The first wave of SEA fintech was wallets. The second was BNPL. Both worked because they did not require anyone to underwrite anything new — they just rewired existing flows. The third act is harder, slower, and where the actual return on capital sits: credit.
Roughly half a billion adults in the region sit outside formal credit bureaus. The banks priced them out because the cost-to-serve never penciled. The first generation of digital lenders tried to brute-force it with social-graph signals and lost a generation of LP money. The current cohort is different.
What changed
The new lenders are embedded. They underwrite inside a marketplace, a payroll stack, or a B2B order ledger where the signal is the transaction itself — not a proxy for it. Their default curves look more like trade credit than consumer lending, and they are quietly compounding.
"Underwriting the invisible is not a product. It is a fifteen-year build with a per-cohort feedback loop."
The honest read: this is not a venture story anymore, it is a balance-sheet story. The teams that survive will look more like Asia-native specialty finance than like SaaS.
Dispatch Newsletter
One operator-grade breakdown every Tuesday. SEA tech, AI economics, execution frameworks. No fluff.
