Why This Problem Is So Hard to Solve

Payment systems have extremely strong network effects.

SWIFT connects over 11,000 financial institutions across more than 200 countries. This network is its primary competitive advantage.

The math problem for alternatives:

Let's say you build a better payment system—faster, cheaper, more transparent. You convince 100 banks to join. Impressive!

But here's the problem: your 100-bank network can only send payments among those 100 banks. The other 10,900 banks in the SWIFT network? Unreachable.

If Maria's mother's bank isn't in your network, Maria can't use your system to send money. It doesn't matter how much better your technology is.

Network coverage beats technology every time.

Every network faces a "cold start" problem: the first users get no network value. Why would Bank #1 join your payment network when there's nobody to send payments to?

SWIFT solved this decades ago. Building that initial network took years of persuasion, relationship-building, and the credibility of major central banks backing the project.

A new entrant must somehow recreate this—while competing against an established network that already reaches everywhere.

Payment systems are two-sided markets: you need both senders and receivers. Even harder than single-sided networks.

To get senders, you need receivers. To get receivers, you need senders. Both sides must be built simultaneously—a chicken-and-egg problem with no easy solution.

None of these approaches has successfully challenged SWIFT's global reach.

Moving money isn't like moving information. It's heavily regulated because:

• Crime prevention: Payments are used for money laundering, terrorism financing, sanctions evasion
• Consumer protection: People can lose their life savings to payment fraud
• Financial stability: Payment system failures can cascade into banking crises
• Tax enforcement: Payments reveal economic activity that governments want to tax

These aren't theoretical concerns. They're why regulations exist, and why regulators are skeptical of new payment systems.

International payments cross borders—each with its own regulatory framework.

To operate globally, a payment system must comply with:

• Banking regulations (in each country)
• Money transmission laws (often state-by-state in the US)
• Currency controls (some countries restrict foreign exchange)
• Anti-money-laundering (AML) rules
• Know Your Customer (KYC) requirements
• Data protection and privacy laws
• Consumer protection regulations
• Tax reporting requirements

Example: Expanding to 50 countries

• 50 regulatory filings (at minimum)
• 50 sets of compliance procedures
• Lawyers familiar with 50 legal systems
• Ongoing reporting to 50 regulators
• The ability to respond to 50 different regulatory changes

SWIFT has spent 50 years building these relationships and capabilities. A startup has... an idea and venture capital.

Large banks don't just comply with regulations—they often help write them. This isn't corruption (usually); it's practical. Regulators need industry expertise to write workable rules.

But it means regulations tend to reflect how incumbents operate. The correspondent banking model is assumed in countless rules. Nostro accounts are recognized regulatory structures. SWIFT messages are the standard format.

New approaches must convince regulators that different is still safe—while incumbents lobby that change creates risk. This is an uphill battle.

Libra was renamed Diem, scaled back dramatically, and eventually sold off in 2022 without ever launching.

The lesson: Even massive corporations can't will their way past regulatory resistance. Any alternative must work with regulators, not around them.

Imagine you're the CEO of a major bank. You're offered a new payment system that could reduce costs. But:

This is a classic coordination game. Everyone might benefit from change, but no individual actor benefits from moving first.

The rational strategy: Wait and see.

But if everyone waits and sees, nothing happens.

Banks have invested billions in systems built around correspondent banking:

• Treasury management software

• Risk management systems

• Reconciliation processes

• Compliance infrastructure

• Staff training and expertise

• Writing off existing investments

• Building new capabilities

• Retraining staff

• Updating compliance procedures

• Managing a transition period where both systems run

Even if the new system is better, the switching costs can exceed the benefits for years.

Bank executives are risk-averse by nature and training. Their incentives favor caution:

As the saying goes: "Nobody ever got fired for buying IBM." In payments: "Nobody ever got fired for using SWIFT."

This asymmetry means banks will stick with incumbents unless the new option is dramatically, obviously, undeniably better—and even then, they'll let someone else go first.

