Does Trading Volume Increase Or Decrease Prior To Earnings Announcements?

Sangwan Kim
University of Massachusetts – Boston

Steve C. Lim
Texas Christian University – M.J. Neeley School of Business

July 16, 2016

Abstract:

This paper reports two empirical regularities regarding trading volume prior to earnings announcements. The literature suggests that discretionary liquidity traders postpone their equity trading until firms publicly announce earnings due to high information asymmetry before anticipated information events. Our first finding is that pre-announcement trading volume increases for firms with high analyst coverage. Our second finding is that preannouncement trading volume decreases for firms with low analyst coverage and trading volume prior to bad news is lower than good news earnings announcements. Our findings suggest that the intensity of analyst activity and the nature of mandatory earnings news jointly determine the direction and magnitude of pre-announcement trading volume. We contribute to the literature by showing that analysts’ information discovery (temporarily pushed back trading demand) prior to earnings announcements may understate (overstate) the magnitude of a short-window trading volume reaction to earnings announcements as measures of information content for firms with high (low) analyst coverage.

Does Trading Volume Increase Or Decrease Prior To Earnings Announcements? – Introduction

Adequate trading volume is essential for the proper functioning of capital markets. The extent to which investors trade shares enables stock prices to reflect available information, market participants to share financial risks, and corporations to raise external funds needed for operating activities. Moreover, volume-based studies have the potential to yield valuable new insights into the information content of earnings announcements because volume reactions to earnings announcements capture changes in the expectations of individual investors, while stock price reaction reflects changes in the expectations of the market as a whole (Bamber, Barron, & Stevens, 2011)

While numerous prior studies offer ample evidence of abnormally high trading volume on and after earnings announcements, only a few studies provide empirical evidence on how investors make equity trading decisions prior to upcoming earnings announcements. In this paper, we investigate how the intensity of analyst activity and the positive vs. negative earnings surprises affect trading volume changes prior to earnings announcements. Using a large sample of U.S. public firms from 1994 to 2013, we find that the timely release of preemptive analysts’ reports and the asymmetric effect of positive vs. negative earnings surprises on preannouncement trading volume (i.e., the tendency of lower pre-announcement trading volume prior to earnings announcements of bad news than good news) contribute to the trajectory of preannouncement trading volume.

Specifically, we find that trading volume in the pre-announcement period increases for firms with a high analyst following but it decreases for firms with a low analyst following. We attribute these results to the differential level of private information discovery performed by analysts before earnings announcements associated with analyst coverage. For example, in high analyst coverage firms, more frequent analysts’ reports preempt a portion of subsequent earnings announcements and the upcoming earnings information becomes partially available to investors in the pre-earnings announcement period (Malatesta & Thompson, 1985). In this case, financial analysts bring forward the upcoming earnings announcements and traders respond to the information, leading to an increase in pre-announcement trading volume. Whether the availability of other information sources (e.g., analyst reports) reduces the relative importance of scheduled corporate announcements is an empirical question and prior research provides mixed results. For example, Ivkovic and Jegadeesh (2004) and Chen, Cheng, and Lo (2010) suggest that investors value analysts’ information discovery role more than the information interpretation role before earnings announcements. Livnat and Zhang (2012) examine corporate announcements beyond scheduled earnings announcements (e.g., 8-Ks) and find that investors value analysts’ public information interpretation role more than the private information discovery role. Francis, Schipper, and Vincent (2002) and Frankel, Kothari, and Weber (2006) find that information content of analyst reports and that of subsequent earnings announcements complement each other, suggesting that analyst research does not necessarily preempt the anticipated information events. Kim and Song (2015) suggest that the literature (e.g., Ivkovic & Jegadeesh, 2004) is likely to have overstated the importance of analysts’ information discovery role once the effect of management earnings forecasts is controlled for. Huang, Lehavy, Zang, and Zheng (2015) find that both information discovery and interpretation roles co-exist in analyst reports around a firm’s conference calls and that several economic determinants affect the relative importance of new information discovery vs. public information interpretation roles.

Earnings Announcements, Trading Volume

In contrast, in low coverage firms, analysts’ information production and dissemination activity is relatively weak, which makes the pre-announcement information environment less rich for these firms. Moreover, analysts release equity reports less frequently prior to earnings announcements for firms followed by fewer analysts. In this case, informed traders would attempt to exploit their private information before earnings announcements, and discretionary liquidity traders may temporarily suspend their equity trading until the earnings announcement resolves the information asymmetry among investors. Thus, consistent with Chae (2005), we expect a decrease in trading volume prior to earnings announcements of firms with low analyst coverage.

Recently, Park, Lee, and Song (2014) find the asymmetric effect of positive vs. negative earnings surprises on pre-announcement trading volume with Korean data. These authors attribute the difference in trading volume pattern to the asymmetric discretionary disclosure of corporate managers with private information about upcoming earnings announcements (Verrecchia, 1983; Trueman, 1997). This line of accounting research shows that managers tend to voluntarily disclose good news early and bad news late (Lev & Penman, 1990; Kothari, Shu, & Wysocki, 2009; Roychowdhury & Sletten, 2012). The asymmetric disclosure incentives of managers make the pre-announcement information asymmetry relatively low for good news firms and relatively high for bad news firms. Therefore, discretionary liquidity traders postpone more of their equity trading until a firm’s earnings announcement resolves the information asymmetry in the case of bad news announcements (Chae, 2005; Park et al., 2014).

In our cross-sectional tests using analyst coverage, we find that the information discovery role of analysts before earnings announcements has important implications for the asymmetric effect of earnings surprises on the pre-announcement trading volume. Specifically, in high analyst coverage firms, trading volume increases in the pre-announcement period for both good news and bad news. These results suggest that when more analysts discover and publish material private information before earnings announcements, analyst reports tend to preempt the information content of both positive and negative earnings surprises. Moreover, for the high coverage firms, the difference in trading volume between positive and negative earnings surprises is statistically insignificant, suggesting that the preemptive disclosure of analysts’ private information dominates the asymmetric earnings surprise impact on trading volume. In contrast, in low analyst coverage firms, we find that trading volume decreases in the pre-announcement period for both good news and bad news, but the bad news firms tend to experience a sharper decrease in trading volume than their good news counterparts. Thus, the asymmetric effect of positive vs. negative earnings surprises on trading volume dominates the analysts’ information discovery role in the low coverage firms. All of our inferences are robust across three alternative metrics of quarterly earnings surprises captured by either standardized unexpected earnings (SUE), analyst consensus forecast-based unexpected earnings (AUE), or price-based unexpected earnings (PUE). In addition, the role of analyst

1, 2  - View Full Page