Welcome to Darren Roulstone’s Research Homepage
Published and Working Papers and Works-in-progress by topic:
Investor Information Demand
Investor Information Demand: Evidence from Google Search around Earnings Announcements (With Michael Drake and Jake Thornock; forthcoming at the Journal of Accounting Research)
The objective of this study is to investigate factors that influence investor information demand around earnings announcements and to provide insights into how variation in information demand impacts the capital market response to earnings. The internet is one channel through which public information is disseminated to investors and we propose that one way in which investors express their demand for public information is via Google searches. We find that abnormal Google search increases about two weeks prior to the earnings announcement and then spikes markedly at the announcement with an order of magnitude at least as large as that of other important corporate events (e.g., acquisition announcements). When investors search for more information in the days just prior to the announcement, pre-announcement price and volume changes reflect more of the upcoming earnings news and there is less of a price and volume response when the news is announced. This result suggests that when investors demand more information about a firm, the information content of the earnings announcement is partially preempted.
What Investors Want: Evidence from Investors’ Use of the EDGAR Database (with Michael Drake and Jake Thornock)
Using a novel dataset that tracks all web traffic on the SEC EDGAR servers, we examine the timing and extent of investors’ revealed preferences for mandatory financial filings. The data reveal that investors request millions of filings from EDGAR each week. The most requested filings include the 10-K, 10-Q, and 8-K, along with insider trading disclosures filed on Form 4. However, many of the filings that are required by the SEC are rarely used by investors. Examining the timing of investor requests, we find that investors commonly request historical disclosures filed in prior periods and that abnormal demand for historical filings is higher when lagged and current abnormal stock returns are lower. Examining firm characteristics associated with the demand for filings, we find that abnormal EDGAR requests are negatively associated with lagged abnormal returns, and positively associated with lagged return volatility, media attention, and the presence of earnings announcements. Thus, we find that investors turn to mandatory financial filings during periods of time when news is released, but more specifically when the news is negative and when there is increased uncertainty about the firm.
The Informativeness of Stale Financial Disclosures (with Michael Drake and Jake Thornock; submitted to Journal of Accounting Research)
In a competitive market, gains to information releases are quickly traded away, rendering the new disclosures less informative to price. Subsequently, only unsophisticated investors would be expected to trade on this “stale” information. This study investigates settings where previously disclosed financial filings are informative to markets. Using a novel dataset that tracks search requests on the SEC EDGAR database, we find evidence that investors acquire stale financial information and trade on it within two hours. Our evidence is consistent with investors using stale information to provide context for current events and disclosures and in cases of high prior information uncertainty. We do not find evidence that trading associated with stale financial disclosures is driven by unsophisticated investors. Thus, our findings are consistent with rational investors finding value in financial disclosures that have been in the public domain for a period of time.
The Impact of Investor Information Demand on the Market Reaction to Mandatory Periodic SEC Filings (With Michael Drake, Bret Johnson, and Jake Thornock; work-in-progress)
Analyst Initiations of Coverage and Stock-Return Synchronicity (With Steven Crawford and Eric So; forthcoming at The Accounting Review)
We examine how the information produced by analysts when they initiate coverage contributes to the mix of firm-specific, industry-, and market-wide information available about the firm. We hypothesize that the first analyst to initiate coverage provides low cost market and industry information allowing him/her to follow more stocks, whereas subsequent analysts provide firm-specific information to distinguish themselves from existing analysts. We use stock return synchronicity to measure the mix of information available about a firm, with higher synchronicity indicating more industry and market information. Coverage initiations of firms with no prior analyst coverage increase synchronicity suggesting that analysts produce industry- and market-wide information. In contrast, analysts initiating coverage on firms with existing coverage appear to focus on producing firm-specific information as these initiations lead to reduced synchronicity. Together, our findings indicate that the type of information analysts produce at initiation depends on the information provided by other analysts.
The Time-Dynamics of Analysts’ Consensus EPS Forecasts (With Steven Crawford and Eric So)
This paper examines the time-varying accuracy of analysts’ consensus EPS forecasts within a firm-quarter. The results identify attributes of the information environment that affect the evolution of analysts’ information acquisition and processing. We find that while analysts appear to learn about the upcoming EPS figure over time, both firm and industry level factors significantly influence the speed and degree of learning. Specifically, consensus accuracy improves to a greater degree within a firm-quarter for firms with volatile net income, high accruals, and prior losses and for firms that issue managerial guidance and which report later than their industry peers. Alternatively, consensus accuracy is less sensitive to time for large firms and firms with high levels of analyst coverage. Consistent with analysts learning from dynamic intra-industry signals, we find that consensus forecast accuracy is increasing in contemporaneous earnings announcements and managerial guidance from industry peers. We also show that intra-quarter changes to consensus forecast errors, incremental to the level, possess information about the market surprise during the announcement window. This result is consistent with intra-quarter consensus changes reflecting analysts’ information acquisition.
