Zambia Revenue Authority v Nestlé Zambia Limited (APPEAL NO. 03/2021) [2025] ZMSC 20 (20 August 2025) | Transfer pricing | Esheria

Zambia Revenue Authority v Nestlé Zambia Limited (APPEAL NO. 03/2021) [2025] ZMSC 20 (20 August 2025)

Full Case Text

IN THE SUPREME COURT OF ZAMBIA HOLDEN AT LUSAKA (Civil Jurisdiction) APPEAL NO. 03/2021 BETWEEN: ZAMBIA REVENUE AUTHORITY APPELLANT AND NESTLE ZAMBIA LIMITED RESPONDENT CORAM: Kajimanga, Kabuka, Chisanga, JJS, On 13th July, 2021 and 20th August, 2025. FOR THE APPELLANT: Mr. F. Chibwe, Legal Counsel, Zambia Revenue Authority FOR THE RESPONDENT: Mr. S. Chisenga, Corpus Legal Practitioners JUDGMENT Kabuka, JS, delivered the Judgment of the Court. Cases referred to: 1. Wilson Masauso Zulu v Avondale Housing Project Limited ( 1982) ZR 172 (SC) 2. Savenda Management Services v Stanbic Bank Zambia Limited, Selected Judgment No.10 of 2018 3. W. T. Ramsay v Inland Revenue Commissioners [1982] AC 300 4. Inland Revenue Commissioners v Duke of Westminster [1936] AC 1 5. Ensign Tankers (Leasing) Ltd v Stokes (HM Inspector of Taxes) [1992] BTC 110 6. icholson v Morris (H. M. Inspector of Taxes) (1) (1973-1978) 51 TC 95 7. National Justice Compania Naveria SA v Prudential Assurance Co. Ltd (The Ikarian Reefer) QB [1993] 2 Lloyd's Rep 68 8. Brady (Inspector of Taxes) v Group Lotus Car Cos Pie & Another 1987 (2) All ER 674 9. Deputy Commissioner of Income Tax, Circle -11 (1) v M/s. Adcock Ingram Limited I. T. (TP). A No. 1039 & 1078/Bang/2015 Legislation and Other Works referred to: 1. Income Tax Act, Cap. 323 SS. 97A, 29A (1), 106, 2. Income Tax Regulations of 2000 S.3(2) 3. Income Tax (Transfer Pricing) Regulations No.24 of 2018 4. British Acts Extension Act, Cap 10 5. English Law (Extent of Applications) Cap 11. of the Laws of Zambia. 6. Arnold, B. J. et al, International Tax Primer, 2 nd Edition 7. United Nations Manual on Transfer Pricing for Developing Countries 8. 2010 OECD Guidelines Introduction 1. When we heard this appeal, we sat with our brother, Kajimanga JS, who has since retired. This judgment is therefore, by the majority. We regret the delay in delivery which was due to circumstances beyond our control. J2 Appellant's appeal 2. The appellant, Zambia Revenue Authority (ZRA), appeals against a decision of the Tax Appeals Tribunal ("the Tribunal"). The Tribunal found that ZRA had justifiable grounds for initiating a transfer pricing audit on the respondent, Nestle Zambia Limited (NZL). The Tribunal however, went on to determine that the assessment was wrongly arrived as ZRA used: (i) inappropriate transfer pricing methods; and (ii) incomparable jurisdictions. 3 . This appeal by ZRA is directed against that determination. Respondent's Cross- appeal 4. The respondent, NZL, has equally cross- appealed against the Tribunal's decision on two grounds. The first ground faults the Tribunal for having found that NZL was a low risk distributor, contending this is contrary to the weight of the evidence on record. The contention in the second ground is that the Tribunal exceeded its jurisdiction when it found that there were transfer pricing issues requiring scrutiny and ordered ZRA to conduct a reassessment. J3 5. For purposes of convenience only, we propose to first deal with the main appeal and thereafter, proceed to consider the cross appeal. We will also for convenience continue referring to the appellant and the respondent respectively, as ZRA and NZL. Background 6. NZL was incorporated sometime in 2010. It was specially established to be a distribution company for Nestle products . The products were intended to be purchased from other related companies around the globe. NZL operations were also financed through loans obtained from the said related entities/parties. 7. It is worthy of note, that at the material time that NZL was incorporated, there was already in existence an established presence of Nestle products on the Zambian market from other independent distributors. The Nestle products were therefore , well known to the general public . 8. After NZL had been trading on the market for about five years, ZRA decided to undertake a transfer pricing audit on the said company. The ZRA decision to conduct an audit was prompted J4 by continuous nil profit that NZL was making and submitting on its tax returns. 9. Accordingly, on 20 th April, 2016 ZRA proceeded to issue an audit notice to NZL. By that notice NZL was informed of the audit intended to be undertaken by ZRA for the period 2010 to 2014. To that end, NZL was requested to make available to ZRA, specific detailed information of its accounts, as well as master and local files, to assist with the said audit. 10. The audit conducted on NZL revealed that NZL was being financed by its shareholders through debt, rather than equity. NZL was also found to be involved in various transactions with associated or related entities or parties (entities) in its general administration, service support, processing of management fees, distribution licence agreements and supplying of Nestle products. 11. It was also established that NZL had a huge volume of transactions with the said related entities. Contrary to circumstances suggesting good business, NZL was found to be making continuous losses from inception, for the period 2010 to JS 2014. The further finding over the same period of time, was that of continuous net operating margins. 12. Apparently, NZL in defence of its position that it was operating unprofitably, contended that it was facing competition from other wholesalers and retailers on the Zambian market. According to NZL, these traders were importing Nestle products directly from other countries. NZL profitability and growth were, as a result, negatively impacted on account of accumulated overhead costs. 13 . The costs were said to be related to the running of its warehouse, royalty payments for use of intellectual property and management fees. NZL also claimed that it was importing 90% of Nestle products from South Africa and the remaining 10% were sourced from Zimbabwe, Brazil and Switzerland. The products were then marketed, sold and promoted, in a bid to advance the brand. 14. Further investigation and interviews held by ZRA in relation to the audit revealed that, according to NZL, it was apparently, experiencing losses due to 'the vagaries and challenges that ravage corporate entities at their infancy stage due to operational costs. ' ZRA found this explanation to be at variance with the J6 initial reasons given by NZL, alluded to earlier in paragraph 12, that the losses were due to competition with other parallel players, who were importing the same products from South Africa into Zambia. 15. On account of the disparity, ZRA rejected the reasons advanced by NZL for the loses. This was on the premise that, an uncontrolled and independent distributor such as NZL would not accept losses for five years running. As Nestle products were already established on the Zambian market and had already gained traction, NZL should have performed better, despite having huge volumes of transactions with related entities. 16. ZRA referred to the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations Regulations, for the position that, when an enterprise is making losses whilst doing business with profitable members of its MNE (Multinational Enterprise) group, this may suggest that the transfer pricing should be examined. The objective for such examination is to rule out the possibility of a scenario suggesting that the loss-making enterprise, is not being dealt with on an arm's length basis by its related entities. J7 17. According to Arnold, B. J. et al, International Tax Primer, 2 nd Edition, a transfer price is defined as a price set by a tax payer when selling to, buying from, or sharing resources with a related entity. The example given is where Company A, manufactures goods in Country A and sells them to its related entity Company B, which is established in Country B. The price at which Company A sells its goods to Company Bis known as a transfer pnce. 18. A transfer price is distinguishable from a market price, in that, a market price is set in the open market, for the transfer and sale of goods or services between unrelated persons or entities. The writers state that, multinational companies often use transfer pricing for sales and transfer of goods and services within their corporate group. 19. The arm's length principle or method simply stated, is that transfer prices should be adjusted to reflect the prices that would have been set between unrelated enterprises, acting independently. 20 . In conducting its audit, ZRA's scope was to determine the arm's length remuneration for the distribution activities of NZL, for the JS fiscal years 2010 to 2014. ZRA proposed to do this by benchmarking NZL using 13 countries in Western Europe. These countries were considered to be the most comparable, and their operating margins were lower than NZL's operating margins. 21. The operating margin selected by ZRA as the most appropriate profit level indicator for the benchmarking study it was undertaking, would be considered underestimated, as Zambia was found to be a higher risk country than its European counterparts. 22. It was contended that ZRA could not have used the transaction by transaction method as it was difficult to do so without comparable entities in the region against which NZL could be benchmarked. Further, NZL had not provided information to ZRA about its uncontrolled transactions with unrelated entities which could be used as comparables. 23 . Granted that position, the only option available to ZRA was to aggregate the transactions, as provided in the UN Manual of Transfer Pricing. According to ZRA, the said manual provides that, where there are no internal comparables, it is justifiable to J9 use entity level comparability, in order to aggregate the transactions so as to arrive at arm's length range. 24. Following the audit, the finding by ZRA was that there were no justifiable adverse economic conditions, inefficiencies or other legitimate business reasons, for the losses incurred by NZL. On further analysis of the information, adjustments were made culminating into an assessment ofZMW13, 860, 103. 00 as gross income tax payable by NZL. 25 . ZRA justified its adjustment due to the disproportionate debt to equity ratio . This was reflected in evidence which revealed that, NZL had less equity and invariably, did not have arm's length equity. This was against the recommendation of the OECD that a company should have at least 10% minimum of the total value of the company in equity and not be thinly capitalised as was found to be the case with NZL. 26. Aggrieved with the ZRA findings, NZL appealed to the Commissioner General, who upheld the assessment made by ZRA. That outcome is what prompted NZL to further appeal to the Tax Appeals Tribunal (the Tribunal). Proceedings before the Tribunal JlO 30. In dealing with the issues raised, grounds one and three were considered together by the Tribunal. Grounds four, five and six, to which we will revert, were dealt with separately. 31. Starting with the issues raised in grounds one and three, the Tribunal considered the points of objection raised by NZL, that the benchmarking was erroneous, and against entities in Europe, that were not comparable to NZL. That ZRA also rejected the transfer pricing analysis submitted by NZL without giving reason. 