Stock Analysis · Bulls vs Bears

Grab Holdings Limited GRAB

🐂 Bulls vs Bears 🐻Report #212 Oct 2026
★ Willem owns this stock · 2.1% of the portfolio
34
moderate risk▼ −12 vs last
🐂 9vs11 🐻
Bears lead
60%
▲ Chance of the biggest tailwind the forecasts miss
35%
▼ Chance of the biggest headwind the forecasts miss
The verdict

The marketplace keeps compounding. Adjusted EBITDA has grown 18 quarters in a row, food share is rising to ~55%, and the stock trades at ~3.5x sales versus 6–9x in prior years. But the accounting profit behind it is thin: GAAP operating profit was $19M in Q2, incentives rose faster than revenue after management called Q1 the peak, and the cash cushion is largely committed to Atome, Taiwan, the buyback and a 2028…

🎙️ Bulls vs Bears podcast — Grab Holdings Limited
Alex and Sam debate every bull and bear argument · 15:13
0:0015:13
⤓ Download

Written by Claude (Anthropic’s AI) from public sources · Willem owns GRAB · AI can make mistakes — numbers in brackets point to the sources · Not financial advice

🐂 Bull arguments 9

  1. 1Adjusted EBITDA up 54% on 22% revenue growthStrong
  2. 2Priced at half its historic sales multiple with ~$5.4B net cashMedium
  3. 3The founder's first open-market buy: $29.9M at $2.89Medium
  4. 4A ~$900M buyback could retire ~7% of shares at today's priceMedium
  5. 5The loan book is small, short-dated and so far cleanMedium
  6. 6Only ~1% of users borrow; Atome takes loans past $6BMedium
  7. 7Food share keeps rising: ~55% regionally, 46% in IndonesiaStrong
  8. 8The 8% cap hurts GoTo more than GrabMedium
  9. 9Grab passes regulatory costs through and wins legal fightsMedium

🐻 Bear arguments 11

  1. 1Indonesia's decree reaches cars, food and fares; rules due 31 DecHigh
  2. 2Accounting profit is thin under the adjusted EBITDAMedium
  3. 3The "incentives peaked in Q1" promise was broken in Q2Medium
  4. 4Credit charges are outrunning revenue, and no NPL ratio is publishedMedium
  5. 5The 2028 targets now lean on an unaudited acquisitionMedium
  6. 6Free cash is ~$2.9B against ~$3.0B of commitments plus a $1.5B putMedium
  7. 7The founder controls ~75% of votes with a small economic stakeMedium
  8. 8Uber is a 13% holder and a likely returning rivalMedium
  9. 9Rides grow 28% but mobility revenue only 12%Medium
  10. 10Vietnam: lost the ride lead, now under competition reviewLow
  11. 11Consensus already sits above guidance, with 25 of 25 analysts at BuyLow

Bull arguments show how well they survive the devil’s advocate (strong / medium / weak); bear arguments show the damage they could do after the counter-argument (high / medium / low).

Bull case

🐂 Bull #1Adjusted EBITDA up 54% on 22% revenue growth

  • The point: the marketplace has crossed into operating leverage. Each extra dollar of revenue now brings more than proportionate adjusted profit.
  • Evidence:
    • Q2-26 results:
      • Revenue was $997M (+22%) and adjusted EBITDA $168M (+54%), the 18th straight quarter of EBITDA growth [1].
      • The adjusted EBITDA margin rose from 15.6% (Q3-25) to 16.9% (Q2-26) over four quarters [1,7].
    • Segment margins:
      • Deliveries segment EBITDA was 2.3% of GMV ($96M on $4,249M), versus 1.8% a year earlier [1].
      • The CFO's steady-state target is "4% plus" for deliveries and "9% plus" for mobility [11].
    • Guidance record:
      • FY25 adjusted EBITDA came in at $500M, the top of the final $490–500M range and above the original $440–470M [3,6,7].
      • FY26 guidance was raised on 4 Aug to $720–740M (+44–48%) [1].
  • Why it matters: if deliveries goes from 2.3% to 4% of GMV on ~$20B of annual deliveries GMV, that is roughly $340M of extra EBITDA (estimate: $4.25B × 4 quarters × 1.7 points). That is the core of the $1.7B 2028 target [4].
  • 😈 Devil's advocate:
    • Part of the FY26 raise is bought, not earned. It includes the consolidation of Superbank (from June) and Stash (from July), and the COO credited "fuel price support" [1,9].
    • Revenue is net of incentives, which grew faster than revenue in Q2 (Bear #3).
    • The 2.3% deliveries margin is still ~1.7 points from target, and management stopped repeating the 4% figure on the Q1 and Q2 calls [9,10].
  • Verdict: Strong The leverage is visible in both core segments, but the FY26 raise flatters it.

🐂 Bull #2Priced at half its historic sales multiple with ~$5.4B net cash

  • The point: the stock has de-rated far more than the business has slowed.
  • Evidence:
    • Price and value:
      • $3.19 on 9 Oct, −49.8% over 52 weeks [16].
      • Market value ~$13.0B; EV ~$8.5B using the company's cash and debt definitions [16].
    • Against its own history:
      • Price/sales is 3.49x, versus 6.05x (FY25), 6.80x (FY24) and 8.65x (FY22) at year-ends [17].
      • EV/sales is 2.28x, versus 4.59x (FY25) [17].
    • Against peers:
      • EV/sales: Sea 1.92x, Uber 2.73x, DoorDash 5.23x [16].
      • Forward P/E: Grab 18.3x, Uber 17.7x, Sea 22.4x [16].
    • On the guided FY26 adjusted EBITDA midpoint of $730M, EV is ~11.7x (my arithmetic: $8.51B ÷ $0.73B).
  • Why it matters: if the 2028 adjusted EBITDA target of $1.7B is only half delivered, today's EV is under 8x 2028 EBITDA (estimate: $8.5B ÷ ~$1.1–1.2B).
  • 😈 Devil's advocate:
    • "Net cash" includes ~$2.5B of bank-customer deposits (Bear #6). On a deposit-adjusted basis EV is ~$11B and ~15x FY26 adjusted EBITDA (my arithmetic).
    • Uber trades at a similar forward P/E with a far larger profit base.
    • The stock's old multiples were bubble-era and are no anchor.
  • Verdict: Medium Cheap against its own history, but only fairly priced against peers once deposits are counted as liabilities.

🐂 Bull #3The founder's first open-market buy: $29.9M at $2.89

  • The point: Anthony Tan bought with his own money at the low, and the monthly selling has paused.
  • Evidence:
    • The buy:
      • On 21 Sep 2026 Tan bought 10,350,000 Class A shares at $2.8866 (~$29.88M), coded "P" (open market) [12].
      • President Alex Hungate bought 299,571 shares at $2.8936 (~$0.87M) [12].
      • These were the only open-market buys in 66 insider filings from April to 7 Oct [12].
    • The selling:
      • Tan's 10b5-1 plan sold 400,000 shares a month from April to August, at $3.51–3.91 (~$7.4M in total) [12].
      • No Tan sale appears for September or October [12].
  • Why it matters: the buy is ~4x the plan sales of the previous five months combined, and it came three days after the low ($2.74 intraday on 18 Sep) [18]. Tan said he "put my money where my mouth is" [18].
  • 😈 Devil's advocate:
    • Other officers kept selling under plans after the buy: CFO Oey 50,000 on 15 Sep at $2.90, Ong 38,000 on 5 Oct at $3.09, Kandal 30,000 on 15 Sep [12].
    • Tan already controls the company through super-voting shares, so $30M is a signal rather than a meaningful change in his exposure.
    • There have been no further buys since 21 Sep.
  • Verdict: Medium It is a real, unplanned buy at the low, but it is a single event.

