Brunel uses AI in stewardship and doubles down on manager misalignment

Pioneering responsible investor Brunel Pension Partnership is using AI to improve stewardship evaluation and has spent much of the last year improving the misalignment of interests between asset owners and managers in relation to climate stewardship.

The team uses an AI-driven tool called a generative pre-trained transformer (GPT) to analyse and compare the voting guidelines of approximately 20 asset managers and owners, according to the fund’s recently published Responsible Investment and Stewardship Outcomes Report 2024.

The £30.8 billion ($38.9 billion) Brunel uses the insights to update its own voting guidelines, ensuring they are ahead of current practices and expectations.

“It’s about understanding the broader shifts in stewardship standards and ensuring our guidelines reflect these,” says Oliver Wright, responsible investment officer.

Brunel has also developed a quarterly report reconciliation tool which uses AI to assess implementation of its voting guidelines. It then uses the reports generated as a tool to engage with its service provider where it sees discrepancies.

“Given the importance of voting implementation, the ability to automatically verify this information has been invaluable,” Wright says.

“It ensures the reliability of our reports and significantly reduces the time and effort previously required for manual checks. This means that we can devote more resources to engaging with investee companies and other core stewardship activities.”

Brunel believes AI’s role in stewardship is only set to grow. As the technology advances, Wright says he expects new tools for more effective engagement with companies, improved monitoring of sustainability factors, and even predictive analytics for identifying potential governance risks to appear.

The future of stewardship will likely involve a greater integration of AI to not only streamline operational tasks but also to enhance the strategic aspects of work like focusing engagements to ensure meaningful outcomes.

“GPT has already proven to be an asset, and its ongoing development will undoubtedly open up further possibilities for enhancing our stewardship practices,” Wright says.

However, whilst AI brings substantial benefits, Wright warns of its risks. Like the potential for social bias in AI algorithms. Given that AI systems are trained on large datasets that may contain societally biased historical data, there’s a concern that these systems could replicate and even amplify existing societal biases.

Alongside this, there’s the challenge of ensuring that the fund’s reliance on AI doesn’t diminish the value of human judgement, and that data privacy and security are rigorously maintained. To mitigate these risks, the team regularly audits AI tools for bias, ensures transparency in operations, and maintains a balanced approach that combines the efficiency of AI with the nuanced understanding of experienced professionals.

Addressing misalignment

In another noteworthy trend, Brunel has spent the last year working with other asset owners to address the misalignment of interests between asset owners and managers in relation to climate stewardship.

The 2023 proxy season provided signs that some asset managers had failed to unequivocally challenge oil and gas companies that were backtracking on their climate commitments, the report says. This contrasted with the positions of large asset owners that shared the view that if climate related risks are not addressed through stewardship activities, this can translate into investment risks for their portfolios, affecting long-term beneficiary interests.

To address the discrepancy, Brunel entered “robust and constructive dialogue” with its managers, identifying how fund managers can be better supported in delivering asset owners’ climate stewardship strategies.

Brunel’s analysis of the misalignment, and conversations with other asset owners, was framed by research findings presented by independent academic, Professor Andreas Hoepner. Using the energy transition in the oil and gas industry as a test case, Professor Hoepner and his team have evaluated the voting records of select managers on oil and gas majors.

This research provided evidence of a misalignment. The full research which was released in November 2023 provided insights on misalignment trends and voting rationales. For example, only a select few asset managers publicly align their reasoning with asset owners.

Some asset managers perceive voting and ESG engagement as mutually exclusive, raising concerns about potential access loss to management if misaligned.

The review also found distinct engagement process types ranging from persistent, long-term engagement with considerable progress to “quick fix” and “jumping the bandwagon” styles, pointing to issues around consistency and a long-term approach to engagement.

The research also put forward a number of rationales for the gap, highlighting that further research is needed to explore these issues in greater detail.

These include cultural misalignment – namely the differences between UK based asset owners and non-UK based asset managers – and resource allocation misunderstanding – aka the potential misunderstanding of the importance of stewardship and voting, leading to insufficient resource allocation.

Other reasons include a misunderstanding fiduciary duty, particularly in terms of risk management related to climate change and financial conflicts of interest.

In the next phase of the project, asset owner participants will initiate one-to-one bilateral conversations with their managers on the basis of the research findings. The next phase will also look into how asset owners can articulate their views on climate stewardship.