Grab Holdings Limited (NASDAQ: GRAB) — A Real First Profit, Flattered by Interest Income and Locked Under a 75%-Vote Founder

Independent equity research — published 2026-07-04. Reporting currency: USD (IFRS). Primary sources: Grab Holdings 20-F FY2025 (filed 2026-03-06), FY2024/FY2023 20-Fs, 6-K earnings releases through Q1-2026 (2026-05-04), Form 4/144 filings, and public market data.

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only. It is not investment advice and not a recommendation to buy or sell any security. The analysis that follows is deliberately position-free and carries no price target — this block is the single exception.

Verdict: HOLD / accumulate-on-weakness into the low-$3s. Great narrative, genuinely improving business, fair-to-full price, unproven quality, and a governance red flag that caps my conviction. Not a short; not a fresh long here at ~$3.90.

Directional valuation zone: fair value ~$3.75–$4.50 (≈3.5–4.5x forward EV/Sales on ~$4.0B FY26 revenue; SOTP central ~$4.0–4.3). I’d start accumulating sub-$3.50 (~3x sales / ~8x EV/gross-profit for a 20% grower with a $5.4B net-cash fortress and a real profit inflection) and would want a margin of safety before paying up. A single hard target would be false precision on a name whose profit is two-thirds treasury yield.

Grab is the rare Southeast-Asian growth story where the operating turn is real: gross margin has gone from −105% (2020) to +43% (2025), Mobility is a genuine cash engine (~$690M segment EBITDA on flat incentives), group Adjusted EBITDA compounded to $500M, and 2025 was the first full-year GAAP profit in the company’s 14-year, ~$17.5B-cumulative-loss history. But three things keep me at HOLD. First, the quality of that profit: audited IFRS operating profit was only $65M (1.9% margin); roughly 76% of pretax income is net finance income — interest on the ~$7.4B cash pile — and the 20-F already notes interest income fell year-on-year as rates eased. About a quarter of owners’ EPS is digital-bank losses shifted to a JV partner. Second, the price: “cheapest-ever” is true only against Grab’s own 10-year history (own-history price/sales 4.5th percentile); cross-sectionally, at ~9x EV/gross-profit it is the richest of its SE/UBER/DASH/MELI/CPNG peer set. Third, the governance: on 24 March 2026 shareholders voted to double the founder’s super-voting Class B shares from 45 to 90 votes, lifting Anthony Tan to ~74.9% of the vote on ~3–3.5% of the economics — an entrenchment ratchet, executed precisely as management is redeploying capital outside its Southeast-Asian moat (foodpanda Taiwan $600M, a US neobank stake, a US remote-driving startup).

Framing: a falling-knife-that-inflected — an abandoned Singapore/EM-beta name (−77% off its 2021 SPAC high, factor DNA that is regional-macro + credit, Momentum zeroed, Quality negative) that just posted its first real profits. This is a show-me story, not a crowded trade. Conviction: medium. The single piece of evidence that flips me bullish: Financial Services reaching sustained Adjusted-EBITDA breakeven and core on-demand segment margin pushing toward the mid-20s% while the share count actually shrinks. The single piece that flips me bearish: revenue growth slipping below 15% ex-FX while operating income goes flat and the “growth” leans further on interest income. Tag: “First profits are real — but you’re paying up for a Singapore macro proxy that still has to prove the fintech.”

📈 Stock Price Action — Five-Year Event Map

Grab is a completed round-trip and then some. Via the Altimeter Growth Corp (AGC) SPAC the stock touched an intraday high of ~$18.11 (Jan 2021) and closed as high as ~$17 (Nov 2021) on merger euphoria, collapsed to an all-time-low close of $2.31 (Oct 2022), base-built in the low-$3s for two years, re-rated to ~$6.6 (late 2025) as GAAP profitability arrived, then gave most of it back. It trades at ~$3.90 (2026-07-02) — ≈−77% off the all-time high, in the lower third of a 52-week range of roughly $3.18–$6.62 (≈−41% off the 52-week high), having bounced ~+8% off the recent low over the last quarter. (Pre-Dec-2021 prices are the AGC SPAC vehicle, not the operating company.)

Cycle narrative. (1–2) The five-year arc opens as a SPAC boom-bust: those early moves are the story of the listing vehicle, not the business — AGC ran to ~$17–18 on hype and cratered on the December-2021 close-and-de-rate. (3–4) The operating company’s first public prints — a ~$1.1B Q4’21 loss into a rising-rate rout — drove the −37% single-day crash and the grind to the $2.31 all-time low. (5) The 2022–24 base and the late-2024 breakout track the fundamentals turning: gross margin climbed from ~5% to ~42% and group Adjusted EBITDA flipped positive. (6) 2025’s run to ~$6.6 priced the first GAAP profit and fintech optimism. (7–8) The −45% give-back into mid-2026 reflects doubts about growth durability and margin quality, with a modest last-quarter bounce. Every price move is a Fact; every attributed cause is Interpretation, cross-referenced to earnings prints and macro sessions.

1. Executive Summary

Grab Holdings is Southeast Asia’s leading “super-app,” offering mobility (ride-hailing), deliveries (food, grocery, parcel), and digital financial services across eight countries and 900+ cities, monetized through a single consumer app plus driver- and merchant-partner apps. FY2025 revenue was $3,370M (+20% YoY; +18% constant-currency), on On-Demand GMV of ~$22.1B (+21%) and ~47.2M monthly transacting users (MTUs, +14%), rising above 50M by year-end and to 52M in Q1-2026.

The headline event is that 2025 was Grab’s first full-year GAAP profit — owners’ net profit of ~$268M, versus a −$105M loss in 2024 and cumulative lifetime losses of roughly $17.5B. This is a genuine milestone and the product of real operating leverage: gross margin has scaled from −105% (2020) to +43% (2025), and group Adjusted EBITDA reached $500M (+60%). But the profit’s quality is mixed. Audited IFRS operating profit was only $65M (a 1.9% margin); approximately 76% of pretax income is net finance income — interest earned on the company’s ~$7.4B treasury — and roughly a quarter of owners’ EPS reflects digital-bank losses shifted to a joint-venture partner (non-controlling interests). Real free cash flow was slightly negative (−$44M) as the digital-bank loan book absorbed working capital; management’s normalized “Adjusted FCF” was ~$290M.

