8/9/2019 Bm Reading Raroc
1/23
Pramod [email protected]
Page 1 of 23
RISK CAPITAL REQUIREMENT & RAROC FRAMEWORK
FOR CORPORATE BANKING RELATIONSHIPS
Confidential for Private circulation only
8/9/2019 Bm Reading Raroc
2/23
Pramod [email protected]
Page 2 of 23
CONTENTS
1. INTRODUCTION ......................... .......................... ........................... .......................... ........3
2. ECONOMIC CAPITAL .......................... .......................... ........................... ......................... 4
3. RAROC .......................... .......................... .......................... ........................... .....................5
4. INPUTS TO RISK MEASURE.............................................................................................6
i. Internal Risk Rating ......................................................................................................6 ii . Rating Mig ration and Probabil ity of Default (PD) ........................................................7 iii. Estimated Exposure at Default (EAD)........................................................................11
iv . Estimated Loss Given Default (LGD) .........................................................................12 v. Spread ........................... ........................... .......................... .......................... ...............13 vi . Fees ........................... .......................... ........................... .......................... ................... 13 vii. Operat ing Cost........................................................................................................13
5. EXPECTED LOSS (RISK PREMIUM)...............................................................................14
6. ECONOMIC CAPITAL MEASUREMENT..........................................................................15
7. APPLICATION OF THE RAROC MODEL: .......................... ........................... .................. 17
ANNEXURE : COMPARISON OF RATING MIGRATIONS.............................. .......................... 19
ANNEXURE : REGULATORY VERSUS IRB CAPITAL ............................. .......................... .....21
ANNEXURE : RAROC VERSUS RETURN ON REGULATORY CAPITAL ............................ ....22
8/9/2019 Bm Reading Raroc
3/23
Pramod [email protected]
Page 3 of 23
1. INTRODUCTION
In a risk-return setting, borrowers placed in a high credit risk category should be priced higher
than those in a low risk category. Thus, risk rating and therefore, expected probability of default is
a key determinant of pricing. As in the case of pricing, risk rating has a bearing also on economic
capital allocation in preference to merely regulatory capital allocation for the underlying asset. A
standard risk adjusted measure of return is a tool for both pricing and making better lending
decisions. The Risk-Adjusted Return on Capital (RAROC) framework, which is already put in
place worldwide by large sized banks, is one such measure. The lender begins by charging an
interest mark-up to cover the expected loss – expected default rate of the rating category of the
borrower. The lender then allocates enough capital to the prospective loan to cover the
unexpected loss – variability of default rates. The primary focus of RAROC is to provide an
apples-to-apples comparison of capital usage and return across business lines and risk types.
Bankers Trust in the 1970s first introduced the RAROC concept. Their original interest was to
measure the risk of the bank’s portfolio, as well as the amount of equity capital necessary to limit
the exposure of the bank’s depositors and other debt holders to a specified probability of loss.
The recent surge among banks to adopt the RAROC approach is attributed to two forces: (i) the
demand by shareholders for improved performance, especially the maximization of shareholder
value and (ii) growth of conglomerates around profit centres. This has forced banks to develop a
measurement of performance, especially when the capital of the bank is both costly and limited.
RAROC is an integrated approach for measuring risk. It is a decision support tool, which enables
the Bank to:
Calculate how much capital is needed to support all risks taken by the enterprise.
Understand where the shareholders’ capital is invested.
Compare risk-adjusted returns earned on that capital across dissimilar business lines and
activities.
Identify opportunities for risk transfer.
The purpose of this document is to lay down the framework and mechanism for computation of
economic capital and Risk-adjusted pricing using internal ratings and internal default and loss
data. The computation of economic capital in this manner will lead us further on the road to
adoption of the Internal Ratings Based (IRB) approach for arriving at capital requirement for credit
risk under the proposed New Basel Accord, when permitted by the regulator. The Risk-adjusted
pricing approach will serve the dual purpose of ensuring that returns generated from a credit
8/9/2019 Bm Reading Raroc
4/23
Pramod [email protected]
Page 4 of 23
exposure is truly commensurate with the extent of risk assumed (capital at stake) and assessing
the profit performance of individual business units relative to the assets booked by them,
effectively a risk-based performance measurement tool.
2. ECONOMIC CAPITAL
Economic Capital is the amount of capital that an activity or business requires to support the
economic risk it faces to a specified solvency standard or default probability. It protects the lender
against unexpected volatility in the economic earnings of the firm. It is calculated from the
aggregated risk distribution at the target solvency standard. Effectively, Economic Capital is a
measure of the unexpected loss or economic capital at risk (VaR) as a result of the lending.
Generally, international banks allocate enough capital so that the expected loan loss reserve or
provision plus allocated capital covers over 99% of the loan loss outcomes.
