I. Issue One
The use of residual profit split models in the global trading of financial instruments (i .e. the reward for research, structuring, sales, trading and back office functions).
II. Issue Two
The determination of the reward for providing capital in financial services businesses, e.g. global trading, including the extent to which it is deemed to be provided from the jurisdiction in which the decisions are made that put it at risk.
III. Issue Three
The appropriate transfer pricing methods to be used for the allocation of investment management fees (i.e. for investment management, sub-advisory, marketing and administration).
IV. Issue Four
The appropriate transfer pricing methods to be used for the allocation of insurance premiums (i.e. between product development, sales, underwriting, captive reinsurance (especially), asset management and claims handlingladministration).
V. Issue Five
The attribution of capital (and its reward) to branches of banks and insurance companies.
VI. Issue Six
Transfer pricing issues arising from the restructuring of banks - for instance, where one subsidiary administers the loans of another subsidiary which is not permitted to make new loans, or where one subsidiary transfers complex financial assets to another subsidiary.
Switzerland
Mohamed Serokh and Karl-Heinz Winder, PricewaterhouseCoopers in Zurich, Switzerland
With regard to local legal or regulatory provisions, administrative practices, the result of tax audits or case law, what is the position in your jurisdiction on the following transfer pricing issues that arise in the financial services sector:
I. Issue One
The use of resiclual profit .split models in the global trading of financ al instruments (i.e. the reward for research, structuring, sales, trading arid back office functions).
Switzerland has not released any specific transfer pricing regulations. Inter-company transactions between related entities need to be determined according to third-party prices under the Federal Law on Direct Federal Tax ("DBG") and the Federal Law on the Harmonisation of the Cantonal and Communal Taxes ("STHG"). These legal principles define the basis for the calculation of taxable profit in Switzerland.
In addition, there are a number of administrative directives (including circulars and circular letters), which implicitly or explicitly refer to the determination of inter-company transfer prices. The circular letter of March 4, 1997 from the Federal Tax Administration regarding the "1995 OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (including 1996 Addendum)" encourages Cantonal Tax Administrations ("CTAs") (e.g., Zurich, Geneva, etc.) to apply the OECD Transfer Pricing Guidelines in the determination of transfer prices for multinational companies but does not specifically address the use of residual profit split models in the context of the global trading of financial instruments.
In general, the cost plus method is applicable for services provided to other related parties (according to circular letter No 4; 2004 Federal Tax Administration) but it cannot be inferred from this that it is applicable to the remuneration for global trading of financial instruments. Where Swiss based traders du not bear or actively (on a daily basis) manage market risks (or execute trades based on orders) or in case of brokering, a cost plus method may be considered.
In practice however, the residual profit split is chosen as a transfer pricing model in the global trading of financial instruments. Typically, a ruling is obtained with the relevant Swiss CTA in order to gain tax certainty for a certain period of time on the residual profit split Ausen and subsequent remuneration for each hinehol" determined.
Back officc functions arc mostly trcatcd as routinc functions and arc remuncratcd through the cost plus method during the inifial step of the residual profit split analysis. The trading function is, if regarded as valne adding, rewarded by a portion of the residual profit. Compensation of other functions, such as sales, research and structuring varies depending on the facts and eireunistanees of the undcrlying casc. Compcnsation of salcs can vary dcpcnding on the type of products traded (e.g., plain vanilla verses structured derivatives, etc.) as well as to what extent sales provides input for tailoring specific products for the bespoke needs of the client. Common approaches include remuneration through sales credits, which may be developed or corroborated through a cost-plus or TNIVIM based method.
Approaches applied to research can also differ. lt is not unusual to see the research functions remunerated through a cost plus method for routine or more lower value added research activities. However, research can participate in a residual profit split for very tailored or high value quantitative research say, for example, related tu high value inveslment advice related to bespoke investment products.
The compensation for capital in cases where a residual profit split model is implemented between separate legal entities needs to be considered, since the capital provider is ultimately bearing risk. The reward for capital is an important issue especially in the era of the global financial crisis where capital and liquidity is very scarce and arguably even more valuable in overall trading activities. Moreover, as legal entities in Switzerland can achieve relatively low effective tax rates compared to other European jurisdictions, the arm's length nature of the compensation for capital is brought into sharper focus.
