臺大管理論叢 NTU Management Review VOL.30 NO.2

177 NTU Management Review Vol. 30 No. 2 Aug. 2020 regulators to disclose ICWs are further supported if such disclosures facilitate firm valuation. Previously, there are no expectations to find similar results to those documented by prior studies for the following reasons. First, Beneish et al. (2008) use data only from firms that disclose at least one ICW, whereas the data in this study include both firms with zero and non-zero ICWs. It is reasonable that investors view the difference between zero and non-zero ICWs to be more salient than the difference between two non-zero ICWs. Second, both Beneish et al. (2008) and Li et al. (2016) base their analyses on U.S. firms, and the U.S. capital market is considered to be highly efficient. Therefore, in such a mature market, the information in ICW disclosures is more likely to be revealed in advance through other channels, whereas in markets that are less mature, such as in Taiwan, ICW disclosures may still be relevant. Most importantly, the present study differs from extant literature by being the first to posit that ICW disclosures may be associated with positive value effects, which are interpreted as the benefit of disclosing ICWs. More specifically, it is less reasonable to expect that a firm would have zero ICWs; therefore, when a firm reports zero ICWs, it is reasonable to expect that such firm does not thoroughly assess its internal control effectiveness, and will not have the opportunity to improve its performance as a result. In contrast, when a firm reports non-zero ICWs, it indicates the presence of deficiency, but also suggests that the firm takes the internal control process seriously. In summary, this study presumes that reporting non-zero ICWs reveals higher level efforts in assessing internal control effectiveness and should be positively valued by the market. This argument may appear counterintuitive because it seems more plausible that ICWs should be priced negatively. This study does not refute that possibility, but when a bank reports the presence of ICWs, investors may perceive: (i) the bank has some deficiencies in operations, and (ii) the bank takes the process of identifying and reporting ICWs seriously, i.e., the bank is diligent in internal control processes. The former would negatively affect how investors value the bank, and the latter would positively affect the bank’s value. The “net effect” of reporting non-zero ICWs depends on which effect is dominating. If it is empirically found that reporting non-zero ICWs is positively valued by investors, it will correspond with the notion that the benefits of reporting non-zero ICWs outweigh its adverse effects. The empirical analysis is conducted in the following ways. First, to avoid endogeneity problems, we employ Heckman’s two-stage method (Heckman, 1979). In the