Mobility, delivery, and financial services

Grab Holdings

GRAB

The lower USD 2.75 entry price raises price attractiveness to 5.5/10 under unchanged USD 4.32 / 1.28 cases. Company strength remains 6.7/10; risks from Atome and the planned USD 750m AI contribution are unchanged.

Analysis and ratings: ChatGPT·Business analysis 11/21/2026·Ratings & valuation 11/21/2026
THESIS STATUSValid · testing returns
CONFIDENCEMediumMedium for the business, low for scenario valuation and the capital bridge.
HORIZON3 years
WHAT MATTERS MOSTTransaction economics, credit quality, and returns on the Atome acquisition.

Business strength & price attractiveness

How AI assesses them

Business strength

Medium confidence
6.7/ 10

Previously 6.7 ·
Unchanged · Assessed

Business strength, independent of the stock price.

All five dimension scores are unchanged. Q3 and binding AI funding elements are already reflected; no new operating result or documented return improvement was added.

Five areas & supporting evidence

Price Attractiveness

Low confidence
5.5/ 10

Previously 4.9 ·
↑ +0.6 · Assessed

At 2.75 USD · close
Three-year horizon, through .

The lower USD 2.75 entry price raises price attractiveness to 5.5/10 under unchanged USD 4.32 / 1.28 cases. Company strength remains 6.7/10; risks from Atome and the planned USD 750m AI contribution are unchanged.

Scenarios & supporting evidence

What changed in the latest update

New information: The lower USD 2.75 entry price raises price attractiveness to 5.5/10 under unchanged USD 4.32 / 1.28 cases. Company strength remains 6.7/10; risks from Atome and the planned USD 750m AI contribution are unchanged.

Strength 6.7 → 6.7Attractiveness 4.9 → 5.5Scores changed

The numerical comparison does not by itself confirm or weaken the investment thesis.

Dated prices & model comparison
Input / modelPreviousCurrent
Reference price3.24 USD2.75 USD
Base modelled return+33.2 %+56.9 %
Adverse modelled return-60.5 %-53.5 %

Quote dates: 11/13/2026 → 11/20/2026. Model horizons end: 11/14/2029 → 11/21/2029. Cumulative returns include modelled dividends; they are scenarios, not guarantees.

Terminal-price and cumulative-dividend assumptions are unchanged; reference-price inputs differ as shown above.

View rating history →

Assessed · Methodology 1.2. Analytical judgments by ChatGPT; price changes do not automatically update these scores.

MARKET OVERVIEW

GRAB · Price chart

Delayed data · English controls.TradingView
Analysis contents · 13 sections
01

Company Snapshot

Grab Holdings Limited | NASDAQ GRAB

Sector: digital platforms, mobility, deliveries and financial services. Main region: Southeast Asia.

Thesis: valid; Q3 strengthened operating evidence, while credit and acquisition risks remain. Confidence: medium for the business, low for valuation. Horizon: three years through November 21, 2029. Review date: November 21, 2026; Q3 covers the period through September 30, with the release dated November 3 in Singapore. Independent ChatGPT assessment, methodology 1.2.

Next quarterly results: date not yet researched; less than two months have passed since the latest results. Active date verification starts no earlier than January 3, 2027.

Grab is demonstrating that transaction growth can generate faster adjusted EBITDA growth. Higher usage, advertising, groceries and financial services are the main opportunities. Atome expands the opportunity and consumer-credit exposure. What matters is shareholder return after credit losses, investment and dilution.

Q3 reached 57.2 million monthly transacting users (+20%). On-demand GMV grew 18% reported / 22% constant currency, revenue 27%, and adjusted EBITDA 73% to $237 million. Operating profit of $108 million and first positive Financial Services EBITDA strengthen the thesis. Gross loans of $2.711 billion and the failed Taiwan transaction require separate risk and capital analysis. Q3 results, pp. 1–9

Priorities: transaction unit economics; sustained Financial Services profitability and cohort quality; conversion into shareholder-available cash; acquisition returns; and net shares after buybacks.

Grab signed Malaysian AI infrastructure agreements with a proposed initial contribution of USD 750 million and returns yet to be demonstrated. Allowing for the potential capital burden lowers company strength to 6.7/10 and price attractiveness to 4.9/10 despite a lower entry price.

02

Investment Thesis

Thesis: Grab can remain a major regional distribution and transaction platform, earning more from the same user, driver or merchant at gradually lower serving costs. Value comes from connecting mobility and delivery with advertising, payments and credit, not just growing rides and deliveries.

1. Growth still reflects service usage. Q3 Mobility transactions grew 21%, faster than GMV (+14% reported). On-demand MTUs grew 16%, GMV per user 2% (5% constant currency). Incentives fell sequentially to 10.7% of on-demand GMV but were 64 basis points higher year over year. Growth is supported, while margins also improve. Q3 results

2. Operating leverage is visible. Adjusted EBITDA grows faster than revenue. Continued volume growth and slower central-cost growth could increase profit without substantial customer price increases.

