Do voluntary disclosures of bad news improve liquidity?

North American Journal of Economics and Finance 40 (2017) 16–29

Contents lists available at ScienceDirect

North American Journal of Economics and Finance j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / e c o fi n

Do voluntary disclosures of bad news improve liquidity? Ajit Dayanandan ⇑, Han Donker, Gökhan Karahan University of Alaska Anchorage, United States

a r t i c l e

i n f o

Article history: Received 5 August 2015 Received in revised form 23 December 2016 Accepted 2 January 2017

JEL classification: G1 M4 Keywords: Profit warning Bad news Voluntary disclosure Liquidity

a b s t r a c t Can managers improve market liquidity and lower the cost of capital by providing voluntary earnings guidance? This study examines the impact of profit warnings on market liquidity and finds that voluntary disclosure of bad news actually improves market liquidity. By conducting an empirical study over the period 1995–2010 on NYSE, NASDAQ and AMEX listed firms, we find that firms that issue profit warnings show enhanced market liquidity during the post-announcement period. We show that profit warnings reduce information asymmetry and lower bid-ask spreads and increase trading volumes. These results are invariant to daily (short run) and monthly (long run) data after controlling for firm specific attributes. The results have major corporate policy implications. By voluntarily disclosing negative earnings guidance by managers, firms will experience significant improvement in market liquidity, thereby lowering the cost of capital. Our results are even more profound for firms that release bad news with extremely negative stock market impact. In other words, voluntary disclosure of bad news is good for market liquidity. Published by Elsevier Inc.

1. Introduction Equity markets are characterized by information asymmetry (Jensen, 2005; Jensen & Meckling, 1976). Firm’s managers have more information about the expected profitability of the firm than investors. Sometimes, the firm’s profit performance falls short of expectations and managers voluntarily choose to issue profit warnings (PWs), causing investors to revise their assessment of firm value. Prior literature have examined the impact of firms’ disclosure on information asymmetry especially on cost of equity (Leuz & Verrecchia, 2000; Levi & Zhang, 2014). Changes in disclosure and consequent changes in information asymmetry are known to affect liquidity. The main aim of this study is to examine the impact of PWs on different measures of market liquidity, using a unique hand-collected dataset from the United States over the period 1995–2010. PWs are issued by management before a quarterly earnings announcement where reported earnings will be less than the market consensus forecast of analysts. PWs differ from earnings announcements in the sense that it occurs irregularly and unpredictably across firms and time and convey unique information and are issued well ahead of earnings announcements. PWs are special forms of earnings guidance, which tend to occur when managers realize that they will not be in a position to meet earnings expectations; they can be considered as revised information (management guidance) about prospective firm’s performance. PWs provide more material to those investors who attempt to process such information to create private benefits. PWs are voluntary disclosures and like other bad news, such as negative earnings surprises, bond rating downgrades, dividend cuts, analyst downgrades, financial restatements, etc. are meant to reduce information asymmetry in the market. ⇑ Corresponding author. E-mail addresses: [email protected] (A. Dayanandan), [email protected] (H. Donker), [email protected] (G. Karahan). http://dx.doi.org/10.1016/j.najef.2017.01.002 1062-9408/Published by Elsevier Inc.

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

17

PWs thus provide valuable information that has not been evaluated and processed by financial analysts (Jackson & Madura, 2003). How stock prices react to negative news – as news or noise – has been a topic of long-standing interest in finance. Skinner (1994, 1997) and Healy and Palepu (2001) argue that litigation risk1 potentially induces firms to disclose forwardlooking information. Thus, if PWs could reduce information asymmetry and reduce the level of adverse selection in the market, then it could have a positive impact on stock market liquidity around the PW news releases (Balakrishnan, Billings, Kelly, & Ljungqvist, 2014; Glosten & Milgrom, 1985; Kyle, 1985; Verrecchia, 2001). Therefore, announcement of PWs could be good news from the point of market liquidity and, this, in turn, could reduce the cost of capital (Levi & Zhang, 2014). Diamond (1985), Diamond and Verrecchia (1991), and Balakrishnan et al. (2014) show that managers may sometimes disclose more than what is required by market regulators. However, the market impacts of such disclosures are uncertain as managers do not know the market participants’ entire information set (Dutta & Trueman, 2002; Dye, 1998; Dye & Sridhar, 2002; Suijs, 2007). Moreover, the release of new information changes market participants’ expectations as well as stock price discovery. Empirical evidence shows that PWs have, on average, a negative abnormal return during the announcement window (Bulkley & Herrerias, 2005; Church & Donker, 2009; Jackson & Madura, 2003, 2004, 2007; Xu, 2008). But, the recent financial crisis has shown that compounding of bad news can affect the market sentiments and liquidity. In this context, an important question is whether negative earnings guidance disclosed by managers, such as PWs, could improve market liquidity or not? Our study is unique as it shows that releasing negative earnings guidance in the form of PWs reduces the information asymmetry in the market and impacts liquidity. The empirical results are based on a large sample of PWs in the United States during 1995–2010. Our study contributes to the literature on the market impacts of bad news by providing rationale for managers to voluntarily disclose lower prospective earnings as it improves market liquidity in the short and medium run. This is important since market liquidity channels have a considerable impact on the discount rate and the cost of equity of firms. The study is organized as follows: Section 2 presents review of the literature. Section 3 presents details on data and methodology. The empirical results are presented in Section 4 and the robustness tests in Section 5. Finally, Section 6 summarizes the results of the study. 2. Review of literature The rationale for issuing PWs rather than waiting until the earnings announcement has been researched extensively. Studies range ranging from litigation risk (Baginski, Hassell, & Kimbrough, 2002; Skinner, 1994, 1997; Soffer, Thiagarajan, & Walther, 2000; and Field, Lowery, & Shu, 2005), reputational costs (Libby & Tan, 1999), entry deterrence (Darrough & Stoughton, 1990; Dontoh, 1989), regulatory risk (Jackson & Madura, 2007; Mikhail, Walther, & Willis, 2004; Skinner, 1994; Verrecchia, 2001) to feedback received from financial markets (Kau, Linck, & Rubin, 2008; Langberg & Sivaramakrishnan, 2010; Luo, 2005). If financial markets are characterized by asymmetrical information, increased disclosure (i.e., PWs) can reduce information asymmetry and reduce insider trading. The asymmetric information theory provides the analytical framework for analyzing the impact of voluntary disclosure on financial markets (Jensen & Meckling, 1976; Leland & Pyle, 1977; Myers & Majluf, 1984; Stiglitz & Weiss, 1981). Investors have imperfect knowledge of firm’s prospects, and they are normally concerned whether they are holding or acquiring an overvalued asset. Research has shown that statements which report an adverse outlook for the future prospects of the firm (PWs) do contain market-relevant information, causing investors to revise expectations about the future profitability of the firm (Bulkley & Herrerias, 2005; Church & Donker, 2009; Jackson & Madura, 2003, 2004, 2007; Xu, 2008). There is considerable literature on voluntary disclosures and its impact on financial outcomes like stock prices, liquidity, etc. Prior literature has examined how high information asymmetry can reduce liquidity (Copeland & Galai, 1983; Glosten & Milgrom, 1985). However, research on the link between voluntary financial disclosure and market liquidity is very limited. Theoretical models of Diamond (1985) and Diamond and Verrecchia (1991) show why managers may choose to disclose more information than mandated by market regulators. Leuz and Verrecchia (2000) argue that increased levels of financial disclosure reduce the information asymmetry and costs of capital. Higher levels of information asymmetry generate transaction costs by including adverse selection between buyers and sellers of shares and increase the discount at which shares are sold and increase the cost of issuing shares. Leuz and Verrecchia (2000) suggest bid-ask spreads and trading volume as proxies for information asymmetry; they find that German firms that commit to higher levels of disclosure experience a 35 percent decrease in bid-ask spreads and a 50 percent increase in stock turnover. There are also studies on analyst evaluations of disclosure quality and liquidity. These studies find that disclosure quality increases market liquidity (Healy, Hutton, & Palepu, 1999; Welker, 1995; Heflin, Shaw, & Wild, 2005). On the other hand, Bardos (2011) find that illiquidity increases after firms restate their financial statements. Further empirical evidence also shows that firms disclose more information when earnings are easier to predict (Chen, Matsumoto, & Rajagopal, 2011). Levi and Zhang (2014) study of U.S. during 1993–2009 found that days before earnings announcements, information asymmetry increases which lead to low liquidity. There is also considerable literature on the role of trading volume in the pre- and post-earnings announcement periods (Bamber, 1987; Chae, 2005; Kim & Verrecchia, 1991). Changes in trading volume facilitate stock price discovery process.

