Grab Holdings Limited
Southeast Asia's everyday app: in eight countries people use it to hail a motorbike or car, order food and groceries, and pay or borrow money.
The scores
Growth3/4 →
About 20% revenue growth for three years with users broadening, but 2026 now includes bought Superbank and Stash revenue, so not purely its own.
Business quality3/4 →
Rising 55% food-delivery share shows a real network advantage, but it has never faced a funded rival; Uber's foodpanda is that test.
Financial position4/4 ↓
Net cash of $5.4B and positive free cash flow fund the deals; Self-funding, though deals may claim up to $4B of it.
Execution record3/4 →
Eighteen quarters of adjusted profit growth and raised 2026 guidance; but the lending breakeven promised in 2022 is still pending and deals are untested.
Per share capture3/4 →
Buybacks held shares flat despite $240M yearly share pay; not Intact, because Stash was partly paid in stock and deals absorb cash.
Potential2/4 ↓
The 2028 target needs Atome approved, credit performing, Uber restrained and Indonesia leaving food alone; several steps outside Grab's control.
Risk profile2/4 ↓
Indonesia's pending food-delivery rule and Uber's non-compete decision are single-step external events inside the horizon; founder controls the vote.
Valuation3/4 ↑
About $22.70 per $1 of profit after share pay prices the 2026 plan with modest growth after it; not Cheap, as growth is not free.
Evidence quality3/4 →
Figures are filed, but market share rests on one research house read through a news site, and Indonesia's pending rule was not read.
If it works 1.6× · if it fails you keep about 100%.
The Verdict
We rate Grab neutral. It is Southeast Asia's everyday app for rides, food delivery and payments, and it keeps winning a bigger share of a growing market, so the business is solid. The shares have roughly halved in a year while sales per share rose. The price is fair: about right if this year's plan arrives, with little credit for the 2028 target, much of it bought. As a bet it is safe, with little upside: rides, delivery and cash cover the price, so lending comes free. The view hinges on competition: Uber is buying a rival in three big markets, and Indonesia may cap food commissions.
The long caseIf people in the region borrow through apps they use daily, Grab can lend to its riders, drivers and merchants and earn on money as well as trips.
The evidence under this - 10 claims
- On guided 2026 adjusted EBITDA of $730M the same company value is about 10.1 times, which reads much cheaper.
- The vendor's EV/EBITDA is 25.0 on a company value of $11.6B that leaves most of the cash out.
Management said the core business is in line with the prior guidance and the raise came from consolidating Superbank and Stash, not from faster organic growth.
- Management says the on-demand business alone is growing over 20%, so some of the raise could be organic.
- The company does not split the $200M uplift between Atome and the rest.
- foodpanda has lost half its regional share since 2021, so Uber buys a declining business.
- Uber may prefer to renegotiate or sell foodpanda to Grab rather than fund a price war.
- Taiwan's regulator is reviewing whether Uber's stake blunts competition, which cuts both ways.
Adjusted free cash flow leaves out the lending and deposit working capital the bank businesses consume, so it flatters a group whose loan book is growing fast.
Part of the growth is bought: the loan book includes $761M from consolidating Superbank in June (C25), and excluding it the book doubled rather than trebled.
The company
One app, used by 53.9 million people a month in eight countries, books a motorbike or car ride, brings food and groceries from nearby shops, and lets people pay, save or borrow. It replaces street-hailed taxis, walking to the market, cash and informal lenders.
Delivery was 53.4% of 2025 sales, rides 36.2% and financial services 10.3%.
The evidence under this - 2 claims
How it makes money
Grab keeps a commission on each order and ride, after paying incentives to drivers, merchants and customers to keep them coming. Those incentives were 10.9% of order value in the June quarter, up 0.72 points in a year.
The profit most analysts quote, $608M over the last four quarters on Grab's adjusted measure, leaves out $240M of staff share pay. Counted in, it is about $325M.
The evidence under this - 2 claims
- The pack's vendor EV/EBITDA of 25.0 implies a larger EBITDA struck on a different EV; the vendor figure is advisory.
- Operating profit includes about $40M of items the company treats as one-off, such as legal provisions and a day-1 credit loss charge, which would lift this figure if excluded.
