Laboratory of Bank Loan Pricing
A.Y. 2023/2024
Learning objectives
Students will gain a thorough understanding of the practices currently used by banks when originating (pricing) a bank loan. The course will offer insight into the various areas of the bank involved when pricing a bank loan (Finance department for funding, Chief Risk Officer area for loans balancesheet provisioning, CFO area for capital allocation). The course will then present students the case for the impact on the Bank Loan Pricing of the new Securitization Framework and advanced Machine Learning techniques for default prediction (internal PD model) and projection (satellite PD model).
Expected learning outcomes
At the end of the course students will gain an overview of the various departments at work within a bank when originating a loan. They will also get a measure of the impact on loan pricing of alternative Machine Learning methodologies for default prediction and the impact of the new securitization framework for loan portfolio capital allocation.
Lesson period: Third trimester
Assessment methods: Giudizio di approvazione
Assessment result: superato/non superato
Single course
This course cannot be attended as a single course. Please check our list of single courses to find the ones available for enrolment.
Course syllabus and organization
Single session
Responsible
Lesson period
Third trimester
Course syllabus
Below a list of subjects to be discussed in class:
Market price (a) vs Origination Price (b):
(a) Interest Rate Curve calibration, Credit Spread Curve calibration, CDS curve calibration.
(b) Bank Loan Origination pricing via target IRR: CoL (cost of liquidity) + CoC (cost of credit) + CoK (cost of capital).
CoC: Lifetime Expected Loss. Rating Transition Matrices. Probability of Default and Loss Given Default.
CoC: Default prediction models. Typical dataset structure, module layout and prediction horizon.
CoC: Default prediction models. Univariate and Multivariate Logistic regressions.
CoC: Default prediction models. Input variables encoding and Cross Validation.
CoC: Default prediction models. Regression and Classification Trees.
CoC: Low Default portfolios PD estimation and backtesting
CoC: Default prediction models. Bank Loan Pricing impact of alternative machine learning and ensemble algorithms (Random Forest: Bagging (Bootstrap aggregating) and Boosting ensemble algorithms.).
CoC: Macro-economic scenario projections. Satellite Credit Portfolio Models as sparse problems.
CoC: Satellite Credit Portfolio Models: Bayesian Model Average.
CoL: Blended curve construction.
CoL: regulatory framework: LCR and NSFR.
CoK: Regulatory capital and Risk Weighted Assets
CoK: Risk Weighted Assets for Standardized and Advanced Internal Rating Based Banks.
CoK: new securitization framework and its impact on the RWA for a securitized portfolio
CoK: Risk Weighted Assets calculation for a securitized portfolio
CoK: Significant Risk Transfer for a securitized portfolio
Market price (a) vs Origination Price (b):
(a) Interest Rate Curve calibration, Credit Spread Curve calibration, CDS curve calibration.
(b) Bank Loan Origination pricing via target IRR: CoL (cost of liquidity) + CoC (cost of credit) + CoK (cost of capital).
CoC: Lifetime Expected Loss. Rating Transition Matrices. Probability of Default and Loss Given Default.
CoC: Default prediction models. Typical dataset structure, module layout and prediction horizon.
CoC: Default prediction models. Univariate and Multivariate Logistic regressions.
CoC: Default prediction models. Input variables encoding and Cross Validation.
CoC: Default prediction models. Regression and Classification Trees.
CoC: Low Default portfolios PD estimation and backtesting
CoC: Default prediction models. Bank Loan Pricing impact of alternative machine learning and ensemble algorithms (Random Forest: Bagging (Bootstrap aggregating) and Boosting ensemble algorithms.).
CoC: Macro-economic scenario projections. Satellite Credit Portfolio Models as sparse problems.
CoC: Satellite Credit Portfolio Models: Bayesian Model Average.
CoL: Blended curve construction.
CoL: regulatory framework: LCR and NSFR.
CoK: Regulatory capital and Risk Weighted Assets
CoK: Risk Weighted Assets for Standardized and Advanced Internal Rating Based Banks.
CoK: new securitization framework and its impact on the RWA for a securitized portfolio
CoK: Risk Weighted Assets calculation for a securitized portfolio
CoK: Significant Risk Transfer for a securitized portfolio
Prerequisites for admission
Students must have taken either the Data Mining and Computational Statistics or the Risk Management exam.
In order to complete the final assignment, students will need to have programming skills in any language of their choice (MatLab, R, Python, C).
In order to complete the final assignment, students will need to have programming skills in any language of their choice (MatLab, R, Python, C).
Teaching methods
In presence classes only. Videosharing with the class material. Seminars on specific topics are going to be taught within the Laboratory by professionals from leading Financial Institutions.
Teaching Resources
Caprioli S., Cogo R. Back-testing credit risk parameters on low default portfolios: a Bayesian approach with an application to Sovereign Risk, working paper 2023
Sala-i Martin, X., Doppelhofer, G., and Miller, R. I. (2004). Determinants of longterm growth: A bayesian averaging of classical estimates (bace) approach. American Eonomic Review, 94(4):813-835.
Schwarz, G. (1978). Estimating the dimension of a model. Annals of Statistics, 6(2):461-464.
Tasche, D. (2013). Bayesian estimation of probabilities of default for low default portfolios. Journal of Risk Management in Financial Institutions, 6(3): pages 302-326.
Torresetti, R. (2021), A comparison of the performance of alternative Machine Learning algorithms on a credit risk dataset, Working Paper, available at www.ssrn.com
Torresetti, R. (2021), Bayesian Model Averaging for Satellite Models in Bank's Credit Risk Projections, Working Paper, available at www.ssrn.com
Sala-i Martin, X., Doppelhofer, G., and Miller, R. I. (2004). Determinants of longterm growth: A bayesian averaging of classical estimates (bace) approach. American Eonomic Review, 94(4):813-835.
Schwarz, G. (1978). Estimating the dimension of a model. Annals of Statistics, 6(2):461-464.
Tasche, D. (2013). Bayesian estimation of probabilities of default for low default portfolios. Journal of Risk Management in Financial Institutions, 6(3): pages 302-326.
Torresetti, R. (2021), A comparison of the performance of alternative Machine Learning algorithms on a credit risk dataset, Working Paper, available at www.ssrn.com
Torresetti, R. (2021), Bayesian Model Averaging for Satellite Models in Bank's Credit Risk Projections, Working Paper, available at www.ssrn.com
Assessment methods and Criteria
Assignment to be submitted individually by each student (no group projects)
SECS-P/01 - ECONOMICS
SECS-P/02 - ECONOMIC POLICY
SECS-P/02 - ECONOMIC POLICY
Laboratory activity: 20 hours
Professor:
Torresetti Roberto
Professor(s)
Reception:
all Fridays 18:00-19:00, from 23rd April to 25th June