🐂 Bull #4A ~$900M buyback could retire ~7% of shares at today's price

  • The point: Grab is shrinking its share count while stock pay runs at ~6% of revenue.
  • Evidence:
    • What has been done:
      • $351M was repurchased under the Feb-26 programme: a $250M accelerated repurchase plus a $101M forward purchase. That retired or put in treasury 68.3M shares in H1 [2].
    • What is planned:
      • A new $750M authorisation on 4 Aug plus $149M unused from February gives ~$899M [1,2].
      • On 15 Sep Grab said it intends to complete it within 12 months [5].
      • CFO: "dislocation between that performance and where our shares trade remains an opportunity we intend to act on" [5].
    • At $3.19, $899M buys ~282M shares, ~6.9% of 4,079M (my arithmetic).
  • Why it matters: total shares were 4,079M at Jun-26 versus 4,076M a year earlier, so buybacks have so far only neutralised stock pay [1]. Doing ~$900M at ~$3 would finally cut the count.
  • 😈 Devil's advocate:
    • Since 4 Aug no open-market repurchases have been disclosed; Grab doesn't publish monthly detail [2].
    • "Within 12 months" is an intention, not a commitment in the authorisation itself [2,5].
    • Stock pay was $240M in the last 12 months, so about a quarter of the buyback just offsets new grants [2].
  • Verdict: Medium The authorisation is real, but execution since August is unproven.

🐂 Bull #5The loan book is small, short-dated and so far clean

  • The point: the numbers that exist don't show a credit problem yet.
  • Evidence:
    • Ageing table, 20-F FY25 [3]:
      • Loans 91+ days past due were $20M, 1.6% of $1,278M gross loans at Dec-25, versus 1.9% at Dec-24 (my arithmetic).
      • Loans 1–90 days past due fell from 10.2% to 8.9% of gross.
      • Write-offs fell from 6.7% to 5.4% of year-end gross loans.
    • Bank subsidiaries (secondary, not checked against their filings):
      • Superbank gross NPL was 1.6% at Jun-26, versus 2.7% a year earlier [21,22].
      • GXS group's expected-credit-loss ratio fell from 6.8% to 4.6% in FY25 [23].
    • Context: driver loans are repaid from daily payouts, and Superbank's book carries credit insurance, which is why its expected losses are lower [2].
  • Why it matters: this is the first evidence on credit quality in numbers. Report #1 found only charts.
  • 😈 Devil's advocate:
    • Most of the book is under two years old, so loss ratios on a fast-growing book look better than they will once growth slows.
    • Superbank wrote off ~Rp217B in H1-26, close to 75% of its full-2025 write-offs, which flatters its NPL ratio (secondary) [21].
    • Impairment charges as a share of revenue rose to 7.2% in Q2 (Bear #4).
  • Verdict: Medium The data is better than feared, but it covers a young book and is a year old at group level.

🐂 Bull #6Only ~1% of users borrow; Atome takes loans past $6B

  • The point: Grab lends using earnings and spending data no bank has, and has barely started.
  • Evidence:
    • Gross loans were $2,317M at Jun-26, up 197% (~100% excluding Superbank). Management expects more than $3B by end-2026 [1,2,9].
    • Financial-services revenue was $134M in Q2 (+59%) [1].
    • Atome adds a ~$1B book, 25M cumulative users and 30,000+ brands in five markets (management accounts, unaudited) [4].
    • 2028 targets (including Atome): loans above $6B and financial-services EBITDA of $500M [4].
    • The yardstick: Sea's Monee has $11.1B of loans at a 1.0% 90-day NPL with adjusted EBITDA of $288M a quarter [42].
  • Why it matters: Monee shows what the model earns at scale. $500M would be ~68% of the FY26 group EBITDA guide.
  • 😈 Devil's advocate:
    • The segment still lost $15M in Q2, and the "break-even in 2H26" promise has been repeated for three quarters [1,9,10,11].
    • The $500M depends on an acquisition that closes only by Q3 2027 [4].
    • Sea is five times bigger in lending and spending 140% more on marketing to grow it [42].
  • Verdict: Medium The runway is real; the profit isn't yet.

🐂 Bull #7Food share keeps rising: ~55% regionally, 46% in Indonesia

  • The point: in Grab's largest profit pool, the leader is pulling further ahead.
  • Evidence:
    • Regional food delivery was $22.7B of GMV in 2025 (+18%). Grab has ~55%, up from 53.8% in 2024. ShopeeFood is ~14% and foodpanda shrinking at ~$2.6B (Momentum Works) [40].
    • Indonesia food delivery is ~$6.4B: Grab 46%, Gojek 31%, ShopeeFood 23% [41].
    • Monthly transacting users hit a record 53.9M (+17%) [1].
  • Why it matters: density lowers courier cost per order, which is how deliveries margins move toward 4%.
  • 😈 Devil's advocate:
    • ShopeeFood gained share faster than Grab (11.9%→14%) [40].
    • The share gains came with incentives rising to 10.9% of on-demand GMV [1].
    • Momentum Works is a single industry source, with no audited share data.
  • Verdict: Strong The direction is consistent across sources and markets.

🐂 Bull #8The 8% cap hurts GoTo more than Grab

  • The point: the motorbike cap lands on GoTo's core product and Grab's smallest one.
  • Evidence:
    • GoTo:
      • On 29 Jul GoTo cut its FY26 on-demand EBITDA guide from IDR 1.7–1.8tn to IDR 1.4–1.5tn because of the cap (secondary summary) [30].
      • Its CEO said the cap "will impact future profitability" [29].
    • Grab:
      • Indonesian motorbike rides are ~6% of Grab's mobility GMV and stay EBITDA-positive after the cap, according to management [9].
      • Indonesia was 21.2% of FY25 revenue and the slowest-growing market (+11%) [3].
    • Consolidation: a DBS note (May, via secondary) argued the cap could ease approval of a Grab–GoTo holding-company merger [53]. Danantara holds under 1% of GoTo and "wants to profit" [31].
  • Why it matters: a weaker GoTo means less incentive spending in Indonesia and keeps a merger possible.
  • 😈 Devil's advocate:
    • There has been no Grab–GoTo news from June to October [31].
    • Indonesia's competition regulator KPPU warned in Dec 2025 that it would cancel a harmful deal.
    • If the implementing rules extend to cars, Grab, which is stronger in four-wheel, loses more (Bear #1).
  • Verdict: Medium True for the motorbike rule; the car rule decides it.

🐂 Bull #9Grab passes regulatory costs through and wins legal fights

  • The point: the record so far is that regulation raises prices for users more than it cuts Grab's margin.
  • Evidence:
    • Fee pass-through: Singapore's platform-worker pension (CPF) rules took effect in 2025, and Grab raised its ride platform fee from S$0.90 to S$1.20 from 1 Jan 2026 [45].
    • Malaysia: the competition regulator MyCC's proposed MYR 86.8M fine was quashed in the High Court. The Court of Appeal dismissed MyCC's appeal and the Federal Court denied leave [3].
    • Philippines: fines from the competition regulator total ~US$1.2M [3].
    • Mobility margin: Q2 mobility was 8.6% of GMV, inside the 8.5–9% band, despite fuel support [1,9].
  • Why it matters: if pass-through keeps working, rules shift costs to consumers and Grab's EBITDA path survives.
  • 😈 Devil's advocate:
    • Indonesia's cap is a ceiling on Grab's share of the fare, which is harder to pass through than a fixed fee.
    • Singapore's CPF rate rises to as much as 17% by 2029; Momentum Works estimates ≥S$493M extra cost for the industry over five years [44].
  • Verdict: Medium

Dropped since report #1:

  • "Hybrid autonomous network": Tan himself says autonomous vehicles are uneconomic across most of the region, because over half of trips are two-wheel at under $1 [9]. Waymo launches in Singapore on its own in 2028 [46].
  • "Competition consolidating": merged into Bull #7 and Bull #8.

