The use of profit splits in commodity trading and banks – a natural alignment?

For many years, the banking sector was the pioneer for using profit split methodologies (PSMs), especially in the global trading context. Outside of this corner of the transfer pricing (TP) universe, PSMs were less common.

Following the OECD Base Erosion and Profit Shifting (‘BEPS’) project the landscape has changed, and profit split is now more of a consideration for many businesses across a range of industries when considering their approach to TP methodologies (TPMs) and, importantly, their discussions with global tax authorities.

This is particularly the case for commodity trading businesses given their similarity to the trading functions in banks which have long adopted PSM as a primary TP method.

This article considers some of the similarities and differences that exist between the industries and how they should be reflected in the respective TPMs.

As discussed, commodity trading desks within mining and resources companies have not traditionally utilised PSM within their TP models. There has been more of a focus on using one of the following two approaches:

• Traditional one-sided methods (cost plus or transactional net margin method (TNMM)) to reward the activities of the ‘traders’ or with an additional ‘commission’ element to reflect the ‘non-routine’ element of their activity; or

• Comparable uncontrolled price/transaction (CUP/T) methodologies – use of the ‘hedge fund model’ (HFM) is widespread across many jurisdictions.

Traditional one-sided methods often elicit little controversy and hence this article focuses on the CUP approach, particularly the HFM where, increasingly, tax authorities are looking to the revised guidance on profit splits which formed part of the revised OECD 2022 Transfer Pricing Guidelines to support the adoption of a PSM.

Hedge funds are arrangements formed between fund managers and investors, usually with the fund managers investing their own capital so that their interests and those of the investor are aligned.

The investment manager makes daily investment decisions for the fund, choosing where and when to allocate capital and managing the portfolio risk. Typically, these funds will operate at the higher risk end of the investment management spectrum, with the goal to achieve ‘alpha’ returns above market (beta) performance, as well as diversification from traditional asset classes such as shares, bonds and property.

The investors are charged a management fee – typically 1-2% of the assets managed and a performance fee of around 15-20% of fund profits, which is only collectable when the fund is profitable. Fee arrangements vary and have changed over time as competition from low-cost fund managers have challenged hedge funds’ market position.

The comparisons with a commodity trading desk are clear at a high level. The traders will typically be in a separate, but connected, entity in the trading location and will make day-to-day investment decisions but will be acting within the constraints imposed by the parent company in terms of capital commitments, risk exposures and internal hedging, similar to how the fund manager is restricted by the rules governing the fund.

Many industry practitioners felt that the similarities between the two models mean that the HFM met the criteria to be a CUP/T. The 2022 OECD Transfer Pricing Guidelines describe the CUP method as follows:

• “The CUP method compares the price charged for property or services transferred in a comparable uncontrolled transaction in comparable circumstance”. (para 2.14)

• “Where it is possible to locate comparable uncontrolled transactions, the CUP method is the most direct and reliable way to apply the arm’s length principle. Consequently, in such cases the CUP method is preferable over all other methods.” (para 2.15)

Using this approach results in an approximate split of profits of 20% to the trading location and 80% to the head office location (after the ‘base’ management fee), with the head office location bearing any losses in the same way as the investor would in the hedge fund context.

For many years many tax authorities accepted this approach and HFMs were the subject of many concluded negotiations in mutual agreement procedure (MAP) cases and advance pricing agreements (APAs) across the globe.

Some tax authorities have raised concerns over the use of the HFM in commodity trading context have sought to replace it with, or at least confirm it with, a profit split model based on a function, assets and risks (FAR) analysis in audits, MAPs and APAs. The result is usually a modification of the HFM as the primary method to reflect an uplift in profit share to the trading location.

Following the BEPS project and subsequent revisions to the Transfer Pricing Guidelines, with its increased emphasis on function, some tax authorities have started to put increasing pressure on using HFMs in the commodity trading industry, citing the revised guidance on the use of profit splits as the ‘most appropriate’ TPM.

They cite para 2.126 as the rationale: “The unique and valuable contributions by each party to the controlled transaction is perhaps the clearest indicator that a transaction profit split may be appropriate…other indicators that the transactional profit split may be the most appropriate method could include a high level of integration in the business operations to which the transactions relate and/or the shared assumption of economically significant risks (or the separate assumption of closely related economically significant risks) by parties to the transaction.”

