Dec 06, 2023

Research analysts face a constant barrage of internal and external data. The sheer volume and diversity of information often leads to information overload, making it difficult and time-consuming for analysts to identify the most relevant and valuable insights. This challenge hampers the decision-making process, delays the identification of potential investment opportunities, and exposes portfolios to unnecessary risks.

In this article, we explore how a modern research management system (RMS) and generative artificial intelligence (AI) can empower research analysts to extract meaningful signals from the noise.

Modern RMS and Generative AI Benefits

An AI-enhanced, configurable system that supports research analysts’ process of sourcing, analyzing, and developing investment ideas, a modern RMS serves as a central store of all the information factored into those investment ideas. Modern RMS already use AI to organize and associate content (referred to as auto-tagging) to deliver advanced search functionality, which allows analysts to quickly find what they’re looking for with precision and perform sentiment analysis on news articles, social media, and transcripts. These AI features can aid analysts in getting through massive amounts of information and data and increase their productivity.

With generative AI, research analysts can further benefit from the following:

  • More auto-tagging: Generative AI can parse the content on an authored internal research note and apply the proper company tags.
  • Summary creation: Generate concise and informative summaries of lengthy artifacts and articles and bring the most important aspects to the top. Generative AI can instantly create a summary to highlight key points and cite sources. Analysts can then review the generated summary and leave it alone or make edits before sharing.
  • Manual input reduction: Note templates within an RMS help capture key structured data consistently and make it easier to see important data in dashboards. Generative AI can extract data from the note’s body and enter it into the required fields in the template.
  • An “assistant” to ask questions and get answers: Use an AI-powered chatbot to ask a question, and the “assistant” retrieves information from stored research artifacts, incorporating both structured and unstructured data, to identify patterns and relevant data points to produce a response.

Incorporating generative AI into an RMS is a high priority for vendors. As a result, vendors have pivoted on their short-term road maps to responsibly integrate generative AI into their systems.

Key Considerations

Generative AI stands out as a notable advancement — poised to deliver considerable value and efficiency. However, as vendors embrace its capabilities for innovation and problem-solving, it’s imperative to approach generative AI with a measured perspective, acknowledging the nuanced challenges it brings. These challenges include the inadvertent biases it might perpetuate, the technology’s limitations, and the critical considerations surrounding security vulnerabilities and ethical implications. Navigating the terrain of generative AI demands a balanced appraisal of its benefits while remaining vigilant to the potential drawbacks.

RMS vendors and clients must collaborate to identify and mitigate biases to ensure fair and equitable outcomes in research and decision-making processes. The “black box” nature of most large language models (LLMs) makes it challenging to understand how the AI models arrive at certain conclusions. To build trust with their users, RMS vendors should strive to make their AI models more transparent and explainable.

Educating users about the capabilities and limitations of generative AI is also critically important. RMS vendors should ensure that users are well-informed to prevent misuse of the technology and foster responsible AI adoption. To be accurate, an LLM must be trained on a large dataset. Requesting insights on niche topics with limited data could lead to inaccuracies or so-called “AI hallucinations.” RMS vendors should collaborate with users to ensure they understand when and how to appropriately utilize the technology.

In addition, it’s essential for both vendors and clients to establish clear security and ethical boundaries. Data privacy and compliance with regulations should be top priorities, with a focus on promoting responsible AI use within the platform. Vendors must continue to implement access controls and encryption mechanisms, as AI models trained on extensive proprietary datasets will capture sensitive information.

Optimize Research Analysts’ Time

The ever-expanding pool of data and information in the asset management industry presents both opportunities and challenges for research analysts. With an abundance of information available, sifting through the noise to identify valuable insights and make informed investment decisions has become increasingly complex and time-consuming. By incorporating generative AI technology into their platforms, RMS platforms can effectively aid research analysts to facilitate more informed investment strategies and make decisions quickly. Don’t get left behind — the time is now to begin exploring how to integrate generative AI into your firm’s investment research process.

Want to discuss this topic with one of our research analysts? Contact us at [email protected].