The Road To Recovery
Goldman analysts today provide a very useful checklist to keep tabs of where we stand on the proverbial road out of hell. As GS says, its US economic forecast calls for stabilization by middle 2009, and after annualized declines of 7% and 3% for Q1 and Q2, the bank expects real GDP to inch up at a 1% rate over the second half of the year. Ironically, even Goldman admits that "many clients are quite skeptical of this forecast, although it is a view widely shared by other forecasters. For example, the median expectation of the 52 forecasts surveyed by the Blue Chip Economic Indicators in February was for third-quarter growth of 0.8%. The bottom ten averaged -1%, indicating that virtually all of those surveyed expected at least a significant moderation in the current down trend. Many of these forecasts have probably been marked down since then, but our guess is that the March survey will still show a consensus among forecasters that the US economy will stabilize by the summer or early fall."
The "roadmap" below should be followed by all market participants to see if Goldman, among other forecasters, is merely hoping for an economic hockeystick with a Q2/Q3 inflection point. Prudently, Goldman hedges itself:
"The recent tightening in financial conditions poses an obvious downside risk to our economic outlook. For example, since early January our Goldman Sachs Financial Conditions Index (GSFCISM) has risen nearly 2½ index points, due largely to the selloff in equities with dollar appreciation and long-term yield increases playing supporting roles. This essentially returns the GSFCISM to its highs of last autumn, indicating extreme tightness in financial conditions. It is difficult to lay out precisely the developments we need to see in financial markets to accompany the anticipated improvement in economic activity, especially given that the index has not captured some key aspects of the current crisis—specifically, the differential effect of tightening in the mortgage market and the seize-up in other markets. Suffice it to say that if financial conditions do not ease materially in coming months we would distrust the stabilization story even if we saw most of the mileposts on the road map."
1. Vehicle sales stop falling. A rebound in vehicle sales is not absolutely essential to our view that real consumer spending is stabilizing, but it would be a promising sign of improvement in confidence and credit availability. In February, sales of lightweight motor vehicles fell to a 9.1-million unit annual rate; a year ago this rate stood at 15.3 million. We would not rule out small further declines in the next month or two, but our base case is that the sales pace six months from now will be somewhat higher than this—closer to the middle of the 9- to 10-million unit range.
2. Core retail sales trend up modestly on balance. The dollar volume of retail sales of goods excluding vehicles, building materials, and gasoline provides the first reasonably reliable read on outlays for non-auto goods in any given month. In January, these sales jumped 1.2%. This was the first large increase since June; during the second half of 2008, core sales fell nearly 4% (not annualized), or about 0.7% per month on average. The January bounce was most likely due to clearance of unsold holiday merchandise; for February we estimate that core sales were about flat (the data are due this Thursday). We need to see more consistent evidence of improvement to confirm the view that consumer spending has stabilized, though the improvement does not need to be large. Increases averaging between 0.1% and 0.2% per month over the next six months would probably be enough.
3. Confidence improves. Confidence indexes provide timely though very loose reads on consumer spending and, in the case of the Conference Board’s index, on labor market conditions. This index plunged to a new all-time low of 25.0 in February; five months earlier it was 61.4, already a low level. Given our outlook we would be surprised if the Conference Board did not report at least some reversal of this loss over the next six months, though disappointment on this score could simply reflect lags in consumers’ perceptions of incremental improvement in labor market conditions, discussed more fully below. The hurdles are less high for the other two main measures—Michigan’s monthly index and the weekly ABC consumer comfort index—as they have not fallen as sharply, though we do expect some improvement from the historically low ranges in which they have fluctuated recently.
The damage to housing activity has been long and deep, pushing both sales and starts of new single-family homes well below the levels consistent with longer-term demographic trends. Although a significant recovery in the sector is unlikely as long as a large excess supply of unoccupied homes persists, we do expect the downward spiral in home building to peter out in coming months. We would not be surprised at a modest rebound in starts and new home sales given the low level to which they have fallen, but we are not counting on this. The more important requirement is evidence that the downward trend in residential construction outlays is slowing. Thus, the indicators in this sector should shape up something like this:
4. Home sales probably continue to languish. Although sales of new homes have fallen even farther than we had estimated, the gap between these sales and underlying demand may persist for a while as foreclosures, credit constraints, and poor economic conditions inhibit recovery in demand for housing and keep the excess supply of unoccupied homes high. Thus, while we have penciled in a bounce from the abysmal 309,000 annual rate to which new home sales fell in January, we would not be surprised if this failed to materialize; in the market for existing homes, which has held up better due to sales of foreclosed units, risks to the 4.49-million sales rate reported for January are to the downside.
