NextGeneration Credit Risk Modeling : From IRB and IFRS 9 to Quantum ML
COURSE OBJECTIVE
Advanced Credit risk modeling course using Artificial Intelligence and Quantum Computing, among many other topics: credit scoring tools, modeling of PD, LGD and EAD parameters of the advanced IRB approach of Basel III, credit risk methodologies for IFRS 9 impairment models, stress testing models of credit risk, economic capital and climate changerelated credit risk.
Among other topics, quantum computing, quantum circuits, important quantum algorithms, quantum mechanics, quantum error and correction, and quantum machine learning are exposed.
Machine and deep learning are used to build powerful credit scoring and behavior scoring tools, as well as to estimate and calibrate risk parameters and stress testing.
A module on advanced data processing is exposed, explaining among other topics: sampling, exploratory analysis, outlier detection, advanced segmentation techniques, feature engineering and classification algorithms.
The course explains the recent final reforms of Basel III regarding the new standard approach and Advanced IRB, IFRS 9 related to credit risk and the new guidelines on estimation of PD and LGD and treatment of exposures in default of EBA.
Predictive machine learning models are shown such as: decision trees, neural networks, Bayesian networks, Support Vector Machine, ensemble model, etc. And in terms of neural networks, feed forward, recurrent RNN, convoluted CNN and adversarial Generative architectures are exposed. In addition, Probabilistic Machine Learning models such as Gaussian processes and Bayesian neural networks have been included.
Advanced methodologies are taught to estimate and calibrate risk parameters: PD, LGD and EAD. The Lifetime PD estimate used in the IFRS 9 impairment models is exposed.
Methodologies and practical exercises of Stress testing in credit risk using advanced techniques of machine learning and deep learning are shown. And a practical exercise with financial statements to understand the impact of stress testing on capital and profits.
We have a global exercise to estimate the expected loss at 12 months and ECL lifetime using advanced credit risk methodologies, including PD, LGD, EAD, prepayment and interest rate curve models.
The course teaches about economic capital methodologies that are used in credit card, mortgage, SME, and corporate portfolios. It also covers capital allocation methodologies.
Quantum Machine Learning is a fusion of quantum algorithms with Machine Learning programs. Machine learning algorithms are used to process huge amounts of data, whereas quantum machine learning employs qubits and specialized quantum systems to enhance the speed of computation and data storage in an algorithm. For instance, some mathematical and numerical techniques from quantum physics can be used in classical deep learning. A quantum neural network has the potential to reduce the number of steps, qubits used, and computation time.
The aim of this course is to demonstrate how quantum computing and tensor networks can be utilized to enhance the accuracy of machine learning algorithms. We will showcase how quantum algorithms can expedite Monte Carlo simulations, which are the most potent tools for constructing credit risk models. This provides a significant advantage for calculating economic capital, lifetime PD, and creating stress testing scenarios.
The course also seeks to compare classical models with quantum models, and make clear the scope, benefits, and opportunities of quantum computing in this field.
WHO SHOULD ATTEND?
The course is designed for financial professionals who want to enhance their skills in developing effective credit scoring models and finetuning their results. It is also suitable for credit risk and data science department managers responsible for managing these models. To better comprehend the course topics, participants are required to have a solid foundation in mathematics and statistics.
Schedules:

Europe: MonFri, CEST 1619 h

America: MonFri, CDT 1821 h

Asia: MonFri, IST 1821 h
Price: 7 900 €
Early Bird Price : 6 900 €
Ending 7 May
Level: Advanced
Duration: 39 h
Material:

