Defending the Residual Profit Split Method in India When the TPO Proposes TNMM for an Intangible-Owning Entity

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Transfer Pricing

Introduction

Transfer pricing methods divide into those that test one side of a transaction and those that look at both. A one sided method selects the less complex party, benchmarks its return against comparables, and leaves whatever remains to the other side. It is the workhorse of Indian transfer pricing because it is simpler to apply and easier to verify. It also carries an assumption that is not always true: that only one party to the transaction is doing anything a comparable cannot replicate.

Indian transfer pricing officers rarely accept the profit split method and commonly propose to substitute the transactional net margin method. Defending the choice is not an argument about which method is more favourable. It is an argument about whether the conditions in the rule are satisfied, and it stands or falls on the functional and value chain evidence.

This note deals with what happens when that assumption fails, and with defending the profit split method when a transfer pricing officer proposes to substitute a one sided method in its place. It covers the statutory conditions and their two gateways, the Indian authority and its limits, how economic ownership of an intangible is established, how a residual split is constructed and documented, the high margin problem, and why the case for rejecting the alternative method has to be made affirmatively rather than left implied. The functional evidence on which the whole analysis rests is dealt with in our note on FAR analysis for transfer pricing audit defence.

1. The statutory conditions

Rule 10B(1)(d) of the Income-tax Rules, 1962 (now rule 79(1)(d) of the Income-tax Rules, 2026) provides for the profit split method and identifies the circumstances in which it applies: transactions involving transfer of unique intangibles, or multiple international transactions which are so interrelated that they cannot be evaluated separately for determining the arm’s length price.

Those are the two gateways, and they are disjunctive. The method is available where unique intangibles are involved, and it is available where the transactions are inseparably interrelated, whether or not intangibles are present. A submission should identify which gateway it relies on, and preferably both where the facts support it.

Rule 10C (now rule 80 of the Income-tax Rules, 2026) sets out the factors governing the selection of the most appropriate method, including the nature and class of the transaction, the functions performed, assets employed and risks assumed, the availability and reliability of data, and the degree of comparability. The method is not chosen in the abstract; it is chosen because it is the one that can be applied most reliably on the available data.

Under the Income-tax Act, 2025 the governing provision is section 165 (section 92C of the 1961 Act, determination of arm’s length price), within Chapter X, sections 161 to 177. There is no current administrative guidance on the method. The Central Board’s Circular No. 2/2013 dated 26 March 2013 on the application of the profit split method was withdrawn by Circular No. 5/2013 dated 29 June 2013 and has not been replaced, so a submission cannot rely on Board guidance here and must argue from the rule and the authorities.

2. The Indian authority

The leading tribunal decision is Infogain India Pvt. Ltd. v. DCIT, pronounced by the Delhi Bench on 19 August 2015 for assessment year 2008-09. The Bench held that under Rule 10B(1)(d) the profit split method applies where a transaction involves unique intangibles or where interrelated international transactions cannot be evaluated separately. On the facts the Indian and United States operations were highly integrated and both contributed significantly, so the profit split method was the most appropriate method. Where comparables are unavailable, residual profit may be split on relative contribution measured through key value drivers, consistent with the guidance published by the OECD and the United Nations. A split accepted by the department in adjacent years was directed to be followed.

The decision should be cited for what it says. It did not hold that the profit split method must always displace the transactional net margin method wherever unique intangibles exist. Its reasoning rested principally on the integration of the two operations and the fact that both contributed materially. A submission that overstates the holding gives the Revenue an easy answer.

On the related question of comparability, CIT v. UT Starcom Inc. (India Branch), Delhi High Court, 25 September 2017, is useful in both directions. The Court found no substantial question of law and dismissed the Revenue’s appeal, affirming the Tribunal’s reasoning that a large risk taking company with significant intangibles and assets is not comparable to a limited risk captive. The corollary is that an entity which does own significant intangibles and does carry entrepreneurial risk cannot be benchmarked against routine service providers either, which is often exactly what a proposed transactional net margin analysis does.

3. Why a one sided method breaks down here

The transactional net margin method works by selecting the less complex party as the tested party, finding comparables for it, and setting its return. Everything else falls to the other party by residue. That is a sound approach where the other party is the only one doing anything unusual.

Consider instead a group in which an Indian entity develops and owns a proprietary software platform, and overseas entities distribute it in their markets. If the overseas distributors are routine, they can be tested and the residual accrues to India, which is the correct outcome. A transfer pricing officer who instead tests the Indian developer against routine software service comparables inverts the structure: it caps the developer’s return at a service margin and leaves the residual with distributors that did not create it.

That inversion is the argument. It is not that the profit split method gives a better answer. It is that applying a one sided method with the wrong tested party allocates the return from the intangible to the party that does not own it, which is the opposite of what the arm’s length principle requires.

4. Establishing economic ownership of the intangible

Everything turns on who owns the intangible in economic terms, and that is a DEMPE question: who performs and controls the development, enhancement, maintenance, protection and exploitation functions, who funds them, and who bears the risk.

Legal ownership is the starting point and nothing more. The evidence that matters is operational.

• Development. Where the engineering teams sit, headcount by location, the roadmap and who sets it, who approves releases. Version control and project management records are more persuasive than an organisation chart because they are contemporaneous and were not created for the file.
• Enhancement. Who decides what gets built next, and on whose commercial judgment.
• Maintenance. Who runs the platform, who resolves defects, who carries the operational burden.
• Protection. Who files and maintains the intellectual property, who funds enforcement.
• Exploitation. Who sets pricing and commercial terms, who decides which markets to enter.
• Funding and risk. Who bears development cost, and in particular who absorbs the cost of a project that fails. The party that carries failure risk is the party that should earn the upside.

The corresponding question has to be answered for the overseas entities. Showing that they perform routine distribution functions, do not develop the product, do not carry development risk and operate within commercial parameters set elsewhere is what establishes them as the routine parties.

5. Constructing the split

A residual profit split has two stages, and both have to be documented.

Stage one: the routine return. Each party is allocated a return for its ordinary functions, benchmarked against comparables performing those functions. For a distributor this is a distribution margin; for a service entity, a service margin. This stage is a conventional benchmarking exercise and should be done conventionally, with a documented search.

Stage two: the residual. What remains after routine returns is allocated by reference to relative contribution to the unique value drivers. The allocation key is where the analysis is most often thin, and it is where the officer will concentrate.

Keys commonly used include cumulative development spend, capitalised and amortised over an appropriate life; headcount or cost of the people performing DEMPE functions; the relative significance of identified value drivers weighted on a reasoned basis; and, where the facts permit, a bargaining analysis of what each party could have obtained from an independent counterparty.

Whichever key is used, three things must be shown: that it measures contribution to the residual rather than merely activity; that the data underlying it is reliable and auditable; and that the result is stable, in the sense that reasonable variation in the inputs does not change the conclusion materially. A sensitivity table is worth more than a paragraph of justification.

6. The high margin problem

An entity that owns a valuable platform and licenses it will often report margins far above anything in a routine comparable set. Where it also claims a unit based deduction, the combination attracts attention, and the officer’s working assumption is frequently that the margin is evidence of mispricing.

The answer is structural. In a model where one party owns the intangible and others distribute, a high margin at the owner is what the arm’s length principle predicts, not a departure from it. The comparison being drawn is with service providers that own nothing, and the margin differential is the return on the intangible. Demonstrating that is done by showing the development investment that produced the platform, the period over which it was made, the risk that it would produce nothing, and the fact that the same entity absorbed that risk.

Where a deduction is also claimed, the two issues should be kept separate in the submission. The availability of the deduction is a question under its own provision. It is not a reason to reallocate profit, and conflating the two in the response invites the officer to do the same.

7. A worked example

Assume, hypothetically, that a group owns a commercially valuable intangible developed and managed in India while overseas affiliates carry out distinct distribution functions. The question is whether the Indian entity makes unique contributions and the dealings are sufficiently integrated to support a profit split, and whether a one-sided comparable can reliably measure the Indian entity’s return. The selection must be supported by actual contracts, risk control and contribution records.

Gateway. Both limbs of Rule 10B(1)(d) are engaged. The platform is a unique intangible, and the development and distribution transactions are so interrelated that neither can be priced without reference to the other.

Economic ownership. Established through development headcount and location, roadmap governance records, the funding pattern across years, and evidence that failed projects were absorbed in India.

Routine returns. Overseas distributors benchmarked against independent distributors in their markets; a documented search for each.

Residual. Allocated on a key reflecting cumulative development contribution, with the computation shown and a sensitivity analysis provided.

Meeting the substitution. The submission has to show affirmatively why the transactional net margin method cannot be applied reliably here, not merely that the profit split method can. The reasons are that no reliable comparable exists for an entity owning a platform of this kind; that the proposed comparables perform services and own nothing, so the differential is the very return in dispute; and that a one sided method applied to the intangible owner allocates the residual to parties that did not create it. On comparability the reasoning in UT Starcom applies.

Consistency. Where the department has accepted a split in an adjacent year, that acceptance should be in the submission. Infogain directed that an accepted split be followed, and consistency arguments carry real weight at tribunal level.

8. The allocation key, in figures

The residual stage is where submissions are thinnest, so it is worth showing what an adequately documented key looks like. Take a group with combined profit of 100 from the platform business, an Indian developer that owns the platform, and two overseas distributors.

Stage one, routine returns. Each distributor is benchmarked against independent distributors in its market and allocated a routine return of 8, giving 16 in total. The Indian entity is allocated a routine return for the service element of its activity, benchmarked against independent software service providers, of 14. Routine returns therefore absorb 30, and the residual is 70.

Stage two, the residual. The residual is attributable to the unique contributions. Suppose cumulative development spend is capitalised over a five year life and the amortised balance by entity is 60 in India and 5 across the two distributors, reflecting local adaptation work they have funded. On that key the residual splits 60 to 5 (about 92.3 to 7.7), giving the Indian entity about 64.6 and the distributors about 5.4 between them.

Totals. India takes 14 plus about 64.6, or about 78.6. The distributors take 16 plus about 5.4, or about 21.4.

Sensitivity. The submission should then show what happens if the key moves. If the amortisation life is three years rather than five, or if a different measure of development contribution is used, how much does the allocation change? Where the answer is a few percentage points, the conclusion is stable and the file should say so. Where a plausible change in assumption moves the allocation materially, the key is not doing enough work and a second key should be tested alongside it.

What makes this defensible is not the elegance of the arithmetic. It is that every figure is traceable to an audited number or to a documented computation, and that the submission has already answered the question an officer will ask about each input.

9. The alternatives, and why they were rejected

A method selection analysis is not complete until it explains why the other methods were not used. For a platform licensing model the realistic alternatives are these, and each should be dealt with in a sentence or two rather than left unmentioned.

Comparable uncontrolled price. Available in principle where a comparable licence of comparable technology to an independent party can be found. In practice it fails for a proprietary platform because the comparability requirements for an intangible licence are demanding, and a licence of different technology on different terms is not a comparable price.

Resale price and cost plus. Both are one sided and both require a reliable gross margin comparison. They are unsuited to a model whose value sits in an intangible rather than in a distribution or service margin.

Transactional net margin method with the overseas distributors as tested parties. This is the alternative that should be taken most seriously, because it can be a correct approach. Where the distributors are genuinely routine and the Indian entity is the only party with unique contributions, testing the distributors and leaving the residual in India produces the same answer as a residual profit split, with less complexity. The honest position is that this is often the better method, and the case for a profit split is strongest where the overseas entities also contribute something unique, such as a market intangible they have developed and funded.

Transactional net margin method with the Indian developer as tested party. This is the method usually proposed by the Revenue, and it is the one that has to be rejected affirmatively rather than merely displaced. The objection is not that it produces an unfavourable number. It is that it treats the party owning the unique intangible as the routine party, which inverts the structure of the transaction.

10. Where these cases fail

First, the gateway is asserted rather than established. Stating that unique intangibles exist is not the same as identifying them, showing what makes them unique and showing who owns them economically.

Second, the allocation key is unsupported. A percentage without a derivation is the single most common weakness, and officers go to it first.

Third, the routine stage is skipped. A profit split that allocates total profit without first benchmarking a routine return for each party is not a residual profit split and should not be described as one.

Fourth, the method changes between years. Where the profit split method appears in a profitable year and a one sided method in a lean one, the selection looks outcome driven.

Fifth, the DEMPE evidence is documentary rather than operational. An agreement recording that the Indian entity owns the intangible is a starting point; the functional evidence is what decides it.

Sixth, the case for rejecting the alternative is not made. The officer does not have to prove the transactional net margin method is better; the taxpayer has to show it cannot be applied reliably.

Seventh, Infogain is overstated. Citing it for the proposition that unique intangibles always require a profit split invites the correction, and the correction weakens everything around it.

11. Documentation

1. A value chain analysis identifying each contribution and the party making it.
2. Identification of each intangible, what makes it unique, and the evidence of economic ownership against each DEMPE function.
3. The functional analysis for every party, including the overseas entities whose routine characterisation the analysis depends on.
4. The gateway analysis under Rule 10B(1)(d), identifying which limb or limbs apply and why.
5. The method selection analysis under Rule 10C, including why each other method was rejected and specifically why a one sided method cannot be applied reliably.
6. The routine return benchmarking for each party, with documented searches.
7. The residual allocation key, its derivation, the underlying data and a sensitivity analysis.
8. The comparability analysis supporting rejection of any comparable set proposed by the Revenue.
9. Evidence of consistency with earlier years, including any departmental acceptance.
10. The accountant’s report under section 172 of the 2025 Act (section 92E of the 1961 Act), furnished in Form No. 48 under rule 85 of the Income-tax Rules, 2026 (Form 3CEB for earlier years), describing the transaction consistently with the analysis.

Practitioner’s perspective

The profit split method is treated with suspicion in Indian practice because it is harder to verify than a one sided method and because it has sometimes been used to justify a predetermined allocation. That suspicion is a reason to document it well, not a reason to avoid it where the rule points to it.

In our experience the cases that succeed share three features. The gateway under Rule 10B(1)(d) is established on evidence rather than asserted. The allocation key is derived from auditable data and tested for sensitivity. And the submission makes the affirmative case that a one sided method cannot be applied reliably, rather than only defending the method chosen.

One further point on sequencing. Where a group expects to apply this method, the evidence should be captured as the business runs: roadmap governance, development cost by entity, decision records. Reconstructing DEMPE evidence three years later from memory and organisation charts is possible but it is never as persuasive as records that were kept because the business needed them.

The profit split method under rule 79 of the Income-tax Rules, 2026

For tax years under the Income-tax Act, 2025, the method is applied through rule 79(1)(d) of the Income-tax Rules, 2026, which implements section 165. As published on the Income Tax Department website, the rule states that the profit split method applies mainly where a transaction involves unique intangibles or unique and valuable contributions by each enterprise, or where multiple transactions are so interrelated that they cannot be evaluated separately. It then describes two approaches in terms: a contribution profit split, which divides combined net profit by relative contributions assessed on functions, assets and risks, and a residual profit split, which first allocates part of the combined profit to each enterprise for contributions that can be benchmarked against uncontrolled transactions and then divides the residual by relative contributions.

The earlier Rule 10B(1)(d) did not spell out the residual mechanics, which is why Infogain, decided under the old rule, remained the main Indian authority for a two-stage split. The explicit rule text gives a taxpayer a firmer footing for the two-stage structure described in this article, and it makes the routine-return stage a stated step rather than a matter of practice. It does not shift the burden: the gateway, economic ownership and the allocation key still have to be proved on evidence, and a transfer pricing officer may still say that a one-sided method is more reliable.

Method selection continues to be tested against rule 80, which asks which method is best suited to the facts and gives the most reliable measure of the arm’s length price, by reference to the nature of the transaction, the functions, assets and risks of each enterprise, the reliability of data, comparability and the reliability of adjustments. A file that walks through each of those factors for the profit split method and for the proposed one-sided alternative is the best protection against substitution.

Frequently asked questions

1. When does Rule 10B(1)(d) permit the profit split method?

Where the transaction involves transfer of unique intangibles, or where multiple international transactions are so interrelated that they cannot be evaluated separately for determining the arm’s length price. The two limbs are disjunctive.

2. What is the difference between a contribution split and a residual split?

A contribution analysis divides the combined profit by reference to relative contributions. A residual analysis first allocates a routine return to each party for its ordinary functions and then divides only what remains by reference to the unique contributions.

3. Does the existence of unique intangibles mean the profit split method must be used?

No. Infogain did not so hold. Its reasoning rested principally on the integration of the operations and the fact that both parties contributed materially. The method must still be shown to be the most appropriate on the facts and the available data.

4. Why can a one sided method be inappropriate here?

Because it tests the less complex party and leaves the residual to the other side by default. Where the officer tests the intangible owner against routine comparables, the method caps the owner’s return at a service margin and allocates the residual to parties that did not create it.

5. How is economic ownership of an intangible established?

By evidence against each DEMPE function: who develops, enhances, maintains, protects and exploits, who funds the activity and who bears the risk of failure. Legal ownership is the starting point only.

6. What allocation keys are acceptable for the residual?

Keys that measure contribution to the residual rather than mere activity, derived from reliable and auditable data. Cumulative development spend, DEMPE headcount or cost, and weighted value drivers are commonly used. Whichever is chosen, the derivation and a sensitivity analysis should be documented.

7. Does a high margin at the Indian entity indicate mispricing?

Not where the entity owns the intangible. In a model where one party develops and owns and others distribute, a high margin at the owner is what the arm’s length principle predicts. The differential against service comparables is the return on the intangible.

8. How does a deduction claim interact with the transfer pricing analysis?

They are separate questions and should be kept separate in the submission. The availability of a deduction is governed by its own provision and is not a reason to reallocate profit between associated enterprises.

9. Does consistency with earlier years help?

Yes. In Infogain the Tribunal directed that a split accepted by the department in adjacent years be followed. Departmental acceptance in any year should be in the submission.

10. Which provisions govern this under the Income-tax Act, 2025?

Chapter X, sections 161 to 177, in particular section 165 (section 92C, determination of arm’s length price), section 171 (section 92D, documentation) and section 172 (section 92E, the accountant’s report). Section 536, the repeal and savings provision, preserves the 1961 Act for earlier years and pending proceedings.

11. Is the residual profit split method accepted in India?

It is accepted where the facts support it, but it is not common. The Delhi Tribunal in Infogain India (19 August 2015) held the profit split method most appropriate for integrated operations where both entities contributed materially. Officers often propose TNMM instead, so acceptance depends on proof of the gateway, economic ownership of the intangible and an auditable allocation key.

12. What does rule 79 of the Income-tax Rules, 2026 say about the profit split method?

Rule 79(1)(d) applies the method mainly to unique intangibles or unique and valuable contributions, or to transactions so interrelated that they cannot be evaluated separately. It describes a contribution split and a residual split, the latter allocating benchmarkable returns first and dividing the residual by relative contributions. It succeeds Rule 10B(1)(d) for tax years under the Income-tax Act, 2025.

13. Can the tax officer replace the profit split method with TNMM?

The officer can propose it, but the choice must follow the most appropriate method factors in Rule 10C (rule 80 of the 2026 Rules). The taxpayer should show that a one-sided method cannot be applied reliably, for example because the proposed comparables own no intangibles and the tested party would be wrongly treated as the routine party.

Related reading from SBC

• When Is the Berry Ratio the Right Profit Level Indicator for a Limited Risk Contract Manufacturer in India?
• Contract R&D or Routine Software Development: How the Functional Profile Decides the Transfer Pricing Method in India
• Digital Platforms and India Tax Once the Equalisation Levy Is Gone: SEP, Treaty and PE Compared
• More articles on the SBC blog

Disclaimer

This article is for general information and does not constitute professional advice. It reflects the law as understood at the date of publication and does not take account of later developments. Statutory references give the Income-tax Act, 2025 section with the Income-tax Act, 1961 section in brackets; section 536 preserves the 1961 Act for proceedings pending on 31 March 2026 and for earlier tax years. The selection of the most appropriate method depends on the functional profile and the data available in each case. No reader should act on the basis of this article without taking advice on their own circumstances.

When Is the Berry Ratio the Right Profit Level Indicator for a Limited Risk Contract Manufacturer in India?

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Transfer Pricing

Introduction

A profit level indicator is a ratio that expresses the tested party’s return against some measure of what it does. The transactional net margin method does not prescribe one. It allows the return to be measured against costs, against sales, against assets, or against any other base that is relevant, and the choice of base is part of applying the method rather than a preliminary to it. Choosing the wrong base produces a comparison between two things that are not alike, however carefully the comparables are screened.

The base that matters here is operating expenses, and the ratio built on it is the Berry ratio. It is the wrong indicator far more often than it is the right one, and a case built on it fails where the functional analysis does not support it. It is not a way of improving a result. It is a conclusion that follows from a particular functional profile, and the profile has to be established first.

This note deals with the base to use where the tested party’s accounts are dominated by purchases it does not control. It covers what the Berry ratio measures and why that matters, the Indian legal authority permitting it and the limits that authority sets, the functional evidence that has to be established before the ratio can be defended, the recharacterisation argument the Revenue usually runs, and the classification reconciliation on which these cases are actually decided. The functional analysis that underpins all of it is dealt with separately in our note on FAR analysis for transfer pricing audit defence.

1. What the ratio measures and why that matters

The Berry ratio is gross profit divided by operating expenses. Expressed differently, it asks how much gross margin the business earns for each rupee it spends on running itself.

What makes it useful is what it leaves out. Cost of goods sold does not appear in the denominator. Where an entity buys components from an associated enterprise, assembles them and sells the output, and where it has no influence over what it pays for those components, the purchases are not a measure of anything the entity does. They are a number passing through its accounts. A cost plus margin computed on total cost treats that number as the base for the entity’s return, which means the return rises and falls with component prices the entity does not set.

Operating expenses, by contrast, are the entity’s own spending: its people, its premises, its plant running costs, its local functions. For a business whose contribution is assembly and local execution, those expenses are a reasonable proxy for the value it adds.

The OECD Transfer Pricing Guidelines describe the Berry ratio as appropriate in narrow circumstances, broadly where the value of the functions performed is proportional to operating expenses, the value of the functions is not materially affected by the value of the products distributed, and the taxpayer does not perform other significant functions that should be remunerated separately.

2. The Indian legal basis

Rule 10B(1)(e) of the Income-tax Rules, 1962 (now rule 79(1)(e) of the Income-tax Rules, 2026) sets out the transactional net margin method and permits the net profit margin realised to be computed “in relation to costs incurred or sales effected or assets employed or to be employed by the enterprise or having regard to any other relevant base”. The closing words are the statutory foundation.

The Delhi High Court settled the point in Sumitomo Corporation India (P.) Ltd. v. CIT, decided on 22 July 2016 in IT Appeal Nos. 381 and 382 of 2013, 702 of 2014 and 738 of 2015, by Muralidhar and Vibhu Bakhru JJ. The Court held that the words “any other relevant base” permit operating expenses to be used as the base, and that the transfer pricing officer’s view that only cost, sales or assets could be used was plainly erroneous.

The Court was equally clear about the limits, and a submission that quotes the first holding without the second is incomplete. The ratio suits what the Court described as stripped down distributors with no financial exposure or risk in the goods. It is unsuitable where the entity uses intangibles or substantial fixed assets, and it is unsuitable where the comparables’ product mix differs materially. The matter was remanded to the Tribunal.

The Delhi Bench had reached a consistent view in Mitsubishi Corporation India Pvt. Ltd. v. DCIT, pronounced on 21 October 2014. For a high volume, low risk trading model the Berry ratio was held appropriate, because comparables carried much higher inventory and inventory risk, which would distort a cost based or sales based margin. The Bench identified the suitable profile as pure trading with no processing, no significant trade or marketing intangibles and low inventory.

Read together, the two decisions give a workable test. The Berry ratio is available where the entity’s contribution is proportional to its operating expenses and is not proportional to the value of the goods flowing through it. Where the entity owns intangibles, carries meaningful inventory or market risk, or deploys substantial fixed assets, the ratio understates what it should be earning and the courts have said so.

3. The functional analysis that has to come first

The method follows the functional profile. Establishing that profile is the work, and it is done on evidence rather than assertion. For a limited risk contract manufacturer the following has to be demonstrated.

Procurement. Who selects the supplier, who negotiates the price, who bears the consequence of a price movement. Where the associated enterprise nominates the supplier and sets the transfer price, and the Indian entity has no ability to source elsewhere or to renegotiate, the purchases are pass through in substance as well as in form. The evidence is the supply agreement, the purchase orders, the pricing mechanism and any correspondence showing who decided.

Product and process design. Who owns the specification, who approves changes, who holds the technical know how. A contract manufacturer executes a specification it did not write.

Customers. Who holds the customer relationship, who negotiates the sale price, who carries the credit risk. In a global contracting arrangement the Indian entity often supplies to customers under terms agreed by the parent, and that should be visible in the contracts.

Inventory. Who bears obsolescence, who decides production volumes, who carries the risk of unsold stock. This is the single most probative risk indicator for a manufacturing entity and it is frequently glossed over.

Capacity and idle cost. Who bears the cost of underutilisation. An entity guaranteed a return on its operating expenses regardless of volume is not carrying volume risk.

Foreign exchange. Where purchases are denominated in foreign currency and sales in rupees, who carries the exposure.

The documentary sources are the inter company agreement, the board minutes, the delegation of authority, the insurance policies, the actual pattern of profitability across years, and the correspondence that shows who actually decided things. A limited risk characterisation contradicted by a pattern of volatile margins is not a limited risk characterisation.

4. The recharacterisation risk, and how it is met

The standard revenue position is that the entity is not a limited risk manufacturer at all but an entrepreneurial one, with the consequence that a total cost base is appropriate and that entrepreneurial comparables should be used. The two halves of that position have to be met separately.

On characterisation, the evidence set out above is the answer, and it has to be assembled from primary documents rather than from the transfer pricing report of the previous year.

On comparables, the principle is well established in a related context. In CIT v. UT Starcom Inc. (India Branch), pronounced on 25 September 2017, the Delhi High Court found no substantial question of law and dismissed the Revenue’s appeal against the Tribunal’s exclusion of large risk taking companies with significant intangibles and assets from the comparable set for a captive unit subject to limited risk. The reasoning, that functional similarity alone does not make such a company a valid comparable, is the Tribunal’s, affirmed on appeal. The reasoning transfers directly to manufacturing: a company that owns its products, carries its own market risk and funds its own development is not comparable to one that assembles to a specification for a guaranteed return.

5. A worked example

Assume, purely for illustration, an Indian processing enterprise whose total costs are dominated by purchases subject to pricing decisions made by an associated enterprise. Its own contribution is confined to identifiable operating functions. Whether those material purchases may be treated as pass-through for benchmarking depends on evidence that the Indian enterprise does not control supplier selection, purchasing risks or principal customer terms; the Berry ratio is not assumed merely from the size of purchases.

The mechanical problem. With a cost base of, say, 100 of which 88 is component purchases and 12 is operating expenses, a benchmark margin of five per cent on total cost demands a return of 5. The same entity earning a Berry ratio in the comparable range on its 12 of operating expenses earns something quite different. The two indicators are not measuring the same thing, and on these facts only one of them is measuring the entity’s actual contribution.

The analysis. Establish through the supply agreement and the purchase documentation that component sourcing and pricing are controlled by the associated enterprise. Establish through the customer contracts that the parent holds the relationship and negotiates price. Establish through the inventory and capacity arrangements who bears obsolescence and idle cost. Identify operating expenses as the value added base. Select comparables on the functional profile, excluding entities with their own products, intangibles or market risk. Compute the Berry ratio for the tested party and the comparable set on a consistent basis.

What decides it. Consistency of classification. The single most common reason a Berry ratio analysis fails is that operating expenses and cost of goods sold are classified one way in the tested party’s accounts and another way in the comparables’ published accounts. A comparable that includes factory overheads in cost of goods sold while the tested party books them as operating expenses produces a ratio that is not comparable to anything. The reclassification work has to be done, documented and disclosed.

6. The computation, in figures

The mechanics are easier to see with numbers. Take an Indian assembler with the following results for a year, in crore.

Line	Amount
Sales	1,000
Cost of goods sold, almost entirely components bought from associated enterprises	880
Gross profit	120
Operating expenses, being the entity’s own people, premises and plant running costs	100
Operating profit	20

On a total cost basis the operating margin is 20 on 980, a little over two per cent. If the transfer pricing officer benchmarks the entity against manufacturers earning five per cent on total cost, the required profit is 49 and the adjustment is 29.

On a Berry ratio basis the measure is gross profit over operating expenses, 120 over 100, or 1.20. If the comparable set of stripped down assemblers shows a Berry ratio range of 1.10 to 1.25, the entity sits inside the range and no adjustment arises.

The two indicators disagree because they are measuring different things. The total cost measure asks what return the entity earns on every rupee that passes through it, including 880 of components whose price it did not set. The Berry measure asks what return it earns on the 100 it actually spends running the business. Where the entity genuinely controls neither the supplier nor the price, only the second question is about the entity.

Two cautions on the arithmetic. First, the result is highly sensitive to where costs sit. Moving 30 of factory overhead from operating expenses into cost of goods sold changes the ratio from 1.20 to 1.29 and changes gross profit at the same time, which is why the classification reconciliation matters more than any other working in the file. Second, a Berry ratio below 1.0 means the entity’s gross profit does not cover its own operating costs, which is a loss and will require explanation whatever method is applied.

7. The conditions the OECD attaches

The OECD Transfer Pricing Guidelines treat the Berry ratio as appropriate only in narrow circumstances, and the three conditions they identify are worth testing explicitly in the file because an officer is likely to raise them.

1. The value of the functions performed is proportional to operating expenses. This is the central condition and it is the one the functional analysis has to establish. It fails where the entity’s contribution is driven by something other than its own spending, such as an intangible or a risk it bears.
2. The value of the functions performed is not materially affected by the value of the products distributed. This fails where the entity earns more simply because the goods passing through it are more valuable, which is characteristic of a distributor that carries price or inventory risk.
3. The taxpayer does not perform other significant functions that should be remunerated by another method. This fails where, alongside the assembly or distribution activity, the entity performs development, marketing or service functions that are not compensated through the same base.

Each condition should be addressed in the documentation by reference to evidence rather than assertion, and where one of them is doubtful the honest course is to say so and explain why the ratio is still the most reliable measure available, rather than to leave the officer to find the weakness.

8. Where a Berry ratio case fails

First, classification is inconsistent between the tested party and the comparables. This is the dominant failure and it is a data problem, not a legal one.

Second, the entity holds intangibles. Any marketing intangible, customer list, process know how or product right makes the ratio inappropriate on the Sumitomo reasoning, because the return on the intangible is not proportional to operating expenses.

Third, fixed assets are substantial. A manufacturer with a significant asset base is deploying capital that a Berry ratio ignores, and a return that compensates operating expenses but not capital is not arm’s length.

Fourth, the purchases are not genuinely pass through. Where the entity can source alternatively, negotiates its prices, or benefits from favourable purchasing, the characterisation collapses.

Fifth, the ratio is selected because of the answer it produces. Where the file shows a total cost analysis prepared first and a Berry analysis prepared after it gave an unfavourable result, the method selection is not credible.

Sixth, the method changes between years on unchanged facts. Consistency matters more than optimality. A change of method invites the question of what changed, and the answer has to be something in the business.

Seventh, the comparable set is too small. The ratio is sensitive, and a narrow set magnifies classification noise.

9. Documentation

A defensible file contains the following, and it is prepared before the return rather than after the notice.

1. A functional analysis evidenced from primary documents, covering procurement, design, customers, inventory, capacity and currency risk.
2. The inter company agreement, with the pricing mechanism and the risk allocation it records.
3. An explanation of why the purchases are pass through in substance, supported by the sourcing and pricing documentation.
4. The method selection analysis, showing why the transactional net margin method was chosen and why operating expenses is the relevant base under Rule 10B(1)(e), with Sumitomo cited for the legal availability of the base.
5. The comparable search, with the screens applied and the reasons for each rejection, including the rejection of entities with intangibles, significant assets or market risk.
6. The classification reconciliation, showing how each comparable’s published figures were adjusted to a consistent basis, and disclosing the adjustments made.
7. The computation for the tested party and the set, with the range and the tested party’s position in it.
8. A note on the alternative, showing what a total cost analysis would produce and why it is not appropriate on these facts, so that the officer’s likely approach has already been addressed.
9. The accountant’s report under section 172 of the 2025 Act (section 92E of the 1961 Act), which is furnished in Form No. 48 under rule 85 of the Income-tax Rules, 2026 (Form 3CEB for earlier years), with the transaction described consistently with the analysis.

10. The statutory framework in the 2025 Act

Transfer pricing now sits in Chapter X of the Income-tax Act, 2025, sections 161 to 177. The provisions most relevant here are section 163 (section 92B of the 1961 Act, the meaning of international transaction), section 165 (section 92C, determination of arm’s length price), section 166 (section 92CA, reference to the transfer pricing officer), section 171 (section 92D, documentation) and section 172 (section 92E, the accountant’s report).

Two changes are worth noting. Section 165 is now headed determination rather than computation of arm’s length price, and it provides for a tolerance band of up to three per cent as notified by the Central Government. Separately, the general anti avoidance rules have been moved out of the transfer pricing chapter into a standalone Chapter XI, sections 178 to 184.

For proceedings relating to earlier years, section 536 preserves the 1961 Act, so current assessments and appeals continue to be argued on the old numbering.

Practitioner’s perspective

The Berry ratio has a reputation in Indian practice as an aggressive position. That reputation is partly deserved, because it is often reached for when a total cost analysis produces an uncomfortable result. Used properly it is the opposite: it is the indicator that reflects economic reality for an entity whose accounts are dominated by numbers it does not control.

The legal question is settled. Sumitomo confirms that operating expenses can be the base under Rule 10B(1)(e), and the same decision sets out the limits. What remains contested in practice is almost always factual: whether the entity really is stripped down, whether the purchases really are pass through, and whether the comparables have been put on a consistent footing.

Two pieces of advice follow. Establish the functional profile from primary documents before choosing the method, and be prepared to abandon the ratio if the profile does not support it. And treat the classification reconciliation as the centre of the file rather than an annexure, because that is where these cases are actually won and lost.

Where the Berry ratio sits under the Income-tax Rules, 2026

For tax years beginning on or after 1 April 2026, the arm’s length price is determined under section 165 of the Income-tax Act, 2025, which replaces section 92C of the 1961 Act. The manner of applying each method is prescribed in rule 79 of the Income-tax Rules, 2026, which the CBDT notified on 20 March 2026 as the successor to Rule 10B. The transactional net margin method in rule 79(1)(e) still measures the net profit margin against costs, sales, assets employed or to be employed, or any other relevant base, so the legal footing for an operating expense base that Sumitomo recognised under Rule 10B(1)(e) carries over in the same words.

Rule 80 replaces Rule 10C and lists the factors for choosing the most appropriate method: the nature of the transaction, the functions, assets and risks of each enterprise, the availability and reliability of data, the degree of comparability, whether reliable adjustments can be made, and the reliability of the assumptions the method needs. A Berry ratio file should be organised against those factors, because that is how an officer or the Dispute Resolution Panel will test it.

Two points on safe harbour. The safe harbour table in rule 89 of the 2026 Rules covers manufacture and export of auto components among manufacturing activities, so a general contract manufacturer outside that category cannot rely on a prescribed margin and must benchmark. And a safe harbour election does not remove the need for documentation under section 171 and the accountant’s report under section 172.

Frequently asked questions

1. What is the Berry ratio?

Gross profit divided by operating expenses. It measures the return the business earns relative to its own operating spending, and it excludes cost of goods sold from the base.

2. Is the Berry ratio permitted under Indian transfer pricing rules?

Yes. Rule 10B(1)(e) permits the net profit margin to be computed in relation to costs incurred, sales, assets employed or any other relevant base. In Sumitomo Corporation India the Delhi High Court held that operating expenses can be that base and that confining the base to cost, sales or assets was plainly erroneous.

3. When is the Berry ratio appropriate?

Where the value of the functions performed is proportional to operating expenses and not to the value of the goods flowing through the entity, and where the entity does not own intangibles, deploy substantial fixed assets or carry meaningful inventory or market risk.

4. When is it inappropriate?

Where the entity owns intangibles or substantial fixed assets, where it carries inventory or market risk, or where the comparables’ product mix differs materially. The Delhi High Court identified each of these limits in Sumitomo.

5. Why exclude cost of goods sold from the base?

Because where the entity neither selects the supplier nor negotiates the price, the purchases measure nothing the entity does. Including them makes the entity’s return move with component prices it does not control.

6. What is the most common reason a Berry ratio analysis fails?

Inconsistent classification between the tested party and the comparables. Where a comparable books factory overheads in cost of goods sold and the tested party books them as operating expenses, the ratios are not comparable. The reconciliation has to be done and disclosed.

7. Can the method be changed from an earlier year?

It can, but the change has to be explained by something that changed in the business. A change of method on unchanged facts, particularly one that improves the result, undermines the credibility of the selection.

8. How should comparables be selected for a limited risk manufacturer?

On the functional profile, excluding entities that own their products, carry their own market risk or fund their own development. The reasoning in UT Starcom, that a large risk taking company with significant intangibles is not comparable to a limited risk captive, applies directly.

9. What evidence establishes that purchases are pass through?

The supply agreement, the purchase orders, the pricing mechanism, evidence of who nominates the supplier, and evidence that the entity cannot source elsewhere or renegotiate. Assertion in a transfer pricing report is not evidence.

10. Which provisions govern transfer pricing under the Income-tax Act, 2025?

Chapter X, sections 161 to 177. In particular section 163 (section 92B, the meaning of international transaction), section 165 (section 92C, determination of arm’s length price), section 166 (section 92CA, reference to the transfer pricing officer), section 171 (section 92D, documentation) and section 172 (section 92E, the accountant’s report). Section 536, the repeal and savings provision, preserves the 1961 Act for earlier years and pending proceedings.

11. Is the Berry ratio accepted by Indian courts and tribunals?

Its legal availability is accepted. The Delhi High Court in Sumitomo Corporation India (22 July 2016) held that operating expenses can be the ‘relevant base’ under Rule 10B(1)(e), and the Delhi Tribunal in Mitsubishi Corporation India (21 October 2014) accepted it for a high-volume, low-risk trading model. Acceptance in a given case depends on the functional profile, and both decisions concerned trading rather than manufacturing.

12. Can the Berry ratio be used for a contract manufacturer in India?

It can, where component purchases are pass-through costs the entity does not control, it owns no significant intangibles, carries no material inventory or market risk and deploys no substantial fixed assets. Those facts must be proved from the supply agreement, purchase records and customer contracts. If the entity owns plant or know-how, a total cost or asset-based indicator is likely to be more reliable.

13. Which rule replaces Rule 10B of the Income-tax Rules under the Income-tax Act, 2025?

Rule 79 of the Income-tax Rules, 2026 prescribes how the arm’s length price is determined under section 165, and rule 80 lists the factors for the most appropriate method. The accountant’s report under section 172 is in Form No. 48 under rule 85. Rule 10B, Rule 10C and Form 3CEB continue to apply to periods and proceedings governed by the 1961 Act.

Related reading from SBC

• Defending the Residual Profit Split Method in India When the TPO Proposes TNMM for an Intangible-Owning Entity
• A Trading Entity Is Not a Service Provider: How to Defend the Resale Price Method Against a TNMM Adjustment in India
• Contract R&D or Routine Software Development: How the Functional Profile Decides the Transfer Pricing Method in India
• More articles on the SBC blog

Disclaimer

This article is for general information and does not constitute professional advice. It reflects the law as understood at the date of publication and does not take account of later developments. Statutory references give the Income-tax Act, 2025 section with the Income-tax Act, 1961 section in brackets; section 536 preserves the 1961 Act for proceedings pending on 31 March 2026 and for earlier tax years. The selection of a profit level indicator depends on the functional profile established on the facts of each case. No reader should act on the basis of this article without taking advice on their own circumstances.

Seconding Staff Between India and the UK: Employer, Permanent Establishment and Payroll Tax Are Different Questions

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Advisory

Introduction

International workforce deployment is often presented as a payroll question: which entity pays the employee and in which country does tax need to be deducted? That is rarely the whole problem. The economic employer may differ from the paying entity, and a group entity that deploys employees may create a permanent establishment through activities that have little to do with the salary remittance.

