Credit Rating Reform, Part II: Forging a Hybrid Model

Regulators tasked with reforming the credit rating industry have been hesitant to rock the ship, taking great pains to explore every possible option.  Furthermore, the fact that structured products are the main target of scrutiny, yet have been the most profitable line of business for credit rating agencies presents a considerable headwind for reforms.  As a result, talks have slipped beneath the waters, floundering along in slow motion.

Recently, we discussed a prominent “rating clearinghouse” proposal that has raised industry objections.  For a brief review of the options before regulators up to this point, please see Table 1.

Table 1: Fixing the Credit Rating Industry

Source: SEC.

The debates of the last few years have put a spotlight on the weaknesses of a range of proposals.  With that in mind, the challenge for the SEC will be to construct a model that effectively erases the conflict of interest problem, yet avoids regulatory overreach.  It must further be operationally feasible and affordable to implement.  By identifying the vulnerabilities shared by the various reform proposals, regulators may forge a more effective model from their midst.  Whatever the eventual model, it should take heed of those design constraints. 

First, it needs to effectively prevent conflicts of interest and “rating shopping,” in which issuers seek out the agency that offers the most favorable appraisal.  In order to do that, it must actually be utilized, unlike the inert Rule 17g-5 program, an early effort which failed to motivate agencies and investors to participate.  Also, solving one conflict of interest could create new ones if incentives are not managed carefully.  The investor-owned models run this risk, as investors have an interest in seeing certain rating results, just as issuers do.  One important distinction to make, when designing a solution, is whether the conflict of interest stems more from the payment mechanism (i.e., who pays the agency), or from the selection mechanism (i.e., who selects the agency).  

In the process, it should take care to avoid regulatory overreach.  As is often the case with legislation and regulation, policies can be too ambitious or aggressive, and have unintended consequences.  In this case, investors’ traditionally free access to ratings should be preserved, competition within the industry should not be undermined, and issuers should not be overly burdened. 

Third, it should be operationally feasible.  The complexity of rating deals under the new regime should not overwhelm regulators’ capacity or an assigned agency’s expertise.  Without the requisite experience, a central clearing party, or assigned agency in any one deal could cause delays, and the quality of service could decline.  Therefore, whatever system is put in place must speedily and efficiently handle the deal type and flow, and adequately compensate the rating agencies.  Preferably, it wouldn’t require legislation to enumerate new powers in order for it to work.  Questions of cost and feasibility have tended to plague the section 15E(w) program, as well as the stand-alone and designation models.    

Last, it should be affordable to implement.  Overly prohibitive costs could include the extensive resources required to form a new oversight body, or even a serious loss in market efficiency.  Even increasing the size of a regulatory agency may be undesirable at a time when government is generally trying to do the opposite.  

One approach that could effectively marshal these lessons is a hybrid model, in which issuers and investors are both stakeholders of a fund that is only responsible for collecting fees from new issuances (from issuers) and secondary transactions (from investors), and for disbursing payments to selected agencies.  Unlike the clearinghouse model, the fund would have no active role in selecting the agencies.  Rather, agencies would simply initially be placed in a continuous queue to receive rating assignments, unless they were unable to rate a particular issue.  

Along with randomization, an agency’s capacity for rating complex structured transactions would be built into the process and determined at the time of NRSRO eligibility, rather than decided later by a centralized body.  In the future, assignments would be ruled by performance.  

Chart 1: Severing the Links that Create Conflict of Interest

This solution would cover quite a few of the constraints listed above, and takes a step towards satisfying the question of feasibility.  It separates many of the links between issuer, agency, payment, and selection that typically generate conflicts of interest; and by bringing together issuers and investors as stakeholders, it balances the various conflicts of interest between raters, issuers, and investors. 

 Although it might tradeoff deal speed and volume because it would be blind to individual NRSRO competencies, it also avoids the burden of having to create a new government oversight board or clearinghouse responsible for determining assignments.  Queuing up agencies that are pre-qualified to rate structured deals would still achieve the random assignment effect.  

Though imperfect, such a hybrid model would be a step forward in a dormant effort, with strengths that reformers can build upon.