Deutsche Telekom AG v The Republic of India [2023] SGHC(I) 7

Introduction

The plaintiff investor (“DT”) is a multinational corporation incorporated under the laws of the Federal Republic of Germany. Through its wholly-owned subsidiary, Deutsche Telekom Asia Pte Ltd (“DT Asia”), it was a shareholder in an Indian company, Devas Multimedia Private Limited (“Devas”).

The defendant State (“India”) is the Republic of India and was the respondent in the relevant arbitration proceedings that took place as detailed below between the plaintiff and the defendant (the “Arbitration”).

A key player in the dispute between DT and India is Antrix Corporation Ltd (“Antrix”), an Indian state-owned entity. Antrix is the commercial arm of the Indian Space Research Organisation (“ISRO”) and administratively controlled by India’s Department of Space (“DOS”).

The Arbitration arose out of India’s annulment of an agreement dated 28 January 2005 (the “Devas-Antrix Agreement”) between Devas and Antrix for the lease to Devas of S-Band electromagnetic spectrum on two satellites to be manufactured and launched by the ISRO. The Devas-Antrix Agreement contemplated (among others) the offering of mobile multimedia and information services to the Indian market via a hybrid satellite-terrestrial communications platform.

DT commenced the Arbitration in Switzerland and obtained an Interim Award (the “Interim Award”) on 13 December 2017 (dealing with jurisdiction and liability) and a Final Award (the “Final Award”) on 27 May 2020 (dealing with quantum) in its favour. DT sought to enforce the Final Award in Singapore and was granted leave to do so on 3 September 2021. India opposes the Singapore enforcement proceedings. India has also since applied for the Federal Supreme Court of Switzerland (the “Swiss Federal Supreme Court”) to review its decision to refuse to set aside the Interim Award. India contends that the Singapore court should stay DT’s enforcement proceedings pending the Swiss Federal Supreme Court’s decision on India’s revision application (the “Swiss Revision Application”). Alternatively, India says that the Singapore court should not recognise or enforce the Final Award. There are consequently four applications before us:

(a) HC/SUM 155/2022 (“SUM 155”) which is India’s application to set aside the order of the General Division of the High Court of Singapore dated 3 September 2021 (the “Leave Order”) granting leave to enforce the Final Award.

(b) HC/SUM 720/2022 (“SUM 720”) which is DT’s application to strike out parts of India’s affidavit evidence in SUM 155.

(c) SIC/SUM 24/2022 (“SUM 24”) which is India’s application to stay SUM 155 and SUM 720 pending the determination of the Swiss Revision Application.

(d) SIC/SUM 45/2022 (“SUM 45”) which is India’s application for leave to adduce further evidence in support of SUM 24.

Background

The arbitration generally

On 10 July 1995, India and Germany entered into a bilateral investment treaty entitled the Agreement between the Federal Republic of Germany and the Republic of India for the Promotion and Protection of Investments (the “BIT”).

DT commenced the Arbitration against India on 2 September 2013, claiming that India’s annulment of the Devas-Antrix Agreement violated the BIT. The Arbitration was governed by the United Nations Commission on International Trade Law (“UNCITRAL”) Arbitration Rules 1976 and seated in Geneva. On 22 May 2014, a tribunal comprising Prof Gabrielle Kaufmann-Kohler, Mr Daniel M Price and Prof Brigitte Stern (the “Tribunal”) issued Procedural Order No 1 in the Arbitration, which bifurcated the Arbitration into an initial phase on jurisdiction and liability followed by a second phase on damages. The hearing of the first phase took place between 6 April 2016 and 11 April 2016.

On 24 October 2016, India wrote to the Tribunal, enclosing a Charge Sheet dated 11 August 2016 (the “CBI Charge Sheet”) issued by India’s Central Bureau of Investigation (the “CBI”), and a Complaint dated 31 May 2016 (the “FEMA Complaint”) filed by the Directorate of Enforcement in India’s Ministry of Finance (“ED”) under section 16(3) of India’s Foreign Exchange Management Act, 1999 (“FEMA”). The CBI brought criminal charges against several government officials, Devas, and certain Devas officers, as reflected in the CBI Charge Sheet. India claimed that the CBI Charge Sheet showed that DT’s investment had not been in accordance with Indian law. India accordingly sought to suspend the Arbitration pending the resolution of those charges.

On 14 November 2016, DT emailed the Tribunal stating that it was too late for India to: (a) object to jurisdiction or admissibility based on the CBI Charge Sheet, (b) seek a suspension of the Arbitration pending resolution of the criminal charges in the CBI Charge Sheet, and (c) introduce new evidence in the form of the CBI Charge Sheet and the FEMA Complaint. DT observed that “India has had knowledge of the key allegations contained in the CBI Charge Sheet for years” and the “alleged facts underlying the accusations in the CBI Charge Sheet are already contained in the evidence before th[e] Tribunal”.

On 20 February 2017, the Tribunal refused India’s application to suspend the Arbitration and deferred its determination on India’s submissions in relation to the CBI Charge Sheet to its forthcoming award on jurisdiction and liability.

On 13 December 2017, the Tribunal issued its Interim Award on jurisdiction and liability. The Tribunal dismissed India’s objections to jurisdiction and found India liable for breach of India’s fair and equitable treatment (“FET”) obligation under the BIT. India’s submissions on the CBI Charge Sheet were also addressed in the Interim Award (see [159] below). The Tribunal then proceeded to the quantum phase of the Arbitration.

India applied to the Swiss Federal Supreme Court to set aside the Interim Award on 29 January 2018. India’s grounds were that: (a) the Tribunal lacked jurisdiction because DT’s investment had been indirectly made through DT Asia, (b) in refusing India’s request to admit the travaux préparatoires of the 1995 India-Netherlands bilateral investment treaty (the “India-Netherlands BIT”) into evidence, the Tribunal had denied India a reasonable opportunity to present its case, (c) the Tribunal lacked jurisdiction to rule on DT’s claims since DT had not made an investment in India but had merely engaged in pre-investment activities which are not protected by the BIT, (d) the Tribunal lacked jurisdiction as the challenged measures were necessary to protect India’s “essential security interests” and thus fell outside the Tribunal’s subject-matter jurisdiction under the BIT, and (e) the Tribunal lacked jurisdiction as the Devas-Antrix Agreement was contrary to Indian law and DT’s investment being based on that agreement was therefore likewise tainted by illegality.

On 11 December 2018, the Swiss Federal Supreme Court rejected India’s application to set aside the Interim Award. It held that: (a) the fact that DT’s investment had been made through DT Asia did not mean that the Tribunal lacked jurisdiction, (b) the Tribunal was entitled to refuse India’s application to introduce the travaux préparatoires of the India-Netherlands BIT, (c) the Tribunal had correctly concluded that the BIT did not contain what India described as an “admission clause” capable of depriving pre-investment activities of substantive protection and, in any event, that DT’s investment was not simply pre-investment activity, (d) India was precluded from raising the issue of essential security interests, and (e) India was precluded from arguing that DT’s investment was purportedly unlawful on the basis of the illegality of the Devas-Antrix Agreement.

Following India’s failure to set aside the Interim Award, the Arbitration’s quantum phase took place between 29 April 2019 and 3 May 2019. The Tribunal rendered its Final Award on 27 May 2020. In the Final Award, the Tribunal ordered that:

For context, the ICC Award named in subparagraph (e) refers to a Final Award dated 14 September 2015 that was issued in related arbitral proceedings. Those proceedings had been commenced by Devas against Antrix in or around June/July 2011, for breaches of the Devas-Antrix Agreement (see [31] below).

The Civil Court of Geneva certified that the Final Award was enforceable and declared the Final Award to be legally binding on 20 August 2020. Thereafter, DT commenced enforcement proceedings in the United States (“US”) and Singapore. As to the US proceedings, on 19 April 2021, DT filed a Petition to Recognize and Confirm Foreign Arbitral Award before the US District Court for the District of Columbia. As to the Singapore proceedings, on 2 September 2021, DT applied ex parte for leave to enforce the Final Award and the Leave Order was granted on the next day. On 11 January 2022, India applied to set aside the Leave Order. On 25 July 2022, the US District Court for the District of Columbia stayed the US enforcement proceedings pending the outcome of India’s Swiss Revision Application.

The Devas-Antrix Agreement

Since 1983, India’s DOS has been responsible for allocating India’s S-band electromagnetic spectrum. In 1997, India’s Cabinet approved a policy framework for satellite communication aimed at attracting foreign investment and encouraging private sector investment in the space industry. In 2000, the Indian Government approved a policy document entitled “Norms, Guidelines and Procedures for Implementation of the Policy Frame-work for Satellite Communications in India” (the “SATCOM Policy”). The SATCOM Policy authorised the DOS to allocate S-band spectrum for commercial use.

After the SATCOM Policy was issued, Forge Advisors LLC (“Forge Advisors”) (a US consultancy) negotiated with the DOS, the ISRO, and Indian Space Commission on a potential collaboration to commercialise some of India’s S-band spectrum. On 15 April 2004, Forge Advisors submitted a proposal to Antrix and the ISRO for the implementation of a “DEVAS System”. DEVAS was short for “Digitally Enhanced Video and Audio Services”. In particular, Forge Advisors proposed a hybrid (satellite-terrestrial) communications platform offering two principal services: (a) an interactive audio-visual service that would deliver television and cable programming to hand-held and mobile terminals, and (b) a broadband wireless access service that would provide internet access to fixed (homes) and nomadic users (PCs, laptops, tablets and mobile devices) in urban areas.

India instructed a committee to review the feasibility of the Devas project, including its technical feasibility, risk mitigation, time schedule and financial and organisation aspects. The committee (the “Shankara Committee”) was headed by Dr K N Shankara, the Director of the Space Applications Centre (the “SAC”), who has since passed away on 17 July 2017. On 14 May 2004, the Shankara Committee issued a report (the “Shankara Report”), concluding that the concept was “attractive” and provided a “significant opportunity to ISRO and Antrix in the development of a new, state-of-the-art satellite application and technology as well as in the broader participation in the international commercial satellite market”.

Based on the Shankara Report, in July 2004, Antrix’s Board of Directors (the “Antrix Board”) approved entry into a partnership with Forge Advisors. The Antrix Board approved the Devas-Antrix Agreement in December 2004. On 17 December 2004, Devas was incorporated in Karnataka, India, for the purpose of entering into the Devas-Antrix Agreement with Antrix.

On 28 January 2005, the Devas-Antrix Agreement was signed. It provided for the lease of 70 MHz of S-band capacity on two satellites to be manufactured and launched by Antrix/ISRO. Antrix wrote to Devas on 2 February 2006 confirming that, the necessary approvals having been received, the Devas-Antrix Agreement had come into effect.

DT’s investment in India

On 2 February 2006, Devas applied to the Indian Foreign Investment Promotion Board (the “FIPB”) for approval of a proposed acquisition of 38% of its shares by Telcom Devas Mauritius Limited and CC/Devas (Mauritius) Ltd, Mauritius affiliates of Telecom Ventures LLC and Columbia Capital LLC respectively. The FIPB approved the application on 18 May 2006.

As described in more detail below at [62]–[64], in 2008, DT invested into Devas through DT Asia. On 19 March 2008, DT Asia signed a share subscription agreement with Devas for 17.2% of Devas’ paid-up share capital in exchange for a US$75m equity contribution (the “Share Subscription Agreement”). On 1 May 2008, Devas applied (through a letter authored by Dua Consulting Private Limited) to the FIPB for approval of DT Asia’s subscription of up to approximately 17% of Devas’ share capital and the possible later acquisition by DT Asia of up to 26% of Devas’ share capital. By way of a letter dated 7 August 2008, the FIPB approved the application.

