Naspers has, for years now, looked like a wholly inelegant swan. Rather than gliding gracefully through the corporate world, it has paddled furiously from one clumsy transaction to another.
Normally, you’d write this off to just another confused corporate strategy. Only, in the case of Naspers, it has so much more meaning given that it and Prosus — the company it spun out in 2019 — have a combined market value of R2.7-trillion. It is, by some distance, the largest local company on the JSE, and a staple in almost every pension fund. In other words, when Naspers dives, so do most retirement savings pools.
The company, which began life in 1915 as the publisher of Afrikaans newspaper De Burger, is almost a victim of its own success, having done the most profitable deal the world has yet seen. In May 2001, Naspers chair Koos Bekker “discovered” a Chinese start-up called Tencent, and bought 49% of it for R266m.
Today, even though Naspers has cashed in billions by selling down its share in the Hong Kong-listed Tencent — the world’s largest video game vendor and owner of WeChat — its remaining 25% is still worth R1.8-trillion. Bloomberg estimates that this is a staggering return of more than 800,000%.
More to the point, Tencent accounts for 93% of Prosus and Naspers’s value. Much of the relentless paddling has been to find something, anything, to reduce its lopsided reliance on China. It has spent an immense $20bn on buying other businesses to prove it’s no one-trick pony. All it did was solidify this impression.
Yet, as Naspers announced each new money-guzzling deal, the investment community looked on with bemusement. For 10 years, it was a story of strange investments, then stranger add-ons followed by write-offs, the listing of Prosus, then the creation of a head-scratchingly complicated cross-holding that everyone hated, followed by the unwinding of that cross-holding.
Signs that this frenetic activity might finally be slowing down was one vital reason Naspers’s latest set of results — for the six months to end-September — was remarkably warmly received last week.
Critically, embers of profit from somewhere else besides Tencent have begun to glow. Is Naspers, investors dared hope, finally repaying the faith?
Narrowing losses
The cold numbers were that for the six months, revenue grew 8.9% to $3bn, while it made an operating loss of $426m. But, throw in its $1.15bn share of Tencent’s profit and the $2.8bn it made when it cut its stake in Tencent from 26% to 25%, and profit comes to $3.3bn. (After foreign exchange movements are factored in, it clocks in at a $717m loss.)
It’ll be a big relief to Bekker and new CEO Ervin Tu, because the grim reality is that the company hasn’t had much to show for the $20bn it has sunk into a series of deals since 2004. It did get some value from its brief investment in Indian e-commerce venture Flipkart, and its Russian investments were looking promising until Vladimir Putin’s 2022 invasion of Ukraine forced an exit.
Yet every year, Naspers management talked up the non-Tencent businesses.
“We delivered another strong year of growth,” said then CEO Bob van Dijk, typically, in 2019. “Revenue from continuing operations on an economic-interest basis was $19bn, up 16% on last year ... Our core focus areas — classifieds, payments and fintech, and food delivery — were key in delivering this growth.”
Always growth, but never money in the bank. As one analyst told the FM, they had such a scattergun approach there was always one area that was doing comparatively well, so Van Dijk was able to focus on that, or refer to “core” performance and exclude new investments or loss-making businesses.
Now there’s a discernible difference in tone. “Naspers is successfully delivering on its commitments,” it crowed last week. “The fundamentals of its operating businesses are improving, driving profitable growth.”
It may have seemed like an AI-generated replica of statements in earlier years but this time there was something new.
Of its five segments — online classifieds, food delivery, fintech, online education and retail — only classifieds has made a trading profit, until now. This year classifieds’ trading profit more than doubled to $94m (though this is flattered by the decision to sell the car classifieds, OLX Auto, for an estimated $181m).
And there’s a change in that food delivery squeezed out a $5m trading profit, from a $70m loss the year before. (Equally, fintech’s trading loss narrowed to $22m from $80m, retail’s loss halved to $27m, while education made a $66m loss.)
Now, $5m isn’t a heck of a lot, and even that meagre amount disappears if you consider total economic interest rather than consolidated. But it’s the direction that’s important. And all five e-commerce businesses seem to be moving in the right direction.
The bottom line: Naspers’s e-commerce trading loss shrank to $38m from $270m for the six months to October. All on track, in other words, to reach its profitability target by its year-end in March.