Where the money comes from, stream by stream
| Stream | Share of sales (FY2025) | How it earns it |
|---|---|---|
| Deliveries | 53.4% | Commissions and fees on food, grocery and parcel orders net of incentives, plus advertising, delivery fees and Malaysian supermarket sales. |
| Mobility | 36.2% | Commissions and fees on car, taxi and motorbike rides net of incentives, plus in-car advertising and renting vehicles to drivers. |
| Financial Services | 10.3% | Interest and fees on loans to users, drivers and merchants, payments, insurance and digital-bank income. |
| Others | 0.1% | Small residual services outside the three businesses. |
Who runs it
Anthony Tan co-founded Grab in 2012 and still runs it. Peter Oey has run the finances since 2020, and Alex Hungate, president since 2022, runs the businesses day to day. Insiders own about 3.5% of the shares, nearly all of it Tan's, but his special shares give him about three-quarters of the vote. He bought about $30M of stock in September. The annual report does not say what his bonus or performance shares pay on.
The evidence under this - 5 claims
- The share plan note says performance awards vest on 'market and non-market performance conditions' without naming them, so the metrics are not disclosed rather than absent.
- As a foreign private issuer Grab files no proxy statement, so no compensation discussion exists to read.
Why it wins, and who could beat it
Why customers come back
More drivers and restaurants in a city mean shorter waits and more choice, which draws more customers, which draws more drivers. Grab is the densest network in the region, so any rival has to pay to catch up.
The outside count agrees. Momentum Works, which tracks the region's delivery market, puts Grab's share up from 53.8% to 55%, while foodpanda's fell from about 22% in 2021 to about 11%.
The evidence under this - 3 claims
- Drivers and merchants use several apps at once, so density on one app can be copied with incentives.
- Incentives rose to 10.9% of GMV in Q2 2026 to hold driver supply, which is paying for density rather than owning it.
The threat
In July Uber agreed to buy Delivery Hero, and with it foodpanda in Singapore, Malaysia and the Philippines, where foodpanda holds 22% to 35%. Uber would then own a funded delivery network in three of Grab's biggest markets, held back only by a non-compete.
In Indonesia, 21.2% of sales, Gojek competes on rides and the government has capped Grab's cut of motorbike fares at 8%.
The evidence under this - 5 claims
- foodpanda has lost half its regional share since 2021, so Uber buys a declining business.
- Uber may prefer to renegotiate or sell foodpanda to Grab rather than fund a price war.
- Taiwan's regulator is reviewing whether Uber's stake blunts competition, which cuts both ways.
The rivals: UBER, SE, SE, RYDE, and 1 more that is private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | Where it competes | How far along | Priced against it |
|---|---|---|---|---|
| UBER | foodpanda under Uber ownershipPriced against it: Yes | Deliveries | On sale | Yes |
| SE | ShopeeFood (Sea Limited)Priced against it: Different product | Deliveries | On sale | Different product |
| Sea's Monee (ShopeePay and its lending) | Financial Services | On sale | ||
| Gojek (GoTo) in Indonesia; Xanh SM in VietnamPriced against it: Listed abroad | Mobility | On sale | Listed abroad | |
| RYDE | Ryde GroupPriced against it: Yes | - | On sale | Yes |
How big it could get
How big this could get
Southeast Asians spent $22.7B on food delivery in 2025, and Grab's 55% share makes about $12.5B of that its own. Its delivery orders, which also include groceries, came to about $14.2B.
The ceiling is set by governments. Indonesia cut what a platform may keep on motorbike rides to 8% in July, and a rule on food delivery is being drafted.
The evidence under this - 4 claims
The results
What growth costs
Growth has been making Grab more profitable: research, marketing and head-office costs grew far more slowly than sales, so each new dollar of sales costs less to win. What still grows with sales are the discounts and payouts to drivers, merchants and customers, and loan losses are growing faster as lending expands.
The evidence under this - 2 claims
- Administration fell partly on a $59M drop in litigation settlement costs in 2025, which will not recur every year.