The Influence of Analysts, Institutional Investors, and Insiders on the Incorporation of Market, Industry and Firm-Specific Information into Stock Prices (With Joseph D. Piotroski; Accounting Review Volume 79 No. 4, 2004)
This paper investigates whether analysts, institutional investors and insiders increase the amount of firm-specific versus industry-specific information found in prices. We measure firm and industry specific information in two ways: stock return synchronicity with market and industry returns (Morck, Yeung and Yu ); and firm and industry-earnings response coefficients (Ayers and Freeman ). Using both measures we find that analyst activity contributes to the incorporation of both types of information; however, industry effects are greater than firm-level effects. Contrarily, insider transactions and changes in institutional holdings have the net effect of increasing the amount of firm-specific information in prices. These results are consistent with the notion that insiders have the greatest access to firm information, while analysts have more limited access to this information.
Analyst Following and Market Liquidity (Contemporary Accounting Research, Volume 20 (Fall) 2003)
This paper investigates the relationship between analyst characteristics (number of analysts following a firm and their forecast dispersion) and market liquidity characteristics (bid-ask spreads and depths and the adverse-selection component of the spread). Prior research has found contradictory results on the relation between analyst following and market liquidity and has offered differing theories on how analysts affect liquidity. While prior research has posited analysts as proxies for privately informed trade or as signals of information asymmetry, I hypothesize that analysts provide public information, implying that analyst following (forecast dispersion) should have a positive (negative) association with liquidity. Cross-sectional OLS and simultaneous estimations provide support for this hypothesis. The results are both statistically significant and economically important. Granger-causality tests indicate that analyst characteristics lead market liquidity characteristics. I also find that depths are negatively correlated with the adverse-selection component of the spread, a finding that reinforces the view that market makers adjust both spreads and depths in reaction to information asymmetry problems. These results clarify the role of analysts in providing information to financial markets and highlight benefits of increased analyst following.
Analyst Coverage Initiations and Institutional Holdings (With Steven Crawford and Eric So; work-in-progress)
Security analysts and institutional owners are two of the major contributors to the information environment of a firm and a long line of research has investigated the benefits and costs of coverage by security analysts and institutions. We investigate the relation between initiations of analyst coverage and changes in the level and breadth of institutional holdings in order to provide evidence on these questions. We are in the preliminary data analysis phase of this project.
Evidence on the Non-linear Relation between Insider Trading Decisions and Future Earnings Information (With Joseph D. Piotroski; Journal of Law, Economics, and Policy, Volume 4 No. 2)
In this paper, we provide evidence that the relations between insider trading decisions and next year's earnings are not strictly linear. We find that insider purchases are positively related to next year's earnings innovation and that this relation is attenuated in the case of extreme positive innovations. We also find that insider selling and option exercises are negatively related to next year's earnings innovation and that these relations are attenuated in the case of extreme positive and negative innovations. We conclude that the observed variation in trading decisions is consistent with the existence of potential legal liability costs. Our estimations also suggest that aversion to trading due to legal liability concerns is stronger for insider selling and option exercises than for insider purchasing. Finally, we investigate the role of earnings persistence as an alternate explanation for our results. We find that insiders only trade on persistent, future earnings innovations, and that, after controlling for persistence, insiders still curtail trading when innovations are extreme.
Do Insider Trades Reflect Both Contrarian Beliefs and Superior Knowledge about Future Cash-Flow Realizations? (With Joseph D. Piotroski; Journal of Accounting and Economics, Volume 39 No.1, 2005, pages 55-82)
This paper examines whether insider trades reflect superior knowledge of future cash flow realizations, as proxied by the firm’s future return and earnings performance. We find strong evidence that insider trades are positively associated with the firm’s future earnings performance. This relation is shown to be incremental to the book-to-market and past return relations documented in Rozeff and Zaman (1998), suggesting that insiders trade on both transitory security misvaluation and private information about future cash-flow payoffs. These results are shown to be robust to several measures of insider trading behavior and future earnings innovations. We show that the relation between insider trades and future earnings performance is amplified (attenuated) as the likely ex ante benefits (costs) to trading on financial performance information increase. Finally, we find that insider trading behavior within book-to-market portfolios varies with the horizon of the subsequent earnings news, with the sign of the relation between insider purchases and contemporaneous earnings being negative (positive) for glamour (value) firms.
