Forex Investment and Currency Trading

The October employment report was disappointing, on balance, with a 190,000 drop in nonfarm payrolls and a surge in the unemployment rate to 10.2%. The establishment survey (which measures payrolls) painted a slightly better picture of the labor market than the household survey (which measures the unemployment rate). The 190,000 decline in nonfarm payrolls was larger than market consensus (-175,000) expectations, but this was countered by net upward revisions of 91,000 to the prior two months. In addition, there were a few encouraging signs among the components, most notably a 34,000 gain in temporary employment and unchanged government payrolls (which had been falling recently). However, this was more than offset by bigger-than-expected declines of 61,000 in manufacturing payrolls, 62,000 in construction jobs and 40,000 in retail sector employment.

There were two other interesting takeaways from the establishment survey. First, the average work week held at a record low of 33.0, which was a disappointment given consensus expectations for an increase of 0.1. This contributed to a 0.2% drop in aggregate hours worked, leaving this measure down 3.1% on a 3m/3m annualized basis. This suggests that the gap between output and hours worked persisted in October, implying continued strong productivity growth. The manufacturing work week edged up to 40.0 from 39.9, and overtime increased slightly, suggesting that the upturn in production is starting to boost manufacturing hours worked and, in time, should translate to hiring. Second, there may have been some seasonal adjustment issues. Going into year end, there is greater-than-usual uncertainty about seasonal adjustment, following the dramatic swing in payrolls at the end of 2008. In October, NSA payrolls rose by 641k - ie, the seasonal adjustment process knocked off 831k jobs, significantly more than usual for the month of October (the average for the past 10 years is -600k). This may suggest that the improving trend in SA payrolls has been understated to some extent if this seasonal component is overstated. For comparison, if we adjust October NSA payrolls with last year’s seasonal factors, nonfarm payrolls would have declined only 83,000. However, we suggest caution when reading too much into the swings in the seasonal factors, as they often create volatility in the data.

The real dismal news came from the household survey with a 0.4pp surge in the unemployment rate to 10.2% (unrounded: 10.196%). This reflects a sharp drop of 589,000 in household employment, which more than offset a small decline of 31,000 in the civilian labor force. Over the past three months, the household measure of employment has declined by an average of 400,000 more than the payroll measure each month. While this is a decidedly large gap, it is not unusual to see such big discrepancies between the two measures; for job growth estimates, we put more weight on the less-volatile payroll measure. The measures of underutilization in the household survey were also weak as the broadest measure of unemployment - which captures marginally attached workers and part-time for economic reasons - jumped 0.5pp to 17.5%. Moreover, the average duration of unemployment increased to 26.9 weeks from 26.2, and the median duration jumped to 18.7 weeks from 17.3, both reaching new record highs.

On balance, this was a disappointing report. However, it does not change the outlook for the economy as market continues to believe that strong cyclical forces should boost growth, and with a lag, the labor market should recover.