Arbitrability of Fraud in India: The Rashid Raza test for Complex Fraud

[Anirban Chanda is a 4th year B.A., LL.B. (Hons.) student and Anujay Shrivastava a 5th Year B.A., LL.B. (Hons.) student, both at the Jindal Global Law School. The authors give special thanks to Dr. V. Niranjan and Mr. Anubhav Khamroi for their inputs.]

The term “arbitrability” generally connotes the capability of a dispute or classes of disputes that can be settled by an arbitrator.[1] As far as the Indian legislation is concerned, neither the erstwhile Arbitration Act 1940 nor the Indian Arbitration and Conciliation Act 1996 expressly classifies any dispute to be not capable of settlement by arbitration. This position did not change in the 2015 Amendment or the recent 2019 Amendment to the 1996 Act. Therefore, in India, the arbitrability of a subject-matter is to be decided by the judicial authorities, who may ultimately choose to interfere with arbitration proceedings on a dispute, should they find a dispute to be non-arbitrable.

The landmark decision by a three-judge bench in Abdul Kadir Shamsuddin Bubere v. Madhav Prabhakar Oak was the first judgment by the Indian Supreme Court on arbitrability of fraud. It was held that a dispute containing fraud of any nature cannot be decided upon by an arbitral tribunal (even where the party accused of committing a fraud does not wish to pursue the case in an open court), since questions of fraud involve complicated factual questions that should be decided in an open court. In such a situation, the jurisdiction of an arbitrator would have to be presumed to be ousted. While this decision was rendered during the era of the 1940 Act, the silence of the erstwhile legislation and the 1996 Act made it a law that continued to apply in the present regime. Subsequently, a division bench of the Supreme Court in the N. Radhakrishnan decision had fully adopted the judgment by Madras High Court in H.G. Oomor Sait, which declared that all disputes involving ‘allegations’ of fraud would be not arbitrable. These decisions in the 1996 Act regime were in line with the decision in Abdul Kadir.

The Supreme Court later in World Sport had clarified that with regard to foreign-seated international commercial arbitrations, an arbitrator is competent to decide disputes involving any type of allegations of fraud in a dispute. Therefore, the ruling in N. Radhakrishnan applied only to domestic disputes or disputes in international commercial arbitrations seated in India (India-seated arbitrations). It is pertinent to note that the decision in World Sport has not elucidated upon the legal basis for differentiating between the standards that apply to judicial intervention in arbitrability of fraud in India-seated arbitrations and foreign-seated arbitrations. This anomaly still persists today and has been earlier discussed in great length here

In some of the earlier posts on this Blog, in particular, here, here, here, here, here and here, the development of law on arbitrability of fraud after N. Radhakrishnan has been discussed. Importantly, N. Radhakrishnan was declared per incuriam by a single-judge decision in Swiss Timing for several reasons. Firstly, the decision in Hindustan Petroleum was not distinguished by the bench. Secondly, the decision in P. Anand Gajapathi Raju was not at all discussed by the court. Thirdly, the provisions contained in section 16 of the 1996 Act which provide that the arbitral tribunal would be competent to rule on its own jurisdiction (including ruling on any objection with regard to existence or validity of the arbitration agreement) were not discussed by the court. Lastly, the 1996 Act emphasized that an arbitration clause that forms part of a contract shall be treated as an agreement independent of the other terms of the contract. Consequently, flowing from the last point, a decision by the arbitral tribunal that the contract is ‘null’ and ‘void’ would not entail ipso jure the invalidity of an arbitration clause. However, as the bench strength in Swiss Timing was lesser than N. Radhakrishnan, it caused haziness before High Courts as to which decision should they follow.

This haziness was ultimately addressed by a Division Bench of the Supreme Court in A. Ayyasamy v. A. Paramasivam which upheld Swiss Timing and overruled N. Radhakrishnan, as well as all cases that held the same holding as the latter. In Ayyasamy, the Court held that a judge must distinguish between ‘fraud simpliciter’ (simple allegations of fraud) and ‘complex fraud’ (serious/complex allegations of fraud). It held that disputes involving fraud simpliciter would be arbitrable, while the disputes that involve complex fraud are non-arbitrable. According to the precedents, the Court in Ayyasamy could not, in principle, overrule the N. Radhakrishnan decision due to both their bench-strength being co-ordinate (same strength) [please refer to the Five-Judge Constitution Bench decision of Indian Supreme Court in Central Board of Dawoodi Bohra v. State of Maharashtra]. However, the ruling in Ayyasamy has been recently affirmed by a three-judge bench decision of the Indian Supreme Court in Rashid Raza v. Sadaf Akhtar, which impliedly overrules N. Radhakrishnan. Interestingly, the bench in Rashid Raza has given a new test for determining presence of ‘complex fraud’ which shall be evaluated in the next segment of this post. However, before delving into a critical analysis of the judgment, a brief note on the judicial origins of the concept of ‘complex fraud’ might be pertinent.   

