Generative AI in Finance: Mapping the Emerging Convergence of Quantitative Trading, Robo-Advisory, and Blockchain Markets (OpenAlex (recherche mondiale))
Summary
- The study analyzes a bibliometric dataset of 695 English-language Scopus records published between 2023 and 2025 to map the intellectual structure of GenAI in finance.p.3
- Annual output expanded rapidly from 46 documents in 2023 to 181 in 2024 and 468 in 2025, with 67.3% of the total three-year corpus published in 2025.p.4
- Publication dissemination is heavily dominated by conference papers, which account for 380 records (54.7%), compared with 170 journal articles (24.5%).p.5
- Author-country participation is concentrated in the United States with 359 occurrences, China with 252, India with 245, and the United Kingdom with 97.p.7
- Global citation volume is concentrated in foundational texts, led by a multidisciplinary GenAI paper with 579 citations, FinBERT with 548 citations, and a ChatGPT finance study with 491 citations.p.10
- Strategic keyword tracking shows portfolio/investment terms rising to 99 occurrences and governance/XAI/regulation reaching 84 in 2025, whereas robo-advisory lagged at only 4 occurrences.p.24
Key insights
- Average citations per publication decreased from 56.6 for 2023 papers to 15.6 for 2024 and 5.1 for 2025 due to shortened citation exposure windows.p.4
- The Proceedings of the Annual Meeting of the Association for Computational Linguistics is the single most frequent publication venue in the corpus with 16 documents.p.6
- Institutional production is highly distributed, led by the University of Oxford with 10 records, while several other top institutions contribute five records each.p.7
- Index keywords highlight core financial application areas, led by investments with 81 occurrences, financial markets with 73, finance with 73, and decentralized finance with 67.p.11
- Author keyword co-occurrence reveals three main knowledge clusters covering technology and governance, quantitative prediction methods, and financial service applications.p.14
- Thematic evolution shows a shift from exploratory artificial intelligence and ChatGPT themes in 2023–2024 toward sentiment analysis, machine learning, and retrieval-augmented generation in 2025.p.24
- Quantitative trading and retrieval-augmented generation/agents each surged from zero or few occurrences in early periods to 60 tracked occurrences in 2025.p.24
Topics and entities
Summary and figures generated from the document and checked automatically: every number shown appears on the cited page. The full document is available from the publisher. Publishers can request removal.