Munibonds.ai launches AI-powered platform for muni bond analysis

Munibonds.ai is set to officially launch its AI-powered platform, which is designed to streamline municipal bond analysis.

The firm was founded in 2024 by a long-time financial technology veteran Robert Kane. The platform uses AI to analyze any municipal bond issuer by centralizing data and delivering easy-to-understand results instantly, Kane said.

"We were seeing the AI focus that's happening and [wondered] how we could apply that to the muni bond market," he said.

The AI-powered credit research tool reviews hundreds, if not thousands, of pages of information, including official statements, material events, rating changes, financials and even news articles, for every new issue, Kane said.

From there, Munibonds.ai creates an easily digestible report, deal and bond summaries, key financial metrics, risk factors and disclosure flagging and an AI-generated podcast, he said.

Its AI-credit sentiment provides three levels of ratings: positive, neutral or negative. The Munibond.ai rating may differ from those of rating agencies, as ratings can become "stale," Kane said. However, a triple-A-rated bond will likely be rated similarly by Munibonds.ai, according to Kane.

The AI-credit sentiment is useful because it offers AI-generated "second opinions" on all new issues, including nonrated bonds, which make up one-third of the muni market, he said.

The firm also has a ChatGPT-like chat function that allows users to ask about new-issue bonds, from basic questions to more specific ones, such as the number of days of cash on hand, Kane noted.

"If you wanted to ask these kinds of questions to ChatGPT, it would need to find the various documents, news sources and other information, including market prices. And you have to pull all that together, load it up, and then start asking questions from the perspective that we created already," he said.

To prevent the chat function from hallucinating, Kane said, the firm used prompt engineering to ensure accuracy. Additionally, all information is updated daily, he said.

Kane likes to think of the user as the "pilot," and Munibonds.ai features as "co-pilots."

"It's not like we are able to do something magical that anyone else can't," he said. "You could do this if you had the time and the capability to monitor all these sources, but up until AI has been here, it's been difficult for anybody to do that."

For example, this tool can assist credit analysts at large funds, who have to review new-issue bonds issued every week. It can be difficult to do a deep dive into all of them, Kane said.

Therefore, "this tool helps in that all these bonds have already been reviewed, [so] credit analysts don't need to start from scratch," he said.

This platform would be valuable to hedge funds in the muni bonds space that buy nonrated or distressed bonds, he said, since when there is no rating, they must do their own legwork, Kane said.

The product provides users with a "starting point" where all the bondsare reviewed and monitored, he said.

Some community banks with bond portfolios of less than $100 million are customers, Kane said.

These institutions don't have a full-time analyst but are required to both assess and monitor the creditworthiness of their portfolio, which goes beyond merely looking at the rating agency's rating, he said.

"It's early days for muni bonds and AI-related products," Kane said, noting most have been focused on pricing.

"There's a lot of opportunity because there's a uniqueness to information in the space: 1.3 million bonds, more or less, each of which has thousands of pages, documents, material and events. I don't know that a lot of attention has been paid to it before because it's just been a Herculean task. But now it's possible to do that," he said.