Model Risk Management - Best Practices

2 days
Prague, NH Hotel Prague
  • Introduction to Model Risk
  • Types and Sources of Model Risks
  • Model Risk Management Framework
  • Practical Approaches to Managing Model Risks
  • Model Risk Inherent in Stress Testing Models
  • Regulatory Issues and Requirements
This two day training seminar covers the key aspects relating to model risk management as practiced by leading financial institutions. The course addresses common issues of model design inadequacy, parameter flaws, interpolating and bootstrapping errors as well as reliance on back testing on out of sample data sets from a practical viewpoint. Issues of regulatory compliance and governance, policies and controls get addressed from a functional perspective with an aim to achieve efficiency with respect to dealing with model risks in banking.

The first day commences by addressing some critical aspects of internal credit rating models while identifying potential areas of malfunction that can lead to disastrous consequences. It follows by examining typical shortcomings of pricing models as well as market risk models. A special case discussion revolves around the traditional paradigm of "single pricing formula" for capital market securities and derivatives, whereby stress situations have shown that pricing can be highly state dependent. The course depicts best practices used by leading institutions aimed at augmenting such deficiencies.

During the second day, we will discuss model risks associated with approximations within capital models and specifically credit portfolio models. Another topic covered represents potential traps inherent within the complex CVA counterparty models that rely on potential future exposure estimations errors which can be conducive to false hedging decisions when applied by CVA desks.

Lastly, we will examine model risk inherent in modern stress testing models and we will propose some mitigating measures as seen in leading international banks.

09.00 - 09.15 Welcome and Introduction

09.15 - 12.00

  • Introduction of Model Risks and illustration of severe losses resulted from inadequate use of models in banks
    • Malfunctions when used for Hedging Purposes
      • Local vs Stochastic volatility Models
      • The instability and non-stationarity of Market Correlations
      • Traps in Dimensionality Reduction Methods
  • Internal Rating Models and challenges:
    • Calibration to Rating Agencies Models (PDs)
    • Validation Issues
    • Reliance on Spreads and other Market Data
    • Reliance on historical back testing
  • Pricing Models and Risk Sources
    • Mapping to Risk Factors
    • Wrong/& Corrupt Datasets on Key Parameters: rates, volatilities, ETF values, etc.
    • Interpolation and Bootstrapping Errors

12.00 - 13.00 Lunch

13.00 - 16.30

  • Market Risk Models
    • Errors resulting from dimensionality reduction
      • Cholesky Decomposition
      • Eigenvalue-eigenvector Analyses
    • Market Liquidity Concerns
      • LVaR; Liquidity at Risk; Stress VaR, Expected Shortfall (CVaR) & other mitigating measures
    • Sensitivity Analyses
      • Incremental, component and Marginal VaR
      • Augmenting Means Best Practices

Day Two

09.00 - 09.15 Brief recap

09.15 - 12.00

  • Credit Portfolio Models and Economic Capital Models
    • Precision vs Utility Discussion
    • Marginal Economic Decisions Examples
  • Capital Allocation Models Reliance on Correlations
  • Governance, Policies and Controls Best Practices
  • Discussion

12.00 - 13.00 Lunch

13.00 - 16.30

  • Model Risk in Counterparty Risk Models
  • Numerical Simplifications and Dimensionality Reduction Techniques
    • American Monte Carlo Methods
    • Reduced Form Models
  • Model Risk in Stress Testing Models

Summary, Evaluation and Termination of the Seminar

back to calendar
detailed program
Printer Friendly Format
E-mail Newsletter
Contact Us
  Site Map    Policy Statements    Mobile Version