The Relation Between Insider-Trading Restrictions and Executive Compensation (Journal of Accounting Research, Volume 41 (June) 2003)
This paper investigates the impact of firm-level, insider-trading restrictions on executive compensation. Using a trading-window proxy for the existence of such restrictions I test theoretical predictions that insiders will demand compensation for these restrictions and that firms will need to increase incentives to restricted insiders. I find that firms that restrict insider trading pay a premium in total compensation after controlling for standard economic determinants of pay. Further, these firms use more incentive-based compensation relative to firms that do not restrict insider trading, e.g., equity grants and bonuses are higher for these firms. Finally, insiders at these firms hold larger equity positions than insiders at “unrestricted” firms. These results hold after controlling for the endogeneity of the decision to impose the restrictions and support the idea that insider trading plays a role in rewarding and motivating executives.
Insider Trading and the Incorporation of Future Earnings into Stock Prices (with Steven Crawford and DuRi Park)
We show that insider trading is associated with a greater incorporation of future earnings news into stock prices. We extend this basic result by showing that trades by CEOs and CFOs increase the association between returns and future earnings news. Our results are incremental to the effects of analyst following, institutional ownership, and management forecasts as documented in prior research (e.g., Ayers and Freeman, 2003).
Early Evidence on Insider-Trading Activity (With Suraj Srinivasan and Joseph D. Piotroski)
We examine insider trades reported to the SEC between 1935 and 1944, a time period marked by lax enforcement of insider-trading laws (Jaffe, 1973). Our goal is to assess the relative ability of insiders during this regulatory era to exploit private information as compared to insiders in the modern era (e.g., after passage of the Insider Trading Sanctions Act of 1984). Insider trading activity in the early time period is hand-collected from the official SEC summaries of insider activity. We document that while insiders appear to not exploit private knowledge of earnings surprises, they do earn roughly 0.5% per month in the six months following their purchases, a number quite close to the abnormal returns earned by insiders in the modern era.
Insider Trading and the Information Content of Earnings Announcements
This paper addresses the question of whether the net effect of insider trading is to promote accurate pricing of stocks by conveying insiders’ private information to market participants. I investigate this question by examining the relation between insider trading and the information content of earnings announcements. I document two main findings: first, decisions to trade are influenced by future earnings surprises and announcement returns. Second, insider purchases and sales executed and disclosed prior to an earnings announcement preempt news in the announcement and have a negative relation with market reactions to the announcement, consistent with insider trading informing the market. These relations hold after controlling for the endogenous relation between trading and market reactions and are of an economically significant magnitude (particularly for insider purchases). These findings are consistent with insiders trading on private information and their trading conveying information to the market, two necessary conditions for insider trading to have a net positive effect on securities pricing.
Effects of Insider-Trading Legislation on Trade Timing, Litigation Risk, and Profitability (With Alan Jagolinzer)
Prior research indicates that insiders avoid trading ahead of major disclosure events such as quarterly earnings announcements and that this avoidance is associated with firm policies restricting the timing of insider trades (Bettis, Coles and Lemmon, 2000; Roulstone, 2003). Garfinkel (1997) provides evidence that this behavior increased in response to the Insider Trading and Securities Fraud Enforcement Act of 1988 (ITSFEA). Using a 24-year sample of insider trades we show that insiders are, over time, increasingly trading after earnings announcements especially since passage of ITSFEA. This finding is robust to controls for insider incentives to trade around earnings announcements. We investigate the economic effects of these changes by documenting a relation between litigation risk (measured by the expected probability of 10b-5 litigation) and insider trade timing. Specifically, litigation risk is decreasing in the extent to which insiders trade following earnings announcements rather than before earnings announcements. However, while insiders are increasingly trading during times of lower litigation risk, we fail to find a decrease in insider-trading profitability over our sample period, suggesting that regulation (economy-wide and firm-specific) has not limited the ability of insiders to exploit private information.
