United States
Review and Preview
June 01, 2010

By Ted Wieseman | New York

Treasuries posted significant long end-led losses over the past week and a notable rise in inflation breakevens in the TIPS market that reversed a portion of the major bull-flattening, deflationary rally of the prior week on some optimism that the situation in Europe, or at least the spillover impacts into US markets, might be stabilizing, though confidence was quickly rattled again Friday afternoon by a downgrade of Spain's debt.  Most important for sentiment prior to that was a significant improvement mid-week in dollar interbank funding conditions, with spot Libor stabilizing after Tuesday and forward Libor/OIS spreads coming down substantially after blowing out to new wides early in the week. Because of the severity of the additional pressures early in the week, the net improvement for the week as a whole wasn't too impressive, but things at least saw some significant movement in the right direction off the wides, and this had broad knock on effects - directly on swap spreads and indirectly on sentiment across a number of areas including equity and credit markets, which at times had been increasingly watching this area as an important gauge of systemic spillovers from the European debt crisis. A few things helped trigger the improvement. First, actual market conditions got bad enough Monday and Tuesday that the terms at the ECB's dollar liquidity injections - OIS plus 100bp plus a collateral haircut plus an additional FX haircut - went from looking very expensive to much less so and thus moved towards becoming a more effective backstop. Second, a secondary recent background domestic concern in interbank markets from financial regulatory reform was put further into the background after S&P said no bank ratings changes were being contemplated any time soon as a result of ongoing legislation. And, third, the increasingly unusual pricing of not all that stressed spot Libor/OIS but a major widening on a forward basis to September followed by not much additional widening after that was reconsidered and repriced in a more reasonable-looking way, with the risk of a much more muted upward move now seen into September and then a bit more upward risk seen into December. Sentiment about the European situation clearly remains fragile, however, as seen in how quickly markets started to swing back in the other direction Friday following Fitch's downgrade of Spain, and there hasn't really been much significant change in a tough underlying fundamental outlook for the eurozone fiscal situation, so it's not clear we've seen the worst of the market fallout.  Having at least contained for now the worst of the spillover from the European stresses into US markets, however, we're increasingly comfortable that this will not significantly impact what had been an increasingly positive US economic outlook before the European situation started badly deteriorating in mid-April. Indeed, with the fallout from the European crisis also causing a move to near-record low mortgage rates and a seasonally very unusual recent decline in gasoline prices, which should lead to a decline in seasonally adjusted CPI in May and a good rise in real income, it's not even clear that to this point the impact on US economic activity has been materially negative. 

On the week, benchmark Treasury yields rose 1-15bp and 2s-30s steepened 14bp, which followed a 4-25bp rally the prior week in which 2s-30s flattened 21bp. The old 2-year yield rose 1bp to 0.75%, 3-year 6bp to 1.24%, old 5-year 7bp to 2.06%, old 7-year 9bp to 2.74%, 10-year 10bp to 3.29% and 30-year 15bp to 4.21%. On top of the partial bear-steepening sell-off, there was some reversal of the prior collapse in inflation expectations in the TIPS market. Interestingly, though, the short end of the TIPS market continued to perform terribly even as front month oil prices surged 5% on the week, with the biggest move in inflation breakevens instead at the long end for some reason. The 5-year TIPS yield fell 2bp to 0.35%, 10-year rose 4bp to 1.28% and 30-year rose 4bp to 1.81%. This moved the benchmark 10-year inflation breakeven up 6bp on the week and 14bp from the low since October to 2.02% on Tuesday, which was a 38bp decline for the month as the 10-year nominal yield fell 37bp and the 10-year TIPS yield rose 1bp. Notwithstanding how much commodity prices have fallen as the European debt crisis has intensified while China tightening fears have also been rising, more of the decline in the 10-year breakeven has actually been in the 5-year/5-year forward portion that the Fed has been known to look at more closely than in the 5-year portion. Interest rate volatility ended up not much changed on the week, and the MBS market performed about in line with Treasuries, which moved current coupon yields up from their lows of the year just above 4% to not much higher at only around 4.1%. Average 30-year conventional mortgage rates fell to a near-record low 4.78% the past week in Freddie Mac's national survey and should be able to hold near that level if the current level of MBS yields hold into the coming week.  Home sales are apparently seeing a big pullback in May (which will be evident in new home sales before existing based on the way they are tallied) judging from mortgage application numbers after extremely strong results in March and April ahead of the homebuyers' tax credit expiration, but with housing affordability at record highs with support from the recent plunge in mortgage rates and consumer incomes and confidence improving, sales should have solid underlying support in the second half. 