3. Financial services use existing distribution. Grab knows some customers' and partners' transaction behavior, potentially reducing acquisition costs and improving credit decisions. Returns after losses and capital costs must confirm the advantage; loan-book size alone is insufficient. Q1 call

4. Atome adds acceleration and risk. It expands consumer lending and the merchant network. Operating synergies are plausible, but higher consolidated revenue alone cannot justify the purchase price. M&A call

What may be underappreciated: resilience under consumer pressure and the long-term benefits of frequency, advertising and financial services. This is an analytical hypothesis, not an established explanation for today's price.

What must be true: customers return without ever-higher subsidies; lending earns through the cycle; capital allocation increases per-share value.

The contribution represents about 15% of consolidated September net cash liquidity of USD 5.044 billion. That ratio does not measure cash available to the parent. The projects compete for capital with Atome and buybacks; contracted demand, potential guarantees and demonstrated returns will matter.

03

The Business

Grab connects customers, drivers, couriers and merchants. Local demand and physical fulfillment form the core; additional monetization is built around them.

BusinessRevenue sourceInvestment relevance
MobilityRide matching and related servicesEstablished operating profit source; sensitive to commissions, incentives and fuel
DeliveriesFood and grocery delivery, retail and advertisingHigher frequency and margin opportunity; some retail is more capital-intensive
Financial ServicesLending, digital banks and other financial productsFurther distribution monetization; credit, regulatory and capital risk

The ecosystem includes GrabFood, GrabMart, GrabPay and banks GXS, GXBank and Superbank. Grab also operates grocery chains Jaya Grocer and Everrise. The entire business is therefore not equally asset-light.

Stash contributes from Q3; the financial segment also includes Superbank. Foodpanda Taiwan is excluded: the October regulatory rejection leads to a $60m Q4 fee. Atome remains conditional, with first-stage completion expected in Q3 2027, not an already consolidated business. Q3 release; M&A call.

04

Competitive Advantages

Network density. More drivers, couriers and merchants can improve availability and capacity utilization. The advantage is local: strength in one city does not automatically imply similar economics elsewhere.

Distribution across services. One app offers rides, food, groceries and financial products. Q1 management reported 1.8× higher frequency among users combining groceries and food than food-only users. This compares user groups; it does not prove cross-selling caused the whole difference. Q1 call

Local data and execution. Maps, payment histories, merchant relationships and country knowledge build over time, helping logistics and credit decisions.

AI as an economic tool. Q3 management reports a 42% sequential reduction in cost per AI interaction alongside 60% interaction growth. These are company operating metrics, not separately demonstrated incremental AI profit. The 21.3% EBITDA margin supports operating leverage, but AI cannot be isolated from volume, advertising, acquisitions and other savings. Prepared remarks, p. 1

Moat strength: meaningful, but not absolute. Customers and partners can use several apps, and regulation can constrain monetization.

05

Growth Drivers

  • Affordable products and more cities: expanding users and transaction frequency.
  • GrabMart: Q3 management says grocery GMV grew 1.8 times faster than food; 16% of food users used GrabMart versus approximately 14% in Q2. This is user overlap, not GMV share. Remarks, p. 2
  • Advertising: quarterly active self-serve advertisers grew 39%, average spending 13%. Advertising and operating leverage contributed to the 2.7% Deliveries EBITDA/GMV margin. Q3, p. 5
  • Financial Services: greater lending penetration and broader products, conditional on underwriting and funding quality.
  • Acquisitions: Superbank and Stash contribute; Atome remains conditional. Foodpanda Taiwan was rejected and is excluded from valuation. Reported financial-segment growth must be separated from organic growth.
  • Autonomous mobility: a long-term option rather than the three-year thesis foundation; cheap two-wheel rides and differing regulation limit region-wide robotaxi adoption.

2028 management targets: Q3 remarks reaffirmed $1.7 billion group adjusted EBITDA and 30%+ revenue CAGR from 2025 to 2028, conditional on Atome. The presentation retains $500 million Financial Services EBITDA and over $6 billion gross loans. The $500 million belongs to the entire financial segment, not Atome alone. The reaffirmed total cannot be treated as a detailed ex-Taiwan bridge; the September breakdown is outdated. Remarks; presentation, pp. 12–13

Sarawak: the October 8 Grab/SoftBank/PETROS framework explores an AI ecosystem. Grab contributes its platform, applied AI and network, but no capital commitment, ownership or expected return was quantified. No value is added to target EBITDA or valuation; an infrastructure-funding commitment would require a fresh capital-allocation assessment. SoftBank

The agreements with SoftBank and PETROS announced on November 9 include development and funding obligations. Of USD 1.85 billion in proposed initial capital, Grab would contribute USD 750 million: USD 100 million in Sarawak and USD 650 million in Peninsular Malaysia. Funding is conditional and staged; the money has not already been spent. The plan for at least 5 GW has no substantiated timeline for returns. SEC 6-K · SoftBank and Grab.

