Bridge over ocean
1 February 2013 CFA Institute Journal Review

Quantifying the Behavior of Stock Correlations under Market Stress (Digest Summary)

  1. Thomas M. Arnold, CFA

Examining 72 years of daily closing prices for the 30 stocks in the DJIA, the authors determine patterns in the average correlations between the stocks relative to the performance of the index. They find that correlation increases when the index performs extremely well or poorly. The effect is more pronounced when the index performs poorly.

What’s Inside?

The authors examine the idea that lower average correlation between securities increases the potential success of diversification strategies. But a concern is whether correlations are stable. Correlations can increase when a market experiences large movements, which, in turn, can negate any potential diversification benefits. To address this concern, the authors examine the average correlations for all pairs of the 30 DJIA stocks and observe that the average correlation tends to increase as the index return becomes larger in absolute value. The effect is more pronounced for negative market returns than for positive market returns.

Their finding is consistent with the idea that extreme market performance tends to increase the systematic risk component of all securities, thus making diversification more difficult when it is most needed.

How Is This Research Useful to Practitioners?

Practitioners can benefit from understanding that extreme market performance increases the systematic component of individual security returns. Consequently, when there is negative performance in the market, fund managers cannot expect as much risk mitigation from diversification strategies as they could under less extreme conditions.

Based on this finding, other forms of risk mitigation need to be implemented when the market exhibits extreme performance or when extreme performance is anticipated. This approach is more critical when markets decline because the increase in the mean correlation between individual securities is larger when markets decline than when markets rise with the same absolute magnitude move.

How Did the Authors Conduct This Research?

The authors use daily closing price data from 15 March 1939 to 31 December 2010 for the 30 DJIA stocks. For different horizons, ranging from 10 to 60 days, the authors compute returns and the average correlations for all pairs between the 30 stocks (with appropriate adjustments for additions and subtractions from the index composition). For a given time horizon, they calculate and normalize index return performance by return volatility. As a result, index performance is measured relative to the number of standard deviations a given return is from average performance.

The average correlation appears to be a function of index performance. Two noticeable and verifiable trends exist in the data, regardless of the time horizon the authors choose. As the index performance becomes more positive, the average correlation between the 30 securities increases. As the index performance becomes more negative, this same effect occurs but at a faster rate, which makes it more pronounced and greater in magnitude for a given absolute market move compared with the effect of the positive case.

The authors conclude that diversification benefits decrease as markets exhibit extreme performance and that the effect is more pronounced in declining markets. Accordingly, diversification begins to fail as a risk mitigation strategy when it is most often needed.

Abstractor’s Viewpoint

The finding that systematic risk becomes more prominent within individual securities during times of extreme market performance (and even more so when the performance is negative) is very interesting. Many researchers in the recent past have alluded to and/or demonstrated this result. But I would like to see this analysis performed on a wider array of stocks and to see an analysis investigating whether such attributes as market capitalization, leverage, or industry classification affect the result.

We’re using cookies, but you can turn them off in Privacy Settings.  Otherwise, you are agreeing to our use of cookies.  Accepting cookies does not mean that we are collecting personal data. Learn more in our Privacy Policy.