Being in rating business in India can be a tricky position, especially in a country where information sharing or accessibility is a huge challenge. But understanding the dynamics of the market and finding your own space in the sun, can be useful for growth. Since 1993, CARE Ratings has been working with a diverse set of industries, despite lack of transparency and information among these entities.
While there were some failures, but there were more successes in terms of its rating, says Executive Director of CARE Ratings, T N Arun Kumar. In an interview with Jescilia Karayamparambil and R N Bhaskar, Kumar talks about the growth and credibility CARE Ratings has achieved over years.
Edited Excerpts:
Where does Care Ratings stands today?
We are one of the seven rating agencies in the country. We are the third rating agency to come into existence. But, today, we are the second largest in terms of revenue.
We have a presence in all industry segments in India. We have around 8,000 – 10,000 outstanding ratings today. We have a strong presence among sectors like banks, NBFCs, infrastructure, metals, etc. Like all ratings companies in India, we too are regulated by both SEBI and RBI.
We have been in existence for last 25 years and during that time frame we have witnessed a lot of cycles— pre-liberalisation and post-liberalisation. While the agency has witnessed a lot of cycles since 1993, currently the fintech sector is gaining prominence.
Does a shift in prominence affect rating agencies?
Ultimately, we are answerable to investors. The entire gamut of regulations has been created to make processes transparent and provide higher disclosure.
The job of rating agencies is not to react post the event, but to find out more about the event before it happens. We are supposed to be fortune tellers. We tell the fortune based on the information that we have access to at that given point of time.
About 99 per cent of time we are correct, but in one per cent of the situations we are wrong. These are measured by transition and stability rates (which currently range from 85-100 per cent). ‘AAA’ is 98-99 per cent stable; ‘AA’ is around 92 per cent and ‘A’ is 85 per cent stable.
How prevalent is rating arbitrage in the ratings business?
Ultimately, all the ratings are in the public domain and that includes some unacceptable ones too. We are measured in terms of what we give and what others give us. Ultimately, it all depends on how reliable you are. That in turn depends on how reliable and stable the rating has been.
How do ratings and pricing work?
Our regulations are such that business development and ratings are in different silos. There is a clear distinction that any person who is involved in revenue cannot be involved in rating. And this differentiation is widely respected. SEBI has come out with more strict rules around it and painted it black and white.
Can a company negotiate the rating and rates?
The end user of rating is the investor and hence, a rating agency assigning higher ratings may not earn credibility with the investors. The investor will rather depend on a credible rating agency and credibility is built over a period of time and not overnight. Out of the seven rating agencies, the investors prefer top 3 agencies including CARE and no wonder, these three have the largest market share.
The Altman Z Score model is known for its accuracy. Have you’ll considered it?
If rating prediction was so simple, we will not need rating agencies with highly- qualified people. The Z score, like many others, is a simple market-based formula and it can work in a perfect market. But we do not have a perfect market, we have a bank loan market which is entirely opaque.
We have situations where a bank gives loans to an entity and we are not even aware about delays if any in debt servicing. Appropriate information is just not available. In our portfolio, around 10 per cent are listed companies and the remaining are unlisted companies. Getting financial results of the latter class is not always easy. Transparency levels are quite low in India.
How critical is human factor in rating?
We look at ability and willingness. You can determine the ability to pay. But it is not easy to determine someone’s inclination to pay.
The willingness factor is a soft aspect and that cannot be done by machines and it needs people like us. Artificial Intelligence (AI) is trying to creep in slowly. AI is an interesting tool but we do not like black boxes. We need to understand what is going on, rather than just getting the product and not know how it is done.
We are finicky about what goes into the model. We need everything to be systematic and cause driven. We are looking at solutions that will breakdown various elements (in rating). We have a proprietor framework for ratings which we use. It gives us some broad range of ratings and then there are human elements too.
All the ratings (in CARE) are done by committees. We have committee-based approach where a minimum of five people deliberate on the ratings to be given. If the issues involved are complex, the number of people in the committee goes up, The more complex the ratings, the larger is the number of members in the committee.
What would your wish list be?
We desire disclosure of the list of defaulters that banks and other lenders have. We would need a repository for such information. It would thus be similar to a Cibil, but for debt and defaults relating to companies, not individuals. This would require a lot of work but I hope it works out.
