The Delhi High Court (Court) recently rendered a decision in GMR v. Doosan (“GMR”) on two critical points related to Indian arbitration– a) joinder of non-signatories to arbitration and b) whether two Indian parties can choose a foreign seat. Both issues have had conflicting decisions from courts leading to confusion in jurisprudence. Did the Court’s decision in GMR help clarify the law? In our view, no. Instead, it addled precedent by issuing a tenuously reasoned decision.
GMR Chhattisgarh (“GCEL”) and Doosan India (“Doosan”) entered into 3 EPC contracts in 2010 (“EPC Contracts”) which provided for SIAC arbitrations in Singapore.
GMR Infrastructure Ltd. (“GIL”) furnished a corporate guarantee to Doosan, on behalf of GCEL in 2013 (“Corporate Guarantee”) containing an arbitration clause (SIAC administered, in Singapore).
Two MOUs were executed between Doosan and GMR Energy (“GE”) in 2015 where GE agreed to repay Doosan in installments for GCEL’s liability under the EPC contracts. The MOUs did not contain arbitration clauses and were terminated before commencing arbitration.
Doosan invoked SIAC arbitration under the EPC contracts and the Corporate Guarantee making GCEL and GIL parties. Doosan sought GE’s joinder based on the MOUs and theories of joinder of non-parties including alter ego, group companies’ doctrine, and common directors, seeking repayment jointly and severally from GCEL, GIL and GE. In response, GE filed a suit seeking a permanent injunction against Doosan from continuing arbitration since GE was not a party to the arbitration agreement in the EPC Contracts and the Corporate Guarantee. The Court stayed the constitution of the SIAC tribunal. Doosan sought vacation of this order and applied under the Arbitration and Conciliation Act, 1996 (“AA”) to compel GE to participate in the arbitration. In this hearing, GE’s motion for injunction and Doosan’s motion for vacation and arbitral reference were heard together and decided.
Was GE’s joinder justified?
The Court examined whether Doosan had made out a prima facie case in its notice of arbitration justifying GE’s joinder. In accepting GE’s joinder was proper, it found as follows:
- That GE, GCEL, and GIL freely “co-mingled” funds and were family run.
Family run businesses are common amongst Indian group companies. There was not much evidence of funds being co-mingled other than the corporate guarantee and MOUs. This, by itself, does not validate piercing the corporate veil to subject a party to the arbitration.
- That the entities had common directors.
In India, common directorship or even the holding company’s control in the appointment of directors in the subsidiary does not remove the juristic and legal independence of such subsidiary. The Supreme Court’s (“SC”) landmark decision in Vodafone International Holdings B.V. v India espouses this principle and holds that directors of subsidiaries have a separate responsibility to the subsidiary. Vodafone observed that in liquidation, such subsidiary’s assets would go to the liquidator, not the parent. The Court has not taken this into account at all.
- That there was no “corporate formality” between the various group companies.
However, it did not explain the absence of “corporate formality” when overlooking distinct corporate personalities. In fact, there are some cases that caution against the overzealous use of this power of piercing the veil (see Balwant Rai Saluja v. Air India). In India, while principles of public interest and single economic entity are accepted to pierce the veil, for the argument of agency, alter ego and control in a parent-subsidiary relationship, a high degree of control needs to be shown (see New Horizons v India).
- The parent-subsidiary relationship and that at the material time GCEL was GE’s 100% subsidiary.
In transactions that are admittedly “sham” where entities are merely created as funnels, the corporate veil should be pierced but not in other circumstances (Balwant Rai). In this case, while GCEL was an SPV, it was not camouflaged. Creation of SPVs is routine for infrastructure projects. There was no written agreement binding GE to arbitration, the MOUs had terminated, and the 100% subsidiary relationship also did not exist.
While this case may be fit for piercing the corporate veil on merits, there was not enough to bind GE to arbitration. This matter could have been resolved by Doosan filing a civil suit against all the entities. While extending the arbitration to GIL was appropriate even in the context of the well-worn Chloro Controls regime, it is hard to find any express or implied term by which GE agreed to submit to arbitration. GE may or may not have breached its commitment to pay, but it made no commitment to arbitrate.
