So, what explains the poor track record of India’s rating agencies? The explanations vary - sheer incompetence to inadequate information access to conflict of interest. The last one is the most disturbing, though all CRAs say the conflict of interest issue does not apply to them as they already have firewalls separating the ratings business from the non-ratings business.
But the allegations persist. CRAs produce ratings used by investors, but obtain most of their revenue from issuers, both as ratings fees and as payment for other services. This leads to a potential conflict of interest.
A paper by Ramin P Baghai and Bo Becker in February 2016 ) employed a detailed panel data set on the use of non-rating services as well as payment flows between issuers and rating agencies in India to test if this conflict affects credit ratings.
The findings were startling. A CRA that receive non-rating revenue from a firm issue a better rating to that firm than agencies that are not paid for such services by the issuer. This effect is increasing in the revenue generated. The figures show that the more an issuer contributes to the total revenue of a rater, the better is the rating that the issuer receives from that agency, on average.
The study also found that, within rating categories, default rates are higher for firms that have paid for non-rating services. This suggests that the better rating that such firms receive does not reflect lower credit risk. Overall, the authors interpreted the results as consistent with a fee-driven conflict of interest between CRAs and security issuers: When an issuer is directly important to an agency through the fees it generates, then ratings are upward biased.
Despite such damning findings, it’s unlikely that the issuer-pays model can, or will, be junked. It didn’t happen in the US where it was widely criticised after the subprime crisis. There are reasons for this. While the issuer-pays model has conflicts of interest baked into it, the fact is that other leading candidates for replacing it also have downsides. For example, the investor-pays model was considered by the Securities and Exchange Commission in the US, but wasn’t adopted.
Here’s why. If investors pay for ratings, only those who pay will get access to ratings. Investors would also not be able to benchmark the quality of their investments against other companies, since they may not be willing to pay for ratings of companies in which they do not invest. The other model of regulator or the government paying for ratings has been found to be impractical.
This means issuer-pays is a model that will stay for some time at least. So the only way to address the concern is to bring in greater transparency to the process. The Securities and Exchange Board of India (Sebi) has done some work on this already, but clearly, more needs to be done.
One option the regulator should look into is increasing the legal liability of CRAs on the lines suggested by the Dodd-Frank Act. For example, Sebi can make CRAs liable for compensating investors for any loss caused to them by negligent or fraudulent rating. But adequate safeguards have to be built in, as CRAs could challenge such a rule in court on the ground that ratings are mere opinions.
The biggest reform, however, will be to ask CRAs to separate their rating entity and their non-rating businesses in order to avoid conflict of interest. The market regulator is reportedly planning to mandate that there should not be anything linking the two other than shareholding and that there should be no repatriation of dividend or profit from the non-rating businesses to the company that owns the rating business. Sebi should implement this — fast.