GRAB Four Records on the Income Statement. Free Cash Flow Fell Anyway.

A Quarter of Records, Except the One That Measures Cash

Grab's second quarter looks, on the surface, like the cleanest beat in this blog's two-year coverage of the company: revenue grew 22% year-over-year (21% constant-currency) to an all-time high of $997 million, On-Demand GMV^» grew 21% YoY to $6.5 billion, Group Monthly Transacting Users hit a record 54 million, and Adjusted EBITDA^» grew 54% YoY to $168 million, a new all-time high that Grab's own materials describe as the company's eighteenth straight quarter of sequential Adjusted EBITDA growth (see Trailing Eight Quarters below). Profit for the period was $235 million, up from just $20 million a year ago. Management raised full-year 2026 guidance on both revenue and Adjusted EBITDA, and the Board authorized a fourth straight share buyback - $750 million this time, bringing cumulative authorizations to $1.75 billion since 2024.

Two things complicate that clean read. First, $307 million of the $334 million increase in finance income - and therefore most of the $215 million YoY profit jump - came from a one-time, non-cash gain Grab itself flags as "one-time in nature": the accounting remeasurement triggered when its Superbank stake crossed 50% and the bank moved from an equity-accounted associate onto Grab's own consolidated balance sheet. Operating profit, the number closest to what the business itself actually earned this quarter, improved by a much smaller $12 million, to $19 million. Second, and less visible in any of the quarter's own headlines: trailing-twelve-month Adjusted Free Cash Flow fell to $450 million from $489 million three months ago - the first sequential decline in this metric this blog has tracked since Grab started reporting it in Q4 2023, arriving in the same quarter as four separate records on the income statement (see Target Valuation Range below).

None of this makes the quarter bad. Deliveries margin expanded for a sixth straight quarter, Financial Services losses kept narrowing toward a management-guided second-half breakeven, and Regional Corporate Costs^» growth fell back below revenue growth after last quarter flagged it as a real problem. But a quarter with four records on the income statement and a declining trailing free-cash-flow figure is a genuinely two-sided result, and the income-statement records are doing most of the talking in Grab's own materials.

The Prescription

Grab should keep running the Financial Services playbook it just executed: distributing through an ecosystem that gives it near-zero customer acquisition cost, then consolidating the two acquisitions (Superbank, Stash) that extend that distribution into full banking and wealth products rather than just lending. Gross Loan Portfolio^» nearly tripling YoY to $2.3 billion, NPL^» ratios holding stable through that scale-up, and the segment's Adjusted EBITDA loss narrowing to $15 million from $26 million a year ago are exactly the ingredients a "profitability inflection in H2 2026" call requires - and the fact that Deliveries kept expanding margin and Mobility kept growing 18% through the same quarter shows the core business isn't being starved to fund the fintech build-out.

What it should stop doing: describing its own ownership stake in the asset that just generated its single largest line-item swing of the quarter only as "over 50%." A stake crossing an accounting threshold worth $307 million to the income statement is exactly the kind of number that deserves a precise figure, not a directional one - especially from a company that, in the same set of materials, discloses a fair-value reconciliation precise to the million dollar. This isn't a new complaint: Q1 2025's post first flagged Grab's Superbank ownership disclosures getting vaguer, not more precise, as the stake grew, and this quarter continues that pattern from the other direction.

Key Financial Metrics

Q2 2026 vs. Q2 2025 - consolidated, reported in USD

Grab reports natively in US dollars, so no FX conversion is needed here.

Metric	Q2 2026	Q2 2025	YoY	Note
On-Demand GMV	$6,463M	$5,354M	✅ +21%	+22% constant-currency
Revenue	$997M	$819M	✅ +22%	+21% constant-currency; new all-time high
Total Segment Adjusted EBITDA	$272M	$201M	✅ +35%	Outpaced revenue growth
Adjusted EBITDA	$168M	$109M	✅ +54%	All-time high; per Grab, its eighteenth straight quarter of sequential growth
Operating profit (IFRS)	$19M	$7M	✅ +186%	The number closest to actual operating improvement
Profit for the period	$235M	$20M	⚠️ NM	$307M of the increase is a one-time Superbank consolidation gain (see below)
Net cash from operating activities	$56M	$64M	⚠️ -12%	Higher working capital requirements
Adjusted Free Cash Flow (quarter)	$73M	$112M	⚠️ -35%	Definition changed this quarter (see Beyond the Usual)
Adjusted Free Cash Flow (trailing 12 months)	$450M	n/a	⚠️ -8% QoQ	First sequential decline since this metric launched in Q4 2023

