1.1 Introduction
The incremental digitalization of the economy – and, therefore, of business transactions – is challenging the traditional territorially based conceptions of jurisdiction under investment treaties.Footnote 1 Assigning a geographic origin or destination to various components of investment law – the home state of the investor; the state in which an investment is made – is familiar ground for counsel and arbitrators in international investment disputes, as is the prevalence of jurisdictional objections arguing that the investor is not covered by the treaty whose protection it is invoking or that no investment has actually been made in the host state whose actions are alleged to have breached the treaty. These hurdles of ‘localization’ are even more prevalent when the quintessential boundary-defiant characteristics of the digital economy are considered. These challenges are presumably only going to become more acute as investments quite literally transcend national jurisdictions, yet they are not altogether new. The question is whether the sometimes-tortured techniques that have been used to assign an origin to an investor or a destination to an investment can be used to reasonable effect in these new contexts.
As an example of a fairly traditional problem, one can think of the issue of territoriality raised before the arbitral tribunal in Cargill v. Mexico: Mexico argued that Cargill’s NAFTA chapter 11 claim fell beyond the jurisdiction of the tribunal given that the facilities allegedly harmed by Mexico’s measures were located in the United States, rather than in Mexico.Footnote 2 Cargill also had operations in Mexico, but it had built a high-fructose corn syrup plant in Nebraska to serve the Mexican market, and thus sought (and won) damages arising from the impact of Mexican measures on the Nebraska facility.Footnote 3 To put this slightly differently, Cargill had an integrated operation that crossed the Mexico–United States border, and thus sought damages for harm to the entirety of that investment, while Mexico argued that only the investment that was physically present in Mexico could give rise to a claim and that any damages had to be limited to that facet of the investment. In that particular case, Mexico’s argument failed, and a Canadian court declined the set-aside application in which Mexico had sought to argue an excess of jurisdiction on the part of the tribunal.Footnote 4
Thus, these cross-border challenges are not entirely new, yet the relatively novel prism of the digital economy gives us the opportunity to think about them differently. Indeed, it appears to us that the digital economy emphasizes what are essentially differences in degree rather than differences in kind. Yet it also highlights some of the conceptual frailties that have accompanied attempts to assign specific locales to investments that are not located in only one jurisdiction or that are only transiently in particular jurisdictions.
Before delving further into the topic, it might be wise to outline what we mean by ‘digital economy’, or at least attempt to offer some clarity about the scope of what we are addressing.
Defining the digital economy is rather difficult, for one of its characteristics is the fuzziness of its boundaries.Footnote 5 Nevertheless, according to one generally accepted definition it is ‘that part of economic output derived solely or primarily from digital technologies with a business model based on digital goods or services’.Footnote 6 In order to ground the discussion that follows, we have identified three types of digital economy assets which investment arbitrators might encounter and that challenge our traditional understanding of jurisdiction.Footnote 7 These are not, of course, exhaustive, but are offered for the purpose of illustration in the discussion that follows in light of likely types of investments for which investors might seek protection.
The first type of asset is cryptocurrency, the most famous example of which is Bitcoin. Cryptocurrency is ‘digital money’ used to conduct transactions; its value can be determined by a ‘mining’ algorithm, market demand, the underlying asset to which it is tied, or by national central banks.Footnote 8
The second sort of digital economy asset is the non-fungible token (‘NFT’). An NFT is a one-of-a-kind, non-interchangeable, and irreplaceable cryptoasset representing things like art or music. An NFT has been defined as ‘a unique tokenised representation of the underlying work with a demonstrable transaction history’.Footnote 9 An NFT’s fixed nature allows for reliable authentication in the digital realm, which is otherwise known for its propensity to replicate works without authorization. However, acquiring an NFT does not entitle one to a property right in the underlying asset, at least not in the non-digital sense. Indeed, holding NFTs is instead akin to bearing property rights over the digital ‘token’ to which the digitalized asset is attached but not the asset itself.Footnote 10 Moreover, while the United Kingdom has recognized NFTs as property, few other jurisdictions have addressed the status of NFTs in their legislation.Footnote 11
A look at these two examples demonstrates that cryptocurrency and NFTs can be equated to or at least linked with more traditional types of investments such as currency, property rights, and even possibly property itself. The status of the asset might depend on the jurisdiction in which it is found or in which it is deemed to be located. This again is not novel – international law does not define property, so that what constitutes property is usually ascertained by reference to the municipal law of the territory in which it is found.Footnote 12 Thus, we think that the digital character of cryptocurrency and NFTs, when compared with non-digital assets, can be understood as facing a change in degree rather than one of kind. Yet it is also the case that the localization challenge is likely to be more complex and perhaps arbitrary in the digital context, which suggests that the requirement that investments be linked to a particular jurisdiction be reconsidered.
