Fabrizio Hernández, Timothy McKenna and Ralph Meghames, NERA Economic Consulting
This is an extract from the 2021 edition of GAR’s The Middle Eastern and African Arbitration Review. The whole publication is available here.
The Middle East and Africa (MEA) region is particularly active in international arbitration. Among these cases, three main characteristics stand out. First, the majority of cases in the region are linked to large infrastructure-related disputes. Second, among recent cases, investor-state disputes appear to be more frequent in the region than commercial disputes. Third, the most frequent allegations for arbitration are expropriation and breach of contract, including breach of shareholders’ agreements. Despite some similarities in the issues and the sectors affected by the cases mentioned above, from a quantum perspective, each displays features that need to be assessed in the specific context in which they arise. In this context, a correct understanding of the role of country risk, foreign exchange (FX) rates and fiscal regimes is key to the proper calculation and discounting of cash flows.
- Recent cases in the MEA
- Country risk and its impact on valuation
- Treatment of FX rates regimes and capital controls
- Fiscal regimes, transfer pricing, and shareholder disputes
Recent cases in the region
The MEA region is particularly active in international arbitration. Out of the 2,498 parties involved in cases filed with the ICC in 2019, 18 per cent were from the Middle East and Africa. Countries such as the United Arab Emirates, Saudi Arabia, and Qatar are among the most frequent nationalities among parties, representing respectively 3.12 per cent, 2.24 per cent, and 1.32 per cent of the total number of parties in 2019 filings (the United States being the number one country with 7.85 per cent share).
Among these cases, three main characteristics stand out. First, according to data collected from GAR covering the past three years, the majority of cases in the region are linked to large infrastructure related disputes, in particular in the energy, mining, and telecom sectors, with an average award value that exceeds US$500 million. Table 1 below illustrates some of the recent cases involving the MEA region.
Table 1: MEA region recent international arbitration cases
GAR, NERA Review
Second, among recent cases, investor-state disputes appear to be more frequent in the region than commercial disputes (around 60 per cent of investor-state cases versus 40 per cent commercial cases). Investor-state disputes in the MENA (Middle East and North Africa) region have, in fact, increased from 5 per cent to 9 per cent of ICSID’s new caseload in the past years (while sub-Saharan new caseload decreased from 21 per cent to 12 per cent). Also within the set of ICSID cases, most are related to energy and infrastructures.
Third, the most frequent allegations for arbitration are expropriation and breach of contract, including breach of shareholders’ agreement. In investor-state disputes expropriation is generally related to project cancellations, particularly in projects related to the exploration and exploitation of natural resources. In Divine Inspiration Group v Democratic Republic of the Congo, the South African oil exploration company was awarded damages over the African state’s failure to honour two oil production-sharing agreements covering an area that contains 6 per cent of the country’s oil reserves. In commercial arbitration, similar issues arise in the context of the early termination of concession agreements. In Damietta International Port Company (DIPCO) v Damietta Port Authority, the dispute related to a long-term concession agreement awarded covering the development, building, and operation of a container facility on Egypt’s Mediterranean coast. The project reportedly encountered several problems, notably in relation to the political upheaval that followed the Arab Spring of 2011. The port authority ultimately terminated the concession agreement in 2015, which gave rise to the claim.
Breach of contract is also a recurring theme in the region. In DP World v Djibouti, the African state was found liable for breaching DP World’s exclusivity rights by pursuing the development of container port facilities with a rival Chinese operator. The dispute revolved around a concession agreement for a container terminal at the port of Doraleh on the Red Sea – a strategically important hub for regional trade. In Unión Fenosa Gas, SA v Arab Republic of Egypt, the tribunal has ordered Egypt to compensate a joint venture between Naturgy (Spain) and Eni (Italy) over the interruption of gas supplies to a LNG plant at the port of Damietta. In Privinvest Group v Greece, the country lost a claim in favour of the Middle Eastern-owned operator of one of the country’s largest commercial shipyards. The dispute started around Greece’s default under its contractual obligations for the construction of submarines and evolved into a number of claims relating to breaches of commitments. In Turkmengaz v National Iranian Gas Company (NIGC), the national company failed to pay for gas it had imported from Turkmenistan, in the context of a long-term supply agreement. The awards in these cases averaged more than US$1.4 billion.