The Mechanisms of Information Transfer (With Michael Drake, Jared Jennings, and Jake Thornock; submitted to Review of Accounting Studies)
The literature on intra-industry information transfers finds that one firm’s information release is associated with market reactions at related firms. While the literature provides evidence that this phenomenon exists, it is relatively silent as to how such transfers occur. We examine the extent to which earnings-related information transfers to related firms through three important information channels: the internet, analyst forecast revisions, and media reports. We find evidence of information transfer around earnings announcements through each channel; in particular, investors greatly increase internet search about peer firms at the time of earnings announcements. Moreover, we find that the level of information transfer varies by the magnitude of the announcing firm’s earnings news. Finally, we find that increases in information transfers through each of these channels are associated with increases in trading volume at peer firms, consistent with investors revising beliefs about peer firms as a result of the transferred information.
Management Earnings Forecasts and Forward-Looking Statements (with Andrew Van Buskirk and Zahn Bozanic)
We examine managers’ disclosures of qualitative forward-looking statements and, in particular, how the decision to issue those statements differs from that of quantitative earnings forecasts. We find that, while positively correlated, forward-looking statements and earnings forecasts are two distinct disclosure decisions – one is not a substitute for the other. We then show that the market assesses qualitative information as credible: forward-looking statements help explain market reactions to earnings announcements and the “surprise” in the tone of forward-looking statements is associated with quarterly stock price performance. In other words, just as forecast accuracy can be evaluated by comparison to realized earnings, the accuracy of forward-looking statements can be evaluated ex post by comparing the tone of forward-looking statements from one period to the tone of non-forward-looking statements in the next period. We then study the determinants of both types of disclosures, focusing on two factors that are widely-believed to affect a manager’s willingness to disclose information about the firm’s future: uncertainty and competition. We find that managers prefer to issue qualitative forward-looking information (relative to forecasts) when uncertainty about future earnings is greater. We find (weaker) evidence that changes in competition are associated with managers’ issuing quantitative versus qualitative forward-looking information. Our findings support the use of qualitative forward-looking statements as a credible tool for managerial disclosure.
Effect of the SEC Financial Reporting Release No. 48 on Derivative and Market Risk Disclosures (Accounting Horizons, Volume 13 (December) 1999)
This study compares the disclosures about derivatives and market risk made by 25 SEC registrants in the years before (1996) and after (1997) the adoption of Financial Reporting Release No. 48 (SEC 1997) (FRR No. 48). FRR No. 48 requires firms to disclose how they account for derivatives and provide quantitative and qualitative disclosures about exposures to market risk. Market risk disclosures, encouraged but not required under FAS No. 119, improved greatly under FRR No. 48 but varied widely in detail and clarity. The majority of registrants provided quantitative and qualitative disclosures of market risk; however, only about half of these firms discussed the details and limitations of their risk measurement models and disclosures. Further, certain required or strongly recommended contextual disclosures were almost completely absent. Firms appear to prefer relatively complicated but more discreet disclosure formats to simpler but more revealing disclosure formats. Overall, while registrants greatly increased their disclosures about market risk, the disclosures leave room for improvement in future filings. These findings have significance for disclosure choice in general and the adoption of FAS No. 133 in particular.
Valuation and Mispricing
Acquirer Valuation and Acquisition Decisions: Identifying Mispricing using Short Interest (With Itzhak Ben-David and Michael Drake; submitted to Journal of Financial and Quantitative Analysis)
We use short interest as a new investor-based measure of over/undervaluation that more sharply distinguishes misvaluation vs. Q theories of mergers and acquisitions. Using this measure, we find that misvaluation is a strong determinant of merger decision making. Specifically, after controlling for known determinants of acquisitions, firms in the top quintile of short interest are 54% more likely to engage in stock mergers within the next month and 22% less likely to engage in cash acquisitions within the next month. Further, post-merger return is strongly correlated with pre-merger short interest. Finally, stock (but not cash) acquirers have higher short interest than their targets. We conclude that overvalued firms self-select to become stock acquirers, while undervalued firms engage in cash acquisitions.
Management’s Information Advantage and Insider Trading (with Itzhak Ben-David and Andy Van Buskirk)
Hutton et al. (2012, forthcoming JAR) analyze management and analyst information advantages in the context of forecast accuracy. We extend this research by investigating whether factors affecting forecast accuracy also affect managers’ trading profitability. Among our preliminary findings is the fact that managers at highly idiosyncratic firms earn higher returns than managers at systematic firms.
Discussion of ‘Large-Sample Evidence of Firms’ Year-over-year MD&A Modifications’(Journal of Accounting Research, Volume 49 No. 2, 347-357)
Discussion of ‘Intangible Investment and the Importance of Firm-Specific Factors in the Determination of Earnings’ (Review of Accounting Studies, Volume 16 No. 3, 574-586)
Accounting, Finance, and Economics Research Links: