U.S. employment data released on Friday January 5 sparked large moves in fixed-income markets, first sending yields higher and then sharply lower as traders reacted to data from the Bureau of Labor Statistics (BLS) and the Institute for Supply Management (ISM). Here is a guide for interest rate traders as they look towards upcoming employment reports, which highlights how markets can overemphasize some pieces of data initially and then readjust to other pieces of data that come subsequently.
Bond yields initially jumped sharply as traders focused on three items:
- The Non-Farm Payroll (NFP) headline number +216K, which was 41K stronger than consensus.
- The unemployment rate, which stayed at 3.7% rather than rise to 3.8% as expected.
- Average hourly earnings, which rose by 4.1%, 0.2% more than expected.
However, each of these supposedly positive surprises came with caveats:
- The two previous NFP numbers were revised downward by 71K, meaning that the NFP jobs total was 30K below consensus rather than 41K above.
- Unemployment stayed at 3.7% for the worst possible reason: labor force participation plunged by 0.3% to its lowest level since February. The household survey, from which the unemployment rate is calculated, showed 753K in job losses in December, and that 676K people left the labor force.
- While average hourly earnings surprised on the high side, the average number of hours worked shrank by 0.3%, negating the income effect of the upside surprise in wages.
An hour and a half after the BLS data, ISM dropped a bombshell. The employment component of the ISM Services Index plunged to 43.3. Any reading below 50 suggests that a plurality of hiring managers at U.S. businesses are considering laying off more workers than they are hiring. 43.3 is a low reading consistent with the pace of layoffs seen during the 2001 recession, which sent unemployment from 3.8% to 6.5%. The employment component of ISM Services has been a solid leading indicator of future changes in employment since it debuted in 1997 (Figure 1).
Figure 1: December ISM shows services businesses contemplating net layoffs for early 2024
Since nearly 92% of Americans in the non-farm sector work in services and only 8.25% work in manufacturing, the ISM Services data is especially important. That said, it’s worth pointing out the employment component of the ISM Manufacturing Index has been below 50 for three consecutive months, suggesting net employment losses could be coming in factories as well.
When it comes to the BLS data, the best approach for both economists and traders is to look at the data holistically rather than piecemeal. Each month the BLS produces data from two separate surveys:
- The Current Employment Statistics survey (CES, also known as the establishment survey) of approximately 120,000 businesses. This survey produces three closely watched groups of numbers: non-farm payrolls and their monthly change as well as revisions to previous months’ data, average hourly earnings, and average number of hours worked.
- The Current Population Survey (CPS, also called the Household Survey) is a survey of 60,000 households from which the unemployment rate and the labor force participation rate are calculated.
Economists and traders often view the three components of the establishment survey in isolation, but they are best observed chained together into a single series: total labor income. Part of the confusion is that average hourly earnings are more commonly presented as a year-on-year percentage change whereas the NFP is typically viewed as a month-to-month change in job numbers, while average hours worked is reported only as a monthly level. If one converts the numbers into the same units, year-on-year (or even month-to-month) percentage changes, the data become much easier to interpret. The total income flowing to the labor force from work (as opposed to entitlement programs or intergenerational transfers) is equal to the total number of people working multiplied by their average hourly earnings times the average number of hours worked. Taken together this is what that data looks like (Figure 2):
Figure 2: Total labor income is a holistic way of looking at the establishment survey data.
As of late 2023, total labor income has been growing at a pace of 5.5% year on year. This is about 1% faster than its average pace between 2011 and 2019, and probably explains in part why inflation hasn’t returned to normal levels.
Unlike the establishment survey, the BLS household survey is not subject to monthly revisions and it may be better at picking up on turning points. Broadly, the establishment and household survey total jobs numbers show similar levels of overall employment, but the two series can diverge and have done so recently. Ever since the pandemic began, the household survey, which does a better job of measuring job creation at small and mid-sized businesses, has shown less employment growth than the NFP. December 2023 was no exception with +216K jobs measured by the establishment survey and -753K by the household survey (Figure 3).
Figure 3: The household survey employment growth has fallen behind that measured in NFP
Interest Rates data
Get comprehensive views of the U.S. dollar-denominated interest rate markets across the entire yield curve with our robust datasets.
All examples in this report are hypothetical interpretations of situations and are used for explanation purposes only. The views in this report reflect solely those of the author and not necessarily those of CME Group or its affiliated institutions. This report and the information herein should not be considered investment advice or the results of actual market experience.