32. Upon analysis of the evidence the Tribunal considered whether, a wrong assessment had been undertaken by ZRA when it failed to test any of the NZL's transactions in order to assess whether the latter, was compliant with the arm's length principle as envisaged by law. The Tribunal also considered whether, ZRA failed to objectively test the related party transactions against which the transfer pricing concerns were raised and, as result, made assumptions and estimates that were unreasonable. 33. The Tribunal referred to the Income Tax Act, Cap. 323 section 97 A, which recognises and permits transactions between or among related parties or entities if conducted at arms' length. J12 According to the Tribunal, the crux of NZL's appeal before it was whether, ZRA was correct at law when it did not test each transaction separately, but opted to aggregated the transactions as a way of determining compliance with the arm's length principle. 34. The Tribunal considered regulation 12 (3) of the Income Tax Regulations of 2000, that set out the methods to be applied in transfer pricing, stated as: (a) comparable uncontrolled price method, (b) resale price method, (c) cost plus method, (d) transaction net margin method, or (e) transaction profit split method. Anchored on regulation12 (3), the Tribunal found that, one of the cardinal steps to be applied in testing transactions for arm's length compliance, is functional analysis. This analysis seeks to identify and compare the economically significant activities and responsibilities undertaken, assets used and risks assumed. It was noted that this analysis is essential, irrespective of the transfer pricing method used. 35. It was also noted that there were no comparables to be found either within the country or on the continent, against which the transactions could be accurately tested and benchmarked. The J13 Tribunal determined that, according to the OECD guidelines, benchmarking involves, amongst other requirements: (i) determining the years to be looked into; (ii) an analysis of the circumstances of the tax payer; (iii) understanding of the controlled comparables; (iv) a review of the existing internal comparables; (v) determination of information on external comparables; (vi) determination and making of comparability adjustments where possible; and finally, (vii) an interpretation of the data collected. 36. The Tribunal, nonetheless, also noted that although the process alluded to in paragraph 34, is considered good practice, it is not compulsory. Any other search process may lead to identification of reliable comparables that may be acceptable since the outcome is said to be more important than the process. The reason being that, in itself, the process does not guarantee that the outcome will be at arm's length. 37. Having considered the whole of the evidence before it, the Tribunal determined the issue as identified in paragraph 33, on the premise that there was no legal requirement for NZL to comply with the request by ZRA to furnish documents relating to J14 transfer pricing information, as this was not provided for in the Income Tax Regulations of 2000. There were, as a result, no internal comparables by which the transactions could be measured. This is what informed ZRA's decision to aggregate the con trolled transactions. 38 . The Tribunal then, proceeded to delve into whether, it was possible to test such controlled transactions for arm's length compliance by aggregation. In that regard, it was noted that one of the key features of aggregation was that the transactions ought to be closely linked or continuous or have a close economic link in purpose, between or among, the controlled transactions. 39. The finding was that for transactions to be aggregated, they cannot be segregated and where they are segregated, they cannot be adequately evaluated on a standalone basis . The Tribunal agreed with ZRA, only to the extent that, if transactions are benchmarked separately, they may not bring out the actual value created which the transactions would yield if placed together. 40 . The determination was however, that on the particular facts in casu, the controlled transactions that were aggregated comprised of unrelated transactions with no close economic link or JlS characteristics as required and set out in the UN Manual on Transfer Pricing. The Tribunal had difficulty accepting how the purchase of finished products for example, could be linked or closely related to payment of royalties, management services or shared services. 41. The finding was that, bundling such transactions together for the purpose of aggregating, would be synonymous to bundling transactions for products such as apples, bread, books and so forth. That doing so, would inevitably, produce inaccurate results in testing for adherence to the arm's length principle. The further finding was that, the use of aggregation, coupled with the failure to test each transaction for arm's length adherence, produced results that were not only erroneous but also excessive. 42. For the reasons given, NZL succeeded on grounds one and three, that the benchmarking was erroneous; and that ZRA did not give reasons why it rejected the transfer pricing analysis that NZL had submitted. 43. On ground two, the allegation was that ZRA had based its assessment on the premise that NZL could not run at a loss since incorporation. The Tribunal found that although loss making is J16 a risk factor to be considered when reviewing transfer pricing, there was nothing wrong with ZRA proceeding with the audit query. However, the Tribunal went on to find that, it was incorrect for ZRA to conclude that the loss making was as a result of non-compliance with the arm's length principle. Accordingly, NZL succeeded on this ground as well. 44. The contention raised in ground four, was that ZRA had erred in categorising NZL as a low risk distributor, when it was shown proof that it was in fact a full-fledged distributor, undertaking all sales and distribution functions and bearing of the risk. On the available evidence, the Tribunal formed the view that Nestle Zimbabwe exercised a huge amount of control over NZL. NZL received all manner of service support in different areas of its business operations. There was very little NZL did on its own without direction and assistance of its related entities. The Tribunal upheld ZRA's classification of NZL as a low risk distributor and dismissed ground four of the appeal. 45. In ground five, it was contended that ZRA had erred when it used benchmarks that are neither comparable to the nature of NZL's business in Zambia nor the macro-economic conditions. The Jl 7 Tribunal reiterated that in as much as there is nothing wrong in taking comparisons between transactions among related parties to those between independent parties, the key is understanding that comparable does not mean identical. None of the differences being compared, if any, could materially affect the pnce or financial indicator being examined in the selected transfer pricing method. The comparison is about the situations of the transactions. The focus should therefore, not be solely on the pnce. 46. Against that backdrop that the issue in benchmarking involves situational comparability, the Tribunal found that ZRA's use of businesses in Europe to benchmark NZL was disproportionate. That this is what led to an inaccurate transfer pricing adjustment. 4 7. Lastly, regarding ground six of the appeal, by which ZRA was faulted for having added back unrealised foreign exchange losses that were attributed to loans, when the same were not included as expenses in NZL's financial statements. The Tribunal determined that section 29A ( 1) of the Income Tax Act, provides for which foreign currency exchange gains or losses J18 other than those of a capital nature, shall be assessable or deductible, as the case may be. 48. In that regard, ZRA could only make adjustment for foreign exchange gains or losses, if it could demonstrate that NZL had claimed these exchange gains or losses, which on the evidence before it, the Tribunal found ZRA failed to do. The adding back of unrealised foreign exchange losses was therefore, reversed and ground six succeeded. 49. NZL having succeeded on all the grounds save for ground four, it was found that the assessment by ZRA that NZL was liable to pay a sum of ZMW13, 860, 103. 00 was wrongly arrived at, as the assessment was based on inaccurate transfer pricing method, disproportionate comparables, and an unjustified add back of unrealised exchange losses. Accordingly, the Tribunal set aside the assessment and ordered that a reassessment be conducted within 180 days. Appellant's Appeal to this Court 50. Dissatisfied with the findings of the Tribunal ZRA launched an appeal to this Court on five grounds, that were couched in the J19 following terms: "1. The tax Tribunal erred in law and fact when it shifted the burden of proof by requiring the appellant to prove that there was something wrong with the respondent's transfer pricing in its transactions with the related parties which resulted in the respondent's loss making. 2. The tax Tribunal erred in law and in fact when it held that the transactions between the respondent and in its related parties could not be aggregated but should have been tested individually contrary to both the 2010 OECD Guidelines, the United Nations Manual on Transfer Pricing for Developing Countries and the evidence tendered at trial. 3. The tax appeals Tribunal erred in law and in fact when it held that at the time the respondent entered into the transactions which were subject of Transfer Pricing there was no legal requirement for the respondent to document its transfer pricing information and, on request by the appellant to provide such transfer pricing documentation to the appellant for purposes of demonstrating that the actual conditions of the respondent's related party transactions were consistent with the arm's length conditions. 4. The tax appeals Tribunal erred in law and in fact when, having properly acknowledged the importance of a functional analysis as part of the comparability analysis to determine the arm's length condition including price and further having correctly characterised the respondent as a low risk distributor, the Tribunal failed to recognise that the characterisation of the respondent as a low risk distributor was crucial in selecting the tested party and the appropriate transfer pricing method. 5. The tax appeals Tribunal erred in law and in fact when, despite having correctly found as a fact that there were no internal comparables in Zambia and on the African Continent, it still went ahead to hold that the appellant used benchmarks that are neither comparable to the nature of the respondent's business J20 in Zambia nor the macro and micro economic conditions in Zambia, thereby contradicting itself and potentially tendering Transfer Pricing compliance in Zambia nugatory". Appellant's Arguments in support of the Appeal 51. For purposes of convenience, we will, in dealing with the appeal and cross appeal, continue referring to the appellant, Zambia Revenue Authority, as ZRA and the respondent, Nestle Zambia Limited, as NZL. 52. Heads of argument 1n support of the appeal were filed by Counsel for ZRA on 22 nd February, 2021. On ground one of the appeal, it was argued that the position of the law in our jurisdiction is that if the tax authority has assessed a taxpayer for a sum of money, the burden lies on the taxpayer to show that the assessment is an excessive amount. Counsel illustrated the argument by examining this principle in three jurisdictions being, the United States of America (USA), the United Kingdom (UK) and our own Republic of Zambia. 