Publication date: October 05, 2026
When does a State act like a merchant? A scholarly and practical analysis
I. The restrictive theory of State immunity and the significance of the 2004 Convention
The United Nations Convention on Jurisdictional Immunities of States and Their Property, adopted by General Assembly resolution 59/38 of 2 December 2004 (hereinafter the “Convention” or “UNCSI”), is the first comprehensive multilateral instrument to regulate the jurisdictional immunity of States before the courts of foreign States. Although the Convention has not formally entered into force – it requires thirty ratifications, and twenty-five States have ratified or acceded to it so far – it exerts a considerable influence on the formation of customary international law, since it expresses the opinio iuris of States and reflects their practice in the field of jurisdictional immunity. The Convention thus constitutes an important point of reference both for domestic courts and for international tribunals. Many States also regard the Convention as binding to the extent that it codifies customary law already existing in this area.
Article 10(1) UNCSI provides that if a State engages in a commercial transaction with a foreign natural or juridical person and, by virtue of the applicable rules of private international law, differences relating to that transaction fall within the jurisdiction of a court of another State, the State cannot invoke immunity from jurisdiction in a proceeding arising out of that transaction. The provision thus lays down four basic conditions:
(i) the participation of the State in the transaction;
(ii) a relationship between the subject matter of the dispute and the transaction;
(iii) the commercial character of the transaction; and
(iv) the jurisdiction of the foreign court under the applicable conflict-of-laws rules.
It is the requirement of a commercial transaction that causes lawyers the greatest argumentative difficulty. The definition of a “commercial transaction” in Article 2(1)(c) UNCSI covers categories such as: any commercial contract or transaction for the sale of goods or supply of services; any contract for a loan or other transaction of a financial nature, including any obligation of guarantee or of indemnity in respect of any such loan or transaction; and any other contract or transaction of a commercial, industrial, trading or professional nature. The International Law Commission (hereinafter the “ILC”), in its commentary to the 1991 Draft Articles, indicated that the very notion of a “transaction” is broader in scope than that of a “contract” and also embraces activity of a non-contractual character, including commercial dealings and negotiations.[1]
It is also worth noting at this point that Article 10(2) of the Convention provides for two “exceptions to the exception”, upon whose occurrence immunity retains its force notwithstanding the commercial character of the transaction. These are two situations:
(i) where the commercial transaction is concluded between States, that is, on a government-to-government basis; and
(ii) where the parties to the transaction have expressly so agreed.
In these situations immunity remains fully operative and cannot be displaced.
The central practical issue, and the one generating the most disputes in the application of the commercial exception, is the criterion by which a given transaction is to be characterised. Put at its simplest, lawyers and judges ask themselves: how does one tell whether a given act is commercial or sovereign? To answer that question, Article 2(2) UNCSI introduces a two-stage test. According to it, in determining the commercial character of a transaction, reference should be made primarily to its nature, while its purpose should be taken into account where, in the practice of the State concerned, the purpose is relevant to determining the non-commercial character of the transaction.
The ILC commentary specifies that the tests are to be applied in sequence. The nature of the transaction is examined first. If it proves to be non-commercial or governmental at that stage, any further inquiry into purpose is unnecessary.[2] If, however, the nature points to a commercial character, the defendant State is entitled to challenge that finding by reference to the purpose of the transaction, provided that the purpose is, in its practice, relevant to characterising the transaction as non-commercial. The scheme may be illustrated as follows: the court first asks “what is this transaction in itself?”, and only where the answer suggests commercial activity does it ask “for what purpose was it concluded?”. The burden of proving the sovereign purpose of the activity rests on the State that invokes it.
State practice in the application of these analytical tools is not uniform. Although a strong emphasis on the nature test as the principal, decisive criterion is discernible,[3] there is no shortage of cases in which courts say expressly that the entire context of the transaction must be examined in order to determine its character.[4] Two main interpretative positions have accordingly emerged among States.
