The Economy Is Growing Faster Than You Think
In the popular view, the economy is not in recovery until the unemployment rate, the most political of lagging indicators, begins to decline. Finally now it has, slipping from 10.2 to 10 percent in November. US output has already been on the rise since the spring, in the face of sagging employment. Increases in productivity are offsetting diminishing man-hours of input. Conclusive signs of a turning point have yet to shake most commentators out of their gloom, and the vigor of the recovery continues to be questioned. All too typically, it's input (i.e. employment) rather than output that receives attention from politicians or ink from journalists.
Already back in 1650, historian Thomas Fuller observed that "it's always darkest just before the day dawneth." That's why the recent positive labor-market news was not greeted with much enthusiasm. One regional Fed president, Charles Plosser, responded that "one month does not a trend make," to which it could be retorted that "every spring must have its first swallow." Today it remains normal for popular and expert sentiment alike to be at odds with market intelligence at economic turning points. Investors swayed by the general pessimism have been missing out on extraordinary gains in risky asset prices since early this year.
The government's preliminary estimate of 2.8 percent annualized growth for the third quarter has done little to dispel the pessimism even though it represented an acceleration of 3.5 percentage points from the 0.7 decline in the second quarter. That in turn was an acceleration of 5.7 percentage points from the severe contraction of the first quarter. Another three months of like acceleration would bring growth up to six or seven percent for this quarter.
That's the kind of growth rate that markets have been signaling. The best market predictor I know of is the yield spread between investment grades in the industrial bond market as defined by Moody's. The sudden widening of these spreads accurately predicted both the magnitude and timing of the downturn last year, and their equally rapid return to normalcy is now predicting an explosive recovery.
This simple market-based indicator has several advantages over the confusing plethora of theoretical arguments being tossed around by forecasters. First, its track record is pristine; during its 90-year history it has faithfully mirrored the economy's cyclical ups and downs. Second, it has credibility; as a derivative of transaction prices, it reflects the objectivity of the financial-market system.
Moreover, it has a natural interpretation as an index of risk tolerance, in that it quantifies the changing uncertainties that influence the willingness with which capital is placed at risk and put to work.
I disagree with the popular idea that the economy is driven by spending or "stimulus" as if politicians could capture and bring in resources from outside the economic system. Mine is a classical perspective in which economies grow simply as the result of capital being put to work. Although capital may be plentiful, it slips away when uncertainty erupts and returns when uncertainty abates. Bond spreads are a market measure of this ebb and flow.
The Baa-Aaa spread has dropped from its end-of-2008 peak of 350 b.p. to just below its long-term average of 120 b.p. Such a dramatic narrowing implies exceptional rates of growth-seven percent or more-in the current and following quarters. The basis for these exceptional growth expectations is the strong correlation between seasonally-adjusted quarterly GDP and the changing spreads. The last 60 years shows that it is rare for the quarter-to-quarter spread change to exceed 40 basis points in either direction; but when it does, dramatic economic results follow.
There were only four occasions during that time when the Baa/Aaa spread narrowed by more than 40 b.p. within a single quarter: one during the recovery from the recession of 1980, two during the recovery from 1982 and one during the recovery from 1974-75. These four incidents were followed by an average of at least seven percent in real GDP growth in the two quarters following. There were also four occasions on which the spread widened by more than 40 b.p. within a single quarter. These incidents were accompanied by real GDP declines at a rate of between 4.5 and 5 percent.
Until recently, there hadn't been any 40-b.p. incidents since1984 because the recessions of 1991 and 2001 were relatively mild. Extreme episodes did occur in the 1930s. Quarterly data are not available, but we do know that the vigorous rebound in real GDP in 1934-36, averaging over ten percent per annum, accompanied the greatest narrowing of spreads on record. When this most extreme case is compared with the postwar relationship, it provides further conformation of a tight fit.
The best should therefore be yet to come. The narrowing of the spread this year has been the largest since the 1930s. From the second to the third quarter the Baa/Aaa spread fell back by 108 b.p., more than twice the 40-b.p average for the four incidents that saw seven-percent growth. This suggests that a forecast of seven percent for the fourth quarter and the first quarter of next year may be conservative.
An important feature of the relationship is a timing difference between the response of the economy, depending on whether spreads are widening or narrowing. When they widen, GDP responds almost immediately with negative growth in the same quarter and that following. But when they narrow, the impetus to growth is strongest in the two subsequent quarters. Analysis reveals an average time lag of less than two months from widening spreads to their maximum impact, but more than four months following a narrowing. This difference in timing has a natural explanation. It is easier for businesses to curtail operations when uncertainty mounts than it is to get them up and running again when fears abate.
This explains why GDP did not show dramatic growth in the third quarter. It was one quarter too early. The historical fact remains that the vigor of US economic rebounds is broadly proportionate to the depth of the preceding downturns. I don't think it will be different this time around, and I don't think the credit should be given to government "stimulus."