The judicial origins of ‘complex fraud’

One of the earliest enunciations of the concept of ‘complex fraud’ may be found in the opinion of Lord Hoffman in the case of Fiona Trust & Holding Corp. v. Privalov. In his opinion, the learned judge, having duly acknowledged the ‘principle of severability’, identified a few exceptional situations where the grounds for invalidating the main agreement are identical to the grounds upon which the arbitration agreement is invalid. These situations would include cases where one of the parties dispute the very existence of any arbitration agreement binding them in the first place or cases where either of the parties claim that their signature on the arbitration agreement has been forged. Such scenarios form the basis of a situation referred to as ‘complex fraud’ where the very foundational basis of the arbitration agreement is under challenge due to a fraudulent act.

Brief Background of Rashid Raza

In Rashid Raza, the appellant and respondents were in a partnership. One of the partners had alleged that there was siphoning of funds and other business improprieties were committed, which constituted fraud. When this case went to the High Court for appointment of arbitrator under section 11 of the 1996 Act, it observed that there are complex allegations of fraud. Therefore, it dismissed the application by the appellant for appointment of an arbitrator.

The test for determining “complex fraud”

The Court agreed with the distinction made by the division bench in Ayyasamy between ‘fraud simpliciter’ and ‘complex fraud’. In Ayyasamy, the Court had noted that whenever serious allegations of forgery or fabrication of documents in support of the plea of fraud are made, or when fraud is alleged against the arbitration provision itself, or the fraud is of such a nature that permeates the entire contract (including an agreement to arbitration), the fraud would impact the very validity of the contract or the arbitration clause itself. This would have implications on not only the internal affairs of the parties inter se, but would also impact the public domain. While dealing with ‘complex fraud’ or serious allegations of fraud, the court should primarily focus on whether or not the jurisdiction of the arbitral tribunal has been ousted. It should not concern itself with whether the court’s jurisdiction is ousted. The inquiry should be on whether the nature of the dispute is such that it cannot be referred to arbitration, even if there is an arbitration agreement between the parties to the contrary. Additionally, it cautioned that wherever fraud is set up by a party to wriggle out an arbitration agreement, a strict and meticulous inquiry into the allegations of fraud must be done.

Building upon the above holding of Ayyasamy, the Court in Rashid Raza laid down a two-step test to determine complex fraud:

[i.] Does this plea permeate the entire contract and above all, the agreement of arbitration, rendering it void?

[ii.] Whether the allegations of fraud touch upon the internal affairs of the parties inter se having no implication in the public domain?

Applying the test to the present matter, the Supreme Court found that there is no allegation of fraud that would vitiate the partnership deed as a whole, or at least the arbitration clause in the said deed. It further held that all the allegations made pertain to the affairs of the partnership and siphoning of funds from the partnership funds. This was not a matter that would fall under the public domain. Therefore, it overruled the decision of the High Court as both the enquiries of the test were failed. At the same juncture, it is important to note that this ruling applies only to India-seated arbitrations. As far as foreign-seated arbitral tribunals are concerned, they would be governed by the World Sport case and would be competent to decide on arbitrability of any type of fraud.

Need for legislative clarity

The lack of an exhaustive legislative list of arbitrable or non-arbitrable subject-matter is beneficial for an arbitration-friendly regime such as India, especially in light of the enormous number of constant legislative developments. An exhaustive list of arbitrable matters would introduce extreme rigidity that would threaten the very purpose of arbitrations. On the other hand, an exhaustive list of non-arbitrable matters would render the judiciary unable to identify other non-arbitrable matters that could affect public policy and interests. However, codification of a non-exhaustive list of non-arbitrable matters by the legislature could be beneficial guide for the judicial authorities. This would permit judicial authorities to be aware of statutorily non-arbitrable matters. Therefore, adding more clarity, while also giving them the scope of identifying and declaring other non-arbitrable matters on the basis of precedents including the landmark Booz-Allen decision by the Indian Supreme Court (which held that all disputes relating to ‘right in personam’ – a right which is an interest protected solely against specific individuals) are arbitrable, while those relating to ‘right in rem’ (a right against a person, unlike a right in rem which is a right that attaches to a particular “thing” rather than a person or creates a legal status such as citizenship, being exercised against the world as a whole) are not. This would ensure a flexibility for judiciary to add non-arbitrable matters and maintain an arbitration-friendly legislative regime in India.