Presentations PDF

Exercises in Excel, R, Python, Jupyterlab y Tensorflow
1. CuttingEdge Techniques Integration: The course offers a unique blend of traditional credit risk modeling methodologies with cuttingedge technologies such as artificial intelligence and quantum computing. By incorporating advanced techniques like quantum machine learning and deep learning, participants gain a comprehensive understanding of modern risk assessment methodologies, equipping them with the skills needed to tackle complex credit risk scenarios in the era of digital transformation.
2. Enhanced Accuracy and Efficiency: Through the utilization of quantum algorithms and tensor networks, participants learn how to enhance the accuracy and efficiency of credit risk models. Quantum computing enables expedited Monte Carlo simulations, a crucial component in constructing robust credit risk models, thereby providing a significant advantage in calculating economic capital, lifetime PD, and developing stress testing scenarios. This enhanced computational capability empowers financial institutions to make more informed decisions and better manage their credit risk exposure.
3. Preparation for Future Challenges: By exploring topics such as climate changerelated credit risk and the latest reforms in Basel III and IFRS 9, participants are equipped with the knowledge and skills needed to navigate evolving regulatory landscapes and emerging risk factors. Additionally, exposure to quantum computing and its potential applications in credit risk modeling prepares participants for the future of finance, ensuring they remain at the forefront of innovation in the field. This forwardlooking approach enables financial professionals to anticipate and address future challenges effectively, thereby enhancing the resilience and sustainability of their organizations.
AGENDA
NextGeneration Credit Risk Modeling :
From IRB and IFRS 9
to Quantum ML
PD, LGD and EAD IRB
Module 1: Guidelines on PD estimation, LGD estimation and the treatment of defaulted exposures

Reduction of parameter variability

Homogenization of the calculation of PD and LGD

Implementation dates in European banks

Data quality

Representativeness of data for model development and for risk parameter calibration

Human judgement in estimation of risk parameters

Deficiency treatment and margin of conservatism (Moc)

PD estimation

Model development

Data requirements

Risk drivers and rating criteria

Processing of external ratings

Rating philosophy

Treatment of Pools


PD calibration

Data requirements

Calculation of oneyear default rate

Calculation and use of observed average default rate

Longrun average default rate

Longrun average default rate calibration


LGD estimation

Methodologies for PD estimation

Data requirements

Recoveries from collaterals

Model development

Risk drivers

Collateral eligibility

Collateral inclusion


LGD Calibration

Definition of economic loss and realized loss

Treatment of commissions, interest and other withdrawals after default

Discount rate

Direct and indirect costs

Longrun average LGD

Calibration of the estimates to longrun average LGD


Estimation of risk parameters for defaulted exposure
 Estimation and calibration of the Expected Loss Best Estimate

Estimation and calibration of LGD indefault

Application of risk parameters

Review of estimates

Accompanying documents

Impact assessment

Problem identification

Policy objectives

Baseline Scenario

Options considered

Costbenefit analysis
IFRS 9
Module 2: Implementation of IFRS 9 in the EU

SICR assessment approaches

Approaches for determining stage transfers

Alignment between the Definition of Default and IFRS 9 exposures in Stage 3

Low credit risk exemption

12month PD as proxy for lifetime PD


Expected Credit Loss Models

Types of Expected Credit Loss models

Model limitations and use of overlays

Effects from the Russian/Ukrainian conflict

ESG including climate risks


IFRS 9 PD variability and robustness

Variability in the IFRS 9 PD

Differences in the use of IRB models for IFRS 9 estimates

Differences in the definition of default

Differences in risk differentiation

Differences in risk quantification

Treatment of 20202021 defaults


Incorporation of forwardlooking information

Macroeconomic scenarios

Variability of the methodological approach for incorporation of FLI and
reflection of nonlinearity
Incorporation of FLI at parameter level

List of macroeconomic variables used for FLI incorporation

Forecasting period and reversion to longterm average


Variability in the impact and different sensitivities from FLI

Effect of nonlinearity and probability framework


Focus on backtesting practices

Staging allocation

ECL measurement

IFRS 9 LGD estimates

IFRS 9 PD estimates

Overlays

Forwardlooking information

CREDIT SCORING
Module 3: Exploratory Analysis

Exploratory Data Analysis EDA

Data sources

Data review

Target definition

Time horizon of the target variable

Sampling

Random Sampling

Stratified Sampling

Rebalanced Sampling


Exploratory Analysis:

Histograms

Q Q Plot

Moment analysis

boxplot


Treatment of Missing values

Multivariate Imputation Model


Advanced Outlier detection and treatment techniques

Univariate technique: winsorized and trimming

Multivariate Technique: Mahalanobis Distance


Over and Undersampling Techniques

Random oversampling

Synthetic minority oversampling technique (SMOTE)

Module 4: Feature engineering

Feature engineering

Data Standardization

Variable categorization

Equal Interval Binning

Equal Frequency Binning

ChiSquare Test


Binary coding

WOE Coding

WOE Definition

Univariate Analysis with Target variable

Variable Selection

Treatment of Continuous Variables

Treatment of Categorical Variables

Information Value

Optimization of continuous variables

Optimization of categorical variables


Exercise 1: EDA Exploratory Analysis

Exercise 2: Feauture Engineering

Exercise 3: Detection and treatment of Advanced Outliers

Exercise 4: Multivariate model of imputation of missing values

Exercise 5: Univariate analysis in percentiles in R

Exercise 6: Continuous variable optimal univariate analysis in Excel
Machine Learning
Module 5: Machine Learning

