counting the costs of investment treaty

8/10/2019 Counting the Costs of Investment Treaty

http://slidepdf.com/reader/full/counting-the-costs-of-investment-treaty 1/12

ar gThe international journal of

commercial and treaty arbitration

This article was frst published in the Global Arbitration Review online news, 24 March 2014

www.globalarbitrationreview.com

NEWS

The expense of bringing a treaty claim is a major concern for investors but the nal

allocation of costs in a case can be hard to predict. In a major survey, Matthew

Hodgson of Allen & Overy in Prague examines the costs ndings in over 200 casesand argues that more transparency on the subject is needed.

Te impressive growth o investment treaty arbitration over the past 15 years showsew signs o exhaustion. In 2013, some 40 new ICSID cases were commenced, secondonly to 2012 when the figure was inflated by a flurry o claims against Venezuela.Meanwhile, the number o bilateral investment treaties in orce continues to grow, andground-breaking projects such as the EU-China and EU-US ree trade agreementsare in prospect.

Te continued growth o investment treaty arbitration brings it into contact with newusers. Other than certain states and a small number o multinational companies, ewentities are requent users o the system. Te aim o this study is to shed empirical lighton a subject o considerable interest to such users, namely the costs o the process,and to examine the propensity o tribunals to adjust costs one way or another. It alsoconsiders the substantive outcomes o proceedings.

Te study involved a review o some 221 cases in which an award or decision waspublicly available on 31 December 2012. Afer making other adjustments (includingremoving decisions reserving costs to a later stage o proceedings), the final data poolwas 176 cases.

Counting the costs of investment treaty

arbitration

CAPTION

24 March 2014

http://globalarbitrationreview.com/news/article/32513/









NEWSCosts

Te costs o an investment treaty arbitration all into two broad categories. First,there are the ees and expenses o the arbitral tribunal itsel, together with those othe institution or appointing authority, i any (tribunal costs). In addition, there arethe ees and expenses o legal counsel as well as any expert or actual witnesses (partycosts).

Party costs

Te study revealed that the average party costs were quite similar, at US$4,437,000or claimants and US$4,559,000 or respondents.

Actual party costs data or the claimant and respondent were available in 73 and 66

cases (out o the total pool o 176 cases) respectively. As a result o the relatively smalldata set, a ew cases with extremely high costs distort the figures. Te median resultor claimants is somewhat lower at US$3,145,000, whereas or respondents there is amuch sharper drop to US$2,286,000. Tese figures are perhaps more reflective o atypical mid-value claim.

Te reason or this divergence is that in a number o cases the respondent’s partycosts were significantly greater than the claimant’s. Te most extreme examples areseen in the table below:

able 1: Party costs – substantial discrepancies

Name o case Respondent’sparty costs (US$)

Claimant’s partycosts (US$)

Discrepancy(US$)

Jan Oostergeteland TeodoraLaurentius vSlovakia

16,330,000 2,231,000 14,099,000

Libananco vurkey

35,702,000 24,382,000 11,320,000

EDF v Romania 18,575,000 8,198,000 10,377,000

It is thereore clear that, excluding a number o extreme results, claimants typicallyincur considerably greater party costs than respondents. Tis is perhaps unsurprisingsince the claimant bears the burden o proo. Further, respondent states are generallymore cost sensitive than claimants, with their tendency to run public tenderswhere price is the driving actor. Certain states also have in-house legal teams withsubstantial investment treaty arbitration experience. In some cases (or example, theUS State Department) these legal teams directly represent the state in proceedings.










NEWSTe study also confirmed that average party costs have increased over time, with a riseo 32 per cent in average party costs or claimants between cases decided beore 2006and those decided in or afer 2006. However, a much sharper rise in the median (192per cent) suggests that the rise in party costs or a representative “mid-value” claimmight in act be higher than this. Respondents have seen a similar average increasein party costs, with an average rise o 32 per cent or post-2006 claims. However, thisrepresented a more modest 96 per cent rise in the median cost

ribunal costs

urning to tribunal costs, the study reveals that the average costs are US$746,000(median US$590,000), or just over US$373,200 or each party (split on a 50:50 basis).Te pool o available cases was 70.

As with party costs, tribunal costs have also shown a noticeable increase over time,with average tribunal costs or 2006 and onwards 56 per cent higher than thosebeore 2006.

Interestingly, the study revealed a disparity between tribunal costs depending on theorum.

able 2: Average tribunal costs – UNCIRAL/ICSID comparison

Forum selected Average tribunal costs

(US$)

Median tribunal cost

UNCIRAL 853,000 714,000

ICSID 769,000 544,000

Te average tribunal costs in ICSID claims are 10 per cent lower than in an UNCIRALclaim. Te median difference is nearly 24 per cent.

