https://arxiv.org/api/ClWMSkqS85htfnfODW6A13JjaXc 2026-03-22T12:03:27Z 3124 135 15 http://arxiv.org/abs/2601.01642v1 2026-01-04T19:15:22Z

Standard rare-event simulation techniques require exact distributional specifications, which limits their effectiveness in the presence of distributional uncertainty. To address this, we develop a novel framework for estimating rare-event probabilities subject to such distributional model risk. Specifically, we focus on computing worst-case rare-event probabilities, defined as a distributionally robust bound against a Wasserstein ambiguity set centered at a specific nominal distribution. By exploiting a dual characterization of this bound, we propose Distributionally Robust Importance Sampling (DRIS), a computationally tractable methodology designed to substantially reduce the variance associated with estimating the dual components. The proposed method is simple to implement and requires low sampling costs. Most importantly, it achieves vanishing relative error, the strongest efficiency guarantee that is notoriously difficult to establish in rare-event simulation. Our numerical studies confirm the superior performance of DRIS against existing benchmarks.

2026-01-04T19:15:22Z Dohyun Ahn Huiyi Chen Lewen Zheng http://arxiv.org/abs/2601.00738v1 2026-01-02T16:39:25Z

This paper examines the impact of reducing Ethereum slot time on decentralized exchange activity, with a focus on CEX-DEX arbitrage behavior. We develop a trading model where the agent's DEX transaction is not guaranteed to land, and the agent explicitly accounts for this execution risk when deciding whether to pursue arbitrage opportunities. We compare agent behavior under Ethereum's default 12-second slot time environment with a faster regime that offers 1-second subslot execution. The simulations, calibrated to Binance and Uniswap v3 data from July to September 2025, show that faster slot times increase arbitrage transaction count by 535% and trading volume by 203% on average. The increase in CEX-DEX arbitrage activity under 1-second subslots is driven by the reduction in variance of both successful and failed trade outcomes, increasing the risk-adjusted returns and making CEX-DEX arbitrage more appealing.

2026-01-02T16:39:25Z Aleksei Adadurov Sergey Barseghyan Anton Chtepine Antero Eloranta Andrei Sebyakin Arsenii Valitov http://arxiv.org/abs/2601.00478v1 2026-01-01T21:16:28Z

Credit risk assessment increasingly relies on diverse sources of information beyond traditional structured financial data, particularly for micro and small enterprises (mSEs) with limited financial histories. This study proposes a multimodal framework that integrates structured credit variables, climate panel data, and unstructured textual narratives within a unified learning architecture. Specifically, we use long short-term memory (LSTM), the gated recurrent unit (GRU), and transformer models to analyse the interplay between these data modalities. The empirical results demonstrate that unimodal models based on climate or text data outperform those relying solely on structured data, while the integration of multiple data modalities yields significant improvements in credit default prediction. Using SHAP-based explainability methods, we find that physical climate risks play an important role in default prediction, with water-logging by rain emerging as the most influential factor. Overall, this study demonstrates the potential of multimodal approaches in AI-enabled decision-making, which provides robust tools for credit risk assessment while contributing to the broader integration of environmental and textual insights into predictive analytics.

2026-01-01T21:16:28Z Zongxiao Wu Ran Liu Jiang Dai Dan Luo http://arxiv.org/abs/2512.24714v1 2025-12-31T08:29:33Z

We first review the convolution fast-Fourier-transform (CFFT) approach for the numerical solution of backward stochastic differential equations (BSDEs) introduced in (Hyndman and Oyono Ngou, 2017). We then propose a method for improving the boundary errors obtained when valuing options using this approach. We modify the damping and shifting schemes used in the original formulation, which transforms the target function into a bounded periodic function so that Fourier transforms can be applied successfully. Time-dependent shifting reduces boundary error significantly. We present numerical results for our implementation and provide a detailed error analysis showing the improved accuracy and convergence of the modified convolution method.