For an India–UK arrangement, one must analyse at least three separate taxpayers or tax positions: the employee, the entity supplying or deploying personnel, and the business that receives the work. The structure becomes especially sensitive where the host directs daily activities while the home entity retains the employment contract, compensation decisions or project risk.

This article provides a practical decision framework for that arrangement. It deliberately separates salary taxation, corporate profit attribution and intercompany-service characterisation, because a favourable answer to one does not settle the others.

1. The actual deployment model takes priority

Describe who recruited the employee, who may terminate the engagement, who decides the work programme, who evaluates performance, who bears employment risk and who receives the deliverables. A secondment agreement alone may not tell the whole story where the parties operate differently in practice.

The arrangement might be a temporary employment transfer, a foreign enterprise delivering services using its personnel, or a split arrangement in which the host has operational supervision while the home employer retains broader control. The answer should emerge from documents and practice, not from the title “secondment”.

2. Salary taxation under the India–UK treaty

The synthesised India–UK treaty places dependent personal services in Article 16. Its short-stay protection requires satisfying the relevant conditions together, including the presence threshold, the paying employer requirement and the treatment of remuneration in the source State.

The familiar 183-day test is not a free-standing exemption. If the conditions concerning employer residence and the bearing of remuneration are not met, salary may be taxable in the State where employment is exercised even for a short presence. Count days as the treaty directs and retain travel documentation.

3. Economic employer and the recharge trail

Payroll may continue with the home entity for administrative convenience. That fact should be tested against who is responsible for the work, who bears compensation cost, who controls the employee’s duties and whether the host is reimbursing salary or purchasing a wider commercial service.

Where a recharge is made, reconcile payroll records to the intercompany invoice and the receiving entity’s expense ledger. A recharge carrying overhead or a service margin may present a different corporate and withholding question from a pure cost recovery; it does not necessarily change the employee’s treaty position.

4. Service PE risk can arise independently

The India–UK treaty contains a service-related PE provision in its Article 5 framework, including a rule for services performed for an associated enterprise. The applicable day-counts and activity tests must be checked against the authenticated treaty wording and periods in question.

An employee may satisfy personal tax conditions without that resolving the deploying company’s PE risk. Track all relevant personnel, connected assignments, client projects and periods, not merely one employee’s travel in isolation. The company’s function in the host jurisdiction must be described.

5. Permanent establishment is not automatically a large profit

If a corporate PE is found, profit attribution remains a distinct exercise. It focuses on the functions, assets and risks of the PE rather than attributing all profits of the foreign employer to the country of deployment.

An engineer seconded to coordinate a discrete task may create an economic contribution very different from a team concluding customer contracts or independently delivering an integrated service. Payroll recharges, project accounting and management records should support the attributable profit analysis.

6. Payroll, social security and immigration cannot be combined into one rule

The tax residence of the employee, domestic salary withholding obligations, available foreign tax relief and any applicable social-security arrangements each require separate testing. A tax treaty article is not an immigration permission, and a visa designation does not determine tax ownership of a function.

Before departure, establish a host payroll policy, employer registrations where required, travel documentation, expected working-day count and the procedure to manage changes in deployment length. Late correction can be costly where work has already been performed.

7. Intercompany billing and indirect-tax characterisation

Where the home entity invoices the host, the supply may be characterised as a service depending on the facts. The GST or VAT analysis is separate from the employee’s income-tax treatment. The existence of an employment relationship does not automatically preclude scrutiny of an intercompany supply.

Where an Indian company receives employees from a related overseas entity, assess the law on import of services, valuation and input tax credit. The CBIC circular on related-party import-of-service valuation may be relevant, but it does not answer every secondment characterisation issue.

8. A deployment memorandum for the business

An effective memorandum should contain an employer and control matrix, compensation and recharge schedule, day-count tracker, corporate PE assessment, salary tax position, treaty evidence, GST or VAT analysis, immigration assumptions and escalation triggers.

Update that memorandum when deployment becomes longer, work shifts from execution to contract negotiation, the receiving company begins exercising substantive control, or recharge arrangements change. A well-drafted initial policy loses force if the actual deployment evolves without review.

What the India-UK treaty tests for salary and for service PE

Article 16(2) of the India-UK treaty exempts remuneration for employment exercised in the other State only if three conditions are met together: the individual is present for not more than 183 days in the relevant fiscal year, the remuneration is paid by or on behalf of an employer who is not resident in the State of work, and the remuneration is not deductible in computing the profits of an enterprise chargeable to tax in that State. A host company that bears the cost through a recharge can therefore defeat the exemption even for a short stay.

Article 5(2)(k) is the service PE rule. On the treaty text published by HM Revenue and Customs, furnishing services through employees or other personnel creates a permanent establishment where the activities continue for more than 90 days within any twelve-month period, and for more than 30 days within any twelve-month period where the services are performed for an enterprise associated with the provider. A secondment to an Indian group company should therefore be counted against the shorter threshold, across all connected personnel and assignments and not only for one traveller.

How have the courts treated seconded employees?

Two decisions show the range. In DIT v. Morgan Stanley & Co. (2007) 292 ITR 416, decided on 9 January 2007, the Supreme Court upheld the Authority for Advance Rulings’ view that employees deputed to an Indian affiliate can create a service PE for the foreign enterprise where it remains responsible for their work and they stay on its payroll or retain a lien on their jobs with it.

In Centrica India Offshore Pvt. Ltd. v. CIT, reported in May 2014, the Delhi High Court treated payments for seconded employees as fees for technical services under the India-UK and India-Canada treaties on the facts, rejected the ‘reimbursement’ label, and held that the host’s day-to-day control did not outweigh the fact that the individuals remained employees of the foreign parties. Neither decision says that every secondment is taxable. Both show that payroll location, the control matrix and the recharge basis will be read together.

What changed for social security from 15 July 2026?

The India-UK Comprehensive Economic and Trade Agreement and the Agreement on Social Security Contributions (the double contribution convention) enter into force on 15 July 2026, according to the Press Information Bureau. For detached workers, the convention is reported to exempt the worker and employer from host-country contributions while they remain covered in the home country, for a period reported as up to five years, provided a certificate of coverage is held. It does not cover workers recruited locally or who localise.

Earlier commentary on UK-India deployments noted the absence of any totalisation arrangement. That position changes for postings the convention covers, although income tax, treaty and permanent establishment analysis is unaffected. Confirm the period, the certificate process and the transitional rules from the agreement text.

A pre-deployment checklist

Before the first day of work, hold the signed secondment or service agreement and a control matrix showing who directs, evaluates, can terminate and bears the risk of the individual’s work. Add a compensation and recharge schedule reconciled to payroll and to the intercompany invoice, and a day-count tracker by individual and by project.

Record the employee’s residential status analysis, treaty position and Indian salary withholding plan under section 392 of the Income-tax Act, 2025 (section 192 of the 1961 Act). Add the corporate PE assessment under Article 5 and the profit attribution approach if a PE cannot be ruled out, the GST analysis of the recharge including the related-party import-of-service valuation position in Circular No. 210/4/2024-GST, Form 3CEB mapping where the deploying entity is an associated enterprise, and the foreign exchange and social security assumptions. Finally, list the events that require re-review: an extension, a shift from execution to contract negotiation, or a change in the recharge basis.

Conclusion

Cross-border staff deployment is a useful illustration of why tax analysis must follow economic facts. An employment contract, travel calendar and recharge invoice may each point to a different aspect of the arrangement.

A sustainable structure makes those records consistent from the outset. It does not treat an exemption under one tax regime as clearance under all the others.

Frequently asked questions

1. Is staying in the UK or India for fewer than 183 days sufficient?

Not by itself. All conditions in the applicable dependent-personal-services article must be satisfied.

2. Who is the employer if payroll is paid abroad?

The answer may require analysis of contractual and economic control; payroll location is one factor.

3. Can one seconded employee create PE exposure?

Potentially, depending on the treaty, activities and duration. There is no blanket exclusion.

4. Does a salary recharge always constitute fees for technical services?

No. Characterisation depends on the real supply, if any, and the applicable statute and treaty.

5. What should be prepared before deployment?

A work-scope and control matrix, day tracker, payroll and recharge plan, treaty assessment, and compliance checklist.

6. Does a seconded UK employee working in India for under 183 days pay Indian tax under the India-UK treaty?

Not necessarily. Under Article 16(2) the exemption applies only if presence does not exceed 183 days in the relevant fiscal year, the remuneration is paid by or on behalf of an employer not resident in India, and the cost is not deductible in computing the profits of an enterprise chargeable to tax in India. If the Indian host bears the cost through a recharge, the third condition can fail.

7. How many days create a service permanent establishment under the India-UK treaty?

Under Article 5(2)(k), services furnished through employees or other personnel can create a permanent establishment where the activities continue for more than 90 days within any twelve-month period, or more than 30 days where the services are performed for an associated enterprise. Check the wording in the MLI-modified text, and track days across all personnel and connected assignments, not one employee.

8. Is GST payable on salary recharged by a UK company to its Indian group company?

It can be. Where the overseas entity supplies services to a related Indian entity, the supply may be an import of services chargeable under reverse charge, depending on the facts. Circular No. 210/4/2024-GST dated 26 June 2024 addresses valuation of such imports where the recipient has full input tax credit, but it does not settle every secondment characterisation question.

Related reading from SBC

• Employee Secondment and GST: Reading Northern Operating Systems with the CBIC Clarifications
• Attributing Profits to an Indian Project Office: Which Head Office Expenses Count?
• Shared Services from Singapore: Applying the ‘Make Available’ Test to Indian Withholding
• Parent Guarantee of an Indian Subsidiary: How Transfer Pricing, Withholding and GST Fit Together
• More articles on the SBC blog

Disclaimer

This article is general professional commentary and not advice on any particular facts. Treaty eligibility, applicable law, procedural requirements and statutory amendments must be verified for the relevant tax period. It is not a description of an identifiable client engagement.

Digital Platforms and India Tax Once the Equalisation Levy Is Gone: SEP, Treaty and PE Compared

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Advisory

Introduction

The withdrawal of India’s equalisation levy is not the same as withdrawal of India’s interest in taxing offshore digital business. It removes a separate levy; it does not repeal the income-tax rules that deal with business connection, royalty, fees for technical services or permanent establishment.

This distinction matters for platform operators, connected television businesses, advertising networks and software suppliers. The customer may be in India, the contract signed abroad, the infrastructure hosted elsewhere and the monetisation split among advertising, licensing, data and services. A single description such as “digital revenue” conceals several different tax questions.

The right starting point is to map the payment and the legal capacity in which it is earned. This article explains how significant economic presence should be assessed after the equalisation levy sunset, and why treaty access often changes the resulting answer.

1. Two levies with different legal foundations

The equalisation levy was a specific charge on defined digital transactions under the Finance Act 2016 regime. India first excluded e-commerce supplies from 1 August 2024 and later withdrew the residual online-advertising levy from 1 April 2025. The Income-tax Department overview reflects the withdrawal.

Significant economic presence is part of the Income-tax Act business-connection framework. It can establish a domestic nexus for a non-resident even without a conventional physical office. The policy may be related, but the tests are not interchangeable. Tax practitioners should not continue treating a platform transaction as automatically subject to an abolished levy.

2. Identify the transaction before measuring nexus

A platform may sell advertising inventory, licence software, provide analytics, arrange third-party services or sell hardware. Each receipt should be described by its contractual rights and obligations. The customer agreement, invoicing chain, allocation of advertising revenue and ownership of the content or technology are fundamental.

For example, a contract granting an Indian customer restricted access to a cloud platform may raise a different characterisation from one transferring commercially exploitable copyright rights. Whether a payment is royalty or business profits cannot be decided by the label “subscription” alone.

3. SEP thresholds and attribution are separate steps

Explanation 2A to section 9(1)(i) of the 1961 Act (now section 9(9)(d) of the Income-tax Act, 2025, with the thresholds in rule 13 of the Income-tax Rules, 2026, formerly rule 11UD) established an SEP test using prescribed payment or user-interaction thresholds. That domestic rule also contemplates attribution only of income connected with the relevant transactions or activities. For current periods, check the relevant statutory provision and operative rules rather than borrowing threshold figures from earlier guidance.

A system collecting India-associated users or payments is not sufficient on its own for a tax computation. The enterprise must determine whether the statutory test applies, what income is connected with the nexus, and how the governing rules treat attribution. Global profit or worldwide advertising revenue is not the default attributable amount.

4. A treaty can change the result

Under many treaties, Article 7 permits source-state taxation of business profits only where a qualifying permanent establishment exists and only to the extent attributable to it. The result requires treaty residency, beneficial entitlement where relevant, the operative treaty text and the facts relating to presence, personnel, agents and infrastructure.

Domestic business connection and treaty PE are related but distinct concepts. A platform can have a domestic SEP analysis requiring reporting even where treaty protection ultimately limits Indian taxation. Equally, a treaty claim should never be assumed where the claimant is not entitled to its benefits or a specific anti-abuse rule applies.

5. Do not overlook service and royalty classification

An advertising or streaming ecosystem may involve software licences, database access, campaign management, creative services, hosting and equipment. Those components may fall under different domestic charging provisions or treaty articles. The contract bundle should be analysed by service and right, with careful attention to who owns the intellectual property and what the recipient may do with it.

A royalty claim should identify the particular right or equipment said to be used, not merely cite the use of technology somewhere in the supply chain. A fees-for-technical-services claim similarly requires the treaty-specific definition and any “make available” requirement.

6. Permanent establishment in a digital operating model

Potential PE questions include who habitually concludes contracts, where decisions are made, whether dedicated infrastructure is at the enterprise’s disposal, and whether a local group company acts independently or on behalf of the foreign principal. Legal separation between entities is relevant but not conclusive.

A local advertising sales team that performs routine marketing is not necessarily the same as one authorised to bind the overseas enterprise. Conversely, a group should not dismiss PE exposure because standard customer terms are accepted through an offshore digital interface. Authority and actual conduct matter.

7. Withholding is a transaction-by-transaction decision

An Indian payer should distinguish payments potentially chargeable to tax under domestic law from those entitled to treaty relief. Before applying withholding, verify the payee, residence documentation, agreements, invoices, scope, PE declarations where relevant and applicable statutory procedures.

Where an agreement contains several revenue streams, a single withholding conclusion for all payments can be indefensible. The file should state which component was reviewed, why a specific treaty article applies and what evidence supports the beneficial-owner and PE assumptions.

8. An operating model that survives review

A defensible digital-tax memorandum maps contractual counterparties, customer location, rights granted, service delivery, software architecture, local functions, user or payment data, invoicing and funds flow. Reconcile the map with indirect tax and transfer pricing positions so the group does not advance inconsistent facts in different filings.

Keep separate a statutory conclusion, a modelling assumption and a commercial forecast. If expected payment volumes may cross the prescribed SEP threshold, record the trigger for a fresh review rather than relying indefinitely on an initial conclusion.

What are the SEP thresholds under the Income-tax Act, 2025?

Significant economic presence is now part of the business-connection rule in section 9 of the Income-tax Act, 2025, with the two thresholds prescribed in rule 13 of the Income-tax Rules, 2026, which carries forward former rule 11UD. Rule 13 prescribes aggregate payments of Rs 2 crore in the tax year from transactions in goods, services or property with any person in India, including downloads of data or software, and 3 lakh users with whom the non-resident systematically and continuously solicits business or who interact with it. The official rule page cites section 9(9)(d)(i) and (ii) for the two limbs, and meeting either is enough to create the nexus.

Attribution is a separate step. Only income reasonably attributable to the transactions or activities that create the nexus is deemed to accrue or arise in India, and transactions confined to the purchase of goods in India for export are excluded. The export-purchase exclusion was added to Explanation 2A of the 1961 Act with effect from 1 April 2026.

How did the equalisation levy end, and which periods remain open?

The 2 per cent levy on e-commerce supply or services stopped applying to consideration received or receivable on or after 1 August 2024, under an amendment made by the Finance (No.2) Act 2024. Section 146 of the Finance Act 2025 then amended sections 163 and 165 of the Finance Act 2016 so that the 6 per cent levy on specified services, mainly online advertisement, does not apply to consideration received or receivable on or after 1 April 2025. The Income-tax Department’s equalisation levy page records that the provisions are not applicable with effect from 1 April 2025.

These are prospective cut-offs, not a retrospective repeal, so assessments, appeals and refund claims for earlier consideration continue under the provisions as they stood. For income received on or after 1 April 2025, commentary reports that the section 10(50) exemption, which protected income already subject to the levy, no longer applies from assessment year 2026-27. The ordinary charging provisions and any treaty therefore become the operative analysis for those receipts.

A decision sequence for a foreign platform with Indian users or customers

First, list each revenue stream and the contract under which it arises, such as advertising inventory, software or content licences, hosting, data and hardware. Second, identify who pays: an Indian customer or group company triggers a withholding question under section 393(2) of the 2025 Act, while a non-Indian payer may not. Third, test the domestic charge: business connection including SEP, royalty and fees for technical services.

Fourth, test treaty entitlement: tax residency certificate, Form 41 (formerly Form 10F), beneficial ownership and any limitation of benefits or principal purpose test. Fifth, if the treaty applies, test permanent establishment on the facts: a fixed place, equipment or servers at the enterprise’s disposal, a dependent agent, and the functions of any local group company. Sixth, attribute profit: only profits attributable to a permanent establishment under Article 7 where a treaty applies, or income attributable to the SEP nexus under domestic law where it does not. Seventh, record the conclusion and the trigger for a fresh review, such as payments or user numbers approaching the rule 13 thresholds.

Common mistakes in digital-tax files after the levy

The first is treating the end of the levy as the end of exposure. The second is quoting SEP thresholds from earlier years or from commentary instead of rule 13. The third is assuming that SEP creates a treaty permanent establishment, which it does not unless the treaty itself provides so. The fourth is calling every digital receipt royalty because of the words ‘platform’ or ‘subscription’, when the Supreme Court in Engineering Analysis Centre of Excellence Pvt. Ltd. v. CIT looked at the rights actually granted in the software.

The fifth is treating the Indian subsidiary’s local activity as irrelevant because customer contracts are accepted through an offshore interface. The sixth is preparing the direct tax conclusion, the GST analysis and the transfer pricing file separately, so that each asserts a different set of facts.

Conclusion

The digital tax question after the equalisation levy is not whether India has ceased taxing offshore platforms. It is which statute and treaty article now governs each identifiable revenue stream.

Commercial reality, contract architecture and evidence of local activities will decide more than broad descriptions of online business. The strongest analysis is built around the transaction, not the business label.

Frequently asked questions

1. Does the equalisation levy still apply in October 2026?

No. The Income-tax Department states that its provisions are not applicable from 1 April 2025.

2. Does SEP automatically create a treaty PE?

No. Treaty PE is determined under the treaty, subject to its application and anti-abuse provisions.

3. Must every digital receipt be characterised as royalty?

No. The rights granted and services performed must be reviewed separately.

4. Can an Indian subsidiary create PE risk for a foreign platform?

It can on particular facts, but common ownership alone is not sufficient.

5. Can historic equalisation-levy periods be ignored?

No. Prior periods may remain relevant for compliance or disputes under the law then applicable.

6. What are the significant economic presence thresholds in India for 2026?

Rule 13 of the Income-tax Rules, 2026 prescribes aggregate payments of Rs 2 crore in the tax year from transactions with persons in India, or 3 lakh users with whom business is systematically solicited or who interact with the non-resident. The rule carries forward former rule 11UD and is linked to section 9(9)(d) of the Income-tax Act, 2025.

7. Is the equalisation levy abolished in India?

It no longer applies to new consideration. The 2 per cent e-commerce levy stopped applying from 1 August 2024, and the 6 per cent levy on specified services such as online advertisement from 1 April 2025. Earlier periods remain subject to the provisions as they stood, so assessments and disputes for those periods can continue.

8. Does significant economic presence create a permanent establishment under a tax treaty?

No. Significant economic presence is a domestic-law limb of business connection under section 9. A treaty permanent establishment is determined by the treaty’s own Article 5, and Article 7 limits India’s right to tax business profits to those attributable to a permanent establishment. A treaty-eligible enterprise therefore needs both analyses, and entitlement to treaty benefit must be documented.

