금융실명법의 경제학적 분석:게임이론의 응용

This research analyzes financial customer identification mechanisms using signaling game theories and draws policy implications. Financial customer identification problem has been an important issue since in particular the launch of FATF (Financial Action Task Force) in 1989 for anti-money laundering & terrorism financing and related predicate offenses such as false identification of financial customers. Recently it becomes more important because of the development of financial instruments such as cryptocurrencies that inherently lacks the customers identification. FATF recommends registration-based identification for cryptocurrencies. Financial transactions with instruments lacking correct customer identification can hurt the Know-Your-Customer policy and market integrity, and put financial markets into illegality traps. This research utilizes game theories such as Spence (1973), Cho and Kreps (1987), Stiglitz and Weiss (1990), Bolton and Dewatripont (2005), Che, Kim and Kojima (2015), Riley (2001), Sobel (2011). The concept of separating equilibrium of game theory can give us useful intuition how to disconnect illegal transactions from sound transactions, and how to evaluate the efficiency and effectiveness of identification mechanisms. This study classifies customer identification mechanisms into two forms: ‘non-registered identification mechanism’ (N-form) and ‘registered identification mechanism’ (R-form). N-form only requires to identify named customer, whereas R-form requires to identify both named customer and beneficial owner. Beneficial owner may be either G-type (good-type that never tries illegal transaction) or B-type (bad type that always tries illegality). Regulator’s monitoring level may be strong or weak. Because of the information asymmetry problem between regulator and beneficial owner about who beneficial owner is and whether he/she intends illegal transactions, N-form fails to reach at separating equilibrium. However, R-form could reach at separating equilibrium by removing the asymmetry problem, and achieve social efficiency of minimizing illegal transactions. Example of N-form includes The Financial Customer Identification Act of 1993 of Korea, whereas that of R-form includes the Customer Identification Program (CIP) of Bank Secrecy Act of the U.S. and ‘the explicit mutual consent among named, beneficial customers and banks’ that was adjudged by the Korean Supreme Court Case of 2019: ADJ 2008-Da-45828. R-form is also supported by FATF, the U.K., Canada, Australia etc. In R-form, signaling can be sharper if penalty scheme is arranged incentive-compatibly and proportionately to the illegality. Penalty scheme includes civil money penalty and/or criminal penalty, and the shift of burden of proof from regulator to the detected and presumed B-type beneficial owner who does not register. In some situations, N-form may not readily transform to R-form because of institutional limitations. In order to minimize the drawbacks in the policy effectiveness of N-form, two policy variables should be reinforced. First, regulatory authority should increase the evidentiary detection probability of unregistered B-type even if it is a harder job in N-form than in R-form. Second, appropriate level of punitive penalty should be charged to disincentivize unlawful motivations by beneficial owners.