8.1 Investment Research Use Cases
Last updated
Last updated
Identifying the technological focus of companies, especially in rapidly evolving industries, is a complex task. Each company often has specialized areas of focus, and with the continuous evolution of technology, new terms and innovations emerge regularly. Traditional methods of categorizing and tracking these technologies can quickly become outdated, leading to gaps in understanding and missed opportunities for analysis. To address this challenge, we leverage Large Language Models (LLMs) to systematically analyze public companies' annual reports and identify any innovative or disruptive technology mentioned. This innovative approach allows us to stay ahead of the curve, ensuring that we capture the most current and relevant information for analysis.
Data Collection: We utilize annual reports directly from our global exchange filings knowledge base. This knowledge base is an extensive repository of publicly filed documents from global exchanges, ensuring that we have comprehensive and up-to-date information at our fingertips. This direct access is a significant time-saver, allowing us to quickly gather the data needed for analysis.
Prompting with LLMs: We then use a specially crafted prompt with LLMs to search through these reports. The prompt we use is: "What are the innovative or disruptive technologies mentioned in the document? Please ignore the general terms and be focused on specific areas, and return these terms as comma separated without any explanations." This prompt guides the LLM to focus on identifying specific technological terms that are of particular interest, filtering out more general, less relevant terminology.
Structuring Results: The LLM outputs the identified terms in a comma-separated format, which is then structured into usable data formats. This structured data forms the basis for further analysis and categorization.
Classification: The next step involves classifying these terms into broader technological themes. This is done using additional prompts that guide the LLM to group related terms into categories such as "Artificial Intelligence," "Blockchain," "Quantum Computing," etc. This classification helps in organizing the data and making it easier to analyze.
Analysis: Once the terms are classified, the data is ready for analysis. Analysts can use this information to assess which technologies are gaining traction across different industries, which companies are leading in certain areas, and how technological trends are evolving over time.
By utilizing LLMs in this innovative way, we are able to capture the fast-paced changes in technology that companies are adopting. This method not only provides a more accurate picture of the current technological landscape but also allows us to predict future trends by identifying emerging technologies early on.
The full result from S&P500 annual reports is available here.
