Journal of Risk Model Validation

Steve Satchell
Trinity College, University of Cambridge

This issue of The Journal of Risk Model Validation covers artificial intelligence (AI), the impact of tariffs, risk model validation within regulatory frameworks, and fund homogenization. It is a pleasure for me and the editorial board to highlight timely papers but never at the expense of academic quality, which can so often be sacrificed in the pursuit of topicality.

The issue’s first paper, by Abraham M. Izquierdo, Francisco Pérez-Hernández and María-del-Mar Camacho-Miñano, has the somewhat lengthy title “Demand deposit balance prediction models under the interest rate risk in the banking book guidelines: an empirical analysis integrating time-series models and machine learning predictions in Mexican banks”. This study is, as far as I can recall, the first paper in this journal to use Mexican data; I am delighted to publish it. Izquierdo et al explore the importance of interest rate risk in the banking book (IRRBB) regulations set forth by the Basel Committee on Banking Supervision, and they emphasize the need for financial institutions to develop robust models for forecasting demand deposit balances while adhering to regulatory guidelines. To ensure these models meet the stringent requirements of risk model validation, the authors implement comprehensive validation techniques that assess predictive accuracy and robustness using a random sample of demand deposit balances in the Mexican banking sector. Their research contrasts the forecasting results used to construct survival rates and attrition curves through traditional time-series methods with the use of AI algorithms. Their results demonstrate that machine learning/AI techniques improve the prediction results for demand deposit balances, allowing the authors to efficiently construct attrition curves and survival rates, providing better risk management tools for financial institutions facing the challenges posed by the IRRBB.

The second paper in the issue, “Validating bank risk models under trade war stress: a framework for adaptive stress testing with AI-driven calibration and crossindustry applications” by Krishan Kumar Sharma, highlights problems in risk modelling generally, and risk validation in particular, caused by the increased use of tariffs globally. The escalation of trade tensions and tariff shocks has created persistent uncertainty for financial institutions by directly impacting the integrity of traditional bank risk models. Conventional frameworks often fail to capture the nonlinear and sector-specific effects of tariffs on credit, market and operational risk, which results in misestimated portfolio losses and distorted capital allocation. This paper develops a comprehensive methodology, focusing on the validation and enhancement of risk models specifically under trade war conditions. A case study of a US regional bank is discussed. By integrating empirical stress testing with robustness checks, AI tools and conservative governance as well as a reproducible calibration protocol, the framework strengthens resilience, supports regulatory compliance and ensures forward-looking model risk management under sustained geopolitical uncertainty. This paper contributes to the growing field of political risk in modelling, which is an area of research I am very pleased to publish.

“Exceedance-based back testing of expected shortfall”, our third paper, is by Andrii Liakhovchenko and Dmitrij Celov. Backtesting is a topic of perennial interest to readers of our journal. The Fundamental Review of the Trading Book encourages financial institutions to shift from the value-at-risk (VaR) risk measure to the expected shortfall (ES) risk measure when measuring market risk capital. Liakhovchenko and Celov examine the application of exceedance-based validation (or “backtesting”) methods, commonly used for VaR model validation, to the validation of ES models. Their approach includes finding the quantile value corresponding to the ES for four different estimation methods. Their paper also investigates the empirical stability of this quantile and proposes an adjustment to the traditional backtesting approaches that helps to accommodate instabilities. The application of the approach is illustrated using real-world Baltic equity index returns data for 90%, 95% and 97.5% confidence levels. Their findings show that the direct application of exceedance-based methods to the validation of ES models can be achieved, even in the case of an unstable ES quantile.

The issue’s final paper, “Quantitative fund homogenization and systemic risk in the stock market” by Mengyu Li, Qian Zheng and Shan Ji, investigates fund homogenization. While this concept is intuitively simple, its technicalities may be unfamiliar to readers. This is important, because if all funds follow the same strategy, markets struggle to trade/clear. Homogenization can be measured in many ways; in this paper a starting point for this measurement is the average correlation of the residuals from capital asset pricing model regressions, averaged over the set of funds under consideration – in this case data from 421 active quantitative funds in China from January 2015 to March 2024. Li et al study the impact of quantitative fund homogenization on systemic risk in the Chinese stock market, with implications for risk model validation in high-frequency trading environments. They design a homogenization measurement method from the perspectives of return rates and Sharpe ratios, validating its robustness through cross-sectional and time series analyses. They also explain the impact of quantitative fund homogenization on the risk to individual funds as well as systemic risk on the aggregate fund, and they provide policy recommendations for promoting the healthy development of quantitative funds and mitigating the systemic risks of the Chinese stock market. Looking at a universe of funds in analysing market behaviour while also investigating the impact of the aggregate on the constituent funds has always struck me as a valuable research approach.