- Grab reports under IFRS, so there is no XBRL numbers card to check against; the ratios are struck by hand on the 20-F's statement of profit or loss.
The last report (4 Aug)
Quarter ended June 2026, filed figures.
The June quarter was strong where it matters most. Delivery order value grew 24% at constant currency to $4.25B, and the profit kept on those orders rose to 2.3% from 1.8%. Adjusted profit grew for the eighteenth quarter in a row, to $168M.
Rides were weaker: trips grew 28% but revenue only 12%, because Grab paid drivers more and sold cheaper rides. And the reported profit leaned on a one-time $307M gain from taking control of Superbank, a digital bank.
What it did to the thesisIt supported the core of the view: delivery keeps growing and earning more on each order. It also showed the pressure points, with rides earning less per trip and reported profit lifted by a one-time gain. Competition and fee rules remain open, so the view stays Neutral.
The quarter in detail
Sales growth from the previous quarter, filed figures.
Earnings per share $0.06 against $0.05 expected.
The evidence under this - 4 claims
What management did with the money
Grab spent 2026 buying: control of Superbank, Stash, foodpanda Taiwan for $600M, and 60% of the lender Atome for $1.49B, with the rest of Atome to follow. Together they claim $2.6B to $4.0B of its $5.4B net cash. Buybacks of $400M in the first half only offset share pay.
Insiders in 2026Q2: 31 purchases and 31 sales, from their Form 4 filings.
The evidence under this - 3 claims
- Phase two is paid only if Atome performs, and part of it can be in shares.
- Adjusted free cash flow of $450M a year refills part of it.
The price, and where we disagree
In short. Fair. The price pays for this year's plan, not the bigger 2028 target, which is about right while Uber's move into its markets and Indonesia's fare rules remain unresolved.
The shares hit $2.74 on 18 September, down from $6.60 a year earlier. Three days later the chief executive bought 10.35 million shares, about $29.9M.
Two decisions are close. Indonesia's food-delivery rule was due in late August or early September and had not appeared by 26 September. And Taiwan's competition regulator rules on the foodpanda Taiwan purchase by 27 October.
The evidence under this - 4 claims
How we read the multiple
Grab holds $5.4B more cash than debt, so the business itself costs about $7.37B. Count the shares it pays staff as a cost and that is about $22.70 for each dollar of cash profit; leave that cost out and it is about 13.9.
That pays for this year's forecast of $720M to $740M of profit and little of the $1.7B targeted for 2028. Most of September's raise to that target came with buying Atome: bought, not grown.
Against its own history. Cheaper than its own normal on sales and on what it owns. The profit measures have no normal to compare against: until 2025 Grab never made a full-year profit.
The evidence under this - 5 claims
- On guided 2026 adjusted EBITDA of $730M the same company value is about 10.1 times, which reads much cheaper.
- The vendor's EV/EBITDA is 25.0 on a company value of $11.6B that leaves most of the cash out.
Management said the core business is in line with the prior guidance and the raise came from consolidating Superbank and Stash, not from faster organic growth.
The raise came the day a $1.49B purchase was announced; the target includes Atome, which is not expected to close until Q3 2027.
- Management says the on-demand business alone is growing over 20%, so some of the raise could be organic.
- The company does not split the $200M uplift between Atome and the rest.
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on price for each $1 of net worth, cheapest first:
| Ticker | Company | price for each $1 of net worth |
|---|---|---|
| RYDE | Ryde Group | 0.8× |
| GRAB | Grab Holdings Limited | 1.9× |
| UBER | Uber Technologies | 5.2× |
The middle of the other 2 is 3.0. Grab Holdings Limited is 2nd cheapest of 3.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
Between tiny, loss-making Ryde and Uber, so this places Grab rather than judging it; two rivals are too few to call the price cheap or dear.
Also checked and set aside: Domino's Pizza (a substitute, not the same product).
Private or foreign rivals that cannot be priced: GoTo Gojek Tokopedia (Gojek, GoPay), Delivery Hero SE (Foodpanda), ComfortDelGro Corporation, Line Man Wongnai, Bolt, Maxim, InDrive, Be Group, Xanh SM, Tada, Lalamove, AhaMove, Transportify, Dana, Touch 'n Go, MAE (Maybank), Google Pay.
