Opinion | A much-needed revamp for rating agencies4 min read . Updated: 12 Nov 2018, 09:00 PM IST
With credit ratings threatening to disrupt markets, the systemic issues with rating agencies must be urgently fixed
Credit ratings are difficult to live with and just as difficult to live without. A lot of ink has been spilled about the fact that the bonds of Infrastructure Leasing & Financial Services Ltd (IL&FS) were rated investment grade as late as July 2018. The firm started defaulting on its debt in September. IL&FS is not unique.
In our analysis of a large sample of bank loan ratings in India, we find that firms that eventually default on their bank loans have an average rating somewhere between BB- and B+ one month before they default.
This matters because according to Reserve Bank of India’s (RBI’s) capital norms, banks have to set aside more capital to cover for losses only if a firm’s rating goes below B-. So, above B-, banks can allocate less capital to cover losses. The fact that Indian banks have less capital to cover loan losses should come as no surprise to any impartial observer of Indian banks.
The sad part about the events such as IL&FS default is that we know exactly what the causes of poor quality ratings are, and have a good sense of potential solutions. It is just that the regulators have been reluctant to embrace these solutions.
One can’t but think that aggressive lobbying by powerful rating agencies has a lot to do with that. In this piece, I outline the sources of the problems and the potential solutions.
Problem 1: Conflicts of interest from non-rating business.
To be effective, rating agencies have to be impartial judges of credit quality. This is not possible if they have other non-rating businesses that they are marketing to the same corporates who they are rating.
Solution: Rating agencies should be forced to divest from non-rating business.
This is an easy solution. I believe the Securities and Exchange Board of India (Sebi) is already trying to implement this, and on this count, kudos to the regulator.
Problem 2: Issuer-pay model.
According to current practice, the borrower picks the rating agency that will rate their debt. Thus, instead of an impartial judge, the rating agency is a seller trying to entice the buyer (the borrower) into buying its rating service. One can very well imagine what will happen.
Rating agencies will compete with one another to offer the best possible service (read: highest possible rating). Such rating shopping is formalized in some rating agency websites wherein the process of how a company can dispute its rating and obtain redressal (read: a higher rating) is laid out. This is crazy, but entirely expected.
Solution(s): a) The regulator should pick the rating agency, or b) greater disclosure.
There is a drastic and less drastic solution to this problem. The drastic solution is for a regulator (say RBI or Sebi) to pick the rating agency.
This can be easily automated with the help of a computer programme. The programme will pick the agency randomly while ensuring a target market share for the different agencies. The target market share, in turn, should depend on the accuracy of the ratings in the past.
Evaluating rating accuracy is easily done and the agencies do this on an annual basis already. The target market shares should be decided in such a way that it rewards good performance and penalizes bad performance.
The less drastic solution is to encourage unsolicited ratings, and also to ensure that the borrowers disclose all ratings and not just the ratings they end up accepting. This will give greater confidence to the rating agencies to tell the truth and not have to cater to the borrower.
Problem 3: Over-reliance on ratings in regulations.
The main source of power for the rating agencies are regulations that are tied to ratings. For example, the Basel capital adequacy norms (which India has adopted) allow banks to decide on their capital allocation based on loan ratings.
Following India’s adoption of these norms, we document a multifold increase in the number of firms with credit ratings. Thus, regulators reward rating agencies with more business.
Solution: Provide alternatives to ratings.
One should try to reduce the reliance on ratings in regulation. For example, banks should have a minimum amount of capital based on the total assets, and not just on risk-weighted assets. This will reduce the importance of ratings for bank capital.
The RBI should also insist that bank borrowers obtain multiple ratings with the regulator picking the agency. Competition combined with reward for good performance will enhance the quality of ratings. While some of the solutions proposed above are based on research in the US, the problems with credit ratings is based on our research in India. In our paper Assessing the Quality of Bank Loan Ratings, we find that credit ratings for bank loans in India are a ticking time bomb.
To assess rating quality, we compare the ratings of unlisted firms with those of listed firms, wherein the public equity market provides an independent check on firm quality. For example, if a firm’s stock price is crashing, it will be very difficult for the rating agency to not revisit the firm’s rating.
On the other hand, no such external discipline exists for unlisted firms. Consistent with this, we find that the ratings of bank loans of unlisted firms are of worse quality than those of listed firms.
Given that a large fraction of bank borrowers are unlisted, this is a problem which the banks and the central bank cannot afford to ignore for long.
Radhakrishnan Gopalan is a professor of finance at Olin Business School.
Comments are welcome at firstname.lastname@example.org