Unsupervised models

K Means

Principal Component Analysis (PCA)

Advanced PCA Visualization

Supervised Models

Ensemble Learning

Bagging trees

Random Forest

Boosting

Adaboost

Gradient Boosting Trees

Exercise 7: Segmentation data using KMeans

Exercise 8: Credit Scoring Support Vector Machine

Exercise 9: Credit Scoring Boosting

Exercise 10: Credit Scoring Bagging

Exercise 11: Credit Scoring Random Forest, R and Python

Exercise 12: Credit Scoring Gradient Boosting Trees
DEEP LEARNING
Module 6: Deep Learning

Feed Forward Neural Networks

Single Layer Perceptron

Multiple Layer Perceptron

Neural network architectures

Activation function

Back propagation

Directional derivatives

Gradients

Jacobians

Chain rule

Optimization and local and global minima


Deep Learning Convolutional Neural Networks CNN

CNN for pictures

Design and architectures

Convolution operation

Filters

Strider

Padding

Subsampling

Pooling

Fully connected

Credit Scoring using CNN

Recent CNN studies applied to credit risk and scoring


Deep Learning Recurrent Neural Networks RNN

Long Term Short Term Memory (LSTM)

Hopfield

Bidirectional associative memory

descending gradient

Global optimization methods

RNN and LSTM for credit scoring

Oneway and twoway models


Generative Adversarial Networks (GANs)

Generative Adversarial Networks (GANs)

Fundamental components of the GANs

GAN architectures

Bidirectional GAN

Training generative models

Credit Scoring using GANs


Exercise 14: Credit Scoring using Deep Learning Feed Forward

Exercise 15: Credit scoring using deep learning CNN

Exercise 16: Credit Scoring using Deep Learning LSTM

Exercise 17: Credit Scoring using GANs
Module 7: Tuning Hyperparameters in Deep Learning

Hyperparameterization

Grid search

Random search

Bayesian Optimization

Train test split ratio

Learning rate in optimization algorithms (e.g. gradient descent)

Selection of optimization algorithm (e.g., gradient descent, stochastic gradient descent, or Adam optimizer)

Activation function selection in a (nn) layer neural network (e.g. Sigmoid, ReLU, Tanh)

Selection of loss, cost and custom function

Number of hidden layers in an NN

Number of activation units in each layer

The dropout rate in nn (dropout probability)

Number of iterations (epochs) in training a nn

Number of clusters in a clustering task

Kernel or filter size in convolutional layers

Pooling size

batch size

Interpretation of the Shap model

Exercise 18: Tuning hyperparameters in Xboosting, Random forest and SVM models for credit scoring

Exercise 19: Tuning hyperparameters in Deep Learning model for credit scoring
Module 8: The development process of the scorecard

Scoring assignment

Scorecard Classification

Scorecard WOE

Binary Scorecard

Continuous Scorecard


Scorecard Rescaling

Factor and Offset Analysis

Scorecard WOE

Binary Scorecard


Reject Inference Techniques

Cutoff

Parceling

Fuzzy Augmentation

Machine Learning


Advanced Cut Point Techniques

Cutoff optimization using ROC curves


Exercise 20: Creating a scorecard using Excel, R, and Python
QUANTUM COMPUTING
Module 9: Quantum Computing and Algorithms

Future of quantum computing in banking

Is it necessary to know quantum mechanics?

QIS Hardware and Apps

Quantum operations

Qubit representation

Measurement

Overlap

Matrix multiplication

Qubit operations

Multiple Quantum Circuits

Entanglement

Deutsch Algorithm

Quantum Fourier transform and search algorithms

Hybrid quantumclassical algorithms

Quantum annealing, simulation and optimization of algorithms

Quantum machine learning algorithms

Exercise 21: Quantum operations
QUANTUM MACHINE LEARNING
Module 10: Development of Credit Scoring using Quantum Machine Learning

What is quantum machine learning?

