NextFin News - Executive pay in South Africa is no longer a conversation boards can politely acknowledge and move on from. Since 22 May 2026, the Companies Amendment Act has converted shareholder votes on remuneration policies and implementation reports into binding legal approvals for public and state-owned companies, and a new "two-strike rule" puts the seats of remuneration-committee directors on the line after two consecutive rejections. The change arrives with pay outcomes appearing broadly supported — shareholders endorsed remuneration outcomes at an 83.8% approval rate last year — but that support was compiled under an advisory regime that boards could safely ignore. The question now is whether the 83.8% holds when a "no" vote carries real consequences.
The Vote That Can No Longer Be Ignored
For years, South African listed companies treated remuneration votes as a disclosure exercise. Under the Johannesburg Stock Exchange listing requirements and the King IV code, companies tabled a remuneration policy and an implementation report, and shareholders cast advisory votes. If dissent crossed 25%, the company was expected to engage with investors. The vote itself changed nothing: pay could be paid, policies could stand, and the annual cycle repeated.
The Companies Amendment Act 2024, No. 16 of 2024 — signed into law in July 2024 after six years of drafting and consultation — ended that arrangement. From 22 May 2026, public companies and state-owned enterprises must obtain shareholder approval by ordinary resolution before implementing or changing a remuneration policy, and must submit the annual remuneration report for binding approval every year. The policy must be tabled every three years, or sooner if materially changed. For JSE-listed issuers, this gives statutory force to what was previously a compliance expectation. The vote now carries legal weight, not just reputational pressure.
The immediate practical consequence is a reversal of timing. Under the advisory system, boards could announce pay, read the vote, and then decide whether engagement was worth the effort. Under the binding system, the vote is decided before the AGM — most votes are cast electronically in advance, and companies often know the result before the meeting begins. Boards that want their pay package to pass must now do the investor-relations work months ahead of the vote, not after it. Courting shareholders is no longer damage control; it is the approval process itself.
The Two-Strike Rule: Personal Consequences for Remuneration Committees
The binding vote is the stick; the two-strike rule is the hand that wields it. Under the new regime, if shareholders reject the remuneration report — more than 50% of votes cast against — every member of the remuneration committee must stand for re-election in their capacity as committee members at the next AGM. If the report is rejected a second consecutive year, those members must stand for re-election as directors and, even if re-elected to the board, are barred from serving on the remuneration committee for two years. The rule applies only to directors who served on the committee during the previous 12 months, but for small boards the threat is acute: losing an entire remuneration committee, or blocking its reconstitution, is a governance disruption no chairman wants to explain to the market.
"At face value, executive pay outcomes in South Africa appear well supported. Shareholders endorsed remuneration outcomes at an 83.8% approval rate last year. But that support was under a non-binding regime. The binding vote changes the equation entirely, and remuneration committees need to be prepared for a more demanding conversation with shareholders."
The 83.8% figure comes from PwC's 2026 executive remuneration landscape report, and it is the number that makes the regime change so consequential. High approval under an advisory system does not mean high approval under a binding one. Advisory votes are cheap for shareholders: they signal displeasure without forcing a confrontation over who gets paid what. A binding vote forces a decision — and forces institutional investors to choose between a pay package they dislike and the governance instability that follows from rejecting it.
History suggests the threshold for real conflict is lower than the headline approval implies. Between 2021 and 2023, Glass Lewis found that 73 of 409 South African remuneration policies — roughly 18% — drew 25% or more opposition, and 10 were opposed by a majority. Of 408 remuneration reports, 101 received 25% or more opposition and 17 were rejected outright. Two companies, Life Healthcare Group Holdings and Northam Platinum Holdings, saw their remuneration reports rejected in consecutive years. Under the old rules, those rejections were reputational events. Under the new rules, the second one would have triggered the second strike.
