Prosus N.V.: A 38.4% NAV Discount, a Tencent-Funded Buyback, and No Conservative Margin of Safety

The current NAV itself quietly expresses skepticism about the acquisition. Prosus’s 2026-08-19 NAV sheet places JET at USD 4.2 billion, about EUR 3.62 billion at the ECB rate used here, below the roughly EUR 4.1 billion acquisition equity value. Six to nine months after delisting, the valuation framework is not capitalizing a large synergy premium. That looks sensible. The acquisition may ultimately work, but current evidence supports “integration under way,” not “value creation proven.”

The market’s main narrative has consequently become a three-way interaction. First is Tencent itself. Second is the discount-closing mechanism, especially the open-ended buyback. Third is whether the ex-Tencent assets can compound rather than consume the value harvested from Tencent. The last two questions are much more Prosus-specific than debating Tencent’s own operating fundamentals.

The buyback deserves especially careful treatment. Prosus started the open-ended programme in June 2022, when the NAV discount was around the mid-50s. The programme has been unusually large. The company’s buyback page said that by 30 September 2025 it had returned close to USD 42 billion, repurchased about 30% of Prosus free float and generated roughly 18% incremental Prosus NAV-per-share accretion compared with a no-buyback case. The FY2026 release later said cumulative buybacks had returned USD 46 billion by 29 June 2026 and referred to 16 percentage points of NAV accretion. The two accretion numbers use different dates or methodologies and should not be combined as one series, but both show that the programme is economically material.

The criticism that Prosus is “selling Tencent to buy itself” is factually correct at the corporate-asset level and incomplete at the per-share level. Prosus disclosed 2.0506 billion Tencent shares in its latest NAV statement. Tencent’s July 2026 return reported 9.083 billion issued shares at 31 July. Combining those slightly mismatched dates gives an ownership estimate of about 22.6%; the exact 14 August percentage may differ marginally because Tencent itself continued repurchasing shares.

At current valuations, Tencent represents 77.7% of Prosus NAV while the Prosus share price represents only 61.6% of NAV. If Prosus sells one euro of Tencent and uses the proceeds to repurchase Prosus at 61.6 cents per euro of NAV, the share count falls faster than Tencent value does. Tencent exposure per remaining Prosus share can rise even while Prosus’s absolute Tencent shareholding declines. Algebraically, a purely Tencent-funded repurchase ceases increasing Tencent value per Prosus share when Prosus’s market price rises above Tencent value per Prosus share. With today’s asset mix, that crossover would occur at a discount of roughly 22%, not today’s 38%. This is why management can accurately say that some accelerated buybacks increase Tencent exposure per Prosus share.

That does not make the programme infinitely scalable. Accretion becomes smaller as the discount closes. Absolute Tencent ownership keeps falling. Prosus itself generally describes associate ownership as typically falling in the 20%–50% range, although significant influence can survive below 20% in particular circumstances. Koos Bekker remains a Tencent non-executive director, which illustrates that influence is not reducible to a single percentage. I found no primary disclosure establishing an automatic loss of board representation, a specific index-treatment change or a forced strategic stop the instant Prosus crosses 20%. Treat the 20% level as an accounting and governance review point, not a known cliff.

Funding is another constraint. Morgan Stanley estimated in March 2026 that since mid-January about 54% of buyback funding had come from Tencent sales and 46% from cash. That is a sell-side estimate, not a Prosus disclosure. Management separately described periods in which proceeds from sales of other assets were used to accelerate repurchases and increase per-share Tencent exposure. The direction is toward a more mixed funding pool, but the programme has not escaped its dependence on the monetization of Tencent and portfolio assets.

A second starting assumption also needs correcting. Prosus and Naspers no longer have the 2021 cross-holding structure. Shareholders approved its removal in 2023 and the cross-holding was unwound. Naspers remains Prosus’s ultimate controlling parent and the companies retain closely linked governance, but the circular cross-holding itself is gone. Naspers shareholders can still suffer a “discount on a discount” because Naspers may trade below the value of its Prosus interest while Prosus itself trades below NAV. A Prosus shareholder directly experiences only the Prosus-level discount; describing Prosus itself as currently carrying two arithmetical holdco discounts would be wrong.

The sharpest expression of the bull/bear dispute is the ex-Tencent valuation. Prosus’s EUR 79.7 billion equity value plus EUR 5.95 billion net debt is roughly EUR 85.7 billion. Its Tencent stake alone is worth about EUR 100.8 billion at spot. A deliberately simple zero-structural-discount calculation leaves an implied value of roughly negative EUR 15.1 billion for every other Prosus asset combined. Against that, Prosus’s own current marks put gross ex-Tencent assets at approximately EUR 34.8 billion and about EUR 28.9 billion after central net debt.

The central valuation dispute is whether that negative implied ex-Tencent value represents mispricing, or merely the mathematical residue of a justified structural discount on Tencent and the holding company. Calling iFood, OLX, JET and PayU “free” skips that distinction. If the market rationally applies a 20%–30% haircut to the whole wrapper for control, China exposure, future central costs, private-asset uncertainty and capital allocation, the implied value of the operating assets becomes positive. The investment case rests on narrowing the justified structural haircut, not on pretending it should be zero.

The market has already rewarded some progress. The buyback began when the discount was around 54%; it is about 38.4% now. Yet the gap remains wide despite tens of billions of dollars of repurchases, the cross-holding unwind and the swing to profitable operating businesses. That persistence is information. It says the market does not view the discount as merely mechanical.

The qualitative portrait is a company in transition. The old Prosus was chiefly a discounted claim on Tencent with a collection of cash-consuming internet options attached. The current Prosus owns a somewhat smaller Tencent stake, has operating platforms that make real money, carries modest net debt after acquisitions and now has to prove it can run a large European delivery network. The next stage will be decided by whether ex-Tencent free cash flow grows fast enough to replace the economic dependence that the buyback is gradually reducing in absolute Tencent ownership.

Vertical history, financial review, and price narrative

Origins and listing path. Prosus was not born in the usual startup sense. It was the listed container created for Naspers’s international internet holdings. The legal lineage includes Myriad International Holdings N.V.; in 2019 the structure was reorganized and international internet assets were contributed into the vehicle ahead of the Amsterdam listing. Its economic DNA predates Prosus itself: Tencent, online classifieds, payments, food delivery and internet investments had been accumulated under Naspers.

Prosus began trading on Euronext Amsterdam on 11 September 2019 with a secondary listing in Johannesburg. Naspers owned 73.84% after listing. This was not a conventional cash-raising IPO in which an operating company sold newly issued shares to fund expansion; it was fundamentally a carve-out/capitalisation structure designed to give investors direct access to Naspers’s international internet assets and broaden their market base. A conventional “IPO proceeds” number is not an economically useful data point here, and I do not force one into the analysis.

