CyberSec Research

Global sensitivity analysis is employed to evaluate the effective dimension reduction achieved through Chebyshev interpolation and the conditional pathwise method for Greek estimation of discretely monitored barrier options and arithmetic average Asian options. We compare results from finite difference and Monte Carlo methods with those obtained by using randomized Quasi Monte Carlo combined with Brownian bridge discretization. Additionally, we investigate the benefits of incorporating importance sampling with either the finite difference or Chebyshev interpolation methods. Our findings demonstrate that the reduced effective dimensionality identified through global sensitivity analysis explains the performance advantages of one approach over another. Specifically, the increased smoothness provided by Chebyshev or conditional pathwise methods enhances the convergence rate of randomized Quasi Monte Carlo integration, leading to the significant increase of accuracy and reduced computational costs.

Solving large-scale, continuous-time portfolio optimization problems involving numerous assets and state-dependent dynamics has long been challenged by the curse of dimensionality. Traditional dynamic programming and PDE-based methods, while rigorous, typically become computationally intractable beyond a small number of state variables (often limited to ~3-6 in prior numerical studies). To overcome this critical barrier, we introduce the \emph{Pontryagin-Guided Direct Policy Optimization} (PG-DPO) framework. PG-DPO leverages Pontryagin's Maximum Principle to directly guide neural network policies via backpropagation-through-time, naturally incorporating exogenous state processes without requiring dense state grids. Crucially, our computationally efficient ``Two-Stage'' variant exploits rapidly stabilizing costate estimates derived from BPTT, converting them into near-optimal closed-form Pontryagin controls after only a short warm-up, significantly reducing training overhead. This enables a breakthrough in scalability: numerical experiments demonstrate that PG-DPO successfully tackles problems with dimensions previously considered far out of reach, optimizing portfolios with up to 50 assets and 10 state variables. The framework delivers near-optimal policies, offering a practical and powerful alternative for high-dimensional continuous-time portfolio choice.

This paper presents a realistic simulated stock market where large language models (LLMs) act as heterogeneous competing trading agents. The open-source framework incorporates a persistent order book with market and limit orders, partial fills, dividends, and equilibrium clearing alongside agents with varied strategies, information sets, and endowments. Agents submit standardized decisions using structured outputs and function calls while expressing their reasoning in natural language. Three findings emerge: First, LLMs demonstrate consistent strategy adherence and can function as value investors, momentum traders, or market makers per their instructions. Second, market dynamics exhibit features of real financial markets, including price discovery, bubbles, underreaction, and strategic liquidity provision. Third, the framework enables analysis of LLMs' responses to varying market conditions, similar to partial dependence plots in machine-learning interpretability. The framework allows simulating financial theories without closed-form solutions, creating experimental designs that would be costly with human participants, and establishing how prompts can generate correlated behaviors affecting market stability.

With the introduction of the PSD2 regulation in the EU which established the Open Banking framework, a new window of opportunities has opened for banks and fintechs to explore and enrich Bank transaction descriptions with the aim of building a better understanding of customer behavior, while using this understanding to prevent fraud, reduce risks and offer more competitive and tailored services. And although the usage of natural language processing models and techniques has seen an incredible progress in various applications and domains over the past few years, custom applications based on domain-specific text corpus remain unaddressed especially in the banking sector. In this paper, we introduce a language-based Open Banking transaction classification system with a focus on the french market and french language text. The system encompasses data collection, labeling, preprocessing, modeling, and evaluation stages. Unlike previous studies that focus on general classification approaches, this system is specifically tailored to address the challenges posed by training a language model with a specialized text corpus (Banking data in the French context). By incorporating language-specific techniques and domain knowledge, the proposed system demonstrates enhanced performance and efficiency compared to generic approaches.

