Handing Over the Shares Is the Easy Part
Employee Share Ownership Plans have become South Africa’s headline transformation instrument — and its most misunderstood one. The percentage on the certificate is not what decides whether workers ever see money. The funding model is.
I have sat in a lot of boardrooms where an ESOP was announced with real warmth — a genuine intention to bring workers into ownership — and I have sat in a lot of canteens three years later where nobody could explain why the dividend was R400. Both rooms were telling the truth. The gap between them is not bad faith. It is structure. An ESOP is a financing transaction wearing the clothes of a benefit, and if you design the financing badly, the goodwill evaporates while the debt keeps compounding.
Those middle and right-hand numbers come from the Competition Commission’s ESOP Impact Study, published in April 2026 — the most detailed look South Africa has yet had at what actually happens inside these schemes after the press release. The R70.3 billion is the Department of Trade, Industry and Competition’s own figure for the combined equity value of worker schemes across 118 companies, presented at the inaugural Worker Share Ownership Conference in April 2024.
Put those two sources side by side and you have the whole story of ESOPs in South Africa: an enormous, sincere, well-supported push to broaden ownership — and a distribution of actual cash to workers that is far thinner and far more uneven than the headline equity value suggests.
This piece sits alongside the earlier walkthrough of the Generic B-BBEE Codes scorecard, but it is deliberately a separate conversation. ESOPs touch B-BBEE, yes. They are not a B-BBEE thing. They are a corporate finance structure, a trust-law structure, a tax structure and an employee-relations structure, all at once — and the B-BBEE points are downstream of getting those four right.
What an ESOP Actually Is — and What It Isn’t
Three different things get called an “employee share scheme”. Only one of them earns broad-based recognition.
Giving five senior managers share options is not an ESOP. Giving most of your workforce units in a trust that owns part of the company is.
The confusion starts with vocabulary. In South Africa the term gets stretched across a management incentive plan, a community trust, and a genuine broad-based worker scheme — three structures with completely different rules and completely different outcomes.
Under the Codes of Good Practice, the distinctions are formal. A Broad-Based Ownership Scheme (BBOS) is a collective vehicle for a broad base of Black beneficiaries — typically communities, youth organisations or educational initiatives — governed by Annexe 100(B). An Employee Share Ownership Programme (ESOP) is a scheme for employees, governed by Annexe 100(C). If either is housed in a trust, the trust rules in Annexe 100(D) apply on top. And any trust, obviously, has to satisfy the Trust Property Control Act and be registered with the Master of the High Court.
A hand-picked scheme for a small group of senior staff will generally be assessed as a conventional trust rather than an ESOP — the point of the ESOP category is breadth. The Competition Commission’s guidance for merger-condition schemes makes the same demand in different language: the scheme must represent a broad base of workers rather than a few highly skilled ones.
A vehicle, not direct shareholding
Shares are almost never issued to individual employees. They are held by a trust or a company, and employees hold units in that vehicle. This is administrative necessity — you cannot run a share register for 8 000 people — but it has consequences all the way down.
Qualifying criteria
Almost every scheme sets an eligibility bar: minimum years of service, permanent employment, and often a grade cut-off. In the Commission’s sample, some schemes covered every grade, others excluded executives, others capped participation at middle management and below.
Funding that the worker never pays for upfront
Workers are not asked for cash. The shares are funded by the company — through a grant, a cash loan, or most commonly notional vendor finance (NVF), which is a bookkeeping loan rather than real money changing hands.
A trickle dividend
Because the vehicle owes money, dividends received are split: part to the beneficiaries, part to servicing the debt. That split — the trickle dividend ratio — is the single most consequential number in the entire structure.
Governance
Trustees are appointed by the beneficiaries and by the company, sometimes with an independent trustee. Whether the scheme also gets a seat on the operating company’s board is a separate design decision — and in the Commission’s sample, only just over half of the firms allowed one.
A duration
Evergreen schemes run indefinitely, so new employees keep joining but nobody realises a capital gain. Buyback schemes run for a set period, then the company repurchases the shares and beneficiaries take the appreciation. Hybrid models try to do both.
Where the B-BBEE Points Actually Come From
Ownership is 25 points. An ESOP is not a shortcut to all of them.
A well-run ESOP earns you three specific points that nothing else can, plus a share of everything else — but it does not remove the need to think about the rest of the ownership scorecard.
Ownership is governed by Statement 100 of the Codes and is worth 25 points, split across three sub-elements: voting rights, economic interest, and net value. It is also a priority element, which means falling below 40% compliance drops your overall level by one — regardless of how well you score everywhere else.