The above approaches are commonly observed as being practices of financial institutions operating in Switzerland and it is our expectation that these should generally remain valid provided that they contirrue to appropriately take into account the contributions of all relevant parties. However, it is our experience that many institutions operating in Switzerland, when setting up global trading structures, apply for a tax ruling via the relevant Swiss CTA, which confirms the application of a residual profit split model for the underlying case and an appropriate (arm's length) return for the activities carried out in Switzerland. Unlike unilateral APA processes in other countries, obtaining a ruling with the relevant Swiss CTA is usually a relatively quick process provided that the relevant documentation supporting the arrangement is produced.
II. Issue Two
The determination of the reward for providing capital in financial services businesses, e.g. global trading, including tlze extent to vvhich it is deemed to be provided from the jurisdiction in which die decisions are mode that put it at risk.
In the absence of Swiss transfer pricing regulations and consistent Nvith the OECD approaches, where global trading takes place between branches of the same legal entitv or in the case that a dependent agent permanent establishment exists, any profit or loss derived from capital may be attributed to the jurisdiction in which the decisions are made that put the capital at risk. However, in the case of separate legal entity capital providers the provision of reward for the capital may be acceptable provided that it can be supported as being arm's length. A key additional consideration for many is to be sensitive to local financial regulatory requirements associated with capital.
In practice, the approaches used to determine the reward for capital vary to take into account the type of risk, the degree of risk borne by the capital provider and also consider the nature of trading activities (e.g. proprietary trading or mere execution of client orders). Each change in fact, can give rise to a potential different method of remuneration for capital.
In relation to market risk, the mere provision of the balance sheet for the booking of trades for a low volatile customer flow book may involve more limited risk and, therefore, capital may be allocated a basic fixed return. Likewise, if the risk bearing is only expected to be temporary, the use of claw back clauses to make the capital provider recoup historic losses when the trading book returns to profit have also been implemented by some institutions. However, if the capital provider may bear more frequent or significant loss(es) (particularly in the case of proprietary trading), other approaches are also employed to determine the compensation for capital provision. In such examples, the capital provider may actually participate in the profit split or receive a form of remuneration that is significantly higher than a routine return on capital.
Though very limited direct practice within Switzerland exists in relation to audits involving such approaches, it is reasonable to assume that the above approaches may be appropriate on the merits of each case, so long as the compensation for capital is adequately aligned with the level and amount of risk assumed by the capital provider and is in line with the behaviour that independent parties would undertake. Ordinarily, for more complex arrangements, financial institutions often obtain a ruling through the relevant Swiss CTA in order to gain tax certainty for a certain period of time.
III. Issue Three
The appropriate transfer pricing methods to be used fr the allocation of investment management fees (i.e. for investment management, sub-advisory, marketing and administration).
As there are no specific transfer pricing regulations in Switzerland, no specific guidance on the allocation of investment management fees exist. However, to the extent that the methodology applied to determine the split of the management fees is acceptable from an OECD Transfer Pricing Guidelines perspective, we would generally expect that it should also be acceptable in Switzerland.
In terms of Swiss practices, methods vary according to facts and specifically, the type of investment management. For example, in the case of traditional asset management relating to retail or institutional fund classes, CUP methodologies based on internal or external CUP data are commonly observed to evaluate either the sub advisory or marketing/sales/distribution functions.
In the case of internal CUPs, this is supported by internal third party arrangements or where these do not exist, external third party data can often be sourced. However, internal third party arrangements are preferred as they provide a more defendable basis of allocation. However, in applying CUP based approaches, there may be sensitivity and necessity to consider the type of fund (retail, institutional), asset dass (equity, fixed income, money market etc.), investment strategy (growth, income, balanced, etc.). Depending on the availability and type of benchmark data, other corroborating approaches have been adopted by some to verify the appropriateness of their transfer pricing.
In the case of alternative investments, such as hedge funds, comparable data for the sub advisory or investment decision making functions are seldom available, although the capital raising function (sales) can often be supported with internal CUP data via arrangements institutions have with third party distributors. In these cases, available comparable data is often used to derive fees payable to sales and distribution, with the residual payable to the other non-routine functions such as investment management. If non-routine investment management functions are located in more than one location, these are usually subject to a profit split.
Fund administration functions are almost always routine in nature, so cost plus or TNMM methods are commonly applied for these functions, but external CUP data can also be used to price these functions. For certain discreet administrative functions such as custody services, third party CUP based benchmarks may be available, although there can sometimes be challenges obtaining highly comparable data and care may also be required to account for differences in volumes, asset size and other key elements that may affect rates on assets under management.
