There were two data releases this morning, the employment report and the housing starts report.
I will touch on each report below and highlight what I see to be the key trends based on the metrics that we study each month.
The economy added 20,000 payrolls in February, far below the expected 175,000 as the unemployment rate fell by two ticks to 3.8% from 4.0% last month.
The participation rate held flat at 63.2%.
Average hourly earnings "AHE" were reported to have increased 3.4% year over year, but as we have discussed in past reports, this is the most misleading indicator in economics. AHE is a ratio of weekly earnings / weekly hours and not an actual earnings measure. If hours worked falls faster than weekly earnings, then AHE goes up due to the fractional math, but the economy may not be any better off. If an employee has less aggregate income but a higher rate of average hourly earnings, that is still a net negative to consumption. Total income growth is what matters, not the growth rate in average hourly earnings.
These headline metrics are not what we focus on but rather the rate of change in various categories of employment and hours worked.
Total employment growth fell to 1.69% year over year from 1.91% last month.
Looking at a longer-term chart of total employment growth shows the peak of the cycle occurring in 2015 and the recent upward trend occurring within a longer-term decline over the past four years.
The following two-panel charts show the three-year trend on the left and the long-term picture on the right.
The growth rate in production workers for the construction sector declined sharply to 2.64% year over year.
While still near the peak, the growth rate in production workers for the manufacturing sector decelerated to 2.01% year over year.
After rising strongly last month, the average workweek length in the construction sector plunged to 39.0.
The average workweek length is the main area that we are focusing on as this is the next step in the economic puzzle that is the business cycle. We have seen a measurable decline in the three early indicators of the business cycle including housing, auto sales, and durable goods or "big-ticket" consumption. The next area that the cycle typically moves through is broader consumption as well as the leading indicators of employment which include the average workweek length.
Confirmation of a move to the ease-off phase in the economy comes from notable declines in various leading employment indicators.
While the report is noisy and subject to a large standard error, the decline in the average workweek is a negative trend to monitor in the context of the three leading sectors already showing declines.
If we multiply the number of production workers times the average workweek length, we can gather the aggregate number of hours worked in the construction sector.
Aggregate hours in the construction sector are up 0.58% compared to the same month last year which represents a significant slowdown. It will take a few more reports to confirm this trend.
The average workweek length in the manufacturing sector is a common leading indicator and used in many of the common "leading indicator" composites that investors follow.
The average workweek in the manufacturing sector, for production workers, has declined since April of 2018. A decline in the average workweek in the manufacturing sector is underway which is not terribly surprising given the weakness in global manufacturing.
This is a negative sign, however, that the slowdown has leaked into the United States.
The aggregate number of hours worked in the manufacturing sector are up 1.05% year over year, down from a growth rate of 2.98% in April of 2018. This is a notable slowdown in growth that is currently being reflected in bond yields but not yet in the stock market.
The cumulative number of hours worked by all production workers in the economy is up 1.27% compared to the same month last year, one of the lowest growth rates in the past five years.
The employment report is noisy and has a large standard error. The trends this month, however, are not inconsistent with the way a typical economic cycle develops.
We have a slowdown in housing, auto sales, and durable goods consumption. As economist Lacy Hunt points out, the next place the slowdown will show up is in initial jobless claims and the average workweek length. Once the slowdown leaks into the employment market, the economic cycle is at great risk.
We will need a few more reports to confirm this trend due to the volatility in the NFP report, but this is consistent with the view that global and domestic growth will continue to slow rather than the commonly held belief that growth will bottom.
The same leading indicators that allowed us to forecast this slowdown in early 2018 are still pointing lower and have not turned up yet. The slowdown continues.
The housing starts report showed a mixed picture, not unlike the last reporting period.
As we know, the month over month reporting of data is silly at best, and we must look at the longer term trend or rate of change in the growth rate of the data. Is the data getting better or worse, not is it "good" or "bad."
The number of housing starts popped sharply in January to 1.230 million units on a seasonally adjusted annualized basis. Housing starts are down 7.80% from the peak in 2018.
Building permits are a better measure than housing starts as housing starts are subject to weather conditions while building permits are just approvals and not impacted by weather conditions.
Building permits increased in January to 1.345 million units. Building permits are down 2.32% from the peak in 2018.
In year over year growth rate terms, building permits declined 1.54%, decelerating compared to last month's 0.45% gain. The trend in the growth rate of building permits remains lower as the housing market stays soft.
Within the data, there is a mixed picture. Multi-family building permits continue to rise in the short-term, although they have been flat for four years.
While the number of building permits for multi-family homes has been rising in recent months, carrying the headline number higher, the number of single-family building permits continues to decline.
The number of building permits for single-family homes is down 6.67% year over year, the sharpest rate of decline since 2011.
The two reports today, while subject to revisions, are consistent with the economic outlook that growth continues to slow. There will be month to month volatility in all data but the trends across the leading indicators, and now some of the coincident indicators point towards a deceleration in domestic economic growth.
A recession is not in the data yet, so this should not be construed as alarm bells, but as we know, asset prices respond to the rate of change in economic growth, not the nominal level of growth. Recession or no recession is not ultimately the most important question. The important factor is the trending direction of growth and inflation, and for now, growth is slowing, and the data suggests growth will continue to decline.
Bond yields are reflecting this economic reality and in time, stocks will as well.