Tu’s management makes much of the fact that if it does this, it’ll be six months ahead of its forecast for profit, in the first half of 2025. But investors — browbeaten by losses — may be justified in thinking it’s 10 years too late.
Still, it’s something. After years of talking about the path to profit, then veering down a lane to check out something in the shadows, the elusive destination seems closer than ever.
SBG Securities says this was “ahead of our expectations” and positions the group well for “value crystallisation over the medium term”, even if that “focus on profitability was, in part, at the expense of revenue growth”.
Admitting mistakes
No doubt the combination of new CEO Tu replacing Van Dijk and glimmers of profitability has led to analysts taking a guardedly more upbeat view of Naspers’s prospects.
Sasfin’s Alec Abraham tells the FM he was encouraged by Tu admitting to the company making some capital allocation mistakes, which implies a more cautious investment approach. Less danger, in other words, that it’ll pile on new cash-hungry acquisitions which will slow the shift out of the red.
“As a new broom it’s easier for Tu to point to past errors, but it also means he is less vested in past decisions,” says Abraham.
In particular, he’s looking forward to more concrete plans about a possible listing, or asset sales. “So far there’s only been talk of a possible listing of PayU in India in the next year or so,” he says.
Anchor Capital’s Mike Gresty says the signs of Naspers’s capital allocation flaws have long been evident in the impairments littered across the group over many years. Gresty is particularly stunned that Stack Overflow — a recent addition to the group’s edtech businesses — has already been impaired, by $340m this year.
Still, Gresty says the steep improvement in Naspers’s cash generation is particularly impressive.
“If [each arm] were cash-generative enough to be stand-alone [operations], then perhaps there could be listings,” says Gresty.
If these listings were to happen, this would make Naspers a more transparent business and help reduce the gaping discount between the value of its shares, and its net asset value.
But as with many weary Naspers shareholders, Gresty is unimpressed with the years of poor deals, which cost billions.
While it’s true it was unlucky to lose some of its most attractive assets in Russia after Putin’s invasion of Ukraine, and with AI overshadowing everything right now, Naspers’s clutch of investments is looking decidedly passé.
And while it may seem the days of throwing spaghetti against a wall are over, Gresty is still a little apprehensive about where the company may be thinking of investing next.
Without constantly adding new projects, it should be easy for some of the e-commerce businesses to grow profits, says Peter Takaendesa, head of equity at Mergence Investment Managers. It is one of three reasons he is upbeat about Naspers’s prospects.
“Some of the e-commerce businesses were profitable but under [Van Dijk], new projects were always being started, which was a big drain on profits and increased the overall losses,” Takaendesa tells the FM. He cites OLX Auto and grocery deliveries as two areas in particular.
A second reason he’s encouraged is that Tu’s team seems less tolerant of poor performers. Had Naspers kept to its traditional classifieds business, it would have done well — but buying OLX Auto at the top of the market and now selling it at the bottom is a symptom of the wider problem.
So much for CFO Basil Sgourdos’s consistent talk of “our disciplined capital allocation”, which littered results presentations for years.
The third cause for optimism, Takaendesa says, is the aggressive cost cutting, with heads rolling not just at the centre, but across the divisions to optimise the cost base.
But he is not persuaded that the improvement is permanent.
“The e-commerce business needs to be way more profitable, not just hovering at break even,” he says.
The discount conundrum
Of course, it wasn’t just improvement in the group’s non-Tencent portfolio that lifted spirits. The latest, and hopefully final, leg of a clumsy restructuring has also been completed, making the group look a little more elegant and easier to grasp.
And this has helped deal with the core affliction ailing Naspers for years — the yawning discount between the value of its assets. To demonstrate this practically, its NAV today is R1.03-trillion (more than 90% of which, predictably, derives from Tencent), yet its market value on the JSE is R640bn.
That’s a discount of 37% to its assets. But the good news is that this discount has been reduced quite heavily in recent months, enhancing the value in the group.
It does seem counterintuitive that Bekker’s team has cut the discount by selling off chunks of Tencent — its most valuable asset — and using these proceeds to buy back shares in Naspers and Prosus.
It seems even less savvy when you consider that Tu believes Tencent’s share price in Hong Kong, where it is listed, is undervalued. Still, if the strategy is to buy back shares to close that discount, it’s the obvious place to find the money.