Related reading from SBC

• Smart TVs, Digital Advertising and Cross-Border GST: Separate the Hardware from the Platform
• Defending the Residual Profit Split Method in India When the TPO Proposes TNMM for an Intangible-Owning Entity
• Attributing Profits to an Indian Project Office: Which Head Office Expenses Count?
• Shared Services from Singapore: Applying the ‘Make Available’ Test to Indian Withholding
• More articles on the SBC blog

Disclaimer

This article is general professional commentary and not advice on any particular facts. Treaty eligibility, applicable law, procedural requirements and statutory amendments must be verified for the relevant tax period. It is not a description of an identifiable client engagement.

Parent Guarantee of an Indian Subsidiary: How Transfer Pricing, Withholding and GST Fit Together

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Advisory

Introduction

A parent company that guarantees its subsidiary’s bank borrowing has done something commercially real. The lender’s risk is reduced, the subsidiary borrows on better terms than it could obtain alone, and the parent has assumed a contingent liability it did not previously carry. Indian tax law responds to that in three places at once, and the three do not coordinate with one another. Each asks its own question, each reaches its own number, and each carries its own penalty and interest consequences.

The commercial amounts involved are usually modest, but the exposure is not, because each regime carries its own penalty and interest consequences, and because a position taken in one regime becomes evidence in the others. Different transfer-pricing and GST values can be justified, but the GST computation must also test the full-input-tax-credit proviso to Rule 28(2) and the documentation for the particular guarantee.

This note deals with how the three fit together on a single guarantee. It covers whether the transaction requires arm’s length pricing at all, how a defensible fee is established and why the familiar one per cent figure is not a legal standard, how the fee is characterised for withholding under a treaty, and how the prescribed GST valuation operates on a related party guarantee whether or not anything is charged. It also explains why the transfer pricing conclusion and the GST figure can legitimately differ, which is the point on which files most often become internally inconsistent. For the related question of how intra group service charges are characterised for withholding, see our note on the make available test in shared services arrangements.

1. Is a guarantee an international transaction at all

Section 163 of the Income-tax Act, 2025 (section 92B of the 1961 Act, the meaning of international transaction) defines the transactions to which the arm’s length requirement attaches. The corresponding provision in the 1961 Act was expanded by the Finance Act 2012, with retrospective effect, by an Explanation that lists guarantees among capital financing transactions. The 2025 Act carries the same treatment forward, so a guarantee remains within the categories the provision names, and the argument described below is about whether naming is sufficient rather than about whether the category exists.

Whether that listing is sufficient by itself has been genuinely contested, and a blog that presents the area as settled will mislead.

In Bharti Airtel Ltd. v. ACIT, the Delhi Bench held in March 2014 that a transaction is an international transaction only if it has a bearing on the profits, income, losses or assets of the enterprise, that this requirement applies even to transactions named in the retrospective Explanation, that the impact must be real rather than contingent or hypothetical, and that the onus lies on the Revenue. On that reasoning a guarantee that costs the guarantor nothing falls outside the provision.

The Ahmedabad Bench reached a similar result in Micro Ink Ltd. v. ACIT, ITA No. 2873/Ahd/10, holding that a parent’s guarantee to its subsidiary is a shareholder activity in the nature of quasi capital rather than the provision of a service, and therefore outside the limb of the provision dealing with the provision of services.

The other line is at least as strong. In CIT v. Everest Kanto Cylinder Ltd., ITA No. 1165 of 2013, pronounced on 8 May 2015 by Dharmadhikari and A.K. Menon JJ, the Bombay High Court proceeded on the footing that the guarantee was within the provision and decided the case on benchmarking. In Advanta India Ltd. v. ACIT, IT(TP)A No. 1643/Ban/12, decided on 30 June 2015, the Bangalore Bench held that a guarantee involving a cost to the assessee does fall within the provision and distinguished Bharti Airtel on that basis.

The practical conclusion for an adviser is straightforward. Arguing that no international transaction exists is available on the right facts, particularly where the guarantee was given purely in a shareholder capacity and cost the guarantor nothing. It is not a position to build a compliance strategy on. Most groups should benchmark and document, and keep the threshold argument in reserve.

2. What an arm’s length guarantee fee looks like

Transfer pricing officers have historically applied a one per cent commission as a default. That figure is an administrative convention. It is not in the statute, it is not in the rules, and it has been displaced on evidence.

Everest Kento Cylinders is the authority that matters, and it should be stated precisely because it is often overstated. The Bombay High Court dismissed the Revenue’s appeal for want of a substantial question of law. It held that the transfer pricing officer’s comparison of a holding company’s corporate guarantee with commercial bank guarantee rates was not a comparison between like transactions, and that the commission the assessee had charged could not be called in question in the manner the officer had done. The Court declined to interfere with the deletion of the adjustment. It did not itself fix an arm’s length rate, and it did not hold that 0.5 per cent is the arm’s length rate. What it establishes is that the administrative benchmark can be displaced, and that bank guarantee pricing is the wrong comparable.

The reasoning matters more than any number. A bank issuing a guarantee is selling a product, carries regulatory capital against it and prices for profit. A parent guaranteeing its own subsidiary is protecting an investment it already owns. The two are not in the same market.

Two methodologies are used to displace the default, and they answer slightly different questions.

The benefit approach measures what the borrower gained. It compares the interest rate the subsidiary obtained with the guarantee against the rate it would have obtained on its own standalone credit profile, and treats some share of the saving as the arm’s length fee. Its strength is that it is grounded in an observable commercial benefit. Its weakness is the counterfactual rate, which has to be established rather than assumed, ideally from a bank’s own indicative pricing or from a credit assessment.

The expected loss approach measures what the guarantor assumed. It estimates the probability that the borrower defaults, estimates the proportion of exposure that would be lost on default, and derives the expected loss the guarantor has taken on. Its strength is that it prices the risk actually borne, which is what a guarantor would charge for. Its weakness is that each input requires support, and an unsupported default probability carries no more weight than a bare assertion.

Either approach can be defended. Neither can be defended as a conclusion without the working. The file should contain the credit analysis, the source of each input, the sensitivity of the result to the main assumptions, and the reason the chosen approach fits the transaction.

3. One further point on quantum: guaranteed versus drawn

A guarantee is frequently issued for a facility limit well above what the borrower actually draws. For transfer pricing, the better view is that the guarantor’s exposure is to the amount outstanding from time to time, and pricing on the drawn amount reflects the risk actually assumed. For GST, as set out below, the prescribed value is expressly tied to the amount guaranteed rather than the amount disbursed. The two regimes diverge on this point by design, and a file should say so rather than leave the inconsistency to be discovered.

4. Withholding on the fee

Where a fee is actually paid to the foreign guarantor, the payer must consider section 393(2), Table serial number 17 of the 2025 Act (section 195, withholding on sums payable to non-residents), and the treaty characterisation that determines the rate.

The characterisation question is live and has produced an important decision that is widely miscited. In Metso Outotec OYJ (earlier Outotec Oyj) v. Dy. CIT, (2023) 203 ITD 79 / (2024) 227 TTJ 715 (Kol Trib), the Kolkata Bench held that a corporate guarantee fee arose in India and was taxable as Other Income under Article 21 of the India Finland treaty, on the reasoning that giving the guarantee was a shareholder obligation protecting the subsidiary’s business interests rather than part of the guarantor’s regular business activity.

Two points about that decision. First, it is a treaty characterisation case. It is not a transfer pricing case, it decided nothing about benchmarking, and it laid down no arm’s length rate. It is cited for the latter proposition with some frequency and should not be. Second, its actual holding is adverse and important: where a guarantee fee is characterised as other income rather than business profits, a taxpayer relying on the absence of a permanent establishment to claim nil taxability may find that the article it is relying on does not apply.

The characterisation therefore has to be worked out against the specific treaty, and the analysis should address the residuary article, not stop at business profits. Where a fee is payable, the treaty documentation requirements apply in the usual way: residency certificate, Form 10F and a beneficial ownership declaration.

5. The GST position

Section 15(1) of the CGST Act values a supply at the transaction value where the supplier and recipient are not related and price is the sole consideration. Where they are related, value is determined under the rules.

Rule 28(2) was inserted by Notification No. 52/2023-Central Tax dated 26 October 2023 and retrospectively amended by Notification No. 12/2024-Central Tax dated 10 July 2024 with effect from 26 October 2023. Circular No. 225/19/2024-GST dated 11 July 2024 sets out how it operates. The points that matter commercially are these.

• The value is one per cent of the amount guaranteed per annum, or the actual consideration, whichever is higher. Where a fee is actually charged above that figure, the fee is the value.
• The base is the amount guaranteed, not the amount disbursed. A guarantee of a facility limit is valued on the limit.
• Rule 28(2) applies to guarantees issued or renewed on or after 26 October 2023. Guarantees issued before that date are valued under the pre-existing Rule 28.
• Each renewal is a fresh supply.
• For a multi year guarantee, the one per cent is multiplied by the number of years, tax is paid at issuance, and part years are pro-rated.
• Where the recipient is eligible for full input tax credit, the invoice value is deemed to be the value of supply.
• Co-guarantors: where the aggregate actual consideration exceeds one per cent of the guaranteed amount, that aggregate is the value; otherwise each pays on one per cent of the amount it guarantees.
• A loan takeover alone is not a taxable event unless a fresh guarantee is issued or an existing one renewed.
• Rule 28(2) does not apply where the recipient of the guarantee service is located outside India, so an outbound guarantee is outside it.

Where the guarantor is outside India and the recipient is the Indian subsidiary, the supply is an import of services and the Indian entity discharges tax under reverse charge. Where the Indian entity is eligible for full credit, the cash effect nets out, but the compliance is not optional and the credit depends on correct discharge and documentation.

6. Why the two references to one per cent are not the same figure

This is the point on which files most often become internally inconsistent, so it is worth stating plainly.

The transfer pricing one per cent is an administrative benchmark with no statutory basis, applied by officers in the absence of better evidence, and displaceable on evidence as Everest Kanto shows. The GST valuation mechanism is subject to the full-input-tax-credit proviso and is not displaced merely by demonstrating a lower arm’s-length fee by showing that an independent party would have charged less.

Accordingly, a defensible transfer-pricing fee can differ from the GST value calculated for the same guarantee, once the full-credit proviso and other applicable conditions have been considered. That is not an inconsistency. The transfer pricing file should record why, so that the GST position is not later produced as an admission that the arm’s length fee was one per cent.

7. A worked example

Assume a foreign related entity guarantees a borrowing facility for an Indian group company, with loan drawdowns differing from the face amount guaranteed. This example is hypothetical; the guarantee terms, eligibility for input tax credit and the documentation of any fee must be examined independently.

Transfer pricing. For periods governed by the Income-tax Act, 1961, the relevant accountant’s report is Form 3CEB under section 92E; tax years governed by the Income-tax Act, 2025 use section 172 and Form No. 48 under rule 85 of the 2026 Rules. Benchmark the transaction by reference to borrowing benefits and guarantee risk using independently sourced inputs. Neither a reported judicial percentage nor a GST percentage is a substitute for this exercise.

Withholding. Where a fee is paid, the treaty characterisation is worked out against the applicable agreement, addressing both the business profits article and the residuary article, with Metso Outotec noted as the reason the residuary article cannot be ignored. Treaty documentation is obtained before payment.

GST. For a covered related-party guarantee issued or renewed on or after 26 October 2023, apply Rule 28(2) and Circular No. 225/19/2024-GST, including the proviso deeming the invoice value to be the supply value where the recipient is eligible for full input tax credit. Otherwise, the prescribed base ordinarily depends on the amount guaranteed, not just the amount drawn; the guarantee term, renewal date, any reverse-charge liability and actual credit eligibility must be documented.

Consistency. A single note records all three positions, explains why the transfer pricing rate and the GST deemed value differ, and is retained in the file. It costs an hour and is worth considerably more than that at assessment.

8. Where these arrangements go wrong

First, no fee is charged and nothing is documented. The absence of a fee is a defensible position on the right facts, but it must be a reasoned position recorded at the time, not a gap.

Second, a rate is adopted without working. A file that states a rate of half a per cent and contains nothing behind it is in a worse position than one that adopts the default, because it has invited the question without answering it.

Third, the GST obligation is missed entirely. Rule 28(2) applies whether or not consideration is charged, and groups that charge nothing frequently assume there is nothing to value.

Fourth, renewals are not tracked. Each renewal is a fresh supply, and a guarantee rolled over annually generates an annual liability that is easy to overlook because nothing new appears to have happened.

Fifth, the base is wrong. GST is valued on the amount guaranteed. Computing it on the drawn amount understates the liability, sometimes by a wide margin where the facility limit is far above utilisation.

Sixth, the treaty analysis stops at business profits. Metso Outotec shows that the residuary article can apply, and a withholding position that has not addressed it is incomplete.

Seventh, the three positions are prepared by three teams and never reconciled. This is the most common failure and the easiest to fix.

9. A review checklist

1. Identify every guarantee, letter of comfort and similar instrument given or received within the group, with issue date, term, amount guaranteed and amount drawn.
2. Determine for each whether Rule 28(2) applies, by reference to the issue or renewal date against 26 October 2023.
3. Compute the GST deemed value on the amount guaranteed, pro-rating part years and multiplying for multi year terms, and confirm the reverse charge discharge and credit position.
4. Diarise renewals as fresh supplies.
5. Benchmark the fee for transfer pricing using a documented benefit or expected loss analysis, with sources for every input.
6. Record why the transfer pricing rate differs from the GST deemed value.
7. Work out the treaty characterisation of any fee actually paid, addressing the residuary article.
8. Obtain treaty documentation before payment.
9. Report the transaction in Form 3CEB under section 172 (section 92E).
10. Consider whether the shareholder activity argument is available on the facts, and record the position either way.
11. Reconcile the transfer pricing, withholding and GST positions in a single note.
12. Review the position at each renewal rather than treating the original analysis as permanent.

Practitioner’s perspective

Corporate guarantees are among the smallest transactions a group reports and among the most reliably mishandled, because they fall between three teams and nobody owns them. The tax cost of getting the fee wrong is usually modest. The cost of missing the GST valuation across several years, or of taking inconsistent positions that are then produced against each other, is not.

Two things are worth emphasising. The one per cent transfer pricing benchmark is an administrative convention and Everest Kanto shows it can be displaced on evidence, but only on evidence; an unsupported lower rate is worse than the default. And the prescribed GST valuation is ordinarily based on the guarantee amount rather than the drawn amount, subject to the full-credit proviso, and afresh on every renewal. Those three features between them account for most of the exposure we see.

Where a guarantee is already in place and none of this has been done, the right sequence is to establish the GST position first, because it accrues with time and interest runs, and then to build the transfer pricing file.

Direction of the guarantee: where Everest Kanto fits and where it does not

In Everest Kanto the assessee was an Indian holding company that charged its foreign subsidiary a commission of 0.5 per cent on a guarantee supporting the subsidiary’s bank borrowing, and the transfer pricing officer substituted three per cent by reference to commercial bank guarantee rates. That is an outbound guarantee: the Indian entity was the guarantor and the question was whether it had earned enough.

The facts in this article run the other way. A foreign parent guarantees an Indian subsidiary’s borrowing, so the Indian entity is the beneficiary and the question is whether it should pay a fee at all, and how much. The comparability reasoning transfers, because a holding company protecting an investment is not selling a bank product. The numerical result does not transfer: a 0.5 per cent rate accepted on one set of facts is not a benchmark for another. For an inbound guarantee the analysis should be built on the subsidiary’s own credit position and the borrowing cost it would have faced without the guarantee, which is why the benefit approach needs evidence of the standalone rate.

Frequently asked questions

1. Is a parent guarantee an international transaction requiring arm’s length pricing?

The position is contested. Bharti Airtel and Micro Ink support the view that a guarantee with no bearing on the guarantor’s profits, income, losses or assets falls outside the provision, treating it as a shareholder activity. Everest Kanto and Advanta India proceed on the footing that it is within. Most groups should benchmark and document, and keep the threshold argument in reserve.

2. Is the arm’s length guarantee fee one per cent?

No. One per cent is an administrative benchmark with no statutory basis. In Everest Kanto the Bombay High Court dismissed the Revenue’s appeal, leaving undisturbed the Tribunal’s deletion of an adjustment based on a three per cent rate derived from commercial bank guarantee pricing, which the Court held was not comparable to a holding company guarantee. It did not itself fix 0.5 per cent as the arm’s length rate.

3. What methodologies are used to price a guarantee?

The benefit approach, which measures the interest saving the borrower obtained and allocates a share of it, and the expected loss approach, which estimates the probability of default and the loss given default to price the risk assumed. Either is defensible with supporting work; neither is defensible as a bare conclusion.

4. Is GST payable where no guarantee fee is charged?

Yes. For covered related-party guarantees, Rule 28(2) specifies one per cent of the guaranteed amount per annum or actual consideration, if higher. Where the recipient is eligible for full input tax credit, the proviso deems the invoice value to be the value of the supply; that proviso must be tested separately.

5. Is GST valued on the amount guaranteed or the amount drawn?

On the amount guaranteed. Circular No. 225/19/2024-GST confirms that the value depends on the amount guaranteed, not the amount disbursed.

6. Does each renewal attract GST again?

Yes. Circular No. 225/19/2024-GST treats each renewal as a fresh supply. A guarantee rolled over annually generates an annual liability.

7. What about guarantees issued before 26 October 2023?

Rule 28(2) applies to guarantees issued or renewed on or after 26 October 2023. Those issued earlier are valued under the pre-existing Rule 28, though a renewal after that date brings the renewal within the new rule.

8. How is a guarantee fee characterised for withholding under a treaty?

It depends on the treaty. In Metso Outotec the Kolkata Bench held a guarantee fee taxable as other income under Article 21 of the India Finland treaty, on the basis that giving the guarantee was a shareholder obligation rather than part of the guarantor’s regular business. A withholding analysis that addresses only business profits is incomplete.

9. Does Metso Outotec establish an arm’s length guarantee rate?

No. It is a treaty characterisation case and decided nothing about benchmarking or rates. It is sometimes cited for a transfer pricing proposition it does not contain.

10. Can the transfer pricing rate and the GST value legitimately differ?

Yes. The transfer pricing rate is the arm’s length price established on evidence. The GST figure is a prescribed deemed value that applies regardless of what an independent party would have charged. The file should record why they differ so that the GST figure is not later treated as an admission.

11. Who pays GST on a guarantee given by a foreign parent to its Indian subsidiary?

The Indian subsidiary pays, under reverse charge, because the supply is an import of services from a related person outside India. Where it is eligible for full input tax credit, the invoice value is deemed to be the value of supply under the proviso to Rule 28(2), so the cash effect can net out, but discharge and documentation must still be correct.

12. Which section of the Income-tax Act, 2025 covers corporate guarantees as international transactions?

Section 163 of the Income-tax Act, 2025, titled ‘Meaning of international transaction’, which corresponds to section 92B of the 1961 Act. Its capital financing limb includes any type of long-term or short-term borrowing, lending or guarantee. Arm’s length pricing follows under section 165, documentation under section 171 and the accountant’s report under section 172.

13. Is tax deducted at source on a guarantee commission paid to a foreign parent?

Where a fee is actually paid to a non-resident guarantor, the payer must consider deduction under section 393(2) of the Income-tax Act, 2025 (section 195 of the 1961 Act). The treaty characterisation, including the residuary article, decides whether and how much tax is chargeable. Treaty documents should be obtained before payment.

Related reading from SBC

• Shared Services from Singapore: Applying the ‘Make Available’ Test to Indian Withholding
• Employee Secondment and GST: Reading Northern Operating Systems with the CBIC Clarifications
• No FTS Article in the India-UAE Treaty: How Offshore Engineering Income Is Taxed in India
• Seconding Staff Between India and the UK: Employer, Permanent Establishment and Payroll Tax Are Different Questions
• More articles on the SBC blog

Disclaimer

This article is for general information and does not constitute professional advice. It reflects the law as understood at the date of publication and does not take account of later developments. Statutory references give the Income-tax Act, 2025 section with the Income-tax Act, 1961 section in brackets; section 536 preserves the 1961 Act for proceedings pending on 31 March 2026 and for earlier tax years. The transfer pricing treatment of guarantees is an area in which tribunal authority is divided, and outcomes turn on the specific facts and documentation. No reader should act on the basis of this article without taking advice on their own circumstances.

Shared Services from Singapore: Applying the ‘Make Available’ Test to Indian Withholding

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Advisory

Introduction

Most tax treaties India has entered into since the 1990s qualify the definition of fees for technical services with a condition borrowed from the United States model: the service must make something available to the recipient. The condition exists because the drafters distinguished between buying expertise and acquiring it. A payment for the first is a payment for a service and belongs to the provider’s business profits. A payment for the second is closer to a payment for knowledge, and the source State is given a limited right to tax it.