In recent years, the Swiss CTAs, especially Zurich and Geneva have typically taken a highly aggressive examination of these types of arrangements, often arguing for a corroborative profit split. The above approaches may continue to be supportable provided that the allocation of fees and/or pricing of individual transactions is in line with arm's length requirements and remain appropriate to each individual set of circumstances. As for other transfer pricing set-ups in the financial services business, a ruling through the relevant Swiss CTA in order to gain tax certainty for a certain period of time is often a preferred approach.
IV. Issue Four
The appropriate transfer pricing methods to be used for the allocation of insurance premiums (i.e. between product development, sales, underwriting, captive reinsurance (especially), asset management and claims handling/ administration). The Swiss CTAs may reasonably be expected to accept transfer pricing methods that are in line with OECD Transfer Pricing Guidelines.
To the extent that insurance premiums are taxable in Switzerland through a branch, the OECD guidance with respect to the Part IV (Insurance) Report on the Attribution of Profits to Permanent Establishments is applicable. In this context, the allocation of insurance premiums and investment income to the Swiss branch is predicated on identifying the Key Entrepreneurial Risk Taking ("KERT") functions undertaken by the branch in order to firstly appropriate investment assets to the branch and thereafter investment income through an OECD based approach consistent with Part IV. Significant People Functions ("SPFs") outlined in Part I which are undertaken by the branch, potentially including underwriting and risk management functions are then usually allocated a portion of generated premiums through a form of residual profit split. How a branch tax analysis is precisely implemented strongly depends upon the type of insurance business being underwritten (i.e. property and casualty ("P&C") versus health), as well as the type of product (e.g. necessitating routine underwriting functions versus those requiring a high degree of speciali sm and judgement).
Where the insurance products subject to transfer pricing consideration do not involve branches, but instead fully regulated subsidiaries, the KERT analysis is not usually relevant and instead a main consideration is whether CUPs exist for sales and marketing functions or whether a profit split needs to be employed for locations undertaking underwriting, product development, sales and marketing and administration functions. The issue of reinsurance can pose additional challenges on pricing especially where excess of loss arrangements are involved, with some insurance groups verifying reinsurance pricing through the use of actuarial models.
Ceding commissions associated with reinsurance arrangements are also sometimes tested via TNMM to test whether a cedent is getting an adequate reward for upfront sales or related support. Often the TNMM in this regard may simply test whether the cedent, by ceding the business in the tested transaction, is financially worse off with no associated commercial benefit. If that were true then it is arguable that the ceding commission is too low relative to an arm's length amount. In other cases, some groups have tested through a TNMM the reinsurer's return on capital.
For captive reinsurance the choice of the transfer pricing method mainly depends on the extent to which the captive acts autonomously and creates better risk diversification for the cedants. In some cases it is important to establish appropriate support for the premiums charged by the captive through retrocession premiums it pays to third party reinsurers. The risk and loss experience of the covered members also needs to be supported. Common approaches include the use of actuarial models as wen as proof of the commercial rationale for the portion of policies ceded. lt is also important to prove that the captive has the experience and skills required to assume and control the covered risks.
Where an entity in a group acts as a captive but then further retrocedes risks externally, for those risks that are simply passed through the captive, it is important to ensure that the costs of the retrocession on these risks are merely passed onto the cedants with a minimal mark-up to compensate the captive for administering the external retrocessions. In cases where the cedent is able to achieve significant discounts through bulk external retrocessions and diversification, some groups have given significant thought as to how the captive should be remunerated for achieving the cost reduction. In cases where the captive assumes some of the risks and passes on a portion externally, then the above principles to compensate the captive for assumed risks need to be considered.
The Swiss taxing authorities do not appear to have a highly significant level of insurance specific experience in this highly technical area, but nevertheless, the documentation of such arrangements is strongly advised, especially in cases where the cedent is based in Switzerland and cedes business to a perceived tax haven. Swiss CTAs do regularly critique cost allocations, expenses and intra-group charges for insurance groups. Some examples include insurance technical support services, e.g. claims handling and administration, as well as head office services.
