Fitch Solutions’ recent extension of its distribution strategy to include a presence on Snowflake is the first step in a programme of leveraging the company’s huge cache of credit ratings and research data.
The move, which saw Fitch add its core credit ratings products on Snowflake Marketplace earlier this year, will be followed by other similar partnerships to make it easier for clients to access the data beyond the company’s APIs and website. The new strategy has also been engineered with an eye on meeting clients where they get their artificial intelligence capabilities, and includes a number of enhancements to its own AI offerings.“Based on client feedback and the presence of Snowflake in the market, we thought it was a good starting point for us to begin the journey of data distribution through not just our own APIs and feeds, but through partnership platforms,” Rachel Lojko, Fitch’s newly appointed president, told Data Management Insight.
“We want to meet customers where they are from a demand perspective, whether they’re consuming through Snowflake, whether they’re consuming through a cloud-service provider, and wherever they are on their AI journey.”
Snowflake Provides Launchpad for Enhanced Distribution Framework
Fitch’s credit data and loan-level performance benchmark datasets from its dv01 business, which covers the US auto, consumer unsecured and non-agency residential mortgage backed securities (RMBS) sectors, are among those first available on Snowflake Marketplace, with Ex-Government Support (XGS) credit ratings for banks and Sustainable Fitch’s Leveraged Finance Scores now live too.
“The first data set that we’ve made available on Snowflake is essentially the ratings data itself across all of the asset classes, but our intentions are that that’s our starting point,” said Lojko. “And then we’re going to fan out our other content sets.”
Peter Kohler, head of solution architects, said that the new strategy was also built on a realisation that clients wanted to “spend all their time getting value from our data, no time having to massage and do the ETL [extract, transform, load] work”.
Doing More With the Data
Lojko said Fitch had recognised that it could do more with its core products as financial clients’ demand for data surges in an age of increased analytical capabilities and AI. That means getting Fitch’s content into the environments where its clients are using their other data.
Snowflake was chosen as the take-off point because many of Fitch’s clients already operate on the platform and because its core competency around data management would make the co-mingling of datasets easier. Lojko said that equity research and low-latency trading providers were already demonstrating the benefits of such a move.“We want to be the credit player who’s there first,” she said.
Fitch sees its distribution strategy as aiding established data use cases and those that have emerged with the growth of AI, among them risk and regulation management, investment workflow and private market participation – the latter of which the company already serves through its Bixby business.
Although Snowflake is a major player in AI provisioning and facilitation, Lojko said that wasn’t the prime reason for beginning its distribution journey through the company.
“We’re not precious about where clients access our data with AI overlaid as long as it’s where they are making a decision,” she said. “That’s the piece that we’re prioritising. Wherever clients go to make a decision, that’s where we want to be. So our insights and data are commingled with their process. We’re delivery agnostic.”
Making the AI Revolution Easier for Clients
Fitch has put new emphasis on AI in a recent slate of product announcements.
The data made available on Snowflake Marketplace is “AI-enabled” so that it can be seamlessly integrated into workflows, and Fitch has added to its Fitch Ratings Pro product Genie, an internal large language model chatbot that is designed for easier retrieval of credit ratings data and source material.
“We’ve got over 2,000 analysts internally using our platform, so as our own analysts are doing their ratings work, they’re essentially helping to train the model and improve our model,” Lojko said.
The new strategy has come with a set of updated future targets too. Chief among them is building out a Model Context Protocol connector to enable clients to access the company’s research troves and deliver it into third-party platforms.
“We’ve been an organisation with a lot of great analytical expertise and really haven’t focused so much on the evolution of distribution strategies and making sure that our data is out there,” said Lojko. “We’re entering a new phase where we really want to prop up our strong analytical expertise with a commensurate distribution strategy and continue to invest in maximising the reach and impact of the insights they deliver.”
Subscribe to our newsletter