Pressure on interbank markets finally started to ease in the middle of the past week. This is the key area we have been watching to judge broader liquidity spillovers from the European debt crisis, so this was certainly an encouraging development, though the reaction to the Fitch downgrade of Spain made clear how sensitive markets remain to headline risk. 3-month Libor rose 4bp on the week to 0.536%, but there was no upside after Tuesday and actually a marginal dip Friday for the first time in a couple weeks. The spot 3-month Libor/OIS spread rose 3bp to 30bp, but forward spreads declined, especially off of their early week peaks. From 0.54% Friday, eurodollar futures now price 3-month Libor to rise to 0.60% in mid-June, 0.845% in mid-September, 0.995% in mid-December and 1.125% in mid-March. Only half of that move reflects Fed expectations, as fed funds futures only show the fed funds rate moving up to near 0.50% over this period. Relative to the 30bp spot rate - up from the normal 11bp seen in mid-April - this Libor path now implies Libor/OIS spreads of 37bp in mid-June, 57bp in mid-September, 60bp in mid-December and 53bp in mid-March - a notably less strained and unusual-looking path (though certainly still not normal by any means) than the June, September, December, March forward spreads of 50bp, 71bp, 72bp and 58bp at the worst closes so far this past Monday. Easing in Libor/OIS pressures allowed front-end swap spreads to also come in off their wides, but improvement was less notable and re-widening Friday when nervousness began to rise again was significant. 

At the time of the early bond market close Friday, risk markets were holding on to minor gains for the week. The S&P 500 was up about 0.5%. Other than a 2% gain by consumer discretionary stocks, moves across major sectors were similarly quite muted. Although there was a bit of a move back up Friday as worries were rising again on the Spain downgrade, for the week equity volatility still saw a big drop, with the VIX down to 32.2 early Friday afternoon from 40.1 at the end of the prior week and the high close since 2008 of 45.8 hit May 20. Credit also posted small net gains for the week, with the investment grade CDX index tightening 2bp to 118bp and the high yield index 45bp to 630bp. The subprime ABX and commercial mortgage CMBX markets lagged badly Thursday when other markets were rallying, but they did a lot better Friday when stocks and credit were pulling back, netting out to a similarly flattish performance on the week for these areas. While the European situation broadly has remained in a state of some turmoil, peripheral versus core debt spreads have been quite stable ever since the huge Euroland/IMF support package was announced a few weeks ago, and the US muni bond credit protection market has apparently flattened out in sympathy with this. The MCDX index has been pretty close to 170bp for a couple of weeks now, ending Friday at about 167bp, towards the upper end of the range seen this year but still below the wides seen in February near 180bp. 

The past week's economic calendar was not too consequential ahead of the more important run of key figures in the coming week, and focus remained on the impact of the European debt crisis. 1Q GDP growth was unexpectedly revised down to +3.0% from +3.2%, as the upward revisions to equipment investment and retail sales implied by the monthly figures were more than offset by BEA's incorporating other source data that resulted in sizable downward adjustments to software investment and consumer spending on utilities. The results didn't have any immediate impact on our outlook for 2Q and after incorporating the underlying details released with the personal income report, we upped our 2Q GDP forecast marginally to +3.5% from +3.4%. Underlying growth should be a lot stronger in 2Q than 1Q than implied by the half-point acceleration we expect in overall GDP. We see final sales (GDP excluding inventories) growth accelerating to +3.9% from +1.4% and final domestic demand (GDP excluding inventories and net exports) growth accelerating to +3.6% from +2.0%. This would be the best gain in final sales since 4Q06 and in final domestic demand since 1Q06. We see consumption posting another solid gain of +3.2% after the +3.5% 1Q rise, business investment accelerating to +4% from +3% as equipment and software investment growth moderates but the pace of decline in structures slows, residential investment turning up sharply to +19% from -11% with the big upswing in home sales a big contributor to the volatility (from the brokers' commissions component tied to home sales), and state and local government construction finally getting on track with a long delay from last year's fiscal stimulus bill helping boost government spending to a 3% gain after the 2% drop in 1Q. With economy-wide inventory/sales ratios still at record lows at the end of 1Q even after the big boosts from inventories to GDP growth in 4Q09 (+3.8pp) and 1Q (+1.7pp), inventory restocking should be an ongoing support to growth over the next year, but we expect inventories to be a temporary modest drag in 2Q, largely as a result of auto sector inventories flattening out after a big restocking move over the past year. 

