Prosus at a 35% Discount: The NAV Gap the Buybacks Are Closing

Prosus trades at roughly a 35% discount to its full net asset value, around 20% if you count only the listed assets. That gap is the entire argument. At €42.60 against a consensus average target of €64.99, the market is pricing in a discount that the company’s own capital allocation has been narrowing for years, not widening. The question is not whether the underlying assets are worth more than the price says. It is whether the mechanism that closes the discount is still working. It is.

Close to 80% of NAV sits in the Tencent stake, a single listed asset that can be marked to market every trading day. That concentration cuts both ways, and I will get to the cut that hurts. But it also means the discount-to-NAV is unusually clean to measure. You are not guessing at the value of private holdings buried in a pyramid structure. You are looking at a listed position, applying a discount, and asking why a stake you could replicate by buying Tencent directly is being offered at a third off. When the gap between look-through value and share price is that observable, the path to closing it is mechanical: buy back shares below NAV, and every euro spent accretes NAV per share for the holders who stay.

That is precisely what the open-ended repurchase programme has been doing. Since mid-2022 the buybacks have driven cumulative NAV per share accretion of about 18%. The arithmetic is not subtle. When your own shares trade at a 35% discount to the assets behind them, retiring them is the highest-return capital deployment available, and management has been running it in continuous tranches rather than episodic announcements. Late March saw multiple tranches, including roughly 2.64 million shares bought around €40.25 over a single week. Each tranche below NAV does the work that a re-rating would do, except it does it whether or not the market cooperates.

The portfolio is also being actively reshaped, which matters more than the buyback alone. Monetization is the second lever that collapses a discount: an outright sale crystallizes value the market was refusing to credit. The disposal of a 4.5% Delivery Hero stake to Uber for around €270 million was a small but clean example. More interesting is the reversal that followed. By late May, Prosus was exploring an increase in its Delivery Hero position and secured an EU antitrust waiver on a prior sell-down commitment, which triggered a session move above 9%. That is management treating the food-delivery holdings as a position to consolidate rather than unwind, and the market repriced the optionality within a day.

The look-through earnings picture underneath supports the patience. Reported diluted EPS of €5.76 on trailing twelve months grew over 70%, and operating margin moved from 2.80% in FY2025 to 3.81% on a trailing basis. Those are thin absolute margins, but the direction is what counts in a portfolio shifting weight from capital-intensive legacy commerce toward higher-margin software assets. Free cash flow of €1,837 million in the latest fiscal year is the number that pays for the buybacks without forcing asset sales at the wrong moment. The discount narrows from two sides at once: NAV per share rising through accretion, and the gap to that NAV closing through monetization.

This is the setup value investors have historically missed in conglomerate discounts. The reflex is to treat the discount as permanent, a structural tax for owning assets through a wrapper, and to walk away. What that reflex misses is the difference between a static discount and a closing one. A discount that sits flat while management does nothing deserves to persist. A discount being compressed by continuous below-NAV repurchase and selective monetization is a different instrument entirely. The error in similar holding-company setups was always extrapolating the discount level instead of tracking its direction.

The backdrop is not frictionless, and it is European-specific. The EU AI Act’s phased obligations, including transparency rules and high-risk system requirements landing around August 2026, raise compliance costs for a group with data platforms and AI-related investments across the continent, with fines that can reach 7% of global turnover. The proposed Digital Omnibus may soften the timeline, but the direction is more regulatory weight, not less. Across the Atlantic the US federal framework is moving toward a lighter national standard that preempts state patchworks, which cuts the other way for the group’s American exposure. Neither regime is a thesis-breaker on its own. Together they mean the operating assets carry a compliance overhead that did not exist three years ago, and that overhead has to be absorbed inside those thin margins.

Now the cut that hurts. The 80% Tencent concentration is the same lever that closes the discount in reverse. If Tencent re-rates down, NAV falls with it regardless of how many shares Prosus retires, and the buyback’s accretion gets swamped by the mark-to-market loss on the dominant holding. The March session where Prosus led the Stoxx 600 on a 7.3% Tencent rally is a reminder that the share price is, on most days, a leveraged proxy for one Chinese internet company and the regulatory mood around it. The second drag is closer to home: the CEO’s warning in May that accelerated iFood investment would pressure FY2027 adjusted earnings sent the stock to its period low. Spending into Brazilian food delivery is defensible, but it consumes the free cash flow that funds the discount-closing buyback. The two uses of cash compete.

If the buyback pace falls and the discount to NAV widens back toward 45% while Tencent stays flat, the mechanism I am relying on has stopped working and this case breaks down. Until then, the discount is being closed from both ends, and €42.60 against a €64.99 average target is paying me to wait while it happens.

Figures reflect the most recent available data and may differ slightly from live market prices. · © Mathstock