Bear case

🐻 Bear #1Indonesia's decree reaches cars, food and fares; rules due 31 Dec

  • The point: the decree that capped motorbike commissions now gives ministries power over car rides, couriers, food-delivery fees and driver welfare. The details are written by 31 Dec 2026.
  • Evidence:
    • The decree: Perpres 27/2026, "Protection of the digital-platform transport ecosystem", was released in full on 8 Oct [24,25,26]. Grab's guidance assumes the commission rule applies to motorbike rides ("OJOL") only [9].
    • Who sets what:
      • The digital ministry (Komdigi) may set service fees and fares for food and goods delivery.
      • The Transport Ministry sets motorbike fares.
      • The MSME Ministry supervises driver–platform agreements [25,26].
    • What platforms must provide:
      • Social security (health, work-accident and death cover) "per prevailing law". Who pays is not found [27].
      • Agreements must cover fee structure, incentives, data access and algorithm use [25].
      • Implementing rules are due by 31 Dec. After that, platforms get six months to comply [25].
    • Cars: several outlets say 8% applies only to two-wheel passenger rides so far, and no car scheme has been set [27].
    • Size of a car extension:
      • Citi estimates a US$35–40M EBITDA hit if cars are included, versus US$5–10M for motorbikes only (secondary) [47].
      • DBS's earlier scenario was revenue −5% and EBITDA −14% (May; not re-verified).
      • Morgan Stanley puts Indonesia at ~20% of group EBITDA (secondary) [48].
  • Why it matters: this is the one rule that can move Grab's take rate in its third-largest revenue market by decree, inside the next three months.
  • 🛡️ Counter-argument:
    • The decree sets no commission number for cars or food. Grab and Gojek both say they will wait for the ministerial rules [24,28].
    • The 2025 draft's 10% cap ended as a narrower 8% motorbike rule, which suggests lobbying works.
  • Severity: High The scope is genuinely open until 31 Dec, and Citi's and DBS's estimates differ by ~3x.

🐻 Bear #2Accounting profit is thin under the adjusted EBITDA

  • The point: the profit that pays shareholders is a fraction of the headline.
  • Evidence:
    • Operating profit:
      • Operating profit was $27M, $52M, $22M and $19M over the last four quarters, a 1.9–5.7% margin [1,6,7].
      • In H1-26 it was $41M, against $323M of adjusted EBITDA [2].
    • Stock pay: $240M over the last 12 months, ~39% of adjusted EBITDA (my arithmetic from [2,3]).
    • One-off gains:
      • Q2 net profit of $235M includes a $307M non-cash gain from consolidating Superbank [2].
      • Q1 profit included a $118M fair-value gain, partly from the convertible's embedded derivative [10].
    • Free-cash-flow definition:
      • Adjusted FCF was redefined in Q2 to also exclude "treasury liquidity" flows. That removed a $95M benefit from operating cash flow [1].
      • Simple FCF (operating cash flow minus capex) was −$168M, −$6M, −$89M and +$11M over the last four quarters. Loan growth is the gap to adjusted FCF of $203M, $76M, $98M and $73M [1,6,7].
  • Why it matters: valuation on adjusted EBITDA treats stock pay and loan funding as free. Neither is.
  • 🛡️ Counter-argument:
    • Loan growth is a balance-sheet investment funded partly by deposits, so excluding it from FCF is standard for lenders.
    • Stock pay has fallen in dollars, from $304M (FY23) to $241M (FY25) [3].
    • The auditor is KPMG, with effective controls over financial reporting at Dec-25 and no restatements [3].
  • Severity: Medium Nothing is hidden, but GAAP earnings power is a fifth of the adjusted figure.

🐻 Bear #3The "incentives peaked in Q1" promise was broken in Q2

  • The point: management guided that Q1 would be the incentive peak, and Q2 spending rose further.
  • Evidence:
    • The promise: on the Q1 call (May) the COO said "We expect this first quarter to be a peak in the driver incentives" [10].
    • Q2 outcome:
      • Partner incentives were $317M (+33%) and consumer incentives $389M (+27%), $706M in total versus ~$650M in Q1. That is +29% year on year against revenue +22% [1,10].
      • On-demand incentives rose to 10.9% of GMV from 10.5% in Q1 [1,10].
    • The explanation: the CFO said on the Q2 call that Grab was "leaning over and above that around partner incentives" because of fuel costs. Grab "leaned in quite hard" in the Philippines, Thailand, Singapore and Vietnam [9].
  • Why it matters: incentives are deducted from revenue, so persistent incentive inflation caps both the margin path in Bull #1 and the credibility of guidance.
  • 🛡️ Counter-argument: fuel is a temporary, external shock, and driver support is targeted. Mobility margin still held at 8.6% of GMV [1,9].
  • Severity: Medium It's a broken specific promise, not yet a trend; Q3 will show which.

🐻 Bear #4Credit charges are outrunning revenue, and no NPL ratio is published

  • The point: the profit-and-loss cost of lending is accelerating faster than the business, while the key ratio stays undisclosed.
  • Evidence:
    • Impairment charges:
      • As a share of revenue: 4.2% (Q3-25), 4.1% (Q4-25), 5.0% (Q1-26), 7.2% (Q2-26) (my arithmetic from [1,2,7]).
      • Net impairment was $72M in Q2-26 versus $33M (+118%), "mainly driven by Digibank expected credit losses" [1].
    • Allowance coverage: fell from 8.5% (Dec-24) to 7.7% (Dec-25) to 6.6% (Jun-26). Management attributes the fall to Superbank's insured book [2,3].
    • Disclosure: quarterly releases say only that 90-day bad loans are "within our risk appetite". Neither the interim accounts nor the 20-F gives an NPL ratio or credit-loss stages [1,2,3].
  • Why it matters: fintech is supposed to supply $500M of the 2028 EBITDA. A loss spike would land just as the book doubles again.
  • 🛡️ Counter-argument:
    • Loans grew faster than impairments: +197% versus +118% in Q2, and +118% versus +104% for the loan-only charge in FY25 [1,3].
    • Bull #5's ageing data shows no deterioration through Dec-25.
  • Severity: Medium High. The ratios are fine on a fast-growing book, but the disclosure doesn't let outsiders check them in time.

🐻 Bear #5The 2028 targets now lean on an unaudited acquisition

  • The point: the target upgrade came with a deal, not with organic outperformance.
  • Evidence:
    • The target change (15 Sep):
      • 2028 adjusted EBITDA went from $1.5B to $1.7B.
      • Revenue CAGR went from ~20% to 30%+ for 2025–28.
      • Both now "including Atome Financial" [4,6].
    • Atome's numbers: gross loans of ~$1B are management accounts, unaudited. Revenue, profit and NPL are not disclosed [4].
    • Price and timing:
      • Phase 1 is $1.49B for 60% (incl. $0.26B new capital), closing by Q3 2027 [4].
      • Phase 2 values 100% at $2.0–4.5B on a formula of 13x adjusted EBITDA and 2.5x revenue. That implies ~$0.8–1.8B for the last 40%, at least half in cash (my reading; the release doesn't say explicitly whether the band is for 100%) [4].
    • The yardstick: the original adjusted-FCF target (>$1.2B, 80% conversion) was not restated [6,11].
    • Market reaction: the stock fell 7.5% to a three-year low on the announcement [18,19].
  • Why it matters: comparing old and new targets now mixes organic and bought growth. The ~$2.4–3.3B total price comes before any audited Atome numbers.
  • 🛡️ Counter-argument:
    • Phase 2 pays on delivered EBITDA, so a weak Atome costs less.
    • Grab says the deal is accretive to adjusted EBITDA on completion [4].
    • Analysts' maths implies ~12x FY28 EV/EBITDA (Maybank, snippet) [53].
  • Severity: Medium High.