Profit for the period improved $215 million YoY to $235 million. Grab's own reconciliation attributes $334 million of that swing to higher finance income, of which $307 million was a non-cash gain from remeasuring its previously-held Superbank equity stake to fair value upon consolidation, plus a $66 million favorable movement from recognizing deferred tax assets - partially offset by a $183 million larger fair-value loss on financial assets and liabilities (largely the embedded derivative in the $1.5 billion convertible notes issued in Q2 2025, a volatility source this blog has tracked every quarter since). Strip out the Superbank gain and the tax item, and the underlying improvement is closer to what operating profit itself shows: $12 million better YoY, not $215 million.

Balance sheet metric	Jun 2026	Dec 2025	Change
Total Assets	$13,445M	$11,983M	✅ +12%
Total Liabilities	$6,242M	$5,226M	⚠️ +19%
Total Equity	$7,203M	$6,757M	✅ +7%

Both sides of the balance sheet grew meaningfully this quarter for a reason unrelated to organic growth: Superbank's full consolidation adds its assets and liabilities - including $2.5 billion of customer deposits across all three digital banks, up from $1.6 billion at year-end - onto Grab's books for the first time (see Beyond the Usual). Gross cash liquidity rose to $7.4 billion from $6.9 billion three months ago, and net cash liquidity rose to $5.4 billion from $5.0 billion (see Target Valuation Range below).

Trailing Eight Quarters

Period	Revenue	Adjusted EBITDA	Profit/(Loss) for the Period
Q3 2024	$716M	$90M	$15M
Q4 2024	$764M	$97M	$11M
Q1 2025	$773M	$106M	$10M
Q2 2025	$819M	$109M	$20M
Q3 2025	$873M	$136M	$17M
Q4 2025	$906M	$148M	$153M
Q1 2026	$955M	$154M	$120M
Q2 2026	$997M	$168M	$235M

Adjusted EBITDA has improved sequentially every quarter shown in this table. Grab's own materials describe this as the company's "eighteenth straight quarter" of sequential Adjusted EBITDA growth - the identical ordinal Q1 2026's materials used for what was, at the time, also described as the eighteenth. This blog can't reconcile that from the disclosed information alone, so it's noted here rather than silently repeated or corrected. Profit for the period has been more volatile: Q4 2025's $153 million was itself driven by a large convertible-notes derivative fair-value swing, this quarter's $235 million by the Superbank gain - two of the three largest profit quarters in this table both trace mostly to non-operating, non-cash items rather than a step-change in the underlying business. Read against that pattern, Q3 2025's more modest $17 million - the quarter closest to pure operating performance - may be a better read on Grab's underlying earnings power than the two much larger, headline-grabbing numbers on either side of it.

Segment Results

Q2 2026, four reportable segments

Grab reports four segments - Deliveries, Mobility, Financial Services, and Others - unchanged since Q1 2024's restructuring.

Segment	GMV (Q2'26)	YoY	Revenue (Q2'26)	Segment Adj. EBITDA (Q2'26)	Margin	YoY Margin
Deliveries	$4,249M	✅ +22% (+24% CC)	$531M	$96M	✅ 2.3% (of GMV)	vs 1.8% in Q2'25
Mobility	$2,214M	✅ +18%	$331M	$191M	⚠️ 8.6% (of GMV)	vs 8.7% in Q2'25
Financial Services	n/a (see Gross Loan Portfolio)	—	$134M	$(15)M	✅ (11.2)% (of revenue)	vs (30.3)% in Q2'25
Others	n/a	—	$1M	~$0M	— NM	roughly flat YoY

Deliveries GMV grew 22% YoY (24% constant-currency) to $4,249 million, and Segment Adjusted EBITDA grew 53% YoY to $96 million as margin expanded to 2.3% of GMV from 1.8% a year ago - a sixth straight quarter of YoY margin expansion in this blog's coverage. GrabMart grew at 1.7 times the rate of core Food Deliveries this quarter, management's clearest single data point yet for where incremental growth in the segment is actually coming from, backed by GrabMart transaction growth of 54% YoY and online SKUs sold up 52% YoY. GrabUnlimited subscribers grew 20% YoY and now account for 35% of total Deliveries GMV, transacting 4x more and spending 5x more than non-subscribers - the loyalty-driven upsell mix this blog has tracked since Q3 2025 continuing to compound.