Here is where an ESOP fits. Within the economic interest sub-element there is a dedicated line: a 3% target for economic interest held by Black Designated Groups, Black participants in an ESOP, a BBOS, or a co-operative, worth 3 points. Those three points are not otherwise reachable. A conventional narrow BEE shareholding does not earn them.
But that is only the beginning. Black participants in an ESOP also count toward the general Black economic interest and voting rights targets, subject to an important cap: they can contribute a maximum of 40% of the total ownership points if the scheme meets the basic qualification criteria in Annexes 100(B) and 100(C) — and up to 100% only where the additional criteria are met. That cap is why “we’ll just do a big ESOP” is rarely a complete ownership answer on its own.
Voting rights — 6 points
25% plus one vote in the hands of Black people (4 points); 10% in the hands of Black women (2 points).
Economic interest — 11 points
25% to Black people (4 points); 10% to Black women (2 points); 3% to Black Designated Groups, ESOP and BBOS participants and co-operatives (3 points); 2% to New Entrants (2 points).
Net value — 8 points
The portion of ownership genuinely free of debt, calculated under Annexe 100(E) against a ten-year scale. This is the line item that punishes badly structured funding, and we will come back to it.
Two Compliance Details People Miss
Flow-through and modified flow-through. Ownership is traced through intermediary entities to the ultimate Black natural persons. The modified flow-through principle allows 100% of a 51%-Black-owned entity’s ownership to be recognised — but only once in the chain. Structuring an ESOP through multiple layers can quietly waste this.
The R25 million reporting threshold. B-BBEE ownership transactions valued at R25 million or more must be disclosed to the B-BBEE Commission. This is not optional and it is not a formality.
Status of the January 2026 Draft Amendments
On 29 January 2026 the dtic gazetted a set of draft amendments to the Generic Codes (Government Gazette 54032), with the public comment period closing on 30 March 2026. Those drafts centre on Enterprise and Supplier Development and the proposed Transformation Fund, and cover Statements 000, 004, 103, 400 and 600 plus the definitions schedule.
Statement 100 — ownership — was not reopened in that gazette. As at the time of writing, the ownership rules described above are the current rules, and the draft amendments remain drafts with no confirmed implementation date. Treat any commentary suggesting otherwise with caution.
The Funding Model Decides Everything
This is the part that gets decided in an afternoon and then governs the next fifteen years.
If the trust borrowed money to buy the shares, the workers are shareholders and debtors at the same time. Every rand of dividend has to choose a side.
Notional vendor finance is elegant on paper. The company “sells” equity to the trust at market value, the trust “owes” that amount, no cash moves, and dividends progressively extinguish the notional debt. When the loan is settled, the workers own the shares outright and receive the full dividend.
The elegance depends on two assumptions: that dividends will be declared, and that the debt will not grow faster than the dividends can pay it down. Both assumptions fail regularly.
On the first: dividends are declared at the discretion of the company, and the Commission’s study found that in 2025, of fifteen implemented schemes, only seven received any dividend at all. Reasons varied — poor performance, reinvestment of profits, or a decision to hold off until the ESOP debt was settled. Which is circular, and workers noticed.
On the second: interest. Where interest is charged on notional debt, the study found the dividend allocated to repayment often did not even cover the year’s interest charge. The capital never moves. The debt grows. The scheme becomes, in the Commission’s assessment, financially unsustainable — while still appearing on the scorecard as a transformation success.
Worked Example — How a Trickle Dividend Actually Lands
A company declares a total dividend of R1 000 000. The ESOP holds 5%, so R50 000 flows to the trust.
The trickle dividend is set at 60% to debt, 40% to beneficiaries. So R30 000 goes to the loan and R20 000 is distributed to workers.
The R30 000 retained in the trust attracts 20% dividends withholding tax. The R20 000 distributed is a pre-tax amount in each beneficiary’s hands, taxed at their own marginal rate.
Split R20 000 across, say, 800 beneficiaries and you can see how a R1 million dividend becomes a R25 payslip line — and why financial literacy training is not a nice-to-have.
There is one more consequence, and it is the one that connects straight back to the scorecard. Net value under the Codes measures the portion of ownership genuinely free of debt, on a ten-year scale. An evergreen debt structure — one where the loan never realistically extinguishes — can therefore prevent the scheme from ever reaching the net value thresholds the Codes require, and locks the Black shareholder in with an encumbered asset. You end up with a structure that is expensive, well-intentioned, scores poorly, and pays badly.