Where that condition applies, it governs the withholding treatment of most regional shared services arrangements into India. It is also routinely applied as though it were a question about the whole agreement, when it is a question about each service. A regional headquarters supplying several categories of support will frequently find that six fall outside the definition and one falls inside, and the correct answer is not a single rate applied to the whole invoice.

This note deals with how that condition applies to an intra group shared services arrangement, where a regional headquarters supplies several distinct categories of support to an Indian subsidiary under a single agreement and a single fee. It covers the treaty text and its three limbs, the exclusions that are frequently overlooked, a service by service framework, the two withholding approaches available and what each requires by way of evidence, the gross up problem, and the transfer pricing analysis that has to agree with whatever conclusion is reached. Where the treaty contains no technical services article at all, a different analysis applies, and that is dealt with in our note on treaties with no fees for technical services article.

1. The treaty text

Article 12(4) of the India Singapore agreement defines fees for technical services as payments of any kind to any person in consideration for services of a managerial, technical or consultancy nature, including the provision of such services through technical or other personnel, if such services:

(a) are ancillary and subsidiary to the application or enjoyment of the right, property or information for which a payment described in paragraph 3 is received; or

(b) make available technical knowledge, experience, skill, know-how or processes, which enables the person acquiring the services to apply the technology contained therein; or

(c) consist of the development and transfer of a technical plan or technical design, but excludes any service that does not enable the person acquiring the service to apply the technology contained therein.

Two drafting features carry most of the weight, and the second is frequently misread.

The limbs are disjunctive, so a service caught by any one of them is within the definition. It follows that the enabling condition is not a gateway through which every payment must pass. It governs limbs (b) and (c), where it appears expressly and, in limb (c), again as an exclusion. Limb (a) contains no enabling requirement at all. A service that is ancillary and subsidiary to the application or enjoyment of a right, property or information for which a royalty under paragraph 3 is paid is within the definition whether or not anything is made available. Where a group licenses technology and also supplies support in relation to it, limb (a) has to be tested first, and a conclusion that nothing was made available does not answer it.

Article 12(2), as substituted by the 2005 Protocol, caps the source state rate at ten per cent of the gross amount where the recipient is the beneficial owner.

2. The exclusions in Article 12(5), which are frequently overlooked

Article 12(5) removes a list of payments from the definition notwithstanding Article 12(4). The exclusions cover payments for services ancillary and subsidiary, and inextricably and essentially linked, to a sale of property; services ancillary and subsidiary to the rental of ships, aircraft, containers or other equipment used in international traffic; teaching in or by educational institutions; services for the personal use of the individual making the payment; payments to an employee of the payer, or to an individual or firm of individuals other than a company for professional services as defined in Article 14; services connected with an installation or structure used for exploration or exploitation of natural resources; and services referred to in paragraphs 4 and 5 of Article 5.

The exclusion for payments to an individual or a firm of individuals for professional services is worth noting in any arrangement where part of the support is delivered by a professional firm rather than by the group entity. So is the sale related exclusion, where support services accompany a supply of goods.

3. What “make available” actually requires

The test is not whether the service is technical. It is not whether the recipient benefits. It is not whether the recipient’s staff learn something incidentally. The question is whether, after the engagement, the recipient is equipped to apply the technology or process itself, without recourse to the provider.

Three propositions follow, and each is contested often enough to be worth stating.

Recurrence is evidence but it is not proof. Where the same service is bought year after year, that is consistent with nothing having been transferred, and it is a useful fact to put forward. It is not conclusive, because a recipient may acquire a capability and still choose to outsource its exercise for commercial reasons.

A deliverable is not itself a transfer of capability. A report, a recommendation, a signed off control assessment or a market analysis is the product of the provider’s expertise, and handing it over transfers the output rather than the method that produced it.

Training is the difficult category. A structured programme that teaches a methodology, supported by manuals, templates and process documentation which the recipient retains and uses afterwards, is the clearest case of a service that does make available. An induction briefing, or an update session on a change of policy, is not. Between those two the answer turns on whether the recipient keeps and uses the materials and can run the process unaided afterwards.

4. A service by service framework

The analysis has to be conducted category by category. The following reflects how a typical regional support arrangement resolves, and the reasoning is more important than the conclusion, because the conclusion depends on how the particular service is actually delivered.

Service category	Typical analysis	Indicative characterisation
Strategic management oversight	Board level direction and review. The provider exercises judgment on behalf of the group; no methodology passes to the Indian entity.	Business profits
Treasury and liquidity advisory	The provider runs the analysis and recommends a position. The Indian entity acts on the recommendation but cannot replicate the modelling.	Business profits
Human resources strategy and leadership development	Coaching and advisory delivered to individuals. Capability accrues to the person, not to the entity as a transferable process.	Business profits
Legal and compliance advisory	Advice on specific questions. The answer is supplied; the ability to answer the next question is not.	Business profits
Internal audit and risk assurance	An assurance opinion on the recipient’s processes. The output is the assurance, and the audit methodology stays with the provider.	Business profits
Sustainability and corporate responsibility reporting support	Assistance in preparing a report to a framework the recipient does not own.	Business profits
Internal control and compliance training programmes	Where structured training is delivered with retained manuals, control matrices and documented process frameworks the recipient then operates itself, the enabling condition is met.	Fees for technical services, in whole or part

The point of the table is not the labels. It is that a single arrangement can straddle the line, and that the characterisation of each category follows from how the service is delivered and what the recipient is left holding.

5. Two withholding approaches, and how to choose

Once the analysis is done, the group has a decision to make, and it is a decision about risk appetite and documentation capability rather than about law.

The conservative approach applies the treaty rate to the whole of the service fee. It is simple, it is certain, and it costs money on the services that are not fees for technical services. Where the recipient is entitled to credit in its own jurisdiction the cost may be partly recovered, but where the contract requires the Indian entity to bear the tax, the cost is real and falls on India.

The segregated approach applies withholding only to the categories that satisfy the enabling condition. It reflects the correct legal position and it is defensible, but only if the supporting record exists. That record is specific.

• A service schedule in the agreement that describes each category separately, with its own scope and its own deliverables.
• Invoices that bill each category separately rather than as a single monthly support fee.
• Time records or an allocation basis that ties cost to category.
• Deliverables retained and filed, so that the nature of what was supplied can be shown rather than asserted.
• For any category treated as fees for technical services, the materials that evidence the transfer, since that category must also be capable of being proved.
• A contemporaneous analysis note recording the characterisation of each category and the reasons for it.

A group that bills a single consolidated support fee and then argues for segregation at assessment stage is in a weak position, because the only evidence of what was supplied is the invoice, and the invoice says “support services”.

6. The gross up problem

Many intra group service agreements require the Indian recipient to bear any withholding tax arising in India. The commercial intention is that the provider receives the agreed fee net. The tax consequence is that the Indian entity funds the tax from its own resources, and the grossed up amount, not the invoice amount, becomes the base.

Two points follow. First, the clause should be read before the first invoice is raised, not after. Where the characterisation analysis shows that most categories are business profits, the gross up exposure is confined to the minority that is not, and the clause can be scoped accordingly. Second, a gross up clause interacts with the transfer pricing position. If the Indian entity bears a cost that an independent party in its position would not bear, the arm’s length character of the arrangement is in question, and the two analyses need to be consistent.

7. The transfer pricing dimension

A shared services arrangement between associated enterprises is an international transaction under section 163 of the Income-tax Act, 2025 (section 92B of the 1961 Act, the meaning of international transaction). It requires arm’s length pricing under section 165 (section 92C, determination of arm’s length price), documentation under section 171 (section 92D, maintenance of information and documents) and the accountant’s report under section 172 (section 92E, the Form 3CEB report).

Three questions arise in sequence and they are often answered out of order. Were the services actually rendered? Did the recipient derive a benefit? Is the charge arm’s length? The first two are evidential and are where most intra group service adjustments are made. A benefit test is not satisfied by a management assertion; it is satisfied by evidence of what was delivered and what the recipient would otherwise have had to do.

Shareholder activity should be identified and excluded at the outset. Costs incurred by the parent in its capacity as shareholder, such as the cost of group consolidation for the parent’s own reporting or investor relations, are not chargeable to the subsidiary at all. A charge that includes them invites an adjustment on a basis that has nothing to do with the mark up.

8. Treaty documentation

Treaty entitlement requires a valid tax residency certificate for the relevant period, Form 10F filed electronically (for tax years under the Income-tax Act, 2025, its successor Form 41, furnished under rule 75 of the Income-tax Rules, 2026 for section 159(8)) with particulars matching the certificate, and a declaration of beneficial ownership and absence of a permanent establishment in India. The permanent establishment question deserves more attention than it usually receives in a shared services context. Where provider personnel spend time in India delivering the services, the service permanent establishment threshold in Article 5 becomes live, and a day count should be maintained for the same reason it is maintained in any cross border service arrangement.

The limitation of benefits provisions of the India Singapore treaty should also be considered, because entitlement under that agreement has historically carried conditions that do not appear in other treaties.

9. A worked example

A Singapore regional headquarters provides seven categories of support to its Indian subsidiary under an agreement effective from the start of a financial year, on a cost plus basis billed monthly. The agreement requires the Indian entity to bear Indian withholding tax. The annual fee is material relative to the Indian entity’s profit.

Step one. Break the fee into the seven categories and establish, for each, what was delivered and what the Indian entity retained. This is done by reading the deliverables, not the agreement.

Step two. Apply the enabling condition to each category. On the facts described above, six categories produce outputs and decisions; one, the structured internal control training with retained manuals and control matrices, transfers a process the Indian entity then operates itself.

Step three. Quantify the two approaches. Under the conservative approach the treaty rate applies to the whole fee and the Indian entity funds it grossed up. Under the segregated approach it applies only to the training category. The difference is the decision the group is actually making.

Step four. Build the record that the segregated approach requires: restructure the invoicing to bill by category, put a service schedule into the agreement, retain deliverables, and prepare a characterisation note before the first payment rather than after the first notice.

Step five. Align the transfer pricing file, excluding shareholder costs, evidencing rendition and benefit, and ensuring the Form 3CEB description of the transaction matches the characterisation adopted for withholding.

10. Where these arrangements are usually weak

First, the agreement describes the services in a single clause. A schedule that lists seven categories with their own scopes is the foundation of the whole position, and it is cheap to draft before execution and impossible to create afterwards.

Second, invoicing is consolidated. A single line for regional support fees makes segregation unprovable whatever the underlying facts.

Third, deliverables are not retained. The reports, control matrices and advice notes that prove what was actually supplied are held by the provider, or by individuals, and cannot be produced three years later.

Fourth, the withholding position and the transfer pricing position are prepared by different teams and do not agree. A transaction described one way in Form 3CEB and another way in the withholding analysis is an invitation.

Fifth, shareholder costs are embedded in the cost pool. This is the most common cause of adjustment in intra group service cases and it is entirely avoidable.

Sixth, no day count is kept for provider personnel visiting India. A service permanent establishment finding would displace the entire analysis, because the income would then be attributable to a permanent establishment and taxable on a net basis under Article 7 regardless of characterisation.

11. A review checklist

1. Read the current text of the applicable treaty, including protocols and the synthesised text, and confirm the wording of the technical services article.
2. List every service category actually supplied, from the deliverables rather than from the agreement.
3. For each, record what the recipient was left holding and whether it can now perform the activity unaided.
4. Check each category against the exclusions in Article 12(5) before concluding.
5. Decide between the conservative and segregated approaches, and record the reasons.
6. If segregating, restructure the agreement schedule and the invoicing before the next payment cycle.
7. Identify and remove shareholder activity costs from the pool.
8. Evidence rendition and benefit for the transfer pricing file.
9. Reconcile the withholding characterisation with the Form 3CEB description.
10. Maintain a day count for provider personnel present in India and monitor it against the treaty threshold.
11. Confirm the residency certificate, Form 10F and beneficial ownership declaration for each year.
12. Model the gross up exposure under each approach and confirm the commercial terms reflect the intended allocation.

Practitioner’s perspective

The enabling condition is one of the more taxpayer favourable provisions in India’s treaty network, and it is frequently under used because groups apply a single rate to the whole arrangement for administrative convenience. The convenience is real, and for a small arrangement it may be the right answer. For a material arrangement it is an avoidable cost.

What determines whether the better answer is available is almost entirely a documentation question, and it is decided at contracting stage. A service schedule, category level invoicing and retained deliverables cost very little to put in place at the start of an arrangement and cannot be created once a notice has issued. In our experience the groups that defend a segregated position successfully are not the ones with the better legal argument. They are the ones whose invoices show what was supplied.

One further point. The analysis should be revisited when the arrangement changes. A category that was advisory when the arrangement began may evolve into something that transfers a methodology, and a characterisation note prepared three years ago and never revisited is a liability rather than a protection.

Treaty status check: protocols, the MLI and anti-abuse rules

The wording of Article 12(4) that this article relies on has not been changed by the Multilateral Instrument. The MLI does not rewrite the definition of fees for technical services. What it adds to the India-Singapore treaty, as a covered tax agreement, is the principal purpose test: treaty benefit can be denied where obtaining it was one of the principal purposes of an arrangement. CBDT Circular No. 1/2025 dated 21 January 2025 explains that the test operates prospectively from the date it takes effect in each treaty, and notes that the Singapore treaty’s own limitation of benefits clause (Article 24A) continues to apply.

The 10 per cent cap in Article 12(2) comes from the Protocol that entered into force on 1 August 2005 and was notified by Notification No. 185/2005 dated 18 July 2005. Before relying on any characterisation, read the current consolidated treaty text on incometaxindia.gov.in, including all protocols, and confirm that no later instrument has altered Article 12 or Article 5.

For a Singapore provider that is a genuine regional headquarters, the entitlement documents (tax residency certificate, Form 10F or Form 41, and a beneficial ownership declaration) are best accompanied by evidence of substance in Singapore: personnel, premises and expenditure. That evidence answers both the limitation of benefits clause and the principal purpose test.

Service permanent establishment: when provider staff visit India

A characterisation of fees as business profits holds only while the provider has no permanent establishment in India. On the published treaty text, Article 5 contains a service PE rule for services furnished through employees or other personnel, with a threshold of more than 90 days in a fiscal year and a shorter threshold of more than 30 days where the services are performed for a related enterprise. Because shared services are by definition supplied to a group company, the shorter threshold is the one to test first.

Keep a day count by individual and by assignment for every provider employee who works in India, including workshops, training sessions and audit visits. If a permanent establishment exists, the income attributable to it is taxed on a net basis under Article 7 and the fees for technical services analysis under Article 12 gives way, because Article 12(6) disapplies Articles 12(1) and 12(2) where the right or contract is effectively connected with the permanent establishment.

Frequently asked questions

1. What does “make available” mean in the India Singapore treaty?

Article 12(4)(b) requires that the service make available technical knowledge, experience, skill, know-how or processes which enables the person acquiring the services to apply the technology contained therein. The recipient must be left able to apply the technology itself, not merely to benefit from the provider having applied it. The condition governs limbs (b) and (c); limb (a) has no such requirement.

2. Is a technical service automatically fees for technical services?

No. The service must be managerial, technical or consultancy in nature and must satisfy one of the three limbs of Article 12(4). A highly technical service that leaves the recipient no more capable than before does not satisfy limb (b). It may still fall within limb (a) if it is ancillary and subsidiary to the enjoyment of a right or information for which a royalty is paid, so limb (a) must be tested separately.

3. Does repeated purchase of the same service prove nothing was made available?

It is good evidence but not conclusive. A recipient may have acquired a capability and still choose to outsource its exercise for commercial reasons. Recurrence should be presented alongside evidence of what was actually delivered.

4. Can one agreement contain both categories?

Yes, and that is the normal case in a regional support arrangement. The analysis is conducted service by service, and it is entirely consistent for most categories to be business profits and one to be fees for technical services.

5. What rate applies where the condition is satisfied?

Article 12(2), as substituted by the 2005 Protocol, caps the source state rate at ten per cent of the gross amount where the recipient is the beneficial owner. Where treaty benefit is not available, the domestic rate applies under section 207 of the 2025 Act (section 115A, the special rate on royalty and fees for technical services in the hands of non-residents).

6. What documentation does a segregated withholding approach require?

A service schedule describing each category separately, invoices raised by category, time or allocation records tying cost to category, retained deliverables, and a contemporaneous note recording the characterisation of each category and the reasons for it.

7. Who bears the tax where the agreement contains a gross up clause?

The Indian entity, from its own resources, and the grossed up amount becomes the base. The clause should be reviewed before the first invoice, and its effect should be consistent with the transfer pricing position.

8. What exclusions apply under Article 12(5)?

Among others, payments for services inextricably linked to a sale of property, for teaching by educational institutions, to an individual or firm of individuals other than a company for professional services as defined in Article 14, and for services connected with an installation used for exploration or exploitation of natural resources.

9. Does the arrangement create a permanent establishment risk?

It can. Where provider personnel are present in India delivering the services, the service permanent establishment threshold in Article 5 becomes relevant. A finding of a permanent establishment would displace the characterisation analysis entirely, because the income would be taxable on a net basis under Article 7.

10. How does the withholding analysis interact with transfer pricing?

The arrangement is an international transaction under section 163 (section 92B) requiring arm’s length pricing, documentation and Form 3CEB reporting. The description of the transaction in the transfer pricing file should match the characterisation adopted for withholding, and shareholder activity costs should be excluded from the charge altogether.

11. Does the MLI change the make available condition in the India-Singapore tax treaty?

No. The Multilateral Instrument does not change the definition of fees for technical services in Article 12(4). It adds the principal purpose test to the India-Singapore treaty, which can deny treaty benefit for arrangements whose principal purposes include obtaining it. The treaty’s own limitation of benefits clause continues to apply alongside it, as CBDT Circular No. 1/2025 notes.

12. Are management fees paid to a Singapore parent taxable in India as fees for technical services?

Not automatically. They are taxable as fees for technical services only if the services fall within a limb of Article 12(4), and for managerial services that usually means the make available test. In Tagit Pte. Ltd. v. Dy. CIT, the Delhi Tribunal held management fees were not fees for technical services where make available was not shown. Each case turns on its facts.

13. What documents are needed to claim India-Singapore treaty relief on service fees paid from India?

The Singapore recipient needs a tax residency certificate for the relevant period and Form 10F, which Form 41 replaces for tax years under the Income-tax Act, 2025 (rule 75 of the Income-tax Rules, 2026). A beneficial ownership declaration and a no-permanent-establishment confirmation are usual practice, and the limitation of benefits clause should be checked.

Related reading from SBC

• No FTS Article in the India-UAE Treaty: How Offshore Engineering Income Is Taxed in India
• Parent Guarantee of an Indian Subsidiary: How Transfer Pricing, Withholding and GST Fit Together
• Seconding Staff Between India and the UK: Employer, Permanent Establishment and Payroll Tax Are Different Questions
• Digital Platforms and India Tax Once the Equalisation Levy Is Gone: SEP, Treaty and PE Compared
• More articles on the SBC blog

Disclaimer

This article is for general information and does not constitute professional advice. It reflects the law as understood at the date of publication and does not take account of later developments. Statutory references give the Income-tax Act, 2025 section with the Income-tax Act, 1961 section in brackets; section 536 preserves the 1961 Act for proceedings pending on 31 March 2026 and for earlier tax years. Treaty positions turn on the specific facts, the contractual documentation and the current text of the applicable agreement. No reader should act on the basis of this article without taking advice on their own circumstances.

Attributing Profits to an Indian Project Office: Which Head Office Expenses Count?

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Litigation / Controversy

Introduction

An engineering contract can be negotiated in one country, designed in another and executed partly through a project office in India. The commercial profit is earned by the combined enterprise, but a tax treaty does not permit the whole amount to be assessed in the source country simply because the work culminates at an Indian site.

The permanent establishment question determines whether India has an Article 7 taxing right. Attribution then asks a different question: what profit is economically connected with the functions, assets and risks of that establishment? The second question is difficult where project-management charges, engineering design, procurement, insurance, finance and central support are booked outside India.

This note concerns the evidence required to defend a claim for expenses incurred by a foreign head office for the purposes of an Indian project office. Its focus is the factual explanation of expenditure and the attribution mechanics, not an assumed right to deduct a percentage of consolidated revenue.