Shareholder disputes are also observed in the region. The Ansbury Investment v Ocean and Oil Development Partners BVI and Whitmore Asset Management case involved shareholder loan repayment and transfer of shares disputes in the context of a joint venture holding stakes in the Nigerian oil and gas company Oando. In PT Ventures v Unitel, a subsidiary of Brazilian telecoms group Oi won a case against its shareholders in Angola’s largest mobile phone carrier Unitel. The dispute emerged after Oi had acquired a majority stake in PT Ventures (which owns 25 per cent of the shares in Unitel) as part of its merger with Portugal Telecom in 2014. The other Unitel shareholders were found liable for breaching the shareholders’ agreement. Shareholder dispute cases are sometimes tainted with corruption claims. In Vale v BSGR, which was related to the largest iron ore deposit in the world, the tribunal found that BSG Resources (a Guernsey mining company) made fraudulent misrepresentations on which Brazilian mining company Vale relied upon when entering into a joint venture to develop an iron-ore mining concession in eastern Guinea.
Challenges for the common valuation approaches
Despite some similarities in the issues and the sectors affected by the cases mentioned above, from a quantum perspective, each displays features that need to be assessed in the specific context in which they arise. The quantification of the damage usually requires the valuation of the companies or projects affected by the liable actions. The three main approaches commonly relied upon to value a business or an asset (the market approach, the income approach and the asset approach ) are often difficult to apply (see Table 2):
Table 2: Limiting factors for valuation approaches in Africa and Middle East
- The market approach is often impractical in countries with underdeveloped or illiquid financial markets. Data might simply not be available, and whenever available, might not be reliable.
- The asset approach could be deemed subjective whenever based on historical costs which depend on management decisions taken at a certain time, and that might not be optimal from the perspective of a rational investor at the time the damage was suffered. In addition, often historical costs are not helpful in estimating the replacement value of an asset or a business.
- The income approach, also known as the discounted cash flow (DCF) method, remains the most widely relied upon method, but it also faces challenges. It requires calculation of the cash flows associated with the project both in the actual scenario, namely after the liable action, and in the counterfactual scenario (ie, but for the liable action). Besides data availability and reliability of prior forecasts, the calculation of cash flows in both scenarios requires appropriate consideration of:
- the country’s uncertain business environment that may make it difficult to measure estimate future performance and expected cash flows; the associated risks are generally referred to as country risk and the results of most valuation exercises are sensitive to how it is treated;
- the fact that cash flows are obtained in a currency different than the one in which the valuation is performed and this requires converting future cash flows into the currency of the valuation, often in countries with multiple rates and capital controls; and
- the applicable fiscal regime, either because it is specific to the country or project in question, or due to the international tax implications on the value of intercompany transactions.
A correct understanding of the role of country risk, FX rates and fiscal regimes is key to the proper calculation and discounting of cash flows. The rest of this chapter is devoted to explaining some of the issues that must be taken into account when dealing with these factors in quantum exercises.
In the DCF approach, the discount rate represents the opportunity cost of the capital employed to finance a business or a project. Calculating an appropriate discount rate for use in DCF calculations requires a logical support for the specific risks affecting the project. Different risk factors stemming from the specific jurisdiction where the project takes place generally imply different risk profiles and activities deployed in jurisdictions with risk factors that can enlarge the variability of potential outcomes for the project (such as changing its future costs or delaying its development) command inherently higher discount rates, reducing - all other things equal - the present value of such projects.
Two key issues to the consideration of country risk relate to: how to consider it in DCF methods; and how to derive a numerical value to quantify the risk.
In relation to the approach to include country risk in the DCF method, many analysts generally add a country risk premium to the discount rate. Besides, analysts often also structure the cash flow projections by modelling scenarios that reflect materialisations of the country uncertainties. It may then appear that country risk is accounted for twice: in the discount rate and in the (undiscounted) cash flows.
The discount rate reflects the time value of money, namely the correction to the current value that investors would attach to a monetary value in the future. The larger the uncertainties that can make such future value vary, the less will be the current value for risk-averse investors, given the same expected future value. As a result, if the conditions in a country A make the expected future value of a project more variable than in country B, even if on average the expected value of the projects is the same, then investors will rationally view the project in country A as less valuable today.
The same country risk premium should, however, not apply to all types of projects in a particular country. Some countries have a more developed and stable regulation and legal environment in some sectors than in others. Often, activities in sectors of strategic importance for the country (such as agriculture and extraction of natural resources) may be subject to more frequent interference by public authorities and more subject to political risk.
