https://www.ml-quant.com/topics/ml-ai-methods/Machine-learning methods applied to finance: deep learning, boosting, RL and new architectures. en https://www.ml-quant.com/papers/arxiv/2610.09048/https://www.ml-quant.com/papers/arxiv/2610.09048/Fri, 09 Oct 2026 07:00:00 +0000The study introduces a trillion-observation U.S. equity dataset and systematically evaluates 18 representation-learning methods across nearly two decades, finding that encoders with similar predictive performance organize market state very differently. https://www.ml-quant.com/papers/arxiv/2610.07006/https://www.ml-quant.com/papers/arxiv/2610.07006/Fri, 09 Oct 2026 07:00:00 +0000A joint-embedding framework learns predictable incremental revisions from market data relative to a financial prior, combining interpretability from classic models with the pattern-capture strength of deep learning. https://www.ml-quant.com/papers/arxiv/2610.09872/https://www.ml-quant.com/papers/arxiv/2610.09872/Fri, 09 Oct 2026 07:00:00 +0000A benchmark evaluates frontier LLMs as live trading agents and finds that realized returns often diverge from capability-specific measurements, revealing outcome-capability gaps through decision traces. https://www.ml-quant.com/papers/arxiv/2610.06856/https://www.ml-quant.com/papers/arxiv/2610.06856/Fri, 09 Oct 2026 07:00:00 +0000Proposes that large-language-model-driven trading agents delegated to autonomous execution define a nascent asset class, mapping infrastructure layers and projecting potential trillions in assets by 2030. https://www.ml-quant.com/papers/arxiv/2610.03128/https://www.ml-quant.com/papers/arxiv/2610.03128/Fri, 09 Oct 2026 07:00:00 +0000Proposes an LLM-guided framework for trading strategy refinement using accumulated experience and cross-period robustness, achieving 27.99% annualized return on Chinese equities with 1.63 Sharpe ratio. https://www.ml-quant.com/papers/arxiv/2610.02225/https://www.ml-quant.com/papers/arxiv/2610.02225/Fri, 09 Oct 2026 07:00:00 +0000Fuses SEC financials with EPA emissions data and conformal machine learning to quantify divergence between self-reported and physical emissions, finding algorithmic divergence predicts lower valuation and profitability. https://www.ml-quant.com/papers/ssrn/7555699/https://www.ml-quant.com/papers/ssrn/7555699/Fri, 09 Oct 2026 07:00:00 +0000An autonomous agent evolved 940 factors over 17 days and finds that reusing backtest data inflates edge by a quarter to a third and in-sample improvement predicts worse performance. https://www.ml-quant.com/papers/ssrn/7575518/https://www.ml-quant.com/papers/ssrn/7575518/Fri, 09 Oct 2026 07:00:00 +0000A natural-language analysis of 13,757 firms finds AI exposure surged after ChatGPT and predicts 1.37 percent higher annual returns, though the effect is concentrated after the launch and driven by firm-specific factors. https://www.ml-quant.com/papers/ssrn/7568638/https://www.ml-quant.com/papers/ssrn/7568638/Fri, 09 Oct 2026 07:00:00 +0000An autonomous research loop in which agents propose factor candidates and code verifies them builds multi-factor alpha models, screening 212 predictors and admitting 14 that raise baseline information coefficient out-of-sample. https://www.ml-quant.com/papers/ssrn/7576913/https://www.ml-quant.com/papers/ssrn/7576913/Fri, 09 Oct 2026 07:00:00 +0000GFlowAlpha uses neuro-symbolic search and internalizes econometric quality control as online rewards to mine white-box factors, delivering mean out-of-sample daily ICIR of 0.485 with Fama-MacBeth t-statistics of 4.99 on Chinese stocks. https://www.ml-quant.com/papers/ssrn/7557458/https://www.ml-quant.com/papers/ssrn/7557458/Fri, 09 Oct 2026 07:00:00 +0000The research proves that query-counted Deflated Sharpe Ratios certify noise in adaptive research and proposes sealed-holdout verification, where validity depends on bits revealed rather than strategy count. https://www.ml-quant.com/papers/ssrn/7576359/https://www.ml-quant.com/papers/ssrn/7576359/Fri, 09 Oct 2026 07:00:00 +0000A graph neural network combining temporal and cross-asset volatility patterns reduces mean squared error and quasi-likelihood loss by approximately 9.9% and 3.4% on Dow stocks. https://www.ml-quant.com/papers/repec/ces-ceswps-13000/https://www.ml-quant.com/papers/repec/ces-ceswps-13000/Fri, 09 Oct 2026 07:00:00 +0000Industries with higher AI intensity show significantly higher returns and valuations when hit by supply and technology shocks but not demand shocks. https://www.ml-quant.com/papers/arxiv/2609.34510/https://www.ml-quant.com/papers/arxiv/2609.34510/Fri, 02 Oct 2026 07:00:00 +0000A benchmark compares machine learning, reinforcement learning, and large language model trading methods across historical backtests, paper trading, and live markets to measure the gap. https://www.ml-quant.com/papers/arxiv/2609.34188/https://www.ml-quant.com/papers/arxiv/2609.34188/Fri, 02 Oct 2026 07:00:00 +0000The research