Qubit and Quantum States

Quantum Automatic Machine Algorithms

Quantum circuits

K means quantum

Support Vector Machine

Support Vector Quantum Machine

Variational quantum classifier

Training quantum machine learning models

Quantum Neural Networks

Quantum GAN

Quantum Boltzmann machines

Quantum machine learning in Credit Risk

Quantum machine learning in credit scoring

Quantum software

Exercise 22: Quantum Support Vector Machine to develop credit scoring model

Exercise 23: Quantum feed forward Neural Networks to develop a credit scoring model and PD estimation

Exercise 24: Quantum Convoluted Neural Networks to develop a credit scoring model and PD estimation
Module 11: Tensor Networks for Machine Learning

What are tensor networks?

Quantum Entanglement

Tensor networks in machine learning

Tensor networks in unsupervised models

Tensor networks in SVM

Tensor networks in NN

NN tensioning

Application of tensor networks in credit scoring models

Exercise 25: Construction of credit scoring and PD using tensor networks
PROBABILISTIC MACHINE LEARNING
Module 12: Probabilistic Machine Learning

Probability

Gaussian models

Bayesian Statistics

Bayesian logistic regression

Kernel family

Gaussian processes

Gaussian processes for regression


Hidden Markov Model

Markov chain Monte Carlo (MCMC)

Metropolis Hastings algorithm


Machine Learning Probabilistic Model

Bayesian Boosting

Bayesian Neural Networks

Exercise 26: Gaussian process for regression

Exercise 27: Bayesian neural networks
ADVANCED VALIDATION
Module 14: Advanced Validation of AI Models

Integration of stateoftheart methods in interpretable machine learning and model diagnosis.

Data Pipeline

Feature Selection

Blackbox Models

Posthoc Explainability

Global Explainability

Local Explainability

Model Interpretability

Diagnosis: Accuracy, WeakSpot, Overfit, Reliability, Robustness, Resilience, Fairness

Model comparison

Comparative for Regression and Classification

Fairness Comparison


Exercise 28: Validation and diagnosis of advanced credit scoring models
IRB PD
Module 15: Probability of Default PD IRB

PD estimation

Econometric models

Machine Learning Models

Data requirement

Risk drivers and credit scoring criteria

Rating philosophy

Pool Treatment


PD Calibration

Default Definition

Long run average for PD

Technical defaults and technical default filters

Data requirement

One Year Default Rate Calculation

LongTerm Default Rate Calculation


PD Model Risk

Conservatism Margin


PD Calibration Techniques

Anchor Point Estimate

Mapping from Score to PD

Adjustment to the PD Economic Cycle

Rating Philosophy

PD Trough The Cycle (PD TTC) models

PD Point in Time PD (PD PIT ) models



PD Calibration of Models Using Machine and Deep Learning

Margin of Caution

Exercise 29: Modeling the Margin of Caution PD
Module 16: Econometric and AI Models of PD IRB

PD estimation

Treatment of Panel data

Econometric models to estimate PD

PD Logistic Regression

PD Probit Regression

PD COX regression of survival

PD Loglog Complementary

PD Regression Data Panel

PD Bayesian Logistic Regression


Machine Learning models to estimate PD

Cox XGBoost

Survival Tree

Random Survival Forest

Deep Learning Survival

PD Neural Networks

PD Quantum Neural Networks


PD Calibration

Calibration of econometric models

Anchor Point Estimate

PD calibration by vintages or vintages

Vintage analysis

PD Marginal

PD Forward

Cumulative PD


Econometric Models

Exercise 30: Using COX regression to estimate the PD

Exercise 31: Using logistic regression with panel data to estimate PD

Exercise 33: Using Bayesian Logistic Regression to estimate PD

Exercise 34: Using PD LASSO regression to estimate PD

Machine Learning and Quantum Machine Learning

Exercise 35: Using Random Forest Survival to estimate PD

Exercise 36: Using Cox Xboost to estimate PD

Exercise 37: Using Deep Learning survival to estimate PD

Exercise 38: Using Feed Forward NN to estimate PD

Exercise 39: Using Quantum Neural Networks to estimate PD
Module 17: PD Calibration IRB

