Payment201 argues that the last decade changed the user-facing side of global payments far more than the layer that actually carries money. Payment methods multiplied, APIs opened up, cross-border transfers got faster, and firms such as Stripe, Adyen, Wise, PayPal and stablecoin networks reshaped the payment experience.
The underlying structure, the article says, moved in a different direction. Access at the payment edge is opening up, while the funding and settlement layer is becoming more concentrated.
It points to several figures to frame that view: the global foreign exchange market was running at about $9.6 trillion in daily volume in 2025, the US dollar was involved in roughly 89% of one side of FX trades, and CHIPS in the United States handled more than $2 trillion in dollar payments a day. The institutions that support those flows are not thousands of payment companies, but a much smaller group of global transaction banks, core clearing systems and liquidity providers.
The article reduces that reality to one question: who can connect to another balance sheet. That, in its view, is why correspondent banking still matters.
Global payments move bank liabilities, not just payment messages
Payment201 says many people start with SWIFT, Visa, Mastercard or payment gateways when they think about global payments. Those rails, however, mostly deal with how payment information is sent. The harder question is where the money is.
The piece draws a basic distinction between the internet and finance: the internet moves information, while banking systems move liabilities issued by financial institutions. If a customer holds $1 million in a JPMorgan account, JPMorgan effectively owes that customer $1 million. If a customer at a Nigerian bank sees $1 million in an account, that bank must also hold corresponding dollar assets or dollar positions on its own balance sheet.
Under that logic, the real challenge in cross-border payments is not simply how Bank A tells Bank B to pay $1 million. It is where that $1 million currently sits and how it gets transferred onto another balance sheet.
The article uses a Nigeria-to-China example. A Nigerian bank may be able to send a SWIFT message to pay a Chinese supplier in dollars. But if it has no US branch, is not a direct participant in the core dollar clearing system, and cannot obtain dollar liquidity directly, it still needs a large bank to connect it to the global dollar network. That is where correspondent banking enters.
Payment201 says it is too simplistic to describe correspondent banking as banks helping one another transfer funds. What a global transaction bank really offers is balance sheet access, along with a full set of financial capabilities.
- Clearing access
- Liquidity
- FX
- Payment routing
- Intraday credit
- AML
- Sanctions screening
- Regulatory infrastructure
The article says the hardest part to replicate is institutional trust. APIs can be bought and systems can be built. A separate issue is whether a global bank is willing to take your funds onto its balance sheet, let your clients use its financial network, and absorb your transaction risk.
Its conclusion from that section is blunt: the scarce resource in global payments is not the existence of an interface, but whether a financial institution is willing to connect with you.
Network value depends on nodes and path length, not only country count
Payment201 then shifts to how companies should think about bank accounts and network value. Many firms treat a bank account as an account number with payment and online banking functions attached. Large enterprises, the author says, see something else: what network that account opens.
The useful questions are which currency systems it touches, which clearing systems it reaches, which banks and liquidity providers sit behind it, and which countries and regions it connects to. By that measure, an ordinary bank account is just an account. A top-tier transaction bank account is an entry point into the global financial network.
The article cites JPMorgan disclosure saying its global clearing network connects more than 4,000 correspondent banking partners across more than 160 countries. The value of that figure, it says, is not simply that JPM works with many banks. The value is the network effect.
If Bank A needs to pay Bank B, a standard path may involve several intermediaries. If both banks connect to JPM, the route gets shorter. If both accounts are inside JPM, the payment may even be completed as a book transfer, with balance sheets adjusted inside the bank rather than across multiple institutions.
That matters because every extra node can add fees, compliance checks, data conversion, repair work, settlement time and liquidity usage. One of the most important metrics in global payments, the article argues, is therefore not only coverage but path length.
Entering a new market requires local financial connectivity, not just map coverage
The article says the payment industry often treats market coverage too loosely. Many companies advertise reach across more than 200 countries, but in infrastructure terms coverage and connectivity are not the same thing.