2025-12-31T08:29:33Z 15 pages, 3 figures, 1 table Xiang Gao Cody Hyndman http://arxiv.org/abs/2512.24621v1 2025-12-31T04:30:05Z

We study short-horizon forecasting in financial time series under strict causal constraints, treating the market as a non-stationary stochastic system in which any predictive observable must be computable online from information available up to the decision time. Rather than proposing a machine-learning predictor or a direct price-forecast model, we focus on \emph{constructing} an interpretable causal signal from heterogeneous micro-features that encode complementary aspects of the dynamics (momentum, volume pressure, trend acceleration, and volatility-normalized price location). The construction combines (i) causal centering, (ii) linear aggregation into a composite observable, (iii) causal stabilization via a one-dimensional Kalman filter, and (iv) an adaptive ``forward-like'' operator that mixes the composite signal with a smoothed causal derivative term. The resulting observable is mapped into a transparent decision functional and evaluated through realized cumulative returns and turnover. An application to high-frequency EURUSDT (1-minute) illustrates that causally constructed observables can exhibit substantial economic relevance in specific regimes, while degrading under subsequent regime shifts, highlighting both the potential and the limitations of causal signal design in non-stationary markets.

2025-12-31T04:30:05Z Lucas A. Souza http://arxiv.org/abs/2512.24526v1 2025-12-31T00:19:41Z

This paper investigates how Large Language Models (LLMs) from leading providers (OpenAI, Google, Anthropic, DeepSeek, and xAI) can be applied to quantitative sector-based portfolio construction. We use LLMs to identify investable universes of stocks within S&P 500 sector indices and evaluate how their selections perform when combined with classical portfolio optimization methods. Each model was prompted to select and weight 20 stocks per sector, and the resulting portfolios were compared with their respective sector indices across two distinct out-of-sample periods: a stable market phase (January-March 2025) and a volatile phase (April-June 2025). Our results reveal a strong temporal dependence in LLM portfolio performance. During stable market conditions, LLM-weighted portfolios frequently outperformed sector indices on both cumulative return and risk-adjusted (Sharpe ratio) measures. However, during the volatile period, many LLM portfolios underperformed, suggesting that current models may struggle to adapt to regime shifts or high-volatility environments underrepresented in their training data. Importantly, when LLM-based stock selection is combined with traditional optimization techniques, portfolio outcomes improve in both performance and consistency. This study contributes one of the first multi-model, cross-provider evaluations of generative AI algorithms in investment management. It highlights that while LLMs can effectively complement quantitative finance by enhancing stock selection and interpretability, their reliability remains market-dependent. The findings underscore the potential of hybrid AI-quantitative frameworks, integrating LLM reasoning with established optimization techniques, to produce more robust and adaptive investment strategies.

2025-12-31T00:19:41Z Alina Voronina Oleksandr Romanko Ruiwen Cao Roy H. Kwon Rafael Mendoza-Arriaga http://arxiv.org/abs/2510.19130v2 2025-12-26T23:43:47Z

Covariance matrices estimated from short, noisy, and non-Gaussian financial time series are notoriously unstable. Empirical evidence suggests that such covariance structures often exhibit power-law scaling, reflecting complex, hierarchical interactions among assets. Motivated by this observation, we introduce a power-law covariance model to characterize collective market dynamics and propose a hybrid estimator that integrates Random Matrix Theory (RMT) with deep Residual Neural Networks (ResNets). The RMT component regularizes the eigenvalue spectrum in high-dimensional noisy settings, while the ResNet learns data-driven corrections that recover latent structural dependencies encoded in the eigenvectors. Monte Carlo simulations show that the proposed ResNet-based estimators consistently minimize both Frobenius and minimum-variance losses across a range of population covariance models. Empirical experiments on 89 cryptocurrencies over the period 2020-2025, using a training window ending at the local Bitcoin peak in November 2021 and testing through the subsequent bear market, demonstrate that a two-step estimator combining hierarchical filtering with ResNet corrections produces the most profitable and well-balanced portfolios, remaining robust across market regime shifts. Beyond finance, the proposed hybrid framework applies broadly to high-dimensional systems described by low-rank deformations of Wishart ensembles, where incorporating eigenvector information enables the detection of multiscale and hierarchical structure that is inaccessible to purely eigenvalue-based methods.

