Author’s note. This commentary was originally intended for publication after the BP AGM. But in working through what each possible outcome would signify, it became clear that the governance lessons remain largely the same whichever way the vote goes. What matters has already happened. I have therefore decided to publish before the result is known — as a lessons-learned from the process, with a framework for reading the four resolutions on the day.
Table of Contents
I. What’s at stake
On 23 April 2026, BP shareholders vote on four contested resolutions at what is likely the most consequential AGM in the UK energy sector for a decade. Three are board-backed: the re-election of Chair Albert Manifold (Res. 4), amendment of the articles to permit virtual-only AGMs (Res. 22), and the retirement of the 2015 and 2019 climate commitments currently embedded in BP’s articles of association (Res. 23). The fourth, Resolution 24, is a shareholder resolution filed by the Australasian Centre for Corporate Responsibility (ACCR) requiring enhanced disclosure on new oil and gas capex — which the board has recommended voting against.
The four contested resolutions at a glance
| Resolution | What it proposes | Board position |
| Res. 4: Chair re-election | Re-elect Albert Manifold, appointed Chair following Helge Lund’s departure. | FOR |
| Res. 22: Virtual AGM articles | Amend BP’s articles to allow virtual-only AGMs. | FOR |
| Res. 23: Retire climate mandates | Remove the 2015 and 2019 climate commitments from the articles. | FOR |
| Res. 24: ACCR capex disclosure | Shareholder resolution requiring disclosure on how new oil and gas capex aligns with transition scenarios. | AGAINST |
The governance lessons themselves, however, do not depend on the 23 April outcome. They are already visible in what has transpired since February 2025: a strategic reversal under pressure from Elliott Investment Management, a CEO and Chair change, the first FTSE 100 exclusion of a Follow This shareholder resolution, a letter before action from Mishcon de Reya, and a full reversal of position by both principal proxy advisors. What the vote will reveal is how institutional investors read those events — not whether they occurred.
One acknowledgement upfront. When Elliott — the activist hedge fund that took an equity stake in early 2025 and pressed BP for a series of strategic and governance changes — arrived, BP was not a healthy company under attack. The board’s commitment to its own energy transition had been quietly eroding for at least two years; the 2030 emissions targets were softened in 2023 without a shareholder vote. Elliott did not create the drift — it read it accurately, took a concentrated stake to exploit it, and forced a question the board had been unwilling to put to its shareholders. That drift was itself the first governance failure.
II. Four governance lessons
1. Sequencing
Backstory. In February 2025, under direct pressure from Elliott, BP reversed its energy transition strategy, pivoting back to upstream oil and gas reinvestment. Over the months that followed, the CEO and Chair were both replaced and capital allocation priorities restructured. None of this was put to shareholders in advance. A full year later, Resolution 23 asks shareholders to retire the 2015 and 2019 climate disclosure commitments embedded in BP’s articles of association — commitments that shareholders themselves previously approved as special resolutions.
Governance issue. Boards must be able to change strategy when conditions change — that is not what is at stake. The question is how a strategic change is communicated, and how shareholder consent is built before the new course is set. When the board failed to engage the wider base that had previously approved its original direction, it allowed an appearance — and perhaps the reality — that the wishes of one concentrated owner had superseded those of the rest. That is the governance tension: not the fact of adjustment, but the process by which it was made.
Lesson. Resolution 23 is not a request for permission to change course. It is a request for approval of a course already taken. A board that changes strategy first and seeks approval afterwards has inverted the accountability relationship. The proper sequence — proposal, engagement, mandate, execution — exists precisely to prevent boards from presenting shareholders with a done deal dressed up as a democratic choice.
Section 172 of the Companies Act 2006 requires directors to promote the success of the company for the benefit of shareholders as a whole, while having regard to long-term consequences and to the interests of employees, the environment, business relationships, and other stakeholders. That duty sits uneasily with a sequence in which a board reverses strategy under pressure from a single concentrated shareholder and only later seeks ratification from the wider shareholder base.
