Bearings January 2015

Liquidity and the Cross Section of Expected Returns

A preference for more liquid investments has led some investors to hypothesize that illiquid securities should command higher expected returns. Our examination of the historical returns and a number of commonly used liquidity and liquidity risk proxies found that the relationship is not robust through time and is mostly isolated to small portions of the market.

Key findings

While it seems obvious that investors would prefer to hold the more liquid of two securities that are otherwise identical, it is less obvious that this preference should have a meaningful impact on expected returns.

The argument for liquidity premiums typically proceeds as follows. Illiquid stocks are costlierto trade, and therefore investors demand higher expected returns to hold them. This hypothesis implicitly assumes that the costs to trade illiquid securities are large for the marginal investor.

The counterargument points out that the costs to trade illiquid securities can be largely mitigated by not trading often or urgently. Long-term investors will be the natural holders of illiquid securities. If they are also the marginal investor in setting prices, trading costs should not be that important for expected returns since trading costs are relatively small when spread over a long holding period. Do investors who trade often and urgently push down the price of illiquid securities to compensate them for their trading costs? Or do these investors simply earn less than their fair share of the premium? The answer to these questions depends on the relative importance of these two types of investors on equilibrium pricing.

We looked for evidence of a liquidity premium by analyzing the most commonly used liquidity and liquidity risk proxies—the Amihud (2002) liquidity measure, which is based on the idea that illiquid stocks will tend to experience larger price changes in response to a given contemporaneous volume of trading than do liquid stocks; the Pastor and Stambaugh (2003) liquidity risk measure, which gives a higher liquidity risk to stocks with returns that covary more strongly with market-wide liquidity shocks; and liquidity measured by past turnover.

The sample set was all common stocks on the NYSE, AMEX, and NASDAQ exchanges listed in the Center

for Research in Security Prices (CRSP) return files during the period July 1962–December 2012. For each liquidity proxy, all stocks were sorted by liquidity into five portfolios and examined, and then another analysis was done after stocks were grouped into microcap, small excluding micro, and large cap, with liquidity sorting done separately within each size group.

Broadly, our results indicate that the relationship between liquidity and returns is not robust through time and is mostly isolated to small portions of the market.

Different variables appear to work in some portions of the market but not in others. From 1968–2012, the Amihud and turnover liquidity variables produce spreads in raw returns and three factor alphas that are reliably different from zero in micro caps only. From 1990–2012, the Pastor and Stambaugh liquidity measure produces economically large spreads for large caps only. Over the past two decades, none of the raw return spreads are reliably different from zero; among alphas there is only a single spread with a t-stat greater than 2.0 and it is obtained within microcaps.

Conclusion

While inconclusive, our analysis of three popular metrics of liquidity suggests these premiums are not robust across different time periods and firm sizes. For a long-only investor, there’s not much evidence that tilting toward illiquid stocks results in higher expected returns. However, market frictions due to illiquidity are still important to consider when managing portfolios. Liquidity driven by the supply and demand for securities affects the prices at which they transact. Dimensional uses patient, flexible trading processes in an attempt not to demand immediacy when trading. This has resulted in security transactions at favorable prices, which has added value to our portfolios. Our research team is always evaluating ways to improve our strategies and will continue to look at liquidity in the future.

Download