The third type of digital economy asset we turn to in our discussion concerns data. Data are made exploitable by data-mining software, which encompasses the digital tools conducting the search, extraction, collection, and storage of personal online information relating to users’ taste, habits, and personal traits, also called ‘data’. Data are in turn commodified into digital assets – and potentially investments – profiting companies that invest to acquire and exploit those data to their benefit.Footnote 13 Although more complex to grapple with, the jurisdictional issues arising out of data-type digital assets are not altogether different from traditional challenges pertaining to jurisdiction. Intellectual property rights, for example, can cross borders, as can integrated operations such as the one described in Cargill. In the digital realm, however, it is possible for even more jurisdictions to vie to attract, tax, and regulate the assets in question.
Because of the increasing growth of investments in digital economy assets,Footnote 14 we foresee that these concerns will be at the forefront of investment arbitration disputes in the coming years. In particular, we seek with this chapter to launch a discussion about three different aspects of jurisdictional considerations that illustrate the complexities of dealing with digital economy investments in a traditional territorially defined jurisdictional framework – two of them related to jurisdiction to adjudicate and the third to jurisdiction to prescribe. This chapter proceeds in the following manner. In the first section we outline two traditional objections to adjudicatory jurisdiction that will likely be at issue in cases involving digital investments. Each of these is familiar to students of investment law. The next section attempts to go beyond traditional jurisdictional considerations to hypothesize about the extraterritorial regulation made almost inevitable by the jurisdiction-defying nature of some digital assets, and the extent to which investment treaties might fail to capture that activity to the extent its effects are felt outside the territory of the regulating state. A short conclusion synthesizes the authors’ observations.
1.2 Adjudicatory Jurisdiction
Most investment cases now contain objections to the jurisdiction of the arbitral tribunal. Sometimes decisions on jurisdiction are bifurcated; other times the hearing – and the decision – on jurisdiction and the merits come at the same time. Frequent jurisdictional objections go to the nature of the investor and the investment. Given the nature of digital assets, as described above, one can only imagine that respondent states will raise challenges to the jurisdiction of tribunals in cases involving investment in digital assets.
1.2.1 Jurisdictional Objections to the Investor: Ratione Personae
A bilateral investment treaty, or the investment chapter of a free trade agreement, protects only investments ‘covered’ by the treaty, and permits claims only by qualified investors. An investor must be a national of one of the contracting states parties to the treaty in order to submit a claim against the other state party to the treaty. The identity of the investor is also linked to the underlying investment – in other words, an investor is an entity that has made an investment or, in the case of a ‘pre-establishment’ treaty, is an entity that seeks to make an investment in the territory of the other contracting party.Footnote 15
1.2.1.1 Origin of Capital vs. Origin of the Investor
When a challenge is brought based on the investor’s ‘origin’, particularly in the corporate context, the focus tends to be on form rather than on substance: does the investor have the requisite nationality based on place of incorporation, principal place of business, or nationality of its controlling shareholders? In most cases little attention is given to substance, the origin of the capital itself. To put this differently, money is fungible; if a multinational enterprise based in the United States invests in Venezuela, it does not have to show that the dollars sent to Venezuela somehow originated in the United States – those monies might have been earned in Vietnam, or in South Africa, and then channelled through the United States to Venezuela. Many treaties specifically permit investors to hold their investments directly or indirectly; these broad standing provisions are meant to ensure that investors can structure their investments as they see fit. Yet they do tend to permit claims by investors related to investments regardless of the origin of the capital involved.
Although perhaps not exactly surprising, for it might make sense in terms of ease of administration and in light of the fungibility of currency, it is nevertheless curious that there is often little inquiry into the provenance of the money at stake. For example, assume that a Swedish businessperson who was born a Romanian citizen invests in a Romanian company. The Swedish business person’s ability to seek the protection of the Romania–Sweden BIT might be challenged on the basis that he is not really Swedish (or that he retains dual nationality) and thus is not entitled to seek the protection of the treaty, but not specifically on the basis that the money used to invest in Romania was not really Swedish money but was Romanian money.Footnote 16 If that money had all been earned in Sweden, the nationality of the person who earned it and who decided to invest it in Romania might seem to be unimportant, particularly if one of the goals of international investment treaties has traditionally been thought to increase investment flows from one treaty party to the other. In practice, it might well be the case that the country importing the capital is less concerned about the origin of the capital (we are assuming no nefarious activity) than about the fact that it arrives. Yet nationality requirements offer the opportunity to contest the jurisdiction of the tribunal, even if the underlying goal of the treaty has been satisfied.