Te obvious explanation or this is ICSID’s cap on arbitrators’ ees. Te currentschedule o ees (in orce since 1 January 2008) states that, in addition to any directexpenses reasonably incurred, arbitrators are entitled to receive a ee o US$3,000

a day or work perormed in connection with the proceedings (together with asubsistence allowance where relevant).

By way o comparison, UNCIRAL has not historically had a similar ee cap, althoughthe 2010 rules (article 41.4) now give the parties a right to appeal the tribunal’s eesto the appointing authority. Te appointing authority has the power o adjustmento ees and expenses, which is binding on the tribunal. It remains to be seen whetherthese changes will reduce the discrepancy between tribunal costs in UNCIRAL andICSID proceedings.

It should be noted that the average costs or the SCC were even lower at US$480,000.However, the available data pool was considerably smaller (13 cases). Further, the










NEWSaverage amount claimed in SCC cases was much lower: approximately US$92.2million, as compared with an overall average o roughly US$491.7 million. TeSCC also provides or caps on the arbitrators’ ees, scaled to the amount in dispute(although costs in cases worth over €100 million are decided by the SCC’s board odirectors on an individual basis).

Substantive outcome: who wins and how much?

Te study also considered the relative success o claimants and respondents in relationto the substantive outcome. For simplicity, success was defined or the claimant asinstances where it received an award o some amount in damages on the merits.

Across the ull pool o 176 cases, respondents were successul in 59 per cent o cases.Tis is a sharp reversal rom the initial trend in early investment treaty arbitrations:

prior to 2001, investors prevailed in 63 per cent o cases.

Chart 1 – Substantive outcome

O these, around 26 per cent o all cases (or nearly hal o total respondent “wins”)were dismissed on jurisdiction. Some our cases (or around 2 per cent o the total)were terminated in circumstances where the claimant was ordered to contribute tothe respondent’s party and tribunal costs.

In cases where the claimant was successul, the average amount awarded wasUS$76,331,000. Again, the largest claims have a substantial distortive effect. Tecombined value o the awards in Occidental Petroleum v Ecuador II, CSOB vSlovakia, CME v Czech Republic and Siemens v Argentina alone was approximately

US$3.12 billion, exerting significant upward pressure on the average. Te median










NEWSamount awarded was substantially lower at US$10,694,000.

It is also interesting to consider the amount awarded in comparison to the amountclaimed. In this regard, the study showed that on average a successul claimant wasawarded 41 per cent o the amount it claimed. In just three cases out o the total 176(CDC v Seychelles, Deutsche Bank v Sri Lanka and SGS v Paraguay) did the investorrecover the ull amount claimed.

Overall, the average amount claimed is US$491,656,000. However, this figure concealsa notable discrepancy. A urther breakdown reveals that in cases where the claimantis ultimately successul, the average amount claimed is US$166,388,000, which issubstantially lower than the average o US$816,924,200 in cases where the claimantis unsuccessul. Tis suggests that the average amount claimed is greatly inflated by anumber o flawed claims. Te three unsuccessul cases in which the largest amounts

were claimed are set out in the table below.

able 3: Largest unsuccessful claims

Case Amount claimed (US$)

Saba Fakes v urkey 19 billion

Libananco v urkey 10.1 billion

Generation Ukraine v Ukraine 7.1 billion

Each o these awards involved misconduct, i not outright bad aith, by the claimant.Somewhat bizarrely, in Generation Ukraine, the claimant apparently sought torecover an alleged US$2.3 billion in legal costs. Needless to say, this request, alongwith all claims, was rejected.

Apportioning costs

When it comes to the apportionment o costs there are two basic philosophies. Te

“pay your own way” principle suggests that a party should bear its own costs. Alsoknown as the “American Rule”, this is the deault position in US litigation and alsoin cases beore the International Court o Justice. A number o investment treatytribunals have endorsed this approach, or example, Romak v Uzbekistan, MCIPower v Ecuador and Anderson v Costa Rica.

In contrast, the “costs ollow the event” or “loser pays” principle suggests that thesuccessul party should ordinarily recover its reasonable costs. It is usual or alosing party to make at least some contribution to the costs o the winner in thenational courts o many jurisdictions, including England & Wales, France, Germany,Hong Kong, Italy, the Netherlands, Poland, Russia, Singapore, Spain and Tailand.According to a 2012 study by Queen Mary, University o London, this approach is










NEWScommon in commercial arbitration, being adopted in around one hal o cases, witha urther 30 per cent o cases seeing some apportionment o costs, leaving just 20per cent o cases where costs were not adjusted. In the investment treaty arbitrationcontext, it has ound support, or example, in the decisions o ADC v Hungary and

Gemplus v Mexico.

