Plugging Into ChatGPT To Improve Qualitative Investment Analysis

# AI

# Technology

A portfolio manager at a forward-thinking tech company seeks an edge by tapping the power of generative AI.

Turning the power of generative artificial intelligence (AI) into a competitive edge by using tools including ChatGPT is a top priority of one member of NeuGroup for Cash Investment . He has begun using the technology’s ability to analyze massive amounts of text using natural language processing, with the goal of improving qualitative investment analysis.

That sets his company apart: While many finance orgs are on the automation bandwagon and want to leverage the wonders of AI and machine learning, not many have committed to finding uses for generative AI right away.

Some finance professionals fear security breaches and other potential risks presented by the technology. Many financial institutions prohibit staff from using ChatGPT and some corporates are following suit .

The member embracing generative AI underscored the need for companies to implement strict controls, guardrails and governance processes.

SEC filings and beyond. “On the qualitative analysis side, this is where I see the most potential for generative AI like ChatGPT,” the member said at the cash investment group’s first-half meeting sponsored by DWS. “You can run SEC filings through a large language model with key topics and themes—you can do that seamlessly.”

He also envisions using the technology to analyze corporate earnings call transcripts, the Fed’s Beige Book and a database of Fed speakers .

The company’s senior leadership is pushing the entire corporation, not just finance, to lean into generative AI. In addition to using tools made by third parties, the member said the company will develop some AI tools itself.

The member does not plan to use generative AI chatbots for quantitative analysis, forecasting or reporting. The reasons include concerns about accuracy and limitations on the technology’s ability to search HTML or PDFs.

Alternatives. Treasury and finance teams without the stomach, budget, skills or leadership support to integrate generative AI into their internal tech systems have options if they want to analyze data like the member planning to mine insights from giant collections of documents. That’s the view of Matt Thomas, head of the cash investment group as well as NeuGroup’s digital assets working group .

Microsoft Office 365 subscriptions, Mr. Thomas noted, include access to Power Automate, which can parse documents for keywords. Microsoft is a major investor in OpenAI, the company behind ChatGPT.

A representative of DWS said the asset manager is blocking some AI tools including ChatGPT due to risk. He said he’s pushing for in-house development of software leveraging AI, which requires having developers embedded within finance teams.

Another member of the cash group said analyzing drivers for forecasting by using generative AI would be a key advantage, “giving you an actual basis for the forecast as opposed to gut feel.” He mentioned interest income forecasts and guidance on OIE (other income/expense).