CyberSec Research

This paper presents a machine learning driven framework for sectoral stress testing in the Indian financial market, focusing on financial services, information technology, energy, consumer goods, and pharmaceuticals. Initially, we address the limitations observed in conventional stress testing through dimensionality reduction and latent factor modeling via Principal Component Analysis and Autoencoders. Building on this, we extend the methodology using Variational Autoencoders, which introduces a probabilistic structure to the latent space. This enables Monte Carlo-based scenario generation, allowing for more nuanced, distribution-aware simulation of stressed market conditions. The proposed framework captures complex non-linear dependencies and supports risk estimation through Value-at-Risk and Expected Shortfall. Together, these pipelines demonstrate the potential of Machine Learning approaches to improve the flexibility, robustness, and realism of financial stress testing.

Every publicly traded U.S. company files an annual 10-K report containing critical insights into financial health and risk. We propose Tiny eXplainable Risk Assessor (TinyXRA), a lightweight and explainable transformer-based model that automatically assesses company risk from these reports. Unlike prior work that relies solely on the standard deviation of excess returns (adjusted for the Fama-French model), which indiscriminately penalizes both upside and downside risk, TinyXRA incorporates skewness, kurtosis, and the Sortino ratio for more comprehensive risk assessment. We leverage TinyBERT as our encoder to efficiently process lengthy financial documents, coupled with a novel dynamic, attention-based word cloud mechanism that provides intuitive risk visualization while filtering irrelevant terms. This lightweight design ensures scalable deployment across diverse computing environments with real-time processing capabilities for thousands of financial documents which is essential for production systems with constrained computational resources. We employ triplet loss for risk quartile classification, improving over pairwise loss approaches in existing literature by capturing both the direction and magnitude of risk differences. Our TinyXRA achieves state-of-the-art predictive accuracy across seven test years on a dataset spanning 2013-2024, while providing transparent and interpretable risk assessments. We conduct comprehensive ablation studies to evaluate our contributions and assess model explanations both quantitatively by systematically removing highly attended words and sentences, and qualitatively by examining explanation coherence. The paper concludes with findings, practical implications, limitations, and future research directions. Our code is available at https://github.com/Chen-XueWen/TinyXRA.

The extreme cases of risk measures, when considered within the context of distributional ambiguity, provide significant guidance for practitioners specializing in risk management of quantitative finance and insurance. In contrast to the findings of preceding studies, we focus on the study of extreme-case risk measure under distributional ambiguity with the property of increasing failure rate (IFR). The extreme-case range Value-at-Risk under distributional uncertainty, consisting of given mean and/or variance of distributions with IFR, is provided. The specific characteristics of extreme-case distributions under these constraints have been characterized, a crucial step for numerical simulations. We then apply our main results to stop-loss and limited loss random variables under distributional uncertainty with IFR.

Financial institutions increasingly adopt customer-centric strategies to enhance profitability and build long-term relationships. While Customer Lifetime Value (CLV) is a core metric, its calculations often rely solely on single-entity data, missing insights from customer activities across multiple firms. This study introduces the Potential Customer Lifetime Value (PCLV) framework, leveraging Open Banking (OB) data to estimate customer value comprehensively. We predict retention probability and estimate Potential Contribution Margins (PCM) from competitor data, enabling PCLV calculation. Results show that OB data can be used to estimate PCLV per competitor, indicating a potential upside of 21.06% over the Actual CLV. PCLV offers a strategic tool for managers to strengthen competitiveness by leveraging OB data and boost profitability by driving marketing efforts at the individual customer level to increase the Actual CLV.

Extending Buehler et al.'s 2019 Deep Hedging paradigm, we innovatively employ deep neural networks to parameterize convex-risk minimization (CVaR/ES) for the portfolio tail-risk hedging problem. Through comprehensive numerical experiments on crisis-era bootstrap market simulators -- customizable with transaction costs, risk budgets, liquidity constraints, and market impact -- our end-to-end framework not only achieves significant one-day 99% CVaR reduction but also yields practical insights into friction-aware strategy adaptation, demonstrating robustness and operational viability in realistic markets.