The business quality is narrow, not broad. The only segment that clearly earns its cost of capital is Mobility (~$690M segment Adjusted EBITDA, a rational Grab/GoTo duopoly with cooling subsidies). Deliveries is large but low-margin (~2% of GMV, multi-homed, contested by foodpanda, ShopeeFood/Sea, and Gojek). Financial Services is the growth flyer and the loss sink simultaneously (revenue +37%, loan book +120% to ~$1.4B, deposits ~$1.6B, but −$110M segment EBITDA, guided to breakeven in H2-2026). Consolidated returns on capital remain weak: ROIC ~2%, ROE ~4% — positive, but well below any reasonable cost of capital.

Capital allocation is mixed with a clear governance negative. Management has begun a serial-acquirer pivot outside its SEA moat (foodpanda Taiwan ~$600M; 50.1% of US neobank Stash Financial, ~$425M EV, closed 1 July 2026; US remote-driving startup Vay ~$55M). Buybacks have restarted ($500M program, $400M accelerated in Q1-2026) but net of stock-based compensation and a June-2025 $1.5B convertible, the share count has not yet shrunk. Most starkly, on 24 March 2026 shareholders approved doubling founder Anthony Tan’s Class B super-voting rights from 45 to 90 votes per share, lifting his voting control to ~74.9% on ~3–3.5% of the economics — an entrenchment ratchet enacted as the company redeploys capital into unproven adjacencies. The long-rumored ~$7B GoTo merger remains stalled on Indonesian antitrust.

At ~$3.90 (EV ~$14–15.7B), Grab trades at ~3.9x EV/sales and ~9x EV/gross-profit — its cheapest-ever price on its own 10-year sales history (own-history price/sales 4.5th percentile), yet the richest EV/gross-profit in its peer set. The market is pricing a durable ~18–20% grower with margins expanding toward the mid-20s%; that is not demanding if execution holds, but it embeds the FinServices turn and continued mobility discipline. This report takes no position (see Claude’s Take for that); the body’s job is to separate the real inflection from the interest-income tailwind, the narrow moat from the marketed one, and a good business from a fully-priced stock.

2. Business Overview

Grab operates an integrated consumer platform (“super-app”) across Cambodia, Indonesia, Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Vietnam (the company describes entering a ninth market in 2026). A single app gives consumers access to four reportable activity clusters; drivers and merchants use dedicated partner apps. The company is Cayman-domiciled, headquartered in Singapore, reports in USD under IFRS, and — as a foreign private issuer — files a Form 20-F annually plus 6-Ks (not 10-K/10-Q).

Reportable segments and FY2025 economics (revenue is reported net of $2.3B of consumer/partner incentives):

Segment margins are of segment revenue; management also frames Mobility/Deliveries margins as a percentage of GMV (Mobility ~8.7% of GMV, Deliveries ~2.0% of GMV), which is the more comparable industry lens. Both denominators appear in this report and are labeled.

How Grab makes money. In Mobility and Deliveries, Grab takes a commission (take rate) on gross bookings (GMV): ~13% of Deliveries GMV (~$14.2B) and a broadly similar ride-hailing commission, plus vehicle-leasing revenue (GrabRentals, ~$194M) and, increasingly, advertising sold to merchants (a high-margin, take-rate-lifting layer). In Financial Services, Grab earns lending net interest margin and fees through its digital banks (GXS Bank in Singapore, GXBank in Malaysia, Superbank in Indonesia), payments/e-money, lending (GrabFin + Validus), and insurance distribution. GrabUnlimited, a paid subscription, drives frequency and accounts for roughly one-third of Deliveries GMV.

Recurring vs. non-recurring. Revenue is transactional but highly repeat: the platform is a habitual daily-life utility for a large, growing MTU base, and subscription (GrabUnlimited) plus digital-bank deposits (~$1.6B) create stickier, quasi-recurring relationships. There is no long-term contracted backlog; “recurring” here means high-frequency habitual usage, not SaaS-style contracts.

Verdict. A diversified, genuinely integrated consumer platform with a real and improving monetization engine, but a revenue base that is still thin in absolute margin (group operating margin <2% on an IFRS basis) and dependent on a single profitable leg (Mobility) subsidizing two developing ones (Deliveries ads-led margin expansion; Financial Services still loss-making).

3. Industry Dynamics

Structural tailwinds are real. Southeast Asia is young, mobile-first, rapidly urbanizing, and structurally underbanked — the classic substrate for on-demand mobility, delivery, and digital-financial-services penetration. The regional digital economy has grown double-digits for a decade, and Grab’s on-demand GMV compounding at ~16–21% is consistent with a still-penetrating market rather than a saturated one.

But “attractive TAM” ≠ “attractive industry structure.” Applying a supply-side (Marathon Capital Returns) lens, each of Grab’s three arenas sits at a different point of the capital cycle:

• Mobility (ride-hailing): the one attractive pocket. After the 2018 Grab/Uber-SEA combination and years of subsidy warfare, Mobility has consolidated into an approximate Grab/GoTo (Gojek) duopoly in most core markets, with on-demand incentives holding flat at ~10% of GMV for three years even as GMV grew — evidence that the capital cycle has turned from over-supply toward pricing discipline. This is where Grab earns real economics (~8.7% of GMV in segment EBITDA). But it is not a fortress: the 20-F itself flags the risk of Uber/Didi re-entry, and the on-again/off-again GoTo merger talks imply that neither incumbent is comfortable competing indefinitely — a “duopoly of mutual exhaustion.”

• Deliveries: structurally difficult. Food delivery is a low-margin, high-frequency, multi-homed business globally; SEA is no exception, contested by foodpanda (Delivery Hero), ShopeeFood (Sea), Gojek, and Line Man Wongnai (Thailand). foodpanda’s regional retreat (and Grab’s pending acquisition of foodpanda Taiwan) shows consolidation is underway, but Grab’s ~2%-of-GMV delivery margin confirms this leg does not, on its own, earn its cost of capital yet. Advertising is the margin lever that could change that.

• Financial Services: greenfield, capital-intensive, regulator-gated. Digital banking in SEA is early and lucrative in theory (underbanked populations, high NIM), but it is loss-making at Grab’s scale, credit-cycle-exposed, and heavily regulated: MAS’s Singapore digital-full-bank regime carries a ~$1.2B minimum-capital requirement; Indonesia caps foreign ownership (85%/49% structures on OVO), imposes a 25% funding cap on P2P lending, and is separately tightening gig-labor rules (see ). This is an option, not yet a business.