5. Housing starts may also remain low, but should not trend down further. For essentially the same reasons, we are not counting on starts of new housing projects to increase materially from the levels posted in January—annual rates of 466,000 overall and 347,000 for single-family structures—even though these are somewhat below the levels at which we thought these key housing indicators would bottom out. However, additional declines in starts would call into question whether the end of the collapse in home building is yet in sight.
6. But declines in construction spending should abate. The data we will watch most closely to verify expected stabilization in home building are for residential construction outlays. They have fallen 2.2% on average (not annualized) over the past six months, as the number of units under construction continues to drop. As completions of these units approach the subdued starts rate, the pace of decline in residential construction outlays should diminish substantially, though it may not disappear altogether given the downward trend in prices in this sector. Monthly declines averaging ½% to 1% by the third quarter would be roughly consistent with stabilization in real outlays, in our view.
As noted, ongoing declines in capital spending will probably continue to weigh on durable goods orders and industrial activity in general (as well as on the private, nonresidential component of construction outlays). In addition, exports are apt to fall further in the near term as the global economy remains weak. Accordingly, while industrial indicators should improve somewhat relative to recent extremely weak trends in response to the anticipated stabilization in consumer spending, we do not look for full stabilization in this sector:
7. Orders for durable goods continue to fall. Bookings for nondefense capital goods, in particular, are apt to continue falling sharply as companies cut back their purchases of equipment and software. Over the past six months these orders have slid at an average monthly rate of 3.9%. A continuation of this trend would not be sufficient grounds to abandon the expectation of stabilization in real GDP, though we would expect the trend in total orders for durable goods—down 4.6% per month over the same period—to abate somewhat reflecting better demand in other sectors.
8. Indexes of industrial activity turn up, though not necessarily to 50. While we expect capital spending and exports to continue to weigh on the US industrial sector, stabilization in consumer spending and in demand for construction materials should soften the downward trend in total output in this sector. Accordingly, we expect indexes of manufacturing activity to close part of the large gap that now exists between the latest readings and benchmarks of unchanged activity in the sector—50 for the Institute for Supply Management’s (ISM’s) index and zero for indexes compiled by various regional banks of the Federal Reserve System. For the ISM index of manufacturing, this means a move into the low to mid 40s (readings of 42-43 are statistically consistent with no change in real GDP) from 35.8 in February; for the Fed indexes, moves into single-digit negative territory would be more than enough. Given the weakness in manufacturing, our outlook would also imply a modest rebound in the ISM’s index of nonmanufacturing activity.
9. And the downward trend in industrial production moderates to single digits at an annual rate. This would mark a significant improvement from the recent trend, which is close to -30% for the first quarter in the Fed’s index of manufacturing output. By the third quarter, we expect this to be -7.5%, though figures around -10% would not be too unsettling if other elements of the road map were in place.
Although the labor market is widely regarded as a lagging sector, and correctly so as the economy approaches a cyclical trough in real GDP, we will need to see some signs that the pace of deterioration is slowing to have any confidence that other signs of stabilization are genuine. This includes, most notably, some easing in the one unquestioned leading indicator in this area, namely initial claims for jobless insurance benefits, but also a clear reduction in the rates at which companies are slashing payrolls and the unemployment rate is rising.
10. Initial claims for jobless benefits fall significantly. This indicator is usually the first in the labor market to show incremental improvement, and persistence of filings at the current rate—about 640,000 per week over the past month—would cast serious doubt on the sustainability of other signs of stabilization, especially those for consumer spending. Although we do not forecast initial claims, we would be concerned about our outlook if they did not drop at least below 500,000 over the next six months.
11. Payroll declines lessen considerably. Stability in real GDP does not imply stability in payrolls, as companies initially meet improvements in demand via productivity gains and increases in work schedules for the existing workforce. Thus, we expect to see payroll losses persist for a while even as real GDP edges up modestly, as occurred in the two most recent “jobless” economic recoveries. That said, the pace of decline should diminish sharply, to 200,000 per month or less (an annualized decline of 1.8% or less) by the third quarter.
12. Increases in the unemployment rate drop by about half. Likewise, while anemic growth in real GDP is not enough to stop the unemployment rate from rising, it should have a noticeable effect on the trajectory of the increase in joblessness. We estimate that a 1% growth rate in real GDP is consistent with a 1-point year-to-year increase in unemployment, or monthly increases averaging just under 0.1 point, less than one-third the 0.3-points per month we have seen on average since mid-2008. However, given the lagging nature of the unemployment rate at this phase of the business cycle, we would expect monthly increases to be slightly larger than this—say, between 0.1 and 0.2 points.
With gratitude to GS whom we make fun of way too much.