This paper examines the empirical failure of uncovered interest parity (UIP) and proposes a structural explanation based on a mean-reverting risk premium. We define a realized premium as the deviation between observed exchange rate returns and the interest rate differential, and demonstrate its strong mean-reverting behavior across multiple horizons. Motivated by this pattern, we model the risk premium using an Ornstein-Uhlenbeck (OU) process embedded within a stochastic differential equation for the exchange rate. Our model yields closed-form approximations for future exchange rate distributions, which we evaluate using coverage-based backtesting. Applied to USD/KRW data from 2010 to 2025, the model shows strong predictive performance at both short-term and long-term horizons, while underperforming at intermediate (3-month) horizons and showing conservative behavior in the tails of long-term forecasts. These results suggest that exchange rate deviations from UIP may reflect structured, forecastable dynamics rather than pure noise, and point to future modeling improvements via regime-switching or time-varying volatility.

Large language model-based agents are becoming increasingly popular as a low-cost mechanism to provide personalized, conversational advice, and have demonstrated impressive capabilities in relatively simple scenarios, such as movie recommendations. But how do these agents perform in complex high-stakes domains, where domain expertise is essential and mistakes carry substantial risk? This paper investigates the effectiveness of LLM-advisors in the finance domain, focusing on three distinct challenges: (1) eliciting user preferences when users themselves may be unsure of their needs, (2) providing personalized guidance for diverse investment preferences, and (3) leveraging advisor personality to build relationships and foster trust. Via a lab-based user study with 64 participants, we show that LLM-advisors often match human advisor performance when eliciting preferences, although they can struggle to resolve conflicting user needs. When providing personalized advice, the LLM was able to positively influence user behavior, but demonstrated clear failure modes. Our results show that accurate preference elicitation is key, otherwise, the LLM-advisor has little impact, or can even direct the investor toward unsuitable assets. More worryingly, users appear insensitive to the quality of advice being given, or worse these can have an inverse relationship. Indeed, users reported a preference for and increased satisfaction as well as emotional trust with LLMs adopting an extroverted persona, even though those agents provided worse advice.

Dynamic hedging is a financial strategy that consists in periodically transacting one or multiple financial assets to offset the risk associated with a correlated liability. Deep Reinforcement Learning (DRL) algorithms have been used to find optimal solutions to dynamic hedging problems by framing them as sequential decision-making problems. However, most previous work assesses the performance of only one or two DRL algorithms, making an objective comparison across algorithms difficult. In this paper, we compare the performance of eight DRL algorithms in the context of dynamic hedging; Monte Carlo Policy Gradient (MCPG), Proximal Policy Optimization (PPO), along with four variants of Deep Q-Learning (DQL) and two variants of Deep Deterministic Policy Gradient (DDPG). Two of these variants represent a novel application to the task of dynamic hedging. In our experiments, we use the Black-Scholes delta hedge as a baseline and simulate the dataset using a GJR-GARCH(1,1) model. Results show that MCPG, followed by PPO, obtain the best performance in terms of the root semi-quadratic penalty. Moreover, MCPG is the only algorithm to outperform the Black-Scholes delta hedge baseline with the allotted computational budget, possibly due to the sparsity of rewards in our environment.

This note gives a bound on the error of the leading term of the $t\to 0$ asymptotic expansion of the Hartman-Watson distribution $\theta(r,t)$ in the regime $rt=\rho$ constant. The leading order term has the form $\theta(\rho/t,t)=\frac{1}{2\pi t}e^{-\frac{1}{t} (F(\rho)-\pi^2/2)} G(\rho) (1 + \vartheta(t,\rho))$, where the error term is bounded uniformly over $\rho$ as $|\vartheta(t,\rho)|\leq \frac{1}{70}t$.

This study investigates enhancing option pricing by extending the Black-Scholes model to include stochastic volatility and interest rate variability within the Partial Differential Equation (PDE). The PDE is solved using the finite difference method. The extended Black-Scholes model and a machine learning-based LSTM model are developed and evaluated for pricing Google stock options. Both models were backtested using historical market data. While the LSTM model exhibited higher predictive accuracy, the finite difference method demonstrated superior computational efficiency. This work provides insights into model performance under varying market conditions and emphasizes the potential of hybrid approaches for robust financial modeling.

