Business Cycle Update - Market Analysis for Oct 24th, 2019
At EPB Macro Research, we often focus on the growth rate cycle or the fluctuations in growth that occur within a business cycle expansion. In this note, we will stay focused on the business cycle and analyze the major coincident data points of the US economy. As a reminder, asset prices typically do not respond to coincident data as it is too lagging for forward-looking markets.
As always, the rate of change or direction of growth is important when analyzing coincident data but when we look at aggregate business cycle indicators, the nominal level in relation to history is also essential in determining the level of recession risk in the economy.
At the end of this note, we will also tackle a question that has been emerging due to the developing theme in the economy: housing vs. manufacturing. Can the economy slip into recession even if housing data remains "strong?"
Let's have a look.
Coincident Economic Data
The four major coincident data points that define the US economy include a measure of production, income, consumption, and employment. These categories are outlined by the NBER in their definition of a recession.
Most commonly, industrial production, published by the Federal Reserve, is used as a coincident gauge of the industrial economy.
It is not a surprise that industrial growth has been decelerating, falling from a rate of growth near 5% to a contractionary reading last month.
The US industrial sector is experiencing decelerating rates of growth and leading data such as new orders and the price of industrial commodities are not suggesting a renewed cyclical upturn in IP growth is imminent.
US Industrial Production Growth:
Source: Bloomberg, EPB Macro Research
Real personal income growth excluding government transfer payments is a data point also cited by the NBER in their recession dating procedure.
Over the long run, consumers can only spend what they earn which makes aggregate income growth an important coincident measure.
When analyzing income growth, we see a primary peak around 2015 before a cyclical downturn and subsequent recovery. Income growth recovered to a lower peak of 4.43% growth before decelerating again to a multi-year low of 2.62%.
If the US and global economy are relying on the US consumer to spark a cyclical recovery, aggregate income growth of just 2.6% makes this a tall order.
US Real Core Income Growth:
Source: Bloomberg, EPB Macro Research
Real consumption growth, published by the Bureau of Economic analysis each month, represents nearly 70% of the entire US economy, measured by GDP. Similar to income growth, (unsurprisingly) consumption growth followed a similar path, experiencing a primary peak of 4.56% at the start of 2015 before decelerating sharply into the 2016 global slowdown.
A cyclical rebound in growth coupled with tax savings pushed consumption growth to a high of 3.75% before falling yet again to a rate of 2.27% growth over the last twelve months.
We often hear narratives about the consumer, characterizing spending growth as "strong," but this is clearly an opinion-based analysis.
A factual characterization of consumption growth is that spending growth is decelerating and in the 40th percentile of readings since this expansion began in late 2009.
US Real Consumption Growth:
Source: Bloomberg, EPB Macro Research
Employment growth holds a long-standing, multi-decade correlation with consumption growth. The growth rate in total employment mirrors the rate of change in consumption growth, charted above.
It should be noted that employment growth is actually weaker than the data graphed below. The BLS published preliminary revisions to the employment data that suggests recent employment data has been overstated.
The revised data will not be fully available until February 2020 but the expected path via Hoisington Management will be posted as a second chart below.
US Total Employment Growth:
Source: Bloomberg, Hoisington Management, BEA, BLS, EPB Macro Research
It is empirically observable that the growth rate in all four critical areas of the economy is decelerating.
When we aggregate the four data points outlined above into one coincident business cycle index, we can measure the growth rate today relative to history to better understand the level of recession risk.
Whenever all four categories are decelerating individually, and when aggregated, are at levels consistent with past recessionary periods, extreme caution is appropriate.
US 4-Factor Coincident Index Growth Rate:
Source: Bloomberg, EPB Macro Research
Furthermore, the employment data, when revised in 2020, will push the coincident index above to an even lower growth rate.
Moreover, leading indicators of industrial production and employment suggest lower growth should be expected.
With all of this data in mind, it is therefore probable to see the coincident index move to the lowest growth rate of this entire economic expansion.
This does not in any way guarantee a recession, this should be clear.
However, it does suggest that the trending growth rate in the economy is near the weakest point of this economic cycle, at a time when all four categories of the economy studied by the NBER are experiencing individual decelerations.
Short leading indicators such as the average weekly hours worked, core durable goods new orders, core capital goods new orders and the price of industrial commodities do not support an imminent and rapid acceleration in these coincident data points.
Market price action aside, it is clearly far too early to remove the risk of a recession entirely from the forecast.
Is A Recession Possible Without A Housing Slowdown?
The aggregate economy is clearly weakening to a level that should rightfully cause some concern. As outlined last week, housing is one sector of the economy that typically responds early to changes in interest rates and monetary conditions.
Housing data has been accelerating for nearly all of 2019 after a "freeze" in housing activity occurred in late 2018.
The housing sector has been a long-standing leading sector of the economy and will likely continue to be an early mover due to the interest rate sensitivity.
As we will see below, however, while housing typically leads most economic cycles, a decline in housing activity is not a necessary condition for a recession.
It is important to track all early moving sectors, including manufacturing, because each cycle is different. There is no single perfect indicator that precedes every recession but tracking the aggregate of all early moving sectors, not discounting any weakness or strength, removes head fakes and false signals.
As the chart of total home sales shows, housing transaction volume typically peaks and declines well in advance of recessionary periods. Ahead of the 2000s recession, however, a pronounced downturn in housing activity was less apparent.
Total US Single-Family Home Sales:
Source: Bloomberg, EPB Macro Research
Ahead of the 2000s recession, total home sales declined just 7% peak-to-trough.
Total US Single-Family Home Sales 1996-2002:
Source: Bloomberg, EPB Macro Research
By comparison, today, we have seen a peak-to-trough decline of 13% for total home sales before an interest rate led recovery.
Total US Single-Family Home Sales 2016-2019:
Source: Bloomberg, EPB Macro Research
Before the turn of the century, single-family building permits had a 12.5% decline before accelerating through the recession.
Total 1-Unit Building Permits 1996-2002:
Source: Bloomberg, EPB Macro Research
Today, an 11% decline in single-family permit activity was seen before a rebound in housing momentum.
Total 1-Unit Building Permits 2016-2019:
Source: Bloomberg, EPB Macro Research
Lastly, national home price growth never fell below 6% on a year over year basis through the 2000s recession.
National Home Price Index Growth Rate (%):
Source: Bloomberg, EPB Macro Research
The 4-factor coincident index, comprised of the factors outlined above, however, was decelerating.
US 4-Factor Coincident Index Growth Rate:
Source: Bloomberg, EPB Macro Research
Short leading industrial data such as durable goods new orders, capital goods new orders and industrial commodity prices all declined in advance of the 2000s recession.
This analysis should not be confused with a forecast for a US recession. Rather, we should not be sounding the all-clear sign due to an acceleration in residential housing activity.
We should be sympathetic to the housing turnaround and it should be characterized as an unambiguous positive sign for the economy.
As noted last week, however, manufacturing payrolls growth holds a far stronger association with total payrolls growth as compared to residential construction.
Year over Year Change In Manufacturing Payrolls Growth Vs. Total Payrolls Growth:
Source: Bloomberg, EPB Macro Research
Therefore, we cannot completely remove the risk of a recession until we see a sustained recovery in short-leading industrial-based data and an easing of the rate of decline in leading indicators of manufacturing payrolls growth.