06

Economics & Financial Profile

MetricQ3 2026Q3 2025Change / meaning
Revenue$1,112m$873m+27%; +30% constant currency
Adjusted EBITDA$237m$136m+73%; 21.3% of revenue
Operating profit$108m$27m+297%
Period profit$18m$17m+5%; revaluations and non-operating items
Profit attributable to Grab owners$27m$37mApproximately −27%; distinct from group profit
Operating cash flow$55m−$127mAffected by lending and deposits
Adjusted FCF$286m$203m+41%; a separate adjusted measure
SBC$49m$55m−11%
Trailing adjusted FCF$533m$283mAlmost doubled
FY2026 revenue guidance$4.15–4.20bnPreviously $4.10–4.15bnManagement raised guidance
FY2026 EBITDA guidance$790–800mPreviously $720–740mMidpoint from $730m to $795m

Q3 release, pp. 1–4, 12–16

Earnings quality: operating profit improved substantially from Q2's $19m. Net profit includes $57m net fair-value losses and a $12m investment impairment. Owner-attributable profit fell from $37m to $27m. Nine-month profit of $373m still includes Q2's one-off $307m Superbank revaluation. Adjusted EBITDA excludes $49m Q3 SBC. The $60m Taiwan fee will affect Q4 IFRS profit but be excluded from adjusted EBITDA. Q3, pp. 3, 9, 12–16

Cash flow improved, but the adjusted measure is not shareholder cash. Q3 OCF was $55m; nine-month OCF $52m versus $10m. Adjusted FCF was $286m in Q3 and $457m year to date. Bridge: 55 − 71 capex + 4 disposals + 445 lending movements − 184 deposits + 37 treasury liquidity = $286m. Deposits and lending materially affect the comparison; the improvement cannot all be attributed to core earnings. Company capex includes leased assets, so this is neither simple cash capex nor owner FCF. Reconciliation and cash-flow statement

For 2025, adjusted EBITDA was $500m and adjusted FCF $290m. Management targets 80% conversion in 2028. Applied to $1.7bn EBITDA, that implies $1.36bn adjusted FCF arithmetically, not a verified owner return. Mobility and Deliveries margins use GMV; group margin uses revenue, so they are not directly comparable. Earlier long-term targets were 9%+ Mobility and 4%+ Deliveries.

Earnings Q&A supplement: the unchanged outlook implies Q4 EBITDA of approximately USD 230–240 million. Management described continued reinvestment, while Financial Services still lacks a detailed profitability breakdown. This adds detail to the reported quarter rather than representing a new realized period.

07

Key KPIs

KPILatest Q3 2026 signalRequired confirmation
Group MTUs / on-demand GMV57.2m (+20%) / $6.835bn (+18%; +22% CC)Habit and affordable organic growth
MobilityTransactions +21%, GMV +14%, revenue +10%Slower growth than Q2 does not impair long-term economics
Mobility EBITDA / GMV$207m / 8.9%; stable YoY, above Q2's 8.6%Resilience to fuel, incentives and regulation
DeliveriesGMV +21%, revenue +21%, EBITDA $123m; margin 2.7%Monetization while sustaining growth
Financial ServicesRevenue $201m (+124%); EBITDA +$12m; loans $2.711bnRepeat profitability after credit risk; separate Superbank/Stash
Credit quality90+ day NPL 2.6%; ECL reserve 6.5%; company coverage ~2.4×Comparable cohorts, write-offs and realized returns
Lending Risk Adjusted Margin7%; steady-state target 10–12%Neither ROE nor fully costed segment profit
Incentives$737m total; on-demand 10.7% GMV, +64bps YoY / −17bps QoQMargin growth without ever-rising support
Per-share capital4.037bn issued shares September 30; $400m buybacks September–OctoberNet dilution including SBC, convertibles and acquisition payments

Q3 results; presentation, pp. 14–15, 34; remarks.

Financial Services includes Superbank and newly Stash, so +124% is not purely organic. Gross loans grew 200%, or 115% excluding Superbank. Lending Risk Adjusted Margin is annualized lending revenue less funding and credit costs divided by ending gross loans: 7% is not ROE and excludes some operating costs, taxes and minorities. Atome's cumulative 25m users cannot be added to Grab's monthly MTUs.

08

Management & Capital Allocation

Anthony Tan is co-founder and CEO, Alex Hungate leads operations and Peter Oey finance. John Rogers left the board and Audit Committee on October 31 after five years and becomes an adviser. Sean Sullivan joins both on December 1. This is a governance change, not proof of better returns. Announcement

Atome

TermMeaning
First stage60% for $1.49bn cash, including $260m growth capital
CompletionExpected by Q3 2027, subject to regulatory approvals
Second stageRemaining 40% approximately two years later
Second-stage pricePerformance-based; whole-company valuation capped at $2–4.5bn; at least half cash

Transaction

The 40% stage implies $0.8–1.8bn, or $2.29–3.29bn combined nominal consideration. $1.49bn / 60% implies $2.48bn but includes growth capital, so is not the price of the existing business alone. Atome only recently became profitable; stand-alone returns after credit losses remain insufficiently quantified. Distribution logic does not prove the purchase price earns attractive returns.

Buybacks: $400m in September–October, $500m remaining; total authorization since 2024 $1.75bn. Issued shares fell from 4.079bn in June to 4.037bn in September; October repurchases are not all in that count. Over 10% cancellation refers to full-program completion at assumed current prices, not an accomplished result. Q3

Operating execution is more convincing than demonstrated returns on the new acquisition wave. Stash must also justify expansion outside the core region.