The Share Subscription Agreement was completed on 18 August 2008. DT Asia then appointed two nominee directors (Mr Alugappan Murugappan and Mr Kevin Copp) to Devas’ Board of Directors (the “Devas Board”).

In 2009, DT Asia and Devas entered into an agreement for DT Asia to make a further equity contribution to Devas. The requisite government approval (through the FIPB) was sought on 14 September 2009. On 17 September 2009, the FIPB approved the increase in Devas’ proposed foreign equity participation, including an increase in DT Asia’s shareholding in Devas to 20.73%.

India’s termination of the Devas-Antrix Agreement

On 8 November 2009, Mr Vijay Anand (Joint Secretary of the DOS) received an anonymous complaint that the S-band spectrum had been leased to Devas as a result of corrupt practices. Representatives of the Space Commission, the DOS and the ISRO met on 8 December 2009 to discuss the anonymous complaint. At this time, Dr K Radhakrishnan had recently become Chair of the Space Commission, Secretary of the DOS, Chairman of the ISRO and Chairman of Antrix. Following the meeting, Dr Radhakrishnan constituted a single-person committee (the “Suresh Committee”) consisting of the Director of the Indian Institute of Space and Technology, Dr B N Suresh. The Suresh Committee was asked to review the “legal, commercial, procedural and technical aspects” of the Devas-Antrix Agreement.

In June 2010, the Suresh Committee sent a report (the “Suresh Report”) to the ISRO and the DOS. The Suresh Report found that there was “absolutely no doubt on the technical soundness” of the Devas System as proposed and that “Antrix ha[d] been following the policy guidelines for leasing the transponder services to private service providers as per the [SATCOM Policy]”. But the report noted that the Devas-Antrix Agreement brought “limitations on spectrum availability for essential strategic and social sectors applications in future”. It therefore recommended that the Devas-Antrix Agreement be “re-visited taking into account all issues” such as the Indian National Satellite System (“INSAT”) Coordination Committee guidelines, the importance of preserving spectrum for essential national needs and international standards, with due weight given to Devas’ upfront payment.

As a result of media reports and public interest in the Devas-Antrix Agreement, Dr Radhakrishnan instructed Mr G Balachandhran, Additional Secretary of the DOS, to prepare a note on annulment of the Devas-Antrix Agreement. On 30 June 2010, Mr Balachandhran issued a Note (the “Balachandhran Note”) identifying concerns with the Devas-Antrix Agreement and recommending its annulment. The Balachandhran Note annexed the Suresh Report and the minutes of a meeting among the Integrated Defence Staff (“IDS”), the Ministry of Defence (“MOD”) and the ISRO dated 15 December 2009. Based on the Suresh Report, Mr Balachandhran prepared a report titled “Report on Dr Suresh Committee Report on ANTRIX-DEVAS Agreement & Issues arising from Therein” (the “Balachandhran Report”) dated 9 January 2011. The Balachandhran Report noted that the Government of India did not have complete information about the Devas-Antrix Agreement at the time of its conclusion and that the Devas-Antrix Agreement did not leave enough spectrum for ISRO/DOS use if required.

At a press conference on 8 February 2011, Dr Radhakrishnan and Dr Krishnaswamy Kasturirangan, a former ISRO Chairman and the Secretary of the DOS, announced India’s decision to terminate the Devas-Antrix Agreement. On 16 February 2011, Dr Radhakrishnan finalised a secret note (the “CCS Note”) to the Cabinet Committee on Security (“CCS”), seeking approval for annulling the Devas-Antrix Agreement. The CCS approved annulment on the next day.

On 25 February 2011, Antrix notified Devas of the termination of the Devas-Antrix Agreement. Antrix stated that termination was due to force majeure, Antrix being unable to obtain the necessary frequency and orbital slot clearance.

The ICC arbitration

In response to the termination of the Devas-Antrix Agreement, Devas and its Mauritius shareholders filed several arbitrations against Antrix and India.

In or around June/July 2011, Devas initiated an Indian-seated ICC arbitration (the “ICC arbitration”) against Antrix. Devas sought specific performance or alternatively approximately US$1.6 billion in damages. The ICC arbitration tribunal issued its Final Award (the “ICC Award”) on 14 September 2015. The ICC Award ordered that Antrix pay damages of US$562.5 million with simple interest at 18% per annum to Devas for wrongful repudiation of the Devas-Antrix Agreement.

On 19 November 2015, Antrix applied to the Indian courts to set aside the ICC Award. On 4 November 2020, the Supreme Court of India ordered that the ICC Award be held in abeyance pending determination of Antrix’s setting-aside application.

On 27 October 2020, the US District Court for the Western District of Washington recognised the ICC Award and on 4 November 2020 entered judgment against Antrix for the full amount of the ICC Award. We have not been apprised of further developments in either the Indian or US proceedings pertaining to the ICC Award.

Indian investigation of the Devas-Antrix Agreement

On 10 February 2011, the Prime Minister of India constituted a High Power Review Committee (the “Chaturvedi Committee”) to look into the decision to enter into the Devas-Antrix Agreement. The Chaturvedi Committee issued its report on 12 March 2011.

On 12 April 2011, the Indian Cabinet Secretary, Mr K M Chandrasekhar, submitted his recommendations (the “Chandrasekhar Note”) to the Prime Minister. He proposed that Devas be investigated for possible violations of the FEMA and India’s Prevention of Money Laundering Act, 2002 (“PMLA”).

Subsequently, a committee headed by Mr Pratyush Sinha, a former Chief Vigilance Commissioner, submitted a “Report of the High Level Team on the Agreement between M/s Antrix Corporation Limited and M/s Devas Multimedia Private Limited” (the “Sinha Report”). The Sinha Report was released by the Government of India on 2 September 2011.

On 1 May 2014, the CBI commenced a Preliminary Enquiry into allegations that Antrix officers had defrauded the Government of India. On 16 March 2015, a First Information Report was made by Mr Sushil Dewan, CBI Inspector of Police. The report implicated certain officers and officials of Devas, Antrix, the ISRO and the DOS in wrongdoing. On 11 August 2016, the CBI issued a Final Report and through the CBI Charge Sheet raised criminal charges against several government officials, Devas, and various Devas officers (the charges are referred to in greater detail at [85] below). On 8 January 2019, the CBI issued a Supplementary Charge Sheet.

On 31 May 2016, the ED filed a complaint under section 16(3) of the FEMA against Devas, its current and former directors and foreign investors, including DT Asia. On 6 June 2016, the ED issued a show-cause notice to the accused persons. Based on the charges, the ED issued a penalty order on 30 January 2019 against the accused persons amounting to INR1,585.08 crores.

On 14 January 2021, Antrix requested authorisation from the Indian Ministry of Corporate Affairs to commence winding-up proceedings against Devas under section 271(c) of the Indian Companies Act, 2013, essentially on the ground of fraud in the conduct of its affairs. On 18 January 2021, Antrix was authorised to file a winding-up petition against Devas. Antrix thereafter commenced winding-up proceedings before India’s National Company Law Tribunal (the “NCLT”). Having heard Antrix’s winding-up petition, the NCLT appointed a provisional liquidator to take over Devas’ affairs on 19 January 2021. The provisional liquidator issued a First Report on 3 February 2021, and on 25 May 2021, the NCLT ordered that Devas be wound up. The NCLT also dismissed a joinder application by Devas Employees Mauritius Private Limited (“DEMPL”), a Devas shareholder, for lack of standing.

On 8 September 2021, India’s National Company Law Appellate Tribunal (“NCLAT”) upheld the NCLT’s winding-up order and its dismissal of DEMPL’s joinder application. On 17 January 2022, the Supreme Court of India issued its judgment (the “SCI Judgment”) dismissing Devas’ and DEMPL’s appeals against the NCLAT’s decision, thereby affirming the decision to wind up Devas.

Discussion

We will deal with the applications that are before this court (as summarised at [5] above) in the following order: (a) SUM 155 (setting aside of leave order), (b) SUM 24 (stay), (c) SUM 45 (adduction of further evidence) and (d) SUM 720 (striking out).

SUM 155 – application to set aside leave order

Jurisdictional objections

In SUM 155, the central issue is whether, as a sovereign state, India is immune from the jurisdiction of the Singapore courts. By section 3(1) of the State Immunity Act 1979 (2020 Rev Ed) (the “SIA”), sovereign states are immune from this court’s jurisdiction unless an exception to state immunity applies. DT contends that the case falls within an exception in the SIA, namely section 11(1): “Where a State has agreed in writing to submit a dispute which has arisen, or may arise, to arbitration, the State is not immune as respects proceedings in the courts in Singapore which relate to the arbitration”. India argues to the contrary.

DT relies on the offer to arbitrate in Article 9 of the BIT. Article 9 reads:

In response, India says that DT’s investment falls outside the scope of Article 9, and contends that there are four reasons for this. First, DT’s investment was not in accordance with India’s national laws (the “illegality argument”). Second, DT’s investment merely amounted to pre-investment expenditure (the “pre-investment expenditure argument”). Third, DT did not make a direct investment but rather an indirect investment, the latter not falling within the ambit of the BIT (the “indirect investment argument”). Fourth, the subject matter of the parties’ dispute falls within the “essential security interests” carve-out in Article 12 of the BIT (the “essential security interests argument”).

(1) The illegality argument

Article 1(b) of the BIT defines “investment” as follows:

Article 9(2) of the BIT therefore only confers jurisdiction on an arbitral tribunal if DT’s investment was invested in accordance with India’s national laws. If DT’s investment was not invested in accordance with India’s laws, DT’s investment would not be protected by the BIT. Neither Article 9 nor the exception in section 11(1) of the SIA would then apply and India would thereby be immune from the Singapore court’s jurisdiction by virtue of section 3(1) of the SIA.

(A) India’s case on illegality

India raises five reasons for why it contends that DT’s investment was contrary to India’s national laws.

First, India says that, in the negotiations leading to the Devas-Antrix Agreement and in the agreement itself, Devas and its original investors fraudulently misrepresented to the ISRO that Devas could deliver technology which in fact did not exist then and has never existed. India alleges that Devas and its investors also fraudulently misrepresented that Devas held intellectual property rights which it did not actually have. India accepts that, when the alleged misrepresentations were made, DT was not an investor. But, according to India, when DT invested in Devas in 2008 (via DT Asia), DT conducted detailed due diligence. That due diligence (India says) would have examined whether Devas had the technology and intellectual property rights to deliver the promised services under the Devas-Antrix Agreement. India infers from this premise that DT must have known that Devas had neither the technology nor the intellectual property rights which Devas purported to have. Nonetheless, DT proceeded with its investment, despite knowing that the same was tainted with fraud and not in accordance with India’s national laws. DT must consequently be regarded as having become complicit in or tainted by the fraud perpetrated by Devas and its original investors on Antrix and India.

In support of its case on fraud, India relies on Devas’ representations in Article 12 of the Devas-Antrix Agreement. Article 12 reads:

India focuses in particular on Devas’ representations in Articles 12(b)(i) to (iv). As evidence that those representations were fraudulent (that is, made dishonestly without any belief in their truth), India cites various passages from the NCLAT’s and the Supreme Court of India’s judgments.