And it does make sense. Look at it this way: when Naspers is trading at, say, a 37% discount to its underlying net asset value, then it costs only 63c to buy back R1 worth of assets, which contains a chunk of Tencent. So as Naspers buys back and cancels its shares, each remaining Naspers share has a higher component of Tencent.
There’s a further kicker. The source of the money to fund the repurchase of heavily discounted Naspers shares comes from the sale of Tencent shares, where there is no discount. In other words, Naspers gets full value for its Tencent shares, but is able to buy back its own shares at a hefty discount. As Takaendesa says, it’s an easy arbitrage: they’re selling Tencent in Hong Kong, where it’s fairly priced, and buying back Naspers in South Africa, where the discount makes it much cheaper.
The open-ended share repurchase programme was launched in June 2022 to an underwhelmed market. Shareholders had enjoyed a Covid-induced spike in the share price in early 2021, but had looked on in dismay as it lost almost 80% of its value by May 2022. Nothing seemed to be going right.
And the biggest thing going wrong was Tencent. The Chinese government had started cracking down on the country’s big tech firms which, until then, had been allowed free rein to build unassailable positions in a strong local market. Then, as the Chinese economy began to slow, tensions with its most important trading partner, the US, heated up.
The non-Tencent portfolio of businesses provided no comfort. Profits seemed as distant a prospect as ever. Making it worse was the growing suspicion that the Naspers board didn’t really know what to do about the discount. Nothing it had done seemed to be working. This was a source of considerable resentment, given the huge fees paid to advisers and the eye-popping remuneration packages of the overstaffed head office staff, particularly Van Dijk.
In 2017, the Geneva-based activist investor Albert Saporta accused Van Dijk of overseeing the destruction of R600bn of shareholder value since his appointment in 2014. Saporta urged Van Dijk to spin off the entire 34% stake in Tencent (which has since been sold down to 25%) to Naspers shareholders.
At the time Naspers’s shares were trading 40% below the value of its Tencent investment. Van Dijk explained this away by referring to structural issues on the JSE — a reasonable argument, given that Naspers had grown to become worth almost 20% of the JSE’s value. It meant institutional shareholders bumped up against restrictions on overexposure to any one share, and had to trim back their holdings of Naspers. This limited demand for the shares.
While Saporta was dismissed by Van Dijk, a version of the recommendations in his hard-hitting open letters was eventually taken up. Saporta had said in 2017: “The only route to effectively reduce the discount is either to sell some Tencent shares one way or another and buy back shares and/or spin out entirely the investment portfolio so Tencent is separated from the rest.”
In 2019 Naspers implemented its plans, listing its global internet assets on the Euronext Amsterdam exchange in a company called Prosus. Naspers held on to 73.2% of that Netherlands-listed company.
From day one, thanks to the Tencent stake, it was the largest listed internet company in Europe. But international investors didn’t arrive in the numbers expected, which meant the discount remained stuck at about 50%. Even the newly created Prosus traded at a discount to 35% of the Tencent stake.
It seemed Naspers had simply created twice the trouble for itself.
In 2021, the Naspers board announced one of the ugliest and most expensive transactions yet seen on the JSE: a Naspers-Prosus share swap designed to create a complex cross-holding between the two companies.
The board argued this was “one of the most feasible ways to sufficiently reduce Naspers’s weighting on the JSE shareholder weighted index”. It was the straw that broke the institutional camel’s back — in an unprecedented move, 36 fund managers got together and signed an outraged letter.
They said the deal “introduces elements which serve to increase complexity in the overall company structures, thereby reducing the likelihood of further value unlock, whether immediate or longer term.”
It was a mess: Prosus would end up with 49% of Naspers, with Naspers owning 57% of Prosus.
Naspers ignored the fund managers — as it could, given that just a tiny group of insiders, including Bekker and director Cobus Stofberg, owned all Naspers’s 970,000 “A” shares. This allowed them to control 68% of the votes. It was a done deal.
In this context, you can imagine the sense of resignation that greeted news in June 2022 that Naspers would be embarking on an “open-ended share buyback” — its latest tactic to close the discount.