Recent crises like the COVID-19 pandemic and geopolitical tensions have exposed vulnerabilities and caused disruptions of supply chains, leading to product shortages, increased costs, and economic instability. This has prompted increasing efforts to assess systemic risk, namely the effects of firm disruptions on entire economies. However, the ability of firms to react to crises by rewiring their supply links has been largely overlooked, limiting our understanding of production networks resilience. Here we study dynamics and determinants of firm-level systemic risk in the Hungarian production network from 2015 to 2022. We use as benchmark a heuristic maximum entropy null model that generates an ensemble of production networks at equilibrium, by preserving the total input (demand) and output (supply) of each firm at the sector level. We show that the fairly stable set of firms with highest systemic risk undergoes a structural change during COVID-19, as those enabling economic exchanges become key players in the economy -- a result which is not reproduced by the null model. Although the empirical systemic risk aligns well with the null value until the onset of the pandemic, it becomes significantly smaller afterwards as the adaptive behavior of firms leads to a more resilient economy. Furthermore, firms' international trade volume (being a subject of disruption) becomes a significant predictor of their systemic risk. However, international links cannot provide an unequivocal explanation for the observed trends, as imports and exports have opposing effects on local systemic risk through the supply and demand channels.

This paper analyzes realized return behavior across a broad set of crypto assets by estimating heterogeneous exposures to idiosyncratic and systematic risk. A key challenge arises from the latent nature of broader economy-wide risk sources: macro-financial proxies are unavailable at high-frequencies, while the abundance of low-frequency candidates offers limited guidance on empirical relevance. To address this, we develop a two-stage ``divide-and-conquer'' approach. The first stage estimates exposures to high-frequency idiosyncratic and market risk only, using asset-level IV regressions. The second stage identifies latent economy-wide factors by extracting the leading principal component from the model residuals and mapping it to lower-frequency macro-financial uncertainty and sentiment-based indicators via high-dimensional variable selection. Structured patterns of heterogeneity in exposures are uncovered using Mean Group estimators across asset categories. The method is applied to a broad sample of crypto assets, covering more than 80% of total market capitalization. We document short-term mean reversion and significant average exposures to idiosyncratic volatility and illiquidity. Green and DeFi assets are, on average, more exposed to market-level and economy-wide risk than their non-Green and non-DeFi counterparts. By contrast, stablecoins are less exposed to idiosyncratic, market-level, and economy-wide risk factors relative to non-stablecoins. At a conceptual level, our study develops a coherent framework for isolating distinct layers of risk in crypto markets. Empirically, it sheds light on how return sensitivities vary across digital asset categories -- insights that are important for both portfolio design and regulatory oversight.

The Diversification Quotient (DQ), introduced by Han et al. (2025), is a recently proposed measure of portfolio diversification that quantifies the reduction in a portfolio's risk-level parameter attributable to diversification. Grounded in a rigorous theoretical framework, DQ effectively captures heavy tails, common shocks, and enhances efficiency in portfolio optimization. This paper further explores the convergence properties and asymptotic normality of empirical DQ estimators based on Value at Risk (VaR) and Expected Shortfall (ES), with explicit calculation of the asymptotic variance. In contrast to the diversification ratio (DR) proposed by Tasche (2007), which may exhibit diverging asymptotic variance due to its lack of location invariance, the DQ estimators demonstrate greater robustness under various distributional settings. We further evaluate their performance under elliptical distributions and conduct a simulation study to examine their finite-sample behavior. The results offer a solid statistical foundation for the application of DQ in financial risk management and decision-making.

Predicting the probability of default (PD) of prospective loans is a critical objective for financial institutions. In recent years, machine learning (ML) algorithms have achieved remarkable success across a wide variety of prediction tasks; yet, they remain relatively underutilised in credit risk analysis. This paper highlights the opportunities that ML algorithms offer to this field by comparing the performance of five predictive models-Random Forests, Decision Trees, XGBoost, Gradient Boosting and AdaBoost-to the predominantly used logistic regression, over a benchmark dataset from Scheule et al. (Credit Risk Analytics: The R Companion). Our findings underscore the strengths and weaknesses of each method, providing valuable insights into the most effective ML algorithms for PD prediction in the context of loan portfolios.