Historically, two approaches have emerged to measure economic capital at risk. The first
approach, following Bankers Trust, develops a market-based measure - maximum adverse
change in the market value of a loan over the next year, as given in the block below:
Capital At RiskMcCauley Loan Duration x Loan Exposure x Expected Discounted
Change in Credit Risk Premium
The second, following Bank of America among others, develops an experiential or historicallybased measure, as follows:
Economic Capital3.4 to 6.0 times Standard Deviation of PD (x Loss Given
Default x Exposure
.
8/9/2019 Bm Reading Raroc
5/23
Pramod [email protected]
Page 5 of 23
3. RAROC
RAROC is the ratio of risk-adjusted income over a period, past or present, to the economic capital
at risk for a business unit as a result of an activity (lending). Thus,
RAROC = Adjusted Income
Economic Capital
Spread It is the direct income earned on the loan, i.e. Loan Rate – Bank’s
Cost of Funds.
Fees Non-interest income directly attributable to the loan over the relevant
period (such as processing charges, commitment fees, LC / BG
charges, fee from ancillary business, etc.).
Expected Loss Probability of Default (PD) x Loss Given Default (LGD)
x Exposure At
Default (EAD)
PD is the default probability over a given horizon for a particular
rating class
LGD is the estimated loss value of the loan or 1 minus recovery
value of a loan on default.
EAD is the estimated exposure at the time of default as a
percentage of commitment
Operating Costs It is the cost of executive time, effort, and resources in originating
and monitoring a loan.
RAROC adjusts the profitability of each lending activity based on the cost of the capital that it
requires. The RAROC calculated for a loan should be compared with the Bank’s expected Return
on Equity (ROE). The ROE will be the Hurdle Rate. Where RAROC exceeds this Hurdle Rate, the
loan will be considered as adding value to Bank’s capital and therefore, eligible for allocation of
capital.
Adjusted Income = Spread + Fees – Expected Loss – Operating Costs
8/9/2019 Bm Reading Raroc
6/23
Pramod [email protected]
Page 6 of 23
4. INPUTS TO RISK MEASURE
For computation of Risk Capital and RAROC, the following will be the fundamental inputs:
i. Internal Risk Rating
ii. Rating Migration and Probability of Default (PD)
iii. Estimated Exposure at Default (EAD)
iv. Estimated Loss Given Default (LGD)
v. Spread
vi. Fees
vii. Operating Cost
We propose to compute Risk Capital and RAROC on a corporate relationship basis, using the
following mechanism for arriving at each of the above inputs:
i. Internal Risk Rating
The Bank has in March 2001, sourced, customized and adopted a corporate credit risk-rating
model from CRISIL, christened by us as the XY-risk Rating System. The model delivers a quick
assessment of a borrower’s credit quality, structured analysis and grading of risk parameters and
key parameters for tracking and controlling asset quality. The model in its basic form is designedto assess credit risk in a structured and comprehensive manner. XY-risk has 10 ratings, not
including the modifiers, to match international best practices. The 10 ratings are as follows (the
first 5 being “pass” grades and the last 5 “problem loan” grades):
Rating Meaning Score Band
XY-AAA Very Strong credit quality 5.50 6.00
XY-AA+ Strong credit quality 4.75 5.50
XY-AA Good credit quality 4.25 4.75
XY-A Above average credit quality 3.75 4.25XY-BBB Average credit quality 3.25 3.75
XY-BB+ Moderate credit quality 2.75 3.25
XY-BB Weak credit quality 2.25 2.75
XY-B Near Default credit 1.75 2.25
XY-CC Default credit 1.50 1.75
8/9/2019 Bm Reading Raroc
7/23
Pramod [email protected]
Page 7 of 23
XY-C Loss credit 1.00 1.50
The above rating is based on an assessment of the borrower only without considering transaction
characteristics and in that respect is uni-dimensional. By default, assets classified as ‘Sub-
Standard’ will be accorded a rating of XY-CC and ‘Doubtful’ assets accorded a rating of XY-C. All
exposures to the same corporate are assigned the same borrower rating, irrespective of any
differences in the nature of each individual transaction.
To introduce the second dimension – transaction characteristics – the Bank to begin with permits
rating of the corporate guarantor to be superimposed on the borrower rating. Further, 100% cash
collateralized exposures (for the individual transaction) are accorded a rating of XY-AAA, while
bank guaranteed exposures are accorded a rating equivalent to the rating of the bank. As and
when sufficient reference data points are generated, the Bank will work towards adopting
transaction characteristic Loss Given Default (LGD) estimates as the second dimension for the
rating system. The transaction characteristics will be security, product type, tenor, etc.
The structure of the rating model, the parameters and risk factors assessed and the process to be
followed for rating a corporate and reviewing the rating is documented separately (copy
appended).
ii. Rating Migration and Probabili ty of Default (PD)
Using the internal risk-rating model, the bank will arrive at a credit rating for each exposure in itsportfolio, every year. The model will be consistently used across the portfolio and over the years.