The contribution represents about 15% of consolidated September net cash liquidity of USD 5.044 billion. That ratio does not measure cash available to the parent. The projects compete for capital with Atome and buybacks; contracted demand, potential guarantees and demonstrated returns will matter.

09

Risks / Bear Case

Credit risk can overwhelm operating progress. Q3 90+ day NPL is 2.6% and ECL reserves 6.5%; management cites a 2–3% NPL range since early 2025. Rapid origination changes loan age, so these metrics do not establish full-cycle resilience. Net impairment losses increased from $37m to $58m. Q3; remarks

Regulation and partner economics: Mobility's 8.9% margin shows resilience after Indonesia's July two-wheel commission framework, although the full current legal framework and possible extensions are not independently verified. Taiwan's regulator rejected the foodpanda transaction in October; the $60m Q4 fee is included, while the business and $600m purchase price are removed. Q3, pp. 6, 9

Malaysia: an October 9 budget speech announced a joint government/Grab MYR160m partner-protection package from 2027, without specifying Grab's share. Minimum-income and social-protection rules are to be completed in early 2027. Retention could improve and costs rise; neither the whole sum nor a margin impact can be assigned to Grab. Budget, item 34

Incentives: on-demand incentives were 10.7% GMV, +64bps YoY but −17bps QoQ. Improving Mobility/Deliveries margins weaken the claim that incentives consume all growth, but incentives still outgrew reported on-demand GMV.

Acquisitions and minorities: consolidated EBITDA can include a whole business despite lower economic ownership. Earn-outs do not eliminate downside from the fixed first payment and price floor.

Cash and capital: bank liquidity, deposits and regulatory capital are not automatically parent-available cash; lending growth can require further capital.

Competition, FX and geopolitics: local rivals pressure prices; weaker currencies reduce USD results and fuel costs burden partners and consumers. Uber's July 16 Delivery Hero offer could strengthen regional foodpanda competition. As of October 10 it was not completed: the August 27 document describes an offer through November 5, with conditional H2 2027 settlement. Uber

Adverse case: the core grows but margins stall, credit losses rise and acquisition returns disappoint; more adjusted EBITDA need not create more per-share value.

After the first offer period Uber was attributed 66.56% of Delivery Hero shares and voting rights, according to the November 10 notice. Settlement still depends on approvals; the impact through foodpanda remains a risk without a change in modeled margins. Eurex.

10

Valuation

ChatGPT produces the calculations and valuation scenarios. Share-price inputs are dated to the analysis; they are not continuously updated market data.

CHATGPT SCENARIOS · November 21, 2026 → November 21, 2029

Different outcomes. Different values.

Historical analysis input
2.75 USD · November 20, 2026 · 16:00 EST · close
Adverse
1.3 USD
Base
4.3 USD
Conditional scenarios from the analysis. The dashed line marks the historical share price used in the calculation. Values are neither current intrinsic value nor a share-price forecast.

Price attractiveness: 5.5/10 at the Nasdaq GRAB Class A regular close of USD 2.75 on November 20, 2026, 16:00 EST. Assessed November 21, 2026 under methodology 1.2, through November 21, 2029. Confidence is low. Base USD 4.32 (+56.9%), adverse USD 1.28 (−53.5%), dividends zero. The decline from the latest completed assessment at USD 3.24 is 15.1%; operating and capital assumptions are unchanged. History.

Current three-year model through November 21, 2029. Operating assumptions are unchanged. Both cases assume the full USD 750 million is funded within the horizon without crediting the equity stake; this is a conservative assumption rather than an expected total loss. Capital adjustments become −0.476 / −1.526 billion USD and equity values become USD 18.124 / 5.624 billion. With 4.20 / 4.40 billion shares, the model gives USD 4.32 / 1.28. Base-case sensitivity at 10× / 12× / 14× core EBITDA is USD 3.70 / 4.32 / 4.93. With no drawdown or full offset from stake value, the same inputs would give USD 4.49 / 1.45. SEC 6-K.

Historical model from November 3–5, 2026, before accounting for the AI infrastructure agreements:

Market context: Nasdaq GRAB Class A, USD. The reference is $2.86, November 2, 2026 regular close at 4:00 p.m. EST, verified November 3. It precedes the results; the later $3.048 trade at 7:59 p.m. EST is excluded. Approximate capitalization of $11.5bn = $2.86 × 4.037bn September issued shares is ChatGPT's calculation, not current fully diluted market capitalization. Price; Q3 shares.

Q3 issued shares were 3.942bn A + 0.095bn B = 4.037bn, versus 4.087bn a year earlier and 4.079bn in June. Quarterly diluted weighted average instead rose to 4.392bn from 4.186bn, mainly through convertible inclusion in diluted EPS. These measures differ; another 551m potential instruments must not be mechanically added. This model retains convertibles as debt and does not also count conversion shares. $5.044bn net cash liquidity is not automatically parent-available cash. Q3, pp. 16–17; presentation, p. 20

Historical multiple, October 6: then-EV of $8bn / then-guidance midpoint of $730m EBITDA was approximately 11×. It is not current. Q3 raised revenue guidance to $4.15–4.20bn and EBITDA to $790–800m. Old EV cannot simply be carried across the changed balance sheet and acquisition obligations. Dividing old EV by the $1.7bn 2028 target gives 4.7× but omits full acquisition costs, minorities, investment and funding; it is no margin of safety.