Why the Binding Vote Changes the Transmission Mechanism
The first-order effect of the reform is obvious: a failed vote blocks the pay package. The second-order effect is what will reshape corporate behavior. When a vote is advisory, the cost of dissent falls almost entirely on the company's reputation. When a vote is binding, the cost is reallocated — to the remuneration committee's composition, to the board's stability, and to the company's ability to execute its stated pay policy. That reallocation changes who has leverage in the room.
Consider the sequence. A remuneration committee designs a package. Under the advisory regime, the committee's primary accountability was to the board and to the King IV principle that pay should be fair and responsible. Engagement happened after a 25% dissent, at the company's discretion and on the company's timeline. Under the binding regime, the committee's accountability runs directly to the shareholder vote, and the penalty for misreading the room is personal: standing for re-election, and potentially a two-year ban from the committee. The committee now has a structural incentive to pre-clear packages with large investors before they are announced — the same "pre-clearance" culture that developed in the United Kingdom after it introduced binding shareholder votes on remuneration policy in 2013.
This is the mechanism that turns pay governance from theatre into negotiation. Boards cannot wait for the AGM to learn what investors think, because by then the votes are already cast. They must engage early, disclose more, and design packages that can survive a binding vote rather than an advisory one. The binding vote does not cap pay; it changes the price of pay. A package that would have passed with a 60% advisory vote may now be withdrawn and redesigned, because 40% opposition signals a first-strike risk the committee cannot afford to carry into a second year.
The Transparency Lever: Pay Gaps and Named Officers
The binding vote arrives bundled with the most far-reaching pay-disclosure requirements South Africa has seen. Companies must now publish, in the annual remuneration report, the ratio between the total remuneration of the top 5% of earners and the bottom 5% of earners. They must also disclose the total remuneration of the single highest-paid employee, the lowest-paid employee, the average, and the median. In a country where income inequality remains a central economic and political concern, the pay-gap ratio is not just a governance metric — it is a public-facing number that employees, unions, regulators, and the media will cite.
Equally significant is the end of anonymous disclosure for prescribed officers. Companies required to have their annual financial statements audited must now identify each prescribed officer by name, with individual remuneration. The previous practice of disclosing prescribed-officer pay in aggregate — a shield that allowed senior executives below board level to remain outside the spotlight — is no longer permitted. Named disclosure means named accountability: an investor unhappy with a specific executive's pay can now target that executive's package directly, rather than voting against a blended pool.
The disclosure regime also interacts with the binding vote in a way that amplifies scrutiny. A pay-gap ratio that looks defensible in a board paper may look indefensible on the front page of a newspaper. Remuneration committees will need a documented, defensible methodology for calculating total remuneration — particularly for long-term incentives, share-based awards, and deferred benefits, where valuation choices can move the numbers materially. The committee that cannot explain its methodology to a shareholder will not survive a binding vote.
The Counter-Thesis: Controlled Companies and the Limits of Shareholder Power
The strongest argument against the transformative power of this reform is that it changes less than it appears for a large slice of the Johannesburg market. South Africa's listed landscape includes many controlled companies — firms where a single shareholder or a tightly aligned group holds a majority of voting rights. An ordinary resolution requires 50% plus one of votes cast. If the controlling shareholder supports the pay package, the binding vote is a formality, and the two-strike rule will never be triggered. For these companies, the reform adds disclosure and process without changing the outcome.
There is a second limit built into the rules themselves. The JSE listing requirements previously required issuers to engage with shareholders where dissent exceeded 25%. Following the Companies Amendment Act, the JSE published an amendments paper proposing to replace that engagement requirement — for foreign applicant issuers — with a clause aligned to the Companies Act's 50% threshold. Even where the 25% engagement trigger remains, the Companies Act itself requires engagement only when 25% or more of votes are cast against, and a company facing 40% dissent still cannot be forced to change its pay package if the ordinary resolution passes. Engagement is a process obligation, not a veto.