The bank will capture ratings year-wise for each corporate in a database and track the upward
and downward migration in rating for each exposure over the years. To compute the transition
rate, each corporate’s rating at the end of a year (or the end of a horizon) is compared with its
rating at the beginning of the year (or the beginning of the horizon). From these individual
transitions, the bank will build a ‘ten by ten’ transition matrix (frequency distribution) for a given
horizon (e.g. 1 year, 2 years, etc.) on a portfolio level. As an illustration, the transition matrix for a
given horizon will look as follows (transition year 1998-1999):
AAA AA+ AA A BBB BB+ BB B CC C Total
AAA 1 - - - - - - - - - 1
AA+ - 6 0 - - - - - - - 6
AA - 1 23 10 - - - - - - 34
A - - 6 69 16 1 - - - - 92
BBB - - - 13 66 17 1 - - - 97
BB+ - - - 1 10 18 10 - 1 - 40
BB - - - - - 2 8 1 2 - 13
B - - - - - - - 1 - - 1
CC - - - - - - - - 0 1 1
C - - - - - - - - - 1 1
1 7 29 93 92 38 19 2 3 2 286
8/9/2019 Bm Reading Raroc
8/23
Pramod [email protected]
Page 8 of 23
As per the above matrix, of 34 exposures (100%), which were rated AA in Year 1998, 23 (68%)
retained their rating in Year 1999 (1 year horizon) while 10 (29%) slipped to A rating and 1 (3%)
moved up to AA+ rating. Similarly of 40 exposures (100%) rated BB+ in Year 1998, 18 (45%)
remained at BB+ in Year 1999, whereas 10 (25%) upgraded to BBB and 1 (3%) to A while 10
(25%) slipped to BB and 1 (3%) defaulted (CC & C).
The bank will build transition matrixes as above for every year (where the horizon is one year).
From these matrixes, a mean (probability) transition matrix and a standard deviation matrix will be
generated. The mean transition matrix will throw up the probability of an exposure with a given
rating migrating upward or downward, including to a default stage (called Probability o f Default
– PD - in the latter case) over a given horizon. For example, assuming the above matrix to be a
mean transition matrix, the Probability of Default (PD) of a corporate rated BB over a one-year
period is 15% (slippage to CC and C), while for a AA rated corporate rate the PD is estimated as
nil. A default will be considered to have occurred with regard to a particular borrower when one or
more of the following events has taken place (in accordance with BIS guidelines):
i. It is determined that the borrower is unlikely to pay its debt obligations (principal, interest,fees) in full.
ii. A credit loss event associated with an obligation of the borrower, such as a charge-off,
specific provision or distressed restructuring involving the forgiveness or postponement of
principal, interest or fees
iii. The borrower is past due more than 90 days on any credit obligation
iv. The borrower has filed for bankruptcy or similar protection from creditors
% AAA AA+ AA A BBB BB+ BB B CC C Total
AAA 100% 0% 0% 0% 0% 0% 0% 0% 0% 0% 100%
AA+ 0% 100% 0% 0% 0% 0% 0% 0% 0% 0% 100%
AA 0% 3% 68% 29% 0% 0% 0% 0% 0% 0% 100%
A 0% 0% 7% 75% 17% 1% 0% 0% 0% 0% 100%
BBB 0% 0% 0% 13% 68% 18% 1% 0% 0% 0% 100%
BB+ 0% 0% 0% 3% 25% 45% 25% 0% 3% 0% 100%
BB 0% 0% 0% 0% 0% 15% 62% 8% 15% 0% 100%
B 0% 0% 0% 0% 0% 0% 0% 100% 0% 0% 100%
CC 0% 0% 0% 0% 0% 0% 0% 0% 0% 100% 100%
C 0% 0% 0% 0% 0% 0% 0% 0% 0% 100% 100%
8/9/2019 Bm Reading Raroc
9/23
8/9/2019 Bm Reading Raroc
10/23
Pramod [email protected]
Page 10 of 23
The last column (PD) gives the Probability of Default over a one-year horizon for each rating. The
PDs will dynamically be revised as further data flows into the transition matrix. A comparison of
CRISIL’s and ICRA’s average one-year transition rates versus our transition rates is given in the
Annexure to this document. For the purpose of the RAROC framework, to begin with, the
following PDs for each performing grade will be used (percentage converted into decimal terms,
i.e. 0.50% stated as 0.0050 and an upward differential correction given to PDs of AA+ and AA
ratings):
Rating PD
AAA 0.0000
AA+ 0.0005
AA 0.0010
A 0.0040
BBB 0.0144
BB+ 0.0634
BB 0.1536
B 0.2500
In order that the ratings migration matrix and the probability of default is upto-date, it is extremely
essential that branches are prompt in annually re-rating their exposures using the XY-risk Rating
model immediately on receipt of the previous year end financials (without waiting for the
periodical review / renewal of the account). Branches should obtain the financials as soon as
these are ready with the company. Delays in populating the ratings afresh could result in the
following:
The Bank may over-price a good credit and end up losing the business to competition
A risky credit could get under-priced and the Bank’s earnings would not be
commensurate with the risk assumed
The risk measure of the individual exposure or portfolio could get under or over-stated
due to which the Bank’s capital could get sub-optimally utilized.
