How AI is unlocking the next wave of efficiency gains in Asset Management

Operational efficiency in asset management has made considerable strides over the last decade through increased automation of standardized, repeatable processes. But as this low-hanging fruit has been harvested, the gains from further investment have become incremental – until now. Predictive and generative AI are laying the ground for the next big leap in operational productivity.

The industry's automation journey

Linedata’s 2023 Global Asset Management Survey found that 57% of asset managers consider increasing automation to be their top strategy to enhance operational efficiency in the upcoming year. This comes as no surprise, considering the industry’s persistent margin pressures.  The focus on operational efficiency is long-running trend,  cited as a major priority in our 2019 and 2021 Global Asset Management Surveys.

Technology has always been a driver of transformation, innovation, and operational efficiency for asset managers. In truth, the process of digital transformation is not something that has an end. It is a constant evolution and asset managers’ lasting focus reflects this. But achieving and benefiting from efficiency gains in investment operations is not a linear progression.

Initially, automation software simplified repetitive tasks, such as data entry, reconciliation, and reporting. These are examples of rules-based solutions that rely on conditional “If This Then That” programming.   As the industry advanced towards exception-based workflows and straight-through processing, the rate of efficiency gains slowed, despite the consistent and growing need for greater efficiency caused by increased competition, declining fees, and rising costs.

Now, with the emergence of relatively accessible generative AI and large language models (LLMs), the industry can explore efficiency opportunities in areas that had previously been overlooked due to their complexity.

AI Unlocks New Doors to Operational Efficiency for Asset Managers

The term “generative AI” is broad, with many applications. At this stage, we see it delivering the biggest operational efficiency gains in augmenting the work of humans undertaking complex tasks. AI tools can automate data extraction from volumes of unstructured and structured data sources to provide insights and recommendations for better and faster decision making.

A prime example of its application is the consumption of large sets of documentation.  LLMs can parse documents, identify specific details, and understand their contextual relevance to the issue at hand, saving time and resources. At Linedata, our generative AI use cases digest large repositories of documents on behalf of our clients and give them insights into how applicable that knowledge is to a particular business issue.

Investment compliance is one area of an asset manager’s operations where generative AI is making a difference. For instance, when setting up an investment instrument with bespoke terms, compliance teams can leverage LLMs to swiftly comprehend the pertinent legal requirements to put the right monitoring in place.

Another area where AI is unlocking greater efficiencies is predictive analytics. While generative AI can be thought of as augmenting the work of humans, predictive AI tools and insights can be applied as a means of changing behaviors or processes before an error or risk incident occurs – to shift the way people think about operational problems.

For example, predictive AI tools can help in trade operations by analyzing vast amounts of data to enable operations teams to better identify anomalies – and take actions ahead of time to prevent potential settlement errors or other incidents.

The use cases and relevant context for generative AI and predictive analytics are different, but in both cases, these tools have the potential to free up staff for higher value-adding activities, whether that be delivering greater transparency to end clients or developing new investment propositions. By deploying both types of AI solutions effectively, asset managers will be positioned to drive improvements to both top and bottom line.

The key to successful AI deployment

To get the most out of AI, however, managers need to have certain foundations in place. AI tools and models need access to relevant and comprehensive information to be trained properly and to produce insights which are relevant to the business. Well-maintained data repositories and maintenance of models are critical to this.

It is also important that AI solutions operate within the right context. This means putting effective guardrails in place around the data that is being accessed and the tasks that tools are permitted to execute. And, of course, AI tools still need humans to provide oversight, evaluation, and context to ensure they are producing useful outputs.

Asset managers will need to consider how AI tools are brought into their technology ecosystems too. Applications should be built in a modular fashion, with APIs used to integrate AI solutions, so that business users can access them as part of existing workflows to augment what they are doing – without stepping outside of the core ecosystem. As well as ensuring ease of access, this should help to engender trust with these nascent tools and create new and continuous opportunity for operational efficiency.