Where insurance groups have a charge recipient in Switzerland for such services, the key issue is adherence to the `benefits test' as enumerated in Chapter VII of the ODECD Guidelines. So, in audits by the Swiss CTAs, there is usually a strong expectation of how the `benefits test' has been adhered to, i.e. if a cost plus is being employed, are there good grounds as to whether a CUP method has actually been fully considered? Have the appropriate costs being recharged under a cost plus method been captured? For indirect costs, have the appropriate allocation keys been used? Are the appropriate direct costs being recharged? How has the appropriate mark-up been determined? Are there any duplicative services being performed by the Swiss charge recipient? Are any of the services being recharged, stewardship or shareholder services? These are the types of questions that the Swiss CTAs usually focus on.
V. lssue Five
The attribution of capital (and its reward) to branches ofbanks and insurance companies. There is no specific Swiss guidance on the attribution of capital to branches of banks and insurance companies in Switzerland. The OECD report has, to date, only had limited impact on these particular issues. Generally, PE profit attribution is required to be consistent with the operations of the branch as reflected in the accounts. The relevant Swiss CTA/Federal Tax Administration would generally seek to apply the principles in the OECD Guidelines proposed by the OECD Report on Attribution of Profits to Permanent Establishments (the OECD Report) except where they are incompatible with specific local provisions.
Vl. lssue Six
Transfer pricing issues arising from the restructuring of banks for instance, where one subsidiary administers the loans of another subsidiary which is not permitted to make new loans.
As no specific Swiss legislation exists, the approach should be in line with the principles outlined in the Chapter IX of the OECD Guidelines.
Where the business restructuring involves a transfer of a loan portfolio or trading book together with associated risks from one entity to the other, the market value of the loan portfolio or trading book needs to be established. Complications may arise in valuing non-performing loans, in particular where the viability and repayment capacity of the borrower is in doubt. In such cases, questions of the commercial rationale behind the transfer of the particular non-performing loan or loan portfolio cannot be ruled out where assets are transferred from overseas to Switzerland as a result of restructuring of overseas operations.
In line with the OECD Guidelines, transactions arising after restructuring should be appropriately considered. As such, transfer pricing methods need to be established taking into account changes in entity functional and risk profiles.
Mohamed Serokh is a Director in Financial Services Transfer Pricing services at PricewaterhouseCoopers in Zurich, Switzerland. Karl-Heinz Winder is a Senior Consultant in Financial Services Transfer Pricing services at PricewaterhouseCoopers in Zurich, Switzerland.
www.pwc.com
The use of residual profit split models in the global trading of financial instruments (i .e. the reward for research, structuring, sales, trading and back office functions).
II. Issue Two
The determination of the reward for providing capital in financial services businesses, e.g. global trading, including the extent to which it is deemed to be provided from the jurisdiction in which the decisions are made that put it at risk.
III. Issue Three
The appropriate transfer pricing methods to be used for the allocation of investment management fees (i.e. for investment management, sub-advisory, marketing and administration).
IV. Issue Four
The appropriate transfer pricing methods to be used for the allocation of insurance premiums (i.e. between product development, sales, underwriting, captive reinsurance (especially), asset management and claims handlingladministration).
V. Issue Five
The attribution of capital (and its reward) to branches of banks and insurance companies.
VI. Issue Six
Transfer pricing issues arising from the restructuring of banks - for instance, where one subsidiary administers the loans of another subsidiary which is not permitted to make new loans, or where one subsidiary transfers complex financial assets to another subsidiary.
Switzerland
Mohamed Serokh and Karl-Heinz Winder, PricewaterhouseCoopers in Zurich, Switzerland
With regard to local legal or regulatory provisions, administrative practices, the result of tax audits or case law, what is the position in your jurisdiction on the following transfer pricing issues that arise in the financial services sector:
I. Issue One
The use of resiclual profit .split models in the global trading of financ al instruments (i.e. the reward for research, structuring, sales, trading arid back office functions).
Switzerland has not released any specific transfer pricing regulations. Inter-company transactions between related entities need to be determined according to third-party prices under the Federal Law on Direct Federal Tax ("DBG") and the Federal Law on the Harmonisation of the Cantonal and Communal Taxes ("STHG"). These legal principles define the basis for the calculation of taxable profit in Switzerland.
In addition, there are a number of administrative directives (including circulars and circular letters), which implicitly or explicitly refer to the determination of inter-company transfer prices. The circular letter of March 4, 1997 from the Federal Tax Administration regarding the "1995 OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (including 1996 Addendum)" encourages Cantonal Tax Administrations ("CTAs") (e.g., Zurich, Geneva, etc.) to apply the OECD Transfer Pricing Guidelines in the determination of transfer prices for multinational companies but does not specifically address the use of residual profit split models in the context of the global trading of financial instruments.