8/9/2019 Bm Reading Raroc
11/23
Pramod [email protected]
Page 11 of 23
For the Bank to benefit from the Internal Ratings Based approach for capital measurement, it is
all the more necessary that the ratings are reviewed promptly every year and the ratings
migration maintained real-time.
Rating initiators should be balanced in their assessment of the risk factors underlying a credit.
Any attempt to play down a risk unreasonably pull down a rating will impact the stability of the
rating system on one hand and on the other distort the migration statistics and the expected
distribution and trend of the portfolio. To illustrate, assume that a i-BBB rated corporate is
“pushed” up to a rating of i-A. In the unfortunate event that this borrower defaults during the year,
this transition from ‘A’ to default grade will resultantly increase the default probability of ‘A’ rating
category. In the process, all assets rated ‘A’ might have to thereafter bear a higher price (risk
premium) and capital requirement compared to existing, thanks to the mis-classification. Level 2
and Level 3 confirmers in the rating system are, therefore, expected to play the balancing role to
ensure that inconsistencies do not occur.
ii i. Estimated Exposure at Default (EAD)
EAD is the estimated exposure at the time of default as a percentage of commitment. The bank
will need to capture data on every exposure, which defaulted in the history of the bank (ideally) or
minimum over one economic cycle. The types of data that the bank will capture for each
exposure are:
a. Nature and size of committed exposures on and off balance sheet at the time of default
b. Actual exposure at the time of default
c. Activity / industry
Principally, (a) and (b) above, for all defaulted credits, will be used for computing the estimated
Exposure At Default (EAD) in percentage terms.
We have captured data on every exposure, which has defaulted (to NPA category) in the history
of the bank. The data captured for each exposure are:
a. Name of Account in default
b. Nature of committed exposures on and off balance sheet at the time of default
c. Quantum of committed exposures on and off balance sheet at the time of default
d. Actual exposure at the time of default
e. Activity / industry
8/9/2019 Bm Reading Raroc
12/23
Pramod [email protected]
Page 12 of 23
From (c) and (d) above for all defaulted credits, the estimated Exposure At Default (EAD) was
computed in percentage terms. As per data as on 31st March 2002, the EAD estimate works out
to 76%. This estimated EAD will be periodically recomputed in the event of fresh defaults.
iv. Estimated Loss Given Default (LGD)
LGD is the estimated loss value of the loan or 1 minus recovery value of a loan on default. For
each of the defaulted exposures in (iii) above, the Bank will additionally capture the following
information:
a. Date of default
b. Amount and date of each recovery made in the credit after default
c. Expenses incurred on the credit after default for legal proceedings, maintenance of security,
recoveries, etc.
d. Nature, seniority, share and value of security
The bank will discount the cash flows and work out the Present Value of the net recovery in the
credit as at the time of default. The net recovery as a percentage of the actual exposure at the
time of default will give the recovery rate (%). 100 minus the recovery rate (%) will yield the Loss
Given Default (LGD), expressed in percentage terms. This will be computed for each exposure
to arrive at an overall estimated LGD.
Besides an overall LGD, it would be also advisable that the bank captures LGD specific to each
industry. This of course provided there is adequate representation of defaults in each industry.
The objective of such fragmentation is that each industry depending on the stage of its economic
cycle, contribution to GDP, phase of technology, etc., will have a different recovery potential out
of the operations of the company or the assets of the defaulting corporate in such an industry. To
illustrate, under current economic conditions, the prospects of recovery in a credit related to a
dyes & pigments industry would be very low compared to recovery in a credit related to a FMCG
industry. LGDs could also be captured by seniority of charge on security and by type of exposure.
For each of the defaulted exposures (NPAs), we have captured the following information:
a. Date of default
b. Amount and date of each recovery made in the credit after default
8/9/2019 Bm Reading Raroc
13/23
Pramod [email protected]
Page 13 of 23
c. Expenses incurred on the credit after default for legal proceedings, maintenance of security,
recoveries, etc.
The LGD was computed for each exposure to arrive at an overall estimated LGD, which as on
31st March 2002 worked out to 88%. This estimated LGD would be periodically computed to
factor in fresh recoveries, costs, defaults, if any.
v. Spread
The difference between the stipulated interest rate on the credit and the cost of funding it is the
spread. Cost of funding used for the RAROC framework will be the prevailing FTP for the
particular exposure horizon.
vi. Fees
On a relationship basis, will include processing charges, non-interest income from non-fund
based credit facilities, forex income, fee income from ancillary services, etc, as per rates
stipulated.
