Active fund managers, including hedge funds, are under increasing pressure to back up their investment ideas and performance with data. But embracing data science can help fund managers justify their fees, in the face of relentless competitive from passive managers, says Gary Paulin, U.K.-based head of global strategic solutions at Northern Trust.
The custodian bank recently surveyed 300 fund management CIOs and portfolio managers around the world and found traditional managers are increasingly combining their qualitative methods with data analytics.
The survey found 58 percent of managers are already exploiting data sets to identify risks and opportunities, 56 percent are using it to automatically execute trades based on signals or trends, and 51 percent are tapping into data to inform or guide an investment theory.
Paulin says the asset-management industry is changing because its leading institutional clients are incorporating data science into their own decisions around manager selection.
“Asset owners are employing sophisticated data analytics to distill the investment skills of managers,” he told DigFin.
This isn’t new, exactly. Businesses always use data to make decisions. But the advance in analytics and the availability of affordable software-as-a-service packages means the nature of these inquiries is changing.
Until recently, even the biggest institutional investors would weigh asset managers on backward-looking performance numbers. They would consider a portfolio manager’s top weightings, the way she sizes her holdings, and intangibles like the team’s experience.
- Read more:
- Investors learning to make Asia’s ESG data more useful
- This is how UBS uses alternative data
- Northern Trust takes custody of blockchain bonds
A manager might be punished for “style drift” – chasing investments against her own strategy – but usually such deviations happened only after the manager secured a client’s mandate.
Such processes mean asset owners typically get managers wrong, investing on the back of a good (or lucky) run and firing managers just before they’re about to perform well.
Fund managers, in turn, have often relied more on marketing, star managers, or gut feel than on a process they can argue and defend. Many active funds end up “benchmark hugging”, charging active-management fees for portfolios that are too similar to the indices that an index or ETF fund will mirror.
Active managers have therefore been forced to cut fees, cut costs, or combine.
Paulin says the growing importance of data analytics should benefit those managers that can invest in tech and data resources – and who also develop a successful, evidence-based process they can show clients.
The flip side is that mediocre performance is going to stand out too.
“You have to show that your decisions are made rationally, and that they are working,” Paulin said.
Double down on investment process
Pressure to perform is old hat, even if data analytics make it easier to measure. But data is also transforming what asset owners are looking at when they evaluate managers.
“This will evolve to the point that asset owners will want to rank managers by skill, by ESG factors, by cybersecurity,” Paulin said. “That’s where the industry is heading, and this trend will accelerate over the next three-to-five years.”
Those firms that can codify their investment process can then make it scalable – and easier to pitch. They can use data to shift from relying on post-mortems on trades to better plan for portfolio transitions.
This requires sticking to an investment process and ignoring the short-term temptations of style drift. It also means, however, that investors can be more confident about building focused portfolios built around conviction, instead of hovering around a benchmark.
You have to show that your decisions are made rationallyGary Paulin, Northern Trust
“High-conviction ideas usually perform, but the alpha gets eroded by the rest of the stocks in a portfolio that are there to justify a manager’s fee,” Paulin said. “Hedge fund managers win mandates by giving asset owners their top five ideas – but at the end of the year, their fund underperforms. Data science can help solve some of these issues by making asset managers more aware of process, and better at matching their conviction to position sizing.”
He says the ESG space is ripe for data science, given its inherently diverse, qualitative nature. Data science will help managers develop ESG rules, and keep their portfolios on track. This will also ease reporting to clients and to regulators.
Ultimately data can be used to help active fund managers grapple with an age-old danger: letting emotions get in the way of investing. Paulin says that those asset management firms that succeed in incorporating data science into their investment processes will stand the best chance of defending their fees against the relentless competition from low-cost passive players.