Technical paper/Model validation
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.
A comprehensive explainable approach for imbalanced financial distress prediction
The authors suggest an explainable machine learning method for imbalanced financial distress prediction which uses extreme gradient boosting.
Model risk quantification for machine learning models in credit risk
This paper analyses bank-specific model risk measurement methods with a focus on implemented model risk rating solutions for MLMs and discusses challenges faced by the validation function.
A three-stage fusion model for predicting financial distress considering semantic and sentiment information
The authors apply sentiment analysis to management discussion and analysis texts to aid the prediction of financial distress with an innovative three-phase fusion model.
The impact of deterioration in rating-model discriminatory power on expected losses
The authors propose a means to estimate the effects on a portfolio’s expected credit loss created by underwriting model risks.
A study of China’s financial market risks in the context of Covid-19, based on a rolling generalized autoregressive score model using the asymmetric Laplace distribution
The authors construct a risk measurement model for the financial market during the Covid-19 pandemic, using data from the Shanghai Stock Exchange for empirical analysis.
Financial distress prediction with optimal decision trees based on the optimal sampling probability
The authors propose and validate a tree-based ensemble model for financial distress prediction which is demonstrated to outperform comparative models.
A new automated model validation tool for financial institutions
The authors put forward a novel automated validation tool, based on US Federal Reserve and Office of the Comptroller of the Currency regulatory guidance, which is used to to validate predictive models for financial organizations.
A modified hybrid feature-selection method based on a filter and wrapper approach for credit risk forecasting
This paper proposes the chi-squared with recursive feature elimination method: a means of feature-selection which aims to improve classification performance using fewer features.
The validation of different systemic risk measurement models
The authors incorporate a capital buffer to the DebtRank model and use data from China's banking industry to compare the proposed model with others.
Performance validation of representative sample-balancing methods in loan credit-scoring scenarios
The authors validate 12 of the most representative sample-balancing methods used for credit-scoring models, finding that a combined SMOTE and Editor Nearest Neighbor method is optimal.
Quantification of model risk with an application to probability of default estimation and stress testing for a large corporate portfolio
This paper discusses the building of obligor-level rather than segment-level hazard rate corporate probability of default models for stress testing.
General bounds on the area under the receiver operating characteristic curve and other performance measures when only a single sensitivity and specificity point is known
Using a single true positive - true negative pair, the author shows how to calculate the area under a ROC curve.
Predicting financial distress of Chinese listed companies using a novel hybrid model framework with an imbalanced-data perspective
In this paper a novel hybrid model framework is constructed to solve the problem of predicting the financial distress of Chinese listed companies using imbalanced data.
Evaluation of backtesting techniques on risk models with different horizons
In this study different value-at-risk (VaR) models are analyzed under different estimation approaches (filtered historical simulation, extreme value theory and Monte Carlo simulation) and backtested with different techniques.
Benchmarking operational risk stress testing models
This paper outlines several approaches to benchmarking operational loss projections under stressed scenarios using both accounting metrics and historical loss experience.
Validation of index and benchmark assignment: adequacy of capturing tail risk
This paper provides practical recommendations for the validation of risk models under the Targeted Review of Internal Models (TRIM).
Optimal allocation of model risk appetite and validation threshold in the Solvency II framework
In this paper, the authors derive an analytical solution for sub-SCR VTs starting with a model risk appetite (MRA) that defines acceptable errors for an insurer’s total SCR.
Evaluating the risk performance of online peer-to-peer lending platforms in China
The objective of this paper is to select effective risk indicators and thus establish a risk index system of P2P platforms so as to evaluate the risk performance of these platforms in China.
Underperforming performance measures? A review of measures for loss given default models
This paper reviews the ways of measuring the performance of LGD models that have been previously used in the literature and also suggests some new measures.
A central limit theorem formulation for empirical bootstrap value-at-risk
In this paper, the importance of the empirical bootstrap (EB) in assessing minimal operational risk capital is discussed, and an alternative way of estimating minimal operational risk capital using a central limit theorem (CLT) formulation is presented.