The Court overstated the factors purportedly evidencing GE’s intention to arbitrate. Illustratively, one of the clauses in the contracts stated that parties could consolidate disputes in one arbitration under the various contracts; but such consolidation was only permissible if all parties consented. Such consent was however absent. The Court also failed to account for lack of an arbitration clause in the MOUs.
GE relied on a clause in the contracts which stated that parties were entering into this agreement on their own behalf and not on behalf of, amongst others, shareholders and agents and that neither party shall have recourse to them including through piercing the corporate veil. The Court ignored the clear mandate of this clause.
Can two Indian parties choose a foreign seat?
There has been judicial divergence on whether two Indian parties can choose a foreign seat. The GMR judgment further muddies the waters. Relying on SC decisions in Atlas Exports and Sassan Power, the Court concluded that two Indian parties can choose a foreign seat. However, the Court’s reasoning is feeble.
Two objections have been made against Indian parties choosing a foreign seat – a) this arrangement runs afoul of Section 28 of the AA which requires the governing law of the underlying contract for all India-seated, domestic arbitrations (between Indian parties) to be Indian law; b) Under Sections 28 of the Indian Contract Act, 1872 (“CA”) two Indian parties must not be prevented from accessing legal proceedings in India. The Court conflates these objections.
The Court misread the Atlas decision where the contract was executed between two Indian and a Hong Kong party. The arbitration clause provided for London arbitration. Relying on Section 28 of the CA, it was argued that denying two Indian parties remedy of Indian courts was contrary to public policy. The SC relied on the arbitration exception under the CA to enforce the award ruling that no public-policy argument could stand merely based on the arbitration being foreign-seated. The Court’s reliance on Atlas is problematic because Atlas does not examine the legality of Indian parties submitting to a foreign seat in the context of substantive law of the contract being foreign law (and therefore hit by Section 28 of the AA). The facts in Atlas do not fit the conclusion that the Court attributes to it.
Court’s reliance on Sasan is misplaced. In Sasan, two Indian parties agreed to arbitrate in London; English law governed. The lower court had ruled that two Indian parties could arbitrate in a foreign country under foreign law. In appeal, SC skirted the issue, ruling that the American parent of the Indian company was also party to the dispute. Since the dispute involved a foreign element, English law could apply. The court made no determination on the seat.
Holding Addhar Mercantile to be per incuriam for its failure to consider Atlas was also incorrect. In Addhar, two Indian parties had agreed to arbitrate “…in India or Singapore and English law to apply.” The Bombay HC held (relying on SC’s decision in TDM Infrastructure) that Indian nationals should not be permitted to derogate from Indian law. It, therefore, ruled that the arbitration would be India seated and accordingly Section 28 of the AA would not be breached. How has the Court misread these precedents? By suggesting that Atlas and Aadhar dealt with the same issue. The decision in Atlas is not a precedent for the proposition that two Indian parties cannot have their disputes determined by foreign law. Atlas dealt with Section 28 of the CA (not, Section 28 of the AA). Aadhar, on the other hand, deals with applicability of Section 28 of the AA to arbitrations involving two Indian parties.
In ruling that two Indian parties can opt for foreign-seated arbitrations, the Court advances party autonomy. The Court’s reasoning, however, is tenuous. The Court could have arrived at the same conclusion as follows:
- Atlas confirms that there are no public policy grounds that prevent two Indian parties from choosing a foreign seat;
- Section 28 of the AA requires Indian law be applied only for India-seated arbitrations between two Indian parties; and
- The SC’s decision in Sassan is not a precedent for either proposition.
In attempting to enforce the arbitration agreement, the GMR decision mirrors a welcome pro-arbitration trend adopted by Indian courts. Yet, its reasoning casts serious doubts on its standing as a precedent. Within their overarching pro-arbitration approach, Indian courts would do well to examine each case critically.
The views expressed herein are solely of the authors and do not represent any organizations or companies to which either author belongs.