GRAB Grab's $600M Bet on the One Deal Uber Couldn't Get Past Taiwan's Regulator

An Expansion Built Entirely on Someone Else's Rejection

On March 23, 2026, Grab Holdings agreed to pay Delivery Hero $600 million in cash - on a cash-free, debt-free basis - for foodpanda's food-delivery business in Taiwan. It's Grab's ninth market and its first outside Southeast Asia in the company's eight-year operating history. On its own, that's a straightforward geographic-expansion story: a profitable Southeast Asian super app taking its playbook somewhere new.

What makes it worth a full post is the fifteen months of history sitting directly underneath it. In December 2024, Taiwan's Fair Trade Commission (FTC) blocked Uber's own $950 million bid for the exact same asset, ruling that it would hand Uber Eats roughly 90% of Taiwan's food-delivery market and eliminate the only real competitive pressure it faced. Uber walked away in March 2025, paying Delivery Hero a reported $250 million termination fee rather than fight the ruling. Grab is now buying the asset Taiwan's own regulator said couldn't go to its biggest rival - and, as it turns out, Uber never fully left the picture (see Beyond the Usual below).

What Grab Is Actually Buying

The deal is narrow and specific: foodpanda's Taiwan delivery operations only, not any other Delivery Hero market, and not any stake in Delivery Hero itself. Per Grab's own announcement:

Deal term	Detail
Price	$600 million cash, cash-free/debt-free basis, subject to customary closing adjustments
Target	foodpanda's Taiwan delivery business only
Seller	Delivery Hero SE
Target's 2025 GMV»	~$1.8 billion (management accounts, before allocation of Delivery Hero group costs)
Target's profitability	Profitable on an Adjusted EBITDA» basis before group-cost allocation, per Grab's disclosure
Expected incremental Adjusted EBITDA	At least $60 million in 2028
Closing	Expected H2 2026, subject to regulatory approval
Post-close footprint	Presence across 21 cities in Taiwan
Transition	A Support Services Agreement under which Delivery Hero continues operating foodpanda Taiwan until Closing, then provides transition support afterward

Grab explicitly reiterated its 2026 Adjusted EBITDA guidance of $700-720 million alongside the announcement - the deal is framed as additive to an already-intact plan, not a bet-the-company move funded by cutting something else.

How Big Is This, Really?

foodpanda Taiwan's disclosed $1.8 billion in annual GMV is a real business, but it's small next to what Grab already runs. Grab's own Q2 2026 results, covered separately, put quarterly Deliveries GMV at $4,249 million - meaning foodpanda Taiwan's full-year GMV is worth roughly 42% of one quarter of Grab's existing Deliveries segment, or in run-rate terms, about a tenth the size of Grab's current Deliveries business. This is a bolt-on, not a merger of equals - Taiwan adds real revenue, but it doesn't meaningfully change Grab's overall scale. What it changes is geographic diversification: every dollar of Grab's GMV today comes from eight Southeast Asian countries with correlated currency, regulatory, and macro exposure. Taiwan is Grab's first data point on whether its operating model - dense-city delivery logistics, driver-partner management, an AI-enabled merchant stack - travels outside that specific region at all.

Why Taiwan Is an Interesting Market to Enter

Grab's own framing, from CEO Anthony Tan: a population of about 23 million with "high demand for mobile-first services, similar to the Southeast Asian consumers Grab serves every day," and cities dense enough that Grab's logistics experience should transfer directly. A few things make that framing more than just talking points:

• Taiwan is a genuine two-player market, not a fragmented one. foodpanda and Uber Eats have run the entire category between them for years - roughly 52% foodpanda to 48% Uber Eats as of the most recent published market-share data (Jan 2022-Aug 2023 window, per Taiwan FTC materials cited in regional reporting). A duopoly with a stable, dominant incumbent is a far more predictable entry than a market with five undercapitalized local apps burning cash to buy share.
• The GDP and urbanization profile fits Grab's existing operating model closely. Grab's entire business is built around dense Southeast Asian cities; Taiwan's largest metro areas (Taipei, Taichung, Kaohsiung) share that density without Grab needing to build a new playbook from scratch, unlike, say, entering a sprawling low-density US market would require.
• It's a stated first step, not a one-off. Grab called this its "first [market] outside of Southeast Asia" in the announcement's own headline framing - language a company doesn't typically use for a single opportunistic asset purchase. Whether Taiwan is actually the first of several such moves, or a one-off experiment that stays exactly at this scale, is the real thing to watch over the next few years, not this deal on its own.

