Content area
Full Text
ABSTRACT
New liquidity measure, based on trading volume induced by order flow as in Pastor and Stambaugh (2002) but estimated with turnover rather than with absolute level of dollar volume, is introduced and analyzed in this paper. Aggregate liquidity measures are found to well track the history of market liquidity problems. However, market price of liquidity risk, estimated as a coefficient of liquidity shock, does not show any systematic timeseries behavior so we could not find the variables which have significant explanatory power for liquidity risk premium.
Keywords: Liquidity, Liquidity risk, Liquidity risk premium, Liquidity shock
(ProQuest: ... denotes formulae omitted.)
INTRODUCTION
In this paper, we introduce a new liquidity measure and investigate the empirical features of this liquidity measure. In addition, we estimate the liquidity risk premium and analyze its time-series behavior using some ex-ante variables. Theoretically, our measure is based on Campbell, Grossman and Wang (1994, CGW thereafter) as is that of Pastor and Stambaugh's (2002), which is constructed as a measure induced by orderflow using absolute level of trading volume (ie, dollar volume of trade). Our new liquidity measure is based on relative trading volume rather than absolute trading volume. We construct our liquidity measure using each individual firm's turnover (relative trading volume). Our liquidity measure is found to track the history of market liquidity problems well and shows the different characteristics by market capitalization of firms, which is now well known: Firms with large market capitalization are more liquid than small firms.
In terms of asset pricing, great concern was shown if the given liquidity risk can explain the cross-sectional variation of stock returns (Amihud 1989, 2002; Pastor and Stambaugh 2002). But to our knowledge, no efforts were shown to explain the behavior of the liquidity risk premium in stock returns1). Though some papers report the estimated value of liquidity risk premium (Pastor and Stambaugh, 2002), they did not try to find out its time-varying behavior by using some pre-specified variables. In the conditional test of multifactor model, Ferson and Harvey (1991) shows some economic variables can track stock market risk premiums. In this paper, we will investigate the economic factors that explain the liquidity risk premium.
We found some evidence that the liquidity, obtained by our new measure, is...