🐻 Bear #6Free cash is ~$2.9B against ~$3.0B of commitments plus a $1.5B put

  • The point: the $5.4B net-cash headline counts money that isn't free.
  • Evidence:
    • Liquidity at Jun-26: gross cash liquidity ~$7.4B, loans and borrowings $2.03B, bank-customer deposits $2.51B [1,2].
      • Net cash liquidity does not subtract deposits [2].
      • Of the total, $205M is legally restricted by jurisdiction [2].
    • Commitments:
      • Atome Phase 1: $1.49B.
      • Buyback: ~$0.9B.
      • foodpanda Taiwan: $0.6B.
      • Stash: deferred payments for the remaining 49.9%, amount not disclosed [2,4,5].
    • The convertible: holders of the $1.5B zero-coupon notes can put them at par on 15 Jun 2028; the notes mature in 2030. Conversion is ~$6.55, about 2x today's price [2,3].
    • The 20-F risk factor says Grab "may not have enough available cash" at a put [3].
  • Why it matters: cash left after deposits ($2.9B, my arithmetic) roughly equals announced commitments ($3.0B). The 2028 put and Atome Phase 2 must come from future FCF (~$450M a year, Q2 trailing 12 months) or new debt.
  • 🛡️ Counter-argument:
    • Deposits are matched by liquid bank assets, not spent.
    • Atome Phase 1 is a year away, giving ~$0.5B of FCF before then.
    • The put can be refinanced, and Grab has only $371M of other bank debt [2].
  • Severity: Medium Not a solvency issue, but buyback and M&A capacity is far smaller than the headline.

🐻 Bear #7The founder controls ~75% of votes with a small economic stake

  • The point: outside shareholders can't block deals like Atome.
  • Evidence:
    • On 24 Mar 2026 shareholders approved doubling Class B votes from 45 to 90 per share (85.9% of votes cast). The stated reason is to keep GXS Bank under Singaporean control, as the Monetary Authority of Singapore (MAS) requires [3,58].
    • Tan's voting control rises from ~59% to up to 74.9% [3].
    • The board has six members after Uber CEO Khosrowshahi left on 6 Jul [13].
  • Why it matters: it raises the cost of Bear #5. Capital allocation is one person's call.
  • 🛡️ Counter-argument: the change has a regulatory reason, and Tan has just bought $30M of stock (Bull #3).
  • Severity: Medium

🐻 Bear #8Uber is a 13% holder and a likely returning rival

  • The point: Uber is buying Delivery Hero, which owns foodpanda in Grab's markets, while it holds a Grab stake that bars it from competing.
  • Evidence:
    • Uber's Grab stake:
      • Uber owns 535,902,982 Grab Class A shares, 13.1% of total [3,13].
      • Its 13D/A (8 Jul) says it is "not currently engaging in discussions" on control or strategy [13].
      • Uber marked the stake down to $2.02B at Jun-26 [39].
    • Delivery Hero bid:
      • €41.50 a share, US$14.8B [38].
      • Acceptance runs to 5 Nov, with closing expected 2H 2027 [38,39].
      • foodpanda's SE Asian units are not in the antitrust carve-out [39].
    • Non-compete: Uber may not compete in Grab's core markets until one year after a full sale of its stake (Tan, Q2 call) [9].
    • Downside scenarios (Maybank, 20 Jul, snippet): ~26% downside if Uber exits its stake to re-enter the region's mobility market, ~5% if it only runs foodpanda [52].
    • Taiwan: the FTC is reviewing whether Uber's stake softens competition in the foodpanda Taiwan deal; ruling due 27 Oct [35,36].
  • Why it matters: a 536M-share block (~10x average daily volume over 2–3 weeks; estimate) is an overhang, and a re-entry would restart incentive wars.
  • 🛡️ Counter-argument:
    • Selling at three-year lows is unattractive for Uber.
    • The non-compete clock only starts after a full sale, so re-entry is >1 year away in any scenario.
  • Severity: Medium

🐻 Bear #9Rides grow 28% but mobility revenue only 12%

  • The point: volume growth comes from cheaper "saver" rides and does not turn into equal revenue.
  • Evidence:
    • Q2 rides +28%, mobility GMV +18%, mobility revenue +12% [9].
    • Segment EBITDA margin was 8.6% of GMV, versus 8.7% a year ago, and below the "9% plus" steady state [1,11].
    • Driver earnings grew only +4%, while monthly active drivers rose +19% [9].
  • Why it matters: mobility is Grab's most profitable segment. Falling revenue per ride limits operating leverage there.
  • 🛡️ Counter-argument: cheaper tiers are a deliberate push into the market, and the margin is still within the guided band [9].
  • Severity: Medium

🐻 Bear #10Vietnam: lost the ride lead, now under competition review

  • The point: Grab's fourth-largest mobility market has a strong local rival and an active regulator.
  • Evidence:
    • Market share: Xanh SM overtook Grab in Vietnamese ride-hailing GMV in 2025 (Mordor data) [34].
    • Regulator: the National Competition Commission met Grab on 8 Sep and requested fare, commission and deduction data. It said it "may" investigate [32].
    • Driver claims: drivers claim deductions of 40–50% on some trips (claim, not verified) [33].
    • Commercial response: Grab raised platform fees by VND 3,000–19,000 in 2026 (secondary) [33].
    • Scale: Vietnam was 7.6% of FY25 revenue [3].
  • Why it matters: it's the clearest case of a market where Grab is not the scale leader in rides, plus a regulator that is copying Indonesia's theme.
  • 🛡️ Counter-argument: no investigation or decree has followed since 11 Sep [32]. Other surveys still show Grab as the most-used app.
  • Severity: Low Medium.

🐻 Bear #11Consensus already sits above guidance, with 25 of 25 analysts at Buy

  • The point: expectations are above the company's own numbers, and nobody is neutral.
  • Evidence:
    • Revenue: FY26 consensus is $4.20B (12 analysts) versus the guide's top of $4.15B. FY27 consensus is $5.24B (+24.75%) [15].
    • Ratings: 20 Strong Buy and 5 Buy, with no Hold or Sell [15].
    • Targets were cut, not ratings: BofA $4.50 on 9 Oct, DBS $5.00, Barclays $5.00, Maybank $5.40 [15].
  • Why it matters: a guide-in-line Q3 could still disappoint, and there are no downgraders left to upgrade.
  • 🛡️ Counter-argument:
    • The average target of $5.76 is ~80% above the price, so the price itself already embeds scepticism [15].
    • Short interest is modest at ~7–9% of float [20].
  • Severity: Low

Tailwinds NOT in the forecasts

Base for all sizing (estimates):

  • Consensus: FY2027 revenue $5.24B and non-GAAP EPS $0.14 [15].

  • Shares and profit: ~4.17B diluted shares [1], which implies ~$584M of adjusted net income.

  • Rates: 15% tax, and 3.5% yield on cash.

  • Missing data: consensus EBITDA and FY2028 figures were not found.

  • foodpanda Taiwan approved and consolidated in FY2027 — 60% chance by end-2026

    • Impact on revenue: not quantifiable. foodpanda Taiwan's revenue is not disclosed.
    • Impact on EPS: +$0.005 per share, +3% vs FY2027 consensus $0.14.
    • How we got the number (estimate): (~$50M EBITDA × 0.85 − $600M × 3.5% lost interest) ÷ 4.17B ≈ +$0.005. The ~$50M EBITDA is a FY27 step toward the ≥$60M 2028 figure from report #1.
    • Expected value: 60% × $0.005 ≈ +$0.003.
    • What it is: the FTC clears the deal (possibly with conditions on Uber's influence), and Taiwan adds an EBITDA-positive market in 2027.
    • Why it's not priced in: Grab's FY26 guide gives no Taiwan contribution and the deal is pending [9]. Whether consensus includes it was not found.
    • Trigger to watch: the FTC decision by 27 Oct [35].
    • Why this probability: the FTC said no law bars foreign shareholdings and a conditional approval is the usual path. But legislators and riders oppose the deal, and the FTC blocked Uber–foodpanda in 2024 [36,37].