What the Competition Commission Found in April 2026
Fifteen schemes, six years, and a fairly blunt set of conclusions.
The regulator went and checked whether the schemes it had ordered were actually working. Its answer was: only partly, and here is exactly what to change.
A note on where these schemes come from, because it matters. Since the Competition Amendment Act came into force in February 2019, section 12A(3)(e) requires the competition authorities to consider whether a merger promotes a greater spread of ownership, particularly by historically disadvantaged persons and workers. Where it does not, the Commission can impose an ownership remedy — and an ESOP is the most common one. Under the Commission’s revised public interest guidelines, such a scheme should generally hold between 5% and 10% of the equity.
The April 2026 impact study reviewed schemes mandated between the 2019/20 and 2022/23 financial years. Twenty-five were identified; fifteen were still standing and analysable, across agriculture, mining, retail and fintech. The findings:
Debt is the default, not the exception
Eleven of the fifteen schemes were funded through notional vendor finance. Two received an outright grant and were debt-free from inception. Six of the eleven NVF schemes were charged interest — five below prime, one at prime — despite the loan being notional in the first place.
Most schemes are small
Eight held between 1% and 5% of the company. Four held 6–10%. Two held 11–20%. One held more than 50%.
Dividends are not guaranteed
In 2025, seven of fifteen schemes received a dividend. Across 159 465 participants, only 128 388 qualified to receive a distribution, and payouts ranged from R360 to R4 734 for the year.
Nearly everything is evergreen
Thirteen of fifteen were evergreen structures, one was a buyback, one a hybrid. Evergreen means beneficiaries never realise capital appreciation — they only ever see dividends, after debt service.
Beneficiaries do not understand the structure
Training exists at most firms but is rarely compulsory. Two firms provided no training at all to their trustees. Stakeholders raised this repeatedly as a source of misaligned expectations and resentment.
Workers are often not consulted on the design
Several firms fixed the design principles with external counsel before workers were involved at all — which is precisely the sequence the Commission now wants reversed.
Charge interest on a notional loan and the debt can outrun the dividend indefinitely — leaving beneficiaries permanently indebted, and the scheme itself unsustainable.
Paraphrasing the Competition Commission’s finding on interest-bearing ESOP debt, April 2026The study is explicit that its recommendations apply to merger-condition schemes — but equally explicit that they may apply just as well to schemes established under the B-BBEE Act and to ESOP best practice generally. If you are designing one voluntarily, this document is now the closest thing South Africa has to a public standard, and I would treat it that way.
The Tax Layer — and the Two-IRP5 Problem
The tax treatment is not the hard part. The administration of it is.
The law offers a real concession for genuinely broad schemes. Whether your workers can actually claim it is a different question.
Two provisions of the Income Tax Act do most of the work here, and they are mutually exclusive in practice.
Section 8B governs broad-based employee share plans. It is deliberately generous, and deliberately conditional: at least 80% of employees must be entitled to participate, the shares must be acquired for no more than a minimum consideration, and the market value of shares received by any one employee must not exceed R50 000 across the current and preceding four years of assessment. Hold the shares five years and the gain on disposal is capital in nature, subject to CGT rather than income tax. There is a corresponding deduction available to the employer under section 11(lA), subject to its own cap.
That R50 000 ceiling has not moved in a very long time. It has quietly narrowed the range of schemes that can use section 8B at all — worth noting, because the 2026 Budget did not propose any change to it.
Section 8C catches schemes that fall outside section 8B, which is most large ESOPs. Under section 8C, any gain made on the vesting of an equity instrument is included in the employee’s income — taxed at marginal rates rather than as a capital gain.
The Administrative Trap Nobody Designs For
Dividends retained in the trust attract 20% dividends withholding tax. Dividends distributed to beneficiaries are treated by SARS as income in their hands.
Here is the part that catches firms out: the trust is regarded as a separate employer. Beneficiaries therefore receive two IRP5s — one from the company, one from the trust. A taxpayer with two IRP5s is not auto-assessed and must file a return manually.
The Commission found this to be a live problem. Where the workforce is largely semi-skilled or unskilled, a meaningful number of beneficiaries simply do not file — and either receive less than they should or nothing at all. A scheme designed to build morale ends up generating exactly the opposite. If you take one operational lesson from this article, take this one: budget for tax support, not just tax advice.
The Companies Act and Trust-Law Layer
Where good intentions meet section 44, section 97, and the Master of the High Court.
If your company is lending money — real or notional — to help someone buy its own shares, there is a specific approval process. Employee share schemes get a carve-out, but only if they qualify for it.
