5.6 Jurisdictional Fragmentation
Why shared principles have not produced consistent national rules, how one token can be a different instrument at each border, why no one agrees which country's law decides who owns it, how insolvency and multi-issuance turn divergence into loss, why compliance moves at the pace of the weakest jurisdiction, and how Chapter 5's six risks map back to Section 2.4.
The ledger is global. The law behind every token on it is not. Section 2.1 recorded that BIS, OECD, FSB, IOSCO and ESMA now agree that tokenization is "a change in representation, not in legal nature." That agreement is why fragmentation matters. If a ledger does not change what an asset is in law, every national rule that applied before still applies: how the asset is classified, who owns it, what happens to it in insolvency, and who must screen its holders. Those rules differ by country, and a token moving between wallets in two countries carries both sets with it. The chain removes the border from settlement. It does not remove it from ownership, enforcement or supervision.
Section 5.5 argued that smart contracts move discretion into keys rather than removing it, and closed where code and documents disagree and someone has to decide, often by a freeze or court order issued in one place against a token held in many. This section asks: under whose law? Section 5.2 examined who owns what inside one legal system, including UNIDROIT's Principle 4 and the FTX customer-property question. Section 4.3 observed that "A wrapper's domicile and form tell a holder which court and insolvency regime would apply if it failed." That holds while issuer, custodian, asset and holder share a country, which the products in Section 4.3 rarely do. Section 1.3 counted it as progress that a firm "can choose among multiple rulebooks"; read from the risk register, that choice is fragmentation. Unlike the fragmentation across chains and cash tokens in Section 3.3 and Section 3.4, this kind is legal.
The Rulebooks Converged on Paper and Diverged in Practice
On paper, the standard-setters agree. In practice, their members do not. The FSB's October 2024 tokenization report already listed "differences in legal and regulatory frameworks, given the cross-border nature of many projects" among the obstacles to scaling1. The FSB's peer review of 37 jurisdictions, published on 16 October 2025, found "significant gaps and inconsistencies that could pose risks to financial stability," warned that "Uneven implementation creates opportunities for regulatory arbitrage," and called cross-border cooperation "fragmented, inconsistent, and insufficient to address the global nature of crypto-asset markets"2. In its summary table, 11 jurisdictions had finalized a framework for crypto-asset activities and 6 had none; for global stablecoin arrangements, 5 had finalized one and 11 had none2. It also named "divergent definitions of crypto-assets" among the threats to information sharing between authorities2. IOSCO's parallel review of 20 jurisdictions concluded that "regulatory divergences not only heighten the risk of regulatory arbitrage but also pose challenges to global financial stability"3.
Both reviews assess crypto-asset and stablecoin regimes, not tokenized funds or securities as such. They bear on RWA tokens because those tokens pass through the regimes assessed: crypto-asset service providers distribute them, stablecoins settle them (Section 4.4), and authorities test them against the definitions the FSB found divergent. Fragmentation therefore works less as a sixth risk than as a multiplier on the other five, as the chapter's closing shows.
The Same Token Is a Different Instrument at Each Border
Classification changes at the border, even inside a single rulebook. ESMA's guidelines on when a crypto-asset is a financial instrument aim at "the common, uniform and consistent application" of MiCA's boundary provision, yet concede that "the term 'share' is not defined by the EU law. As a result, Member States interpret this concept differently"4. ESMA's compliance table of 17 July 2026 shows the result5. The Czech National Bank marked the guidelines "Not applicable" for tokenized securities under domestic law, because applying them is "prevented by legislative barriers, specifically the formal concept of securities as defined in Sections 514 et seq. of the Civil Code." The National Bank of Slovakia did the same, since under current law market participants "cannot classify any crypto assets issued under Slovak securities legislation as transferable securities"5. The ESCB's September 2026 summary: tokenized financial instruments, "even if regulated under MiFID, they are still subject to national laws"6.