The connection is clear - commodity trading businesses need to manage portfolios, capital and risks holistically across multiple jurisdictions, leading to a high level of integration. The trading function is (usually) a sophisticated one that requires fast paced decision making and therefore, to a certain extent, traders must have the ability to assume and manage the risk related to individual trades. There are often several layers of risk management, especially where the business is more than just a hedged ‘flow’ commodity trader and is holding speculative positions, perhaps as part of a wider business. In that regard, the investor/hedge fund manager analogy is somewhat different to the actual role of the commodity traders.

However, the attraction to the tax authorities is clear. It allows for some necessarily subjective inputs into the modelling to measure the value of the activity of the often-extensive range of functions involved in the business (again, especially where there is an underlying physical trading business).

However, tax authorities don’t often go very far in explaining why the use of the HFM CUP is inappropriate other than the “it isn’t a hedge fund” argument.

Given that the CUP, where appropriateness is established, is the method of choice as per the OECD TPG, a more detailed push-back may be appropriate, especially given the example at paragraph 1.70 (TPG 2022):

“Assume that an investor hires a fund manager to invest funds on its account. Depending on the agreement between the investor and the fund manager, the latter may be given the authority to make portfolio investments on behalf of the investor on a day-to-day basis in a way that reflects the risk preferences of the investor, although the risk of loss in value of the investment would be borne by the investor. In such an example, the investor is controlling its risks through four relevant decisions: the decision about its risk preference and therefore about the required diversification of the risks attached to the different investments that are part of the portfolio, the decision to hire (or terminate the contract with) that particular fund manager, the decision of the extent of the authority it gives to the fund manager and objectives it assigns to the latter, and the decision of the amount of the investment that it asks the fund manager to manage.

“Moreover, the fund manager would generally be required to report back to the investor on a regular basis as the investor would want to assess the outcome of the fund manager’s activities. In such a case, the fund manager is providing a service and managing his business risk from his own perspective (e.g., to protect his credibility). The fund manager’s operational risk, including the possibility of losing a client, is distinct from his client’s investment risk. This illustrates the fact that an investor which gives another person the authority to perform risk mitigation activities such as those performed by the fund manager do not necessarily transfer control of the investment risk to the person making these day-to-day decisions.”

This is useful as it outlines why the HFM works in the way that it does, and it is consistent with reward matching the active risk management functions. All the four ways in which the investor is seen to be managing risk in this example are typically present in the commodity trading set up with the head office location supplying the capital with limits and applying risk-based restrictions as to how that capital is utilised to the trading location. The monitoring of this is often daily (or even hourly) and regular interactions with risk teams and managers can often lead to trading positions being exited, reduced or increased. The head office location will also make key decisions relating to hiring and firing.

The other issue often raised by the tax authorities is the similarity between a typical commodity trading business and the trading desks observed within banking organisations which have commonly adopted PSMs to reward the trading desks. Superficially, there are clear comparisons but there are also differences around regulation and capital management etc. However, the value in this challenge pre-supposes that the approach taken by the banks is ‘more appropriate’ than that taken by the commodity traders in the use of the HFM.

This article now looks at the history of PSMs in the banking context to try and identify why they have evolved differently, and the challenges they are currently facing.

Commodity trading (or trading in what is known as ‘physicals’) in the investment banking industry grew over the 1990s and 2000s to be a significant part of the business of many investment banks. This commonly involved proprietary trading (where banks traded positions using their own money rather than client’s funds). However, the Volcker rule (part of the US Dodd-Frank Act) which was brought in after the 2008 global financial crisis and which prohibited proprietary trading, significantly changed banks’ participation in these markets. As a result, trading on a proprietary basis with significant capital placed at risk is much less common in the financial services industry than it used to be. This is potentially a key departure point when considering differences with less regulated commodity trading businesses outside of the investment banking industry.

In the banking and investment banking context, as noted above, profit splits are the default methodology to split profits where traders operate in a multi-jurisdiction basis to trade on the same book. Where the traders are in the same jurisdiction as a capitalised entity where the trades are booked, a profit split is usually not necessary at the book level. This is assuming the other contributions such as marketing and middle/back offices can be categorised as ‘routine’ and rewarded by a transactional method such as cost plus, or a share of the initial “Day 1” profit and loss. However, where there is integration, such as multiple traders trading the same book using a “follow the sun” approach, for example, a profit split can be an appropriate methodology to apply.