The original capital-markets proposition contained the same tension that remains today. Naspers possessed an extraordinary asset in Tencent, but the value of that asset had become so large relative to the rest of Naspers that the parent’s own market value persistently lagged its underlying holdings. Prosus was supposed to create a cleaner global listed vehicle. The listing improved access and free float. It did not solve the discount.

The history since then divides naturally into five stages.

The inherited-portfolio stage. Before 2019, what later became Prosus was effectively Naspers’s international internet capital-allocation arm. The decisive success was Tencent; the other lasting strategic choice was to build or acquire local internet platforms in classifieds, payments and delivery rather than attempt to create one global consumer brand. This architecture still defines the company: its operating assets are mostly locally dominant networks, not one integrated worldwide product.

The Amsterdam-carve-out stage. The 2019 listing transformed a collection of Naspers holdings into a separately priced security. The market initially understood Prosus mainly as a more accessible Tencent proxy with venture assets attached. That framing was rational because Tencent dwarfed the rest of NAV. It also created an awkward benchmark for management: a prospective investor could often ask whether buying Tencent directly was cleaner than paying for a controlled holding company with private assets and central costs.

The structural-complexity stage. In 2021 Prosus acquired an economic interest in listed Naspers shares as part of a share-exchange arrangement that created a cross-holding. The transaction was intended to increase Prosus free float and reduce Naspers’s concentration on the Johannesburg market, but it made the ownership map harder to understand. Shareholders later voted to remove the structure, and the cross-holding was unwound in 2023.

In hindsight, this was an important capital-market lesson. The group attempted to solve a discount partly through structural engineering, but the extra complexity itself became one reason investors demanded a discount. The subsequent unwind was economically sensible simplification, yet current pricing shows that simplification alone did not eliminate the problem.

The discount-as-strategy stage. The strategic turn in June 2022 was more consequential. Prosus began selling small amounts of Tencent and using the proceeds to repurchase discounted Prosus shares, with Naspers conducting a linked programme. The discount at inception was around 54%. Rather than hoping the market would re-rate the wrapper, management started harvesting the spread directly.

The buyback turned the NAV discount from a symptom into an operating variable management could exploit. Repurchasing EUR 1 of NAV for materially less than EUR 1 mechanically raises NAV per remaining share. That explains why the programme can create value even if the headline discount never closes. It also explains its internal limit: every point of discount compression reduces the spread available for future accretion.

The operator stage. The latest phase began with the push to make the ecommerce portfolio profitable and accelerated after Fabricio Bloisi became group CEO in July 2024. Bloisi came with a credible operator’s résumé: he acquired iFood in 2013 when the business had roughly 20 people and built it into Prosus’s strongest directly operated consumer platform. Under his group tenure, Prosus has stressed local ecosystems, AI-enabled operations, tighter portfolio management and greater operating integration.

The acquisitions of Despegar and JET make that shift tangible. Despegar gives Prosus a travel platform in Latin America that can cross-sell into iFood’s customer base; by FY2026, 21% of Despegar Brazil B2C net revenue came from iFood customers. JET is a much bigger test because Prosus is now responsible for the economics of European food delivery rather than merely marking a minority investment.

Prosus has simultaneously reduced some public food-delivery exposure. The European Commission’s JET review focused partly on Prosus’s large Delivery Hero holding. Prosus subsequently reduced that interest; it sold another 5% of Delivery Hero in May 2026, 15.2 million shares at EUR 22 each for roughly EUR 335 million, after an earlier sale to Uber. Its current NAV still includes about USD 2.2 billion of Delivery Hero stock.

The important nodes can be condensed into a data chronology:

The chronology is drawn from Euronext, Prosus filings and offer disclosures; currency conversions use the 2026-08-19 ECB rate.

Financial vertical review. A conventional five- or ten-year consolidated revenue-and-margin series obscures more than it reveals here. Prosus has changed consolidation scope repeatedly through disposals and acquisitions; Tencent is equity accounted rather than consolidated; some assets moved from investment status into consolidation; JET adds six months of a very large revenue base in FY2026. The financially relevant history is the change in cash economics, not a mechanically stitched revenue CAGR.

The first era, through roughly FY2023, combined Tencent’s large equity-accounted profits with losses across many ecommerce investments. Accounting net income could also be dominated by disposal gains. In FY2022, for example, a Tencent stake sale generated approximately USD 14.6 billion of proceeds and a reported gain of USD 12.34 billion. Those numbers did not reflect recurring operating earnings from the ecommerce portfolio.

The second era is the profitability turn. Prosus reported that ex-Tencent free cash flow moved from negative USD 235 million in FY2024 to positive USD 36 million in FY2025. FY2025 core headline earnings reached USD 7.4 billion, up 47%, while ecommerce revenue was USD 6.2 billion, up 21%. The more important signal was that the operating businesses had ceased collectively consuming cash.

FY2026 moved beyond breakeven. Ecosystem adjusted EBITDA reached approximately EUR 1.12 billion, with profitability across the major ecosystems, while total Prosus free cash flow reached about EUR 1.29 billion. The composition matters more than the headline growth rate: OLX is a high-margin marketplace; iFood is becoming a profitable commerce-and-finance ecosystem; PayU is only just above EBITDA breakeven; JET adds huge revenue but relatively little EBIT.

The balance sheet remains sound but is no longer an argument for treating all M&A as costless. Current pro-forma cash is USD 9.5 billion, about EUR 8.2 billion, against USD 16.4 billion, about EUR 14.1 billion, of debt, leaving approximately EUR 5.95 billion of net debt. That is just 4.4% of gross asset value, a modest level for a portfolio with more than EUR 100 billion of listed Tencent stock. Prosus has plenty of financial flexibility; the economic risk lies in how that flexibility is spent.

Returns on consolidated equity or invested capital are similarly poor guides. Tencent’s equity-accounted carrying economics, large market-value movements, private investments and acquisition gains or impairments can dominate the denominator and numerator. Segment profitability and portfolio return against capital deployed are more useful. OLX’s 48% EBITDA margin and iFood’s EUR 345 million EBITDA demonstrate high-quality operating economics in their strongest assets. PayU’s 2.3% margin and JET’s near-breakeven adjusted EBIT show how far the weaker assets still have to go.

Management’s capital-allocation record is consequently mixed rather than uniformly excellent. Tencent is one of the great investment outcomes in public-company history, but that success belongs to the Naspers lineage rather than the current Prosus team alone. More recent investing has included expensive bets whose returns are harder to establish. Prosus paid roughly USD 1.74 billion for Stack Overflow in 2021; the current NAV sheet does not separately disclose a Stack Overflow value, instead grouping it with other unlisted holdings. That opacity is one reason a private-asset haircut is appropriate.

Bloisi’s operating record deserves more credit than the historical venture portfolio. iFood’s margin, loyalty penetration and fintech extension are concrete achievements. Despegar’s early cross-sell is evidence that a “local ecosystem” can create revenue beyond a slide-deck concept. JET will determine whether that model transfers to a mature and competitive European market.

