Quantification of Operational Risk

1
Quantification of Operational Risk

• Eric Rosengren
• Senior Vice President
• Federal Reserve Bank of Boston
• FRB Chicago Bank Structure Conference
• May 9, 2002

2
Operational Risk Basels Definition

• The risk of loss resulting from inadequate or
  failed internal processes, people and systems or
  from external events
• This definition includes legal risk
• Strategic and reputational risk are excluded

3
Operational Risk Can Be Significant

• Damage to Physical Assets / Business disruption
  and system failures
• World Trade Center attack Publicly announced
  insurance exposure to date - 27 billion
• Internal fraud
• Allied Irish Banks Rogue trading - 690 million
• Clients, Products Business Practices
• Prudential Insurance Sales practices - 2
  billion settlement of class action lawsuit
• Providian Financial Unfair sales / collection
  practices - 400 million settlement restitution
• Execution, Delivery Process Management
• Lehman Brothers Erroneous UK equity trade 425
  million estimated loss

4
The Basel Proposal

• 1) Basic Indicator Approach (the alpha)
• Capital agross income
• 2) Standardized Approach (the betas)
• Capital bgross income, by business line
• Strength
• Simplicity
• Limitations
• Blunt charge, not risk-sensitive
• One size fits all
• Risk does not increase linearly with gross income
• Fails to capture effect of banks management of
  operational risk
• No incorporation of qualitative factors
• No incentive for banks to invest in op risk
  infrastructure

5
The Basel Proposal

• 3) Advanced Measurement Approach (AMA)
• Capital firm specific calculation,
  statistically based methodology
• Intended to overcome the lack of risk sensitivity
  in the simpler approaches by setting regulatory
  capital based on the banks internal risk
  measurement models
• These models use the banks own metrics to
  measure operational risk, internal loss data,
  external loss experience, scenario analysis, and
  risk mitigation techniques to set capital
  commensurate with the operational risk posed by
  the banks activities

6
AMA - Key Characteristics

• Risk Sensitive - capital reflects operational
  risks for size and scope of banks activities
• Low risk activities require less capital for
  operational risk
• Banks with low risk mix of business or less
  activity need less capital for operational risk
• Banks with better control environments require
  less capital for operational risk
• Banks with well developed risk mitigation hold
  less capital for operational risk
• Flexible
• Banks choose supportable methodologies reflective
  of their business
• Banks use a combination of internal data,
  external data, and scenario analysis to determine
  appropriate capital
• Capital allocation can be integrated into
  scorecards, risk indicators, warning systems and
  audit scores used to measure and monitor
  operational risk

7
AMA - Key Characteristics

• Rewards investment in better control
  environments
• Actions that reduce loss experience also reduce
  capital
• Actions that reduce the likelihood or severity of
  extreme events can reduce capital
• Actions that mitigate risk can reduce capital
• Results in Appropriate Capital
• AMA is not a capital tax
• Capital allocation changes with risk profile of
  organization
• Capital allocation changes as industry improves
  the measuring, monitoring, and mitigation of
  operational risk

8
AMA - Implementation

• Internal statistical model is basis for
  calculating OpRisk exposure and capital charge
• Three Components of an AMA
• Operational Loss Data
• Internal / external
• Quantification Methodologies
• Loss Distribution Approach (LDA)
• Qualitative Risk Assessments and Key Risk
  Indicators
• Scorecards
• Early warning systems

9
Quantification Methodologies the LDA

• An operational loss distribution takes the form

10
AMA - Challenges

• Greater complexity / resource commitment than
  exposure indicator approaches
• Numerous modeling issues / decisions need to be
  made by bank
• Incorporation of external data
• Appropriate distributional assumptions
• Incorporation of risk mitigation
• Scenario Analysis
• Qualitative assessments require improved rigor
• Identification of risk indicators that highly
  correlate with op losses
• Combining quantitative techniques and qualitative
  factors into a comprehensive methodology

11
AMA - Benefits

• Banks investing in AMA methodology are already
  seeing benefits
• Reduce both expected losses and volatility of
  earnings
• Measuring losses allows identification of causal
  factors for operating losses
• AMA provides framework for addressing extreme
  outcomes
• AMA allows comparison of investment in controls,
  investment in technology, investment in
  insurance, or self insuring with capital
• AMA reduces distortions in decision making and
  performance evaluation from omitting capital for
  operational risk

12
Conclusion

• Exposure indicator approaches are relatively easy
  to implement but lack risk focus and proper
  incentive structure
• AMA requires additional effort but reinforces
  banks existing risk management objectives /
  practices and results in a more accurate
  allocation of capital
• Risk-sensitive
• Flexible
• Rewards investment in controls / reducing op risk
• Well integrated with banks existing risk
  management processes
• Not a capital tax