Working 2028 operating scenarios

ScenarioGroup adjusted EBITDAAssumptions
Weaker$0.9–1.1bnSlower growth, margin pressure or worse lending
Middle$1.3–1.5bnContinued growth and partial synergies, short of the full plan
Strong execution$1.7bnTimely acquisitions and management target delivery

The first two are analyst assumptions; the third uses management's target, not a certain outcome. These are operating sensitivities, not the complete three-year stock valuation through November 2029.

Valuation framework: ChatGPT's independent sum of parts. Core excluding Taiwan, after regional corporate and lease costs, is valued on normalized EBITDA; Financial Services including Stash and assumed Atome on cash profit attributable to Grab after funding, taxes, credit losses and minorities. A capital bridge starts at September 30, 2026, followed by scenario shares. Deposits and loans are not added as free cash. Confidence is low; this is neither detailed DCF nor a bank-ROE model.

Price attractiveness: 5.6/10 at $2.86. Quote November 2, 2026, 4:00 p.m. EST regular close; assessed November 3. Horizon November 3, 2029; methodology 1.2. Approximately fair with significant loss risk. Operating evidence improved, but the higher entry price and conservative ex-Taiwan revision lower attractiveness from the source's prior 5.7/10 at $2.71. Historical analyst consensus of $5.76, last updated September 25, was not reverified and is not this analysis's fair value.

ItemBase caseAdverse case
Business2029 revenue $6.2bn; core EBITDA after leases $1.30bn, Financial Services $0.45bn, total $1.75bn. Mobility margin 8.8–9.0% GMV, Deliveries 3.0–3.5%. Stable cohorts and Atome completion; no Taiwan.Revenue $5.0bn; core $0.75bn, Financial Services $0.15bn, total $0.90bn. Mobility 7.5–8.0%, Deliveries 1.8–2.2%. Weaker demand, more incentives, losses and capital; weaker Atome synergies; no Taiwan.
Core valuation1.30 × 12 = $15.60bn EV0.75 × 9 = $6.75bn EV
Financial Services valuationGrab-attributable profit after funding, risk, taxes and minorities 0.25 × P/E12 = $3.00bn equityProfit 0.05 × P/E8 = $0.40bn equity
Capital adjustment+$0.274bn, below−$0.776bn, larger reserves/weaker cash generation
Equity value15.60 + 3.00 + 0.274 = $18.874bn6.75 + 0.40 − 0.776 = $6.374bn
Scenario shares4.20bn4.40bn
Three-year priceApproximately $4.49Approximately $1.45
Three-year dividends$0/share, analyst assumption$0/share, analyst assumption
Total three-year return+57.1%, ~16.3% annualized−49.3%, ~−20.3% annualized

Q3 supports profitability and reaffirms management's $1.7bn 2028 target with Atome; the scenarios are not guidance. Removing Taiwan reduces previous revenue assumptions $6.5/$5.2bn to $6.2/$5.0bn and core EBITDA $1.35/$0.80bn to $1.30/$0.75bn. These are conservative analyst allocations of previously unseparated Taiwan contributions, not segment guidance. Strong Q3 does not yet raise long-term multiples or Financial Services profit. Core includes advertising/corporate costs; Financial Services includes Stash and assumed Atome. Core after leases differs from company IFRS16 adjusted EBITDA.

12× core EBITDA is a working midpoint for a growing platform with regional/regulatory risks; 9× reflects slowdown and margin pressure. P/E12 and 8 reflect credit/capital risk, not a verified peer median. Normalized financial cash profit is after credit losses, funding, tax and minorities, but before SBC represented through dilution: $450m EBITDA less $200m gives $250m; stress $150m less $100m gives $50m. Superbank's one-off consolidation gain is not capitalized.

Capital bridge to 2029

USD billions, from September 30, 2026 through November 3, 2029. June cash and Q3 cash flow are not added again. Reported figures are separate from analyst reserves. Missing debt allocation, Stash consideration and September/October buyback split remain estimates.

ItemBaseAdverseInput type
September 30 net cash liquidity+5.044+5.044Reported: gross cash 7.114 less loans/borrowings 2.070
Bank debt allocation+0.370+0.370Analyst split: 1.700 non-bank / 0.370 bank. Bank debt is in Financial Services equity valuation; segment split undisclosed.
Exclude bank/restricted liquidity−2.890−2.890Analyst reserve: reported deposits 2.703 + cushion 0.187; not a verified parent-cash breakdown
Stash payments after September 30−0.300−0.425Reserve for future consideration at unknown fair value; paid amounts already in starting balance; not total historical cost
Taiwan termination fee−0.060−0.060Company Q4 fee, assumed cash settlement; $600m purchase and business removed
Atome first stage−1.490−1.490Announced price including $260m growth capital, not deducted twice
Remaining Atome 40%−1.200−0.800Analyst assumption of all-cash payment within horizon; contractual range $0.8–1.8bn
Buybacks after September 30−0.750−0.750Assumed $0.250bn October plus announced $0.500bn remaining; company $0.400bn covers September–October, split unknown; no repeat deduction of September purchases
New core cash after all cash costs from Q4 2026+2.050+1.025Q4 2026/2027/2028/part-2029: 0.10/0.50/0.70/0.75; stress 0.025/0.30/0.35/0.35. After interest, tax, cash capex, leases and working capital; no Taiwan.
Additional bank capital/integration uncertainty−0.500−0.800Reserves beyond Atome first-stage growth capital; future Stash consideration separate
Total+0.274−0.776Analyst bridge, not reported net debt

