Group of Companies Doctrine: Judgments Implicitly Overruled in Cox & Kings

[Varun Srinivasan is an advocate practising in New Delhi and is a junior in the chambers of Mr. Gaurav Pachnanda SA]

The five-judge decision of the Supreme Court in Cox & Kings Ltd v. SAP India Pvt Ltd 2023 INSC 1051 has already been well summarized here and some of its implications are discussed here. The focus of this post is to recognize some of Cox & Kings’ more subtle contributions by implicitly erasing the effect of a few judgments that had been troubling India’s arbitration landscape. Two of these are worthy of discussion.  

Arupri Logistics

On 24 July 2023, a single judge of the Delhi High Court handed down the judgment in Arupri Logistics Pvt Ltd v. Vilas Gupta 2023 SCC OnLine 4297. In Arupri, the Court held that arbitral tribunals could not implead a non-signatory to the arbitration agreement based on group of companies doctrine (para 91). There were two threads of reasoning offered in Arupri to justify this conclusion:

First, although joinder of non-signatories through group of companies doctrine had already been recognized by the line of judgments beginning with Chloro Controls India (P) Ltd v. Severn Trent Water Purification Inc (2013), this did not empower the arbitral tribunal to implead non-signatories. That was because, according to Chloro Controls, the group of companies doctrine was read into the phrase “any party or persons claiming through or under” appearing in sections 8 and 45 of the Arbitration and Conciliation Act, 1996 (“A&C Act”). These provisions empowered a court, and not arbitral tribunals, to exercise jurisdiction over such parties (para 91).

Second, there is no other source granting arbitral tribunals the power to implead. It could not be traced to any statutory provision in the A&C Act. Arbitral tribunals, being creatures of contract, did not exercise inherent powers of impleading like civil courts either (paras 67, 68, 93, 95).

At first blush, the judgment in Arupri appeared puzzling in its ultimate effect. It was not immediately clear whether it meant that arbitral tribunals could not implead non-signatories even if they were otherwise parties to the arbitration agreement by conduct (a more extreme conclusion). Some paragraphs could certainly suggest so (para 93). A more conservative reading would imply that Arupri’s ratio was only that non-signatories could not, by law, be parties to arbitration agreements in the first place (para 94)—which would make the distinction from Chloro Controls weaker. Either way, Arupri led to a flood of applications being filed under section 37 of the A&C Act, or grounds being taken in section 34 in a frantic bid to stall arbitrations by unwilling respondents.

Both threads of reasoning offered in Arupri are unlikely to survive Cox & Kings. The Supreme Court has now made one crucial clarification: although the existence of the group of companies doctrine is affirmed, its legal basis is not the phrase “any party or persons claiming through or under” appearing in section 8 and section 45. That much was wrongly held in Chloro Controls. Instead, the Supreme Court read this doctrine as being inherent within a reading of section 7 (arbitration agreement) and the definition of “parties” in section 2(1)(h) A&C Act (paras 143 to 149). This means that the distinction from Chloro Controls and that was attempted in Arupri has been erased. By reading it into section 7 and section 2(1)(h), the group of companies doctrine exists to be invoked equally by an arbitral tribunal as it could by a court.

Further, Chloro Controls makes it clear that to answer the question of whether a non-signatory can be joined to an arbitration proceeding, the relevant question is not “is there a statutory power to implead?”, but rather “does the arbitral tribunal have jurisdiction over a party?” Several paragraphs of the Cox & Kings judgment give the impression that, once it is established that a court or tribunal has jurisdiction over a party, the power to join or implead is a natural consequence of the exercise of that jurisdiction (paras 122, 149, 163). This is logical—it would be odd for a tribunal to have jurisdiction over a party, but not the power to implead it, if that joinder was necessary for proper adjudication. With respect, Arupriwas wrong to focus its analysis on a statutory power to implead. Cox and Kings has corrected this error and put an end to the six month legacy of Arupri.

Magic Eye Developers

The implicit overruling of some judgments is more obvious. On 12 May 2023, a two-judge bench of the Supreme Court handed down the judgment in MagicEye Developers Pvt v. Green Edge Infrastructure Private Limited 2023 SCC OnLine SC 620. The Court had held that any issue concerning identifying who the parties to an arbitration agreement are must be conclusively answered by the referral court under section 11 of the A&C Act. Courts could not leave the issue open to be determined by the arbitral tribunal at all (para 16). This was a strange conclusion—despite section 11(6A) of the A&C Act being categorical in stating that any examination as to the existence of an arbitration agreement could only be “prima facie”, MagicEye was equally categorical in ensuring no room was left to leave this determination to the arbitral tribunal. It would be bound by the findings of the referral court.

In Cox & Kings, this judgment has been impliedly overruled. The five-judge bench held that the referral court should only make a prima facie examination under section 11 and, where that would be complex, the court should necessarily leave it to the arbitral tribunal to determine whether an arbitration agreement exists between two parties (para 164). In doing so, the Court explicitly departed from an earlier decision in Deutsche Post Bank Home Finance v. Taduri Sridhar (2011) 11 SCC 375 holding that it was no longer relevant after Chloro Controls (para 163). It appears to have missed noting that even post-Chloro Controls, the judgment in MagicEye had been perpetuating the exact same mischief. Either way, MagicEye is no longer good law, for good measure.


At the time of the reference to the five-judge bench, it was often asked if the judicial expense of a constitution bench was necessary for examining the group of companies doctrine. It had, after all, recently enjoyed a spate of three-judge decisions that analysed the doctrine in detail. It is now clear that while affirming the doctrine itself is a surprise to no one, the exercise of reformulating its legal basis within section 2(1)(h) read with section 7 of the A&C Act was a necessary course correction. Chloro Controls had unfortunately left room for distinctions like Arupri to grow, and the five-judge bench will effectively stamp out such distinctions in the future.

Varun Srinivasan

About the author

Add comment

Top Posts & Pages


Recent Comments


web analytics

Social Media