Across the Atlantic the gap widens. Robinhood Europe, authorized by the Bank of Lithuania, offers "2,000+ Classic Stock Tokens linked to US stocks and ETPs" and states that they "are offered under MiFID II as derivatives"7. SEC staff describe a model in which a third party issues a token providing "synthetic exposure to the underlying security, such as a tokenized linked security or a tokenized security-based swap," and state that the swap version may not be sold to anyone who is not an eligible contract participant "unless a Securities Act registration statement is in effect as to the crypto asset, and the transactions in the crypto asset are effected on a national securities exchange"8. How the SEC would classify any EU product is a question for Chapter 6 and the companion paper on regulation. The point here is narrower: a token referencing a US share is sold under one legal category in the EU, while the US framework for synthetic exposure attaches conditions of its own. xStocks show the pattern from the issuer's side, with the Jersey issuer, Liechtenstein-approved prospectus, Swiss-law certificates and four excluded countries that Section 4.3 documented.
The Law Commission of England and Wales puts it plainly: "There is no accepted taxonomy in the market, regulatory contexts, or legal contexts for labelling these different types of crypto-tokens"9. A buyer does not buy one legal instrument, but whichever instrument each jurisdiction the token touches decides it is.
A Global Ledger, a National Court
The Law Commission frames the subject as three questions: "In which country's courts should the parties litigate their dispute? Which country's law should be applied to resolve it? How can the judgment be enforced in another country?"10 For most assets, location answers the second question: where the land lies, where the register is kept. Tokens break that premise. In the Commission's phrase, they exist "nowhere and everywhere at the same time"9. IOSCO warns that token ownership and transfers may be "invalidated or not recognized as envisaged," a risk "further complicated in cross-border transactions given the wide variation in jurisdictional approaches"11.
The most developed answer is UNIDROIT's Principle 5, adopted in May 2023. Its waterfall applies first the law "expressly specified in the digital asset," then the law specified in the system recording it, then, where there is an issuer, the law of its statutory seat, "provided that its statutory seat is readily ascertainable by the public," with a forum-based fallback after that12. The commentary rejects location-based rules as "incoherent and futile" for "intangibles that have no physical situs," while conceding that "a conflict-of-laws rule will always be imperfect"12.
Three limits stop this from settling the question for RWAs. The Principles bind no one until a State adopts them; they take effect as law only "once they have been adopted and implemented into the law of a State"12. They choose a law, not a court: Principle 5 "does not address the question of the jurisdiction of any tribunal"12. And Principle 4, as Section 5.2 showed, leaves the link between token and asset to "other law"12. A court could apply the law named in the token to decide who owns the token, and another country's law to decide whether owning it means owning the fund share or gold bar behind it.
Elsewhere the work is unfinished. The Hague Securities Convention of 2006, the only treaty on the law governing securities held with an intermediary, has three Contracting Parties: Mauritius, Switzerland and the United States13. The Hague Conference's Experts' Group on Digital Tokens works on the basis that its mandate excludes securities14, and in March 2026 the Conference asked for 2027 reports from that group and, on jurisdiction and applicable law in securities markets using distributed ledgers, from its Permanent Bureau15. The Law Commission's final report was still pending on 28 September 2026, more than a year after its consultation closed10. The European Commission's MiCA review consultation of 20 May 2026 notes that ownership "remains a matter of (national) property laws," asks "Should EU law introduce a conflict of law rule for tokens?", and lists nine candidate connecting factors, from the law specified in the token to "the primary residence of the encryption private master keyholder"16. The ESCB counts five Member States that have already legislated and warns that "differences in national approaches are emerging, notably as regards corporate law, property law and insolvency law"6.
The honest qualifier favors most RWA products. The Law Commission's consultation paper states that "Centralised and/or permissioned applications of DLT are less problematic for private international law," because a system operator or "gatekeeper" can serve as the connecting factor9. A tokenized fund with a named issuer, transfer agent and whitelist has one, and UNIDROIT's issuer-seat rule fits it. The qualifier weakens once the token leaves the permissioned venue: xStocks can be withdrawn to self-custody (Section 4.3), and a token moving through wallets and DeFi protocols carries fewer anchors for a court.
When the Issuer Fails, Two Courts Can Claim the Same Assets
Insolvency and resolution are built on national foundations. Tobias Adrian of the IMF writes that tokenization "challenges crisis management and resolution frameworks that are built around nationally domiciled institutions, territorially bounded infrastructures, and jurisdiction-specific legal authority," and that the failure of a systemically important tokenized institution could bring "jurisdictional conflict, delayed intervention, or paralysis"17. Even UNIDROIT's choice-of-law rule gives way to insolvency rules "such as ... the ranking of categories of claims"12.
