USE OF LARGE LANGUAGE MODELS IN THE ASSET MANAGEMENT INDUSTRY: EVIDENCE REVIEW

Abstract

Purpose Large Language Model (LLM) applications can play an important role in the asset management (AM) industry, which is a critical sector within finance and investment. For successful implementation of the technology, there must be a match between the industry challenges and the technological capabilities.

Design/ methodology/ approach A review of evidence supported by an LLM-based evidence and review tool was completed to identify key challenges of the industry and current LLM applications within the sector. Content analysis was used to match the key categories of the challenges with the areas of LLM applications.

Findings The challenges in the AM industry include regulatory and compliance issues, technological disruption, market dynamics, performance measurement, and adaptation to consumer preferences. Despite a scarcity of specific literature on LLMs in AM, the research points to potential applications such as sentiment analysis, market predictions, and regulatory compliance. LLMs can enhance customer interaction, improve risk management, and develop informed investment strategies by processing and analyzing large datasets. A matrix of the industry challenges and LLM application is created to indicate well addressed areas and those requiring further development. There is a need for more diverse training datasets, better integration, and scalability of LLMs, and improvements in global applicability.  

Practical implications While LLMs hold significant potential for transforming AM by addressing various challenges, further investigation is needed to overcome existing gaps and limitations. Future research should focus on enhancing data diversity, improving model integration, ensuring regulatory compliance, and addressing the environmental impact of LLM technologies.  

Originality/value This study contributes valuable insights into LLM applications, supporting the advancement of automated tools in financial services.