1 Rule 10b-5 of the Securities Exchange Act of 1934 could be used against a non-PW firm of either failing to disclose adverse information or disseminating misleading information (Dyl, 1999).

18

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

Chae (2005) provides an analysis of pre-announcement trading volume, which gives evidence into how informed and uninformed investors behave. According to Chae (2005), trading volume increases asymmetry of information in financial markets as uninformed investors postpone their liquidity trading demands when they anticipate increased market participation by informed investors. Chae (2005) also argues that uninformed investors do not adjust the timing of their trades, if they are not aware of impending disclosures. If informed traders increase their trading demands, while uninformed investors do not adjust their trading patterns, the event will drive up the total volume, thus, possibly decreasing information asymmetry. However, given the fact that PWs are bad news, one would expect increased trading activity (volume) by informed traders thereby accentuating the negative impacts of stock returns. Based on the extant literature discussed above, we hypothesize that profit warnings will reduce information asymmetry and improve market liquidity in the post-PW announcement period (s).2 The present study examines not only the stock market impacts but also the market liquidity impacts of profit warnings. Stock market impacts have been extensively studied, but very limited number of studies exists on the impact of voluntary disclosure on market liquidity. Reducing information asymmetry has beneficial effects on the cost of capital. The public policy implication of this study is that issuing negative earning guidance (profit warnings) will improve the efficiency and price discovery process of the market. 3. Data source and methodology The present study analyzes the impact of profit warnings (n = 1945) on the market liquidity of listed firms during the period 1995–2010. We use firms listed on the New York Stock Exchange (NYSE), American Stock Exchange (AMEX), and the National Association of Securities Dealers Automated Quotations (NASDAQ) from 1995 to 2010 as our cohort. Profit warnings were manually collected by searching the Wall Street Journal, New York Times, Washington Post, Financial Times, and newswires from the financial markets. The keywords search used were ‘‘profit warning” and ‘‘earnings warning.” We do not extend our sample period to before 1995 since profit warnings were infrequently acknowledged in newspapers before 1995. Our data gathering process resulted in a total of 1945 observations of firms issuing profit warnings. The stock price data, adjusted for dividends and splits, was compiled from the Center for Research in Securities Prices (CRSP). Financial data were obtained from COMPUSTAT files. The event date (t = 0) is the announcement date in the newspapers. We use the event study methodology to analyze the effect of a profit warning on stock prices (Brown & Warner, 1985; Campbell, Lo, & Mackinlay, 1997; MacKinlay, 1997). The abnormal returns for the profit warnings from each firm were calculated using Fama-French 3 factor model (Eq. (3)), based on a value-weighted total-market index from COMPUSTAT. An event window of 30 days before and 30 days after the day of the profit warning was used for this event study. The preestimation period for the event study was 240 days before the event date, equivalent to approximately one year of trading. The case for event studies has been made by Brown and Warner (1985), and this technique is widely used in the empirical investigations (Kothari & Warner, 2005). Based on the framework of Brown and Warner (1985) and Campbell et al. (1997), let t = 0 represent the time period t relative to the profit warning event and actual return (Rit), is:

Rit ¼ Kit þ eit

ð1Þ

where Kit is the ‘‘normal” (i.e., expected or predicted return of a particular model) and eit is the component of the return which is abnormal or unexpected. Thus the abnormal return (AR) is the difference between the observed return and the predicted return:

eit ¼ Rit  Kit

ð2Þ

In the econometric investigation, we need to specify a model of normal returns (i.e., expected returns unconditional on the event but conditional on other information). To calculate the daily abnormal returns we used the Fama and French (1993) three-factor model (FF3):

FF3 :

^ HMLt Þ ^i MKT mt þ ^si SMBt þ h ^i þ b ARit ¼ Rit  ð/ i

ð3Þ

where MKTmt is the rate of return of the equally weighted NYSE/AMEX/NASDAQ market index from the Center for Research in Security Prices (CRSP) to proxy for the U.S. market index on day t; SMBt is the average return on small market-capitalization portfolios minus the average return on three large market-capitalization portfolios on day t; HMLt is the average return on two high book-to-market equity portfolios minus the average return on two low book-to-market equity portfolios on day t. For the U.S. factors, we use the data posted on French’s website.3 2 Voluntary disclosures, good or bad, are expected to improve liquidity even though stock market reactions are more adverse for bad news. It is also expected that bad news such as PWs likely have better liquidity effects than good news such as profit updates. There is also the theoretical possibility that proprietary costs of voluntary disclosures may be such that disclosed information may lead to such competitive disadvantages that liquidity may be reduced. However, the likelihood of this happening with ‘‘voluntary” disclosures may be very small. 3 http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

19

The concept of liquidity is elusive, so there is no single measure that fully captures liquidity (Aitken and Comerton-Forde (2003),4 Goyenko, Holden, & Trzcinka, 2009; De Jong & Rindi, 2009). We consider a range of measures including breadth and price impact of trading measures. The simplest is the volume-based measures, which primarily measure breadth and depth of the market. Among the pricing impact of trading measures, we use Amihud’s (2002) illiquidity measure (ILLIQ) which is estimated as the daily ratio of absolute stock return to its absolute order flows (buy plus sell orders).

AmihudðILLIQ it Þ ¼

i Days 1 Xt jReturnt;n j  106 Dayst n¼1 Turnov erit;n

ð4Þ

where n = 1, 2, 3,. . ., Dayst represent the days of the month t, Returnit;n is the ex-dividend stock return of security i on day n of month t and Turnov er it;n is the value of shares traded (measured in millions of dollars). We use daily data available from CRSP. The Amihud measure (ILLIQ) is called an illiquidity measure; a high estimate indicates low liquidity (high price impact of trades). Amihud (2002) shows this measure is strongly priced in the cross-section of stock returns. This result is also confirmed by Chordia, Huh, and Subramanyam (2009) and Goyenko et al. (2009). Although we prefer to use Amihud (ILLIQ) in as much as existing literature clearly favors this one, our results are invariant to the choice of liquidity measure (Quoted Spread and Effective Spread) or measures of illiquidity-Amihud measure (ILLIQ). The literature also discusses various measures to measure the depth of the market and the common liquidity is the bid-ask spread. Based on bid-ask spread, we consider (a) Quoted spread (QSpread) and (b) Effective Spread (ESpread). The Quoted spread (QSpread) is the difference between the asking price (pAt ) and the bid price (pBt ). For the monthly quoted spread, we calculate the average quoted spread over each month,

jSt j ¼

1 t

days=month X

ðptA  pBt Þ

ð5Þ

t¼1

where t = 1, 2, 3,. . ., Dayst/montht Effective spread (ESpread) is given by

Eff jSt j ¼

1 t

days=month X

ðpt  pM t Þ

ð6Þ

t¼1 ðp A þpB Þ

t t where pt is the last traded price before time t and pM , where pAt and pBt are the ask and bid price, respectively. The t ¼ 2 effective spread measures the actual trading cost. The lower the spread better is the liquidity. A smaller effective spread than the quoted spread means that trading mainly occurs within the quoted spread (Chordia, Roll, & Subramanyam, 2000). A simplest measure of volume is the turnover (TURN) which is the average ratio of daily volume divided by the average volume in the prior twenty days. We also estimated the mean abnormal relative volume (AVOL) by conducting a volume event study using the methodology proposed by Campbell and Wasley (1996). The volume event study is conducted similar to an event study that uses returns. The main difference is that log-transformed relative volume replaces the returns for each security. The market model abnormal trading volume is given below:

