Day One

Operational risk assessment: effectiveness over elegance

Tuesday 22 September 2009

8:30 Registration and coffee

9:00 UTILISING LOSS-RELATED AND CAUSE-RELATED INDICATORS TO ENHANCE PREDICTIVE CAPACITY

  • Adapting performance indicators to become risk indicators and using senior staff to ratify targets
  • Aggregating risk indicators across business lines
  • Establishing a common risk language through summarising findings on a heat map
  • To what extent do loss-related indicators lack a forward looking ability and risk sensitivity?
  • Back-testing operational risk indicators against actual loss experience

Tutor

Christopher Lotz, Head of Quality Risk Management, BAFIN

10:30 Morning break

11:00 SCENARIO ANALYSIS: TRANSFORMING EXPERT OPNION INTO MITIGATION ACTIONS

  • Combining top down and bottom up approaches to scenario analysis
  • Defining sources of information for assessment of risk scenarios
  • Balancing qualitative and quantitative information
  • Creating an IT environment supporting the easy assessment of risk scenarios
  • Linking scenario analysis with actual loss data and risk indicators

Tutor

Günther Helbok, Head of Operational Risk & Risk Integration, BANK AUSTRIA, member of UNICREDIT GROUP

12:30 Lunch

13:30 INTEGRATION DATA SOURCES

    • Integrating internal losses, external data, expert opinion and key risk indicators
    • How does the value of internal data increase as its volume rises and variation between it decreases?
    • Improving a firms loss experience credibility as more losses are observed over a longer time period
    • Using credibility theory to ensure capital charge is largely determined by internal loss experience

    Tutor

    Peter Hoflijk, Director of Operational Risk, FORTIS BANK

    15.00 Afternoon break

    15.30 MEASURING ECONOMIC CAPITAL FOR OPERATIONAL RISK

    • Estimating loss distribution by modelling frequency and severity losses
      • Modelling frequency of losses using discrete distributions
      • Modelling severity of losses using continuous distributions
    • Calibrating frequency and severity distributions to ensure good quality model input
    • Accounting for parameter uncertainty: Impact on the capital requirement
    • Using a goodness-of-fit test to determine the appropriate of a distribution

    Tutor

    Joerg Fritscher, Risk Analytics & Instruments, DEUTSCHE BANK

    17.00 End of day one