The Use of Relative Valuation Methods in Investment Arbitration

In summary

This chapter sets out how relative valuations methods, otherwise known as comparables or multiples valuation methods, are applied, and considers their usefulness in arbitration contexts. It goes on to review investment arbitration awards in which tribunals considered the use of relative valuation methods.

Discussion points

• Differences in the use of relative valuation methods in commercial contexts and in investment arbitration
• The reliability of valuation conclusions based on comparables valuation methods relative to discounted cash-flow analysis
• Practical difficulties in the application of relative valuation methods
• Reliance on relative valuation methods by investment arbitration tribunals

Referenced in this article

• Tidewater Investment SRL and Tidewater Caribe, CA v the Bolivarian Republic of Venezuela (ARB/10/5)
• Sistem Mühendislik Insaat Sanayi ve Ticaret AS v Kyrgyz Republic (ARB(AF)/06/1)
• Waguih Elie George Siag and Clorinda Vecchi v The Arab Republic of Egypt (ARB/05/15)
• Yukos Universal Limited (Isle of Man) v The Russian Federation (PCA Case No. AA 227)
• OI European Group BV v Bolivarian Republic of Venezuela (ARB/11/25)
• Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2)

Introduction

Relative valuation methods, also known as comparables or multiples valuations, are perhaps the most commonly used business valuation methods in investment contexts. Relative valuation analysis is, however, less commonly relied on by Tribunals in investment arbitration. This chapter explores why it is that relative valuation methods play a less prominent role in investment arbitration than in other contexts, and whether the usefulness of relative valuation methods is sometimes overlooked in disputes.

The chapter first describes how relative valuation analysis is applied. It then explains some of the strengths and weaknesses of relative valuation methods, principally by comparing them to discounted cash flow (DCF) analysis, which, in our experience, is the method most commonly used to calculate ‘forward looking’ market values in investment arbitration. The chapter then considers how certain investment tribunals have approached relative valuation analysis. Finally, the chapter sets out views as to how relative valuation analysis can usefully be relied on in the context of arbitration.

What is relative valuation?

Relative valuation is the valuation of an asset by reference to observed prices at which other, similar assets have been traded. We are undertaking a very simple form of relative valuation analysis when we note that the identical house next door to ours was recently sold for €150,000 and observe that our house is likely also worth around €150,000. A slightly more sophisticated analysis might note that a house on the same street as our own was recently sold for €100,000 and had floor area of 100 square metres, or €1,000 per square metre. As our house has floor area of 150 square metres, it must be worth around €150,000. In the latter case we are applying a ‘multiple’ of €1,000 per square metre to the floor area of our house.

In business valuation, the most frequently encountered forms of relative valuation involve the application of multiples of financial metrics derived from transactions in comparable companies. For this reason, in business valuation, relative valuation is often known as ‘multiples’ or ‘comparables’ valuation.

For example, we may know that a business similar to one we are seeking to value was recently acquired for €150 million. We may also know that that business generated earnings before interest, tax, depreciation and amortisation (EBITDA) of €15 million in the twelve months leading up to the date of acquisition. If so, that business was acquired at an enterprise value (EV) to EBITDA multiple of 10x (€150 million/€15 million = 10). Applying that multiple to the business we are seeking to value, which we know generated EBITDA of €20 million in the past 12 months, we would arrive at a valuation of €200 million (10 x €20 million = €200 million).

In principle, relative valuation is straightforward and intuitive. This is one reason that relative valuation is used widely in commercial and investment contexts. For example, a survey of European valuation practitioners found that around 80 per cent made use of relative valuation approaches^[1] and a survey by Duff & Phelps found that valuation by reference to quoted comparables was the most widely used approach in fairness opinions.^[2] Multiples are an important part of the ‘language’ of valuation – participants in mergers and acquisitions and stock market investors will often express deal values or share prices in terms of profit multiples.

While the principles are straightforward, complexities arise in practice. In the following section we briefly consider two of the key decisions required in a relative valuation: first, the choice of comparables, and second the choice of metric to which the multiple is applied.