V it ¼ ait þ bi V mt þ 2it

ð7Þ

where V mt is computed as the sum of the trading volume of all securities in CRSP equally weighted index. The abnormal trad^ it þ b^i V mt Þ. ing volume is v it ¼ V it  ða We employed an event study framework to analyze the link between profit warnings and liquidity. Through the analytical framework of event study and choosing a shorter event window, one is able to isolate the market liquidity impacts of profit warnings. The empirical results are presented on (a) daily and (b) monthly basis. The daily data presents the estimates of various liquidity measures ten days prior and ten days subsequent to PWs. A dummy variable is used for the postannouncement period which equals one for the post-announcement period, and zero otherwise. In the regression analyses, we use firm specific variables, such as relative size (SIZE), return on assets (ROA), leverage (LEV), cash flow from operations (CFO), and market-to-book ratio (MTB). We also present a detailed analysis of determinants of liquidity/illiquidity using monthly data to bring into focus firmspecific attributes as well (apart from the post-announcement effects such as relative firm size (SIZE), return on assets (ROA), leverage (LEV), cash flow from operations (CFO), market-to-book ratio (MTB), which are treated as control variables). The list of variables, description, and source used in the empirical investigation is presented in Appendix 1. The preannouncement window consists of months six before the announcement of the profit warning and six subsequent months. Larger firms typically are more diversified and have greater financial strength and financial analyst coverage. Therefore, it is hypothesized that profit warnings of large firms will improve market liquidity. Similarly, profitability enables the firms to be

4

According to Aitken and Comerton-Forde (2003), there are 68 measures of liquidity in the market.

20

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

resilient in the midst of bad news brought by emitting a profit warning. We anticipate the liquidity response to profit warnings will be smaller for firms with higher profitability. Furthermore, we included several other control variables, such as leverage (LEV), cash flows from operating activities (CFO), and market-to-book (MTB). Based on these variables, we estimate a fixed-effects panel regression (which account for individual heterogeneity of firms) based on the following equation:

Amihud ðILLIQ it Þ ¼ a0i þ b1i POST it þ b2i SIZEit þ b3i ROAit þ b4i LEV it þ b5i CFOit þ b6i MTBit þ 2it

ð8Þ

Theoretically, profit warnings may not be entirely exogenous. It might be endogenously related to liquidity issues. To account for the possibility of endogeneity of profit warnings we also estimate market liquidity as the dependent variable (ILLIQ), using 2SLS. In addition, we present results based on various portfolios depending on the stock market impacts (cumulative abnormal returns – CARs). We prepared five portfolios depending on the magnitude of stock market impacts ranging from large negative CARs and report estimates varying from the extreme left hand side of the CAR distribution (highly negative CAR portfolio) to extreme right hand side of the distribution (less negative CAR-portfolio),highlighting any heterogeneous behavioral impacts, if any. We refrained from conducting behavioral analysis based on business cycles as there is considerable overlap between preand post-announcement periods especially for events, which are located closer to the end period of the business cycle. For example, if a profit warning is in the last month of the expansion phase, the post-announcement impacts lies in the contraction period and this by itself confounds the real liquidity impacts. However, we have conducted tests for daily data and, the main results are confirmed by such analysis. 4. Empirical results The stockholder wealth impact of the PW announcements on firms is presented in Table 1. As expected, firms trumpeting bad news are hit with negative cumulative average abnormal returns (CAARs). During the event window [1,0,+1] around the announcement day, firms experience negative abnormal returns of 14.08 percent which is statistically significant at the 1 percent level. If one takes an event window of ten days prior and post PWs, the stock market impact is slightly larger (17.81 percent). Our results are consistent with Jackson and Madura (2003, 2004, 2007) for the United States. Our results are much stronger when compared to studies on earnings surprises (Feldman, Govindaraj, Livnat, & Segal, 2010; Johnson & Zhao, 2012). Table 2 provides estimates of 5 liquidity measures on a daily basis–10 days prior and 10 days after PWs: Amihud illiquidity (see Eq. (4)), quoted spread (see Eq. (5)), absolute effective spread (see Eq. (6)), turnover, and mean abnormal relative volume (see Eq. (7)). Amihud illiquidity measure (ILLIQ) shows that the market liquidity has improved (lower Amihud illiquidity) in the post-PW period, corroborating evidence of improvements in turnover (TURN) and decline in spread-based indicators of liquidity. The difference in pre- and post-PW periods for Amihud illiquidity measure is positive (0.003) and also statistically significant. These results validate the hypothesis that profit warnings improve the market liquidity of firms. Both the quoted spread (QSpread) and absolute effective spread (ESpread) have decreased during the post-PW period implying market liquidity has improved during the post-announcement period. For both these spread-based measures, the difference between pre and post is expectedly positive implying that spread was higher during pre-PW period as opposed to post-PW period; the pre-post difference in QSpread is 0.0128 and in ESpread is 0.006. The results are also statistically significant at the 1 percent level. It is evident that turnover (TURN) records an almost three-fold increase during the event date and slowly decays over the ten day event window implying volume or liquidity has shown marked changes after PWs. The changes in turnover over the pre and post-PW period is negative (0.518) and statistically significant at the 1 percent level implying that the turnover in the post-PW period is significantly higher than pre-PW period. Similarly, the abnormal relative volume (AVOL) has increased substantially during the post-PW period and the difference between pre and post is expectedly negative (0.337) and statistically significant at 1 percent level. Table 2 reports the results of the analysis based on daily measures of liquidity for the event window [10,0,+10] for the entire period. The post-announcement period is captured with a dummy variable of 1 and 0 otherwise with market-to-book value acting as the control variable. The empirical results reported in Table 3a shows that there is a negative relationship between a post-announcement variable and all liquidity measures. The slope coefficients for the post-announcement period (POST) are statistically significant at the 1 percent level and the overall fit of the equation is also good as measured by R2. We divided our sample into 5 portfolios of equal size based on the cumulative average abnormal returns CAAR [1,0,+1] using Fama-French three factor model. Table 3b shows the results of the portfolio with the largest negative cumulative average abnormal returns of the total sample (mean FF3 = 39.4%), whereas Table 3c shows the results of the portfolio with the lowest negative cumulative average abnormal returns of the total sample (mean = 2.1%). Our regression analysis based on extreme portfolios shows similar results-negative coefficient associated with post-announcement drift. However, the slope coefficient associated with the post-announcement dummy variable is higher for the extreme negative portfolio as compared with the positive portfolio. This holds for all three liquidity measures. This implies that improvement in market liquidity is more pronounced for extreme negative portfolios as compared with positive portfolios. This result confirms that the

