The perpetually wrong ADP report for August has come, printing at 91K, on expectations of 100K, and down from a downward revised 109K (previously 114K). How much of a leading indicator to Friday's NFP this is is anyone's guess: historically the ADP's error factor has been between -100% and +100%. Also as a reminder, the report does not account for the Verizon employee strike which impacted initial claims last week and which according to Goldman will likely be an addition downward wildcard in Friday's NFP.
According to the report "Employment in the service-providing sector rose by 80,000 in August, marking 20 consecutive months of employment gains. Employment in the goods-producing sector rose by 11,000 in August, up from a loss of 2,000 jobs last month. Manufacturing employment slipped 4,000 in August. Employment on small payrolls (up to 49 workers) rose 58,000 in August, almost equal to that of last month. Employment on medium payrolls (50 to 499 workers) rose 30,000, while employment on large payrolls (500 or more workers) created only 3,000 jobs." More importantly, employment in the construction industry rose 7,000 in August, following three consecutive monthly declines, bringing the total decrease in construction employment since its peak in January 2007 to 2,128,000. Employment in the financial services sector remained essentially unchanged in August. The total decrease in financial services employment for the same period remains 687,000. This is particularly amusing considering that everyone knows by now that Wall Street plans to shed a massive amount of people. We are confident that despite the miss this report will be dead wrong as usual when predicting the actual even more manipulated NFP. But futures are up on the news which is all that matters.
Breakdown of private payrolls data: