درآمدهای پیش بینی در یک محیط اطلاعاتی نامناسب

Financial intermediaries, such as analysts, play an important role in providing information to investors. However, a large segment of the market (about 39% of CRSP firms between 1992 and 2009) is not served by financial analysts, leaving investors in a poor information environment. In this paper, we examine whether other publicly available information signals, such as insider trades, institutional holdings, and firms’ stock repurchases, can be used to predict information about earnings for these firms. We find that CFOs’ trading decisions are associated with new information contained in the annual earnings reports for firms with no or scant analyst coverage. In contrast, for firms with multiple analyst coverage, insider trading decisions are not predictive of new information in earnings reports. Our results suggest that some public information signals, such as insider trades, can be used to alleviate the poor information environment faced by investors. However, the market may not have fully priced the information contained in these signals.

Financial intermediaries, such as analysts, play an important role in providing information to investors. However, a large segment of the market (about 39% of CRSP firms between 1992 and 2009) is not served by financial analysts, leaving investors in a poor information environment. In this paper, we examine whether other publicly available information signals, such as insider trades, institutional holdings, and firms’ stock repurchases, can be used to predict information about earnings for these firms. Information about a firm’s earnings is arguably the most important information for equity investors. Earnings are used in many valuation metrics (e.g. P/E ratio and the residual income model à la Feltham and Ohlson, 1995) and are a good predictor of future cash flows ( Dechow et al., 1998). A firm’s periodic earnings release is highly anticipated by the capital market and can cause significant price movements. 2 Given its importance, market participants and intermediaries have invested considerable resources into predicting earnings. Of these market intermediaries, financial analysts play a very prominent part by conducting research on firms, informing the market about earnings expectations, and making buy/sell recommendations. 3 The presence of financial analysts has made information economically accessible to the average market participant. 4 It reduces the need for costly private information-gathering and processing by average investors and facilitates more efficient investment decision-making. However, not all segments of the market are equally served by analysts. Between 1992 and 2009, on average, about 39% of the firm years on CRSP are not covered by financial analysts on I/B/E/S. Even though these are mostly small firms, they serve an important role in the diversified portfolios of investors and constitute a large segment of the investing market. Firms with no analyst coverage consist of about 13% of total market value on CRSP. In addition, a larger proportion of the shares of these firms are held by individual investors as opposed to institutional investors. In our sample period, non-institutional (individual) investors held an average of 80% of the shares for firms with no analyst coverage compared with an average of 46% of shares for firms with analyst coverage. Since individual investors do not have the resources to engage in private information-gathering and often lack the expertise to process information (SEC, 2012), the absence of financial intermediaries makes the information problem even more acute.5 Following accounting literature that uses properties of analyst coverage to proxy for information environment (e.g. Lang et al., 2003, Gebhardt et al., 2001, Horton et al., 2013 and Frankel and Li, 2004), we define firms that lack financial analyst coverage as being in a poor information environment.6 Although investors can overcome a poor information environment by procuring information privately, individual procurement of information is not economically efficient. For small investors in particular, it is usually not economically feasible. In this paper, we examine the usefulness of several sources of public information that could be used by investors to predict new information about future earnings in the absence of financial analyst coverage. Specifically, we examine whether insider trades, stock repurchases, and institutional holdings are predictive of abnormal returns during annual earnings report windows, which we use as a measure of new information in earnings. Following theories in psychology and behavioral finance, we expect that for firms in a poor information environment, public information may not be fully incorporated into prices. The psychology literature finds that attention is a scarce cognitive resource (Kahneman, 1973), and decision makers have limited attention and processing power for information signals. In a limited attention regime, only information that is salient and easily available is processed (Tversky and Kahneman, 1973 and Hirshleifer, 2001). Therefore, information signals that are not salient, readily available, or in an easily processed form may not be (completely) processed by investors (Hirshleifer, 2001). In Hirshleifer and Teoh’s (2003) model, even highly attentive investors could still fail to absorb non-salient information due to costly time and attention constraints, and the presence of even a portion of investors that have limited attention and processing power will affect firms’ accounting choices (method of accounting for employee stock options, pro forma earnings disclosure, and segment reporting) and distort stock prices. The finance literature has found empirical evidence of such limited attention in investor behavior. For example, Barber and Odean (2008) find that investors focus on attention-grabbing stocks only and that information may not be incorporated into prices until it attracts investor attention. Peng and Xiong (2006) show that investors with limited attention will use simple decision rules that utilize market and sector information at the expense of firm-specific information. In this paper, we extend the literature to the context of a poor information environment, where the limited attention problem is exacerbated. In a poor information environment, information signals, even if publicly available, are not salient due to a lack of analyst attention to these firms. Whereas analysts may call attention to insider trades, share repurchases, or institutional trading when following a firm and incorporate such information into their buy/sell recommendations and earnings forecasts, firms that lack analyst coverage do not enjoy such attention. In addition, these information signals are not easily accessible to average investors.7 Therefore, we expect that the information about future earnings contained in these signals is not fully incorporated into prices. We focus on trading decisions made by informed investors (insiders and institutional investors) that are publicly disclosed prior to earnings reports. Insider trades can serve an information role in the market (Manne, 1966). Insiders (e.g. executives and directors) likely possess private information about firms’ earnings. When they trade, such information is incorporated and thus revealed to the market (Piotroski and Roulstone, 2005, Leland, 1992 and Hu and Noe, 2001). However, the market may fail to fully incorporate the