For years, India’s transfer pricing Safe Harbour regime was like a restaurant menu where everything looked simple until you saw the prices. The concept was attractive: accept a prescribed margin and avoid transfer pricing litigation. But the actual margins were often so high that many taxpayers politely pushed the menu away and chose the longer, costlier route of benchmarking, audits, disputes or Advance Pricing Agreements.

That may be changing now.

From FY 2026-27, India’s new Safe Harbour framework under the Income-tax Act, 2025 and the Income-tax Rules, 2026 gives IT services companies, software development captives, ITeS units, KPO centres and software-related contract R&D centres a materially more practical option. The headline number is hard to ignore: a 15.5% operating profit margin on operating expenses, with the eligibility threshold increased to ₹2,000 crore of aggregate operating revenue for eligible IT services transactions. The government also announced that Safe Harbour for IT services will be approved through an automated, rule-driven process, without the need for a tax officer to examine and accept the application.

This is not a small technical amendment. It can change how Indian captives plan margins, evaluate APA strategy, price intercompany contracts, and close books for the next five years. But, like most favourable tax options, the real question is not “Is it attractive?” The real question is: “Are we eligible, and does it actually suit our business model?”

What Safe Harbour really means in transfer pricing

Transfer pricing rules require related-party transactions to be priced at arm’s length, meaning the price or margin should be comparable to what independent parties would have agreed in similar circumstances. In normal cases, this requires benchmarking, comparable company analysis, method selection, documentation and, sometimes, years of litigation.

Safe Harbour works differently. Under section 167 of the Income-tax Act, 2025, Safe Harbour means circumstances in which the income-tax authorities shall accept the transfer price or deemed income declared by the assessee.

That word “shall” is the commercial value. It converts transfer pricing from a debate into a rule-based outcome, provided the taxpayer satisfies the prescribed conditions.

For a typical Indian captive software development centre, this can mean something very practical. Instead of defending whether its 13.8%, 14.6% or 16.2% cost-plus margin is arm’s length based on comparable data, the company may elect the Safe Harbour route and maintain at least the prescribed 15.5% margin, assuming it qualifies. The tax department should then accept the declared transfer price for that eligible transaction.

What has changed: from narrow and expensive to broader and more realistic

The earlier Safe Harbour regime was often seen as too restrictive for many IT and ITeS taxpayers. For FY 2024-25 and FY 2025-26, the threshold was expanded from ₹200 crore to ₹300 crore for certain service categories, which itself was a useful step.

But the new regime is a much larger reset. The government’s Budget 2026 announcement stated that software development services, IT enabled services, knowledge process outsourcing services and contract R&D services relating to software development would be clubbed under one broad category of Information Technology Services, with a common Safe Harbour margin of 15.5% and a threshold enhanced from ₹300 crore to ₹2,000 crore.

This matters because the old regime often forced taxpayers to first argue classification. Is the entity a software development service provider? Is it ITeS? Is it KPO? Is part of the work contract R&D? Classification disputes could become almost as painful as margin disputes.

The new rules reduce that friction by bringing these connected service models into one IT services bucket, provided the taxpayer satisfies the prescribed conditions. Rule 87 of the Income-tax Rules, 2026 recognises eligible assessees engaged in providing information technology services consisting of software development services, ITeS, KPO, and contract R&D services wholly or partly relating to software development, with insignificant risk, to a non-resident associated enterprise.

That is a meaningful simplification for Indian GCCs and captive centres whose internal service descriptions have evolved over time. A centre that started as a coding unit may now perform analytics support, platform maintenance, product testing and process automation. Under the old regime, tax teams often worried whether the profile had drifted into another category. The new bucket is more commercially aligned with how IT captives actually operate.

The 15.5% margin: favourable, but not automatically right for everyone

The new Safe Harbour condition for IT services is clear: the operating profit margin in relation to operating expenses must be not less than 15.5%, where aggregate operating revenue from the eligible transaction during the tax year does not exceed ₹2,000 crore.

This is why the regime deserves serious attention. A 15.5% cost-plus margin may be more realistic than earlier margins for many limited-risk Indian IT captives. It may also be more efficient than a prolonged TP audit cycle, particularly where the company’s comparable set routinely produces a range not far from this level.

But it is not automatically the best choice.

For example, if an Indian software captive historically earns a stable 13% or 14% mark-up and has strong benchmarking support, electing Safe Harbour at 15.5% may mean voluntarily offering more taxable income in India. That may still be worthwhile if the group values certainty, lower litigation risk and administrative simplicity. But it is a conscious commercial choice, not a default compliance step.

On the other hand, if the company’s normal tested margins are already around 15% to 17%, and the tax team spends every year defending comparables, filters, risk adjustments and working capital adjustments, Safe Harbour may offer a cleaner answer: pay tax on a known margin and stop converting every year-end close into a mini transfer pricing battlefield.

Why this may become a real alternative to APA

For many Indian captives, the Advance Pricing Agreement route has been the gold standard for certainty. APAs are particularly useful where the business model is complex, high-value, strategically important, or where bilateral certainty is needed to avoid double taxation.

But APAs come with cost, effort, time and negotiation. They require detailed functional analysis, extensive filings, discussions with the APA authorities, and in bilateral cases, coordination between competent authorities. For large or complex groups, that cost is justified. For simpler limited-risk IT services companies, the new Safe Harbour regime may now become a serious alternative.

The Budget announcement also recognised this relationship by announcing a fast-track unilateral APA process for IT services, with an endeavour to conclude it within two years, extendable by six months at the taxpayer’s request.

This creates a useful strategic choice. If the taxpayer has a plain-vanilla, low-risk, cost-plus IT services model and can live with 15.5%, Safe Harbour may be faster and less expensive. If the taxpayer has unique intangibles, complex value creation, substantial onshore decision-making, multiple transaction streams, or needs relief from double taxation through treaty mechanisms, APA may remain the better route.

In simple terms: Safe Harbour is like choosing a fixed fare for a familiar route. APA is like negotiating a long-term contract for a more complex journey. Both can be valuable. The wrong choice is assuming they are interchangeable without evaluating facts.

Eligibility is the real gatekeeper

The most important step is not filing Form 49. It is first testing whether the Indian entity is actually eligible.

The rules repeatedly focus on insignificant risk. For IT services, the foreign principal should perform most of the economically significant functions, including conceptualisation, product design and strategic direction. The foreign principal or its other associated enterprises should provide capital, funds, significant assets and intangibles, while the Indian entity is remunerated for the work assigned to it. The Indian entity should work under direct supervision, should not assume economically significant realised risks, and should not have ownership rights over intangibles or outcomes generated during service delivery.

This is where many taxpayers need to be careful.

A contract saying “limited-risk service provider” is not enough. The rules make it clear that if the contract says the foreign principal controls the risk but conduct shows the Indian entity is actually doing so, the contractual terms are not the final determinant.

That is a major practical point. If the Indian team is making product roadmap decisions, controlling key technology architecture, owning customer outcomes, managing strategic risks, or developing valuable intangibles in substance, Safe Harbour eligibility can become questionable. The tax department may accept simplicity only where the facts support simplicity.

The five-year block: attractive, but it creates commitment

One of the most commercially useful features of the new IT services Safe Harbour is the five-year period. Rule 91 provides that once the option is validly exercised for eligible IT services, it continues for five consecutive tax years. The ₹2,000 crore aggregate operating revenue threshold is tested in the first of those five consecutive tax years.

This can be extremely valuable for planning. If a company qualifies in year one, it can build a five-year TP operating model around that certainty. Finance teams can set pricing policies, true-up mechanisms, group reporting models and Indian tax provisions with much greater confidence.

But this also means companies should not elect casually. A five-year choice is useful only if the business model is expected to remain broadly consistent. If the Indian centre is about to move up the value chain, take on entrepreneurial functions, build IP, or absorb new risks, electing Safe Harbour too early may create mismatch later.

The rules do allow withdrawal, but withdrawal has its own consequences. The option cannot be withdrawn after six months from the end of the first tax year, and once withdrawn, the assessee cannot again exercise the option until the five-year period expires.

That is why the evaluation should happen now, not after March 2027 when the year is already over and the numbers are locked.

Form 49 and the automated process: less discretion, not zero compliance

The new procedure requires the assessee to furnish Form No. 49 electronically for the first of the five consecutive tax years, to the Director General of Income-tax (Systems), on or before the due date for furnishing the return of income for that first tax year. The verification of eligibility, eligible transaction and valid exercise of option is to be done electronically, and the taxpayer must be informed of acceptance or rejection within two months from the end of the month in which the option is exercised.

The automated process is welcome. It reduces uncertainty and officer-level subjectivity. But it does not mean companies can treat Safe Harbour as a checkbox.

Form 49 itself asks for details of eligible transactions, operating revenue, associated enterprise details and operating profit margins. The form requires certification, and for IT services the Rules require the form to be certified by the CEO or chairman and managing director and verified by the person authorised to verify the return of income.