21

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29 Table 1 Cumulative average abnormal returns using Fama-French Three Factor Model. Event Window

CAAR

p-value

CAAR CAAR CAAR CAAR

3.167*** 14.084*** 0.560 17.811***

0.00 0.00 0.04 0.00

[10,2] [1,0,+1] [+2,+10] [10,0,+10]

This table present the cumulative average abnormal returns using Fama-French three factor model (1993) as the return-generating process, where:

Rit ¼/ þbi Rmt þ ui SMBt þ li HMLt þ eit Rmt is the rate of return of the S&P 500 market index on day t; SMBt (Small Minus Big) is the average return on three small marketcapitalization portfolios minus the average return on three large market-capitalization portfolios; HMLt (High Minus Low) is the average return on two high book-to-market equity portfolios minus the average return on two low book-to-market equity portfolios; eit is a random variable. Data on the portfolios of SMBt and HMLt are obtained from the website of French. Note: ***, **, * indicate statistical significance at 1%, 5% and 10% respectively, two-sided tested.

Table 2 Descriptive statistics. Days

Amihud

QSpread

ESpread

TURN

AVOL

10 9 8 7 6 5 4 3 2 1 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10 D(pre-post)

0.1112 0.1100 0.1086 0.1135 0.1056 0.1055 0.1062 0.1050 0.1105 0.1122 0.1319 0.0926 0.0994 0.1096 0.1112 0.1109 0.1030 0.1020 0.1074 0.1019 0.1165 0.003**

0.1349 0.1402 0.1343 0.1431 0.1399 0.1413 0.1388 0.1360 0.1386 0.1309 0.1259 0.1234 0.1254 0.1243 0.1273 0.1235 0.1229 0.1248 0.1258 0.1238 0.1289 0.0128***

0.0886 0.0928 0.0942 0.0955 0.0950 0.0957 0.0903 0.0922 0.0944 0.0938 0.0968 0.0910 0.0912 0.0875 0.0860 0.0882 0.0858 0.0867 0.0883 0.0827 0.0882 0.006***

1.0366 1.0239 1.0390 1.0599 1.0778 1.0730 1.0774 1.0988 1.1311 1.3361 3.6296 2.5936 1.9555 1.7275 1.5873 1.4913 1.4196 1.3922 1.3749 1.3386 1.2995 0.518***

0.06537 0.04947 0.05604 0.05743 0.08944 0.07717 0.07606 0.09214 0.12274 0.26000 1.60646 0.97914 0.63975 0.51270 0.42591 0.37531 0.32568 0.29847 0.28418 0.25315 0.22365 0.337***

This table presents descriptive statistics of the liquidity measures and volume variables in the sample. The statistics are cross-sectional means averaged over each day. The sample contains 1945 profit warning of firms listed on NYSE, AMEX, and NASDAQ from 1995 to 2010. Data on liquidity are obtained from the CRSP database. Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each day. QSpread is the average ratio between the difference of the daily ask and bid quote as percentage of the daily ask quote. ESpread is the absolute effective spread. TURN is the average ratio between the daily volume and the average volume of days (t-11) to (t-31). The mean abnormal relative volume (AVOL) is estimated by conducting a volume event study using the methodology proposed by Campbell and Wasley (1996). The volume event study is conducted similar to an event study that uses returns. The main difference is that log-transformed relative volume replaces the returns for each security. Note: ***, **, * indicate statistical significance at 1%, 5% and 10% respectively, two-sided tested.

announcement of negative, unanticipated news ahead of the earnings announcement, measured by extreme negative CAARs, contains new information that reduces information asymmetry between management and investors and, as a result, improves market liquidity. The empirical analysis using panel regression method is conducted using (a) monthly and (b) daily data. Table 4 reports the panel regression results based on Eq. (8) for Amihud measure of illiquidity (ILLIQ) with monthly data (as opposed to analysis based on daily data reported in Tables 2 and 3). The empirical results corroborate the results derived from daily data analysis (Table 3). The relationship between post-announcement period (POST) and Amihud measure of illiquidity is negative (0.040) and statistically significant at the 1 percent level, which confirms our hypothesis that market liquidity improves during the post-announcement period. The relationship between relative firm size (SIZE) and Amihud measure of illiquidity is negative, which indicates that larger firms have higher market liquidity (lower illiquidity). In general, large firms have better analyst coverage and higher trading volumes and will show higher market liquidity. Similarly, the performance measure (ROA) is also negative with Amihud illiquidity measure which implies that firms with higher performance

22

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

Table 3a The impact profit warning announcements on daily liquidity measures. Variables

Amihud

Quoted Spread

Abs. Effective Spread

Intercept

0.512 (3.123) 0.006*** (0.001) 0.047 (0.106) 0.297*** (0.085) 0.034 (0.135) 0.174 (0.405) 0.003*** (0.001) 0.556 24.21*** 35,781

2.956 (2.310) 0.012*** (0.001) 0.037 (0.074) 0.112** (0.058) 0.020 (0.092) 0.240 (0.283) 0.005*** (0.001) 0.689 45.52*** 38,853

8.176*** (2.471) 0.005*** (0.001) 0.004 (0.080) 0.009 (0.062) 0.190* (0.101) 0.155 (0.306) 0.004*** (0.001) 0.324 10.59*** 37,956

Post-Announcement period Firm Size (SIZE) Return on Assets (ROA) Leverage (LEV) Cash Flows from Operations (CFO) Market-to-Book ratio (MTB) Adjusted R2 F-statistic Observations

This table shows panel data regressions (fixed effects) of the influence profit warning announcements on daily liquidity measures (Amihud, quoted spread, absolute effective spread) during the period of 10 days before and 10 days after profit warning announcements. Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each day. Quoted spread is the average ratio between the difference of the daily ask and bid quote as percentage of the daily ask quote in each month. Absolute effective spread is the absolute effective spread. The post-announcement period is a dummy variable (POST) that equals one for the post-announcement period and is zero otherwise. The variable firm size (SIZE) is defined as the relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement. The variable is scaled in the range [0,1]. The variable return on assets (ROA) is defined as earnings before interest and taxes (EBIT) as percentage of the book value of total assets. The variable leverage (LEV) is defined as total debt as percentage of total assets. The variable cash flow from operating activities (CFO) is defined as the cash flows from operating activities as percentage of total assets. Market-to-Book ratio (MTB) is defined as the market capitalization divided by total assets. Note: ***, **, * indicate statistical significance at 1%, 5% and 10% respectively, two-sided tested. Standard errors are in brackets.