Concept of adjustment to central tendency

Bayesian approach

PD calibration in developed countries

PD calibration in emerging countries

Scaled PD Calibration

Scaled Likelihood ratio calibration

Smoothing of PD curves

Quasi moment matching

Approximation methods

Scaled beta distribution

Asymmetric Laplace distribution


Rubber function

Platt scaling

Broken curve model

Isotonic regression

Gaussian Process Regression

Exercise 40: PD calibration using Platt scaling and isotonic regression

Exercise 41: PD calibration using Gaussian Process Regression

Exercise 42: Calibration of the PD asymmetric Laplace distribution
Module 18: Bayesian PD and Gaussian Process

Bayesian and deterministic approach

Expert judgment

Prior distributions

Bayes' theorem

Posterior distributions

Bayesian PD Estimation

Markov Chain–Monte Carlo MCMC approach

Credibility intervals

Bayesian PD in practice

Calibration with Bayesian approach

Process Gaussian regression

Exercise 43: Logistic Model Bayesian PD in Python

Exercise 44: PD using MCMC in R

Exercise 45: PD using Process Gaussian Regression
Module 19: Low Default Portfolio PD IRB

Confidence interval approach for PD LDP

PD estimation without correlations

PD estimation with correlations

Oneperiod and multiperiod estimation


Bayesian PD estimation for LDP

Neutral Bayesian

Conservative Bayesian

Expert judgment


Real analysis of PD of Corporate, Sovereign and Retail portfolios

LASSO regression to measure corporate default rate

Generating synthetic data for LDP using GAN

Exercise 46: PD LDP confidence interval approach in R

Exercise 47: Multiperiod confidence interval approach PD LDP

Exercise 48: Neutral Bayesian PD in R

Exercise 49: Conservative Bayesian PD in R

Exercise 50: Generating synthetic data with GAN for estimating PD
IFRS 9 PD
Module 20: IFRS 9 PD Forecasting

IFRS 9 requirements

Probability Weighted Outcome

Forward Looking


Lifetime PD modeling

PD Forecasting Modeling

PD Point in Time Forecasting

PS TTC Forecasting

Markov models

PIT PD Forecasting Models

ARIMA

VAR

VARMAX

ASRF

Traditional LSTM

Bayesian LSTM

Quantum LSTM


Exercise 51: Forecasting PD using VARMAX in R

Exercise 52: PD forecasting using quantum LSTM
Module 21: Lifetime PD

Lifetime PD in consumer portfolio

Lifetime PD in mortgage portfolio

Lifetime PD in Credit Card portfolios

Lifetime PD in SMEs portfolio

Vintage model

Exogenous Maturity Vintage EMV Model

Decomposition analysis

COVID19 Pandemic Application

Advantages and disadvantages


Basel ASRF model

Matrix ASRF model

Leveraging IRB in IFRS 9

Advantages and disadvantages


Regression Models

Logistic Multinomial Regression

Ordinal Probit Regression


Survival Models

Kaplan–Meier

Cox Regression

Advantages and disadvantages


Markov models

MultiState Markov Model

Advantages and disadvantages


Machine Learning Model

Support Vector Machine


Deep Learning Models

Neural network architecture


Lifetime PD Extrapolation Models

Lifetime PD Calibration

Exercise 53: using multinomial regression for estimating Lifetime PD

Exercise 54: using Multi State Markov model for estimating Lifetime PD

Exercise 55: using matrix ASRF model for estimating Lifetime PD

Exercise 56: using Quantum SVM for estimating Lifetime PD

Exercise 57: using traditional SVM in Python for estimating Lifetime PD

Exercise 58: using traditional Deep Learning for estimating Lifetime PD

Exercise 59: using Quantum Deep Learning for estimating Lifetime PD
IRB LGD
Module 22: LGD IRB in Retail Portfolios

Impact of COVID19 on LGD

Definition of default

Moratoriums

Renovations and restructuring

Default Cycle

Real Default Cycles


Expected Loss and Unexpected Loss in the LGD

LGD in Default

Default Weighted Average LGD or Exposureweighted average LGD

LGD for performing and nonperforming exposures

Treatment of collaterals in the IRB

Workout approach

Techniques to determine the discount rate

Treatment of recoveries, expenses and recovery costs

Default Cycles

Recovery expenses


Downturn LGD in consumer portfolios

Downturn LGD in Mortgages

LGD in consumption

LGD in Mortgages

LGD in companies

LGD for portfolios with replacement
Module 23: Econometric and machine learning models are used to estimate the Loss Given Default (LGD)