Shares and double-counting protection: start 4.037bn issued shares. Base buybacks $0.750bn / $3.10 retire 0.242bn, plus ~0.405bn employee dilution gives ~4.20bn shares. Stress $0.750bn / $2.80 retires 0.268bn, plus 0.631bn dilution gives 4.40bn. These are scenarios, not current diluted averages. Convertibles remain debt; SBC is represented by dilution, not another cash deduction. Future Stash/Atome consideration is cash-equivalent, not also new shares. Leases are already in core profit/cash flow; the full liability is not deducted again. Non-bank debt is already in net cash; bank debt/funding is in bank equity value.

Calculation and sensitivity: price = (core EV + Financial Services equity + capital adjustment) / shares. Return = (terminal price + dividends) / 2.86 − 1, without dividend reinvestment, before tax, fees and FX. Returns use unrounded 18.874/4.20 and 6.374/4.40. Core multiples 10×/12×/14× yield $3.87/$4.49/$5.11. Another $1bn capital burden subtracts ~$0.24/share; zero Financial Services value lowers base to $3.78. A $100m payment/buyback change shifts value ~$0.024/share. Decimals do not imply forecast precision.

Conditional price implication: with the same Financial Services value, +$0.274bn capital adjustment, 4.20bn shares and 12× multiple, zero three-year return from $2.86 corresponds to ~$0.73bn core EBITDA after leases versus base $1.30bn. This is scenario algebra, not an observed market expectation.

Rationale: operating leverage and the financial-segment milestone support the business, but +57.1% base versus −49.3% stress supports only 5.6/10. Taiwan is removed from both operations and purchase cost; capital reserves remain material. Main weakness: Atome/banks must turn lending into Grab-attributable profit after losses, capital, minorities and dilution. Confidence: low. Profit normalization, parent-available cash, debt allocation, future multiples/shares are explicit, sensitive analyst assumptions rather than a finished detailed model.

Conditions for change: a lower price with the thesis intact, repeated Financial Services profitability and progress toward $1.30bn core EBITDA, $250m Grab-attributable financial profit and no more than 4.20bn shares support a higher rating. Delinquencies, acquisition payments or capital needs above reserves require lower value/rating. Actual parent cash, Stash consideration and buyback timing supersede assumptions.

Q3 results; presentation; prepared remarks; price; Atome terms; Stash. The earlier Taiwan agreement is historical and superseded by the rejection.

Price reassessment as of 5 November 2026: the latest regular close was USD 3.36 on 4 November. Unchanged terminal values of USD 4.49 / USD 1.45 now imply three-year returns of +33.7% / −56.9%. The higher entry price further reduced the cushion, so attractiveness falls from 5.2 to 5.0/10; the business scenarios and company strength are unchanged.

Current return and interpretation: return = (terminal price + dividends) / 2.75 − 1. Unrounded equity/share values of 18.124 / 4.20 and 5.624 / 4.40 imply +56.9% (16.2% annually) and −53.5% (−22.5% annually). The lower price improves upside and reduces modeled loss, but credit risk and unverified AI-project returns prevent a strong buy assessment. At the −USD 0.476bn capital adjustment, unchanged Financial Services value, 4.20bn shares and 12× core multiple, zero return implies approximately USD 0.75bn core EBITDA versus USD 1.30bn base. This is scenario algebra, not an observed market expectation. The older model and its returns above remain explicitly historical.

11

What Would Change the Assessment?

These are analyst thesis-monitoring thresholds, not company guidance.

SignalImplication
Two quarters of slower on-demand growth and rising incentivesWeaker demand argument; Q3 growth slowed but incentives fell QoQ and margins rose
Mobility margin below 8% GMV for two quarters without recoveryReassess structural economics; Q3's 8.9% does not meet this condition
Repeated Financial Services losses after its first positive Q3Check organic profitability, Stash/Superbank mix and credit costs
Rising NPL/write-offs at comparable cohort mix and ageLower confidence in distribution/data advantage; baseline 90+ day NPL 2.6%
Atome completion beyond Q3 2027 or weaker synergiesRecalculate 2028 plan and valuation
Strong EBITDA without better economic cash flow and EPSQuestion shareholder benefit
Further regulatory rejection or capital costs above reservesRecalculate bridge; Taiwan already removed and fee included

Several quarters of growth, better margins, stable credit cohorts and acquisition returns after all costs would improve confidence. Plan delivery and earnings quality must be assessed together.

Watch the AI project contracts, guarantees, actual drawdowns and contracted demand; without it the model adds no project earnings.