In general, the cost plus method is applicable for services provided to other related parties (according to circular letter No 4; 2004 Federal Tax Administration) but it cannot be inferred from this that it is applicable to the remuneration for global trading of financial instruments. Where Swiss based traders du not bear or actively (on a daily basis) manage market risks (or execute trades based on orders) or in case of brokering, a cost plus method may be considered.
In practice however, the residual profit split is chosen as a transfer pricing model in the global trading of financial instruments. Typically, a ruling is obtained with the relevant Swiss CTA in order to gain tax certainty for a certain period of time on the residual profit split Ausen and subsequent remuneration for each hinehol" determined.
Back officc functions arc mostly trcatcd as routinc functions and arc remuncratcd through the cost plus method during the inifial step of the residual profit split analysis. The trading function is, if regarded as valne adding, rewarded by a portion of the residual profit. Compensation of other functions, such as sales, research and structuring varies depending on the facts and eireunistanees of the undcrlying casc. Compcnsation of salcs can vary dcpcnding on the type of products traded (e.g., plain vanilla verses structured derivatives, etc.) as well as to what extent sales provides input for tailoring specific products for the bespoke needs of the client. Common approaches include remuneration through sales credits, which may be developed or corroborated through a cost-plus or TNIVIM based method.
Approaches applied to research can also differ. lt is not unusual to see the research functions remunerated through a cost plus method for routine or more lower value added research activities. However, research can participate in a residual profit split for very tailored or high value quantitative research say, for example, related tu high value inveslment advice related to bespoke investment products.
The compensation for capital in cases where a residual profit split model is implemented between separate legal entities needs to be considered, since the capital provider is ultimately bearing risk. The reward for capital is an important issue especially in the era of the global financial crisis where capital and liquidity is very scarce and arguably even more valuable in overall trading activities. Moreover, as legal entities in Switzerland can achieve relatively low effective tax rates compared to other European jurisdictions, the arm's length nature of the compensation for capital is brought into sharper focus.
The above approaches are commonly observed as being practices of financial institutions operating in Switzerland and it is our expectation that these should generally remain valid provided that they contirrue to appropriately take into account the contributions of all relevant parties. However, it is our experience that many institutions operating in Switzerland, when setting up global trading structures, apply for a tax ruling via the relevant Swiss CTA, which confirms the application of a residual profit split model for the underlying case and an appropriate (arm's length) return for the activities carried out in Switzerland. Unlike unilateral APA processes in other countries, obtaining a ruling with the relevant Swiss CTA is usually a relatively quick process provided that the relevant documentation supporting the arrangement is produced.
II. Issue Two
The determination of the reward for providing capital in financial services businesses, e.g. global trading, including tlze extent to vvhich it is deemed to be provided from the jurisdiction in which die decisions are mode that put it at risk.
In the absence of Swiss transfer pricing regulations and consistent Nvith the OECD approaches, where global trading takes place between branches of the same legal entitv or in the case that a dependent agent permanent establishment exists, any profit or loss derived from capital may be attributed to the jurisdiction in which the decisions are made that put the capital at risk. However, in the case of separate legal entity capital providers the provision of reward for the capital may be acceptable provided that it can be supported as being arm's length. A key additional consideration for many is to be sensitive to local financial regulatory requirements associated with capital.
In practice, the approaches used to determine the reward for capital vary to take into account the type of risk, the degree of risk borne by the capital provider and also consider the nature of trading activities (e.g. proprietary trading or mere execution of client orders). Each change in fact, can give rise to a potential different method of remuneration for capital.
In relation to market risk, the mere provision of the balance sheet for the booking of trades for a low volatile customer flow book may involve more limited risk and, therefore, capital may be allocated a basic fixed return. Likewise, if the risk bearing is only expected to be temporary, the use of claw back clauses to make the capital provider recoup historic losses when the trading book returns to profit have also been implemented by some institutions. However, if the capital provider may bear more frequent or significant loss(es) (particularly in the case of proprietary trading), other approaches are also employed to determine the compensation for capital provision. In such examples, the capital provider may actually participate in the profit split or receive a form of remuneration that is significantly higher than a routine return on capital.