Governance. Naspers remains the ultimate controlling shareholder. Prosus and Naspers share governance and strategic oversight, although the 2021 circular cross-holding has been eliminated. The control structure can support long-duration capital allocation without pressure from short-term activists, but minority shareholders have limited ability to force a different capital-return policy. That governance reality deserves some discount even after simplification.

Price and valuation narrative. Since listing, the stock’s fundamental valuation language has shifted less than the business has. Investors first priced it as a Tencent wrapper; during the China technology and regulatory selloff of 2021–2022, Tencent weakness, risk aversion and structural complexity pushed the Prosus discount wider. The June 2022 buyback created a second source of return: NAV-per-share accretion independent of a market re-rating. The 2023 cross-holding unwind removed one structural objection. By 2025–2026, operating profitability and M&A gave the market a third variable to price.

A market-price series cited by TradingView places the 52-week high near EUR 63.94 in November 2025; the 19 August 2026 close of EUR 37.895 is roughly 41% below that level. I would not attribute that move to a single cause. The relevant current facts are that Prosus is integrating JET, management is reinvesting in growth, and the discount remains close to 40% despite large buyback accretion.

I do not assign a numerical “historical discount percentile.” The public primary materials reviewed do not provide a clean, daily, adjusted Prosus NAV-discount history spanning listing, the cross-holding and subsequent share-structure changes. The defensible comparison is narrower: around 54% near the buyback’s 2022 launch versus 38.4% today. That is meaningful progress, but it falls well short of discount elimination.

Business model, moat, industry, and horizontal comparison

The economic balance sheet is the best starting point for understanding the business. Using Prosus’s 19 August NAV disclosure and the ECB’s 19 August exchange rate gives the following bridge. Prosus itself publishes EUR 129.4 billion of NAV and EUR 61.5 NAV per share using its own FX convention; small differences from the converted totals below are FX and rounding.

Prosus’s unlisted marks total USD 32.8 billion and are based on a mixture of analyst averages, post-money transaction values and internal valuations as of 30 June 2026. They are estimates, not executable bids. That is an important distinction when comparing the published EUR 61.5 NAV per share with a tradable share price.

How the operating machine works. Prosus’s operating businesses make money in different ways. Food delivery collects marketplace commissions, consumer fees, advertising, logistics revenue and increasingly financial-services revenue. Classifieds monetize listing visibility, dealer subscriptions, lead generation and adjacent services. Payments earns fees on processed transactions and, in parts of the portfolio, credit or financial-service spreads. The group is not an integrated product company in the conventional sense. Its common layer is ownership, capital, technology and customer-distribution infrastructure.

Current disclosed economics show the disparity:

† JET figures cover the six months consolidated by Prosus, so they should not be annualized without caution. ‡ Prosus disclosed iFood Pago revenue separately but not a clean like-for-like total-revenue figure suitable for this table. Currency conversions use EUR 1 = USD 1.1605.

OLX is currently the cleanest operating asset. Revenue of USD 992 million supported USD 481 million of adjusted EBITDA. Motors generated a 59% adjusted-EBITDA margin, real estate 46% and jobs 43%. Those margins are evidence of genuine marketplace economics: once a local classifieds network has enough buyers, sellers and professional advertisers, incremental digital inventory costs very little.

iFood has the strongest combination of growth and network density. Core food orders grew 8%; GMV rose 17% on a reported basis and 12% in local currency. Clube represented 45% of delivery volume by March 2026. iFood Pago revenue reached USD 463 million, about EUR 399 million, up 219% reported and 93% in local currency, and represented roughly a quarter of iFood revenue. This moves iFood from a simple restaurant-delivery marketplace toward a consumer-and-merchant financial ecosystem.

The strongest Prosus moats are local network liquidity at OLX and order density, loyalty and merchant integration at iFood; “the Prosus ecosystem” itself is still an emerging capability rather than a proven group-wide moat.

OLX’s moat has survived competitive periods because both sides of the marketplace care about local liquidity. Sellers list where buyers already search, and buyers search where inventory is broad. The 48% EBITDA margin is the financial signature of that effect. iFood adds logistics density: more orders can improve courier utilization, merchant economics and delivery times, while a loyalty programme makes customer frequency harder to dislodge. The extension into payments can deepen those economics if credit losses remain controlled.

PayU has a different advantage. Local payment rails, merchant relationships and regulatory infrastructure create switching friction, especially in India and other high-growth markets. Yet an economic moat should ultimately appear in returns. PayU’s first positive EBITDA year at only a 2.3% margin means that profitability evidence remains early. Adyen provides a useful quality benchmark rather than a direct business-model twin: in H1 2026 Adyen reported EUR 1.303 billion of net revenue, EUR 803.8 billion of processed volume and roughly a 50% EBITDA margin excluding one-offs. PayU has scale and local relevance, but it has not demonstrated Adyen-like unit economics.

JET’s moat is weaker because European consumers and restaurants can multi-home. Brand recognition and restaurant selection matter, but switching between delivery apps has little friction. Subscription programmes and cross-category ecosystems can improve retention, which gives Uber a structural advantage in markets where customers use both mobility and food. JET has to win through local density and better marketplace economics rather than pure lock-in.

The competitive operating backdrop is demanding. Delivery Hero reported FY2025 adjusted EBITDA of EUR 903 million, up 30%, and EUR 250 million of free cash flow; its 2026 target is EUR 910–960 million of adjusted EBITDA and more than EUR 200 million of FCF. Q1 2026 GMV was EUR 12.5 billion, up 8.8% like for like. It has already passed the “can delivery generate cash?” threshold that JET is still proving under Prosus.

Uber is the most dangerous cross-category reference. In Q2 2026 it produced USD 2.8 billion, about EUR 2.41 billion, of adjusted EBITDA, equal to 4.9% of gross bookings at group level. Its mobility and delivery businesses share users, subscriptions, payments and acquisition channels. That makes a marginal delivery order potentially cheaper to acquire than for a standalone food app.

DoorDash is the product-and-logistics benchmark. Q2 2026 adjusted EBITDA was USD 914 million, about EUR 788 million, or 2.8% of Marketplace GOV. Following its Deliveroo transaction, it also has a larger European strategic presence. So JET is being integrated while two global operators can spread technology and customer-acquisition costs over broader businesses.

Prosus nevertheless has a useful position in food delivery because it does not need every asset to win globally. iFood can dominate locally in Brazil while JET is managed as a European turnaround/growth asset and Delivery Hero remains a minority listed holding. The risk is that portfolio breadth becomes an excuse for accepting mediocre returns. Capital has to migrate toward networks with proven local density rather than perpetually funding share wars.