  • Buyback completed at ~$3.2, retiring ~7% of shares — 65% chance of ≥$800M executed by Sep-2027

    • Impact on revenue: none.
    • Impact on EPS: +$0.003 per share, +2% vs FY2027 consensus.
    • How we got the number (estimate): ($584M − $900M × 3.5% × 0.85) ÷ (4.17B − 0.28B) = $0.143 vs $0.140 ≈ +$0.003.
    • Expected value: 65% × $0.003 ≈ +$0.002.
    • What it is: the remaining ~$899M authorisation is spent near today's price.
    • Why it's not priced in: consensus is typically built on the current share count. No open-market repurchases since August have been disclosed [2]. This is partly priced: the intention is public.
    • Trigger to watch: the share count and repurchase line in the Q3 release.
    • Why this probability: Grab stated the 12-month intention [5], but it also has to fund Atome and Taiwan and keep ammunition for the 2028 put.

  • Credit holds and financial services beats its EBITDA path — 25% chance in FY2027

    • Impact on revenue: +$80M, +1.5% vs FY2027 consensus.
    • Impact on EPS: +$0.010 per share, +7%.
    • How we got the number (estimate): +$50M segment EBITDA (faster Superbank cost/income below 50%, stable losses) × 0.85 ÷ 4.17B ≈ +$0.010.
    • Expected value: 25% × $0.010 ≈ +$0.0025.
    • What it is: loans exceed $3B at end-2026 with impairments growing slower than revenue, and FS turns solidly positive in 2027.
    • Why it's not priced in: consensus treats FS as roughly break-even moving up. Management's numbers jump to $500M only in 2028, with Atome [4].
    • Trigger to watch: FS segment EBITDA ≥$0 in Q3/Q4-26, and impairment below 5% of revenue.
    • Why this probability: Superbank's guided cost/income ratio fall is concrete [9], but Q2's 7.2% impairment ratio points the other way.

  • Grab–GoTo combination — 10% chance within 36 months

    • Impact on revenue: +$1.3B, about +20% vs an estimated FY2028 consensus of ~$6.3B. That figure is my extrapolation, as FY28 consensus was not found.
    • Impact on EPS: +$0.013 per share, +9% vs the FY2027 consensus base (used as a proxy).
    • How we got the number (estimate): (GoTo EBITDA ~IDR 3.3tn ≈ $185M + $150M incentive savings) × 0.8 = $268M, added to $584M, ÷ (4.17B + ~1.4B shares issued in a swap) ≈ $0.153 vs $0.140.
    • Expected value: 10% × $0.013 ≈ +$0.0013.
    • What it is: a holding-company or share-swap merger, possibly with a Danantara golden share.
    • Why it's not priced in: nothing has been announced since the January "snag". There has been no news June–October [31].
    • Trigger to watch: Danantara raising its GoTo stake; Telkomsel's position; KPPU comments.
    • Why this probability: the government backed it in Nov 2025 and the cap squeezes GoTo, but talks have failed since 2020, and KPPU plus a ~90% combined share are big obstacles.

  • Uber sells foodpanda SE Asia to Grab — 8% chance within 36 months

    • Impact on revenue: +$300M, ~+5% vs the estimated FY2028 base.
    • Impact on EPS: +$0.009 per share, +6%.
    • How we got the number (estimate): ($30M acquired EBITDA + $50M incentive savings) × 0.85 − ($1B price × 3.5%) = ~$33M ÷ 4.17B ≈ +$0.008–0.009.
    • Expected value: 8% × $0.009 ≈ +$0.0007.
    • What it is: to keep its non-compete and its Grab stake intact, Uber sells foodpanda's SE Asian units after closing Delivery Hero. Maybank values them below $1B (snippet) [52].
    • Why it's not priced in: Uber–Delivery Hero closes in 2H 2027, and there is no public SE Asia plan [39].
    • Trigger to watch: Uber's disclosures after the 5 Nov acceptance deadline.
    • Why this probability: Singapore's CCCS and Malaysia's MyCC would resist Grab adding share. Uber may prefer to keep foodpanda.

Headwinds NOT in the forecasts

  • Incentives stay elevated through 2027 — 35% chance

    • Impact on revenue: −$62M, −1.2% vs FY2027 consensus $5.24B.
    • Impact on EPS: −$0.013 per share, −9% vs FY2027 consensus $0.14.
    • How we got the number (estimate): +0.2 points of incentives × ~$31B FY27 on-demand GMV ($6.5B × 4 × 1.2) = $62M × 0.85 ÷ 4.17B ≈ −$0.013.
    • Expected value: 35% × −$0.013 ≈ −$0.0045.
    • What it is: on-demand incentives stay at or above 10.9% of GMV instead of falling back. Causes would be fuel, ShopeeFood's push, or Vietnam and Philippine competition.
    • Why it's not priced in: consensus FY27 revenue growth (+24.75%) is above FY26's guided +22–23%, which implies easing incentives [1,15].
    • Trigger to watch: incentives as a share of on-demand GMV in Q3 (Q1 10.5%, Q2 10.9%).
    • Why this probability: the Q1 "peak" call already failed once [10]. Fuel was "still elevated" going into Q3, per the CFO [9].

  • Credit losses jump as the book doubles — 25% chance in FY2027

    • Impact on revenue: −$40M, −0.8%, from slower loan growth.
    • Impact on EPS: −$0.014 per share, −10%.
    • How we got the number (estimate): +2 points of credit cost on a ~$3.5B average FY27 book (ex-Atome) = −$70M × 0.85 ÷ 4.17B ≈ −$0.014.
    • Expected value: 25% × −$0.014 ≈ −$0.0035.
    • What it is: losses on consumer and Indonesian loans rise as the book seasons and the rupiah stays weak, around 17,900 per dollar [49].
    • Why it's not priced in: guidance gives no credit-cost assumption. Coverage fell to 6.6% [2].
    • Trigger to watch: impairment above 7% of revenue for a second quarter. Any NPL disclosure, or its continued absence.
    • Why this probability: a young book that tripled in a year is the textbook setup, but early ageing data is clean (Bull #5).

  • Indonesia's rules extend to cars, food fees or platform-paid social security — 30% chance by mid-2027

    • Impact on revenue: −$90M, −1.7% vs FY2027 consensus.
    • Impact on EPS: −$0.008 per share, −6%.
    • How we got the number (estimate): Citi's −$35–40M EBITDA for cars [47] × 0.85 ÷ 4.17B ≈ −$0.008. Revenue: ~$1.3B of Indonesian four-wheel GMV × ~7 points lower take ≈ −$90M. That uses Morgan Stanley's 17–19% Indonesian share of mobility GMV [48], minus the ~6% from motorbikes [9].
    • Expected value: 30% × −$0.008 ≈ −$0.0024.
    • What it is: ministerial rules due 31 Dec cap car commissions, set food/courier fees, or make platforms pay drivers' social-security premiums.
    • Why it's not priced in: Grab's guide explicitly assumes motorbikes only [9].
    • Trigger to watch: Komdigi and Transport Ministry drafts in November–December, and driver protests.
    • Why this probability: the decree names cars, couriers and food, and gives Komdigi fee powers [25,26]. But it set no number for them, and early reporting says the 8% is two-wheel only [27]. DBS's harsher scenario (EBITDA −14%) would make this ~2.5x larger.