Selection of comparables

A key determinant of a multiples valuation is the selection of comparable quoted companies and transactions from which an appropriate multiple is derived. Consider our house valuation example – if we are valuing a suburban residential house located five minutes’ walk away from a railway station with direct access to the city centre, it is intuitively clear that applying a price per square metre multiple derived from the sale of an office building in the financial district or from a rural warehouse building will not lead to a reliable valuation. Likewise, care needs to be taken when selecting comparables in a business valuation.

Despite the prevalence of relative valuation methods, and the importance of comparable selection to the result of any relative valuation, there is no universally agreed list of criteria for selecting comparables. Factors that are commonly used to select comparables in business valuation include:

• industry;
• country;
• size;
• profit margin;
• leverage; and
• product diversification.

In considering the relevance of these factors, it is useful to recall that risk and growth are generally considered to be the key underlying drivers of the value of a stream of cash flows. All else equal, investors prefer streams of cash flows that are subject to less risk. And all else equal investors prefer streams of cash flows that are expected to grow faster. As a result, a higher multiple will generally be appropriate for a business with a less risky stream of expected cash flows or for a business whose profits are expected to grow more quickly.

For that reason, the most important factor to consider when considering the comparability of a particular business is the extent to which it is expected to grow at the same rate as the target business and whether it is exposed to a similar level of risk.

Choice of metric

In business valuation the mostly commonly applied multiples are multiples of the profits earned by the business, for example EV/EBITDA multiples or price earnings (P/E) multiples in which the value of a company is expressed as a multiple of its profits after tax. This reflects that the value of a business is generally considered to depend on the profits or cash flows it will earn for its owners.

When calculating profit multiples, it is important to ensure that the profit metric for the comparable business is calculated consistently with that for the target business. For example, the recent introduction of IFRS 16, an accounting standard that affects how leases should be accounted for in many jurisdictions, can have a substantial impact on the calculation of EBITDA. Applying an EV/EBITDA multiple derived from a business whose EBITDA is accounted for under IFRS 16 to a target business that does not report under IFRS 16 could result in misleading conclusions.

Less commonly, businesses are valued at a multiple of revenues. EV/revenue multiples can be useful for businesses that are currently unprofitable and to which profit multiples cannot be applied. However, as profit margins achieved by businesses vary, EV/revenue multiples provide a less reliable link between the value of a business and the cash it will earn than do profit multiples, and so typically allow less precise valuation conclusions to be drawn.

Certain valuation multiples are applied in particular industries. For example, in extractive industries, investors consider EV/tonne of resource; investors in oil exploration businesses sometimes consider EV/barrel; retail businesses can be valued at a multiple of square feet of store space; the EV/berth metric is sometimes used in the valuation of cruise lines; and EV/key (or EV/room) is sometimes used in hotel valuation. As the value of a business ultimately depends on the cash it will earn, the precision of a valuation performed using these multiples generally depends on whether there is a close link between the metric and cash flow. In some cases, multiples of industry specific metrics are used as rules of thumb rather than as standalone valuation methods in their own right.

Multiples valuation and DCF

Arbitral tribunals are increasingly familiar with the use of the DCF valuation approach. The starting point of a DCF valuation is a projection of the cash flows that the subject asset of the valuation will generate in the future. Those cash flows are then discounted back to their present value at a discount rate that reflects the relevant risk to which they are exposed.

As we explain above, the key factors that affect the value of a stream of cash flows are its expected growth and the risk to which it is exposed. In a DCF valuation, expected growth in cash flows is reflected in the preparation of the cash flow projection. That cash flow projection typically depends on a number of judgments and assumptions on the part of the valuer. If profits are expected to grow more quickly, all else equal this will lead to a higher cash flow projection and a higher resulting valuation. Risk is reflected in the discount rate, which also depends on the valuer’s judgment. All else equal more relevant risk should be reflected in a lower discount rate and a lower resulting valuation.

