Stewarding the last war

Lessons in where power lies from the BP skirmish

The news shortly before Christmas that BP chief executive Murray Auchincloss will be replaced by Meg O’Neill from April 2026 confirms the company’s continuing strategic shift away from renewables and back towards its traditional activities.

There are undoubtedly numerous factors at play, and I am doubtful that any investor strategy aimed at preserving BP’s previous commitments could have decisively altered its direction of travel. However, the way in which that direction was ultimately set, and the channels through which influence was exercised, are worth close attention.

In this case, stewardship as practised by most traditional investors was largely orthogonal to the strategic outcome that emerged. The most consequential changes at BP did not flow from the AGM, nor from stewardship centred on it. It was urged through a different channel altogether, utilising different tactics and timescales that do not map neatly onto the conventional stewardship framework.

It was not a story about whether stewardship “worked” or “failed”. It is about two forms of power operating on different axes. BP offers a particularly clear example of what happens when those axes diverge.

Two models of investor power

To understand what happened at BP it is useful to distinguish between two quite different forms of investor activity that might be described as “stewardship”. Both typically seek changes in companies, but in practice they operate very differently.

1. Stewardship as institutionalised ‘ownership’

The first model of stewardship practiced by most large institutional investors such as long-only asset managers, pension funds and other asset owners.

This model is built around a relatively stable conception of ‘ownership’¹ arising from their shareholding. Investors hold shares on behalf of beneficiaries, exercise voting rights at meetings, and engage with companies through structured dialogue with boards and management.

The AGM is often a focal point, both symbolically and procedurally. The more US-style terminology for the items on the agenda - ‘proposals’ - is perhaps better than ‘resolutions’. Essentially the AGM provides an opportunity for investor assent or dissent on the board’s proposals. This obviously has a tendency to pull engagement towards reactive activity.

The escalation of stewardship under this model follows a familiar path: private engagement, public statements, voting against proposals to elect directors or approve their remuneration, and, in rare cases, the filing of shareholders’ own proposals.

Time horizons in this model are relatively long, and this is seen by participants (at least on the investor side) as legitimation. Investors do not move straight to signalling intentions to vote against, or file resolutions. They typically seek to build up a paper trail that demonstrates that they have clearly communicated their position and expectations throughout. Any investor seeking to file a resolution will inevitably be asked ‘what previous engagement have you had with the company previously?’ In any case, the process of filing a resolution, or even running a ‘vote no’ campaign, typically runs for many months. Therefore under this model legitimate influence is understood to be cumulative. No shortcuts allowed!

The tools available to practitioners of this model are formal, rule-bound and visible. Importantly, legitimacy is closely tied to the ownership of shares: the right to meet with companies to get a hearing, and the right to vote are understood to rest on holding equity. And in practice the real control rights that investors have are attached to the shares themselves.

This is the model that is underpinned by the Stewardship Code, voting disclosure and other policy architecture. Policymakers and traditional investors created this world and most traditional investors agree to operate within its constraints. It does not really matter what objectives investors seek to pursue within it, but this is where almost all ‘responsible investment’ sits too.

2. Strategic intervention via economic exposure

The second mode of investor activity in play at BP looks very different. Here, influence is exercised not primarily through formal governance channels, but through strategic positioning in markets combined with targeted pressure on boards.

This model does not even require a shareholding: economic exposure matters more than legal ownership. Positions can be built quickly, adjusted rapidly, and maintained without holding meaningful quantities of physical shares. Instead those positions are often built using equity derivatives.

Voting rights may be acquired later or not at all. This matters for a number of reasons. Firstly, it clarifies that the investor is not focused on ‘set pieces’. Although events do matter, the AGM is rarely central and often it is incidental. Secondly, it means that the investor can spring on the company by surprise.

It is not clear if BP knew that Elliott had a acquired a sizeable derivative position before this emerged in press reports. I suspect that the best advisers are looking out for unusual market activity. But it certainly is the case that other market participants won’t know of the existence or nature of a derivative position without voting rights until someone chooses to make this public. Until Elliott acquired voting rights and triggered a TR-1 even the existence of its position was not officially confirmed.

This model is overwhelmingly proactive. By the time they launch their strike the investor has a clear sense of what they want to achieve - how the board should react to them - and will have no doubt gamed potential responses. (Again, good advisers in turn arm company targets with defence strategies, some prepared in advance of activists even appearing.)

Engagement, where it occurs, is typically private and backed by credible threats to the company’s reputation, strategy, leadership or a combination of these. Voting rights can be brought into play, but are not necessarily required as part of this.

This second model also seeks decisive change over a short timescale, rather than building legitimacy over time. Tactics are opportunistic rather than procedural, and timing is driven by market conditions, corporate vulnerability or transaction windows rather than the traditional calendar that most governance and stewardship people are used to.

Contrast this with the way the traditional stewardship model would typically require clear communication, often in public, and a gradual ratcheting up of pressure over a prolonged period.

This form of activity sits largely outside the stewardship frameworks that apply to long-only investors, and it does not rely on the slow accumulation of consensus among diversified shareholders. Yet as the BP example appears to show it can be highly effective in shaping corporate outcomes.

Two axes, not a spectrum

These two models are not simply different points on a single spectrum of “more” or “less” stewardship. They are different axes of power.

One is rooted in a conception of ownership, and relies on process and legitimacy. Getting the process ‘right’ is an important consideration and is judged almost as an end in its own right. The other is grounded in leverage, timing and strategic pressure and is entirely outcome focused. Process is only ever a means to an end. This model barely nods to formal notions of ownership or accountability (certainly not accountability of the investor).

At BP, both were present. But they were not operating through the same channels, to the same timescale or even using the same instruments. Despite acquiring voting rights, Elliott appears to have only ever held a token number of shares. The strategic outcome was shaped primarily through the second channel, while the first continued to operate largely as policy designed it to, adhering to the formal mechanisms and chronology of governance and stewardship. And that was not where the decisive action was taking place. The vote against the chair, whose departure had already been announced, triggered a formal response. But it was largely irrelevant given what else was already happening and going to happen.

This does not mean that traditional stewardship is meaningless, but that it is sometimes operating on a different plane from the forces that ultimately determine strategic outcomes at companies. That raises questions about how the mainstream stewardship world relates to this other model - does it challenge it, align with it, learn from it or some combination?

It also raises questions about policy and regulation. The modern stewardship ecosystem structurally favours the first model because it is organised around visible ownership, formal voting rights and process-based accountability. It cannot easily accommodate the second, because influence exercised through economic exposure rather than durable shareholding largely escapes the disclosure, engagement and legitimacy mechanisms on which stewardship relies.

The regulatory architecture is almost entirely focused on the first model - the Stewardship Code does not even mention derivatives and firms like Elliott can and do simply decline to comply with it in any case - even as significant changes such as those at BP are driven by the second.

This increasingly feels like an important gap in the governance system we have. Understanding how power actually operates across it is likely to matter more as ‘ownership’ fragments further as a result of market practices.

1. Shareholding is not legal ownership of the company, but in practice this rarely matters. Shareholders do have control rights and both companies and investors act as if the latter have a legitimate right to influence the direction of the former.