Is Risk Really as Low as You Think?—Revisited
Our latest paper revisits a topic we wrote about earlier this year – with all the political and economic uncertainty around the world, why is volatility remaining so stubbornly low? In the paper, we start with the notion that investors don’t know what to do, so they are not doing anything. We expected to see that trading volume had fallen substantially since, say, the US election. What we found is that 1) volume (as measured by the sum of the average daily volume of stocks in the FTSE Global Developed index) has risen this year and 2) volume as a percent of total market capitalization has been falling since at least 2010. Both observations led us to conclude that investors are not just sitting on their hands waiting for the storm to pass.
We next set out to determine how dangerous the current low level of risk might be – in other words, does it portend a market meltdown, because low levels of risk don’t last forever? Using some risk-based studies we have performed in the past we found that three signals have tended to precede market tops, none of which we see currently. 1) Short-horizon risk has tended to rise substantially in the months before the peak, but in the current environment short-horizon risk has continued to fall. 2) style factors became more correlated in advance of a bear market, but they have remained relatively low this year, and 3) alpha-generating style factor returns tend to increase in magnitude, and often in the opposite direction to what is “expected.” Again, style factors have remained muted this year.
Finally, we noted that risk (as measured by the short-horizon predicted volatility of the FTSE Developed index, but confirmed by other indices as well) has bumped along, at or near low levels since at least 2000. We dug into the sources of this low risk and found that the major driver was low market risk. Country and currency risk are currently above their 17-year low levels, and industry risk has climbed substantially over the past few months. So, depending on the kinds of bets managers are making, they may not be seeing the low level of total risk filtering down into their active portfolios. We also noted that the median pairwise sector correlation has come down substantially since the election (although it has popped more recently). And, in good news for stock pickers, specific risk as a percent of total risk has climbed to a seven-year high and pairwise asset correlations have been extremely low. This means that less of a stock’s risk is coming from factors, and more is specific to the company, and therefore theoretically the reward to good stock picking should be higher.
We concluded that managers still need to be paying attention to the underlying risks in their portfolios, as they are not all equally low. And although the risk-based market-forecasting signals are not flashing red right now, risk is more likely to increase than to decrease going forward.
Please click here for the paper.