우리나라의 복수 신용평가 의무제도는 신용평가 시장의 경쟁을 제한함으로써 신용평가 산업의 발전을 저해시키고 있다는 주장이 제기되고 있으며 이에 따라 정부도 동 제도의 폐지 여부를 구체적으로 검토하고 있다. 그러나 최근 글로벌 금융위기를 초래한 원인 중 하나로 국제 신용평가사들의 역할부재론이 지적되고 있으며 특히 채권발행사의 신용등급 쇼핑 등이 신용평가가 부실해지는 가장 큰 이유로 분석되고 있어 복수평가 의무제도의 존폐에 대해서는 보다 정밀한 검토가 필요하다. 이를 위해 본 연구는 Bolton et al.(2011)의 모형을 기초로 하여 복수평가제도와 등급 쇼핑 간의 관계에 대해 이론적으로 분석한다. 분석 결과 복수평가의무제도의 도입은 신용평가사들이 항상 인플레된 등급을 발표하는 균형의 가능성을 높이는 반면 신용평가사들이 정직하게 신용등급을 발표하는 경우에는 발행기업의 등급 쇼핑을 제한함으로써 오히려 사회후생을 증가시킨다는 결과를 발견하였다. 늘 인플레된 신용등급을 발표하는 신용평가사는 현실에서 유지되기 어려우므로 복수평가제도의 존재가 사회후생을 증진시킬 가능성이 더 현실적인 것으로 보이며 결국 복수평가 의무제도의 존폐 문제에 보다 신중하게 접근할 필요가 있다. 이와 같이 복수평가가 단수평가보다 발행사의 등급 쇼핑을 감소시킬 수 있다고 하더라도 등급 쇼핑 유인을 완전히 제거할 수는 없으며 이에 따라 본 연구는 보다 정확하고 중립적인 신용평가 서비스가 이루어질 수 있기 위한 단기 및 장기 정책과제를 제시한다.
In Korea, all issuers of corporate bonds are required to obtain credit ratings from two or more credit rating agencies. Since the Korean credit rating market is dominated by three big agencies, many industry commentators have argued that the requirement can stifle competition in the credit rating market. Reflecting these critiques, the Korean government has been reported to consider repealing the requirement. The global financial crisis, however, showed that rating shopping and conflict-ridden rating processes could threat the stability of financial systems. Furthermore, according to Benmelech and Dlugosz (2010), tranches that are rated by just one agency, a characteristic that is consistent with ratings shopping, are more likely to be downgraded, and more severely at that. In the light of these experiences, the Korean government should be careful in deciding whether to repeal the requirement of multiple credit ratings. This paper provides a theoretical analysis about the effects on rating shopping of the requirement based on the model of Bolton et al. (2011), a credit ratings game among issuers, investors and credit rating agencies. We find the introduction of the requirement of multiple credit ratings can increase the possibility of the equilibrium where all the credit rating agencies always announce inflated ratings. However, the requirement can increase the social welfare in the equilibrium where all the CRAs always tell the truth. Since the former type of equilibrium is not sustainable dynamically, the latter type of equilibrium is more realistic and sheds more light on the decision about the repeal of the requirement of multiple credit ratings. The direct policy implication from these results of our theoretical analysis is simple. The Korean government needs to retain the requirement of multiple credit ratings in order to reduce rating shopping by issuers. We also discuss other relevant policy issues such as credit rating fees, disclosure of rating process and reliability of the credit ratings. First of all, issuers need to be prevented from paying for specific ratings and they should pay the CRA upfront before it does its initial analysis. If bond issuers pay CRAs upfront, not contingent on the report, the CRAs are likely to estimate a more correct rating. An independence of credit rating is hardly established in the current issuer-pay model in which CRAs are paid on ratings issued. Furthermore, an innovative method for fee payment through rating platform is recommendable in that the central platform can solve the conflict of interests between issuers and investors. The platform would be paid pre-issue fees, assign CRAs, and allocates fees to CRAs regardless of outcomes. In other words, the platform would be completely in control of the rating process. Second, issuers should be required to disclose all credit ratings including preliminary ratings from other agencies. If the supervisory authority enforces issuers to announce all CRAs who are requested to evaluate the credit rating irrespective of their acceptance and to pay for even preliminary or unofficial rating, it will decrease the incentives of the issuers to contact CRAs unofficially and solicit better credit rating. This requirement can prevent the issuers from shopping ratings and allow investors to take rating shopping information into account in pricing the corresponding securities. Third, transparency of rating process should be enhanced. For this objective, the supervisory authority needs to set the standards that will make CRAs provide more exact information related to rating evaluation, including applied models, assumptions, sensitivity with respect to rating variability, reasons of rating change Finally, relative performance statistics for CRAs must be made transparently and clear. Social welfare is observed to increase only if CRAs announce truthful rating, according to the results of the paper. An obvious path to this end would be for the government to create a centralized data repository and to standardize performance histories so that the track record of CRAs could be easily compared by investors at one centralized site.