Mobility GMV grew 18% YoY to $2,214 million, with segment margin essentially flat at 8.6% of GMV versus 8.7% a year ago - a small decline management attributes to "recalibrating incentive spend towards driver-partners to strengthen supply" amid a regional fuel-price crisis that persisted through the quarter (Grab committed over $7 million in localized driver support during the period). Mobility transactions grew 28% YoY, continuing to outpace GMV growth, the same "more, cheaper trips" pattern this blog has tracked since mid-2025 - reinforced this quarter by Saver and Group ride penetration, with 1 in 4 new Mobility users entering via a Saver ride. Indonesia's new commission structure for two-wheel ride-hailing (ojol) took effect in July, a regulatory thread first flagged last quarter with no resolution yet disclosed.

Financial Services revenue grew 59% YoY (62% constant-currency) to $134 million, again the fastest-growing segment, and the segment Adjusted EBITDA loss narrowed to $15 million from $26 million a year ago and from $17 million last quarter - the third straight quarter of sequential improvement. Total loans disbursed grew 72% YoY to an all-time high $1.2 billion in the quarter (a $5.0 billion annualized run-rate), and Gross Loan Portfolio nearly tripled YoY to $2,318 million; excluding Superbank's contribution, the portfolio still doubled YoY to $1.6 billion across GrabFin, GXS Bank, and GXBank. Management again stated that NPL ratios "have been stable" and that expected-credit-loss^» provisioning improved even as the book scaled - the same claim flagged last quarter as unverifiable in dollar terms, and it remains so this quarter. The big structural change: beginning June 2026, Grab began consolidating Superbank's results after its shareholding crossed 50% (see Beyond the Usual), and in July - after quarter-end - Grab completed its acquisition of Stash Financial, a US-based wealth-management platform with $5.5 billion in Assets Under Management^», first disclosed as a pending deal in the FY2025 20-F covered last quarter but one. Superbank's own Q2 2026 report as a newly-consolidated but separately-listed entity - it trades independently on the Indonesia Stock Exchange as SUPA - shows standalone net income up 415.5% quarter-on-quarter, a business genuinely improving even as its own stock lost more value over the same period.

Others posted roughly $1 million of revenue, immaterial to the group total.

Every segment either grew or held margin this quarter, but the standout shift is compositional: Financial Services' contribution to the group - in revenue growth rate, in balance-sheet size, and now in headline profit - jumped discontinuously this quarter because two acquisitions landed in the same three-month window, not because the underlying fintech business itself re-rated. That's worth separating from the segment's genuine, gradual operating improvement (loss narrowing three straight quarters running) when reading this quarter's numbers.

Beyond the Usual

Grab's only Q2 2026 source documents are the earnings press release (also furnished to the SEC as a 6-K exhibit with the full unaudited financial statements attached, identical in substance), the supplemental investor presentation, and CEO/COO/CFO prepared remarks - no annual 20-F was filed between last quarter's coverage and this one, so the dormant legal/governance threads this blog tracks (the GrabMart trademark suit, the Singapore tax assessments, the Toyota framework agreement) have no new company-disclosed developments this quarter.

A $307 Million Accounting Gain, Not a $307 Million Improvement In The Business

Grab discloses clearly, in both the press release and CFO Peter Oey's prepared remarks, that the $307 million gain from consolidating Superbank is "one-time in nature" and that profit "in the second half may continue to reflect a degree of variability tied to fair value measurements and other non-operating items." That disclosure is genuine and specific - this isn't a hidden item. But it sits several paragraphs below three headline bullets (revenue, GMV, profit) that don't distinguish between operating and non-operating drivers, and a reader skimming the top of the release would reasonably conclude the business itself grew earnings twelvefold this quarter. Operating profit - the closest IFRS line item to what the business itself actually earned - improved by $12 million, not $215 million.

