تاثیرات قیمت گذاری دارایی از شوک نقدینگی بازار انگلیس: شواهدی از داده تیک
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|13776||2014||10 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Review of Financial Analysis, Volume 32, March 2014, Pages 85–94
Using tick data covering a 12 year period including much of the recent financial crisis we provide an unprecedented examination of the relationship between liquidity and stock returns in the UK market. Previous research on liquidity using high frequency data omits the recent financial crisis and is focused on the US, which has a different market structure to the UK. We first construct several microstructure liquidity measures for FTSE All Share stocks, demonstrating that tick data reveal patterns in intra-day liquidity not observable with lower frequency daily data. Our asymptotic principal component analysis captures commonality in liquidity across stocks to construct systematic market liquidity factors. We find that cross-sectional differences in returns exist across portfolios sorted by liquidity risk. These are strongly robust to market, size and value risk. The inclusion of a momentum factor partially explains some of the liquidity premia but they remain statistically significant. However, during the crisis period a long liquidity risk strategy experiences significantly negative alphas.
One of the striking features of the recent financial crisis was the abrupt drop in aggregate liquidity across global financial markets. This drop in liquidity is a market failure that led to a large increase in trading costs through wider spreads and greater price impact. The financial crisis has heightened awareness among investors of the importance of considering liquidity (Brunnermeier, 2008; Longstaff, 2010). In this paper we make three key contributions to the literature on liquidity and stock returns. We are the first paper to examine the pricing of liquidity risk in stock returns in the UK market. Second, we employ a high frequency intra-day data set unprecedented in depth for a UK study. Finally, as we specify a sample period which incorporates this crisis incidence of market illiquidity, our paper provides much needed additional evidence on the role of liquidity in asset pricing. Trading on the UK stock market is quite different to the US, where prior research on high frequency data has focused. In the UK all trading takes place on the London Stock Exchange (LSE) whereas in the US stocks trade primarily on two main exchanges, the Nasdaq and NYSE. On the LSE trading is a mix of order book driven (SETS) and a hybrid quote/order book driven system (SETSmm), whereas in the US trading on Nasdaq is order book driven and the NYSE has a hybrid system. The differing market structure of UK and US exchanges leads to differences in liquidity characteristics (Huang & Stoll, 2001). By providing evidence on the pricing of liquidity in the UK market we are able to assess whether these differences in market structure and liquidity characteristics affect conclusions on the relation between liquidity and stock returns as documented in the predominantly US literature. Using an extensive data set of over 1.2 billion tick and best price observations covering the period January 1997 to February 2009 we are able to construct several microstructure stock liquidity measures for the UK for the first time. Our tick data enable the calculation of liquidity measures, some of which cannot be calculated using lower frequency, even daily, data. Others can be estimated with daily data but we find such estimates risk biasing results.1 We construct time series of seven liquidity measures for each of the FTSE All Share constituent stocks over our sample period. We examine a large number of measures as different aspects of liquidity risk may not all be captured by one measure. For each liquidity measure we use asymptotic principal component analysis to capture commonality in liquidity across stocks in order to develop a systematic market liquidity factor. We also develop a systematic market liquidity factor across all seven measures combined which draws on the commonality in liquidity across assets as well as the commonality across liquidity measures. We construct liquidity risk mimicking portfolios based on stocks' sensitivity to shocks to our systematic market liquidity factors. We examine several related questions: Is there a return premium for UK market or systematic liquidity risk? If so, is this return premium compensation specifically for the stock's systematic liquidity risk or the liquidity characteristics of the stock generally? What is the degree of commonality across liquidity measures among UK stocks? Are liquidity shocks persistent? Briefly, we find that liquidity risk confers a significant premium in normal market conditions. There is evidence that the liquidity risk premium is related to momentum, consistent with Sadka (2006), but is unrelated to market, size and value risk. However, our new evidence around the recent financial crisis indicates that liquidity risk sensitive portfolios suffered significant abnormal negative returns during the period, highlighting the skewed nature of the pricing of liquidity risk. The paper is organised as follows: Section 2 provides a brief discussion of the theory and empirical methods of the surrounding literature. Section 3 describes the extensive data set used. Section 4 outlines the methodology for estimating the liquidity measures from the data while Section 5 presents the methodology and results of tests for the cross sectional pricing of liquidity risk. Section 6 concludes.
نتیجه گیری انگلیسی
In this studywe employa high frequency intra-daydataset, unprec- edented in scale for the UK equity market, to investigate the asset pric- ingeffectsofmarketliquidityshocks.Ourtickandbestpricedatapermit a richer analysis of liquidity by enabling the construction of liquidity measures which could not be calculated using lower frequency daily data. We construct time series of seven liquidity measures for each of the FTSE All Share constituent stocks during our sample period. Wethen construct systematic market liquidity factors for each measure as well as an across measure factor which captures commonality both across stocks and across liquidity measures. Our preliminary data anal- ysis indicates strong commonality across liquidity measures and also shows that market liquidity shocks persist for up to one year. In our main results, liquidity risk mimicking portfolios exhibit a statistically signi fi cantreturn premium amonghighliquidity riskstocks.Controlling forliquiditylevelasastockcharacteristicdoesnotalterourconclusions. These are strongly robustto market, sizeand value risk.The inclusion ofa momentum factor partially explains some of the liquidity premia but they remainstatistically signi fi cant. Weextendthe literature by provid- ingevidence on thepricingof liquidity risk duringthe fi nancial crisis. In contrast to more normal market conditions our fi ndings highlight that liquidityriskmimickingportfolios experienced signi fi cantlosses during the crisis period. Overall, our results suggest that liquidity risk makes a signi fi cant contribution to asset pricing and point to a need to examine the liquidity exposure and liquidity risk adjusted returns of actual UK equity funds.