RAROC – An aid for credit risk evaluation

Reported by: |Updated: February 20, 2020

Siddharth Mahanty, faculty and senior manager at Sir Sorabji Pochkhanawala Bankers’ Training College of Bank of Baroda

Credit risk is the most popular form of risk by virtue of it being the most pervasive and pronounced among other forms of risk. The understanding of credit risk in the most basal form is limited to the concept of risk emanating from default in payments. While credit risk is vast, imposing and has many facets to it, one of the most challenging aspects within credit risk is the risk based pricing.

 The earlier concept of ‘cost plus profit pricing strategy’ adopted by banks may have worked in the shorter run but not in the longer run as borrowers looked for alternatives. Such pricing worked when banks had the competitive advantage on array of services and not only on price alone. In a developing market economy like ours, banks have never been price leaders but price takers.

In the guidance note prepared by RBI on Credit Risk Management, it is inferred that dual rating systems may be currently in practice due to the familiarity and ease but will certainly be overhauled on account of a major shortcoming – the condition of either default or no default. There is no intermediate level to suggest the divergent attributes in an account. Furthermore, there are various rating models available across banks that assign a risk premium on the basis of the rating achieved through their internal rating models. These models are partly subjective in pricing the borrower accounts. The purpose of pricing is diluted on account of interest rate concessions that are allowed to sustain in the business of lending.

So, the situation could be likened to ‘addressing the elephant in the room’ ie, are we actually pricing the borrowers on the basis of the inherent risk they carry? We observe that an account considered riskier is priced higher, but the question is whether a riskier* customer can service the higher costs in the long run. Another question is whether the pricing takes care of the provisions, overhead expenses, cost of funds etc. Thus, the interplay of so many uncertainties has led the banks to embrace the concept of Risk Adjusted Return on Capital (RAROC). This concept is widely used across banks today for performance measurement, wealth maximization, capital allocation and is also a prerogative in the EASE framework of RBI

The concept has gained huge popularity because of its reliance on Economic Capital (EC). While it is even challenging to foresee the expected loss in the normal course of business, EC ascertains the loss that is beyond the realms of expectations ie, unexpected loss.

What exactly is RAROC?

RAROC= Adjusted Income/ Economic Capital (Unexpected Loss).

(In SME accounts, the trend across industry is to rate them just above cut-off grade.)

Where Adjusted Income= Income- Expenses (including Expected loss)

Income=Interest Income + Non Interest Income+ Ancillary income*

Expenses= Cost of Funds + Operating Expenses+ Expected Loss

Expected Loss= Probability of Default (PD)* Loss given default (LGD)

Economic Capital = Capital at Risk at a certain confidence level (VaR based*)-Expected Loss

– VaR based capital at Risk is the maximum worst loss one can incur at a certain confidence level say 99% & at a certain period say 1 year. It incorporates the various correlations between portfolios, thereby diversifying the risk.

– The ancillary income is the income incidental to sanctioning a loan (up- selling, cross-selling) and it may be at facility level and customer level depending upon the data sufficiency.                                        

PD= The ratio of standard accounts at the beginning of the year to accounts out of those many standard accounts that defaulted at the end of year. This is computed rating category wise.

LGD=1- Recovery rate

Recovery Rate is calculated in the form of a ratio of recoveries made after the account turned NPA/ The balance outstanding at the time of account being declared NPA.

Economic Capital=Max loss @ some confidence level-Expected loss

The hurdle rate/threshold rate could be the Weighted Cost of Capital (WACC) or Return on Equity (ROE). The wide acceptance of ROE as the hurdle rate is because (a) WACC is generally lower than ROE as the debt cost is tax deductible and (b) ROE denotes shareholder maximization ie, whatever is left after meeting all expenses and costs is for the equity shareholders to savor. Any loan with a RAROC higher than the Hurdle rate is considered a value addition (Economic Value Added) and taken onto the loan book.

There is a polarization in the opinion on the use of a single hurdle rate versus multiple hurdle rates for different product lines like SME, Large Corporate, Agriculture, Retail etc. It is for a simple reason that a single hurdle rate would fail to discriminate on cost efficiency between product lines, thus over penalizing or under penalizing the product lines. However a single hurdle rate simplifies the common objective of the lending institution.

While RAROC has drawn lot of appreciation, it also has its fair share of limitations .For eg, let’s say we get an EVA of 5% and the prevailing risk free rate is also 5%., then it would be better if we invested in a government security rather than lending funds and assuming greater risk for the same return. The problem of overlooking the correlation between the bank’s portfolio and the market is another shortcoming. To counter this, Adjusted RAROC is in place.

Adjusted RAROC=RAROC-risk free rate/ Beta of Equity.

Also, while we have emphasized on the Var Based Capital at Risk in computation of Economic Capital, it is a proven fact that VaR is not a coherent risk measure as it fails the principal of sub-additivity for non-linear trends. (Sub Additivity implies it is possible, when combining portfolios, that the overall VaR of the new portfolio could be greater than the sum of the individual VaRs.)

While RAROC as a concept has commanded huge acceptance, it is to be kept in mind that in the times of business climate dynamism, sometimes the most efficient models might be a tad ineffective due to their assumptive nature and RAROC will be no exception. Hence its use in discretionary lending will be with certain disclaimers.

  • Siddharth Mahanty is faculty and senior manager at Sir Sorabji Pochkhanawala Bankers’ Training College of Bank of Baroda