No RWA issuer has yet failed across borders at scale. The largest documented test of a digital-asset intermediary failing in two legal systems at once was FTX, a crypto exchange rather than an RWA issuer, and it shows what happens when two courts reach for the same wallet. Section 5.2 used FTX for customer property; here the conflict is between two countries. On 12 November 2022, acting "under the authority of an Order made by the Supreme Court of The Bahamas," the Securities Commission of The Bahamas directed "the transfer of all digital assets of FTX Digital Markets Ltd." to a wallet it controlled, and stated that "It is not the understanding of the Commission that FDM is a party to the US Chapter 11 Bankruptcy proceedings"18. The US debtors said cryptocurrency belonging to them had been moved "after the commencement of the chapter 11 cases" and "without their authorization"19. The two sides did not even agree on what had moved: about $296 million at spot prices at the time of transfer, by the debtors' count, assuming all of the FTT could be sold at spot, against "more than $3.5 billion" by the Commission's19. No rule of recognition ended the conflict. A bespoke agreement did. On 19 December 2023, about 13 months after the transfer, the parties announced a global settlement "subject to the approval of" both the US Bankruptcy Court for the District of Delaware and the Supreme Court of the Bahamas, letting most FTX.com customers choose which proceeding paid their claims20. The debtors called it "a novel and mutually-beneficial solution"20. The next failure will need its own.
The standing machinery is partial. The UNCITRAL Model Law on Cross-Border Insolvency has been used by "63 States in a total of 66 jurisdictions," to varying extents21. The United States and the British Virgin Islands, domicile of BUIDL and JTRSY in Section 4.3, appear on the list; the Bahamas, Jersey, Luxembourg and Switzerland do not21. The 2019 Hague Judgments Convention has 33 Contracting Parties, counting EU member states bound through the Union's approval; the United States signed it in March 2022 and has not ratified it22. For an xStocks holder, with a certificate from a Jersey issuer under Swiss law and custody split between Swiss and US firms, Section 4.3's domicile test points to several courts at once, and nothing guarantees they agree.
One Token, Two Issuers, Two Sets of Rights
Section 4.4 showed that one coin can mean several tokens once it crosses a bridge. Multi-issuance is its jurisdictional version, already live in the cash leg most tokenized funds settle in. Circle's USDC Terms state that "From July 1st 2024, Circle Internet Financial Europe SAS will become a second, or 'dual issuer' of USDC," and that the Terms "only apply to holders of USDC located outside of the European Economic Area"23. EEA holders follow a separate MiCA redemption policy instead24. One fungible token carries two sets of redemption rights, divided by where the holder is, a fact the token itself does not record.
The FSB's peer review explains why this matters under stress. Many FSB members, it reports, see higher liquidity risk where "each co-issuer could be liable for the entire stock of tokens in circulation while holding only a fraction of the overall reserve of assets"2. Holders under stress have an incentive to redeem from "the co-issuer obliged to pay back with the shortest timeline and without charging fees," and "Contractual obligations between co-issuers (e.g., reserve rebalancing) may not be enforceable, particularly in stress scenarios or may be halted by measures adopted by a local authority"2.
EU authorities do not even agree on whether such schemes are permitted. The European Systemic Risk Board recorded that some national authorities consider them allowed under MiCA, "while others consider that they are not allowed," so that "potentially risky schemes rejected in one Member State may be approved in another"25. Noting "the authorisation of two entities as of September 2025" and that "third-country authorities may prioritise national interests, restricting the transfer of funds," it recommended that the Commission read MiCA "as not permitting such schemes"25. By September 2026 the question was still open: the Commission's May 2026 consultation asked respondents to rate "Regulatory arbitrage due to fungibility of tokens" as a risk16, and the ESCB wrote that "amendments to MiCAR are urgently needed to clarify the permissibility of third-country multi-issuer stablecoin schemes"6.
The stablecoin mechanics stay with Section 4.4; what carries over to the asset side is the template. A fund or note issued as one fungible token by affiliates in two domiciles would inherit the same questions about which issuer pays, from which reserve, and under which court.
