Building A Leading Indicator Of Employment

|
Includes: AGZ, BBSA, BIL, BKT, CLTL, DFVL, DFVS, DLBL-OLD, DLBS, DMRL, DTUL, DTUS, DTYL, DTYS, EDV, EGF, EPS, FIBR, FLAT, GBIL, GNMA, GOVT, GSY, HYDD, IEF, IEI, IPE, ITE, IVV, LDSF, LTPZ, MBB, MBG, OPER, PBTP, PLW, PPLC, PST, RINF, RISE, RSP, RVRS, RYARX, SCHO, SCHP, SCHR, SDS, SFLA, SH, SHV, SHY, SPDN, SPLX, SPTL, SPTS, SPUU, SPXE, SPXL, SPXN, SPXS, SPXT, SPXU, SPXV, SPY, SSO, STIP, STPP, STPZ, TAPR, TBF, TBT, TBX, TDTF, TDTT, TFLO, TIP, TIPX, TIPZ, TLH, TLT, TMF, TMV, TTT, TUZ, TYBS, TYD, TYNS, TYO, UBT, UPRO, USFR, USMC, UST, USTB, VFINX, VGIT, VGLT, VGSH, VMBS, VOO, VRIG, VTIP, VUSTX, ZROZ
by: Eric Basmajian
Summary

A useful tool in the context of a basket of composite cycle indicators.

From an economic cycle standpoint, the indicators within the composite logically lead the business cycle.

What is the index saying now?

Building A Leading Indicator Of Employment

Geoffrey H. Moore was famously called "the father of leading indicators" for his composite index approach to forecasting business cycles.

In this research note, I want to walk through an example of a leading indicator that Moore described for the employment market in a 1983 working paper.

The composite index approach to analyzing economic data is critically important in gaining information about the risk of an economic inflection point. Many economic forecasts work well during the linear phase of the economy or when the business cycle is trending, either higher or lower. Most economic forecasts fail to identify the inflection points consistently or at least note the elevated risk of an economic inflection point.

When creating a composite index, the benefit lies in the ability to objectively identify when a basket of indicators that are correlated in terms of their sequence in the economic cycle collectively turn higher or lower.

The construction of a leading indicator starts with an understanding of the sequence of the economic cycle and finding indicators that should turn before the broader economy.

We will use Moore's leading employment indicator as an example.

In this leading indicator of employment, Moore uses the average manufacturing workweek, average overtime manufacturing hours, initial jobless claims and the ratio of voluntary to involuntary part-time labor (the part-time ratio).

These four indicators make a good leading indicator basket for several reasons. First, the data comes from three different data sources (two economic reports) so there is diversity in the reporting. Secondly, all of these indicators logically and over time have empirically been shown to lead the business cycle. Tracking the number of hours worked leads actual hiring trends because it is easier to increase or decrease employee hours based on demand compared to hiring or firing employees with every change in output. Initial claims are a widely understood leading indicator and tracking the cyclical movements in part-time employment can provide foresight into future changes in economic activity.

When speaking about the part-time ratio in his 1983 paper, Moore said:

The cyclical movements in this ratio are attributable primarily to the denominator, which reflects employers' decisions to shorten work hours in response to current or anticipated adverse business conditions. It behaves as a leading indicator at peaks and is roughly coincident at troughs. It is based on data from the Current Population Survey of households and hence is statistically independent of the other series in the index which are based on the Bureau of Labor Statistics establishment survey (average workweek and overtime hours) or unemployment insurance records (initial claims). Also, it covers all sectors of the economy, not just manufacturing.

-Moore

When we take this group of indicators that come from different sources, measure different things, and all generally lead the business cycle, we can have increased conviction that risk of an inflection point is elevated compared to looking at many indicators individually.

There are no indicators that are perfect and even composite indicators will provide false signals over time, but the probability of false signals is lower and the accuracy in assessing the level of risk in a turn of the economic cycle increases as diversified indicators are all moving in the same direction on average.

As mentioned above, as outlined by Moore, we will take a composite look at the risk of turning points in the employment market by aggregating the average manufacturing workweek, the average weekly manufacturing overtime hours, initial jobless claims and the part-time ratio. Each component is adjusted for the standard deviation of the monthly change.

The result is an index, graphed below, that leads broad nonfarm payrolls. The magnitude of the decline in the index is highly relevant. There are small declines that coincide with pockets of economic weakness but the large cyclical turning points were all clearly captured.

Leading Employment Index:

Source: YCharts, EPB Macro Research

Now, this employment index must be used in conjunction with other composite indices which we use at EPB Macro Research to further weed out false signals about major economic inflection points.

Another interesting note in the chart above is the lack of movement in this composite index during the 2015-2016 economic slowdown. This index declined in rate of change terms but never showed a material decline which would have suggested the risk of a loss of jobs was elevated. In the 2015-2016 slowdown, the employment market never contracted and no recession occurred, properly identified by the leading employment index.

The leading employment index also troughs before the bottom in non-farm payrolls.

If we look at a longer-term chart of this index and nonfarm payrolls, you can clearly see how all four recessionary periods of job loss were identified early by the leading employment index.

Leading Employment Index:

Source: YCharts, EPB Macro Research

There are several periods in which the leading employment index declined in which no recession occurred but those periods, such as 1994, were periods of sharp economic deceleration where the risk of an inflection point was, in fact, elevated. If the risk of a turning point is high, but the cycle doesn't turn, that does not mean that risk did not exist. The magnitude of the move in the leading employment index is equally as important as the direction.

If we look at a close-up chart of the leading employment index in year over year percentage change terms, we can clearly see how the leading employment index properly identified the three cyclical moves in growth we have experienced this economic cycle. The index also has not gone negative this entire cycle and there has not been any job loss.

Leading Employment Index Year over Year (%):

Source: YCharts, EPB Macro Research

The leading employment index did dip into negative territory last month but I watch a smoothed average or the trend in this index rather than each value as the month to month results can be volatile. It is curious, however, that this index has been declining for several months and we just had a nearly flat jobs report of just 20,000.

If the six-period moving average of this index moves into negative territory and continues to decline, I would then suggest the risk of a turn in the employment market is very high. Until then, the bias is to the downside as economic growth has been decelerating but it is premature to say that declines in the nonfarm payrolls number are likely.

Remember, these are not random data points but a basket of employment indicators that logically move before broader employment trends. If the confluence of these indicators moves in the same direction, there is certainly no guarantee, but the probability of an economic inflection point goes up dramatically.

We are trying to use a data-driven approach to economic cycles and to always be prepared for economic inflection points. These inflection points have the largest implications for changes in asset prices and alpha lies in the ability to spot economic cycle trends with accuracy before the consensus arrives at the same conclusion.

When the employment report comes out every month, this is why we look strongly at the hours worked component across various sectors. We are looking for the sequence of economic data to be as informative in the direction of the cycle as the magnitude of the change.

In the 1980s, when Moore wrote this paper, the average workweek was reduced prior to major job losses. The same sequence holds true today.

While false positives are always possible, when there are several distinct indicators that are all moving in one direction on average, the risk of a cycle turn is elevated.

As far as the current status of the leading employment indicator; the recent weakness in payrolls growth is curious while more data is needed to make a strong call on this economic slowdown leaking from the manufacturing sector into the employment sector.

Currently, the economic slowdown is very clearly in the manufacturing sector, the housing market is stabilizing, the auto market is suffering along with manufacturing and services remain strong. If these indicators deteriorate further, and the employment market starts to reverse, that would increase the risk of a major economic inflection point compared to the current environment of just slowing economic growth.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.