Lighting up a country on a map does not mean a company truly has payment capability in that market. The key issue is not whether one transaction can be collected. It is whether funds can enter the local financial system reliably and complete settlement.
Payment201 lays out three common ways a payment company can enter a new market.
The first is to build local capability directly, including licensing, local teams, bank connections and operational infrastructure. That gives stronger control, but it brings high cost, long timelines and regulatory complexity.
The second is to rely on a large international banking network, such as using a global transaction bank to enter the market. The advantage is stability, compliance strength and funding capacity. The limitation is narrower coverage and less flexibility.
The third is to find a local partner that already has real financial connectivity. The article points to global payment infrastructure companies such as Thunes and Nium, saying their expansion model is not simply about adding another country on a map. What matters is finding direct local banking links, local clearing capability, local-currency liquidity, regulatory understanding and stable settlement capacity.
By that standard, a market is only truly covered when money can reliably come in, go out and settle. The value of a global payment network comes not from the number of flags on a slide deck, but from the financial infrastructure attached to each country.
If a provider says it covers a market but collection goes through several intermediaries, routing is complex and settlement depends on multiple third parties, the commercial value of that coverage is limited. The article says direct connectivity is what matters. In its framing, the old competition was about more countries; the next phase is about deeper connectivity, because payments are a network business rather than a map business.
Chinese banks are pushing further abroad, and correspondent roles are changing rather than disappearing
The piece also folds in an observation about the overseas expansion of Chinese banks. Payment201 says that in July the author spoke over dinner with a deputy executive from the head office of a domestic Chinese bank, who described overseas positioning that was deeper than many people assume.
The author says that when people discuss banks going global, they often think first about opening overseas branches, serving local clients and supporting Chinese companies abroad. But in the global financial network, banks also play another major role: correspondent bank.
According to the article, the banking executive said the institution was advancing correspondent-related business in Afghanistan, Iraq and parts of Africa. These are not markets with the maturity of Europe or the US as financial centers. Even so, they still need links into the international financial system for trade and capital flows.
The article distinguishes between two different operating positions. In some cases, a local branch is needed to establish a local institution, obtain regulatory approval and serve local clients. In others, physical presence is not required. A bank can operate as a liquidity provider or correspondent partner, offering clearing capacity, liquidity in currencies including RMB, cross-border settlement capability and risk management.
The conclusion from that section is that strength in the global financial network is not determined by who opens the most branches. It depends on where a bank sits in the funding network. Some institutions connect local markets. Some provide clearing. Some provide liquidity. Some handle global treasury reallocation. Together, those roles form the infrastructure of current capital flows.
The article also notes that more overseas banks are joining the RMB cross-border clearing system. It mentions Standard Bank as an example of a large regional bank strengthening its connectivity tied to RMB internationalization. The author says the point is not merely the addition of another payment currency, but that more financial nodes are connecting to a new funding network.
From there, Payment201 argues that future competition in global payments will not be limited to links inside the dollar system. It will also center on which institutions can build more efficient and more stable settlement networks across different currency systems. Correspondent banking has not disappeared, in this view. It is evolving into a connection layer between global monetary systems.
Transaction volume is rising while the number of core nodes shrinks
The article then addresses an apparent contradiction. If correspondent banking is so important, why have large banks not kept expanding their correspondent networks?
Its answer is de-risking. Over the past decade and more, many large banks have reduced exposure to higher-risk markets, smaller financial institutions and correspondent relationships with weak commercial value.
Maintaining a correspondent relationship is expensive, the article says. It requires KYC, AML, sanctions screening, transaction monitoring, audit work, data governance, regulatory review and operating resources. Much of that cost is fixed.
For a small market with a population of a few hundred thousand, annual revenue may be limited. Banks then calculate how much income the relationship brings, how much regulatory risk it introduces, and whether the risk-adjusted return makes sense. If returns are small and risk is high, the rational decision may not be to raise fees, but to exit.
