Double-Dips in the US Are Rare
During the 19th and early part of the 20th century, double-dips in the US were fairly common because economic cycles were closely tied to agricultural conditions. But it's important to keep in mind that double-dips are quite rare for an industrialized economy like the US. If we define double-dip as a recession that begins less than two years after the prior recession has ended, then the US has experienced only one double-dip since the Great Depression. That occurred in the early 1980s, and the catalyst for the dip was clear-cut: short-term interest rates rose more than 1,000bp and long-term yields jumped by more than 400bp between the end of the 1980 recession and the beginning of the 1982 recession. Obviously, there is no such easily identifiable catalyst at this point.
Most Agree That the Key to Sustaining Recovery in the US Is the Labor Market
Even with the modest uptick in employment seen over the past several months, the level of private sector jobs is below the mid-2003 trough. Yet, the economy has recovered just about all of the output that was lost during the Great Recession of 2008-09. The recent divergence between output growth and labor input reflects robust gains in productivity. History tells us that an economy is capable of registering extremely strong growth in productivity during the early stages of economic recovery. But history also suggests that this sort of performance is not sustainable indefinitely. If output is going to continue to grow, then companies are going to have to add labor. We still see signs of such a handoff from an export- and productivity-led recovery to a more mature expansion that is sustained by job and income growth. For example, layoff announcements are running at the slowest pace in a decade, temporary workers continue to be added at a rapid rate, and (despite a bit of a blip in June) the length of the workweek still appears to be gradually lengthening. Most importantly (as highlighted in our previous Growth Scare note), there has been a steady underlying improvement in federal government withheld taxes over the course of the past few months. Historically, such a development has linked up quite well with movements in job and income growth.
Fading Stimulus Fears Are Overblown
Many observers cite the withdrawal of fiscal stimulus as a potential catalyst for a 2H slowdown - or even an outright double-dip. There have been numerous fiscal stimulus initiatives enacted in recent years, but the one that justifiably attracts the most interest is the American Recovery and Reinvestment Tax Act of 2009 (ARRA) adopted in February 2009. ARRA contained three key components - tax cuts, infrastructure spending and aid to state & local governments. Original estimates claimed $787 billion of so-called ‘stimulus' over a 10-year period, while more recent official estimates put the figure at $862 billion (see CBO's latest Budget and Economic Outlook). However, much of the feared fading of stimulus is highly misleading because ARRA included items such as an extension of the alternative minimum tax (AMT) fix, which is routinely adopted every year. In fact, the AMT fix accounts for nearly $100 billion of ‘stimulus' in F2010 in the ARRA budget scorekeeping. Also, the withholding changes related to the "Making Work Pay" component of ARRA went into effect in spring 2009, resulting in a temporary boost to after-tax income. Shortly after the withholding changes were implemented, after-tax income resumed its decline, reflecting the severe job loss that was occurring at that point. More recently, spurred by the relative improvement in the labor market over the course of 2010, after-tax income has begun to grow. So, the boost to after-tax income - the goal when implementing a tax cut - is now underway.
Meanwhile, infrastructure spending has finally started to show signs of life in recent months. In fact, public expenditures on construction projects plummeted through most of 2009 and early 2010 following passage of ARRA, and it is only in the last few months that we have started to see signs of a turnaround. This is likely tied to the stimulus funding finally beginning to kick in (although it doesn't really matter whether it is tied to stimulus or not - the important new development is the rise in public construction activity).
Admittedly, the sizeable chunk of financial support for state and local governments that was included in ARRA is slated to phase out later this year. But we (and many Washington observers) suspect that there will be additional federal funding - primarily tied to Medicaid relief - in the pipeline.
There are some concerns tied to the loss of other types of fiscal stimulus. In particular, the recent plunge in new home sales in the wake of the expiration of the homebuyer tax credit appears to be the source of the latest round of stimulus-related angst. The homebuyer tax credit did pull forward some sales, and thus a payback is to be expected. But the homebuyer tax credit also helped to drive inventories of unsold homes to near-record lows, which helps to alleviate some of the stress for homebuilders.