Classifieds provides a more attractive industry structure. Scout24’s H1 2026 private-customer ordinary operating EBITDA margin was about 60%, while its guidance points toward roughly 61% at group level. Scout24 is a narrower vertical property marketplace; OLX spreads across autos, property and jobs and across more varied countries. Scout24’s higher margin shows what a focused vertical can earn when lead value is high. OLX’s 48% margin is already excellent, with arguably more reinvestment and geographic growth optionality.

The industry-cycle exposure is consequently mixed. European food delivery has moved from the penetration-at-any-cost phase into a mature contest over density, subscriptions and profit. Latin American delivery still has more room for penetration and adjacent finance. Classifieds is less tied to one technology cycle but is exposed to housing, auto and employment transaction volumes. Payments benefits from long-run cash-to-digital conversion but is sensitive to transaction growth and credit conditions. Tencent adds China consumer, advertising and technology exposure that dominates NAV regardless of what the operating portfolio does.

Prosus is not highly cyclical in the industrial sense. Its main cycles are consumer activity, advertising, housing and auto transactions, capital-market risk appetite, rates and technology regulation. A weaker economy can depress OLX transactions and delivery order frequency, but the larger valuation risk historically has come from changes in the market’s required discount for China and holding-company exposure.

Regulation enters in several places. Tencent brings long-duration Chinese platform and geopolitical risk. PayU operates inside payments and credit regimes. JET and Delivery Hero are subject to European competition and worker/regulatory debates. The European Commission’s JET review already affected Prosus’s capital allocation by requiring a substantial reduction in the Delivery Hero position. These are recurring structural constraints rather than one-off headlines.

There is no single direct Prosus peer. Naspers is the parent and cannot be treated as an independent comparable. Operating peers answer narrower questions: Delivery Hero, Uber and DoorDash show what food-delivery scale can earn; Scout24 shows what a concentrated classifieds network can earn; Adyen shows what a payments platform with mature unit economics can earn. None provides a valid consolidated multiple for Prosus.

This is also why Prosus’s ecological niche is unusual. It is becoming an operator of local internet networks while retaining the balance sheet of an investment company. Its competitive advantage is capital plus existing local density. Its weakness is the inverse: one management team has to allocate capital across businesses with very different economics while a single Chinese associate still determines almost four-fifths of NAV.

Current fundamentals, valuation, and risk

What is happening now. FY2026 is the first reporting period in which the “new Prosus” is visible at meaningful scale. Operating profitability is no longer the main unknown. The harder question is how much of FY2026’s improvement can continue while Prosus spends on JET and other growth initiatives. The FY2026 release reported ecosystem adjusted EBITDA of USD 1.3 billion, up 84% on the headline basis and roughly 44% on Prosus’s comparable organic basis.

A public transcript of the FY2026 discussion indicates that management expects heavy FY2027 reinvestment and approximately flat EBITDA rather than another year of dramatic profit growth. Because that source is a secondary transcript rather than a primary guidance document, I treat the exact wording cautiously; the strategic message is consistent with the company’s stated investment programme. The market now needs proof that lower near-term profit growth is financing higher-quality growth rather than reopening the old cash-burn model.

JET is central to that proof. Before Prosus acquired it, JET reported FY2024 adjusted EBITDA of EUR 460 million versus EUR 339 million in 2023 and free cash flow before working capital of EUR 104 million, while its large net loss was heavily influenced by a EUR 1.002 billion impairment related to Grubhub. In early 2025, JET guided for EUR 360–380 million of adjusted EBITDA and approximately EUR 100 million of free cash flow before working capital, with a long-term adjusted-EBITDA target above 5% of GTV.

The first Prosus-consolidated period does not yet prove a return to that trajectory. Six-month adjusted EBITDA of roughly EUR 71.5 million and adjusted EBIT of only EUR 6.9 million leave little room for competitive mistakes. The reported selected-city pilots with order growth up to 25% indicate that higher investment can stimulate demand, but the economic test is whether those orders produce acceptable contribution margins after promotions and logistics.

Prosus is already pruning. JET announced its exit from Bulgaria effective September 2026, while Roberto Gandolfo succeeded founder Jitse Groen as CEO from January 2026. Those actions fit an operator-led portfolio rather than passive ownership, but there is not yet enough post-acquisition history to judge integration against a mature synergy run-rate.

The buyback remains the second live fundamental. The latest weekly disclosures show that repurchases are continuing; for example, Prosus bought about 2.19 million shares during 20–24 July 2026 at an average EUR 37.8363 for EUR 82.8 million. Euronext continued publishing Prosus repurchase notices into August, confirming both the programme and Amsterdam listing remain active.

The current balance-sheet constraints are modest: EUR 5.95 billion of net debt against EUR 135.6 billion of gross assets. This gives Prosus room to bridge temporary operating investment. It does not mean additional large acquisitions would be free. The appropriate monitoring variable is net debt as a percentage of gross asset value, currently about 4.4%.

What the market is trading. The current share price reflects four expectations simultaneously: Tencent’s market value, continued repurchase accretion, a structural holding-company discount, and skepticism about whether JET and the broader operating portfolio deserve Prosus’s marks. It is no longer enough for iFood and OLX simply to report better EBITDA. The market needs their cash generation to become large relative to the value being sold out of Tencent.

The Morgan Stanley view published in March 2026 is useful as expectation context rather than a conclusion. It argued that the discount had likely peaked and expected material upside, while estimating that about 54% of recent buyback funding came from Tencent and 46% from cash. Current pricing shows that a durable re-rating has yet to become fact.

I did not find a robust, date-matched series of aggregate analyst estimate revisions that would justify saying “consensus has been raised” or “consensus has been cut” for the entire group. Given Prosus’s changing consolidation perimeter, such a statement without a source-consistent dataset would create false precision.

Historical valuation. NAV discount is the relevant historical metric. Current price/NAV is 61.6%, equivalent to a 38.4% discount. That is materially narrower than the roughly 54% discount around the 2022 buyback launch. The re-rating reflects at least three real improvements: mechanical NAV-per-share accretion, elimination of the cross-holding and profitable ex-Tencent operations. The discount remaining near 40% indicates that the market still prices control, asset opacity, China concentration and capital allocation as structural rather than temporary.

Peer valuation and ex-Tencent marks. Since direct Prosus peer multiples are inappropriate, one useful sanity check is to divide Prosus’s own asset marks by current operating earnings. iFood’s USD 6.4 billion mark equals roughly 16 times FY2026 adjusted EBITDA. Combined OLX marks of about USD 8.7 billion equal roughly 18 times its USD 481 million EBITDA. Despegar is marked at roughly 13 times FY2026 adjusted EBITDA. PayU’s USD 3.3 billion mark is about 4.2 times revenue because its first-year EBITDA is too small to support a sensible earnings multiple. JET’s USD 4.2 billion mark is roughly 25 times annualized six-month Prosus-period EBITDA, although seasonality and the short consolidation period make that last figure only a rough diagnostic.

