Gatekeepers Without A Key: How Cox and Kings II left non-signatories stranded at the gates of Sections 9 and 11 of the Arbitration and Conciliation Act, 1996
- 3 hours ago
- 9 min read
By- Anchal Kanthed[1]
The judgment of the Cox and Kings Ltd. v. SAP India Pvt. Ltd.,[2] settled around 30 years of doctrinal confusion regarding the Group of Companies Doctrine (hereinafter “GOCD”) by affirming it as a consent-based principle enshrined in Sections 2(1)(h) and 7 of the Arbitration and Conciliation Act, 1996 (hereinafter “A&C Act”). The judgment adjudged the “claiming through or under” in Chloro Controls[3] as erroneous, confirmed that non-signatories can be parties in their own right and established a five-factor test for invocation of arbitration. Though this is a remarkable judgment, it has left open two loopholes. First, at the Section 11 stage, directing courts to conduct only a prima facie review of the non-signatory party, leaving the full determination at the behest of the tribunal, making it a double-adjudication, which increases procedural costs. Secondly, regarding Section 9, imposing a condition on the non-signatory for a prior tribunal determination for an interim relief. The paradox is that a non-signatory cannot invoke Section 9 until the tribunal has ruled, as Section 9(3) bars the court. This article examines these loopholes and argues that both can be resolved through purposive judicial interpretation, without any legislative amendment.
Introduction
The basis of arbitration is consent between the parties. However, it becomes complicated when questions like whose consent and how to identify consent arise. The law answers these questions through the doctrine of GOCD, which means that when a non-signatory party has participated so substantially in the negotiation, performance or termination of a contract that circumstances reveal a mutual intention to be bound, then that party can be drawn into the arbitration even without signing the agreement.
The judgment in Cox and Kings Ltd. v. SAP India Pvt. Ltd.[4] [hereinafter “Cox & Kings II”] resolved long-standing disputes on the validity of the doctrine, corrected errors from Chloro Controls India (P) Ltd. v. Severn Trent Water Purification Inc.[5] (hereinafter “Chloro Controls”) and provided a clear framework for the application of the doctrine. But the judgments left two fault lines open, one at the Section 11 referral stage and second around Section 9 interim relief, which collectively undermine the very purpose of the doctrine of GOCD.
I. The Framework: What Cox & Kings II Settled
Non-signatories were not included in arbitration before Chloro Controls. The cases like Sukanya Holdings (P) Ltd. v. Jayesh H. Pandya[6] and Indowind Energy Ltd. v. Wescare (I) Ltd.[7] held that a written agreement was required to initiate arbitration proceedings/ arbitration was possible amongst the signatories of the agreement, as was mentioned in Section 7. Chloro Controls changed this position by introducing a phrase, “any person claiming through or under” as given under Section 45 of the A&C Act, including non-signatories to the agreement in a derivative capacity. But the judgment created a doctrinal error by treating this derivative capacity phrase as the very basis for the doctrine of GOCD itself. It was Cox & Kings II that later reversed this position, holding that a non-signatory joined through GOCD is bound as a principal party in its own right, by its own consent and not by derivation.
Cox & Kings II corrected this error by holding that a conjoint reading of Sections 2(1)(h) and 7 of the A&C Act accommodates non-signatories as parties without requiring their signatures. A written agreement under Section 7(4)(b) ensures that there is a record of consent, but it does not limit the forms consent may take. Consent to arbitrate can be gathered from conduct, participation and surrounding circumstances. Moreover, the Court also listed five factors, citing ONGC v. Discovery Enterprises Pvt. Ltd.,[8] to determine consent: (i) mutual intention between the parties; (ii) the relation between signatory and non-signatory; (iii) commonality of the subject matter; (iv) composite nature of the transaction; and (v) performance of the contract. Mere membership of a corporate group was not enough on its own. The Court held that the concept of ‘single economic unit’ could not be used on its own to justify pulling a non-signatory into arbitration under GOCD. Moreover, the Court also explained that Canara Bank[9] had not relied solely on that reasoning and to the extent it had been interpreted otherwise. This interpretation was expressly overruled. The Court made a differentiation between the doctrine of GOCD and veil piercing, while the former is consent-based and preserves the separateness of the parties, the latter is equity based that overrides it. Now the doctrine has a principle-backed foundation that it previously lacked.
The chart below places the doctrine of GOCD post Cox & Kings II within a broader spectrum of doctrines that extend arbitration to non-signatories, illustrating the focus of the judgment on consent as an anchor and its rejection of veil-piercing as a basis, representing a non-technical/principled position.