Regulatory landscape. Beyond digital-bank capital rules, the binding sector risks are (i) gig-worker classification / minimum-earnings rules — e.g., Indonesia’s early-2026 move toward an 8% commission cap on two-wheel (O2O) ride-hailing (management says two-wheel is <6% of its Mobility GMV, limiting direct exposure), and EU-style reclassification pressure globally; (ii) antitrust, which has repeatedly blocked or complicated consolidation (Singapore’s CCS on the 2018 Uber deal and the abandoned Trans-cab acquisition; Indonesia’s KPPU on the GoTo merger); and (iii) e-money / payments and data-localization rules across eight jurisdictions.

Verdict: structurally improving, not structurally good. Only Mobility currently exhibits the consolidated, discipline-restored structure that supports durable returns. Deliveries is competitive and thin; Financial Services is greenfield and loss-making. The secular TAM is genuine, but a diversified position across three arenas of very different quality means the blended industry attractiveness is mediocre — one good business carrying two developing ones.

4. Competitive Position

Name the moat — and its limits. In Greenwald’s Competition Demystified taxonomy, Grab’s genuine advantage is economies of scale combined with local demand-captivity via network density — concentrated in Mobility. The financial proof is unambiguous: Mobility produces ~$690M of segment Adjusted EBITDA (≈8.7% of its GMV) on flat incentives, a level of profitability that only the scale leader in a two-sided liquidity market, with the densest driver/rider network per city, can sustain. Density is self-reinforcing (more drivers → shorter wait times → more riders → more drivers), and it is local — a competitor must rebuild it city-by-city, which is why the SEA market rationalized into a Grab/GoTo structure rather than fragmenting.

Pressure-testing the broader “super-app moat.” Management’s framing (echoed on the Q1-2026 call) is that 14 years of proprietary data — “20 billion transactions,” a hyper-local mapping and payments layer, an “AI intelligence layer,” and a physical fulfillment network — constitute “our biggest competitive moat.” Treat this as a hypothesis, and apply the test: does the claimed moat tie to a financial outcome that would deteriorate without it?

• Mobility: PASSES. The scale/density advantage is directly visible in segment economics and in stable/rising take rates on flat incentives.
• Deliveries: FAILS (mostly). The two-sided network is real, but food delivery is multi-homed (consumers and riders use multiple apps) with low switching costs, and the ~2%-of-GMV margin shows the “moat” does not translate into pricing power. The lever that could build one — high-margin advertising and GrabUnlimited lock-in — is promising (advertiser spend +44% YoY, GrabUnlimited ~⅓ of Deliveries GMV) but unproven at scale.
• Financial Services: NO MOAT YET. GXS/GXBank/Superbank are subscale challengers competing against entrenched banks and other fintechs; the segment loses money, and any advantage rests on ecosystem cross-sell (cheap customer acquisition from the app), which is a distribution edge, not yet a durable economic moat.

Cross-sell flywheel — real but modest. The super-app thesis is that owning the consumer relationship lets Grab acquire fintech/delivery customers at near-zero marginal cost. There is evidence for it (deposits gathered cheaply, lending disbursed to platform users, GrabUnlimited frequency uplift), but the headline metric is sobering: GMV per MTU rose only ~+4% — growth is coming more from new users than from deeper wallet share per user, which is what a powerful cross-sell flywheel would produce.

Direct comparison. Versus GoTo (Indonesia-centric, also improving toward profitability), Grab is the broader regional leader with the stronger balance sheet. Versus Sea (Shopee/ShopeeFood/SeaMoney), Grab is smaller and growing slower but with a comparable gross-margin structure; Sea’s e-commerce-led model is a different animal. Versus global on-demand peers (Uber, DoorDash), Grab’s take rates and gross margins are broadly comparable, but its scale, absolute margins, and market maturity lag.

Verdict: a genuine but NARROW moat. Durable competitive advantage exists and is measurable — in Mobility, via density-driven scale economics. Deliveries is a crowded, low-switching-cost market with a potential (not yet realized) advertising/subscription moat; Financial Services has no durable moat yet. The group multiple blends one advantaged business with two that must still prove they can earn their cost of capital. Critically, much of the Mobility position was acquired (the 2018 Uber-SEA deal, local licenses), and the live 20-F risk of well-capitalized re-entrants means even the good leg is defended, not impregnable.

5. Growth History and Forward Opportunities

History — high-quality, organic, penetration-led. The multi-year record is one of durable top-line compounding alongside a dramatic margin repair:

(Adjusted EBITDA figures are management non-IFRS; operating profit and net income are audited IFRS. Early-year revenue growth is partly optical — the 2022 jump reflects a change in how consumer incentives are recognized against revenue — but the post-2023 ~20% growth is clean and organic.)

Growth has been entirely organic and penetration-driven, not acquisition-manufactured, and — importantly — accompanied by operating leverage: Adjusted EBITDA has been growing roughly 3x as fast as revenue. That is the single strongest data point in the bull case.

Forward opportunities.

1. Mobility depth: EV transition (drive-to-own programs with BYD/GAC, 30,000+ EVs on-platform in Thailand), affordability products (group rides, GrabShare), and AI-driven marketplace efficiency (management cites a 23% earnings uplift for drivers using its “Turbo” AI mode, and Mobility transactions growing 28% — faster than GMV).
2. Deliveries → advertising + grocery: advertising is the margin lever (advertiser spend +44% YoY); GrabMart (grocery) is ~10% of Deliveries GMV but growing ~1.7x faster than food, with a much larger TAM and higher basket sizes — management points to global peers at 20–40% grocery penetration.
3. Financial Services: the growth flyer — loan disbursals +67% YoY to >$1B/quarter, loan book +120% to ~$1.4B, deposits ~$1.6B, guided to segment Adjusted-EBITDA breakeven in H2-2026 and a $2B loan book by year-end 2026. If the digibanks turn cash-generative, this is the segment that re-rates the story.
4. AI and autonomous vehicles: management is investing in an “AI intelligence layer” (with a conscious step-up in regional corporate costs to ~$114M/quarter) and running small AV pilots (WeRide in Singapore, plus May Mobility and Vay partnerships/investments), while explicitly signaling AV is not a near-term disruptor in SEA.

Guidance. FY2026: revenue $4.04–4.10B (+20–22%), Adjusted EBITDA $700–720M (+40–44%); a 2028 target of $1.5B Adjusted EBITDA was reiterated. The compounding is real, but note the caveat below.

Verdict: high-quality growth — with an incentive asterisk. Organic, penetration-led, margin-accretive growth is genuinely high quality. The asterisk: On-Demand incentives rose to ~10.5% of GMV (+46bps) in FY2025/early-2026 on a regional “fuel crisis,” so recent growth is not incentive-free, and spend-per-user is rising only modestly. Growth is real; it is not yet self-evidently cheap growth.

