Wednesday, December 9, 2015

The Future is Bright . . . or perhaps not

 - by New Deal democrat

Bill McBride a/k/a Calculated Risk has reiterated his case for optimism, in "The Future's So Bright . . ."  While I like and respect Bill, and both of us believe the housing market is crucially important, both of us in important part due to a terrific 2007 paper by Professor Edward Leamer, "Housing IS the business cycle,"  this is one point where I part company with him.  It's not that I am pessimistic. I just do not believe that his argument supports his conclusion.

To show you why, let me take each of his points in order, frequently using his own graphs.

First of all, he cites housing, using the following graph of single-family (blue) and total (red) starts:

Notice the huge bulge in multi-family starts from the beginning of the graph (1968) through the late 1980s, as the Boomer generation hit young adulthood.  Bill focuses on this demographic argument, saying "Demographics and household formation suggests starts will increase to around 1.5 million over the next few years."

But note that even that bulge did not prevent huge downturns in 1970, 1974, and 1980-82.  So there is a secular tailwind -- but turbocharged volatility to both the upside and the downside.

Next, Bill cites growth in state and local government jobs, which are clearly lagging indicators for the economy. The reason is, these follow revenue collection, and revenue collection only starts to fall/rise after the recession/recovery starts:

Next, he cites deficit reduction:

But note when the "peaks," or at least smallest deficits have occurred:  in 1989, 2000, and 2007 - in other words, just before the onset of each recession.  This is a coincident indicator.  That the deficit continues to fall simply means that the expansioin is continuing, and doesn't really tell us anything about the future.

Next, he cites household debt levels.  These have fallen to 35 year lows. but are at levels they were in 1981 when the Fed raised rates and engineered a recession anyway! And while they bottomed in mid-expansion in the 1990s and 2000s, they peaked in 1986 and fell off in the years leading to the 1991 recession:

Bill also cites total household debt:

But notice when this graphic last peaked:  in the 3rd Quarter of 2008!  This is a coincident to slightly lagging indicator.  So I don't see where we can make any useful forecasts based on household debt levels.

Next, Bill cites the architectural billings index.  But this is for commercial construction (blue), which lag residential construction (red):

 Further, the index is current at similar levels to where it was in 2000 and 2007:

So I don't see how it can serve as a useful economic forecasting tool.

Finally, he cites the renewed growth in the prime working age 25 - 54 demographic.  If that were true, then real YoY GDP growth ought to correlate well with YoY population growth in this demographic.  But as shown in the  graph below, which subtacts the YoY% change in this demographic from the YoY% growth in GDP, while the deceleration in growth of the prime working age demographic has generally correlated with a deceleration in real GDP since 1985, the relationship does not hold true at all for the 1965-85 period when the prime working age demographic was growing at an accelerated rate - but real GDP growth was decelerating:

UPDATE:  Here is a graph of the two components broken out separately, so you can see the acceleration of growth in the prime age demographic as the Boomer generation hit (blue), but the simultaneous deceleration of real GDP growth (red):

Bottom line: I am not persuaded by Bill's fundamental analysis, which relies mainly on coincident and even one lagging indicator.   I think there is considerable merit to his demographic argument (but see my caveat above), but I think there are two other long-term trends that will prove more important.

First, the long-term decline in interest rates is almost certainly over.  Interest rates will either go sideways or up from here.  Periodically refinancing debt at lower interest rates has been a vital middle class strategy since the early 1980s, but that well has prbobaly run dry:

as supported by Bill's own graph of mortgage refinancing since 1990:

Secondly, the entry of a large demographic into the labor force tends to depress wages (as did the Boomers between 1974-95).  Even without that headwind, wages have basically stagnated since 2000:

The American middle class will only make progress for the next few decades to the extent that its real income rises - something that, with the exception of the late 1990s tech boom, and the oil price crash of 2008, has been largely elusive for 40 years.  Whether the future is bright or not will depend most of all on how that wage issue plays out.