The first assumes a rigorous stance, manifested in the application of the “nature test” alone, regardless of the purpose of the activity. On this approach, for example, a commercial contract for the supply of boots and uniforms for the army will be characterised as economic activity, and breaches arising from it – provided they fall within the jurisdiction of a foreign court – will be adjudicated there, irrespective of the sovereign purpose of the transaction, namely the provisioning of the armed forces.[5] The very nature of a purchase of uniforms from a private supplier is commercial, and in that relationship the State behaved like a private party. The State’s motivation, by contrast, is irrelevant to the assessment of the transaction.
The second position adopts a somewhat more flexible approach, taking into account, in assessing the transaction, its purpose and the situational context of its conclusion (including, where contracts concluded by State-owned enterprises are concerned, the level of State control and the ownership structure). This conception is also endorsed by the text of the UNCSI, which permits an analysis of purpose where it is relevant to determining non-commercial character.
In Certain Iranian Assets (Islamic Republic of Iran v United States of America) (Judgment of the ICJ of 30 March 2023), the International Court of Justice, addressing the question of jurisdiction over claims concerning Bank Markazi – the central bank of Iran – applied reasoning functionally convergent with the nature-and-purpose test, even though the case did not formally concern the jurisdictional immunity of the State but rather the interpretation of the notion of a “company” within the meaning of the 1955 Treaty of Amity. The Court made clear that the question of “company” status and the rules on sovereign immunity constitute two distinct normative orders, and then held that an entity performing exclusively sovereign functions cannot be characterised as a “company” within the meaning of the Treaty, whereas an entity combining commercial activity with sovereign functions may acquire that status only to the extent that it actually carries on commercial activity “alongside” its sovereign functions. A key element of the Court’s reasoning was its rejection of an approach that examines the character of a transaction “as such”, in isolation from its context – the Court stressed that a transaction must be placed in its context, having regard to its links with the exercise of a sovereign function, and found that the purchase and management of securities by Bank Markazi was merely a way of exercising its essential sovereign function of managing the State’s currency reserves, and not commercial activity carried on independently of those functions. Although this reasoning was formulated on a treaty basis rather than in the context of the commercial exception in Article 10 of the 2004 UN Convention, its logic is structurally identical to the moderate approach to the nature-and-purpose test embodied in the Convention: the nature of a transaction cannot be assessed in the abstract but requires an examination of its functional embedding in the entirety of the entity’s activity, and in particular of its connection with the exercise of sovereign prerogatives – an argument against a mechanical application of the nature criterion in isolation from the institutional and functional context of the act in question. The Court found in that case that acts undertaken “within the framework and for the purposes of” a sovereign objective do not constitute commercial activity but a means of carrying out sovereign functions.[6]
Republic of Argentina v Weltover, Inc (504 US 607 (1992)) is one of the foundational decisions of the Supreme Court of the United States on the application of the commercial exception under the Foreign Sovereign Immunities Act of 1976 (FSIA), and thus bears directly on the displacement of the jurisdictional immunity of the State. The facts concerned Argentine bonds known as “Bonods”, which the Republic of Argentina issued in the early 1980s as part of a governmental currency stabilisation programme: unable to cover its obligations under foreign exchange insurance contracts (FEICs), Argentina issued bonds denominated in US dollars, payable at the creditor’s election on the markets in London, Frankfurt, Zurich or New York. When in 1986 Argentina unilaterally extended the payment deadlines and offered creditors substitute instruments in place of redemption of the overdue bonds, two Panamanian creditors and a Swiss bank – refusing the restructuring and designating New York as the place of performance – brought an action for breach of contract before the federal court in New York. Argentina contested the court’s jurisdiction, invoking jurisdictional immunity.