Advantages and disadvantages of LGD Predictive Models

Forward Looking models incorporating Macroeconomic variables

Parametric and nonparametric models and transformation regressions

Typology of LGD Multivariate Models

Linear regression and Beta transformation

Linear Regression and Logit Transformation

Linear regression and Box Cox transformation

Logistic and Linear Regression

Logistic and nonlinear regression

Censored Regression

Generalized Additive Model

Beta regression

Inflated beta regression


Support Vector Regression

Support Vector Classification

Random Forest Regression

XGBoosting Regression

Neural networks

Deep learning

Exercise 60: econometric models for estimating LGD:

Logistic and Linear Regression

Beta regression


Exercise 61: Machine Learning and Deep Learning Models for estimating LGD:

Random Forest Regression

XGBoosting Regression

Deep learning Regression


Exercise 62: Comparison of the performance of the models using Calibration and precision tests.
IFRS 9 LGD
Module 24: LGD for IFRS 9

Comparison of IRB LGD vs. IFRS 9

Impact on COVID19

IFRS 9 requirements

Probability Weighted

Forward Looking


IRB LGD adjustments

Selection of Interest Rates

Allocation of Costs

floors

Treatment of collateral over time

Duration of COVID19


LGD PIT modeling

Collateral Modeling

LGD IFRS 9 for portfolio companies

LGD IFRS 9 for mortgage portfolio

LGD IFRS 9 for corporate portfolios

IFRS 9 LGD using LASSO Regression

Machine Learning Models

Support Vector Machine

Random Forest

Xgboost

Neural Networks


Exercise 63: Estimation and adjustments for LGD IFRS 9 using Random Forest regression

Exercise 64: Estimation and adjustments for LGD IFRS 9 Beta Regression and Neural Networks

Exercise 65: Estimating IFRS 9 LGD using Xgboost regression in Python

Exercise 66: Estimating IFRS 9 LGD using traditional SVM

Exercise 67: Estimating IFRS 9 LGD using quantum SVM

Exercise 68: Estimating IFRS 9 LGD using traditional neural networks

Exercise 69: Estimating IFRS 9 LGD using Bayesian neural networks
IRB EAD
Module 25: Development of models to estimate the Exposure at Default (EAD) and the Credit Conversion Factor (CCF) for the Internal RatingsBased (IRB) approach

Impact of COVID19 on credit lines

Guidelines for estimating CCF

Guidelines for Estimating CCF Downturn

Temporal horizon

Transformations to model the CCF

Approaches to Estimating CCF

Fixed Horizon approach

Cohort Approach

Variable focus time horizon


Econometric Models

Beta regression

Inflated beta regression

Fractional Response Regression

Mixed Effect Model


Machine Learning Models

Neural networks

SVM


Intensity model to measure the withdrawal of credit lines

Exercise 70: Neural network regression model for estimating CCF

Exercise 71: Support Vector Regression Model in Python for estimating CCF

Exercise 72: Neural Networks and beta regression in R for estimating CCF

Exercise 73: Quantum SVM and Classic Regression SVM for estimating CCF
IFRS 9 EAD
Prepayment Rate
Module 26: Contractual Options

Prepayment and other options

IFRS 9 requirements

Probability Weighted

Forward Looking

IFRS 9 prepayment modeling

Cox Regression

Logistic regression


Survival rate estimate

Joint Probability Model with PD Lifetime

Exercise 74: prepayment model based on random forest regression, implemented using R and Excel.
Module 27: Lifetime EAD for Credit Lines

The pandemic has significantly affected the utilization of credit lines.

Lifetime measurement in credit cards

Lifetime EAD

IFRS 9 requirements

Probability Weighted

Forward Looking

Adjustments in the EAD

Interest Accrual

CCF PIT Estimate

CCF Lifetime Estimate

EAD lifetime modeling

Model of the use of credit lines with macroeconomic variables

Credit card abandonment adjustment

EAD Lifetime model for pool of lines of credit

vintage model

Chain Ladder Approach


Exercise 75: Neural network model for line of credit

Exercise 76: EAD Lifetime model for line of credit
BACKTESTING VALIDATION
Module 28: PD Backtesting

Validation of PD

Backtesting of PD

Statistical test:

Hosmer Lameshow test

Normal test

Binomial Test

Spiegelhalter test

Redelmeier Test

Traffic Light Approach


Traffic Light Analysis and Dashboard

Stability Test for PD

Comparing Predicted Probability of Default (PD) with Default Rate over Time.