The situation is analogous to the cash-for-clunkers program, which provided a big boost for car sales in the summer and fall of 2009. The follow-on production effects of cash-for-clunkers have lingered to today. For example, some automakers announced last month that they would forgo the plant shutdowns that typically occur around the July 4 holiday. Instead, factories remained up and running to help bolster lean inventories. Based on assembly schedules published by the automakers, we estimate that vehicle production, on a seasonally adjusted basis, will jump about 15% in July. This situation has helped to drive jobless claims much lower over the last couple of weeks, and the boost to July vehicle assemblies should also help to bolster industrial production, payroll employment and 3Q GDP.
In the case of the auto sector, reining in excess inventories with some short-term demand stimulus helped to lay the groundwork for a prolonged recovery in vehicle production. In our view, residential construction is unlikely to show the same degree of upside going forward, but the lack of an overhang of unsold new homes should - at the very least - help to limit the downside in homebuilding going forward.
Moreover, the housing sector is not a critical part of our sustained recovery story for the US. Residential investment directly accounts for less than 2.5% of GDP at this point - down from a recent peak of 6% in 2005. Also, among participants in the Blue Chip survey of forecasters, our estimate of an 11% rise in housing starts in 2010, moderating to a 5% increase in 2011, is among the most pessimistic.
Finally, we would note that there is justifiable concern regarding potential fiscal drag going forward. But that is a story for 2011 - not 2010. Indeed, as addressed in a recent note, there is probably too much emphasis on the economic spillover effects of the fiscal pressures confronting state and local governments (see State & Local Fiscal Problems: How Big a Headwind? June 25, 2010). It's the federal government that really matters. Our current baseline assumption is that Congress will extend expiring income and capital gains tax cuts for all but the highest-income individuals (i.e., households with more than $250,000 annual income). But there are a wide range of possible outcomes, and the situation may not be resolved until after the November elections. If tax rates were to rise across the board in 2011, there would certainly be some downside risk relative to our current economic forecast. Conversely, if tax rates remain at current levels for everyone, there is upside risk.
Treasuries posted sizeable gains, led by the intermediate part of the curve over the past week, just about fully reversing the bout of modest weakness seen from the release of the employment rate on July 2 through Tuesday, as economic data were weaker than expected and pointed to slower growth in 2Q GDP, risk markets were unable to sustain an early week extension of the big improvement seen in the first half of July despite a continued run of strong earnings reports, mortgage-backed securities yields plunged to more record lows, and investors moved to put money to work after the run of mid-month auctions was finished Wednesday with a strong 30-year reopening. We lowered our 2Q GDP forecast substantially to +3.2% from +4.0%, in response to the much wider-than-expected May trade deficit and the soft June retail sales report. While big gains in imports and exports in May were clearly not suggestive of weak domestic or overseas demand, incorporating the relatively bigger upside in imports, we now see net exports subtracting nearly a full percentage point from 2Q GDP, double our prior estimate. The smaller-than-expected gain in underlying retail sales in June and downward revisions to May and April led us to lower our 2Q consumption forecast to +2.3% from +2.9%. But with an expected acceleration in business capital spending in particular - which looked more robust after the biggest gain in capital goods imports in five years in May - and also residential investment and government spending, the outlook for 2Q domestic demand still looks strong even with this muted outlook for consumer spending, and we see final domestic demand (GDP excluding trade and inventories) rising over 4% in 2Q after muted gains near 1.5% over the prior two quarters when most of the GDP upside was instead led by inventories. Meanwhile, the stock market didn't react well in the second half of the week to seemingly much better-than-expected earnings results from major banks, supporting the big Treasury market gains seen starting after Tuesday, apparently disappointed with revenue trends and some softness in capital markets activity. The forward-looking economic read-through from the results looked quite positive, however, with non-performing loans and charge-offs for consumer and business loans showing substantial sequential improvement, supporting our banking analyst team's expectation coming into the quarter that large-cap bank reports would show that 1Q marked the peak in non-performing loans. A second quarter of surprisingly fast improvement in consumer credit losses was particularly encouraging, as it indicates that ongoing consumer balance sheet repair that is expected to restrain consumer spending for some time more is proceeding more quickly than expected. We think that a consumer debt service ratio of 11% to 12% (compared with 12.5% in 1Q and a peak of 14.0% in 1Q08) and a debt to income ratio of 80% to 100% (compared with 126% in 1Q and a peak of 136% in 1Q08) would represent long-term sustainable norms and that we are on pace to reach the former by the end of this year and the latter by early 2012 (see US Economics: Credit Demands: Stealth Revival, Richard Berner, July 16, 2010).
