Centrotrade Minerals v. Hindustan Copper: how the Supreme Court settled the validity of two-tier arbitration in India

The Supreme Court's judgment of 15 December 2016 in Centrotrade Minerals & Metal Inc. v. Hindustan Copper Ltd, reported as (2017) 2 SCC 228, closed a decade-long doctrinal impasse on the question of two-tier or appellate arbitration in India. A three-judge bench of Madan B. Lokur, J. (authoring), R.K. Agrawal, J. and Dr D.Y. Chandrachud, J. held that nothing in the Arbitration and Conciliation Act 1996 forbids parties from contractually agreeing to a two-tier arbitral structure in which an appellate second-tier proceeding sits in review of a first-tier award. The judgment is the operative authority on the permissibility of two-tier clauses; it reaffirms party autonomy as the lodestar of the 1996 Act; and it is the foundation on which the subsequent Centrotrade III enforcement decision of 2 June 2020 — which held the resulting foreign appellate award enforceable under Part II of the 1996 Act — was built.

The judgment is doctrinally consequential at two levels. At the level of arbitral architecture, it confirms that the 1996 Act, like its UNCITRAL Model Law parent, is a permissive statute that admits a wide range of contractual designs unless something in the statutory text positively prohibits them. At the level of practice, it validates a drafting option — the two-tier clause with a domestic first tier and an international appellate tier — that international counterparties to Indian contracts had been seeking since the late 1990s. After Centrotrade, the two-tier clause is a recognised feature of cross-border contracting with India-resident or India-incorporated parties.

The statutory and contractual architecture

To see what Centrotrade decided, the contract and the statutory backdrop must be set out.

The Arbitration and Conciliation Act 1996 is divided into two principal parts. Part I, in Sections 2 to 43, governs arbitrations seated in India — covering the arbitration agreement (Section 7), commencement, conduct of proceedings, the award, recourse against the award by way of Section 34 set-aside, and the enforcement of domestic awards as decrees under Section 36. Part II, in Sections 44 to 60, gives effect to the New York Convention 1958 and the Geneva Convention 1927 — providing for the enforcement in India of "foreign awards" rendered in convention-country seats. The architecture mirrors the international ordering: domestic arbitrations and domestic awards in Part I; foreign-seated arbitrations and foreign awards in Part II.

The statutory question that Centrotrade posed was structural. The 1996 Act contemplates that an arbitration produces a single award — Section 31 speaks of "the arbitral award", Section 34 of "an arbitral award", Section 35 of the finality of "an arbitral award", and Section 48 of "the foreign award" sought to be enforced. The statute does not, on its face, anticipate or address a two-tier arbitral structure in which a first-tier award is itself subject to appellate review by a second-tier arbitral tribunal sitting under a different institutional umbrella and at a different seat. The interpretive question was whether the statute's textual orientation to a single award foreclosed the two-tier clause by implication, or whether the absence of express prohibition left the two-tier clause within the residual space of party autonomy.

The contractual clause in Centrotrade was a sophisticated two-tier design. Centrotrade Minerals & Metal Inc. — a US-incorporated trader in non-ferrous metals — had contracted with Hindustan Copper Ltd — a Government of India undertaking — for the supply of copper concentrate. The contract provided, in its arbitration clause, that disputes would first be referred to arbitration in India under the rules of the Indian Council of Arbitration. If either party was dissatisfied with the first-tier award, the dispute would, on notice given within a contractually prescribed period, be referred to a second-tier appellate arbitration conducted in London under the Rules of the International Chamber of Commerce. The contract was clear that the second-tier ICC award would be final and binding.

The disputes arose, the first-tier ICA arbitration produced a "nil" award in favour of Hindustan Copper, and Centrotrade invoked the second-tier ICC arbitration in London. The ICC tribunal, on the merits, made an award in favour of Centrotrade in a substantial sum. Centrotrade sought enforcement of the ICC award in India; Hindustan Copper resisted on the ground that the two-tier clause was itself impermissible under the 1996 Act, and that the second-tier award could not be given effect.

The factual matrix and the 2006 two-judge split

The procedural history of Centrotrade is itself part of the doctrinal record. The matter reached the Supreme Court in 2006, before a two-judge bench of S.B. Sinha, J. and Tarun Chatterjee, J. The bench split. Tarun Chatterjee, J. held that the two-tier clause was valid under the 1996 Act — that party autonomy was the central principle, that the statute did not prohibit appellate arbitration, and that the second-tier ICC award was capable of enforcement as a foreign award under Part II. S.B. Sinha, J. disagreed — holding that the architecture of the 1996 Act did not contemplate two-tier arbitration, that the second-tier proceeding amounted to a private appellate forum that the statute did not authorise, and that the second-tier award could not stand.