vii. Operating Cost
This will be computed on a look-back basis or budgets for the current year. The annual operating
expenses as a percentage of aggregate of deposits, advances, investments and off-balance
sheet exposure, will be multiplied by the total exposure to arrive at the overheads for the related
account. For corporate banking relationships, the operating cost per Rs 100 of business
(advances + contingent + deposits) has been computed at Rs 0.23 (i.e. 0.23%) using the
following data:
Budget as on
Average Advances Rs 3538 crore
Average Contingent Rs 1669 crore Average Deposits Rs 3750 crore
Operating Expenses Rs 20.79 crore
8/9/2019 Bm Reading Raroc
14/23
Pramod [email protected]
Page 14 of 23
5. EXPECTED LOSS (Risk Premium)
The Expected Loss (EL) corresponding to each rating will be the product of the Probability of
Default (PD) for that rating, the estimated Exposure at Default (EAD) and the estimated Loss
Given Default (LGD). Based on the information for each of these variables given earlier in this
paper, the rating-wise EL is as follows:
Rating PD Expected Loss (Risk Premium)
AAA 0.0000 0.00%
AA+ 0.0005 0.03%
AA 0.0010 0.07%
A 0.0040 0.27%BBB 0.0144 0.96%
BB+ 0.0634 4.24%
BB 0.1536 10.27%
B 0.2500 16.72%
The Expected Loss should be recovered from pricing as a risk premium. Alternatively,
suitable specific provisioning should cover this. The Bank could consider carving out the risk
premium from each exposure’s pricing and building up the provision cover.
8/9/2019 Bm Reading Raroc
15/23
Pramod [email protected]
Page 15 of 23
6. ECONOMIC CAPITAL MEASUREMENT
To arrive at the Economic Capital, the risk related capital requirement formula as proposed in the
Basel Capital Accord II under Internal Ratings Based approach, will be used. However, as this
factor includes Expected Loss component, the Expected Loss as worked out above separately for
each rating will be deducted from the capital requirement computed using the IRB formula to give
the Economic Capital for the RAROC framework. The formula in the Accord is as follows (the
confidence level used in 99.90%)
Capital Requirement (K) = LGD x M x N[(1-R)^-0.5xG(PD)+(R/(1-R))^0.5xG(0.999)]
LGD = Loss Given Default
M = Maturity Factor = 1+0.047x((1-PD)/PD^0.44)
PD = Probability of Default
N ( ) = Standard Normal Cumulative Distribution Function, with mean 0 and standard
deviation 1.
G ( ) = Inverse Standard Normal Cumulative Distribution Function, with mean 0 and
standard deviation 1.
R = Correlation = 0.10 x (1-EXP(-50xPD))/(1-EXP(-50))+0.20x[1-(1-EXP(-50xPD))/(1-
EXP(-50))]
EXP ( ) = Natural Exponential Function
Using the above formula, the rating wise Capital Requirement and the Economic Capital after
netting off Expected Loss is as follows:
Rating Expected Loss IRB Capital Req. Economic Capital(IRB Cap – Exp Loss)
AAA 0.00% 0.56% 1.40% @
AA+ 0.03% 2.52% 2.49%
AA 0.07% 3.61% 3.54%
A 0.27% 7.15% 6.88%
BBB 0.96% 12.22% 11.26%
BB+ 4.24% 22.12% 17.88%
BB 10.27% 35.13% 24.86%
B 16.72% 44.52% 27.80%
@ Minimum requirement
8/9/2019 Bm Reading Raroc
16/23
Pramod [email protected]
Page 16 of 23
A comparison of capital requirement as per the above capital measures versus current regulatory
requirement of 9% for the corporate banking portfolio as on 31st March 2002 is given in the
Annexure.
8/9/2019 Bm Reading Raroc
17/23
Pramod [email protected]
Page 17 of 23
7. APPLICATION OF THE RAROC MODEL:
Using the above inputs, RAROC will be computed as follows for each relationship:
Spread + Fees – Operating Costs – Risk Premium (Expected Loss)RAROC = --------------------------------------------------------------------------------
Economic Capital
The targeted RAROC will be 25% (hurdle rate). To achieve this target, and considering the
Economic Capital, Risk Premium, Operating Cost worked out above, Corporate Banking will need
to leverage a minimum Spread plus Fees, rating-wise as follows, from the relationship:
Rating Spread + Fees
(as percentage of exposure) AAA 0.58%
AA+ 0.93%
AA 1.23%
A 2.23%
BBB 3.98%
BB+ 8.98%
BB 16.73%
B 23.98%
(Exposure for the above purpose, will be: Fund Based : 100% of limit sanctioned, Non-fund
Based : 50% of limit sanctioned till March 2003, 100% thereafter and Forex Limit : 100% of limit
sanctioned)
To illustrate, as per the RAROC framework devised above, the rating-wise pricing structure will
be as follows, considering 29th July 2002 FTPs, for fund based facilities:
