Even well-designed Remuneration Policy proposals can attract significant shareholder dissent. Across the 2024 to 2026 AGM seasons, nearly a quarter of Remuneration Policy resolutions received less than 90% support and 11% fell below the 80% ‘low vote’ threshold. Yet, assessments of likely investor reactions are often primarily qualitative: they consider how investors may view particular design choices and the influence of proxy advisers such as ISS, but do not quantify the likely voting impact in the context of the company’s entire shareholder base.
This can make it difficult to distinguish between an issue that may prompt concern and one that could materially affect the outcome. A more quantitative, company-specific assessment can help translate investor attitudes, voting history and proxy adviser influence into an evidence-based estimate of likely support, while proposals are still evolving and changes are easier to make. This Lens explains how Ellason’s Investor Voting Probability Model adds a quantitative layer to investor analysis, supporting Remuneration Committees, Investor Relations and HR teams throughout the policy cycle, from initial design to shareholder engagement and AGM preparation.
The model
The Investor Voting Probability Model uses multiple reference points to estimate the likely voting outcome for a company’s remuneration proposals. The process of running the model includes:
Map the shareholder base - use the company’s full share register to identify investors eligible and likely to vote, providing a complete picture of the shareholders behind the resolution.
Assess investor independence - assess each investor’s decision-making independence, including the extent to which they typically follow a proxy adviser.
Analyse voting behaviour - examine the company’s voting history to identify patterns in past investor behaviour, then benchmark the remuneration proposal against comparable resolutions, considering how and why investors voted and what influenced their decisions.
Estimate the likely outcome – taken together, these inputs produce an investor-by-investor assessment that can be aggregated into a company-specific probability vote estimate across the entire shareholder base.
We have demonstrated the model against ten FTSE350 companies from the 2026 AGM season. The sample included companies that received either a ‘For’ or an ‘Against’ recommendation from ISS. It also covered policy changes ranging from limited to significant, including uplifts to variable pay and the adoption of hybrid LTI vehicles by both US-focused and non-US-focused companies.
How to read the chart: the grey dotted horizontal line represents each company’s actual vote on its Remuneration Policy at the 2026 AGM, rebased to 0%. Each green diamond shows how far the model’s prediction differed from that result. For example, if a company received 90% support and the model predicted 92%, the green diamond would appear two percentage points above the grey line.
Across the companies analysed, the model’s predictions were, on average, within three percentage points of the actual voting outcome, with individual differences ranging from zero to five percentage points.
Takeaways
The Ellason Investor Voting Probability Model gives companies an early, evidence-based view of how investors are likely to respond to specific policy choices and where material risks to support may arise. By translating investor behaviour into a quantitative assessment tailored to the company’s shareholder base, it enables remuneration committees and management teams to act before positions become fixed, thereby refining proposals, informing the shareholder engagement strategy and building stronger support. Its output identifies shareholders to prioritise for engagement and a data-led estimate of the likely voting outcome. A qualitative judgement remains essential but combining it with quantitative analysis provides a clearer basis for decision-making, more focused engagement and greater confidence heading into the AGM.
Please do not hesitate to contact any of the Ellason team should you wish to discuss this topic further.