53 . Starting with the USA, reference was made to the cases of Sunstrand Corp. v Commissioner, 96 T. C. 226, 353-54 (1992); Seagate Tech Inc. v Commissioner, 102 T. C.149. 164 (1994); Amazon. Com v. Commissioner, 148 T. C. No. 8 at page 68 (2017) J21 107 T. C. and Welch v Helvering 290 U. S. 111) 115 (1933). Counsel submitted that, it was in those cases held that, determinations of the Internal Revenue Services (IRS) are generally presumed correct and taxpayers bear the burden of proving them erroneous. 54. In the United Kingdom, this point was said to have been made clear in the case of Nicholson v Morris (HMIT) (1977) 51 TC 95, where the Court held that even where an amount assessed is manifestly incorrect, it is the taxpayer that bears the onus to provide the correct amount and not the duty of the revenue authority as it would be a very onerous and costly exercise. 55. In our jurisdiction, Counsel relied on the well-known position as stated in the case of Wilson Masauso Zulu v Avondale Housing Project Limited1, that he who alleges must prove the allegation. Section 106 of the Income Tax Act was further cited in support of the proposition that, every assessment by the Commissioner General under the Act, shall stand good unless proved otherwise by the person assessed, upon objection or on appeal. 56. Counsel urged that section 106 must be read in relation with section 97 A (2), as a taxpayer is not only required to J22 demonstrate that the assessment was wrong, but to also compute its taxable income based on the arm's length standard. 57. It was contended that the Tribunal had departed from its own previous decisions stating that the onus is on a tax payer to discharge or disprove a tax assessment, failing which it shall stand valid. The case of Twikatane African Art Company Limited v Zambia Revenue Authority 1999/ RAT/ 36 was amongst numerous other decisions of the Tribunal, cited in support of the submission that, in casu, the Tribunal had misdirected itself when it shifted the burden of proof on to the tax authority, ZRA. 58 . On ground three of the appeal, it was submitted that at the time ZRA conducted an audit on NZL, there was in fact a legal requirement to document transfer pricing. This documentation was for purposes of demonstrating that the actual conditions of the NZL related party transactions were consistent with the arm's length conditions. 59 . On the findings of the Tribunal that there was no legal requirement for ZRA to request for transfer pricing information, reference was made to the Income Tax Act (Transfer Pricing) Regulations No. 18 of 2018 and section 47(1) of the Income J23 Tax Act. The latter, provides that the Commissioper General may by notice in writing require any person to furnish him within a reasonable specified period, further returns or particulars in relation to any matter contained in a return under the Act. The request could also be in relation to any transactions or matters that appear to the Commissioner General, to be relevant to the ascertainment of the income of that person for tax purposes. 60 . Section 58 (a) of the same Act further empowers the Commissioner General to request any person to produce for examination, accounts, books of accounts, and any other documents as specified. 61. It was argued that ZRA's letter to the NZL of 20th April, 2016 was categorical as to the information required. NZL failed to provide all the requested information which would have assisted ZRA in selecting the appropriate comparables and the appropriate transfer pricing method. That, lack of mention of the term 'local file ' in the law, does not negate the statutory authority conferred on the Commissioner General to request for any information from a tax payer, to enable the Commissioner General verify the correct income for tax purposes . J24 62. Counsel submitted that the appropriate response to a taxpayer that is evasive and fails to provide necessary information is the one found in the case of Nicholson v Morris (HMIT) (1977) 51 TC 95. It was in that case held that, it is the duty of every taxpayer to make his own return. Where there is failure to support such return or to come completely clean and the taxpayer gives no evidence, he cannot be surprised if he is lumbered with more than he in fact received. Respondent's Arguments in opposition to the appeal 63. Counsel for NZL filed Heads of argument on 7 th April, 2021 1n which he took no issue on the settled position of the law that it is indeed the party who alleges who must prove the allegation. The argument by Counsel for NZL was rather that the Tribunal never shifted the burden of proof to ZRA. It merely considered the facts and evidence led by both parties and made findings on issues raised by ZRA. In the event, the Tribunal cannot be faulted for finding that the evidence adduced by NZL does not show any wrong doing on its part. 64. Counsel equally relied on our decision Wilson Masauso Zulu 1 for the contention that, a trial court has a duty to adjudicate upon JZS every aspect of the suit between the parties, so that every matter in controversy is determined in finality. It was argued that before the Tribunal made the findings allegedly, shifting the burden of proof, it made a finding that NZL having incurred losses for five years was not, in itself, an indictment of the ZRA's wrong findings with regards to the NZL's transfer pricing. Similarly, the Tribunal's finding on the benchmark used by ZRA to assess NZL was made after assessing the evidence of both parties. 65. The submission on the point was that, the obligation to discharge and prove that the amounts assessed by ZRA were excessive, was discharged by the NZL. NZL had filed its annual tax returns for the period under assessment, as a self-declaration of its tax affairs. All documentation requested by ZRA was provided by NZL. This position was acknowledged by a witness of ZRA who stated that they had carried out tests on all related party transactions. That evidence goes to confirm that all requested documentation was provided to the ZRA, which assertion was not challenged by ZRA. 66. It was submitted that ZRA admitted to having received NZL's documentation on related party transactions together with the J26 master file . ZRA, erroneously, and without explanation, disregarded that documentation and substituted NZL's transfer pricing documents and methodology with that of its own. 67. It was argued that, in accord with the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrators, it was incumbent on ZRA to analyse NZL's documentation on transfer pricing. In the event that it did not agree with the same, ZRA should have provided reasons for rejecting the methodology applied, or determination of non-arm's length compliance to the transactions, before it proceeded to use its own methodology. 68. The submission was that, once all the necessary documentation as proof of the arm's length nature of the transactions was provided by NZL, the burden of proof shifted to ZRA to explain its rejection of NZL's documentation. This is the obligation that ZRA failed to discharge. 69. In response to ground three that at the material time, there was no legal obligation on the tax payer to document its transfer pricing information, the submission on the part of NZL was that, this ground is misconceived as the Tribunal did not err in its findings. NZL maintained that it provided ZRA with its transfer J27 pnc1ng documentation in the form of the master file, which contained data for all related party transactions within its group. The Tribunal had correctly determined that ZRA did not query the master file upon production and that ZRA, in fact, relied heavily on it. Further, that besides the master file, ZRA made no other demands for other information and it was in fact provided with all the information needed for assessment. 70. NZL submitted that the Tribunal was on firm ground when it found that the law was inadequate at the time as did not require Multinational Companies (MNCs) to maintain a local file. ZRA had admitted that the Income Tax (Transfer Pricing) Regulations No. 24 of 2018 specified the information that must be kept by MNCs in the local file. These regulations only came into effect after the fact and did not apply at the material time. Prior to the said regulations, the only requirement by law, was for MNCs to transact at arm's length with their related entities. 71. NZL contested ZRA's submission that sections 47 (1) and 58 (a) of the Income Tax Act, can be interpreted to mean that NZL was expected to maintain records that it had no obligation to maintain. According to NZL, the only reasonable interpretation J28 of section 47(1) in relation to section 97A of the Act is that, a tax payer is expected to have documentation that can show the arm's length nature of its transactions. 72. NZL argued that it would be absurd for the sections to be interpreted to mean that the Commissioner General can ask for information which a tax payer is not legally obligated to keep. The submission was that, the cases cited by ZRA in support of this ground are irrelevant. The Tribunal was on firm ground when it held that there was no legal requirement for NZL to maintain a local file. Submissions at the Hearing of the appeal 73. When the appeal came up for hearing, learned Counsel for ZRA sought to augment his arguments by placing reliance on foreign authorities that were filed into Court on 20th April, 2021. 74. In augmenting grounds one and three, Counsel cited the South African case of ABC Pty Limited v Commissioner for the South African Revenue Service, before the Tax Court of South Africa, in which the respondent raised an additional assessment on 11th January, 2016. This was intended to effect an adjustment made to the applicant's taxable income relating to transactions it J29 considered not to have been conducted at arm's length. Counsel argued that the circumstances were similar to the transfer pricing audit that was being undertaken by ZRA on NZL in the appeal in casu. 75. The submission, in that regard, was that like in the South African case, the request for information prior to the audit was made pursuant to the principal Act. In our case, these are sections 48 and 58 of the Income Tax Act. It was not any regulations as held by the Tribunal in its finding that there were no regulations in place compelling NZL to store information in a local file. In the South African case, it was found that the duty to keep documents rests on the tax payer and that it is in its own best interests to document its transactions. 