Those numbers do not make the private portfolio obviously cheap. OLX and iFood have sufficient margins and growth to support high values. Despegar has credible current earnings. PayU requires continued margin expansion. JET requires a large improvement in absolute profit. A blanket assertion that the entire ex-Tencent portfolio should trade at Prosus’s current marks overstates the evidence.

Cash-flow passthrough. The framework normally asks for five-year operating cash flow divided by net income and an owner-earnings P/E. I do not calculate those ratios for Prosus because they are economically misleading. Tencent earnings are equity accounted in Prosus net income while only Tencent dividends arrive as operating cash. Gains from Tencent share sales can enter accounting profit while the sale proceeds appear in investing cash flow. Acquisitions and changes in consolidation further break comparability. The FY2022 Tencent sale, with USD 14.6 billion of proceeds and a USD 12.34 billion accounting gain, illustrates why a consolidated OCF/net-income ratio would not measure cash conversion of the underlying businesses.

Maintenance versus growth capex is also not disclosed at a portfolio level with enough consistency to support a reliable split, particularly after JET. I do not invent an owner-earnings number or owner-earnings P/E. For Prosus, the closest useful equivalents are ex-Tencent free cash flow, central cash costs, net debt and the cash generated or consumed by each operating ecosystem. The shift from negative USD 235 million of ex-Tencent FCF in FY2024 to positive USD 36 million in FY2025 is consequently more informative than consolidated P/E.

Tencent disposal tax. Prosus’s current NAV marks Tencent at gross market value and does not show a separate disposal-tax liability. Historical disclosure also suggests that Tencent sales have not produced tax leakage large enough to dominate the economics: in FY2022, despite the very large Tencent disposal and accounting gain, reported group tax expense was only USD 97 million. That is evidence, not a definitive tax opinion, because group tax expense is not the same thing as a legal ruling on every future sale.

My base SOTP therefore assumes zero material incremental Tencent disposal tax, consistent with how Prosus presents NAV. I separately stress it. A hypothetical 2% leakage on the current Tencent market value would reduce pre-discount NAV by about EUR 2.02 billion, or EUR 0.96 a Prosus share. A 5% leakage would cost about EUR 5.04 billion, or EUR 2.40 a share. If future tax disclosure shows material leakage, the SOTP should be reduced immediately.

Absolute SOTP. I start from current spot values rather than consolidated earnings. The conservative case deliberately haircuts both assets and the residual holding-company valuation. This double layer is intentional: an illiquid asset can be worth less than its stated mark while the wrapper can separately deserve a discount for central costs, control and capital allocation. The central-cost reserves below are my assumptions, not management guidance.

† The EUR 0.28 FY2026 distribution was recommended by the board and remained subject to shareholder approval at the 26 August 2026 AGM as of the research base date.

This is valuation-scenario analysis within a research framework, not investment advice.

The conservative case asks what happens if Tencent falls 10%, listed minority interests deserve a 25% haircut, the private portfolio realizes only 60% of current marks, and Prosus still deserves a 30% wrapper discount. That produces EUR 34.75 a share, below the current EUR 37.895.

The base case is not aggressive about Tencent: it assumes no increase from the current spot mark. Other listed assets receive a 10% haircut, unlisted assets 20%, EUR 1.72 billion is reserved for the present value of central leakage, and the structural discount narrows only to 25%. Fair value is EUR 43.34.

The optimistic case needs several things to go right together: Tencent rises 15%, the private marks are broadly realized or modestly exceeded, and the Prosus discount falls to 15%. That produces EUR 58.65. The current company-published NAV of EUR 61.5 should not be mistaken for my base fair value; it is an asset-value starting point before applying uncertainty and holding-company economics.

At the current price, the market-implied gross value of the ex-Tencent portfolio is negative by roughly EUR 15.1 billion under a zero-structural-discount calculation, yet the conservative SOTP still produces only EUR 34.75 per share. Both statements can be true. The first shows how much skepticism is embedded in the wrapper. The second acknowledges that some of that skepticism is economically justified.

Expectation gap. The metric most capable of changing the market’s mind is no longer simply “ecommerce EBITDA positive.” That milestone has been reached. The next gap is between accounting profitability and distributable ex-Tencent cash. If JET moves from roughly EUR 72 million of six-month EBITDA and almost zero adjusted EBIT toward several hundred million euros of sustainable EBITDA, while iFood and OLX keep compounding, the private-asset haircut can fall. If FY2027 reinvestment instead pushes ex-Tencent FCF back below zero, investors will treat FY2026 as a temporary high-water mark.

The second expectation variable is the discount itself. A 38% discount creates large buyback accretion. A discount around 20% would be good for the stock’s re-rating but would make Tencent-funded repurchases much less powerful. Prosus has a self-liquidating source of excess return: successful discount closure reduces the attractiveness of the mechanism that helped close it.

The third variable is the Tencent ownership boundary. At about 22.6%, another few percentage points of absolute selling brings the holding near the 20% accounting presumption level. Prosus could still retain significant influence below that point, depending on governance and facts, but the market would need a clear accounting and strategic explanation.

Margin-of-safety recheck. Current EUR 37.895 is about 9% above the conservative EUR 34.75 fair-value point. Because the price sits above that point rather than below it, the margin of safety against the conservative case is zero.

The most fragile base-case assumption is discount compression. Today’s discount is 38.4%; my base valuation assumes 25%. If only 70% of that expected narrowing occurs, the ending discount would be about 29%. Applied to the EUR 57.79 pre-discount base NAV, fair value falls to roughly EUR 41.0, only about 8% above current price before distributions.

For the flat-three-year test, the cleanest holding-company equivalent is flat NAV per share and an unchanged discount. Under that deliberately static case, the shareholder receives essentially the distribution yield. EUR 0.28 divided by EUR 37.895 is about 0.74% annually before tax, versus a roughly 3.6% reference yield on long-dated euro-area government debt around the research date. Buyback accretion could improve that return, but including it would require assuming continued asset sales and an unchanged wide discount. On the strict static test, there is no margin of safety at this buy price.

This is not a simple “good company, bad price” case because Prosus is not one homogeneous operating company. It is a high-quality Tencent stake, two increasingly attractive operating assets, several less-proven businesses and a controlled holding structure. The current price is reasonable against my base case but does not satisfy the framework’s conservative buying discipline.

Margin-of-safety sufficiency verdict: none.

Permanent-loss risks. Tencent concentration remains the largest risk. Probability is medium and impact is high. A 30% decline in Tencent alone would remove roughly 23% of Prosus NAV before any change in the holding-company discount because Tencent is about 78% of NAV. If a China or geopolitical shock simultaneously widened the Prosus discount, price damage would exceed the NAV effect. The observable indicators are Tencent’s market value, regulatory developments and Prosus’s absolute ownership percentage.

