ZBA
A Zero Balance Account (ZBA) is a bank account that automatically maintains a closing balance of zero each day by sweeping surplus funds to, or pulling deficit funds from, a designated concentration master account. ZBAs let multi-entity corporates centralise liquidity without forcing every business unit to share a single operating account.
A Zero Balance Account is a bank account designed to end every business day with a balance of exactly zero. Any credit balance that builds up during the day is swept automatically into a linked master, or concentration, account. Any debit balance is funded from that same master account so the ZBA never goes overdrawn. The account remains fully operational for receipts and payments during the day, but its overnight balance is always nil.
Corporates use ZBAs to keep operational accounts open at the business-unit level while concentrating actual liquidity in one place. A subsidiary, division or location can run its own account for payroll, lockbox collections or supplier payments, yet treasury still sees and controls a single pool of cash at the end of each day.
The setup involves three components: one or more ZBA sub-accounts, a master concentration account and an automated sweep instruction held by the bank. During the day, transactions hit the ZBA as normal. Customer receipts post as credits, supplier payments and payroll runs post as debits. At a cut-off time defined in the bank agreement, the bank calculates the net position of each ZBA and moves that amount to or from the master account.
The transfer is a real movement of funds, not an accounting entry. That movement creates an intercompany position when the ZBA and the master account belong to different legal entities. Treasury teams document this with an intercompany loan agreement, an interest rate reference and a clear audit trail so the relationship stands up to audit and transfer-pricing scrutiny.
The core benefit is liquidity concentration. Without a sweep structure, cash sits idle in dozens or hundreds of operating accounts. With a ZBA structure, the entire balance lands in one place where treasury can invest it, use it to reduce short-term borrowing or fund disbursements elsewhere in the group.
The secondary benefits are operational. Cash visibility improves because treasury only needs to track the master account to know the group position. Bank reconciliation gets cleaner because each ZBA has a predictable zero closing balance, so any non-zero result is an exception worth investigating. Interest income improves because surplus cash is consolidated into a single balance large enough to earn a meaningful yield or offset a credit facility.
Three structures appear most often in practice. The first is the payroll ZBA: a dedicated account funded just-in-time on payroll dates, which keeps employee-facing transactions ring-fenced from general operating cash. The second is the disbursement ZBA: one account per division or country used for supplier payments, funded from the master account as payment files are released. The third is the lockbox or collection ZBA: an account that receives customer payments from a lockbox or virtual-account product and immediately sweeps the balance up to concentration.
Larger groups combine all three. Collection ZBAs feed cash up the structure, the master account holds the consolidated balance, and disbursement and payroll ZBAs draw funds back down on demand. Some structures add a second tier, with regional master accounts feeding a global header account.
Three issues need attention before going live. First, bank charges: every sweep is typically billed, so high-frequency or low-value transfers can erode the interest benefit. Treasury should model expected sweep volumes and negotiate a tariff that fits the structure. Second, intercompany documentation: each sweep creates a payable or receivable between legal entities, which means a written intercompany loan agreement, an arm's-length interest rate and monthly reconciliation of the intercompany balances. Third, cut-off timing: a ZBA only works if all expected receipts and payments are visible before the bank's sweep cut-off. Late files, manual payments or unexpected card settlements can leave a ZBA short and trigger an automatic draw on the master account that treasury did not plan for.
Tax and regulatory considerations matter too. In some jurisdictions, automated cross-border sweeps face withholding or thin-capitalisation rules, so multi-entity structures usually run domestic ZBA pools that feed into a separately documented cross-border arrangement.
A static ZBA structure delivers the basic concentration benefit. An AI-native treasury platform turns it into a continuously optimised process. Forecasting models predict each ZBA's intra-day position from open receivables, payment runs and historical patterns, so treasury knows before the cut-off which accounts will need funding and which will deliver excess.
Agentic workflows monitor the actual flows during the day, compare them to the forecast and flag variances early enough to act on them. The same workflows post the intercompany entries triggered by each sweep, attach the supporting bank confirmation and reconcile the master account automatically. Over time the platform learns the seasonal and weekly patterns of each entity, sharpens the forecast and reduces the working-capital buffer the group needs to hold against unexpected funding gaps. The result is a ZBA structure that captures the full concentration benefit without burning treasury time on manual sweep monitoring or month-end intercompany clean-up.
A sweep account moves excess funds above a defined target balance into an investment vehicle or a linked account, often overnight. A Zero Balance Account is a specific form of sweep where the target balance is exactly zero and the linked account is a concentration master inside the same banking relationship. Every ZBA is a sweep account, but not every sweep account is a ZBA.
A ZBA physically transfers cash between the sub-account and the master account each day, which creates intercompany payables and receivables. Notional pooling leaves balances where they are and calculates a combined position on paper for interest and overdraft purposes. ZBA gives treasury direct control of the consolidated balance; notional pooling avoids intercompany positions but is restricted or unavailable in several jurisdictions.
Cross-border ZBAs are possible but more complex. Automated sweeps between entities in different countries can trigger withholding tax, transfer-pricing requirements and local cash-pooling restrictions. Most groups run domestic ZBA structures inside each country, then move the net position cross-border through a separately documented intercompany loan or a regulated cash-pooling arrangement.
If an outgoing payment would push the ZBA into overdraft, the bank pulls the shortfall from the master account as part of the sweep mechanism. The payment goes through and an intercompany payable is created from the ZBA entity to the master account entity. If the master account itself lacks funds, the structure falls back on the group's committed credit facility or the payment is rejected, depending on the bank agreement.
Yes. Because each sweep is a real cash movement between legal entities, it creates intercompany loans that must carry an arm's-length interest rate and be documented under transfer-pricing rules. Some jurisdictions also apply thin-capitalisation limits that cap how much intercompany debt an entity can hold. Treasury should align the structure with tax and legal advisors before launch and review it whenever the group restructures.
An AI-native platform forecasts each ZBA's closing position before the bank cut-off using open invoices, scheduled payments and historical flows. It flags accounts that will be short or long, posts the intercompany entries automatically and reconciles the master account in real time. Over time it learns each entity's patterns, tightens the forecast and lowers the precautionary buffer the group needs to hold, turning a static sweep structure into a continuously optimised liquidity engine.
