International Banking Strategy

Designing a banking structure that holds under pressure.

Why international banking requires a strategy

Domestic banking rarely requires design. One institution handles the payroll, the receipts, the overdraft and the mortgage, and the arrangement works because everything sits inside one legal system, one currency and one settlement environment.

Cross the first border and that stops being true. Currencies multiply. Settlement runs through correspondent chains the client never sees. Compliance obligations arrive from several directions at once. Counterparties in unfamiliar jurisdictions insist on instruments the domestic bank has never issued. The arrangement that worked for fifteen years starts to fail in ways that look operational but are structural.

Most clients respond by opening another account. It is the fastest available answer and almost always the wrong one.

Strategy, in this context, means something narrow and practical. It means deciding what each banking relationship is for before opening it, choosing the jurisdiction on legal and operational grounds rather than proximity, and reviewing the whole structure periodically rather than allowing it to accrete.

Three common structural mistakes

Building complexity instead of structure
Relationships opened without a defined purpose do not add resilience. They add administration. Each new account carries onboarding, ongoing due diligence, periodic review, reporting obligations, signatory management and a further reconciliation line. Beyond a certain point the marginal account consumes more management attention than it returns in capability.

The question to ask of any existing relationship is simple: what would break if it were closed tomorrow? Where the answer is nothing, the relationship is cost without function.

Confusing diversification with relocation
Clients frequently arrive believing that international banking means moving away from domestic banking. It rarely does.

Established domestic relationships often provide access to local payment infrastructure, lending on terms an international institution cannot match, familiarity with the local regulator, and a credit history that took years to build. Abandoning that in favour of a single offshore relationship does not diversify anything. It substitutes one concentration for another, usually a weaker one.
Diversification means holding capability in more than one place, so that the failure, withdrawal or de-risking of any single institution is inconvenient rather than existential. It is additive. Relocation is not.

Selecting banks instead of banking strategies
A bank is a component. The strategy is the assembly.

Choosing an institution before deciding what it is for produces a structure organised around whatever the chosen institution happens to be good at. The correct sequence runs the other way: define the functions the structure must perform, then identify which institutions can perform each of them, then negotiate.

Banking architecture

An architecture assigns function. Six functions cover most internationally active structures.

Operating. Commercial receipts and payments in the currencies the business actually trades in. Judged on transaction throughput, reporting quality and cost per payment.

Treasury. Reserve, working capital and term liquidity, held apart from operating flow so that the position is legible at any moment. Judged on segregation, deposit terms and foreign exchange access.

Settlement. Cross-border execution. Judged on correspondent depth, access to SWIFT GPI, SEPA Instant, RTGS and real-time payment rails, and on the institution’s willingness to explain a delayed payment rather than issue a reference number.

Trade. Documentary credits, standby letters of credit, demand and performance guarantees. Judged on issuance capability and on whether the institution’s paper is accepted by the client’s counterparties.

Custody. Securities, funds and precious metals, segregated from transactional accounts, with an audit trail that survives contact with an auditor.
Structural. Holding companies, special purpose vehicles, escrow and fiduciary arrangements, banked separately so that ring-fencing is real rather than notional.

A structure that maps cleanly onto these six functions is usually smaller than the one it replaces. That is the point.

Jurisdiction selection

Jurisdiction is frequently the more consequential decision, and it is frequently made last.

Each jurisdiction offers a different combination of regulatory standard, legal certainty, tax treatment, banking infrastructure, international standing, investment access and political stability. No jurisdiction optimises all of them. The relevant question is which of those attributes the specific function requires.

Custody demands legal certainty and enforceable property rights above all else. Settlement demands correspondent reach and infrastructure. Trade finance demands that counterparties accept the jurisdiction’s paper. Operating accounts demand proximity to the commercial activity and to the payment systems that serve it.

A jurisdiction that is excellent for one of these can be actively unsuitable for another. Selecting a single jurisdiction for the entire structure is therefore a compromise in every function simultaneously.

Two further considerations are habitually underweighted. The first is reputational standing, which determines how the client’s counterparties, auditors and other banks will treat the arrangement. The second is correspondent stability, which determines whether payments will still clear in three years. Both are more consequential than headline tax treatment, and both are harder to reverse.

Diversification principles

Diversification in banking is not about the number of relationships. It is about the independence of the failure modes.