Table 3b The impact profit warning announcements on daily liquidity measures. Portfolio with the largest negative CAARs (mean = 39.4%): Variable

Amihud

Quoted Spread

Abs. Effective Spread

Intercept

19.557* (10.383) 0.010** (0.005) 0.309 (0.293) 0.488*** (0.188) 0.023 (0.410) 0.218 (1.163) 0.003** (0.001) 0.450 14.32*** 6353

8.894 (5.828) 0.029*** (0.003) 0.141 (0.157) 0.107 (0.104) 0.082 (0.204) 0.409*** (0.001) 0.004*** (0.002) 0.587 29.50*** 7846

12,923** (5.982) 0.017*** (0.003) 0.050 (0.162) 0.051 (0.106) 0.137 (0.215) 0.577 (0.648) 0.004*** (0.001) 0.330 10.71*** 7599

Post-Announcement period Firm Size (SIZE) Return on Assets (ROA) Leverage (LEV) Cash Flows from Operations (CFO) Market-to-Book ratio (MTB) Adjusted R2 F-statistic Observations

This table shows panel data regressions (fixed effects) of the influence profit warning announcements on daily liquidity measures (Amihud, quoted spread, absolute effective spread) during the period of 10 days before and 10 days after profit warning announcements. The total sample is divided into 5 portfolios of equal size based on the cumulative average abnormal returns CAAR [1,0,+1] using Fama-French three factor model. This table shows the results of the portfolio with the largest negative cumulative average abnormal returns of the total sample (mean = 39.4%). Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each day. Quoted spread is the average ratio between the difference of the daily ask and bid quote as percentage of the daily ask quote in each month. Absolute effective spread is the absolute effective spread. The post-announcement period is a dummy variable (POST) that equals one for the post-announcement period and is zero otherwise. The variable firm size (SIZE) is defined as the relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement. The variable is scaled in the range [0,1]. The variable return on assets (ROA) is defined as earnings before interest and taxes (EBIT) as percentage of the book value of total assets. The variable leverage (LEV) is defined as total debt as percentage of total assets. The variable cash flow from operating activities (CFO) is defined as the cash flows from operating activities as percentage of total assets. Market-to-Book ratio (MTB) is defined as the market capitalization divided by total assets. Note: ***, **, * indicate statistical significance at 1%, 5% and 10% respectively, two-sided tested. Standard errors are in brackets.

23

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29 Table 3c The impact profit warning announcements on daily liquidity measures. Portfolio with the lowest negative CAARs (mean = 2.1%): Variable

Amihud

Quoted Spread

Abs. Effective Spread

Intercept

5.319 (5.719) 0.005* (0.002) 0.078 (0.202) 0.118 (0.203) 0.220 (0.242) 0.872 (0.845) 0.002 (0.003) 0.557 24.80*** 7255

6.956 (5.105) 0.007*** (0.002) 0.226 (0.160) 0.282* (0.168) 0.220 (0.194) 0.348 (0.662) 0.005** (0.002) 0.664 40.09*** 7653

18.303*** (5.405) 0.004 (0.002) 0.238 (0.174) 0.001 (0.178) 0.062 (0.213) 0.625 (0.719) 0.008*** (0.002) 0.305 9.521 7261

Post-Announcement period Firm Size (SIZE) Return on Assets (ROA) Leverage (LEV) Cash Flows from Operations (CFO) Market-to-Book ratio (MTB) Adjusted R2 F-statistic Observations

This table shows panel data regressions (fixed effects) of the influence profit warning announcements on daily liquidity measures (Amihud, quoted spread, absolute effective spread) during the period of 10 days before and 10 days after profit warning announcements. The total sample is divided into 5 portfolios of equal size based on the cumulative average abnormal returns CAAR [1,0,+1] using Fama-French three factor model. This Table shows the results of the portfolio with the lowest negative cumulative average abnormal returns of the total sample (mean = 2.1%). Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each day. Quoted spread is the average ratio between the difference of the daily ask and bid quote as percentage of the daily ask quote in each month. Absolute effective spread is the absolute effective spread. The post-announcement period is a dummy variable (POST) that equals one for the post-announcement period and is zero otherwise. The variable firm size (SIZE) is defined as the relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement. The variable is scaled in the range [0,1]. The variable return on assets (ROA) is defined as earnings before interest and taxes (EBIT) as percentage of the book value of total assets. The variable leverage (LEV) is defined as total debt as percentage of total assets. The variable cash flow from operating activities (CFO) is defined as the cash flows from operating activities as percentage of total assets. Market-to-Book ratio (MTB) is defined as the market capitalization divided by total assets. Note: ***, **, * indicate statistical significance at 1%, 5% and 10% respectively, two-sided tested. Standard errors are in brackets.

show a higher market liquidity. This variable is also statistically significant at the 1 percent level. Firms with larger cash flows from operations (CFO) also experienced lower illiquidity and this coefficient is statistically significant at the 1 percent level. The variable market-to-book value is positively related to measure of illiquidity but its coefficient is small. The model overall is significant. These results are invariant to the choice of other liquidity measures like QSpread and ESpread and are not reported in order to conserve space. We also performed panel regression models using daily liquidity measures and these are reported in Table 5. Table 5 reports results of the panel fixed effects model of determinants of Amihud measure of illiquidity taking the direction of the stock market impact as a threshold – firms experiencing positive stock market impacts compared with firms with negative stock market impacts. We divided our sample into five portfolios of equal size based on the cumulative average abnormal returns CAAR [1,0,+1] using Fama-French three-factor model. Table 5 shows the results of the portfolio with the largest negative cumulative average abnormal returns of the total sample (mean FF3 = 39.4%) and the results of the portfolio with the lowest negative cumulative average abnormal returns of the total sample (mean = 2.1%). The post-announcement effects as captured through the dummy variable (POST) were relatively higher for negative CAR firms compared with positive CAR firms. This result confirms our hypothesis that the releases of profit warnings reduce information asymmetry between managers and investors and improves market liquidity. This outcome is even more pronounced for firms with very negative abnormal returns. Similar coefficients for other firm-specific variables, such as firm size (SIZE), firm performance (ROA), leverage (LEV), cash flow from operations (CFO) and market to book ratio (MTB) coefficients were relatively higher for negative CAR firms compared with positive CAR firms. The overall fit of the fixed effects models are high (adj. R2 > 0.75) and the models overall are statistically significant at the 1 percent level. 5. Robustness tests To test the robustness of our empirical results, we analyzed whether the dot-com bubble period (2000–2001) and the financial crisis (2007–2008) characterized by low levels of return and high levels of volume turnover and volatility are driving our results. We ran separate panel data regressions for expansion and contraction periods within 1995–2010 to estimate the impact on our daily liquidity measures. The regression estimates show that the daily liquidity measures (ILLIQ) improved during the post-announcement period of profit warnings, which are consistent with the empirical results reported in our study. The coefficients of the post-announcement period are higher during the expansion periods (0.007) compared to contraction periods (0.005), which indicate that impact of profit warnings on market liquidity is higher during expansion period.

24

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29 Table 4 Panel data regression of profit warnings on Amihud illiquidity. Variable

Exp. Sign

Total Sample

Intercept

±

Post-Announcement period (POST)

()

Firm Size (SIZE)

()

Return on Assets (ROA)

()

Leverage (LEV)

±

Cash Flows from Operations (CFO)

()

Market-To-Book (MTB)

±

6.439*** (0.102) 0.040*** (0.008) 6.718*** (0.102) 0.512*** (0.165) 0.212*** (0.077) 0.643*** (0.095) 0.004*** (0.002) 0.804 47.17*** 24,016

Adjusted R2 F-statistic Observations

This table shows panel data regressions (fixed effects) of the influence of profit warning announcements on the monthly Amihud illiquidity measure during the period of 6 months before and 6 months after profit warning announcements. The sample contains firms listed on NYSE, AMEX, and NASDAQ. Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each month. The post-announcement period is a dummy variable (POST) that equals one for the period after the profit warning announcement and is otherwise zero. The variable firm size (SIZE) is defined as the relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement. The variable is scaled in the range [0,1]. The variable return on assets (ROA) is defined as earnings before interest and taxes (EBIT) as percentage of the book value of total assets. The variable leverage (LEV) is defined as total debt as percentage of total assets. The variable cash flow from operating activities (CFO) is defined as the cash flows from operating activities as percentage of total assets. The market-to-book ratio (MTB) is defined as the market capitalization divided by the book value of total assets. Note: ***, **, * indicate statistical significance at the 1, 5, 10 percent level, two-sided tested. Standard errors are in brackets.

