Journal of Risk Model Validation

Steve Satchell
Trinity College, University of Cambridge

This issue of The Journal of Risk Model Validation consists of three papers that will greatly please any readers who enjoy acronyms. All three papers use sophisticated techniques to model and validate risk, and they are technically challenging. They do, however, make an interesting read and I hope you will learn much from them.

The issue’s first paper, by Ruhao Chen, Tong-Yu Lu, Jiyuan Min and Wenfu Xu, is “A comprehensive explainable approach for imbalanced financial distress prediction”. As the title suggests, this paper looks at financial distress prediction (FDP) models, which are important risk prevention and control tools for monitoring the financial status of companies. To simultaneously address the issue of imbalanced data and the explanation of “black box” models in FDP, Chen et al propose an explicable approach for imbalanced FDP through focal-aware extreme gradient boosting (FL-XGBoost). XGBoost is a framework used for building machine learning models. The authors’ approach addresses the difficulty of identifying minority-class samples in imbalanced data sets by introducing the focal loss function into the XGBoost framework. Validation tests of different FDP models show the strong performance of their proposed model. Moreover, they use three explanatory methods to generate explanations for different FDP model users. The experimental results based on different kinds of imbalanced data sets show that FL-XGBoost has strong classification performance. Further, the explanatory methods can increase the transparency and credibility of their proposed black box model. The focus on model identification and the fact that the approach can be easily generalized make this paper of particular value to readers.

Our second paper is “Failure mode and effects analysis–analytic hierarchy process (FMEA-AHP) model in supplier risk management” by Marija Panić and Que Xiaojun. I was intrigued to discover that the FMEA approach has been around for more than 70 years, and it has been validated via application in many areas. This paper uses the FMEA–AHP risk model to identify and manage supplier risks in a mining firm. First, the FMEA approach is used to identify and evaluate the supplier’s risks and to determine the risk priority. The AHP is then used to suggest several potential solutions for reducing the risks identified by the FMEA, with each solution being evaluated based on several criteria. Validation of the risk model in this paper confirmed its usefulness. The authors focus on a particular application, but their approach also has great generality; Panić and Xiaojun claim that their approach “provides an enterprise risk management model that is universally applicable and adaptable, no matter the type of risk or the size or activity of the company”.

The final paper in the issue is “Systemic importance identification and risk supervision of banks: evidence from China” by Juan Chen, Qiuli Hu, Lilan Tu and Si Liu. The authors start from the fact that systemic banking risk is related to the interconnectedness of the various banks in the banking system. Indeed, they say: “The systemic importance of banks is closely related to the topology of banking networks.” They construct a banking network based on tail dependence, conduct an analysis of its node characteristics, and explore the network structure and systemic importance of individual banks using the Laplacian matrix method to compute the smallest eigenvalue of the grounded matrix. (For noninitiates, a grounded matrix represents the relationship between changes and levels when various banks have been omitted from the analysis.) In addition, Chen et al quantify systemic risk in China using the t-copula conditional value-at-risk approach. They argue that the method of taking the smallest eigenvalue of the grounded Laplacian matrix accurately identifies banks that have a significant systemic impact. Further analysis by the authors reveals that large banks – such as the Industrial and Commercial Bank of China, the Agricultural Bank of China, the Bank of China and the China Construction Bank – occupy central positions in the network, serving as key nodes for information dissemination and risk transmission, thereby exhibiting notable systemic importance. This study provides new perspectives on, and theoretical support for, the identification and regulation of financial network risks, which is of particular importance for strengthening the prevention and control of systemic financial risks. It is worth noting that such models have been used to model changes in opinion in systems where some agents keep their opinions constant, and this application to banking systems is of great interest.