In a judgment dated 25 April 2012 (Lombard North Central plc and another v GATX Corporation  EWHC 1067 (Comm)), Judge Andrew Smith of the English High Court (Commercial Court) underlined the potential risks that might arise from arbitration clauses which have a limited scope. Although this judgment is relevant to a number of different procedural and substantive aspects (including the statutory construction of the arbitration exception as enshrined in English law (Section 9 of the Arbitration Act 1996 (the “Act”)), this blog post considers issues of scope of arbitration clauses.
In 1998, Lombard and GATX concluded an agreement concerning the financing of train vehicles. In 2000, Lombard and GATX concluded another agreement. Clause 9.4 of the 2000 agreement allowed GATX, in some circumstances, to force the sale of the trains and realise their share of the profits. The 2000 agreement, including parts of its Clause 9.4, was amended by a further agreement entered into in 2004.
Clause 9.4(i), as amended in 2004, provided that if no agreement to extend the lease of the train vehicles was put into place with the UK Strategic Rail Authority by 30 March 2004, then, by no later than 30 June 2011, the parties were to establish a joint venture “in a form and manner acceptable to the parties thereto.” This provision added that “[if] the parties have been unable to establish a [joint venture] by 30 June 2011, the parties agree to negotiate in good faith to achieve such objectives through other means at the earliest opportunity.”
Amended Clause 9.4(x) (the relevant arbitration clause) provided as follows:
“Any disputes relating to the creation of the [joint venture] pursuant to this Clause 9.4 that cannot be resolved by the good faith efforts of the parties shall be referred to and finally resolved by arbitration in London. Such arbitration shall be decided pursuant to the Rules of the London Court of International Arbitration from time to time in force.”
No joint venture was established within the agreed deadline and Lombard brought legal proceedings against GATX seeking two declarations: (1) the obligation to negotiate in good faith under Clause 9.4(i) is unenforceable for want of legal content and therefore GATX is not entitled to share any profit; and (2) the amendments agreed in 2004 take effect notwithstanding the unenforceability of Clause 9.4(i). GATX, in turn, relying upon Clause 9.4(x), requested that the English High Court stay the proceedings under Section 9 of the Act.
The parties disagreed about the scope of the arbitration clause. Lombard contended that the arbitration clause was intended only to resolve disputes as to how the joint venture mentioned in Clause 9.4 should be constituted. GATX argued that the scope of the arbitration clause was broader:
“[B]y way of illustration only, [GATX] say that it covers a dispute about whether the relevant parties have been ‘unable to establish a [Joint Venture] by 30 June 2011’ within the meaning of clause 9.4(i) and any disputes not only about the terms of a Joint Venture but also about terms of any alternative arrangement or structure constituting ‘other means’.” (para 7 of the Lombard judgment).
In the light of Section 9 of the Act, the key question to be decided by the High Court was whether the proceedings commenced by Lombard were “in respect of” a matter which under the agreement to arbitrate was to be referred to arbitration. Andrew Smith J considered that the way in which the claims are formulated is immaterial to this inquiry. Instead, the right approach is for the court to “consider what questions will foreseeably arise for determination in the proceedings and whether they include referred matters.” (para 14 of the Lombard judgment).
Andrew Smith J also stated that in cases where the parties have clearly agreed to refer only certain disputes to arbitration, “the risk of proceedings before both the courts and an arbitral tribunal is inherent in the agreement.” (para 16 of the Lombard judgment). This can lead to the partial stay of legal proceedings, giving rise to the fragmentation of related disputes.
Such fragmentation of disputes did not take place in Lombard. It was common ground between the parties that if GATX was entitled to a stay in relation to the first declaration sought by Lombard, then the claim for the second declaration should also be stayed. Smith J concluded that the first declaration pursued by Lombard fell within the scope of the arbitration clause and therefore he stayed the whole of the proceedings.
The Lombard judgment highlights some of the potential risks arising from concluding an arbitration clause with a limited scope or inserting carve-out clauses into an otherwise broad arbitration agreement. Contrary to what Lombard and GATX did in their agreement, the prevailing practice in international arbitration is to include arbitration clauses that have a broad scope. A broad arbitration clause provides distinct benefits, including:
• The parties prevent the fragmentation of disputes between different fora (particularly due to the nature of the claims advanced). Obviously, having to litigate related claims before national courts and arbitral tribunals is not only expensive and time-consuming but can also give rise to inconsistent results.
Historically, the potential fragmentation of disputes before different fora has been an issue affecting disputes where some public bodies have exclusive jurisdiction to decide certain matters or where the jurisdiction of national courts was preserved on public policy grounds (e.g. antitrust and intellectual property disputes). As a result of the increasing narrowing down of the doctrine of inarbitrability, at present, the prevailing practice in leading jurisdictions is to allow most disputes (including those on the invalidity of intellectual property rights) to be resolved by arbitration.