From the NCLAT’s judgment, India relies on the following passages (at [66]–[70]):

From the SCI Judgment, India relies on the following passages (at [12.8]):

India does not allege any other fraudulent representations by Devas.

Second, India says that the Devas-Antrix Agreement was illegal because it was entered into without a tender or auction process. India says that this was unusual for the allocation of spectrum rights as its national law requires the allocation of natural resources to be conducted in a fair, non-arbitrary and transparent manner. India having found out that there had been irregularities in the allocation of spectrum rights to Devas, formal charges for criminal conspiracy were brought on 11 August 2016 against government officials as well as directors and other officers of Devas. In relation to DT, which only came onto the scene in 2008, India makes the same point (as made in [48] above). DT, having conducted detailed due diligence, must have known that spectrum had been allocated to Devas in contravention of national law. As DT nonetheless proceeded with its investment in Devas shares, it follows (India submits) that DT’s investment in Devas became tainted with illegality.

In support of its case on the illegality in the allocation of spectrum, India cites the following passage from the SCI Judgment (at [12.8]):

Third, India says that the Devas-Antrix Agreement was illegal because it contravened India’s SATCOM Policy. Contrary to what the Devas-Antrix Agreement purportedly authorised Devas to carry out, India’s SATCOM Policy at the material time prohibited a party from engaging in both telecommunications and broadcasting services. In support, India cites the following from the SCI Judgment (at [12.8]):

Again, India submits that, having conducted due diligence prior to its investment, DT would have known that the Devas-Antrix Agreement was contrary to India’s SATCOM Policy and hence illegal. Because DT nevertheless decided to proceed with its investment in Devas despite such knowledge, DT’s investment was also illegal.

Fourth, according to India, the FIPB approved DT’s investment in Devas subject to several conditions, including that Devas would be operating as an Internet Services Provider (“ISP”). India says that, as the services to be provided by Devas under the Devas-Antrix Agreement were different in nature from ISP services, DT’s investment was contrary to the conditions of the FIPB approval and hence not in accordance with India’s national law.

On 2 February 2006, Devas applied for FIPB approval to operate as an ISP. Devas’ application stated:

The FIPB approval dated 18 May 2006 (FC.II.107(2006)/43(2006)) stipulated:

According to India, on its terms, the FIPB approval was confined to Devas only providing internet services.

On 1 May 2008, Dua Consulting Private Limited wrote to the FIPB on Devas’ behalf, stating that:

A letter dated 27 April 2008 from Devas and DT Asia to the FIPB explained DT’s investment in Devas through DT’s wholly-owned Singapore subsidiary DT Asia. Initially, DT Asia would acquire up to approximately 17% of Devas’ paid-up share capital. Thereafter, it would have the option, in accordance with a shareholders’ agreement among Devas, some of Devas’ existing shareholders and DT Asia, to acquire up to 26% of Devas’ paid-up share capital at a consideration “not less than that determined as per the prevailing RBI [ie, Reserve Bank of India] guidelines”. In any event, the aggregate non-resident shareholding of Devas would not exceed 74% of Devas’ total paid-up share capital.

By letter dated 7 August 2008, the FIPB approved DT Asia’s acquisition of 17.2% of Devas’ paid-up capital. The letter referred to the FIPB approval of 18 May 2006 “read with amendment letter of even number dated 19.5.2008” and stated:

India’s case is that condition 4 of the letter means that DT’s investment in Devas’ shares was approved subject to Devas complying with the conditions in the earlier FIPB approval of 18 May 2006. As Devas purported to provide more than just internet services, Devas was in breach of the conditions in the FIPB approval of 18 May 2006. In consequence, DT’s investment in Devas must likewise have breached the conditions imposed by the FIPB approval letter of 7 August 2008.

Fifth, India says that the FIPB approved of money coming into India for Devas’ ISP business. But 85% of that money (India complains) was promptly transferred by Devas to Devas Multimedia America Inc (“Devas Delaware”) in the US. India submits that the money which was approved to come into India had to be used there, ie, within the jurisdiction. DT must have known that the moneys were transferred out, because by then DT had two nominee directors on the Devas Board. The transfer out of funds was thus contrary to the FIPB’s conditions, and was therefore an illegal act by Devas. DT through its nominee directors would have been party to this illegality.

In support, India observes that Devas’ 2 February 2006 application letter and the letter dated 27 April 2008 from Devas and DT Asia (at [59] and [63] above) stated that the technologies for the ISP services to be provided by Devas would be developed indigenously in India. India argues that the FIPB approval was given on that basis, which would consequently also have been a condition of DT’s investment in Devas shares. On the illegality of the transfer out of the moneys, India relies on the NCLAT’s judgment where the tribunal stated (at [195]–[197]):

It will be seen from the foregoing summary that India’s entire case on illegality, in essence, hinges on an inference that DT knew or must have known about Devas’ alleged fraud by reason of the due diligence which DT conducted prior to making its investments and, in respect of the fourth and fifth illegality submissions, because DT had two nominee directors on the Devas Board. By way of tying up its case, India refers to the following passage from the SCI Judgment (at [13.5]–[13.6]):

India submits that the passages quoted above amount to a finding of fraud on the part of DT (among others) by India’s highest court and should therefore be regarded by this court as conclusively establishing DT’s complicity in the fraud perpetrated on Antrix and India.

(B) DT’s case on illegality

In essence, DT’s response is that the alleged illegality does not concern DT’s making of its investment, namely the acquisition of shares in Devas. When the Devas-Antrix Agreement was concluded in 2005, DT did not have any investment in Devas, nor did it have plans for any. DT (through DT Asia) only acquired shares in Devas from 2008, after: (a) extensive negotiations with Devas and representatives from the ISRO and the DOS; (b) due diligence on the advice of Indian counsel; and (c) the receipt of requisite approvals (including those of the FIPB). According to DT, there is no evidence that DT had any knowledge of the alleged illegality when it acquired the shares in 2008 and 2009.

DT contends that India’s arguments are therefore entirely speculative. That DT performed due diligence into aspects of Devas’ business did not mean that it learnt about Devas’ prior negotiations with Antrix. That DT had two nominee directors on the Devas Board also did not mean that those directors managed to obtain information about Devas in a complete and timely manner, and in turn that DT had full knowledge of all of Devas’ dealings and operations.

Further, some of India’s allegations (such as India’s fifth objection that money was being transferred abroad) concern the performance of the investment and not its making, and as such do not fall foul of the legality requirement in Article 1(b) of the BIT.

Other allegations (such as India’s fourth objection that the FIPB only approved the provision of ISP services) would, in any event, only constitute trivial breaches that would not deprive DT of substantive protection under the BIT.

(C) Analysis of India’s case on illegality

We are not persuaded by India’s case on illegality.

First, we are unable to draw the inference upon which India’s first three submissions depend. We are prepared to proceed on the footing that Devas engaged in fraudulent misrepresentation. We are, however, unable to infer that, merely because DT undertook due diligence of Devas prior to its investment, DT must have known of Devas’ fraud so that, by proceeding with its investment in Devas, DT must be taken to be complicit in the deceit on Devas’ part. The evidence is that DT and India both conducted due diligence on Devas. If India (as seems to have been the case) was nonetheless deceived by Devas, it is difficult to see why DT was not or could not have been likewise taken in by Devas, especially where government officials were (according to India) acting in concert with Devas’ original investors.

That India itself scrutinised Devas’ proposed investment, not just once, but several times over the years, is evident from the chronology of events.

As stated above at [18], in May 2004, a High Power Committee (the Shankara Committee) was constituted to evaluate the technical feasibility, risk mitigation, time schedule, and financial and organisational aspects of the then proposed Devas-Antrix Agreement. In June-July 2004, the Antrix Board, having reviewed the findings of the Shankara Committee, accorded in-principle approval to the proposed Devas-Antrix Agreement. In October 2004, there was a presentation in Vancouver to the chairperson of the ISRO, a director of the ISRO Satellite Centre (“ISAC”) and Antrix (among others) about Devas, the competitive landscape, international trends, Devas’ technological developments, receiver pricing aspects and other matters. Between August and December 2004, the Shankara Committee reviewed and recommended terms for the proposed Devas-Antrix Agreement. A briefing to the Technical Advisory Group took place in November 2004. It was only thereafter that the draft Devas-Antrix Agreement was approved by the Antrix Board at its 57th meeting on 24 December 2004.

In June 2010, the Suresh Report was submitted to the chairperson of the ISRO, who was also the secretary of the DOS (see [26] above). In his review, Dr Suresh was assisted (among others) by Mr K R Sridhara Murthi and Mr S Parameswaran of the Antrix Board, and Mr Vijay Anand and Mr S K Jha of the DOS. The Suresh Report commented extensively on the Devas-Antrix Agreement. The following are excerpts:

India initiated other reviews of the Devas-Antrix Agreement. It was not until 30 June 2010, when the Balachandhran Note was produced (as described at [27] above), that India identified irregularities in the arrangements between Devas and Antrix. The Balachandhran Note was then considered by the Space Commission at its 117th meeting on 2 July 2010, and the Commission resolved that the DOS “may take actions necessary and instruct Antrix to annul the Antrix-Devas Agreement”. Further investigation led to the decision, announced in a press conference on 8 February 2011, to revoke the allocation of spectrum to Devas and annul the Devas-Antrix Agreement.

On 12 March 2011, the Chaturvedi Committee issued its report on aspects of the Devas-Antrix Agreement (see [34] above). In the Chandrasekhar Note dated 12 April 2011 commenting on that report (see [35] above), Mr K M Chandrasekhar (India’s Cabinet Secretary) observed:

The Chandrasekhar Note suggests that, far from being parties to fraud, new investors in Devas (such as DT) were themselves the victims of the fraud by Devas’ original investors, the latter having offloaded their shares (at a “huge premium”) onto the newcomers despite Devas’ alleged shallow asset base.

The next report in the chronology, the Sinha Report (see [36] above), concluded (among other matters):

The Sinha Report recommended that disciplinary action be taken against certain individuals, including Mr G Madhavan Nair, Mr A Bhaskaranarayana, Mr K R Sridhara Murthi, and Dr K N Shankara. The Sinha Report then stated in its final paragraph:

On 11 August 2016, in the CBI Charge Sheet, the following individuals were charged in connection with the Devas-Antrix Agreement: Mr K R Sridhara Murthi (former Managing Director of Antrix), Mr Ramachandran Viswanathan (Devas’ President and Chief Executive Officer), Mr Muthgadahali Gangarudaraiah Chandrasekhar (a director of Devas, and a different individual from the Indian Cabinet Secretary referred to at [35] and [81] above), Mr G Madhavan Nair (former DOS Secretary and chairperson of the ISRO and of Antrix), Mrs Veena Ram Rao (former Additional Secretary of the DOS), Mr Appana Bhaskarnarayana (former Director of the Satellite Communication & Navigation Programme Office of the ISRO), Mr Desaraju Venugopal (a director of Devas), and Mr Marike Umesh (a chartered accountant). Despite the Sinha Report’s recommendation that Dr Shankara be investigated, no charges were ever levied against him prior to his death in mid-2017. No charges have ever been brought against Dr Suresh either. By way of summary of the crime committed by those charged, the CBI Charge Sheet stated at paragraph 16(2):

The gain of Rs 578 crores mentioned in paragraph 16(2) just quoted is the total of what Devas received for allotting shares to new investors (including DT). Of these Rs 578 crores, DT paid up Rs 325,25,00,000 and 106,53,81,167 on 18 August 2008 and 29 September 2009 respectively into Devas’ account. The CBI Charge Sheet then set out (at paragraphs 16(152) and (153)):

Devas and certain of its officers and shareholders were essentially accused of allotting Devas shares to foreign investors (such as DT) at inflated premiums and then pocketing the funds so injected into Devas by remitting the same into their foreign accounts. If so, DT would be the victim, rather than a perpetrator, of fraud. It is telling in this respect that India has never brought charges against DT. Indeed, with several high-ranking government officials in the ISRO and the DOS and certain members of Devas’ management apparently involved in fraud, on the material before us, we cannot rule out the real possibility that, much as was apparently the case with India, DT was also duped by Devas’ original shareholders, notwithstanding the due diligence which DT carried out. We are therefore not satisfied that DT was complicit in any fraud.

