Finn
GRAB Consumer Internet · Southeast Asia · SuperApp · Ride Hailing · Thesis updated August 4, 2026

Grab is scaling, but the price is demanding

01 Running thesis

Scale is starting to show up

Grab is no longer only a growth story. The ecosystem hit 54 million monthly transacting users, which shows the app is still adding people after years of heavy competition. The company reported a net profit for 2025, and group adjusted EBITDA continues to expand.

The bull case is that scale and AI now feed better margins. Proprietary AI tools are proving their worth, with the cost per AI interaction halving year-over-year. Financial services is on track to achieve adjusted EBITDA breakeven in the second half of 2026 after fully consolidating Superbank and Stash. GrabMart continues to outpace food delivery growth, driving higher basket sizes and lifetime value.

The bear case is that this is still a delicate marketplace. Grab must keep riders cheap enough, drivers paid enough, and merchants willing to stay. Recent fuel price volatility forced a $7 million targeted incentive spend to protect driver supply. Foreign exchange translation also remains a structural headwind on reported U.S. dollar financials.

Finn gives Grab a balanced overall score, weighed down by valuation. The company is executing better, but the stock needs a lot to go right. The next clear tests are the Foodpanda Taiwan deal getting approval by the end of 2026, advertising penetration rising, and autonomous vehicle operations launching commercial fares in Punggol in Q4 2026.

Aug 2026Q2 2026 confirmed strong growth with MTUs hitting 54 million. Stash and Superbank were fully consolidated, and AV operations in Singapore advanced toward commercial fares in Q4.
May 2026Q1 2026 showed AI gains turning into operating proof. Group MTUs reached 52 million, Turbo lifted driver earnings 23%, and Mai lifted merchant GMV 15%.
Mar 2026The 2025 Form 20-F confirmed Grab's first net profit year and added Infermove, an AI robotics deal for first and last-mile delivery.
Feb 2026Q4 2025 confirmed a full year of net profit and a $1.3 billion gross loan portfolio. Financial Services breakeven moved to the second half of 2026.
Nov 2025Q3 2025 set a record $136 million in adjusted EBITDA and reached 48 million MTUs. Management framed AVs as a long-term hybrid fleet.
Jul 2025Q2 2025 showed profitable growth at scale, including 46 million MTUs and $20 million of net profit.
Apr 2025Q1 2025 raised full-year adjusted EBITDA guidance to $460 million to $480 million. Chope and Validus added dining reservations and supply invoice financing.
Mar 2025The 2024 Form 20-F added Everrise and an insurance license, but also disclosed a DOJ self-reporting matter tied to potential anti-corruption issues.
02 Business model

A toll booth on daily spending

Grab connects consumers with drivers, restaurants, grocers, stores, banks, and lenders. It earns money from commissions and fees on rides and deliveries, advertising sold to merchants, and lending spreads, which are the gap between what it earns on loans and what funding and credit losses cost.

The model gets stronger when one user does many things in the same app. A person may use Grab for a ride, order dinner, buy groceries through GrabMart, pay with a wallet, then use a digital bank. That repeat use can lower marketing cost per transaction over time.

The weak point is incentives. Grab reports revenue after many partner and consumer incentives. If fuel, wages, or competition rise, Grab may need to spend more to keep the network healthy. In early 2026, the company spent $7 million in targeted subsidies just to offset fuel price spikes and protect driver earnings.

Advertising and AI are the high-margin levers to watch. Advertising has tracked at over 1.7% of deliveries GMV, and self-serve tools can bring in more merchants. AI helps driver routing and merchant sales, while Grab tests a future hybrid fleet of human drivers, EVs, and autonomous vehicles.

03 Product portfolio

The app is becoming a stack

Growth engine

Deliveries

This includes food, grocery, parcel delivery, dine-out deals, and ads. GrabMart GMV is growing 1.7x faster than food deliveries.

Cash cow

Mobility

This is ride hailing, including Saver rides and premium trips. Mobility maintained 8.5% to 9% margins despite recent fuel subsidies.

Growth engine

Advertising

Grab sells promoted listings and banners to merchants and brands. The appeal is simple: ads sit close to the moment when users decide what to eat or book.

Option

Financial Services

This includes lending, digital banks like Superbank, and the U.S.-based investing platform Stash. The segment expects breakeven in late 2026.

Steady

Grocery and offline retail

Grab owns grocery assets such as Jaya Grocer and a majority interest in Everrise. These stores help GrabMart supply and can raise basket size.

Option

Autonomous and robotics

Grab is testing a future fleet that mixes human drivers with autonomous vehicles and robots. Singapore operations move to point-to-point commercial fares in Q4 2026.

04 Business segments

Deliveries leads the mix

Deliveries53%growing fast
Mobility36%modest
Financial Services10%growing fast
Others0%flat

Segment shares use 2025 revenue from Grab's 2025 Form 20-F. Deliveries and Mobility still drive most revenue, while Financial Services is smaller but growing faster.

05 Risk factors

What could break the story

Incentive spiral

High impact · Medium odds

Grab has to balance prices for users with pay for drivers and merchants. If fuel or competition rises, Grab must spend more to keep supply and demand in balance, as seen with the $7 million in recent targeted fuel subsidies.

We watchPartner and consumer incentives as a percentage of on-demand GMV.

Credit losses in lending

High impact · Medium odds

Financial Services revenue grew, and the portfolio now includes Superbank and Stash. But lending growth brings credit risk. Impairment losses can rise as the loan portfolio expands.

We watchLoan portfolio growth, net impairment losses, and the H2 2026 adjusted EBITDA breakeven target.

Regulatory drag

Medium impact · High odds

Grab operates in many countries, and each market can change rules on commissions, driver treatment, or deals. Indonesia capped some driver commissions, while antitrust regulators are reviewing the pending Foodpanda Taiwan deal.

We watchFoodpanda Taiwan approval, Indonesia driver rules, and antitrust developments.

Foreign exchange pressure

Medium impact · High odds

Grab earns mostly in Southeast Asian currencies but reports in U.S. dollars. The company expects a 2% to 3% structural headwind in the second half of 2026. If local currencies weaken, reported growth looks worse.

We watchReported revenue growth versus constant-currency growth commentary.

AI infrastructure costs

Medium impact · Medium odds

To power its new tools, Grab is stepping up investments in AI tokenization and cloud capacity. This raises regional corporate costs and puts near-term pressure on margins.

We watchCorporate cost trends and the cost per AI interaction.
06 Quick answers

In one breath

Is Grab profitable?

Yes, Grab reported a net profit for 2025 after prior years of losses. The more useful test is whether it can keep growing adjusted EBITDA while using fewer incentives per dollar of GMV.

What is Grab's biggest business?

Deliveries is the largest revenue segment. In 2025, it produced $1.800 billion of revenue, followed by Mobility at $1.219 billion and Financial Services at $347 million.

Why does Grab care so much about AI?

AI can improve the marketplace without only using discounts. Turbo helped driver earnings per online hour, while Mai helped merchants raise GMV for engaged users.

What is the main risk for GRAB stock?

The main risk is that the business improves but the stock already prices in too much success. Watch incentives, Financial Services losses, regulatory news, and foreign exchange.

Get started with Finn today