 

On the capex outlook, the past week's durables report continued to show positive underlying results. Durable goods orders surged 2.9% in April but only because of a tripling in the volatile civilian aircraft category. Non-defense capital goods ex aircraft orders, the key core gauge, fell 2.4%, but this followed an upwardly revised 6.5% surge in March and left core orders up a record 21% over the past year. There has been a repeated pattern over the past couple of years of extreme volatility around quarter-end in machinery orders that continued in March (+10.5%) and April (-5.9%), but a very strong underlying trend has emerged during the past year's rebound off the spring 2009 lows. Non-defense capital goods ex aircraft shipments ticked up 0.2% in April, pointing to some near-term moderation in capital spending growth in 2Q after the big gains in 1Q and 4Q09.  Note, though, that even after the big recent gains, equipment and software spending was near a 40-year low as a share of GDP in 1Q and would need to grow about 25% over the next year just to return to a more average level. 

The holiday-shortened upcoming week has a very busy economic data schedule containing the key initial run of numbers for May - manufacturing ISM Monday, motor vehicle sales Wednesday, chain store sales and non-manufacturing ISM Thursday, and the employment report Friday. The next round of supply will be coming up quickly, with the 3-year, 10-year reopening and 30-year reopening that will be auctioned June 8, 9 and 10 scheduled to be announced on Thursday. We look for the 3-year to be cut another $1 billion on top of the $2 billion cut last month to $37 billion, and we expect the 10-year reopening to be cut $1 billion to $20 billion to match the size of the new issue reduction. The 30-year will likely be unchanged, continuing to extend the average maturity of issuance a bit. Other data due out next week include construction spending Tuesday and revised productivity and factory orders Thursday:

* The results from all the regional surveys point to a moderation in the pace of manufacturing growth in May, so we look for about a 3-point dip in the ISM to 57.5  from the six-year high hit in April. The orders category is expected to show significant slippage, reflecting concerns related to the potential spillover of events in Europe. Also, the regional results suggest that the inventory component will move lower. Finally, the recent slide in commodity prices - especially within the energy complex - should lead to a sharp pullback in the price gauge.

* We look for a 0.5% rise in April construction spending. The recent rise in housing starts points to another gain in the residential sector. Also, infrastructure spending appears to finally be kicking in at the state and local level. So, we look for upside in the residential and public components to more than offset a further slide in non-residential activity.

* Industry reports suggest that May motor vehicle sales will register a slight uptick relative to the 11.2 million unit pace seen in April, and we forecast a gain to 11.4 million.  However, much of the anticipated gain appears to be attributable to a pick-up in fleet activity. Fleet sales plummeted last year when production was slashed and automakers shifted their focus from volumes to margins. Thus, there is some pent-up demand from rental car companies that is now being filled. We continue to look for about 12.0 million sales in 2010.

* We expect 1Q productivity growth to be revised down to +3.2% and unit labor costs up to -1.3%. The GDP data pointed to a slight downward revision to 1Q productivity relative to the originally reported reading of +3.6%. This implies a corresponding upward revision to unit labor costs (since compensation growth was unrevised) from the original -1.6%. Meanwhile, compensation was revised quite a bit lower in 4Q (from +0.9% to -1.2%), which should push unit labor costs all the way down to a new all-time record low of -8.0% (versus the original reading of -5.6%). 

