October 18, 2017, Jean-Jacques Ohana, CFA and Dr. Christian Witt (both YCAP Asset Management)
Stock market volatility typically jumps during US economic recessions (see Figure 1). Recessions in turn tend to occur when the labor market is weak (see Figure 2). Accordingly, the current investment outlook looks rosy when focusing on headline labor market data. But under the surface a more rattled picture emerges. For instance, the breadth of the labor market, a reliable recession indicator by historical standards, suggests the US labor market has considerably slowed (see Figures 3 and 4). Yet, investors seem to ignore the fact. For stock market volatility resides near all-time lows––a troubling dislocation. Will stock market volatility spike soon?
On the surface, the US labor market is well. The unemployment rate is near multi-decade lows (4.2%) and non-farm payrolls have been steadily rising for years (see Figure 2).
But by another measure, labor market breadth, the US labor market looks much more exhausted. We define labor market breadth in two ways. Gross labor market breadth (GLMB) is the percentage of US states with rising unemployment rates. Net labor market breadth (NLMB) denotes the difference between the percentage of US states with rising and falling unemployment rates (negative values are set to zero). Thus, like popular stock market breadth indicators, the two metrics estimate how widespread labor market weakness is across the United States.
Notably, both indicators depict a more tainted picture than headline numbers (see Figures 5 and 6). The GLMB has risen to 47.1% in August; just shy of the 50% threshold that frequently announces a recession is in the making. Interestingly, some of the false alarms nonetheless correctly identified imminent volatility spikes (e.g. June 1995, September 2002). Meanwhile, the NLMB has risen to 19.6% which is right in the middle between white-noise and critical levels (NLMB ? 10% and NLMB ? 30%, respectively). But again, such « no-man’s-land » readings have repeatedly preempted volatility spikes. Therefore, our two indicators suggest a rally of volatility might be in the making.
This is all the more important, as volatility currently is near all-time lows. For episodes of extreme dislocation between the GLMB and VIX (GLMB << VIX or GLMB >> VIX) have historically marked turning points in the direction of volatility. For hidden labor market weakness points towards economic vulnerability which is eventually reflected by stock markets.
But how does the indicator relate to « fear » beyond the borders of the US? To respond to this question, we compare labor market breadth to the Riskelia Risk Aversion index (see Figure 7) which tracks risk premia across global equity, sovereign bond, credit, money and commodity markets. What we find is that both indicators typically jump at the same time, if a recession is in the making. Perhaps, because lasting turmoil needs to be founded in verifiably poor fundamentals. But currently the two indicators do not line up. So, investors may prefer to wait until this is the case.
Finally, as a litmus test, we run two strategies based on the GLMB against the S&P 500. The first strategy is long S&P 500 unless the latest GLMB reading is below 50%. Then, the strategy invests in the money market (average effective FED funds rate). As shown above, a GLMB reading above 50% tends to distinguish expansion from recession (see Figure 3). The second strategy also takes Risk Aversion into account. Therefore, the strategy hedges its equity exposure only if the GLMB stands above 50% and Risk Aversion is positive, too. To avoid curve-fitting, both strategies rebalance in the middle of the month (about a week after both signals are observed).
The results generally corroborate the predictive power of the GLMB indicator (see Figure 8 and Table 1). However, the best results are obtained when the GLMB and Risk Aversion indicators are simultaneously used. Since the 1970s, the indicator helps to alleviate drawdowns while conserving long-term performance. As a result, the Sharpe increases. Both volatility and drawdowns decline. Further taking Risk Aversion into account (since January 1997) improves upon these results. The return rises while the Sharpe increases. It is the combination of both indicators––when « fear » meets economic vulnerability––that makes for the most reliable hedging signal.
How could investors respond to the current situation? One is to reduce the risk-taking of their portfolios in anticipation of supposedly further deteriorating labor market breadth. This could be achieved by taking profits (especially on short volatility strategies), purchasing protection or mitigating leverage. Another one would be to closely monitor the next state-level US jobs report which will be published on October 20, 2017 (not to be confused with the standard national BLS report) and act accordingly. If the GLMB continues to deteriorate, US stock market volatility will likely spike soon. If the opposite scenario comes to pass, any stock market weakness would be an opportunity to buy into the dip. But the first-best strategy might be to wait until both indicators will have jumped above critical levels. In the current set-up, this means Risk Aversion would need to escalade violently.
FIGURE 1: US STOCK MARKET VOLATILITY (VIX) VS. RECESSIONS
FIGURE 2: UNEMPLOYMENT RATE AND NON-FARM PAYROLLS (CHG) VS. RECESSIONS
FIGURE 3: GROSS LABOR MARKET BREADTH (% STATES WITH RISING UNEMPLOYMENT RATES) VS. RECESSIONS
FIGURE 4: NET LABOR MARKET BREADTH (% STATES WITH RISING LESS FALLING UNEMPLOYMENT RATES) VS. RECESSIONS
FIGURE 5: GROSS LABOR MARKET BREADTH (% STATES WITH RISING UNEMPLOYMENT RATES) VS. VOLATILITY (VIX)
FIGURE 6: NET LABOR MARKET BREADTH (% STATES WITH RISING LESS FALLING UNEMPLOYMENT RATES) VS. VOLATILITY (VIX)
FIGURE 7: GROSS LABOR MARKET BREADTH (% STATES WITH RISING UNEMPLOYMENT RATES) VS. RISKELIA RISK AVERSION INDEX
FIGURE 8: INVESTMENT STRATEGIES BASED ON GROSS LABOR MARKET BREADTH & RISKELIA RISK AVERSION VS. S&P 500. MONTHLY REBALANCING.
TABLE 1: SUMMARY STATISTICS OF INVESTMENT STRATEGIES