As emerging technologies become more embedded in the world of corporate actions processing, new risks and challenges are moving to the fore, highlighting the importance of high-quality data. Ben Pumfrett, Director of Product Management, explains.
The move to modernize
Corporate actions processing—including securities-related events such as dividend payments, buybacks, stock splits, mergers, spin-offs and de-listings—has traditionally been a manual-intensive activity, creating additional costs and risk for asset managers and asset owners. This important activity is expected to come under significant pressure as markets move to shortened settlement windows. The increased volume and complexity of corporate actions is further compounding matters.
Corporate actions processing has traditionally been manual-intensive
Swift data reveals that corporate action volumes have doubled since the pandemic, while research by The ValueExchange found that inefficiencies in corporate actions processing are costing each market participant an average of between USD 3 million and USD 5 million annually.1 As a result, participants are on the hunt for opportunities to improve the efficiency and effectiveness of managing these transactions.
Leaning into digitization
Efforts to digitize corporate actions processing are now in full swing. Some providers are looking to automate through cloud-based technologies that facilitate cost savings, scalability and ease of system upgrades. Swift recently completed a successful pilot that employs blockchain to improve efficiency when communicating corporate actions to investors.2 And many believe that generative AI tools such as ChatGPT could transform corporate actions processing, helping to achieve true straight-through-processing (STP) by streamlining and standardizing data, automating communications and improving exception handling.3 For example, AI could be used to identify and resolve linkage issues between SEDOL, CUSIP and ISIN identifiers and reduce manual intervention and look ups.
AI has the potential to reshape the corporate actions space
Wider adoption of AI, along with other emerging technologies, has the potential to reshape the corporate actions space.
The importance of robust data
Effective use of these technologies requires access to high-quality data. If the data consumed by tools like generative AI is incomplete, inaccurate or less than real time, the outputs will be of similarly poor quality. It’s a case of “garbage in—garbage out!” Although generative AI can consume vast quantities of data, this data must be relevant and refined. Failure to get the basics right could result in corporate actions processing errors, potentially leading to suboptimal investment and operational decisions by investors. In other words, asset managers, owners and participants alike need to ensure that their data foundations are solid.
Effective use of emerging technologies requires high-quality data
And finally, highly-sensitive data requires strong security safeguards, especially in an environment of increasingly sophisticated cyberattacks. Shortcomings around data security can have serious repercussions, both reputationally and legally.
Balancing the risks and rewards
Emerging technologies, particularly generative AI, pose a number of challenges for asset servicers, who initially were skeptical about AI. Many banned it outright to avoid sharing proprietary data with publicly available tools such as ChatGPT. Not only was it difficult to see what data employees were putting into the AI platforms, but it was even harder to monitor the outputs.4
However, times are slowly beginning to change. Some asset servicers, conscious of generative AI’s potential benefits, are now working with technology companies that offer private AI “sandboxes” with the ability to incorporate multiple large language models (LLMs) from the likes of Open AI, Google and Microsoft.5 As the name suggests, these “sandboxes” can be isolated within a company and limited to a handful of users.6
Concerns remain around AI’s propensity to “hallucinate”
While progress has been made on addressing privacy issues, concerns remain around the technology’s propensity to “hallucinate”—present false information as fact. When it comes to generative AI, human guardrails continue to be an essential component of effective oversight and error prevention. This is especially true for some of the more complex or bespoke corporate actions, which will continue to require human intervention, irrespective of AI advances.
Developing an exceptions-based approach
RBC Investor Services is actively researching how Horizon 2 technologies, such as generative AI, could be used to achieve business objectives such as a more exceptions-based approach to the management of complex, sometimes unique corporate actions events. While being mindful of inherent risks and requirements for robust data, this work has the potential to deliver significant benefits to investors in the form of improved processing efficiency and effectiveness.