Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,742 papers · 148 categories

Trend · papers per month

4068121,2171,623 · Jun 202019922001200920172026
← all fields·60 papers on deep learning in Quant Finance · 1 year

AlphaZeroBeta uses deep reinforcement learning for market-neutral portfolios, outperforming traditional methods.

problem Traditional portfolio management methods often fail during market regime shifts or when assumptions break down.
method Combines a composite reward function and CNN-GRU policy trained end-to-end via Recurrent PPO.
result Achieves higher Sharpe ratios than baselines while maintaining near-zero benchmark correlations.

CEFOL uses deep learning for dynamic programming with recursive utility.

problem Challenges in solving dynamic programming problems with recursive utility.
method Introduces a separate neural network for certainty equivalent, uses first-order optimality conditions to learn value and policy functions.
result CEFOL achieves high accuracy in learning value and policy functions, matching VFI benchmarks.

Proposes a method to learn adaptive ambiguity sets for robust optimization.

problem Misspecification in distributionally robust optimization (DRO).
method Learned predictive ambiguity sets (LPAS) using deep contextual models.
result Significantly improves portfolio optimization performance compared to baselines.

Deep learning solves dynamic programming with recursive utility.

problem Challenges in solving high-dimensional discrete-time dynamic programming problems with recursive utility.
method Certainty Equivalent Learning (CEL) algorithm that learns certainty-equivalent value directly with neural networks.
result Accurate value and policy approximations in high-dimensional problems, comparable to VFI in some cases.

Deep forecasting models show output heads significantly improve performance on fat-tailed financial returns.

problem Improving deep learning models for forecasting fat-tailed financial returns.
method Comparison of backbone architectures and output heads (point, Gaussian, Gaussian mixture) on S&P 500 monthly log-returns.
result Switching from point to Gaussian heads improves CRPS by about 1.3 percent, and from Gaussian to mixture adds another 2.4 percent.

Enhances cryptocurrency pair trading with DRL, outperforming classical methods.

problem Rigidity and divergence risks in traditional pair trading strategies in crypto markets.
method Hierarchical pair selection, Fixed Risk, Adaptive Mean execution model, PPO with LSTM.
result DRL outperformed heuristic baseline by a statistically significant margin.

FinStressTS creates synthetic benchmarks for financial forecasting, revealing model weaknesses.

problem Limited failure attribution in real-world financial benchmarks.
method Synthetic benchmark with 30 diagnostic environments linked to six mechanism families.
result Model performance varies by mechanism type, with autoregressive models often outperforming Transformers.

HANET combines LSTM and attention mechanisms for better financial forecasting.

problem Lack of distinct macroeconomic regimes in financial datasets.
method Hierarchical Cross-Attention mechanism integrating long-run macro contexts with high-frequency market dynamics.
result HANET outperforms neural forecasters, especially during turbulent periods.

Study forecasts U.S. bond index using deep learning, finding persistence is key.

problem Forecasting U.S. aggregate bond index with deep learning methods.
method Constructed a stationary but maximally persistent representation of the bond index, evaluated using MLPs and CNNs.
result Deep learning models outperform traditional methods in short-horizon forecasting of bond indices.

Paper presents deep LSMC method for efficient variable annuity pricing.

problem Efficiently pricing variable annuities with guarantees using simulation methods.
method Modifies least-squares Monte Carlo (LSMC) algorithm for optimal stochastic control problems.
result Deep LSMC provides more stable and robust pricing performance for higher-dimensional problems.

Deep learning models reconstruct volatility surfaces from noisy data under no-arbitrage constraints.

problem Reconstructing implied volatility surfaces from sparse and noisy option quotes.
method Compared multiple neural architectures including Transformers, U-Nets, and variational autoencoders.
result Transformer and U-Net architectures achieve strong reconstruction accuracy, especially under sparse observation regimes.

End-to-end framework optimizes financial metrics using neural networks.

problem Difficult portfolio optimization in financial markets due to non-stationarity and high costs.
method Directly optimizes differentiable financial metrics via neural networks, incorporating realistic costs and rebalancing.
result Best model achieves +7.86% total return, outperforming S&P 500 by 12.38 percentage points.

Paper uses VAEs to model yield curves without arbitrage violations.

problem Forecasting yield curves across diverse macroeconomic regimes leads to arbitrage violations.
method Proposes a two-stage architecture with CVAEsT+LS and Neural SDEs penalized by No-Arbitrage PDE.
result Significantly reduces forecasting errors and overcomes HJM model limitations.

