From Port of Felixstowe to BlackRock

Research in support of sacked workers turned up pension scheme reports suggesting US equities in a large BlackRock fund were not voted from 2014 to mid 2023, a second fund was also affected

This is a brief report on a voting failure at BlackRock. There’s a good news piece about it on IPE here for a wider overview. My own link to it is work I have done with the UK trade union Unite.

By way of background, Unite has been involved with a lengthy fight with Hong Kong listed CK Hutchison over the dismissal of four workers, including three reps, at the Port of Felixstowe (which is owned by CK Hutchison).

Unite believes that the treatment of the four members at the Port of Felixstowe represents a failure to observe international human rights principles, including those contained within the UN’s Universal Declaration of Human Rights and International Bill of Human Rights and the International Labour Organisation’s 1998 Declaration on Fundamental Principles and Rights at Work.

As such Unite has engaged with a number of CK Hutchison's investors to seek to resolve these issues and last year attended its AGM.

Unite had tried a number of times to get a response from BlackRock regarding its engagement with CK Hutchison. Capital IQ shows it is currently the company’s largest external shareholder. 

At the same time, related research that included collating information on pension fund stewardship activity turned up reports from a number of funds that stated that US equity holdings in the Aquila Life MSCI World Fund had not been voted for a number of years. 

Here are some excerpts from pension fund implementation statements that covered periods ending in 2023:

Fund 1

Due to reasons beyond the Trustee’s control, BlackRock were unable to exercise their delegated voting rights for US companies held in the Aquila Life MSCI World Fund during the reporting period. The Trustee’s advisor engaged with the manager to understand the issue, and to limit the possibility of this reoccurring in the future. The advisor will monitor progress by the manager to resolve this issue ahead of publishing the Fund’s next annual Implementation Statement.

Fund 2

On 6 June 2023 BlackRock, together with the custodian for the Aquila Life MSCI World Fund (‘the Fund’), learned that they had not been receiving proxy voting ballots from their custodian for US securities held in the Fund. As a result, no voting rights on US securities were exercised during the Scheme year.

Fund 3

BlackRock have informed the Trustee of an issue in relation to proxy voting on the equity fund in which the Scheme invests. The manager has confirmed that the issue related to votes not being made in relation to US securities, dating back to 2014. they have also confirmed that no significant votes were missed, and that the issue has not impacted the outcome of any vote (given the size of the positions held by the fund).

I understand that the error was first spotted by investment consultant Redington as a result of work with a pension fund client, and hence why reports of the failure started to appear in some pension fund implementation statements. Here’s a snip from Redington’s latest Sustainable Investment and Impact Report which does not mention BlackRock by name but does highlight the issue:

So a big hat-tip to Redington for spotting this in the first instance. I think this is the primary reason that the error has become public. It is also great that they sought to ensure that other funds that were not their clients were informed.

So it appears US equities in the Aquila Life MSCI World Fund were not voted from some point in 2014 to early June 2023. According to the fund factsheet the fund launched on 25 October 2010 so the failure to vote US companies encompasses the larger part of the fund’s history.

This must represent a significant proportion of that fund’s holdings. According to the current fund factsheet, the US allocation in this fund, which had about £3.6bn in assets at end Q3, is currently almost 72%. The top 10 holdings are all US stocks and account for 23.5% of the portfolio in total.

Whilst the geographic allocation will have undoubtedly moved around during the period, given the importance of the US this means that for 9-10 years a large proportion of the fund’s holdings by both number and market cap were not voted.

A couple of pension fund reports I have reviewed provide corrected voting data. So where previously the ‘% of eligible resolutions voted’ figure relating to the Aquila Life MSCI World Fund for the year ended 31/3/2023 was reported as 88%, the ‘corrected’ figure taking account of non-voted US stocks is reported as 55%.

There is also an example of a pension fund that invested via the fund reporting ‘significant votes’ at a US company during the affected period. Here are a couple of snips from the Marubeni UK Pension and Life Assurance Fund implementation statement for the year ending March 2021.

(Note the allocation to the BlackRock funds is listed as £0m because the fund reallocated money during the period.)

Given what has subsequently been revealed regarding the voting failure, it’s possible that actually no votes were cast by at the company by that specific BlackRock fund.

Notably a second BlackRock fund, the Aquila Life Global Developed Fundamental Weight Index Fund, was also affected by the same problem. In this case one of the investors in the fund is the pension scheme of another trade union, the NUJ, so Unite has ensured that the fund is aware the issue.

It’s clear from some of the implementation statements that some trustees and advisers are aware of the problem. Half a dozen or so implementation statements report on the issue. But there are also numerous pension funds that report voting data relating to one or other of the funds that seems to use a ‘% of eligible resolutions voted’ number for periods that included the voting failure that do not appear to include US equities. Most of the implementation statements I have read for pension schemes investing via either of the funds (in a few cases both) that cover the affected period do not report the voting failure.

To bring it back to the issue that started the research, the duration of this voting failure will have included meetings of companies where there was a resolution on the ballot relating to workers’ freedom of association and collective bargaining rights.

The period also encompasses some other really notable votes at US companies, like the first vote on Elon Musk’s pay, the Engine No. 1 campaign for board seats at Exxon and so on. BlackRock has stated that the error will not have affected any vote outcomes.

Coincidentally, at this year’s BlackRock AGM there was a shareholder proposal on the ballot from Mercy Investment Services. Here is the resolved clause:

Resolved: Shareowners request that the Board of Directors initiate a review of both BlackRock’s 2023 proxy voting record and proxy voting policies related to climate change, prepared at reasonable cost, omitting proprietary information.

The 2023 proxy record would include the period up to early June 2023 when the two funds were not voting US equities. BlackRock unsurprisingly recommended that shareholders oppose the proposal. Here’s a snip from the concluding paragraph in its opposition statement in the DEF14A:

given the Board’s existing oversight and the transparency that BIS already provides regarding the team’s voting and stewardship approach, a review of BlackRock’s proxy voting guidelines and record would not yield meaningful new information for shareholders.

To be crystal clear: this appears to have affected two funds only. Inevitably in an organisation of BlackRock’s scale things will go wrong somewhere, sometime. That said, it looks to me to be a meaningful stewardship failure. The exercise of voting rights is a core part of investors’ stewardship activity. Failing to exercise those rights for a large proportion of holdings over a prolonged period is obviously not good.

In one sense the system has worked. Because pension funds are required to produce implementation statements voting data was provided. When the data was checked by an adviser the error was identified. Once the error was identified it was immediately rectified by BlackRock.

On the other hand, this had been going on for almost ten years at the world’s largest asset manager. It also means there is a lot of wonky voting level reporting out there. Numerous pension funds have produced implementation statements, sometimes over several years, that are unintentionally misleading because the voting data they have been provided with was incomplete. It is not the fault of the pension funds that the data is wrong but the reports are misleading as they stand.

Finally, and most importantly, the four guys that lost their jobs at Port of Felixstowe still deserve justice. Investors who hold stock in CK Hutchison should engage with the company as the owner of the port to help them achieve this. We don’t need any more stewardship failures.