6. Financial Quality

This is the section where the bull and bear cases are decided, because the FY2025 GAAP-profit headline is real but heavily flattered, and the distinction matters enormously for valuation.

The normalized earnings bridge (IFRS, audited 20-F, US$M):

Two flatters, quantified.

1. Only ~$65M (24%) of pretax profit is operating (a 1.9% margin); ~$203M (76%) is net finance income — interest earned on the cash pile, which is rate-sensitive and which the 20-F notes fell ~$19M YoY “driven by lower interest rates.” As rates in the region normalize down, this tailwind fades.
2. Owners’ profit of $268M exceeds group profit of $200M because ~$68M of digital-bank losses are pushed to JV partners (non-controlling interests, principally Singtel’s share of GXS/Superbank). So ~25% of owners’ EPS is loss-shifting, and that NCI drag grew (−$53M → −$68M) as the digibanks scaled losses.

Reconciliation note: Third-party aggregators (ROIC.ai) reclassify net finance income into “operating income” and show ~$222M operating income and ~$268M net income; the audited IFRS statements show $65M operating profit and $200M total profit / $268M to owners. This report uses the audited IFRS figures as primary. The honest one-line summary: operating profitability inflected to a thin positive; the bulk of the reported profit is treasury yield and accounting geography.

Segment economics confirm the structure (see table): Mobility (~$690M) funds the group; Financial Services (−$110M) is the loss sink (loss margin narrowed from −41% to −32%, but the absolute loss widened). “Group Adjusted EBITDA $500M” is a management non-IFRS metric that adds back $241M of real SBC (≈48% of the metric) plus ~$140M of impairment — i.e., the headline profitability figure is roughly half stock-based compensation add-back. IFRS/GAAP correctly expenses SBC; investors should weight the IFRS numbers.

Cash flow and the float. Reported free cash flow was −$44M (operating cash flow $79M − capex $123M); OCF collapsed ~91% from $852M in 2024. The swing is bank-book working capital: the loan book grew ~$691M (a use of cash) while deposit inflows and float move the line around year-to-year (2024’s $852M OCF was itself inflated by an ~$843M deposit inflow). Management’s normalized “Adjusted FCF” was ~$290M (TTM $489M by Q1-2026), which reasonably strips bank-book working capital — a defensible adjustment. The takeaway: IFRS OCF is nearly useless as a corporate-cash proxy for a company that is now part-bank; true underlying corporate FCF is modest and positive (~$150–290M), not the $739M that 2024’s headline implied and not the negative $44M that 2025’s headline implies.

Balance sheet — a genuine fortress. Gross cash and liquidity of ~$7.4B; net cash liquidity ~$5.4B after debt. The 2023 Term Loan B (~$2B) was fully repaid in 2024 and replaced by a ~$1.5B June-2025 zero-coupon convertible (conversion ~$6.55, a ~228.9M-share overhang). Roughly $1.6B of “cash” is matched by e-wallet float and digital-bank customer deposits — customer money, not distributable, already netted from the $5.4B. Even so, this is one of the strongest balance sheets in the peer group and a real source of downside support (net cash is ~⅓ of the market cap).

Dilution and SBC. Share count rose from ~3,742M (2020) to ~4,089M (2025), +9.3% over five years but decelerating to <2%/year; SBC has fallen from $412M (2022) to $241M (2025, ~7.2% of revenue) and is now partly offset by buybacks. Directionally good, but SBC still consumes roughly half of Adjusted EBITDA and, combined with ESPP and the convertible, the share count has not yet begun to shrink.

Returns on capital. ROIC ~2.1%, ROA ~2.5%, ROE ~4% — positive for the first time, but far below cost of capital. A business can be GAAP-profitable and still be destroying economic value if it earns less than its WACC on invested capital; Grab is at that threshold. The bull case requires these to climb materially.

Verdict: economics improve with scale — partially and thinly. The operating leverage is real (gross margin 43%, Mobility ~57% of segment revenue), but the profit inflection is modest ($65M operating), ~76% of reported profit is rate-driven, ~25% is JV loss-shift, and returns on capital are still sub-WACC. The inflection is durable only if operating leverage compounds faster than the interest-income tailwind fades and Financial Services reaches breakeven.

7. Capital Allocation

Verdict up front: mixed, with a clear and worsening governance negative. Management has run the operating business with genuine cost discipline, but the capital-allocation posture — serial M&A drifting outside the core moat, buybacks that don’t yet shrink the count, a convertible overhang, and above all a founder-entrenchment vote — is not unambiguously shareholder-friendly.

M&A. Historically sensible SEA bolt-ons: Everrise (~$54M grocery), Validus (~$48M, feeding GXS lending), Jaya Grocer, and — to its credit — Grab walked away from the Trans-cab taxi acquisition when Singapore’s competition authority objected. But a distinct serial-acquirer pivot outside the SEA moat is now underway:

• foodpanda Taiwan (~$600M) — its first market outside Southeast Asia (pending regulatory approval);
• Stash Financial (US neobank), ~$425M EV, 50.1% closed 1 July 2026 — a US-domiciled digital-investing/neobank stake, a strategically odd move away from the regional platform. (Note: this is Stash Financial, Inc., not the Singapore robo-advisor “StashAway” — a common confusion.)
• Vay (~$55M) — a US remote-driving/AV startup.

This is real capital ($1B+) flowing into unproven adjacencies and geographies where Grab has no density moat. It may be prescient optionality; it may be empire-building. Given the governance backdrop (below), outside holders have limited ability to check it.

Buybacks vs. dilution — only marginally “real.” The Feb-2024 $500M program was “substantially completed” ($226M in 2024 + $274M in 2025), but its largest single tranche (58.45M shares @ ~$4.68) was an off-market block purchased from convertible-note holders to facilitate their hedging — issuance-linked, not a classic open-market return. SBC + ESPP still outran repurchases in share terms (count crept from ~4,070M to ~4,098M). A new $500M program (Feb-2026) — a $250M JPMorgan ASR plus up to $150M Morgan Stanley contingent forward, and ~$400M accelerated in Q1-2026 — is a genuine step-up, and management explicitly frames it as conviction “at these dislocated prices.” But net of dilution and the 228.9M-share convertible overhang, the buyback is not yet reducing the count. Verdict: improving intent, not yet accretive in practice.