  • Vietnam caps commissions or fines Grab — 20% chance within 24 months

    • Impact on revenue: −$60M, −1.1% vs FY2027 consensus.
    • Impact on EPS: −$0.006 per share, −4%.
    • How we got the number (estimate): Vietnam ~7.6% × $5.24B ≈ $400M revenue. ~60% of it is mobility, and a quarter of that take is lost, giving −$60M. About half drops through after pass-through: −$30M × 0.85 ÷ 4.17B ≈ −$0.006.
    • Expected value: 20% × −$0.006 ≈ −$0.0012.
    • What it is: the NCC opens a formal case, or the government sets a commission ceiling copying Indonesia.
    • Why it's not priced in: this is a September 2026 event with no outcome yet [32].
    • Trigger to watch: an investigation notice from the Ministry of Industry and Trade; new driver boycotts.
    • Why this probability: the regulator "may" investigate, and the press is calling for an Indonesia-style cap [33]. But nothing has happened in a month.

  • Uber sells its 13% block — 20% chance within 36 months

    • Impact on revenue: none in the first year. Re-entry would come at least one year after a full sale.
    • Impact on EPS: none directly. It is a share-price overhang, not an earnings effect.
    • How we got the number: not quantifiable on EPS. Maybank's scenario is ~26% price downside if Uber exits to re-enter mobility (snippet) [52].
    • Expected value: ≈ $0 on EPS.
    • What it is: a block sale or exchange of ~536M shares, possibly to fund Delivery Hero.
    • Why it's not priced in: Uber's 13D/A shows no intention to sell [13].
    • Trigger to watch: a new Uber 13D/A or a block trade; Uber's Delivery Hero financing after 5 Nov.
    • Why this probability: Uber has debt-funded a $14.8B deal, but selling at three-year lows destroys value and keeps the non-compete running anyway.

Investment Risk Score: 34/100 (moderate)

# Factor Weight Measured value Score Why
1 Liquidity & funding runway 15% Gross liquidity ~$7.4B; debt $2.03B; deposits $2.51B → ~$2.9B "free" vs ~$3.0B announced commitments; TTM adj FCF $450M; simple FCF negative (loan growth) 25 FCF-positive with net cash, but nearly all free cash is spoken for, plus the $1.5B put in Jun-2028. Just outside the bottom band.
2 Leverage 10% Net cash ~$5.4B (company) or ~$2.9B after deposits; bank debt $371M secured on vehicles; no ratio covenants disclosed 15 Net cash on either definition.
3 Dilution (12 months + overhang) 10% Total shares 4,076M→4,079M (+0.1%); convertible 229M (5.6%, conv. $6.55); unvested RSUs 3.3%; Stash deferred consideration may be paid in shares; swap option 33.5M shares 30 Flat count, modest overhang mostly out of the money. Buyback execution would push this lower.
4 Profitability 10% TTM adj FCF margin ~12% ($450M/$3.73B); GAAP op margin 1.9–5.7%; SBC 6.4% of revenue 40 Inside the 5–15% FCF band, but at its weak end because GAAP profit is near break-even.
5 Revenue quality & concentration 10% No customer is material; transactional fees; Malaysia 31% / Singapore 22% / Indonesia 21% of FY25 revenue 28 Diversified and repeat-use, but not contracted and concentrated in three countries.
6 Execution & track record 10% FY25 EBITDA top of range, revenue $10M below the raised low end; FY26 raised (partly inorganic); incentive "peak" broken; FS break-even still −$15M/qtr; four deals in integration; KPMG, effective ICFR 38 Mostly hits its money guidance; misses on specific promises.
7 Competitive position & disruption 10% ~55% regional food share (rising); Indonesia food 46% vs Gojek 31%; behind Xanh SM in Vietnam rides; Sea Monee ~5x larger in lending 38 A strong regional leader with contested pockets.
8 Cyclicality & commodity/macro sensitivity 10% Consumer discretionary; mobility GMV fell ~30% in COVID lockdowns; fuel shock forced $7M+ driver support; rupiah ~17,900/$; consumer credit book 50 Cyclical demand plus a growing credit book and FX translation.
9 Regulatory, legal & geopolitical 5% Indonesia rules by 31 Dec; Vietnam NCC review; Taiwan FTC 27 Oct; Singapore CPF rising to 17% by 2029; Malaysia Gig Workers Act; KPPU cases open 65 Large pending rulings in several markets at once, set by governments that treat drivers as a political constituency.
10 Valuation & expectations 10% P/S 3.5x vs 6–8.7x history; EV/FY26 adj EBITDA 11.7x (15x deposit-adjusted); fwd P/E 18x; short interest ~7–9%; consensus above guidance 30 Fair to cheap against its own history, fair against peers.

Overall: weighted average = 0.15×25 + 0.10×15 + 0.10×30 + 0.10×40 + 0.10×28 + 0.10×38 + 0.10×38 + 0.10×50 + 0.05×65 + 0.10×30 = 3.75 + 1.5 + 3.0 + 4.0 + 2.8 + 3.8 + 3.8 + 5.0 + 3.25 + 3.0 = 33.9 → 34.

Change vs last: −12 (46 → 34).

  • Method (most of it): the old 6-area method had no separate liquidity, leverage or profitability lines, where Grab scores 15–40. It put 25% on "business & competition" (42) and 10% on regulation (72).
  • Substance, worse:
    • The incentive "peak" promise was broken.
    • Impairment rose to 7.2% of revenue.
    • Adjusted FCF was redefined in Q2.
    • Free cash after deposits roughly equals commitments.
  • Substance, better:
    • Atome's $1.49B is due by Q3 2027, not imminently.
    • First hard credit data is clean: 91+ days at 1.6%.
    • The auditor is KPMG with effective controls.
    • Stock pay is verified at $240M, flat.
    • The Perpres text gives no number for cars yet.
  • Net substantive change: about zero.

Ranking check: Grab (34) sits level with AVGO (34) and just above NU (33). Its lending book is far smaller relative to the business than Nubank's, and it holds net cash, but it ranks above NU because regulators in three countries can set its take rate and its GAAP profit is thinner. It is clearly safer than VIST (41), which is 99.7% exposed to Argentina and to oil.

Conclusion

  • What you are really betting on: that a profitable, share-gaining marketplace keeps compounding while governments decide how much of each fare it may keep, and that management turns ~$3B of committed cash (Atome, Taiwan, buyback) into per-share value rather than just scale.
  • The 2–3 things that will decide the outcome:
    1. Indonesia's implementing rules (by 31 Dec): motorbike-only versus cars, food and social security.
    2. Whether incentives and credit costs grow slower than revenue from Q3 on. Those two lines are where the margin story is won or lost.
    3. Capital allocation: buyback execution near $3, versus more deals before Atome is digested.
  • Signposts:
    • Bull confirmed:
      • Q3 incentives below 10.5% of on-demand GMV.
      • FS segment EBITDA ≥$0 in Q3 or Q4-26.
      • The share count falls below 4.0B by Q4-26.
      • Indonesian rules leave car commissions and food fees untouched.
    • Bull broken:
      • Impairment above 7% of revenue for a second quarter, with no NPL disclosure.
      • The Indonesian rules set a car cap or platform-paid social security.
      • FY26 adjusted EBITDA below $720M, or FS break-even slipping into 2027.
      • Another acquisition before Atome closes, or an Uber block sale.
  • New since the last report:
    • Hard numbers replaced charts: the 91+ days bucket is 1.6% (Dec-25), allowance coverage 6.6% (Jun-26), and impairment rose from 4.2% to 7.2% of revenue in four quarters.
    • Corrections:
      • The 2028 target is $1.7B including Atome (not "organic").
      • Atome closes by Q3 2027.
      • The convertible matures in 2030, with the put in 2028.
      • The auditor is KPMG.
      • The new $750M buyback dates from 4 Aug.
    • Q1's "incentives peak" was broken in Q2. GoTo cut its on-demand guidance because of the cap, while Grab's motorbike exposure is ~6% of mobility GMV.