Digital investments might be even more intractable with respect to this question. If cryptocurrency were to be used as the means for making an investment, where does the cryptocurrency originate? If a Swedish investor uses Bitcoin to fund his investment in Romania, and the Bitcoin is held on computers around the world, but no one knows exactly where, is that a Swedish investment into Romania? On the one hand that money could have been invested somewhere else, so presumably Romania is pleased to receive the investment and could be viewed as receiving its part of the grand bargain – the receipt of capital from a Swedish investor in return for having signed the investment treaty.Footnote 17 But is it really ‘Swedish’ money? Does that matter? Should it matter? If a US data mining company invests in Romania, and the US-based company making that investment earned all its money from data collected in other Eastern European countries, is that US investor entitled to the protection of the Romania–United States BIT? Again one might assume that Romania might be indifferent to the ultimate origin of the capital, but it nonetheless might raise objections in an individual case, particularly if it seemed the US company was engaged in treaty shopping and/or nationality planning.
How, then, do digital investments fit in that framework? The acquisition of cryptocurrency is a good example. What does it mean to be an investor in cryptoassets? If the cryptocurrency company is based in one jurisdiction, say Delaware or the United Kingdom, but all of its assets (for example, the software creating the digital currency) are found in computers operating in some dozens of jurisdictions, is it good enough to look at the place of incorporation? Indeed, does the place of incorporation of a cryptocurrency company carry the same significance as it ordinarily does with companies that are primarily non-digital, even if all or most of the activity happens elsewhere? Where is Bitcoin generated? What about individual holders of Bitcoin? Should we look, instead, at the origin of wire transfers emerging from cryptocurrency investments – where the assets are deemed to originate and thus ‘exist’ for at least some period of time?
1.2.1.2 Veil Piercing
Difficulties stemming from the trans-territorial and/or non-territorial nature of digital investments could push us to move away from the form and to delve into the substance. Doing so, however, would require piercing the corporate veil, something that is generally disfavoured in investment tribunal practice. It was nevertheless done in Loewen v. United States, albeit in the disaggregated fashion that treats the investor as distinct from the capital it is supplying. Loewen provides an excellent illustration of the complexities of localization.Footnote 18
The claimants were The Loewen Group Inc. (‘TLGI’), a Canadian company conducting business on both sides of the Canada–US border, and Raymond Loewen, ‘founder of TLGI and its principal shareholder and chief executive officer’.Footnote 19 TLGI’s principal US-based subsidiary was Loewen Group International, Inc. Most of the Loewen conglomerate’s business was in the United States – to put it differently, the revenues were mostly earned in the United States and reinvested in the United States – and thus did not come from Canada. Yet, the ‘Canadian-ness’ of TLGI was not contested for the mere reason that it was incorporated in Canada, nor was the fact that most of the investment in TLGI was in fact reinvestment,Footnote 20 with the capital flows remaining within the United States.
Once TLGI filed a petition under chapter 11 of the United States Bankruptcy Code and ceased to exist, and the Loewen group was reorganized as a US company with a Canadian subsidiary (‘Nafcanco’) created to hold the arbitral claim, the United States renewed its jurisdictional objections.Footnote 21 The United States claimed that the continuous nationality rule required that ‘there must be continuous national identity from the date of the events giving rise to the claim … through the date of the resolution of the claim’.Footnote 22 The tribunal accepted the United States’ argument and dismissed the case.Footnote 23 It found that the Canadian subsidiary was a shell company with no asset aside from the NAFTA claim, and that the beneficial owner was the US company.Footnote 24 To determine that the Canadian company was an empty vehicle, it had to pierce the veil. However, the tribunal did not offer any rationale as to why it pierced the corporate veil, though one could infer it was concerned about an abuse of process of the sort usually captured by denial of benefits clauses.Footnote 25
The Tokios Tokelés case also aptly exemplifies the reluctance of tribunals to pierce the corporate veil, and also the focus on the nationality of entities and persons rather than of capital. The Claimant, Tokios Tokelés, was a company incorporated under the laws of Lithuania. Tokios Tokelés created and wholly owned Taki spravy, its Ukraine-incorporated subsidiary. The Claimant alleged that its subsidiary’s activities had been damaged by various actions undertaken by agents of the Respondent, Ukraine.Footnote 26 Ukraine alleged that the Claimant could not launch proceedings against it under the ICSID Convention because it was actually not a Lithuanian corporation entitled to benefit from the Ukraine–Lithuania BIT protections but was instead a Ukrainian company.Footnote 27 To back its claim, Ukraine argued that the Claimant had no substantial business activities in Lithuania and retained its headquarters in Ukraine.Footnote 28 The Respondent thus demanded that the tribunal pierce the corporate veil and determine the Claimant’s nationality ‘according to the nationality of its predominant shareholders and managers’.Footnote 29