Second, we are not persuaded that DT would have known through its nominee directors on the Devas Board that:

The gist of India’s complaint in relation to the Devas services was that all that the FIPB had approved for Devas to provide were ISP services and those would not have required use of a satellite. The Devas services were, however, purporting to provide satellite-related services. Our difficulty is that, on its face, the FIPB plainly approved the provision of satellite-based services also. We have highlighted above (at [59] and [60]) explicit references to the delivery of internet services (including multimedia and information services) via landline, satellite, and terrestrial wireless systems in Devas’ application for FIPB approval and in the FIPB’s approval itself. In its oral submissions, India invited us to take judicial notice that ISP services can be provided without resort to satellite systems. That may possibly be so. But that proposition begs the question whether internet services can also be provided by satellite. In the absence of technical evidence to the contrary (of which none was placed before us), we do not accept that, by definition, ISP services are limited only to non-satellite systems of delivery. It is difficult to see why, reading Devas’ application to the FIPB and the FIPB’s approval, DT or its nominee directors ought to have realised that, in contracting to provide ISP services by satellite, Devas was purporting to act in a way which it had not been authorised to.

India has suggested that the FIPB’s approval for DT’s investment was contingent on Devas meeting the conditions imposed by the FIPB on Devas’ investment. But we are unable to read the FIPB’s approval of DT’s investment as importing such a condition. We explain further below.

In relation to the remittance by Devas of funds abroad, as already pointed out (at [75]–[87]), it can be surmised from various reports as well as the CBI Charge Sheet that the remittance, if done in fraud, was done as much in fraud of Devas’ new investors (including DT) as it was in fraud of India. In those circumstances, it is difficult to see how DT could have been complicit in a fraud against itself. What is more plausible is that, at the highest, DT and its nominee directors were duped by those involved. In any event, given the gravity of the accusations being levied against DT and its nominee directors (ie, fraud), far more evidence, and cogent evidence, is required. India cannot rely on a tenuous inference based on the presence of nominee directors on the Devas Board.

We would also add these observations:

(a) India suggests, in connection with the Devas services, that Devas never actually acquired the right to use satellite spectrum. This is because (India says) Devas never obtained the WPC licence needed to use spectrum terrestrially. It is alleged that Devas never even applied for a WPC licence. It is suggested that DT ought to have realised this, either because of its due diligence or the presence of nominee directors on the Devas Board. In fact, according to the Suresh Report (at p 11):

The Interim Award picks up the story (at para 386):

It is clear to us that no final WPC licence application was ever made by Devas for the simple reason that before any such step was taken, India annulled the Devas-Antrix Agreement in February 2011.

(b) We have assumed in India’s favour that its arguments on the illegality of the fund transfers abroad go towards the Tribunal’s jurisdiction. That is, the contention is that the alleged illegality meant that DT’s investment was not invested in accordance with India’s national laws and so outside of the scope of the offer to arbitrate in Article 9 of the BIT. However, in our view, India’s arguments concern the performance or use of DT’s investment and not the making of it. As we note below (at [97]), DT’s investment was the purchase of Devas’ share capital through DT Asia. That investment was within the terms of the BIT and so within the scope of the offer to arbitrate (and in turn, the arbitration agreement). At best, the question of the illegality or otherwise of the fund transfers abroad and DT’s complicity (if at all) in such illegality only involves the performance of a covered investment under the BIT or the use to which that investment is put.

Third, in respect of India’s second and third reasons for asserting illegality (as described at [54] and [56] above), we find that these lack any evidential foundation. The norms, guidelines and procedures for satellite communications approved by India in 2000, and thus constituting the SATCOM Policy at the time of the Devas-Antrix Agreement, stipulated:

In respect of India’s second objection (that the Devas-Antrix Agreement was illegal because it was entered into without a tender or auction process), paragraph 2.6.2 of the SATCOM Policy clearly contemplates the possibility of satellite capacity being allocated to commercial users on a “first come first served” basis. In respect of India’s third objection (that the SATCOM Policy prohibited an entity like Devas from engaging in both telecommunications and broadcasting services), the document does not on its face proscribe a party from engaging in both telecommunications and broadcasting services. Accordingly, it is not apparent to us how, solely by reason of it having conducted due diligence, DT would have become aware that Devas’ engagement in both telecommunications and broadcasting services was contrary to the national laws of India at the time the investment was made. This would especially be the case as, to all intents and purposes, Devas’ proposals had received FIPB approval for, among others, the provision of satellite services.

(2) The pre-investment expenditure argument

Article 3(1) of the BIT provides: “Each Contracting Party shall encourage and create favourable conditions for investors of the other Contracting Party and also admit investments in its territory in accordance with its law and policy.”

(A) India’s case on pre-investment expenditure

India says that by virtue of Article 3(1), the BIT is an “admission clause” treaty. That means that mere pre-investment activity is outside the scope of the BIT and consequently outside of the Tribunal’s jurisdiction. Execution of the Devas-Antrix Agreement (India submits) was only the first of many steps to Devas’ investment being admitted as a covered investment within the terms of the BIT. For instance, Devas had to obtain numerous approvals and authorisations before it could implement the Devas Services. It failed to do so. It never acquired the right to use the S-band spectrum. It never obtained the requisite WPC licence for terrestrial use of such spectrum. Nor did it ever receive approval for the frequency and orbital slot coordination required to operate a satellite. Under Article 7 of the Devas-Antrix Agreement, Devas and Antrix could respectively terminate their agreement if the other failed to obtain requisite regulatory approvals. It follows (so India argues) that DT’s acquisition (through DT Asia) of shares in Devas amounted at best to pre-investment activity falling outside the scope of the BIT. DT’s acquisition of shares constituted nothing more than “corporate activities surrounding the formation of Devas”. According to India, the shares “could only become eligible for protection under the BIT if and when Devas had obtained the requisite licences and approvals to offer Devas Services (which it failed to do)”.

(B) DT’s case on pre-investment expenditure

DT observes that this is the third time India is re-litigating the same arguments on pre-investment expenditure, having already done so (unsuccessfully) before the Tribunal and the Swiss Federal Supreme Court. DT disagrees that Article 3(1) makes the BIT an “admission clause” treaty. Even if it was, DT’s investment was duly admitted by India.

(C) Analysis of India’s case on pre-investment expenditure

We respectfully disagree with India’s submissions. In our judgment, Article 3(1) of the BIT is not a permissive clause authorising India to decide whether to “admit” something as an investment protected by the BIT. Article 3(1) instead obliges India and Germany as contracting states to the BIT to admit investments into their territories subject only to their respective laws and policies. Here DT’s investment fell within the definition of “investment” in Article 1(b) of the BIT. DT (through DT Asia) bought shares in Devas, an Indian company. DT made an investment in “shares in ... a company” and thereby acquired an interest in that company. There was nothing more that needed to be done to qualify as an “investment” within the meaning of Article 1(b) of the BIT. If (contrary to our view) an official “admission” of DT’s investment was required, such “admission” came in the form of the FIPB’s approval of DT’s purchase of 26% of Devas’ paid-up share capital (see [62]–[64] above).

(3) The indirect investment argument

Article 2 of the BIT defines its scope: “This Agreement shall apply to all investments made by the investors of either Contracting Party in the territory of the other Contracting Party, whether made before or after the coming into force of this Agreement”.

(A) India’s case on indirect investment

India says that DT’s investment did not fall within Article 2 of the BIT. This is because it was not DT, but DT Asia (DT’s wholly-owned subsidiary), which acquired the Devas shares. DT Asia, as a Singapore company, was not entitled to protection under the BIT. India refers, in support, to its treaty practice as evidenced in the travaux préparatoires of the India-Netherlands BIT (which was later terminated on 1 December 2016). The travaux include the following communiqué dated 26 October 1994 from India to the Dutch embassy in India in relation to the draft Article 2 of the India-Netherlands BIT:

(B) DT’s case on indirect investment

As with India’s arguments on pre-investment expenditure, DT observes that this is the third time India is re-litigating the same arguments on indirect expenditure. India had failed before the Tribunal and the Swiss Federal Supreme Court, and rightly so, because nothing in the definition of “investment” and “investor” in Article 1 of the BIT excludes indirect investments.

(C) Analysis of India’s case on indirect investment

We disagree with India’s position. Nothing in the wording of Article 2 of the BIT, or of Article 1(b) under which “investment” and “investor” are in turn defined, limits investors to those making direct investments, as opposed to indirect investments through wholly-owned subsidiaries incorporated in a third country. There is no basis for reading such a limitation into the plain words of Article 1(b) in the absence of words to that effect. The communiqué is also of little help in construing Article 1(b). It is difficult to see how a bilateral instrument such as the BIT can be construed by only looking at one contracting state’s position. Further, it is not apparent to what extent the contents of the communiqué were public knowledge when the BIT was being negotiated. It is more likely that at the time, the communiqué was a confidential note between India and the Netherlands. In fact, Article 2 of the India-Netherlands BIT that was eventually concluded provided:

If anything, that outcome suggests that, despite the communiqué, India was amenable to foreign investments being made indirectly irrespective of where the investor’s direct investing company was located. India invites us to infer from the express mention of indirect investment in Article 2 of the India-Netherlands BIT that the absence of a similar reference in the BIT means that indirect investments were excluded from the latter. In our view, the available material does not justify us drawing such a conclusion.

(4) The essential security interests argument

Article 12 of the BIT provides:

(A) India’s case on essential security interests

India says that, by Article 12, the BIT is inapplicable where India’s “essential security interests” are invoked and therefore an arbitral tribunal has no jurisdiction under the arbitration agreement stemming from the BIT over claims touching on those interests. Further, according to India, it must be for India to determine what its “essential security interests” are and when those interests are at risk or in jeopardy. In February 2011, India concluded that, given India’s military and security needs, Devas should not be allocated an orbital slot in the S-band for commercial purposes. Antrix accordingly terminated the Devas-Antrix Agreement based on those security needs. In the circumstances, the termination of the Devas-Antrix Agreement came within the scope of the carve-out from the BIT in Article 12.

(B) DT’s case on essential security interests

As with the preceding arguments on pre-investment expenditure and indirect investment, DT points out that India had failed to persuade the Tribunal and the Swiss Federal Supreme Court on its arguments concerning its essential security interests. Those arguments were moreover couched as substantive objections before the Tribunal, not jurisdictional ones. The Swiss Federal Supreme Court decided that they could not be improperly recast as jurisdictional objections, and DT urges this court to decide the same. In any event, on the facts, the Tribunal found that the Devas-Antrix Agreement had not been annulled with a view to protecting India’s essential security interests.