1. The two-stage Article 7 enquiry

First determine whether the enterprise has a permanent establishment under the applicable treaty. The site, fixed-place, service or agency clauses can differ materially across treaties. Only after a PE is established should the practitioner consider profits attributable to it; the mere existence of a PE does not make all receipts of the non-resident taxable in India.

Article 7 generally focuses on business profits attributable to activities of the PE. Read the authentic treaty text, including any protocol, alongside domestic statutory limits. The Income-tax Department treaty repository is a suitable starting point, but the specific convention controls.

2. Map the actual project before selecting a formula

Project records should show where design approvals took place, who procured specialised components, who bore warranty and performance risk, who supervised construction, and who controlled client variations. A head office may perform central technical or commercial functions of real value even when the project office interacts with the customer each day.

An allocation based only on time sheets or employee location can miss responsibility for engineering risk, contract negotiations and technology. Conversely, a foreign head office may hold contractual authority without performing the functions attributed to it. The record must distinguish legal labels from operational conduct.

3. Differentiate direct and shared expenditure

Direct costs can often be traced to project-specific design drawings, subcontractor bills, travel schedules or procurement orders. Shared costs, such as global engineering support and management, require a principled allocation key. Project hours, headcount, contract value or engineering deliverables may each be suitable in particular circumstances; none is automatically correct.

The allocation should be applied consistently, exclude unrelated group costs and reconcile to the foreign entity’s financial statements. A management certificate describing a single arbitrary percentage, without ledger detail and evidence of benefit, invites disallowance.

4. The treaty clause and domestic limitations must be read together

Article 7 treaties do not all contain identical wording on executive and general administrative expenses, reimbursements or dealings between a PE and its head office. Some treaty provisions expressly address expenses incurred outside the source State. The terms of the specific treaty must therefore be set out in the submission rather than inferred from a general OECD description.

Domestic provisions on head office expenditure may impose limits and definitions. The treaty benefit analysis must distinguish a substantive permission to recognise relevant expenditure from any domestic limitation that continues to apply. It is particularly important not to transplant language from another convention into the treaty that actually governs the dispute.

5. Cost reimbursement is not automatically profit attribution

A head office debit note may reimburse a cost, apply a mark-up or purport to remunerate a function. These are different economic claims. A cost charged at actual value needs evidence it was incurred and relates to the PE. A mark-up requires a separate explanation of function, value and the treatment permitted under the relevant treaty.

Intercompany-style transfer pricing principles may inform the functional assessment, but the PE and head office are legally parts of one enterprise. An accounting charge between them does not necessarily have the same tax consequences as a payment between distinct companies.

6. Avoid gross-receipt attribution by default

An assessment sometimes begins with the Indian contract value and applies an estimated profit rate without examining where value was created or what costs were incurred. A presumptive rate, where a particular statutory regime applies, has its own conditions. Outside that regime, an estimation should not replace available project accounts and contemporaneous evidence without explanation.

The taxpayer should reconcile site revenue, milestone billing, advances, change orders, retention amounts and all project costs. The allocation narrative must explain any differences between invoicing location, accounting recognition and functions actually performed.

7. Construct an auditable project file

Useful primary evidence includes the signed EPC or services contract, scope split, tender and approval records, purchase orders, engineering deliverables, project-office trial balance, head office ledgers, time allocation records, travel documents, invoices, inter-office charge notes and proof of payment.

For each claimed expenditure, an indexed schedule should capture cost centre, vendor, service description, period, Indian project connection, allocation method and documentary reference. Do not allow duplicated charges at the project office and head office level; the reconciliation should specifically detect them.

8. Procedural representation and outcome implementation

Explain the governing Article 7 interpretation before debating amounts. Present the functional matrix, followed by reconciled schedules and supporting documents. A short narrative on why a chosen allocation key follows the commercial facts is usually more useful than a large set of vouchers submitted without index.

If an adverse adjustment reaches the appellate stage, distinguish rejection of a legal method from defects in the evidence. Preserve alternate computations where supportable and seek consequential interest and credit relief. After an appellate decision, verify the giving-effect order and the resulting demand ledger.

Section 44C and the Supreme Court’s December 2025 ruling

Section 44C of the Income-tax Act, 1961 limits the deduction a non-resident can claim for head office expenditure against its business or professional income in India. The cap is the lower of five per cent of the adjusted total income (the average adjusted total income where there is a loss) and the head office expenditure attributable to the Indian business. For years under the 1961 Act, any treaty argument for deducting head office costs has to be read against this ceiling.

In Director of Income Tax v. American Express Bank Ltd., decided on 15 December 2025, a Division Bench of the Supreme Court held that head office expenditure incurred outside India falls within section 44C even where it is incurred exclusively for the Indian branches, and that the section restricts deductions rather than conferring them. The earlier Bombay High Court line distinguishing common from exclusive expenditure was not followed. Reports differ on whether the Court also remitted the question of which expenses fall within the enumerated categories, so the judgment should be read before it is relied on. Whether a given treaty’s Article 7 paragraph on expenses overrides the ceiling depends on that treaty’s wording and is a separate argument.

Documentation checklist for a project office attribution file

The file should allow a reviewer to follow each rupee from the Indian contract to the foreign ledger. At a minimum: the signed contract with the scope split and pricing of each component; the Indian project office trial balance and ledgers; the foreign entity’s ledgers and audited accounts; a schedule of every inter-office charge note with the period, the function, the allocation key and proof of payment; engineering deliverables and design approvals showing where the work was performed; purchase orders and subcontractor invoices; and time and travel records for head office personnel who worked on the project.

Add a one-page functional matrix that lists each function (design, procurement, project management, finance, warranty risk) against the entity that actually performed it, with a reference to the supporting documents. The matrix is the bridge between the legal argument on Article 7 and the numbers, and it exposes inconsistencies before the Revenue finds them.

Common mistakes in attribution assessments and appeals

The first mistake is claiming a percentage of contract value as head office cost without ledger detail. The second is treating the project office accounts as the whole answer when the same cost appears in a head office charge, so that it is claimed twice or not at all. The third is arguing the legal method and the evidence in one undifferentiated submission, so that an appellate authority that rejects the method also dismisses the evidence. The fourth is ignoring section 44C, or assuming a treaty deduction removes it, without reading the treaty text.

A fifth mistake is taking inconsistent positions on whether the contract is composite. A taxpayer that argues the contract is indivisible to support a cost deduction cannot comfortably argue elsewhere that the supply and the installation are separate. The Tribunal decisions in Voith Paper and Andritz AG illustrate how the composite nature of a contract cuts in both directions.

How a dispute usually moves forward

For a foreign company, or a non-resident that is not a company, a proposed variation prejudicial to the assessee is first set out in a draft assessment order and can be taken to the Dispute Resolution Panel under section 144C of the 1961 Act (section 275 of the 2025 Act). Where that route does not apply, the first appeal lies to the Commissioner (Appeals) and then to the Tribunal. At each stage, ask for the computation to be redone on a net basis if any profit is held attributable, and keep an alternative computation ready.

Conclusion

PE attribution is not an exercise in choosing the most plausible percentage of project turnover. It is a disciplined account of the business carried on through the Indian establishment, set against the functions retained outside India.

The most persuasive allocation is one a reviewer can reconstruct from contemporaneous records. It identifies the cost, the benefit, the allocation key and the connection with the project, while remaining faithful to the words of the relevant treaty.

Frequently asked questions

1. Does every project office constitute a PE?

No. The treaty definition and the actual duration, premises and activities must be examined.

2. Can engineering work outside India reduce Indian attribution?

It can be relevant where genuinely performed outside India and supported by evidence, subject to the governing treaty and attribution method.

3. Are all global management charges deductible?

No. Purpose, project relevance, allocation, treaty terms and domestic restrictions must be tested.

4. Is a head office invoice sufficient evidence?

No. The underlying functions and actual costs should be substantiated.

5. What is the most useful document in an attribution defence?

A reconciled project-specific cost and function matrix with source records is often the strongest starting point.

6. Does section 44C apply to head office expenses of a permanent establishment in India?

Section 44C of the 1961 Act limits the deduction of head office expenditure by non-residents to the lower of five per cent of adjusted total income and the amount attributable to the Indian business. In American Express Bank (15 December 2025) the Supreme Court held that it applies to expenditure incurred outside India, whether common or exclusive. Treaty wording may add a separate argument.

7. How are profits attributed to a permanent establishment in India under Article 7?

Article 7 allows India to tax only the profits attributable to the permanent establishment, generally computed on a net basis by looking at the functions performed, assets used and risks assumed through the Indian establishment. The OECD 2010 report on the attribution of profits to permanent establishments sets out the commonly used approach, but the specific treaty text and Indian law govern.

8. Can a foreign head office charge a mark-up to its Indian project office?

Not as a matter of course. A head office and its permanent establishment are parts of the same enterprise, so an internal charge does not have the same tax effect as a payment between separate companies. Actual costs incurred for the project can be relevant if proved; a mark-up needs a separate basis under the specific treaty and Indian law.

Related reading from SBC

• No FTS Article in the India-UAE Treaty: How Offshore Engineering Income Is Taxed in India
• Digital Platforms and India Tax Once the Equalisation Levy Is Gone: SEP, Treaty and PE Compared
• Seconding Staff Between India and the UK: Employer, Permanent Establishment and Payroll Tax Are Different Questions
• More articles on the SBC blog

Disclaimer

This article is general professional commentary and not advice on any particular facts. Treaty eligibility, applicable law, procedural requirements and statutory amendments must be verified for the relevant tax period. It is not a description of an identifiable client engagement.

Faceless Reassessment After Section 147A: Which Officer Can Issue the Section 148 Notice?

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Litigation / Controversy

Introduction

For some years, the first question in reassessment was sometimes not whether income had escaped assessment but which officer had issued the notice. The faceless reassessment scheme changed how jurisdiction was allocated, while notices continued to be issued through jurisdictional channels. That mismatch produced litigation across several High Courts. A procedural point which looked straightforward in one jurisdiction could produce a different result in another.

The Finance Act 2026 materially altered that landscape by inserting section 147A in the Income-tax Act, 1961. It would be unsafe either to repeat earlier High Court conclusions without addressing that amendment or to assume that retrospective legislation has disposed of every pending challenge. A challenge now has to identify the exact notice, the statutory year, the purported defect and the effect of the amended text on that defect.

This article deals with the jurisdictional examination of a reassessment file. It does not substitute a generic claim that a notice was issued by the wrong officer for a review of jurisdiction, limitation, approval, service and the material said to disclose escapement. The separate problem of automated return processing is considered in our note on the limits of section 143(1).

1. Where the jurisdiction question came from

Section 151A of the 1961 Act enabled a scheme for reassessment and the issuance of reassessment notices that was intended to reduce direct interaction and allocate functions electronically. The relevant scheme and notifications matter because they give practical effect to the enabling section. Courts were asked whether a notice issued by a jurisdictional Assessing Officer, instead of being assigned through the faceless mechanism, was competent.

That is a question of legal power, not merely the choice of portal used to upload a document. A taxpayer challenging jurisdiction needs the notice itself, the digital signature, the authority named on the document, the applicable notification, and the relevant dates. A label on an e-proceedings screen is not necessarily conclusive evidence of the statutory identity of the issuing authority.

2. Why the Finance Act 2026 cannot be overlooked

The Finance Act 2026 inserted section 147A into the 1961 Act, with retrospective effect from 1 April 2021. As reported, it provides that for sections 148 and 148A the “Assessing Officer” means an officer other than the National Faceless Assessment Centre or an assessment unit under section 144B(3), so that a jurisdictional officer can issue those notices. Its retrospectivity and interaction with the existing schemes are central to the dispute. The precise enacted language and commencement must be quoted rather than paraphrased from a departmental summary.

In Jyoti Sareen v Union of India, decided by a Division Bench of the Punjab and Haryana High Court on 10 September 2026 (CWP No. 15791 of 2024 and connected petitions), the Court held the retrospective operation of section 147A unconstitutional and set aside the reassessment notices before it. On 18 September 2026 the Supreme Court stayed that judgment in Union of India v. Bharat Industrial Enterprises Pvt. Ltd., SLP (C) No. 33196 of 2026, and the final hearing is reported as listed for 3 December 2026. The Supreme Court has therefore not yet decided the validity of section 147A, and the High Court judgment should not be treated as settled law. Read the judgment and procedural history before relying on it in a particular jurisdiction.

3. Start with the date and governing statute

Two reassessments with similar facts can be governed by different legal frameworks. Record the assessment year or tax year, the original return date, the alleged year of escapement, the first information or risk-management input, any section 148A show-cause notice, the order on objections and the section 148 notice. Identify whether an earlier notice was subject to transitional treatment.

For years governed by the 1961 Act, analyse the repeal-and-savings provisions alongside later validating legislation. For periods governed by the Income-tax Act, 2025, use the corresponding operative provisions as enacted. A submission which cites only the 2025 Act for a historical assessment year risks answering the wrong statutory question.

4. Jurisdiction, sanction and limitation are distinct objections

The officer permitted to issue a notice is one issue. Whether the designated authority validly sanctioned that notice is another. Whether the notice is within the outer time limit is a third. A jurisdictional submission should put them in separate grounds because an amendment addressing one defect does not necessarily answer the others.

Limitation also demands a quantified enquiry into the kind and value of income alleged to have escaped assessment where the law uses a higher threshold for an extended period. The file should identify the evidence on which the authority relies. A large figure mentioned in a report is not automatically a finding that the statutory threshold has been satisfied for the year under examination.

5. Distinguish a procedural irregularity from want of authority

Courts distinguish between defects that can be cured and acts performed without statutory competence. It is tempting to label every processing failure a jurisdictional nullity, but the characterisation must follow the section and the relevant amendment. DIN defects, deficient internal approvals, service disputes and the competence of the notice issuer should not be presented as interchangeable.

When legislation purports to validate earlier actions, counsel should test exactly what the provision validates, which periods it covers and whether the asserted defect lies within its terms. A broad constitutional challenge can require a different forum and remedy from a conventional statutory objection made before the Assessing Officer or appellate authority.

6. Develop the merits defence while reserving jurisdiction

A serious dispute strategy has two tracks. One preserves threshold objections without acquiescence. The other answers the alleged escapement with records, transaction history and reconciliations. Procedural litigation may fail, become moot after a legislative change or be decided differently in another jurisdiction. The underlying factual file must still be able to withstand scrutiny.

For example, where the information relates to an alleged unexplained investment, preserve the original bank entries, sources of funds, counterparties and property records. Where the allegation relates to income from a non-resident payer, obtain the agreement, invoices, ledger, withholding certificate and treaty documentation. The legal objection is stronger, not weaker, when the facts have been investigated.

7. Evidence index for a jurisdictional challenge

A useful index has the following order: original assessment and return; all reopening communications; electronically delivered notices and acknowledgements; digital signatures and officer identification; sanction records to the extent supplied; information relied on for reopening; response to section 148A; final notice; statutory limitation working; relevant notifications and judgments.

Put each procedural objection in a matrix showing the provision, factual event, record proving it, Revenue response anticipated and requested relief. Avoid sweeping statements that the entire reassessment is invalid without identifying the event that attracts the objection.

8. Relief, strategy and preservation of rights

The right next step depends on the stage. Before an order, record objections in time and answer fact-specific enquiries. On receipt of an adverse order, preserve appeal periods and consider demand stay. A writ is not a universal substitute for the statutory appeal: its appropriateness depends on the asserted jurisdictional defect and available alternative remedy.

An objection which may ultimately succeed does not by itself suspend collection. Protective applications for stay, penalty abeyance and correction of outstanding demand should be considered separately. The matter is not finished when an appellate order is received: implementation and consequential refunds or demands must be verified.

Which Act governs the notice: the 1961 Act or the 2025 Act?

The CBDT’s FAQs on interplay and transition state that reassessment for tax years beginning before 1 April 2026 continues under the Income-tax Act, 1961, and that sections 279 to 286 of the Income-tax Act, 2025 apply from tax year 2026-27. The mapping given there is: section 147 to 279, section 148 to 280, section 148A to 281, section 149 to 282, section 150 to 283, section 151 to 284, section 152 to 285 and section 153 to 286. Section 147A, being a 1961 Act provision, is the operative text for notices relating to the earlier years that are the subject of most current disputes.

A file should therefore record both numberings from the outset. A submission for a 2020-21 notice that cites section 280 of the 2025 Act answers the wrong statute, and a submission for tax year 2026-27 that cites section 147A may be answering a provision that does not apply to it.

How the JAO-FAO dispute reached section 147A

The revised reassessment procedure under sections 147, 148 and 148A applied from 1 April 2021. The e-Assessment of Income Escaping Assessment Scheme, 2022 and the faceless jurisdiction scheme were notified on 29 March 2022, and the Supreme Court dealt with the transition for old notices in Union of India v. Ashish Agarwal on 4 May 2022. High Courts then divided: the Bombay High Court in Hexaware Technologies (3 May 2024) held that a jurisdictional officer and a faceless officer cannot hold concurrent jurisdiction, and other High Courts took a similar approach, while the Delhi High Court in T.K.S. Builders (28 October 2024) took the contrary view. The Supreme Court dismissed the Revenue’s special leave petition in Deepanjan Roy on 16 July 2025 without a reasoned order.

The Finance Act 2026 then inserted section 147A, with effect from 1 April 2021 and a non-obstante clause. In April and May 2026 the Supreme Court remitted a batch of pending appeals to the High Courts in the light of the new section, leaving its validity open. The Punjab and Haryana High Court decided the lead petition, Jyoti Sareen v. Union of India, on 10 September 2026 together with a large batch of connected petitions, and held the retrospective operation of section 147A unconstitutional. Reported grounds include separation of powers, because Parliament may cure a defect identified by a court but may not simply nullify the judgment, and the fact that the amendment did not address the randomised, automated allocation that the faceless scheme requires. The Supreme Court stayed that judgment on 18 September 2026.

Practical steps while the stay operates

First, do not treat the High Court judgment as an operative authority in your own proceeding while it is stayed; commentary on the stay suggests it cannot be relied on to support a case, but the position in your High Court should be checked. Second, read the Supreme Court’s stay order itself, because reports describe its effect on pending reassessment proceedings differently, and an officer who proceeds when the proceedings are meant to remain stayed creates a separate ground. Third, keep the objection alive in the reply to the section 148A notice and in any writ petition, and record that the reply is given without prejudice to it.

Fourth, diarise the appeal periods and any demand-stay application independently of the constitutional question, as an objection that may eventually succeed does not suspend recovery. Fifth, keep the merits file current. If the final hearing in December 2026 upholds section 147A, the jurisdictional ground in many pending cases will fall away and the case will be decided on the material said to show escapement, the sanction and the limitation position.

Conclusion

The useful question for a reassessment team in 2026 is no longer simply whether a faceless officer issued the notice. It is whether the complete chain of statutory authority remains valid after the latest amendment and the judicial decisions governing the relevant period.

The strongest file preserves threshold objections precisely and prepares the substantive defence independently. A challenge to procedure should never become a reason for leaving the underlying facts unexplained.

Frequently asked questions

1. Are all notices issued by jurisdictional officers invalid?

No. That proposition ignores section 147A inserted by the Finance Act 2026, the Punjab and Haryana High Court judgment of 10 September 2026 and the Supreme Court stay of 18 September 2026. The validity of a particular notice requires a period-specific review.

2. Does a defect in sanction raise the same issue as faceless jurisdiction?

No. They are legally distinct. The notice file must be tested against each applicable requirement.

3. Can a taxpayer rely only on a High Court decision from another State?

Its persuasive value depends on the legal issue, amendments, and subsequent judicial developments. The jurisdictional High Court and binding Supreme Court law must be checked.

4. Should merits be addressed while the jurisdiction objection remains pending?

Usually the safer approach is to preserve the objection expressly while complying with applicable requirements and substantiating the facts.

5. Is a reassessment dispute closed when the appeal is allowed?

Not necessarily. Giving-effect action, interest, refunds, recovery status and linked penalty proceedings require separate verification.

6. Has the Supreme Court decided whether section 147A is valid?

Not yet. The Punjab and Haryana High Court declared the retrospective operation of section 147A unconstitutional on 10 September 2026, and the Supreme Court stayed that judgment on 18 September 2026. The final hearing is reported as listed for 3 December 2026. Until then the validity of the section remains undecided at the apex level.