FTP 6.58% 6.65% 6.75% 6.70% 6.78% 6.84% 7.57% 7.90% 8.66% 8.82%
RATING 7-14 days 15-30 31-45 46-90 91-180 181-270 271-3651 yr 1 day-
1.5 1.5 yr 1 day-22yrs 1 day- 3
yrs
AAA 7.16% 7.23% 7.34% 7.28% 7.36% 7.42% 8.15% 8.48% 9.25% 9.40%
AA+ 7.51% 7.58% 7.69% 7.63% 7.71% 7.77% 8.50% 8.83% 9.60% 9.75%
AA 7.81% 7.88% 7.99% 7.93% 8.01% 8.07% 8.80% 9.13% 9.90% 10.05%
A 8.81% 8.88% 8.99% 8.93% 9.01% 9.07% 9.80% 10.13% 10.90% 11.05%
BBB 10.56% 10.63% 10.74% 10.68% 10.76% 10.82% 11.55% 11.88% 12.65% 12.80%
BB+ 15.56% 15.63% 15.74% 15.68% 15.76% 15.82% 16.55% 16.88% 17.65% 17.80%
BB 23.31% 23.38% 23.49% 23.43% 23.51% 23.57% 24.30% 24.63% 25.40% 25.55%
B 30.56% 30.63% 30.74% 30.68% 30.76% 30.82% 31.55% 31.88% 32.65% 32.80%
8/9/2019 Bm Reading Raroc
18/23
Pramod [email protected]
Page 18 of 23
Returns based on the above pricing on Current Regulatory Capital are given in the Annexure
(RAROC versus Return on Regulatory Capital). Based on the comparison, it is suggested that till
a Risk Based Capital framework (such as the Internal Ratings Based Approach) is accepted by
the Regulator, pricing based on Scenario III in the annexure is accepted, i.e. pricing where both
RAROC and Risk Adjusted Return on Regulatory Capital is 25% minimum.
Risk premiums are normally based on Default Probabilities over a 1-year horizon. Therefore, in
the case of exposures with a tenor of over one year, the Bank needs to stipulate a pricing-reset
clause linked to either the internal rating of the borrower or certain Events of Default (EODs),
which have a direct bearing on the rating, so that at the end of each year the pricing can be
revised to factor in the changed risk premium corresponding to the fresh rating of the corporate.
Credit proposals will henceforth use this model for fixing risk related pricing and determining
account profitability both actual and projected, to understand the value of the relationship. Value
of collateral, market forces, perceived value of accounts, future business potential, portfolio /
industry exposure and strategic reasons may play an important role in pricing. However, any
attempt at price-cutting for market share would result in mispricing of risk, adverse selection, and
sub-optimal capital utilization.
Ideally, the RAROC model should be applied across the portfolio of the loan assets and
aggregated. It should also be aggregated for the various business activities of the Bank, to arrive
at the enterprise-wide risk-adjusted return on economic capital. To begin with the same shall be
applied to individual corporate credit relationships. Branches can then work out the RAROC for
the branch portfolio by extending the framework to the entire portfolio on an aggregate, to
benchmark it against the targeted RAROC – in the process it will ensure that higher earnings in
another asset duly compensate any shortfall in earnings in an otherwise preferred low yield asset,so that targeted RAROC is achieved at branch level. The same principle should be followed at
the Region Level and at the Country Level.
*******
8/9/2019 Bm Reading Raroc
19/23
Pramod [email protected]
Page 19 of 23
Annexure : COMPARISON OF RATING MIGRATIONS
Bank’s Average One-Year Transi tion Rates (1997-2001)
Crisi l Average One-Year Transit ion Rates (1993 – 2000)
Rating DataPoints
AAA AA A BBB BB B C & Below
AAA 154 96.8% 3.2%
AA 500 2.8% 85.2% 10.0% 1.0% 0.4% 0.4% 0.2%
A 759 3.3% 82.3% 8.8% 3.2% 0.3% 2.1%
BBB 317 0.3% 5.7% 73.2% 11.0% 1.9% 7.9%
BB 117 2.6% 58.1% 2.6% 36.7%
B 16 62.5% 37.5%
C & Below 32 100%
ICRA’s Average One-Year Transit ion Rates (1992 – 2001)
Rating Data Points AAA AA A BBB NI
AAA 142 92.3% 7.0% 0.7%
AA 287 1.7% 86.4% 10.1 0.3% 1.4%
A 279 1.4% 83.5% 7.5% 7.5%
BBB 133 0.8% 85.7% 13.5%
NI 2.6% 58.1%
NI – Non-Investment (BB+ to D)
Remarks:
The long-run stability of our rating model is yet to be proved given its short history.
% AAA AA+ AA A BBB BB+ BB B CC C PD
AAA 100.00% 0.00%
AA+ 58.33% 30.21% 11.46% 0.00%
AA 7.90% 59.60% 31.10% 1.39% 0.00%
A 0.26% 14.05% 66.33% 18.29% 0.67% 0.40% 0.40%
BBB 0.26% 30.34% 55.30% 11.44% 1.22% 1.44% 1.44%
BB+ 4.68% 26.16% 44.52% 18.29% 6.34% 6.34%
BB 14.37% 55.85% 14.42% 15.36% 15.36%
B 75.00% 25.00% 25.00%
CC 61.06% 38.94% 100.00% C 100.00% 100.00%
8/9/2019 Bm Reading Raroc
20/23
Pramod [email protected]
Page 20 of 23
CRISIL’s and ICRA’s transition matrix cover almost a decade, which could be considered
as one economic cycle. This rating model is yet to go through a full economic cycle.