* We forecast a 0.9% rise in April factory orders. The durables data showed a sharp jump in order volume that was largely tied to the volatile aircraft category. So, even assuming some energy price-related slippage in the non-durables category, we expect to see a gain in overall factory orders. Meanwhile, shipments should also post a modest gain, and inventories are expected to be +0.5%.

* We look for a 525,000 gain in May non-farm payrolls - +420,000 temporary census workers and +105,000 ex census - and a 0.2pp drop in the unemployment rate to 9.7%. A surge in Census-related hiring should bolster the headline payroll gain, but we look for a bit of moderation in the pace of private sector job growth in May as the impact of special factors that provided a boost in April fades away. In particular, we believe that the April job tally received some modest support from a relatively late survey period combined with unusually mild weather across much of the country. Also, while we would caution against reading too much into the recent rise in initial unemployment claims, federal government withheld tax collections appear to have flattened out in recent weeks. So, we expect to see a slight downshift in private sector job gains this month. The temporary elevation in census hiring should contribute to a dip in the unemployment rate for May. We calculate that under the extreme assumption that all Census workers are job finders (i.e., individuals previously counted as unemployed), the impact on the unemployment rate would be about -0.50 percentage points. Under the more realistic assumption that about one-third of the workers are job finders, one-third are new entrants and one-third are multiple job holders, the impact on the unemployment rate should be about -0.15 percentage points. Finally, note that the average workweek and hourly earnings figures exclude the public sector and thus are unaffected by the influx of Census workers.

United States
Why Are Jobless Claims so High?
June 01, 2010

By David Greenlaw | New York

Although initial jobless claims registered a modest pullback in the latest week, the level of filings remains quite elevated in relation to other indicators of labor market activity. Claims have become disconnected from payroll employment data. The same is true of the relationship between claims and a wide range of other labor market indicators, such as the household survey measure of employment, the BLS business employment dynamics data, and the Challenger layoff survey.

Increase in ineligible filers may play a role. We suspect that the root cause of the divergence between claims and other labor market data is the dramatic extension in the duration of benefit payments - up to 99 weeks in some states. While benefits have been extended beyond the standard 26 weeks many times in the past, there has never been an extension anywhere near as long as at present. The increase in the duration of benefits implies an increase in the present value of unemployment insurance, which raises the incentive to file.

The initial claims data are a tally of applications for benefits - not approvals. Historically, about half of all claims for unemployment insurance are rejected because the applicant is deemed to be ineligible (note: we derive an estimate for the rejection rate by comparing initial claims to actual first payment of unemployment insurance). Unfortunately, the data that we use to calculate the rejection rate are quite volatile and are only available with a lag of several months. We see some indication that the rejection rate began to rise in March and April, meaning that some of the elevation in filings seen over the past few months might merely reflect an increase in ineligible filers. However, given the noise in these data, it is too early to draw any firm conclusions.

Also, it's worth noting that even in the best of times, about one-third of all beneficiaries exhaust their 26 weeks of regular state benefits. Up until last year, the range for the so-called exhaustion rate for regular benefits was 25-44%. At present, the exhaustion rate is at a record 54%. To be sure, the exhaustion rate is high today partly because the labor market is still quite soft. However, in 1982 the unemployment rate was about one full percentage point higher than it is now, yet the exhaustion rate never got above 41%. One of the big differences between then and now is that extended benefits are available for as long as 73 weeks today versus a maximum of only 29 weeks in the aftermath of the 1982 recession. So, the recent rise in the exhaustion rate might represent another example of behavior being influenced by benefit extensions.

Construction workers tend to file more frequently. Another potential source of distortion in the relationship between claims and other labor market data is related to conditions in the construction sector. At present, construction workers account for about 13.5% of those receiving unemployment insurance. But construction workers account for only about 4.5% of overall employment (note: there are more workers involved in commercial construction than in the residential sector). No other industry group has anywhere near as large a divergence between current employment levels and benefit recipients. We suspect that construction workers may enter and exit the unemployment insurance system with a greater frequency than workers in other industries (in other words, they tend to experience more periods of short-term employment and unemployment). This churning could lead to higher levels of initial jobless claims than in periods when the relationship between the economy and the construction industry is more normal.

