Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword or section
Like this
2Activity
P. 1
Do Measures of Liquidity Measure Liquidity

Do Measures of Liquidity Measure Liquidity

Ratings: (0)|Views: 13|Likes:
Published by calibrateconfidence

More info:

Published by: calibrateconfidence on Mar 31, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

04/04/2013

pdf

text

original

 
Do Measures of Liquidity Measure Liquidity?*
Ruslan Y. GoyenkoMcGill UniversityCraig W. HoldenIndiana UniversityCharles A. Trzcinka**Indiana University January 25, 2008
Abstract
Liquidity plays an increasingly important role in empirical asset pricing, market efficiency, andcorporate finance. Identifying high quality proxies for liquidity based on daily data only (notintraday data) would permit liquidity to be studied over relatively long timeframes and acrossmany countries. We introduce new liquidity measures. We run horseraces of both monthly andannual liquidity measures. We compare to effective spread, realized spread, and price impactbased on both TAQ and Rule 605 data, including the decimals era. We identify the best proxiesin each case and find that the new liquidity measures win the majority of horseraces.
 JEL classification:
C15, G12, G20.
Keywords:
 
Liquidity, transaction costs, effective spread, price impact, asset pricing.
* We thank Utpal Bhattacharya, Andrew Ellul, Joel Hasbrouck, Christian Lundblad, DariusMiller, Xiaoyun Yu, and seminar participants at Indiana University and the Frontiers of FinanceConference in Bonaire, The Netherland Antilles. We also thank Charles Jones for making Dowspreads available. We are solely responsible for any errors. This paper was previously circulatedas “Horseraces of Monthly and Annual Liquidity Measures.”** Corresponding author: Charles A. Trzcinka, Kelley School of Business, Indiana University,1309 E. Tenth St., Bloomington, IN 47405-1701. Tel.: +1-812-855-9908; fax: +1-812-855-5875: email address: ctrzcink@ indiana.edu.
 
2
1. Introduction
  The role of liquidity in empirical finance has grown rapidly over the past five years andhas begun to have an effect on conclusions in asset pricing, market efficiency and corporatefinance. A number of studies have proposed liquidity measures derived from daily return andvolume data as proxies for the liquidity and transactions costs experienced by investors.Researchers rarely test the hypothesis that the liquidity proxies are related to actual transactionscosts. The maintained assumption of most studies is that the available liquidity proxies capturethe transactions costs of market participants. This assumption has not tested because of thelimited availability of actual trading costs. In U.S. markets transactions data are only availablesince 1983; in many markets transactions data are not available at all. The consequences of nottesting liquidity proxies on actual trading data is that there is little consensus on which measuresare better and little evidence that any of the proposed measures are related to investor experience.In short, not much is known about whether transactions cost proxies measure what researchersclaim they measure.A handful of studies, Lesmond, Ogden, and Trzcinka (1999), Lesmond (2004),Hasbrouck (2006), test whether a few of the available liquidity proxies as constructed on anannual or quarterly basis from daily return data are related to annual or quarterly liquiditycomputed from transactions data. Yet the vast majority of the literature that
uses
liquidity proxiesis based on
monthly
(or finer) data. This suggests the need to test
monthly
proxies. Furthermore,the literature has proposed different types of liquidity proxies designed to capture differentliquidity benchmarks (e.g. effective spread, realized spread, or price impact). Given the limitednumber of liquidity proxies tested, the limited set of liquidity benchmarks compared to, and theabsence of monthly proxies, it is not surprising that there are conflicting views about which
 
 3measure is better and that there is little assurance that these measures actually capture thetransactions costs of market participants. The purpose of this paper is to address this omission in the literature. We provide acomprehensive study of available measures and introduce three proxies for effective spread andnine price impact measures. We run “horseraces” of liquidity measures testing all widely usedproxies and our new ones. We conduct our tests relative to multiple liquidity benchmarks, usingmultiple high frequency datasets (TAQ and Rule 605 data), using multiple performance metrics,and over a long testing period that includes the decimals regime. We find a close associationbetween many of the measures and actual transactions costs. Some measures are able to preciselyestimate the magnitude of effective spread and many are highly correlated with the effectivespread and price impact. We can safely assert that the literature has generally not been mistakenin the assumption that liquidity proxies measure liquidity. The new class of measures weintroduce in this paper consistently wins a majority of our horseraces. Furthermore, a measurecommonly used in the literature, Pastor-Stambaugh Gamma, is clearly dominated by muchsimpler measures. Further, we find a better alternative to the widely used Amihud measure. The paper is organized as follows. Section 2 details the empirical design of the paper.Section 3 explains the high-frequency liquidity benchmarks in the horserace. Section 4 explainsthe low-frequency measures of effective spread in the horserace. Section 5 explains the low-frequency measures of price impact in the horserace. Section 6 describes the datasets andmethodology used. Section 7 presents the horserace results. Section 8 concludes the paper.
2. Empirical Design
Our basic hypothesis is that useful monthly and annual liquidity measures going back intime can be constructed from low frequency (daily) stock returns and volume data, where such

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->