Adjusted Free Cash Flow Was Redefined Again, For The Third Time In This Blog's Coverage

Starting this quarter, Grab's Adjusted Free Cash Flow now subtracts "net changes in treasury liquidity positions in the Financial Services segment" - a new $95 million deduction with a $0 comparative in the prior-year column, meaning the reported $73 million figure isn't computed the same way as the $112 million Q2 2025 comparator sitting next to it. This follows a definition change in Q1 2025 (adding PP&E disposal proceeds) and the metric's initial launch with its own methodology in Q4 2023. Each change has a stated rationale - this one specifically to exclude treasury-liquidity movements now that Superbank and Stash bring meaningfully larger balance-sheet cash management into the mix - but three redefinitions in roughly two and a half years of reporting the metric means the trailing-twelve-month figure ($450 million) blends quarters computed under at least two different formulas, not one consistent methodology.

Superbank's Consolidation Left Its Fingerprints All Over The Balance Sheet

Intangible assets and goodwill jumped to $1,575 million from $1,057 million (the new goodwill recognized on the acquisition), non-controlling interests in equity rose to $437 million from $29 million (Grab's roughly-49% partners in a bank it now consolidates at just over 50%), and the non-IFRS reconciliation shows a new $15 million "Day-1 expected credit loss on an acquired loan portfolio" line - the standard accounting charge for provisioning against a loan book on the day it's absorbed onto the balance sheet, added back in the Adjusted EBITDA reconciliation but a real IFRS cost this quarter nonetheless.

Last Quarter's Unexplained $105 Million Outflow Now Reads Like It's Been Answered

The consolidation resolves a thread this blog left open last quarter: Q1 2026's post flagged a $105 million "acquisition of associates and joint venture" outflow with no stated purpose. This quarter's cash flow statement shows that outflow landed entirely in Q1 (the six-month figure is $105 million, the standalone Q2 figure just $1 million) - consistent with it being exactly the share purchase that pushed Grab's Superbank stake past the 50% consolidation threshold in the following quarter. Grab never explicitly confirmed this connection in either quarter's materials, but the timing and the subsequent consolidation announcement line up closely enough to treat the question as effectively answered.

A Long-Tenured Uber Board Seat Changed Hands Amid An Active Cross-Border Deal

In July, Uber board representative Dara Khosrowshahi stepped down from Grab's Board of Directors - a seat he'd held since 2018, when Uber sold its Southeast Asia operations to Grab in exchange for a stake and board representation. CEO Anthony Tan addressed the departure directly on the call, thanking Khosrowshahi and noting Grab "maintain[s] ongoing dialogue with Uber in their capacity as a shareholder." The timing sits alongside Grab's still-pending acquisition of Delivery Hero's foodpanda Taiwan business - first disclosed last quarter as Grab's first operating expansion outside Southeast Asia - which remains unclosed as of this call, with Tan noting only that Grab "continues with its efforts to complete" it. Neither document draws an explicit line between the board change and the pending deal, and nothing here suggests one exists, but a long-tenured board seat tied to Grab's regional-competitor history changing hands during an active cross-border acquisition is worth watching rather than dismissing as routine board turnover.

Regional Corporate Costs Growth Fell Back Below Revenue Growth, Resolving Last Quarter's Concern

Regional Corporate Costs grew 13% YoY to $104 million - below revenue's 22% growth for the first time in several quarters, and $10 million lower than last quarter's $114 million on lower staff costs and professional fees. This directly resolves the concern this blog raised last quarter about RCC growth outpacing revenue with no dollar-level breakdown of the AI-infrastructure spend driving it; CFO Oey now guides RCC to stay "broadly stable around second-quarter run-rate levels" heading into the second half.

