Why Grab paid $1.5b for Atome instead of building lending

Last month, Grab made its most expensive bet yet – not on rides or food, but on credit. Acquiring 60% of buy now, pay later player Atome Financial for US$1.5 billion is the company’s acknowledgement that real, scaled, multimarket consumer lending is too slow and difficult to build from scratch.

A few months ago, I wrote about Grab’s three banks and how Superbank – the smallest in the group – is the only one that’s profitable. The question I left open was whether the company could replicate that across its other markets.

Image credit: Ulla

The Atome deal indicates that Grab decided it’s not waiting for an answer.

I call this the “rebundle season.” In the last decade, fintech was about unbundling: one company, one product, one pain point solved better than any bank could.

But unlike in other regions, the platforms in Southeast Asia own the distribution channels. They have built super apps, digital wallets, and ride-hailing networks while integrating restaurants, retailers, and supermarkets into their ecosystems.

Now, the platforms are stacking these products together, creating flywheel effects that are hard, slow, and expensive to replicate.

Opting for M&A

Since going public in 2021, Grab has spent more than US$1.2 billion in adjusted EBITDA losses on its financial services unit. This segment generally has one of the highest margins across industries, but it comes with trade-offs.

While revenue for Grab’s financial services arm grew 38% in 2025, its losses barely moved from US$105 million to US$110 million. That’s the cost of learning how to lend at scale: underwriting customers who have no formal credit history, managing risk across five markets with different regulators, and building the models that separate good borrowers from bad ones.

See also: The one Grab digital bank that’s actually profitable

Atome has already solved this. The BNPL firm’s AI-powered underwriting models have been trained on years of data from consumers without formal bank histories, and its credit losses have stayed low.

For years, platforms like Grab burned cash to build distribution. Now that their core businesses are generating positive EBITDA, they have the firepower to buy what they can’t develop fast enough. At the same time, fintech valuations in Southeast Asia have normalized from their 2021 peaks.

Grab COO Alex Hungate put it plainly: This Atome acquisition lets the super app “leapfrog the timeline” for consumer credit in the Philippines, Indonesia, and Thailand, rather than spending years developing underwriting models market by market.

I expect to see more deals of this kind in the coming quarters, as tighter regulation, higher funding costs, and the need for scale push stand-alone BNPL and lending players to find stronger distribution partners.

Is a BNPL player worth US$1.5 billion?

What the combined business could look like

Where the play works and where it doesn’t

What needs to happen next