SWIFT isn't passive. When facing potential disruption, they've responded with:

• Faster tracking of payments
• More transparency on fees
• Same-day completion for many corridors
• Launched 2017, now covers 90%+ of SWIFT volume

gpi doesn't solve the nostro problem, but it addresses many customer pain points. It's "good enough" for many users, reducing pressure to switch.

SWIFT's strategy is classic incumbent defense: Make enough improvements to satisfy most customers, raising the bar for alternatives without fundamentally changing the model.

These companies have had real success—billions in transfers, millions of customers. But notice what they've achieved:

Consumer-facing transfers: Wise made it cheaper and easier for individuals to send money internationally. Significant improvement for Maria sending $500 to the Philippines.

Last-mile delivery: Remittance companies improved the experience of getting money to recipients in developing countries.

Developer tools: Stripe made it easier for businesses to accept payments online.

The underlying rails: Wise still uses correspondent banking to move money. They've optimized the customer interface and found clever routing, but they're building on SWIFT infrastructure, not replacing it.

Large-value transfers: Corporate and institutional payments still flow through traditional channels. When a company moves $50 million, they use their bank.

The nostro problem: Fintech companies don't have their own global correspondent network. They partner with banks who have nostro accounts.

Analogy: Fintech built a better taxi app. But the roads (correspondent banking) and the cars (bank infrastructure) remain unchanged.

This isn't failure—these founders often get rich. But it's not disruption. The system absorbs challengers rather than being replaced by them.

Given everything we've discussed, what would a true payments disruptor need?

1. Comparable Network Reach
It must connect enough endpoints that users can actually send payments where they need to go. Not 100 banks—thousands.

2. Regulatory Legitimacy
It must satisfy regulators across major jurisdictions. This means compliance infrastructure, not just good technology.

3. Speed and Cost Advantages
It must be noticeably better—not 10% cheaper, but fundamentally different economics.

4. Settlement Mechanism
It must solve the trust problem differently than pre-funding. If it still requires nostro accounts, the core inefficiency remains.

5. Institutional Credibility
Banks won't adopt technology from an unknown startup. The solution needs backing from entities banks trust.

6. Coexistence Strategy
Realistically, it must work alongside existing infrastructure during a transition period. "Rip and replace" isn't viable.

This is where blockchain technology enters the conversation. The hypothesis:

• Trustless settlement (no pre-funding needed)

• Real-time finality (no waiting for batch processing)

• Low cost (no intermediary extraction)

• Transparency (visible on public ledger)

• Programmable compliance (built into the system)

• Network effects still matter (need adoption)

• Regulatory legitimacy not automatic

• Technical scalability questions

• Volatility of native assets

• Energy and environmental concerns (for some designs)

We'll explore blockchain's potential and limitations in Lesson 5. For now, the key point is: understanding the problem's difficulty is essential before evaluating any solution.

The international payment system is hard to disrupt because it's designed to be hard to disrupt. Network effects, regulatory complexity, and coordination problems form interlocking barriers that have defeated many challengers.

This doesn't mean disruption is impossible—but it means anyone evaluating a potential disruptor should look for evidence that these specific barriers can be overcome, not just that the technology is good.

Network Effect: A phenomenon where a product becomes more valuable as more people use it. Payment systems have strong network effects.

Two-Sided Market: A platform that serves two distinct user groups who provide each other with network benefits. Payment systems serve senders and receivers.

Coordination Problem: A situation where collective action would benefit everyone, but individual incentives prevent anyone from acting first.

Regulatory Capture: When regulatory agencies act in favor of the industries they regulate, often because incumbents help shape the rules.

Switching Costs: The costs (financial, time, effort) of changing from one system to another. High switching costs protect incumbents.

SWIFT gpi: SWIFT's Global Payments Innovation—improvements to tracking, speed, and transparency announced in 2017 to address competitive pressure.

We've established that payments are broken, capital is trapped, and fixing it is incredibly hard. But we haven't yet examined the deepest layer: the trust problem. Why do we need nostro accounts in the first place? What makes sending money fundamentally different from sending email? Lesson 4 explores the trust mechanics that underlie all payment systems—and sets the stage for understanding how blockchain attempts to solve them differently.

Lesson 3 Complete. Continue to Lesson 4: The Trust Problem in Global Payments →