Technical paper/Credit risk modelling
Interpretable machine learning for default risk prediction in stress testing
This paper proposes a benchmark model which can be used to predict the forward-looking probability of default of a real-world credit card portfolio.
Model risk quantification for machine learning models in credit risk
Enhancing default prediction in alternative lending: leveraging credit bureau data and machine learning
Analyzing credit risk model problems through natural language processing-based clustering and machine learning: insights from validation reports
Instabilities in Cox proportional hazards models in credit risk
The authors explore possible instabilities in applying Cox PH models and conduct numerical studies to demonstrate the same linear specification error from APC models an occur in Cox PH estimation.
Collateralised exposure modelling: bridging the gap risk
Concentration, leverage and correlations may affect a collateralised equity swap portfolio
Modeling credit risk in the presence of central bank and government intervention
In this paper a simple approach for including central bank and government intervention in credit models is developed and illustrated using the Fed’s data for the CCAR 2021 stress test.
Incorporating small-sample defaults history in loss given default models
This paper proposes a methodology for estimating loss given default (LGD) that accounts for small default sample sizes.
Forecasting consumer credit recovery failure: classification approaches
This study proposes an advanced credit evaluation method for nonperforming consumer loans, which may serve as a new investment opportunity in the post-pandemic era.
NLP and transformer models for credit risk
News feeds are factored into models to predict credit events
Explaining credit ratings through a perpetual-debt structural model
This paper calibrates a perpetual-debt structural model (PDSM) by using Moody’s historical credit ratings.
Credit migration: generating generators
A stochastic time change helps the modelling of rating transition
From incurred loss to current expected credit loss: a forensic analysis of the allowance for loan losses in unconditionally cancelable credit card portfolios
The authors analyze the performance of the CECL framework under plausible assumptions about allocations of future payments to existing credit card loans, a key implementation element.
The effects of customer segmentation, borrower behaviors and analytical methods on the performance of credit scoring models in the agribusiness sector
The main aim of this study is to analyze the joint effects of customer segmentation, borrower characteristics and modeling techniques on the classification accuracy of a scoring model for agribusinesses.
The economics of debt collection, with attention to the issue of salience of collections at the time credit is granted
This paper considers the role of policies that protect consumers from aggressive debt collection tactics.
Bankcard performance during the Great Recession: a consumer-level analysis
This paper investigates factors associated with high credit card loss rates during the period 2008–11 associated with the Great Recession.
Finding the corporate credit cycle for IFRS 9
Decomposing corporate default rates helps identify credit cycles
A joint model of failures and credit ratings
The authors propose a novel framework for credit risk modeling, where default or failure information and rating or expert information are jointly incorporated in the model.
Current expected credit loss procyclicality: it depends on the model
This work looks at a wide range of models to test the degree to which CECL is procyclical for different types of model.
The efficiency of the Anderson–Darling test with a limited sample size: an application to backtesting counterparty credit risk internal models
This paper presents a theoretical and empirical evaluation of the Anderson–Darling test when the sample size is limited.
A statistical technique to enhance application scorecard monitoring
Application scoring plays a critical role in determining the future quality of a lender’s book. It is therefore important to monitor the performance of an application scorecard to ensure it performs as expected.
Counterparty trading limits revisited: from PFE to PFL
The potential future loss is proposed as a replacement for PFE
Smoothing algorithms by constrained maximum likelihood: methodologies and implementations for Comprehensive Capital Analysis and Review stress testing and International Financial Reporting Standard 9 expected credit loss estimation
In this paper, the author proposes smoothing algorithms that are based on constrained maximum likelihood for rating-level PD and for rating migration probability.
Counting processes for retail default modeling
The article discusses the use of counting processes for retail (mortgage) default modeling.