A fundamental role of accounting information in financial markets is to serve as a basis for capital allocation. As a result, the determinants and consequences of the quality of accounting information are of interest to investors, managers, regulators and standard setters. Arthur Levitt, former Chairman of the Securities and Exchange Commission (SEC), has remarked that “an important benefit of high quality accounting standards is improved liquidity and lower cost of capital.” This issue has acquired even greater prominence in recent times as increased information asymmetry due to lack of financial reporting transparency likely contributed to the near break-down in the interbank lending market in 2008. Dechow and Schrand (2004) call for academic research to better understand the costs and consequences of poor quality earnings. Although, a handful of recent papers have investigated the adverse consequences of poor earnings quality, the evidence remains controversial and incomplete. In this study, we investigate the relation between earnings quality and the information asymmetry among market participants during earnings announcement and non-announcement periods.
There is a significant debate in the theoretical literature on how earnings quality may affect asset prices. Various analytical models suggest different paths through which earnings quality might affect a firm’s cost of capital. Figure 1 illustrates these possible links. In the model proposed by Lambert, Leuz and Verrecchia (2007a), earnings quality affects a firm’s cost of capital via its impact on the firm’s CAPM beta as accounting information has a direct effect on the assessed covariance of a firm’s cash flows with cash flows of other firms in the economy (Linkage 1). By contrast, Easley and O’Hara (2004) propose a framework in which information asymmetry induced by poor earnings quality may give rise to a non-diversifiable “information risk,” as less informed investors are always at a disadvantage relative to informed investors in adjusting their portfolio weights (Linkage 2). Consequently, investors would expect higher returns from a firm with high information asymmetry.
However, Hughes, Liu and Liu (2007) question the validity of this link by arguing that the pricing effect obtained in the Easley and O’Hara’s framework is driven primarily by under-diversification and would disappear in large economies. Further, Lambert, Leuz and Verrecchia (2007b) argue that the level of information precision (the degree of uncertainty about future payoffs), and not information asymmetry, affects expected returns. We posit that earnings quality can affect the cost of capital due to its effect on information asymmetry and firm liquidity (Linkage 3). Analytical models (e.g., Kyle, 1985; Glosten and Milgrom, 1985) predict that information asymmetry increases the adverse selection risk for liquidity providers, who respond by widening the spread and lowering the quoted depth, thereby decreasing liquidity. Prior empirical research provides evidence that higher liquidity cost generally translates into higher cost of capital. Thus, earnings quality can affect cost of capital via its impact on information asymmetry and liquidity cost.
Although two of the three linkages described above posit that earnings quality can impact information asymmetry in financial markets, there is no direct empirical evidence in the literature, at least to our knowledge, on this important linkage. Using a large sample of NYSE and NASDAQ firms, we document that poor earnings quality is significantly and incrementally (over the known determinants of trading costs) associated with higher information asymmetry. Our primary measure of earnings quality is the accruals quality metric used by FLOS (2004, 2005) and the direct measure of information asymmetry is the component of the bid-ask spread that is associated with adverse selection risk (price impact). Our results are robust to several alternative measures of earnings quality and information asymmetry. Furthermore, when we decompose earnings quality into innate (associated with the firm’s operations and business model) and discretionary (associated with managerial reporting choices) components, we find that each component has significant and incremental impact on information asymmetry.
Lee, Mucklow and Ready (1993) find that information asymmetry increases around earnings announcements. In recent work, Rees and Thomas (2008) and Rajgopal and Venkatachalam (2008) document that return volatility increases around earnings announcements. These results prompt us to probe further whether the change in information asymmetry around earnings announcement (relative to the non-announcement period) is associated with earnings quality. After controlling for firm characteristics and the magnitude of earnings surprise, we find that firms with relatively poor earnings quality experience a greater increase in information asymmetry around earnings announcement. Further earnings quality has a more pronounced impact for firms where earnings represent the principal source of information. These results suggest that poor earnings quality exacerbates the uncertainty surrounding earnings news and provides greater advantage to informed traders.
Download
PDF Ebook Earnings Quality and Information Asymmetry