12

Current View

Company strength 6.7/10, assessed November 21, 2026 under methodology 1.2. All five dimensions remain unchanged; Q3 and AI agreements are already reflected. No new operating result or documented improvement in financing or project returns was added. The AGM notice and management-sale filings alone do not change the balance of evidence.

Price attractiveness: 5.5/10 at the Nasdaq GRAB Class A regular close of USD 2.75 on November 20, 2026, 16:00 EST. Assessed November 21, 2026 under methodology 1.2, through November 21, 2029. Confidence is low. Base USD 4.32 (+56.9%), adverse USD 1.28 (−53.5%), dividends zero. The decline from the latest completed assessment at USD 3.24 is 15.1%; operating and capital assumptions are unchanged. History.

Historical assessment from November 3–5, 2026:

Business view: cautiously positive and stronger after Q3. $108m operating profit, 2.7% Deliveries margin, 8.9% Mobility margin and first positive Financial Services EBITDA support operating leverage. Credit reserves, low actual consolidated cash flow and capital allocation remain limitations. Taiwan shows planned expansion need not happen.

Stock view: 5.6/10 price attractiveness at the $2.86 November 2, 2026 close, assessed November 3 under methodology 1.2. The ex-Taiwan model gives $4.49 (+57.1%) base and $1.45 (−49.3%) stress, no dividends. Better business strength coexists with uncertain Atome returns, cash availability and dilution. Valuation confidence is low; the close precedes the results.

The thesis remains valid and subject to review. A price decline alone does not change business strength. Next tests: sustained Q4 Financial Services profitability, cohorts, EPS and Taiwan's $60m IFRS fee. Q3 is already published; the next earnings date is not yet actively researched under the two-month rule. Prepared Remarks is not a full Q&A transcript. Strong business performance is not automatically a buy recommendation.

Company strength

6.8/10, assessed November 3, 2026, methodology 1.2. Increased from 6.6 on Q3 execution/economics evidence, not price. Each dimension has 20% weight.

DimensionScoreEvidence
Competitive advantage7.0/1057.2m MTUs and cross-service distribution; switching costs remain low and competition persists; scale alone does not strengthen moat
Demand durability8.0/10On-demand MTUs +16%, GMV +18% reported/+22% CC; slower Mobility and higher YoY incentives limit an increase
Management execution6.3/10Up from 6.0: first +$12m Financial Services EBITDA meets H2 milestone; guidance raised and buybacks documented; Taiwan and Atome returns limit upside
Business economics5.5/10Up from 5.0: operating profit $108m versus $27m, margins 2.7%/8.9%, SBC $49m versus $55m; OCF $55m versus adjusted FCF $286m and owner profit $27m versus $37m still lag
Financial resilience7.0/10Net cash liquidity $5.044bn below June; $2.703bn deposits are not free capital. Avoided Taiwan purchase saves capital, but $60m fee, Atome and dilution remain

(7.0 + 8.0 + 6.3 + 5.5 + 7.0)/5 = 6.76 → 6.8. Q3 margin improvement, the profitability milestone and lower SBC support the increase. Owner returns lag adjusted EBITDA, while lending and acquisitions add capital risk. Main weakness: cash-flow quality and unproven returns on financial/acquisition exposure. Confidence: medium: several periods are available but the business scope is changing and full-cycle credit returns remain unproven.

Further quarters of better margins and economic per-share cash flow with stable cohorts support an increase. Two quarters of slowing growth with more incentives, worsening cohort losses or major capital commitments without returns support a reduction.

Q3 release; presentation; remarks; Q2, Q1, FY2025 and September 15 Atome calls. Reported metrics and management targets are distinguished from ChatGPT's independent ratings.

Assessment as of 5 November 2026: Company strength remains 6.8/10. Price attractiveness falls to 5.0/10 at the 4 November closing price of USD 3.36. The price is 17.48% above the 2 November close and met the two-session price condition; the price move itself does not change company strength.

13

Thesis History / Updates

Date or periodChangeThesis implication
Q4 2025, February 11, 2026 callAnnual adjusted EBITDA $500m; original 2028 target $1.5bnFrom survival/breakeven to profitable growth
Q1 2026, May 4 call52m MTUs, on-demand GMV +24%; guidance retainedDemand withstands early fuel shock
Q2 2026, August 3 US-time callMore users/profitability; higher guidance and Superbank consolidationStronger operations, more complex earnings quality
September 15, 2026Atome agreement; 2028 EBITDA target $1.7bnMore financial opportunity, capital and risk-management demands
October 7, 2026Consolidated analysis updateThesis valid, valuation/acquisition returns conditional
October 8, 2026Grab/SoftBank/PETROS AI framework in SarawakNo quantified commitment or return; no valuation uplift
October 10, 2026H1 cash flow, incentives, shares and Uber offer; methodology 1.0First strength 6.6; price unassessed at October 9 $3.19 close
October 10, supplementalMalaysian package and price/source checkStrength 6.6 (7/8/6/5/7), price unassessed; Grab's share unquantified
October 10, price modelThree-year scenarios, capital reserves and sharesAttractiveness 5/10 at $3.19; $4.60/+44.2%, $1.53/−52.1%; low confidence; strength 6.6
October 17, 2026October 16 close $2.94, Q3 date, methodology 1.2; scenarios retainedStrength 6.6; attractiveness 5.4; +56.4%/−48.1%; old 1.0 remains historical
October 24, 2026October 23 close $2.79; no new earnings evidenceStrength 6.6; attractiveness 5.6; +64.9%/−45.3%; low confidence
October 31, 2026October 30 close $2.71; board/Audit Committee changeStrength 6.6; attractiveness 5.7; +69.7%/−43.7%; Q3 then due November 2
November 3, Q3Revenue $1,112m, EBITDA $237m, operating profit $108m; first positive FS EBITDA, higher guidance, rejected Taiwan/$60m fee; September-balance-sheet model excluding TaiwanStrength 6.8 (7/8/6.3/5.5/7); attractiveness 5.6 at $2.86 November 2 close; base $4.49/+57.1%, stress $1.45/−49.3%; low confidence; earlier scenarios historical