2025-10-21T23:33:27Z Andres Garcia-Medina 10.1142/S0129183127500458 http://arxiv.org/abs/2512.22109v1 2025-12-26T18:46:06Z

We study the construction and rebalancing of sparse index-tracking portfolios from an operational research perspective, with explicit emphasis on uncertainty quantification and implementability. The decision variables are portfolio weights constrained to sum to one; the aims are to track a reference index closely while controlling the number of names and the turnover induced by rebalancing. We cast index tracking as a high-dimensional linear regression of index returns on constituent returns, and employ a sparsity-inducing Laplace prior on the weights. A single global shrinkage parameter controls the trade-off between tracking error and sparsity, and is calibrated by an empirical-Bayes stochastic approximation scheme. Conditional on this calibration, we approximate the posterior distribution of the portfolio weights using proximal Langevin-type Markov chain Monte Carlo algorithms tailored to the budget constraint. This yields posterior uncertainty on tracking error, portfolio composition and prospective rebalancing moves. Building on these posterior samples, we propose rules for rebalancing that gate trades through magnitude-based thresholds and posterior activation probabilities, thereby trading off expected tracking error against turnover and portfolio size. A case study on tracking the S&P~500 index is carried out to showcase how our tools shape the decision process from portfolio construction to rebalancing.

2025-12-26T18:46:06Z Dimitrios Roxanas http://arxiv.org/abs/2512.22001v1 2025-12-26T11:59:30Z

We present a scalable, hardware-aware methodology for extending the Variational Quantum Eigensolver (VQE) to large, realistic Dynamic Portfolio Optimization (DPO) problems. Building on the scaling strategy from our previous work, where we tailored a VQE workflow to both the DPO formulation and the target QPU, we now put forward two significant advances. The first is the implementation of the Ising Sample-based Quantum Configuration Recovery (ISQR) routine, which improves solution quality in Quadratic Unconstrained Binary Optimization problems. The second is the use of the VQE Constrained method to decompose the optimization task, enabling us to handle DPO instances with more variables than the available qubits on current hardware. These advances, which are broadly applicable to other optimization problems, allow us to address a portfolio with a size relevant to the financial industry, consisting of up to 38 assets and covering the full Spanish stock index (IBEX 35). Our results, obtained on a real Quantum Processing Unit (IBM Fez), show that this tailored workflow achieves financial performance on par with classical methods while delivering a broader set of high-quality investment strategies, demonstrating a viable path towards obtaining practical advantage from quantum optimization in real financial applications.

2025-12-26T11:59:30Z Irene De León Danel Arias Manuel Martín-Cordero María Esperanza Molina Pablo Serrano Senaida Hernández-Santana Miguel Ángel Jiménez Herrera Joana Fraxanet Ginés Carrascal Escolástico Sánchez Inmaculada Posadillo Álvaro Nodar http://arxiv.org/abs/2512.21791v1 2025-12-25T21:43:16Z

Data scarcity and confidentiality in finance often impede model development and robust testing. This paper presents a unified multi-criteria evaluation framework for synthetic financial data and applies it to three representative generative paradigms: the statistical ARIMA-GARCH baseline, Variational Autoencoders (VAEs), and Time-series Generative Adversarial Networks (TimeGAN). Using historical S and P 500 daily data, we evaluate fidelity (Maximum Mean Discrepancy, MMD), temporal structure (autocorrelation and volatility clustering), and practical utility in downstream tasks, specifically mean-variance portfolio optimization and volatility forecasting. Empirical results indicate that ARIMA-GARCH captures linear trends and conditional volatility but fails to reproduce nonlinear dynamics; VAEs produce smooth trajectories that underestimate extreme events; and TimeGAN achieves the best trade-off between realism and temporal coherence (e.g., TimeGAN attained the lowest MMD: 1.84e-3, average over 5 seeds). Finally, we articulate practical guidelines for selecting generative models according to application needs and computational constraints. Our unified evaluation protocol and reproducible codebase aim to standardize benchmarking in synthetic financial data research.