uses reinforcement learning with multi-objective rewards to discover formulaic alphas that work well together despite evolving reward functions and shifting environments. https://www.ml-quant.com/papers/arxiv/2609.30316/https://www.ml-quant.com/papers/arxiv/2609.30316/Fri, 02 Oct 2026 07:00:00 +0000A new method adapts financial language models to avoid look-ahead bias using low-rank adapters instead of expensive annual retraining, matching performance of full pretraining. https://www.ml-quant.com/papers/arxiv/2609.31013/https://www.ml-quant.com/papers/arxiv/2609.31013/Fri, 02 Oct 2026 07:00:00 +0000Cross-model rank correlations for LLM-extracted sentiment, clarity and risk average only 0.52, and model choice significantly alters coefficient signs and significance in downstream analysis. https://www.ml-quant.com/papers/arxiv/2609.34507/https://www.ml-quant.com/papers/arxiv/2609.34507/Fri, 02 Oct 2026 07:00:00 +0000Diffusion transformer foundation model reformulates candlestick prediction as conditional trajectory generation, achieving 0.057 mean return RankIC across markets and timescales. https://www.ml-quant.com/papers/arxiv/2609.30705/https://www.ml-quant.com/papers/arxiv/2609.30705/Fri, 02 Oct 2026 07:00:00 +0000The study tests whether extra reasoning in large language models improves portfolio returns net of trading costs across multiple model families, finding no reliable gains. https://www.ml-quant.com/papers/arxiv/2609.33470/https://www.ml-quant.com/papers/arxiv/2609.33470/Fri, 02 Oct 2026 07:00:00 +0000An evaluation framework tests language model agents on structured option trading tasks, finding current systems underperform in most real-world scenarios. https://www.ml-quant.com/papers/arxiv/2609.33524/https://www.ml-quant.com/papers/arxiv/2609.33524/Fri, 02 Oct 2026 07:00:00 +0000A framework tests whether self-evolving research capabilities improve alpha discovery over time, finding no consistent gains from accumulated experience and tools. https://www.ml-quant.com/papers/arxiv/2609.34474/https://www.ml-quant.com/papers/arxiv/2609.34474/Fri, 02 Oct 2026 07:00:00 +0000A hybrid approach combining neural networks with kernel methods produces robust derivative hedges in low-data regimes and scales via random Fourier features with convergence guarantees. https://www.ml-quant.com/papers/arxiv/2609.36914/https://www.ml-quant.com/papers/arxiv/2609.36914/Fri, 02 Oct 2026 07:00:00 +0000Post-training Qwen3-4B via supervised fine-tuning and policy optimization more than doubles direction-magnitude score on chronological stock-price predictions to 43.31. https://www.ml-quant.com/papers/ssrn/7528558/https://www.ml-quant.com/papers/ssrn/7528558/Fri, 02 Oct 2026 07:00:00 +0000Deep residual networks outperform shallow models in asset pricing, achieving a long-short Sharpe ratio of 2.07 versus 1.92 for shallow versions by preserving and refining earlier layers. https://www.ml-quant.com/papers/ssrn/7522878/https://www.ml-quant.com/papers/ssrn/7522878/Fri, 02 Oct 2026 07:00:00 +0000The research proposes a temporal graph neural network with explainability tools for real-time systemic risk surveillance, achieving early warning signals 3-4 quarters ahead of financial distress on bank data. https://www.ml-quant.com/papers/repec/ecb-ecbwps-20263262/https://www.ml-quant.com/papers/repec/ecb-ecbwps-20263262/Fri, 02 Oct 2026 07:00:00 +0000Large Language Models extract signals about potential trigger events from financial news, improving forward-looking estimates of downside risks and helping monitor financial stability threats ahead of major events. https://www.ml-quant.com/papers/repec/zbw-glodps-1821/https://www.ml-quant.com/papers/repec/zbw-glodps-1821/Fri, 02 Oct 2026 07:00:00 +0000AI automation reduces occupational employment by 21 percent with little wage effect, while AI augmentation raises wages by 8 percent, revealing that AI's labor impact depends on the balance between these opposing channels. https://www.ml-quant.com/papers/arxiv/2609.27051/https://www.ml-quant.com/papers/arxiv/2609.27051/Fri, 25 Sep 2026 07:00:00 +0000Proposes a statistical referee that judges investment factors proposed by language-model agents using out-of-sample market outcomes, ensuring false-discovery control at any stopping time. https://www.ml-quant.com/papers/arxiv/2609.29014/https://www.ml-quant.com/papers/arxiv/2609.29014/Fri, 25 Sep 2026 07:00:00 +0000Proposes a multi-agent system with post-training that automates alpha factor mining locally, using diverse research paths and joint optimization to broaden exploration while maintaining prediction quality. https://www.ml-quant.com/papers/arxiv/2609.28876/https://www.ml-quant.com/papers/arxiv/2609.28876/Fri, 25 Sep 2026 07:00:00 +0000Introduces a replayable environment combining 1,568 resolved prediction-market questions with 18.8M dated news articles to benchmark and train language-model forecasting agents on historical data.