یک تخمین ساده از گسترش معاملات طول روز با استفاده از داده روزانه
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|13769||2014||27 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Financial Markets, Volume 17, January 2014, Pages 94–120
This study examines the relation between the bid-ask spread from the daily CRSP data and the bid-ask spread from the intraday TAQ data. We show that the CRSP-based spread is highly correlated with the TAQ-based spread across stocks using data from 1993 through 2009. The simple CRSP-based spread provides a better approximation of the TAQ-based spread than all other low-frequency liquidity measures in cross-sectional settings. However, the CRSP-based spread is highly correlated with the TAQ spread in time-series settings only for NASDAQ stocks. Overall, our results suggest that the simple CRSP-based spread could be used in lieu of the TAQ-based spread in academic research that focuses on cross-sectional analysis.
The bid-ask spread is a measure of stock market liquidity that has been frequently employed in market microstructure studies. For example, previous studies (e.g., Christie and Schultz, 1994, Huang and Stoll, 1996 and Bessembinder, 2003a) use the bid-ask spread to perform inter-market comparisons of trading costs. In addition, market regulators implement various rules and regulations to reduce the cost of trading, and subsequently assess the efficacy of these rules and regulations by analyzing their impact on the bid-ask spread. For instance, the U.S. Securities and Exchange Commission (SEC) reduced the minimum price variation (i.e., tick size) from $1/8th to $1/16th in 1997 and again from $1/16th to one cent in 2001 with the specific purpose of reducing the bid-ask spread in the U.S. securities markets. Similarly, the SEC enacted the Limit Order Display Rule in an effort to reduce the inside spread of NASDAQ-listed stocks.2 Most prior market microstructure research relied on the Trade and Quote (TAQ) data provided by the New York Stock Exchange (NYSE), which involves a tedious process of data downloading, error filtering, and variable calculation. In this study, we propose an alternative method of calculating the bid-ask spread that requires only daily data and minimal computational efforts. Our simple liquidity measure would be useful to those who do not have an access to the TAQ database and/or those who want to incorporate stock market liquidity in their research without having to go through the process that is required for the TAQ database. Our simple liquidity measure is readily available for NYSE/AMEX/NASDAQ stocks from 1993 onwards, which coincides with the time period covered by the TAQ database. Another possible advantage of our simple liquidity measure is its availability beyond the time period covered by the TAQ database: one can easily obtain our low-frequency bid-ask spread measure from 1925 to 1942 for all NYSE/AMEX stocks and from 1982 to 1992 for most NASDAQ stocks.3 Despite its usefulness as a measure of stock market liquidity and information asymmetry, the usage of the TAQ-based bid-ask spread in other research areas has been limited due, at least in part, to data availability problems. Our study is mainly motivated by the need for readily available liquidity measures in other research areas, such as corporate finance, financial accounting, and asset pricing. Researchers in these areas show that liquidity plays an important role in many financial decisions and the pricing of assets. For example, prior research underscores possible interactions between stock market liquidity and (1) capital structure (Frieder and Martell, 2006 and Lipson and Mortal, 2009), (2) dividend payout and stock repurchase decisions (Banerjee et al., 2007 and Brockman et al., 2008), (3) ownership structure (Heflin and Shaw, 2000, Sarin et al., 2000 and Brockman et al., 2009), (4) firm value (Fang, Noe, and Tice, 2009), (5) corporate governance (Chung, Elder, and Kim, 2010), (6) executive compensation (Jayaraman and Milbourn, 2012), (7) corporate innovation (Fang, Tian, and Tice, 2013), (8) institutional investors' stock selection decisions (Falkenstein, 1996, Chung and Zhang, 2011 and Huang, 2013), and (9) asset pricing (Amihud and Mendelson, 1986 and Spiegel and Wang, 2005).4 Readily available liquidity measures would be useful to researchers in these areas, especially when they need liquidity measures for a large cross-section of firms in the post-1993 period. Previous studies have developed low-frequency liquidity measures using daily closing prices from the Center for Research in Security Prices (CRSP). They include Roll (1984), Lesmond, Ogden, and Trzcinka (1999), Hasbrouck (2009), and Holden (2009). Goyenko, Holden, and Trzcinka (2009) compare the TAQ-based effective spread with various low-frequency liquidity measures using a sample of 400 randomly selected stocks over the period from 1993 through 2005. They show that the TAQ-based effective spread is highly correlated with these low-frequency liquidity measures. Other studies estimate stock liquidity using Ask or High Price and Bid or Low Price in the CRSP database. Ask or High Price is the highest trading price during the day or the closing ask price on days when the closing price is not available. Likewise, Bid or Low Price is the lowest trading price during the day or the closing bid price on days when the closing price is not available. Eckbo and Norli (2002) and Holden (2009) use these variables on no-trade days to estimate the monthly bid-ask spread of stocks that have at least one no-trade day in a given month. However, as Holden (2009) shows, only 26% of the 62,100 stock-months contain one or more no-trade days during 1993–2005. Corwin and Schultz (2012) develop a bid-ask spread estimator using the CRSP's Ask or High Price and Bid or Low Price. Using a sample of U.S. stocks from 1993 through 2006, Corwin and Schultz (2012) show that their spread estimates are highly correlated with the TAQ-based effective spread–the average cross-sectional correlation coefficient between the two variables is 0.930 during 1993–1996 and 0.732 during 2001–2006. In the present study, we propose a simple bid-ask spread measure that can be calculated using the daily data from the CRSP. In contrast to previous studies that use the CRSP's Ask or High Price and Bid or Low Price to obtain low-frequency liquidity measures, we use data in the two new fields (i.e., Ask and Bid) that were added to the CRSP database in December 2005. 5 To our best knowledge, this study is the first to analyze the usefulness of these variables in academic research. In addition, our study differs from previous studies (e.g., Lesmond et al., 1999 and Corwin and Schultz, 2012) in that our low-frequency liquidity measure requires neither a sophisticated estimation procedure nor large computational efforts (in both time and space), making it easier for both researchers and practitioners to use. We show that the CRSP-based spread is highly correlated with the TAQ-based spread using data from 1993 through 2009. For instance, we find that the annual average of monthly cross-sectional correlation coefficients between the CRSP spread and the TAQ spread ranges from 0.8267 in 1996 to 0.9603 in 2003 for NYSE/AMEX stocks. For NASDAQ stocks, the annual average of monthly cross-sectional correlation coefficients between the CRSP spread and the TAQ spread ranges from 0.9193 to 0.9729. The cross-sectional correlations between the CRSP spread and TAQ spread are quite robust across firms with different characteristics, especially for NASDAQ firms. We also provide evidence that the simple CRSP-based spread provides a better approximation of the TAQ spread than all other low-frequency liquidity measures in cross-sectional settings. However, the CRSP spread is highly correlated with the TAQ spread in time-series settings only for NASDAQ stocks. Overall, our empirical results suggest that the simple CRSP spread could be used in lieu of the TAQ spread in academic research that focuses on cross-sectional analysis. The rest of the paper is organized as follows. Section 2 provides the definition of each liquidity measure and presents descriptive statistics. In Section 3, we analyze the cross-sectional relation between the CRSP spread and the TAQ spread. Section 4 provides evidence on how the CRSP spread compares with other low-frequency liquidity measures as an approximation of the TAQ spread in both cross-sectional and time-series settings. Section 5 provides a brief summary and concluding remarks.