For this reason, an accepted principle in DCF analysis is that cash flow projections should reflect the characteristics of the project at hand. They will therefore reflect features of the sector where the project operates, the contracting arrangements it has signed to sell its production, the labour contracts in place for its workforce; many of these features may be sector-specific or project-specific, particularly when the project is one of a kind. The cash flows the project delivers may then be more or less subject to change as compared to those in the average sector in the country.
As long as the cash flow projections reflect the project characteristics, they must consider all the likely scenarios for the project in the future, which reflect all the uncertainties related to the project, including those related to the country where it operates. Such scenarios would reflect the practical consequences on the specific project of the interaction between all risk factors, project-specific and country-specific. The scenarios are then weighted according to their probability of occurrence and the resulting average values reflect the expected stream of cash flows going forward.
While cash flow projections reflect all risks affecting the project, the inclusion of country risk in the discount rate must consider only the systematic (or non-diversifiable) part of country risk. Standard asset valuation models such as the capital asset pricing model (CAPM) attach a premium related only to the part of the risk that investors cannot avoid with diversification. The degree of diversification of the relevant investors must be analysed case by case as often global investors are well diversified and local investors are not, particularly in relation to investments in real assets and if restrictions to access capital markets exists.
Often country risk premiums are derived by comparing US dollar-denominated bond yields in the local country with those in reference countries (such as the United States or Germany) for the same maturities. Adding such a premium to the risk-free rate implicitly assumes that: all country risk is systematic, and the risks related to the local government defaulting on its bonds is a relevant risk for the project under valuation. These assumptions do not hold in many cases and not correcting for these factors may lead to using country risk premiums that are unrealistically high for the project in question.
When country risk premiums are derived from bond rates, the question arises as to whether further adjustments are necessary to account for the greater risk inherent in equities than in debt. Often such equity risk differentials are reflected in the determination of equity discount rates, for instance by using market risk premiums that are then multiplied by the relative volatilities (the betas) of specific samples of companies with similar risk profiles relative to the overall market. Attempts to use data from companies with comparable risk (such as multinational companies with a similar percentage of activity in countries with similar risk assessments) with equity-to bond country risk adjustments may lead to double counting country risk. In fact, for beta estimates to be reliable, local industry stock are often not very liquid or have a short history of public trading. These conditions are common in many emerging markets.
In sum, the assessment of country risk must be tailored to each particular environment. Cash flow projections need to include realistic assumption about all risks on a project if such risks exists at the time of valuation. This includes political risk, such as expropriation risk. However, it should avoid the temptation to excessively weigh bad outcomes to ‘illustrate’ how certain elements of country risk would impact the project if they were to materialise. Once expected cash flows reflect the market expectations at the valuation date, they need to be discounted with a rate that includes the systematic component of country risk and considers the effects of potentially illiquid local equity markets on the volatilities of companies used as benchmarks for risk purposes. This is because, in emerging markets, country risk is project-specific and not fully systematic, and it may not correlate well with the spread of government bonds of the country concerned.
Foreign exchange rates
In many cases project cash flows are earned in a currency that is different from the currency of the valuation. The question then arises as to whether it is preferable to:
- discount the cash flows using foreign discount rates and convert the resulting discounted value using the spot exchange rate; or
- convert the future cash flows into domestic currency values using expected future exchange rates and then discount using a domestic discount rate.
The two methods may seem equivalent and, in fact, they generally provide the same answer as this equivalence is the well-known interest rate parity result. The results might not, however, be exactly the same and there are many reasons why this may be, ranging from slight imperfections in market prices to theoretical issues.
However, the valuation issues we address below go beyond questions related to the accuracy of interest rate parity. Rather, we consider how the non-existence of foreign exchange markets or the existence of currency controls affects the valuation of cash flows. These are significant concerns and apply to valuation in many countries around the world, including in the MEA region.
In the MEA region, there are a variety of FX regimes, with pegged currencies, floating currencies, and a variety of arrangements that seek to approximate a fixed exchange rate (see Table 3).
Table 3: FX regimes in the MEA region
Moreover, capital controls (ie, restrictions on the ability to access FX) can lead to ‘black market’ exchange transactions, whereby participants exchange currency at rates other than the official rates. Capital controls can coexist with managed and stabilised currency regimes. For example, Iran has a black-market FX rate, an official FX rate, and another FX rate called the NIMA rate, which is the rate at which exporters are required to transact (when converting foreign currency into domestic currency). This is a complicated structure that has features of a floating exchange rate regime along with a controlled currency regime where conversion is restricted. Valuation of an asset in a home country like Iran would require considering not only these FX rules at present but how they are expected to evolve as well.
