Blog Posts — May 2, 2019

Are European Stocks Really Less Volatile? Or Are Correlations Driving Volatility Lower?

An article headlined “Risk Takers Take Early Holiday in Europe” appeared in the May 2, 2019 edition of The Wall Street Journal. In the piece, the authors noted that the European market, as defined by the Stoxx Europe 600, “set a record for apparent investor apathy, and as of Wednesday (May 1) hit the 18th consecutive trading session when its price moved by less than 0.5%.”

Since the drivers of volatility (up or down) are always of interest, we decided to look a little deeper into the causes of this apparent apathy. First, we confirmed the low volatility according to our own risk models, and using the FTSE Developed Europe index. The market has steadily ratcheted up this year, matched by a corresponding steady downward drift in forecast risk. The current level of risk for the index (in euros) is in the bottom quartile, compared with its history since at least 2001.

Exhibit 1. FTSE Developed Europe Index Risk and Return

Source: FTSE Russell, Axioma

Of course, index volatility is a function of both correlation and asset volatility. We also confirmed that correlations among assets within the European market have fallen. The median 60-day realized correlation of stocks fell into the 15th percentile relative to history as of the end of April, so quite low in the longer historical context, but also down by quite a bit this year.

Exhibit 2. Median 60-Day Rolling Correlation, Stocks in FTSE Developed Europe

Source: FTSE Russell, Axioma

The risk data we showed in Exhibit 1 is euro-based, but the UK comprises more than 25% of the FTSE Developed Europe index. Therefore, the correlation between the euro and the pound could also have an impact on volatility.  Sure enough, the realized correlation between the currency pairs has fallen substantially since last August, and is now close to its 20-year low.

Exhibit 3. Rolling 60-day Correlation, GBP vs. EUR

Source: FTSE Russell, Axioma

Finally, we simulated a full dense covariance matrix to determine the impact these correlations have on overall volatility, and how the impact of lower correlation compares with a decrease in stock volatility. The analysis shows that more than 3 percentage points of the 5 percentage point decrease in the short-horizon forecast was the result of lower correlations, with the other 2% coming from lower stock volatility.

Exhibit 4. Decomposition of the Change in Risk, Last 3 Months, FTSE Developed Europe

Note: This chart depicts the drivers of the change in risk. Of the 5% decrease over the past three months, about 3/5 was the result of lower correlations, while 2/5 could be traced to lower volatility in individual stocks. Please contact your Axioma representative for more detail on the methodology.
Source: FTSE Russell, Axioma

As the WSJ article noted, there have been numerous reasons to be concerned about the European market, including an upcoming potential trade war involving European cars, budget issues in Italy and, of course, the potential disruption from Brexit, not to mention generally lackluster economic growth. But these are not necessarily things that would normally drive a drop in market volatility. In fact, one could reasonably assume they would drive volatility up.

Given the sharp decline in asset correlations and its impact on overall volatility, along with the sharply lower currency correlation, we see this situation in a different light. Lower correlations suggest that investors believe these issues and others will affect assets and currencies quite differently. So, auto tariffs may have the biggest impact on German car makers, for example, while Brexit could potentially hurt UK companies but benefit others in the EU.

It is clearly presumptuous to assume we can get inside investors’ heads, but this simple analysis suggests that differentiation across assets is a substantial driver of overall lower volatility in Europe—and elsewhere.