The Supreme Court, deciding the case unanimously through the pen of Justice Scalia, applied the nature test flowing directly from § 1603(d) FSIA, which expressly requires the commercial character of an activity to be determined by reference to its nature rather than to the purpose for which the State undertook it. The Court squarely rejected Argentina’s argument that the context of the bond issue – serving the sovereign objective of currency stabilisation and management of the foreign exchange deficit – should be decisive of the sovereign character of the transaction, stressing that the statute unambiguously moves away from assessing commerciality through the prism of the motive for the act. In doing so, the Court formulated the key criterion of distinction: a State acts “commercially” where it does not perform the role of a regulator of the market but behaves like a private player within the market, undertaking acts which private parties could perform in the same way.[7] The Bonods were held to be instruments corresponding in every respect to standard debt instruments of private commerce: they were negotiable, dollar-denominated, capable of being held by private parties, and provided a stream of monetary income – none of these features exhibited characteristics reserved exclusively for sovereign activity. The Court further rejected the proposition that an analysis of the entire context of the transaction was necessary in order to establish its commerciality, observing laconically that even viewed in their full context the bonds contained nothing that was not analogous to a private commercial transaction – except precisely their purpose, which by force of the statute remains irrelevant. It is worth noting that the approach adopted in Weltover represents the rigorous variant of the nature test, close to the Anglo-American approach: the purpose of the State’s act is excluded from the analysis altogether, which contrasts both with the moderate position adopted by the UNCSI (which allows purpose to be taken into account as an auxiliary criterion) and with the functional, contextual approach applied by the ICJ in Certain Iranian Assets (2023).
Beijing Urban Construction Group Co Ltd v Republic of Yemen (ICSID Case No ARB/14/30, Decision on Jurisdiction, 31 May 2017) admittedly concerned the jurisdiction ratione personae of an arbitral tribunal in an investment dispute under the China–Yemen bilateral investment treaty, rather than the jurisdictional immunity of the State in the strict sense; nevertheless, the tribunal’s reasoning on the characterisation of the activity of a State-owned enterprise displays significant structural affinities with the nature test applied in the context of the commercial exception. The facts concerned the Chinese State-owned enterprise BUCG which, having won an international tender, concluded a contract in 2006 for the construction of a new terminal at Sana’a airport. In 2009 Yemen was alleged to have used military force to expel BUCG from the construction site and to have expropriated its investment. Yemen contested the tribunal’s jurisdiction, arguing that BUCG, as a wholly State-owned company, was an emanation of the Chinese State and therefore lacked the status of a “national of another Contracting State” within the meaning of Article 25(1) of the ICSID Convention.
The tribunal, rejecting that objection, applied the so-called Broches test, under which a State-owned enterprise is excluded from the personal scope of the Convention only where it acts as an agent of the government or performs essentially governmental functions. What matters for present purposes is that the tribunal clearly separated the purpose of the activity from its nature, stressing that the test must be applied to the specific transaction in dispute, not to the general ownership structure or the strategic objectives of the enterprise. Although BUCG operated in line with general State policy guidance, and the performance of the contract might serve governmental objectives, the tribunal held that the construction activity itself – performed under a contract concluded as the result of an open tender, on market terms and in direct competition with other contractors – was of an “essentially commercial, rather than governmental” character. This reasoning is functionally identical to the logic of the nature test applied within the commercial exception to immunity: the purpose of the entity’s activity (the implementation of State policy) is irrelevant to the legal characterisation of the act if its external form and the manner of its performance correspond to the model of activity of a private party – in this case, a general contractor operating on market terms. The case thus supplies an argument for the proposition that the distinction between the nature and the purpose of an act is a principle of a general character, extending beyond the framework of State immunity and present in the wider system of international investment law.
In the judicial and doctrinal practice of States, a series of criteria has crystallised which assists lawyers in properly characterising the activity of a State. These criteria do not form an exhaustive catalogue or a rigid hierarchy – they are applied jointly and contextually.
The fundamental question is: could the act in question have been undertaken by a private party? If the answer is affirmative, the activity is commercial in character. The purchase of office supplies by a ministry, the conclusion of a lease of a building for an embassy, the issue of bonds on the financial market – in all these cases there exists a private counterpart in the form of a business or an investor performing analogous acts. If, by contrast, the act by its very essence can be performed only by a sovereign – for example, the grant of a licence for the extraction of mineral resources, the establishment of an economic zone, or the imposition of an embargo – then no private counterpart exists and the act must be characterised as acta iure imperii.
