The Stock / Housing Divergence Since QE1

Howard Simons
  JUN 13, 2011 7:45 AM

It used to be the stock market led the PCH index by about three months. Here, a look at what's happened since quantitative easing began.


If you find yourself parked on the sofa watching some overblown sporting event from an aerial view courtesy of a blimp owner, ask yourself whether the ancient Romans would have understood. The answer is they would have no idea what was going on in an American football or baseball game but they would feel right at home with the spectacle. Theirs was a culture of panem et circenses, of bread and circuses. They might guess as well the event was being sponsored by the local authorities as a distraction; sports do provide an alternative to being ripped to shreds by an angry mob.

Now switch your mental frame of reference from the Super Bowl to the S&P 500 and ask whether the monetary authorities also understand panem et circenses. If they have demonstrated their incapacity to stimulate output and employment via money-printing the same way the fiscal authorities have with their runaway deficits, they certainly could and have pointed to the rally in the stock market as evidence of their magical powers.

The Chicago Federal Reserve produces a personal consumption & housing (PCH) index dating back to March 1967; it has declined every month since January 2007. The duration and extent of this decline, 52 months and counting, is unprecedented in the series and most likely is the longest and deepest downturn since the Great Depression.

Note, however, what its relationship to the S&P 500’s year-over-year changes has been. It used to be the stock market led the PCH index by about three months. More important than the lead-time was the congruence of direction: If the stock market rose, so did the PCH and vice-versa.

What happened after QE1 began? Let’s run a simple in-sample regression model with the PCH index as a function of the S&P 500’s year-over-year changes lagged three months over the June 1967 – February 2009 period and save the coefficients. The fitted values of that model are shown in red below. If we then apply those coefficients to the March 2009-forward out-of-sample period, we see the fitted values in green. They are, starting one year after the 2009 rally began, much higher than the actual PCH index in blue. For example, the actual value of the PCH index at the end of April was -0.39; the out-of-sample fitted value was 0.04.

This is equivalent to saying the change in policy produced a change in financial markets without producing a change in the underlying real economy. There is your circus; you are on your own for the bread part of the deal.
No positions in stocks mentioned.