Liquidity Level or Liquidity Risk? Evidence from the Financial Crisis

Although generally considered safe assets, liquid stocks underperformed illiquid stocks during the financial crisis of 2008–2009. The performance of stocks during the crisis can be better explained by their historical liquidity betas (risk) than by their historical liquidity levels. Stocks with different historical liquidity levels did not experience different returns after controlling for liquidity risk. The authors’ findings highlight the importance of accounting for both liquidity level and liquidity risk in risk management applications.In our study, we highlighted the difference between liquidity level and liquidity risk and showed that the latter is a better predictor of performance during a crisis. We defined the level of a stock’s liquidity as the ability to trade large quantities of its shares quickly and at low cost, on average. In contrast, we defined the liquidity risk (beta) of a stock as the covariation of its returns with unexpected changes in aggregate liquidity. The two measures capture different attributes of a stock’s liquidity profile. For example, liquidity level may be considered a mean effect, whereas liquidity beta may signify a volatility or correlation effect. Although generally considered safe assets, liquid stocks underperformed illiquid stocks during the financial crisis of 2008–2009. The performance of stocks during the crisis can be better explained by their historical liquidity betas than by their historical liquidity levels. Furthermore, stocks with different historical levels of liquidity did not experience different returns after controlling for liquidity risk. With some benefit of hindsight, these results are perhaps not particularly surprising. After all, which stocks are more likely to suffer during a liquidity crisis? We suggest that portfolio managers should worry about liquid stocks with high liquidity risk because their liquidity is likely to dry up during a crisis whereas the illiquid stocks will continue to be illiquid. Liquidity beta offers a way to measure this type of risk. Moreover, because variances are more persistent than means, liquidity beta could provide more accurate out-of-sample signals for risk management than could liquidity level. Finally, the U.S. SEC and the U.S. Commodity Futures Trading Commission recently published a report on the “flash crash” of 6 May 2010. Once again, large companies, such as Procter & Gamble and Accenture, were among the most affected by that crisis, which lends our argument further support. Because the correlation between stock liquidity variations has been increasing over the past few decades, these results highlight the importance of accounting for liquidity risk in risk management applications.