The fundamental starting point of a profit multiples valuation is therefore the same as that for a DCF valuation: in each case, we are seeking to value a stream of profits or cash flows. In each case, that stream of profits or cash flows will be more valuable if it is expected to grow more quickly, and will be less valuable if it is considered to be more risky.

The key difference between the approaches arises from how the growth and risk in the cash flows is reflected. In the case of a DCF, growth is reflected explicitly in the cash flow forecast and relevant risk is reflected explicitly in the discount rate. In the case of a multiples valuation both risk and growth are reflected implicitly in the multiple applied. More specifically, the risk and growth profile of the subject company is assumed to be equivalent to that of the comparable business or businesses from which the multiple has been derived.

The strengths of multiples valuation relative to DCF flow from this difference. A key advantage of a multiples valuation is that it is rooted directly in market evidence. A multiple is derived directly from transactions between willing buyers and sellers in the market. A multiples valuation does not therefore rely on as many judgments and assumptions on the part of the valuer as a DCF valuation. This strength is particularly pertinent in the case of a dispute. As a multiple is derived from prices set by parties unconnected to the dispute, it can provide persuasive evidence of value.

A multiples valuation can also be quicker to prepare as it does not require a detailed projection of cash flows over a multi-year period, although the analytical work required to identify comparators with equivalent risk and profit growth to the subject asset should not be underestimated.

A key weakness of a multiples valuation relative to DCF flows from the same difference. The key assumption underlying any profit multiples valuation is that the risk and growth profile of the subject asset’s profits is equivalent to that of the comparator businesses from which the multiple has been derived. In practice, it can be difficult to identify businesses with the same risk and growth prospects. This can limit the precision of the results of a multiples valuation relative to a DCF valuation in which the cash flow projections and discount rate can be tailored to the particular characteristics of the subject asset.

Given the strengths and weaknesses of each of the DCF and multiples valuation approaches, valuers often seek to apply both methods. Since multiples valuation and a DCF valuation are both exercises in valuing an asset by reference to the growth and risk profile of their future profits or cash flows, the two approaches should, at least in principle, arrive at the same result.

In practice, however, that is rarely the case. That is because the explicit assumptions about growth and risk made by the valuer in preparing a DCF valuation will rarely match exactly the implicit assumptions reflected in the multiple at which transactions in the comparable businesses have taken place, which are in any event often difficult to observe.

Nonetheless, the results of the two approaches should be broadly comparable, and it is good practice to seek to understand and rationalise the difference between the results. A substantial difference between the two could mean that the preparer of the DCF has made different assumptions from those of market participants. It could mean that the risk and growth profile of the comparable businesses is materially different from that of the subject asset. Whatever the reason, a comparison of the results of one valuation approach to those of the other, followed if necessary by revisiting the analysis and assumptions from which each flows, will generally lead to a more robust overall conclusion.

It is for this reason that, when there is adequate information and it is proportionate to do so, a valuer should generally seek to pursue both approaches. In disputes in which different valuers have adopted different valuation approaches and have arrived at different conclusions, it can be more constructive to seek to understand why the difference in results arises and triangulate between them than to attempt to determine whether a multiples or DCF approach is more appropriate and reject outright the results of the approach that is not preferred. Of course, when arbitration tribunals are faced with such a situation, as they often are, the information required to do so is not always available.

Relative valuation and arbitration

There have been a number of investment treaty arbitration awards in which tribunals have considered the appropriateness of relative valuation approaches, examples of which are set out below.

Awards in which relative valuation was rejected

Certain tribunals have rejected relative valuation approaches. For example, in Tidewater vs Venezuela the value of Semarca, the claimant’s operating subsidiary in Venezuela, was at issue. The tribunal rejected the use of multiples on the basis that the comparables presented were insufficiently similar to Semarca:

The Tribunal is not persuaded that the other transactions and companies referenced by Claimants’ experts in their reports are sufficiently comparable to SEMARCA’s business, and to the market in which it operated, so as to assist it in arriving at a fair market value of SEMARCA. It accepts the views of Respondent’s experts that the differences between the business contexts in which such other companies were operating are simply too great for them to be comparable.^[3]