Deep learning models price convertible bonds with complex reset and call features.

problem Pricing convertible bonds with path-dependent reset and call provisions.
method Formulated as a PPDE, deep learning approximates conditional expectations.
result Deep learning produces stable and accurate prices across various model specifications.

This paper uses deep learning to price American options under stochastic volatility.

problem Pricing American options with a time-varying exercise boundary under the Heston model.
method Coupled PINNs with curriculum learning and adaptive resampling.
result Demonstrates the effectiveness of the proposed deep learning framework for American option pricing.

Enhanced volatility forecasting using options data and rough volatility model.

problem Improving realized volatility forecasting accuracy.
method Infer spot volatility from options data using rough stochastic volatility model, accelerate estimation with deep learning, benchmark against traditional models.
result Augmented HAR-RV-RHeston model outperforms traditional models in daily and long-term forecasting.

The paper values variable annuities using complex stochastic models and deep learning.

problem Valuation of variable annuities with early surrender options under non-Markovian models.
method Developed a deep signature Least Squares Monte Carlo approach to handle path-dependent continuation values.
result Fair fees increase with Hurst parameters of stock volatility and mortality force.

Paper proposes deep learning model for dynamic stock repurchase forecasting.

problem Complex temporal dependencies in corporate financial conditions.
method Hybrid Temporal Convolutional Network (TCN) and Attention-based LSTM.
result Model significantly outperforms static baselines in stock repurchase forecasting.

ARTEMIS combines deep learning and symbolic reasoning for financial predictions.

problem Lack of interpretability and economic principles in deep learning models in finance.
method Neuro-symbolic framework combining neural operators, stochastic differential equations, and symbolic distillation.
result ARTEMIS achieves state-of-the-art directional accuracy, outperforming all baselines on synthetic crash regime.

Deep neural networks improve portfolio construction by jointly modeling returns and risks.

problem Traditional portfolio construction methods fail under time-varying market conditions.
method Jointly modeling dynamic expected returns and risk structures using deep neural networks.
result Deep forecasting model achieves competitive predictive accuracy and economically meaningful directional accuracy.

Benchmarking deep learning models for financial time series, focusing on risk-adjusted performance.

problem Optimizing risk-adjusted performance in financial time series prediction.
method Evaluation of various deep learning architectures including linear models, RNNs, transformers, state space models, and sequence representation approaches.
result Hybrid models like VSN with LSTM and xLSTM achieve the highest overall Sharpe ratio and superior downside adjusted characteristics.

Study compares nine deep learning architectures for multi-horizon financial forecasting.

problem Evaluating the performance of deep learning architectures for multi-horizon financial forecasting.
method Conducted 918 experiments across cryptocurrency, forex, and equity markets using nine architectures.
result ModernTCN achieves the best mean rank (1.333) with a 75 percent first-place rate.

This review analyzes deep learning methods for electricity price forecasting across different markets.

problem Insufficient analysis of deep learning methods in electricity price forecasting.
method Unified taxonomy of deep learning components, analysis of trends across markets.
result Shift toward probabilistic, microstructure-centric, and market-aware designs.

Deep model improves option pricing for CSI 300 index with sentiment and volatility features.

problem Challenges in real market option pricing, especially with constant volatility assumption.
method Deep Forward-Backward Stochastic Differential Equation (FBSDE) framework with dual-network architecture.
result Significant reduction in MAE and MAPE compared to BSM model.

Deep learning improves portfolio optimization in volatile markets.

problem Challenges in long-only, multi-asset strategies across market cycles.
method Training DL models with limited regime data using pre-training techniques and transformer architectures.
result Models show resilience and improved predictive accuracy in volatile markets.

Simple feature engineering beats complex models in financial prediction.

problem Understanding when complex models outperform simple alternatives in financial prediction.
method Independent Component Analysis (ICA), Wavelet Coherence, Long Short-Term Memory (LSTM) networks with attention mechanisms.
result A simple linear model using normalized flows achieves superior returns compared to complex models.

DeePM is a deep-learning portfolio manager that outperforms classical strategies in diversified futures markets.

problem Maximizing risk-adjusted returns in financial markets with low signal-to-noise ratios and asynchronous data.
method Structured deep learning with a Directed Delay mechanism, Macroeconomic Graph Prior, and distributionally robust optimization.
result DeePM achieves net risk-adjusted returns roughly twice those of classical strategies and passive benchmarks.

CB-APM uses analyst consensus as a bottleneck to interpret stock returns.

problem Tackles the challenge of understanding and predicting stock returns using professional beliefs.
method Embeds analyst consensus as a structural bottleneck, treating it as a sufficient statistic for market information.
result CB-APM portfolios exhibit strong monotonic return gradients and robust across different economic conditions.