Management's Framing Of A Quarter It Called Strong, And What The Board Note Didn't Fully Address

CEO Anthony Tan built his opening remarks around three structural arguments for why On-Demand growth can keep accelerating rather than plateau: user penetration is still nascent relative to developed-market frequency, product-tier affordability keeps unlocking users at price points Grab couldn't previously serve profitably, and engagement is deepening (Daily Transacting User growth outpacing MTU growth, retention stable). He framed Financial Services as "approaching the inflection point we have been building toward for years," explicitly contrasting Grab's ecosystem-distribution model with "standalone fintechs," and credited the Grab intelligence layer - halved AI-interaction costs since June 2025, autonomous coding agents cutting product time-to-market by up to 30% - as a margin lever rather than a cost center. None of this framing addressed the profit-quality question directly; the Superbank gain's one-time nature was disclosed by CFO Oey in the financial-results section, not folded into Tan's strategic narrative.

On the board departure, Tan's remarks were more direct than this blog has seen management address a governance change in prior quarters - naming Khosrowshahi, the 2018 Uber transaction, and Grab's continuing relationship with Uber as a shareholder explicitly, rather than leaving it to a filing footnote. What the remarks didn't do is connect that change to the foodpanda Taiwan deal's still-open status, or explain why the departure happened now rather than at a scheduled board cycle - a gap this blog flags in Beyond the Usual above without asserting a connection that the materials themselves don't support.

Target Valuation Range

DCF-implied fair value of roughly $2.55-$4.35 per share (bear to bull), against a $3.77 close - the base case (~$3.45) now sits below the actual price, only the bull case clears it, a reversal from last quarter's convergence where the base case landed almost exactly on the price. The reversal is driven by a declining trailing free-cash-flow base, not a rising stock price - the more cautious of the two ways this gap could have opened up.

Grab's stock closed the second quarter at $3.77, up 3% from $3.66 three months ago and still well below the $4.99-$6.02 range this blog tracked through most of 2025 (see Q1 2026's coverage of that decline). Using the 4,084,860 thousand weighted-average basic shares Grab discloses for the quarter:

Market cap → enterprise value	Q2 2026
Share price (period-end)	$3.77
Shares outstanding	~4.085B
Market capitalization	~$15.4B
Total liabilities	n/a (net cash basis used)
Less: net cash liquidity	$5.4B
Enterprise value	~$10.0B

Market cap and enterprise value both moved only slightly from Q1 2026's ~$15.0B and ~$10.0B - a quiet quarter for the stock sandwiched between a headline-record income statement and a declining trailing free-cash-flow base, which is itself a data point: the market didn't reward the records much, and (based on the metric that matters most to a DCF) may have been right not to.

Peer-multiple sanity check	Q1 2026	Q2 2026	Change
Enterprise value	~$10.0B	~$10.0B	— roughly flat
EV/Revenue (annualized run-rate)	~2.6x	~2.5x	✅ down slightly
EV/Revenue (FY guidance)	~2.5x (FY2026 guidance, prior range)	~2.4x (FY2026 guidance, raised range)	✅ down
EV/On-Demand GMV (annualized run-rate)	~0.41x	~0.39x	✅ down slightly

Multiples compressed modestly again, continuing the pattern from the last several quarters, driven this time by revenue growth outpacing a roughly flat enterprise value rather than a falling stock price. Updating the DCF built last quarter with the new $450 million trailing Adjusted Free Cash Flow base (down from $489 million three months ago - see Beyond the Usual above for the methodology change contributing to that decline), the same 13% discount rate, and five-year fade paths still anchored to management's 2028 outlook ($1.5 billion Adjusted EBITDA, 80% cash conversion):

Scenario	Year 1-5 Adjusted FCF path	Terminal growth	Implied equity value per share
Current (Q2 2026 close)	— actual market price, for reference	—	$3.77
Bear	$320M → $520M	2%	~$2.55
Base	$420M → $980M	3%	~$3.45
Bull	$500M → $1,380M	4%	~$4.35

All three scenarios shifted modestly lower than last quarter's $2.68-$4.63 range, tracking the lower trailing Adjusted Free Cash Flow base the model now starts from. The base case (~$3.45) sits below the $3.77 close for the first time in two quarters, a reversal from Q1 2026's convergence where the base case landed almost exactly on the price. Read plainly: the stock held roughly flat this quarter while the cash-flow metric a DCF actually depends on moved backward - not a large gap, but a genuine one, and the opposite direction from every other headline number in this report. This remains a directional model built on management's own multi-year targets, not a definitive one, as noted every quarter it's been run here.

---