How much related-party debt is too much before interest is denied?
The test is structural. Legislation sets a permitted proportion of related-party debt to equity, and interest on the excess portion is denied — not all of the interest, and not by reference to what the borrowing costs. Check the permitted proportion, the definition of equity and the measurement dates in the rules that apply to the borrower for the year in question, because all three differ between countries and all three move the answer. The useful way to think about it is capacity: for a given equity base there is a quantum of related-party debt the company can carry deductibly, and funding is far cheaper to plan before the loan is drawn than to unwind afterwards.
Does charging a commercial interest rate protect the deduction?
No, and this is the point most often missed. There are two separate questions about an intercompany loan: whether the rate is what an unrelated lender would have charged, and whether the amount of related-party debt sits within the permitted proportion of equity. The first is a transfer pricing question, the second is a structural one, and a loan can pass one and fail the other. A well-benchmarked rate on a company funded almost entirely by related-party debt still has interest denied, because the rules test how the company is capitalised rather than what its capital costs.
What counts as equity for a thin capitalization test?
It is a statutory figure, not the equity line in the accounts. Broadly the components are contributed capital and accumulated retained earnings, adjusted as the legislation directs, and it is usually measured as an average or at specified points in the year rather than at the year end. Both features catch people out. A distribution or a share redemption shortly before the year end reduces the base the test uses, and so can a loss year, which means deductible capacity falls at the moment the company is least able to respond. Establish the equity figure on the statutory definition before deciding what related-party debt the company can carry.
What happens to the interest that gets denied?
In most systems the interest on the excess is denied outright rather than carried forward, so the deduction is lost rather than deferred. Some go further and treat the denied amount as a distribution to the lender for withholding purposes, which means one payment produces two costs: no deduction for the borrower, and tax withheld on the payment. That combination is what makes an error here expensive relative to its size. It also reaches the payer's reporting obligations, so a denial identified after the year end usually means correcting what was reported at source as well as the return.
Does a parent standing behind a bank loan make it related-party debt?
It can. Back-to-back rules exist precisely to stop the structural test being sidestepped by routing funds through an unrelated lender. Where a third-party loan is supported by a related-party deposit, security or other credit support, the arrangement can be treated as related-party debt to the extent of that support. The question to ask of any facility is where the money and the risk actually sit: who would bear the loss on default, and what the lender relied on in advancing the funds. Support given informally within a group is the common source of the problem, because nobody records it as funding.
Do earnings-based interest limits replace the thin capitalization rules?
They generally sit alongside rather than instead. One limit tests the capital structure, the other tests interest against a measure of earnings, and a company can be comfortably inside one and outside the other: a profitable, heavily related-party-funded subsidiary fails the structural test, while a thinly profitable and modestly geared one fails the earnings test. The deductible amount is the more restrictive outcome, so both have to be modelled, and modelled across the forecast rather than for a single year, because the earnings limit moves with trading while the structural one moves with the balance sheet.
Do I pay tax twice on a foreign dividend?
Not at full rates if the relief is claimed. The paying country usually withholds at source, capped by treaty where one applies and the paperwork is in place; your residence country then taxes the dividend and credits the foreign withholding against its own charge. Where the withholding exceeded the treaty rate because no declaration was filed, the excess is recovered from the paying country, not credited at home. See the dividends article.
How do you avoid double taxation?
You claim relief once, in the right country, in the right order. Usually the source country taxes first, the residence country then gives a credit for that tax against its own charge on the same income, and a treaty caps the source-country rate. Getting the order wrong is what produces a double charge you then have to unwind. The mechanism differs by income type, which is why we map the whole position before filing either return. See how to avoid double taxation.