Founder control — the sharpest negative. Founder-CEO Anthony Tan controlled ~59–60% of the vote on just ~3.2–3.5% of the economics via proxies over departed co-founders’ Class B super-voting shares. On 24 March 2026, an EGM passed a resolution doubling Class B voting power from 45 to 90 votes per share, lifting Tan to ~74.9% of total voting power. He can elect or remove the entire board. This is an entrenchment ratchet, executed precisely as management redeploys capital into unproven, out-of-region bets — a governance structure that concentrates control while diffusing accountability. For an outside minority holder, this is a material, permanent negative that no operating result fully offsets.

Insider behavior. The 2026 Form 4/144 corpus shows heavy insider selling, but 100% routine Rule 10b5-1 planned sales (CHRO Ong Chin Yin, President/COO Alex Hungate, others) — diversification, not signal — with zero open-market discretionary purchases (code P). A March-2026 cluster of Form 3s reflects newly-designated Section 16 officers, not trading. The meaningful capital signal is the company buyback, not insider trades; there is no bullish insider-conviction tell.

Verdict: has management allocated capital intelligently? Operationally, yes — cost discipline drove the margin turn, and the balance sheet is pristine. At the capital-allocation and governance level, only partially — the M&A is drifting out of the moat, the buyback doesn’t yet shrink the count, and the founder just voted himself supermajority control. This is the weakest pillar of the thesis.

8. Changes and Headwinds — Last Two Years

Strengthening the thesis:

• Profitability inflection — first full-year GAAP profit (2025), 17 consecutive quarters of Adjusted-EBITDA growth, self-funding underlying FCF.
• Financial Services scaling — loan book +120% to ~$1.4B, deposits ~$1.6B, guided to breakeven H2-2026.
• Superbank (Indonesia) consolidation (announced 2026-05-20): GXS acquired Singtel’s stake, taking Grab above 50% — this enlarges the FinServices loan book and the reported losses, and shifts accounting from associate to consolidation (a modeling item for FinServices trajectory).
• Restarted capital return — $500M buyback program and $400M ASR in Q1-2026.
• AI-driven efficiency — measurable driver/merchant productivity gains and more efficient incentive spend.

Weakening the thesis:

• Governance entrenchment — the March-2026 doubling of founder voting power to ~74.9% (see ).
• Capital leaving the moat — foodpanda Taiwan, US neobank Stash, Vay.
• Convertible overhang — $1.5B June-2025 zero-coupon convertible, ~229M potential shares.
• Persistent FinServices losses and rising credit provisions (ECL overlays added in Q1-2026 for macro uncertainty).
• Regulatory friction — Indonesia’s 8% two-wheel commission cap (management: <6% of Mobility GMV, so contained but a precedent), gig-labor rules region-wide, and a stalled ~$7B GoTo merger blocked by Indonesian antitrust (KPPU) — an on-again/off-again overhang rather than a catalyst.
• Macro/FX — a regional “fuel crisis” (Middle East conflict, higher fuel) pushed driver incentives up in Q1-2026 (management expects Q1 to be the peak); FX added ~2 points to reported FY2025 growth and would reverse on USD strength.

Verdict: net neutral-to-slightly-negative. The business is unambiguously stronger than two years ago; the ownership/capital-allocation wrapper around it is worse. On balance, a genuinely improving operating story carrying more governance and capital-deployment risk than it did before.

9. Risk Analysis (Risk Matrix)

Catastrophic-loss assessment. The risk of a total loss is low: Grab has ~$5.4B net cash, no near-term debt wall, and positive (if modest) underlying FCF, so bankruptcy is not a realistic scenario. The realistic downside is value erosion — the stock de-rating toward net cash + a modest multiple on a business that grows more slowly and stays sub-WACC — not a wipeout.

10. Valuation Discussion (Embedded Expectations)

No price target and no recommendation appear in this section (see Claude’s Take for the single labeled exception). This is an analysis of what the current price embeds.

Where it trades. At ~$3.90 (2026-07-02), market cap ~$15.9B and enterprise value ~$14–15.7B (the range reflects how much of the ~$3.4B+ short-term investment portfolio is netted as cash). On trailing figures: ~3.9x EV/sales, ~9x EV/gross-profit, ~30–40x EV/EBITDA, and a barely-meaningful ~45–76x P/E (first profitable year).

Own-history vs. cross-sectional — the central valuation tension. On its own 10-year history, Grab is at its cheapest-ever price/sales (own-history 4.5th percentile; composite valuation percentile 26th) — the “great narrative at a washed-out multiple” optic. But cross-sectionally it is not cheap:

EV/sales is not apples-to-apples (CPNG/MELI/SE book gross-er retail revenue; GRAB/UBER/DASH book net take-rate revenue). EV/gross-profit normalizes for that, and there GRAB (~9x) is the RICHEST of the cohort, roughly double its nearest business-model twin SE (3.9x, growing faster). “Cheapest-ever” is an own-history phenomenon, not a cross-sectional bargain.

Sum-of-the-parts (approximate). The blended multiple mixes one good business with two weaker ones, so a SOTP is the right lens:

• (a) Core on-demand (Mobility + Deliveries; profitable, ~18–22% growth; ~$0.8–1.0B core segment EBITDA): ~4–5x sales or ~15–18x segment EBITDA → ~$11–14B
• (b) Financial Services (digibanks + lending + payments; ~$0.35B revenue, still EBITDA-loss, narrowing): 3–5x revenue / option value → ~$1.0–1.75B
• © Net cash + investment portfolio → ~$1.4–3.5B (treatment-dependent)
• Equity value ≈ $14.7–19B → ~$3.60–$4.65/share; central ~$4.0–4.3. Current ~$3.90 screens fair-to-slightly-cheap; the upside is FinServices scaling and core-margin expansion, not the current run-rate.

Embedded-expectations / reverse-DCF sketch. At ~$14B EV on ~$3.55B TTM revenue (growing ~20%) and ~$469M TTM EBITDA, the price implies roughly 18–20% revenue CAGR with EBITDA margin expanding toward a ~25% steady state — i.e., revenue to ~$8B by 2030 at ~25% margin ≈ ~$2.0B EBITDA, or ~7x 2030 EBITDA. That is undemanding if execution holds; the genuinely demanding valuation was at $6+ in late 2025. The market appears to be underwriting correctly: the real operating inflection, the SEA on-demand duopoly, and the net-cash fortress. It is underwriting uncertainly: Financial Services turning cash-generative, GoTo competition staying rational, and the path to a mid-20s% margin as the interest-income tailwind fades.

Scenario analysis:

What the market is pricing correctly vs. incorrectly. Correctly: that this is a real, improving, well-capitalized franchise — hence a premium EV/gross-profit. Possibly incorrectly (in either direction): the durability of the margin path (the P/S at its 4.5th own-percentile plus a negative Quality factor loading says the market is skeptical of durability, not asleep) and the option value of Financial Services (which the SOTP arguably under-credits if the digibanks inflect).

