Internal Model Approach

Internal Model Approach
Market Risk
Corporation Bank
Structure
Introduction
Broad Principles for RBI Approval
Qualitative Criteria for IMA Approval
Quantitative Criteria
Risk Measurement
Back Testing
Stress Testing
2014 KPMG, an Indian Partnership and a member firm of the KPMG network
of independent member firms affiliated with KPMG International Cooperative
(KPMG International), a Swiss entity. All rights reserved.
Scope of IMA
Only Held for trading will come under IMA
AFS to continue under SMM, as markets are illiquid and market prices may not be
available
If other group level entities are not ready, IMA for parent, and those entities which
are ready, and SMM for others
Insignificant positions, minor currencies, negligible business areas, etc. can be
under SMM
Expected to follow IMA to all market risk positions and entities in the future
Trading Book risk covers
Interest rate-related instruments and equities in trading book
Exchange rate risk trading book
Risk relating to investments in MFs in trading book
Internal model and its purpose
What is an internal model?
What is the purpose of an internal model?
A risk management system developed by the Bank to analyze the overall

risk position, to quantify risks and to determine the economic capital
required to meet those risks
To fully integrate processes of risk and capital management within the

Bank
IMA and VAR?
Risk Management models in use far more advanced than rigid rules
Banks can use their own VaR models as basis for capital requirement for
Market Risk
VaR is a robust Risk Measurement and Management Practice
Market Risk Governance Structure

Board of Directors
Integrated Risk
Management Department
Internal Audit
Unit
Mid office/Market Risk

Unit
Risk Control, Capital

Computation and
Reporting team
Model Development
Unit
Model Validation Unit
IMA - Qualitative Criteria
Board and senior management should be actively involved
Documentation of Policies, Procedures and model parameters
Daily reports for Risk Unit to be reviewed by Senior Management in order to take appropriate
remedial action, if required
Banks risk measurement system must be well documented to describe the basic principles of
risk management system and to provide empirical techniques used to measure market risk.
Maintenance of Market Risk Model Dossier
Ongoing, updated Dossier - to keep a record of the details of the model and of the changes /
refinements,
Can be built by references and links to other policies, operating manuals
Independent Risk Control Unit responsible for design and implementation of Banks risk
management systems
Regular Back-Testing and stress testing
Initial and on-going Validation of Internal Model
Banks Internal Risk Measurement Model must be integrated into Management decisions
Should be used in conjunction with Trading and Exposure Limits.
Should be well documented
Independent review of risk measurement systems by internal audit
Model Development Unit

Either the bank can develop the model in house or use an external vendor. As the bank
is using an external vendor the bank should :
document and explain the role of the vendor model and the extent to which it is used within
the market risk measurement system of the bank;
demonstrate a thorough understanding of the vendor model;
ensure that the vendor model is appropriate for measuring the market risk of the bank,
given the nature of the portfolio and the capabilities of the staff; and
have clearly described strategies for regularly reviewing the performance of the vendor
model.
Kastle Risk Management Software Architecture
Internal Model Validation Unit
It comprise qualified members independent of model development. The role would

include
ensuring that the current systems setup is capable of supporting the models; all
changes made to the models being used, or to the modelling process, should be
validated and approved;
maintaining previous versions of the model being altered; and
Ensuring models are subjected to change-control procedures, so that computer
codes cannot be changed except by authorized staff.
External Model Validation Unit

This unit will comprise the suitably qualified external team who were not involved in
model development process. The role would involve validating:
Formulae used in calculation and pricing are validated by Risk Control Unit, independent of
traders
Checking that structure of model is adequate for banks activities and geographical
coverage
comparing back-testing results

Data flows are transparent and accessible to all auditors
10
Internal Audit
Purpose of Internal Audit Team is to ensure:
The adequacy of the documentation of risk management system and process;
The organization and functioning of the risk control unit;
The approval process for risk pricing models and valuation systems used by front
and back-office personnel;
The validation of any significant change in the risk measurement process;

The scope of market risks captured by the risk measurement model;
The integrity of management information system;
The accuracy and completeness of position data; verification of sources

KPMG has shared an internal audit checklist with the bank
11
Market Risk Dossier
The content of Market Risk Dossier would include:

Authors responsible for the contents, date updated
Description of the scope of application of the model

Risk exposures and levels
Policies and organization
Risk measurement system
Stress analysis and back testing program and results of the tests
Technological environment and information integrity controls
12
Market Risk File
Market Risk File content should include:

Scope of application of model
Description of Exposure
Estimation of Regulatory Capital
Policies and organization
Measurement Systems
Stress Analysis Program
Back testing program

Structure of Limits
13
IMA RBI filing documents

Bank need to provide written intention to RBI for migration to IMA. Documents which
need to be submitted are:
Preliminary and detailed application to RBI for prior approval
Internal audit report of the model
A MR File
MR Model Dossier
14
RBI assessment
RBI performs an assessment based on the following parameters:
Accuracy of documentation
Model scope
Qualitative review
Technological environment and information integrity
Quantitative review
Model monitoring (any modification proposed need prior intimation and approval from RBI)
15
IMA - Quantitative criteria
IMA capital a function of
Normal VAR
Stressed VAR (for positions subject to interest rate specific-risk
IMA to be modeled as given below

Normal VAR General market risk
Stressed VAR general market risk
Specific Risk Charge as per the Standardized Measurement
Method
16
Value at Risk (VaR) - Basics
The value at risk (VAR) of a portfolio is the loss in value in the portfolio that can be expected
over a given period of time (e.g., 1-Day) with a probability not exceeding a given number (e.g.,
5%)
Probability (Portfolio Loss VAR) = K
K = Given Probability
A VaR statistic thus has three components:

Time Period
Confidence level
Loss Amount
a time period (a day, a month or a year) and

a relatively high level of confidence (typically either 95% or 99%),
an estimate of investment loss (expressed either in rupees or percentage terms)
17
Parameters for Normal VAR Computation
Computed daily
VaR computation be based on following inputs :
Horizon of 10 Trading days Can use daily VAR and scale to 10 days
99% confidence level
Observation period at least 1 year historical data
Update data sets at least quarterly
Can use any type of model - variance-covariance matrices, historical simulations, or
Monte Carlo simulations.
Portfolio VaR for the bank has been provided in the attached PDF
18
VaR The Historical Simulation Approach

The Approach
The historical method simply re-organizes actual

historical returns, putting them in order from
worst to best
The Approach
The VaR is taken as the worst 1% (at 99%

confidence level) of all daily returns
The simulation trials assume that the percentage

changes in all market variables are as on the
previous day i.e. the history will repeat
VaR is arrived at by tabulation of results at the
desired confidence level
5%
1.645 Std Dev
-10MM
Possible Profit/Loss
Strengths
Easy to describe and justify

Fat tails, asymmetric responses, outliers are all incorporated automatically
The assumptions are realistic as the scenarios occurred in the past are considered
Easily accommodates stress simulations
Captures all of the markets previous hits
19
Example Historical VaR
20
Parameters for Stressed VAR computation
Intended to replicate a VaR calculation that would be generated on the banks

current portfolio
The stressed-VaR should be calculated at least weekly
The model inputs for the stressed VaR should be calibrated to historical data
from a continuous 12-month period of significant financial stress relevant to the
banks portfolio
Example 2007-2008 period of crisis
The said period of stress will be approved by the RBI as part of its approvalfor
the IMA model submitted by the bank and would be regularly reviewed.
21
IMA Market Risk Charge
C = max {VaRt-1; (mc +pc)*VaRavg} + max {sVaRt-1; (ms+ps)*sVaRavg }
The capital requirement C is calculated according to the following

formula:
mc and ms - multiplication factors to be set by the RBI on the
basis of their assessment of the quality of the banks risk
management system, subject to absolute minimum of three for
both the factors; and
pc and ps - the plus / add on factor, generally ranging from zero
to one, to be decided by the bank based on the results of
the
back testing of its VaR model.
22
Capital computation
Total Market Risk Capital is sum of :
HFT general Market Risk Capital(IMA capital)
HFT specific Capital( standardized approach)
AFS general market Risk (standardized approach)
AFS specific capital (standardized approach)
23
Market Risk Capital example
24
Multiplication and Plus factor