The bet: what it pays if it works
The bet is on lending and the digital banks. They still lose money, years after breakeven was first promised, so rides, delivery and the cash are counted separately, at what the shares cost today, a price this thesis calls fair.
Counted that way, the rest of Grab covers today's price, so lending comes free. If lending grows as management plans, it adds a little over half again; if borrowers fail to repay, little is lost on it.
| Financial Services revenue in two to three years | Dollars of gross loans needed | At the stated multiple |
|---|---|---|
| $1.40B (the big win) | about 6.1B at 100% kept | 1.6× |
| $2.00B | about 8.7B at 100% kept | 1.8× |
To pay 2× (a Fair bet) it would need about $2.42B of Financial Services revenue (about 10.5B dollars of gross loans); to pay 5× (a Cheap bet), about $9.88B of Financial Services revenue (about 42.8B dollars of gross loans), at the same multiple.
Each row is Financial Services alone, plus the rest of the company (Deliveries and Mobility, with the net cash) counted at about $12.75B.
The steps it has to take
- Done Gross loan book above $2B
- In progress Financial Services at adjusted EBITDA breakeven due 2026-12
- In progress Atome's 60% bought and consolidated due 2027-03
- Not started Financial Services earning $500M of adjusted EBITDA due 2028-12
The detail behind the label
The evidence under this - 5 claims
Part of the growth is bought: the loan book includes $761M from consolidating Superbank in June (C25), and excluding it the book doubled rather than trebled.
Q2 2026 was still a $15M loss and credit charges more than doubled.
- The segment lost $15M in Q2 2026 and breakeven is still a second-half 2026 promise.
- Atome's $1B loan book and delinquency trend are unaudited management figures.
Adjusted free cash flow leaves out the lending and deposit working capital the bank businesses consume, so it flatters a group whose loan book is growing fast.
Where we disagree with the market
Analysts are uniformly positive, yet the shares have halved, as if the 2028 targets were out of reach.
The core business still grows about 20% and gains share, but the 2028 raise is mostly bought, so the price is fair.
The shares have fallen steadily while profit rose. Coverage, meanwhile, leans on adjusted profit that leaves out $240M a year of share pay, and on a 2028 target partly bought.
What would prove us too cautious: delivery growth above 20% after Uber owns foodpanda, incentives holding near 10.9% of order value, and Indonesia leaving food commissions alone.
What analysts say, and the evidence
The evidence under this - 7 claims
- The pack's vendor EV/EBITDA of 25.0 implies a larger EBITDA struck on a different EV; the vendor figure is advisory.
- Operating profit includes about $40M of items the company treats as one-off, such as legal provisions and a day-1 credit loss charge, which would lift this figure if excluded.
- Management says the on-demand business alone is growing over 20%, so some of the raise could be organic.
- The company does not split the $200M uplift between Atome and the rest.
What could go wrong, and how we would know
The risks
- Uber returns, funded. Once it owns foodpanda, only a non-compete stops Uber spending against Grab in three big markets.
- A fee cap in Indonesia. A cap on food commissions would hit the market that was 21.2% of 2025 sales.
- Drivers as employees. A draft Indonesian decree would make Grab fund drivers' accident, health and pension contributions.
- Lending losses. Its loans grew 197% in a year to $2.32B, and the charge for loans going bad more than doubled to $72M.
- A founder's vote. Anthony Tan owns about 3.2% of the shares but would control about 75% of the vote.
The evidence under this - 6 claims
- foodpanda has lost half its regional share since 2021, so Uber buys a declining business.
- Uber may prefer to renegotiate or sell foodpanda to Grab rather than fund a price war.
- Taiwan's regulator is reviewing whether Uber's stake blunts competition, which cuts both ways.
What we still do not know
- What Indonesia's food-delivery rule will say, and when.
- Whether financial services, which lost $15M in the June quarter, breaks even in late 2026, a goal first set in 2022.
- How sound Atome's $1B of loans are: the figures come from unaudited management accounts.
The evidence under this - 4 claims
Q2 2026 was still a $15M loss and credit charges more than doubled.
