Koos Bekker ran Naspers for fourteen years without a salary or bonus, taking his compensation entirely in share options. Under that structure he made a modest investment in a Chinese internet start-up in 2001 that became worth many times the entire company that made it — the single most consequential investment decision in South African corporate history, and a demonstration of what incentive design actually produces.
The most valuable decision ever made by a South African executive was made by one paid nothing to make it. This story covers the compensation structure, the pay television foundation, the Chinese investment, the failures alongside it, the discount problem and what the record actually shows — part of the South Africa Company Stories hub.
What was unusual about the compensation?
He took no salary, no bonus and no benefits for his entire tenure as chief executive, receiving only share options that were worth nothing unless the share price rose substantially.
What was the investment?
A stake acquired in 2001 in a then-small Chinese internet company, which grew into one of the world’s largest technology businesses and came to represent the overwhelming majority of the group’s value.
What is the honest assessment?
One extraordinary success alongside a long list of expensive failures — which is exactly the return distribution venture investing produces, and which few listed companies have the structure to tolerate.
Why does the pay structure matter?
Because it removed the asymmetry that shapes most executive behaviour. A chief executive drawing a large salary and an annual bonus is rewarded for not losing money this year, which makes long-horizon bets with uncertain outcomes personally irrational.
An executive paid only in options has the opposite incentive: modest performance produces nothing, so the rational strategy is to pursue outcomes large enough to move the entire share price.
Critics note that options are a one-way bet, since the holder loses nothing if they expire worthless. That is true, and it is also why the structure produces risk appetite that conventional pay does not — the question is whether the shareholder wants that appetite.
What was the business before the internet?
A newspaper and magazine publisher that moved into pay television in the 1980s, building a subscription business across South Africa and then across the continent, described in the MultiChoice story.
Pay television generated exactly the characteristics an investor in early-stage technology needs: recurring subscription revenue, strong cash conversion and a customer base that grew with African urbanization.
That cash flow funded the internet investments. Without a profitable subscription business behind it, the venture programme would have required raising capital from shareholders who would not have approved the strategy.
How did the Chinese investment happen?
Through a deliberate search for internet businesses in emerging markets, on the reasoning that consumer internet models proven in developed markets would replicate where internet adoption was beginning, and that local operators would win those markets.
The specific investment was a minority stake acquired for a modest sum in a business that was then unprofitable and largely unknown outside its home market.
The decision that mattered almost as much was holding it. A stake that appreciated many hundredfold offered countless opportunities to sell, and resisting them for two decades required conviction that most institutional holders would not have sustained.
What about the investments that failed?
There were many. Internet businesses acquired across several continents were written down, closed or sold at losses, and the aggregate of those failures is a large number that receives far less attention than the one success.
That distribution is not a criticism — it is how venture returns work. A portfolio in which one investment returns a thousand times and most return nothing outperforms a portfolio where everything works modestly.
The problem is that listed companies are poorly structured to tolerate it. Public shareholders see the write-downs quarterly and the compounding winner only in retrospect, which is why very few listed groups run genuine venture programmes.
What is the discount problem?
That the group’s market value has persistently traded far below the value of its stake in the Chinese business alone, implying that the rest of the company — and the management running it — carries negative value in investors’ assessment.
The causes are structural and behavioural: index constraints limiting how much any fund can hold, tax on any disposal, and scepticism that management will return value rather than reinvesting it in businesses shareholders did not choose.
Attempts to solve it through corporate structure largely failed, and the mechanism that has worked is simply selling small amounts of the asset and buying back discounted shares — covered in the Naspers and Prosus story.
What management approach produced this?
A deliberately small head office, decentralized operating businesses, and a willingness to enter markets and categories without established local expertise on the reasoning that internet businesses were being built everywhere simultaneously.
Long absences from the business were part of the method — extended periods away, on the argument that a chief executive constantly present makes decisions that subordinates should be making and loses perspective on which decisions actually matter.
Whether that approach is replicable or simply idiosyncratic is impossible to settle from one example, which is the central difficulty in learning anything transferable from exceptional outcomes.
How much was luck?
A substantial amount, and any honest account says so. The investment was made in a company whose eventual scale nobody predicted, in a market whose growth exceeded every forecast, in a category that turned out to be extraordinarily profitable.