76. The further argument was that, NZL had failed to prove its transactions were at arm's length as required by section 97 of the Income Tax Act. Its fate was, as a result, sealed. It was submitted that the approach of the court in the South African case is in consonance with our holding in the Savenda Management Services v Stanbic Bank Zambia Limited2 , where we stated that a judge must interrogate the submissions of a J30 party and not merely reproduce them. Counsel further submitted that, one of the documents requested for was a local file, but the Tribunal was of the view that this was not provided for in the regulations at the time the request was made. It was contended that what the 2018 regulations provided for was simply nomenclature, but that the requirements remained the same and could be requested under the Act. 77. The other issue that learned Counsel for ZRA addressed was the burden of proof. It was submitted that the Tribunal had erred when it had found that ZRA had failed to prove that NZL had not transacted at arm's length. A Nigerian case, Prime Plastichem v Federal Inland Revenue Service was cited for the holding that, the burden of proof lies on a tax payer to prove that the transaction was at arm's length. It was submitted that the finding of the Tribunal in this matter, that the burden of proof lay on ZRA, the tax authority, was fundamentally flawed. 7 8. We were directed to page 10 of the said case and page 6 7 of the UN Manual on Transfer Pricing which provides that the burden of proof in tax matters typically fallows domestic rules of a particular country. The submission was that, the issue of burden J31 of proof in this jurisdiction is well settled and not up for debate. The onus in the matter lay on NZL to prove that its transactions with related parties, and the services rendered, were at arm's length, as well as, when these services were rendered and by whom. 79. The case of W. T. Ramsay v Inland Revenue Commissioners3, was relied upon in which the tax payer company had sold land and realized a chargeable gain for the sole purpose of reducing the amount of capital gains tax payable. This was through a scheme known as the capital loss scheme that was designed to create artificial capital losses on share transactions so as to set off against chargeable gains, with the tax payer paying little to no tax, as a result. In the earlier English case of Inland Revenue Commissioners v Duke of Westminster4 this was said to be a genuine scheme. However, under the recent 'step doctrine' as developed in the United States, which allows for considerations of a scheme or transaction as a whole and not a step by step examination, the House of Lords found that, although separate steps were genuine and had to be accepted under the Westminster doctrine, the court could on the basis of the findings J32 made and its own analysis, consider the scheme as a whole. It was not confined to a step by step examination of the transactions made. 80. The submission was that, where a taxpayer has engaged in a series of transactions, if those transactions were intended to avoid taxation, they will be collapsed into a single composite unit and then taxed accordingly. 81. Counsel also referred us to a Canadian case in OSFC Holdings Ltd. v Canada, 2001 DTC 5471, in which the appellant had appealed to the Tax Court of Canada. The appeal was from an assessment of the Minister of National Revenue, disallowing a non-capital loss deduction in the computation of the appellant's income in its 1993 taxation year, and the deduction of a related non-capital loss. Professor Krishna explained the step transaction in the following words: " ... for the purpose of the GAAR (general anti-avoidance rules) a series of transactions: refers to the integration of individual and separate steps into a composite transaction. The linkage of the separate steps into a series results from their inter-dependence and the manner in which the transactions are structured" . 82. Learned Counsel for ZRA argued that, since the step transaction applies in the United States, United Kingdom and Canada, it also J33 applies in this jurisdiction by virtue of Article 7 of the Constitution, The British Acts Extension Act, Cap. 10, and English Law (Extent of Application) Act, Cap. 11 of The Laws of Zambia. He further referred to Statutory Instrument No. 20 of 2000, regulation 3 (2) which provides that: "For the purpose of subsection (3) of section ninety-seven C, a series of arrangements shall not be prevented from being regarded as a series of arrangements by means of which conditions have been made or imposed as between any two persons by reason of either or both of the following matters: (a) that there is no arrangement in the series to which both those persons are parties; and (b) that there is one or more arrangements in the series to which neither of those persons is a party". 83. Counsel argued that in all the cases brought to our attention, the taxpayers were engaging in a series of transactions trying to avoid taxation, but the Courts decided to collapse all the transactions and then regard them as a composite unit, thus allowing the authority to tax them accordingly. 84. Reference was made to the three tests used 1n the step transaction doctrine, being (i) the mutual interdependence test; (ii) end result test; and (iii) binding commitment test; which tests, apparently, need not be used at the same time. J34 85. In reference to the mutual dependency test, Counsel explained that in relying on this test, the Court will look at whether, the transactions are mutually interdependent. In applying the test to the case in casu, the controlled transactions considered, such as royalties, purchases and loans, it was submitted that the same are connected. The reason was that the products sold are patented. When purchased, one eventually has to enter licensing to pay for the royalties. 86. The other case ref erred to was Ensign Tankers (Leasing) Ltd v Stokes (HM Inspector of Taxes)5 in which Lord Wilberforce was quoted in the Westminster case, where courts are cautioned not to consider a document or transaction in blinkers, isolated from any context to which it belongs. If it can be deduced that a document or transaction was intended to have effect as part of a nexus or series of transactions, or an ingredient of a wider transaction, there is nothing in the doctrine to prevent it from doing so. Proceeding in that manner, is not to prefer form to substance or substance to form. It is rather, the task of the court to ascertain the legal nature of any transaction to which it is sought to attach a tax or a tax consequence. Where it emerges J35 from a series or combination of transactions intended to operate as such, it is that series or combination which may be regarded. 87. Counsel submitted that the role of the court is to interpret a document before it and consider whether the vanous transactions such as purchases and royalties are interdependent and what the end result is. It was submitted that there was a significant departure from the position in the Westminster case to that obtaining in the Ramsey case. That the House of Lords had taken the stance that, although it appeared that the courts were legislating beyond their sphere, the courts were not obliged to stand still and do nothing as the facts in each case must be established and a legal analysis made in these tax aggressive activities. 88 . In that regard, Counsel relied on the reasoning of Lord Geoff of Chieveley, who stated that there is a fundamental difference between tax mitigation and unacceptable tax avoidance . The former involved the taking advantage of the law to plan one's affairs so as to minimise the incidence of tax, whilst the latter, typically involves the creation of complex artificial structures by J36 which a taxpayer conJures of thin rur a loss or a gain or expenditure which would have never existed. 89. In relating the said legal arguments to the case in casu, now before us, Counsel for ZRA argued that NZL claimed to have been making losses for five years yet at group level profits were being made. It was submitted that, the so-called losses were self inflicted and not actually incurred. He went on to underscore the point that, a taxpayer has an obligation to furnish information and failure to do so, places a revenue authority in a precarious position as they must guess and such a tax payer should not complain. 90. Counsel cited the case of Nicholson v Morris (H. M. Inspector of Taxes)6 for the further submission that the onus is on the taxpayer to show when an assessment is wrong, as it is the taxpayer who has intimate knowledge of the transaction and is in the position to provide the right answer. 91 . According to Counsel, it is idle for any taxpayer to say to the revenue authority that, 'somewhere hidden in your vaults is the right answer go and dig them out of the vaults.' Counsel further alluded to and placed emphasis on the case of National Justice J37 Compania Naveria SA v Prudential Assurance Co. Ltd (The Ikarian Reefer)7, which he submitted, was important for its pronouncement on the duties of an expert witness. That expert evidence presented to the Court should be seen to be independent and uninfluenced as to form or content by the exigencies of litigation. 92. It was Counsel's contention that the testimony of the expert called to testify by NZL was highly flawed, such that even the Tribunal found that part of the factual information given was inaccurate. Counsel urged for rejection of the NZL's contention that, as ZRA did not call its own expert witness, the evidence of NZL's expert witness had to be admitted by this Court. It was argued that contrary to the Ikarian Reefer case, irrespective of his/her competence, the evidence of an expert cannot usurp the powers of the court. The fact that NZL's expert had tendered evidence, did not mean that the Tribunal had to automatically accept it. The Tribunal was duty bound to interrogate such evidence. 93. Lastly, on ground five, when asked to comment on functional analysis, Counsel for ZRA responded that it refers to a situation J38 where a tax payer or tax authority, analysed the functions, assets employed and risks of an entity. He argued that, this was the reason that ZRA used comparables from Western Europe. The focus was on the function the entities performed and not the age of the entities or the length of time spent in business. 94 . The net effect was said to be that if a particular business entity performs routine functions , employs the same assets, assumes routine risks, then at the end of the day, they would have routine profits. An example given was that if you assume higher risks and employ capital intensity, it is expected that you must be compensated for the investment that you have made, commensurate with assets employed. 95 . On the question of comparables, Counsel maintained the position that there were no comparables in Zambia or in the sub Saharan Africa. He conceded that the comparables picked had a different economic situation to that obtaining in Zambia, but asserted that alternatives had to be found outside the continent. This was the reason ZRA had sourced comparables in Western Europe, just like NZL itself had used comparables from Asia Oceanic countries . He submitted that it was inevitable that both J39 parties had to move outside the continent to find comparables and NZL could not turn around and have issue with ZRA doing the same . 