flows are more likely to invest in negative net present value projects or waste them on organisational inefficiencies rather than distribute them to shareholders (Jensen, 1986). Jensen’s free cash flow hypothesis predicts a significant valuation effect for signals of changes in wasteful investments for firms with poor investment opportunities or those characterised by organisational inefficiencies. Lang and Litzenberger (1989) argue that these firms are reluctant to subject negative net present value projects to the monitoring associated with external financing. Therefore, announcements of dividend changes by overinvesting firms reflect management’s chosen course of action regarding its future investment policy. Specifically, an increase in dividends mitigates the overinvestment problem and increases the market value of the firm as markets adjust their expectation about the outcome of the overinvestment problem. Conversely, markets react negatively to a reduction in dividends by overinvesting firms as this sends a signal that management would continue in the path of overinvestment. While bank regulation, aimed at mitigating the potential agency problem associated with banks, may suggest a weak monitoring role for external financing and dividends, evidence and the recurring banking crisis also suggest that regulation is not a perfect substitute for capital market monitoring. In fact, evidence suggests that regulatory efforts, such as the deposit insurance schemes, are associated with other unintended agency problems, suggesting a potential role for market monitoring of banks.3
The agency cost theory is pervasive in finance and accounting, in particular, as an explanation for firms’ capital structure and dividend decisions. However, Berger and Bonaccorsi di Patti (2006) note that the inconclusive evidence regarding the role of agency cost in capital structure may be due to the problem of measurement. They argue that the lack of convincing evidence on the direction of the relationship “could be partly explained by the intrinsic difficulty in defining a measure of performance that is close to the theoretical definition of agency costs.” In line with their argument, we conjecture that the inability to find consistent empirical evidence in support of the role of agency cost in explaining market reactions to dividends or further distinguishing this explanation from the competing cash flow signalling explanation may also be due to the problem of measurement. For example, Lang and Litzenberger (1989) adopt the Tobin’s Q ratio that incorporates both accounting and market data in the investigation of cash flow signalling
and free cash flow hypotheses as explanations for market reactions to dividends. They find support for the role of the latter but not the former. However, Denis et al. (1994) use similar measure for their study of dividend change announcements for a sample of 6777 US firms and find no support for the free cash flow hypothesis. One potential problem with the use of Tobin’s Q ratio is that it relies on stock prices and thus incorporates any potential misvaluation of shares and exogenous variables, which are outside the control of management and thus may not reliably reflect the agency problem faced by firms.
Recent empirical evidence also suggests that markets incorporate the level and changes in firm efficiency in the assessment of market value of firms. For example, Nguyen and Swanson (2009) develop a simple model that relates firm inefficiency to expected returns. They argue that in an efficient market with rational expectation market prices and stock returns should reflect firm efficiency, such as agency problem and financial distress. They provide evidence that firm efficiency exhibits significant explanatory power for average equity stock returns and further suggest that an (in)efficiency factor be incorporated in asset pricing models.
Alam and Sickles (1998) investigate the relation between stock market returns and changes in technical efficiency in the US airline industry and find a significantly positive relation between these variables. Consistent with this result, Eisenbeis et al. (1999) find that the stock returns of US bank holding companies are a decreasing function of cost inefficiency. In addition, Kohers et al. (2000) find that abnormal returns for US bank holding companies engaged in merger and acquisitions are positively related to (the target’s) X-efficiency as well as the difference in bidder/target efficiencies relative to their peer group (see, also, Wu and Ray, 2005; Cummins and Xie, 2009). Fiordelisi (2007) shows that banks lose around one-third of potential shareholder value due to inefficiency. Finally, the relation between firm efficiency and stock prices is not limited to the US. For example Chu and Lim (1998) analyse six publicly traded Singaporean banks and find that changes in stock prices are positively related to changes in profit efficiency but not related to changes in cost efficiency (see, also, Kirkwood and Nahm, 2006).
Our bank efficiency score that we obtain from a data envelopment analysis of banks provides a simple, yet direct, measure of management use of firm resources to generate desired outputs, relative to the best firm(s) in the same industry. Stigler (1976) first sug- gests the conceptual link between (productive) efficiency and agency cost. Later studies show that firms that seem profitable relative to other firms in the same industry are not necessarily efficient in applying resources to generate optimal outputs. For example, Sherman and Zhu (2006) note that studies of benchmarking practices with frontier ana-
lysis, such as DEA have identified resources inefficiencies in some of the most profitable firms.
Frontier analysis evaluates how close a firm is to the best-practice firm facing similar exogenous conditions and utilising similar inputs to produce similar outputs. We argue that the performance of a best-practice firm under the same exogenous conditions is a reasonable benchmark for the expected performance of the firm if agency costs were at their minimum level. While agency problem would not fully explain the differences in efficiency among firms and while all banks are likely to face some level of agency cost, we argue that inefficient banks are more likely to exhibit higher overinvestment problem that suggests a role for dividends. Managers of inefficient banks are more likely to engage in excessive consumption of perquisites (expense preference behaviour) that increases the inputs of the firm while at best producing the same level of outputs.
The effect of this preference behaviour is the reduction in the efficiency of the firm relative to the value maximising bank. Management of inefficient banks may signal any improvement in the level of efficiency or a desire for a change of course of action by increasing the level of dividends. Similarly, a reduction in the level of dividends signifies management’s desire to continue or exacerbate the agency problem. Either way, in an efficient market, we expect a positive (negative) market reaction to the “efficiency news” contained in the dividend increase (dividend decrease). Therefore, if resolution of the agency problem motivates bank dividend payments, we would expect less efficient banks that are characterised by high agency problem to increase their level of efficiency by reducing their inputs, for example, by disgorging excess cash through higher dividends.
The above discussion suggests the main hypothesis tested in this study stated in its alternative form:
H1: There is a positive (negative) relation between change in bank efficiency and market response to dividend increases (decreases)