Similarly, in Sistem Mühendislik Insaat v Kyrgyzstan, the tribunal rejected the use of multiples to value a hotel, stating:

In this case the Tribunal is not persuaded that there is an adequate basis for the application of the ‘multiple deals’ approach. Conditions in the Kyrgyz hotel sector in the period in question are not so obviously comparable with the comparators in Ireland, Sweden, the UK, the USA, and the other States to which the first Raymond James report makes reference as to persuade the Tribunal, without more, that these comparators offer a sound basis for comparison and extrapolation. The fact that the Raymond James report states that ‘there are almost no completed transactions in the Central Asian region with accurate transaction data’ and that ‘in addition, 30% discount is applied on the deal multiples to normalize the multiple for the Kyrgyzstan market’ reinforces the view that the valuation based on this approach involves a large measure of speculation.^[4]

While both of these tribunals rejected the use of comparables, they did so on the basis of the lack of comparability of the particular multiples put to them. Indeed, it is easy to see why it was difficult to find reliable evidence of the prices paid for comparable assets, given the assets in question were a marine oilfield services business in Venezuela and a hotel in Kyrgyzstan. It is likely that the particular nature and circumstances of assets that are the subject of investment arbitration, and the resulting difficulty in finding comparables, is a key reason for the lack of reliance by investment tribunals on multiples valuations.

However, neither award appears to rule out the use of multiples as a valuation methodology more generally. In fact, the tribunal in Sistem made that point explicitly:

Accordingly, while the ‘multiples’ approach would in principle be an appropriate approach to valuation in the circumstances of this case, there is inadequate information to which that approach could be applied.^[5]

Awards in which relative valuation was relied on by the tribunal

Siag & Vecchi v Egypt concerned the value of land on which the claimants had planned to build a hotel but which, they alleged, Egypt expropriated. The claimants presented a DCF valuation prepared on the assumption that the hotel would have been completed and would have commenced operations.^[6] The tribunal rejected the DCF in light of the early stage of the development of the hotel.^[7] Instead, it relied on a valuation of the land by reference to transactions in similar properties. It recognised there was some imprecision in the comparable valuation and reflected that imprecision by applying a discount to the valuation:

[The surveyor said] that ordinarily he would hope to be within 5% either side of an exact or precise figure. However, in view of the uniqueness of this Property and the difficulties which he acknowledged were attendant upon conducting this particular valuation, he stated that in the present case: ‘I believe that percentage should be wider, and it could be at least 10% on either side of my figure.’ In all the circumstances, the Tribunal increases that margin to 20% and will apply that discount to the value of the Property as assessed by [the surveyor].^[8]

In Yukos v Russia, the tribunal also relied on a relative valuation. In that case, the claimant’s expert put forward a valuation that depended in part on a DCF calculation and in part on two multiples valuations, one prepared by reference to trading in listed stocks (the ‘comparable companies’ approach) and the other by reference to transactions in comparable privately-owned businesses (the ‘comparable transactions’ approach).^[9],^[10]

The respondent’s expert put forward criticisms of the claimant’s expert’s DCF calculation. He also stated that certain of the comparable companies and none of the comparable transactions relied on by the claimant’s expert were comparable to Yukos.

The tribunal agreed with the respondent’s expert’s criticisms.^[11] However, as the respondent’s expert had not put forward his own alternative valuation, the tribunal adapted the claimant’s expert’s valuation to reflect the respondent’s expert’s criticisms. In its own valuation of Yukos, the tribunal rejected DCF, rejected multiples valuation by reference to comparable transactions and relied entirely on valuation by reference to comparable companies, after adjusting the claimant’s calculation to remove those comparables identified as unreliable by the respondent’s expert.^[12]

The Yukos award is interesting in that it is a rare investment treaty award in which the tribunal relied entirely on multiples valuation in arriving at its valuation conclusions.^[13] However, the tribunal appeared to do so primarily due to the lack of a reliable alternative presented to it.

