Agenda

Agenda

Credit Risk Modelling - Agenda

14:0015:00

Regulatory risk implications of credit risk models

08:30 - 09:30

  • Introduction to credit risk models 
  • Objective, purpose, and utility of credit risk models 
  • Credit risk models and capital adequacy requirements- A-IRB approach (Basel II) 
  • Credit risk models and expected credit loss measurements- ECL provisions (IFRS9) 
  • Applicable models and risk-based pricing for low default portfolios 
Maria Kostova

Lead quantitative specialist

CRISIL Limited

Maria Kostova has over 16 years of experience in regulatory risk with specific focus on expected credit loss methodologies and internal ratings based approach valuation techniques.

Her significant focus of expertise resides in model development, validation, risk governance, capital allocation mechanisms and financial stability frameworks across Pillar 2A and 2B types of risk- credit, market, liquidity, concentration, IRRBB and model risk management. During her career, she has been predominantly engaged in top tier banking groups in the United Kingdom, the Netherlands and Spain, allowing her to acquire a comprehensive IFRS9 and A-IRB credit risk modelling expertise in PiT/TTC PD, LGD, EAD development and excellent understanding of internal capital adequacy requirements (ICAAP/ ILAAP) alongside risk weighted asset adjustments, Liquidity Restructurings and Stress-testing techniques.

She is specialized in banking and finance with strong emphasis on statistics and microeconomics in the United States, where this substantial exposure on logistic and linear regression methodologies in statistics for retail and corporate portfolios has been developed and practiced in accordance with Basel/ BIS/ ECB/EBA regulatory frameworks and the PRA requirements in the UK.

 

15:0015:15

Break

09:30 - 09:45

15:1516:15

Model Governance

09:45 - 10:45

  • Regulatory expectations of model risk
  • Model validation techniques (Basel III, IFRS9, CECL) 
  • Current model validation challenges
  • Monitoring and reporting requirements 
  • Counterparty credit vs credit risk
  • Forward looking trends
Ushnish Banerjee

Vice President, QAG- EMEA Model Risk

Morgan Stanley

14:0015:00

Best practices for credit risk modelling

08:30 - 09:30

  • Effect of strengthening Basel regulations on credit risk modelling 
  • Example of rating systems based on default data and portfolios 
  • Modelling for low default portfolios 
  • Modelling macroeconomic drivers for IFRS9: probability weighted adjustments 
  • Estimating rating migration matrices 
  • Connecting credit risk portfolio and LGD modelling 

15:0015:15

Break

09:30 - 09:45

15:1516:15

Stress testing credit risk portfolios

09:45 - 10:45

  • General principles of forecasting models and model design

    • Types of scenarios

    • Scenario severity

    • Scenario stress selection

  • Regulatory expectations

  • Utilising past data 

  • Credit risk losses in a downturn economic cycle

  • Use in income stress testing

14:0015:00

Developments for credit risk modelling

08:30 - 09:30

  • Adjusting for the new normal 
  • Realistic and worst-case scenarios 
  • Evolution of modelling approaches aligned with default frequencies 
  • How to interpret the changes of frameworks 
  • Policy updates in credit risk models and capital adequacy release 

15:0015:15

Break

09:30 - 09:45

15:1516:15

ESG and impacts of climate on credit risk

09:45 - 10:45

  • Emerging climate risk policies and their impact on credit risk 
  • Climate risk in your credit risk assessment  
  • ESG factor considerations 
    • Transition risk 
    • Building resilience  
  • Credit risk model amendments and their Impact on minimization of Model Risk  
  • Socio- economic implications and shifts towards a greener economy 

14:0015:00

AI/ML in credit risk modelling

08:30 - 09:30

  • Forecasting techniques in AI/ML methodologies
  • Regulatory requirements and model risk
  • Challenges
    • Modelling with imbalanced classes
    • Model compound error
    • Identifying biases
    • Implications of model transparency and explainability
  • Overcoming challenges  

15:0015:15

Break

09:30 - 09:45

15:1516:15

Future of AI/ML in credit risk modelling

09:45 - 10:45

  • Business model impact of COVID 19 – Procyclicality of models 
  • Algorithm’s optimisations in credit risk – decision tree neural networks  
  • Applying NLP in traditional model approaches  
  • Identifying problems in financial information 
  • Importance of transactional data
    • Identify future happens
    • Future characteristics