Who's Already There, and How Grab Could Screw This Up

Grab enters as a pure new-brand acquirer of an existing #1 asset, not a challenger building share from zero - which is a real advantage, but not a guarantee. The risks fall into three buckets:

Competitive. Uber Eats isn't going anywhere - it's the #2 player with roughly 48% share already, deep pockets from a $14.8 billion parent-level acquisition of the rest of Delivery Hero (more below), and every incentive to fight hard for share against a new operator learning the market. A brand transition - foodpanda Taiwan's app, riders, and merchant relationships migrating onto Grab's platform - is itself an execution risk on any acquisition like this; if the migration is clumsy, Uber Eats has a specific, time-limited window to pick off switching-fatigued users and merchants.

Regulatory. The FTC has already extended its review once (see Beyond the Usual), and a drawn-out or conditioned approval - remedies, behavioral commitments, a longer timeline than Grab's own H2 2026 closing target - is a live possibility, not a hypothetical one, given the regulator already blocked the previous attempt at this exact asset.

Structural. foodpanda Taiwan's disclosed profitability is stated "before the allocation of Delivery Hero's group costs" - a caveat in Grab's own press release, not an external criticism. That means the standalone business, once it's carrying its own corporate overhead as part of Grab rather than sharing Delivery Hero's, could look less profitable on day one than the pre-close numbers imply. Whether Grab's own Deliveries-segment infrastructure can absorb Taiwan without adding a comparable overhead burden is the real test of whether the deal's economics hold up as disclosed.

Why Is Delivery Hero Selling?

Delivery Hero's own framing, from CEO Niklas Östberg, in the announcement: this is "a key first step in our ongoing strategic review" toward "a more focused global footprint." That review didn't stay small. Roughly four months after this Taiwan deal was announced, Uber agreed to acquire essentially all of the rest of Delivery Hero - reportedly around $14.8 billion, excluding Taiwan - with Delivery Hero divesting operations in over a dozen other markets to clear antitrust hurdles on that much larger transaction. Read against that later announcement, the Grab deal wasn't really Delivery Hero selling one underperforming market - it was the opening move of Delivery Hero's own full breakup, and Taiwan was carved out to Grab specifically because the exact same asset going to Uber (as part of the bigger deal) would have revived the same near-monopoly concern that already killed Uber's standalone 2024 bid for it.

That reframes the whole transaction: Grab isn't opportunistically picking up a market Delivery Hero didn't want. It's the designated non-Uber buyer for the one piece of Delivery Hero that competition law wouldn't let Uber have twice.

This also isn't the first time a Grab-Delivery Hero tie-up has been rumored. On February 21, 2024, Delivery Hero officially terminated talks over a much larger regional sale - foodpanda operations across Singapore, Malaysia, the Philippines, Thailand, Cambodia, Myanmar, and Laos - citing an inability to agree terms with "the main negotiating counterpart." Delivery Hero's own announcement never names that counterpart. Contemporary reporting speculated it was Grab, at a value north of €1 billion, but that traces to unofficial press speculation, not a confirmation from either company - as far as this post's research can establish, neither side has ever officially confirmed Grab was the party at that table. Worth knowing as context, not something this post treats as established fact.