While the tribunal, in the award, noted that the Claimant’s company had conducted business with what appeared to be ‘sham enterprises’,Footnote 30 it declined to pierce the corporate veil for purposes of the decision on jurisdiction and noted that ‘the only relevant consideration is whether the Claimant is established under the laws of Lithuania. We find that it is. Thus, the Claimant is an investor of Lithuania under Article 1(2)(b) of the BIT’.Footnote 31 Furthermore, the tribunal reasoned that piercing the corporate veil would run against the core aims of the Treaty, namely, to fulfil the parties’ expectations, ensure predictability in dispute settlement processes, and enable ‘investors to structure their investments to enjoy the legal protections afforded under the Treaty’.Footnote 32
How does piercing the corporate veil intersect with the way that cryptocurrency is owned? Cryptocurrency is necessarily held in pseudonymous crypto wallets, which provide an additional layer of cryptographic protection for the asset. The pseudonym is a series of numbers tied to the entity or person holding the key, but is not necessarily traceable. This would seem to add an additional layer of complication if one is considering piercing the veil over ownership of assets, and might indeed render traditional approaches inapplicable.
Traditionally, we see that inquiries into nationality and origin tend to stop when there is a sufficient anchor – say the existence of some kind of link between a company and a given jurisdiction – without necessarily scrutinizing the actual flow of capital involved in the business. Furthermore, in those rare circumstances when a tribunal agrees to pierce the veil, it does not provide full-bodied explanations as to why it did so. Cryptocurrency holdings might not lend themselves to veil-piercing, even were tribunals minded to try.
1.2.1.3 Denial of Benefits Clauses
This line of thought brings to mind so-called denial of benefits clauses, a feature of some but not all investment treaties. These clauses prevent ‘sham’ companies from benefitting from investment treaty protections where such businesses are incorporated in a jurisdiction in which they do not have substantial business activities and are owned or controlled by non-nationals of treaty parties.Footnote 33 The corporation might or might not have been set up to take advantage of an investment treaty; the motivation is not one of the requirements for the application of the provision. In cases where there is no denial of benefits clause, or when the denial of benefits clause does not capture the factual situation in question, the question of abuse of process would govern any assessment and the investor’s motivation would be at issue.Footnote 34
What constitutes a ‘sham’ corporation is not always clear. Digital companies – ones that operate exclusively or primarily online – might seem like ‘sham’ companies from a traditional viewpoint but that would not necessarily be an accurate evaluation. Given the lack of clarity about how much activity within one jurisdiction could serve to negate a ‘sham’ characterization in the case of a rather typical bricks-and-mortar business, the question of whether a digital corporation is a ‘sham’ enterprise vis-à-vis a specific jurisdiction might well be difficult to answer. Could they be understood to be companies that chose to incorporate in a particular jurisdiction specifically to benefit from particular protections and advantages even though their operations are primarily, and even by definition, a-jurisdictional or trans-jurisdictional? If so, they might fall within ‘sham’ territory. Yet if too-strict criteria are applied, there is a danger that digital companies could effectively forfeit these protections because their business model diverges from a more traditional bricks-and-mortar business and strays into territory that many find hard to understand.
If so, do these strategic business choices disentitle digital economy companies from treaty protection on the basis of denial-of-benefits clauses? This should not be the case. It would be odd to interpret the clauses in that way insofar as it would essentially leave digital companies with no jurisdictional anchoring. It seems most likely that one of the traditional jurisdictional anchors would be selected, in company with an assumption that ‘substantial’ activities occur in that jurisdiction given that by traditional conceptions they are not more reasonably viewed as occurring anywhere else. At this stage, this is mere speculation; particular facts could lead to different results.
How should tribunals, then, consider the particularities of digital economy investors to account for the unprecedented novelty – and difficult localization – of their assets? While the current framework likely affords digital economy companies and investors in digital economy assets the protections of broadly worded treaties given the hard-to-pierce corporate veil, how should tribunals address concerns about treaty-shopping and sham companies? Furthermore, is it wise to treat the economic innovations of the digital economy just like non-digital investments? Would one answer lie in directly addressing such specificities in new treaties and investment chapters? One could try to amend investment treaties to take into account digital investors, and their investments, as discussed further in the next section. Yet changes to treaties for digital investments raise questions that could be applicable to more conventional investments as well regarding the origin of the capital behind the investments, as illustrated above.