11. Variant Perception

Consensus. The sell-side is broadly constructive (targets clustered ~$5–6): “profitable-growth Southeast-Asian super-app, first GAAP profit, net-cash balance sheet, fintech optionality, rational mobility duopoly.”

Strongest bull case. SEA digitization is a decade-long secular tailwind; Grab is the regional scale leader in a Mobility/Deliveries duopoly with cooling subsidies; it has proven operating leverage (gross margin −105% → +43%, operating loss → +$65M, Adjusted EBITDA to $500M); it sits on a $5.4B net-cash fortress; and it now trades at its cheapest-ever price/sales. If Financial Services inflects to profit and core margins push to the mid-20s%, the earnings power is multiples of today’s — and at 4.5th-percentile own-history valuation, that is an asymmetric setup.

Strongest bear case. Cross-sectionally the stock is not cheap (~9x EV/gross-profit vs. SE’s 3.9x growing faster); the GAAP profit is ~76% interest income and ~25% JV loss-shift, with underlying operating margin still <2%; Financial Services is an unproven-economics cash sink exposed to a regional credit cycle; returns on capital are sub-WACC; the founder just entrenched himself at ~75% voting control while deploying $1B+ into out-of-moat US/Taiwan bets; the convertible adds a 229M-share overhang; and the price sitting at its 4.5th own-percentile with a negative Quality factor loading says the market is deliberately skeptical of durability. This is a “show-me,” not a mispricing.

The 3–5 assumptions that matter most, each with a falsification test:

1. Core on-demand margin expands toward the mid-20s% (EBITDA/segment revenue). Falsify: segment-EBITDA margin stalls or declines for two consecutive quarters.
2. Financial Services reaches Adjusted-EBITDA breakeven in ~2026–27 and stays there. Falsify: FinServices loss widens YoY, or credit costs (ECL) spike through the cycle.
3. Revenue growth stays ~18–20% ex-FX. Falsify: growth drops below 15% constant-currency.
4. The Grab/GoTo duopoly stays rational. Falsify: on-demand incentives re-accelerate above ~11% of GMV or take rates compress.
5. Interest income is a bridge, not the story — operating profit carries the baton. Falsify: operating income goes flat while “net income grows” only on treasury yield.

Factor-positioning read (multi-factor risk model, All-Factors specification, R²≈0.33). Grab’s factor DNA is regional-macro + credit, not US quality-growth: dominant loadings are Market (+0.91) and Country:Singapore (+0.57), with a super-app mix of Social-Media/Online-Retail/Transportation/Fintech and a positive CreditRisk (+0.27) tilt; crucially, Quality is negative (−0.14), LowVol is negative (−0.25, i.e., high-vol), Growth is negligible (+0.03), and both Momentum and Value are zeroed (absent). Idiosyncratic vol is ~30% and roughly two-thirds of variance is stock-specific — it trades on its own news. Its nearest factor peer is SE (0.83), then a wall of EM/Singapore/Pacific-ex-Japan and EM-high-yield-bond ETFs. Track record: y5 −19.8% annualized (max drawdown −86.5%), y3 +6.3%, y1 −20.4%, m6 −38.9%, m3 +34.7% annualized (~+7.7% raw) — a multi-year falling knife that base-built, sold off again, and just bounced. The synthesis: an abandoned Singapore/EM-beta name at its cheapest-ever sales multiple that just posted its first real profits — but with factor DNA screaming regional-macro-and-credit, Momentum zeroed, and Quality negative. That is the profile of a contrarian show-me story, not a crowded momentum trade — which is exactly why the burden of proof sits on execution, not on the multiple.

12. Fact vs. Interpretation Table

13. Open Questions

1. How fast does interest income fade? With ~76% of pretax profit from treasury yield, the trajectory of regional rates and the reinvestment yield on the $7.4B pile materially swings reported earnings. What is the run-rate net finance income at normalized rates?
2. Financial Services unit economics through a full credit cycle. The loan book grew +120%; ECL overlays were added for macro uncertainty. What is the through-cycle net charge-off rate, and is the H2-2026 breakeven guide robust to a downturn?
3. Superbank consolidation mechanics. Post-May-2026, how much additional loss and loan book does full consolidation bring on-balance-sheet, and how does that reshape the FinServices trajectory and NCI line?
4. Rationale for the out-of-moat M&A. What is the strategic logic and return hurdle for a US neobank (Stash) and foodpanda Taiwan? Is this optionality or empire-building?
5. Does the buyback ever shrink the count? Given SBC (~$241M), ESPP, and the 229M-share convertible, at what buyback pace does the diluted share count actually decline?
6. Advertising as the delivery moat. Can advertising/GrabUnlimited lift Deliveries margin durably above ~2% of GMV, or does competition cap it?
7. GoTo endgame. Is the merger dead, dormant, or delayed — and what does the antitrust remedy set look like if it revives?

14. What Must Be True

For the BULL case to be right:

• Core on-demand (Mobility + Deliveries) operating leverage compounds, pushing group operating margin from ~2% toward high-single/low-double digits — independent of interest income.
• Financial Services reaches sustained Adjusted-EBITDA breakeven (H2-2026 guide) and demonstrates healthy through-cycle credit, turning the loss sink into a growth engine.
• Revenue growth stays ~18–20% ex-FX with incentives normalizing back toward ~10% of GMV as the fuel shock passes.
• Buybacks + decelerating SBC finally shrink the diluted share count.
• Falsification test: if, over the next 2–4 quarters, operating income (ex-finance income) goes flat or down while headline “net income grows,” and/or FinServices losses widen YoY, the bull thesis is broken — the profitability would be revealed as a rate-cycle artifact, not a durable operating turn.

For the BEAR case to be right:

• The GAAP profit proves rate-dependent and fades as interest income falls, with operating margin stuck near breakeven.
• Financial Services stays loss-making / suffers a credit-cycle hit; Deliveries margin stays ~2% under competitive pressure.
• Out-of-moat M&A destroys value; the founder’s ~75% control enables continued undisciplined deployment; the count keeps creeping up.
• Falsification test: if group Adjusted EBITDA and audited operating income and underlying FCF all compound together for several quarters, FinServices crosses into sustained profit, and the diluted share count declines, the bear thesis is broken — the business would be demonstrating genuine, self-funding, operating-driven compounding that justifies a premium multiple.

15. Source Appendix

Key primary sources below.