For the week, benchmark Treasury yields fell 4-15bp, with the intermediate part of the curve leading the gains. The 2-year yields fell 4bp to 0.60%, 5-year 15bp to 1.69%, 7-year 15bp to 2.36%, 10-year 11bp to 2.94%, and 30-year 9bp to 3.95%. Upside came in a run of three straight good rallies from Wednesday to Friday that totaled 6-20bp and mostly reversed a fairly modest period of weakness from July 1 (ahead of the employment report) to Tuesday of 3-23bp. At Friday's close, the 2-year, 3-year, 5-year and 7-year yield were at new lows in over a year (and an all-time low for the 2-year yield), while the 10-year and 30-year yields were a bit above the prior lows on July 1. Even with some upside in core inflation reported for June and steady oil prices on the week, TIPS relative performance was terrible, with benchmark inflation breakeven more than reversing the improvement seen after the strong 10-year TIPS auction on July 8 to move to new lows since October. The 5-year TIPS yield fell 5bp to 0.28%, 10-year yield fell 2bp to 1.24%, and the 30-year yield rose 5bp to 1.87%. This moved the benchmark 10-year inflation breakeven down 10bp on the week to 1.71% after it had moved up to 1.85% at Tuesday's close from the prior low for the year of 1.73% just ahead of the auction. This rough performance came after weakness was reported in headline PPI (-0.5%) and CPI (-0.1%) on a continued sharp pullback in energy prices (and for PPI food also). But core measures showed some acceleration. Core PPI ex motor vehicles has accelerated to +0.2% over the past four months from a prior trend of +0.1%. More important were growing indications that core CPI has bottomed as shelter costs have started to gradually turn up in line with rising rents across the country. The core CPI accelerated to +0.2% in June, keeping the year-on-year pace steady at +0.9%, with shelter - which makes up over 40% of the core and accounted for all of the deceleration in core inflation the past few years - up 0.1%. Owners' equivalent rent (+0.1%) and rent (+0.1%) both extended a recent move higher after an unprecedented run of prior declines, likely marking the trough in core inflation in 2Q. Moderating shelter costs have been responsible for all of the deceleration in core CPI seen over the past several years. Core CPI inflation slowed from +2.7% in January 2007 to +0.9% in June, with shelter inflation down from +4.3% to -0.7% over this period and the core ex shelter up to +2.1% from +1.5%.
As strong as the renewed Treasury market gains were over the past week, mortgages continued to do even better, which has brought current coupon MBS yields low enough that a refinancing wave in the large supply of high credit quality 4.5% MBS is becoming a meaningful risk. On the week, Fannie 4% MBS outperformed Treasuries by more than a quarter point, and 4.5% issues did nearly as well, which moved the 4.5% dollar price all the way up to almost 4 points over par as current coupon MBS yields plunged more than 15bp to another all-time record low below 3.55%. Average 30-year conventional mortgage rates have been at record lows a bit above 4.5% over the last few weeks, and while they have been stickier than MBS yields (widening margins significantly for originators), the further rally in recent days should, if sustained, start to move 30-year rates down towards 4.375%. At this level, refinancings in 4.5% MBS (with underlying 30-year mortgage rates around 5.25%) could start to accelerate notably further from a meaningful pick-up already seen in the June prepay figures. This large chunk of the MBS market is made up of mortgages largely written in 2009 when house prices were at their lows and credit standards at their peaks. So there should not be any issues with negative equity or tighter credit slowing refis in these issues that have been so key in sharply slowing refinancings of higher coupon mortgages to this point (though our trading desk and interest rate strategy team have become more cautious on higher coupon issues also, since if Fannie and Freddie were to ease underwriting standards and costs to allow homeowners stuck in higher rate mortgages to refinance, the huge premia these issues trade at make them vulnerable to significant losses). Interestingly, the Fed actually owns most of the 4.5% MBS supply, and its holdings of these issues account for more than a quarter of its entire portfolio of mortgages, Treasuries and agencies, so if there is a significant refi wave in these issues and the Fed sticks to its policy of not reinvesting MBS principal repayments, there could be a significant draining of excess reserves and a meaningful passive start to the Fed's exit strategy. From a near-term market perspective, since the Fed doesn't hedge its mortgage convexity risk, there is still much less need for mortgage duration hedging as rates fall to levels that threaten a 4.5% MBS refi wave. But private sector mortgage investors were still more actively moving to hedge their convexity in the latest week as yields plunged, which added to the upside and the strength in the intermediate part of the curve. This was also apparent in a good narrowing in swap spreads on top of the Treasury yield declines, as mortgage-related receiving in swaps picked up notably. On the week, the benchmark 10-year swap spread fell another 3.5bp to 2bp and the benchmark 2-year spread 6.5bp to 24.5bp. It's important to note that even as strong as rates markets continue to be, with volatility plunging, risk-adjusted carry still looks quite attractive in many cases. Normalized 3-month X 10-year swaption volatility fell another 5bp over the past week to 93bp after run-ups to only about as high as 120bp in early May and again in early June when European fears were most elevated. Friday's close was nearly all the way back to the two-year lows near 90bp hit in the spring and less than half of the highs above 200bp hit a year ago in mid-2009. Our interest rate strategy team calculates that hedged carry on mortgages has actually broken out to new cycle highs recently, and the team sees carry at the short end of the Treasury and swaps market as remaining attractive (see the July 15 report, US Interest Rate Strategist: Lowered Mortgage Convexity Needs - Implications).
