When a captive creates a taxable presence for its parent

Short answer: every other question a GCC faces adjusts its own markup. This one makes the parent taxable in India, on profits attributable to India. And every test is applied to what actually happens, not to what the intercompany agreement says.

Why this is the bigger number

A transfer pricing adjustment moves a markup by a few points on a cost base. It is unwelcome and it is bounded.

A permanent establishment finding is a different category of problem:

• the parent becomes taxable in India, not just the captive;
• the measure is profits attributable to the Indian operation, tested against the parent's India-linked revenues rather than the captive's costs;
• it brings filing and compliance obligations for the parent; and
• it generally surfaces for several years at once, because the facts that create it are rarely confined to one.

That asymmetry is why the PE question deserves attention out of proportion to how often it is raised.

The three routes in

The applicable treaty matters — thresholds and definitions differ between them — so the parent's jurisdiction is part of the analysis rather than a detail.

Conduct beats recitals, every time

The most common reassurance a group gives itself is that the agreement describes the captive as an independent service provider bearing its own risks.

That recital does no work on its own. Each test asks a factual question:

• Is space actually at the parent's disposal?
• Do its people actually render services here, and for how long?
• Does someone here actually conclude or negotiate contracts that bind it?

Worse, an agreement that describes an arrangement the facts contradict is weaker than no agreement, because the contradiction is itself evidence. A services agreement reciting arm's length independence, alongside appraisals running to the parent and commercial terms being settled in Bengaluru, does not present as two consistent things.

The arm's length defence, and its limit

There is a well-known argument that where the Indian entity is remunerated at arm's length for the functions it performs, little or no further profit is attributable to any permanent establishment — because the return for the Indian activity has already been taxed in the captive's hands.

It is a real argument and it has succeeded. It also has a precise limit, which is frequently missed:

It works only where the transfer pricing analysis actually captures everything the Indian operation does.

If functions are being performed for the parent that the markup was never priced for — business development, decision-making, customer-facing work that nobody put in the cost base — then the premise of the defence fails, and it fails exactly where the exposure is largest. The defence is only as good as the functional analysis behind it, which makes the transfer pricing documentation a PE document as well as a TP one.

What to look at, in order

1. Contract authority. Who signs, who negotiates, whose email finalises commercial terms. This is the most establishable of the three routes and the hardest to explain away, so it is where to start.
2. The functional analysis. Does it describe what the Indian team actually does today, or what it did when the entity was set up?
3. Visit patterns. Regular structured presence of parent personnel, and what they do while here.
4. Space. Whether anything is consistently at the parent's disposal.
5. Reporting lines, which bear on this and on the secondment question — and the two should not be answered inconsistently, because control is central to both.

Point 5 is where groups most often trip themselves up. Arguing that the parent controls seconded staff, in order to explain a payroll arrangement, sits awkwardly beside arguing that the parent has no presence in India. Both answers may be individually arguable; together they are a problem.

For the whole picture, see GCC and captive unit taxation.

The permanent establishment question is decided on facts; the transfer pricing filings run every year regardless of how it is answered. Those are covered in what a captive actually files.

Where an adjustment of ₹1 crore or more does follow, the cash has ninety days to reach India before interest starts running on it. The mechanics are in secondary adjustment under Section 92CE, now Section 170.

This is a working reference, not the statute, and outcomes here are fact-specific and treaty-specific. For anything you are relying on, confirm the position directly.