Technology special report: To automation and beyond

AI and other automation technology is being used to propel growth in the private credit sector, but risks remain. Kathryn Gaw investigates…

Technology has an increasingly important role to play in the private credit sector, particularly in the era of artificial intelligence (AI). Automation has been a buzz word in the industry for several years now, and the mainstreaming of AI tech solutions has dovetailed with a boom in private credit funds, for better or for worse.

Just about every private credit fund manager uses technology such as AI to cut costs, speed up due diligence and data collation processes, and monitor investment portfolios for compliance risks. The extent to which automation is used varies from fund manager to fund manager, and doubts persist about the reliability of the technology.

However, like it or not, AI is very much a part of the private credit ecosystem, and it is just the tip of the technological revolution that is disrupting the private markets, and capturing the attention of the regulators.

Law firm Macfarlanes believes that cutting-edge technology – including AI, automation and contract management systems – is becoming increasingly critical to credit funds.

Read more: TPG in talks to buy fund administrator Alter Domus

“Whilst technology cannot replace the human skills and relationships that are essential for successful deal-making and risk management, it is playing an increasingly important role,” says Adam Caines, a partner at Macfarlanes.

“The consensus is that the private credit industry will need to balance the benefits and risks of technology, and invest in it alongside continued investment in talent and culture, to remain competitive and resilient in the future.”

Fund managers are acutely aware of the importance of maintaining that human element of portfolio management. Investors want to place their money with people, not algorithms. This is an industry where investors will follow individual fund managers and credit teams from one company to another; where dynamic reputations are rewarded with easier access to funding and new investment opportunities. To remove the human element completely risks alienating long-term investors.

Fund manager Pollen Street has been vocal in its commitment to automation and new technologies, but partner Michael Katramados believes that there are some functions that simply cannot be performed by technology.

“As things stand, I would not be comfortable removing the human element from monitoring and from data ingestion,” says Katramados. “If you are monitoring a portfolio of multiple assets with multiple degrees of freedom in the risks that you need to assess and understand, you don’t want to completely remove the human element from looking and understanding the data, and the trends that are generated from them.”

That human element is also necessary when it comes to building relationships with clients and investors. This means that automation is best used behind the scenes, in those parts of the business which are not client-facing.

“Careful thought must be given to replacing human involvement in any part of the credit and underwriting processes,” says Macfarlanes’ Caines. “But that freeing up deal team time to focus on originations and relationships will drive value creation.”

Indeed, many of the technological changes which have been rolled out recently have been inspired by investors. Recent economic turmoil has led to an increased focus on transparency and compliance by the institutional investors who fund the private credit sector.

These institutions are professional investors. They choose and monitor their allocations extremely closely, and they know exactly which data points to pay attention to. This means that fund managers must be able to meet these high expectations and provide data not just on their own operations, but on the operations of their clients too.

Pollen Street invests in a number of direct lending platforms, which means that it is not unusual for the company to have tens of thousands of loans representing millions of data points to monitor.

“As an asset based lender, data digestion, data manipulation, and data accuracy have been integral parts to our strategy,” says Katramados.

“We have an in-house tech team that is leading the development of a proprietary tech stack that comprises of a data lake that sits at the core of what we do.

Read more: “Compelling” opportunity for new capital in direct lending

“That data lake will ingest information from all the businesses we’re working with, will communicate with our finance and accounts department, and will act as the clean source of accurate information on everything that we do from the loans, to the collateral that are securing our loans, and the returns for our funds in one place”.

AI is already being used to manage asset inventories, for relationship mapping, and to prioritise data for marketing and research purposes. Its potential is almost unimaginable, but even in the short term AI promises to save time and money, while satisfying investor demand for more data.

Like many fund managers, Pollen Street also uses technology as part of its reporting to investors, and uses “a high degree of automation in that process”. However, Katramados is still wary of AI.

“There’s an element of data cleansing and making sure that there’s as much automation and automatic checks on the quality of that data,” says Katramados. “And there are certain providers out there that we’ve spoken to that use AI for that purpose.

“There’s going to be around the edges more and more of the human intervention that can be automated and therefore give more leverage to the team. And I think that’s really valuable. I don’t believe we’re at the point yet that we can just kind of close our eyes and let AI do our job well.”

Concerns around over-reliance on AI have been bubbling across the industry lately. In a recent paper for UK Finance, James Watts, sector lead, banking, financial services and operational resilience at Armis, warned that AI “directly feeds into external risk factors.”

Read more: Covenant breaches decline for third quarter in a row

“Its rapid development, commoditisation and proliferation will see it settle in the hands of those that choose to operate outside the controls of the global regulatory system,” Watts wrote.

“Regulation will struggle to keep pace with AI and the pending acceleration in innovation. AI’s power will grow, cyber ‘incidents’ will become ‘existential events’ for some, with the potential to become ‘systemic events’ for all.”

The Alternative Investment Management Association (AIMA) has even created a checklist for credit fund managers which aims to help them safely and ethically use generative AI – a subset of AI which creates content such as software codes and product design.

This checklist has been seen by Alternative Credit Investor and includes warnings around data privacy and the quality of the data produced. AIMA has also cautioned that the incorrect use of generative AI could present an increased risk of cyber security threats.

“A wide array of threat actors have already used the technology to create ‘deep fakes’ or copies of products, and generate artifacts to support increasingly complex phishing scams,” AIMA told its members. “Investment managers must develop robust internal policies on cyber security risk management.”

Pollen Street’s Katramados says that cybersecurity has been a big point of diligence for the fund manager.

“It’s a risk we need to cover specifically on every deal,” he says. “We have a cybersecurity risk framework and a checklist of things we want our borrowers to do and a risk scorecard that we have developed in house. If there are any vulnerabilities, they will be flagged and we will insist upon any gaps being closed.”

Several investment firms have already been the subjects of attempted cyber security attacks, which have been swiftly contained and sparsely publicised.

But while risks are inevitable with any emerging technologies, private credit fund managers are expert risk managers.

At present, AI and other forms of automation are used primarily on back-office processes such as background diligence on sectors, sponsors and potential portfolio companies, investor reporting, portfolio monitoring and environmental, social and governance benchmarking. If used correctly, it can be a powerful tool which can speed up many labour-intensive elements of the portfolio management process.

“Credit funds are very focused on optimising the application of AI without introducing additional risk – driving efficiencies where possible whilst maintaining tight controls and human oversight, particularly when it comes to credit analysis and decision making,” says Macfarlanes’ Caines.

It is this prudent approach towards new technologies that will serve private credit managers well as the sector continues to grow. However, challenges will persist.

More and more credit funds are seeking to target retail investors, in addition to institutional investors. Retail money comes with enhanced regulatory requirements which could either be streamlined or stymied by the use of automation.

In a recent speech, Jessica Rusu, the Financial Conduct Authority’s (FCA’s) chief data, information and intelligence officer asked: “Just because we have the ability to process the data, should we?”

The FCA has indicated that it will enhance its regulation of AI and similar technologies in the near future. Meanwhile, across the pond, the US Securities and Exchange Commission (SEC) has proposed new rules to address the risks of AI, particularly around using predictive data analytics which could potentially place a firm’s interest ahead of its investors’ interests.

The challenge for fund managers is finding the balance between investor requests for data transparency, and safeguarding those same investors from cyber attacks and data leaks. Somewhere out there, someone is working on a piece of software that does exactly that.