Schroders has seen success through utilizing alternative data in its own decision-making, particularly in tools such as geo-spatial analysis, natural language processing and web scraping. Company analysts met with a Brazilian pharmacy chain that claimed it would double its store footprint over the next five years. Schroders’ data scientists built a model that examined all the locations and service areas of the client’s current stores, as well as those of its competitors. It then examined the various service areas relative to demographic data, applied a number of growth scenarios that allowed the analyst to get real-world data that challenged management’s assumptions.
Natural language programming allows Schroders to effectively look at patent databases, said Chisholm, noting it would be near impossible for an analyst to mine through one patent database, let alone databases in multiple countries. But he says large language models can map those linguistic similarities in a way that quickly identifies the most meaningful patents. “And when you look at all the intangible value on balance sheets today and understanding where that intellectual capital sits, that’s another really powerful tool in the toolkit for analysts to work with.”
The investment manager also uses web scraping to comb through e-commerce platforms, track data around user engagement statistics and get an idea whether clients’ investment theses are panning out in real time. This tactic has been helpful in monitoring industry pricing trends and pricing discipline, said Chisholm. Indeed, when the firm held a position in Bridgestone Corp. a number of years ago, the investment thesis was predicated on its ability to deliver better margin progression, relative to its peer group. He said Schroders was concerned about the company’s lack of execution and margin progression.
Read: 2023 Defined Benefit Investment Forum: The value of active management in emerging markets
After the management team told investors it was losing market share in the ASEAN region because other competitors were pricing more aggressively, Schroders’ investment insights unit and global equities team used real-time web data to gauge industry-pricing discipline, which showed the opposite — that it was actually the most aggressive pricer, but it still wasn’t gaining market share from this strategy. “It really raised questions about the credibility of that management team in addition to the execution issue.”
Schroders sold their position and the company continued to underperform for a period of time, leading to many members of that management team being subsequently replaced and setting up for an interesting restructuring story.
Chisholm noted Schroders uses generative artificial intelligence to validate and test risk models, adding that it can efficiently analyze transcripts, detect language patterns and identify valuable insights. Several of the firm’s teams are starting to experiment with AI-generated stock-screening tools, which use algorithms similar to those used by Netflix Inc. or Amazon.com Inc. to make recommendations for their viewers based on stock characteristics that resemble some of their more successful performing stock picks at the time of purchase.
Read: When does active management make sense for large pension funds?
But ensuring that a prospective company is sustainable is critical, said Chisholm, noting great companies have good return on their invested capital, well above their cost of capital, as well as great re-investment opportunities, well-managed franchises with competitive advantages and an innovative culture. But understanding how those companies manage their stakeholder relationships is critical to understanding the durability of their earnings power and ultimately is what makes those ‘great’ companies stay ‘great.’
Effectively incorporating non-financial factors and alternative data into the research process provides a far more robust lens for an analyst and can help fill in many of those blind spots where applicable or material. “If you can just find four different ways in the course of a year to get a decision incrementally more correct through these types of tools, it obviously gives you a very powerful lever on your alpha generation.”
Read more coverage of the 2024 Global Investment Conference.
Read More: 2024 Global Investment Conference: How institutional investors can re-think