No audited figures or delinquency rates were published with the deal.
What would change our view
- Neutral→AttractiveThe price falls on an unchanged business
It needs all of:
- ○The company's value, net of cash, falls to 16 times its profit after share pay, from about 22.7.about ≈$2.21 today, 29% below the last close
- ○Delivery order value keeps growing 20% or more a year at constant currency.checked at the next results, due by 3 Nov
- Neutral→AttractiveThe two threats resolve in Grab's favour
It needs all of:
- ○Indonesia issues its delivery rule without capping food commissions below 15%.watched as it happens
- ○Two quarters after Uber owns foodpanda, incentives are still at or below 10.9% of order value.checked at the next results, due by 3 Nov
- Neutral→UnattractiveThe price war arrives
It needs any of:
- ○Incentives on rides and delivery rise above 12% of order value.checked at the next results, due by 3 Nov
- ○Indonesia caps food-delivery commissions at 10% or less.watched as it happens
We would also be wrong if
- Grab's share of the region's food-delivery spending falls below 53% in Momentum Works' 2026 count.
- Financial services is still losing money in the second half of 2027.
- Shares outstanding pass 4.2 billion because deals or staff pay are paid in stock.
What is coming
- 27 October 2026 Taiwan Fair Trade Commission decision on foodpanda Taiwan: Whether the $600M purchase clears while the regulator reads Uber and Grab as linked; a block or conditions would test how regulators treat Uber's stake.
Every dated event ahead
- 27 October 2026 Taiwan Fair Trade Commission decision on foodpanda Taiwan: Whether the $600M purchase clears while the regulator reads Uber and Grab as linked; a block or conditions would test how regulators treat Uber's stake.
- 3 November 2026 Q3 2026 results: Deliveries GMV growth against 20% at constant currency, on-demand incentives against 10.9% of GMV, and Financial Services against its second-half breakeven promise.
- February 2027 Q4 and full-year 2026 results: Revenue and adjusted EBITDA against the 2026 guidance, whether lending broke even in the second half, and the first 2027 outlook after Uber's foodpanda purchase.
- September 2027 Atome 60% acquisition closes: Whether $1.49B of cash buys the loan book the raised 2028 target leans on, and how much net cash is left before the last 40%.
Since we wrote this
- 11 OctThesis reviewLending declared an early line, with a bet on itThe owner asked every thesis to name the lines its story rides; Financial Services is a segment losing money and $500M of the 2028 target.
- 5 OctThesis reviewThe long case now sits under the verdictAfter the five-name pilot, the owner asked for the long case on every thesis.
- 4 OctThesis reviewOwner's checklist: revenue streams, cost of growth, management, catalystsThe owner's checklist found the method silent on who runs the company, what growth costs, where the money comes from and what is dated ahead; this pass fills those four blocks from the 20-F.
- 26 SepThesis writtenNeutral, medium convictionA good, growing delivery business at a fair price, with Uber about to own a rival and Indonesia's food-delivery rule still pending.
Also checked 1 time with nothing new to act on.