This counter-thesis has real force. It means the binding vote will bite hardest at widely held companies with no dominant shareholder — precisely the firms where shareholder activism was already most effective. At controlled companies, the reform's main effect is transparency: the pay-gap ratio and named-officer disclosure will be public even where the vote is not competitive. That is not nothing. Public pay data changes the politics of pay even when it does not change the vote.
But it does mean investors should not assume the binding vote will produce a wave of rejected pay packages across the JSE. The mechanism works where ownership is dispersed and where institutional investors hold the balance of votes. Elsewhere, the binding vote functions more as a disclosure and documentation regime than as a veto.
Is This Cyclical or Structural?
This is a structural shift, and it will not revert. The binding vote and the two-strike rule are statutory, not a listing requirement that a board can negotiate around or a code provision that can be watered down at the next review. Reversing them would require Parliament to amend the Companies Act again — a politically difficult move when pay inequality is a live public issue. The disclosure requirements are equally statutory and equally hard to unwind.
The structural nature of the change matters for how investors should read it. A cyclical reform — a temporary tightening of sentiment, a one-year surge in activist campaigns — would eventually mean-revert, and boards could wait it out. A structural reform changes the permanent cost of capital for pay: every remuneration package now carries an approval risk that did not exist before, and that risk is priced into how boards design incentives, how they sequence engagement, and how much they disclose. The companies that adapt are those that treat shareholder approval as a design constraint, not as a post-hoc ratification.
South Africa is also not alone in this direction of travel. Australia introduced a two-strike rule in 2011, with a lower 25% dissent threshold and a "spill meeting" mechanism that gives shareholders a separate vote on whether directors should stand for re-election. The United Kingdom has required binding shareholder votes on remuneration policy since 2013, though London keeps the annual implementation report advisory while Pretoria makes both binding. South Africa's version is stricter in one respect — the second strike has immediate consequences rather than triggering an additional vote — and more lenient in another, since it does not remove directors from the board outright. The regional precedent suggests the regime, once introduced, tends to persist and to be tightened rather than relaxed.
What to Watch: The First AGM Season Under the Binding Regime
The rules apply to remuneration votes taken after 22 May 2026, so the first full AGM season under the binding regime will unfold across late 2026 and 2027. That is when the market will get its first read on how the rules work in practice. Investors should watch three signals.
First, the dissent rate on remuneration reports relative to the 50% rejection threshold. A rise in 30%–49% "against" votes — dissent that would have been survivable under the advisory system but now signals acute first-strike risk — would indicate that boards are struggling to pre-clear packages. Second, the number of companies that withdraw or revise pay proposals ahead of the AGM after investor feedback; an increase would show the binding vote working through engagement rather than through rejection. Third, the pay-gap ratios disclosed in annual reports, and whether companies with extreme ratios face coordinated voting opposition.
The base case is that most pay packages pass, but with more pre-AGM engagement and more withdrawn or revised proposals than under the advisory era. The upside case for shareholder power is a cluster of first-strike re-elections at widely held companies, which would establish the two-strike rule as a credible threat. The downside case is that controlled companies dominate the statistics, engagement obligations prove unenforceable, and the binding vote proves largely symbolic outside a narrow set of issuers — with transparency, not accountability, as its lasting legacy.
The falsifying signal for the view that this reform meaningfully shifts power to investors is specific: if, across the first two AGM seasons, fewer than five companies face a remuneration report vote above 40% against and no remuneration committee member is forced to stand for re-election under the first-strike rule, then the binding vote has not altered the balance of power in practice, and the reform's effect is disclosure rather than accountability.
For boards, the lesson is plain. Under the advisory regime, the annual remuneration vote was a ritual. Under the binding regime, it is an approval gate with personal consequences. The companies that survive the transition are those that start courting investors before the proxy is printed — because once the votes are cast, the conversation is over.
Explore more exclusive insights at nextfin.ai.
