(+) The ISM manufacturing index for September was a bit stronger than anticipated, rising from 0.5 points from August, coming in at 56.2 versus an expected 55.0.  This is the highest point for the ISM since April 2011.  The details weren’t quite as dynamic as the headline, with declines in new orders but better numbers from employment, production and inventories.

(-) Conversely, the ISM non-manufacturing survey was weaker than expected, falling from 58.6 in August to 54.4 for September, and fell short of an expected 57.0 reading.  Business activity and employment both fell, as did new orders slightly, and there were no improving areas.  This pushes the gauge back down towards the middle of its typical range of the past several years, but remains well above 50, which is an expansionary signal.

(-) The final September Markit PMI report was a bit weaker than expected, falling from 53.1 in August to 52.8 relative to an expected unchanged 53.1 figure.  New orders and employment were down for the month, while output rose a bit.

(+) The Chicago PMI survey for September was slightly better than expected, at 55.7 vs. an expected 54.0—this was an increase of 2.7 points from August, so a decent reading and along the lines of what the ISM told us.  Production and new orders rose, while employment fell—again, similar to the national data.  Employment has continually been a sticking point in these surveys and appears to be having a hard time gaining significant and consistent traction.

(-) Unit motor vehicle sales for September came in at 15.2 mil. units, which fell short of the 15.6 expected.  The domestic component also fell short, at 11.7 mil. compared to the anticipated 12.1 mil.—domestic drops were the primary group affecting the total.  Some of this was apparently due to an early Labor Day, but other seasonal factors came into account.

(+) Initial jobless claims for the September 28 week were essentially unchanged, at 308k, which outperformed expectations of 315k, and, more importantly, claim reporting from states with computer changeovers appears to be back to normal.  This release brought the closely-watched four-week moving average down (which of course includes those strange weeks of sporadic data, which have to be mentally discounted).  However, everyone likes round numbers and when claims get below 300k, it generally means good things are happening—like better economic growth.  Continuing claims for the September 21 week came in at 2,925k, which was a bit higher than the 2,805k expected, but also included some claims from California that were logged during system problems so missed the initial claims reporting completely.  Too much detail perhaps, but these little things add up.

(-) The ADP employment report that serves as an appetizer to the big government release later in the week showed job gains of +166k, which underwhelmed the expectations of +180k.  August jobs gain was revised down by -17k, so ended up being more similar to the government report of the month.  In the details, job growth declined in a few different areas, including professional/business services, but improved in trade/transportation/utilities as well as construction slightly.  Again, based on past experience, this can be interesting data, but doesn’t always fit well with the government BLS figures.

Sadly, those were our only employment pieces of the week.  Due to shutdown, the government’s Bureau of Labor Statistics claimed to not have enough manpower to get these figures cranked out in time for the usual Friday morning employment situation release.  If this is anything like the Jan. 1996 situation, these may come out in a few weeks, depending on the status of things in Washington.  The claims numbers will continue to come out, however, as this is a separate division (currently running on a skeleton crew of 30 out of over 1,000 normal employees handling the stats).