【正文】 ch integrated model. We build on their work by developing a dividend model that incorporates both asymmetric information and free cashflow problems. Particularly, we consider a dual role for dividends. Dividends may provide a signal of current ine to investors (hence the manager is motivated to choose a high dividend to provide a positive signal). However, in our analysis, a new project is available to the firm. If the firm wishes to invest in this project, it must get the funds from current ine. Hence, in addition to the current inesignaling role, the level of dividends may also affect the manager39。s ability to invest in the new project. Hence, the manager may wish to cut dividends (to take a good, valueadding project), or he may wish to payout high dividends (to reduce the free cash flow in order to mit not to take a bad, valuereducing project). Our model contributes to Fuller and Thakor39。s (FT 2020) analysis in the following ways. First, we develop the analysis, and derive the equilibrium of the dividend game, in a formal and rigorous manner. Second, FT only considers the possibility that a negative NPV project exists. Hence, they focus on Jensen39。s (1986) free cashflow problem. In contrast, we consider the possibility that the project may have a negative or positive NPV. This enables us to consider the following underresearched area. Although the majority of theoretical and empirical work suggests that the relationship between dividends and share price is positive (dividend increases are ―good news‖), Wooldridge and Ghosh (1998) have suggested that, when a firm has strong growth opportunities available, dividend cuts may not always be bad news. Conversely, dividend increases may be bad news. For example, Allen and Michaely (2020) argue, ―However, we note that with asymmetric information, dividends can also be viewed as bad news. Firms that pay dividends are the ones that have no positive NPV projects in which to invest‖, and according to Black (1976), ―Perhaps a corporation that pays no dividends is demonstrating confidence that it has attractive investment opportunities that might be missed if it paid dividends‖. Hence, the relationship between dividends and firm value is plex. The market may view an increase in dividends favorably, either as a positive signal of current ine (that is dividends reduce asymmetric information problems), or as a means of mitigating free cashflow problems (that is dividends reduce agency problems). However, a dividend increase may be seen as a negative signal (the firm lacks growth opportunities), while a dividend cut may be seen as a positive signal (the firm has significant growth opportunities available). Our integrated signaling/excess cashflow model of dividends attempts to analyses all of these affects. Our third contribution is to consider two types of inefficiency (or agency problem) relating to managerial incentives regarding dividend policy. Firstly, we analyses a moral hazard problem in which empirebuilding managers may cut dividends in order to invest in a negative (valuereducing) NPV project due to managerial private benefits. Secondly, we consider an adverse selection problem, whereby managers may refuse to cut dividends, hence passing up a positive NPV project[4]. Indeed, Cohen and Yagil (2020) identify ―a new type of agency cost of dividend: the sum of positive NPV projects that are abandoned in order to pay dividends‖. Frankfurter and Wood (2020) observe that, in addition to analyzing the effects of agency and asymmetric information problems on dividend policy, researchers are beginning to consider behavioral models. They observe that, Investor behavior is substantially influenced by societal norms and attitudes [… ]. According to Shiller (1989), including these influences in modeling efforts can enrich the development of a theory to explain the endurance of corporate dividend policy. Our fourth contribution is that we include a behavioral dimension in our model. That is, we consider investors who have been conditioned to believe that high dividends signal high quality. In our first case, the investors39。 beliefs are indeed correct, as the highquality firm can separate from the lowquality firm by paying a high dividend. In the second case, these beliefs drive the adverse selection problem (the highquality firm is unwilling to reduce the dividend below that of the lowquality firm, since the market then mistakenly values this firm as low quality)[6]. As an extension to our second case, we consider whether the adverse selection problem can be mitigated by munication to investors, reinforced by managerial reputation effects. Indeed, Wooldridge and Ghosh (1998) argue that firms may be able to avoid a negative market reaction by municating to investors that the reason for dividend cuts is to invest in future growth opportunities. It may be argued that this may be chea