Agenda

Agenda

Credit Risk Management and Modelling - Frankfurt

Course agenda

Day One: Wednesday, November 13, 2019

08:30

Registration and refreshments

09:00

Credit risk management post IFRS 9

  • Update on IFRS 9 implementation and supervisory actions 
  • Challenges and remediation activity
  • Classification and measurement of financial instruments
  • Comparison of methodologies and outputs across Europe, the UK and US
  • Challenges arising from IFRS 9;
  • Data management challenges
  • Back testing and P&L attribution  - overcoming data alignment issues
  • The question of volatility in P&L figures
  • The importance of precision and quality

Speaker: Andreas Koutras, Independent Consultant 

10:30

Morning break

11:00

Basel IV, IRB and TRIM 

  • Revisions to the credit risk standardised approach under Basel IV
    • Motivations for a change for the credit risk standardised approach
    • Introduction of the SME factor and change in the asset class segmentation
    • Real estate lending, general treatment and credit splitting
  • Future of IRB and the introductions of the new modeling requirements and components
    • Changes of the IRB scope within the asset classes
    • Exposure level parameter floors
    • Modifications in the credit risk mitigation techniques
    • Output floor and its impact on the RWA optimisation
    • The new partial use requirements and prospective reversal options for the standardised approach
    • The new default definition requirements and its impact on the risk parameter model development
    • Rating philosophy and how to account for the effects
    • Downturn LGD and the MoC framework related differences
  • TRIM and its immediate consequences; model review standards and data qality management frameworks

Speaker: Kaan H. Aksel, director, regulatory management - quant team, PwC

12:30

Lunch

1:30

Machine learning and AI for credit risk modelling

  • Applications for machine learning
  • How machine learning can help banks rate credit risk
  • Using AI and machine learning to build credit risk models
  • Collaborative portfolios
  • Capturing changes in behaviour to incorporate in models

Speaker: Stefano Bonini, executive, Accenture

3:00

Afternoon break

3:30

Best practice credit risk modelling

  • Example rating systems based on default data
  • Modelling for low-default portfolios
  • Modelling macroeconomic drivers for IFRS 9
  • Estimating rating migration matrices
  • Connecting credit risk portfolio and LGD modelling

Speaker: Jörg Lemm, privatdozent, Institute of Theoretical Physics at WWU Münster and Credit risk manager, DZ HYP (part time job in credit risk after retirement)

5:00

End of day one

Day Two: Thursday, November 14, 2019

08:30

Refreshments

09:00

Model validation and assurance

  • Model validation framework
  • Building and maintaining a framework to validate credit portfolio models
  • Macro-economic variable choices
  • Vendor models
  • Future considerations

Speaker: Radka Margitova, senior manager, financial services risk consulting, credit risk, PwC

10:30

Morning break

11:00

Application of explainable AI in a regulated market

  • Digitalization and machine learning (ML)
  • Capabilities and risk of ML methods for banks
  • Regulatory requirements vs. explainability of ML
  • Technical challenges of ML infrastructures for banks
  • Practical example in Python of an explainable AI model

Speakers: Marcel Jäger, manager credit risk and lead of ML lab, Deloitte & Dominik Zabel, expert in ML and development of cloud microservices services, Deloitte

 

12:30

Lunch

1:30

Stress testing credit risk portfolios

  • Basics of credit risk stress testing
  • Stress testing at the macro portfolio level
  • Challenges of forecasting
  • Potential differences under IFRS 9
  • Pros and cons of the new and old approaches

Speaker: Dr Zhilin Yao, head of risk controlling, Agricultural Bank of China

3:00

Afternoon break

3:30

ECB observations on IFRS 9 implementation

  • Banks’ methodologies to assess “significant increase in credit risk”
  • Incorporation of forward-looking information in ECL models

Speaker: Luca Ciavoleillo, Principal supervisor, supervisory policy division, European Central Bank

4:30

End of course