Thursday, June 17, 2010

Picture-Perfect May CPI But Initial Claims Raise Some Eyebrows

The May CPI report is near-perfect in that it confirms the picture of uneventful price trends, putting to rest any lingering concern about either inflationary, or deflationary, impulses in the current environment.

The 0.2% decline in the overall index last month was the result of a 2.9% fall in the energy component- the latter largely due to a 5.2% drop in gasoline prices. The core CPI's increase of 0.1% for the month, follows a flat reading in April, and represents only the second gain in the series since the beginning of the year. In fact, in the last 6 months to May, the core CPI has been up 0.8%, while in the last three months, it has risen by only 0.4%. Year-on-year, core inflation is up 0.9%.

Source: Bureau of Labor Statistics

In the last three months, the key housing component (which represents 40% of the overall CPI) has been flat (with a 9% drop in fuel oil prices helping offset a 15.8% spike in the "lodging away from home" category), while apparel prices have declined by 4.2% and, even medical costs have risen by a relatively moderate -by the standards of that component- 2.9%.

The impressively benign price picture is hardly surprising, given the abundant amount of slack in the economy and its slow absorption rate in in the midst of a moderate economic recovery, with these factors likely to continue taming price trends over the next 12 months or so.

Inasmuch as the favorable price dynamic should appropriately be viewed as providing the Fed with ample space to delay the onset of the tightening process, it is important to recognize that the monthly inflation data are not likely to be the primary reason that will determine the timing of the Fed's exit strategy. Instead, that is more likely to be shaped by a combination of the Fed's assessment of three factors: a) the economic recovery's prospects -particularly, in the wake of the fiscal turmoil in Europe- b) the ability of the global financial system to withstand the stress that a turnaround in the U.S. interest rate cycle would entail, and c) the degree of restlessness on the ground (i.e financial markets) about the need to see that the Fed remains vigilant vis-a-vis the longer-term inflation risks posed by the liquidity currently in the system.

The 12,000 rise in initial unemployment claims in the week of June 12 to 472,000 would not disconcerting per se, given the inherent volatility of the series, but it does validate a disappointing pattern of an essentially stalled downtrend in claims since the beginning of the year. The 4-week moving average of the series is now at 464,000, not much different compared to five months ag0.

The puzzle with the behavior of claims in recent months is that it stands in sharp contrast with the significant overall turnaround we have seen in the monthly payroll data and most other labor market measures. Although, we should not expect claims and payrolls to go hand-in-hand over the short-term, one would have thought that a nearly six-month period is long enough to have allowed the two series to send a more consistent message.

One explanation for the disconnect is that the payroll data may ultimately be revised downward for the first part of the year during the annual benchmark revisions of the series next spring. Another possible, but not fully satisfactory, explanation is that the last recession has caused profound dislocations among the various sectors in the economy, where some industries continue to shed off jobs at a strong pace (therefore accounting for the still elevated level of lay-offs), while other industries are turning around in a more robust fashion, accounting for the bulk of the hiring reflected in the improved payroll data.

Today's initial claims data were for the survey week of the June employment report and there is usually an attempt to use claims as a hint for what the monthly employment report may look like. The correlation between initial claims during the employment survey week and payrolls for that month is non-existent, but, at times, some loose relationship may exist between new filings and the unemployment rate. Still, there is no meaningful hint that can be derived from today's data, as the 472,000 claims number today was nearly identical to the 474,000 claims number for the survey week in May.

The rise by 88,000 to 4.571 million in the continuing claims for the week of June 5th is also consistent with the broader theme of lack of progress on the front of both initial filings and claims recipients.

Anthony Karydakis