The two-judge split necessitated a reference to a larger bench. The matter, after the usual delays inherent in the constitution of larger benches and the rotation of judges, was eventually placed before the three-judge bench that delivered the December 2016 judgment. The decade-long gap between the 2006 split and the 2016 resolution is itself a comment on the cost of doctrinal indeterminacy in Indian commercial arbitration: a sophisticated cross-border contract — with substantial sums in dispute, with parties from two jurisdictions, with a contractual mechanism designed precisely to provide certainty — was left in legal limbo for ten years while the question of the clause's validity awaited authoritative determination.

The three-judge bench in 2016 was constituted in the express terms of the reference: to decide whether the two-tier arbitration clause was valid and permissible under the 1996 Act. The bench answered the question in the affirmative, broadly along the lines of Tarun Chatterjee, J.'s 2006 view, and held that the separate question of the enforceability of the second-tier ICC award was to be listed for later hearing. That separate question was decided four years later, on 2 June 2020, in what has come to be called Centrotrade III — holding the foreign ICC award enforceable under Part II.

The Court's reasoning

The 2016 bench's reasoning rested on three connected propositions.

Party autonomy as the foundational principle of the 1996 Act. The bench emphasised that the Arbitration and Conciliation Act 1996, modelled on the UNCITRAL Model Law, is a permissive statute that gives effect to the contractual choices of the parties subject only to express statutory prohibitions and the requirements of public policy. Where the statute is silent — where it neither expressly authorises nor expressly forbids a particular contractual design — the default rule is that the parties' chosen design is given effect. The architecture is consonant with the international consensus on commercial arbitration, where party autonomy is the central organising principle and the courts intervene only to the limited extent that the statute permits.

The absence of express prohibition on two-tier arbitration. The bench engaged the textual orientation of the 1996 Act to a single award and held that the textual orientation did not amount to a prohibition by implication. The references in Sections 31, 34, 35 and 48 to "the arbitral award" or "an arbitral award" are descriptive of the ordinary case; they do not foreclose, by negative implication, contractual arrangements in which the parties have agreed that the award of a first-tier tribunal will be subject to appellate review by a second-tier tribunal. The two-tier structure operates in the contractual space: the parties' agreement defines what constitutes the binding "arbitral award" for purposes of the statute. Where the parties have agreed that the binding award is the second-tier award, the statute gives effect to that contractual definition.

The conceptual coherence of appellate arbitration. The bench addressed the conceptual objection — pressed by S.B. Sinha, J. in the 2006 split — that an appellate arbitration amounts to a private appellate forum that the legislature has not authorised. The bench held that the objection rests on an inapt analogy. The "appellate" character of the second-tier proceeding is not a judicial appeal in any constitutional sense; it is simply a further round of arbitration that the parties have contractually agreed to. The proceeding is consensual, the tribunal is constituted by party agreement, and the second-tier award is the product of the same arbitral process — under a different institutional umbrella, at a different seat — that the statute generally authorises. There is no constitutional or statutory disability that prevents the parties from designing a two-stage arbitral process where they regard a single-stage process as commercially inadequate.

The bench was also careful to record what it was not deciding. The validity of the clause did not, of itself, decide the enforceability of the foreign appellate award. The enforceability question — whether the second-tier ICC award delivered in London qualified as a "foreign award" under Part II of the 1996 Act, and whether enforcement was subject to any of the Section 48 grounds of refusal — was separate. The bench listed that question for separate hearing.

The doctrinal contribution

Centrotrade contributed three propositions to Indian arbitration doctrine.

It established that the Arbitration and Conciliation Act 1996 admits two-tier or appellate arbitration as a permissible contractual design. The position is now uncontroversial: parties may, by contractual agreement, structure their dispute resolution as a first-tier arbitration followed by an appellate second-tier arbitration, with the second-tier award being the binding outcome for purposes of recognition and enforcement.

It reaffirmed party autonomy as the lodestar of the 1996 Act. The reaffirmation is structurally important because it operates as an interpretive presumption: where the statute is silent on a particular contractual design, the default is to give effect to the parties' agreement rather than to read in a prohibition by implication. The presumption has been deployed in the post-Centrotrade arc to validate other previously contested designs — including the foreign-seated arbitration between Indian parties endorsed in PASL Wind Solutions Pvt Ltd v. GE Power Conversion India Pvt Ltd (2021) 7 SCC 1.

It resolved the cost of the 2006 indeterminacy. The decade-long limbo between the Sinha–Chatterjee split and the three-judge resolution is itself a doctrinal lesson: the cost of unresolved Constitution-Bench-level questions in commercial arbitration is borne by international parties and by the institutional reputation of the Indian arbitration regime. The 2016 resolution closed the question and removed the residual uncertainty that had constrained the use of two-tier designs in India-anchored contracts.

What the judgment did not decide

Three limits should be flagged.

First, the judgment in 2016 decided only the validity of the two-tier clause; it did not decide the enforceability of the second-tier ICC award. The enforceability question — whether the London-seated ICC award qualified as a "foreign award" under Part II of the 1996 Act and was enforceable in India — was reserved for later hearing. It was answered, in the affirmative, by the Supreme Court on 2 June 2020 in what is colloquially called Centrotrade III. The 2020 decision is what completed the doctrinal saga and gave the validity holding of 2016 its practical consequence.

