Multilateral netting is a treasury process in which three or more companies in the same group offset what they owe each other through a central netting centre, so each participant makes or receives one net payment per cycle instead of settling every intercompany invoice separately. DBS describes netting as a way to cut intercompany settlement down to one payment per month on a common settlement date. The result is fewer cross-border wires, fewer currency conversions, and a cleaner month-end close.
​

This article covers corporate treasury netting between subsidiaries: how the netting cycle runs, a worked example with gross versus net flows, how FX netting fits in, how multilateral differs from bilateral netting, where local rules restrict it, and which software vendors sell netting modules. Readers looking for the onchain version, where settlement happens in stablecoins, can go to stablecoin netting and clearing instead.
​

Multilateral netting is the offsetting of intercompany payables and receivables across three or more group entities through one central hub. Each entity reports what it owes and is owed, the hub calculates a single net position per entity, and only those net amounts move. Offsetting between just two entities is bilateral netting, a simpler and more limited version.
​

The central hub is usually called the netting centre. It is often a function inside group treasury or a treasury subsidiary, and it acts as the counterparty to every participant for the cycle. Subsidiaries stop paying each other directly. They pay the centre if they are net payers and receive from the centre if they are net receivers.
​

The flows that go through a netting system are typically trade-related: intercompany sales of goods, management fees, royalties, shared-service recharges, and similar invoices. The German treasury software firm Technosis, which sells a product called ATAQ Netting, frames the goal as optimising intercompany payment flows at both a national and international level, bundling all receivables and payables into a single cash flow amount per participant.
​

Netting is distinct from cash pooling, although the two often sit side by side. Cash pooling concentrates balances to fund deficits and invest surpluses. Netting settles invoices. Many groups run netting through an in-house bank, where the net amounts are booked to intercompany current accounts rather than paid out externally at all.
​

Multilateral netting works in a fixed monthly or periodic cycle. Subsidiaries submit intercompany invoices by a cut-off, the netting centre reconciles payer and payee records, disputes get resolved, FX is dealt on the net currency positions, and every entity then settles one net amount with the centre on a shared settlement date.
​

Treasury Today breaks the cycle into four phases: data submission, reconciliation, information distribution, and settlement. In the first phase, participants forward details of all intercompany invoices to the netting centre before the cut-off date. In reconciliation, the netting software matches what each payer offers to pay against what each payee expects to receive, and someone investigates the discrepancies. The centre then publishes each subsidiary's netted balance, often in its own operating currency. Finally, all credits and debits collapse into a single amount per subsidiary and payments are released.
​

DBS lays out a sample calendar for the same flow, spread across roughly a week:

The gap between reporting and settlement matters. It gives subsidiaries time to dispute invoices before cash moves, which is where much of the value sits. Netting forces both sides of every intercompany invoice to agree on the amount, so mismatches that would otherwise surface as reconciliation breaks at month-end get caught inside the cycle.
​

Once settlement is done, the netting system generates posting data so each subsidiary can clear the underlying open items in its ERP. Technosis notes that its system connects to SAP via RFC for exactly this purpose, clearing customer and vendor items automatically after the cycle closes.
​

A worked example shows how multilateral netting compresses flows. Four illustrative subsidiaries with eight intercompany invoices between them would make eight gross payments. Bilateral netting cuts that to six. Multilateral netting through a centre leaves only three payments and moves a small fraction of the gross value.
​

The figures below are illustrative only, not drawn from any real company. Assume a group with four subsidiaries: US, UK, Germany (DE), and Japan (JP). For simplicity, every invoice has already been converted to US dollar equivalents. At month-end, the open intercompany invoices look like this.
​

Without netting, each invoice is paid on its own. That means eight payments with a combined value of 2,500. Most of them cross a border and several need a currency conversion, each carrying bank fees and an FX spread.
​

With bilateral netting, each pair of entities offsets what they owe each other. US and DE owe each other 300 and 600, so DE pays US a net 300. UK and JP owe each other 200 and 150, so UK pays JP a net 50. The other four pairs have flows in one direction only and stay as they are. The result is six payments totalling 1,600.
​

With multilateral netting, the centre adds up each entity's total receivables minus its total payables across all counterparties:
​

The net positions sum to zero, which is the check every netting run must pass. Only three payments occur: JP pays the centre 100, and the centre pays US and UK 50 each. Total value moved falls from 2,500 gross to 200, and Germany does not touch its bank account for intercompany settlement at all this cycle. Every invoice is still settled in the books. Only the cash movement shrinks.
​

FX netting means the netting centre aggregates every subsidiary's currency needs after offsetting, then trades only the net amount of each currency pair. Rather than each entity buying and selling currency for individual invoices, the group executes fewer, larger FX trades centrally, which cuts spread costs and gives treasury one consolidated view of currency exposure.
​

In the example above, all flows were shown in dollars for simplicity. In practice, the UK entity invoices in sterling, Germany in euros, and Japan in yen. A common design is for each subsidiary to pay or receive in its own functional currency, with the netting centre absorbing the conversion. Treasury Today notes that the centre typically shows each subsidiary its netted balance in its operating currency, which keeps FX risk off the local books.
​

Kyriba's case study on Wahl Clipper illustrates the mechanics. The hair clipper manufacturer, which Kyriba says operates in 165 countries with six manufacturing facilities and 11 sales offices, set up two netting centres, one in Sterling, Illinois and a regional centre in China. Intercompany payables and receivables are collected mid-month from each business unit, matched, and disputes resolved. The netting output then shows treasury what it needs to buy and sell in each currency across the group, and trades are executed through 360T. The company reported fewer FX trades and lower currency costs as a result.
​