We are now left with two questions that encapsulate the tension inherent in objections ratione personae as applied to digital investments: on the one hand, if we find that looking beyond the place of incorporation and into the origin of the investor is important in the digital realm, why would that not be the case in non-digital investments as well? On the other hand, if we investigate the origin of the investor for digital economy assets, do we not risk jeopardizing protections afforded to a booming sector of the economy by subjecting them to scrutiny under denial of benefits clauses, or under abuse of process standards if there is no denial of benefits clause, in a way that is arguably unfair given the nature of their business?
1.2.2 Jurisdictional Objections about the Investment: Ratione Materiae
Jurisdictional objections related to investments are manifold. What constitutes an investment is all the more complex with digital economy assets given their very nature, which tends to be intangible and delocalized. Are digital investments even investments in the sense prescribed by international investment law?
1.2.2.1 Definitional Categories
The traditional approach to defining what constitutes an investment is to list – exhaustively or not – what qualifies in the BIT. An illustrative list is found in the UK Model BIT from 2008:
Investment means every kind of asset, owned or controlled directly or indirectly, and in particular, though not exclusively, includes:
(i) movable and immovable property and any other property rights such as mortgages, liens, or pledges;
(ii) shares in and stocks and debentures of a company and any other form of participation in a company;
(iii) claims to money or to any performance under contract having a financial value;
(iv) intellectual property rights, goodwill, technical processes, and know-how;
(v) business concessions conferred by law or under contract, including concessions to search for, cultivate, extract, or exploit natural resources.Footnote 35
These categories, no matter how broad and encompassing they might seem to be at first glance, do not make it easy to ‘frame’ digital economy assets as investments under such a BIT, or especially to meet the requirement that the investment be in the territory of the treaty party whose measures are being challenged.
Cryptocurrencies are digital monies; they can therefore be understood as ‘claims to money’, but only to the extent that the specific cryptocurrency is recognized as ‘money’ by the parties. Indeed, cryptocurrencies can vary, with some resembling traditional money in a digital format and others being digital money whose use is restricted – say to buying only a given sort of product or the products of a specific company, in which case cryptocurrency is not unlike Monopoly money. Yet ‘every kind of asset’ is extraordinarily capacious – and cryptocurrency is unquestionably an asset. Thus, whether any given cryptocurrency can be recognized as an investment under a particular treaty is unclear, yet the breadth of the definitions makes it likely.
NFTs do not automatically confer an exclusive right of exploitation on the owner – although this can be done if provided for in the contract; they do not entitle the owner to copyright over the art piece, but they confer something akin to a lien. Recognizing NFTs as investments under a BIT list is probable, assuming said list recognizes such property rights as liens or ‘claims to money or to any performance under a contract having a financial value’, assuming NFTs could be analogized to claims to performance under a contract with financial value.
Again, data pose the trickiest questions. Can commercially exploitable personal data be covered by intellectual or industrial rights if they are mined by software the patent of which is owned by the company that seeks rights over the data? If data have economic value, could they not be considered as ‘resources’, although not natural ones? Could data mining and exploitation be regarded as a technical process?
If one has an open list, such as the one excerpted above, digital assets might be added as another kind of asset. If, on the other hand, the BIT gives only a closed list, digital assets would have to be shoehorned into one of the existing categories.Footnote 36 This seems possible but not inevitable.
1.2.2.2 Territoriality
Other difficulties surround questions about territory: would the investment be regarded as being ‘in the territory’ of the host state and thus be entitled to the protection of the treaty?Footnote 37 Regarding cryptocurrency, the difficulty lies primarily in assessing where the cryptocurrency is held, and its nominal location in a financial institution or even a foreign exchange market. Whereas the currency of a country can be traced to an emitting governmental body, a central bank, and therefore to a state, localizing cryptocurrency remains a challenge. While some states are developing their own cryptocurrency (e.g., the Marshall IslandsFootnote 38), other cryptocurrencies (such as Bitcoin) are neither the fruit of a company nor a state, but instead stem from individuals creating data-mining software and ‘letting the currency exist’ on an online, decentralized blockchain that is not overseen by any specific entity.Footnote 39 In investing in that type of cryptocurrency, where is one actually investing? Similar questions arise with respect to NFTs. If one were to buy an NFT on an exchange in London, but the NFT was connected to a tangible asset in France, would the NFT constitute an investment in the United Kingdom, in France, or in neither jurisdiction?
