Thank you, Mr. Chairman. Many of the spending indicators that we’ve received over the past few weeks have been coming in to the weak side of our expectations. As you can see by comparing the red and blue striped bars in the top left panel of exhibit 1, in the Greenbook we responded to the weaker numbers by revising down our projection for second-quarter growth of real GDP to a 2 percent annual rate, about 1¾ percentage points less than in the May Greenbook. We interpreted some of the unexpected softness as reflecting weaker underlying demand, and we let that part feed through beyond the current quarter as well. But as I’ll discuss shortly, the softness of the spending data wasn’t the only factor leading us to temper our forecast for the out quarters.
A key component of the downward revision to the current quarter has been a slower-than-anticipated pace of consumer spending—the panel to the right. Based largely on the available data for retail sales, motor vehicles, and the CPI, we estimate that real PCE in the second quarter rose at an annual rate of only 2.2 percent. The slowing reflects a step-down in purchases of light motor vehicles as well as essentially flat real outlays for goods other than motor vehicles in the first two months of the quarter.
The other big surprise in the recent data is the sharper-than-expected drop-off in housing activity, as illustrated in the middle left panel. Despite the May uptick in housing starts, the levels of both starts and permits are well below where we thought they would be at the time of the last Greenbook.
In contrast to the household sector, the business sector indicators generally have been favorable. For example, orders and shipments for nondefense capital goods, plotted to the right, continue to point to strengthening demand. In addition, business spending for new structures, not shown, appears to have posted another sizable gain this quarter. The scant data available for June do not suggest that the slowdown in spending earlier in the quarter has been cumulating. As shown in the bottom left, initial claims for unemployment insurance, after retracing the rise that was associated with the effects of the temporary government shutdown in Puerto Rico, have remained in a range consistent with further moderate employment increases. Meanwhile, as shown to the right, the Empire State and Philadelphia business surveys, which were less favorable in April and May, were considerably more positive this month.
Thus, while many of the indicators of activity have lately been to the downside of our expectations, the news has not been entirely negative. All told, we still see economic growth as being in a transition—to a pace somewhat below that of potential. This can be seen in exhibit 2, which presents the longer-run outlook. We now think that real GDP, as shown by the striped blue bars in the top left panel, will increase at a 2¾ percent rate in the second half of this year and through 2007.
As I mentioned earlier, several developments besides the incoming spending indicators influenced our thinking about the economic outlook. First, revised data from the BEA now put the 2005:Q4 change in wage and salary disbursements, plotted by the black line in the top right panel, considerably below the figure published at the time of the last Greenbook, the red dashed line. The reduction in labor income suggests less consumption going forward. Regarding our assumption for monetary policy, the funds rate is assumed to be increased to 5¼ percent at this meeting and to remain there through the end of 2007. As shown in the middle right panel, the Wilshire 5000 stock market index is about 7 percent below the level we had assumed in the May Greenbook. We assume that equity prices will rise at a 6½ percent annual rate going forward—a pace that roughly maintains risk-adjusted parity with the returns on Treasury securities. The latest reading on house prices from the OFHEO index, illustrated in the bottom left panel, was in line with our expectations. We continue to expect an appreciable deceleration in house prices over the projection period. Finally, after several Greenbooks in which we revised up our forecast of crude oil prices, this round we revised down our forecast. When we prepared the Greenbook, the spot price of WTI was just under $70 per barrel, about $5 less than at the time of the May Greenbook. And with futures prices also lower, we reduced the path of crude oil prices throughout the projection period.
Exhibit 3 presents some details of the outlook for business fixed investment. As illustrated in the top left panel, total real outlays for equipment and software, excluding the volatile transportation equipment component, are projected to increase 7½ percent over the four quarters of 2006 and to grow nearly as fast in 2007. You can see from the red portion that the bulk of the support comes from spending for high-tech equipment.
Much of the recent strength in the high-tech sector has reflected a spurt in capital spending for communications equipment by telecom service providers, whereas demand for servers and PCs, the top right panel, has faltered. However, our forecast calls for real outlays for computing equipment to pick up later this year and to be sustained in 2007. Industry analysts cite several upcoming technological developments that should boost demand for new servers and PCs. These are summarized in the middle left panel. With regard to servers, several manufacturers plan to offer new generations of servers this autumn that offer significantly faster computing power and, just as important, lower electricity consumption. In addition to ongoing demand for large-scale servers from financial services companies, demand for clusters of small-scale servers—so-called server farms—by Internet content providers, for example, also appears to be robust. With regard to PCs, Intel will be introducing a fundamentally new chip design in the second half of this year that reportedly will increase performance and significantly reduce power consumption. With the new generation of chips on the horizon, prices on old chips have been plummeting. We project that demand for PCs should step up later this year, spurred by the combination of the new generation of chips for high-end users and falling prices on old chips for middle- and low-end users.