7. What is section 147A of the Income-tax Act, 1961?

Section 147A was inserted by the Finance Act 2026 with effect from 1 April 2021. In substance it provides that, for sections 148 and 148A, the Assessing Officer means an officer other than the National Faceless Assessment Centre or an assessment unit under section 144B(3), which allows a jurisdictional officer to issue reassessment notices. Read the enacted text before relying on this summary.

8. Which sections of the Income-tax Act, 2025 replace sections 147, 148 and 148A?

According to the CBDT’s FAQs on interplay and transition, section 147 corresponds to section 279, section 148 to section 280 and section 148A to section 281. These sections apply from tax year 2026-27. Reassessment for earlier tax years continues under the 1961 Act under the repeal and savings provisions of section 536.

9. Can a section 148 notice issued by a jurisdictional officer be challenged today?

Yes, but the challenge must address section 147A and the Supreme Court stay. Preserve the officer-identity objection, and also test the sanction, limitation, service and reasons separately. The right forum, whether a writ petition or the statutory appeal route, depends on the stage and on the law in the relevant High Court.

Related reading from SBC

• Limits of Summary Processing: Why a Section 143(1) Intimation Cannot Decide Treaty or Forex Questions
• No FTS Article in the India-UAE Treaty: How Offshore Engineering Income Is Taxed in India
• Attributing Profits to an Indian Project Office: Which Head Office Expenses Count?
• More articles on the SBC blog

Disclaimer

This article is general professional commentary and not advice on any particular facts. Treaty eligibility, applicable law, procedural requirements and statutory amendments must be verified for the relevant tax period. It is not a description of an identifiable client engagement.

Limits of Summary Processing: Why a Section 143(1) Intimation Cannot Decide Treaty or Forex Questions

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Litigation / Controversy

Introduction

Processing a return is a mechanical step. The return is received, checked against what it contains and against prescribed sources, and an intimation issues. It is not an assessment, no one examines the taxpayer’s position, and nothing in the process is designed to resolve a question on which two views are possible. The statute recognises that by placing the enquiry route in the same provision, immediately after the processing route, for use when a question needs one.

Two fact patterns now generate a large share of these appeals. The first is the denial of a treaty claim during processing, where the return was filed on the footing that business profits are not taxable in India absent a permanent establishment and the system has instead applied the domestic special rate. The second is the taxation of notional foreign exchange gains which the taxpayer disclosed, correctly, as an income computation standard adjustment in the tax audit report. Although the two appear unrelated, they raise the same argument, and it is an argument about jurisdiction rather than about the merits.

This note deals with what follows when an intimation has decided something it had no power to decide. It sets out the categories of adjustment the provision actually permits, the authority confining them, and two fact patterns that now generate a large share of these appeals: treaty claims disallowed during processing, and notional foreign exchange gains taxed because they were disclosed in the tax audit report. It also covers how the submission should be structured and why the rectification route is usually the wrong one. For the underlying treaty question in the first of those fact patterns, see our note on treaties with no fees for technical services article.

1. What the provision actually permits

Section 270(1) of the Income-tax Act, 2025 (section 143(1) of the 1961 Act, summary processing of a return) permits the total income or loss to be computed after specified adjustments. Those adjustments are confined to the six categories listed in section 270(1)(a):

1. any arithmetical error in the return;
2. an incorrect claim, if such incorrect claim is apparent from any information in the return;
3. any such inconsistency in the return with respect to the information in the return of any preceding tax year, as may be prescribed;
4. disallowance of a loss claimed by way of set off, where the return for the year in which the loss arose was furnished beyond the due date under section 263(1);
5. disallowance of expenditure, or increase in income, indicated in the audit report but not taken into account in computing the total income; and
6. disallowance of a deduction under section 144 or under Chapter VIII-C where the return was furnished late.

The third category deserves particular attention, because it is recent. An equivalent power was inserted into the 1961 Act as section 143(1)(a)(iia) by the Finance Act 2025 with effect from 1 April 2025, and the 2025 Act carries it forward, and the scope of what is prescribed under it will determine how far it reaches. It also bears directly on the point made later in this note about taking consistent positions across years, because an inconsistency that would once only have invited a scrutiny notice may now support an adjustment in processing.

It is worth noting what is no longer there. Section 143(1)(a)(vi) of the 1961 Act permitted an adjustment for income appearing in a statement or form but not included in the return, but a proviso bars any such adjustment for returns for assessment year 2018-19 onwards, and the 2025 Act has no equivalent. Neither Act now permits that adjustment, and a submission that proceeds on the footing that it does is answering a case the Revenue is not making.

Two features of that list decide most appeals. First, every category is tied to something ascertainable from the return, the accompanying documents or a prescribed source. None invites an enquiry. Second, an incorrect claim must be apparent from any information in the return. A claim that requires the reader to construe a statute, read a treaty or evaluate a legal doctrine before deciding whether it is correct is not apparent from anything.

The 2025 Act consolidated the processing provision with scrutiny. Sections 143(1), 143(2) and 143(3) of the 1961 Act are now one section: processing and intimation are section 270(1) to (5), and the notice for scrutiny and the assessment order are section 270(8) to (10). The structural point this makes is useful in argument. The statute itself places the enquiry route in the same section, immediately after the processing route. Where a question needs enquiry, the statute supplies the mechanism, and the officer is required to use it.

2. The authority

The leading case remains Khatau Junkar Ltd. v. K.S. Pathania, Bombay High Court, 1992, decided by Sujata Manohar and S.H. Kapadia JJ. The Court held that adjustments in summary processing are confined to arithmetical errors and to deductions prima facie admissible or inadmissible on the face of the return and the accompanying documents alone. Where a disallowance requires enquiry, or turns on a debatable point, the officer must issue a notice for scrutiny; he cannot decide it in processing.

The principle has been applied consistently since. Its practical force is that it converts a substantive dispute into a jurisdictional one. The appellate authority does not have to agree with the taxpayer on the treaty question or the accounting question. It has only to agree that the question was debatable, which is a much lower bar, and that finding disposes of the appeal.

3. First application: treaty claims disallowed in processing

A non resident files a return claiming nil taxability on service income under the business profits article of the applicable treaty, supported by a tax residency certificate, Form 10F and a declaration that it has no permanent establishment in India. The system disregards the treaty position and applies the domestic rate under section 207 of the 2025 Act (section 115A, the special rate on royalty and fees for technical services in the hands of non-residents), generating a demand.

Consider what the system had to decide in order to reach that result. It had to determine that the income was fees for technical services rather than business profits. It had to determine that no permanent establishment existed, or that the existence of one was irrelevant. It had to determine that section 159(4) (section 90(2), treaty relief where more beneficial) did not displace the domestic charge. Each of those is a question of characterisation and treaty construction. The Supreme Court in Union of India v. Azadi Bachao Andolan, (2003) 263 ITR 706 (SC), held that a treaty overrides the Act where it is more beneficial. Whether it does so on particular facts is not a matter capable of determination on the face of a return.

The submission follows from that. The adjustment is not within any category in section 270(1)(a). It is not an arithmetical error. It is not an incorrect claim apparent from the information in the return, because the information in the return supports the claim and the certificate and declaration were filed. Determining otherwise required an exercise in treaty interpretation for which the provision supplies no power.

It is worth adding the alternative. Even if jurisdiction existed, the claim is correct on the merits, and the appellate authority should say so to avoid a second round. A ground that wins only on jurisdiction protects the year but leaves the same adjustment available in the next.

4. Second application: notional foreign exchange gains

A company with foreign currency trade receivables restates them at the closing rate on the balance sheet date, as the accounting standards require, and the restatement produces a gain. The restatement produces an unrealised gain. The company discloses the gain in its tax audit report as an income computation standard adjustment, identifying it as unrealised. The system reads the disclosure as a declaration of income and taxes the full amount.

The adjustment is unsustainable on two grounds, and both should be run.

Jurisdiction. Deciding whether an unrealised restatement is taxable requires reading section 43 of the 2025 Act (section 43AA, the treatment of gain or loss arising from change in foreign exchange rates), the relevant income computation and disclosure standard, and the judicial doctrine on what constitutes real income. That is interpretation, not correction. An audit report disclosure is a reconciliation item. It exists so that the difference between the accounting treatment and the tax treatment is visible. Treating the disclosure as an admission of taxable income inverts its purpose and penalises the transparency the disclosure requirement was designed to produce.

Substance. The real income doctrine is long settled. In CIT v. Shoorji Vallabhdas & Co., (1962) 46 ITR 144 (SC), the Supreme Court held that if income does not result at all, there cannot be a tax. The Court restated the principle in CIT v. Excel Industries Ltd., Civil Appeal No. 125 of 2013, judgment of 8 October 2013, adding that income tax cannot be levied on hypothetical income. In Godhra Electricity Co. Ltd. v. CIT, (1997) 225 ITR 746 (SC), the Court held that although the company followed the mercantile system and had made the book entries, no real income had accrued, and that accrual must be judged by the probability or improbability of realisation in a realistic manner.

One authority needs careful handling. CIT v. Woodward Governor India (P.) Ltd., (2009) 312 ITR 254 (SC), is frequently cited in this context. What it actually decided is that an unrealised exchange loss on restatement of liabilities at the closing rate is allowable expenditure, that expenditure is not confined to cash outgo, and that a consistently followed method of accounting is presumed correct unless the officer gives reasons for rejecting it. It is not a direct holding on the taxability of notional exchange gains. What it does supply is the symmetry principle: among the factors the Court set out is that losses and gains must receive the same treatment. That is a sound argument, but it should be presented as what it is rather than as a holding the case does not contain.

5. The shape of the submission

A submission on either fact pattern should run in this order.

1. State the adjustment made, with the amount, the intimation date and the precise clause of section 270(1)(a) (section 143(1)(a)) under which it purports to fall. Often the intimation does not identify one, which is itself worth saying.
2. Set out what the processing authority had to determine in order to make the adjustment, as a list of discrete questions.
3. Show that each question is debatable, by reference to the statutory provisions that have to be construed and, where available, to decisions going both ways.
4. Conclude on jurisdiction, relying on Khatau Junkar and the structure of section 270.
5. Argue the merits in the alternative, fully, so that a favourable finding disposes of the issue for later years.
6. Annex the documents: the return, the audit report with the relevant clause, the certificate and declaration where treaty benefit is claimed, the reconciliation schedule, the ledger extracts and the accounting entries.

6. A worked example

A manufacturer records unrealised exchange gains on year end restatement of foreign currency trade receivables. The gain is shown in the tax audit report as a disclosure item, described as unrealised. Processing adds the full amount and raises a demand at the applicable rate of tax.

What the file should contain. The tax audit report with the relevant clause; the reconciliation schedule showing how the accounting profit was adjusted to arrive at returned income; the accounting entries recording the restatement; ledger extracts for each foreign currency receivable; the invoice level listing with the transaction rate and the closing rate; and evidence of subsequent realisation, which is often the most persuasive document in the file because it shows what the receivable was actually worth.

What the submission argues. That the adjustment required a determination under section 43 (section 43AA) and the relevant income computation standard, read with the real income doctrine, and that such a determination is outside the processing power. That, in substance, a restatement entry is not a receipt or an accrual, and the charge attaches to income, not to an accounting measurement. That the disclosure was a reconciliation item, not a declaration of income. And that, on the symmetry principle, a position that taxes unrealised gains while disallowing unrealised losses cannot be right.

What usually resolves it. The appellate authority accepting that the question was not capable of summary determination, and directing deletion subject to verification of the disclosure. The verification direction is routine and should not be resisted; it is the mechanism by which the authority disposes of the appeal without conducting an enquiry itself.

7. Where these appeals go wrong

First, the appeal is argued only on the merits. A submission that spends its length on treaty characterisation and never mentions jurisdiction has given away the easier ground.

Second, the jurisdictional ground is stated but not developed. Asserting that the officer exceeded his powers is not an argument. The argument is the list of questions he had to answer, each shown to be debatable.

Third, the rectification route is used first and the limitation period for appeal is allowed to run. A rectification application does not extend the time for appeal, and a debatable question is by definition not a mistake apparent from the record, so the application is likely to fail on the very ground that makes the appeal strong. File the appeal; pursue rectification alongside if it is worth pursuing at all.

Fourth, the documents that make the claim apparent were not filed with the return. Where a certificate or declaration was obtained late, the taxpayer is arguing that the claim was apparent from information that was not before the processing system. That is a harder case, and it should be met directly rather than left for the Revenue to point out.

Fifth, the position across years is inconsistent. Where the same restatement was offered to tax in an earlier year, or the same treaty claim was not made, the inconsistency has to be explained.

Sixth, no stay application is filed. An intimation is a demand, and recovery can follow while the appeal is pending. The jurisdictional ground is usually a strong basis for a stay, and it is wasted if it is not deployed for that purpose.

8. Preventing the adjustment in the first place

Several of these adjustments are avoidable at filing. Where a treaty claim is made, the residency certificate and Form 10F should be in place and filed before the return, and the return should be completed in a way that makes the basis of the claim visible rather than leaving it to inference. Where an income computation standard adjustment is disclosed, the disclosure should state in terms that the amount is unrealised and has not been offered to tax, and the reconciliation should be capable of being read on its own. Where a position is novel, a note in the computation costs nothing and converts a silent claim into a stated one.

None of this guarantees that the system will not make the adjustment. It does mean that when the appeal is filed, the file supports the argument that the claim was apparent on the face of the return, which is precisely the question the appeal turns on.

Practitioner’s perspective

The common thread in these appeals is that an automated process is making determinations the statute reserves for an assessing officer exercising judgment. That is a structural feature of a digitised administration rather than an error in any individual case, and it will continue to produce demands of this kind.

For taxpayers the practical consequence is that the jurisdictional ground should be the first ground in every such appeal, not an afterthought, and the merits should be argued in full alongside it. For non residents in particular, the risk is systemic: a treaty position that is correct, documented and accepted in scrutiny can still generate a demand in processing, and the cost of resolving it falls entirely on the taxpayer.

Two practical habits reduce the exposure materially. Make the basis of every claim visible on the face of the return rather than leaving it to be inferred. And treat the tax audit report as a document that will be read literally by a system that does not distinguish between a disclosure and an admission.

Conclusion

Summary processing cannot adjudicate contested treaty and income-characterisation questions. A procedural ground should be supported by the statutory adjustment limits and a complete alternative argument on the merits.

Frequently asked questions

1. What adjustments are permitted in summary processing?

Section 270(1)(a) of the 2025 Act (section 143(1)(a)) permits six: an arithmetical error; an incorrect claim apparent from any information in the return; a prescribed inconsistency with the information in a preceding year’s return; disallowance of a set off loss where the loss year return was late; an audit report disallowance or increase in income not taken into account; and disallowance of a deduction under section 144 or Chapter VIII-C where the return was late. The third was inserted in the 1961 Act by the Finance Act 2025 (from 1 April 2025) and is carried into the 2025 Act. The power to adjust for income appearing in a statement but not in the return is barred by a proviso for returns for assessment year 2018-19 onwards and has no successor in the 2025 Act.

2. Can a treaty claim be denied during processing?

Denying a treaty claim requires characterising the income, considering whether a permanent establishment exists and applying the treaty override. Those are interpretational questions and are outside the processing power. The denial should be challenged first on jurisdiction and then on the merits.

3. What does it mean for a question to be debatable?

That two views are reasonably possible on it. The taxpayer does not have to show that its view is correct in order to succeed on the jurisdictional ground, only that the question was not capable of determination on the face of the return.

4. Are foreign exchange gains disclosed in the tax audit report automatically taxable?

No. The disclosure is a reconciliation item identifying a difference between the accounting treatment and the tax treatment. It is not a declaration that the amount is taxable income, and treating it as one reverses the purpose of the disclosure requirement.

5. Does Woodward Governor decide that unrealised exchange gains are not taxable?

No. It decided that an unrealised exchange loss on restatement of liabilities is allowable expenditure, and that a consistently followed method of accounting is presumed correct. It supports a symmetry argument, because the Court said losses and gains must receive the same treatment, but it is not a holding on the taxability of unrealised gains and should not be cited as one.

6. Should a rectification application be filed instead of an appeal?

Rectification is confined to a mistake apparent from the record. A debatable question is by definition not such a mistake, so the application is likely to fail on the same ground that makes the appeal strong. Filing rectification also does not extend the limitation period for appeal. File the appeal.

7. Can recovery be enforced while the appeal is pending?

An intimation that raises a demand is enforceable, so a stay application should be filed. A well developed jurisdictional ground is usually a strong basis for a stay, because it goes to the validity of the demand rather than only to its quantum.

8. What if the residency certificate was obtained after the return was filed?

The argument becomes harder, because the taxpayer is contending that the claim was apparent from information that was not before the processing system. The substantive entitlement survives, but the point should be addressed directly in the submission rather than left to be raised against you.

9. Does winning on jurisdiction settle the issue for later years?

No. A finding that the question could not be decided in processing says nothing about the correct answer. That is why the merits should be argued in full in the alternative, so that a favourable finding on substance is available for subsequent years.

10. Which Act applies to an intimation for an earlier year?

Section 536 of the 2025 Act, the repeal and savings provision, preserves the 1961 Act for proceedings pending on 31 March 2026 and for proceedings relating to tax years beginning before 1 April 2026. Most current appeals are therefore argued on 1961 Act numbering, with the 2025 provisions relevant going forward.

11. Is there a time limit for issuing a section 143(1) intimation?

Yes. Under the 1961 Act the intimation must be issued within nine months from the end of the financial year in which the return is furnished (as amended by the Finance Act 2021). Section 270(4) of the Income-tax Act, 2025 carries the same nine-month outer limit. An intimation issued beyond that period is open to challenge.

12. Does the Income-tax Act, 2025 allow an adjustment for inconsistency with the previous year’s return?

Yes. Section 270(1)(a)(iii) allows an adjustment for an inconsistency, as prescribed, with the information in the return of a preceding tax year. The 1961 Act received the same power as section 143(1)(a)(iia) through the Finance Act 2025, effective 1 April 2025. Positions taken across years should therefore be consistent.

Related reading from SBC

• No FTS Article in the India-UAE Treaty: How Offshore Engineering Income Is Taxed in India
• Faceless Reassessment After Section 147A: Which Officer Can Issue the Section 148 Notice?
• Attributing Profits to an Indian Project Office: Which Head Office Expenses Count?
• More articles on the SBC blog

Disclaimer

This article is for general information and does not constitute professional advice. It reflects the law as understood at the date of publication and does not take account of later developments. The Income-tax Act, 2025 came into force on 1 April 2026, and section 536 preserves the Income-tax Act, 1961 for proceedings pending on 31 March 2026 and for earlier tax years, so the applicable numbering depends on the year and the stage of the proceeding. Outcomes turn on the specific facts and documentation. No reader should act on the basis of this article without taking advice on their own circumstances.

No FTS Article in the India-UAE Treaty: How Offshore Engineering Income Is Taxed in India

By CA Mithilesh Reddy, Founder and CEO, and CA Rajesh Vaishnav, Partner, Steadfast Business Consulting LLP | Direct Tax Litigation / Controversy

Introduction

A double taxation avoidance agreement does not impose tax. It allocates the right to tax between two States, and it limits what the source State may do. The domestic Act creates the charge; the treaty decides how far that charge may reach a resident of the other State. Where the two differ, section 159(4) of the Income-tax Act, 2025 (section 90(2) of the 1961 Act, under which the treaty applies to the extent it is more beneficial) gives the treaty priority. That ordering is the whole of the argument in this area, and most disputes turn on a point of fact within it rather than on the ordering itself.

Where a treaty omits an article dealing with fees for technical services, the argument that follows is simple to state and routinely lost on execution. Assessing officers do not generally dispute the absence of the article. They dispute the facts that make Article 7 the operative provision: whether a permanent establishment exists, where the work was actually performed, and whether the taxpayer’s documentary position was in place at the time the payments were made rather than assembled afterwards. This note sets out the statutory framework, the authorities that govern each limb, and the evidence an assessment file needs to carry.

This note deals with the source State side of that question for technical and engineering services: how income is characterised when the treaty contains no article dealing with fees for technical services, what the permanent establishment analysis requires, where the work has to have been performed, and what the assessment file must contain before any of it can be argued. It also covers the procedural ground that frequently decides these assessments before the merits are reached. A companion note on the limits of summary processing under section 143(1) deals with the related problem of treaty claims disallowed without an assessment at all.