When the order is reversed, the vote becomes ratification, not authorisation.
2. Procedural gatekeeping
Backstory. The Follow This shareholder resolution, backed by 23 institutional co-filers representing $1.74 trillion in assets, asked BP to disclose its long-term strategy under scenarios of declining oil and gas demand. BP confirmed in January 2026 that the resolution had met the statutory threshold for submission. When the AGM notice was published in March, the resolution was absent — the first time a FTSE 100 company had excluded a Follow This resolution. BP advanced two grounds for the exclusion: that the resolution had not been designated as a special resolution, and that its use of the word request rather than a directive form rendered its effect upon passage uncertain. Both grounds were novel. Follow This had filed resolutions under the same procedure at BP and at Shell across multiple prior AGM cycles, and both boards had tabled every one without objection. The objection surfaced only once the substantive content of the 2026 resolution had become inconvenient. Shell, faced with a virtually identical resolution for its 19 May AGM, accepted it. The exclusion triggered Glass Lewis’s recommendation against Chair Manifold’s re-election, public opposition from Legal & General and LAPFF, and a letter before action from the law firm Mishcon de Reya.
Governance issue. Exclusion is not, by itself, a governance failure. Boards can legitimately refuse to circulate resolutions that are outside their powers, inconsistent with the articles, or already rejected at a recent AGM. The governance tension arises when eligibility arguments become a tool to prevent shareholders from voting on resolutions that are inconvenient rather than invalid. Where that line falls is rarely visible from outside the boardroom — except when a peer board, faced with the same question, lands on the other side of it. Then the issue is no longer whether the law permitted exclusion, but whether a board acting in good faith would have reached the same conclusion. That is the governance question the Shell–BP comparison forces.
Lesson. The Shell comparison is decisive. Same resolution, same filer, two London-listed oil majors under the same Companies Act — opposite procedural outcomes. One board chose to meet the challenge on the ballot; the other chose to prevent the ballot from being needed. One reading of the law permitted exclusion; another did not. BP chose the first. Shell chose the second. A resolution that fails at the ballot is democratic; one that never reaches it is procedural closure. When boards deploy eligibility arguments to prevent inconvenient resolutions from facing a vote, they substitute process management for democratic legitimacy. A board will always look compliant under conventional screening when it wins on procedural grounds. Whether it is actually governing is a different question.
3. The climate governance ratchet
Backstory. At the 2019 BP AGM, 99% of shareholders — and the board itself — supported a binding special resolution committing BP to annually review and disclose the consistency of each new material capex investment with the Paris climate goals. At the 2022 AGM, 88.5% of investors endorsed BP’s climate plan, including the 40% hydrocarbon production reduction target for 2030. Those commitments were enshrined in BP’s articles of association — its corporate constitution — and would require a 75% supermajority to repeal. Resolution 23 is the first board-initiated attempt to remove them.
Governance issue. The tension here is between the mutability of corporate strategy — the fact that strategy must bend as circumstances bend — and the durability of shareholder-approved commitments, which are meant to hold across political and market cycles. Boards must adapt as conditions change, but shareholders have reasonable expectations that commitments they have explicitly approved, particularly those adopted as special resolutions and embedded in the articles, will endure beyond any single political or market cycle. A board’s move to repeal such commitments forces a question: were they real governance architecture, or merely management preferences that secured shareholder endorsement at the time? That distinction matters far beyond BP. Institutional investors supported climate-linked constitutional provisions at many issuers during 2019–2022 believing those commitments would endure. The BP case is the first occasion on which that understanding is being tested at scale.
Lesson. Either outcome is instructive — and each signals something different, not just about BP, but about the durability of climate commitments at every issuer. The table below sets out what each outcome would establish, including what it would signal about institutional stewardship — meaning the active oversight that asset managers and asset owners exercise on behalf of their beneficiaries through voting, engagement with management, and shareholder resolutions.