Tax authorities for the major financial hubs have commonly accepted (in APAs and audits) the profit split method as the best approach to split profits and losses arising from financial trades (e.g. FX, interest rates or commodity derivatives). In line with the OECD 2010 permanent establishment attribution guidance, traders are usually considered to undertake non-routine functions and their contributions are valued based on relative compensation, usually including bonuses. The main controversy has been on approaches to reward the capital provider, where these differ from the trading entity/jurisdiction e.g., where trades are ‘remote booked’. Over time, various methodologies have been accepted including a routine (e.g. benchmarked) return to capital, such as a benchmarked return on equity. Sometimes, capital is considered non-routine and as such is valued as a non-routine contributor to the profit and compared with the traders’ non-routine contribution. However, different tax authorities have different views on this topic.

Further issues which can cause differences of view include reward to the marketing function (i.e. the customer facing activity to introduce trades) and particularly whether that activity is so integrated to the trading function and decision making (e.g. in structuring complex trades) that it should be part of the trading profit split. Often, the marketing and origination function is accepted to be separate from the trading function, and therefore rewarded through a transactional TP method such as a sales credit or fee based on costs plus an arm’s length mark-up. This is particularly the case where the hedging of market risk or the central management of credit counterparty risk is undertaken by a different team.

The natural question that arises is: why is there an apparent difference in accepted TP methodology for certain types of trading activity between commodity traders outside of banks and investment banks, and trading activity within banks? As noted, this is an area where tax authority challenges are increasing, particularly in the resources industry.

Ultimately, the selection of the most appropriate TP methodology must have regard to the function, asset and risk profile of the related parties that are engaging in cross border activity, the wider commercial business and economic context, and availability/reliability of comparable data.

A key criterion to explain why different methodologies arise is the degree of proprietary risk taken by the trading function. Essentially this relates to the role of the capital provider in the business. To the extent that the trading activity involves assumption of significant trading risk (i.e., unhedged positions), the entity housing the trades (the capital provider) is subject to significant upside and downside profit/loss risk. As noted, regulation in the banking industry, as well as internal approaches to risk assumption/management, has significantly limited or completely prohibited the assumption of proprietary trade risk.

Where the trader and the booking of trades are in the same legal entity and jurisdiction, then the profits and losses arising from those trades will fall in that location. Where traders are in multiple jurisdictions and separated from the booking entity/capital provider, it will be necessary to determine the appropriate reward to capital, and consider where on the spectrum of ‘routine’ vs ‘non-routine’ capital contribution should lie. To the extent that capital is considered routine, a benchmarked return on capital may be appropriately used. Where capital is a vital driver of profit or loss then an assessment of the relative contribution of capital and trader activity is required. Where they are considered joint non-routine contributions, then a relative valuation of capital and trader contribution can be applied, and profits/losses on the book apportioned between both. To the extent that capital is the prime contributor, such that the investor/fund manager analogy in the HFM is the best comparable to apply, then this can be the most appropriate TP methodology.

For commodity businesses, a further consideration is the role of the trading function in the broader business, and whether it should be considered as a supportive procurement/supply or sales function for the broader global mining or production business or a key profit hub in its own right.

Reliability of comparable data to benchmark rewards and/or split profits is another critical consideration. There is a wide variety of hedge funds in the market and leverage models (even where trading is undertaken on a wholly proprietary basis), resulting in several different risk and reward profiles across the relevant industries.

Finally, the management of other issues such as whether separation of traders from the booking entity creates permanent establishment risks of that entity in the trader location is another complexity that has arisen over the years, particularly in the banking industry.

Given the scale and profitability of large commodity and banking businesses in the global economy, particularly in the large trading hubs and commodity producer economies, and with the changes in the TP guidance arising from the BEPS project, it is only natural to see tax authority activity increase in this area. The inquiries into the appropriateness of the TP methodology will necessarily be initially factual to understand the role and contribution of each part of the enterprise. As highlighted in this article, the degree of risk taken by a trading business will be a critical fact to assess, and naturally businesses will vary in their approach to this, resulting in the application of a range of TP methodologies.

Where there is subjectivity and potential for multiple tax authorities to take differing views, the use of APAs can be an effective tool, as has been demonstrated in the banking industry where bilateral and multilateral APAs have been entered into across key jurisdictions. Even then, tax authorities’ views can fundamentally differ, and a spirit of cooperation and focus on achieving resolution and protection from double taxation is required to overcome such differences.

Given this evolving landscape, it is timely to review the appropriateness of the TP methodologies applied to trading activity in the commodity and financial industries, and consider risk mitigation strategies such as pro-active engagement with tax authorities, or enhanced TP documentation.