A third and more nuanced approach (really a variation on the “costs ollow the event”principle) seeks to apportion costs based on the relative success o the different issuesin the arbitration. Examples o this approach are the decisions in Nagel v CzechRepublic, International Tunderbird Gaming v Mexico, EDF v Romania, Occidental

v Ecuador II, and Helnan v Egypt. An unusual application o this “relative success”approach came in Eastern Sugar v Czech Republic,where the tribunal required thesuccessul investor to pay part o the respondent’s share o tribunal costs (that is,ultimately “winner pays”), apparently on the basis that the claimant had submitted a

large claim, o which it was awarded only around 25% in damages.

Te ICSID Convention does not express a preerence as between these approaches.Article 61(2) o the Convention leaves costs to the broad discretion o the tribunal,without offering an indication o the actors a tribunal may take into account inreaching its decision. An earlier draf o the Convention established a deault “payyour own way” approach, except in the event o claims commenced “rivolously”or in “bad aith”. Ultimately, the question o costs was lef entirely to the tribunal’sdiscretion.

A different approach is taken in the 2010 UNCIRAL rules, which state at article42.1 that “[t]he costs o the arbitration shall in principle be borne by the unsuccessulparty or parties” subject to the proviso that “the arbitral tribunal may apportion eacho such costs between the parties i it determines that apportionment is reasonable,taking into account the circumstances o the case”. Te 1976 UNCIRAL rulescontain similar, albeit more limited, provisions: article 40.1, which applies only totribunal costs, and article 40.2, which expressly gives the tribunal discretion as to theparty costs. As will be seen, this has led to divergent approaches between ICSID andUNCIRAL tribunals on the subject o costs.

Apportionment in practice

A tribunal has three basic options in terms o costs: to make each party responsibleor its own party costs and 50 per cent o the tribunal costs (an unadjusted costsorder); or to order one party to pay the other side’s party costs and tribunal costs inull (a ully adjusted costs order) or in part (a partially adjusted costsorder).

A slight majority o arbitration tribunals (56 per cent) have required each party tobear its own costs, that is, an unadjusted costs order. Just 10 per cent o tribunalsmade a ully adjusted costs order and 34 per cent a partially adjusted costs order.










NEWSChart 2: Adjustment of costs

Te chart below examines the ways in which tribunals elect to make an adjustmentmore closely. In nearly two-thirds o these cases, tribunals elected to adjust someportion o both tribunal and party costs. In the remaining cases, tribunals were morelikely to make the adjustment to the tribunal costs, rather than the party costs.

Chart 3: Breakdown of adjustments

Successul claimants are more likely to recover their costs than successul respondents.In a slight majority o cases (53 per cent), upon finding or the claimant, tribunalsmade an adjusted costs order. By comparison, tribunals make an adjusted costs order










NEWSin avour o a successul respondent in only 38 per cent o awards.

Chart 4: Adjustment of costs according to successful party

Tere are a number o possible explanations or this disparity. Where a claimantis successul, the tribunal has necessarily ound an unlawul act by the respondentand, ollowing the principles o compensation set out by the Permanent Court oInternational Justice in the Chorzow Factory case, may be mindul o the need tomake whole the investor or losses flowing rom the unlawul act. In contrast, wherethe respondent is successul, while concluding that it lacks jurisdiction or that therehas been no technical breach o the relevant treaty, the tribunal may still perceive thatthe claimant has a legitimate grievance and was justified in pursing the claim. An

example o this was seen in Azinian v Mexico, where the tribunal, in finding or therespondent but making an unadjusted costs order, noted that the respondent “maybe said to some extent to have invited litigation”. Similarly, the tribunal in Salini vJordan noted that, while the claimant was unsuccessul, it remained sceptical o therespondent’s case.

It is a common perception among arbitration practitioners that investment tribunalsare increasingly avouring the “cost ollows the event” approach. Te charts belowshow that this perception is justified. In awards decided beore 2006, tribunalsadjusted costs in just one-third o cases (35 per cent). However, in cases decided afer

2006, this number has risen sharply and tribunals made some adjustment in aroundhal (49 per cent) o awards.










NEWSChart 5: Adjustment of costs by date

Tere is a notable divergence between UNCIRAL and ICSID tribunals. Inarbitrations conducted under UNCIRAL rules, tribunals made some orm oadjusted costs order in over two-thirds (69 per cent) o cases, while ICSID tribunalsonly did so in just over one-third (36 per cent). Tis demonstrates the impact oguidance in the arbitral rules.

