Internal Risk Rating – an antibiotic badly administered

Internal risk rating (IRR) exercise undertaken annually by banks is like an antibiotic as their credit officers can thereafter conscientiously discard borrowers with a poor prognosis.

Rating process

The IRR exercise is intended to rate all standard accounts and the Rating Managers at banks begin the process of Rating as soon as audited financials are received from the borrowers after tax returns are filed for the year. I would suggest on the basis of experience that Rating  Managers can schedule the rating work to begin from September and complete by December every year. 

The step by step procedure for rating to be followed by  Rating Manager  is detailed (or expected to be) in the Rating manual of banks. (see below). It is also advisable to take up individual proposals and finish them in one go since information from various sources has to be pieced together to perfect the result. 

Ratings are given on the basis of the score and are given in sequential number.  Initials of the bank are prefixed to the number to distinguish the proprietary nature of the code used for rating. e.g. A1, A2,..A15 etc where A indicates the name of the bank. The rating range starts from the lowest risk category to the highest, with the lowest risk category being assigned the top most rating (e.g. A1) and vice versa. The bottom-most rating (say A15) may indicate a bad and doubtful case.

Once the rating is over, the rating has to be kept valid for a period of one year since rating data is used for analysis has to be constant. Any drastic development since the rating process is over, that calls for urgent downgrade has to be managed at the user level rather than at the database or the bank level.

Making the best use of IRR exercise

No doubt, the IRR forms an integral part of the Risk Management system of banks and ratings are linked ideally to the risk premium in the rate of interest ultimately offered to the borrower. Not only that, Credit officers consider the rating for risk sharing options such as higher margin requirement, additional collaterals or consortium arrangement with other banks.

Rating at the whole bank level finds many uses. The data is used for studying something called ‘migration analysis’ and for preparing Portfolio quality index that helps drafting the Annual credit policy.

Migration analysis

Briefly, migration analysis helps in finding the trend in the movement of outstanding credit from one risk category to another. This is done by comparing the aggregate outstanding credit against each rating for two or more consecutive years and studying the variation in their proportion to the total credit outstanding. If the proportion of low risk outstanding shows an increasing trend, it indicates an improvement in the quality of credit outstanding. Alternately, increase in the proportion of high-risk category credit outstanding indicates an alarming trend that calls for appropriate change in strategy for sourcing new business.  

Portfolio Quality index

After isolating outstanding credit on the basis of various industries, geographical regions, sectors etc., outstanding against each rating is worked out. The result can be further analysed in detail to find which industry, region or sector is providing accounts with top ratings. Large borrowers can be studied in isolation. A lot more can be done with the data comparing with previous years. The results can provide the management with insight on the industry, sector or region that are performing well. Accordingly, banks can draw strategy for balanced growth and income. 

When management is apprised of the pattern of distribution of credit into various industry, sectors, regions etc and the profitability there from, they can prepare the annual credit policy guidelines for the ensuing year after taking also into account the government directives and national priorities. 

Sources of data

Three most significant contributors for a successful risk management system are the data annually provided by the borrower, historical data available with the bank and the skill level of Risk raters. Therefore, Risk raters must collect data promptly and complete the rating exercise for all eligible accounts on time. Similarly, banks should maintain at least five to six-years data of borrowers (need not of every borrower) useful for analysis. Here are some of the main data sources:

Audited financials and annual reports to shareholders of borrowers are in their custody and can be easily obtained. However, delay in receipt can cause an overload to the Rater and impact the quality of rating. 

Second set of information can be gathered from regulators and other monitoring authorities. Information relating to business units are available in India for public use from RBI, SEBI, stock exchange, Govt sources etc. Risk rater’s sixth sense while going through audit remarks on litigations, sovereign debts, group concerns, directors etc. helps him to do full justice to the job.

Third set of information is on the borrowers’ management skill. Risk raters evaluate their policy, structure, commitment, ethics and social standards. It is ideal for Risk raters to maintain a file for market intelligence on an ongoing basis. Supplementing this, banks can centrally store such information and provide access to all Risk raters. 

Fourth set of data can be gathered from reports from various management associations, chamber of commerce and authentic social media.

External rating agencies store vast data and have a lot of experience in data analysis and forecasting skills. The rating done by external rating agencies as of now are applied for working out the risk weighted value of assets of banks to ascertain the bank’s Capital Adequacy. The rationale for rating awarded is a relevant important input for IRR.

Rating manual 

Having discussed about the importance of IRR, process, uses and data sources, let me mention a few important points about the document that plays a pivotal role in risk management. The manual should contain the details of the rating, score, the process of arriving at the score and the authority of approve the rating and review it. Since a lot depends on the manual, I emphasise that the content has to be comprehensive and must not leave opening for frequent revision. Every revision makes the historical data useless for comparison. 

The style of writing adopted for the manual must be simple, self-contained and shall not call for third party interpretation. The consistency of data is as important for data analysis as regularity, integrity and accuracy. If consistency has to be ensured, the structure of rating also must be left unaltered. Too technical a manual would be abandoned altogether without reading it and the whole IRR fails to return the expected results. 

Coming to the end of this discussion on IRR, readers would agree that IRR is like an antibiotic that removes the harmful bacteria.  It is for the readers to find out if there exists a correlation between banks struggling with NPAs and their quality of annual IRR exercise. IRR is meant for commercial banking. Therefore, one would find banks exclude personal banking and investment banking from IRR exercise. 

Some practical issues with data for rating, appraisal and external rating have been discussed in my book ‘Banking India’ in chapter 10: ‘Uncertainties’. A lot more details I feel could have made this story more interesting. The purpose being to familiarise readers with some rarely discussed but very significant, pains in our banking system, I presume I have covered it to generate interest on the topic. I shall look for your comments and respond to your comments gladly. Thank you.