Normal VAR
Stressed VAR
Mc Factor Range
3 -4
Mc How used
Used to multiply 60 day Average

normal VAR
Used to multiply 60 day Average

stressed VAR
McWho determines this

factor
RBI
RBI
Plus Factor )Pc) Range
0-1
0-1
Pc How used.
Added to Mc and used for

multiplying 60 days Average
normal VAR
Added to Mc and used for

multiplying 60 days Average
stressed VAR
Pc Who determines
Banks based on back testing

results of NORMAL VAR MODEL
and confidence in the model.
Banks based on back testing

results of NORMAL VAR MODEL
and confidence in the model.
3-4
25
Stress Testing - Scenarios Analysis
Evaluating the portfolios under various states of the world

Running simulations of the current shocks e.g. 1987 crash, etc
Scenarios requiring simulations
Bank specific scenario driven by the current position of the bank than
historical simulation
Can help to identify undetected weakness in the bank's portfolio
Much more subjective than VAR portfolio subject to large historical scenario
26
Back Testing
Statistical testing that consist of checking whether actual trading losses are in
line with the VAR forecasts
The Basel back testing framework consists in recording daily exception of the 99% VAR
over the last year
Even though capital requirements are based on 10 days VAR, back testing uses a daily
interval, which entails more observations
A bank should also report to the RBI the results of their back-testing exercise every
quarter before the last day of the month. In addition to exceptions, report to include:
Classification of exceptions and proposed investigations.

Action already taken or proposed to be taken for model improvement.
Number of exceptions observed during each of the last three back-testing results.
Too many exceptions indicate that
The banks systems are not simply capturing the risk of the positions themselves.
Model volatilities and/or correlations were calculated incorrectly
Models accuracy could be improved
Bad Luck or market moved in unanticipated manner
Loss due to intra-day trading
27
Back Testing
28

Internal Model Approach

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Internal Model Approach

Uploaded by

Copyright:

Available Formats

Internal Model Approach

Exchange rate risk trading book

Risk relating to investments in MFs in trading book

Internal model and its purpose

What is an internal model?

What is the purpose of an internal model?

A risk management system developed by the Bank to analyze the overall

To fully integrate processes of risk and capital management within the

IMA and VAR?

Market Risk Governance Structure

Mid office/Market Risk

Risk Control, Capital

Model Validation Unit

IMA - Qualitative Criteria

Board and senior management should be actively involved

Documentation of Policies, Procedures and model parameters

Maintenance of Market Risk Model Dossier

Regular Back-Testing and stress testing

Initial and on-going Validation of Internal Model

Should be used in conjunction with Trading and Exposure Limits.

Should be well documented

Independent review of risk measurement systems by internal audit

Model Development Unit

Kastle Risk Management Software Architecture

Internal Model Validation Unit

It comprise qualified members independent of model development. The role would

External Model Validation Unit

comparing back-testing results

The validation of any significant change in the risk measurement process;

The accuracy and completeness of position data; verification of sources

Market Risk Dossier

The content of Market Risk Dossier would include:

Description of the scope of application of the model

Technological environment and information integrity controls

Market Risk File

Market Risk File content should include:

Back testing program

IMA RBI filing documents

IMA - Quantitative criteria

IMA capital a function of

IMA to be modeled as given below

Value at Risk (VaR) - Basics

A VaR statistic thus has three components:

a time period (a day, a month or a year) and

Parameters for Normal VAR Computation

VaR computation be based on following inputs :

VaR The Historical Simulation Approach

The historical method simply re-organizes actual

The VaR is taken as the worst 1% (at 99%

The simulation trials assume that the percentage

1.645 Std Dev

Easy to describe and justify

Example Historical VaR

Parameters for Stressed VAR computation

Intended to replicate a VaR calculation that would be generated on the banks

The stressed-VaR should be calculated at least weekly

Example 2007-2008 period of crisis

IMA Market Risk Charge

C = max {VaRt-1; (mc +pc)*VaRavg} + max {sVaRt-1; (ms+ps)*sVaRavg }

The capital requirement C is calculated according to the following

HFT general Market Risk Capital(IMA capital)

HFT specific Capital( standardized approach)

C = max {VaRt-1; (mc +pc)VaRavg} + max {sVaRt-1; (ms+ps)sVaRavg }