United States: Review and Preview
Ted Wieseman and David Greenlaw (New York)
The Treasury market saw modest long end led losses over the past week, as a significant 5- and 10-year led sell-off Friday reversed some small earlier gains. Trading was very slow all week as investors appear completely focused on the upcoming employment report to provide direction. Friday's weakness came in sympathy with a similar sized reversal in JGBs on continued strength in Japanese stocks and positive signs for the economy that partially carried over to stronger U.S. stocks. U.S. economic data released the past week were mixed -- lower capital shipments in February and downward revisions to prior months led us to cut our Q1 GDP estimate to +4.0% from +4.6%, but home sales and consumer confidence rebounded, and positive signs from various employment indicators led us to raise our estimate for March payrolls slightly to +125,000.
Meanwhile, a meaningful upward revision to core PCE prices in the fourth quarter put the annual rate of advance in the latest month back into the bottom end of the range Fed officials have indicated they are comfortable with. Although the revision was based on a change in assumptions for unmeasured phantom prices and therefore did not imply any real change in the inflation picture, it did provide further evidence of a bottoming in underlying inflation that had previously been seen in the core CPI. A bottoming in core inflation is one prerequisite for a change in Fed policy. The other is significant improvement in labor market conditions, so Friday's employment report will likely remain the overriding market focus in the coming week.
Benchmark Treasury yields rose 1 to 5 bp over the past week, and the curve steepened, with 2’s-10’s and 2’s-30’s (looking at the old 2-year) rising 4 bp on a 1 bp rise in the 2-year yield to 1.54% and 5 bp gains in the 10-year and long bond yields to 3.84% and 4.77%. The new 2-year was awarded Wednesday at 1.52%, but backed up to 1.58% at Friday’s close. After leading the sell-off Friday as some mortgage related selling emerged (with its yield up 11 bp), the 5-year yield ended up 4 bp at 2.795%. The market’s largest losses on the week were in off-the-run bonds, as much of the selling appeared to be bond futures led, possibly reflecting month and quarter end stock/bond asset reallocations as well as mortgage related selling. The February 2023 issue, which was around the cheapest to deliver at week end, saw its yield rise 7 bp to 4.68%. Following the release of the monthly PCE price measures on Friday, the futures market shifted forward the timing of the first rate hike forward slightly. At 1.145%, the October fed funds contract is now putting about 60% odds on a rate hike by the September FOMC meeting, up from 50% a week prior. The December 2004 eurodollar contract was off 3 bp on the week to 1.57%, but the December 2005 contract only weakened 0.5 bp to 2.835%
Economic data released the past week were mixed. While nondefense capital goods ex aircraft orders posted a solid 1.1% gain in February, January results were revised down significantly, and shipments fell for a second straight month. The unexpected 0.6% drop in nondefense capital goods shipments in February on top of the downwardly revised 0.3% drop in January pointed to significantly slower investment growth in Q1 than we had previously estimated. Real equipment and software investment now appears on track for a gain near 12%, compared with the 21% gain we estimated previously, leading us to cut our Q1 GDP estimate to +4.0% from +4.6%. Obviously, these data are highly volatile, so the picture could change significantly again after the release of the March report, revisions to recent data, and information on international trade in capital goods.
On the positive side, consumer sentiment and home sales rebounded (with new and existing home sales both rising significantly more than expected in February), weekly chain store sales reports indicated further strength in spending in March, and various labor market indicators continue to point to improvement in labor market conditions. The worrisome plunge in consumer sentiment seen in February and early March appears to have stopped in the latter half of the month. The University of Michigan’s gauge for all of March was revised up to 95.8 from the preliminary reading of 94.1 (compared with 94.4 in February and 103.8 in January). This was also seen in the ABC/Money weekly poll, which after plunging from -3 in the four weeks through January 18 to -22 in the period through March 14, recovered slightly to -21 in the four weeks through March 21.
Improvement in labor market conditions may be helping to offset the negative impact of surging gasoline prices on consumers’ moods. For some time now, the unemployment claims data have been pointing to a reduced pace of layoffs and a general improvement in labor market conditions. For example, initial claims dipped again in the latest report, with the four-week moving average hitting another new three-year low of 341,500. However, the payroll employment figures have been quite disappointing for the past several months. As the FOMC indicated in its latest official statement, the lack of any significant job growth to this point appears to reflect the fact that the drop in layoffs has yet to be accompanied by a corresponding improvement on the hiring side. The latest Manpower survey may be reason for optimism on this front. Net hiring intentions posted an impressive gain – to a new three-year high. Also, the Conference Board’s Help Wanted Index is showing signs of at least a mild turnaround. So we expect to see some underlying improvement in the labor market figures going forward. This week, we decided to make a slight upward adjustment to our estimate for March payrolls. We now look for Friday’s report to show a 125,000 rise in employment (versus our prior estimate of +100,000). And we continue to expect the job growth in excess of 200,000 per month to materialize before the end of the year.