• Grab Holdings Limited, Form 20-F for FY2025 (filed 2026-03-06), income statement, balance sheet, cash-flow statement, segment note, share-capital/voting note, risk factors. https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001855612
• Grab Form 20-F FY2024 (filed 2025-03-14) and FY2023 (filed 2024-03-28) — multi-year trend.
• Grab 6-K earnings releases through Q1-2026 (2026-05-04 earnings call/transcript); Q4-2025 (2026-02-11); Superbank consolidation 6-K (2026-05-20); Stash Financial completion 6-K (2026-07-01).
• Grab Form 4 / Form 144 corpus (2024–2026) — insider-transaction read.
• EGM voting results (24-Mar-2026) — Class B super-voting amendment.
• ROIC.ai — standardized financials, ratios, enterprise value, valuation multiples, Q1-2026 transcript (cross-checked to 20-F; IFRS 20-F is primary).
• Own-history valuation percentiles — price/earnings, price/book, price/sales vs. the stock’s own multi-year range.
• Factor / price-action data — factor loadings, risk-adjusted track record, and 5-year daily price history.
• Peer comparables (public filings): Sea Ltd (SE), Uber (UBER), DoorDash (DASH), MercadoLibre (MELI), Coupang (CPNG).

APPENDIX A — Standard Diligence Questionnaire — Grab Holdings Limited (NASDAQ: GRAB)

Supplemental to the report above. Fact/Interpretation/Assumption labels applied where material. Report date 2026-07-04.

General

What thoughtful questions have other investors asked about this company? From the Q1-2026 call, the buy/sell-side pressed on: (1) the fuel-crisis impact on driver incentives and margins (management: Q1 was the incentive peak; FY guide reiterated); (2) Indonesia’s 8% two-wheel commission cap (management: two-wheel is <6% of Mobility GMV, four-wheel unaffected); (3) Financial Services operating leverage and loan-book/deposit dynamics (breakeven H2-2026, $2B loan book target); (4) buyback vs. SBC dilution (the $400M ASR ≈2% of count, “more than offsets” SBC — though the diluted count has not yet actually fallen); (5) AI monetization (bundled into take rate, not a separate SaaS line yet); and (6) regional corporate cost step-up for AI infrastructure (~$114M/quarter, expected to stabilize). The deeper questions worth adding: how much of the profit is interest income, and what is FinServices’ through-cycle credit?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: early-cycle for the operating business (first profitable year, margins expanding from a low base) but potentially cyclical-high on the finance-income component — ~76% of pretax profit is interest on the cash pile, and rates are easing. So operating earnings are near a cyclical low; total reported earnings benefit from a rate tailwind that is likely near its peak.

Driven by external environment or internal actions? Both: internal (cost discipline, incentive rationalization, AI efficiency) drove the operating turn; external (interest rates, FX, fuel prices) drives a large share of the reported bottom line.

How stable are revenues? Transactional but highly repeat — a daily-life utility with ~50M+ MTUs, one-third of Deliveries GMV under GrabUnlimited subscription, and a growing deposit base. Discretionary delivery/mobility demand is somewhat cyclical, but the base is habitual.

Outlook for products/services; how big will the market be? Large and growing — SEA on-demand and digital-financial-services penetration is early. Management guides FY2026 revenue $4.04–4.10B (+20–22%) and a 2028 Adjusted-EBITDA target of $1.5B. Growth is domestic-to-SEA-regional, now edging outside the region (foodpanda Taiwan, US Stash).

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Interpretation: Mobility is less competitive (Grab/GoTo duopoly, cooling subsidies); Deliveries remains highly competitive (foodpanda, ShopeeFood, Gojek, Line Man); Financial Services is a crowded greenfield vs. banks and fintechs.

How profitable is the business (ROIC, ROE)? Just-positive and sub-WACC: ROIC ~2.1%, ROE ~4%, ROA ~2.5% (FY2025). Positive for the first time, but not yet value-creating on a cost-of-capital basis.

How profitable is the industry; barriers to entry? Mobility earns real economics for the scale leader (density moat); Deliveries is structurally thin; FinServices is capital- and regulation-gated (MAS ~$1.2B digital-full-bank capital; Indonesia ownership caps). Barriers are local density (Mobility) and licensing/capital (FinServices), not brand or IP.

Can the business be easily understood? Moderately — the super-app model is intuitive, but the financials are complicated by IFRS-vs-non-IFRS metrics, bank-book working capital, JV/NCI loss-shifting, and finance-income geography. It takes work to see the real operating profit ($65M) behind the headline ($268M).

Undermined by foreign low-cost labor? No — it is the local-labor-marketplace; the relevant labor risk is regulatory (gig-worker classification, minimum earnings), not offshoring.

Do brands matter? Yes, moderately — “Grab” is a trusted regional consumer brand that lowers fintech/deposit customer-acquisition cost, but brand alone is not the moat (density is).

Nature of competition; switching costs? Two-sided liquidity competition. Switching costs are low for consumers and riders (multi-homing is common in delivery), moderate where GrabUnlimited and deposits create lock-in. The durable edge is density/scale in Mobility, not switching costs.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Interpretation: the density network, brand, data (“20B transactions”), and digital-bank licenses are internally-generated intangibles carried at little/no book value — a genuine but unquantified off-balance-sheet asset. Offset by ~$0.9B goodwill and ~$1.1B intangibles from past M&A that are on the books.

Off-balance-sheet liabilities? Principal items are the ~$1.5B convertible (on-balance-sheet, but with a ~229M-share dilution overhang) and customer float / e-wallet / deposit obligations (~$1.6B, on-balance-sheet as liabilities). No material hidden operating leases beyond disclosed.

How conservative is the accounting? Interpretation: mixed. IFRS treatment is standard and SBC is correctly expensed, but the presentation leans promotional — “Adjusted EBITDA” adds back ~48% SBC, and “Adjusted FCF” normalizes bank-book working capital favorably. Read the audited IFRS lines, not the adjusted metrics.

How CapEx-hungry? Light on physical capex (~$123M, ~3.6% of revenue — asset-light platform) but capital-hungry in a different sense: the digital-bank loan book absorbs working capital/regulatory capital, which is the real cash draw (loan book +$691M in 2025).

Capital Allocation & Management

How much FCF; how is it used; philosophy? Underlying corporate FCF ~$150–290M (normalized); reported FCF −$44M. Uses: reinvestment in FinServices loan book, out-of-moat M&A ($1B+), and a restarted buyback ($500M program, $400M ASR). Philosophy (stated): “highly disciplined, long-term shareholder value, diversification.” Interpretation: disciplined operationally, but the M&A drift and founder entrenchment temper the shareholder-friendliness.

