MSCI, the research and analytics firm best known for market indexes, is now rolling out tools that use generative artificial intelligence aimed at the middle offices of asset management companies.
Jorge Mina, New York-based head of analytics, says that in addition to improving productivity in areas such as risk management, genAI gives mid-sized investment firms the chance to do things that until now required the resources that only the biggest firms in the industry could access.
“The biggest organizations build an infrastructure that ensures everyone has access to the same information,” Mina told DigFin. “They embed risk management in the portfolio-construction process, not just keep that information siloed away in a report. Not all organizations have been able to build that kind of infrastructure. But genAI can help democratize that process.”
Middle-office musts
Risk management is considered a middle-office function in asset manager, between the portfolio managers and client-facing sales teams, and the operations in the back office.
Its role includes making sure portfolio managers are sticking to the parameters of any given fund’s mandate and complying with both regulations and client agreements. For example, a portfolio may have bespoke limits on position size, sector allocations, ESG-related exposures, and so on.
Portfolio managers sometimes breach these rules for the sake of better performance, but more often breaches occur because managers are responsible for more funds than they can humanly handle.
While these parameters are one-size-fits-all for a mutual fund, asset managers typically manage thousands of portfolios, each tweaked for various institutional client needs.
Index to intel
MSCI’s claim to fame is creating and managing indexes. Like its rivals (S&P, FTSE-Russell, Dow Jones Indexes), it offers both widely followed benchmarks (eg the MSCI World for equities) and indexes customized to the preferences of asset managers, including hedge funds and asset owners.
Licensing agreements for these indexes are the foundation for a trove of data about markets and portfolios, which MSCI and its competitors have spun into lucrative businesses around analytics. MSCI also has large businesses around climate- and sustainable-investing strategies, factor investing (used for many ETFs), and services for private capital and real assets.
Given the data-heavy nature of the business, index vendors have been big users of big-data analytics and artificial intelligence. The firm uses AI for assuring data quality, parsing documents, and finding data points for products such as climate benchmarking.
Enter genAI
But since ChatGPT exploded on the scene in early 2023, the possibilities have expanded. “The models became more powerful by an order of magnitude,” Mina said.
One of the challenges of being a data firm is that, well, there’s too much data. Risk management teams have to oversee thousands of portfolios, each of which requires regular reports to PMs or clients. “Portfolio managers tend to focus on the largest funds or the flagship products, so there are gaps,” Mina said.
MSCI began to address this first by moving the storage of data to cloud provider Snowflake, where it or its clients could use AI tools to search the data for salient information and produce a narrative around portfolio performance and compliance.
This summer it is introducing genAI-layered services on top, one aimed at specific portfolios and the other a more aggregate tool for asset allocators. The first tool lets clients use natural-language questioning to source and analyze portfolio data to make better investment decisions. The second leverages MSCI’s own financial modeling and stress-testing capabilities to run economic scenarios across various asset classes.
Asset manager’s friend?
Mina says he hasn’t seen any asset managers using AI – generative or otherwise – to actually build investment models. But they are using to manage their processes more effectively, and to help with their own coding.
“Too often people in middle offices are doing data management, not risk management,” Mina said. “I hope these tools allow them to spend more time making decisions.”
He thinks genAI will encourage more change within asset management companies, which to date have not been fast movers when it comes to technological change. But competition (especially from low-fee ETFs), fee compression, and other challenges continue to force asset managers to scale and be more efficient. This has driven M&A in the industry; the biggest firms, unable to grow meaningfully by acquisition, are now undergoing bigger digital overhauls.
Mina believes genAI-infused products will have a big impact in two of MSCI’s product areas. One is within traditional asset management on climate-related investments. The other is the potential to bring more data-led services to private assets, such as benchmarking, risk management, performance attribution, and portfolio valuation.
But these changes will ripple throughout the entire industry, Mina said. “Everybody will use AI to make their products better.”