Table 5 Panel data regressions of different CAR portfolios on Amihud illiquidity. Variable

Exp. Sign

Largest Negative CARs (mean = 39.4%)

Lowest Negative CARs (mean = 2.1%)

Intercept

±

Post-Announcement Period (POST)

()

Firm Size (SIZE)

()

Return on Assets (ROA)

()

Leverage (LEV)

±

Cash Flows from Operations (CFO)

()

Market-To-Book (MTB)

±

7.601*** (0.228) 0.110*** (0.033) 8.164*** (0.248) 0.540* (0.420) 0.442*** (0.171) 0.932*** (0.267) 0.015*** (0.006) 0.768 35.71*** 4600

6.348*** (0.219) 0.038*** (0.013) 6.689*** (0.214) 0.670** (0.375) 0.048 (0.172) 0.503*** (0.186) 0.006* (0.004) 0.828 55.23*** 4752

2

Adjusted R F-statistic Observations

This table shows panel data regressions (fixed effects) of the influence of profit warning announcements on the monthly Amihud illiquidity measure during the period of 6 months before and 6 months after profit warning announcements. The sample is divided into 5 portfolios of equal size based on the cumulative average abnormal returns CAAR [1,0,+1] using Fama-French three factor model. The panel data regressions show the results of the portfolio with the largest negative cumulative average abnormal returns of the total sample (mean = 39.43%) and the portfolio with the lowest negative cumulative average abnormal returns (mean = 2.1%). The sample contains firms listed on NYSE, AMEX, and NASDAQ. Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each month. The post-announcement period (POST) is a dummy variable that equals one for the 6-months period after the profit warning announcement and is otherwise zero. The variable firm size (SIZE) is defined as the relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement. The variable is scaled in the range [0,1]. The variable return on assets (ROA) is defined as earnings before interest and taxes (EBIT) as percentage of the book value of total assets. The variable leverage (LEV) is defined as total debt as percentage of total assets. The variable cash flow from operating activities (CFO) is defined as the cash flows from operating activities as percentage of total assets. The market-to-book ratio (MTB) is defined as the market capitalization divided by the book value of total assets. Note: ***, **, * indicate statistical significance at the 1, 5, 10 percent level, two-sided tested. Standard errors are in brackets.

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

25

The information content of profit warnings during expansion periods is much higher than during contraction periods. Investors do not expect profit warnings during expansion periods. Therefore, profit warnings reduce information asymmetry between managers and investors and increase market liquidity. Bessembinder and Kaufman (1997) suggested that bid-ask spreads are larger for NASDAQ-listed firms than for NYSElisted stocks. Gao and Ritter (2010) suggested that trading volume of NASDAQ shares needed to be adjusted for institutional features. Therefore, we computed the ratio between the daily volume and the average volume of days (t-11) to (t-31). In addition, we tested our panel regressions for different stock exchanges separately (NYSE, NASDAQ, and AMEX). We could not find significant differences in market liquidity between these stock exchanges. We also ran the panel regressions for quoted spread and absolute and effective spreads. The results remain the same. And, we performed a robustness test to determine whether Sarbanes-Oxley (SOX) regulation would impact our empirical results. We ran separate panel regressions for the period before SOX and the period after the implementation of SOX. Our results show that market liquidity significantly improved after the announcement of profit warnings regardless of SOX regulations. 5.1. Endogeneity tests In order to test whether profit warning are endogenous, we also conducted a multivariate analysis (Eq. (8)) using 2SLS to account for the possibility of endogeneity and are presented in Table 7. The empirical results report in Table 7 indicates (as in Table 6) the post-announcement period dummy (POST) negatively impacts measure of illiquidity (ILLIQ). Moreover, we further conducted causality tests using the instrument that explain the decision to disclose PWs. We chose measure of performance (ROA) as a potential instrument and conducted a bi-variate Granger causality of ROA on the measure of illiquidity (ILLIQ) as was done in by Næs, Skjeltorp, and Ødegaard (2011). The results indicate that ROA Granger cause ILLIQ but not otherwise (no reverse causality from ILLIQ to ROA). The results rule out the possibility of endogeneity in our panel regression analysis. Further, we conducted a comparative examination of firms that did not report profit warnings but surprise the market with negative earnings during regular earnings announcement dates. We obtained actual and forecast earnings per share (EPS) for the years 1995 through 2010 from the Institutional Brokers Estimate System (IBES). We measured the standardized unanticipated earnings surprises (SUE) as the difference between actual earnings minus the mean earnings analysts’ forecast (analyst consensus) divided by the standard deviation of the earnings analysts’ forecast. Our control sample consists of firms

Table 6 Panel data regressions of Amihud illiquidity during business cycles. Variable

Expansion Period

Contraction Period

Intercept

1.501 (3.614) 0.007*** (0.002) 0.093 (0.129) 0.307*** (0.109) 0.118 (0.162) 0.213 (0.477) 0.003*** (0.001) 0.583 24.27*** 28,048

2.069*** (6.090) 0.005* (0.003) 0.089 (0.187) 0.254* (0.136) 0.234 (0.241) 0.027 (0.744) 0.013*** (0.003) 0.577 28.70*** 8227

Post-Announcement Period (POST) Firm Size (SIZE) Return on Assets (ROA) Leverage (LEV) Cash Flows from Operations (CFO) Market-To-Book (MTB) Adjusted R2 F-statistic Observations

This table shows panel data regressions (fixed effects) of the influence of profit warning announcements on daily Amihud illiquidity measure during the period of 10 days before and 10 days after profit warning announcements. Data on expansion and contraction periods are obtained from NBER (National Bureau of Economic Research). The contraction periods are March 2001 – October 2001 and December 2007 – May 2009. Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each day. The Post-Announcement Period is a dummy variable that equals one for the post-announcement period. The variable firm size (SIZE) is defined as the relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement. The variable is scaled in the range [0,1]. The variable return on assets (ROA) is defined as earnings before interest and taxes (EBIT) as percentage of the book value of total assets. The variable leverage (LEV) is defined as total debt as percentage of total assets. The variable cash flow from operating activities (CFO) is defined as the cash flows from operating activities as percentage of total assets. The market-to-book ratio (MTB) is defined as the market capitalization divided by the book value of total assets. Note: ***, **, * indicate statistical significance at the 1, 5, 10 percent level, two-sided tested. Standard errors are in brackets.

26

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29 Table 7 Panel two-stage least squares (2 SLS) of profit warnings on Amihud illiquidity. Variable

Exp. Sign

2 SLS

Intercept

±

Post-Announcement period (POST)

()

Firm Size (SIZE)

()

Return on Assets (ROA)

()

Leverage (LEV)

±

Cash Flows from Operations (CFO)

()

Market-To-Book (MTB)

±

55.415* (32.363) 0.112* (0.074) 0.565* (0.309) 0.454* (0.314) 0.021 (0.037) 0.469 (1.474) 0.250 (0.183) 49.76*** 22,123

F-statistic Observations

This table shows the 2 SLS regression of the influence of profit warning announcements on the monthly Amihud illiquidity measure during the period of 6 months before and 6 months after profit warning announcements. Instrument variables are dependent and independent variables lagged by one period. The sample contains firms listed on NYSE, AMEX, and NASDAQ. Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each month. The post-announcement period is a dummy variable that equals one for the period after the profit warning announcement and is otherwise zero. The variable firm size (SIZE) is defined as the relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement. The variable is scaled in the range [0,1]. The variable return on assets (ROA) is defined as earnings before interest and taxes (EBIT) as percentage of the book value of total assets. The variable leverage (LEV) is defined as total debt as percentage of total assets. The variable cash flow from operating activities (CFO) is defined as the cash flows from operating activities as percentage of total assets. The market-to-book ratio (MTB) is defined as the market capitalization divided by the book value of total assets. Note: ***, **, * indicate statistical significance at the 1, 5, 10 percent level, two-sided tested. Standard errors are in brackets.