A rough Friday for risk markets reversed a positive start to the week to lead to modest net losses, ending the big recovery off the recent lows that began after the July 4 holiday. The S&P 500 lost 1.2% on the week after dropping 2.9% Friday. Financials lost 4% Friday and about 2.5% on the week to lead the sell-off. Credit saw similar swings, with the investment grade CDX index widening 5bp Friday to end 3bp wider on the week at 113bp. High yield was less volatile than stocks or IG, with about a 20bp widening Friday to near 600bp resulting in a slightly larger net widening for the week. The recently badly struggling muni bond CDS market managed to hold in fairly well late in the week after an early week improvement. Late Friday the 5-year MCDX index was near 225bp, about 10bp wider than the recent low hit Tuesday but about 10bp tighter on the week and 50bp tighter than the wides for the year hit early in the month. There seems to have been an increased focus among investors recently in thinking about whether the MCDX index at 225bp really makes sense compared to corporate credit and other sovereign CDS. Short-term pressures remain intense in some cases and long-term pension and other imbalances huge for some states. But current state debt burdens are negligible compared to problems of the most stressed countries in Europe (and the US federal government too), which often seems not to have been understood in the flurry of recent press and market commentaries drawing comparisons between fiscal problems in peripheral Europe and American states with budget problems.
Economic data over the past week were mostly worse than expected, with the international trade report and retail sales reports each cutting our 2Q GDP forecast by 0.4pp and June IP and early July regional surveys showing some slowing in manufacturing sector growth, which has led the recovery over the past year.
The trade deficit widened $2 billion in May to $42.3 billion on big gains in both imports (+2.9%) and exports (+2.4%) that pointed to robust domestic and overseas demand but a sizeable drag on 2Q GDP growth from net exports. Most of the export upside reflected a surge in capital goods led by machinery, but there was also good upside in consumer goods, autos and industrial materials. Aside from a pullback in petroleum products, the jump in imports was broadly based, with the most notable gain in capital goods imports, which posted their biggest monthly rise in five years, boosting expectations for a sharp rise in 2Q business investment. Incorporating these results, we now see net exports subtracting 0.9pp from 2Q GDP growth instead of 0.4pp, but we raised our forecast for business investment in equipment and software to +21% from +18%. Meanwhile, retail sales fell 0.5% in June, largely as a result of a 2.3% drop in auto sales. Ex auto sales dipped 0.1%, and while much of the weakness was in a price-related drop at gas stations (-2.0%), the key ‘retail control' grouping (sales ex autos, gas stations and building materials) rose a less-than-expected 0.2% in June, and May (-0.2% versus +0.1%) and April (-0.4% versus -0.2%) were revised lower. There was notable weakness in the heavily weighted grocery store category (-0.5%), furniture (-1.1%), and sports, books, and music (-1.4%). General merchandise (+0.2%) and clothing store (+0.6%) gains were also smaller than seemed to be suggested by the solid chain store sales reports, though the unusually late Memorial Day made those results difficult to interpret. Based on this weakness in retail control, we lowered our 2Q consumption forecast to +2.3% from +2.9%. These two reports combined to lower our 2Q GDP forecast to +3.2% from +4.0%. Even with the downward revision to our consumption estimate, however, we see final domestic demand growth accelerating to +4.1% from only 1.8% annualized growth in the first three quarters of the recovery thanks largely to what should be another strong quarter for business investment, an upswing in residential investment, and a pick-up in government spending with help from an estimated 20% increase in state and local government construction spending.
