July 26th, 2012

In the first half of 2012, despite two or three anomalous results, the average week-to-week change in initial jobless claims was exactly zero. The first three weeks of the second half of 2012, however, have been marked by a decline of 24k, a rise of 36k and a fall of 35k. The Department of Labor specifically commented this week that, due to the recent volatility owing to the mid-week timing of the July 4 holiday and modified plans among automobile manufacturers to shut down for summer retooling, the 4-week moving average is the better gauge to monitor. The 4-week moving average for seasonally adjusted initial claims fell 8,750 in the July 21 week. This marked the fifth consecutively week with a decline and at 375,500 was the lowest level since the March 31 week. Dealers recognize that special factors continue to distort weekly US jobless claims data, but that the underlying trend appears to be favorable to the labor market. Accordingly, the dealer community at large expects July job growth to closely resemble that in June.

Barclays Economic Research suggested interpreting the week-to-week changes in the headline claims data “with a pinch of salt for the time being.” Already believing the four-week moving average is a better indicator of conditions in the labor market, BarCap wrote, “By that measure we suggest there is some reason for optimism in this report. The four-week moving average has now fallen for five consecutive weeks after peaking at a recent high of 388k in the week of June 23. For the July employment report, BarCap looks for headline nonfarm payrolls to increase by 100k and for private payrolls to rise by 110k; they expect the unemployment rate to hold steady at 8.2%. “The claims data during the survey week suggested that the picture on layoffs had changed very little since last month, although the data appear marginally better than in the June survey week. Altogether we see an employment picture that has changed little, but the gradual decline in the four-week moving average for initial claims is a positive development for future payroll reports should the trend continue in the coming weeks.”

The Trading Desk Strategy team at RBS Securities Economics would prefer to wait “for another few weeks” to get an improved sense of the underlying tendency in the number of new filings. They wrote, “That being said, it seems unlikely that the trend has changed dramatically from the 380,000 to 385,000 level seen prior to July. In fact, we would not be surprised to see a bounce in next week’s reading to around 380,000. In any case, we are looking for nonfarm payrolls to rise by 80,000 in July.” While last week’s data coincided with the reference week for the July BLS Establishment survey, Jefferies and Co. cautioned that the seasonal adjustment distortions this month make it difficult to use claims to forecast payrolls. The JEF Economics team submitted, “Be that as it may, the 4-week moving average is 11.5k lower than it was during the June survey. So, although their predictive value this month is limited, the comparisons with last month will argue for a solid increase in payrolls and probably a decline in the unemployment rate. We expect July NFP to rise 80K with private payrolls up 90K and the unemployment rate steady at 8.2%.”

Naroff Economic Advisors warned, “When you consider that the vehicle sector changeovers to “new models” did not occur in the same orderly pattern as they did in the past, you can see how the claims numbers can be very glitchy. Don’t be surprised if new claims jump next week. Smoothing out the ups and downs, the trend is clearly down. It seems we have stabilized from the spring and early summer surge at a level that points to job growth a lot better than the 80,000 we saw in June. We should be adding something closer to 150,000 workers each month and that may start appearing when the July data are released at the end of next week.” Indeed, Naroff calls for July job growth of 157k, making them the high in the Reuters economist survey and putting them more than two standard deviations above the current survey median (i.e., consensus) of 100k.

BNP Paribas Securities reiterated the BLS analyst’s caveat to the press on Wednesday that the “massive jump” in unadjusted claims in the July 14 employment survey week, aside from auto retooling shutdowns, was fed mainly by an unexpected increase of 26,244 claims reported by California and +11,948 by North Carolina. BNP wrote, “Claims have been particularly volatile this summer and could be masking the true underlying trend as of late. While initial claims seem to be moving in the right direction, we believe that July was not the end of the sluggish readings on payrolls in the absence of a solid pick-up in hiring. We look for nonfarm payrolls to remain subdued at 75k in July and for the unemployment rate to edge up to 8.3%.”

Economists at MFR, Inc. broke it down this way. “This week, the seasonal adjustment process factored in a decline in unadjusted claims about equal to the cumulative increases that had been reported in the preceding two weeks. In the event, the unadjusted decline was considerably sharper than that, hence the reported 35K drop in seasonally adjusted claims. All of this is statistical noise, and what will be more important than these short-term gyrations is where claims settle down after the distortions end. We suspect that the data will point to a soggy labor market.”

Economists at J.P. Morgan North America Economic Research characterized the 20k drop in the four-week average over the past four weeks as “potentially a good sign for the labor market, though a very tentative one.” They anticipate the volatility in the claims data could continue until August reports start rolling in. As BMO Capital Markets Economics sees it, “Many pieces of the July payrolls puzzle will not be released until next week (i.e., ADP, the ISMs and the Conference Board survey), but suffice it to say that any improvement from the prior month’s +80k is expected to be modest, consistent with a still-sluggish U.S. labor market.”

First Trust Economics wrote, “Plugging recent data into forecasting models suggests nonfarm payrolls expanded 80,000 in July, while private payrolls grew 95,000.” FT advised that these forecasts are subject to change next week “as we get more claims data as well as private jobs reports from Intuit and ADP.” Intuit will release its Small Business Employment Index for July just after midnight on Monday, July 30. ADP will publish its National Employment report on Wednesday August 1 at 08:15.

Deutsche Bank Global Markets Research is also “inclined to put greater faith in the [4-week] moving average but said, “We are not rushing to interpolate the claims numbers into a July payroll forecast, and we are maintaining our forecast of +75k. To be sure, if the four‐week moving average can maintain its current level through August, then we will become more optimistic regarding August payrolls relative to July.” We are currently projecting the unemployment rate to remain unchanged at 8.2% in July, but if continuing claims fall further it could be an important indication that the unemployment rate is poised to resume its downtrend.”

Dealers would never rely solely on any weekly claims print or even the four-week moving average, but the data usually do provide dependable anecdotes for the jobs report. Getting the seasonal factors right for July is always tough, however, and this July was particularly challenging. What’s encouraging is that dealers have not materially lowered their NFP forecasts based on a series of downside surprises in recent jobs reports or from the uncertainty connected to the claims data.