Beyond the Usual

The same regulator that blocked Uber is now reviewing Grab - because Uber owns a piece of Grab

Taiwan's FTC extended its review of the Grab-Delivery Hero Taiwan transaction by 60 working days in July 2026, pushing the decision deadline from late July to October 27, 2026. The stated reason isn't Grab's own market position in Taiwan - Grab has none, having never operated there before - but Uber's roughly 13% equity stake and 3.7% voting rights in Grab, a legacy of Uber selling its Southeast Asian ride-hailing and delivery business to Grab in 2018 in exchange for equity. Regulators are explicitly examining whether that stake could affect "Uber Eats' and Grab's incentive and ability to compete" once Grab controls foodpanda Taiwan - the same underlying competitive concern that killed Uber's own 2024 bid for the identical asset, now resurfacing one ownership layer removed. This is worth watching independent of how the deal itself performs: a regulator that already said no once to a Uber-Taiwan-delivery combination has a real basis to ask hard questions about a structure where Uber still has a minority claim on the winner.

Taiwan delivery workers are protesting the deal on competition grounds

foodpanda and Uber Eats delivery workers protested outside the FTC's offices during the review, according to regional reporting, on the theory that consolidating Taiwan's two-player food-delivery market under effectively one set of commercial incentives - even without a full merger - could let both platforms move together on rider commission rates rather than compete for riders on pay. It's not a financial finding, but it's a real, publicly visible signal that stakeholders outside the two companies see this less as "new competitor enters market" and more as "the market's already-thin competitive pressure gets thinner" - directly relevant to how the FTC's own review concludes.

Grab's guidance explicitly treats this as additive, not offsetting

Grab reiterated its full-year 2026 Adjusted EBITDA guidance of $700-720 million in the same announcement disclosing the acquisition, rather than revising guidance to reflect deal costs or integration spend. Combined with the disclosed $60 million of incremental Adjusted EBITDA expected only by 2028 - two years after the deal is expected to close - this reads as a company treating Taiwan as a genuinely separate growth initiative funded outside its existing operating plan, not a transaction requiring the market to lower near-term expectations elsewhere in the business.

The Prescription

Grab should treat Taiwan as a real strategic test, not a rounding-error acquisition to quietly bolt on and move past. At roughly a tenth the scale of Grab's existing Deliveries business, Taiwan is small enough that a stumble wouldn't threaten the core Southeast Asian franchise - but that's exactly what makes it a clean, low-stakes place to prove out (or disprove) whether Grab's operating playbook travels outside the specific regulatory, currency, and demographic environment it was built in. If Grab wants a second, third, and fourth market outside Southeast Asia later, how cleanly it migrates foodpanda Taiwan's brand, riders, and merchants onto its own platform - without giving Uber Eats a multi-quarter opening - is the evidence future expansion decisions should actually be built on, not the deal's press-release framing.

What Grab should stop doing, if the pattern from this announcement continues: leaning on "we reiterate guidance" language to imply a deal is costless, when the real signal investors need is a specific accounting of what integration actually costs and when Taiwan crosses from GMV-accretive to Adjusted-EBITDA-accretive on a fully-loaded basis - including whatever share of Grab's own corporate overhead the market eventually has to absorb, not just the pre-close, group-cost-excluded profitability Delivery Hero disclosed. And given a regulator has already shown it's willing to block this exact asset once over a competitor's ownership ties, Grab has a real incentive to get ahead of the Uber-stake question publicly rather than let the FTC's review be the only place it's being addressed - silence here reads as hoping the issue resolves itself, not managing it.

What to Watch Next

Whether the FTC's extended review resolves clean or with conditions attached (see Beyond the Usual). Whether Grab discloses Taiwan as a standalone reporting line once it closes, the way it already breaks out Deliveries, Mobility, and Financial Services, or folds it silently into existing segment totals - the choice will say something about how much Grab wants investors watching this specific bet. And whether Uber's own $14.8 billion acquisition of the rest of Delivery Hero closes cleanly, since a stumble there could change the competitive and ownership picture around Taiwan before Grab's own deal even closes.

The memorable version: Grab is paying $600 million for the one asset Taiwan's regulator already ruled Uber couldn't have alone - and the same regulator is now checking whether Uber still gets it anyway, one ownership layer removed, through the 13% of Grab it never sold off.

---