Paper uses deep learning to price and hedge options in incomplete markets.

problem Incomplete markets lack unique no-arbitrage solutions for pricing and hedging European options.
method Constrained deep learning approach with a single neural network representing option prices and hedging strategies.
result Constrained networks produce superior P&L distributions compared to unconstrained networks.

Hybrid LSTM-PPO optimizes dynamic portfolios with better performance.

problem Dynamic portfolio optimization under non-stationary market conditions.
method Combines LSTM for forecasting and PPO for adaptive portfolio adjustments.
result Hybrid framework outperforms single-model and equal-weight approaches in various metrics.

Deep learning framework for bond and yield curve forecasting with no-arbitrage constraints.

problem Arbitrage-free yield curve and bond price forecasting.
method Combines Kalman, extended Kalman, and particle filters with LSTM/CLSTM, and introduces AER term.
result Arbitrage regularization improves forecast accuracy, especially at short maturities.

Study uses deep learning for pairs trading in Polish equities, achieving profits in 2017-2019.

problem Statistical arbitrage in Polish equities market using traditional methods.
method Deep learning (LSTMs) for asset replication, PCA for risk factor analysis, Ornstein Uhlenbeck process for residual modeling.
result Deep learning methods, especially LSTMs, show promise for profitable trading in Polish equities.

Deep RL ensemble strategy outperforms individual algorithms in stock trading.

problem Designing profitable stock trading strategies in a complex market.
method Ensemble of three deep reinforcement learning algorithms (PPO, A2C, DDPG) for stock trading.
result Deep ensemble strategy outperforms individual algorithms and traditional min-variance portfolio.

ETCNN uses neural networks to price American options accurately.

problem Accurately pricing American options with inequality constraints.
method ETCNN framework solving BSM equations with exact terminal condition.
result ETCNN achieves high accuracy and robustness across various scenarios.

Hybrid model combines deep learning and agent-based methods for synthetic LOB generation.

problem Generating realistic financial time series data for model training.
method Combining TABL model with Chiarella model for intraday trading activity simulation.
result Hybrid model generates realistic price dynamics but fails to accurately recreate market microstructure.

Investigates portfolio selection with transaction costs and stochastic volatility, using deep learning for computation.

problem Optimal portfolio selection with transaction costs and stochastic volatility.
method Two-factor stochastic volatility model, option-implied utility function, deep learning policy iteration.
result Deep learning method effectively computes optimal investment decisions under transaction costs and stochastic volatility.

Deep learning model optimizes portfolios by integrating news sentiment, stock relationships, and price data.

problem Optimizing portfolio weights using traditional methods introduces instability.
method Combines LSTM, GAT, and sentiment analysis in a unified pipeline.
result Delivers higher cumulative returns and Sharpe ratios compared to benchmarks.

Study finds long-range dependence in financial markets, but deep generative models struggle to replicate it.

problem Long-range dependence in financial markets and challenges of deep generative models.
method Empirical analysis of financial data from three sectors, including LRD through various statistical methods and deep learning models.
result Deep generative models can reproduce stylized features but fail to capture long-range dependence structures.

LEMs extend transformer-based architectures for complex execution problems.

problem Handling flexible time boundaries and multiple execution constraints in deep learning.
method Decouples market information processing from execution allocation decisions using TKANs, VSNs, and multi-head attention mechanisms.
result LEMs achieve superior execution performance compared to traditional benchmarks.

EMDLOT predicts bond defaults better than traditional methods.

problem Lack of interpretability and irregular temporal dependencies in financial data.
method Integrates time-series and textual data, uses Time-Aware LSTM, soft clustering, and multi-level attention.
result EMDLOT outperforms traditional and deep learning benchmarks in recall, F1-score, and mAP.

DeepAries optimizes rebalancing intervals and asset allocations for better portfolio performance.

problem Fixed rebalancing intervals lead to unnecessary transactions and poor risk-adjusted returns.
method Adaptive deep reinforcement learning with Transformer state encoder and PPO.
result DeepAries outperforms traditional strategies in risk-adjusted returns, transaction costs, and drawdowns.

Study portfolio selection with exogenous and endogenous transaction costs using deep learning.

problem Portfolio selection with both exogenous and endogenous transaction costs.
method Deep learning-driven policy iteration scheme for high-dimensional HJB equations.
result Proposes a scheme to address the curse of dimensionality and adapt to high-dimensional control spaces.