Chapter 24: Assessment of operational risk Flashcards
Why operational risk should be assessed URE DEO MICRO
• Upside potential of risk is absent
• Reputational risk is significant
• Emphasis on quantitative models for all risk types
• Day-to-day losses minimized
• Extreme event losses minimized
• Objectives of business met with greater certainty
• Main driver behind financial disasters
• Interlinked with credit and market risk
• Consistent treatment of ops risk not done
• Regulatory capital requirement
• Overall ERM process and framework strengthened
Data used to assess operational risk SHES ER ICU
• Skewness to right (many small low impact events)
• Heavy tailed
• External data not applicable internally
• Statistical methods hard to apply
• Expertise to assess the data required
• Root cause analysis must be derived
• Infrequent incidence
• Cyclical
• Uncertain severity
Components to consider for modelling operational risk VELLIES
• Volumes of transactions used as proxy for incurred losses
• EVT approach to model tail losses
• Lines of business
• Loss events considered - everyday losses
• Internal data for repetitive high frequency losses
• External data on low frequency losses
• Stress scenarios and simulations with Monte Carlo
Merits of bottom-up approach BRAD
• Breakdown of losses into risk components may be difficult
• Robust picture of risk profile
• Application of external data may be difficult
• Data may lack
Benefits of scenario analysis CARBO
• Cause and effect relationship better understood
• Arbitrage opportunities reduced
• Reliance on data is small
• Black swan events identified
• Opinions of risk experts considered
Shortfalls of Top-down models SCARES I COM
• Specific risks not anticipated and mitigated
• Cause and effect scenarios not considered – source of risk not identified
• Aggregated assessment
• Relationship between risks ignored
• Extreme events not captured properly
• Soft measures of ops risk not captured
• Improvement in future ops risk management not considered
• Complexity of businesses differ
• Overseas trading
• Mitigation put in place not considered
Operational risk assessment process PIC TOM
• Policy creation
• ID and assess risks
• Capital allocation and performance management
• Transfer and finance
• Organizational structure
• Mitigation and control
Tools and techniques to assess, measure and manage operational risks SLIRP CRAME
• Self-assessment
• Loss incident data bases created
• Indicators of risk
• Risk mapping
• Performance triggers
• Controls put in place
• Root cause analysis done
• Analyse trends loss data
• Mitigation strategies formulated
• Prioritise risks
• EWTs
Basel assessment of operational risk: Advanced measurement approach (AMA) DECS
• Data sources considered (internal, external, stress scenarios)
• Events of loss and lines of business considered
• Confidence level: 1 year assessment at 99%
• Statistical and scenario analysis
Basic indicator approach FOG
• Fixed multiplier for whole business applied
• Only profitable years considered
• Gross income across business
Basel operational risk assessment: Standardised approach MLS
• Split gross income into different business lines
• Multiplier for each line
• Loss making years are considered
Different to down models to consider IVEA
• Implied capital model (balancing item)
Ops risk capital = total risk capital – non-ops risk capital
• Income volatility model
Ops IV = Total IV – non-ops IV
• Economic pricing model CAPM
Assess movement in share price to determine impact of publicized operational losses
• Analogue model
Model operational losses based on data from similar companies
Operational risk assessment: Economic pricing model (CAPM) APOS
o All market information is included in a company’s share price
o Published operational losses’ effect on a company’s share price assessed
o Overall market movement stripped out
o Softer elements and specific events of ops risk management can be allowed for
Compare data for operational and credit risk VERRAS HSR
• Volumes of credit risk data is higher
• Economic cycles have a greater impact on credit risk
• Reliability of external credit risk data is higher
• Relevance of external ops risk data is low
• Availability of credit risk data is higher
• Sources for ops risk data is less
• Heterogeneity of ops risk data is higher
• Subjectivity in ops risk data
• Random occurrence of ops risk events