(C) Analysis of India’s case on essential security interests

We reject India’s submission.

First, we agree with DT that India advanced its “essential security interests” argument in the arbitration as a substantive objection, not a jurisdictional one. This is apparent from the Tribunal’s summary of India’s position on “essential security interests” at paragraph 201 of the Interim Award:

There was good reason for India taking such an approach in the arbitration. That is because the question of whether “essential security interests” are involved in a dispute and (if so) with what consequence, does not go towards jurisdiction. To the contrary, where essential security interests are at stake and it is established pursuant to Article 12 of the BIT that India only acted “to the extent necessary for the protection” of those interests, India would have a complete answer to claims that it breached a substantive obligation under the BIT. Where the conditions in Article 12 are met, India’s substantive obligations under the BIT yield to the interest protected by Article 12. In other words, India’s “essential security interests” trump any substantive rights that DT might otherwise have enjoyed under the BIT. This does not mean that the Tribunal’s jurisdiction was ousted in relation to matters said by India to constitute “essential security interests”. The Tribunal remained duty-bound to assess (as it did) whether the conditions in Article 12 had been met, in particular whether India had acted in a manner which was no more than necessary to protect the interests invoked.

Before us, India referred to Continental Casualty Company v The Argentine Republic ICSID Case No ARB/03/9, Award (5 September 2008) (“Continental Casualty”) in support of its contention that Article 12 goes towards jurisdiction. In Continental Casualty, Argentina invoked Article XI of the 1991 Argentina-US BIT, the equivalent of Article 12 of the BIT. Article XI of the Argentina-US BIT provided:

The ICSID award in Continental Casualty stated (at [164]):

The italicised sentence in the foregoing paragraph was footnoted as follows:

For completeness, Article 25 of the ARSIWA, which addresses the circumstance precluding wrongfulness of necessity, states:

India cited the underlined words from footnote 236 in support of the argument that Article 12 of the BIT deals with jurisdiction. But, read in their proper context, the words do not, in our judgment, support that contention. Significantly, Argentina advanced jurisdictional challenges in Continental Casualty. However, those challenges were not premised on Article XI. Just as India before the Tribunal, Argentina relied on Article XI as a substantive defence and the award in Continental Casualty was consequently addressing that substantive defence in the passage cited above. The award concluded that, where interests protected by Article XI were rightly invoked (as the award found was potentially the case), Argentina’s substantive obligations under the Argentina-US BIT had to be construed as limited by Article XI. None of this meant that Argentina ceased to be bound by its procedural obligations under the Argentina-US BIT (including the obligation to resolve disputes with US investors through arbitration). The tribunal in Continental Casualty therefore had jurisdiction to evaluate (as it did) whether, in relation to each of the investor’s specific complaints, Argentina had acted proportionately (that is, in a manner that was no more than necessary to protect its essential interests).

The net result of the foregoing analysis is that the “essential security interests” argument having been mounted (as it had to be) by way of a substantive defence in the arbitration and that defence having been rejected by the Tribunal, India is bound by the Tribunal’s determination that the conditions of Article 12 have not been met. There is no basis upon which this court can review the merits of the Tribunal’s substantive determination on the issue.

Other grounds relied on by India

In addition to state immunity, India submits that the leave granted to enforce the Final Award should be revoked:

(a) pursuant to the International Arbitration Act 1994 (2020 Rev Ed) (the “IAA”), specifically sections 31(2)(b) and 31(2)(d), as the Tribunal lacked jurisdiction for the reasons canvassed in [42]–[112] above;

(b) pursuant to section 31(4)(b) of the IAA, as enforcement of the Final Award would be contrary to the public policy of Singapore due to the fraud and illegality identified in [47]–[69] above; and

(c) because DT did not make full and frank disclosure when applying to court ex parte for leave to enforce the award.

On the latter ground on disclosure, India complains that, despite DT’s duty when applying ex parte to disclose India’s potential arguments against enforcement, DT failed to make such disclosure. In particular, DT did not mention India’s claim to state immunity or the NCLT’s adverse findings against DT as a shareholder in Devas. To this, DT’s response is that the duty to provide full and frank disclosure only extends to facts which it could reasonably ascertain and potential defences that it could reasonably anticipate. In this regard, for instance, DT could not have reasonably anticipated India’s arguments based on the NCLT decision, since DT’s position has always been that the NCLT, NCLAT and Supreme Court of India’s decisions are irrelevant to the present proceedings. In any case, even if there had been a breach, DT invites us to exercise our discretion not to set aside the leave order.

In our view, none of the additional grounds has merit. We have already rejected India’s jurisdictional challenges and the submission that DT must be treated as complicit in Devas’ fraud. Recasting those allegations as grounds to set aside the leave order or to refuse enforcement of the Final Award under the IAA does not alter the analysis. Nor do we believe that there is substance to the complaint of non-disclosure. Dr Ina Roth’s 1st Affidavit filed with DT’s ex parte application referred to India’s Swiss setting aside application. It mentioned the Swiss Federal Supreme Court’s finding that the Tribunal had jurisdiction and had conducted the arbitration proceedings fairly. Dr Roth exhibited the Swiss Federal Supreme Court’s Judgment which (among others) dismissed India’s submissions on state immunity. It would have been evident from a perusal of the Swiss Federal Supreme Court Judgment that India was likely to raise (as it has done) similar submissions on immunity and DT’s complicity in fraud before the Singapore courts.

Estoppel, res judicata and waiver

We have so far considered India’s illegality arguments on their merits de novo in light of the evidence before us, without reference to what other fora have previously determined. We have concluded on that basis that India’s arguments should be rejected. However, there were extensive arguments advanced before us by Ms Koh Swee Yen SC (for DT) and Mr Cavinder Bull SC (for India) on the extent to which DT or India was estopped or precluded:

(a) from raising matters contrary to findings in:

(i) the judgments of the NCLT, the NCLAT and the Supreme Court of India; or

(ii) the judgment of the Swiss Federal Supreme Court; or

(b) from raising arguments not previously made before the Tribunal.

For completeness, we will now address the parties’ respective submissions and assess the preclusive effects, if any, of the various determinations in these fora.

(1) The NCLT, NCLAT and Supreme Court of India judgments

On 18 January 2021, pursuant to section 271(c) of the Indian Companies Act, 2013, Antrix applied before the NCLT to wind up Devas based on allegations of fraud (see [39] above). The NCLT allowed the application and accordingly ordered Devas to be wound up. The NCLT, the NCLAT and the Supreme Court of India disallowed Devas’ application to cross-examine witnesses. It is to be noted that neither DT nor its subsidiary DT Asia were parties to the proceedings before the three tribunals/courts. On appeal by Devas, the NCLAT and the Supreme Court of India upheld the NCLT’s decision that Devas should be wound up.

India does not say that the three judgments should formally be recognised by this court. Nor does India claim that the three judgments have res judicata effect before us. India instead submits that the three judgments constitute a compelling basis for this court to conclude that DT’s investment was not in accordance with India’s national laws by reason of fraud or illegality. In any event, India argues that the common law rules for recognition of a foreign in rem judgment are met in this case. By in rem judgment, India means a judgment that by its nature is binding on the whole world. In particular, India says that the SCI Judgment is final and conclusive. India further suggests that there are no valid defences to recognition.

India also submits that we should accord not just weight, but binding effect, to passages in the SCI Judgment (such as those quoted in [52], [55] and [56] above) as findings that DT was involved in Devas’ fraud. For the alleged in rem effect of the SCI Judgment, India relies on the following evidence of its Indian law expert:

In response, DT contends that the three judgments do not have legal effect in the present proceedings. Among other reasons, they do not satisfy the requirements for the recognition of foreign judgments, because there is neither identity of parties nor identity of issues with the present proceedings. Crucially, DT was not a party to those proceedings. There are also concerns that questions of fraud had been determined in a manner that cannot be regarded as being final and conclusive. Even if they are, there are limits to whether findings of fact would bind non-parties to the proceedings.

DT also contends that the judgments were obtained in breach of natural justice, because (among others) Devas was deprived of the opportunity to cross-examine Antrix’s witnesses despite the serious allegations made of fraud and illegality, and further as the NCLT had rushed the appointment of the provisional liquidator (a government employee) without affording Devas an opportunity to be heard on the same. Recognising these decisions would be a contravention of Singapore’s public policy.

We are not persuaded by India’s submissions for a number of reasons.

First, DT was not a party to the proceedings that resulted in the three judgments and took no part in them. We therefore do not see how they can constitute evidence of fraud or illegality on the part of DT. The three judgments concern whether Devas should be wound up for fraud, and are not judgments in criminal proceedings. Granted, the judgments contain remarks about fraud and illegality on the part of Devas’ shareholders and directors. But these are, in our judgment, at best broad-brush obiter remarks (as opposed to actual findings of fact) which, especially in the absence of DT as a party, can hardly be regarded as binding on DT before this court. To proceed as if those remarks were binding and conclusive as findings of fact involving fraud and illegality on DT’s part for the purposes of these proceedings, would, in our judgment, constitute a denial of elementary notions of natural justice and due process. In civil proceedings, persons accused of fraud and serious illegality normally have a right to be heard before being found liable.

It cannot be gainsaid that a finding of fraud against a party is a very serious finding and one which requires cogent evidence to be brought to the fore to undergird it. It has not escaped our attention that none of the judgments in fact implicates DT in any way. Having carefully examined the SCI Judgment in particular, we do not consider the Supreme Court of India’s comment (see [68] above) that if “the seeds of the commercial relationship between Antrix and Devas were a product of fraud perpetrated by Devas, every part of the plant that grew out of those seeds, such as the [Devas-Antrix] Agreement, the disputes, arbitral awards, etc, are all infected with the poison of fraud” to be a specific finding directed at DT, DT’s investment in Devas, or the Final Award obtained by DT against India.

India argues that DT could have applied to be joined as a party to the proceedings before the NCLT, the NCLAT and the Supreme Court of India. That submission is, in our view, untenable. DT had no prior notice that it (as opposed to Devas) stood accused of fraud as part of the winding-up proceedings and that DT risked being found by the relevant tribunals/courts to have acted criminally in its absence. In those circumstances, it is unclear why DT would be under an obligation to apply to be joined as a party. In any event, neither the Indian Companies Act, 2013, nor the Indian Companies (Winding Up) Rules, 2020 (“Winding Up Rules”) gives a shareholder of a company which is the subject of winding-up proceedings the right to intervene or participate in those winding-up proceedings. The NCLT dismissed the impleadment (joinder) application brought by DEMPL, another shareholder of Devas, for lack of locus standi because “[it] has no grievance against the affairs of Devas and its Directors”. DEMPL’s appeal to the NCLAT was dismissed as “not at all maintainable” because Devas’ “shareholders have no role in the instant proceedings at the present stage, as their liability is limited to their share-holding”. In contrast, the SCI Judgment stated (at [11.9]) that “the dismissal of the appeal filed by DEMPL, by NCLAT on the ground of maintainability may not be correct” because “[t]o say that DEMPL cannot be taken to be a person aggrieved, may be farfetched”. But, regardless of the position on “maintainability”, the Supreme Court of India refused the appeal against the rejection of DEMPL’s impleadment application on the ground that “there is no scope either in the [Companies] Act or in the [Winding Up] Rules for the impleadment of any shareholder as a respondent to the petition for winding up” (at [11.3]).

