The Market Rout by the Numbers
Market concerns finally came home to roost in recent days. The STOXX USA 900 index fell about 6.5% from February 24 through 25, with other markets around the world seeing similar declines.
So how did this affect total risk and other measures we use to determine managers’ ability to achieve their risk goals? Even if the market recovers, volatility is likely to remain higher than it has been for quite some time. Bottom line, it is time for managers to examine both total and active portfolio risk to determine the impact of market moves.
For starters, a chart of risk forecasts from the beginning of 2020 for several markets looks like a hockey stick—relatively flat through February 20 and then straight up (Exhibit 1). Short-horizon fundamental forecasts rose 30%-40% over the past two days for the US, Europe (STOXX Europe 600), and global markets (STOXX Global 1800). In Japan (STOXX Japan 600), risk has been rising all year, and in the past two days only jumped about 10%.
The sharp increase in risk was almost exclusively the result of higher market risk. In Exhibit 2 we show the four factor blocks for the STOXX USA 600 index. Style, Industry and Specific risk have remained relatively flat all year, save for an upward blip in specific risk at the end of January. The pattern for the market risk element, however, clearly drives the increase in total benchmark risk. The relative flatness of the style factors does not mean that no style factors experienced big changes in risk. Just since last week risk for the Market Sensitivity factor rose more than 11%, but changes for other factors were much more muted, with several, including Medium-Term Momentum, seeing risk fall. The outsized returns of some factors in certain models may also mean that style risk has and/or will also increase in the near term.
The list of industries for which risk rose the most over the past few days is not surprising. Transportation Infrastructure, Airlines and especially Hotels, Restaurants and Leisure experienced big jumps in their volatility. Most industries, however, saw only small changes.
Exhibit 1. Short-Horizon Fundamental Risk Forecasts, Selected Indices
Exhibit 2. Short-Horizon Components of Risk, STOXX USA 600 Index
Factor returns for February 24 were not particularly outsized relative to expectations, with the exception of Volatility and Market Sensitivity. Adding data for February 25 shows far more factors with returns that fall outside of a two standard deviation range, some in the desired direction and some that may have had a larger-then-expected negative impact on returns.
In the US, Volatility’s two-day return was almost three standard deviations below expectations, based on risk forecasts on February 21, and Market Sensitivity’s was more than four standard deviations. (Many managers have negative tilts on these factors, so the highly negative return should have boosted active returns. Those who make bets in the other direction may have seen sharp active performance shortfalls.) Two-day returns were also highly negative for these two factors in US Small Cap, Canada, UK, Japan and in the Worldwide model.
It is interesting that returns to the Size factor were negative in almost every region, meaning that smaller stocks fared better. That may not be too surprising, however, as investors tend to focus on more liquid and hence bigger stocks to sell when they are in panic mode. Size’s return in the Worldwide model was almost six standard deviations below expectations though!
The sell-off did not necessarily signal a big shift in leadership, however, at least in a few markets, as signified by the positive (albeit small) Momentum return. Only the UK—where the factor had been unusually positive this year—did the return look highly negative. Nor did it signal a renewed appreciation for cheaper stocks, as both Value and Earnings Yield returns were generally negative (and Value was roughly three standard deviations below average in the Worldwide and Emerging Markets models).
Overall, the highest incidence of outsized returns was in US Small Cap and the Worldwide model, although none of the models was spared any bigger-than-expected returns.
Exhibit 3: Recent Factor Returns and Deviation from Expectations
It obviously has been a tough few days for investors trying to sort out the impact of the coronavirus, which included a sharp increase in volatility and some outsized factor returns that may have led to unintended portfolio active returns. At times like these, concerns about negative returns need to be accompanied by a renewed focus on portfolio risk, which has likely changed considerably, or will shortly.