Chart 1: Doctrinal Spectrum from fully consensual (agency) to fully non-consensual (veil piercing). GOCD sits at the midpoint by preserving corporate identity and finding consent through conduct. Cox and Kings II confirmed that it cannot migrate rightwards into the territory of veil-piercing.
II. Section 11 Problem: Prima Facie Review And Double Adjudication
If the agreed procedure to appoint arbitrators fails, Section 11 of the A&C Act empowers the courts to appoint arbitrators. The scope of examination at this stage has been narrowed down through judgments like Dakshin Haryana Bijli Vitran Nigam Ltd. v. Navigant Technologies Pvt. Ltd.[10] and NTPC v. SPML Infra Ltd.[11] The Supreme Court consistently held that the examination is limited to only whether a valid arbitration agreement exists, on a prima facie basis and the rest of everything is examined by the tribunal under the kompetenz-kompetenz principle under Section 16 of the A&C Act.
Cox and Kings II applied this to the doctrine of GOCD and held that the question of whether a non-signatory can be impleaded is to be determined by the tribunal and not the referral court. At the Section 11 stage, the court must only determine that the non-signatory is not a complete stranger to the transaction. This understating unknowingly creates a double-layer procedure that Cox and Kings II did not fully acknowledge.
This is because when either of the parties seeks to add a non-signatory at Section 11 stage, the court makes a prima faciefinding and constitutes a tribunal based on the agreement between the signatory parties. The non-signatory is now involved in the arbitration, but with an unresolved status. Then, before any substantial hearing, the tribunal must take up the joinder question as a preliminary issue, which requires both signatory parties to brief the facts about the involvement of the non-signatory. If tribunal rejects the joinder, the non-signatory question is litigated twice and if it accepts the joinder, the preliminary hearing is an additional cost without any benefit. Either way, the outcome results in more litigation than necessary, defeating the purpose of the doctrine of GOCD. The chart depicted below explains this double-adjudication loop with a proposed solution to this problem.
Chart 2, which highlights the double-adjudication loop, indicates that two-stage litigation occurs regardless of the tribunal’s prima facie finding on joinder. The proposed solution is that a prima facie review under Section 11 creates a baseline for the tribunal that can be built upon, rather than a restart.
The Bombay HC in Cardinal Energy and Infra Structure Pvt. Ltd. v. Subramanya Construction and Development Co. Ltd.,[12] held that a tribunal may implead a non-signatory even without a prior court order, offering some flexibility. However, it does not address the fundamental question of the evidentiary threshold at Section 11 stage.
In author’s view, a better approach is a ‘structured prima facie’ review. Rather than just answering whether arbitration agreement exists or not, the Section 11 court should also assess whether at least two of the five factors to determine mutual consent exist, specifically, whether there is a direct relationship between signatory & non-signatory and whether the subject matter of the dispute is common to both. This becomes consistent with J. Narasimha’s concurring observations in Cox & Kings II that existence of arbitration agreement and identification of parties to the agreement are “inextricably connected.”[13] Thus, logically, if parties are inextricably connected to the existence of agreement, the Court’s prima facie review cannot leave the question of party-identification untouched.
III. Section 9 Paradox: Interim Relief and the Frozen Window
Cox & Kings II held that a non-signatory, found to be a party by the tribunal, may seek an interim relief under Section 9 of the A&C Act. This observation is based on the assumption that there will be a prior tribunal determination. The reason was to fix the anomaly created by the ‘claiming through or under’ framework under Chloro Controls where a non-signatory could be referred to arbitration under Sections 8 or 45, but could not access Section 9, as it applied only to the “parties”. By re-categorising the non-signatory as a party in its own right under Section 2(1)(h), the Court made Section 9 available to the non-signatory. However, the language created a new and arguably worse problem.
Section 9 is designed to provide interim protection of assets and subject matter during the period of pre-constitution of an arbitral tribunal because at this stage, the risk of dissipation is at its highest. The utility of this provision is undermined if a party is required to wait for the constitution of tribunal and to decide the issue of jurisdictional competence before being entitled to invoke Section 9 relief, since by that stage the very window of protection that the provision contemplates has already lapsed. This difficulty increases due to the fact that Section 9(3) bars the jurisdiction of the Court to entertain a Section 9 application once the tribunal has been constituted except the relief under Section 17 would be rendered inefficacious. The result is a jurisdiction paradox: the determination of party status is condition precedent to invocation of Section 9 by non-signatory, which can only render post-constitution of tribunal by which time Section 9(3) has already divested the Court of jurisdiction necessary to grant such relief. The non-signatory, thus, is trapped in an irresolvable procedural loop, wherein the determination is required invoking of Section 9 jurisdiction simultaneously operating to extinguish it.