This criterion is related to the one described above, but places the emphasis on the manner in which the State behaves. The questions asked here are the following: does the State enter the market on equal terms, like a private party? Does the State act on the market in a manner characteristic of private actors? Does the State refrain from taking advantage, on the market, of any privileges characteristic of a sovereign? In Republic of Argentina v Weltover, Inc (504 US 607 (1992)), already mentioned above, the US Supreme Court held that an act is commercial in character where the State “does not act as a market regulator, but in the manner of a private player within the market”.[8] A good example of commercial activity in the light of the market participant criterion is the situation analysed by the ICSID tribunal in Beijing Urban Construction Group v Republic of Yemen. The arbitral tribunal held that participating in a tender on an equal footing with other private bidders and negotiating the terms of the contract are behaviours typical of a market participant.[9] Where, by contrast, the State makes use of instruments of public authority unavailable to private parties (statute, decree, concession), it cannot then be treated as a market participant.
The ILC commentary indicates that, according to one of the conceptions proposed in the course of the work on the Convention, the profit motive constitutes the most important criterion of the commerciality of a transaction.[10] In practice, however, the criterion is merely auxiliary in character: its presence reinforces a characterisation as acta iure gestionis, but its absence does not automatically determine the sovereign character of the act. A profit motive is, by its very essence, what we associate with the conduct of economic and commercial activity. It must be remembered, however, that many State transactions – such as the purchase of office supplies for an embassy, or agreements signed with scientific and research institutions covering, for example, State funding of research – are commercial in nature even though their immediate objective is not pure profit in the economic sense of the word.
Of fundamental importance is the distinction between regulatory acts (the issue of licences, administrative decisions and permits, the application of customs provisions, the imposition of sanctions) and contractual acts based on a genuine contractual relationship in which the parties have equal rights and neither of them wields any attribute of public authority. Acts in the regulatory sphere are an attribute of sovereignty and are invariably classified as iure imperii. This is so because the State then occupies a privileged position, having a greater or exclusive influence on the shape of the legal relationship, but also because, as a regulator, the State decides – in the name and on behalf of its citizens – about their common goods, for example natural resource deposits. The example of resource extraction also perfectly illustrates the interconnection between the various criteria. Under the criterion here described, when a government grants a concession for the extraction of crude oil, it acts as a regulator. When, on the other hand, it concludes a contract with a contractor for drilling in the territory of another country, this may be characterised as the activity of a contracting party which occupies a position equal to that of the contractor in the transaction. Pursuing the mining industry example, let us then imagine that a State grants a concession to a private company for the extraction of resources in its territory. Knowing the market regulator criterion, we will conclude that it acts in the sphere of iure imperii, since by its act it conditions access to a good belonging to all its citizens – something only a sovereign endowed by the Nation with the appropriate competences can do. In the same situation we may also ask whether a private party could have done this, or whether in this transaction the State is taking advantage of some special privileges giving it a better position. Is the State behaving like a private player on the market? As this example shows, all the criteria listed above interlock and complement one another.
The commercial exception requires the dispute to arise out of a commercial transaction. The commercial character of an act may therefore also be indicated by the existence of a real contractual relationship within which the State has assumed an obligation similar to that of a private person. Simplifying, what is at issue is the existence of a contract binding on the State whose clauses, subject matter and guarantee mechanisms show that the parties acted as equal participants in commerce. Where the dispute arises precisely out of a contract so shaped, immunity cannot serve as a shield against contractual liability.
For the sake of completeness, it is worth identifying the categories of activity which – irrespective of their external form – remain within the sphere of iure imperii and are not subject to the commercial exception. There is no coherent and binding catalogue in this respect, but it is emphasised in the literature that such acts include, above all: the issue of licences and permits; the imposition of taxes and customs duties; the enactment of legislation and implementing acts; the conduct of foreign and defence policy (including in the field of economic security); the conclusion of intergovernmental agreements creating public-law institutions and funds; and the issue of administrative decisions restricting or rationing access to the market. These acts are regulatory and sovereign in character; their essence is the exercise of authority to which a private party has no access.