Differences in the FX regimes and in the rules specific to any asset may affect the best way to approach the valuation. To illustrate the challenges that arise, we refer to a numerical example below. An asset generates foreign currency (FC) cash flows and the aim is to derive its value in terms of the domestic currency (DC). Specifically, an asset generates a FC cash flow of FC 20 one year in the future. The FC risk-free rate is 2 per cent and the DC risk-free rate is 10 per cent (including country risk of the domestic country), and the spot FC/DC exchange rate is five (ie, one unit of FC is valued at five units of DC). For simplicity the cash flows are assumed to be risk-free.
Data on expected future FX rates are often not available and, even when they are, the reliability of such data may be questionable. While it depends on the specifics of why future FX rates are not available, a straightforward approach that can be appropriate would be to discount the FC cash flows of the asset using FC risk-free rates and then convert the result to DC units using the spot exchange rate. In the example, that results in FC 19.6, which is the result of dividing FC 20 by (1 + 2 per cent). When converted using the spot exchange rate, this is equal to DC 98.0. The domestic risk-free rate of 10 per cent does not enter this computation. This is because the cash flows are earned in the future in FC.
When valuing an asset in a country with currency controls like Iran, the rules that apply to the cash flows of the asset must be considered. As a stylised example, again consider the valuation of an asset with FC flows of FC 20. The black-market FX rate remains 5 DC = 1 FC. If the expectation is to be allowed to convert FC into DC at the black-market rate, then the valuation remains the same as in the example above, that is, the value of the asset is DC 98. However, if the FC cash flows must be converted at a less favourable rate, the valuation in DC would be lower.
Each valuation must assess the rules that are expected to apply to the asset in question. In addition, what is important are the rules that are expected to apply when the cash flows are realised. While current rules can serve as guidance, what is relevant is the rules that are expected to apply in the future for conversion of FC into DC. If in a year all FC cash flows are expected to be converted at a less advantageous rate, then such a fact should be incorporated into the valuation.
In the example, if FC earned by the asset in one year is forced to convert at the official rate and by then the official rate is such that each unit of FC gets 20 per cent less DC than the black-market rate, then an expectation of a discount to the value of the asset should be built into the valuation. This is because 20 per cent of the value is diverted due to the use of the official exchange rate. If covered interest parity holds for black market rates, the value of the asset is worth exactly 20 per cent less than the value without the forced use of the official exchange rate. In this case the forced use of the official exchange rate had a meaningful impact on the valuation.
Naturally there may not be easy answers for every situation. Market data on domestic assets that are expected to generate FC cash flows may provide some guidance. Analyst and market commentary may also help to understand what expectations are about conversion of FC into DC.
Under other FX regimes the valuation may require other approaches. Some countries operate with an exchange rate that is allowed to adjust only slowly. Other countries have a currency that is pegged to either the euro or the US dollar. What is important for valuation of future cash flows is the expectations for these regimes in the future, whether the peg will remain in place, or what FX regime will be adopted if the peg collapses. These are not necessarily easy questions to answer, but a proper valuation should consider them.
As an extreme example, since 2019 Lebanon has placed unofficial restrictions on the withdrawal of currency from the country. This makes it challenging to value an asset that generates cash flows in Lebanese pounds for a US investor. One approach could be to shift cash flows into future periods when the restrictions on transfers are expected to cease. Such an estimate is, however, difficult to make and might be highly subjective. In addition, several different currency exchange rates emerged in the country: the Lebanese pound remains officially pegged at 1,517 to US$1, banks allow cash withdrawal at a rate closer to 3,900, the black-market rate at the time of writing stands at 13,000, and is subject to fluctuation, and businesses randomly apply their own exchange rate, which depends on the payment means (cash, foreign or local credit cards, etc). In such circumstances, a careful examination of the factors that are relevant to the project cash flows and to the valuation is necessary to decide on the most appropriate approach.
Fiscal regimes and transfer pricing
The applicable tax regime is a third potential driver of project or business value in both the actual and ‘but for’ scenarios.
