Despite the Labor Day-shortened week, several important releases kicked off the fall market season (traditionally the week where everyone on Wall Street gets ‘back to business’).

(+) The ISM manufacturing index for August came in better than expected for August, rising from 55.4 to 55.7 (while expectations called for a 54.0 reading) and bringing the index to its highest level since June 2011.  In the underlying detail, new orders gained while employment and production both declined (however, all three remained over 50—a positive indicator in a diffusion index).

(+) Non-manufacturing ISM also came in strong for August, rising from July’s reading of 56.0 to 58.6, versus a median forecasted level of a slightly worse 55.0.  In fact, this is the strongest reading for the non-manufacturing index since 2005.  Under the hood of the report, most items were solidly stronger, with new orders, overall business activity and employment gaining several points to above or near the 60 reading.

Certainly, the higher manufacturing and non-manufacturing ISM numbers represent a positive from the standpoint of near-term economic growth—therefore, could serve to act as a factor in the Fed’s decision later this month regarding the ‘taper’ question.  However, other factors play a role as well, including progress in employment, which is discussed in more detail below.

(+) Construction spending rose +0.6% for the month of July, compared to an expected +0.4% gain, and prior months’ spending was also revised upward.  By segment, private residential construction rose an identical +0.6% (up +17% year-over-year), private non-residential rose +1.3% (+2% year-over-year), while public spending fell again, by -0.3% (-4% for the twelve-month period, unsurprising with sequester cuts).

(0) Factory orders fell -2.4% in July, but were slightly better than the expected drop of -3.4%.  Both core capital goods orders and shipments were revised downward somewhat for the previous months as well, but manufacturing inventories rose a few tenths of a percent.

(+) Motor vehicle sales for August rose more than expected to a new post-crisis high point.  Total sales, at a seasonally-adjusted 16.0 million units surpassed the 15.8 mil. forecast, as did domestic sales, at 12.4 million versus an anticipated 12.3 million.  Interestingly, when looking at truck sales through a sample of the popular Ford F-Series (new trucks being somewhat indicative of construction and other industrial needs), monthly sales represented the best August since 2007 and the year-to-date figures point to the best such period since ’06.  Full-size pickup are now 12% of the industry, versus 11% last year.  Those levels still didn’t represent peaks, however.  Perhaps part of the driving force here—pardon the pun—is that auto fleets are averaging a record 11.4 years in age, which is quite long relative to history.

(-) The trade deficit for July widened as expected, from June’s -$34.5 billion to -$39.1 billion (slightly worse than the forecast -$38.6 bill.)  Non-petroleum factors explained most of the change, as exports fell six-tenths of a percent, notably from an 8% decline in consumer goods, and overall imports rose 2%.

(+/0) The Fed Beige Book released this week showed continual economic activity expanding at a ‘modest to moderate pace’ during early July to late August period—so very little new here.  On the positive side, it appeared that the bulk of Fed districts reported better consumer spending, a modest expansion in manufacturing activity, while hiring gains were little changed—all on par with recent trends.  In terms of industry detail, auto production in the Chicago district was especially strong, driven by pent-up demand and attractive financing terms; while several districts noted weaker defense-related activity on account of the sequester.  Housing activity increased at a moderately but at a more tempered pace than the previous book in July, in line with other flattening data as of late, and lending activity weakened, perhaps due to the impact of higher interest rates.  Employment improved modestly, again in line with other data and businesses seem more willing to prefer full-time to part-time employment (a subtle change but one worth noting).

(+) Non-farm productivity growth for the 2^nd quarter was revised up to +2.3% compared to an initial +0.9% and forecasted update to +1.6%.  The bulk of the change was the result of a stronger increase in business output (a percent better than the original +2.6% estimated) and a downward revision to hours worked by a few tenths of a percent.  Unit labor costs (compensation / output, per hour) was flat during the quarter, despite calls for a gain of almost a percent—this is consistent with negligible inflation levels.

(+) Initial jobless claims for the August 31 week fell to 323k, which bested a median forecast 330k, bringing the four-week moving average to a new post-recession low of 329k.  Continuing claims for the August 24 week came in at 2,951k—lower than the 2,985k expected.  It’s sometimes difficult to get a sense of these by simply looking at the numbers week to week but the charts below provide a picture of the improvement that’s taken place since the peaks in claims in the financial crisis period.  Things are clearly better on that front, even if employment overall is not where policymakers would quite like it to be.

(-) The ADP private employment report showed a gain of +176k jobs for August, which somewhat underwhelmed the median forecast of 184k.  By industry, the largest gains originated from professional/business services (+50k) and trade/transports/utilities (+40k), while manufacturing and construction added a few thousand each (slightly better than last month overall).  As we’ve mentioned previously, the ADP report hasn’t shown consistent correlation to the government job numbers released a day later; but, at the same time, the data is a bit more ‘real’ so to speak, with no birth/death modeling or other statistical adjustments that can cause government modeled figures to diverge from reality at times.  However, in this month’s case, it was quite close.

As it is, government numbers are provided with a room for error of approximately +/- 100k jobs—a wide band of error by anyone’s estimation.

(-) The big government employment situation report for August showed mixed results, but were generally a disappointment.  Payrolls rose 169k, which was a bit of a disappointment versus the expected 180k, and the two previous months’ figures were revised downward by a total of 74k.  The slowdown in job growth originated from the private service sector, including business services and leisure/hospitality.  Construction employment was unchanged.  Growth elements included manufacturing, which added 14k, and government jobs gained by 17k (state/local were the engine—federal jobs were unchanged, which was a positive in itself somewhat).

The unemployment rate ticked downward a tenth to 7.3% (an unchanged 7.4% was expected), but this was again explained by a decline in labor force participation.  Household survey employment declined by 115k, although the ‘payroll-consistent’ measure saw a rise of +286k jobs.  The labor force participation rate may be a factor in the Fed’s ongoing assessment of the 6.5% threshold as an appropriate target (they take it lower, such as 6%, etc.).

The ‘long-term unemployed,’ defined as those out of work for at least six months, remained at a 4.3 million level, representing about 38% of the total unemployed pool.  This number has declined by 733k over the past year.  Additionally, the number of part-time workers ‘involuntarily’ in that status (would prefer full-time work) fell by 334k in August.  The number of discouraged workers remained around the same level as a year ago, at 866k.  While these are small changes relative to the size of the overall economy and labor pool, unemployment across all segments is slowly improving.  Those in the most difficult situation, as you might expect, are younger workers, as well as those without technical training or college education—but this is typically the case.

Average hourly earnings rose +0.2% for the month, which was on par with forecast, and the average workweek length was also stable at 34.5 hours.  This was alluded to in the Fed’s Beige Book rather than the employment report, but health care benefit costs have been pressuring non-wage compensation costs a bit higher as of late, due to more expensive premiums.  Time will tell the effect this has as Obamacare takes hold further in the next several years.