Another interesting aspect of the claims data is the rise in the relative share of filings in the state of California. Starting in late 2009, California experienced some processing problems which led to backlogs and some temporary spikes in filings. But even sorting through this unusual weekly volatility, claims have been relatively high in California for the past several months. This may be related to our construction worker thesis. Or it's also possible that state budgetary cutbacks have contributed to a jump in fraudulent filings.

Potential Census effect. Finally, we suspect that another special factor may contribute to some ongoing elevation in claims ahead. Specifically, there is a strong possibility that some of the more than 600,000 temporary Census workers will attempt to file for unemployment benefits when they leave their positions (just do an internet search for ‘census' and ‘unemployment benefits' and you will see that this is a hot topic on some bulletin boards frequented by census workers). Eligibility requirements vary from state to state, and most Census workers are unlikely to qualify. But, as noted above, the claims data count applications for benefits - not approvals. Since Census Bureau data show that the number of claims workers may have peaked in early May (at nearly 600,000), this effect may already be starting to show up in the claims figures.

United States
The 'Strong' Dollar Doesn't Threaten the Outlook
June 01, 2010

By Richard Berner | New York

Concerns overdone. In the wake of the European sovereign credit crisis, the dollar has jumped significantly against the euro; it has risen by 9% in little more than a month, and 18% since December 1. That surge has triggered worries that a stronger dollar will undermine US exports, will prove deflationary and will depress corporate earnings.  Indeed, markets have vented those qualms in sharp declines in risky asset prices, in a flight-to-quality bid for Treasuries, and in a plunge in inflation breakevens measured in the TIPS market. Unless the dollar retreats significantly, many fear it will raise the odds of much slower growth.

We think those fears are overblown. First, while the euro has weakened a great deal versus the dollar, the dollar has strengthened only modestly on a broad, trade-weighted index basis (the TWI). The Fed's broad index has risen by about 4% since mid-April and by 6% since December 1. Second, growth is far more important than currency movements in driving trade, prices and profits, and we see only modest risks to global growth. Finally, much of the dollar's recent strength reflects a flight-to-quality bid symptomatic of - and idiosyncratic to - the European sovereign debt crisis. While our FX team expects the EUR to decline to 1.16 against the dollar this year, it expects the dollar to weaken again versus AXJ and other currencies, leaving the broad dollar TWI little changed for the balance of the year and into 2011. 

Exchange rate pass-through has declined. The influence of currency movements on the macro outlook is best analyzed in terms of trade-weighted indexes that cover trading partners. Regardless of the metric used, the ‘pass-through' of nominal exchange rate changes to import prices and thus to trade and inflation has declined over the past two decades. Several Fed economists in a 2006 paper found that the decline is a global phenomenon. Across G7 currencies, their work suggests that exchange rate pass-though to import prices has declined to 40% over the 1990-2005 period from 70% in the 1970s and 1980s, and in the US, it has declined to only 30%. That is, a 10% sustained depreciation in the dollar might, other things equal, boost the import price level by 3% over a 2-3 year time period. 

Declining exchange-rate ‘pass-through' may reflect good monetary policies and the globalization of markets and production that has promoted ‘pricing to market'. That is, exporters not wanting to surrender market share have absorbed currency changes in their profit margins or hedged their currency risk. Pass-through also varies with the cyclical state of the global economy and inflation expectations; as the economy strengthens, pass-through may increase. 

The ‘stronger' dollar won't undermine the trade outlook. In turn, the influence of import prices on US trade and inflation is less important than economic growth at home and abroad and other factors influencing inflation expectations and slack in the economy. In an earlier note, we estimated that the elasticity of real exports with respect to non-US GDP alone is about 2 - implying that, other things equal, 5% growth abroad will yield 10% growth in US exports. In contrast, we estimated the elasticity of exports with respect to the real effective exchange rate to be one-third as large. 