In this paper we study the pricing and hedging of nonreplicable contingent claims, such as long-term insurance contracts like variable annuities. Our approach is based on the benchmark-neutral pricing framework of Platen (2024), which differs from the classical benchmark approach by using the stock growth optimal portfolio as the num\'eraire. In typical settings, this choice leads to an equivalent martingale measure, the benchmark-neutral measure. The resulting prices can be significantly lower than the respective risk-neutral ones, making this approach attractive for long-term risk-management. We derive the associated risk-minimizing hedging strategy under the assumption that the contingent claim possesses a martingale decomposition. For a set of nonreplicable contingent claims, these strategies allow monitoring the working capital required to generate their payoffs and enable an assessment of the resulting diversification effects. Furthermore, an algorithmic refinancing strategy is proposed that allows modeling the working capital. Finally, insurance-finance arbitrages of the first kind are introduced and it is demonstrated that benchmark-neutral pricing effectively avoids such arbitrages.

We consider an economy composed of different risk profile regions wishing to be hedged against a disaster risk using multi-region catastrophe insurance. Such catastrophic events inherently have a systemic component; we consider situations where the insurer faces a non-zero probability of insolvency. To protect the regions against the risk of the insurer's default, we introduce a public-private partnership between the government and the insurer. When a disaster generates losses exceeding the total capital of the insurer, the central government intervenes by implementing a taxation system to share the residual claims. In this study, we propose a theoretical framework for regional participation in collective risk-sharing through tax revenues by accounting for their disaster risk profiles and their economic status.

This article demonstrates the transformative impact of Generative AI (GenAI) on actuarial science, illustrated by four implemented case studies. It begins with a historical overview of AI, tracing its evolution from early neural networks to modern GenAI technologies. The first case study shows how Large Language Models (LLMs) improve claims cost prediction by deriving significant features from unstructured textual data, significantly reducing prediction errors in the underlying machine learning task. In the second case study, we explore the automation of market comparisons using the GenAI concept of Retrieval-Augmented Generation to identify and process relevant information from documents. A third case study highlights the capabilities of fine-tuned vision-enabled LLMs in classifying car damage types and extracting contextual information. The fourth case study presents a multi-agent system that autonomously analyzes data from a given dataset and generates a corresponding report detailing the key findings. In addition to these case studies, we outline further potential applications of GenAI in the insurance industry, such as the automation of claims processing and fraud detection, and the verification of document compliance with internal or external policies. Finally, we discuss challenges and considerations associated with the use of GenAI, covering regulatory issues, ethical concerns, and technical limitations, among others.

Conditional Value-at-Risk (CVaR) is a risk measure widely used to quantify the impact of extreme losses. Owing to the lack of representative samples CVaR is sensitive to the tails of the underlying distribution. In order to combat this sensitivity, Distributionally Robust Optimization (DRO), which evaluates the worst-case CVaR measure over a set of plausible data distributions is often deployed. Unfortunately, an improper choice of the DRO formulation can lead to a severe underestimation of tail risk. This paper aims at leveraging extreme value theory to arrive at a DRO formulation which leads to representative worst-case CVaR evaluations in that the above pitfall is avoided while simultaneously, the worst case evaluation is not a gross over-estimate of the true CVaR. We demonstrate theoretically that even when there is paucity of samples in the tail of the distribution, our formulation is readily implementable from data, only requiring calibration of a single scalar parameter. We showcase that our formulation can be easily extended to provide robustness to tail risk in multivariate applications as well as in the evaluation of other commonly used risk measures. Numerical illustrations on synthetic and real-world data showcase the practical utility of our approach.

The rapid expansion of cross-border e-commerce (CBEC) has created significant opportunities for small and medium-sized enterprises (SMEs), yet financing remains a critical challenge due to SMEs' limited credit histories. Third-party logistics (3PL)-led supply chain finance (SCF) has emerged as a promising solution, leveraging in-transit inventory as collateral. We propose an advanced credit risk management framework tailored for 3PL-led SCF, addressing the dual challenges of credit risk assessment and loan size determination. Specifically, we leverage conditional generative modeling of sales distributions through Quantile-Regression-based Generative Metamodeling (QRGMM) as the foundation for risk estimation. We propose a unified framework that enables flexible estimation of multiple risk measures while introducing a functional risk measure formulation that systematically captures the relationship between these risk measures and varying loan levels, supported by theoretical guarantees. To capture complex covariate interactions in e-commerce sales data, we integrate QRGMM with Deep Factorization Machines (DeepFM). Extensive experiments on synthetic and real-world data validate the efficacy of our model for credit risk assessment and loan size determination. This study represents a pioneering application of generative AI in CBEC SCF risk management, offering a solid foundation for enhanced credit practices and improved SME access to capital.