In this paper, we investigate the Markovian iteration method for solving coupled forward-backward stochastic differential equations (FBSDEs) featuring a fully coupled forward drift, meaning the drift term explicitly depends on both the forward and backward processes. An FBSDE system typically involves three stochastic processes: the forward process $X$, the backward process $Y$ representing the solution, and the $Z$ process corresponding to the scaled derivative of $Y$. Prior research by Bender and Zhang (2008) has established convergence results for iterative schemes dealing with $Y$-coupled FBSDEs. However, extending these results to equations with $Z$ coupling poses significant challenges, especially in uniformly controlling the Lipschitz constant of the decoupling fields across iterations and time steps within a fixed-point framework. To overcome this issue, we propose a novel differentiation-based method for handling the $Z$ process. This approach enables improved management of the Lipschitz continuity of decoupling fields, facilitating the well-posedness of the discretized FBSDE system with fully coupled drift. We rigorously prove the convergence of our Markovian iteration method in this more complex setting. Finally, numerical experiments confirm our theoretical insights, showcasing the effectiveness and accuracy of the proposed methodology.

This research proposes a cutting-edge ensemble deep learning framework for stock price prediction by combining three advanced neural network architectures: The particular areas of interest for the research include but are not limited to: Variational Autoencoder (VAE), Transformer, and Long Short-Term Memory (LSTM) networks. The presented framework is aimed to substantially utilize the advantages of each model which would allow for achieving the identification of both linear and non-linear relations in stock price movements. To improve the accuracy of its predictions it uses rich set of technical indicators and it scales its predictors based on the current market situation. By trying out the framework on several stock data sets, and benchmarking the results against single models and conventional forecasting, the ensemble method exhibits consistently high accuracy and reliability. The VAE is able to learn linear representation on high-dimensional data while the Transformer outstandingly perform in recognizing long-term patterns on the stock price data. LSTM, based on its characteristics of being a model that can deal with sequences, brings additional improvements to the given framework, especially regarding temporal dynamics and fluctuations. Combined, these components provide exceptional directional performance and a very small disparity in the predicted results. The present solution has given a probable concept that can handle the inherent problem of stock price prediction with high reliability and scalability. Compared to the performance of individual proposals based on the neural network, as well as classical methods, the proposed ensemble framework demonstrates the advantages of combining different architectures. It has a very important application in algorithmic trading, risk analysis, and control and decision-making for finance professions and scholars.

Quantitative investment (quant) is an emerging, technology-driven approach in asset management, increasingy shaped by advancements in artificial intelligence. Recent advances in deep learning and large language models (LLMs) for quant finance have improved predictive modeling and enabled agent-based automation, suggesting a potential paradigm shift in this field. In this survey, taking alpha strategy as a representative example, we explore how AI contributes to the quantitative investment pipeline. We first examine the early stage of quant research, centered on human-crafted features and traditional statistical models with an established alpha pipeline. We then discuss the rise of deep learning, which enabled scalable modeling across the entire pipeline from data processing to order execution. Building on this, we highlight the emerging role of LLMs in extending AI beyond prediction, empowering autonomous agents to process unstructured data, generate alphas, and support self-iterative workflows.

This study explores the integration of a representative large language model, ChatGPT, into lending decision-making with a focus on credit default prediction. Specifically, we use ChatGPT to analyse and interpret loan assessments written by loan officers and generate refined versions of these texts. Our comparative analysis reveals significant differences between generative artificial intelligence (AI)-refined and human-written texts in terms of text length, semantic similarity, and linguistic representations. Using deep learning techniques, we show that incorporating unstructured text data, particularly ChatGPT-refined texts, alongside conventional structured data significantly enhances credit default predictions. Furthermore, we demonstrate how the contents of both human-written and ChatGPT-refined assessments contribute to the models' prediction and show that the effect of essential words is highly context-dependent. Moreover, we find that ChatGPT's analysis of borrower delinquency contributes the most to improving predictive accuracy. We also evaluate the business impact of the models based on human-written and ChatGPT-refined texts, and find that, in most cases, the latter yields higher profitability than the former. This study provides valuable insights into the transformative potential of generative AI in financial services.