Te UNCIRAL figures also reveal that tribunals adjusted both tribunal and partycosts in 42 per cent o cases, adjusting only tribunal costs in just 22 per cent o awards.Tis is interesting because all cases surveyed were decided under the 1976 ruleswhich, as set out above, provide guidance only with regard to the tribunal costs andexpressly reserve the flexibility o the tribunal to award party costs as they considerappropriate. Yet despite this, UNCIRAL tribunals, when making an adjusted costsorder, tended also to shif party costs.










NEWSChart 6: Adjustment of costs by institution

Annulment proceedings

Te study also examined costs adjustment in the context o publicly available ICSIDannulment decisions, with a total data pool o 24 annulment cases.

Chart 7: Adjustment of costs in ICSID annulment proceedings

Perhaps surprisingly, annulment committees show a greater tendency towardsunadjusted costs orders than ordinary tribunals. In 16 cases (67 per cent o annulmentdecisions) there was an unadjusted costs order. O the remaining eight decisions,some 21 per cent o the decisions made a partially adjusted costs order whereas just12 per cent involved a ully adjusted costs order. O these eight adjusted costs orders,

our decisions involved an adjustment o the ad hoc committee’s costs only, and our










NEWSdecisions involved an adjustment o both the committee costs and party costs.

Chart 8: Breakdown of cost adjustments in ICSID annulment proceedings

In all six cases (25 per cent o annulment decisions) where the ad hoc committeedecided to partially annul the tribunal’s award, there was no adjustment o eitherparty costs or committee costs. In the five cases (21 per cent o decisions) in whichthe previous award was entirely annulled, there was no adjustment o party costs.

However, the “losing” party was required to bear committee costs in two o thosecases.

In the 13 cases (54 per cent o annulment decisions) where the application orannulment was dismissed, one might have expected tribunals more readily to requirethe unsuccessul applicant to bear the ull costs o the process, consistent with theexceptional nature o annulment proceedings and the desire to discourage dilatoryapplications. However, o these 13 decisions, there were only three cases in which thecommittee made a ully adjusted costs order. Tere was no adjustment o party costsin 69 per cent o decisions where the application was dismissed. An adjustment o

committee costs was made in 46 per cent o such cases.

It is interesting to note that the institutional appointment o the ad hoc committeedoes not increase their tendency to award costs to the successul party. Whereasthe ailure o tribunals to adopt a “costs ollow the event” approach is sometimesattributed to compromises among arbitrators to maintain the unanimity o thetribunal, this does not appear to be borne out by the evidence.









www globalarbitrationreview com

NEWSA need for transparency

One o the challenges o conducting the underlying study is that ull and clear costsinormation is provided only in a minority o cases. Tis is an area where, in theinterests o transparency, there is room or improvement on the part o tribunals.

Despite the absence o ull data, combining average party costs (or both parties)and average tribunal costs suggests a typical claim may cost in the region oUS$9,743,000; or, looking at the combined median figures, US$6,021,000. Given thatthe average award is US$76,331,000(median US$10,694,000) it is clear that costs arean important aspect o many proceedings. In cases where the respondent succeeds,costs will amount to the entire financial outcome o the dispute. At the same time,tribunals devote little attention to the subject, with just six paragraphs addressingcosts in the average award.

Te likely costs o a proceeding, and how these costs may be adjusted by a tribunal,are a part o the risk/reward assessment an investor makes beore commencingan investment treaty case. It is also relevant to a state considering its approach tosettlement discussions. As it stands, and notwithstanding the discernible shiftowards “costs ollow the event” approach in recent years, it is extremely difficultor parties to predict the approach a tribunal will take to costs. Te UNCIRALRules at least provide guidance and a deault position in this regard. However, theresulting divergence in approaches between ICSID and UNCIRAL tribunals is arrom satisactory.

I have argued elsewhere that the “costs ollow the event” approach is an appropriatestarting point in investment arbitration proceedings. One way to achieve this wouldbe to amend the ICSID arbitration rules in line with the UNCIRAL rules. Tere are,o course, differing views on this and the debate is beyond the scope o this article.However, on any view, the allocation o costs is an important subject or the partiesand deserves greater attention and a more principled approach rom tribunals.

Te author would like to thank a number of colleagues at Allen & Overy who haveassisted with this study, including Sophy Cuss, Alistair Campbell, Michal Stanek and Alistair Wrench. Te views represented in this article are those of the author and do

not necessarily reflect those of the firm




counting the costs of investment treaty

Documents