Second, in any event there are significant difficulties in taking the remarks within the three judgments as final and conclusive findings of fraud on DT’s part for the purpose of the Singapore proceedings.

The NCLT decided the winding-up petition on a summary basis by reference to documents alone, without cross-examination of witnesses. It stated in its judgment (at [19(12)]):

For context, under section 271(c) of the Companies Act, 2013, a company may be wound up on three grounds, namely that: (a) the affairs of the company have been conducted in a fraudulent manner; (b) the company was formed for a fraudulent and unlawful purpose; or (c) persons concerned in the formation or management of its affairs have been guilty of fraud, misfeasance or misconduct in connection therewith.

On appeal on this issue, the NCLAT held as follows (at [238] of the concurring judgment of Mr V P Singh):

In the ordinary course, it would be a matter of some significant concern to this court, that parties accused of fraud or analogous conduct imputing serious criminality were not allowed to adduce oral evidence or cross-examine witnesses on what is stated in documents. To find a party guilty of fraud without affording that person an opportunity to adduce oral evidence or cross-examine witnesses, including as to the contents of allegedly incriminating documents would, in the absence of compelling reasons, constitute a breach of natural justice and due process.

It appears from the foregoing passages that the NCLT and the NCLAT regarded themselves as exercising a summary jurisdiction to determine whether, on the face of what were described as undisputed documents, there was a case for winding up Devas on one or more of the grounds in section 271 of the Indian Companies Act, 2013. Having found that there was such a case, the NCLT ordered that Devas be wound up and the NCLAT upheld that decision. As Mr V P Singh of the NCLAT stated in the passage quoted, the question of “who was responsible for committing the fraud” was not one with which the NCLT or the NCLAT were concerned. That understanding of the effect of the NCLT and NCLAT judgments is seemingly confirmed by M Venugopal J’s lead judgment in the NCLAT (at [331]):

The fact that no cross-examination of witnesses had been allowed was again raised before the Supreme Court of India as a ground of appeal. The Supreme Court of India reasoned:

The reference in M Venugopal J’s judgment to only a prima facie case for winding up having been made out was also brought to the Supreme Court of India’s attention. On that, the Supreme Court commented:

Then, having reviewed what it referred to as the “undisputed facts” which emerged from the documents before the NCLT and the NCLAT, the Supreme Court of India continued:

While the Supreme Court of India considered that M Venugopal J had mischaracterised the findings of the NCLT and the NCLAT, the Supreme Court of India did not explain how a judge who expressly described his finding as prima facie can nonetheless be treated as having come to a final conclusion. For the purposes of the proceedings before us, we do not think that M Venugopal J’s description of what he believed he was doing can be so readily disregarded. Nor did the Supreme Court of India say anything about Mr V P Singh’s observation that the NCLT and the NCLAT judgments were not concerned with who was responsible for fraud. Even if Mr V P Singh’s remark is also treated as erroneous, we would, with respect, be uncomfortable using the view expressed (ie, that cross-examination would have been pointless) as justification for us to accept the three judgments as providing the necessary compelling evidence of DT’s complicity in fraud for the purposes of the proceedings before us – that would require us to take a quantum leap that we are not prepared to. In light of the problems with India’s case against DT highlighted in [42]–[112], it seems to us that oral evidence from DT as to what it knew or did not know at any given time and cross-examination of the makers of the documents adduced before the NCLT as to their precise ambit and underlying circumstances, would be material in establishing fraud (if any) on DT’s part. The documentary evidence adduced may have been final and conclusive for the purposes of winding up Devas on the ground of fraud as a matter of India’s national law. But we do not think that the three judgments can be regarded as sufficient evidence, compelling or otherwise, of fraud and illegality for the purposes of the proceedings before us.

Given our conclusion on the evidentiary effect of the three judgments as far as these proceedings are concerned, it is unnecessary for us to examine whether and (if so) the extent to which the SCI Judgment operates in rem under India’s national law. It suffices for us to register our difficulty with accepting the proposition that the SCI Judgment, even if it were regarded as containing a finding of fraud in which DT was complicit, would operate in rem to bind DT as to that matter before us.

(2) The Swiss Federal Supreme Court judgment

DT submits that India is precluded from now raising arguments on fraud and illegality that were rejected by the Swiss Federal Supreme Court. DT further maintains that India is estopped from raising arguments in these proceedings that could have been (but were not) raised before the Swiss Federal Supreme Court.

India contends to the contrary. India says that its arguments on fraud and illegality were not fully canvassed before the Tribunal and the Swiss Federal Supreme Court did not make final and conclusive findings from which estoppel on those issues may arise. India suggests that the Swiss Federal Supreme Court was constrained by the facts established by the Tribunal in the Interim Award which India was then challenging. Consequently, the Swiss Federal Supreme Court did not have the benefit of the findings of fraud in the NCLT, NCLAT and Supreme Court of India judgments. The Swiss Federal Supreme Court explicitly noted (India argues) that the criminal allegations relied on by India “were not yet decided” and a different outcome may be justified where a final criminal decision has been issued. According to India, we are consequently fully entitled to reconsider India’s jurisdictional objections based on fraud and illegality, despite the Swiss Federal Supreme Court’s refusal to set aside the Interim Award on those grounds.

More particularly, cause of action estoppel (India argues) does not arise because an application to enforce an award in one country is a different cause of action from an application to enforce the same award in another country. It follows a fortiori that an application to set aside an interim award before the seat court is a different cause of action from an application to resist enforcement of the final award in another jurisdiction. Nor (India reasons) does issue estoppel arise. This is because, for a foreign judgment to give rise to issue estoppel, the decision on the specific issue must be final and conclusive under the foreign law (here, Swiss law). The Singapore Court of Appeal recognised that “in certain jurisdictions, binding effect might be accorded to the result arrived at in a judgment, but not to the reasons, intermediate steps or other elements that led to that result even if they are stated in the judgment”: see Merck Sharp & Dohme Corp (formerly known as Merck & Co, Inc) v Merck KGaA (formerly known as E Merck) [2021] 1 SLR 1102 at [43]. India suggests that Switzerland is such a jurisdiction and cites in support the expert evidence of its Swiss law expert. The latter takes the view that in Swiss law there is “no equivalent to the common law doctrine of issue estoppel, whereby a finding by a court on a specific issue may bind a future court” and “the general principle is that only the operative part of a decision, to the exclusion of the reasoning, has res judicata effect”. This means that the factual findings and legal considerations underlying the Swiss Federal Supreme Court’s judgment do not constitute its operative part. All that is operative is the Swiss Federal Supreme Court’s conclusion that “[t]he appeal is rejected insofar as the matter is capable of appeal”.

Even if issue estoppel is pertinent, India submits that it should not be precluded from raising arguments of illegality, because of the Arnold exception to issue estoppel. The Arnold exception as identified in The Royal Bank of Scotland NV (formerly known as ABN Amro Bank NV) and others v TT International Ltd (nTan Corporate Advisory Pte Ltd and others, other parties) and another appeal [2015] 5 SLR 1104 (referencing Arnold v National Westminster Bank plc [1991] 2 AC 93) applies where material relevant to the correct determination on the issue of illegality has become available. India says that, because of the three judgments and evidence discovered after the Swiss Federal Supreme Court’s judgment, none of which could reasonably have been adduced earlier, it is plain that the Swiss Federal Supreme Court’s judgment was wrong on illegality and fraud.

India lastly asserts that there would be great injustice if it were precluded from arguing illegality and the Final Award were enforced.

We accept that, for the reason highlighted by India, cause of action of estoppel does not arise here. The issue is whether the Swiss Federal Supreme Court’s judgment gives rise to issue estoppel or (if not) some other form of preclusion.

India raised the indirect investment, the pre-investment expenditure and the “essential security interests” arguments before the Swiss Federal Supreme Court. All were rejected on reasoning similar to that of the Tribunal.

India additionally argued the illegality of the Devas-Antrix Agreement as it has done before us. The Swiss Federal Supreme Court summarised India’s submissions on illegality thus:

For context, Article 190(2) of the Swiss Private International Law Act (“PILA”) (cited in the extract’s last sentence) provides that:

The Swiss Federal Supreme Court found (at [4.4.1]) that, in the absence of contrary indication, whether an investment complied with India’s national laws was “a condition relating to the jurisdiction of the Arbitral Tribunal”. It followed that the Swiss Federal Supreme Court “may, in principle, freely review the relevance of the reasons given on this point by the arbitrators”.

However, referring back to [3.2.3.3.1] of its judgment (which is set out more fully at [170] below), the Swiss Federal Supreme Court held at [4.4.2] that India “ha[d] forfeited the right to argue … lack of jurisdiction … in connection with the compliance clause, unless it could [not] reasonably have raised such an objection before the time when it did so”. The Swiss Federal Supreme Court was not persuaded by India’s “assertion” that it had raised the non-compliance objection at the earliest moment that it reasonably could have done. The Swiss Federal Supreme Court stated:

India lastly argued that its right to be heard in relation to illegality had been prejudiced due to the Tribunal’s refusal to admit the CBI Charge Sheet as evidence. The Swiss Federal Supreme Court likewise rejected this argument.

Having disposed of all of India’s contentions, the Swiss Federal Supreme Court pronounced that India’s “appeal [is] rejected insofar as the matter is capable of appeal”.

It will be seen that, by reason of Article 186(2) of the PILA, which governed the Arbitration, the Swiss Federal Supreme Court viewed India as having forfeited (that is, waived) the right to raise non-compliance with India’s national laws and “essential security interests” as jurisdictional objections. The effect of Article 186(2) of the PILA is further discussed below, from [156]. For present purposes, the focus is on the extent to which India is estopped or precluded from raising (a) jurisdictional objections that the Swiss Federal Supreme Court rejected and (b) jurisdictional arguments that it did not raise before the Swiss Federal Supreme Court.

The Swiss Federal Supreme Court itself explained the juridical effect of its decision:

Given that explanation, we conclude that the Swiss Federal Supreme Court’s judgment does have negative res judicata effect to the extent that it rejected the grounds of review raised by India. That negative effect precludes India from raising the same grounds of review in later proceedings. In other words, India may not later challenge the Tribunal’s jurisdiction on the indirect investment, pre-investment expenditure and “essential security interests” grounds that it unsuccessfully raised before the Swiss Federal Supreme Court. On illegality, India is, in our judgment, likewise precluded by negative res judicata from later challenging the Tribunal’s jurisdiction on grounds of illegality, that is, non-compliance of DT’s investment with India’s national laws. But the Swiss Federal Supreme Court left open the possibility of a challenge to the Interim Award in the circumstance of “a final, criminal decision, likely to affect the final award not yet issued as of this date, ... taken after the pronouncement of the award, ... if necessary and all other conditions being met”. Although not elaborated upon, that is presumably a reference to the possibility of revision under Swiss law. Under Article 190a(1)(a) of the PILA, an applicant may seek revision where it becomes aware of significant facts or uncovered decisive evidence (which it could not have produced in the earlier proceedings despite exercising due diligence) after the relevant award was rendered.