In many cases involving joint ventures between local (often state-owned) entities and international companies, damage claims refer to the difference between the actual project value, which was reduced by the liable actions, and the continuation value of the project for the claimant. In these cases, the ability to restructure the project by the local entity or by the state after expropriation or after the cancellation of the concession contracts with the international investor, may lead to substantial value being diverted to the state in the form of fiscal revenues. If the possibility of these alternative arrangements is not considered, the continuation value of the project may appear low, when in fact the distribution of profits between the local entity and the state implicit in the original project structuring does not need to apply after cancellation of the concession. Post-cancellation values would be affected by changes in the fiscal treatment of the project and treatment does not necessarily remain the same after cancellation of the contract with the original investor or partner.
Many projects in the energy and infrastructure sectors span several segments of the value chain. Because different activities in different segments may face different effective tax rates, different project structures may imply differences in value. When, for example, upstream production is sold to a downstream affiliate at a transfer price, the distribution of profits along the value chain is affected. Often such transfer pricing rules are designed to: abide with international tax regulation; guarantee sound resource allocation decisions within the company, which maximise profits; and provide objective divisional performance measures for management purposes. By increasing the transfer price, the upstream affiliate would increase its post-tax profits and its net present value (NPV). Of course, if upstream activities face lower taxes, the profits and NPV of the downstream business would be reduced but by less than the increase in the upstream value because of differences in tax rates. The calculation of future cash flows must therefore assume the expected fiscal regime and consider any expected change that could affect the optimal structuring of an integrated project.
A second set of disputes in relation to transfer pricing relates to shareholder disputes about internal transactions among entities within the same multinational company. These include post-M&A disputes, breach of shareholder agreements, financial instruments valuation, compensation differences and dividend-related dispute.
In commercial arbitration, transfer pricing issues may arise in the context of shareholder disputes, whereby shareholders disagree on the level of profit made by the entities they own, as a result of intercompany transactions. This can for example be the case when an entity controlled by one shareholder sells products or charges service fees or intangible related fees to an affiliate it co-owns with other shareholders.
In investor-state arbitration, transfer pricing issues may be linked to the current wave of regulations aiming at countering tax avoidance and tax optimisation schemes following the Base Erosion and Profit Shifting initiative led by OECD. While measures taken by tax authorities are legitimate, parties can call upon investment treaties to challenge fiscal measures.
A common topic of recent disputes in commercial arbitration relates to minority shareholders viewing the level of management fees paid by the entity they own to the majority shareholder as abusive. Subsidiaries pay management fees to their parent company in return for centrally performed activities. The latter can be administrative and of low value, as well as strategic or operational and of high value. Moreover, such activities can lead to the development of intangibles, which need to be taken into consideration in assessing the right remuneration of the parent company at arm’s length. The diversity of headquarter activities and organisation structure translates into a diversity of transaction structure. Management fees can thus be charged on a cost-plus basis, on a revenue basis (royalty), on a lump-sum basis or following other mechanisms.
Minority shareholders are not indifferent to the level of management fees paid by the entity they own shares in, since those fees have an influence on the profit level of the entity, and in turn on dividends paid.
To assess the correct level of management fees, the OECD guidelines on transfer pricing clearly refer to the arm’s-length principle and suggests the remuneration must be in line with what independent third parties would have agreed upon in similar circumstances. In this context, the assessment of the level of management fees that should be paid at arm’s length – and that should satisfy both the minority and majority shareholders’ interests – would likely account for the following factors:
- the roles and responsibilities of the parent company in performing centralised activities and of the subsidiary, in line with the functions they perform, the risks they assume and the assets they own, within the overall framework of the group’s value creation;
- the actual benefits obtained by the subsidiary from the centralised activities, relative to its next best alternative;
- the contractual arrangements governing the overall relationship between the parent company and the subsidiary (this is not necessarily limited to the management services agreement between the parent company and the subsidiary, but could include other agreements that could have an impact on pricing);
- a comparison with similar market practices, and a specific pricing analysis, notably in cases where comparability is limited.
The last factor is often the case in the MEA region, where data availability can be an issue. In this case, whenever transactions amongst unrelated parties are not available to serve as a point of reference, it might possible, under certain conditions, to rely on similar transactions with other parties with different ownership interests. Such analysis requires, however, a careful review of the facts and circumstances surrounding each transaction.
Subscribe here for related content, breaking news and market analysis from Global Arbitration Review.
Covering commercial and investment arbitration, Global Arbitration Review is the unofficial broadcaster for members of the international arbitration community, keeping them feeling up to date and informed.