The corporate finance literature, starting with Modigliani and Miller, studies how managerial decisions can increase firm value. Specialized management leads to better decision‐making but it forces the separation of ownership and control and generates asymmetric information between managers and shareholders. This leads to an agency problem that destroys firm value. Therefore, reducing this asymmetric information problem by finding credible ways to transmit information to investors becomes a crucial part of managers’ jobs.
In particular, managers will have to make decisions about two different types of information that is generated inside the firm and must be communicated to outside investors—hard and soft information. On the one hand, there are pieces of information about the company that can be codified in a systematic way and that can be verified by the receivers of this information. This is the so‐called hard information such as sales volume or board appointments. On the other hand, there exists soft information related to intangibles like managers’ ability, firm strategy, employee morale, etc., that a third party cannot verify. Soft information can be as important as hard information for investors to monitor (Cornelli, Kominek, & Ljungqvist, ). Remarkably, chief executive officers (CEOs) enjoy large discretion over the disclosure of soft information and they can use it in a strategic way in order to pursue their own interests. Consequently, understanding what makes soft information credible is particularly important.
Managers can credibly transmit soft information in two ways. First, they can use signals with direct and explicit costs for the firm, like leverage or dividends, which make these signals credible. Second, they can use cheap talk as explained by Bhattacharya and Dittmar () and Almazán, Banerji, and Motta ().
Our paper is connected to the stream of literature on the transmission of soft information through apparently costless actions (cheap talk). Here cheap talk is understood as a call for attention that at a glance does not seem to require any effort or cost. Unlike traditional signals that require doing something costly for the firm in order to attract attention, using cheap talk works because attracting attention has a cost connected to the pressure on CEOs’ reputation.
Theories related to cheap talk are well established in the economics literature (see, e.g., Chakraborty & Harbaugh, ; Crawford, ; Farrell & Rabin, ; Stein, ), whereas the application in finance is rather limited, with Almazán et al. () being one of the few examples. This paper is the first to test empirically and comprehensively the cheap talk theory in the context of finance as an effective transmission of information mechanism and taking into consideration CEOs’ incentives and reputational costs.
A clear visual example of how cheap talk works is as follows: A manager sits in a crowded market floor with many other managers. Our manager suddenly raises a red flag to make his firm stand out and attract investors’ attention. But, if cheap talk is really as cheap as raising a red flag, all managers could do it. And, if all managers raise flags, this fails to make the firm stand out. This implies that cheap talk can only work if it is not cheap in some subtle way. In fact, raising a red flag has significant indirect costs for managers. This happens because raising the flag attracts analysts’ and investors’ attention and induces them to engage in a costly revision of their valuation of the firm. If, after revisiting the firm, the market is unconvinced and the firm's market value does not increase, managers will bear some costs for improperly attracting such attention given that remuneration and tenure are linked to stock market prices. Hence, not all managers will raise flags. Managers will disclose soft information truthfully, and only raise flags when they think that their firms are undervalued. Therefore, the mechanism for costly transmission of soft information through cheap talk requires that market players understand the use of cheap talk as a call for attention, revisit the firm, revise their previous analysis, and update their expectations about the firm. Investors will take the call for attention seriously because in case the updated expectations fail to improve on the previous ones, CEOs will be punished with lower compensation or removal from their positions.
We depart from recent papers that provide partial evidence of CEOs’ incentives and consequences of using certain cheap talk actions (e.g., stock splits) for the transmission of information (Devos, Elliott, & Warr, ). These authors show that the decision to split is directly related to managerial remuneration and specifically to the delta of the CEO's compensation portfolio. Our evidence confirms that CEOs with more high‐powered incentives will have more incentives to announce not only stock splits, but also to make more voluntary earnings announcements and press releases. Moreover, we contribute to this line of research by explaining why there can exist an equilibrium where the market reacts positively to the calls for attention of managers that have incentives to increase market prices. We provide empirical evidence showing that the market reacts to these announcements, not because it cannot see the incentives of the CEO, but because these incentives are fine‐tuned to punish CEOs that try to attract attention without having any positive soft information to communicate, making cheap talk a costly and credible signal. This situation may arise when there is a decoupling between CEOs’ words (call for attention to revise the firm's valuation) and their actions to generate value. Levit () shows the tension between CEOs’ communication (words) and intervention (actions). Ananchotikul, Kouwenberg, and Phunnarungsi () show evidence of decoupling in firms’ corporate governance. In our case, we do not study the reasons for the eventual disconnection between the call for attention of cheap talk and the eventual lack of positive market returns, but we show that if the returns after cheap talk are negative, CEOs are penalized (they are fired or their compensation is reduced). Moreover, we show that our analysis is robust to the use of different proxies for characterizing cheap talk, namely: stock splits, CEO voluntary announcements, and press releases.
We test empirically the mechanisms that motivate and allow managers to credibly transmit soft information to the market (Almazán et al., ). We first present evidence confirming that managers whose compensation is more sensitive to stock market prices engage more in cheap talk measures, including stock split announcements, voluntary earnings forecasts, and press releases. We then go on to show that cheap talk does in fact attract analysts’ attention, increasing the number of earnings forecasts issued by analysts. Moreover, we find that managers are also more likely to face dismissal and/or a reduction in their pay packages when, after engaging in cheap talk, stock returns do not increase. This is consistent with the idea that soft information can be credibly transmitted through the use of cheap talk because managers are punished when their attempts at attracting attention fail to produce an upwards price reaction. Our results are stronger for smaller firms, firms that are less followed by analysts, and firms that are more undervalued and therefore are in greater need of attention.
The empirical analysis relies on a sample of 3,332 US firms for the 1992–2016 period and analyzes three different proxies for cheap talk: stock split announcements, CEO voluntary earnings forecasts, and press releases issued by the firm to the media. Our three main results are consistent with the predictions from the theoretical framework.
The first result indicates that an increase in the proportion of variable compensation in CEOs’ pay packages that is linked to stock market prices induces managers to use cheap talk strategies. In particular, it turns out that a 1% increase in the proportion of the variable part in CEOs’ remuneration packages is correlated with (a) 10.3% increase in the probability of announcing stock splits; (b) 6.08 more annual CEO earnings forecasts; and (c) around 73.25 more press releases initiated by a firm each year.
Our second main finding is that the use of these measures serves to attract the market's attention, as there is an increase in the analyst following of the firm, both in terms of the number of earnings per share (EPS) forecasts issued by analysts and the proportion of upward revisions in their estimates. In particular, annual cumulative number of EPS forecasts issued by analysts for a given firm increases by (a) around 122 with one standard deviation increase in the number of stock split announcements (0.24 more announcements); (b) 1.53 for upward revisions with one standard deviation in CEO forecasts (almost 10 more annual CEO forecasts); and (c) by 34.27 with one standard deviation increase in the number of press releases per year (close to 317 more articles).
Finally, we find evidence consistent with the idea that cheap talk works as a credible signal because it is costly for managers, that is, CEOs are punished with a higher probability of turnover and/or a reduction in their total subsequent compensation if cheap talk is not followed by a positive stock price reaction. More specifically, assuming that the different measures of cheap talk move up one standard deviation from their respective mean values but there is no change in returns and they stay at their median value, we would observe: (a) for an increase in stock splits, a 13.97% higher probability of CEO turnover and a $2.16 million decrease in CEO total compensation; (b) for an increase in CEO forecasts, a 4% increase in the probability of CEO turnover and a $0.69 million decrease in subsequent CEO compensation; and (c) for an increase in...