The foregoing conclusions we have reached are consistent with the evidence of India’s Swiss law expert, Professor Christoph Müller. He stresses (and we accept) that Swiss law has “no equivalent to the common law doctrine of issue estoppel, whereby a finding by a court on a specific issue may bind a future court”. He distinguishes instead between formal and substantive res judicata in Swiss law. Formal res judicata means that “no ordinary means of recourse can be brought against [a] decision”, leaving only extraordinary means such as a request for revision. Substantive res judicata, on the other hand, means that “the matters decided in a decision cannot become the object of later proceedings between the same parties which concern the same subject-matter”. In so far as the Swiss Federal Supreme Court’s judgment is concerned, he takes the view that, whether looked at substantively or procedurally, it can at best only give rise to negative res judicata. He writes:

Accordingly, apart from an application for revision under Swiss law based on material evidence that could not previously have been adduced, India is precluded from challenging the Tribunal’s jurisdiction on the grounds that the Swiss Federal Supreme Court rejected. This is by operation of the doctrine of res judicata, rather than because it is estopped by the Swiss Federal Supreme Court’s findings on discrete issues. As Prof Müller stated: “Consequently, the effect of the Federal Supreme Court’s rejection of India’s Application against the Interim Award is simply that India cannot file another application in Switzerland to set aside the Interim Award or the Final Award on the same grounds as traversed in India’s Application against the Interim Award.”

Applying res judicata principles under Singapore law (and specifically, those relating to issue estoppel), we take the view that India must now be treated as barred from raising the very jurisdictional objections that the Swiss Federal Supreme Court has already rejected. Considering the elements of res judicata under Singapore law (see BAZ v BBA and others and other matters [2020] 5 SLR 266 (“BAZ v BBA”) at [30]): (a) the Swiss judgment is final and conclusive in respect of its rejection of the arguments run before the Swiss Federal Supreme Court; (b) there is identity between the parties in the Singapore and Swiss proceedings; and (c) the subject matter of both proceedings is the same, namely, the Tribunal’s jurisdiction under the BIT. It follows from propositions (a), (b) and (c) that India is now bound by the outcome of the Swiss Federal Supreme Court (that is, the rejection of India’s jurisdictional objections) and the same arguments cannot be re-ventilated before us. The rationale underpinning res judicata generally (including the doctrine of issue estoppel) is that a party should not be twice troubled on the same matter: BAZ v BBA at [32]. It would be contrary to the public policy of finality if, India’s jurisdictional objections having been rejected by the seat court, India should nonetheless be permitted to resurrect the same objections in an application to resist enforcement (see also BAZ v BBA at [39] and PT First Media TBK (formerly known as PT Broadband Multimedia TBK) v Astro Nusantara International BV and others and another appeal [2014] 1 SLR 372 at [75]).

India relies on MAD Atelier International BV v Manès [2020] 3 WLR 631 (“MAD Atelier”) as support for its case that it is not so barred. In that case, MAD sued Manès in England on claims analogous to those brought against Manès’ companies in France. The Paris Commercial Court had decided the latter claims against MAD. But the evidence in the English proceedings was that French law had no equivalent to the doctrine of issue estoppel and that res judicata under French law only attaches to the operative part of a decision. Thus, if a fact found by the Paris Court did not appear in the dispositif of its judgment, “the relitigation of that fact is not prevented by the preclusive effect of that judgment” (MAD Atelier at [89]). It followed that, if MAD “were to bring a new civil claim in France against [Manès] claiming a different remedy or based on a different cause of action, [it] would not be prevented from relitigating any of the issues in the new proceedings” (MAD Atelier at [91]). There was the additional difficulty that the parties to the French and English actions were not the same so that the French decision could not be final and binding anyway as between the parties to the English action (MAD Atelier at [105]). As a result, it seems to us that MAD Atelier involved a different factual situation and is of little assistance here. In MAD Atelier there was neither issue estoppel nor res judicata in the cause of action estoppel sense. In contrast, there is here negative res judicata in the issue estoppel sense, subject only to the possibility of revision based on material evidence that was not available at the time of the setting aside application before the Swiss Federal Supreme Court. On whether the three judgments constitute such new material evidence, see our discussion at [156]–[171] below.

Finally, on the assumption that the Arnold exception applies to cases of cause of action res judicata in the same way that it applies to issue estoppel res judicata, for the reasons discussed at [117]–[135] above, we do not think that the NCLT, the NCLAT and the Supreme Court of India’s judgments constitute any evidence (let alone compelling evidence) of fraud or illegality in respect of DT’s investment.

(3) The Interim and Final Awards

DT submits that India is precluded from raising arguments that could have been (but were not) raised before the Tribunal. DT observes that, as the Arbitration was seated in Geneva, it was subject to chapter 12 of the PILA, and in turn, Article 186(2) of the same. Article 186(2) of the PILA provides: “Any objection to [the tribunal’s] jurisdiction must be raised prior to any defence on the merits.” The upshot (DT argues) is that a party which did not raise a jurisdictional objection which it could have raised to the tribunal in a Swiss arbitration, has forfeited (waived) its right to raise that objection in later enforcement proceedings. Reference may also be made to Article 30 of the UNCITRAL Arbitration Rules 1976 which governed the Arbitration. Article 30 stipulates:

India contends that an enforcement court is “entitled, indeed obliged, to undertake a fresh examination” of whether grounds for refusing enforcement (including lack of jurisdiction) have been established. This court is therefore not bound (India says) by the Tribunal’s conclusions on India’s jurisdictional challenges. It follows (India contends) that there can be no bar to introducing new arguments not previously raised before the Tribunal.

As described at [7]–[11] above, the Arbitration between DT and India commenced on 2 September 2013 with the filing of a Notice of Arbitration. The Tribunal issued the Interim Award on 13 December 2017. By its Procedural Order No 1 dated 22 May 2014 the Tribunal had bifurcated the Arbitration into an initial phase on jurisdiction and liability followed by a second phase on damages. The Interim Award thus addressed all issues of jurisdiction and liability. There was the prior ICC arbitration (seated in Delhi) between Devas and Antrix which started in or around June/July 2011 and concluded on 14 September 2015 with the issue of a final award (the ICC Award). The ICC Award ordered (among other matters) that Antrix pay damages of US$562.5 million (with simple interest at 18% per annum) to Devas for the wrongful repudiation of the Devas-Antrix Agreement. Antrix filed an action before the Indian courts for annulment of the ICC Award. In the Arbitration, the Tribunal’s position was that, in the absence of a compelling reason to the contrary, it “should accord deference to the findings of the ICC tribunal, being the forum entrusted with the settlement of contract disputes” (at [114] of the Interim Award).

By letter dated 24 October 2016, India attempted to adduce the CBI Charge Sheet into the Arbitration as an important recent development. In the Interim Award, the Tribunal commented as follows on the CBI Charge Sheet (at [115]–[119]):

The Tribunal then considered India’s three preliminary objections to the Tribunal’s jurisdiction. Those were that (a) the investment had been made indirectly through DT Asia and so was not covered by the BIT, (b) DT’s investment was effectively only a pre-investment expenditure, and (c) the BIT was inapplicable because the dispute concerned India’s essential security interests. Much as this court and the Swiss Federal Supreme Court have done, the Tribunal rejected all three arguments on their merits. We note in passing that there was a belated application by India on 14 March 2017 to introduce the travaux underlying the India-Netherlands BIT. The Tribunal denied that application on 20 March 2017. Having rejected the jurisdictional objections, the Tribunal went into the question of liability and concluded that there was a breach of India’s obligation of fair and equitable treatment.

In brief, of the four grounds raised by India as placing the parties’ dispute outside of the BIT, three (indirect investment, pre-investment expenditure, and essential security interests) were argued before the Tribunal. Of India’s illegality grounds, three illegality allegations (namely that: (a) DT was complicit in Devas’ fraudulent misrepresentations, (b) DT’s investment was contrary to FIPB approval conditions, and (c) DT’s investment was contrary to India’s SATCOM Policy) were not raised before the Tribunal. It is those three allegations that DT claims that India is now estopped from raising by reason of Article 186(2) of the PILA. DT also complains that India is precluded from running its “essential security interests” argument as a jurisdictional objection. We have already dealt with the substantive nature of the “essential security interests” argument (see [102]–[112] above), so there is no need for us to consider whether there is preclusion as far as concerns the re-packaging of that argument as a jurisdictional objection.

In our view, the failure to raise the three illegality allegations before the Tribunal has the consequence that India must be deemed to have waived the same as jurisdictional objections.

Article 186(2) of the PILA is analogous to Article 16(2) of the UNCITRAL Model Law on International Commercial Arbitration 1985 (the “Model Law”), which applies in Singapore (by virtue of section 3 of the IAA) and states:

As the Court of Appeal held in Rakna Arakshaka Lanka Ltd v Avant Garde Maritime Services (Pte) Ltd [2019] 2 SLR 131 at [51], the drafters of the Model Law intended Article 16(2) to have preclusive effect. As to the nature of this preclusive effect, in BAZ v BBA, Belinda Ang Saw Ean J (as she then was) contrasted the extended doctrine of res judicata with that of waiver. She stated (at [63]–[64]):

The question is therefore not one of preclusion by estoppel, but of preclusion due to waiver. The issue is whether a party should be treated as having waived the right to raise a jurisdictional objection by failing to raise it in an arbitration. As a matter of fairness, waiver will only apply if a party was aware of the matters underlying the jurisdictional objection so that it could have objected in a timely fashion during the arbitration. As Belinda Ang J observed in BAZ v BBA at [59], “waiver by a party under Art 16(2) only applies if the objection was clear to the party and the party knew of the objection”. Given the similarity in the wording of Article 16(2) of the Model Law and Article 186(2) of the PILA, it would be odd if the two provisions functioned differently and it has not been suggested by either party that they do function, or should be interpreted, differently. In the present context of the Arbitration, we are also fortified in our view by Article 30 of the UNCITRAL Arbitration Rules 1976, which likewise requires an objection to be made promptly.

India disagrees with this analysis. India submits that Article 16(2) should only be engaged when a party raises no plea at all that the tribunal lacks jurisdiction but instead participates in the arbitration on the merits. In contrast, India did not accept DT’s invocation of arbitration but pleaded that the Tribunal lacked jurisdiction from the outset. In any event, Article 16(2) (India submits) is “considerably tempered” as a tribunal may “admit a later plea if it considers the delay justified”. Such discretion (India suggests) may also be exercised by an enforcement court. India contends that its actions in relation to its illegality objection were justified because, when it filed its Counter-Memorial in the Arbitration in February 2015, it was still carrying out confidential investigations and had yet to uncover evidence that only came to light later. India moved to bring the CBI Charge Sheet to the Tribunal’s attention as soon as it was published. The facts are now known as a result of the Supreme Court of India’s conclusive ruling on DT’s involvement in pervasive fraud.

We do not accept India’s submissions.

It is not apparent why Article 16(2) of the Model Law or the analogous Article 186(2) of the PILA should be read as being limited to the situation where a party does not object to jurisdiction at all and simply proceeds with an arbitration. On their plain and ordinary meaning, the two provisions are categorical in requiring a party to raise an objection or plea against jurisdiction prior to defending an arbitration on the merits. If the party fails to do so, then it will be deemed to have waived the unargued jurisdictional point in the absence of valid justification. There is no reason why a party should be allowed to keep some jurisdictional objections up its sleeve, for later deployment in setting aside or enforcement proceedings if it should lose the arbitration. The raison d’etre of provisions like Article 16(2) is to have parties raise their jurisdictional objections at the earliest possible time: Hunan Xiangzhong Mining Group Ltd v Oilive Pte Ltd [2022] SGHC 43 at [42]–[45]. They are indicative of an “up-front” or “cards on the table” approach to dispute resolution whereby a party participating in an arbitration is required to put forward its entire case on the lack of jurisdiction at the outset to enable a tribunal to rule comprehensively on all objections.