These questions are of paramount importance because an investor will need to localize the investment (and the damage) in order to hold a state responsible for banning the use, holding, transfer, or sale of cryptocurrency – the investment itself – if the investor wishes to sue under a BIT. Should tribunals look, instead, at the origin of the funds that initially effected the acquisition of the cryptocurrency, or at the origin of any wire transfers emerging from cryptocurrency investments – where the assets are deemed to originate and thus ’exist’ for at least some period of time?
1.2.2.3 Investment under the ICSID Convention: The ‘Salini’ Criteria
Even satisfying the definition found in a traditional BIT might be insufficient if one has a dispute under the ICSID Convention, which imposes a separate test.Footnote 40 Building on the ICSID dispute Fedax v. Venezuela,Footnote 41 the tribunal presiding over the Salini v. MoroccoFootnote 42 dispute used four criteria to determine whether an investment qualifies as such under the ICSID Convention, which does not define itself ‘investment’.Footnote 43 Thus, the tribunal in Salini listed (1) duration; (2) assumption of risk; (3) substantial commitment; and (4) contribution to the host state’s development as required characteristics to qualify as an ‘investment’ under the ICSID Convention.Footnote 44 This test was later taken up by many other tribunals, and has been included in some newer investment treaties, though some tribunals have been less willing to treat the criteria as definitive.Footnote 45 Still, it is included in some newer treaties and is discussed if not always definitive; it is thus useful for our purposes of assessing the types of jurisdictional objections states might raise in disputes involving digital assets.
How does one frame digital economy assets under the Salini test? The first criterion, duration, poses difficulties from the outset. The duration of the investment in cryptocurrency and NFTs might very well vary according to the given investor and asset. Is investing in such assets a one-time action that occurs at the instant of purchase? Does it, inversely, only truly become an investment when sold and the benefits reaped, materializing in ‘traditional’ money form? Is it sufficient that the asset is capable of being held long-term, or must it be almost inevitable that it be of a significant duration? Similar questions arise in the realm of claims by minority shareholders, whose assets (‘shares’) are almost inevitably included in the list of investments covered by a treaty but are easily alienable and thus not necessarily long-term in the sense of having a likely duration.Footnote 46 Yet if they have been held for a certain period they have usually been deemed to satisfy the ‘duration’ requirement.Footnote 47
Full consideration of the assumption-risk criterion would demand a special assessment for each of the discussed types of assets (cryptocurrency, NFTs, and data) and for any specific investment falling within each of those categories. However, broadly put, it seems we might see two broad yet opposing conclusions. Either we assume that digital economy assets are high-risk due to their intangible, delocalized nature – especially in the era of cyberattacks and other hacks – or, conversely, we assume that digitalization implies lower risks given the automation and technological protections that erase human and other exogenous errors.
The substantial commitment of capital criterion seems straightforward but would exclude those investors investing in digital assets on a micro scale – the billionaire investing hundreds of thousands in Bitcoin might make an ‘investment’ whereas you and I would not qualify, with our few tens of dollars or euros placed as experiments. Again, of course, this is not a new question, but arose in cases involving Italian bondholders in Argentina, for example.Footnote 48 What about the proportion of the investment held by an individual? Indeed, at the time of writing, a single Bitcoin is worth around US$ 107,000.Footnote 49 Would an investment in a dozen Bitcoin units, for a total sum above one million dollars, be a substantial investment, or should the investor buy a substantial number of the units available (likely more than a simple dozen) for the investment to be sizeable? The volatily of cryptocurrency is also likely to further complicate the assessment of damages. Again this is not a new question. Thus, how one should approach investments in cryptocurrency in relation to the Salini criteria remains unclear.
Finally, the criterion of a contribution to the economic development of the host state brings the territorial complexities back. What anchors a digital investment to a particular state and how is it contributing to that state’s economic development if the activities and benefits are conducted online, in multiple jurisdictions at the same time, through dozens of computers? Strikingly, some recent findings suggest that crypto-mining is in fact an energy drain in local states, with an electricity consumption that far exceeds that of some countries.Footnote 50 Perhaps in placing demands on the energy industry, crypto-mining bolsters that sector of the economy. On the other hand, if the crypto-mining consumes more than it gives to the host state, or if one cannot really say where the economic development is being delivered, is it truly an investment? The link is tenuous, and it would be a stretch to frame investment in the crypto-mining industry as an investment in the energy sector. Beyond the energy concern, the delocalized nature of digital investments does not permit a firm answer on the contribution of such assets to the host state’s economy.
NFTs raise similar difficulties with respect to localization of any contribution to economic development. Perhaps the company selling the NFTs is the beneficiary of the BIT, and the jurisdiction in which it is located is the most likely to reap the benefits for activity in the sector, yet the underlying asset to which the NFT is tied might be in a different jurisdiction. Does that matter? Should it matter?