That senior certification changes the tone. This is not just a tax department election. It needs CFO, business and leadership alignment.

Documentation still matters

A common misconception is that Safe Harbour eliminates transfer pricing documentation. It does not.

Rule 89(6) states that sections 171 and 172 continue to apply in respect of international transactions even where the assessee exercises the Safe Harbour option.

This means companies still need documentation and accountant reporting. The practical difference is that the TP study should support eligibility, transaction classification, operating margin computation and consistency of conduct with the limited-risk model, rather than only building a traditional comparable-company defence.

This is where the new Safe Harbour and the new Form 48 environment connect. The future of TP compliance is moving toward more structured, data-backed reporting. Safe Harbour may reduce disputes, but it does not excuse weak books, vague agreements or incomplete intercompany reconciliations.

The trade-offs: no comparability adjustment, no MAP

The rules also contain two important limitations.

First, where Safe Harbour is accepted, no comparability adjustment and allowance under section 165(3)(a)(ii) can be made to the accepted transfer price.

Second, Rule 93 provides that once the transfer price for an eligible international transaction is accepted under Safe Harbour, the assessee is not entitled to invoke the Mutual Agreement Procedure under the relevant tax treaty for that transaction.

This is important for multinational groups. If the overseas jurisdiction does not accept the corresponding outcome, or if the group expects a double taxation concern, Safe Harbour may not be the best answer. APA, particularly bilateral APA, may still be preferable where treaty-level certainty matters.

Also, Safe Harbour does not apply where the associated enterprise is located in a country or territory notified under section 176, or in a no-tax or low-tax country or territory.

So the decision is not simply “15.5% looks good, let’s apply.” The decision is: does the whole fact pattern, including jurisdiction, treaty position, group policy and risk allocation, support this route?

Why FY 2026-27 needs action now

Although the filing for FY 2026-27 will happen after the year closes, the Safe Harbour decision cannot wait until then.

A company that wants to elect Safe Harbour must run its books during the year in a way that supports the prescribed margin. That means intercompany invoices, cost base definitions, stock-based compensation treatment, operating and non-operating classification, foreign exchange treatment, true-up mechanisms and year-end provisions must be aligned before closing.

A 15.5% Safe Harbour margin is simple only if the accounting is simple. If the Indian captive discovers in April 2027 that its actual operating margin is 14.8%, the discussion becomes awkward. The CFO will ask whether a true-up can be booked. The auditor will ask whether the agreement permits it. The tax team will ask whether the adjustment is defensible. The group controller will ask why this was not modelled earlier. By then, everyone will have a spreadsheet, but nobody will have peace.

The better approach is to evaluate eligibility now, model the margin quarterly, align intercompany agreements, and decide whether Safe Harbour, APA or regular benchmarking is the right route.

A practical decision framework

For a low-risk software development or ITeS captive with stable functions, no local IP ownership, clean cost-plus invoicing and margins already close to 15.5%, Safe Harbour may be highly attractive.

For a high-value development centre with product ownership, complex functions, unique intangibles or significant India-side decision-making, Safe Harbour may be risky because the eligibility test depends heavily on actual conduct.

For a company already in APA discussions, the new regime should not automatically derail the APA strategy. It should, however, trigger a cost-benefit review. If unilateral certainty at 15.5% is enough, Safe Harbour may be simpler. If bilateral certainty or bespoke treatment is needed, APA may remain superior.

For a company below the ₹2,000 crore threshold but with volatile margins, the key question is whether it can operationally maintain 15.5% every year for the five-year block without distorting commercial reality.

Conclusion: Safe Harbour is finally worth a serious boardroom discussion

The new 15.5% Safe Harbour for IT services is one of the most practical transfer pricing developments for Indian captives in recent years. It brings together related IT service categories, expands the threshold to ₹2,000 crore, introduces a five-year framework and moves toward an automated, rule-driven approval process. For many companies, that may convert Safe Harbour from a theoretical option into a real alternative to recurring TP disputes or even unilateral APA.

But favourable does not mean automatic. The real work is in eligibility testing, conduct review, margin modelling and documentation readiness.

FY 2026-27 has already begun. The companies that evaluate Safe Harbour now will treat it as a strategic choice. The companies that wait until after March 2027 may discover that a favourable rule is useful only when the business was prepared to use it.

In transfer pricing, certainty is never free. But this time, for the right IT services captive, it may finally be available at a price worth considering.