Resolution 23: Two Paths, Two Signals
| If Resolution 23 passes | If Resolution 23 fails to reach 75% | |
| What the outcome establishes | A precedent is set. Climate commitments made during strong ESG periods can be unwound when conditions change — as long as the board starts the process and wins a supermajority. | The 75% threshold has teeth. It has blocked a simple majority — which may include activist-aligned holders with shorter horizons — from undoing what an earlier coalition approved. |
| What it says about other issuers | Companies whose climate commitments sit in ordinary resolutions are more exposed than BP. Commitments in special resolutions are now in range as well. | The higher the repeal bar, the more durable the commitment. Boards considering constitutional-layer commitments have evidence the structure can hold. |
| What it says about stewardship | Institutional stewardship could not stop a board-backed repeal — despite $1.74 trillion of co-filed investor support lined up against it. An uncomfortable signal about stewardship’s real influence. | Stewardship held — but held on a vote about reputation, which is where stewardship is strongest. The harder tests, where real money is at stake, still lie ahead. |
The underlying point. The structural weakness the BP case has surfaced is visible whichever way the vote goes: climate commitments embedded in a company’s articles of association can be tested by board-initiated repeal, and the mere filing of such a resolution forces the shareholder base into a defensive posture. The architecture around a commitment — the resolution type, the repeal threshold, the safeguards against board-initiated reversal — matters at least as much as the commitment itself. Climate governance is, in the end, governance architecture. Every FTSE 100 board that embedded climate commitments during 2019–2022 should be re-reading its articles of association today.
4. The missing stakeholders
Backstory. The entire governance conversation around the BP AGM has been framed almost exclusively around shareholders. Yet the strategic decisions that flowed from Elliott’s campaign have profound implications for the creditors and bondholders who fund BP’s balance sheet. ACCR’s 10 March 2026 investor bulletin calculated the net present value of BP’s $22 billion of new oil and gas developments since 2020 at just $0.9 billion under forward prices (Brent approximately $58 per barrel). Oil prices have since moved materially higher following the Iran conflict, but the concern ACCR raises is not primarily about the current spot price: BP’s portfolio sits on the higher end of the global cost curve, and what matters for the NPV of multi-decade projects is where prices settle once the current spike recedes.
Governance issue. The tension here is structural rather than procedural. The UK listed-company governance architecture is built around the shareholder franchise: votes, resolutions, the AGM itself are equity-holder mechanisms. Creditors, bondholders, and long-duration stakeholders are structurally absent from these mechanisms — yet they bear much of the downside risk when strategic decisions redistribute value and risk across the capital structure. A board exercising genuine independent judgement should test major decisions against the interests of all capital providers, not just those with formal voting rights. The failure to do so is not visible to conventional governance screening — and it is made worse, not better, when boards become more tactically skilled at the equity dialogue. Shareholder mapping, investor psychology, and narrative construction in the face of activist challenge are necessary capabilities in 2026. But capability and orientation are different things.
Lesson. When activist pressure drives decisions that redistribute value across the capital structure — from creditors with claims on long-term stability to equity holders with shorter horizons — a board should test those decisions against the interests of all capital providers, even where the formal governance mechanisms require only the equity view.
The Companies Act section mentioned above requiring directors to promote the success of the company for the benefit of shareholders as a whole, while having regard to long-term consequences and to the interests of employees, the environment, business relationships, and other stakeholders clearly applies here as well.
A board that masters the capital-markets arena without widening its stakeholder aperture has just become more efficient at the wrong conversation.
Elliott was the most effectively engaged stakeholder in the BP process. The creditors who fund the capital pivot appear not to have been part of the dialogue at all. This is the creditor blind spot — and the reason the most dangerous boards, on this reading, are sometimes the ones that look best on paper.
III. How to read the vote
The lessons above do not depend on the 23 April outcomes. What the vote will show is how the wider institutional system — asset managers, asset owners, stewardship teams, and the proxy advisors that guide them — has read what has happened.