Moreover, we think strong growth in overseas domestic demand will boost US net exports and add materially to growth in US output in 2010 and 2011. Domestic demand in the fast-growing economies is driving their growth, and US exporters will be increasingly leveraged to that rapidly growing pie. Indeed, the share of US exports going to Asia ex Japan, Latin America and Canada has risen significantly in the last decade, thanks in part to these countries' faster growth. Despite the global recession, the share of US goods exports going to those destinations rose from 69% of the total to 73% over the past five years, while the share of US services exports going to those economies also rose 4 percentage points to 46% over that period. 

To be sure, weakness in European and Japanese growth will restrain overall overseas demand, but those areas are less important than in the past: The share in US exports of goods delivered to Europe and Japan has been steadily shrinking; recently to 27%. Moreover, we have long expected sluggish growth in both Europe and Japan, and if anything, growth in each at the dawn of the sovereign crisis has proved more robust than expected. 

Dollar not a deflation threat; inflation still poised to bottom soon. The recent collapse in inflation breakevens signals growing concerns that a stronger dollar in concert with the sovereign debt crisis would promote deflation. There is broad agreement that a stronger dollar can contribute to lower inflation, both through its direct influence on import and commodity prices and via its indirect impact on inflation expectations. The interplay among them is important: While a stronger dollar and falling commodity prices will primarily change relative prices, like those of imports and energy goods and services, they can also influence both inflation expectations and inflation itself.

In any case, we continue to think the fundamentals for pricing power are gradually improving, even though retail core inflation is edging lower, courtesy mostly of the deceleration in US apartment and owner equivalent rents. Slack in housing, goods and labor markets is peaking or has peaked, companies continue to cut capacity, demand continues to improve, and the Fed remains committed to fighting disinflation. The combination of those factors has lifted operating rates 500bp from their lows, and more is clearly coming. As a result, the data at early stages of the processing pipeline support the story of eventually higher, not lower, inflation. Core intermediate wholesale prices rose at an 8.3% annual rate over the past six months. And in spite of the dollar's recent move, strong global growth is pushing up import quotes for consumers and businesses.  Non-petroleum import quotes rose by 3.3% in the year ended in April, and import prices for consumer goods excluding autos rose by 0.4%. 

Dollar doesn't threaten near-term earnings outlook. With one-third of corporate earnings coming from overseas, the dollar's influence on corporate profits has never been more important. There is no mistaking the channels: A weak euro will cut profits earned in Europe when translated back into dollars. Moreover, just as the dollar's long decline boosted the earnings and cash flow of US affiliates abroad by making them more competitive in the markets they serve, a stronger dollar will erode that edge. 

But several factors reduce the importance of a stronger dollar versus the euro on earnings.  First, earnings from European affiliates of US companies are less important than in the past. According to our strategy colleague Jason Todd, such results in 1995 comprised 14% of offshore earnings, but at the end of 2008, the share had fallen to only 8%. Second, limited exchange rate pass-through suggests that the change in end-market pricing in Europe will be significantly less than the extent of the exchange rate change.

In addition, just as with exports, currency translation effects are less important than the influence of growth abroad on overseas earnings. We continue to expect strong global growth of about 4.7% this year. US domestic earnings fundamentals, moreover, are getting stronger as the push from global growth has spurred the jobs and income growth that is essential for a sustainable recovery. Finally, rising operating leverage continues to boost margins as expansion allows companies to spread fixed costs across a wider revenue base. Margins have risen by 100bp over the last year, reflecting significant corporate capacity reductions that have enabled companies to exploit that operating leverage. Industrial capacity declined by a record 1.3% over the past year, and the boost that gave to operating rates helped sustain margins.

The bigger challenge to earnings will come in 2011, when we expect overall growth to slow and operating leverage to fade. In part, our forecast of slower growth reflects the assumptions that policymakers will begin the exit from accommodative monetary policies and that European fiscal austerity takes effect. In addition, it reflects the effects on growth of the increase in long-term yields we expect, as private credit demands clash with still-massive Treasury issuance.