Though very limited direct practice within Switzerland exists in relation to audits involving such approaches, it is reasonable to assume that the above approaches may be appropriate on the merits of each case, so long as the compensation for capital is adequately aligned with the level and amount of risk assumed by the capital provider and is in line with the behaviour that independent parties would undertake. Ordinarily, for more complex arrangements, financial institutions often obtain a ruling through the relevant Swiss CTA in order to gain tax certainty for a certain period of time.
III. Issue Three
The appropriate transfer pricing methods to be used fr the allocation of investment management fees (i.e. for investment management, sub-advisory, marketing and administration).
As there are no specific transfer pricing regulations in Switzerland, no specific guidance on the allocation of investment management fees exist. However, to the extent that the methodology applied to determine the split of the management fees is acceptable from an OECD Transfer Pricing Guidelines perspective, we would generally expect that it should also be acceptable in Switzerland.
In terms of Swiss practices, methods vary according to facts and specifically, the type of investment management. For example, in the case of traditional asset management relating to retail or institutional fund classes, CUP methodologies based on internal or external CUP data are commonly observed to evaluate either the sub advisory or marketing/sales/distribution functions.
In the case of internal CUPs, this is supported by internal third party arrangements or where these do not exist, external third party data can often be sourced. However, internal third party arrangements are preferred as they provide a more defendable basis of allocation. However, in applying CUP based approaches, there may be sensitivity and necessity to consider the type of fund (retail, institutional), asset dass (equity, fixed income, money market etc.), investment strategy (growth, income, balanced, etc.). Depending on the availability and type of benchmark data, other corroborating approaches have been adopted by some to verify the appropriateness of their transfer pricing.
In the case of alternative investments, such as hedge funds, comparable data for the sub advisory or investment decision making functions are seldom available, although the capital raising function (sales) can often be supported with internal CUP data via arrangements institutions have with third party distributors. In these cases, available comparable data is often used to derive fees payable to sales and distribution, with the residual payable to the other non-routine functions such as investment management. If non-routine investment management functions are located in more than one location, these are usually subject to a profit split.
Fund administration functions are almost always routine in nature, so cost plus or TNMM methods are commonly applied for these functions, but external CUP data can also be used to price these functions. For certain discreet administrative functions such as custody services, third party CUP based benchmarks may be available, although there can sometimes be challenges obtaining highly comparable data and care may also be required to account for differences in volumes, asset size and other key elements that may affect rates on assets under management.
In recent years, the Swiss CTAs, especially Zurich and Geneva have typically taken a highly aggressive examination of these types of arrangements, often arguing for a corroborative profit split. The above approaches may continue to be supportable provided that the allocation of fees and/or pricing of individual transactions is in line with arm's length requirements and remain appropriate to each individual set of circumstances. As for other transfer pricing set-ups in the financial services business, a ruling through the relevant Swiss CTA in order to gain tax certainty for a certain period of time is often a preferred approach.
IV. Issue Four
The appropriate transfer pricing methods to be used for the allocation of insurance premiums (i.e. between product development, sales, underwriting, captive reinsurance (especially), asset management and claims handling/ administration). The Swiss CTAs may reasonably be expected to accept transfer pricing methods that are in line with OECD Transfer Pricing Guidelines.
To the extent that insurance premiums are taxable in Switzerland through a branch, the OECD guidance with respect to the Part IV (Insurance) Report on the Attribution of Profits to Permanent Establishments is applicable. In this context, the allocation of insurance premiums and investment income to the Swiss branch is predicated on identifying the Key Entrepreneurial Risk Taking ("KERT") functions undertaken by the branch in order to firstly appropriate investment assets to the branch and thereafter investment income through an OECD based approach consistent with Part IV. Significant People Functions ("SPFs") outlined in Part I which are undertaken by the branch, potentially including underwriting and risk management functions are then usually allocated a portion of generated premiums through a form of residual profit split. How a branch tax analysis is precisely implemented strongly depends upon the type of insurance business being underwritten (i.e. property and casualty ("P&C") versus health), as well as the type of product (e.g. necessitating routine underwriting functions versus those requiring a high degree of speciali sm and judgement).
Where the insurance products subject to transfer pricing consideration do not involve branches, but instead fully regulated subsidiaries, the KERT analysis is not usually relevant and instead a main consideration is whether CUPs exist for sales and marketing functions or whether a profit split needs to be employed for locations undertaking underwriting, product development, sales and marketing and administration functions. The issue of reinsurance can pose additional challenges on pricing especially where excess of loss arrangements are involved, with some insurance groups verifying reinsurance pricing through the use of actuarial models.
