Accounting quality, information risk and implied volatility around earnings announcements
ABSTRACT
We examine the impact of accounting quality, used as a proxy for information risk, on the behavior of equity implied volatility around quarterly earnings announcements. Using US data during 1996–2010, we observe that lower (higher) accounting quality significantly
relates to higher (lower) levels of implied volatility (IV) around announcements. Worse accounting quality is further associated with a significant increase in IV before announcements, and is found to relate to a larger resolution in IV after the announcement has taken place. We interpret our findings as indicative of information risk having a significant impact on implied volatility behavior around earnings announcements.
1. Introduction
Past research has associated earnings quality with idiosyncratic return volatility (Rajgopal and Venkatachalam, 2011). This is because uncertainty about the future profitability of firms is expected to influence their stock return volatility (Pastor and Veronesi, 2003; Wei and Zhang, 2003), while the quality of earnings has been considered as a proxy for so-called ‘information risk’ or ‘information uncertainty’. This risk refers to the likelihood for firm-specific information important for investor decisions to be of poor quality (Francis et al., 2005), or the degree to which corporate value can be reasonably estimated by the most knowledgeable investors at an acceptable cost (Jiang et al.,
2005). In this direction, Rajgopal and Venkatachalam (2011) distinguish between sources of uncertainty about the future profitability of firms i.e., uncertainty about future cash flows from an operating point of view, vs. information about future cash flows stemming from the quality of accounting information, and confirm that lower earnings quality is associated with higher idiosyncratic stock market volatility, even after controlling for volatility in firm operating performance.
In this paper, we examine the association between financial reporting quality, measured by assessing the quality of accounting accruals, and levels and changes in equity option implied volatility around quarterly earnings announcements. We employ accruals quality as a proxy for firm information risk, in accordance with past research (Francis et al., 2005; Ecker et al., 2006). This is because the quality of accruals is expected to inform investors about the mapping of accounting earnings into cash flows, and since investors value securities by assessing future cash flows (and their risk), poor accruals quality is expected to weaken this mapping and, as a result, increase information risk (Francis et al., 2005).
Unlike previous research, that investigates the association between information risk and stock return historic volatility (e.g., Rajgopal and Venkatachalam, 2011), this study focuses on stock return volatility as implied by the prices at which investors transact in the equity options market. Even though many studies hypothesize and conclude that more knowledgeable investors prefer to trade through the options market and that option market traders are more sophisticated than investors in the stock market (Diavatopoulos et al., 2012; Jin et al., 2012; Goodman et al., 2012), this is not the main reason that makes the association between accounting quality (hereafter AQ) and implied volatility an important one to investigate.
Equity implied volatility (hereafter IV) is a forward-looking measure of investor expectations about the risk and future economic performance of firms, and since the seminal work of Black (1975), it has been considered to reflect new information earlier than stock markets and to contribute greatly to price discovery.1 It has been established as a good predictor of future stock returns and return volatility (Latane and Rendelman, 1976). As corporate earnings announcements are rich in information, the idiosyncratic risk associated with the reported information (proxied by the quality of the information provided) affects the expectations formation of market participants regarding the (overall, i.e., information and performance-related) future uncertainty of a firm’s equity. This last one is manifested in the (annualized implied) volatility level at which market participants are willing to buy or sell a firm’s stock, up to and
including a future point in time. Earnings announcements represent a channel through which firms resolve idiosyncratic uncertainty about the value of their equity (Barth and So, 2011), and the quality of the accounting information disclosed makes understanding past and predicting future firm performance more or less difficult for outside investors, directly affecting the forward-looking, equity volatility estimate they attach (through trading) to a firm's stock shortly before or after its earnings are released.
We employ all US firms from Compustat that have equity options data on Ivy DB Option Metrics between 1996 and 2010, and testify that the negative and significant association between AQ and stock return variability observed by past research for longer time horizons (Rajgopal and Venkatachalam, 2011) also holds in the event study time window, i.e., around quarterly earnings announcements. Closer to the aim of this paper, by focusing on at-the-money (ATM), short-term (30-day) implied volatility (that should be the quickest to respond to new information released via earnings announcements, see for example Truong et al., 2012; Donders et al., 2000; Beckers, 1981), we first testify that lower (higher) accounting quality is associated with significantly higher (lower) implied volatility in the days around quarterly earnings announcements. Much like in previous research, we also confirm a tendency for a sharp increase (resolution) in short-term, ATMimplied volatility in the days immediately before (after) the earnings announcement day (see Truong et al., 2012; Whaley and Cheung, 1982; Donders and Vorst, 1996; Dubinsky and Johannes, 2005).
However, we extend previous contributions, by further showing that the increase (decrease) in implied volatility immediately before (after) the quarterly earnings announcement day is more pronounced for firms whose information risk is high (as proxied by poor accounting quality). Poor (good) quality of quarterly accounting information is found to be associated with larger (smaller) changes in implied volatility around earnings announcements, and this finding is robust to the use of different proxies for accounting quality, different day windows surrounding the earnings announcement day and different implied volatilitymeasures (from calls only, fromputs only, calls and puts averaged or delta-interpolated). At the same time, this evidence is robust to isolating our sample for firms experiencing positive vs. negative earnings surprises, indicating that the effect of information risk on the behavior of IV around earnings announcements is
incremental to and distinct from the observed impact of good vs. bad news announcements observed by past research (Truong et al., 2012).
Furthermore, in order to completely isolate the effect of investor uncertainty about firms’ expected economic performance from uncertainty purely arising from the quality of accounting information, we construct twenty five (five-byfive, double-sorted) portfolios of mean implied volatility levels and changes around EA days, by sorting sample firms first on accounting quality metrics and then on historic volatility of operating cash flows (that serves as a control for firm performance uncertainty). Results indicate a clear and significant tendency of implied volatility levels to increase when operating performance volatility increases, and accounting quality deteriorates. Even after controlling for volatility in operating performance, our initial findings are confirmed in that one observes a significant trend for implied volatility to increase before EA as AQ decreases, while implied volatility is observed to significantly dissolve more for lower vs. higher AQ firms, after the announcement of earnings has taken place.
The main finding of this paper, that information risk, as proxied by poor quality of quarterly accounting information, is positively associated with more pronounced changes in implied volatility around earnings announcements, is further confirmed by panel regression estimation, which permits to explicitly control for the effect and significance of a number of factors possibly associated with implied volatility levels and changes, such as firm size, profitability, cash flow generation, leverage, stock market performance and historic volatility, variability of operating cash flows, firm-specific liquidity, and analyst following and forecasts.
It should be noted that past research has differentially examined the impact of good vs. bad operating performance news. Truong et al. (2012) for earnings announcements and Rogers et al. (2009) for management forecasts have found that option markets do not uniformly respond to good vs. bad news or surprises. However, the focus of our study is on the impact ofinformation risk (rather than performance risk) on option implied volatility. This is because, for example, a firm may be a poor performer, however, its financial statements may not pose difficulties and uncertainty with respect to forecasting its future prospects, no matter how poor those may be. In this way, the focus of this study is on the impact of information risk, for which accounting quality (i.e., the extent to which accounting accruals map and translate into cash flows, changes in revenues and tangible assets) is employed as a proxy, on option implied volatility levels and changes around earnings announcements (hereafter, EA), after controlling for firm performance, as well as volatility in operating performance. Moreover, we control for the impact of AQ on the behavior of IV arou