift in that relationship could cause a sell-off in both asset classes.

One last comment on the JP Morgan bulletin, and on market commentary in general:  Everyone treats the rise in interest rates as a Fed failure, which is understandable given the Fed's own stated goal of lowering rates.  But, the Fed's unfortunate policy of communicating through rate targets notwithstanding, I consider the higher rates to be clearly a positive, if for no other reason than that the greater distance from the price floor of the zero lower bound allows fixed income markets to clear with less distortion.  In the current context, higher rates should be taken as a sign of Fed success unless there is clear evidence to the contrary.

Addendum:  I need to do some more work on this.  Here is a chart that shows the correlations (weekly, 30 periods, rolling) between TIPS, nominal bonds, and expected inflation, all correlated with the S&P 500.  The correlations here are positive because this is based on yields, not bond prices.  Some of my intuition about the recent context is still possible.  A higher correlation between inflation and equities in general suggests that the Fed has been too tight.  I think real (TIPS) rates rising roughly coincident with the QE phases suggests that periods of loosening did create positive real effects in the economy.  But, I need to wrap my head around the relationship of changes in real rates to equity returns.  I have a graph here of nominal rates regressed against equity returns, again weekly in 30 week rolling periods, and the positive correlation of the past decade is anomalous.  It could be that the correlation itself is negatively correlated to the general level of inflation, so that the bond yield to equities correlation is only positive in low inflation contexts.  Or, there could be something structural about long term changes in the source of American equity earnings.
In any case, I can make an argument for the apparent drop in correlations being a product of reduced inflation uncertainty.  But, the case seems weaker if I can't tie it into a fuller explanation of long term rate correlations.

Signs of stagnation in the JOLTS data lead me to worry that we are in the midst of a period like the 70's, where either demographics or political structural issues create poor real economic growth, low real interest rates, and persistently high unemployment.  If the Fed remains hawkish, sticky wages and prices could worsen these problems (especially since some of the structural political problems create high inflation in non-wage benefits).  If the Fed turns dovish, they might push inflation to uncomfortably high levels if structural unemployment proves persistent.  The Fed seems extremely hawkish right now, but a paradigm shift could certainly happen if unemployment bottoms out above 6%.  Of course, while loose monetary policy can mitigate against unemployment at low inflation rates, the lesson we learned from the 70's was that this relationship breaks down at high inflation rates.  In other words, tight money can exacerbate unemployment problems, but loose money can't solve structural unemployment problems.

No comments:

Post a Comment