Ceding commissions associated with reinsurance arrangements are also sometimes tested via TNMM to test whether a cedent is getting an adequate reward for upfront sales or related support. Often the TNMM in this regard may simply test whether the cedent, by ceding the business in the tested transaction, is financially worse off with no associated commercial benefit. If that were true then it is arguable that the ceding commission is too low relative to an arm's length amount. In other cases, some groups have tested through a TNMM the reinsurer's return on capital.
For captive reinsurance the choice of the transfer pricing method mainly depends on the extent to which the captive acts autonomously and creates better risk diversification for the cedants. In some cases it is important to establish appropriate support for the premiums charged by the captive through retrocession premiums it pays to third party reinsurers. The risk and loss experience of the covered members also needs to be supported. Common approaches include the use of actuarial models as wen as proof of the commercial rationale for the portion of policies ceded. lt is also important to prove that the captive has the experience and skills required to assume and control the covered risks.
Where an entity in a group acts as a captive but then further retrocedes risks externally, for those risks that are simply passed through the captive, it is important to ensure that the costs of the retrocession on these risks are merely passed onto the cedants with a minimal mark-up to compensate the captive for administering the external retrocessions. In cases where the cedent is able to achieve significant discounts through bulk external retrocessions and diversification, some groups have given significant thought as to how the captive should be remunerated for achieving the cost reduction. In cases where the captive assumes some of the risks and passes on a portion externally, then the above principles to compensate the captive for assumed risks need to be considered.
The Swiss taxing authorities do not appear to have a highly significant level of insurance specific experience in this highly technical area, but nevertheless, the documentation of such arrangements is strongly advised, especially in cases where the cedent is based in Switzerland and cedes business to a perceived tax haven. Swiss CTAs do regularly critique cost allocations, expenses and intra-group charges for insurance groups. Some examples include insurance technical support services, e.g. claims handling and administration, as well as head office services.
Where insurance groups have a charge recipient in Switzerland for such services, the key issue is adherence to the `benefits test' as enumerated in Chapter VII of the ODECD Guidelines. So, in audits by the Swiss CTAs, there is usually a strong expectation of how the `benefits test' has been adhered to, i.e. if a cost plus is being employed, are there good grounds as to whether a CUP method has actually been fully considered? Have the appropriate costs being recharged under a cost plus method been captured? For indirect costs, have the appropriate allocation keys been used? Are the appropriate direct costs being recharged? How has the appropriate mark-up been determined? Are there any duplicative services being performed by the Swiss charge recipient? Are any of the services being recharged, stewardship or shareholder services? These are the types of questions that the Swiss CTAs usually focus on.
V. lssue Five
The attribution of capital (and its reward) to branches ofbanks and insurance companies. There is no specific Swiss guidance on the attribution of capital to branches of banks and insurance companies in Switzerland. The OECD report has, to date, only had limited impact on these particular issues. Generally, PE profit attribution is required to be consistent with the operations of the branch as reflected in the accounts. The relevant Swiss CTA/Federal Tax Administration would generally seek to apply the principles in the OECD Guidelines proposed by the OECD Report on Attribution of Profits to Permanent Establishments (the OECD Report) except where they are incompatible with specific local provisions.
Vl. lssue Six
Transfer pricing issues arising from the restructuring of banks for instance, where one subsidiary administers the loans of another subsidiary which is not permitted to make new loans.
As no specific Swiss legislation exists, the approach should be in line with the principles outlined in the Chapter IX of the OECD Guidelines.
Where the business restructuring involves a transfer of a loan portfolio or trading book together with associated risks from one entity to the other, the market value of the loan portfolio or trading book needs to be established. Complications may arise in valuing non-performing loans, in particular where the viability and repayment capacity of the borrower is in doubt. In such cases, questions of the commercial rationale behind the transfer of the particular non-performing loan or loan portfolio cannot be ruled out where assets are transferred from overseas to Switzerland as a result of restructuring of overseas operations.
In line with the OECD Guidelines, transactions arising after restructuring should be appropriately considered. As such, transfer pricing methods need to be established taking into account changes in entity functional and risk profiles.
Mohamed Serokh is a Director in Financial Services Transfer Pricing services at PricewaterhouseCoopers in Zurich, Switzerland. Karl-Heinz Winder is a Senior Consultant in Financial Services Transfer Pricing services at PricewaterhouseCoopers in Zurich, Switzerland.
www.pwc.com