How to Drive Effective AI Adoption in Investment Management Firms

By Chandini Jain, CEO of Auquan

Artificial intelligence (AI) has the potential to augment the work of investment management firms to unprecedented levels, powering decision-making, driving efficiencies, and ultimately improving performance. In fact, the market for AI in asset management is expected to grow to an astounding US$13.43 billion by 2027, expanding at a CAGR of 37.1% between 2020 and 2027. Innovative firms are applying AI across the industry value chain and transforming the ways in which they use the ever-expanding amounts of data that are available to them.

However, that’s not to say that there aren’t challenges and obstacles involved in leveraging the technology. AI adoption is not a ‘magic bullet’ that can solve inefficiencies without the right set-up, nor should it be treated as a simple ‘add-on’ that portfolio managers (PMs) can tap into when they see fit. AI implementation in an investment management firm requires a number of prerequisites in order to have maximum impact. But first, let’s take a look at exactly how AI can boost the performance of investment management firms.

How AI adds value

Implementing data analytics into the investment management value chain holds a number of benefits. For example, when it comes to front office operations, AI can supplement investment decisions by drawing insights from alternative sources of data such as satellite imagery or social media, while also automating the analysis of large datasets. Data science teams working within investment management can build simulations to allow PMs to predict the performance of new investment ideas. They can also use AI for trading – to optimize trade execution and automate trading decisions.

One example of using AI to power alpha generation comes from Man Group, which saw a five times increase in assets between 2014 and 2018, and whose funds that incorporate AI total more than US$12 billion. Front office operations are arguably the business area where AI holds the most potential.

When it comes to distribution and marketing, AI can improve prospect and sales targeting using segmentation, predict and reduce attrition, support personalization, and help develop pricing algorithms. Data analytics can also be implemented into the areas of operations, tech, and support to automate processes, improve talent targeting, predict team member performance, and strengthen compliance, amongst other uses.

Going beyond simply reducing costs and driving efficiencies, AI is providing new opportunities for investment management firms to transform how they use data to operate and inform decisions. But despite all of this, adoption levels are still relatively low: A 2019 survey by the CFA Institute found that only 10% of PMs responding had used machine learning (ML) techniques during the year prior. Furthermore, a 2019 report by BCG found that less than 30% of asset management firms are actively leveraging data analytics. Evidently, launching an AI project is not an overnight process – nor is it one that guarantees success without the right prerequisites in place.

Here’s how investment management firms can set themselves up for success and ensure readiness for AI implementation.

Embed a data culture 

Before steaming ahead with any AI project, investment management firms need to ensure that the entire organization appreciates the value of data-driven decision making. A firm may have already hired a data science team or gained access to alternative data sets, but if it doesn’t have a culture of systematic decision making that permeates across the organization, the success of any AI project will be limited.

How can firms ensure that this is the case?

Ultimately, building data-driven must start at the top: the CEO, CIO, and all other executives must lead by example and evidence of their own commitment to data-based decisions. If leaders want their teams to leverage data at all points of decision-making, they must make the data accessible for non-technical employees and provide training on how to use any relevant tools. Teams must feel comfortable with the why of data analytics solutions, so management must make them explainable while ensuring they are aware of the capabilities and limitations of AI. And finally, the data science team must avoid working in a silo, away from the other business functions of the firm.

Reconfigure the team structure

The core investment process must be re-thought, from the ground up. Data science teams must be driven by a business need which is provided by the PM, and then the two must work together to co-develop the right solution.

In addition to having a centralized data science team, the firm should have decentralized data scientists that sit within the business unit. The central team should focus primarily on data acquisition, cleaning, and ensuring reliability. The rest of the work should be done by data scientists on the PMs team – this will ensure the work is in-line with the business needs and will actually be used by the PM. With the clean, reliable data coming from the data acquisition team, the data scientists can rapidly prototype ideas for the PM.

Invest in the right software

Too many investment management firms attempt to build all of their AI software in-house. While the software that’s required for core operations and stems from core finance expertise should be developed internally, this does not apply to all other solutions being used.

For example, data analysis and automation tools that leverage ML domains such as language processing, big data processing, or image processing should not be built in-house. Constructing these systems internally is expensive, time-consuming, and means hiring for skills that would otherwise not be required within the firm. Not to mention, such systems would need a large and active development force to continuously maintain them.

That’s why it’s advisable for firms to find a third-party vendor who can take care of building the feature set that’s required, update the software with its latest version, and scale according to needs. This vendor will also take measures to ensure that the firm’s standards are consistent with its peers, and importantly, keep the system stable and secure. By integrating with a third party vendor, data science teams can focus on the core business objectives and maximize the use of overall resources.

While AI offers countless opportunities for investment management firms to augment and power decision-making and is already setting apart the top-performing firms from those that lag behind in adoption. With so much potential to enhance portfolio performance, AI adoption should be viewed as non-negotiable for forward-looking and innovative firms. It is paramount, however, that these firms embed a data-driven approach across all teams – not just PMs – and provide the structures and tools necessary for results to flourish.