Why Cross-Border Payments Are Still Broken in 2019

Swift was built in 1973. We landed on the moon in 1969. Yet moving $100 across borders still takes 3-5 days and costs an average of 6% in fees. Something is fundamentally wrong with how we’ve accepted this as normal.

I’ve spent the last two years in the weeds of payment infrastructure — building acquiring pipelines, negotiating with card networks, watching FX spreads eat into every transaction. And the more I see, the more convinced I am that the core problem isn’t technical. It’s structural.

The Correspondent Banking Problem

When you wire money internationally, your bank doesn’t have a direct relationship with the recipient’s bank. Instead, your money hops through a chain of correspondent banks — sometimes 3, sometimes 5 — each taking a cut and adding latency. It’s like routing an email through five different servers, except each server charges you a fee.

The correspondent banking chain. Every hop = fees + latency.

SWIFT’s GPI initiative has improved tracking visibility, but the fundamental architecture remains the same. You can now watch your money sit in each bank’s nostro account. Great. But it’s still sitting.

The average cost of sending $200 internationally is 6.3% — that’s $12.60 gone before the recipient sees a cent. For migrant workers sending remittances home, this isn’t a rounding error. It’s a month of groceries.

Acquirer Limitations in Emerging Markets

We’ve been building payment rails for Southeast Asia and the problems are even more acute here. Traditional acquiring in markets like Indonesia, Vietnam, Thailand — you’re dealing with:

• High domestic interchange rates (often 1.5-2% vs 0.2% in Europe)
• Limited multi-currency settlement options
• T+3 to T+5 settlement cycles
• Compliance requirements that change with little notice

Local payment methods dominate Southeast Asia — GrabPay, GCash, OVO, PromptPay — but no unified rail.

The businesses that suffer most are the ones moving money for people who most need it — migrant workers sending remittances home, small exporters getting paid by international buyers, freelancers working for global clients.

What’s Actually Changing

A few things give me genuine optimism:

Stablecoins for settlement rails. Not consumer payments — settlement. Using USDC or USDT as a bridge asset between corridors makes mathematical sense. No FX risk during settlement, near-instant finality, programmable compliance. The trust problem is real, but it’s solvable at the infrastructure level.

Local payment method aggregation. GrabPay, GCash, OVO, PromptPay — the real payment method innovation is happening at the local level. The infrastructure play is aggregating these rails and exposing them through a single API. That’s what we’re building.

Regulatory sandboxes. MAS in Singapore, OJK in Indonesia — regulators in Southeast Asia are actually moving faster than their Western counterparts on enabling fintech innovation. The sandbox model is working.

The Uncomfortable Truth

The biggest barrier to fixing cross-border payments isn’t technology. It’s the fact that banks make enormous margins on FX and correspondent fees. The incumbents have little incentive to disrupt themselves.

Which is why I think the disruption will come from the edges — from fintech infrastructure companies quietly building better rails underneath the existing system, from stablecoins that make FX a software problem, from local payment networks that bypass correspondent banking entirely for regional corridors.

We’re in the early innings. But for the first time in decades, I think the architecture is actually going to change.