The table to the right presents our forecast for real spending on nonresidential structures. As shown in line 2, outlays for drilling and mining have been growing briskly. Reflecting the sharp increases in prices of oil and natural gas over the past few years, the number of drilling rigs in operation, the bottom left panel, has been climbing steadily, with much of the increase for the natural gas component. We see the growth of activity slowing somewhat next year as the prices of natural gas and crude oil begin to flatten. Outside drilling and mining, investment in new buildings has strengthened recently, as vacancy rates (illustrated for office buildings by the black line in the bottom right panel) have been trending down and rents received by building owners (the red line) have been climbing. That said, we think the current rate of investment growth is unsustainable, given our projection of decelerating employment and business sales, and consequently the growth rate of construction spending slows in our forecast.
Turning now to the household sector, the subject of exhibit 4, we expect real PCE, the red bars in the top left panel, to increase at a rate of about 3 percent in the second half of this year and to stay close to that pace in 2007. The forecast reflects two important crosscurrents. On the one hand, real income growth, the blue bars, is projected to be robust reflecting, in part, a waning drag from higher energy prices. On the other hand, the wealth-to-income ratio, plotted by the black line in the top right panel, falls in our forecast as house prices decelerate. With slower gains in wealth and rising interest rates, we expect that spending will be restrained relative to income and, accordingly, the saving rate will rise. The remainder of the exhibit examines the housing market. Looking through the monthly volatility, you can see that sales of new homes and existing homes, the middle left panel, are well off their peaks, whereas backlogs of unsold homes, the panel to the right, have increased significantly. In putting together the forecast, we factored in the recent data on starts, sales, and inventories, which led us to mark down the profile of activity throughout the forecast period. All told, as shown in the bottom left panel, we expect that real residential investment spending will drop 5¼ percent this year and fall another 1¾ percent next year.
Widespread anecdotal reports suggest that the drop in demand is being driven partly by a withdrawal from the market by investors and purchasers of second homes. Data on mortgage originations by investors and those purchasing second homes, which begin on a monthly basis in mid-2003 and are available only through March, are plotted to the right. The first thing to note is that these groups are a relatively small part of the overall market. Moreover, although the data are quite noisy, neither group, at least through March, was leaving the housing market in droves. Nevertheless, the recent spate of reports and a jump in the rate of contract cancellations for new homes, which homebuilders attribute largely to a retreat by investors, pose a downside risk to the housing outlook.
Another risk to the forecast is the possibility of a pronounced deterioration in the financial conditions of some vulnerable households, which could cause them to retrench significantly on spending. As shown in exhibit 5 in the top left panel, the financial obligations ratio for homeowners has risen sharply over the past year or so. It reached a record high in the second half of last year, and we estimate that it rose further in the first quarter of this year. (As an aside, the rate for renters has actually been falling since mid-2001, though that could be changing.) Some analysts have expressed concern about the high level of the financial obligations ratio—especially in light of the potential for further increases in obligations related to variable-payment mortgages, which represent more than one-fifth of all outstanding first-lien mortgages. Nowadays, variable-payment mortgages typically carry a fixed interest rate for a few years before converting to floating-rate, and most aren’t scheduled for a payment change for some time. The top right table presents some evidence. As you can see in the last column, the bulk of variable-rate mortgages—both ARMs and interest-only—that have yet to reprice won’t begin to reprice until after 2007. Moreover, the end of the interest-only term for nearly all I-O mortgages that are awaiting the end of the I-O term is well in the future—line 3 of the table.
Given these patterns, scheduled mortgage payment changes should have only a limited effect on the aggregate mortgage burden—adding just a few tenths to the homeowners’ financial obligations ratio this year and next. As a result, we have not seen—and don’t expect—a broad deterioration in mortgage credit quality. For example, as shown in the middle left panel, delinquency rates for prime borrowers, who account for 85 percent of the mortgage market, have been relatively flat for some time—as illustrated by the black line.
That said, there are some indications of stress among subprime borrowers. This group presumably is at greater potential risk for financial stress generally: Note the higher levels of their delinquency rates, the red line, compared with the prime market. In particular, we are seeing a deterioration among subprime borrowers with variable-rate mortgages—the red line in the panel to the right. This type of loan is far more prevalent in the subprime market, representing about two-thirds of all subprime mortgages. The fixed-rate period for subprime variable-rate loans is considerably shorter than that for prime loans—typically one or two years versus five to seven years—so more subprime borrowers with variable-rate mortgages are likely to see their monthly payments rise over the projection period. And the increases for those subprime borrowers experiencing resets could be striking: We estimate increases of something like 25 to 30 percent of their original payment.
More generally, households that are likely to be more financially vulnerable appear to have become decidedly more pessimistic. The bottom left panel plots the Michigan consumer sentiment index stratified by income. Consumer sentiment for both upper- and lower-income groups plunged last autumn. Currently, both groups appear more concerned than they had been before mid-2005, but the lower third appears especially nervous. Our baseline projection for the household sector incorporates these developments. Nevertheless, the greater stress evident among the most financially vulnerable segment of the household sector presents a downside risk to the forecast. David will now continue our presentation.