Sources 20
Sources - 20
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-03-06business description and segment mix; revenue by country; risk factors naming competitors, the Uber non-compete and the Indonesian commission decree; Class B voting; major shareholders; convertible note terms; share plansbacksC01C02C03C06C10C12C19C22C24C29C36C44C45C46C47C48C49
- S02primarycompany filing2026-08-04Q2 and H1 2026 income statement, balance sheet, cash flow, segment revenue and adjusted EBITDA, incentives, MTUs, loan portfolio, deposits, raised 2026 guidance, $750M buybackbacksC01C03C07C08C09C12C14C25C28C35C37C50C52C53
- S03primarycompany filing2026-02-11FY2025 figures, share-based pay, initial 2026 guidance and the first 2028 target of $1.5B adjusted EBITDAbacksC02C09C10C15
- S04primarycompany filing2026-09-15deal terms for 60% and the remaining 40%, Atome loan book and users, raised 2028 targetsbacksC15C17C26C27
- S05primarycompany filing2026-08-13Superbank acquisition accounting and $307M remeasurement gain, loan portfolio growth, subsequent events including the Stash closing paid partly in about 21 million sharesbacksC17C25C37
- S06primarycompany filing2026-02-11Stash enterprise value of $425M for 50.1%, paid in cash and stockbacksC17
- S07primarycompany filing2026-03-23$600M cash price for foodpanda Taiwan, about $1.8B of 2025 GMVbacksC17
- S08primarytranscript2026-08-03Q2 2026 earnings call transcript (FMP)data/cache/transcripts/GRAB-2026Q2.jsonmanagement answers on Indonesian two-wheel commission (6% of Mobility GMV), mobility take rate, Financial Services breakeven in H2 2026, Uber and foodpanda. Questions were submitted by email and read out by investor relations, so there was no live cross-examination
represents 6% of our total mobility GMV, so relatively small
backsC08C14C20C22C28C38C43 - S09primarycompany filing2026-03-24approval of 90 votes per Class B sharebacksC29
- S10secondaryindustry data2026-01-28Southeast Asian food-delivery GMV of $22.7B in 2025, up 18%, and Grab's share of it. Momentum Works' own site returned 403 to both curl and an agent fetch, so the report is read through TNGlobal's account of it
Southeast Asia's food delivery gross merchandise value (GMV) grew 18 percent year on year to $22.7 billion in 2025. Grab increased its regional market share from 53.8 percent to 55 percent in 2025.
backsC04C40 - S11secondaryanalyst2026-07-22Maybank Investment Banking on Uber's return via foodpanda, as reported by TNGlobal Maybank Investment Bankingfoodpanda's regional size and share trend, and its 22% to 35% share in Singapore, Malaysia and the Philippines
Foodpanda's regional gross merchandise value (GMV) has remained broadly flat at around $2.5 billion to $2.7 billion since 2022, while its market share nearly halved from about 22 percent in 2021 to around 11 percent in 2025. Foodpanda commands around 22 percent to 35 percent market share in Singapore, Malaysia and the Philippines.
backsC05 - S12secondaryanalyst2026-09-09Maybank Investment Banking competition scenarios, as reported by TNGlobal Maybank Investment BankingGrab's 69% share in Singapore; scenarios in which Deliveries GMV growth slows to about 11% a year and EBITDA falls 23%
in Malaysia Foodpanda had a 22 percent share, ahead of ShopeeFood's 11 percent
backsC05C41 - S13primarycompany filing2026-02-13Uber's own statement of the Southeast Asia non-compete, and its Grab stake carried at $2,529MbacksC19
- S14primarycompany filing2026-07-16which Delivery Hero markets Uber keeps: foodpanda in Singapore, Malaysia, the Philippines, Cambodia, Laos and MyanmarbacksC20C39
- S15primarycompany filing2026-04-17Sea's own description of its Monee payments and lending business; no mention of ShopeeFoodbacksC40C51
- S16primarycompany filing2026-04-24a Singapore rival naming GrabExpressread, but no claim rests on it
- S17secondaryregulator2026-05-07East Java regional parliament (DPRD Jawa Timur) note on Presidential Regulation 27 of 2026 DPRD Provinsi Jawa Timurthe regulation's number and the 92% minimum driver share, i.e. an 8% cap on the platform's cut. The Presidential Palace page and the BPK regulation register both returned 403
Pengemudi ojol berhak mendapatkan minimal 92 persen dari total pendapatan per order
backsC22 - S18secondarynews2026-08-18that a regulation covering food and goods delivery was being prepared, with the 8%/92% ride rule in force since 1 July 2026
Perpres Ojol yang ditargetkan terbit pada akhir Agustus atau awal September 2026 mendatang
backsC23 - S19secondaryvendor data2026-09-26GRAB research pack (FMP, FINRA, SEC Form 4), frozenprice and return series, the price-move split, analyst rating history, short interest, insider transactions with 10b5-1 flags, headcount, cohort multiplesbacksC30C31C32C33C36C52
- S20secondarynews2026-07-23the Taiwan review extended to 27 October 2026, citing Uber's stake in Grab
Taiwan's Fair Trade Commission extended its review of Grab Holdings' proposed acquisition
backsC39