In legal practice, the most serious error is to conflate the purpose of an act with its nature. For example, a court, or counsel for the State party, may argue that the purchase of soldiers’ boots for a military unit serves the sovereign functions of national defence and is therefore not subject to the commercial exception, its nature being sovereign. But is that really so? In such a situation one should return to the auxiliary criteria of the nature test set out above. Could a private person have made such a purchase? Essentially yes: there are no prohibitive regulations, nor any special entitlements a person must hold, in order to purchase soldiers’ boots, even in large quantities. Did the State act like a private party in this relationship? Yes: the State concluded a contract of sale on terms equal to those of other market participants (we assume, for the purposes of the hypothetical, that it enjoyed no special privileges in this respect). And finally, is there a contract whose content points to a commercial character? A contract of sale is without doubt commercial in nature, being the basic tool of economic exchange. The nature of the transaction under discussion is therefore commercial. Once the nature has been determined, it becomes easier to separate it from the purpose, which in this situation was the provisioning of the army and, in the broader context, no doubt action in the interest of a sound state of national defence. The further characterisation of that purpose, and the determination of how far it matters in the specific factual situation for the characterisation of the transaction, is for the court.
Another frequent error is the automatic attribution of a sovereign character to the activity of companies owned by the State. As follows from the ILC Draft Articles, State enterprises are covered by a presumption that they are not entitled to perform governmental functions.[11] Under Article 10(3) UNCSI, where a State enterprise or other entity established by a State has independent legal personality and is capable of suing or being sued and of acquiring, owning and disposing of property, including property which the State has authorised it to operate or manage, and it is involved in a proceeding relating to a commercial transaction in which it is engaged, the immunity from jurisdiction enjoyed by the State itself remains unaffected. Such a company contracts obligations on its own account, while State immunity remains intact; no “piercing of the corporate veil” occurs. Only if it is proved that a particular State-owned company in fact exercised prerogatives of sovereignty can one speak of its acquiring the status of an “instrumentality of the State” and of the State’s immunity extending to that company as well. The burden of proof, however, rests on the company, or on the State invoking immunity.
A third error consists in confusing the scope of the notion of a “transaction” with that of a “contract”. As already indicated, a transaction also embraces dealings and pre-contractual conduct. This means that claims derived from the negotiation stage may fall within the commercial exception even if no contract was ultimately concluded.
The restrictive doctrine of State immunity is an important element of contemporary public international law, although the text of the Convention itself is in need of “refreshing”, particularly as regards laying down clear criteria for the characterisation of State conduct, consistent with the general practice of States. Is that practice, however, currently uniform enough to justify a further attempt at codification? That question remains, unfortunately, difficult to answer, since States are as a rule reluctant to agree to any limitations of their sovereignty – above all to limitations which, in the economic sphere, could entail the necessity of submitting to the jurisdiction of a foreign State. It is not inconceivable, however, that the dynamic development of business and advancing globalisation will compel countries to accept certain restrictions in the name of mutual trust in economic transactions and the harmonious development of the market.
The commercial exception (Article 10 UNCSI), although seemingly simple in conception, gives rise to interpretive difficulties in practice. The keys to its interpretation are: avoiding the automatic transposition of the purpose of an act onto the assessment of its nature; a harmonious weighing of the criteria in assessing the character of the State’s activity, so as to arrive at the result closest to the truth; and refraining from automatic determinations concerning State-owned companies based solely on their formal structure, in isolation from the acts the company actually performs. The case law, from Weltover to Certain Iranian Assets, provides the lawyer with solid tools for conducting such an analysis. The practitioner should ask, in turn: could this act have been performed by a private party? Did the State behave like a market participant? Is there a genuine contractual relationship? Only a negative answer to these questions opens the way to considering the sovereign character of the act. In the author’s view, the drafters of the Convention did a genuinely good job in codifying the commercial exception, taking an intermediate position which also allows the purpose of a transaction to be taken into account in its analysis.