In Conclusion: Proposing an idea to elevate arbitral tribunals to the same standards as courts of law

The decision in Rashid Raza is a welcome one. It removes the haziness over how courts should approach disputes involving allegations of fraud and lays down a simpler as well as concrete test to determine whether a dispute loses arbitrability on accounts of ‘complex fraud’ or serious allegations of fraud. The court in Rashid Raza had also held that judicial interference into determination of complex fraud must be made after strict factual scrutiny and should not be made an instrument for nefarious parties trying to wriggle out of the arbitration agreement by using frivolous allegations. This decision not only strengthens the holding of Ayyasamy, but would result in lesser judicial intervention into arbitral proceedings. While the judicial approach to arbitrability is appreciable, the creation of a non-exhaustive list of non-arbitrable matters could be beneficial for the judiciary and an arbitration-friendly regime.

Interestingly, the catena of judicial precedents in India and the silence of Indian Constitution on settlement of domestic disputes by arbitration would not permit it to hold a domestic arbitral tribunal to the same standard as a judicial body, unlike a foreign-seated tribunal for which Article 51(c) of the Indian Constitution encourages settlement by. Arguably, the fact that rules of evidence and procedure are not binding on arbitrators, arbitrators are not necessarily law-aware persons or legal professionals, substantive questions of law in ‘complex fraud’ and other inadequacies in arbitration would indeed be valid grounds to differentiate an arbitrator or arbitral tribunal from a court of law. A similar concern was echoed in the judgment of H.G. Oomar Sait where the learned judge, having identified the deficiencies of arbitration as stated above, further opined that even if the arbitrator was to take assistance from the court under section 27 of the 1996 Act – she would still fail to critically appreciate the demeanour of the witnesses which is an essential feature of analysing oral evidence.

The above critique is more important in the present juncture as Rashid Raza upheld paragraph 25 of the Ayyasamy judgment which expressly states as an additional qualification that the serious allegations of fraud must also lead to a virtual criminal offence. The need to qualify for a virtual criminal offence marks a difference with cases with involve complicated commercial misdoings.

There can be several reasons why an arbitral tribunal may be ill-suited to preside over disputes concerning criminal offences. Firstly, an arbitrator does not have the benefit of examining if an accused had been previously convicted. This is pertinent in criminal trials to ascertain the character of the accused. Since the arbitral tribunal cannot access previous arbitral awards it cannot determine if the same person had committed similar fraudulent acts in the past. Secondly, an arbitral tribunal cannot render a criminal conviction while it may have the ability find if the criminal offence has been committed. Similarly, it cannot render a custodial sentence while it can award damages (the damages can have a punitive element based on findings of a criminal act). Lastly, while the court has the power to directly instruct the police or investigating agency the same is not available to arbitrators putting them at a clear disadvantage.

However, if the Indian Constitution is amended to provide for settlement by arbitration under ‘procedure established by law’ and sufficient legislative safeguards (including appeal mechanisms) are enacted for an arbitral tribunal to follow, India could become a jurisdiction that permits arbitral tribunals to adjudicate on fraud in the same manner that a court of law can.

Anirban Chanda & Anujay Shrivastava

[1] Joseph Mante, Arbitrability and public policy: an African perspective, 33 OXFORD ARBITRATION INTERNATIONAL 277 (2017). 

About the author


  • Excellent article. By far, one of my the best posts I’ve read on this blog.

    I like the portion where you’ve carved out the arguments in favour of having legislative amendments to define or provide for, arbitrability of non-exhaustive subject matters. The short and to the point observation on the anomaly brought by Ayyasamy is also well placed.

    As a suggestion, I think that you should explore deeper into the arguments in favour of making India-seated arbitral tribunals as equally competent as courts. Maybe you could do this in another post, and even go into Constitution and Administrative Law angles in this.

    Once again, kudos! Hope to see more posts from you.

  • Dear Chanda and Shrivastava,

    Very good article, as someone above also indicated. However, I differ from your view with regard to making non foreign-seated arbitral tribunals as equally competent as courts.

    You see, most arbitration practitioners in India are terrible when it comes to understanding procedure and are unlikely to follow the evidentiary rules at the same level that commercial courts can. I do take cognizance of the fact that you’ve indicated a need for “APPEAL MECHANISMS”, which means that High Courts and the Apex Court could be potentially be above an arbitral tribunal. This would eliminate the need to approach Trial Courts for remedy and eliminate a level of court. While it may be arbitration-friendly, a removal of level may lay more burden on the high courts and may eliminate a more appropriate level of review by trial court, which is generally better at adducing facts, especially in cases where arbitrability is to be decided.

    However, I note that this was not the primary focus of your article. Also, it would take a research paper to analyse the pros and cons of making non foreign-seated arbitral tribunals as equally competent as courts. But I think that it’ll be a good topic for a future research. Your analysis here can be taken further on.

    Best regards,


Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Top Posts & Pages


Recent Comments


web analytics

Social Media