96. In his response, Counsel for NZL relied on the heads of argument filed on record. His brief oral augmentation was that, the cases relied on by Counsel for ZRA were not on all fours with the circumstances and the facts of the appeal before this Court. He noted that his counterpart did not specifically zero in on specific portions of the record when ref erring to the foreign cases relied on. 97. Counsel reiterated that NZL had provided ZRA with all relevant documentation pertaining to all related party transfer pricing. That ZRA's own witness confirmed to the Tribunal that the documentation relating to the group transfer pricing was availed. It was further contended that section 99B as read with section 98 of the Income Tax Act, provides that any taxpayer who fails to provide information requested by the Commissioner General commits an offence, and that at no time did any of the ZRA's witnesses protest that certain documentation was not availed. J40 98. Counsel for NZL further countered that the reason ZRA did not find it necessary to test all the underlying transactions made by NZL was that ZRA had failed to undertake a bench-marking exercise as confirmed by its own witness RWl, who also agreed that it was appropriate to test each transaction. It was submitted that a bench-marking exercise is required in order to come up with suitable comparables and ground two, should fail for failure to meet a condition precedent. 99. Counsel maintained that there was no requirement to maintain transfer pricing documentation until the coming into effect of the Income Tax (Transfer Pricing) Regulations No. 28 of 2018. That position notwithstanding, NZL did keep certain information which it availed to ZRA. He reiterated that the Tribunal did not shift the burden of proof to ZRA. What it did, was to merely consider the evidence of both parties when making its findings. Counsel submitted that benchmarking is mainly used to test the arm's length nature of the related party transactions carried out between related parties. The Nigerian case of Prime Plastichem v Federal Inland Revenue Service, does not actually assist ZRA's case. At page 10 of the said case, it was emphasised why a J41 benchmarking study was a critical part of transfer pnc1ng documentation or policy. 100. On the selection by ZRA of comparables from Western Europe, Counsel for NZL was of the view that the more suitable selection would have been countries from East Asia whose economy was similar to the Zambian one. It was argued that the European economy was more advanced with companies that were in existence for as long as sixty years in comparison to NZL that had only been operating for five years. That this was actually a rebuttal as regards the suitability of the comparables used by ZRA. 101. In his oral reply, Counsel for ZRA reiterated that the Prime Plastichem v Federal Inland Revenue Service case, supports its position as the court in that case held that, preparation of a benchmarking study is obligatory for taxpayers as part of transfer pricing documentation. He conceded that ZRA's witness did say there was no benchmarking exercise undertaken by it, as the obligation was that of NZL, as the taxpayer. Evidence on record was referred to where the expert witness confirmed that ZRA had analysed some controlled transactions. J42 102. On NZL's contention that it had submitted the local file to ZRA, Counsel for ZRA asserted that it was actually the master file that was submitted as shown at page 936 in Volume 2 of the record . That lack of relevant information interfered with the bench marking exercise which, in any event, was supposed to be undertaken by the NZL. Consideration and Determination of the main appeal 103. We have read the record of appeal, considered heads of argument together with oral augmentation made at the hearing by Counsel for the parties on both sides. 104. The record shows common ground evidence is that ZRA commenced a transfer pricing audit of NZL covering the years 2010 to 2014, in order to determine arm's length remuneration for the distribution activities of NZL. A benchmarking study was thereafter undertaken of independent companies operating in Western Europe. 105. In that regard, ZRA resorted to employing the use of operating margins, as a means of measuring NZL's pricing strategy and operating efficiency. ZRA found that NZL's transactions were, J43 apparently, not undertaken at arm's length and required re computation. This resulted in a re-assessment and adjustment of profit for the years 2010 to 2014 in the total sum of K56, 579, 048. 00. NZL was found liable to pay a final levy of K 13, 860,103.00 as gross income tax. 106. The grievance of NZL with that amount was directed at the process resorted to by ZRA in arriving at the assessment and the final figure. This was the premise on which NZL appealed the assessment to the Tribunal. 107. Starting with the issue raised in ground one of the appeal, on who bears the burden of proof where an assessment has been made by the tax authority, and the tax payer disputes the correctness of such assessment. Learned Counsel for ZRA has correctly, pointed out and his counterpart Learned Counsel for NZL has magnanimously, conceded the settled position of the law, that the burden of proof generally, lies on the person making the assertion. In our view that legal position is no different in this particular instance, where an assessment has been made by the tax authority and the tax payer disputes the assessment asserting it is incorrect. The burden to prove the assertion with J44 cogent evidence, clearly, lies on the taxpayer. We reiterated that position in the case of Agro Fuel Investment Limited v Zambia Revenue Authority 8 in the following words: "We agree with Mr. Locha, that mere verbal assertions that the goods in question were in transit and did exit Zambia, was not sufficient to enable the appellant to discharge the burden of proof cast on it". 108. The above position is fortified by section 106 of the Income Tax Act, relied upon by Counsel for ZRA. The section provides as follows: "Subject to the Commissioner-General's power relating to assessment, every assessment under this Act shall stand good unless proved otherwise by the person assessed upon objection or appeal under this Part". (Underlining for emphasis only) 109. The words in section 106 as quoted above, are unambiguous and go to support the position that, where an assessment is made and the tax payer disputes such assessment through an objection, the tax payer bears the burden of proving the incorrectness of the assessment by providing evidence to support the objection. When the issue arose in the English case of Brady (Inspector of Taxes) v Group Lotus Car Cos Pie & Another9 it was opined as follows: "It is common ground between the parties, and they so directed themselves, that perhaps surprisingly, the burden of proof in an J45 appeal to the commissioners against an assessment made by the Revenue is not on the Revenue to justify the assessment but on the taxpayer to show that the assessment is not justified. That is established by the decision in T Haythornthwaite & Sons Ltd v Kelly (Inspector of Taxes) (1927) 11 TC 657 at 658, in which it was held by the Court of Appeal- 'that the lay on Commissioners upon sufficient evidence that the assessment was erroneous, that as such evidence was not produced, the Commissioners had proceeded correctly the assessment' .. . ... " (bold facing and underlining for emphasis only) the Appellants in confirming to satisfy the onus 110. In that case the taxpayer had deliberately produced no evidence to show that the assessment was wrong and Sargant, W, had this to say: "'In that state of things it seems to me that the very first thing that a person having to prove that the assessment made by the Inspector was wrong had to do-the very first thing that was requisite-was to produce the books of the Company; and besides that-because the books probably would not in any way be conclusive; it might very well be that these outside transactions were kept outside the books of the Company-to produce the managing director and other high officials of the Company for the purpose of showing that there were in fact no such outside profits, no such gambling on the part of the Company. Silence, or the absence of evidence of that kind, was in my judgment evidence-very cogent evidence too-to show that the assessment made by the Inspector could not be displaced on the part of the (boldfacing and underlining Company." supplied) (at 671-672) 111. Learned authors of Halsbury's Laws of England, 4 th Edition, paragraph 1677 echo the same position of the law when they J46 state that: "Upon an appeal to the General or Special Commissioners against an assessment, the onus is upon the taxpayer, by evidence given ...... to demonstrate that the assessment ought to be reduced or set aside. If the appellant fails to lead evidence before the commissioners, he cannot have the assessment reduced or displaced". 112. Against that backdrop of the law as recounted in paragraphs 108 to 110, we note that ZRA contention before the Tribunal was to the effect that NZL neglected to produce evidence to counter its findings on assessment. The Tribunal acknowledged that ZRA made a request for various documents and information from NZL. It also made a finding of fact, that NZL had failed to com ply with the said request. 113. ZRA having made the assessment, the burden of proof shifted to NZL as the taxpayer that was dissatisfied and had raised objection to the assessment, to prove that there was nothing wrong with the transfer pricing in the transactions with its related parties. Those are the reasons that we find merit in ground one of the appeal, faulting the Tribunal as having shifted the burden of proving that the assessment was wrong from NZL to ZRA. J47 114. On ground two, ZRA faults the Tribunal's findings that transactions between NZL and its related parties should not have been aggregated but should have been tested individually contrary to the OECD Transfer Pricing Guidelines. In that regard, we have noted that, chapter 3, A. 3.1, paragraph 3. 9 at page 110 of the said Guidelines provides that, the ideal scenario in order to arrive at a precise approximation of arm's length conditions is to apply the conditions on a transaction by transaction basis. It is nonetheless, acknowledged that there are situations where transactions are separate but so closely linked, that they can only be evaluated adequately by aggregating them. 115. By way of illustration reference is made to some long-term contracts for the supply of commodities or services, rights to use intangible property, and pricing a range of closely linked products, such as in a product line where it is not practical to determine the price of each product or transaction. 