Then, out of the blue, Naspers announced this year that it agreed that the cross-shareholding was an ugly mess, so it would unwind this structure. And, what a bonus! It wouldn’t even cost so much this time. The new cleaned-up structure sees Naspers holding 43% of Prosus and Prosus owning nothing of Naspers.
Accident or design?
All of this meant that when everyone sat down to hear Naspers’s half-year results last week, it all looked far less blue.
Tu told shareholders that since the June 2022 share repurchase programme began, the discount between the value of Naspers and Prosus stock had been reduced by about 17 percentage points.
“The narrowing of the discount and the increase in NAV per share has created $25bn of value for shareholders,” he said. The company remains committed to this programme, he added, as it creates value “while increasing our exposure to Tencent and our e-commerce portfolio on a per share basis”.
Gresty says as the repurchase programme soaks up natural sellers in the free float, it’s likely to have greater impact on the discount. Naspers says it has brought back 14% of the free float.
So the biggest value enhancement in Naspers hasn’t required any insight into technology or even well-honed capital allocation skills. The most important skill required was a deep understanding of financial engineering, to recalibrate the relationship between three distinct entities.
Could the present slim-lined, tax-efficient group structure have been the end plan all along? What appeared chaotic and ill-considered, like a wildly flapping swan, may have been just the right path after all.
It’s something Protea Capital’s Jean Pierre Verster has given some thought to. “Naspers is in a very strong position today despite all the sideshows that have distracted the investment community’s attention over the past seven or so years. It’s difficult not to suspect it was planned, but you never know,” he says.
He points out that, back in 2017, when Saporta demanded action, Naspers was a hugely wealthy company trapped on the JSE. Almost anything dramatic it might have done would have hit regulatory obstacles and involve a crippling tax bill.
Today it’s no longer trapped and its most valuable asset is in the tax-efficient jurisdiction of the Netherlands, where it’s far easier to sell the Tencent shares to fund the repurchase of Prosus shares.
South Africa-based investors, of course, won’t escape tax liabilities. And while the JSE might feel uncomfortable about the steady shrinking of its once most valuable company, the fact is that Naspers, through repurchasing activity, still contributes an important chunk of business to the exchange.
Does it even matter if the end result was by design or accident?
Naspers, under Bekker’s leadership, has created an enormously valuable asset which has made millions of South Africans richer, including the Government Employees Pension Fund clients, and has bumped up tax coffers along the way. And unlike Sasol and ArcelorMittal (formerly Iscor), no taxpayers’ funds went into its creation. All in all, remarkably elegant.
Share repurchases, which were illegal in South Africa until a Companies Act amendment in 1999, are controversial affairs. A growing band of supporters claim the practice is a useful tool for managing a company’s balance sheet, while opponents contend it’s an unimaginative use of a company’s capital and fraught with danger, including conflict of interests.
So when Naspers/Prosus announced an open-ended share repurchase programme in June 2022, investors paid close attention. Just as well they did: it turns out it’s a far more imaginative and value-creating exercise than most buybacks. It’s also much more complicated.
Fortunately, the Dutch authorities require detailed information on a weekly basis. The JSE has nothing similar.
According to Naspers, in the 15 months between June 2022 and September 2023, Prosus bought back 483-million of its ordinary N shares for $13.9bn. The purchase, which it says was at “a significant discount to NAV”, was funded by the sale of Tencent shares and resulted in a 7% increase in Prosus NAV per share.
Over the same period, Naspers sold nearly 68-million Prosus shares for $4.4bn and used most of the proceeds ($4.3bn) to buy back 31-million of its ordinary shares.
The company will not reveal the transaction fees nor name the brokers used. But a spokesperson tells the FM the fees paid include “a standard brokerage fee taking into account the size and open-ended nature of the buyback, and a customary required securities transfer tax in South Africa as well as in Hong Kong related to the sale of Tencent shares”.
However, there are restrictions, imposed by the Amsterdam Stock Exchange, on the volume of shares that can be bought back on an ongoing basis. No more than 25% of the 20-day average daily trading volumes can be repurchased on that exchange, which limits the number of Naspers shares that can be bought back because of the need to balance the two buybacks. So, while Prosus’s buybacks account for just below 25% of 20-day average trading volumes in that share, Naspers’s buybacks account for below 20% of average daily volumes.
— The buyback that worked
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.