Utilizing GenAI for Asset Management

Investing.com — According to Bernstein, Generative AI (GenAI) is becoming an essential tool in asset management, serving as a “technical scaffold” that helps non-technical professionals engage with models and gain deeper insights.

During the Bernstein “Quant Speaker Series,” a portfolio manager from GIC pointed out that large language models (LLMs) are helping to bridge the gap for individuals who previously found traditional machine learning daunting due to significant technical challenges.

Bernstein notes that LLMs are particularly adept at handling unstructured data, such as text and natural language, where creativity and context play a critical role. Unlike traditional models, which are better suited for more formulaic tasks, GenAI thrives in environments requiring variability and adaptability.

Additionally, LLMs are valuable tools for summarizing documents, extracting key information, and translating images into text, enabling users to effectively differentiate between relevant signals and background noise.

However, challenges persist. Bernstein highlighted that GenAI models are still in a developmental stage and come with substantial fixed costs, making them more advantageous for larger firms that can benefit from economies of scale. Smaller companies may find it difficult to adopt these technologies broadly due to the significant investments necessary. Moreover, organizational readiness is vital, necessitating a strong data infrastructure and alignment of legal and compliance aspects.

The firm also noted the importance of human involvement, stressing that as AI reduces the costs associated with cognition, humans must adapt by either becoming specialists in the field or learning to integrate GenAI tools into their workflows.

Despite its potential, Bernstein cautions that GenAI still requires human oversight due to its limitations, such as forward-looking bias and inconsistent outputs. This means that, for the time being, human input remains a crucial component in the decision-making process within asset management.

In summary, while GenAI presents substantial opportunities for asset management—particularly in areas like feature extraction and summarization—Bernstein believes that its widespread adoption will hinge on overcoming various organizational and technical hurdles.