• Including non-contractual claims arising from an underlying agreement within the scope of the arbitration clause may help prevent technical difficulties down the line. This is the case, in particular, where the respondent has asserted in arbitral proceedings that the underlying agreement is invalid. If the arbitration clause only concerns contractual disputes, a finding of invalidity of the underlying agreement would deprive the relevant arbitral tribunal from the possibility of deciding the case under a torts theory. This, in turn, can give rise to two unwelcome scenarios for a claimant. First, it might have to file a tortious claim before national courts. Second, the underlying tortious claim might be time-barred.
• A broad arbitration clause may permit the parties to bring fresh claims arising from breaches of the agreement to arbitrate or the relevant rules of arbitration before an existing arbitral tribunal. This may include, for example, breaches of duties of confidentiality if so provided in the agreement to arbitrate, arbitration rules (e.g. Article 30.1 of the LCIA Rules, Article 43.1 of the DIS Rules and Article 73 of the WIPO Rules) or applicable law (e.g. English, New Zealand, Singaporean and Spanish law). Further, depending upon the scope of the agreement to arbitrate and applicable law, an arbitral tribunal might be in a position to award damages for a breach of the agreement to arbitrate itself (see in this respect the recent West Tankers judgment  EWHC 854 (Comm)).
In the light of these common benefits of broad arbitration clauses, national courts and administering institutions usually assume that the parties to an arbitration agreement would wish to have all their disputes resolved by arbitration. Particularly, national courts in many leading jurisdictions take the view that agreements to arbitrate should be construed liberally (which is the case, for example, in England, Germany, Switzerland and the USA).
Further, the language in the model clauses of many leading institutions aims at capturing most potential disputes arising from an agreement (e.g. ICC, LCIA, SCC and WIPO).
In fact, common wisdom suggests that entering into an arbitration clause with a limited scope or inserting carve-outs into an otherwise broad arbitration clause is a bad idea. Should this conclusion be elevated to a dogma of international arbitration? Is concluding an arbitration clause with a limited scope always a bad idea? Is inserting carve-outs into an arbitration clause simply a no-no?
One could, at least at first sight, be tempted to answer that limiting the scope of an arbitration clause is always a bad idea. However, for the reasons discussed below, when different factors are taken into account, the answer to the questions set out above appears to be less clear-cut.
Setting aside potential partisan considerations (where it is clear that one of the parties is likely to be the respondent, that party, for tactical reasons, may seek to agree to an arbitration clause of narrow scope, to the effect that potential disputes are fragmented – a sort of “divide and conquer” approach), parties might have legitimate reasons (quite often technical) to limit the scope of an arbitration clause.
For example, parties sometimes carve out from the scope of an arbitration agreement discrete technical or valuation disputes to the effect that these are finally resolved by means of expert determination. This can be a cost-effective and efficient solution.
In some specialist fields, the parties may have particular reasons for limiting the scope of an arbitration clause. This is the case, for example, of arbitration clauses inserted into agreements that involve intellectual property rights. If the parties wish to obtain an erga omnes ruling on the validity of the underlying intellectual property, they would have to indicate in the relevant arbitration clause that disputes on the validity of the underlying rights will be resolved by national courts or, where applicable, by a specialist national body.
Further, in the investment arbitration context there might be situations where the parties may consider limiting the scope of an arbitration clause. Suppose, for example, that a concession agreement between a State and an investor includes a model ICC arbitration clause. In this case, it would be possible to conclude that non-contractual claims which are related to the underlying investment fall within the scope of the relevant arbitration clause. Where applicable, depending upon the language of the relevant investment treaty, in theory, a potential investment claim by an investor could be caught within the scope of the arbitration clause. In such a case, with a view to avoiding potential controversies on the scope of the arbitration clause, if the investor has a preference for the framework in the relevant investment treaty, it may wish to include a carve-out in the arbitration clause.
The above options, however, have potential drawbacks. For example, combining arbitration with final expert determination can give rise to protracted (and expensive) disputes if the issues that are to be resolved by these different mechanisms are not clearly delimited. It does not help that predicting the nature of all the disputes that might arise under an agreement (particularly if it is a long-lasting one) is notoriously difficult.
In addition to the fragmentation of disputes, carving out issues of invalidity regarding intellectual property rights from the scope of an arbitration clause is likely to cripple the functioning of an arbitral tribunal if issues of invalidity arise: it is likely that the arbitral tribunal would have to suspend the conduct of the proceedings until a decision on the validity of the asserted rights is made by a national court or relevant public body. As a result, if the parties want an erga omnes declaration on the validity of the rights in issue, they would probably be better off by resorting only to litigation.
On balance, although in some circumstances limiting the scope of an arbitration clause might provide some benefits, unless the parties are in a position to undertake a painstaking drafting exercise, arbitration clauses with a limited scope and carve-out clauses are better avoided. When the stakes are high, it is often better to err on the side of caution.