Ebook Information Asymmetry around Earnings Announcements during the Financial Crisis
Neoclassical theories in information economics model public information and private information as substitutes and find that earnings announcements reduce information asymmetry by reducing the informational advantage of informed traders. However, recent theoretical models find that earnings can increase information asymmetry by providing informed traders with an ability to better interpret the earnings announcement than uninformed traders. In a recent study, Ng, Verrecchia and Weber (2009) show that earnings that are not informative about underlying volatility increase information asymmetry. We use the financial crisis as an experimental setting which resulted in a shock to firms’ underlying volatility and examine whether earnings decrease information asymmetry by providing information about underlying volatility or whether they exacerbate information asymmetry by providing some traders with a better ability to interpret these announcements.
Financial reporting for banks has come under increased scrutiny during the recent financial crisis and has been subject to intense criticism. Motivated by concerns in the media and the regulators about the failure of accounting information to provide timely information about bank performance, we examine changes in information asymmetry around bank earnings announcements during the financial crisis period. Several factors have been argued to have aggravated the crisis, such as executive compensation schemes which encouraged bank managers to move into more opaque and risky transactions and the lack of relevant, reliable and understandable financial reporting information by firms (Rajan (2006), Ryan (2008), OECD Report (2009), Schwarcz (2009)). We investigate the role of exposure to sub-prime assets, equity-based managerial incentives, risk-management disclosure, bank size, loss recognition and fair value hierarchy information in alleviating or exacerbating information asymmetry around bank earnings announcements.
Using bid-ask spreads to measure information asymmetry, we compute changes in spreads after earnings announcements and compare these changes during the crisis period with those in the non-crisis period. Following Laeven and Huizinga (2009), we use the period from 2002 to denote the non-crisis period, as 2001 marked the end of a business cycle. As banks were widely believed to have initiated the financial crisis in 2007 which then spread to the entire economy, we examine changes in bid-ask spreads for 2007 and 2008 separately and also differentiate between banks and industrial firms using an indicator variable. We use firm-fixed effects in all our specifications thus enabling us to use a differences-in-differences approach (see Bertrand, Duflo and Mullainathan (2004)).
Our empirical results provide several interesting insights. While bid-ask spreads increase after earnings announcements for all firms in the non-crisis period, these increases are pronounced during 2007. In particular, the increase in spreads in 2007 is higher by 10.5% relative to the non-crisis period for industrial firms. Further, consistent with the prominent role of banks in the advent of the financial crisis, we find that increases in spreads in 2007 are 13.1% higher for banks relative to the non-crisis period. These results suggest that earnings announcements in 2007 provide informed traders with an increased ability to interpret the earnings relative to other market participants and increase information asymmetry.
Interestingly, we find that earnings announcements in 2008 decrease information asymmetry relative to earlier periods and that this decrease exists only for banks and not for industrial firms. In particular, we find that bank earnings announcements reduce spreads by around 11% in 2008 relative to other periods. In subsequent tests, we find that these decreases are driven by large banks which are more likely to be the focus of regulatory intervention. In contrast, small banks experience increases in bid-ask spreads in 2008. We find that large banks recognize more unrealized losses on securities available for sale, realize losses on securities available for sale and have higher levels of loan loss provisions. These results are consistent with the political cost hypothesis of Watts and Zimmerman (1978, 1986), and suggest that regulatory interest in bank financial statements in 2008 is associated with greater loss recognition and higher loss provisions which result in lower information asymmetry after earnings announcements.
Download
PDF Ebook Information Asymmetry around Earnings Announcements during the Financial Crisis
Posted in :