[Mish note: these guys have been overly optimistic for at least 6 months]
A number of Fed officials the past week again noted that for the Fed to become less “patient” about keeping rates on hold, they would need to see significantly stronger payroll gains for several months and a reversal of disinflationary pressures, with a modest move higher in core price measures. Despite some significant flaws, the Fed has made clear its preference for the core PCE price index, and a number of officials have indicated that they would prefer to see this measure rising 1% to 2% at an annual rate, compared to the previously reported +0.8% gain in January. A major problem with the core PCE gauge was on display in the most recent releases, when the gain in Q4 was revised up to +1.2% annualized from +0.7% partly based on new data relevant to the calculation of financial services provided without charge. This gauge is supposed to account for consumption of financial services that do not have any direct cost, such as check processing, safekeeping, money transfers, etc. Prior to the latest revision, the implicit financial services price index was reported to have declined 13.5% over the 12-month period ending in January – fully accounting for the gap that had been evident in the core CPI and core PCE price measures. This change was revised to -9.3% in January followed by a further move higher to -7.6% in February.
The revision means that this gap has now been largely eliminated, with both core inflation measures running in the +1.1% to +1.2% zone, having shown slight acceleration in recent months. As a result, to the extent that the Fed focuses on the core PCE price index, the underlying inflation picture has now been fundamentally altered by a technical shift with little if any economic meaning. To the extent that the Fed still considers the core PCE valid, the annual rate of change is now three-tenths of a percent closer to the mid-point of their presumed 1% to 2% target range. To the extent, however, that they rightly discount the revisions and focus instead on the core CPI or the experimental “market based” core PCE price index (which excludes phantom price changes such as financial services provided without charge), which were both at +1.2% year/year in February, we don’t really have any idea what their preferred target range is.
Key focus in the coming week will be on Friday’s employment report. In the meantime, the Fed calendar continues to be very busy, with additional appearances in the coming week scheduled for Governors Bernanke, Kohn, and Gramlich and regional bank Presidents, Parry, Guynn, Poole, and Moskow. An interesting aspect of the remarks in the past week was further signs of a divergence of views among Fed officials about the risks to the inflation picture. Richmond Fed President Broaddus said he sees risks weighted towards further disinflation, while St. Louis Fed President Poole reiterated his view that risks were tilted towards an upside surprise. Governor Kohn took a middle path, acknowledging signs of some bottoming in inflation, but calling the evidence “inconclusive” and arguing that the Fed should err on the side of fighting disinflation rather than tightening prematurely. On the supply calendar, Treasury will announce a 5-year note and 10-year TIPS reopening on Thursday for auction the next Tuesday and Wednesday. The auction schedule was moved forward a day to avoid problems with the early close on April 8 ahead of the Good Friday holiday. We expect the 5-year size to be held at $16 billion and expect the TIPS reopening to be raised $1 billion to $10 billion in line with the $1 billion increase in the new issue size to $12 billion in January. In addition to the employment report, key data releases in the coming week include consumer confidence Tuesday, factory orders Wednesday, and ISM, construction spending, and motor vehicle sales Thursday:
* We expect the Conference Board’s consumer confidence index to fall to 86.0 in March. The University of Michigan sentiment gauge stabilized in March following a sharp pullback in February. However, the neutral reading for the month as a whole reflected some further deterioration in the early part of the month offset by improvement later on. This is the same pattern evident in the weekly ABC consumer confidence measure. We look for the Conference Board index to post a slight decline (relative to the 87.3 reading seen in February) because it is conducted by mail and thus has an earlier cut-off point than the other surveys.
We expect factory orders to rise 1.5% in February, as a sharp 2.5% gain in the durables goods component implies a solid rise in overall bookings. Even though much of this elevation in February is related to the volatile aircraft category, it appears that underlying order activity continues to trend higher with the key core category -- nondefense capital goods excluding aircraft -- now up 12% on a year/year basis.
* We forecast a March ISM reading of 61.0. The regional business conditions surveys from the Philly Fed and NY Fed showed some slippage in headline diffusion indexes, but little change on an ISM-weighted basis. Therefore, we look for ISM to post only a slight dip relative to the 61.4 reading seen in February. To get a sense of how high the recent ISM readings have been from a historical perspective, note that our estimate of 61.0 has historically been consistent with about a 6.5% pace of GDP growth.
* We look for some weather-related slippage in construction activity during February and expect overall spending to dip 0.2%. Indeed, based on the housing starts figures, it looks like the residential component of construction spending should post an outright decline for the first time since last June.
* Preliminary surveys point to a modest pickup in the sales pace during March. We look for an overall selling rate of 16.5 million units -- versus an average of 16.2 in January and February. March appeared to start off on a strong note, but we are building in some softness around mid-month to account for the effects of the winter storms that battered parts of the nation. The final tally for March will be somewhat dependent on whether automakers introduce new incentives before month end. We will be updating our estimates accordingly.
* We expect nonfarm payrolls to rise 125,000 in March. For some time now, the unemployment claims data have been pointing to a reduced pace of layoffs and a general improvement in labor market conditions. However, the payroll figures have been quite disappointing for the past several months. As the FOMC indicated in its latest official statement, the lack of any significant job growth to this point appears to reflect the fact that the drop in layoffs has yet to be accompanied by a corresponding improvement on the hiring side. The latest Manpower survey may be reason for optimism. Net hiring intentions posted an impressive gain -- to a new 3-year high. Also, the Conference Board’s Help Wanted Index is showing signs of at least a mild turnaround. So, we expect to see some modest underlying improvement in the payroll figures over the next few months. The March payroll data are also expected to show a small net gain (+10,000) related to the resolution of the California grocery worker strike. Finally, the unemployment rate is likely to hold steady in March at 5.6%, with the household survey showing a bounceback in employment following a sizeable decline in February. |