Persistent discount and control risk has high probability and medium-to-high impact. Tens of billions of dollars of buybacks, the 2023 structural simplification and profitable ecosystems have narrowed but not eliminated the discount. Naspers remains the controlling parent. A discount staying above 45% despite good operating results would suggest the market regards control and structure as permanent rather than fixable. The transmission mechanism is almost entirely valuation: underlying assets can compound while Prosus shareholders capture substantially less than their marked value.

JET integration has medium probability and high impact. The direct asset is only a few percent of NAV, but failure would damage more than its own valuation because JET is the flagship proof of Prosus’s new operator strategy. The key indicators are adjusted EBITDA relative to revenue, adjusted EBIT, order growth outside pilots and cash flow. Two Prosus reporting periods with JET EBITDA margins below 3%, or renewed negative adjusted EBIT after heavy investment, would indicate that competitive intensity against Uber, Delivery Hero and DoorDash is consuming the intended return.

Capital-allocation risk has medium probability and high impact. Prosus can currently fund investments easily because net debt is only 4.4% of gross assets, but that flexibility creates temptation. Another multibillion-euro acquisition before JET generates convincing returns would weaken the case that management has changed from venture-style expansion to return-disciplined capital allocation. Net debt above roughly 8% of gross asset value, combined with slower buybacks, would be an early warning.

Private-asset valuation risk has medium probability and medium-to-high impact. USD 32.8 billion of unlisted holdings are not quoted markets. Prosus uses analyst estimates, recent financing values and internal marks. A series of asset sales below those marks would reduce NAV directly and cause investors to increase the discount applied to the remainder. The most useful observable variable is realized sale value versus the immediately preceding NAV mark.

Foreign exchange adds volatility but is secondary to those permanent-loss channels. Prosus is quoted in EUR, reports in USD, owns a Tencent stake priced in HKD, and generates operating earnings in currencies including BRL and INR. Currency can move reported NAV materially; the underlying economic risk becomes permanent mainly when FX reflects deterioration in local business economics or capital controls rather than normal translation.

Catalysts, tracking dashboard, and cross-synthesis

Positive catalysts are easy to identify but need to be judged by economic consequence. A sustained discount below 30% would show that buybacks and operating evidence are changing the market’s required holding-company haircut. JET order gains accompanied by a rising EBITDA margin would validate the acquisition thesis. Continued double-digit local-currency growth at OLX with margins around the high-40s, and iFood order growth above mid-single digits while EBITDA grows, would increase the value of the ex-Tencent portfolio without needing higher valuation multiples. Asset disposals at or above current NAV marks would validate private valuations. A stronger Tencent price would amplify all of these because Tencent remains the dominant NAV component.

Negative catalysts are the mirror image only in part. FY2027 ecosystem EBITDA falling materially below FY2026 would undermine the profitability turn. JET generating higher orders through subsidies while EBIT falls would imply bought rather than profitable growth. A major acquisition before the current integration cycle matures would reopen capital-allocation concerns. A drop below 20% Tencent ownership without clear accounting and governance guidance could create uncertainty even if no economic event occurs automatically. A Tencent selloff plus discount widening remains the fastest path to large mark-to-market loss.

The tracking dashboard below focuses on variables that can distinguish those outcomes.

† Revenue-margin measure, not directly comparable with JET’s historic EBITDA/GTV target. ‡ Prosus’s current investor calendar had not published a confirmed HY2027 results date in the material reviewed. HY2026 results were released on 24 November 2025, so approximately 24–25 November 2026 is a cadence-based expectation, not company guidance.

The discount and Tencent ownership should be tracked together. A falling discount is positive for immediate market value but reduces buyback accretion. A falling Tencent percentage can be positive per Prosus share while the discount is wide, yet it raises strategic questions as the absolute holding approaches 20%. Looking at either metric alone misses that interaction.

JET EBITDA needs to be read alongside orders. Higher orders with a lower margin would show that Prosus is buying volume. Higher orders with a stable or rising margin would show that investment is improving density. OLX has the opposite burden of proof: at a 48% margin the risk is over-harvesting, so moderate reinvestment that keeps local-currency revenue in double digits can be value creating even if the margin does not expand every year.

PayU’s next few points of margin matter disproportionately. A payments business with EUR 673 million of revenue and 2.3% EBITDA margins has far more operating leverage than one already at 40%–50%. Positive EBITDA is only the first threshold; the investment case needs to show that scale produces cash rather than simply more transaction volume.

Cross-synthesis. Vertically, Prosus has proven two different capabilities, but at different levels of confidence. The first is inherited capital allocation: the Naspers lineage recognized extraordinary internet-platform economics early enough for Tencent to become the dominant source of group wealth. No other Prosus investment has approached that outcome, and the current company should not be valued on the assumption that another Tencent can be found.

The second capability is more recent and more relevant to current management: taking local internet networks to profitability. iFood and OLX are the evidence. They are not merely “promising assets.” iFood produces hundreds of millions of euros of EBITDA while still growing orders, and OLX earns a 48% EBITDA margin. Despegar is already profitable and is beginning to show cross-platform customer acquisition. PayU has crossed breakeven. These facts justify valuing the ex-Tencent portfolio above zero independently of venture optimism.

Past success came from a mix of capital-allocation skill, an extraordinary era of global internet penetration and a highly asymmetric Tencent outcome. Luck mattered in the sense that no allocator can repeatedly plan a multi-decade return of Tencent’s magnitude. Skill mattered because Naspers was willing to hold the position long enough for compounding to dominate the group. The relevant question now is whether Prosus can convert the resulting financial capital into operating returns without destroying the scarcity value of that original asset.

The buyback is a rational answer while the discount remains wide. At today’s price/NAV relationship, selling some Tencent to repurchase Prosus does not necessarily reduce Tencent exposure per Prosus share. That is genuine arithmetic value creation, not financial marketing. The criticism becomes stronger as the discount narrows and the absolute ownership approaches accounting and strategic thresholds.

A useful way to frame the endpoint is to separate corporate liquidation from shareholder economics. At the corporate level, Prosus is slowly liquidating part of Tencent. At the per-share level, it can increase Tencent value and total NAV per remaining share because it retires even more discounted claims. Both descriptions are simultaneously true. The policy ceases being clearly attractive when the repurchase price approaches the Tencent value embedded per Prosus share, when material tax leakage appears, when the Tencent holding loses strategically useful influence, or when alternative uses of cash offer better risk-adjusted returns.

The present crossover is far away but no longer theoretical. Tencent is about 77.7% of NAV; Prosus trades at about 61.6% of NAV. The rough pure-Tencent-funding threshold corresponds to a Prosus discount near 22%. If the discount approached that level without changes in the asset mix, management should increasingly fund repurchases from operating FCF or non-core asset sales rather than Tencent.