Table 8 Market liquidity of profit warnings vs negative earnings announcements. Window

Profit Warning Announcements

Negative Earnings Announcements

AIM [1] AIM [3] AIM [10]

17.469 7.965 3.106

1.147 4.670 2.978

This table shows the daily Amihud illiquidity measure (AIM) around profit warning announcements and negative earnings announcements. Amihud is the average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each day. AIM [1] is the average AIM(t+1) one day after the announcement minus the average AIM(t-1) one day before the announcement divided by the average AIM(t-1) one day before the announcement as percentage. AIM [3] is the average AIM over the period of one day through three days after the announcement minus the average AIM one day through three days before the announcement divided by the average AIM one through three days before the announcement as percentage. AIM [10] is the average AIM one day through 10 days after the announcement minus the average AIM one through 10 days before the announcement divided by the average AIM one through 10 days before the announcement as percentage.

with the most negative SUE score (n = 1945). We calculated the Amihud illiquidity Measure (ILLIQ) around earnings announcement dates. Table 8 reports the comparative results of our robustness test. First, we estimated the difference of the mean ILLIQ one day after the negative earnings announcement minus the mean ILLIQ one day before the negative earnings announcement divided by the mean ILLIQ one day before the negative earnings announcement as percentage. We followed the same procedure for profit warnings. Table 8 shows that ILLIQ is 17.469 for firms that warn and 1.147 for firms that do not warn. The market liquidity is significantly higher for firms that announce a profit warning ahead of the regular earnings announcement date. We repeated this procedure for different time horizons. The results for the average AIM of three days before and three days after the earnings/profit warning announcement day show that the market liquidity is almost twice (8.0% vs 4.7%) for firms that announce profit warnings. The difference slightly disappears for longer time horizons (10 days), but the liquidity is still higher for firms that warn compared to firms that do not warn. To measure information asymmetry, we computed an event study on the trading volume using Campbell and Wasley (1996) methodology. Our control sample consists of firms with the most negative SUE score (n = 1945). Table 9 reports

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

27

Table 9 Trading volume of profit warnings vs negative earnings announcements. Window

Profit Warning Announcements

Negative Earnings Announcements

MARV[0] MARV[10,+10] MARV[3,+3] MARV[10,1] MARV[+1,+10]

1.606*** 6.870*** 4.213*** 0.946*** 4.317***

0.513*** 2.520*** 1.928*** 0.149* 2.155***

The cumulative mean abnormal relative volume (CMARV) is estimated by conducting a volume event study using the methodology proposed by Campbell and Wasley (1996). The volume event study is conducted similar to an event study that uses returns. The main difference is that logtransformed relative volume replaces the returns for each security. Note: ***, **, * indicate statistical significance at the 1, 5, 10 percent level, two-sided tested.

Fig. 1. Trading volume around profit warning announcements.

the mean abnormal relative trading volume (MARV) for firms that announce profit warnings and firms with negative earnings surprises at earnings announcement dates using different windows. The results show that the mean abnormal trading relative volume (MARV) is higher for PW announcements compared to negative earnings surprises at earnings announcements dates for all event windows. The results are statistically significant. Fig. 1 shows that the mean abnormal relative trading volume peaks on the announcement day and stays high during the post-announcement period. Fig. 1 shows similar results for firms that surprise investors with negative earnings without warning, but the mean abnormal relative trading volume is much lower than for firms that warn. These results lend support to the hypothesis that the announcement of profit warnings reduces the information asymmetry between management and investors. 6. Conclusions This study examines the impact of voluntary disclosure (profit warnings) on market liquidity using a large cohort of PWs in the United States during 1995–2010. The empirical results presented show evidence that voluntarily issuing PWs leads to enhanced market liquidity although the stock markets impacts are generally negative. These results are invariant to the measures of liquidity adopted for empirical investigation. We also show empirical evidence by using Amihud’s measure of illiquidity, quoted spread and effective spread as well as trading volume. Our analysis (using both daily and monthly data), controlling for firm-specific attributes and analyzing the impact of PWs in the post-announcement period, shows clearly that it improves liquidity (reducing Amihud measure of liquidity) and increases trading volume. Enhanced disclosures such as PWs lead to a reduction in informational asymmetries and thereby enable investors to trade intensively and strategically. This improves the market liquidity, which is a positive development for the stock market. The results have major public policy implications that firms voluntarily disclosing potentially negative news may experience improvement in market liquidity, thereby potentially enjoying a lower cost of equity and capital. Although the stock market impacts are negative in general, the firms experience higher liquidity. In other words, bad news such as profit warnings is good from the point of view of market liquidity.

28

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

Appendix 1. List of variables used in the study. Variable

Description

Amihud

Average illiquidity measure of the daily ratio of absolute stock return to its dollar volume, averaged over each day; Average ratio between the difference of the daily ask and bid quote as percentage of the daily ask quote; Absolute effective spread; Average ratio between the daily volume and the average volume of days (t-11) to (t-31); The post-announcement period is denoted by a dummy variable that equals one for the post-announcement period and is zero otherwise; Relative rank of market capitalization based on all firms listed on NYSE, AMEX, and NASDAQ in the year of the announcement; Earnings before interest and taxes (EBIT) as percentage of the book value of total assets; Total debt as percentage of total assets; Cash flows from operating activities as percentage of total assets; Market capitalization divided by the book value of total assets.