Thinking about data collection takes us into even more nebulous territory. The self-building nature – meaning that the algorithms are somewhat autonomous in the way they function – of data-storing and data-mining algorithms bolsters the complexities that arise when thinking about jurisdiction in traditional ways. Indeed, the automation of the systems might push a given algorithm to relocate a set of data to another, or many other, computers in different places, which would effectively displace the assets without the investor’s volition and therefore complicate the localization efforts. ‘Cloud’ storage, which despite the name means that our data are stored on hard drives in physical data centres, nonetheless permits our data to be shared and spread among multiple machines at once; our Facebook photos and other Big Tech-sourced data do not have a ‘permanent’ place on a particular chip, but rather they move across computers that can be in different jurisdictions.Footnote 51 If data, which can be owned by businesses that profit from their exploitation, can be considered to be an investment – something yet to be confirmed – finding the origin of the investment by looking at the origin of the investor may not be accurate. Does this make digital economy investments fundamentally different in kind from traditional ones?
Determining the contribution to the economic development of the host state was also at issue in the Philip Morris v. Uruguay case. The state of Uruguay contended that Philip Morris’s manufacturing of cigarettes was an investment that did not contribute positively to its economic development but actually burdened it with additional costs, given the adverse health consequences of tobacco consumption.Footnote 52 Yet, the Philip Morris v. Uruguay tribunal dismissed Uruguay’s argument on the grounds that determining the contribution of an investment to the economic development of a host state would require an ex post facto analysis with the potential of generating ‘a wide spectrum of reasonable opinions’.Footnote 53 Nevertheless, the tribunal recognized that the ICSID Convention sets out an outer limit on what can qualify as an investment capable of benefitting from the protection of the treaty.Footnote 54 The tribunal also relied on the Pey Casado dispute, and more precisely on the tribunal’s qualification of the criterion of ‘contribution to economic development’ as a consequence, rather than as a condition, of investment.Footnote 55
These criteria might all be at issue in cases involving digital assets, as shown by the examples in the discussion, but questions related to territorial location are especially likely to arise. The UK-Mexico 2006 BIT seems to emphasize the territorial nexus between the parties and the investment by repeatedly referring to ‘the territory’ of one or both states.Footnote 56 The overall definition of ‘investment’ under Article 1 of the Agreement reads: ‘“investment” means an asset acquired in accordance with the laws and regulations of the Contracting Party in whose territory the investment is made…’Footnote 57 Thus, as Eric de Brabandère has noted, if we are to consider digital investments as different in kind entirely – with the criterion that they are a-territorial to distinguish them from traditional investments – these would likely not be covered by the BIT at all because they lack the territoriality required by the very language of the agreement.Footnote 58
Faced with the ‘territorial nexus’ requirement, tribunals have accepted that the end-use of the benefits stemming from the investment can be determinative of the place in which the investment is actually made.Footnote 59 In Fedax v. Venezuela, the tribunal considered that if the funds corresponding to the investment are enjoyed by or at the disposal of the host state, the investment can be deemed to be ‘made’ in that state.Footnote 60 In Abaclat v. Argentina, the majority of the tribunal took the same approach as the tribunal in Fedax v. Venezuela with the exception of arbitrator Abi-Saab, who issued a dissenting opinion reaffirming the intrinsically territorial nature of an investment.Footnote 61
If the desire to engage in localization and to focus on territoriality poses problems in the so-called traditional context of international investment arbitration, what should we expect if we are to consider digital assets as distinct because of their trans-territorial nature? Is it possible to recognize their online, digital, almost ethereal characteristics without detaching them completely from the considerations brought to bear for common material and financial investments?
The tendency to ‘divorce’ investor from the investment might need to be revisited in order to ascertain whether there is really an investment entitled to coverage. As we have seen in the discussion above, if there is a covered investor, and a covered investment, those facts are often sufficient to ground jurisdiction. To be sure, in certain cases this is not true; if there is significant doubt about the amount of the investment, or about an investment having been made at all, then more scrutiny might be given to tracing the capital. This is generally done in the guise of an objection to the existence of an investment, however, rather than to the propriety of a particular investor seeking the protection of the treaty. These questions are even more likely to arise in the context of a decision on the merits of the case and appropriate compensation, if any.
The likely quest will be for an economic anchor in a particular territory. For example, one might ask where (if anywhere) are the profits taxed?Footnote 62 Is it the home of the investor? Can the investor be disaggregated from the investment? It seems we could be back to the question of funds, but looking towards the destination, rather than the source.
One approach that might help overcome the difficulties in locating digital investments within a particular territory is to focus on the jurisdiction (or jurisdictions) that are seeking to regulate or otherwise control those investments.