116. Other examples given include the licensing of manufacturing know-how and the supply of vital components to a manufacturer, routing a transaction through another associated enterprise, or where a tax payer wishes to take a portfolio approach, by J48 bundling certain transactions for purposes of earning an appropriate tax return across the portfolio rather than through a single product. When such a scenario arose in the Indian case of Deputy Commissioner of Income Tax, Circle -11 (1) v M/s. Adcock Ingram Limited I. T. (TP) 10 it was held that: "In order for transaction to be closely linked, it is not necessary that the transactions need to be identical or even similar, for example, a collaboration agreement may provide for import of raw materials, sale of finished goods, provision of technical services and payment of royalty. Different methods may be chosen as the most appropriate methods for each of the above transactions when considered on a standalone basis. However, under particular circumstances, one single method maybe chosen as the most appropriate method covering all the above transactions as the same are closely linked". 117. We are persuaded by the above cited case, where aggregation of transactions, that were neither identical or similar, was found to be suitable in a benchmarking exercise. On that premise, we find the Tribunal narrowly construed the circumstances in which aggregation of transactions can be undertaken. 118. In another Indian case of Sony Ericsson Mobile Communications India (P) Ltd v CIT10 , it was held that the international transaction of expenses should be aggregated with other international transactions carried out by the assessed taxpayer, as a distributor, who either simply acts as an agent of J49 manufacturer or purchases goods from the manufacturer for resale on his own account. 119. The cited decisions although not binding appear to support the position taken by ZRA, that aggregation is actually a much wider concept. It encompasses transactions emanating from a common source being an order or a contract and that the transactions substantially flow from the said common source. Ground two of the appeal is sustained for those reasons. 120. Coming to ground three of the appeal, NZL's failure to provide the transfer pnc1ng information as requested by the Commissioner General of ZRA inevitably, affected NZL's ability to sufficiently demonstrate, that the controlled transactions conducted with its related parties, were done at arm's length as required by section 97 A of the Income Tax Act. The issue that a taxpayer was not at the material time required to maintain records of such information, in terms of the Income Tax Act (Transfer Pricing) Regulation No. 20 of 2000, is settled. Notwithstanding that position, section 58 of the Income Tax Act, adequately covers and gives powers to the Commissioner General to request any type of information, in the form required, JSO for the specific inquiry in question. 121. ZRA claimed information provided by NZL was inadequate. That being the case, the findings of the Tribunal that placed the onus of resolving NZL's objection to the transfer pricing assessment on ZRA, most certainly shifted the burden from NZL. NZL as the taxpayer bore the responsibility to provide comparables to its own business dealings. It conceded that there were no internal comparables within the country and on the continent. The Asia Oceania-Africa Countries (AOA), were, as a result, resorted to for comparison. 122. We acknowledge NZL contention that the law at the material time was inadequate as the Income Tax (Transfer Pricing) Regulations No. 24 of 2018 which now specify the information that must be kept by multinational in a local file only came into effect after the fact. Prior to those regulations, the law only required multinationals to transact at arm's length with their related entities, which was the only obligation to be met. In order to satisfy its tax obligations however, the tax payer was still obligated to maintain a record of all such relevant information as would be necessary to support compliance with tax obligations. JSl 123. ZRA could only obtain information from a taxpayer only where a document was specifically mentioned, in this case , a local file. Nonetheless, sections 47(1) and 58 (a) of the Income Tax Act are couched in general terms with regards to allowing the Commissioner General sufficient powers to obtain any information required for assessing tax payable on one's income. It is for that reason that we reject NZL's assertion that it was not required to provide the information requested by the Commissioner General for it was not obliged to keep certain records by law. 124. It is clear to us that that provisions of sections 47(1) and 58(a) of the Income Tax Act were sufficient for the Commissioner General to request whatever documentation was considered to be relevant for audit purposes, as argued by ZRA. Ground three equally succeeds. 125. In ground four of the appeal, the Tribunal is said to have properly acknowledged the importance of the functional analysis as part of the comparability analysis in determining arm's length conditions. It further, correctly characterised NZL as a low risk distributor, yet failed to recognise that this characterisation was J52 crucial in selecting the tested party and appropriate transfer pricing method. 126. Chapter III of the OECD Guidelines, at page 107, under Comparable Analysis, defines a comparison, to mean an examination of a controlled transaction under review and the uncontrolled transaction that is regarded as potentially comparable. It is advised that a methodical and consistent approach should provide some continuity or linkage in the whole analytical process and maintaining a constant relationship from the preliminary analysis of the conditions of the controlled transaction, to the selection of the transfer pricing method. This is achieved through the identification of potential comparables, which leads to a conclusion as to whether the con trolled transactions being examined are consistent with the arm's length principle. 127. Part of the process 1s said to require selecting the most appropriate transfer pricing method and applying it by finding reliable comparables. Where there are some uncontrolled transactions with a lesser degree of comparability than others, these should be eliminated. It is acknowledged in the Guidelines J53 that the search for comparables can be burdensome. This is on account of limitations in availability of information and an exhaustive search of all possible sources of comparables. 128. On the purpose of a functional analysis, the Guidelines at D.1.2.2., page 45, paragraph 1.42 provide that, in transactions between two independent enterprises, comparison usually will reflect consideration of the functions that each enterprise performs, assets used and risks assumed. The functional analysis is said to aim at identifying and comparing the economically significant activities and responsibilities undertaken, assets used and risks assumed by the parties to the transactions . 129. The discourse from paragraphs 126 to 128 goes to support the classification of NZL by ZRA as a low risk distributor that was upheld by the Tribunal. Paragraph 1.47 of the Guidelines provides that the functions carried out will determine to some extent, the allocation of risks between the parties, and therefore, the conditions each party would expect in arm's length transactions. An example given is where a distributor takes on responsibility for marketing and advertising by risking its own J54 resources, in such activities, it is expected that the return from the said activities will be commensurately higher and the conditions of the transaction will be different from when the distributor acts merely as an agent, being reimbursed for its costs and receiving income appropriate to its activity. 130. In its arguments ZRA demonstrated that the transaction method or transactional net margin method may be applicable in cases where one of the parties makes all the unique contributions whilst the other party does not, and that the tested party should be the less complex one . 131 . For its part NZL takes great exception at being characterised as a low risk distributor by the Tribunal. As already pointed out and accepted, it is clear that NZL's risk was most certainly properly classified as low risk. The Tribunal revealed its mind on how it reached that determination which was inf armed by the functions, assets inventory and know- how risks of the products that NZL sales emanate from its related parties and do not lie with it. NZL argued that whether or not ZRA used the correct tested party is not the issue but the fact that ZRA aggregated unrelated transactions. JSS 132. We find the challenge disparaging the selection process of the correct tested party, unsustainable. This is in light of the fact that the review of the controlled transaction directly influences the selection of the tested party and the application of the most appropriate transfer pricing method, in trying to determine whether a transaction or transactions were at arm's length. We have already determined in accordance with the Guidelines that it is not in every circumstance that transactions will have to be tested separately. As long as they flow from a common source, aggregation is permissible. Ground four of the appeal succeeds. 133. Finally, on ground five of the appeal, ZRA faulted the Tribunal as having erred in law and fact when despite having correctly found as a fact that there were no internal comparables in Zambia or on the African continent, it found that the benchmarks used by ZRA, were neither comparable to the nature of the business nor the macro and micro economic conditions in Zambia. That this rendered the transfer pricing nugatory. According to ZRA the Tribunal's ruling on the use of Western European comparables has far reaching consequences. That it sets a dangerous precedent where transfer pricing in Zambia can only be based on J56 local comparables. ZRA found it to be a gross misdirection for the Tribunal not to have analysed the comparables it had presented. As a result, The Tribunal also failed to compel NZL to present its own preferred comparables. This was despite having previously done so in its global transfer pricing report from Europe, Japan, Korea, Latin America etc. 134. ZRA contended that the Tribunal's holding that the use of Western comparables was neither comparable nor suited to the macro and micro economic conditions was contrary to the evidence tendered before it. This evidence revealed that NZL had perpetuated this position by not disclosing its economic activities at local level, but only did so at global level, leaving ZRA at large in the search of comparables. 135. ZRA maintained that the burden of proof is on the tax payer to prove compliance on arm's length basis in so far as sections 97 A and 106 of the Income Tax Act are concerned. 136. In response, NZL countered that ZRA has misconstrued the holding of the Tribunal as the issue was the comparison of NZL that had been trading for only five years to European companies that were well established for decades and made huge profits . J57 NZL asserted that the only contention it had was that its profits could not be compared to those in Europe, in view of the micro and macro-economic conditions that were different from Zambia. The Tribunal did not, at any point, hold that comparables from outside the country can never be used in Zambia but rather, that the comparable country or market must operate in similar economic conditions to Zambia for it to be comparable . 