2025-12-25T21:43:16Z 23 pages, 7 figures, 6 tables. Submitted as a preprint. This work presents a unified multi-criteria evaluation framework for synthetic financial data, applied to ARIMA-GARCH, VAEs, and TimeGAN models Christophe D. Hounwanou Yae Ulrich Gaba Pierre Ntakirutimana http://arxiv.org/abs/2512.16115v2 2025-12-25T14:20:44Z

The increasing need for rapid recalibration of option pricing models in dynamic markets places stringent computational demands on data generation and valuation algorithms. In this work, we propose a hybrid algorithmic framework that integrates the smooth offset algorithm (SOA) with supervised machine learning models for the fast pricing of multiple path-independent options under exponential Lévy dynamics. Building upon the SOA-generated dataset, we train neural networks, random forests, and gradient boosted decision trees to construct surrogate pricing operators. Extensive numerical experiments demonstrate that, once trained, these surrogates achieve order-of-magnitude acceleration over direct SOA evaluation. Importantly, the proposed framework overcomes key numerical limitations inherent to fast Fourier transform-based methods, including the consistency of input data and the instability in deep out-of-the-money option pricing.

2025-12-18T03:08:56Z Liying Zhang Ying Gao http://arxiv.org/abs/2512.21539v1 2025-12-25T07:15:44Z

It was recently established that the formalism of the generalized transfer operator (GTO) of dynamical systems (DS) theory, applied to stochastic differential equations (SDEs) of arbitrary form, belongs to the family of cohomological topological field theories (TFT) -- a class of models at the intersection of algebraic topology and high-energy physics. This interdisciplinary approach, which can be called the supersymmetric theory of stochastic dynamics (STS), can be seen as an algebraic dual to the traditional set-theoretic framework of the DS theory, with its algebraic structure enabling the extension of some DS theory concepts to stochastic dynamics. Moreover, it reveals the presence of a topological supersymmetry (TS) in the GTOs of all SDEs. It also shows that among the various definitions of chaos, positive "pressure", defined as the logarithm of the GTO spectral radius, stands out as particularly meaningful from a physical perspective, as it corresponds to the spontaneous breakdown of TS on the TFT side. Via the Goldstone theorem, this definition has a potential to provide the long-sought explanation for the experimental signature of chaotic dynamics known as 1/f noise. Additionally, STS clarifies that among the various existing interpretations of SDEs, only the Stratonovich interpretation yields evolution operators that match the corresponding GTOs and, consequently, have a clear-cut mathematical meaning. Here, we discuss these and other aspects of STS from both the DS theory and TFT perspectives, focusing on links between these two fields and providing mathematical concepts with physical interpretations that may be useful in some contexts.

2025-12-25T07:15:44Z Phys. Scr. 100 125233 (2025) Igor V. Ovchinnikov 10.1088/1402-4896/ae2296 http://arxiv.org/abs/2511.12391v3 2025-12-24T13:37:04Z

Risk allocation, the decomposition of a portfolio-wide risk measure into component contributions, is a fundamental problem in financial risk management due to the non-additive nature of risk measures, the layered organizational structures of financial institutions, and the range of possible allocation strategies characterized by different rationales and properties. In this work, we conduct a systematic review of the major risk allocation strategies typically used in finance, comparing their theoretical properties, practical advantages, and limitations. To this scope we set up a specific testing framework, including both simplified settings, designed to highlight basic intrinsic behaviours, and realistic financial portfolios under different risk regulations, i.e. Basel 2.5 and FRTB. Furthermore, we develop and test novel practical solutions to manage the issue of negative risk allocations and of multi-level risk allocation in the layered organizational structure of financial institutions, while preserving the additivity property. Finally, we devote particular attention to the computational aspects of risk allocation. Our results show that, in this context, the Shapley allocation strategy offers the best compromise between simplicity, mathematical properties, risk representation and computational cost. The latter is still acceptable even in the challenging case of many business units, provided that an efficient Monte Carlo simulation is employed, which offers excellent scaling and convergence properties. While our empirical applications focus on market risk, our methodological framework is fully general and applicable to other financial context such as valuation risk, liquidity risk, credit risk, and counterparty credit risk.

