Earlier today we noted that while the headline Services ISM number came slightly better than expected, if still damn ugly, it is the Employment index which stuck out, coming at an almost 2 year low and which, as the chart below demonstrates has an uncanny correlation with the NFP number. In fact, based on the two series' 5 Year rolling correlation of 0.89, the September NFP is expected to print at just about ~0, unless the establishment survey has somehow joined the Chicago PMI in decoupling from the rest of the US economy. But that's only half of it. As BNY's Nicholas Colas reminds us, a far more important and fundamental driver is the trend in monthly tax receipt withholdings, which actually indicate not correlation (which never implies causation), but true causation: i.e., if less tax withheld, then less people employed - simple. To wit: "If employment is improving on a monthly basis, it should show up the Treasury data pretty quickly. New hires – and existing employees, for that matter – usually receive their compensation in the form of a paycheck. The monies withheld for items like Federal and state taxes as well as Social Security go directly to Treasury from a payroll processing company or employer. There are always adjustments to be made as you analyze the data, of course, as withholding tables are a favorite political tool to juice the economy when things are slow." Unfortunately, the data is far from pretty, and in this case causation does imply correlation.
But speaking of simple correlation, here first is just NFPs compared to the Employment Index in the Services ISM. As can be seen, NFP would be expected to come in not only below 60,000 which is the consensus (and Goldman at 50,000), but negative.
But, far more importantly, the actual causation of tax withholdings as proxy for actual jobs. From BNY ConvergeEx:
We’ve attached several charts to highlight the most recent trends and what they say about the both the upcoming Friday Jobs Report and the current U.S. employment picture. Here are a few summary points on the topic:
- September’s total receipts (adjusted for changes in withholding tables) were up 5.9% from last year. That’s slower than the August data, which was up 9.5%. When you only look at deposits to Treasury for paycheck-type withholding, the results are largely the same. Taxes and withholding from this source is up 5.9% as well, as lower than the 8.8% increase year on year in August.
- The three month rolling average for increases in tax and withholding receipts is certainly moving lower, and peaked in the Winter/Spring of 2011. This corresponds very well to the lower jobs growth numbers we’ve seen in recent months, punctuated by the August zero reading.
- The summary message here is that we shouldn’t expect very much from the Friday Jobs Report, the current +60K headline estimate notwithstanding. Labor market conditions do not appear to have strengthened much in September. The incremental tax receipts we are seeing point to employers either hiring more part-time labor or paying select employees an incrementally better salary. Real labor market growth is likely still on the back burner.
So when NFP comes in far worse than expected in order to make the case for the Goldman wet dream of $2 trillion in LSAPs, please do not be surprised.