137. In as much as we agree that the Tribunal found that the use of a Western European entity as a comparable was inappropriate as the two countries or markets were incomparable on account of the complete disparity in the economic conditions. The Tribunal was not stating that comparables must in any circumstance be local. The finding was rather, that where no local comparables are found whether in Zambia or on the continent, it was incumbent on ZRA to find suitable and compatible comparables that represent as closely as possible, the economic conditions in Zambia to avoid a disparity in the analysis . 138. For the reasons given, we do not accept the proposition that transfer pricing would be rendered nugatory in Zambia, when all that was being requested for was fairness in making the JSB comparables between the entities. This could be achieved by ensuring that the market set up and the economy are at par or close enough in economic terms. This is provided for in the Guidelines at D.1.1., paragraph 1.33 as states that, in order for comparisons to be useful, the economically relevant characteristics of the situations being compared must be sufficiently comparable. 139. It goes on to define comparable to mean, that none of the differences if any, between the situations being compared could materially affect the condition being examined 1n the methodology such as pnce or margin, or that reasonably accurate adjustments could be made to eliminate the effect of any such differences. 140. As alluded to is some authorities referred to earlier in this judgment, it is clear that comparable although not meaning identical, should be close enough to not alter the outcome of the transaction being compared. We find merit in ground five of the appeal. J59 Respondent's cross appeal and Arguments in support of cross appeal 141. As recounted in paragraph 4 of this judgment, NZL filed a cross appeal and Heads of argument in support, on 8 th April, 2021. This was following a consent order entered into with ZRA to do so. The cross- appeal is anchored on two grounds faulting the Tribunal as having erred in law and fact when: 1. 2. it made a finding in the absence of supporting evidence on low risk distributor record that the appellant was a notwithstanding that there was sufficient evidence on record that clearly demonstrated that the appellant was a fully-fledged distributor; and it exceeded its jurisdiction when it ordered that the reassessment be conducted within 180 days by the respondent on account that there could be transfer pricing issues requiring scrutiny which conclusion prejudiced the appellant. further 142. On ground one of the cross-appeal, Counsel submitted that the Tribunal's finding that NZL is a low risk distributor based on the evidence that there is quite little that it does on its own, without direction and assistance from its related party, Nestle Zimbabwe, is not supported by evidence. Our decisions in Communications Authority v Vodacom Zambia Limited11 and Attorney General v Marcus Kampumba Achiume 12 were cited in aid of the J60 submission. The Tribunal also premised its finding on the understanding that Nestle South Africa retains ownership of the products, and the attendant risks in the know-how of the products distributed by NZL. It further made a finding that the functions, assets, inventory risks and know-how risks of the products that NZL sells do not lie with NZL but its related parties. 143. NZL argued that according to Advances in International Accounting Volume 18, a low risk distributor is an entity that only undertakes operational and tactical sales and marketing activities whose inventory risks are managed by its principal. Essentially, this makes it a low risk distributor of its principal that owns the goods distributed by the agent and bears all the risks attendant to the distribution of his goods by the agent. According to NZL, the real question before the Tribunal was who bore the inventory risk in the products it distributed? 144. The submission in that regard, was that NZL is a private company limited by shares and duly incorporated under the Laws of Zambia. This makes it a separate legal entity from its shareholders and sister companies in the larger Nestle group. Those facts considered, the Tribunal ought to have taken judicial J61 notice that multinational companies get general business direction from the Group. As a Group, they have continental, regional and sub-regional management units, which oversee the strategic direction of subsidiaries within the Group, but that this does not render the independence of the subsidiaries futile. 145. It was contended that the administrative arrangement between NZL and the unit in Zimbabwe was not strange. Such types of arrangements were common throughout the Southern region to allow ease of internal administration. NZL contracts and pays for the services it receives. They are not paid by a related party on its behalf. Further, that NZL finances the purchase of the products it distributes, pays rentals for the warehouse where the products are stored, and delivers them to its customers. These financials were never queried by ZRA and they reveal that the distribution, selling and marketing expenses lie with NZL, and not its related parties. 146. It was contended that ZRA did not challenge this evidence and that failure to do so amounted to acceptance of such evidence as stated by the authors of Murphy on Evidence. The submission was that, we ought to interfere with the finding that it is a low J62 risk distributor as this is not supported by the evidence on record. It was further submitted that the General Licensing Agreement, (GLA) between Nestle South Africa and NZL, allows the latter, strictly for its own purpose, use of intellectual property 1n storing, supplying, importing, exporting, marketing, promoting, distributing and selling of the products. 14 7. It was contended that the provision of technical support and other management services provided by the related parties were on an on-call basis to other entities in the region not only NZL. They are essential functions that any company needs to survive. The fact that NZL's Finance Manager was sent from Zimbabwe as an expatriate was not anything peculiar. This, in fact, is a common arrangement amongst multinational companies where employees are sent to other markets to provide necessary services and to gain experience. That NZL as a distributor obtains its products from all over the world, with majority of the products procured from Nestle Zimbabwe and Brazil. In the event, it could not be classified as a low risk distributor. 148. On ground two, the submission was that, contrary to the case of Anderson Kambela Mazoka & Others v Levy Patrick J63 Mwanawasa & Others 13 , the Tribunal went beyond its jurisdiction and the issues NZL set out in its pleadings when it made a finding that there was conduct on the part of NZL worth investigating. That the main issue before the Tribunal was simply, whether or not ZRA correctly carried out the tax assessment on NZL. That the Tribunal's finding that there were transfer pricing issues requiring scrutiny and directing ZRA to conduct a reassessment are legally flawed as it had no jurisdiction to make such pronouncements. That under our Income Tax Act, only the Commissioner General can legally, order a reassessment. The Tribunal therefore, assumed an accusatory role, thereby exceeding its jurisdiction. 149. ZRA in its heads of argument in response to NZL's cross-appeal, submitted that there is abundant evidence on the record to support the notion that NZL is a low risk distributor. The cases of Nkhata & Four Others v The Attorney General1 4 ; Wilson Masauso Zulu v Avondale Housing Project Limited 15 ; Nkongolo Farms Limited v ZANACO Limited & Others 16 , where it was held that an appellate court will only reverse findings of fact made by a trial court, if it is satisfied that the J64 findings in question were either perverse or made in the absence of any relevant evidence or a misapprehension of facts. 150. The submission on the point was that the Tribunal made it clear it had considered all the evidence presented by the parties. It was in that regard, particularly, noted that there was a huge amount of control exercised by Nestle Zimbabwe over the NZL in relation to the strategic direction of the company and its operations. There was testimony from one witness, AWl, who knew more about the operations and business of NZL than any local employee. More importantly, the Tribunal considered all the services that NZL receives from its related entities such as sales support, marketing support, procurement and supply chain management, amongst others. 151. On that evidence, the Tribunal found that there is very little that NZL did on its own without the direction, assistance and control from its related entity. Counsel submitted that NZL has misconstrued what constitutes a low risk distributor when it cited the authors of Advances in International Accounting Volume 18, who state that a low risk distributor is an entity that only undertakes operational and tactical sales and marketing J65 activities whose inventory risks are managed by its principal, who in turn owns the goods distributed by the agent. ZRA placed a caveat on this authority relied upon by NZL in that this definition is more for accounting purposes than those of transfer pnc1ng. 152. Reference was made to the 2017 OECD Guidelines which provide that, full-fledged distributors are enterprises with a relatively higher level of functions. Conversely, limited risk distributors, marketers, sales agents or commissioners are enterprises with relatively lower levels of function and risk. Further, that the Guidelines also use the term relatively, in terms of the level of functions and risks undertaken. The submission was that, the Tribunal was correct in finding that NZL carries out very limited functions and does very little on its own without the direction, assistance and control from its related entities. 153. ZRA also addressed the contention of its alleged failure to question NZL's financial statements appearing at pages 345 to 384 of the record. The submission was that, the financial statements showing that NZL was paying huge sums of money for the services supplied by its related entities, is the strongest J66 evidence that it is a low risk distributor. That it is the analysis of the functions performed by an entity that matters and not whether it pays for those services. 154. To illustrate the point, ZRA referred to the financial statements showing that NZL's initial investment was K2, 317, 500 and yet in the same year, it paid out K4, 503, 539 in management fees and royalties. A payment out that was twice what it invested. Other examples given were of large sums of payments made in respect of management fees and royalties totalling K46, 540, 594. 155. It was submitted that a careful analysis of the financial statements on the record reveals that the losses made by NZL were artificial, deliberate and premeditated to avoid paying tax. We were urged to disallow the said payments. Section 95 of the Income Tax Act, was cited as authority for the submission. 