Investment managers stand at a crossroads today. Faced with a rapidly changing digital world, they must determine which path to take to help them transform their business models and respond to the needs of a younger generation of investors.
One key aspect to this transformation story is how investment managers successfully integrate people, processes and technology, as part of a sophisticated digital strategy, to adapt to the digital age. Data, and the way that is ingested, processed, stored and managed, is one of the main pressure points to overcome.
As SEI points out in its recent white paper, Evolution in Asset Management, “virtually everyone is familiar with the potential value of data, but it is still not treated like a precious commodity by many firms. Through force of habit, data acquisition, integration, management, protection, analysis and disposal still often occur in an ad hoc way”.
Colleen Ruane is Director of Analytics at SEI Investment Manager Services. When asked if she thinks the penny is beginning to drop with respect to firms using transformational technologies such as natural language processing and other AI-driven tools, she responds in the affirmative: “At a minimum, firms are beginning to lay down the groundwork from which to do analytics. Based on what we are hearing in conversations with clients, there’s a heavy emphasis on data in terms of security, management and governance across the board. The relationship that investment managers have with data is definitely changing.”
That being said, there is still a fair degree of immaturity around the use of AI and machine learning. A lot of effort is being made but there appears to be no consensus as to particular direction. A lot of research and data is being applied to how fund managers think about transforming their investments, but there is still a long way to get to a place where the industry, at large, is fully mature at using data-focused technologies in a more joined up fashion.
Front-office focus among US managers
In SEI’s white paper, reference is made to how fund managers are prioritising the deployment of advanced analytics. Interestingly, some 70 per cent of US managers are focusing on the front-office, to improve portfolio management and marketing/fund raising activities. This compares to approximately 50 per cent for European managers and 40 per cent for Asia Pacific fund managers.
“Whether the US is leaving a little bit on the table in terms of applying data analytics to the back-office or firms have already dedicated resources to RPA, we’ve seen efforts have largely been made to implement predictive analytics and AI to the front-office,” says Ruane. “There are lots of technology start-ups building plug-and-play products, which are enabling asset managers to implement advanced analytics and AI in the portfolio management end of the spectrum.”
In her view, the competitive landscape is fuelling a lot of this interest around enhancing the front-office experience. “Portfolio performance is still key, but it isn’t the only thing that attracts and retains investors. Everybody is competing to build the best client experience, and that means investing in some of these newer transformational technologies,” she adds.
Understanding what good data management means to firms is helping to focus minds and come up with ways to overhaul legacy systems that were not built, or designed, to handle the volume or sheer variety of data sets – both structured and unstructured – being used today.
A more holistic approach to data within a system architecture that supports AI technology tools is the ultimate goal but that does not mean that existing systems need necessarily be overhauled completely.
“There are different ways to solve for the disaggregation or segmentation of data sets,” says Ruane. “Data lakes are still an option, the cloud enables much more openness around the data landscape, while APIs and extraction layers can be built on top of legacy systems without having to move data into too many different places.
“However, doing it well is not easy and requires careful planning. As much as structured data may be fragmented in some fund management firms, unstructured data is much more fragmented. So making that data available – organising it, categorising it etc – so that these AI systems can learn from it, is no small effort.”
Indeed, the pace of transformation has arguably been slower in the financial industry in part because of this complex, fragmented data landscape in which they operate.
No silver bullet
It is easy to be seduced by the potential of new technology advances but investment managers should take care to understand exactly how far they can incorporate transformational change, especially in the front-office. Some of the larger quant shops are embracing AI but for the average fund, being able to clearly explain to investors how the portfolio makes money remains vital; AI has its uses but we are a long way from suggesting the algorithms will take over.
“There are myriad of different ways that firms are thinking of using AI, in some capacity,” says Ruane. “Some of those applications might have more to do with making their portfolio managers and analysts more efficient; some might have to do with searching for more relevant information to facilitate better investment decision-making; some might have to do with better aggregating data to allow them to be more readily consumed.
“The role of the PM and analyst is certainly not about to be outsourced to an algorithm. But the way they use algorithms and AI-based technologies is going to change. Everything is becoming more data-driven. Processes that were previously driven by large amounts of research are, I think, now being supported by a lot more data. This allows for PMs to uncover nuggets of opportunities they might have previously missed and in a much more efficient manner.”
Having an algorithm sift through enormous amounts of unstructured data and reduce it to a set of relevant information will become increasingly important in front-office portfolio management teams, as may NLP applications such as the ‘research assistant’.
Culture is key
For this to work in practice, regardless of how far a firm embraces technology transformation, it will be incumbent upon them to have the right people in place.
As SEI’s white paper states, “Unlocking this potential requires not only the right algorithms, but also the right people. This means more data specialists. Their contributions will not be maximised if they are set apart in dedicated data teams.”
Going forward, this could mean that a bigger focus is placed on ensuring fund management teams are AI literate to lead that transformation…building the analytics and the AI framework to do a lot of the heavy lifting. This will require a cultural change.
As Ruane states: “Even if you hire the best data scientist or CTO, if the firm as a whole isn’t willing to change their processes or ways of doing business, it’s not going to work out or at the very least will be sub-optimal. The right environment will be needed to attract and grow talent, as well as support the growth and development of the existing workforce. Fostering adoption across the enterprise will be key.”
SEI further points out that “a growing number of companies are taking the additional step of hiring high-profile senior data scientists, ensuring visibility and representation in the C-suite so data initiatives are considered strategically and prioritised accordingly”.
Looking forward, Ruane believes that for transformation to succeed, investment firms will need to have a sound foundation for how they store, manage and access/aggregate different data sets. “That is going to be really important in terms of how AI tools make sense of data.
“It’s a classic People, Process and Technology problem. You don’t necessarily need everything to be perfect all at once but firms will need to think about transformational change in respect to all three aspects to succeed.”