Payment201 argues that de-risking changed more than the price of individual payments. It changed the structure of the global financial network. In the past, many banks were directly connected to many global banks. Now the network is becoming one with fewer edges and stronger hubs.
That strengthens the position of banks such as JPM, Citi, HSBC and Standard Chartered because they hold global clearing capacity, multi-currency liquidity, deep correspondent networks and long-established financial trust.
The article describes this as one of the biggest tensions in global payments today: access at the edge is decentralizing, but financial settlement is re-centralizing. More Stripe-like companies exist, more PSPs exist, and payment methods keep expanding. Yet the number of institutions with dollar liquidity, core clearing access, global balance sheets and trusted standing has not expanded at the same pace. Technical barriers have fallen. Institutional trust barriers have not.
The real contest underneath fees is balance sheet efficiency
Payment201 next turns to what banks and large financial institutions care about most: the cost of capital tied up in payments.
The article asks a simple question. If a bank needs to complete $10 billion in payments each day, what matters more: saving $1 on each wire, or avoiding the need to lock up an extra $2 billion in cash? For large financial institutions, it says, the second issue usually matters more because capital itself has a cost.
That is why, in its view, many common measures used to judge payment infrastructure only scratch the surface. Fees, settlement speed and API stability matter, but at the level of global transaction banks the real competition is liquidity efficiency.
The article returns to CHIPS, saying its significance is not only that it processed more than $2 trillion in dollar payments per day on average in 2025. What matters is how limited liquidity can support settlement volumes far larger than the cash actually parked for the day.
A bank that needs to process $10 billion in daily payments does not necessarily want to prefund $10 billion in cash. That money could otherwise be lent out, invested, used for market making, used to support other clients or held as a group liquidity buffer.
If large sums sit for long periods in nostro accounts, settlement accounts or prefunding accounts, asset utilization declines. That is why major payment infrastructure focuses on prefunding cost, netting efficiency, intraday liquidity, settlement finality and balance sheet usage.
Payment201 says fintech companies and transaction banks are often looking at two different worlds when they talk about payments. Fintech looks at APIs, conversion, checkout and payment fees. Transaction banks look at liquidity, balance sheet management, settlement and risk.
The article states plainly that banks are the principal side in cross-border payments because the visible business earns fees, while the deeper competition is over balance sheet access. Once transaction volumes reach the billions or tens of billions, reducing capital usage by a few percentage points can matter more than shaving a few basis points off the fee line. At that level, payments start to look more like treasury than just payment processing.
In emerging markets, collection is often easier than exit
The article treats Africa as one of the clearest examples of this mismatch. Global attention over the past few years has centered on African payment innovation: mobile money, wallets, instant payments, QR payments and local acquiring.
Payment201 says the picture looks different from the perspective of a multinational treasurer. Imagine a Chinese company operating in Africa. Consumers may pay in Kenyan shillings, Nigerian naira or South African rand. The payment succeeds, the order closes and the consumer experience looks strong.
That is where the harder part starts for headquarters. The company now has to decide what to do with those local currencies, whether they can be converted into dollars, at what FX rate, whether enough dollar liquidity exists, whether foreign exchange restrictions apply, whether funds can legally leave the country, where the money should end up, whether in Hong Kong, Singapore or a treasury center, and whether the local bank has stable correspondent relationships.
The article’s point is that a successful payment does not mean the capital cycle is complete. Many payment companies focus on the first leg, while the more difficult issue is how local currency re-enters the global financial system. Collection is the first mile. Treasury exit is the deep-water section.
It gives a simple example. A wallet company may have 10 million users and process large daily volumes in local currency. From the consumer side, that looks like a huge payment platform. But if an enterprise client asks to convert the equivalent of $10 million in local currency into dollars each day and send that back to headquarters, the problem shifts immediately to who provides the FX, who provides dollar liquidity, who carries the local-currency inventory, who connects to international banks and who completes final settlement.