2025-11-15T23:52:47Z 38 pages (main) + 12 pages (appendixes), 16 figures, 9 tables, 36 references. Fixed some typos Marco Scaringi Marco Bianchetti http://arxiv.org/abs/2512.20515v1 2025-12-23T17:03:04Z

The growing economic influence of the BRICS nations requires risk models that capture complex, long-term dynamics. This paper introduces the Bank Risk Interlinkage with Dynamic Graph and Event Simulations (BRIDGES) framework, which analyzes systemic risk based on the level of information complexity (zero-order, first-order, and second-order). BRIDGES utilizes the Dynamic Time Warping (DTW) distance to construct a dynamic network for 551 BRICS banks based on their strategic similarity, using zero-order information such as annual balance sheet data from 2008 to 2024. It then employs first-order information, including trends in risk ratios, to detect shifts in banks' behavior. A Temporal Graph Neural Network (TGNN), as the core of BRIDGES, is deployed to learn network evolutions and detect second-order information, such as anomalous changes in the structural relationships of the bank network. To measure the impact of anomalous changes on network stability, BRIDGES performs Agent-Based Model (ABM) simulations to assess the banking system's resilience to internal financial failure and external geopolitical shocks at the individual country level and across BRICS nations. Simulation results show that the failure of the largest institutions causes more systemic damage than the failure of the financially vulnerable or dynamically anomalous ones, driven by powerful panic effects. Compared to this "too big to fail" scenario, a geopolitical shock with correlated country-wide propagation causes more destructive systemic damage, leading to a near-total systemic collapse. It suggests that the primary threats to BRICS financial stability are second-order panic and large-scale geopolitical shocks, which traditional risk analysis models might not detect.

2025-12-23T17:03:04Z 22 pages and 7 figures Haibo Wang http://arxiv.org/abs/2512.20216v1 2025-12-23T10:16:00Z

This research introduces a novel quantitative methodology tailored for quantitative finance applications, enabling banks, stockbrokers, and investors to predict economic regimes and market signals in emerging markets, specifically Sri Lankan stock indices (S&P SL20 and ASPI) by integrating Environmental, Social, and Governance (ESG) sentiment analysis with macroeconomic indicators and advanced time-series forecasting. Designed to leverage quantitative techniques for enhanced risk assessment, portfolio optimization, and trading strategies in volatile environments, the architecture employs FinBERT, a transformer-based NLP model, to extract sentiment from ESG texts, followed by unsupervised clustering (UMAP/HDBSCAN) to identify 5 latent ESG regimes, validated via PCA. These regimes are mapped to economic conditions using a dense neural network and gradient boosting classifier, achieving 84.04% training and 82.0% validation accuracy. Concurrently, time-series models (SRNN, MLP, LSTM, GRU) forecast daily closing prices, with GRU attaining an R-squared of 0.801 and LSTM delivering 52.78% directional accuracy on intraday data. A strong correlation between S&P SL20 and S&P 500, observed through moving average and volatility trend plots, further bolsters forecasting precision. A rule-based fusion logic merges ESG and time-series outputs for final market signals. By addressing literature gaps that overlook emerging markets and holistic integration, this quant-driven framework combines global correlations and local sentiment analysis to offer scalable, accurate tools for quantitative finance professionals navigating complex markets like Sri Lanka.

2025-12-23T10:16:00Z 08 Pages, 9 Figures, Published in the Proceedings of the 32'nd IET Sri Lanka Annual Technical Conference 2025 of the IET Sri Lanka Network; Second Runner Up Paper of the Conference Proc. 32nd IET Sri Lanka Annual Technical Conference (IETSL ATC 2025), Colombo, Sri Lanka, 2025, pp. 70-77. ISSN 2345-9564 Linuk Perera