Groundbreaking ideas and research for engaged leaders
Rotman Insights Hub | University of Toronto - Rotman School of Management Groundbreaking ideas and research for engaged leaders
Rotman Insights Hub | University of Toronto - Rotman School of Management

To sway corporate governance, institutional investors need to use their voice, not just their vote

Read time:

Irene Yi

Every vote counts — even in matters of organizational governance.

Institutional investors — firms that invest money on behalf of others, which include pension fund managers and investment banks — control a very large number of the shares in publicly traded corporations, and that means they have sway in shareholder votes. But the extent to which they use that influence to shape corporate governance has been a matter of some debate. Research by Irene Yi, an assistant professor at the Rotman School of Management, shows that institutional investors do use their voice to effect change — and that it yields results.

Working with Rony Michaely of the University of Hong Kong and Silvina Rubio of the University of Bristol, Yi’s analysis looked at a large data set of institutional investor voting rationales for board of director elections, which showed that these bodies use their vote to shape corporate governance in a variety of ways. The research looked at more than 600,000 voting rationales on director elections provided by institutional investors globally, which showed that institutional investors sought to address a number of concerns, including improved board diversity, limiting director tenure and reducing busyness (that is, how many boards a director sits on).

The most consequential finding occurs when institutional investors vote “no” and provide reasoning for that decision. With more than 70 per cent of outstanding shares in the United States managed by institutional investors, a threat to divest their fund from a given equity could have considerable implications for its stock price.

“Diversity was one of the main reasons institutional investors cited when voting against director candidates,” says Yi. “It was the second most cited concern in voting rationales, and this has not previously been recognized as a major motivation for votes against directors.”

The study also sought to understand whether institutional investors’ concerns actually reflected a company’s weaknesses. Researchers looked at director tenure, busyness and gender diversity.

“We found that when institutional investors cited these concerns in their voting rationales, they were well grounded. For example, gender diversity was more likely to be cited in a voting rationale when a company lacked women on its board,” says Yi. “There has been debate as to whether institutional investors seek to influence corporate governance with their votes. This is evidence that they do care about it, and that they seek to influence governance with their votes.”

The data set this research analyzed was extensive, and to understand this information at such a large scale, Yi used a supervised machine learning technique called bidirectional encoder representations from transformers — or BERT for short. Yi categorized more than 1,000 of the voting rationales manually. This was used to train the machine learning model, which classified the rest of the data with an accuracy that exceeded 95 per cent.

But the data was not a representative sample of institutional investors. Voting rationales are voluntarily disclosed, and regulations recommending disclosure vary across countries. Environmental, social, governance (ESG) investors are much more likely to disclose this information voluntarily, and European investors are more likely to disclose a rationale for their votes than American ones.

So, Yi adjusted for this skewing with a technique called propensity score weighting. For example, U.S. institution investors are underrepresented, so the authors gave them additional weight to account for the discrepancy. And while this adjustment did change the numbers, it did not change the conclusions. Diversity and board independence were still the most important factors cited in voting rationales.

It’s worth noting, however, that votes alone do not necessarily sway corporate governance. Yi tested whether organizations saw a change to director makeup based on gender in the years following a significant dissent over director elections, determining that it wasn’t statistically significant. The authors concluded that companies are better able to address investor’s concerned when they are explicitly voiced in a voting rationale.

“The effect of rationales go above and beyond vote. We found that when institutional investors voiced their concern, the companies did make changes. They listened to investors’ concerns by adding women to their board, reduced average director busyness, and tenure,” says Yi. “So, companies are willing to listen to investors. And we argue that providing a voting rationale can be a meaningful channel to bring about changes to corporate governance.”


Irene Yi is an assistant professor of finance at the Rotman School of Management.