information into prices (e.g. Wang et al., 2012). As a result, such information may still surprise the market and generate abnormal returns when earnings are reported, especially in a poor information environment. The information possessed by insiders may also not be of equal quality. Wang et al. (2012) find that CFOs’ trades are more informative than CEO’s trades (i.e. CFOs, but not CEOs, earn abnormal returns after trades). This is consistent with CFOs possessing superior information due to their roles in firms’ financial policy and financial reporting processes (Jiang et al., 2010 and Geiger and North, 2006). Therefore, we expect that CFOs’ trades in particular are predictive of new information in earnings. In addition to trading from insiders’ personal accounts, share repurchases (trading on corporate accounts) can also reflect insiders’ information about future earnings. Firms can buy back shares when future earnings prospects are good and refrain from stock repurchases to conserve financial resources when future earnings prospects are bad. To the extent that investors fail to fully incorporate such information into prices (e.g. Ikenberry et al., 1995 and Peyer and Vermaelen, 2009), we expect that firms’ share repurchase is predictive of new information in future earnings. Institutional shareholders possess the resources to engage in private information gathering in a poor information environment. Although the information they procure is not available to the public, their trading and holding decisions could convey relevant information to the market (Chakravarty, 2001). If investors fail to attend to and price such information, then we expect institutional trading of stocks to be predictive of new information in future earnings as well. Using a sample of firms traded on the major exchanges (NYSE/AMEX/NASDAQ) between 1992 and 2009, we find that CFOs’ trading decisions are predictive of new information in future earnings for firms in a poor information environment.8 We measure new information in earnings as the cumulative abnormal returns (CARs) around the announcement windows of annual earnings reports. We find that for these firms, average CARs are higher by 62–82 basis points where a CFO purchases stocks during the one year period prior to the release of its earnings report, compared with firms whose CFOs sell stocks. CEOs’ trades, on the other hand, are not predictive of new information in future earnings. For comparison, we also examine the predictive power of insider trades for firms in a good information environment, defined as one in which a firm has two or more analysts following it. We find that insider trades (by CFOs or CEOs) of these firms do not predict future earnings information. Also as expected, we find that stock repurchases predict new information in future earnings. However, the informativeness of stock repurchases is not exclusive to poor information environment firms. Stock repurchases by firms with a good information environment are also predictive of future earnings. Contrary to our predictions, we find that institutional trading decisions are generally not predictive of future earnings for firms in poor information environments. Our paper is related to and contributes to several streams of literature. First, the paper is related to the extensive body of research that uses informational leading indicators to predict earnings or earnings-related fundamentals (e.g. Fairfield et al., 2009, Bratten, 2009 and Beneish and Vargus, 2002). Our primary contribution to this literature is the focus on firms in a poor information environment, which constitute a large segment of the market. Most prior research relating information signals to earnings either focus on firms that are covered by financial analysts9 or do not separately examine firms with no analyst coverage. We complement this stream of research by showing that certain information signals (e.g. CFO trades) are useful only in a poor, rather than good, information environment. Second, we add to the literature on imperfect information processing by market participants (e.g. Barber and Odean, 2008, Peng and Xiong, 2006 and Hirshleifer and Teoh, 2003). Built on Kahneman’s (1973) seminal theory, this literature shows that, generally, investors with limited attention cause price inefficiencies. We extend the literature to the context of a poor information environment, where the limited attention problem is even more acute. We identify the source of the information (e.g. earnings information contained in CFO trades) that investors fail to incorporate in prices. Third, our paper adds to the understanding of the informativeness of insider trades (e.g. Seyhun, 1998). Recent research (Wang et al., 2012) finds that CFOs rather than CEOs earn abnormal profits in their trades, suggesting an information advantage in CFO trades. Our results support the notion that the source of the information advantage is that CFO trades reflect information about future earnings. Our evidence could also be useful for market participants, especially those who trade small cap stocks not covered by financial analysts. We show that some public signals can be used to overcome the information deficiency in a market segment underserved by financial intermediaries. Investors in a poor information environment can benefit from trading decisions that originate from insiders, especially CFOs, and extract useful information about future earnings that the market has not fully incorporated into prices. The rest of the paper is organized as follows. In Section 2, we describe data and methodology. In Section 3, we present our main findings. In Section 4, we conclude.

In this paper, we investigate whether public information signals, such as insider trades, corporate stock repurchases, and institutional trades, are useful for predicting new information in future earnings for firms that are not, or are scantly, served by financial analysts. We find that CFOs’, but not CEOs’, trading of company stocks is predictive of new information in future earnings for these firms. In contrast, insider trades are generally not informative of future earnings in a good information environment, where firms are covered by multiple analysts. Another public signal, stock repurchases, is also found to be informative about new information about future earnings for poor information environment firms. However, the results are not as robust as those from CFO trades. Our finding that CFOs’ trading is informative of future earnings news for firms in a poor information environment provides insights into a source of information advantage as well as the environment in which their trades can be most informationally useful. Of all information sources, including institutional investors, share repurchases, and other insiders, CFOs probably possess a unique information advantage due to their roles in forming firms’ financial strategies and in financial reporting. Our results also complement previous research (e.g. Wang et al., 2012) showing a general information advantage in CFO trades. Our study is potentially useful for investors interested in trading in a poor information environment. Firms with no financial analyst coverage constitute a significant portion of the investing market. For these firms, investors (especially small investors) suffer a large information asymmetry. We identify a robust information signal that could help investors predict future earnings information and perhaps make better investment decisions.