Asset quality and bank loan ratings
The high level of non-performing loans (NPLs) in banking is the most important issue facing the Indian economy currently. One aspect of the NPL issue that has escaped attention is the role played by bank loan ratings. Indian banks extensively use ratings provided by external (rating) agencies as a part of the loan sanctioning and monitoring process. Any assessment of risk management in banks will be incomplete if we do not critically evaluate the role of these loan ratings.
Under the Basel II regime, banks are allowed to use external ratings—those provided by approved rating agencies—for their loans. Such ratings are expected to be a key factor in determining whether a loan is approved or not and also the pricing, i.e. the interest rate to be charged on the loan. There is another important use of the loan ratings—they determine the capital banks need to provide for the loans. Current capital adequacy norms link the “risk weight” of a loan—and hence the capital banks must provide for it—to the rating of the loan. The higher the rating, the lower the risk weight.
In India, the risk weight for AAA-rated loans is 20%, for AA-rated loans is 30%, while for BBB-rated loans, it is 100%. This means that for a BBB- and a AAA-rated loan of the same amount, capital provided for the latter would be 20% of that for the former. The regulator has also mandated that all loans over Rs.5 crore must be rated. Consequently, all loans over Rs.5 crore and a significant proportion of those with even lower amounts are rated. In addition to the rating done at the time of sanctioning the loan, rating agencies also provide what are called “surveillance” ratings, which are annual updates on the loan rating. My conservative estimate is that around 30% by volume and 70% by value of all commercial loan accounts (i.e. non-consumer loans) of Indian banks are rated. These ratings are a sizeable business for rating agencies and earn them revenue of around Rs.300 crore annually.
It is reasonable to expect that at the time of sanctioning the loan, it has acceptable rating, indicating solvency of the borrower. Loan quality typically deteriorates slowly and there are clear signals, such as delayed payment of loan dues and changes in the business environment that impact the borrower adversely. The surveillance rating of such a loan, if done well, should pick up these signals and reflect in a lowered rating. This is similar to a downgrading of bond rating that takes place as the financial condition of the bond issuer worsens. Just as the pricing in the bond market reflects change in the rating of a bond—reduction in rating results in reduction in the price of the bond—we should expect a similar impact on the bank. Banks should try to adjust the pricing to the new rating. Importantly, we should expect the regulator to reassess the capital requirement on the re-rated loans, and demand higher capital in case the loan rating has gone down.
If on the other hand, the surveillance ratings are not picking up quality deterioration in loans or if the banks are doing nothing on such ratings downgrades, then we must question the whole purpose of using ratings in the first place, including the use of rating in the initial sanction decision. More importantly, the regulator must examine the logic of linking risk weights to the ratings of loans.
The entire edifice of ratings rests on their credibility, which in turn depends on the credibility of the agencies that provide the ratings. In the bond market, this credibility is achieved through disclosures. Bond issuers not only have to disclose their ratings and changes in them but also the agencies that have provided the ratings. Based on their experience, bond investors form their own view on the relative credibility of the rating agencies.
However, unlike bond ratings, bank loan ratings are not uniformly disclosed by banks. This raises the question—why should the bank’s investors (and depositors) not know about the ratings and rating changes as do the bond investors? After all, they are investors in a portfolio of loans that banks make. Investor presentation of banks sometimes describes the rating profile of the sanctions made in the preceding quarter or year. Occasionally, they also give the rating mix of their loan portfolio—the percentage of loans in each rating bucket at the time of sanctioning of the loan. There is no disclosure of the rating migration—how the ratings of the loan portfolio have changed since the sanction. Nor is there any disclosure of what agencies are providing these ratings.
To ensure credibility of loan ratings to justify linking capital requirement to them, it is imperative that disclosures of loan ratings and changes in them and the agencies providing the ratings are made mandatory.
Specifically, the regulator should ask the banks to disclose every quarter:
l Rating mix of all outstanding loans that are rated
l Changes in the preceding quarter in the overall portfolio ratings, essentially a rating migration matrix
l Details of the agencies that the bank has used to rate the loan portfolio—the number of loans rated by each agency and agency-wise rating mix; and
l For all the loans that become non-performing, a history of their rating along with the names of the agencies that provided the ratings
An important pillar of the Basel risk framework is “market discipline”, which is the pressure exerted by the market on the management of banks to manage risks better. Such market discipline can exist if there are adequate disclosures.
Let us hope that the current non-performing asset crisis encourages the regulator to seriously look at the role of bank loan ratings and, at the very least, compel significantly better disclosures about them.
Harsh Vardhan is partner, Bain & Company. These are his personal views.
Comments are welcome at firstname.lastname@example.org