The first time a bank lost my money, I was in Hong Kong and the amount was around fourteen thousand dollars. It was not stolen. Nobody did anything criminal. A payment to a supplier went out, cleared our account, and then existed nowhere. Not in our account, not in theirs. It sat in an intermediary bank somewhere between the sending institution and the receiving one, tagged with a reference field that did not match what the beneficiary bank expected, and there it stayed for eleven days while three institutions each told me, politely, that the issue was on someone else’s side.

Eleven days. In that window I had a factory in Guangdong holding tooling, a shipping slot I was going to lose, and a founder’s very specific realisation that all my planning had been about the wrong things. I had a lawyer. I had an entity. I had a hiring plan. I had a marketing budget with a media agency lined up. I had not spent one hour thinking about how money would physically travel from a customer’s bank to mine, and from mine to a supplier’s.

I have since reviewed a lot of expansion plans, my own and other people’s. Almost every one budgets for legal, hiring and marketing. Almost none budget for money movement. And money movement is the thing that decides how fast you can actually execute.

Nobody discovers this until after they sign the first customer

The pattern is remarkably consistent. A European or American company decides to enter an Asian market. They do the sensible things. They pick a jurisdiction, usually Singapore or Hong Kong, on advice from a corporate services firm. They incorporate. They get a country manager or a distributor. They spend six to nine months on business development. And then they win.

That is when the trouble starts, not before. The first invoice goes out. The client, a large one, has a procurement process. The invoice enters approval, then treasury, then their bank’s compliance review because you are a new counterparty in a new jurisdiction with a name their system has never seen. Nineteen days later, the payment initiates. Then it travels. If it is going through correspondent banking, it may touch two or three institutions on the way, each of which can hold it for its own screening. Add two to five business days on a good route, more if it hits a name-match flag or a public holiday you did not know about.

Meanwhile, the supplier in Shenzhen has a policy: production starts on cleared funds, not on a payment confirmation screenshot. And payroll is Friday.

That gap between “the customer has approved payment” and “I have usable money in the right currency in the right account” is the real cash conversion cycle in cross-border business. It is routinely thirty to sixty days longer than founders model. I have seen teams with strong revenue, real contracts and good margins get into genuine distress purely because the money was in transit rather than in hand. Nothing was wrong with the business. Everything was wrong with the plumbing.

Why the obvious approach fails

The obvious approach is to treat banking as an administrative task you delegate. You hire a corporate services provider, they incorporate the entity, they introduce you to a bank, you sign forms, done. This fails for three structural reasons, and I say this having spent time on the inside of a large international bank, working on lending and personalisation, watching how the machine actually makes decisions.

First, banks do not underwrite your business. They underwrite their own regulatory risk. A brand new entity, foreign shareholders, no local operating history, a business model the relationship manager cannot summarise in one sentence, and revenue expected from three countries: that file is expensive to approve and cheap to decline. Account opening for a fresh offshore entity can run six to twelve weeks, and it can simply end in a no with no meaningful explanation. I have watched founders assume that because they have money, a bank wants them. It does not work that way. The bank is asking whether your account will generate enough revenue to justify the compliance cost of maintaining it. For an early stage entity, often the honest answer is no.

Second, the structure you chose for tax or legal reasons is frequently the worst structure for banking. A holding company in one jurisdiction, an operating entity in another, invoicing clients in a third, with a director who is resident in none of them, looks perfectly rational on a lawyer’s slide and looks like a risk pattern in a compliance system. The people who designed your structure were not thinking about how it would score in an onboarding review, because that was not their job. It becomes your problem, later, at the worst moment.

Third, currency. If you invoice in your home currency and your client pays in theirs, someone is taking the conversion, and by default it is the bank, at a spread you never see quoted. Two to three percent on the exchange rate, plus fixed fees per transfer, plus intermediary deductions, is normal for standard corporate banking. On a business running twenty percent net margin, moving money twice, that is a meaningful share of your profit disappearing into infrastructure you never negotiated. In the years I spent paying Chinese factories for consumer hardware, currency handling was not a finance detail. It was a line item comparable to a junior salary.

What operators do instead

The shift is to treat payments as an operating constraint, on the same level as manufacturing lead time, and to make a small number of decisions before market entry rather than after.

Start with where the entity sits, judged by bankability, not only by tax. Before you incorporate anywhere, talk to two or three banks in that jurisdiction and describe your actual business honestly: shareholders, revenue sources, expected flows, counterparties. Ask them directly whether they would open the account and what they would need. Some will tell you plainly. That conversation, which costs you a few weeks, is worth more than any structuring memo, because a structure that cannot hold a bank account is not a structure, it is a shell.

Then decide who your banking sponsor is. Not which bank, which human. In Asia more than anywhere, the file that gets approved is the file a relationship manager decides to champion internally. I have seen identical applications go different ways because one had someone inside willing to answer the compliance team’s questions. Build that relationship before you need it, meet the person, bring them real information, and give them something to defend. This is not networking, it is risk management.

Then choose your invoicing currency deliberately, and put it in the contract. Whoever holds the currency risk should be whoever can hedge it or absorb it, and in a small company that is usually not you. If you can invoice in your home currency, do it, even at the cost of a small discount, because certainty on 100 percent of the invoice beats optimism on 103 percent. If the client insists on local currency, price the conversion in explicitly rather than discovering it in the settlement.

Then build redundancy. One bank account is a single point of failure. Run at least two rails: a traditional bank for credibility and local settlement, and a modern payments provider for speed and better foreign exchange. Different corridors behave very differently, and the ability to switch route when one is jammed has saved me more than once. Also negotiate payment terms that reflect physics, not politeness. If your money takes forty days to arrive, your terms need a deposit, a milestone, or an invoice date that starts earlier than delivery.

And write the cash timeline into the plan. Not “we expect payment in thirty days” but “customer approval fifteen to twenty days, transfer three to five days, conversion and clearing two days, so ninety days of runway must exist between signing and usable cash.” Then hold enough working capital to survive it. Every expansion budget I would now sign off has a specific line for money movement: banking setup, foreign exchange cost as a percentage of flow, and float.

The decision you cannot defer

Here is the uncomfortable part. Almost every failure I have described is reversible in principle and brutally expensive in practice. Moving an entity after you have signed clients means re-papering contracts. Changing banks after eighteen months means starting onboarding again with the added question of why you are leaving. Renegotiating currency terms with a customer who already agreed a price means giving something up.

So the sequence matters more than the sophistication. Payment infrastructure is not the reward for winning customers. It is the precondition for serving them. In hardware, in healthcare expansion into mainland China, in climate technology, the pattern has been identical: the constraint on speed was never the strategy, and rarely the talent. It was whether the money could move when the business needed it to move.

Spend two weeks on this before you incorporate. It is the least exciting fortnight of an expansion, and the one that decides whether the rest of the plan gets to happen.

I write from twenty years of building businesses between Europe and Asia. If your company is facing this, start a conversation.