Here's a situation that exists in more groups than anyone admits.
The UK parent buys euros on Tuesday. The German subsidiary sells euros on Tuesday. Both go through their own banking relationships, both pay a spread, and nobody in the group knows the other transaction happened.
That's not a treasury function. It's several businesses sharing a logo.
How groups end up here
It's almost always a product of growth rather than a decision.
An acquisition comes with its own bank, its own finance team and its own way of doing things. A new market gets a local entity with local banking because that's what was practical at the time. A subsidiary's finance director has a relationship with a particular manager and sees no reason to change it.
Each of those decisions is locally sensible. Collectively they produce a group with five banking relationships, no consolidated view of currency exposure, and several people independently buying the same currencies at different prices.
What it costs
Duplicated spreads. Each entity converts separately, each at its own rate, each with its own margin. Smaller entities convert smaller amounts and are therefore priced worse — sometimes substantially worse than the parent.
Lost netting. A group's gross currency flows are almost always larger than its net flows. If entities can't offset against each other, the group hedges and converts gross, which costs more than it needs to.
No consolidated exposure view. If no one can state the group's net position by currency, nobody can hedge it properly. The parent may be hedging a position that a subsidiary has already partially offset.
Idle balances. Cash sitting in one entity's currency account while another entity borrows to fund a payment in that same currency. It happens more often than it sounds.
Fragmented control. Multiple payment authorisation processes, multiple sets of beneficiary records, and multiple points of exposure to payment fraud. Fraudsters specifically target the entity with the weakest controls.
What centralisation actually means
The term covers a spectrum, and you don't have to go all the way along it.
Level 1: visibility. All entities report currency flows and balances in a common format on a common cycle. Nothing moves, nothing changes operationally — but someone can finally see the group position. This alone usually identifies netting opportunities nobody knew existed.
Level 2: coordinated execution. Entities keep their own banking but currency conversions are executed centrally or through a single provider at group-negotiated pricing. The small subsidiary gets the parent's rate.
Level 3: a payment factory. One central function executes payments on behalf of all entities. Single platform, consistent controls, consolidated beneficiary management, one authorisation framework.
Level 4: an in-house bank. The centre acts as banker to the group, with intercompany current accounts, internal funding, and all external currency activity handled centrally.
Most mid-sized groups get the overwhelming majority of the benefit from levels 1 and 2, and should probably stop there. Levels 3 and 4 involve real complexity — tax, transfer pricing, local regulatory requirements and systems work — and are justified by scale rather than principle.
Starting at level 1
The first exercise is unglamorous and it is where the value is.
Ask every entity for twelve months of data: currency flows in and out by month, foreign currency balances held, and existing hedges. Build one table for the group.
Three things reliably fall out of it:
Natural offsets nobody had seen, because the two sides sat in different entities
Pricing inconsistency — the same currency bought at materially different rates by different entities in the same week
A net exposure figure that differs significantly from the sum of the parts
That table is the business case. It's also the thing a board can act on without authorising a project.
Intercompany flows deserve their own look
A specific and common inefficiency: intercompany invoices converted individually, in both directions, each one paying a spread.
A UK parent invoicing its German subsidiary in sterling while the subsidiary invoices back in euros, with both conversions happening separately, is paying twice to move money within the same group.
Intercompany netting — settling the net position between entities on a regular cycle rather than invoice by invoice — removes most of that. It needs agreement on process and attention to transfer pricing documentation, but mechanically it isn't complicated.
Where to be careful
Three genuine cautions:
Local requirements vary. Some jurisdictions restrict cross-border cash movements, intercompany lending or centralised payment arrangements. Check before designing anything.
Tax and transfer pricing. Intercompany funding and netting arrangements have tax implications. Involve advisers early rather than retrofitting documentation.
Local relationships have value. A subsidiary's local bank may provide credit, local payment access or regulatory familiarity that a central arrangement can't replace. Centralising currency execution doesn't require abandoning local banking.
The one question for a group board
"What is our net exposure by currency across the group, and who is managing it?"
If the answer involves gathering information from several people over several days, that is the finding. The exposure exists either way; what's missing is anyone's ability to see it.
William Fuller, Co-Founder of Orbis Exchange Group contact on 0203 918 5622
Ben James, Co-Founder of Orbis Exchange Group contact on 0203 918 5621