Jobs, Jobs, Jobs

The US employment data is the main event before the weekend.  The Federal Reserve's QE3+ is conditioned on sustained and substantial improvement in the labor market.  That means regardless of today's report, the Fed's long term asset purchases will continue.  In fact, they are likely to continue through at least next year and may even increase early 2013 to make up for the completion of Treasury purchases under Operation Twist. 

Today's employment report may be more important from a political point of view than policy.  The consensus is that the Presidential debate has breathed fresh life into Romney's campaign and changed the momentum.  A soft jobs report may see Romney press his advantage.  That said, we think the polls in the swing states are more important than national polls.

The monthly employment report is always a bit of a wild card.  It is regarded as one of the most difficult numbers for economists to forecast.  Remember it is a net figure.  Hundreds of thousands of jobs are created and destroyed every month.  The few inputs that economists use, like weekly initial jobless claims, ISM reports and other surveys, were mixed.

ADP is not very reliable on a month-to-month basis.  Over the past 12-months the average miss has been about 55k and in the August, its initial estimate was 98k too much.  In four of the past six months, ADP's estimate has been on average 82k above the BLS initial estimate.  The two times that it was too low, the average miss was 10k.  

Work by Stone and McCarthy found that the BLS itself typically under-estimates the August report, perhaps due to faulty seasonal adjustments.  They found the BLS revises up the initial August estimate by an average of 62k over the past decade.   

The US needs to create around 170-200k net new jobs a month to drive the unemployment rate toward 7%, which is not, to be clear the target.  The Fed has adopted a formal inflation target, but no formal target for its other mandate.  Nevertheless, that is the kind of jobs growth that economists appear to have settled on. 

Canada will report employment data has well today.  Like the US, the Canadian employment data tends to be volatile.  Look for some payback after the out sized 34.3k increase in August.   Recall there was a 12.5k loss of full time jobs and a 46.7k increase in part time jobs.   The consensus is for a 10k increase in September.  In August there was a 36.4k loss of goods producing jobs.  This seems a bit much.  Similarly, in August the transportation sector lost 37k jobs this also out sized. 

The correlation between the Canadian dollar and the S&P 500 has eased from the summer's tight fit near 0.88 to a still noteworthy 0.76 over the past 60-days (on a percentage change basis) and slightly less over the past 30-days (0.74).