Significant acquisitions recently? Yes — foodpanda Taiwan (~$600M), Stash Financial US neobank (50.1%, ~$425M EV, closed 2026-07-01), Vay (~$55M), plus prior SEA bolt-ons (Everrise, Validus, Jaya Grocer). Superbank consolidation (2026-05).

Buying back shares? Yes, but not yet net-accretive — SBC + ESPP + convertible have kept the count flat-to-up despite ~$500M repurchased in 2024–25 and a new $500M program in 2026.

Issuing large amounts of stock to insiders? SBC ~$241M (declining, ~7.2% of revenue); still ~48% of Adjusted EBITDA. Not egregious by early-stage-tech standards and improving, but material.

Compensation / motivations of management? Key negative: founder-CEO Anthony Tan holds ~74.9% of votes on ~3–3.5% of economics after the March-2026 super-voting amendment — extreme control/economics mismatch. Other insiders sell routinely (10b5-1); no open-market buying.

Valuation & Market Data

ADR, MLP, or K-1 issuer? Fact: Grab is not an ADR — it is a Cayman-domiciled foreign private issuer with ordinary shares (Class A) listed directly on NASDAQ; it files a 20-F. No K-1; no MLP structure.

Dividend policy? None — no dividend; capital return is via buyback only.

How profitable is the business? GAAP-profitable for the first time (FY2025), but thinly on an operating basis (1.9% margin) and sub-WACC on returns.

Is net income diverging from cash from operations? Yes, materially, and in both directions across years — a QoE flag. FY2025: net income +$268M but OCF only $79M / FCF −$44M (loan-book drag). FY2024: net loss −$105M but OCF +$852M (deposit inflow). Bank-book working capital makes IFRS OCF an unreliable proxy; use normalized/adjusted FCF (~$290M).

Risks & Downside

What would cause the stock to decline? Fading interest income exposing thin operating margins; FinServices losses widening or a credit-cycle hit; growth decelerating below 15%; competitive re-intensification (GoTo/Sea/Uber-Didi); value-destructive out-of-moat M&A; multiple compression (already the richest EV/gross-profit in its peer set).

Risk of catastrophic loss? Low — $5.4B net cash, no near-term maturities, positive underlying FCF.

Chance of total loss? Very low — solvency is not the risk; value erosion toward net-cash-plus-modest-multiple is the realistic downside.

Recent News & Events

Has the business environment changed recently? Yes: (i) first GAAP profit + reiterated FY2026 guidance; (ii) Superbank (Indonesia) consolidation (2026-05); (iii) Stash Financial (US) 50.1% completion (2026-07-01); (iv) founder voting power doubled to ~74.9% (2026-03-24); (v) regional fuel-cost shock lifting driver incentives (Q1-2026 peak); (vi) Indonesia 8% two-wheel commission cap (contained); (vii) foodpanda Taiwan acquisition pending; (viii) $500M buyback / $400M ASR. (Timeline compiled from 6-K filings and the earnings-call transcript.)

Significant acquisitions? See above — foodpanda Taiwan, Stash Financial, Vay, Superbank consolidation.

Change in accounting policies? No material change flagged; Superbank moves from associate to full consolidation (a scope, not policy, change).

Recent changes — new markets, facilities, management? Entering a ninth market in 2026; first market outside SEA (foodpanda Taiwan) and first US fintech (Stash); a March-2026 cluster of newly-designated Section 16 officers.

APPENDIX B — Source Appendix — Grab Holdings Limited (NASDAQ: GRAB)

Report date 2026-07-04. Primary sources (SEC filings, company disclosures) listed before third-party/aggregated data. Every material claim traces to a source below. IFRS 20-F figures are primary; third-party aggregated data is used only as a cross-check, reconciled to filings.

1. Company SEC Filings (primary) — CIK 0001855612

EDGAR filing index: https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001855612&type=&dateb=&owner=include&count=40

2. Company Investor Materials

• Grab IR site — earnings press releases, supplemental presentations, shareholder-letter framing. https://investors.grab.com
• Q1-2026 earnings-call transcript (cross-checked to the 6-K filing) — management commentary on fuel, Indonesia commission cap, FinServices operating leverage, AI, buyback, GrabMart, deposits/securitization, foodpanda Taiwan.

3. Quantitative Data Sources (public / third-party; cross-checked to filings)

• Aggregated fundamental data (ROIC.ai) — standardized income statement / balance sheet / cash flow (FY2020–FY2025), profitability/credit/liquidity ratios, enterprise value, valuation multiples, per-share data, earnings-call list + transcripts. Note: ROIC reclassifies net finance income into “operating income” (shows ~$222M op income / ~$268M NI); the audited IFRS 20-F shows $65M operating profit / $200M group profit / $268M to owners — IFRS used as primary.
• Own-history valuation percentiles (price/sales, price/earnings, price/book vs. the stock’s own ~10-year range): P/E ~15th percentile, P/B ~59th, P/S ~4.5th (near cheapest-ever), composite ~26th (as of 2026-07-02).
• Multi-factor risk model — factor loadings (Market +0.91, Singapore +0.57, Quality −0.14, Momentum/Value zeroed), risk-adjusted track record (5y −19.8% ann, max drawdown −86.5%; 3m +34.7% ann), idiosyncratic vol ~30%, nearest factor peer SE.
• 5-year daily price history — OHLCV, moving averages, beta/alpha for the price-action event map.

4. Peer Comparables (public filings)

• Sea Ltd (SE) — closest SEA super-app peer; comparable and moat framing.
• Uber (UBER) — global mobility/delivery unit economics.
• DoorDash (DASH) — delivery margin/advertising benchmarks.
• MercadoLibre (MELI) — LatAm platform + fintech analog.
• Coupang (CPNG) — Asian e-commerce/logistics comparable.

5. Industry / Regulatory Context

• Monetary Authority of Singapore (MAS) — digital-full-bank framework and capital requirements (GXS Bank).
• Bank Negara Malaysia (BNM) — digital-bank licensing (GXBank).
• Indonesia OJK / KPPU — digital-bank rules, foreign-ownership caps, gig-labor/commission-cap developments, GoTo-merger antitrust review.
• Trade press (Reuters, Bloomberg, Tech in Asia, DealStreetAsia) — GoTo merger status, foodpanda Taiwan, Stash Financial, Superbank, buyback and EGM coverage (used for event dating; validated against primary filings).

Where any third-party figure and a Grab filing disagree on a material number, the filing governs and the discrepancy is noted in the text.