The dealing step is where timing matters. DBS places FX execution on a dedicated dealing day after confirmation, so trades happen once net positions are final and are not reopened by late invoice disputes.
​

Bilateral netting offsets obligations between two entities only, while multilateral netting offsets obligations across many entities at once through a central hub. Bilateral netting is simple and needs no central function, but it leaves more payments and more gross value moving. Multilateral netting compresses flows further but needs a netting centre, shared cut-offs, and software.
​

Offsetting receivables and payables with a single business partner inside an ERP is effectively bilateral netting. Groups that want to net across the whole entity map tend to add a dedicated netting module or a treasury management system on top.
​

The main benefits of multilateral netting are fewer payments, lower bank and FX costs, less float lost in transit, stronger intercompany reconciliation, and centralised visibility of currency exposure. The main risks are regulatory, since some countries restrict net settlement, plus operational dependence on a timely cycle and clean invoice data from every subsidiary.
​

Treasury Today's list of netting benefits includes less administrative work and fewer payment errors, known payment dates and better payment discipline, lower payment and interest costs, bank account consolidation, hedging efficiencies, and reduced float and FX exposure. DBS adds reduced fraud and error risk and economies of scale in foreign exchange.
​

The reconciliation benefit is often underrated. Zanders documented a netting project at Inmarsat, which settles intercompany netting through SAP In-House Cash. After a configuration fix, auto-matching of intercompany inflows against open invoices rose from 14 to 16 percent to 85 percent, and Zanders reports it removed 3 to 4 hours of monthly manual payment allocation per FTE during month-end close.
​

On the risk side, netting shifts work rather than removing it. Every subsidiary must submit accurate invoices on time, and a single late or disputed invoice can hold up confirmation for everyone. Treasury Today notes that even groups that outsource netting management keep roughly 20% of the workload in-house. There is also concentration risk: the netting centre becomes a single point of failure for intercompany settlement, so its controls and segregation of duties need to hold up to audit scrutiny.
​

Multilateral netting is allowed without restriction in many major markets, but countries with exchange controls often limit or condition it. Common constraints include central bank approval, limits on which flows can be offset, or a requirement to settle gross. Treasury teams usually handle restricted entities with gross-in, gross-out settlement or by excluding them from the cycle.
​

Treasury Today lists Hong Kong, Singapore, Australia, New Zealand, and Japan as markets without netting restrictions in Asia-Pacific. The same source notes that Thailand allows netting only with restrictions, Indonesia requires central bank permission, and the Philippines imposes notably heavy restrictions. Ripple Treasury (formerly GTreasury) publishes a netting restrictions reference table that sorts countries into full netting, netting with restrictions, and central bank approval required.
​

Where net settlement is not permitted, groups can still run the entity through the netting cycle for reconciliation and FX purposes, then settle gross. DBS describes this gross-in, gross-out structure as two flows, one in and one out, instead of a single net payment. The entity keeps the operational benefits of the shared calendar and matching, while each leg complies with local rules. DBS also flags that exchange controls can introduce payment delay risk.
​

India is a useful example of rules changing. The Reserve Bank of India's Foreign Exchange Management (Export and Import of Goods and Services) Regulations, 2026, notified on 13 January 2026 and effective 1 October 2026 according to EY India, now allow export receivables for goods to be set off against import payables for services, and vice versa. Set-off still runs through the Authorised Dealer bank and is tied to counterparty relationships, so it is not the same as full participation in a group netting cycle. Treasury teams should confirm treatment with their AD bank and local counsel.
​

China manages cross-border group flows mainly through regulated cash pooling structures. As of 16 September 2026, the cross-border cash pooling pilot expanded nationwide from its earlier base in Beijing and Guangdong, with more than 260 multinationals registered by 30 June 2026, according to Global Times. Whether a given structure permits net settlement of intercompany trade flows depends on the specific registration and bank arrangement, so groups usually treat Chinese entities as a separate design decision.
​

Multilateral netting software is usually either a module inside a treasury management system or a specialist netting application. Both collect intercompany invoices from ERPs, run matching and dispute workflows, calculate net positions, feed FX dealing, and generate payment and posting files. The right choice depends on entity count, ERP landscape, and whether an in-house bank already exists.
​

For a broader comparison of treasury platforms, including which ones bundle netting with cash positioning and in-house banking, see best treasury management software for 2026. Smaller groups with a handful of entities often start with ERP-level AR/AP offsetting and a spreadsheet calendar, then move to a dedicated module once FX volume or entity count makes the manual process error-prone.
​

Setting up a netting centre starts with mapping intercompany flows and checking local rules for every entity, then choosing a netting currency policy, settlement calendar, and software. Most groups pilot with entities in unrestricted jurisdictions, prove the reconciliation and FX process, and then extend the cycle to more subsidiaries, using gross settlement where required.
​

A practical rollout sequence looks like this:

Groups already running an in-house bank can book net amounts to intercompany accounts instead of making external payments, which removes even the remaining wires for many entities. That combination of netting plus in-house banking is the model Inmarsat uses through SAP In-House Cash.
​

Some treasury teams are also exploring whether the same netting logic can run on shared ledgers, with net positions settled in stablecoins rather than through correspondent banks. Eco covers that model separately in stablecoin treasury netting explained, for readers comparing the onchain approach with a traditional netting centre.
​