Recession Risks are No Longer Dormant (but not yet acute)

John P. Hussman, Ph.D.
All rights reserved and actively enforced.

Time to update the economic picture.

The latest data on May housing starts indicates that housing activity improved more than expected last month, to 1.957 million starts, while permits for future construction fell by 2.1%. While the permits figure confirmed a cooling market, the seemingly firm level of housing starts painted a confusing picture for analysts.

In my view, even the May gain in housing starts was an important negative, because in the context of how housing starts have behaved over economic cycles, we've already got a breakdown. You don't get much information from housing starts just by looking at whether they're up or down. You have to do a little bit of signal processing to extract the cyclical component. The chart below displays what I think is the proper window. Often, what appears in the data to be just a slight decline in housing activity is actually an informative signal. The blue line is based on housing starts (see the Geek's note below for technical details). Recessions are shaded in gray.

[Geek's note: the blue line depicts the 12-month stochastic (the current point as a percentage of the past 12-month range) of the 24-month exponential smoothing (essentially a moving average) of housing starts. Notice that binary shifts (back and forth between 1 and 0) in this index have historically done a good - though not perfect - job of containing economic recessions.]

Despite the increase in May housing starts, the relevant cyclical component of housing starts has already rolled over, and such rollovers have historically tended to continue (1995 was an exception). Given current economic and interest rate conditions, a continuation appears likely here. Also, though not all periods of housing weakness have resulted in recessions (so a downturn in housing isn't a "sufficient" condition), new recessions have always been preceded or accompanied by a drop in the stochastic to zero (so it does appear to be a "necessary" condition). I expect to be presenting this graph again in the months ahead.

Housing, gross domestic investment, and the current account

From a broader economic perspective, I continue to view the enormous U.S. current account deficit as a major challenge to U.S. economic growth ahead. This deficit reflects the combination of huge U.S. fiscal deficits combined with a historically low domestic savings rate. With no net growth in U.S. gross domestic savings since 1998, the U.S. has depended on large and growing inflows of foreign savings in order to finance our domestic investment (housing, factories, capital spending, etc). We observe those inflows as a current account deficit.

To understand what's happening here, it's important to recognize that by accumulating what is now more than half of the entire float of U.S. Treasury securities, foreign countries have implicitly funded the entire U.S. budget deficit in recent years. This has allowed us to run fiscal deficits without crowding out domestic investment (which is nice while it lasts). In fact, foreign capital inflows have actually financed more than the entire U.S. fiscal deficit. The result is that all of the growth in U.S. gross domestic investment since 1998 has been financed by foreign capital inflows. The chart below tells the story (a good portion of the recent spike represents the U.S. housing boom). The huge and growing gap between domestic investment and domestic savings represents foreign capital inflows (i.e. the current account deficit).

In a reasonable world, one would look directly at the huge U.S. government deficit and paltry U.S. savings rates, combined with a housing boom that's creating a growing inventory of unsold houses, and say, "Wow, we're not saving enough, and we're investing in the wrong stuff, and as a result, we're becoming dangerously dependent on foreign savings that will be hard to repay with future productivity." Unfortunately, we're living in a world with no introspection, so we actually see analysts arguing with a straight face that our current account deficit reflects - I'm not making this up - a "foreign savings glut." That's like the Garth Brooks song where he looks at the beer he's holding and sings "lo-ooong neck bottle, let go of my hand..."

Ultimately, the cause of the current account deficit doesn't matter as much as its sustainability does. The concern here is that even a stabilization - not a reversal, but just a plateau - in foreign capital inflows (i.e. a failure of the current account deficit to continue growing indefinitely) will be enough to halt the growth of U.S. gross domestic investment here, including growth in things like housing investment, factory expansion, capital spending, and so forth.

In my view, this is precisely what's likely to happen over the coming years. That's not to say that U.S. gross domestic investment will collapse, but rather that it is likely to stagnate overall. As a result, any growth we observe in capital spending is likely to come at the expense of housing investment. Of course, a recession would result in an actual decline in gross domestic investment, and would therefore allow us to reduce the amount of capital imported from foreigners. For that reason, a recession would virtually ensure an "improvement" in the current account deficit. Unfortunately, that "improvement" would be nothing more than a reflection of weak domestic investment.

As always, the U.S. would benefit from fiscal discipline, particularly on the spending side, and well-targeted tax incentives, particularly in the area of R&D investment. The best growth policies are those that encourage a combination of saving and productive investment. Such policies don't seem to be much in focus right now, but those are the policies to advocate.

Recession risks - no longer dormant, but not yet acute

On the employment front, it's notable that while the unemployment rate seems acceptably low at 4.6%, part of this reflects unusually low growth in the labor force. Historically, the U.S. labor force has grown by about 1.5-1.6% annually. Over the past 5 years we've only seen the total labor force grow by about 1% annually (due to a falling percentage of the population that is actively working or looking for work). Put another way, if the U.S. labor force had grown at its normal rate over the past 5 years, the U.S. would either have had to create far more jobs than were actually created, or we would currently observe an unemployment rate 2.5% to 3% higher than it is currently (i.e. 7.1-7.6% instead of 4.6%). The relatively tepid growth in actual employment is consistent with the fact that the help wanted advertising index has dropped to its lowest level since the 1950's. Though there's some argument that this index should be expected to fall since it's largely based on print advertising, the current weakness is evident even adjusting for the general downtrend.

In the context of slowing employment growth, a stall in aggregate weekly hours, an emerging widening in credit spreads, and other factors, my impression is that recession risks have increased considerably.

That said, we are still not at the point where I would view a recession as imminent. The main factors that would create that expectation would be a further flattening in employment growth (not necessarily a downturn, just growth in non-farm employment of less 0.5% on a 6-month lookback - which would require employment growth to average about 100,000 jobs per month or less over the next quarter or so), a weakening of the ISM figures toward or below 50 (not all declines below 50 indicate recession, but at present, such a decline would be a strong confirmation of negatives in other indicators already suggesting caution), and a further widening of credit spreads, particularly between 6-month commercial paper and 6-month Treasury bills.

For now, suffice it to say that recession risks are no longer dormant, but aren't yet acute. While we don't yet have enough evidence to anticipate a significant and impending economic downturn, the trends are continually turning in that direction, so it is becoming more a question of "when" than "if."