CRISIL’s and ICRA’s reports on the transition acknowledge that deteriorating general
economic conditions have resulted in greater volatility / lesser stability in the ratings in the
recent years – this is reflected in our matrix which covers the recent years only.
Our ratings transition rates may not be truly comparable with that of CRISIL and ICRA
due to possible differences in the underlying characteristics of the statistical population
(number of data points, geographical spread of exposures, size of corporate entities, etc).
8/9/2019 Bm Reading Raroc
21/23
Pramod [email protected]
Page 21 of 23
Annexure : REGULATORY VERSUS IRB CAPITAL
Corporate Banking Exposure (Performing Assets) as on 31st March
Remarks
Aggregate Risk Capital requirement is lower than Regulatory Capital requirement as on 31st
March
Aggregate Risk Capital requirement relative to Regulatory Capital requirement will change
with change in the composition of the portfolio.
Rs in Crore
Rating Risk Capital FB O/S NFB O/S
Regulatory
Capital (9%) Risk Capital
i-AAA 1.40% 11.38 86.06 4.90 0.76
i-AA+ 2.49% 112.95 15.10 10.84 3.00
i-AA 3.54% 530.12 278.12 60.23 23.70
i-A 6.89% 1360.98 544.17 146.98 112.48
i-BBB 11.25% 400.32 224.48 46.13 57.68
i-BB+ 17.88% 65.19 39.01 7.62 15.14
i-BB 24.86% 11.18 6.13 1.28 3.54
i-B 27.80% 0.00 0.00 0.00 0.00
Total 2492.12 1193.44 277.98 216.31
8/9/2019 Bm Reading Raroc
22/23
Pramod [email protected]
Page 22 of 23
Annexure : RAROC VERSUS RETURN ON REGULATORY CAPITAL
Portfolio Return on Regulatory Capital will change depending on the portfolio mix. Since the
pricing is higher at the higher risk end, the Returns will expectedly increase as Risk
increases. The Returns steeply increase in view of the risk premium required to be earned
from the exposure, reflective of the propensity to default.
As the Risk Premium (Expected Loss) is netted from Returns, the Risk Adjusted Return on
Regulatory Capital increases gradually as risk increases.
Related to targeted RAROC of 25%, the Risk Adjusted Return on Regulatory Capital
obviously is much lower than the RAROC figure in the lower risk categories and then
surpasses the RAROC figure as one moves up the risk band.
The pricing scenarios for RAROC target 25% (scenario I), Risk Adjusted Return on
Regulatory Capital (Scenario II) target 25% and Minimum of Scenario I and Scenario II
(Scenario III) are given below rating-wise. This is followed by the tenure-wise pricing structure
for each Scenario. The Bank is required to maintain Capital as per current regulatory
requirement till permission is acceded by the Regulator to move into a Risk-based Capital
(Internal Ratings Based Approach, etc) framework. In order that the Bank generates the
Rs in Crore
Rating Risk Capital FB O/S NFB O/S
Regulatory
Capital (9%) Risk Capital
Pricing
Proposed
Return on
Regulatory
Capital
Risk Adjusted
Return on
Regulatory
Capital RAROC
i-AAA 1.40% 11.38 86.06 4.90 0.76 0.58% 6% 4% 25%
i-AA+ 2.49% 112.95 15.10 10.84 3.00 0.93% 10% 7% 27%
i-AA 3.54% 530.12 278.12 60.23 23.70 1.23% 14% 10% 26%
i-A 6.89% 1360.98 544.17 146.98 112.48 2.23% 25% 19% 25%
i-BBB 11.25% 400.32 224.48 46.13 57.68 3.98% 44% 31% 25%
i-BB+ 17.88% 65.19 39.01 7.62 15.14 8.98% 100% 50% 25%
