Implications for multi-asset investors
What emerges from our empirical exploration of macro and market volatility is a better understanding of the current tension between the structural factors pushing vol down and the cyclical, late-cycle factors pushing upward. Given that this economic expansion in the U.S. may, in fact, end up being the longest in postwar history, we expect this tension to continue featuring prominently in investors’ asset allocation discussions. From here to the end of the current macro and market cycles, our results suggest that vol will move higher from the unusually low levels observed in 2017, but also that it will ultimately revert to a lower trend level than in prior cycles.
In contemplating the influence on near-term asset allocation of vol edging higher (assuming that the move occurs in the absence of recession), it is important to recognize that the influence will be felt more in position sizing and less in directional views on risk assets. A state of persistently higher volatility can still be consistent with a pro-risk stance and a preference for equities vs. bonds in a portfolio. Indeed, that would be the default allocation amid low recession risk and with a global growth outlook that continues to be supportive. However, higher volatility implies that smaller position sizes would be required to meet volatility targets in portfolios, and hence the absolute size of deviations from a given strategic benchmark allocation would get smaller as vol rose.
Even as vol picks up, the cross-asset covariance of returns would not necessarily rise.
A related point is that even as vol picks up, the cross-asset covariance of returns would not necessarily rise. It depends on the nature of the shocks that are sparking the market volatility in the first place. In the bout of vol in February of this year, the root cause of the disturbance was the rapid ascent of bond yields. Amid the market volatility, the correlation between equity and bond returns swung temporarily from negative to positive as both asset classes sold off, and the covariance of stocks and bonds over the course of the month was roughly zero. As markets continue to balance growth concerns (which drive negative stock-bond correlation) with inflation and monetary policy uncertainty (which drive positive correlation), we expect cross-asset covariance to make a somewhat smaller contribution to portfolio volatility, on average, through the remainder of the cycle.
In an environment of higher expected portfolio vol and reduced cross-asset diversification, how can risk be tactically managed? What are appropriate hedging strategies for higher vol? In addition to the mechanical effect on position sizing mentioned above, we also note that higher vol environments shift the focus of tactical asset allocation incrementally away from taking directional bets—on equities, duration, credit and so on—toward relative value opportunities within asset classes.
Options are another way to address tail risk in portfolios, to the extent that the cost-benefit trade-off makes sense. In general, we advocate sticking with the asset allocation that matches our central expectation of outcomes and then overlaying considerations of sizing, relative value bets and more explicit hedges for tail risks.
From a longer-term perspective, a decline in vol is an inherently good thing insofar as it is driven by smoother economic variation over time. The decline in economic volatility that follows from advances in consumers’ ability to smooth consumption and businesses’ ability to manage inventories is a welcome development. These structural economic trends are persistent in nature and likely imply smoother and longer business cycles in the future. Our study of the Treasury term premium suggests that declining bond yields have been exacerbated by falling inflation volatility (a positive development in that low inflation volatility reduces economic uncertainty in people’s lives). And our finding that equity vol is linked to financial system leverage suggests that the deleveraging at the end of the leverage cycle of the ’00s may underpin a sustained period of lower trend market vol. All in all, these results imply that long-term volatility forecasts may be overestimating the risk of future volatility based on historical experience.