1.3 Jurisdiction to Prescribe
The third category of difficulty with respect to digital investments concerns jurisdiction to prescribe – or what might be viewed as the other side of the coin. When can a state be held accountable for taking a measure that affects investments or investors outside its territory, or whose effects are inextricably linked to harm both within and without the territory? Investment treaties proscribe a state’s conduct that causes harm ‘within its territory’, with territory generally delineated to include land and sometimes specific areas over which the state exercises sovereignty.Footnote 63 What if, given the nature of a particular type of investment, a state’s conduct almost inevitably has extraterritorial as well as territorial effects? To put it a different way – what is the effect of state measures that affect investments that might – or might not – be physically present in the territory of that state? Is a state engaging in extraterritorial regulation, and if it does so is that activity by definition outside the scope of an investment treaty? But does that mean a state’s acts can escape scrutiny, because if a digital asset cannot be localized within a state’s territory then even the most targeted and arbitrary act might escape scrutiny because of its extraterritorial effects. If the extraterritorial application of law is sometimes thought to be unlawful, but investment treaties do not protect against a state’s exercise of its laws extraterritorially, does this mean that unlawfulness begets unlawfulness?
Again, cryptocurrency mining provides a useful potential model. Digital assets have been the focus of discussions among legislators, some of whom have raised the possibility of banning digital currencies.Footnote 64 If, hypothetically, the United States banned the use of cryptocurrency within the United States, would that ban be limited to mining occurring in the United States, or to transactions using cryptocurrency within the United States? Even if it were limited in language, could it be limited in practice if cybercurrency mining were occurring on computers linked across national boundaries? Indeed, what if the parent company were located in Mexico, but owned crypto-mining farms in the United States and elsewhere, with those farms linked in such a way as to defy tracing those links or the jurisdiction(s) in which they are found?
Another point of importance concerns the reasons for an eventual ban. Should reasons matter? Take climate change: crypto-mining is highly energy-consuming, so if the ban is a measure taken to protect the environment, are we more inclined to accept its consequences? What if the ban is implemented with an express trans-border mindset (global emissions, etc.) – would that have different effects than a ban made for purely domestic reasons? Might it fall within the state’s police powers, a question presumably for the merits rather than jurisdiction?
Addressing jurisdiction to prescribe makes it possible to dig further into our reflections on the challenges posed by territorial delimitations. Indeed, if a state’s acts violate the standards currently found in an investment treaty, and the investor has some legitimate expectation to have been free from that state’s coercive authority, then perhaps that host state should be responsible for the consequences of its acts.
1.4 Conclusion
In sum, it is not altogether clear to us that there are differences in kind, as opposed to differences in degree, between traditional jurisdiction-focused investments and digital economy investments. We can see apparently arbitrary limitations due to the bilateral nature of investment treaties and at times rather strained attempts at localization. There is often a focus on form with assumptions about substance that tends to avoid more difficult questions. On the other hand, it is also possible that we are trying to fit new realities into old paradigms.
Perhaps the challenges posed by applying traditional localization techniques to cases involving digital assets could be a catalyst for change. They could give us another opportunity to think of ways to transcend the jurisdictional limitations imposed by the current regime. One possibility would be to amend the investment regime along the lines of human rights treaties, where a state is responsible for all acts that harm investors, whether they be foreign or domestic.Footnote 65 In doing so, one could avoid at least some jurisdictional challenges ratione personae, and some of the concerns that investment law unfairly favours foreigners over locals.
This approach does not, however, respond to ratione materiae problems raised by the difficulties inherent in localizing digital assets. Perhaps it is time to revisit the question of a multilateral agreement on investment that would focus more on the reach of government regulation and less on the specific territory(ies) in which the investment is found. Such a device might avoid several of the jurisdictional problems that bedevil arbitration by ensuring that a state should have to afford fair and equitable treatment not only to all those within its jurisdiction but also to those subject to its authority. With more global coverage, the focus would be more on the merits rather than on jurisdictional questions.
Of course, in any given case one might still have some questions about standing, and whether a particular investor was injured sufficiently to submit a claim, but coverage questions would diminish. This would represent a shift from the ‘grand bargain’ of promising specific investor protection in return for increased capital flows towards a more overtly regulatory approach. A multilateral agreement on investment focused on a governance approach would avoid tribunals and counsel – and investors and states – getting lost in the morass of highly technical and perhaps inaccurate, or at the least evanescent, assessments necessary to localize an investor, or an investment, within its territory. The alternative is likely to be continuation of the strained but familiar attempts to localize trans-territorial – and perhaps a-territorial – investments.