Manufacturing activity has apparently hit a soft patch recently, but a big rise in July motor vehicle assemblies should provide a big near-term renewed boost. IP ticked up 0.1% in June, supported by a surge in utility output (+2.7%) and a modest gain in mining (+0.4%), which hasn't shown any disruption since the Gulf oil spill. Manufacturing output declined 0.4% after surging 3% over the prior three months. Motor vehicle and parts production declined 1.9% in June but should be up sharply in July as a major automaker did not take normal summer downtime. Ex motor vehicle manufacturing dipped 0.3%, with weakness in heavily weighted food and chemicals largely offset by good gains in some key durables categories, notably machinery and fabricated metals. Looking to July, softer results from the Empire State and Philly Fed manufacturing surveys led us initially to forecast a decline in the national ISM to 55.0 after the 3.5-point drop to 56.2 in June. This would be a notably better result than the ISM-comparable Philly Fed (50.0 versus 53.3) or Empire State (52.3 versus 55.8) results and certainly not a weak number in absolute terms. A key positive in July should be a big rise in motor vehicle assemblies, as one of the major producers worked through late June and early July instead of taking normal summer downtime to rebuild depleted inventories. This should provide a major boost to industrial production in July. It has also shown up in a big way in greatly improved jobless claims over the past two weeks - though the Labor Department stressed that the latest decline was more broadly based and not just in the auto sector - and should be reflected in a boost to payroll growth in the July employment report.
The economic data calendar is very light in the coming week, but a couple of key events have the potential to be significant market movers - Fed Chairman Bernanke's semiannual monetary policy testimony Wednesday and Thursday and the release of the European bank stress-test results on Friday. The economic projections that will be contained in the report Chairman Bernanke will submit to Congress and discuss in his testimony were released with the minutes from the June FOMC meeting, and revisions were minor, with small downward adjustments to the prior April forecasts for GDP growth and inflation over the next couple of years. Certainly, however, investors will be looking for any change in tone after the softer recent economic data results. As discussed in the weak University of Michigan consumer confidence report, public discontent with government economic policies has been sharply rising recently, which members of Congress surely heard from their constituents when they traveled home during the recent Congressional recess, and this might be reflected in the hearings in some discontent among Committee members with the Fed's firmly on hold policy stance. A notable aspect of the FOMC minutes release was how little discussion there seems to have been about the possibility of renewed policy easing measures. We do not expect or advocate such steps, but numerous press reports ahead of and since the June FOMC meeting suggested a significantly higher level of Fed concern about the economic outlook and a greater focus on what additional policy actions could be taken against downside risks than was reflected in the actual meeting discussions. Economic data releases due out include housing starts Tuesday and existing home sales and leading indicators Thursday:
* We forecast a small rise in June housing starts to 600,000 units annualized. During the first few months of 2010, starts trended higher (from an extremely depressed level) as the homebuyer tax credit stimulated sufficient demand to drive the inventory of unsold new homes to a multi-decade low. But demand for new residences has tailed off in the wake of the expiration of the tax break, and new construction registered a pullback in May. We look for starts to be little changed in June (+1.2%) as builders sit tight and attempt to assess the state of the market.
* We look for June existing home sales to fall to a 5.20 million unit annual rate. To qualify for the homebuyer tax credit, contracts had to be signed by April 30, and closings originally had to take place by June 30. But closings were experiencing delays as the backlog of deals swelled. So before adjourning for July 4 recess, Congress extended the closing deadline to September 30 (although contracts still had to have been in place by the original April 30 deadline). So, the impact of the homebuyer credit on the existing home sales series (which is based on closings) will linger for another few months. However, the NAR's pending home sales index plummeted 30% in May, so we look for a significant decline in June resales.
* The index of leading economic indicators is likely to fall 0.3% in June, which would be the first decline since early 2009, with significant negative contributions from the manufacturing workweek, supplier deliveries and stock prices. The yield curve and money supply should be partially offsetting positives.
Read Full Article »