This is illustrated below:
Stage | Position of non-signatory |
Pre-tribunal | Since, the party status is not determined, it cannot seek Section 9 relief. |
Post-constitution of tribunal | Tribunal takes up preliminary issue of joinder |
Tribunal holds the non-signatory as a party | Theoretically, Section 9 relief can be sought but it is barred by Section 9(3) |
Result | Procedural limbo during a most critical stage |
A partial solution to this problem is Section 17, wherein the tribunal can order interim protection after finding the non-signatory to be a party and such orders are enforceable as court decrees under Section 17(2). But the issue is: first, Section 17 does not cover the pre-tribunal window and second, since post Arbitration and Conciliation (Amendment) Act, 2015, orders under Section 17 are in parity with orders under Section 9,[14] a tribunal’s jurisdiction under Section 17 extends only to parties to the arbitration agreement, so relief against non-signatory third parties, such as financial institutions, may still require recourse to Section 9.
Moreover, Emergency arbitration, as upheld in Amazon.com NV Investment Holdings LLC v. Future Retail Limited,[15]provides another solution, but the problem is that it applies only in institutional arbitrations with emergency procedure provisions.
In the author’s view, Cox & Kings II held that Section 2(1)(h) includes non-signatories as parties. Moreover, Section 11 court makes a prima facie determination of the party status of a non-signatory. Thus, a non-signatory who has received a prima facie recognition of party-status from the Section 11 court has, in effect, been provisionally identified as a party by a court. Now, there is no reason why that provisional recognition should not be sufficient to satisfy the “party” requirement under Section 9 during the pre-tribunal constitution. The conditioning language of the judgment should be read to establish the party-status and not as a jurisdictional bar to access Section 9 for interim relief.
Conclusion
Cox & Kings II is a landmark judgment that provided the doctrine of GOCD the statutory recognition it had always required. The five-factor test, the rejection of veil-piercing as a basis for invoking the doctrine of GOCD and the correction of erroneous reliance on “claiming through or under” in Chloro Controls, these are durable contributions to arbitration jurisprudence. However, the judgment left two structural loopholes that, in combination, jeopardise the purpose of the doctrine.
Section 11 gap produces double adjudication on non-signatory joinder and Section 9 gap deprives non-signatories of interim protection when they need it most. This article argues that both can be addressed without legislative amendment through a structured prima facie review at Section 11 and a purposive reading of Section 2(1)(h) for Section 9. This would restore the doctrinal coherence that Cox & Kings II sought to achieve. Until a subsequent judgment fills these gaps, contracting parties would do well to name all intended arbitration participants as signatories from the outset, because the doctrine of GOCD, for all its post-Cox & Kings II clarity, still leaves non-signatories requiring recognition from the court without the procedural key to make use of it.
[1] Advocate.
[2] Cox and Kings Ltd. v. SAP India Pvt. Ltd., (2023) INSC 1051.
[3] Chloro Controls India (P) Ltd v. Severn Trent Water Purification Inc., (2013) 1 SCC 641.
[4] Cox and Kings Ltd. v. SAP India Pvt. Ltd., (2023) INSC 1051.
[5] Chloro Controls India (P) Ltd v. Severn Trent Water Purification Inc., (2013) 1 SCC 641.
[6] Sukanya Holdings (P) Ltd. v. Jayesh H. Pandya, (2003) 5 SCC 531.
[7] Indowind Energy Ltd. v. Wescare (I) Ltd., (2010) 5 SCC 306.
[8] ONGC v. Discovery Enterprises Pvt. Ltd., (2022) 8 SCC 42.
[9] MTNL v. Canara Bank, (2020) 12 SCC 767.
[10] Dakshin Haryana Bijli Vitran Nigam Ltd. v. Navigant Technologies Pvt. Ltd., (2021) 7 SCC 657.
[11] NTPC v. SPML Infra Ltd., (2023) 9 SCC 385.
[12] Cardinal Energy and Infra Structure Pvt. Ltd. v. Subramanya Construction and Development Co. Ltd., MANU/MH/2164/2024
[13] Cox and Kings Ltd. v. SAP India Pvt. Ltd., (2023) INSC 1051. The relevant extract from J. Narasimha’s concurring opinion is quoted as under:
“15(iv). ‘Party’ is defined in Section 2(1)(h) as “a party to an arbitration agreement”. The determination of the arbitration agreement and its parties are inextricably connected with one another, their existence is based on the written agreement.” (emphasis supplied)
[14] Amazon.com NV Investment Holdings LLC v. Future Retail Limited, (2022) 1 SCC 209.
[15] Amazon.com NV Investment Holdings LLC v. Future Retail Limited, (2022) 1 SCC 209.