i-BB 24.86% 11.18 6.13 1.28 3.54 16.73% 186% 69% 25%
i-B 27.80% 0.00 0.00 0.00 0.00 23.98%
Total 2492.12 1193.44 277.98 216.31 28% 20% 25%
8/9/2019 Bm Reading Raroc
23/23
Pramod Vaidyapramodvaidya@gmail com
targeted minimum Return of 25% on Capital (satisfying both Current Regulatory Capital and
Risk Capital), Scenario III pricing structure is recommended:
Rating Risk Premium
Note : (1) Pricing = Spread + Fees from the exposure (2) Overheads = 0.23%
i-AAA 0.00%
i-AA+ 0.03%
i-AA 0.07%
i-A 0.27%
i-BBB 0.96%
i-BB+ 4.24%
i-BB 10.27%
i-B 16.72%
3.98%
8.98%
16.73%
23.98%
0.58%
0.93%
1.23%
2.23%
Scenario IIScenario I
2.75%
2.77%
2.80%
3.00%
3.70%
7.00%
13.05%
23.98%
16.73%
Scenario III
19.50%
Mark up for RAR on
Regulatory Cap 25%
Mark up for RAR onRegulatory Capital AND
RAROC 25%
2.75%
2.77%
2.80%
3.00%
3.98%
8.98%
Mark-up for RAROC 25%
ExposureTenure 7-14 days 15-30 31-45 46-90 91-180 181-270 271-365 1 yr 1 day-1.5 1.5 yr 1 day-2
2yrs 1 day- 3yrs
FTP 6.58% 6.65% 6.75% 6.70% 6.78% 6.84% 7.57% 7.90% 8.66% 8.82%
i-AAA 7.16% 7.23% 7.34% 7.28% 7.36% 7.42% 8.15% 8.48% 9.25% 9.40%
i-AA+ 7.51% 7.58% 7.69% 7.63% 7.71% 7.77% 8.50% 8.83% 9.60% 9.75%
i-AA 7 .81% 7.88% 7.99% 7.93% 8.01% 8.07% 8.80% 9.13% 9.90% 10.05%
i-A 8.81% 8.88% 8.99% 8.93% 9.01% 9.07% 9.80% 10.13% 10.90% 11.05%
i-BBB 1 0.56% 10.63% 10.74% 10.68% 10.76% 10.82% 11.55% 11.88% 12.65% 12.80%
i-BB+ 15.56% 15.63% 15.74% 15.68% 15.76% 15.82% 16.55% 16.88% 17.65% 17.80%
i-BB 23.31% 23.38% 23.49% 23.43% 23.51% 23.57% 24.30% 24.63% 25.40% 25.55%
i-B 30.56% 30.63% 30.74% 30.68% 30.76% 30.82% 31.55% 31.88% 32.65% 32.80%
Scenario I Effect - Overall pricing on Fund Based exposure
ExposureTenure 7-14 days 15-30 31-45 46-90 91-180 181-270 271-365 1 yr 1 day-1.5 1.5 yr 1 day-2
2yrs 1 day- 3yrs
FTP 6.58% 6.65% 6.75% 6.70% 6.78% 6.84% 7.57% 7.90% 8.66% 8.82%
i-AAA 9.33% 9.40% 9.50% 9.45% 9.53% 9.59% 10.32% 10.65% 11.41% 11.57%
i-AA+ 9.35% 9.42% 9.52% 9.47% 9.55% 9.61% 10.34% 10.67% 11.43% 11.59%
i-AA 9.38% 9.45% 9.55% 9.50% 9.58% 9.64% 10.37% 10.70% 11.46% 11.62%
i-A 9.58% 9.65% 9.75% 9.70% 9.78% 9.84% 10.57% 10.90% 11.66% 11.82%
i-BBB 10.28% 10.35% 10.45% 10.40% 10.48% 10.54% 11.27% 11.60% 12.36% 12.52%
i-BB+ 13.58% 13.65% 13.75% 13.70% 13.78% 13.84% 14.57% 14.90% 15.66% 15.82%
i-BB 19.63% 19.70% 19.80% 19.75% 19.83% 19.89% 20.62% 20.95% 21.71% 21.87%
i-B 26.08% 26.15% 26.25% 26.20% 26.28% 26.34% 27.07% 27.40% 28.16% 28.32%
Scenario II Effect - Overall pricing on Fund Based exposure
Exposure
Tenure 7-14 days 15-30 31-45 46-90 91-180 181-270 271-365 1 yr 1 day-1.5 1.5 yr 1 day-2
2yrs 1 day- 3
yrs
FTP 6.58% 6.65% 6.75% 6.70% 6.78% 6.84% 7.57% 7.90% 8.66% 8.82%
i-AAA 9.33% 9.40% 9.50% 9.45% 9.53% 9.59% 10.32% 10.65% 11.41% 11.57%
i-AA+ 9.35% 9.42% 9.52% 9.47% 9.55% 9.61% 10.34% 10.67% 11.43% 11.59%
i-AA 9.38% 9.45% 9.55% 9.50% 9.58% 9.64% 10.37% 10.70% 11.46% 11.62%
i-A 9.58% 9.65% 9.75% 9.70% 9.78% 9.84% 10.57% 10.90% 11.66% 11.82%
i-BBB 10.56% 10.63% 10.74% 10.68% 10.76% 10.82% 11.55% 11.88% 12.65% 12.80%
i-BB+ 15.56% 15.63% 15.74% 15.68% 15.76% 15.82% 16.55% 16.88% 17.65% 17.80%
i-BB 23.31% 23.38% 23.49% 23.43% 23.51% 23.57% 24.30% 24.63% 25.40% 25.55%
Scenario III Effect - Overall pricing on Fund Based exposure