Credit has become a key growth pole for private investors as loans and similar instruments have benefited from rising rates and wider public credit spreads. Interest in the sector is so strong that Morgan Stanley forecasts its value will rise to $5tn by 2029 from $3tn this year.
The difficulty for institutional investors clamouring to claim a piece of the action, though, has been prising open the black boxes of information on those assets. Credit is among the most illiquid and opaque corners of a part of the economy that has been traditionally lacking in easily accessible data. And for participants used to the data efficiency of public markets, this can prove a sticking point to competitiveness.New York-based Alkymi has long specialised in bringing transparency to private markets and has become the latest artificial intelligence-led workflow provider to shine a light into the space. Its recently launched Alkymi Private Credit offering seeks to automate the extraction and deployment of credit data locked in unstructured sources such as PDFs, email attachments, financial reports, covenant documents, and investor portals It also provides for close risk monitoring on a platforms that’s designed to reduce the operational burden on market participants.
Data Needs of Smaller Market Players
Alkymi, which services clients that have $20tn of total assets under allocation, is betting that the rapid growth of the market will draw in smaller participants who don’t have the same data capabilities, said co-founder Harald Collet.
“When you already have a large market growing astronomically, you start seeing the infrastructure coming under strain – what worked yesterday doesn’t work tomorrow,” Collet told Data Management Insight.
“If you’re already participating in the market today and your firm experiences operational headaches, many more firms will struggle with those operational headaches when volume doubles, triples – we expect there will be lots of other participants. Our large clients have the volume today, others will have the volume tomorrow.”
Returns Profile Boosts Appeal of Diversification Strategy
Private credit has become attractive to institutions as they diversify their investment strategies to hedge against macro-economic volatility and uncertainty. Credit has been paying compelling returns since the days of easy monetary policy came to an end after the Covid-19 pandemic.
Large data providers including Bloomberg have entered the space, leveraging their size and capacity to offer data and indexes to navigate the market. But the idiosyncratic nature of private markets has also opened the door to specialised data and technology providers, including Alkymi.
Data Extraction Challenge
The complexity of loans and other credit structures doesn’t lend itself to easy information sharing. Loans, unlike equity and other relatively straight-forward assets, have more moving parts – terms, covenants and rates, to name just a few. Extracting that information and then presenting it in a useable format is complicated.
“When you make investments in private credit, there’s a lot of critical signals to monitor… and lo and behold, all of this data is unstructured and creates a huge headache for everyone who’s involved,” said Collet.
“In the complex private credit market that has all of these different facets, can you transform that horrible underlying data mess into clean structured data that can work well?”
His answer, predictably, is yes.
Alkymi has deployed AI to get around those hurdles. It’s Private Credit offering is focused on three key documents that credit investors look for: loan agent notices, the fundamental document outlining details of credit agreements; compliance certificates; and, financial statements.
“Within loan agent notices, that is first off our fundamental and, frankly table stakes pattern, where it’s the largest operational burden… over the life of a single loan facility, you may have dozens of documents coming in,” said Alkymi senior data solutions architect Nate Byerly. “For many of our clients in the private credit space, they may have hundreds of loans. Some will have thousands. And so this becomes quite quickly an operational burden.”
Time Saving Equals Cost Saving
From loan agent notices, the technology mines information on amounts to be wired, banking details, facility type and fee sizes. Byerly said that automating this process saves 10 to 15 minutes per notice, and provides early warnings for new borrowing requests and deteriorating market conditions.
Alkymi’s extraction of compliance certificate data is considered its most sophisticated feature and processes documents issued by borrowers at even the most erratic of cadences. It finds identifying details, detects defaults and categorises covenants. It can also track changes in covenant metrics.
These processes would traditionally have been laboured over manually and logged in spreadsheets, Collet said. Only technology has enabled access to these troves of information.
“Without the rapid advancements that’s happened the last two, three years, in AI, this would be nearly impossible to do,” Collet said. “The workhorses in these systems for us are large language models (LLMs) that now allow us to reason over these complex documents, to understand the different types of covenants, and then of course to be able to do this at scale across a complex universe.”
Alkymi has enriched its offering through its Enterprise Solution, which knits a variety of large language models into its cloud-based generative AI offering, permitting secure and more flexible processing. The addition was built in “close partnership with some of the world’s most sophisticated investment firms”, the company said.
Being unable to detect and extract data to help monitor private markets puts firms at a competitive and strategic disadvantage, Collet said.
“What is it worth if you have a $2bn private credit exposure to have an early warning on deterioration in a market that has a lot of bells going off?” he said. “What is it worth to have a better radar to detect early signals that your competitors may not?”
Subscribe to our newsletter