QSpread ESpread TURN POST SIZE ROA Lev CFO MTB

References Aitken, M., & Comerton-Forde, C. (2003). How liquidity should be measured? Pacific-Basin Finance Journal, 11(1), 45–59. Amihud, Y. (2002). Illiquidity and Stock Returns: Cross section and time series effects. Journal of Financial Markets, 5(1), 31–56. Baginski, S. P., Hassell, J., & Kimbrough, M. (2002). The effect of legal environment on voluntary disclosure: Evidence from management earnings forecasts issued in U.S. and Canadian markets. Accounting Review, 77(1), 25–50. Balakrishnan, K., Billings, M. B., Kelly, B., & Ljungqvist, A. (2014). Shaping liquidity: On the causal effects of voluntary disclosure. Journal of Finance, 69(5), 2237–2278. Bamber, L. (1987). Unexpected earnings, firm size and trading volume around quarterly earnings announcements. Accounting Review, 62(3), 510–532. Bardos, K. S. (2011). Quality of financial information and liquidity. Review of Financial Economics, 20(2), 49–52. Bessembinder, H., & Kaufman, H. (1997). A comparison of trade execution costs for NYSE and NASDAQ-listed stocks. Journal of Financial and Quantitative Analysis, 32(3), 287–310. Brown, S. J., & Warner, J. B. (1985). Using daily stock returns; the case for event studies. Journal of Financial Economics, 14(1), 3–31. Bulkley, G., & Herrerias, R. (2005). Does the precision of news affect market under-reaction? Evidence from returns following two classes of profit warnings. European Financial Management, 11(5), 603–624. Campbell, C. J., & Wasley, C. E. (1996). Measuring abnormal daily trading volume for samples of NYSE/ASE and NASDAQ securities using parametric and nonparametric test statistics. Review of Quantitative Finance and Accounting, 6(3), 309–326. Campbell, J. Y., Lo, A. W., & Mackinlay, A. C. (1997). The econometrics of financial markets. Princeton, New Jersey: Princeton University Press. Chae, J. (2005). Trading volume, information asymmetry and timing information. Journal of Finance, 60(1), 413–442. Chen, S., Matsumoto, D. A., & Rajagopal, S. (2011). Is silence golden? An empirical analysis of firms that stop earnings guidance. Journal of Accounting and Economics, 51(1–2), 134–150. Chordia, T., Huh, S. W., & Subramanyam, A. (2009). Theory-based Illiquidity and Asset Pricing. Review of Financial Studies, 22(9), 3629–3668. Chordia, T., Roll, R., & Subramanyam, A. (2000). Commonality in liquidity. Journal of Financial Economics, 56(1), 3–28. Church, M., & Donker, H. (2009). Profit warnings: Will openness be rewarded? Applied Economic Letters, 17(7), 633–637. Copeland, T. C., & Galai, D. (1983). Information effect on the bid-ask spread. Journal of Finance, 38(5), 1457–1469. Darrough, M. N., & Stoughton, N. M. (1990). Financial disclosure policy in an entry game. Journal of Accounting and Economics, 12(1–3), 219–243. De Jong, F., & Rindi, B. (2009). The microstructure of financial markets. New York: Cambridge University Press. Diamond, D. W. (1985). Optimal release of information by firms. Journal of Finance, 40(4), 1071–1094. Diamond, D. W., & Verrecchia, R. E. (1991). Disclosure, liquidity and the cost of capital. Journal of Finance, 46(4), 1325–1359. Dontoh, A. (1989). Voluntary disclosure. Journal of Accounting, Auditing and Finance, 4(4), 480–511. Dutta, S., & Trueman, B. (2002). The interpretation of information and corporate disclosure strategies. Review of Accounting Studies, 17(1), 75–96. Dye, R. (1998). Investor sophistication and voluntary disclosures. Review of Accounting Studies, 3, 261–287. Dye, R. A., & Sridhar, S. (2002). Resource allocations effects of price reactions to disclosures. Journal of Accounting Research, 19(3), 385–410. Dyl, E. A. (1999). Estimating economic damages in class action securities fraud litigation. Journal of Forensic Economics, 25(1), 1–11. Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56. Feldman, R., Govindaraj, S., Livnat, J., & Segal, B. (2010). Management’s tone change, post earnings announcement drift and accruals. Review of Accounting Studies, 15(4), 915–953. Field, L., Lowery, M., & Shu, S. (2005). Does disclosure deter or trigger litigation? Journal of Accounting and Economics, 39(3), 487–507. Gao, X., & Ritter, J. R. (2010). The marketing of seasoned equity offerings. Journal of Financial Economics, 97(1), 33–52. Glosten, L. R., & Milgrom, P. R. (1985). Bid, ask and transaction prices in a specialist market with heterogeneously informed traders. Journal of Financial Economics, 14(1), 71–100. Goyenko, R. Y., Holden, C. W., & Trzcinka, C. A. (2009). Do liquidity measures measure liquidity? Journal of Financial Economics, 92(2), 153–181. Healy, P. M., Hutton, A. P., & Palepu, K. G. (1999). Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research, 16(3), 485–520. Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical literature. Journal of Accounting and Economics, 31(1–3), 405–440. Heflin, F. L., Shaw, K. W., & Wild, J. J. (2005). Disclosure policy and market liquidity: Impact of depth quotes and order sizes. Contemporary Accounting Research, 22(4), 829–865. Jackson, D., & Madura, J. (2003). Profit warnings and timing. Financial Review, 38(4), 497–513. Jackson, D., & Madura, J. (2004). Profit warning and signaling. Managerial Finance, 30(9), 20–31. Jackson, D., & Madura, J. (2007). Impact of regulation fair disclosure on the information flow associated with profit warnings. Journal of Economics and Finance, 31(1), 59–74.

A. Dayanandan et al. / North American Journal of Economics and Finance 40 (2017) 16–29

29

Jensen, M. C. (2005). Agency costs of overvalue equity. Financial Management, 34(1), 5–19. Jensen, M. C., & Meckling, W. H. (1976). Theory of firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360. Johnson, W. B., & Zhao, R. (2012). Contrarian share price reactions to earnings surprises. Journal of Accounting Auditing & Finance, 27(2), 236–266. Kau, J. B., Linck, J. S., & Rubin, P. H. (2008). Do managers listen to the market? Journal of Corporate Finance, 14(4), 347–362. Kim, O., & Verrecchia, R. (1991). Trading volume and price reactions to public announcements. Journal of Accounting Research, 29(2), 302–321. Kothari, S. P, & Warner, J. B. (2005). Handbook of corporate finance: Empirical corporate finance chapter, Handbook in Finance Series, North Holland: Elsevier. Kyle, A. A. (1985). Continuous auctions and insider trading. Econometrica, 53(6), 1315–1335. Langberg, N., & Sivaramakrishnan, K. (2010). Voluntary disclosure and analyst feedback. Journal of Accounting Research, 48(3), 603–646. Leland, H. E., & Pyle, D. H. (1977). Informational asymmetries, industrial structure and financial intermediation. Journal of Finance, 32(2), 371–387. Leuz, C., & Verrecchia, R. E. (2000). The economic consequences of increased disclosure. Journal of Accounting Research, 38(Supplement), 91–124. Levi, S., & Zhang, X. J. (2014). Do temporary increases in information asymmetry affect the cost of equity. Management Science, 61(2), 354–371. Libby, R., & Tan, H. (1999). Analysts’ reactions to warnings of negative earnings surprises. Journal of Accounting Research, 37(2), 415–435. Luo, Y. (2005). Insiders learn from outsiders? Evidence from corporate takeovers. Journal of Finance, 60(4), 1951–1982. MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of economic literature, 35(1), 13–39. Mikhail, M. B., Walther, B. R., & Willis, R. H. (2004). Earnings surprises and the cost equity capital. Journal of Accounting, Auditing, and Finance, 19(4), 491–513. Myers, S. C., & Majluf, N. S. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13(2), 187–221. Næs, R., Skjeltorp, J. A., & Ødegaard, B. A. (2011). Stock market liquidity and the business cycle. Journal of Finance, 66(1), 139–176. Skinner, D. (1994). Why firms voluntarily disclose bad news? Journal of Accounting Research, 32(1), 38–60. Skinner, D. (1997). Earnings disclosures and stockholder lawsuits. Journal of Accounting and Economics, 23(3), 249–282. Soffer, L., Thiagarajan, S. R., & Walther, B. (2000). Earning pre-announcement strategies. Review of Accounting Studies, 5(1), 5–26. Stiglitz, J. E., & Weiss, A. (1981). Credit rationing in markets with imperfect information. American Economic Review, 71(3), 393–410. Suijs, J. (2007). Voluntary disclosure of information when firms are uncertain of investor response. Journal of Accounting and Economics, 41(2–3), 391–410. Verrecchia, R. (2001). Essays on disclosure. Journal of Accounting and Economics, 32(1–3), 97–180. Welker, M. (1995). Disclosure policy, information asymmetry and liquidity in equity markets. Contemporary Accounting Research, 11(2), 801–827. Xu, W. (2008). Market reactions to warnings of negative earnings surprises: Further evidence. Journal of Business, Finance and Accounting, 35(7–8), 818–836.