What Intercompany Accounting Software Misses

Month end is running late again, intercompany balances don't agree, your finance lead is tracing why, and someone’s suggesting buying intercompany accounting software to fix the problem.

Intercompany accounting software is the category of tools, usually add-ons to your ERP, that match the two sides of every transaction between your entities, flag the pairs that disagree, and remove the internal balances when the group accounts are put together.

It's good at that job. What it can't do is to tell you whether the figure on the invoice was right in the first place.

What the software does for a group with volume

Past a few hundred intercompany invoices a month, matching by hand stops working. Each local finance team posts its side of a recharge or management fee separately, in a different currency and to a different month-end calendar. The software sits across the ledgers, pairs each receivable with its payable, applies the same exchange rate to both sides and lists the differences. Some tools go further: they net the balances so cash moves once a quarter instead of forty times, and they post the elimination journals at consolidation.

If you're consolidating entities out of several ERPs, buying it is probably the right call.

Where the number comes from

The software starts from the invoice. Something upstream decided the mark-up on the management fee or the royalty on the licence. That decision is your transfer pricing policy, and it has to be written into an intragroup agreement before the first invoice goes out. The financial attributes of each agreement then get coded into the accounts.

Matching confirms that the payable in one entity equals the receivable in the other. It has no view on whether the mark-up reflects what the subsidiary actually does, or whether anyone revisited it when the business changed. The test a tax adviser applies is whether an independent company would have agreed to that price, and the ledger can't answer it.

A loan with two balances

A customer of ours has its engineering team in Finland and its sales team in the US (countries changed for privacy). The US entity held the cash, so for a year it paid the Finnish team's software, travel and contractor bills from its own cards and bank accounts. The Finnish accountants did the right thing on their side: every one of those payments went against the intercompany loan account, but nobody told the US side. The US team coded the same payments as its own expenses, so one ledger showed a loan growing every month and the other showed none of it. By year end the balance the parent expected was about three times what the subsidiary's books showed, spread over thousands of transactions, with a statutory audit a few weeks away.

Matching software would have flagged that gap in month one instead of month twelve, and that's the case for it. But the fix here wasn't a matching fix. What mattered was which entity should have been paying for what: development costs sit with the entity that owns the IP, delivery costs with the entity that sells, and once that's settled each entity pays its own bills and there's nothing left to match. The finance lead had been keeping as much cost as possible in the US on advice that was out of date.

Before you buy it

1. Determine the pricing method for each type of charge and write it into an intercompany agreement. Cost plus a mark-up for services, a defensible interest rate on the loan, cost with no mark-up on a pure recharge such as a shared software subscription or a conference the parent paid for. This is a transfer pricing job.
2. Decide which entity bears which cost, and have each entity pay its own bills. Most intercompany volume at this stage is one entity paying another's suppliers: payroll, cloud hosting, travel. Move the payment to the entity that benefits and the intercompany entry disappears.
3. Check both sides against the agreement, not just against each other. Once a quarter, your group financial controller takes the agreement, the invoice and both ledgers and confirms all three say the same thing.

Then buy the software, if the volume that's left justifies it. It makes step three faster, and that's the job it was built for.