Second, the judgment did not decide the position where the contract is structured as a two-tier arbitration with both tiers seated in India. The contractual design before the bench was a domestic first tier and a foreign-seated appellate tier; the application of the Centrotrade validity principle to a purely domestic two-tier structure — for instance, where parties contract for an ad hoc first-tier arbitration in Delhi and an institutional second-tier arbitration in Mumbai — was not engaged on the facts. The structural logic of the bench's reasoning would support the validity of such a design, but the question awaits a separate, on-facts decision.

Third, the judgment did not address the question of judicial recourse against the first-tier award where the parties' contract requires them to elect between proceeding to the second tier and seeking Section 34 set-aside. The interaction between the contractual appellate route and the statutory set-aside route — whether a party can pursue both, whether election of one waives the other, whether the second-tier appellate proceeding renders the first-tier award immune to Section 34 challenge — is a residual practitioner question that subsequent decisions will need to refine.

The doctrinal arc

Centrotrade sits in a doctrinal arc that traces the Indian Supreme Court's progressive embrace of party autonomy as the central principle of the 1996 Act.

Bhatia International v. Bulk Trading SA (2002) 4 SCC 105 had read Part I as applying to foreign-seated arbitrations unless excluded — an expansive territorial reach that had constrained the parties' freedom to design their dispute resolution. Bharat Aluminium Co. v. Kaiser Aluminium Technical Services Inc. (2012) 9 SCC 552 — the five-judge Constitution Bench commonly called BALCO — overruled Bhatia International and confined Part I to India-seated arbitrations, restoring the territorial separation that the 1996 Act had contemplated.

Centrotrade (2016) is the next major step in the arc. By validating the two-tier clause, the bench reinforced the architectural proposition that the 1996 Act is a permissive statute that admits a wide range of contractual designs unless something in the text positively prohibits them.

Indus Mobile Distribution Pvt Ltd v. Datawind Innovations Pvt Ltd (2017) 7 SCC 678 — decided four months later — extended the party-autonomy framework to the seat-as-exclusive-jurisdiction question, holding that the designated seat operates as an exclusive jurisdiction clause that displaces the residual jurisdiction of courts that would otherwise have competence under the Code of Civil Procedure 1908.

PASL Wind Solutions Pvt Ltd v. GE Power Conversion India Pvt Ltd (2021) 7 SCC 1 — the further reach — held that two Indian parties may validly designate a foreign seat for their arbitration, and that the resulting award is a "foreign award" under Part II. The decision is the high-water mark of the party-autonomy reading.

Centrotrade III (2 June 2020) — the enforcement coda to the 2016 validity decision — held the London-seated ICC second-tier award enforceable under Part II of the 1996 Act, completing the saga and removing the residual practical doubt about two-tier designs.

What practitioners take from the judgment today

For transactional counsel structuring international contracts with Indian counterparties, Centrotrade validates a recognised drafting option. Parties who require an appellate review mechanism — typically because the substantive stakes are high, or because the first-tier institution is ad hoc or domestic and the second-tier institution offers procedural reassurances valued by the international party — may design a two-tier clause with confidence that the structure will be upheld in Indian courts. The drafting discipline is to make the structure express, to define the time and notice requirements for invoking the second tier, to specify the seat and institutional rules of each tier, and to record clearly that the second-tier award is the binding outcome.

For arbitration counsel resisting enforcement of a two-tier award, the Centrotrade validity holding closes the structural objection. The resisting party must, after Centrotrade, contest enforcement on the merits of the Section 48 grounds — public policy, due process, scope of the arbitration agreement — rather than on the structural ground that the two-tier design is impermissible. The strategic burden has shifted.

For Indian counsel advising on domestic two-tier designs — where both tiers sit in India — the Centrotrade framework supports validity but the on-facts question has not been authoritatively answered. The cautious drafting course is to anchor the design in the Centrotrade reasoning, to record the party-autonomy basis expressly, and to assume that an attempted challenge to the structure will engage the Centrotrade validity principles in the parties' favour.

For arbitral institutions developing institutional rules for appellate proceedings — including the International Chamber of Commerce, the Singapore International Arbitration Centre, the London Court of International Arbitration, and Indian institutions such as the Mumbai Centre for International Arbitration and the recently launched Chandigarh International Arbitration Centre — the Centrotrade doctrine is the validating authority for offering appellate arbitration as a service. The institutional market has, in the post-Centrotrade period, developed precisely such offerings.

Related editorial pieces

• Bharat Aluminium Co. v. Kaiser Aluminium: the territorial reset that confined Part I to India-seated arbitrations
• PASL Wind Solutions v. GE Power Conversion: two Indian parties, a foreign seat, and a Part II award
• Hindustan Construction Company v. Union of India: how Section 87 fell and the no-automatic-stay regime was restored
• Osterreichischer Lloyd: Section 9 interim relief in support of foreign-seated arbitration