Performing a validation through Monte Carlo simulation

Exercise 77: Backtesting PD in Excel
Module 29: LGD Backtesting

LGD Backtesting

Accuracy ratio

Absolute accuracy indicator

Confidence Intervals

Transition analysis

RR Analysis using Triangles

Advanced LGD Backtesting with a vintage approach

Backtesting for econometric models:

Test calibration

Ttest

Wilcoxon signed rank test

Accuracy Test

F Test

Ansari–Bradley Test

Exercise 78: Comparison of the performance of the models using Calibration and precision tests.
Module 30: EAD Backtesting

EAD Performance

R squared

Pearson coefficient

Spearman correlation

Validation using ROC, KS and Gini

Exercise 79: Comparison of the performance of EAD models
STRESS TESTING IRB
Module 31: Deep Learning for Modeling Macroeconomic Dynamics

Macroeconomic models

Neoclassical growth model

Partial differential equations

DSGE Stochastic Dynamic General Equilibrium Models

Deep learning architectures

Reinforcement Learning

Advanced Scenario Analysis

Exercise 80: using neural networks for Bellman equation macroeconomic model
Module 32: Deep Learning models for macroeconomic projections

Multivariate Models

VAR Autoregressive Vector Models

ARCH models

GARCH models

GARCH Models Multivariate Copulas

VEC Error Correction Vector Model

Johansen's method


Machine Learning Models

Supported Vector Machine

Neural network

Forecasting market time series yields

NN and SVM algorithms for performance forecasting

Forecasting volatility NN vs. Garch


Development and validation base


Deep learning

Recurrent Neural Networks RNN

Elman's Neural Network

Jordan Neural Network

Basic structure of RNN

Long short term memory LSTM

temporary windows

Development and validation sample

Regression

Sequence modeling


Quantum Deep Learning

Time series analysis with Facebook Prophet

Prediction of the spread of Covid19

Exercise 81: Chargeoff model with VAR and VEC

Exercise 82: Forecasting PD using financial series and Bayesian LSTM indices in Python

Exercise 83: Pandemic Forecasting using Multivariate RNN LSTM in Python

Exercise 84: Forecasting PD using quantum neural networks
Module 33: Stress Testing PD and LGD

Temporal horizon

Multiperiod approach

Data required

Impact on P&L, RWA and Capital

Macroeconomic Stress Scenarios in consumption

Expert

Statistical

Regulatory


Stress Testing PD :

Credit Portfolio View

Multiyear Approach

Reverse Stress Testing

Rescaling

Cox Regression


Stress Testing of the Transition Matrix

Credit Portfolio View Approach

Credit cycle index

Multifactor Extension


Stress Testing LGD:

Downturn LGD : Mixed Distribution Approach

PD/LGD Multiyear Approach modeling

LGD stress test for mortgage portfolios


Stress Testing of:

Net Charge Off

Rating/Scoring transition matrices

Recovery Rate and LGD


Exercise 85: using Credit Portfolio Views for Stress Testing PD

Exercise 86: using Bayesian LSTM for Stress Testing PD

Exercise 87: using Variational Quantum Regression for Stress Testing PD

Exercise 88: using MARS Model for Stress Testing PD

Exercise 89: using LASSO regression for Stress Testing PD
Module 34: Stress Testing in corporate portfolios

Temporal horizon

Data required

Main Macroeconomic variables

Impact on P&L, RWA and Capital

ASRF model

Creditmetrics model

Using Transition Matrices

Use of the credit cycle index

Default forecasting

Stress Test Methodology for corporate portfolios

Impact on RWA and Capital

Exercise 90: Performing stress tests for Probability of Default (PD) in corporate portfolios using a transition matrix and the ASRF model
STRESS TESTING ECL IFRS 9
Module 35: Stress Testing under ECL IFRS 9

Stress testing under IFRS 9 and COVID19

Pandemic scenarios applied to the ECL calculation

Stress Testing of IFRS 9 parameters

EBA Stress Testing 2023

Treatment of the moratorium

Possible regulatory scenarios

Impact on P&L

PIT starting parameters

PIT projected parameters

Calculation of nonproductive assets and impairments

Changes in the stock of provisions

Changes in the stock of provisions for exposures phase S1

Changes in the stock of provisions for exposures phase S2

Changes in the stock of provisions of exposures phase S3

Sovereign Exposure Impairment Losses

Impact on capital

Internal Stress Testing Model for ECL IFRS 9

Stage Migration

Stage Transition Matrix

Stress testing of PDs and credit migrations

Stress testing of exhibitions

Stress testing of recoveries

Exercise 91: An internal global exercise was conducted for ECL stress testing using R and Excel
Module 36: Quantum Stress Testing

