Deep Mukherjee of Indiaratings, writing recently in the Economic Times, proposes three ways in which Credit Rating Agencies (CRA) may be better regulated. He likens the CRAs to Universities (because both models rely on the evaluated paying the evaluator) and proposes the following reforms in regulation of CRAs.
First, he argues that the debt-issuers' ability to forum-shop for a "good" rating should be taken away. He pitches for regulation that would make it mandatory for the debt-issuer to disclose all the rating outcomes, regardless of whether the issuer has "accepted" the rating.
Second, he argues that the end-users of the rating should be able to "rate" the ratings through ways which are publicly accessible. Investors investing in the debt instrument should be able to provide public feedback to the CRAs; for instance, they might benchmark the CRA-designated rating with their own internal rating on the security. (thus presumably, if repeated defaults and recurring spreads between the internal ratings and CRA ratings become evident, the CRA in question will lose credibility and potentially go out of business over time)
Finally, he argues regulators should mandate disclosure of default by issuing entities because Institutional investors in debt have perverse incentives NOT to disclose the default, because (contingent on extent of exposures), the market may discount the investors' stock were the default to be made public.
These are all valid points and the analogy to the Universities very apt. The first and the third appear arguably feasible to be implemented; however, I have doubts about the second proposal. Investors inn debt, like insurance companies, pension funds etc rarely have the internal expertise to evaluate the probability of default and loss given default, rigorously. This is because, prudential limits constrain their exposure to the issuer and not having "skin in the game", they 'd rather buy a rating from a CRA than develop internal rating expertise. Banks are much more likely to develop the expertise necessary, but then they 'd rather invest in the issuer through loans than through debt paper (in part, owing to the dichotomy in the accounting rules that treat loans and bonds differently).
One interesting model for mitigating conflict of interest at the CRA end based on the "skin in the game" idea, is for regulation to mandate that a certain portion of the fees paid to the CRA be paid in the securities it is rating. The proportion of the total fee-mix that will be paid in the "rated paper" will increase as the risk related to the cashflows underlying the debt increases. Thus, the CRA may be paid, say, 10 % of the fees in the form of paper if the paper is senior. On the other hand, if the paper that it has rated is mezzanine, 20 % (say) of the fees will be paid in terms of rated paper. Because it will be paid in the paper that it is rating, this proposal arguably creates the incentive to rate the paper genuinely and mitigates the conflict of interest that the "issuer-pays" model generates. An analogy for this model is Independent Directors. IDs are not unlike CRAs in that they also "signal" the corporate governance risk in the companies. Further, they too are paid by the company concerned and appointed and fired by the Company. However, one way regulators ensure they send out the "right" corporate governance signal is by permitting the company to compensate them in stock. (though some countries like India do prevent companies from compensating them in stock options, whether rightly or wrongly). However, compensation in stock per se is permitted and this partly ensures that the ID watches over the management of the Company objectively because having her own personal wealth locked in the company through stock, the ID is doubly motivated to ensure that the Company does not commit any act of malfeasance.
No comments:
Post a Comment