Horizontally, Prosus’s edge is not superior technology across every business. Uber has a stronger cross-category consumer network in global delivery. DoorDash has a formidable logistics/product culture. Delivery Hero already generates meaningful delivery cash flow. Adyen has vastly better payment margins. Scout24 has a more focused and higher-margin vertical-classifieds model. Prosus’s advantage is the ability to own locally dense networks, supply capital and increasingly share distribution between them.

That advantage is clearest in Latin America. iFood can send customers to Despegar and potentially to financial services; 21% of Despegar’s Brazilian B2C net revenue already coming from iFood customers is evidence of real distribution value. It is much less proven in Europe. A German or British JET customer does not automatically become part of a Prosus ecosystem in the way an iFood user can become a Pago or travel customer. Europe is where the “ecosystem” thesis must survive contact with mature competition.

The market is probably right to demand proof. JET’s current six-month adjusted EBIT of roughly EUR 7 million gives Prosus little room for an expensive European price war. The acquisition can succeed if higher order density lifts contribution profit and central efficiencies emerge. It can also become a persistent capital sink if Uber, DoorDash/Deliveroo and Delivery Hero are willing to sacrifice near-term margin to defend consumers and restaurants.

The Naspers control structure is the second reason the discount should not go to zero. The 2023 cross-holding unwind solved circularity, not control. Minority Prosus shareholders still delegate capital allocation to a board linked closely with Naspers. A 10%–20% structural discount could therefore remain even if every operating asset is valued transparently. My optimistic valuation assumes 15%, not zero.

The market’s likely misjudgment is subtler than “ex-Tencent is free.” At current price, a zero-wrapper-discount calculation makes the ex-Tencent portfolio look worth negative EUR 15.1 billion. Yet the operating evidence from OLX, iFood and Despegar makes a literally negative value implausible. The market is effectively imposing a large combined haircut across the Tencent wrapper, central structure and private assets. The opportunity exists if that combined haircut proves too punitive.

The counterweight is that a structural haircut is warranted. Current unlisted marks total EUR 28.3 billion and are not liquid market prices. JET is marked below purchase cost but still at a valuation requiring a meaningful operating recovery. PayU’s current profitability is tiny. Central costs do not disappear simply because they are not allocated to an asset mark. Tencent itself carries geopolitical and China-platform risk. The entire 38% discount cannot be called irrational.

Over the next 12 months, three variables dominate: JET integration economics, the trajectory of ecosystem EBITDA during the reinvestment year, and the buyback’s funding mix. A narrow market debate over Tencent’s daily share price misses those Prosus-specific variables.

Over three years, the emphasis moves to Tencent ownership around the 20% level, sustainable ex-Tencent free cash flow and leverage. Prosus should be able to demonstrate by then whether the absolute reduction in Tencent is being replaced by higher per-share value from operating businesses. A three-year period in which Tencent ownership falls materially while JET consumes cash and OLX/iFood growth slows would expose the buyback as harvesting past value faster than new value is created.

Over five years, the question is corporate identity. A successful Prosus would still own a very valuable Tencent stake but would no longer need Tencent to explain almost all of its NAV growth. iFood, OLX, fintech, travel and a repaired European delivery operation would generate enough FCF to fund a meaningful share of capital returns. A failed version would hold less Tencent, a collection of merely average operating assets and the same persistent governance discount.

The best long-run Prosus outcome requires the company to use today’s wide discount to increase per-share value while converting the ex-Tencent portfolio into a source of cash before the absolute Tencent holding becomes strategically smaller.

Core bull reasons.

• The 38.4% discount to published EUR 61.5 NAV gives Prosus considerable buyback accretion, and the programme has already retired about 30% of free float according to the company’s September 2025 measurement.
• The ex-Tencent portfolio has crossed the profitability threshold: OLX earns a 48% EBITDA margin, iFood produced about EUR 345 million of FY2026 adjusted EBITDA, and ex-Tencent FCF had already turned positive in FY2025.
• A zero-structural-discount calculation assigns roughly negative EUR 15.1 billion to the ex-Tencent assets even though Prosus marks them at roughly EUR 28.9 billion after net debt, creating a large expectation gap if operating marks prove broadly realizable.
• Because Tencent’s 77.7% NAV weight exceeds Prosus’s 61.6% price/NAV ratio, Tencent-funded repurchases can still increase Tencent value per Prosus share while absolute ownership declines.

Core bear reasons.

• EUR 37.895 is roughly 9% above the EUR 34.75 conservative SOTP point, so current price has no cushion against the deliberately conservative asset and discount assumptions.
• Tencent still represents about 78% of NAV while Prosus’s absolute holding is already around 22.6%, leaving only a few percentage points before the commonly used 20% associate presumption becomes a live accounting and governance issue.
• JET’s first six Prosus-consolidated months generated only about EUR 72 million of adjusted EBITDA on EUR 1.64 billion of revenue and almost no adjusted EBIT, leaving its EUR 3.6 billion NAV mark dependent on a large future improvement.
• Naspers remains the controlling parent and the discount survived both the cross-holding unwind and massive buybacks; part of the discount is structural, not simply an arbitrage waiting to close.
• Prosus’s EUR 28.3 billion of current unlisted marks are partly analyst, financing-round or internal estimates rather than tradable prices, so simultaneous markdowns can hit both NAV and the market’s confidence in NAV.

Pre-mortem. One plausible three-year failure script begins in European delivery. During 2027–2028, Uber, DoorDash/Deliveroo and Delivery Hero intensify subscription and restaurant promotions in JET’s major markets. JET’s pilots continue producing order growth but only through customer incentives. Its adjusted-EBITDA/revenue margin falls from the current roughly 4.4% to around 1%, and the market value assigned to JET falls from EUR 3.6 billion toward EUR 2 billion. At the same time Tencent declines 30%, Prosus ownership crosses below 20%, and private-asset marks are cut 20%. If the Prosus discount then widens to 50%–55%, the shares could plausibly trade around EUR 20–22, roughly 42%–47% below the current price. This is a stress script, not a forecast.

A second script is capital-allocation driven. Prosus commits another EUR 4–5 billion to a large acquisition before JET has proven sustainable cash generation. Net debt rises above 8%–10% of gross asset value, the buyback is slowed to protect liquidity, and ex-Tencent FCF turns negative. Tencent falls only 15%, but investors conclude that monetizing the best asset is financing lower-return acquisitions. A discount above 50% could again place the stock in the low EUR 20s even without a severe Tencent bear market. This scenario is speculative; the trigger to watch is acquisition spending combined with leverage, not acquisition headlines alone.

Final research conclusion. Prosus at EUR 37.895 is a much better business than the “Tencent plus cash-burning ventures” vehicle it was several years ago. OLX and iFood have proven economic moats, the broader operating portfolio has crossed into positive cash generation, the cross-holding has been removed and the buyback has created real NAV-per-share accretion. The market still values Prosus at only about 62 cents per published euro of NAV. That is a meaningful discount.