Nor do we think that India had good reason for holding back on the three allegations due to lack of evidence and still-ongoing investigations. As we have pointed out (at [74]–[93] above), India’s illegality case against DT hinges on nothing more than a tenuous inference that DT was complicit in the fraud because it conducted due diligence and had two nominees on the Devas Board. We have queried the validity of such an inference. But, even taking India’s case at face value, it is difficult to see why it could not have adduced the facts and matters said to support such an inference by the time of its Counter-Memorial in February 2015. We have seen that by the time of the Sinha Report in September 2011, all facts underlying the alleged fraud and the charges in the 2016 CBI Charge Sheet were already known. Certainly, all facts said to support the inference that we are invited to draw were already manifest. It was known that Devas and its original investors had offloaded their shares at substantial premiums to new investors. It was known that moneys had been transferred abroad. It was known that Devas lacked the technology and intellectual property rights that it claimed to have. It was known that several high-level government officials had engaged in serious irregularities and failed to observe proper procedures. It was known that the SATCOM Policy had not been observed in a transparent fashion. Yet none of these matters was deployed before the Tribunal as part of a jurisdictional objection premised on fraud and illegality that also implicated DT and its investment. All that India did was to adduce the CBI Charge Sheet, belatedly, in October 2016. But the CBI Charge Sheet was at best (as the Tribunal observed in the Interim Award at [119]) a series of allegations. India could have (but did not) adduce the underlying facts and matters to the Tribunal (all of which had by then been reviewed on several occasions by various committees) (see [80]–[84] above) in support of a jurisdictional objection. India relies heavily on the NCLT, NCLAT and Supreme Court of India judgments as material evidence. For the reasons discussed above at [122]–[135], we disagree that the three judgments provide any evidential basis for finding that DT was complicit in fraud or that its investment was illegal. However, similarly taking India’s case in respect of the three judgments at face value, the facts and matters underlying those judgments were available and known to India as early as the Sinha Report in September 2011. It could therefore have drawn the Tribunal’s attention to those matters in support of a jurisdictional objection, but for whatever reason decided not to do so.

We note further in support of our reading of the effect of Article 186(2) of the PILA, the following passage from the Swiss Federal Supreme Court judgment:

In those premises, we do not think that India’s failure to raise the three illegality allegations in the Arbitration was justified or excusable. In our judgment, India is to be treated as having waived its right to raise the three allegations pursuant to Article 186(2) of the PILA, which governed the Arbitration.

Conclusion on SUM 155

In summary, none of India’s illegality grounds is tenable. Neither are its arguments on pre-investment expenditure, indirect investment, or essential security interests. India’s jurisdictional objections to the Final Award being enforced are thus rejected. In our judgment, the exception to state immunity in section 11(1) of the SIA applies and the Final Award is enforceable against India. None of the grounds for refusing to enforce the Final Award under the IAA applies. As such, we dismiss SUM 155 in its entirety.

SUM 24 – application to stay hearing of summonses

By SUM 24, India submits that we should stay our determination of SUM 155 and SUM 720 pending the outcome of the Swiss Revision Application. It framed its application as a case management stay, grounded upon the court’s inherent jurisdiction and section 18I(1) of the Supreme Court of Judicature Act 1969 (2020 Rev Ed) (the “SCJA”) (read with section 18(2) and paragraph 9 of the First Schedule to the SCJA). India says that the Swiss Revision Application has a realistic prospect of success. A stay (it is contended) will save judicial resources and avoid potentially inconsistent judgments between the Swiss and Singapore courts, as the same arguments being run before us by India are essentially also being pursued in the Swiss Revision Application. India claims only to have become aware of significant facts or uncovered evidence recently, so that it could not have raised those matters earlier. In particular, India points out that the findings of fraud by the Supreme Court of India, India’s highest court, were only handed down on 17 January 2022. Within the stipulated time limit for applying for revision, India took out the Swiss Revision Application following the handing down of the SCI Judgment.

We are not persuaded by India’s submissions.

First, as the Supreme Court of India noted (see [132] above), it does not find facts. Its role is constrained to considering whether the NCLT and the NCLAT erred in the application of the law to the facts as found by those two tribunals. Therefore, all relevant facts would have been known to India by 8 September 2021 at the latest, when the NCLAT handed down its judgment. The contention that India only became aware of relevant material facts or evidence through the SCI Judgment therefore does not withstand scrutiny.

To bolster its case, India refers to an extract from a 2021 report by Devas’ provisional liquidator. According to the report, Devas Delaware contracted “to internally develop product design which will be transferred to [Devas] for use in production of satellite system to support telecom in India” and Devas Delaware was “actually retaining control” of Devas, rather than Devas controlling Devas Delaware (the subsidiary). India suggests that the fraud by Devas’ shareholders was thereby only revealed with clarity at that juncture, such that “the fraud was finally pieced together with the benefit of fresh evidence, leading the Supreme Court to hold that Devas’ shareholders could not ‘feign ignorance and escape the allegations of fraud’”. To this, DT argues that the 2021 report does not present new evidence, but merely observations from a provisional liquidator who had been appointed under “highly unusual circumstances”. Leaving the circumstances aside, we are unable to discern how the 2021 report comes anywhere near to demonstrating DT’s involvement in fraud. In any event, India did not take out a revision application before the Swiss Federal Supreme Court at any point in 2021 following the provisional liquidator’s report. The reality is that India’s case before us and the evidence adduced in support of it have essentially remained as those that were before the Tribunal and the Swiss Federal Supreme Court. We are asked to infer DT’s culpability from the slender facts that DT conducted due diligence before making its investment and thereafter had nominee directors on the Devas Board. As we have already explained, it is not possible to do so without more. But nothing more has been forthcoming by way of material new evidence, whether through the three judgments or otherwise.

Consequently, in our assessment, the Swiss Revision Application has minimal prospect of success before the Swiss Federal Supreme Court. As explained above at [150], under Article 190a(1)(a) of the PILA, for an applicant to succeed on revision, it must demonstrate that it became aware of significant facts or uncovered decisive evidence (which it could not have produced in the earlier proceedings despite exercising due diligence) after the relevant award was rendered. A revision application must be taken out within 90 days (excluding holidays) of a party becoming aware of new facts or evidence. India has known of the relevant facts and matters, and the evidence that it says supports the same, from 2011 onwards. For the reasons discussed at [122]–[135] above, the SCI Judgment is not fresh material evidence in respect of alleged fraud on DT’s part. It follows that the threshold requirement in Article 190a(1)(a) has not been met.

For completeness, we note that there are two other grounds for revision under Article 190a(1) of the PILA, but India has not suggested that they are pertinent in this case nor advanced any arguments on those grounds. They are namely that the award was influenced by a felony or misdemeanour, and that a ground for challenging a tribunal member’s independence or impartiality only came to light after the arbitration proceedings.

Second, there has been no saving of judicial resources. That is largely due to India having only belatedly issued its stay application on 17 May 2022. This is despite India being aware of the SCI Judgment since its handing down in January 2022. By the time that SUM 24 was issued, SUM 155 and SUM 720 had been fixed for substantive hearing on 30 June and 1 July 2022. The result was that this court heard SUM 24, SUM 155 and SUM 720 together.

Third, given our view on the minimal prospects of the Swiss Revision Application succeeding, we are also of the view that the risk of inconsistent conclusions being reached by the Singapore and Swiss courts is likewise low at best.

For these reasons, we decline to exercise our discretionary case management powers to stay SUM 155 and SUM 720 pending the determination of the Swiss Revision Application. SUM 24 is accordingly also dismissed.

At the hearing of SUM 24, India’s counsel, Mr Bull, informed us that in the event of the rejection of SUM 24, India would immediately apply for leave to appeal against such refusal. Mr Bull submitted that we should not proceed to deal substantively with SUM 155 as that would render nugatory any application for leave to appeal against the refusal of a stay. We disagree. A decision on the merits of the stay application requires this court to form a view on, inter alia, the prospects of the Swiss Revision Application. That in turn involves consideration of counsel’s submissions before us on SUM 155 so far as relevant to the prospects of the Swiss Revision Application, the two having been heard together in consequence of India’s belated filing of SUM 24. Nor do we believe that any application for a stay would necessarily be rendered nugatory if SUM 24 and SUM 155 are decided together. In the worst case scenario from India’s point-of-view, it would be open to India, if so advised, to appeal against the dismissal of SUM 155, apply for a stay of execution of any enforcement order in relation to the Final Award pending such appeal, and to apply to the Court of Appeal for a stay of the substantive hearing of the appeal pending the outcome of the Swiss Revision Application. India’s concern on the alleged nugatory effect on SUM 24 is thus, in our view, overstated.

SUM 45 – application to submit further evidence

Three weeks after the substantive hearing of SUM 155 and SUM 24 had concluded on 1 July 2022, the US District Court for the District of Columbia granted a stay of proceedings brought by DT in the US for the enforcement of the Final Award. The stay of US proceedings was granted pending the outcome of the Swiss Revision Application. By SUM 45, India sought leave to adduce the US court’s stay order and the papers filed in the US stay application as fresh evidence that will assist us in deciding SUM 24. It says that the decision forms part of the circumstances of the case, which includes the “overall shape of the proceedings worldwide to enforce the Awards”. India notes that the US court held that a stay “avoids the possibility of expensive and duplicative litigation to unwind the arbitral award”, and had observed that the Swiss Federal Supreme Court was expected to decide within six to ten months, with no possibility of appeal. In response, DT’s principal argument is that the US court’s decision is not relevant to the exercise of our discretion.

Having considered the parties’ submissions, we dismiss SUM 45.

First, the US stay application papers were available before the hearing of SUM 24 and SUM 155. India chose not to adduce them in connection with SUM 24 or SUM 155. Thus, the stay application papers do not constitute evidence or events that only arose after the hearing before us.

Second, the fact that the US court granted a stay may be a factor that we can consider in deciding whether to grant a stay of SUM 155, but the fact that a stay was granted by the US court does not, whether considered alone or in all the circumstances, necessitate that we also exercise our discretion in favour of granting a stay. Further, the reasons underpinning the US court order are not evidence of facts or matters to be considered in the exercise of our discretion on whether to grant a stay. The reasons are simply submissions advanced by India in the US proceedings which found favour with the US court. It was open to India to make similar arguments (as it did) before us at the substantive hearing of SUM 24. How the US court weighed India’s arguments is not, in our judgment, pertinent to how we should exercise our discretion whether to grant a stay in light of the totality of evidence and submissions before us. Thus, in so far as India wished to adduce the US court order as evidence of the reasons underpinning the US court’s decision, we found that evidence to have no relevance or probative value to our deliberations on SUM 24.

SUM 720 – application to strike out affidavit evidence

Finally, SUM 720 is DT’s application to strike out, inter alia, parts of India’s expert affidavit evidence on Indian law referring to the NCLT, the NCLAT and the Supreme Court of India’s judgments. DT’s argument was that, since the three judgments have no binding or legal effect and should not be recognised nor given weight, India’s Indian law expert’s evidence on the legal effect of the three judgments should be struck out as irrelevant.

At a Case Management Conference on 10 May 2022, we pointed out that, in practical terms, there was no need for SUM 720. The parties could simply make their submissions on the relevance or irrelevance of the Indian law expert evidence, and we would then decide whether such evidence was material and to what extent we would accept or reject the evidence so adduced.

Conclusion