Jurisdictional immunity is not an unlimited privilege of the sovereign. And although the proponents of the principle par in parem non habet imperium – as an expression of the sovereignty of States, which is the foundation of the international order – oppose the codification of exceptions to immunity, especially in the context of State-owned enterprises, it must be remembered that the world does not stand still, and States engage in far more extensive economic activity than they did a few decades ago. It is in the interest of States and private-sector actors alike to provide judicial authorities with the tools that will enable them to resolve disputes arising out of commercial transactions and to decide before which court a given State will answer. In today’s business realities, the sovereign must reckon with accountability before the courts of the States of the parties with which it deals. That is the price of the global activity of States in the twenty-first century.
Foreign Sovereign Immunities Act of 1976, 28 USC §§ 1603(d), 1605 [USA].
Regulation (EU) No 1215/2012 of the European Parliament and of the Council of 12 December 2012 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters (Brussels I bis).
State Immunity Act 1978 [United Kingdom].
State Immunity Act, RSC 1985, c S-18 [Canada].
Treaty of Amity, Economic Relations, and Consular Rights between the United States of America and Iran, Tehran, 15 August 1955.
United Nations Convention on Jurisdictional Immunities of States and Their Property, adopted by UN General Assembly resolution 59/38 of 2 December 2004 [UNCSI], arts 2, 10.
Beijing Urban Construction Group Co Ltd v Republic of Yemen, ICSID Case No ARB/14/30, Decision on Jurisdiction, 31 May 2017.
Certain Iranian Assets (Islamic Republic of Iran v United States of America) (Judgment of 30 March 2023).
I Congreso del Partido [1983] 1 AC 244 (HL) (Lord Wilberforce).
Jurisdictional Immunities of the State (Germany v Italy: Greece intervening) (Judgment) [2012] ICJ Rep 99.
Republic of Argentina v Weltover, Inc 504 US 607 (1992) [Supreme Court of the United States].
Draft Articles on Jurisdictional Immunities of States and Their Property, with Commentaries, Yearbook of the International Law Commission 1991, vol II, pt 2 (UN 1994), Commentary to art 2, paras 15, 20, 25.
UN Doc A/C.6/45/SR.25.
UN Doc A/C.6/59/SR.13.
Fox H and Webb P, The Law of State Immunity (revised and updated 3rd edn, Oxford University Press 2015).
Thomas KR, The Commercial Activity Exception to State Immunity: An Introduction (Edward Elgar 2024).
[1]Draft Articles on Jurisdictional Immunities of States and Their Property, with Commentaries, Yearbook of the International Law Commission 1991, vol II, pt 2 (UN 1994), Commentary to art 2, para 20.
[2]Draft Articles on Jurisdictional Immunities of States and Their Property, with Commentaries, Yearbook of the International Law Commission 1991, vol II, pt 2 (UN 1994), Commentary to art 2, para 25.
[3]Beijing Urban Construction Group Co Ltd v Republic of Yemen, ICSID Case No ARB/14/30, Decision on Jurisdiction, 31 May 2017, [35].
[4]I Congreso del Partido [1983] 1 AC 244, 267 (Lord Wilberforce).
[5]KR Thomas, The Commercial Activity Exception to State Immunity: An Introduction (Edward Elgar 2024) 12.
[6]Certain Iranian Assets (Islamic Republic of Iran v United States of America) (Judgment of 30 March 2023) [50].
[7]Republic of Argentina v Weltover, Inc 504 US 607, 614 (1992).
[8]Weltover (n 7) 614.
[9]Beijing Urban Construction (n 3) [40].
[10]ILC Draft Articles (n 1), Commentary to art 2, para 20.
[11]ILC Draft Articles (n 1), Commentary to art 2, para 15.