Open questions & data gaps

  • NPL ratio: the 90-day NPL ratio by book is still not disclosed. The group ageing table is from Dec-25, and the Superbank and GXS figures are secondary. This matters most for Bear #4 and for factor 8.
  • Perpres text: the full text of Perpres 27/2026 was not read, because the government legal-document sites (JDIH) were blocked. Unknowns: who pays the social-security premium, the sanctions, and whether a car clause exists. This matters a lot for the biggest regulatory headwind.
  • Holdco cash: the split of free cash between the holding company and the bank subsidiaries is not disclosed. Only the $205M restricted figure is known.
  • Consensus gaps: consensus adjusted EBITDA, FY28 figures, and whether Atome or Taiwan are included were all not found. The tailwind and headwind sizing uses FY27 EPS of $0.14 as the base.
  • Atome: revenue, profit and credit history are not found, and the figures are unaudited.
  • Buyback: repurchases since 4 Aug were not found.
  • Uber: its intentions are known only from a 13D/A boilerplate. Its own Q2 call comments on Grab were not read.
  • Vietnam: share data comes from one research firm (Mordor). The NCC outcome is not found.
  • Call quotes: quotes come from third-party transcript summaries, not verified word for word.
  • Q3 results date: not yet announced. Last year it was 3 Nov, and Unusual Whales estimates 3 Nov.
Sources
  1. Grab Reports Record Second Quarter 2026 Results (6-K Ex. 99.1) — Grab/SEC EDGAR — 2026-08-04 — https://www.sec.gov/Archives/edgar/data/1855612/000185561226000123/a2026q2-earningspressrelea.htm
  2. Grab Condensed Consolidated Interim Financial Statements, H1 2026 (6-K) — SEC EDGAR — 2026-08 — https://www.sec.gov/Archives/edgar/data/1855612/000185561226000131/ck0001855612-20260630_d2.htm
  3. Grab Holdings Ltd, Annual Report on Form 20-F FY2025 — SEC EDGAR — 2026-03-06 — https://www.sec.gov/Archives/edgar/data/1855612/000185561226000020/ck0001855612-20251231.htm
  4. Grab to acquire majority stake in Atome Financial (6-K Ex. 99.1) — SEC EDGAR — 2026-09-15 — https://www.sec.gov/Archives/edgar/data/0001855612/000185561226000133/exhibit991-atometransactio.htm
  5. Grab Intends to Complete the Remaining $900 Million of its Share Repurchase Authorization (6-K) — SEC EDGAR — 2026-09-15 — https://www.sec.gov/Archives/edgar/data/1855612/000185561226000135/grab-sharerepurchasepr.htm
  6. Grab Q4 and FY2025 earnings press release — SEC EDGAR — 2026-02-11 — https://www.sec.gov/Archives/edgar/data/1855612/000185561226000011/a2025q4-earningspressrelea.htm
  7. Grab Q3 2025 earnings press release — SEC EDGAR — 2025-11-04 — https://www.sec.gov/Archives/edgar/data/1855612/000185561225000058/a2025q3-earningspressrelea.htm
  8. Grab Q1 2026 earnings press release — SEC EDGAR — 2026-05-05 — https://www.sec.gov/Archives/edgar/data/0001855612/000185561226000077/a42026q1-earningspressrele.htm
  9. Grab Q2 2026 earnings call transcript (third-party; summary excerpts) — Platform Aeronaut — 2026-08-04 — https://transcripts.platformaeronaut.com/transcripts/GRAB-2Q26-transcript
  10. Grab Q1 2026 earnings call transcript (third-party; summary excerpts) — Platform Aeronaut — 2026-05-05 — https://transcripts.platformaeronaut.com/transcripts/GRAB-1Q26-transcript
  11. Grab Q4 2025 earnings call transcript (third-party; summary excerpts) — Platform Aeronaut — 2026-02-11 — https://transcripts.platformaeronaut.com/transcripts/GRAB-4Q25-transcript
  12. Grab insider Form 4 filings, 2026-04-02 to 2026-10-07 (incl. Tan and Hungate purchases 2026-09-21) — SEC EDGAR — 2026 — https://www.sec.gov/Archives/edgar/data/0001855612/000189649726000009/wk-form4_1790037630.xml (index: https://data.sec.gov/submissions/CIK0001855612.json)
  13. Uber Technologies Schedule 13D/A on Grab — SEC EDGAR — 2026-07-08 — https://www.sec.gov/Archives/edgar/data/1855612/000155278126000379/primary_doc.xml
  14. Toyota Motor Corp Schedule 13G/A on Grab (0 shares at 2026-09-09) — SEC EDGAR — 2026-09-14 — https://www.sec.gov/Archives/edgar/data/0001855612/000119312526389761/primary_doc.xml
  15. Grab stock forecast and analyst ratings — stockanalysis.com (S&P Global data) — 2026-10-12 — https://stockanalysis.com/stocks/grab/forecast/
  16. Grab statistics and peer statistics (Sea, Uber, DoorDash) — stockanalysis.com — 2026-10-12 — https://stockanalysis.com/stocks/grab/statistics/
  17. Grab historical valuation ratios — stockanalysis.com — 2026-10-12 — https://stockanalysis.com/stocks/grab/financials/ratios/
  18. Grab stock hits three-year low on Atome deal; execs buy shares (Reuters) — DealStreetAsia — 2026-09-23 — https://www.dealstreetasia.com/?p=495937
  19. Grab shares slide 7.5% on $2.39B cash outlay for Atome and buyback — stockti — 2026-09 — https://stockti.com/grab-shares-slide-7-5-on-2-39b-cash-outlay-for-atome-and-buyback
  20. GRAB short interest history (FINRA data) — curvedtrading — 2026-10 — https://curvedtrading.com/stocks/grab/short-interest/
  21. Superbank bukukan laba Rp182,6 miliar pada semester I 2026 — Kontan — 2026 (exact date not verified) — https://keuangan.kontan.co.id/news/superbank-bukukan-laba-rp-1826-miliar-pada-semester-i-2026
  22. Melesat di semester I, nasabah Superbank tembus 7 juta, kredit Rp13,4 triliun — Kabarbursa — 2026 (exact date not verified) — https://www.kabarbursa.com/market-hari-ini/melesat-di-semester-i-nasabah-superbank-supa-lebih-tujuh-juta-kredit-tembus-rp134-triliun
  23. GXS Group posts sharp cost-to-income improvement in FY2025 — The Asian Banker — 2026 (exact date not verified) — https://www.theasianbanker.com/press-releases/gxs-group-posts-sharp-cost-to-income-improvement-in-fy2025-as-lending-portfolio-grows
  24. Perpres Ojol 2026 terbit, Gojek bilang begini — Detik Oto — 2026-10-09 — https://oto.detik.com/berita/d-8699729/perpres-ojol-2026-terbit-gojek-bilang-begini
  25. Perpres Ojol terbit, Menhub: beri kepastian hukum pada pengemudi — Katadata — 2026-10 — https://katadata.co.id/berita/industri/6ac75160ca217/perpres-ojol-terbit-menhub-beri-kepastian-hukum-pada-pengemudi
  26. Perpres Ojol terbit: atur potongan hingga status driver — Uzone — 2026-10-08 — https://uzone.id/perpres-ojol-terbit-atur-potongan-hingga-status-driver-ojek-online
  27. Perpres Ojol berlaku, ini hak driver dan kewajiban aplikator — Harian Jogja — 2026-10 — https://news.harianjogja.com/r-1274472/perpres-ojol-berlaku-ini-hak-driver-dan-kewajiban-aplikator
  28. Soal Perpres Ojol, begini respons Grab soal komisi driver — Liputan6 — 2026-10-08 — https://www.liputan6.com/tekno/read/8309220/soal-perpres-ojol-begini-respons-grab-soal-komisi-driver