That is no longer just a payment problem, the author says. It is a financial infrastructure problem. This is why any serious look at African payments eventually reaches banks such as Standard Bank, Standard Chartered, Citi, HSBC and other large regional institutions. Wallets solve how consumers pay. Transaction banks solve how local money gets back into the global funding network.
Stablecoins improve settlement, but they do not automatically solve liquidity
On stablecoins, the article takes a more segmented view. Payment201 says stablecoins have been one of the biggest variables affecting traditional cross-border payments in recent years. They brought around-the-clock settlement, faster movement of funds, less dependence on cut-off times, fewer intermediaries in some cases, and more programmability.
Even so, the author does not think stablecoins simply replace correspondent banking. The reason is that correspondent banking contains several layers, and stablecoins do not solve them all.
The first layer is transport: how value moves from A to B. Stablecoins are strong here. The second layer is the settlement asset itself, meaning what asset both sides use to settle. Stablecoins also have obvious advantages there.
The third layer is liquidity and FX, and the article says this is a different problem entirely. If a company holds NGN 1 billion and wants to obtain $10 million in USDC, the real questions are who is willing to buy the NGN, at what price, who will warehouse the NGN, who supplies the dollars and who takes the FX risk. Those are not problems blockchain technology solves on its own. They are market structure problems.
The fourth layer is access and compliance: who is allowed into the system, whether funds are lawful, whether regulators permit the transaction and whether local rules are met. For that reason, the article says stablecoins are changing the settlement layer, but they do not automatically resolve the liquidity layer.
Payment201 also warns against a common mistake in current stablecoin discussions: treating settlement innovation as if it were liquidity innovation. Blockchain can move an asset quickly. It does not create liquidity in another currency by itself. USDC may travel from Singapore to Dubai in seconds. But the conversion of NGN 1 billion into USDC, and then the final conversion of USDC into RMB, still depends on who provides the intermediate liquidity.
The article therefore frames stablecoins less as a replacement for the financial system and more as a redesign of the settlement layer inside it. It points to BIS-backed Project Agorá, a SWIFT-led ledger, Citi token service and JPMD as examples of efforts aimed at making bank assets move more efficiently.
In the model laid out by the author, banks will continue to provide balance sheet capacity, liquidity, FX, compliance and trust, while newer technology handles faster asset movement, lower settlement friction and greater automation. Banking relationships remain. Settlement rails are what get redesigned.
The scarce resource is connection
The article closes by returning to the figures at the start: $9.6 trillion in daily FX volume, dollar participation at close to 90%, and more than $2 trillion in daily CHIPS payments. The key nodes that connect those flows, it says, are still few in number.
At the same time, payment methods keep multiplying, APIs are becoming more open, stablecoins are maturing and cross-border payments are getting faster. That means the competition in global payments is no longer only about how money moves. It is about who provides liquidity, who has the balance sheet, who carries the risk, who has access and who is trusted.
Payment201 reduces the next stage of global payments to three words: liquidity, trust and access. Without liquidity, technology cannot complete value exchange. Without trust between financial institutions, speed alone does not bring adoption into the mainstream system. Without access to core currency systems, even a strong product remains limited to local markets.
The article adds that stablecoins, tokenized deposits and instant payment networks are all likely to reshape payments over the next decade, but more as a redesign of the global funding network than as a force that makes the network disappear.
Its final conclusion is that the central shift is not tech companies replacing banks. It is payment companies becoming more bank-like and banks becoming more technology-driven. Stripe is moving closer to financial infrastructure. Banks are moving closer to tech companies. Both sides are competing for the same capability: moving money across different financial systems with the lowest funding cost, the shortest path and the highest certainty.
Payment methods are only the surface, the article says. APIs are only the entry point. The forces that actually shape the global payment order are balance sheet capacity, liquidity and financial connectivity. That is why correspondent banking, more than a century after it emerged, still deserves close study.