Awards in which the tribunal used multiples valuation as a relevant benchmark

The awards in OI European v Venezuela and Crystallex v Venezuela are particularly interesting in this regard, because the tribunals took the evidence of multiples valuations into account in their overall conclusions without relying on them exclusively. That approach is consistent with how investors tend to make use of multiples valuations.

In OI European v Venezuela, the claimant’s expert proposed a valuation of the subject asset calculated as the weighted average of his DCF, comparable companies and comparable transactions valuations.^[14] Those valuations arrived at broadly similar results.

However, according to the tribunal, the respondent’s expert argued that:

the market value should be calculated by applying the DCF methodology only, while the other methodologies should be used as a simple ‘sanity check,’ in the words of the expert, in order to corroborate the result achieved.^[15]

The tribunal agreed with the respondent’s expert.^[16] Nonetheless, it recognised the difficulties that arise with a DCF valuation:

Any DCF model is simply the result of plugging some estimated future parameters, as determined by the expert, into a mathematical formula. If the estimates of these parameters turn out to be inaccurate, the results will not reflect the true market value of the expropriated asset. Small changes in the parameters can cause very significant differences in the outcomes.^[17]

As a result, the tribunal considered that it would be appropriate to cross-check the results of its DCF valuation by reference to a multiples approach:

[The Tribunal] will develop its own DCF model, determine the resulting valuation, and then check this valuation by comparing it to the value obtained through the comparable-companies and comparable-transactions methodologies.^[18]

An approach of seeking to rely on more than one valuation methodology is consistent with the approach generally adopted by investors. However, as an aside, the practical effect of the tribunal’s decision that DCF should be adopted as a primary valuation methodology, while the other valuation methodologies should only be used as a ‘sanity check’ is unclear. In the case of both the claimant’s and the tribunal’s valuations, the results of the different valuation approaches were broadly consistent with one another and so the choice of valuation methodology had little effect on the overall conclusion.^[19] Had the relative valuation arrived at a different conclusion to the DCF, in applying the ‘sanity check’ the tribunal considered necessary, presumably it would have been necessary to revisit the assumptions behind the DCF. If it had done so, the tribunal would have been relying, at least indirectly, on the relative valuation.

Crystallex v Venezuela concerned the value of certain gold-mining rights. The claimant’s experts put forward four valuation methodologies,^[20] of which the tribunal adopted two: valuation by reference to the stock market capitalisation of the company whose principal asset had been expropriated and a comparable companies valuation.^[21]

As regards the comparable companies valuation approach, the tribunal’s view was as follows:

The Tribunal considers that such method is widely used as a valuation method of businesses, and can thus be safely resorted to, provided it is correctly applied and, especially, if appropriate comparables are used.^[22]

The tribunal recognised the difficulties in finding appropriately comparable companies, but considered this was no obstacle to relying on a multiples approach in the circumstances of the case:

… the Tribunal is of the view that the Claimant and its experts have identified sufficiently comparable companies to find their EV/Resource multiple. While the Tribunal has noted the Respondent’s criticism that some of the 73 companies used as comparables seem to bear little resemblance to Crystallex, no two companies will ever be exactly alike. This is a given that must be accepted when using this kind of methodology. After all, ‘to compare’ is a process made with objects similar to the subject rather than with identical objects – if those even exist.^[23]

The tribunal did not express a preference between the two methodologies that it adopted, and, noting that they arrived at similar results, awarded damages by reference to the average of the two.^[24]

In arriving at its decision as to the appropriate valuation methods, the tribunal recognised that the choice will depend on the facts of the case:

Valuation is not an exact science. There often is no single value of a business. Rather, there are typically a range of values. Similarly, there is no one methodology best suited for determining the fair market value of the investment lost in every situation. Tribunals may consider any techniques or methods of valuation that are generally acceptable in the financial community, and whether a particular method is appropriate to utilize is based on the circumstances of each individual case.^[25]

That the appropriate valuation method is fact specific is borne out in the tribunal’s own decision-making. While it chose to rely on the comparable companies approach, and in doing so recognised the difficulties in finding good comparables, it rejected the comparable transactions approach. The tribunal did so because it considered it had not been provided with sufficient evidence to rely on it:

The Tribunal has no difficulty in accepting that in theory such method could yield reasonable results and would thus be an appropriate valuation method to value an investment in an international arbitration. However, its particular application by the Claimant’s expert presents uncertainties and speculative elements so that the Tribunal cannot consider it in this particular case.^[26]

Conclusions

Relative valuation is widely used in the investment community and valuation multiples are an important part of the language of valuation. Multiples can provide particularly useful evidence in disputes as they provide direct market evidence of value and rely less on the assumptions and judgment that underpin a DCF valuation. However, the identification of properly comparable companies and transactions is a key challenge when adopting multiples valuations.

This is reflected in the practice of investment treaty tribunals that have considered the multiples valuation methodology. They generally appear to consider that relative valuation can be an acceptable valuation methodology. Where they have rejected it, they have done so because the valuations they have been presented with do not rely on sufficiently comparable companies.

Where information allows and it is proportionate to do so, relative valuation should be used alongside other valuation methods to improve the reliability of overall valuation conclusions. That approach is commonly adopted by investors and has been adopted by certain tribunals.

Notes

^[1] The Gap between Theory and Practice of Firm Valuation: Survey of European Valuation Experts.

^[2] In Defense of Fairness Opinions: An Empirical Review of Ten Years of Data.

^[3] Tidewater Investment SRL and Tidewater Caribe, CA v the Bolivarian Republic of Venezuela (ARB/10/5), Award: 166.

^[4] Sistem Mühendislik Insaat Sanayi v Ticaret AS v Kyrgyz Republic (ARB(AF)/06/1), Award: 162.

^[5] Sistem Mühendislik Inşaat Sanayi ve Ticaret AS v Kyrgyz Republic (ARB(AF)/06/1), Award: 163.

^[6] Waguih Elie George Siag and Clorinda Vecchi v The Arab Republic of Egypt (ARB/05/15): 549.

^[7] Waguih Elie George Siag and Clorinda Vecchi v The Arab Republic of Egypt (ARB/05/15): 566.

^[8] Waguih Elie George Siag and Clorinda Vecchi v The Arab Republic of Egypt (ARB/05/15): 575–6.

^[9] Yukos Universal Limited (Isle of Man) v The Russian Federation (PCA Case No. AA 227), Final Award: 1782.

^[10] The Claimant’s expert also advanced a number of other valuation methods, but the tribunal considered these were ‘secondary’ valuation methods put forward to support the primary valuation methods. See Final Award paragraph 1786.

^[11] Yukos Universal Limited (Isle of Man) v The Russian Federation (PCA Case No. AA 227), Final Award: 1785.

^[12] Yukos Universal Limited (Isle of Man) v The Russian Federation (PCA Case No. AA 227), Final Award: 1787.

^[13] The tribunal determined that the appropriate valuation date was different from that adopted by the claimant’s expert. It therefore relied entirely on the comparable companies approach to arrive at a valuation as at the valuation date adopted by the expert, and then adjusted it for the valuation dates it had determined.

^[14] OI European Group BV v Bolivarian Republic of Venezuela (ARB/11/25), Award: 655.

^[15] OI European Group BV v Bolivarian Republic of Venezuela (ARB/11/25), Award: 656.

^[16] OI European Group BV v Bolivarian Republic of Venezuela (ARB/11/25), Award: 657.

^[17] OI European Group BV v Bolivarian Republic of Venezuela (ARB/11/25), Award: 662.

^[18] OI European Group BV v Bolivarian Republic of Venezuela (ARB/11/25), Award: 666.

^[19] The tribunal noted this itself – see paragraph 667 of the award.

^[20] Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2), Award: 887.

^[21] Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2), Award: 916.

^[22] Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2), Award: 901.

^[23] Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2), Award: 901.

^[24] Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2), Award: 917.

^[25] Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2), Award: 886.

^[26] Crystallex International Corporation v Bolivarian Republic of Venezuela (ARB(AF)/11/2), Award: 886.