  29. GoTo Q2 2026 prepared remarks (third-party transcript) — Platform Aeronaut — 2026-07-29 — https://transcripts.platformaeronaut.com/transcripts/GGTKY-2Q26-transcript
  30. GOTO profit surges, revising outlook as commission capped at 8% — IDNFinancials — 2026-07/08 (search summary only) — https://www.idnfinancials.com/news/66773
  31. Danantara eyes bigger stake in GoTo — The Jakarta Post — 2026-05-12 — https://www.thejakartapost.com/business/2026/05/12/danantara-eyes-bigger-stake-in-goto
  32. Vietnam regulator reviews Grab's commission policies — The Edge Malaysia — 2026-09-12 — https://theedgemalaysia.com/node/817840
  33. Nhiều nước khống chế trần chiết khấu xe công nghệ — CafeF — 2026-09-11 — https://cafef.vn/nhieu-nuoc-khong-che-tran-chiet-khau-xe-cong-nghe-188260911201945397.chn
  34. Xanh SM overtakes Grab to dominate Vietnam ride-hailing (Mordor Intelligence data) — Tuoi Tre News — 2025-08-12 — https://news.tuoitre.vn/vietnamese-ride-hailing-app-xanh-sm-turns-the-tables-overtakes-grab-to-dominate-market-103250812172910459.htm
  35. Grab 收購台灣 foodpanda 案續審 (FTC extends review to 10-27) — TechNews — 2026-07-22 — https://finance.technews.tw/2026/07/22/fair-trade-commission-20260722/
  36. FTC extends Grab-foodpanda review — Taipei Times — 2026-07-23 — https://www.taipeitimes.com/News/biz/archives/2026/07/23/2003861199
  37. Grab-Foodpanda deal still under review despite media report, Taiwan watchdog says — MLex — 2026-08-18 — https://www.mlex.com/articles/2514542/grab-foodpanda-deal-still-under-review-despite-media-report-taiwan-watchdog-says
  38. Uber agrees $14.8bn Delivery Hero takeover at €41.50 — The Next Web — 2026-07-16 — https://thenextweb.com/news/uber-delivery-hero-agreed-takeover-41-50
  39. What will Uber gain from buying Delivery Hero (carve-out; close 2H27) — Inside Retail Asia — 2026-07-17 — https://insideretail.asia/2026/07/17/what-will-uber-gain-from-buying-delivery-hero/ ; Uber Form 10-Q Q2 2026 (Grab stake $2.02B) — SEC — 2026-08 — https://www.sec.gov/Archives/edgar/data/0001543151/000154315126000032/uber-20260630.htm (summary only)
  40. Southeast Asia food delivery GMV grew 18% to US$22.7B in 2025 — Momentum Works / The Lowdown — 2026-01-28 — https://thelowdown.momentum.asia/new-report-southeast-asia-food-delivery-gmv-grew-18-to-hit-us22-7b-in-2025/
  41. Grab kuasai layanan pesan antar makanan di RI 2025, kalahkan Gojek dan ShopeeFood — Bisnis.com — 2026-01-28 — https://teknologi.bisnis.com/read/20260128/84/1947981/grab-kuasai-layanan-pesan-antar-makanan-di-ri-2025-kalahkan-gojek-dan-shopeefood
  42. Sea Limited Reports Second Quarter 2026 Results — Business Wire — 2026-08-11 (summary only) — https://businesswire.com/news/home/20260810021050/en/Sea-Limited-Reports-Second-Quarter-2026-Results
  43. Thailand to tighten enforcement across ride-hailing services from Oct 1 — The Star — 2026-09-24 — https://www.thestar.com.my/aseanplus/aseanplus-news/2026/09/24/thailand-to-tighten-enforcement-across-ride-hailing-services-from-oct-1
  44. Platform Workers Act: why your rides and deliveries are going to cost more — DollarsAndSense — 2025/26 (summary only) — https://dollarsandsense.sg/business/platform-workers-act-why-your-rides-and-deliveries-are-going-to-cost-more/
  45. Grab to raise ride platform fee in Singapore, adding 30 cents from Jan 1 — Malay Mail — 2025-12-24 — https://malaymail.com/news/singapore/2025/12/24/grab-to-raise-ride-platform-fee-in-singapore-adding-30-cents-from-jan-1/203067
  46. Waymo to launch operations in Singapore in 2028 — Electrive — 2026-10-06 — https://www.electrive.com/2026/10/06/waymo-to-launch-operations-in-singapore-in-2028/
  47. Grab's Indonesia shock: driver-fee cap hits GRAB stock days before earnings (Citi estimate) — TS2 — 2026-05 (secondary) — https://ts2.tech/en/grabs-indonesia-shock-driver-fee-cap-hits-grab-stock-days-before-earnings/
  48. Grab Holdings faces hurdles, but upside potential is hard to ignore (Morgan Stanley: Indonesia 17–19% of mobility GMV, ~20% of EBITDA) — Barchart via Nasdaq — 2026 (secondary) — https://www.nasdaq.com/articles/grab-holdings-faces-hurdles-upside-potential-hard-ignore
  49. Indonesian rupiah news — Trading Economics — 2026-10-12 — https://tradingeconomics.com/indonesia/currency/news/567062
  50. Grab looks like a monopoly. Atome is the execution test — DC Capital (Substack) — 2026-09-21 — https://dccapital.substack.com/p/grab-looks-like-a-monopoly-atome
  51. Grab bets $1.49B on Atome as COO buys in — Basis Report — 2026-09-22 — https://www.basisreport.com/news/grab-grab-atome-acquisition-coo-insider-buy
  52. Grab Holdings — Maybank Research report, Buy, target $6.25 (Uber overhang scenarios) — minichart (snippet of paywalled note) — 2026-07-22 — https://www.minichart.com.sg/2026/07/22/grab-holdings-maybank-research-pte-ltd-report-buy-target-price-6-25-75-upside/
  53. Maybank keeps BUY on Grab as Atome deal raises execution stakes; DBS on Grab–GoTo — minichart (snippets of paywalled notes) — 2026-09-28 and 2026-05-05 — https://www.minichart.com.sg/2026/09/28/maybank-keeps-buy-on-grab-holdings-as-atome-deal-raises-execution-stakes/ ; https://www.minichart.com.sg/2026/05/05/grab-goto-merger-indonesian-take-rate-cap-paves-way-for-synergies-and-regulatory-approval/
  54. Grab fair value edges lower as analysts split (Barclays $7→$5, Morgan Stanley $6.25) — Simply Wall St via Yahoo Finance — 2026-09-19 — https://finance.yahoo.com/markets/stocks/articles/grab-grab-stock-fair-value-191243090.html
  55. Singapore's Grab lifts annual forecasts (Reuters) — New Straits Times — 2026-08 — https://www.nst.com.my/amp/business/corporate/2026/08/1503771/singapores-grab-lifts-annual-forecasts-ai-incentives-drive
  56. Grab to Announce Second Quarter 2026 Results on 3 August 2026 — Grab IR — 2026-07-07 — https://s205.q4cdn.com/179588156/files/doc_news/Grab-to-Announce-Second-Quarter-2026-Results-on-3-August-2026-2026-0nRL6bYnpf-2026.pdf
  57. Grab Q2 2025 earnings press release — SEC EDGAR — 2025-07-31 — https://www.sec.gov/Archives/edgar/data/1855612/000185561225000045/a2025q2-earningspressrelea.htm
  58. Grab Holdings EGM Proxy 2026 (Class B 45→90 votes) — Grab IR — 2026-03-19 — https://s205.q4cdn.com/179588156/files/doc_document_list/2026/Mar/19/Grab_Holdings_Limited-EGMProxy2026-FULL-SET-1-79a3a9.pdf
All GRAB Bulls vs Bears reports
Disclaimer. AI-generated research and podcast (Claude by Anthropic, voices by Google Gemini) for educational purposes — not financial advice and not a recommendation to buy or sell. Willem (MR.INVEST) owns this stock. Check the company’s own filings before making any decision.
← All Bulls vs Bears reports