Quantum economics

Classic Monte Carlo simulation

Quantum Monte Carlo

Coding Monte Carlo problem

Breadth Estimation

Acceleration applying the amplitude estimation algorithm

DGSE model using neural networks

Quantum Monte Carlo Simulation vs Normal Monte Carlo Simulation

Exercise 92: DGSE model using deep learning

Exercise 93: Quantum Monte Carlo Simulation vs. Classical Monte Carlo Simulation
CREDIT RISK PORTFOLIO
Module 37: Economic Capital Models

Regulatory Capital

Economic Capital Methodologies

Correlation of Assets and Default

Unexpected Tax Loss

ASRF Economic Capital Models

Business Models

Kmv

creditmetrics

Credit Portfolio View

Credit risk +


Economic capital management

Allocating Economic Capital

Exercise 94: Portfolio Approach: Estimation of EL, UL, ULC, Correlation and Economic Capital in Excel

Exercise 95: Creditrisk + in SAS

Exercise 96: Creditmetrics in Excel and R

Exercise 97: Singlefactor model in Excel
QUANTUM COMPUTING FOR ECONOMIC CAPITAL
Module 38: Quantum computing for Economic Capital

Distribution of credit risk losses

Quantum uncertainty model

Circuit Definition

Quadratic acceleration over the classical Monte Carlo simulation

Expected loss

Cumulative distribution function

VaR

Expected Shortfall

Exercise 98: Estimation EL, VAR, ES of quantum credit risk
Quantifying climate changerelated credit risk
Module 39: Climate Risk in Credit Risk

Credit Risk Transition risk

Scenario analysis to assess the transition risk component of a portfolio's credit risk

Introduction: preparing banks for the low carbon transition

A growing need for climate scenario analysis

The challenge for banks

Take advantage of and integrate the resources available to banks


An integrated approach to transition risk assessment Transition scenarios

Understand transition scenarios and their sources

Using scenarios for transition risk assessment

Closing the gap between climate scenarios and financial risk assessment


Borrower Level Calibration

Portfolio Impact Assessment

Link expected loss to transition impacts on portfolios

Assessment of probability of default (PD)

Loss Given Default (LGD) Assessment


Putting the Approach to Work: Lessons Learned from Banking Pilots

Piloting the transition risk methodology

Definition of sectors and segments

Evaluate the relative sensitivities of the segments

Determination of calibration points at the borrower level Case studies and results

The pilot transition scenario

pilot results


Transition Opportunities: Exploring an Institutional Strategy

evaluating the market

Grounding Opportunity Assessments in Scenario Analysis

Assessing the market attractiveness of the segment

Identification of banking capabilities

Discovering the opportunities with the greatest potential


Future Directions: Developing the Next Generation of Transition Risk Analysis


Physical risks and opportunities

An Integrated Approach to Physical Risk Assessment

Borrower Characteristics

Insurance as a risk mitigator due to extreme climatic and meteorological events

climate change scenarios


Impacts of climate change on the probability of default PD

Evaluation of changes in the productivity of the sector

Adjustment of income statement metrics

Determination of changes in the probability of default


Real Estate: Climate Change Impacts on LTV LoantoValue

Estimation of the impacts of extreme events on the value of properties.

Determining Changes in LTV LoantoValue Ratio


Physical Opportunities: Exploring an Institutional Strategy

Taxonomy of opportunities and data sources

evaluating the market

Evaluation of the financing demand of the sector

Sector evaluation

Assess the institutional capacity and market positioning of a bank

Evaluate opportunities


Future Directions: Towards the Next Generation of Physical Risk and Opportunity Analysis

Develop internal analytics and capabilities within banks

Strengthening the research base

Develop analytical platforms and tools to support physical assessments of risks and opportunities.

Improve information flows on physical risk and adaptation between banks and borrowers

Improve dialogue with governments and insurers



Exercise 99: Estimating PD and DD adjusted for climate change in transition risk

Exercise 100: Estimating PD and DD adjusted for climate change in physical risk