AI/ML is finally moving the dial on operational efficiency in private markets: This is part 2 in a 5-part blog series

New pressures arise amid changing market dynamics

Linedata’s 2023 Global Asset Management Survey (GAM survey) revealed that cost efficiency and increased automation are at the top of alternative asset managers’ strategic agenda over the next 12 months.

The complex and often bespoke nature of alternative investment processes has made it a challenge for private markets firms to digitalize their operations at the same rate as other segments of the asset management industry.

And while heavy reliance on manual data processing, Excel spreadsheets, and legacy IT systems have long been a source of frustration, the combination of new market pressures and the emergence of more sophisticated technologies is creating an inflexion point for alternative managers.

The end of the low interest rate era has perhaps brought the most acute cost pressures for private equity firms, as debt finance has become a lot more expensive to access. But PE shops also face new risk management demands, as portfolio companies in struggling sectors of the economy have taken a hit to earnings – adversely impacting their overall risk profile.

For private credit managers, higher rates can be seen as a positive for returns, but they also bring increased default risks to bear, which demands more scrutiny of risk exposures and closer monitoring of interest rate coverage ratios across portfolios.

Against this backdrop of more expensive finance, heightened risk management demands, and growth pressures, alternative managers are taking a hard look at their operating models.

Efficiency doesn’t come easily in private markets

The relative lack of standardization in private market investments has long been a barrier to simplifying data management and streamlining operational processes in the sector.

Private credit arrangements tend to be bespoke and thus are difficult to codify in a standardized way. PE assets often require many different data sets and customized calculations to determine valuations.

Furthermore, unlike in public markets, private markets data is often provided to managers in multiple different templates and formats. This means a big chunk of the investment teams’ time is spent on collecting, organizing, and transcribing data before they can do any meaningful analysis. And, in the absence of a unified data provider for the industry, managers need to repeat these time-consuming processes each earnings season.

Machine learning is transforming the data landscape

Thanks to machine learning and both ‘traditional’ and generative AI, we are now seeing a path to solving some of these data processing challenges. Within private credit, for example, we can teach machine learning tools to contextualize information contained in financial or contractual documents, extract and standardize relevant data, and aggregate this into data portals that can be accessed by investment teams.

This lifts the data processing burden away from credit analysts and standardizes processes for monitoring credit instruments.

While the initial aim of this approach will be to standardize data coverage across credit instruments and ensure new market data is updated seamlessly, firms will also be able to layer on additional capabilities such as dynamic modelling further down the line.

The same principles can be applied within private equity. Machine learning tools can extract information contextually from the multitude of data formats portfolio companies use to report their financials – saving time and enhancing predictability and informed decision making.

Alternative managers will need the right data infrastructure to support these new tools, whether they host this internally or work with an external partner. Yet in our GAM survey, only 23% of alternative managers said they have already invested in a cloud data warehouse.

Centralized data repositories are key to the effectiveness of a machine learning-driven approach to data processing and standardization. But looking beyond that, as generative AI and large language models unlock more sophisticated capabilities, it will be increasingly important that these are hosted in the public cloud.

About the author, Anup Kumar

Anup Kumar is a seasoned leader with 30 years of industry experience in Investments and Insurance. He is currently EVP & Global Head of Services at Linedata and leads all aspects (Operations and Technology) of Linedata’s Global Services business. Linedata’s full Services portfolio for buy-side firms includes Front Office, Middle and Back Office, Advisory, Cybersecurity, and Managed Services Provider (MSP). Anup’s career experience spans the Asset Management, Healthcare, Insurance, and Retirements industries, where he has held executive roles including President and CEO, Board Member, JV Head, and Business Head. He has led businesses of up to $350M in annual revenue at leading outsourcing services firms, including Capgemini, Hewitt, EXL, Patni and Tech Mahindra. Before joining Linedata, Anup was doing CXO Advisory for Institutional buy-side clients and mid-market private equity firms.