156. We were also invited to consider the approaches in Pacific Management Group, BSC Leasing Inc., Tax Matters Partner, et Al Petitioners v Commissioner of Internal Revenue August, 2018; Frank Lyon Co. v United States, 435 U. S. 561 (1978); Cooper v Commission, 877 F. 3D 1086, 1091 (9th CIR. 201 7) in which the transactions under consideration were said J67 to be a device to extract profits in the guise of tax deductible payments or the creation of tax benefits. That in assessing the reasonableness of the 'management fees' the question is whether those sums represented reasonable compensation for the personal services rendered, and that whether compensation for personal services is reasonable is a question of fact and circumstances. The US Tax Court is said to have found that the payment of the management fees were not reasonable and were unnecessary. It was submitted that the test used in deciding whether a particular payment is reasonable and has economic substance is an objective one. 157. The case of Metro Leasing & Dev. Corp v Commissioner, 376 F 3D 1015, 1019 (9th CIR. 2004) AFF'G 119 T. C. 8 (2002), was referred to where the Circuit Court considered whether the amounts paid would be regarded by a hypothetical independent investor. Further reference was made to the cases of United States v Basye, 410 U. S. 441, 450 (1973) and Lucas v Earl, 281. U. S. 111, 115 (1930) in which the longstanding principle that he who earns income may not avoid taxation by entering contractual arrangements diverting such income, was J68 enunciated. 158. It was submitted that the payment for management fees and royalties by NZL kept on increasing exponentially, and that this sort of behaviour of deliberate stripping or diversion of corporate profits to another entity was frowned upon by the United States Tax Courts. We were urged to adopt the same stance and confirm the Commissioner Generals' powers under section 95 of the Income Tax Act to disallow what were considered unreasonable and unnecessary expenses with no economic benefit to NZL, apart from eroding its corporate profits. That the total payment relating to management royalty fees for the five years being, K46, 540, 594 which was under tax audit, should not be allowed to be paid out by a hypothetical investor. 159. Counsel concluded his submissions in opposition on ground one, by urging that as NZL had failed to show that the finding of the Tribunal that it was a low risk distributor was perverse or made in the absence of any relevant evidence, the ground lacked merit and ought to be dismissed. 160. On the second ground of the cross-appeal by which NZL is aggrieved by the Tribunal decision to order reassessment within J69 180 days to be undertaken by ZRA, on account of the possibility that there could be more transfer pricing issues requiring further scrutiny. ZRA argued that the issue of further scrutiny actually arose from some of NZL's own witnesses, particularly the expert witness, who during trial testified that where an entity is incurring losses, this may be indicative to the tax authority that there may be an issue of the pricing of goods and that special scrutiny or an audit is required of the related party transactions . 161. ZRA submitted that section 9(1) (b) of the Tax Appeals Tribunal No. 1 of 2015 gives power to the Tribunal to take any other course which may lead to the just, speedy and inexpensive settlement of any matter before it. In any event, that ZRA has powers under section 63 of the Income Tax Act to reassess taxpayers if it so wishes . Respondent's Oral submissions on the cross-appeal 162 . In augmentation made at the hearing, Counsel for NZL 1n addressing the grievance subject of ground one of the appeal, that NZL was a low risk distributor, rehashed the arguments as formulated in the heads of argument. He underscored the fact J70 that one of the cardinal tests to consider when categorizing an entity, is who bears the inventory risk? The submission was that, evidence that the NZL bore the inventory risk and insured the goods purchased from Nestle South Africa and Nestle Brazil was unchallenged. 163. In his response Counsel for ZRA submitted that before you can even characterize an entity as a low risk distributor, a functional analysis must be done, as explained earlier, this involves scrutinising the functions performed by the entity, the assets employed by the entity in its business and the risk borne by an entity. One could not simply rely on inventory risk as NZL was proposing, as that was only one facet of the characterization. Counsel went on to elaborate that the reason NZL was classified as a low risk distributor was that the functions that it ought to have performed were actually outsourced and done by its related entities or associates. He argued that, this was proved by the financials submitted by NZL which showed and proved that it was paying for these functions and did not perform them itself. 164. The other argument was that NZL did not manufacture any of the products but simply distributed them. It was similarly, not J71 involved in branding, manufacturing or any value addition to the products distributed. Neither, did it have any assets such as heavy equipment other than, forklifts and vehicles for distribution of the goods. Even the warehouse that NZL used to store its goods was, apparently, not owned by it but was rented from a third party. 165. Counsel questioned how NZL could have a loan of U$ 8 million and not have any assets that could be used as collateral to secure the loan as required by usual business practice. It was submitted that in ZRA's view, the business risk lay with different parties and not NZL, particularly strategic risk, marketing risk, legal risk and IT risk, which all lay with the other related entities. Even the risk of purchasing the product to be distributed was said not to lie with NZL, as Nestle East Africa purchased products on its behalf. All these factors were said to have contributed to NZL being categorized as a low risk distributor by ZRA and ultimately, by the Tribunal. 166. Counsel for NZL in his reply defined a low risk distributor as an agent of a principal who owns the goods distributed by the agent and bears all the attendant risks related to the distribution of his J72 goods by the agent. He submitted that this was not the relationship NZL had with its related entities as it was an independent entity. As already demonstrated in its heads of argument, it was stated that NZL had its own board of directors responsible for strategic direction, budget and vision of the company, assertions not rebutted by ZRA. 167. Counsel reiterated that although NZL received the var10us services mentioned by the Counsel for ZRA from its related parties, it paid for the services. Further, that NZL bears its own inventory risk as evidenced by its financial records appearing on pages 345 to 384 of the record . Counsel went on to submit that there was no evidence before the Tribunal that NZL was controlled by any affiliated company of the Nestle Limited group. The record of proceedings was also clear, that no such evidence was adduced by ZRA. 168. According Counsel for NZL the issue was, whether or not, it was correct for ZRA to categorise NZL as a low risk distributor in its functional analysis . Whether, it was also correct, on the strength of the evidence before the Tribunal to deem NZL as an agent of any Nestle affiliated entity. Counsel rested his submissions by ]73 urging that ZRA had wrongfully categorized NZL a low risk distributor instead of a fully-fledged distributor that was independent, paid for its services and assumed inventory risk. Counsel for NZL prayed that the cross-appeal be upheld with costs. Consideration and determination of Cross-appeal 169. We have read the Heads of arguments and submissions on the cross- appeal by Counsel for the parties. We have similarly, considered the oral augmentation from Counsel. From the record, it is clear that informed by the evidence before it, the Tribunal made a finding of fact that NZL was a low risk distributor. 170. A number of our previous decisions were cited where we have reiterated that, an appellate Court will not reverse findings of fact made following the hearing of a matter unless, it is satisfied that the findings in question were either perverse or made in the absence of relevant evidence or misapprehension of the facts or that on a proper view of the evidence, they were not findings that a reasonable tribunal acting properly, could have made. J74 1 71 . The record of appeal at pages 39 and 40, reveals that the Tribunal clearly applied its mind to the evidence before it and also to the very arguments that NZL has rehashed before this Court. As recounted earlier in this judgment, the level of reliance placed by NZL on its related entities for the services, including sales support, finance, accounting and so forth, was not an issue in dispute. 172. As was properly reasoned by the Tribunal, it is indeed odd that a fully-fledged and apparently, independent entity such as NZL does very little administration on its own, without interference from its related parties such as Nestle Zimbabwe and Nestle South Africa. The Tribunal was not oblivious to the fact that NZL, bore some risk in its day to day operations. However, from its analysis of the evidence before it, the ultimate risk as stated at page 41 of the record, lies with Nestle South Africa and the risk to NZL was inevitably low. Even from a functional analysis perspective, when one considers the functions carried out by NZL, it is clear that its related parties bore the greater risk. 173. In that regard, we note that one of the excerpts quoted from a General Licence Agreement reads that, "as LICENSOR shall from J75 time to time be in a position to render or make available on an as needed basis to LICENSEE strictly for or in relation to its own or contract-manufactured manufacture, processing control, packaging, storage, supply, import, export, marketing, promotion, distribution and/ or selling of the Products which have been deployed by or for or are otherwise available to the LICENSOR." 174. On a holistic consideration of the facts, evidence, applicable laws and other. authorities we were referred to by Counsel for the parties on both sides, we find no basis to reverse the findings of fact made by the Tribunal as they are supported by the evidence on record. Those are the reasons that we find no merit in ground one of the cross-appeal, faulting the finding of the Tribunal that NZL was a low risk distributor. 175. Proceeding with ground two, in which NZL is aggrieved with the order for reassessment within 180 days by ZRA on the premise that there could be transfer pricing issues requiring further scrutiny. We have considered the pronouncement made by the Tribunal and agree with NZL that such reassessment goes beyond the scope of the initial assessment and would be beyond the scope of the issues escalated to this Court by the parties in J76 this appeal. Nonetheless, as alluded to in paragraph 60 and pursuant to section 63 of the Income Tax Act, ZRA is not precluded from initiating any reassessment. We for those reasons find ground two of the cross-appeal equally lacks merit. Conclusion 176. To recap, the respondent, Nestle Zambia Limited (NZL) has failed on its cross-appeal. The appellant Zambia Revenue Authority (ZRA) having substantially, succeeded on four out of its five grounds of appeal, is the successful party. Accordingly, we allow this appeal with costs to the appellant, to be taxed in default of agreement. Appeal allowed J. K. KABUKA SUPREME COURT JUDGE F. M. CHISANGA SUPREME COURT JUDGE J77