1. Why the treaty and the Act answer different questions

Section 9(7) of the Income-tax Act, 2025 (section 9(1)(vii) of the 1961 Act, which deems fees for technical services to accrue in India) brings within the Indian charge consideration for managerial, technical or consultancy services paid by a resident, subject to the exclusions in the provision. Section 207 (section 115A, the special rate for royalty and fees for technical services received by non-residents) then fixes the rate. Section 393(2), Table serial number 17 (section 195, withholding on sums payable to non-residents) imposes the deduction obligation on the payer.

None of this decides whether India may actually tax the income. Section 159(4) of the 2025 Act (section 90(2), under which the treaty applies to the extent it is more beneficial to the assessee) makes the treaty the governing instrument wherever it is more favourable. The Supreme Court settled the position in Union of India v. Azadi Bachao Andolan, (2003) 263 ITR 706 (SC), holding that a treaty operates even where it is inconsistent with the Act and overrides the Act in computing total income and tax liability.

One provision deserves attention. Section 159(6) of the 2025 Act, which carries forward section 90(2A) of the 1961 Act, provides that the general anti avoidance rules in Chapter XI apply even where they are not more beneficial to the assessee. The treaty override in section 159(4) is therefore expressly made subject to the anti avoidance chapter on the face of the statute. For a genuine operating business with real activity in the treaty state this changes nothing. For a structure whose only substance is the treaty claim, it narrows the room considerably.

2. Which treaties have no fees for technical services article

The India UAE treaty, signed on 29 April 1992 and in force from 22 September 1993, runs from Article 1 (Personal Scope) to Article 32 (Termination). Its articles include Article 5 (Permanent Establishment), Article 7 (Business Profits), Article 12 (Royalties), Article 14 (Independent Personal Services) and Article 22 (Other Income). There is no article headed Fees for Technical Services or anything equivalent. Article 12 covers royalties, including payments for information concerning industrial, commercial or scientific experience, but it is not a technical services provision.

India has several treaties in the same position, the UAE and Thailand among them. The position varies and has in some cases been altered by protocol, so the absence of a technical services article should be confirmed for the particular treaty rather than assumed from a list. The point to take from this is procedural rather than substantive: before building an argument on the absence of an article, read the operative text of the treaty as currently amended, including any protocol and the synthesised text reflecting the Multilateral Instrument. A treaty article that was absent in the original text may have been inserted later.

3. Where the income goes when the treaty is silent

If the treaty has no technical services article, the income must be allocated under some other article. Two candidates present themselves: Article 7, which deals with business profits, and the residuary Other Income article, which in the India UAE treaty is Article 22.

On the India UAE treaty the residuary route is closed by the text itself. Article 22(1) allocates income not expressly dealt with in the other articles exclusively to the State of residence, wherever it arises, subject only to the carve out in Article 22(2) for income effectively connected with a permanent establishment or fixed base. A Revenue argument that technical service income is taxable in India as other income therefore fails on the face of the article, without needing to go further.

That is not true of every treaty, because residuary articles are not uniformly drafted. Where the residuary article permits source State taxation, the question becomes live, and the answer is supplied by the Madras High Court in Bangkok Glass Industry Co. Ltd. v. ACIT, (2013) 34 taxmann.com 77 (Mad), dealing with the India Thailand treaty, which likewise contains no technical services article. The Court held that such income cannot be taxed under the residuary article; it falls to be considered under Article 7 and is taxable in India only where the non resident has a permanent establishment to which it is attributable.

The logic is one of structure. Technical service income earned by an enterprise in the course of its trade is, as a matter of ordinary characterisation, business profit. The residuary article is designed for items of income that fit nowhere else, not for the central receipts of an operating business. An authority that treats a residuary article as a general source taxing power would make the business profits article largely redundant in any treaty that has both.

4. The permanent establishment question

Once Article 7 applies, everything turns on the permanent establishment. For a UAE engineering company contracting with an Indian customer, the exposures are the usual four.

• Fixed place. An office, workshop, yard or other fixed place of business in India at the disposal of the enterprise.
• Project or installation. A building site, construction, assembly or installation project in India, where the treaty’s duration threshold is exceeded. Under the India UAE treaty, Article 5(2)(h) and Article 5(2)(i) each set a threshold of more than nine months, and in the case of Article 5(2)(i) that period is measured in the aggregate within any twelve month period for the same or a connected project. The threshold is therefore expressed in months and it aggregates, which makes the computation of duration, and the question of which activities form part of a connected project, the usual battleground.
• Service presence. Where the treaty contains a service permanent establishment clause, the furnishing of services through employees or other personnel present in India beyond the stated period.
• Dependent agent. A person in India habitually concluding contracts, or habitually playing the principal role leading to their conclusion, on behalf of the enterprise.

Most offshore engineering disputes are decided on the second and third of these. Where the contract is for fabrication, conversion or manufacture performed wholly at the contractor’s own yard outside India, with Indian presence limited to short supervisory or commissioning visits, the question is whether the visits cross the threshold and whether they are properly characterised as a separate installation project or as incidental to an offshore supply.

5. Where the work was done: offshore supply and offshore services

The Supreme Court’s decision in Ishikawajma-Harima Heavy Industries Ltd. v. DIT, (2007) 288 ITR 408 (SC), remains the starting point. The Court held that only income attributable to operations carried out in India is taxable; where property in the goods and payment both passed outside India, the offshore supply was not taxable in India, and the place at which the contract was signed was immaterial.

The decision has a second limb that must be handled carefully. On offshore services the Court held that, for section 9(1)(vii) to apply, the services had to be both rendered in India and utilised in India. That limb was legislatively reversed. The requirement of rendition in India was removed by the Explanation below section 9(2) of the 1961 Act, inserted by the Finance Act 2010 with retrospective effect from 1 June 1976, and the position is now carried in section 9(11)(b) of the 2025 Act, which deems income of the kinds described in sections 9(5), 9(6) and 9(7) to accrue in India whether or not the non resident has rendered services in India. Citing Ishikawajma for the territorial nexus proposition without flagging the reversal is a common and serious error in written submissions. The surviving and still powerful part of the decision is the offshore supply analysis and the principle that only operations carried out in India can generate Indian source profit under the treaty.

The Court followed the same approach in CIT v. Hyundai Heavy Industries Co. Ltd., (2007) 291 ITR 482 (SC). The installation permanent establishment came into existence only after the sale was concluded, so profits of the Korean design and fabrication operations had no economic nexus with the Indian permanent establishment and were not taxable in India. Only the profit attributable to Indian installation and commissioning was brought to tax.

More recently the Delhi Bench of the Tribunal considered related questions in Andritz AG v. Dy. CIT (IT), (2024) 165 taxmann.com 76 / (2025) 122 ITR 479 (Delhi Trib), a case under the India Austria treaty. The Bench held that consideration for drawings and designs was inseparable from the offshore supply and could not be carved out and taxed separately as fees for technical services, and that cost reimbursements without mark up were not fees for technical services. The decision should be cited with care rather than as a clean win on every point. The non taxability of the offshore supply was a premise already accepted in the proceedings rather than a holding in that order; the net basis treatment of onshore supervisory receipts under Article 7 was remanded rather than decided; and the Bench also observed that supervisory fees fall within the technical services article irrespective of a permanent establishment, which cuts against part of the proposition the decision is often cited for.

6. The contrary line, and why a submission should address it

A submission that presents this area as settled invites the Revenue to produce the other half of the jurisprudence. In Voith Paper GmbH v. Dy. DIT, (2020) 80 ITR (Trib) 589 (Delhi Trib), the Tribunal held that the contracts could not be viewed independently of one another, that the buyer’s dominant purpose throughout was the installation of a plant under the supplier’s supervision, that a business connection was therefore established, and that a service permanent establishment existed with thirty five per cent of profits attributed to it. Interest was held exigible because the assessee had conceded the permanent establishment before the assessing officer.

Two lessons follow. First, the composite nature of a contract cuts both ways. A taxpayer who argues that a contract is commercially indivisible in order to defend an expense deduction cannot simultaneously argue that it is severable in order to keep an offshore supply outside the Indian charge. The positions must be reconciled across the whole file. Second, a concession on the existence of a permanent establishment made at assessment stage is very difficult to withdraw later, and carries consequences for advance tax and interest that are not always considered when the concession is made.

7. What the treaty position requires by way of documentation

Treaty entitlement is substantive, but it is proved by documents, and the documents must exist contemporaneously. For a UAE resident claiming Article 7 treatment the file should carry the following.

• A valid tax residency certificate issued by the UAE authority for each relevant year, covering the period in which the income arose.
• Form 10F, filed electronically, with the particulars matching the residency certificate.
• A no permanent establishment declaration from the non resident, specific to the contract and the year rather than generic.
• Evidence of where the work was performed: yard records, fabrication logs, production records, shipping and delivery documents, and the contractual terms governing passage of title and risk.
• A day count record for every employee or representative who travelled to India in connection with the contract, with dates, purpose and location. This is the single document most often missing and most often decisive.
• The contract itself, with the scope of work, the division of responsibilities, the pricing of each component and the delivery terms.

On the timing of Form 10F, a procedural lapse should not be allowed to defeat a substantive entitlement. Where residency and the absence of a permanent establishment are established on the evidence, the position is that treaty benefit flows from the treaty and the statute, not from the form. That argument is considerably stronger where the certificate itself was in existence at the relevant time and only the filing was late.

8. The procedural limb: finalising an assessment before the Panel has spoken

For an eligible assessee, which under section 275 of the 2025 Act means a person in whose case a variation arises from an order of the transfer pricing officer, a non resident who is not a company, or a foreign company, section 275 of the 2025 Act (section 144C, the Dispute Resolution Panel) prescribes a mandatory sequence. The assessing officer must pass a draft order. The assessee then has the option of filing objections before the Panel. The Panel issues directions, which bind the officer, and only then may the final order be passed.

Where an officer passes a final order without first issuing a draft, or passes it before the Panel’s directions are received, the weight of tribunal and High Court authority treats the defect as jurisdictional rather than procedural, so that the order is void rather than merely irregular and the defect is not cured by the eventual issue of directions. The consequence, on that line of authority, is that the entire assessment falls and the substantive merits never arise. The point should be pleaded with the authorities relevant to the particular bench, because the consequence claimed is a strong one and the officer will resist it.

Two practical points. First, the defect must be pleaded as a ground; it is not something a tribunal will usually take up unprompted. Second, pleading it does not require abandoning the substantive grounds, and a well constructed appeal runs both tracks. A taxpayer who wins on the jurisdictional ground alone secures the year but establishes nothing for the years that follow, which is why the substantive treaty argument should be fully developed even where the procedural ground looks strong.

9. A worked example

Consider a hypothetical UAE-resident engineering enterprise providing an offshore work package to an Indian customer. The substantive work and supporting records are overseas, and only limited activities occur in India. The legal enquiry is whether any Indian presence satisfies Article 5 and, if so, which profit can properly be attributed to it. Documentary evidence of where the work was done, visits and contract performance is essential.

The Revenue’s likely position. The payments are consideration for technical services, section 9(7) (section 9(1)(vii)) applies, section 207 (section 115A) fixes the rate, and the Indian payer was obliged to withhold under section 393(2), Table serial number 17 (section 195). The Indian visits are relied on as evidence of a taxable presence, and the single lump sum is relied on to resist any splitting of the consideration.

The taxpayer’s position. The treaty has no technical services article, so the domestic deeming provision cannot be imported into it. Article 22(1) allocates other income exclusively to the State of residence, so the residuary route is closed on the text, and the income is business profits under Article 7. The aggregate presence falls well below the nine month threshold in Article 5(2)(i), and the visits are incidental to delivery of the offshore supply rather than a separate project of their own. Under Ishikawajma and Hyundai Heavy Industries, no profit attributable to the UAE fabrication operations has an economic nexus with India. Under Andritz AG, even if some element of the consideration were attributable to an Indian presence, it would be taxed as business profits on a net basis, not on a gross basis as technical service fees.

What decides it. Not the legal proposition, which is reasonably well settled, but the presence record, the yard documentation, the delivery terms and whether the contract on its face supports the characterisation. Where those four hold, the position is defensible. Where presence is reconstructed after the event from passport stamps and the contract describes the scope in terms of services rather than supply, it is not.

10. Where these files are usually weak

First, the treaty text relied on is not the current text. Protocols and the Multilateral Instrument have altered a number of Indian treaties, and a submission built on an outdated article invites a short answer.

Second, the presence record is incomplete. Visits by subcontractor personnel, by personnel of group companies, and short trips not recorded in the project file are routinely omitted and then produced by the Revenue from immigration data. Because the India UAE threshold aggregates across connected projects, an omission in one project file can affect the position in another.

Third, the contract describes the work in service language throughout. Commercial teams draft for commercial clarity, and a contract that speaks of providing engineering services will be read as what it says, whatever the underlying activity.

Fourth, the positions taken across years are inconsistent. A permanent establishment conceded or returned in one year is very difficult to disclaim in the next on materially identical facts.

Fifth, the residuary article argument is not anticipated. On the India UAE treaty it is answered by the text of Article 22(1), but under a treaty whose residuary article is drafted to permit source State taxation it is a live point, and a submission built only on the absence of a technical services article has nothing further to say.

Sixth, the withholding and the recipient’s position are argued separately. The payer’s obligation under section 393(2) (section 195) and the recipient’s liability are related but distinct, and a submission that conflates them loses the benefit of the arguments available on each.

11. A pre assessment review

1. Obtain the current text of the treaty, including protocols and the synthesised text, and confirm the article structure.
2. Identify every article under which the Revenue could seek to tax the receipt, and prepare a position on each, not only on the technical services characterisation.
3. Reconstruct the presence record from primary sources, for every person connected with the contract, aggregate it across connected projects within the relevant twelve month period, and reconcile it to immigration records before the Revenue does.
4. Read the contract as an assessing officer would, and identify the clauses that work against the characterisation being advanced.
5. Confirm that the residency certificate, Form 10F and the no permanent establishment declaration exist for every relevant year and are internally consistent.
6. Map where title and risk passed, and assemble the shipping and delivery evidence that proves it.
7. Check the positions taken in earlier and later years, including in the payer’s withholding filings, and resolve any inconsistency before it is put to you.
8. If the matter is at draft order stage, verify that the Panel sequence has been followed and record the dates.
9. Where any part of the receipt is attributable to an Indian presence, prepare the net basis computation under Article 7 as an alternative position.
10. Consider whether the anti avoidance chapter could be invoked, and document the commercial substance of the arrangement in the treaty state.

Practitioner’s perspective

The absence of a technical services article is a strong position, but it is a position about the law, and assessments are decided on facts. In our experience the cases that fail do not fail because the tribunal disagrees that Article 7 governs. They fail because the day count was incomplete, the contract was drafted in the language of services, or the taxpayer had taken a different position in an earlier year and could not explain the change.

Two developments should be factored into any current review. The Income-tax Act, 2025 came into force on 1 April 2026, and while section 536 preserves the 1961 Act for pending proceedings and for earlier tax years, submissions filed from here on will increasingly need to carry both sets of references. Separately, section 159(6), which carries forward section 90(2A) of the 1961 Act, makes the treaty override expressly subject to the anti avoidance chapter, which raises the premium on demonstrable substance in the treaty state.

For UAE engineering and technical service providers with Indian contracts, the practical conclusion is that the defence is built at contracting stage and maintained through the project, not assembled when the notice arrives. The documents that decide these cases are generated in the ordinary course of the project, and they are only available later if someone decided to keep them.

Frequently asked questions

1. Does the India UAE treaty contain a fees for technical services article?

No. The agreement runs from Article 1 to Article 32 and contains no article headed Fees for Technical Services or any equivalent. Article 12 deals with royalties, which includes payments for information concerning industrial, commercial or scientific experience, but it is not a technical services provision.

2. If the treaty has no technical services article, can India still tax the income under its domestic law?

Only to the extent the treaty permits. Section 9(7) of the 2025 Act (section 9(1)(vii)) creates the domestic charge, but section 159(4) (section 90(2)) applies the treaty where it is more beneficial. The treaty allocates the income to Article 7, which confines Indian taxing rights to profits attributable to a permanent establishment.

3. Why does it matter whether the income falls under Article 7 or the residuary article?

Article 7 requires a permanent establishment before India may tax, while a residuary article may, depending on its drafting, permit source State taxation without one. Under the India UAE treaty the point does not arise, because Article 22(1) allocates other income exclusively to the State of residence. For treaties whose residuary article is drafted differently, Bangkok Glass holds that technical service income falls under Article 7 rather than the residuary article.

4. Is Ishikawajma-Harima still good law?

In part. The offshore supply analysis and the territorial nexus principle survive. The holding that services had to be both rendered in India and utilised in India was reversed by the Explanation below section 9(2) of the 1961 Act, inserted by the Finance Act 2010 with retrospective effect from 1 June 1976, and now carried in section 9(11)(b) of the 2025 Act. A submission that cites the decision without noting the reversal is vulnerable.

5. What presence in India creates a permanent establishment under the India UAE treaty?

Article 5(2)(h) and Article 5(2)(i) each set a threshold of more than nine months, and under Article 5(2)(i) that period is measured in the aggregate within any twelve month period for the same or a connected project. The threshold is in months, not days, and it aggregates across connected projects.

6. What happens if Form 10F was filed late?

A late filing does not by itself extinguish treaty entitlement where residency and the absence of a permanent establishment are established on the evidence. The position is stronger where the residency certificate was in existence during the relevant period and only the filing of the form was delayed.

7. Can the Indian payer be treated as in default if the recipient is ultimately held not taxable?

The payer’s obligation under section 393(2), Table serial number 17 (section 195) and the recipient’s substantive liability are distinct questions. Where the income is not chargeable in India, the foundation for the deduction obligation falls away, but the payer should be in a position to show the basis on which it formed that view at the time of payment.

8. What is the effect of an assessment order passed before the Dispute Resolution Panel issues directions?

Section 275 of the 2025 Act (section 144C) prescribes a mandatory sequence of draft order, objections and binding directions. An order passed outside that sequence is void rather than irregular, and the defect is not curable. The ground must be specifically pleaded.

9. If part of the income is taxable in India, how is it computed?

Under Article 7 the attributable profit is computed on a net basis, with deduction of expenses incurred for the purposes of the permanent establishment. This is materially different from gross basis taxation at the special rate under section 207 (section 115A), and the distinction was applied in Andritz AG.

10. Does the general anti avoidance rule affect a treaty claim?

Section 159(6) of the 2025 Act provides that Chapter XI applies even where it is not more beneficial to the assessee, so the treaty override in section 159(4) is expressly subject to the anti avoidance chapter. A genuine operating business with real activity in the treaty state is unaffected; a structure whose only function is to support the treaty claim is exposed.

11. Is Form 10F still used after the Income-tax Act, 2025 came into force?

For tax years beginning on or after 1 April 2026, the Income Tax Department lists Form 41 under the Income-tax Rules, 2026 (self-declaration under section 159(8)) as the counterpart of Form 10F under the 1962 Rules. Form 10F continues to be relevant for earlier years. Both are filed electronically with the tax residency certificate.

12. Can fees for technical services under the India-Thailand treaty be taxed as other income?

Tribunal decisions say not where the income is business profits. The Madras High Court in Bangkok Glass Industry and the Delhi Tribunal in Denso (2024 and 2026 orders) held that a residuary article cannot be used for such receipts if the services form part of the recipient’s business and there is no permanent establishment. Each case turns on its facts.

Related reading from SBC

• Limits of Summary Processing: Why a Section 143(1) Intimation Cannot Decide Treaty or Forex Questions
• Attributing Profits to an Indian Project Office: Which Head Office Expenses Count?
• Shared Services from Singapore: Applying the ‘Make Available’ Test to Indian Withholding
• More articles on the SBC blog

Disclaimer

This article is for general information and does not constitute professional advice. It reflects the law as understood at the date of publication and does not take account of developments after that date. The Income-tax Act, 2025 came into force on 1 April 2026 and section 536 preserves the Income-tax Act, 1961 for proceedings pending on 31 March 2026 and for proceedings relating to earlier tax years, so the applicable numbering depends on the year and the stage of the proceeding. Treaty positions turn on the specific facts, the contractual documentation and the current text of the applicable agreement. No reader should act on the basis of this article without taking advice on their own circumstances.