{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T15:53:36Z","timestamp":1791388416478,"version":"4.3.3"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":0,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>The rapid expansion of multimodal financial data and the inherent non-stationarity of markets have exposed critical limitations of traditional linear quantitative frameworks. In response, the integration of Deep Reinforcement Learning and Large Language Models has catalyzed a paradigm shift from manual feature engineering to autonomous financial intelligence. To systematically synthesize these cross-disciplinary advancements, this paper reconceptualizes modern quantitative investment not as a collection of isolated predictive tasks, but as an Four-Step Multi-Factor Investing Framework organized along four computationally interconnected stages: Factor Mining, Factor Combination, Risk Control, and Portfolio Management. Through this unified framework, we trace the evolutionary trajectory of each stage, showing how technical innovations propagate through the pipeline: factor mining has progressed from stochastic symbolic search to LLM-driven semantic reasoning; factor combination has shifted from simple linear weighting to synergistic aggregation that resolves multicollinearity; risk control has evolved from static exposure limits to the active prediction of forward-looking tail-risk indicators; and portfolio management has advanced from static convex solvers to DRL orchestrated dynamic allocation under real-world market frictions. We further provide a structured taxonomy of multimodal datasets and benchmark simulation environments, and present a systematic comparative summary of state-of-the-art performance across diverse market regimes. This survey establishes a structural blueprint for evaluating cross-modal AI models in finance, identifies critical bottlenecks including semantic hallucinations and Out-of-Distribution fragility, and outlines open research directions toward fully autonomous multi-agent trading systems.<\/jats:p>","DOI":"10.2139\/ssrn.7568958","type":"posted-content","created":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T15:06:54Z","timestamp":1791385614000},"source":"Crossref","is-referenced-by-count":0,"title":["From Factors to Portfolios: A Survey of AI-Driven Quantitative Investing"],"prefix":"10.2139","author":[{"given":"Zuohao","family":"Liu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T15:06:54Z","timestamp":1791385614000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7568958"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":0,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7568958","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}