The current price nevertheless fails the conservative margin-of-safety test. My SOTP reaches EUR 34.75 in the conservative case, EUR 43.34 in the base case and EUR 58.65 in the optimistic case. New money at EUR 37.895 relies on some combination of discount compression, stable Tencent and successful operating execution. JET is the largest unresolved operating test, while the gradual approach toward 20% Tencent ownership is the largest strategic question created by the buyback itself.

For an existing shareholder, the arithmetic supports continuing to own Prosus while the buyback remains strongly accretive and ex-Tencent cash generation improves. For a new investor applying a strict balanced-risk margin-of-safety discipline, the current quote is not sufficiently low. The valuation becomes distinctly more attractive in the high EUR 20s, where the price would stand more than 20% below my conservative fair-value point and would leave room for both operating disappointment and a persistent structural discount.

【Company-profile scores】

• Fundamental quality: medium
• Growth: medium
• Moat: medium
• Financial soundness: strong
• Management credibility: medium
• Valuation attractiveness: medium
• Risk level: medium
• Suitable investor type: value

【Investment rating】

• Rating: Hold
• One-line thesis: At EUR 37.895, the 38% NAV discount is real, but conservative SOTP offers no cushion while JET integration and Tencent-funded buybacks remain unresolved.
• Current-price classification: acceptable hold.
• Whether to wait for a better price: yes for a new position. The disciplined entry zone is EUR 26.0–27.8 provided net debt remains below 8% of gross assets, ex-Tencent FCF stays positive and JET does not deteriorate materially. The opportunity cost is that the base case offers roughly 15% one-year total return and the stock may never trade into the buy zone.
• Target holding horizon: 3–5 years.
• Expected annualized return: if the scenario values are reached in roughly one year, conservative -7.6%, base +15.1%, optimistic +55.5%, including the proposed EUR 0.28 distribution. If convergence instead takes three years and the distribution stays at EUR 0.28 annually, the rough annualized returns are approximately -2%, +5% and +16%, respectively.
• Max-loss risk: approximately 45%–50%, toward roughly EUR 19–21, if a Tencent drawdown, JET failure/private-asset markdowns and a 50%–55% NAV discount occur together.
• Reassessment trigger: NAV discount above 45% for a sustained period despite continuing repurchases and satisfactory operating results.
• Reassessment trigger: Tencent ownership below 20% without clear disclosure on associate accounting, board influence and future sale policy.
• Reassessment trigger: JET adjusted-EBITDA/revenue below 3% for two Prosus reporting periods, or adjusted EBIT turning materially negative.
• Reassessment trigger: net debt rising above 8% of gross asset value.
• Reassessment trigger: ex-Tencent free cash flow returning to negative territory.

【Ideal Buy Price】26.0–27.8 EUR

Basis: 20%–25% below the EUR 34.75 conservative SOTP point, providing the explicit margin of safety required by the framework.

Acceptable hold price: 37.0–49.8 EUR, corresponding approximately to ±15% around the EUR 43.34 base value.

Clearly overvalued price: 64.5–70.0 EUR, beginning roughly 10% above the EUR 58.65 optimistic fair-value point.

【Valuation Range】

• current: 37.895 EUR (close as of 2026-08-19)
• bear (conservative · ideal buy zone): [26.0, 27.8]
• base (fair · acceptable hold zone): [37.0, 49.8]
• bull (optimistic · above the clearly-overvalued line): [64.5, 70.0]

Research uncertainties and sources

Five uncertainties could materially change the conclusion.

First, the tax treatment of future Tencent disposals is not stated in the latest NAV material as a simple per-sale tax rate. Prosus’s current NAV is presented without a disposal-tax reserve, while historical sales appear not to have incurred large visible tax leakage. I use zero incremental leakage in the base case and provide 2% and 5% sensitivities rather than claiming a legally certain zero rate.

Second, the roughly 22.6% Tencent ownership estimate combines Prosus’s 14 August 2026 holding with Tencent’s 31 July 2026 issued-share denominator. Tencent’s own buybacks mean the exact contemporaneous percentage may be marginally different. The economically important fact is that the holding is in the low-22% range and approaching, but remains above, 20%.

Third, no automatic consequence at precisely 20% was identified in the primary materials reviewed. Prosus’s own accounting language treats 20%–50% as a typical associate range, while significant influence depends on facts and governance. Any exact consequence for accounting classification, board representation or index treatment should be rechecked when the stake gets materially closer to that level.

Fourth, USD 32.8 billion of unlisted assets are marked through a combination of analyst estimates, transaction values and internal methods. Those are reasonable NAV inputs but not guaranteed realizable values. The base case’s 20% private-asset haircut and conservative case’s 40% haircut are intended to absorb some of that uncertainty.

Fifth, the often-cited 54% Tencent / 46% cash split for recent buyback financing is a Morgan Stanley estimate rather than a Prosus-reported funding schedule. Prosus has separately confirmed using proceeds from other asset disposals to accelerate repurchases, so the direction toward diversified funding is supported, while the exact split is not a company fact.

The main primary research base is Prosus’s 19 August 2026 NAV statement, which supplies current asset marks, Tencent shares held, net debt, NAV per share and the net-share denominator.

FY2026 operating analysis is based primarily on Prosus’s 29 June 2026 results release and accompanying results materials, including the disclosed economics of iFood, OLX, PayU, Despegar and JET.

Capital-return analysis uses Prosus’s open-ended buyback disclosure and CEO communication, supplemented only where explicitly identified by secondary sell-side reporting.

The corporate-history and ownership review relies on Euronext’s 2019 listing materials and Prosus’s disclosures on the later Naspers/Prosus structure and its unwind.

JET analysis uses the original Prosus offer, final acceptance/delisting notice, JET’s pre-acquisition financial disclosures and the European Commission’s competition review.

The operating-peer work is based on current company disclosures from Delivery Hero, Uber, DoorDash, Scout24 and Adyen rather than third-party multiple databases.

Current EUR/USD and EUR/HKD conversions use the ECB’s 19 August 2026 reference rates. The current Prosus close is the 19 August 2026 Amsterdam close, corroborated across market-data sources; Euronext confirms the continuing primary listing.

Other tickers mentioned

• 0700.HK: Tencent, the dominant listed asset and roughly 78% of Prosus NAV.
• NPN.JSE: Naspers, Prosus’s controlling parent and the vehicle in which a second-level “discount on a discount” can arise.
• DHER.XETRA: Delivery Hero, food-delivery competitor and a remaining listed Prosus investment.
• UBER.US: global mobility-and-delivery competitor whose shared consumer ecosystem is a strategic benchmark for JET.
• DASH.US: delivery and logistics competitor with expanded European exposure through Deliveroo.
• G24.XETRA: Scout24, high-margin European classifieds reference for OLX.
• ADYEN.AS: payments benchmark showing the margin potential of a scaled global payments platform relative to PayU.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.