Credit ratings are widely used by investors as a screening device. We introduce and study several natural notions of risk consistency that promote prudent investment decisions in the framework of Choquet rating criteria. Three closely related notions of risk consistency are considered: with respect to risk aversion, the asset pooling effect, and the benefit of portfolio diversification. These notions are formulated either under a single probability measure or multiple probability measures. We show how these properties translate between rating criteria and the corresponding risk measures, and establish a hierarchical structure among them. These findings lead to a full characterization of Choquet risk measures and Choquet rating criteria satisfying risk consistency properties. Illustrated by case studies on collateralized loan obligations and catastrophe bonds, some classes of Choquet rating criteria serve as useful alternatives to the probability of default and expected loss criteria used in practice for rating financial products.

Blockchain-based decentralised lending is a rapidly growing and evolving alternative to traditional lending, but it poses new risks. To mitigate these risks, lending protocols have integrated automated risk management tools into their smart contracts. However, the effectiveness of the latest risk management features introduced in the most recent versions of these lending protocols is understudied. To close this gap, we use a panel regression fixed effects model to empirically analyse the cross-version (v2 and v3) and cross-chain (L1 and L2) performance of the liquidation mechanisms of the two most popular lending protocols, Aave and Compound, during the period Jan 2021 to Dec 2024. Our analysis reveals that liquidation events in v3 of both protocols lead to an increase in total value locked and total revenue, with stronger impact on the L2 blockchain compared to L1. In contrast, liquidations in v2 have an insignificant impact, which indicates that the most recent v3 protocols have better risk management than the earlier v2 protocols. We also show that L1 blockchains are the preferred choice among large investors for their robust liquidity and ecosystem depth, while L2 blockchains are more popular among retail investors for their lower fees and faster execution.

We explore credit risk pricing by modeling equity as a call option and debt as the difference between the firm's asset value and a put option, following the structural framework of the Merton model. Our approach proceeds in two stages: first, we calibrate the asset volatility using the Black-Scholes-Merton (BSM) formula; second, we recover implied mean return and probability surfaces under the physical measure. To achieve this, we construct a recombining binomial tree under the real-world (natural) measure, assuming a fixed initial asset value. The volatility input is taken from a specific region of the implied volatility surface - based on moneyness and maturity - which then informs the calibration of drift and probability. A novel mapping is established between risk-neutral and physical parameters, enabling construction of implied surfaces that reflect the market's credit expectations and offer practical tools for stress testing and credit risk analysis.

We introduce the Circular Directional Flow Decomposition (CDFD), a new framework for analyzing circularity in weighted directed networks. CDFD separates flow into two components: a circular (divergence-free) component and an acyclic component that carries all nett directional flow. This yields a normalized circularity index between 0 (fully acyclic) and 1 (for networks formed solely by the superposition of cycles), with the complement measuring directionality. This index captures the proportion of flow involved in cycles, and admits a range of interpretations - such as system closure, feedback, weighted strong connectivity, structural redundancy, or inefficiency. Although the decomposition is generally non-unique, we show that the set of all decompositions forms a well-structured geometric space with favourable topological properties. Within this space, we highlight two benchmark decompositions aligned with distinct analytical goals: the maximum circularity solution, which minimizes nett flow, and the Balanced Flow Forwarding (BFF) solution, a unique, locally computable decomposition that distributes circular flow across all feasible cycles in proportion to the original network structure. We demonstrate the interpretive value and computational tractability of both decompositions on synthetic and empirical networks. They outperform existing circularity metrics in detecting meaningful structural variation. The decomposition also enables structural analysis - such as mapping the distribution of cyclic flow - and supports practical applications that require explicit flow allocation or routing, including multilateral netting and efficient transport.