Market simulator tries to create high-quality synthetic financial data that mimics real-world market dynamics, which is crucial for model development and robust assessment. Despite continuous advancements in simulation methodologies, market fluctuations vary in terms of scale and sources, but existing frameworks often excel in only specific tasks. To address this challenge, we propose Financial Wind Tunnel (FWT), a retrieval-augmented market simulator designed to generate controllable, reasonable, and adaptable market dynamics for model testing. FWT offers a more comprehensive and systematic generative capability across different data frequencies. By leveraging a retrieval method to discover cross-sectional information as the augmented condition, our diffusion-based simulator seamlessly integrates both macro- and micro-level market patterns. Furthermore, our framework allows the simulation to be controlled with wide applicability, including causal generation through "what-if" prompts or unprecedented cross-market trend synthesis. Additionally, we develop an automated optimizer for downstream quantitative models, using stress testing of simulated scenarios via FWT to enhance returns while controlling risks. Experimental results demonstrate that our approach enables the generalizable and reliable market simulation, significantly improve the performance and adaptability of downstream models, particularly in highly complex and volatile market conditions. Our code and data sample is available at https://anonymous.4open.science/r/fwt_-E852

We study the existence of equilibrium when agents' preferences may not beconvex. For some specific utility functions, we provide a necessary and sufficientcondition under which there exists an equilibrium. The standard approach cannot be directly applied to our examples because the demand correspondence of some agents is neither single-valued nor convex-valued.

To quantify the changes in the credit rating of a bond is an important mathematical problem for the credit rating industry. To think of the credit rating as the state a Markov chain is an interesting proposal leading to challenges in mathematical modeling. Since cumulative default rates are more readily measurable than credit migrations, a natural question is whether the credit transition matrix (CTM) can be determined from the knowledge of the cumulative default probabilities. Here we use a connection between the CTM and the cumulative default probabilities to setup an ill-posed, linear inverse problem with box constraints, which we solve by an entropy minimization procedure. This approach is interesting on several counts. On the one hand, we may have less data that unknowns, and on the other hand, even when we have as much data as unknowns, the matrix connecting them may not be invertible, which makes the problem ill-posed. Besides developing the tools to solve the problem, we apply it to several test cases to check the performance of the method. The results are quite satisfactory.

Graph Neural Networks have significantly advanced research in recommender systems over the past few years. These methods typically capture global interests using aggregated past interactions and rely on static embeddings of users and items over extended periods of time. While effective in some domains, these methods fall short in many real-world scenarios, especially in finance, where user interests and item popularity evolve rapidly over time. To address these challenges, we introduce a novel extension to Light Graph Convolutional Network (LightGCN) designed to learn temporal node embeddings that capture dynamic interests. Our approach employs causal convolution to maintain a forward-looking model architecture. By preserving the chronological order of user-item interactions and introducing a dynamic update mechanism for embeddings through a sliding window, the proposed model generates well-timed and contextually relevant recommendations. Extensive experiments on a real-world dataset from BNP Paribas demonstrate that our approach significantly enhances the performance of LightGCN while maintaining the simplicity and efficiency of its architecture. Our findings provide new insights into designing graph-based recommender systems in time-sensitive applications, particularly for financial product recommendations.

We introduce the deep multi-FBSDE method for robust approximation of coupled forward-backward stochastic differential equations (FBSDEs), focusing on cases where the deep BSDE method of Han, Jentzen, and E (2018) fails to converge. To overcome the convergence issues, we consider a family of FBSDEs that are equivalent to the original problem in the sense that they satisfy the same associated partial differential equation (PDE). Our algorithm proceeds in two phases: first, we approximate the initial condition for the FBSDE family, and second, we approximate the original FBSDE using the initial condition approximated in the first phase. Numerical experiments show that our method converges even when the standard deep BSDE method does not.