Manage in-house bank accounts: what each role knows and where it breaks

In an in-house bank, treasury knows the cash positions and the bank mechanics. Subsidiaries know why each payment exists. Group accounting knows how balances must reconcile at close. Trouble starts where these meet: unidentified receipts, disputed interest, payments released on stale data, and intercompany balances that never agree.

What central treasury sees

Treasury runs the accounts. It sees every movement across the internal ledger and the external bank accounts behind it. It knows which subsidiary is long cash and which is borrowing. It knows the interest rates, the fee schedule and the cut-off times at each external bank.

What it usually lacks is context. A payment file arrives from a subsidiary, and treasury can check that the format is valid and the account has funds. It cannot tell whether the supplier is genuine, whether the invoice was approved, or whether the same bill went out last week through a local bank account nobody closed.

What subsidiary finance teams know

Local teams hold the commercial story. They know which customer pays late, which supplier insists on a particular bank, and which local regulation forces some payments outside the central structure. They also know their own forecasts, often better than the figures they send upward.

Their blind spot is the internal account itself. Many local accountants treat the in-house bank statement as something that simply happens to them. They see a balance and an interest charge. They rarely see how the rate was set, why a receipt landed on another entity's account, or when netting will settle.

What group accounting and tax hold

Group accounting owns the chart of accounts and the period-end close. It knows which intercompany accounts must eliminate on consolidation and what happens when they do not. Tax knows the transfer pricing rules that govern internal interest, plus the withholding and thin capitalisation limits that apply in certain countries.

Neither sits close to daily operations. A rate change agreed in treasury may reach tax only after the year has closed.

Where the gaps cause trouble

Incoming payments nobody claims. Customers pay into a central collection account with poor references. Treasury can see the cash but not who it belongs to. Subsidiaries know their open invoices but cannot see the central receipts. Unmatched items sit in a suspense account, and each side assumes the other is chasing them. Returned items such as bounced cheques or reversed direct debits make this worse, because the original credit may already have been passed to a subsidiary.

Interest and fees that surprise people. Treasury calculates interest on internal balances using a method set up long ago. Local controllers budget without knowing it. When the charge arrives, they dispute it, and the conversation turns on value dates and day count conventions that only treasury understands. If tax was never consulted on the margin, the dispute becomes a compliance issue as well.

Intercompany balances that do not match. The in-house bank posts one side of a transaction on its ledger. The subsidiary posts the other side from its statement, sometimes late, sometimes to the wrong account. At close, group accounting finds differences and sends them back to both. Timing is the usual culprit. Treasury closes its period on one basis while subsidiaries close on another, and manual journals written to fix the gap rarely carry enough explanation for anyone to reverse them cleanly.

Netting run on incomplete data. Netting depends on each entity confirming its payables and receivables. When one party confirms from an old aging report, the settlement is wrong for everyone involved, and unwinding it takes far more effort than getting it right first.

Borrowing without a paper trail. Intercompany loans are sometimes agreed by phone between a local finance director and a treasury manager. The cash moves. The loan agreement, the interest terms and the repayment schedule follow later, if at all. Auditors and tax authorities then ask questions nobody can answer quickly.

Questions to ask the people who run it

Documentation for this process tends to describe the design. The people doing the work can describe what actually happens. Useful questions:

• When a receipt arrives that cannot be matched, who owns it, and how does the subsidiary learn it exists?
• Which payments still leave through local bank accounts, and why?
• How was the internal interest rate set, and when did tax last review it?
• Do subsidiaries read the in-house bank statement, or do they post from it without checking?
• What manual journals get written at every close to make intercompany balances agree?
• Which netting confirmations arrive late or need correcting after settlement?
• Who can approve an intercompany loan, and where is the signed agreement kept?
• What does treasury check on a payment file, and what does it assume the subsidiary has already checked?
• If a returned payment hits the central account after the credit was passed on, what happens next?

Ask these of treasury and of a couple of subsidiaries separately. The differences between their answers show where the process is held together by habit and goodwill. Those are the points most likely to fail once anything changes.

APQC's Process Classification Framework® (PCF) is an open standard developed by APQC, a nonprofit that promotes benchmarking and best practices worldwide. To download the full PCF or to view definitions and measures, please visit www.apqc.org/pcf.