What was not luck was the decision to look for such investments at all, the willingness to invest before the thesis was obvious, and the discipline to hold through a decade of opportunities to take a profit.
The useful framing is that luck determines the size of an outcome and process determines whether you are positioned to receive one — which is the whole argument for venture-style allocation.
What is the succession question?
Whether a group defined by one investment decision can generate value from operating businesses in the way its original strategy assumed, now under management without the same latitude or compensation structure.
The operating portfolio — classifieds, food delivery, payments and e-commerce — has produced at least one clear success and a number of businesses whose path to profitability remains the central question for shareholders.
The uncomfortable truth is that the group will be judged for years on whether it can do anything else, against a benchmark that was itself a once-in-a-generation event.
What is the lesson?
That incentive structure determines which decisions get made. A salaried executive optimizing for annual results would not have made a speculative minority investment in an unprofitable foreign start-up, and would not have held it for twenty years.
The second lesson is about return distributions. Most bets fail; the arithmetic works because the winners are unbounded, and any structure that cannot tolerate visible failures cannot access those returns.
The third is humility about attribution. The outcome was extraordinary, the process was sound, and separating the contribution of each is beyond what a single case can honestly support.
Why is a subscription business a good venture funder?
Because it produces predictable cash monthly, requires limited incremental capital once the network is built, and its results do not depend on the outcome of the venture programme it is funding.
That separation is what allows a listed company to make speculative investments without threatening its own solvency or its dividend, which is the constraint that stops most corporates from trying.
It also gives management time. A venture thesis needs a decade to be tested, and only an underlying business that can fund that decade without external capital gives the board the patience to allow it.
What does the emerging markets internet thesis assume?
That consumer internet business models proven in developed markets — classifieds, marketplaces, payments, food delivery — will be repeated wherever internet adoption grows, and that local operators with local knowledge usually win those markets over foreign entrants.
The thesis has been broadly correct on the first point and mixed on the second, since global platforms with capital have taken several markets that local operators were expected to hold.
Where it has worked best is in categories requiring dense local operations — delivery logistics, payment rails, classified liquidity — which are precisely the businesses a foreign platform cannot run remotely.
Why did the group’s other internet bets fail?
Mostly for the ordinary reasons ventures fail: entering markets where a better-funded local competitor already had liquidity, buying at valuations that required flawless execution, and underestimating how long profitability would take.
Marketplace businesses in particular are winner-takes-most, so a strong second place has very little value, and capital spent competing for a market that consolidates around someone else is capital lost entirely.
The strategic response has been consolidation into fewer categories where the group holds genuine leadership, accepting that a portfolio of subscale positions is worse than a smaller number of dominant ones.
What should shareholders watch now?
Three things: whether the buyback programme continues at a scale that meaningfully narrows the discount, whether the operating businesses reach sustainable profitability, and how much of the Chinese stake remains as a share of total value.
Each is measurable from the accounts, which is a change from a decade ago when the investment case rested largely on an asset the group did not control and could not influence.
The strategic question underneath is what the group is for. A vehicle that exists to hold one appreciating asset and return capital is a very different proposition from an operator building businesses, and shareholders are entitled to a clear answer about which one they own.
Is the compensation model replicable?
In principle yes, and almost no board does it, because an executive who accepts no salary must already be wealthy enough to work for years without income, which narrows the candidate pool dramatically.
Remuneration committees also resist option-only structures on governance grounds, since large option grants dilute shareholders and can reward share price movements unrelated to management performance.
The transferable part is the principle rather than the mechanism: compensation weighted heavily toward long-dated equity, with vesting periods measured in many years, produces different decisions from cash bonuses tied to annual results.
Frequently Asked Questions
Why did he take no salary?
He was compensated entirely in share options, which are worthless unless the share price rises substantially, aligning his outcome with long-term shareholder returns rather than annual performance.
What was the 2001 investment?
A minority stake in a then-small Chinese internet company acquired for a modest sum, which grew into one of the world’s largest technology businesses.
Did most of the investments succeed?
No. Many internet acquisitions were written down, closed or sold at a loss. The programme worked because one outcome was large enough to outweigh all of them.
What is the holding company discount?
The gap between a holding company’s market value and the value of its underlying assets, driven by index limits, tax on disposals and scepticism about capital allocation.
Discover more from Kurums | Business Intelligence
Subscribe to get the latest posts sent to your email.