Section 44 of the Companies Act regulates financial assistance given by a company for the subscription or purchase of its own securities, or those of a related company. The default position is demanding: a special resolution of shareholders adopted within the previous two years, plus a board resolution satisfied that the company will pass the solvency and liquidity test immediately afterwards, and that the terms are fair and reasonable to the company.
Section 44 also provides a carve-out where the assistance is given pursuant to an employee share scheme that satisfies the requirements of section 97. That carve-out is genuinely useful — and it is conditional. A scheme that does not meet the section 97 requirements does not get it, and the full approval machinery applies. This is a question to put to your corporate lawyers before the structure is drawn, not after.
If the vehicle is a trust, a further layer applies. The trust must be registered with the Master of the High Court and comply with the Trust Property Control Act. Trustees carry real fiduciary duties, which is why the Commission’s finding that two firms provided their trustees with no training whatsoever should be read as a governance exposure and not merely a shortfall in employee relations. Beneficiary-nominated trustees are frequently ordinary employees who have never read a set of financial statements — and who are now legally responsible for an asset worth tens or hundreds of millions of rand.
Annexe 100(C) reinforces this from the B-BBEE side. Among its requirements are rules on the appointment of fiduciaries and the extent to which participants appoint them, rules of participation, mechanisms for distributing economic interest, an annual presentation of the scheme’s financial reports to participants, and availability of the scheme’s constitution on request in a language the participant understands. Those are not box-ticking clauses. They are the difference between a scheme that survives verification and one that gets characterised as a fronting arrangement.
A Note on Skills Development Spend
The Commission makes a practical suggestion worth flagging to any finance director reading this: the training of employee beneficiaries and employee trustees can be funded through B-BBEE skills development spend. You earn recognition points on a scorecard element you are already spending against, while fixing the single most consistently reported weakness in South African ESOPs. There are not many places where the compliance incentive and the right thing to do line up this neatly.
The Questions to Settle Before Anyone Drafts Anything
An ESOP is not a document you commission. It is a set of about eight decisions, most of which are irreversible once the trust deed is signed, and each of which quietly allocates value between the company, the current workforce and the workforce of 2040.
Is it funded by grant or by debt? A partial grant, a discount on the share price, or interest-free notional funding will do more for beneficiary outcomes than any subsequent adjustment you can make. Is interest charged? If yes, model what happens in a year when no dividend is declared, and again in a year when the dividend does not cover the interest. What is the trickle dividend ratio — and have the workers been shown what each option means for them in year three versus year twelve?
Evergreen or buyback? Evergreen sounds more generous and often is not, because it locks capital appreciation permanently inside the structure. Where does the scheme sit — holding company or operating subsidiary? The Commission leans toward the operating level, so that workers own the business their effort actually affects, with the caveat that a struggling subsidiary is a poor place to park a worker’s only equity. Who are the trustees, and are they trained? Who is a good leaver? Someone who resigns after eleven years is not obviously “bad”, and a pro-rata payment costs you far less than the story that circulates afterwards.
And then the one that never makes it onto the agenda: who explains this to people, in their own language, every year, for as long as the scheme exists? Every serious weakness in the Commission’s study traces back to that question going unanswered. Ownership that nobody understands is not ownership. It is paperwork with a dividend attached.
The regulatory direction of travel is clear enough. Zero interest, real discounts, worker consultation before the structure is fixed, compulsory training, board representation, and monitoring with teeth. Whether or not those recommendations are formally adopted into merger conditions, they now describe what a defensible scheme looks like. Structuring one below that standard is a decision — and increasingly a visible one.
A Note on the Numbers in This Article
The scheme-level figures are drawn from the Competition Commission’s ESOP Impact Study published in April 2026, which covers fifteen merger-mandated schemes and is not a census of all South African ESOPs. The R70.3 billion equity value and the count of 125 schemes across 118 companies come from the dtic’s presentations at the April 2024 Worker Share Ownership Conference.
One caveat worth stating plainly: reported beneficiary counts from that 2024 conference vary considerably across contemporaneous accounts — from roughly 307 000 to over 550 000, depending on whether the count is of workers covered by newly established schemes or of all participants in all existing schemes. I have therefore not used a single headline beneficiary figure. Where you see a number in this piece, it is one I could tie to a specific document.
This article is general commentary, not legal, tax or financial advice. ESOP structuring sits at the intersection of company law, trust law, tax and the B-BBEE Codes, and every one of those layers needs a properly qualified adviser on your side of the table.