- Res. 4 (Manifold re-election). Compare dissent to the 24% opposition to former chairman Helge Lund in 2025. Higher = deeper scepticism. Lower = benefit of the doubt given to new leadership. But whatever the margin, the fact that a board chose to silence a validly filed shareholder proposal — and the backlash that followed — now stands as a precedent every other board will have to weigh carefully before considering the same.
- Res. 22 (Virtual AGM articles). Defeat would signal shareholders want their channels for voice preserved, not narrowed.
- Res. 23 (Climate mandates). The test of whether shareholder-approved commitments embedded in the articles of association can be unwound by a later board under different conditions. Either outcome demonstrates the vulnerability; the difference is whether it has been closed off or merely shown.
- Res. 24 (ACCR disclosure). The margin matters more than pass or fail. A narrow defeat would tell observers that disclosure-based activism on capital allocation sits close to winning at a major UK oil and gas company; a decisive pass would confirm that investors now expect this class of disclosure, even from resistant boards.
IV. Seven principles for directors and boards
The lessons above describe what the BP case has exposed. The seven principles below translate them into direct takeaways — spanning from preventing strategic drift in the first place to handling the architectural and stakeholder questions that surface once a crisis has already arrived. They apply at BP, and at every publicly-traded company whose climate commitments, activist risk profile, or capital-allocation decisions may next come into question.
Seven Principles for Directors and Boards
| Principle | What it means in practice |
| 1. Prevent strategic drift | A board that loses conviction in its own strategy should address it as a priority and transparently reopen the question directly with shareholders — not let commitments erode quietly through softened targets, delayed milestones, or silent retreat. Drift creates the vacuum that concentrated activist voices will eventually fill. The best time to revisit a strategy is before someone else with potentially more narrow interests forces the revisit. |
| 2. Mandate before execution | On a major strategic reversal, secure the shareholder mandate before acting — not after. Ratification is not authorisation. Shareholders know the difference, and activists will exploit it. |
| 3. Exclusion is a governance judgement, not a technicality | Before refusing to circulate a valid shareholder resolution, ask: would a similarly placed peer board reach the same decision? If not, procedural closure will escalate rather than settle the matter — and the cost will spread to unrelated agenda items. |
| 4. Visible practice is not actual governance | Proxy advisor ratings measure what is visible: structure, committees, policies, disclosure. They cannot see what is actually happening in the boardroom until consequences surface — and by then it is often too late. Use them as one signal among several, never as confirmation of substantive governance quality. |
| 5. Architecture protects commitments | A shareholder-approved commitment is only as durable as the mechanism required to repeal it. If a commitment matters, embed it at the constitutional layer with a supermajority threshold — and assume it will eventually be tested. |
| 6. Widen the stakeholder aperture | Test major strategic and capital-allocation decisions against the interests of all capital providers, not just the loudest equity voice in the room. Value can be destroyed in ways conventional governance screening does not see. |
| 7. Capability is not orientation | Mastering the equity dialogue — shareholder mapping, investor psychology, and shaping the narrative when challenged by activists — is necessary but not sufficient. It does not substitute for a stewardship orientation that looks past the loudest voice in the room and considers the voices of other stakeholders. |
These are not abstract principles. Each is a response to a specific moment in the BP case when a different decision would have produced a different outcome — for BP, and for the governance norms every board now operates under.
When a board pivots quickly under activist pressure, the question is not whether the activist was right. It is whether the board’s own governance processes were robust enough that the pivot, if warranted, would have happened anyway. If the answer is no, the board did not exercise governance. It surrendered it.
BP is the cautionary tale. A comparative assessment anchored on the Shell AGM will follow in Blog 2 — Shell’s 19 May vote on a virtually identical Follow This resolution will provide the direct governance comparator this case demands, testing whether a board that chose to meet the challenge on the ballot can carry the argument where BP chose to prevent the ballot from being needed. A longer research paper applying the Governance Substitution Effect framework — which I am developing as an assessment tool for understanding and identifying the governance failures that lie beneath surface-level compliance — is forthcoming in June–July.
Recent Comments