Historical September 2028 target bridge: $1.5bn + $60m Taiwan + $300m Superbank/Atome combined − $160m reinvestment = $1.7bn. Taiwan rejection supersedes the breakdown. Q3 reaffirmed $1.7bn and 30%+ revenue CAGR conditional on Atome, without a full replacement bridge. The total is neither confused with the old breakdown nor mechanically reduced by $60m. Q3 remarks

Sources and interpretation

  • FY2025 call, February 11, 2026: 2025 results and original 2028 plan.
  • Q1 call, May 4: demand, fuel, lending and repurchases.
  • Q2 call, August 3 US time: operations, Superbank, Stash and outlook.
  • September 15 M&A call: Atome and changed plan.
  • Q3 release, dated November 3 in Singapore, period through September 30; presentation, liquidity/credit; prepared remarks, not full Q&A. Q2 release remains comparison.
  • Atome terms, September 15: both stages and pricing.
  • Price history, verified November 3: $2.86 November 2 close, before results. $3.048 at 7:59 p.m. EST is separate and excluded. Statistics and forecast are historical; September consensus not reverified.
  • Original Q3 date is historical; actual release is dated November 3 Singapore. October 30 board change.
  • Sarawak, October 8: framework, not quantified investment.
  • Uber offer document, August 27: Delivery Hero offer terms.
  • Malaysia Budget 2027, item 34, October 9 speech: partner measures from 2027.
  • Atome investor presentation, September 15: transaction framework/selected cohorts, historical rather than new earnings.
  • Video, transcript added October 9: secondary commentary, upload date not a verified video publication date.
  • Investment Analysis Rules, methodology 1.2, read November 3; independent authorship and two-month earnings-date rule; unchanged, earlier ratings remain historical.

Reported figures are company data; forecasts/synergies are management statements. Scenarios, ratings and monitoring thresholds are independent analysis. Confidence is high in reported data and medium in future plan delivery.

| 4 Nov 2026 | The first post-results close reached USD 3.12; the price rose 15.13% over two sessions. Added the earnings Q&A and implied Q4 EBITDA of USD 230–240 million. | Company strength remains 6.8/10; price attractiveness falls from 5.6 to 5.2/10. Base USD 4.49 / +44.0%, adverse USD 1.45 / −53.6%. |

| 5 Nov 2026 | The 4 November close reached USD 3.36; no new evidence of a change in the business. | Company strength remains 6.8/10; attractiveness falls from 5.2 to 5.0/10. Base USD 4.49 / +33.7%, adverse USD 1.45 / −56.9%. |

November 7, 2026: November 6 close of USD 3.41; added management sale filings under 10b5-1 plans. Strength of 6.8/10 and attractiveness of 5.0/10 unchanged; scenarios of USD 4.49 (+31.8%) and USD 1.45 (−57.5%).

Company strength is 6.7/10, assessed November 14, 2026 under methodology 1.2. Financial resilience falls from 7.0 to 6.5/10 because of binding funding elements and the size of the potential contribution alongside Atome. The other four dimensions are unchanged; price is excluded. At the Nasdaq regular close of USD 3.24 on November 13, 2026 (4 p.m. EST, Class A), assessed on November 14 under methodology 1.2, price attractiveness is 4.9/10 with low confidence. The base case is USD 4.32 (+33.2%) and the adverse case USD 1.28 (−60.5%), both without dividends. The lower entry price did not offset greater capital uncertainty. The agreements with SoftBank and PETROS announced on November 9 include development and funding obligations. Of USD 1.85 billion in proposed initial capital, Grab would contribute USD 750 million: USD 100 million in Sarawak and USD 650 million in Peninsular Malaysia. Funding is conditional and staged; the money has not already been spent. The plan for at least 5 GW has no substantiated timeline for returns. SEC 6-K · SoftBank and Grab.

November 21, 2026 — methodology 1.2. The lower USD 2.75 entry price raises price attractiveness to 5.5/10 under unchanged USD 4.32 / 1.28 cases. Company strength remains 6.7/10; risks from Atome and the planned USD 750m AI contribution are unchanged.

Was this analysis useful to you?

Have a suggestion about this analysis?

Suggestions are not displayed publicly. They help inform future updates. No name or email is required.

Sources and scenarios reflect the date stated in the analysis.Back to all companies