Three relationships in the same jurisdiction, clearing through the same correspondent, subject to the same regulator, are one relationship in three envelopes. Two relationships in genuinely distinct legal and settlement environments are diversification.

The practical test is a set of questions. If this institution withdrew from the client’s market segment, what would stop working, and how quickly could it be replaced? If this jurisdiction imposed capital controls, what proportion of liquidity would be affected? If this correspondent chain broke, is there a second route for the same payment? If the answers cluster around a single institution or a single jurisdiction, the structure is concentrated regardless of how many accounts it contains.

Redundancy costs money. Concentration costs more, but only occasionally, which is why it is consistently underpriced.

Domestic and international banking

The two are complementary in most cases and competitive in very few.
Domestic banking typically provides local payment infrastructure, credit on relationship terms, familiarity with local regulatory practice and administrative simplicity. International banking typically provides multi-currency capability, cross-border settlement, instruments for international trade, jurisdictional diversification and access to counterparties the domestic institution cannot reach.

A client asked to choose between them has usually been asked the wrong question. The structure should hold both, with each performing what it performs well, and neither depended upon for what it does badly.

Treasury considerations

Treasury is where a poorly designed architecture shows first.
The symptoms are consistent. Liquidity has to be assembled manually across entities and institutions before anyone can state a position. Reconciliation absorbs finance staff who should be forecasting. Foreign exchange happens by default, at the point of payment, at whatever rate the receiving institution applies. Funding pressure in one entity is invisible to the group until it becomes urgent.

None of these are banking product problems. They are structural problems that no product will fix.

A structure designed around function produces the opposite. Position is reported rather than reconstructed. Foreign exchange becomes a decision taken in advance. Intra-group funding is visible before it is needed. The finance function moves from explaining the past to managing the present.

Cross-border payments

Payment execution is the most common point of failure, and the least well understood by clients before it fails.

An international payment traverses a chain of correspondents, each applying its own compliance screening, cut-off times and fee structure. The originating bank’s capability determines only the first step. What determines the outcome is the depth and quality of the correspondent network behind it, and whether the institution will trace a stalled payment on request.

Clients evaluating an institution on payment capability should ask three things. Which rails does it reach directly, and which through intermediaries. What is the escalation path when a payment does not arrive. And who, by name, answers when it is escalated.

Cost matters less than clients expect. Certainty matters considerably more.

Family office considerations

Family offices carry an additional constraint: the structure must survive people.

Principals change. Advisers retire. Beneficiaries inherit. A banking structure that depends on one person’s memory of why an account exists will fail at precisely the moment it is most needed, which is during transition.

The remedy is documentation and separation. Every relationship has a stated function recorded somewhere other than in a founder’s head. Operating, investment, custody, settlement and structural holdings are separated so that responsibility can be delegated without delegating everything. Signatory arrangements are designed for the family as it will be, not as it is.

Sophisticated family offices treat their banking architecture as part of the governance documentation, reviewed on the same cycle as the family constitution. Most do not, and discover the omission during probate.

Corporate structures

Group structures introduce a specific hazard: the banking arrangement drifts out of alignment with the legal one.

Entities are formed, acquired, dormant or wound up, while the accounts attached to them persist. Cash sits in subsidiaries that no longer trade. Guarantees remain outstanding against entities that have been restructured. Intra-group flows run through routes designed for a group configuration that no longer exists.

Alignment is restorable but it requires deliberate work: a map of legal entities against banking relationships, a stated function for each, and closure of what no longer serves one. The exercise is unglamorous and consistently produces the largest single improvement in treasury clarity.

Frequently asked questions

Is more banking relationships safer?

Beyond a small number, no. Additional relationships add administration and compliance surface without adding independent capability, unless each occupies a distinct function or a distinct failure mode.

Should international banking replace domestic banking?

Rarely. Domestic relationships usually retain advantages in local payments, credit and regulatory familiarity that international institutions cannot replicate. Diversification is additive.

What matters most when selecting a jurisdiction?

The requirement of the specific function. Custody needs legal certainty. Settlement needs correspondent reach. Trade needs counterparty acceptance. There is no single best jurisdiction, only jurisdictions that suit particular purposes.

How often should the structure be reviewed?

Annually, and on any material change: new market, acquisition, restructuring, change in ownership, or the loss of a correspondent relationship.

What does IFB actually do here?

We advise on the design and we operate the functions for which we are the appropriate institution. Those are separate things, and we are content to be told the second does not apply.