Compliance Moves at the Pace of the Slowest Jurisdiction
Section 2.3 noted that "Sanctions screening, jurisdictional restrictions, and qualification thresholds have to be maintained over the life of the token." Those obligations are national and unevenly enforced. The FATF's July 2026 update found that 91 of 109 surveyed jurisdictions had passed Travel Rule legislation, up from 85 of 117 a year earlier, but that "almost half" of those with legislation had not yet taken "Travel Rule-related supervisory or enforcement action"26. Of 149 jurisdictions assessed against the FATF standard for virtual assets, "only one jurisdiction continues to be fully compliant"26. FATF President Giles Thomson warned of criminals exploiting "weak links in the global system"27.
Permissioned RWA tokens can respond by writing jurisdiction into the token. The ERC-3643 standard's identity registry links each wallet to "a country code corresponding to the investor's country of residence," and its compliance rules can impose "a limited number of holders per country"28. The issuer's jurisdictional map then travels with the token, as intended, but is only as current as the registry: a holder's move abroad, or a new sanctions listing, reaches the token when someone updates it. Section 5.5 treats the freeze and forced-transfer powers behind this as an attack surface. Here they are a jurisdictional choice exported through code.
At the other end sits the token outside any whitelist. The FSB notes that unhosted wallets can prevent "proper identification of token holders' geographic location"2, and the FATF reports that "Jurisdictions continue to report challenges with mitigating the risk of offshore VASPs"27. A token that can reach a venue in a jurisdiction that has not implemented the standards is, in practice, held to that venue's standards. On a global ledger, compliance moves at the pace of the slowest jurisdiction the token can reach.
Fragmentation Resolves as Arbitrage or as Silos
Divergence resolves in one of two directions, and both cost something.
The first is arbitrage. The FSB warned in 2024 that tokenization's reach across borders could worsen regulatory arbitrage where rules are inconsistent, "which could ultimately lead to a race to the bottom and threaten financial stability"1. IOSCO notes that businesses can exploit different paces of regulation by "choosing a jurisdiction with lighter or no regulation in which to structure their operations"3, and its tokenization report warns that gaps in regulatory treatment "expose investors to the risks of forgoing protection under the regulatory frameworks for dealing with non-regulated entities"11. For a holder, arbitrage means a product whose protections are those of its lightest jurisdiction.
The second is silos, in which the same exposure is split by residence. Section 4.3's eligibility table put BUIDL behind a "U.S. Qualified Purchaser" gate and USDY behind a "Non-U.S. Investor" one, though rwa.xyz counts both in the same Treasury class (Section 4.5), and kept xStocks out of the United States, the United Kingdom, Canada and Australia. Section 5.4 treats what that does to the buyer pool. Adrian describes the end state: without coordination, tokenization "risks fragmenting the global financial system into jurisdictionally bounded digital silos," which "would also undermine the efficiency gains that motivate tokenization in the first place"17.
Neither outcome is a technical failure, and no protocol upgrade chooses between them. In Adrian's words, "policymakers face a trade-off between preserving sovereign control and enabling global integration"17.
What Chapter 5 Establishes
Section 2.4 read an RWA as a dependency chain and asked who controls the asset, the facts, and execution. Chapter 5 took each question as a risk and added credit, liquidity and law.
| Section | Risk | Place in Section 2.4's chain | Question | Where the answer is decided |
|---|---|---|---|---|
| 5.1 | Oracles | Information layer | Who controls the facts? | The provider whose report reaches the chain |
| 5.2 | Custody and legal title | Control layer | Who controls the asset? | Property law and the product's terms, usually after a failure |
| 5.3 | Counterparty and credit | Around the chain | Who must pay, and in what order? | Every obligor standing in front of the holder |
| 5.4 | Liquidity | Around the chain | Who provides the exit? | The issuer's redemption capacity |
| 5.5 | Smart contracts | Execution layer | Who controls execution? | The holders of privileged and upgrade keys |
| 5.6 | Jurisdiction | Beneath the whole chain | Under whose law? | National courts applying laws that differ |
Read down the last column. Tokenization moved each dependency without removing it: the facts to a reporter, title to documents and courts, credit to a longer list of obligors, the exit to the issuer, discretion to keys. The last row differs in kind. Every other row ends in a dispute, and every dispute ends before a court applying national law to a token that has no nationality. Jurisdictional fragmentation is the ground on which the other five risks are resolved, and that ground shifts at every border the token crosses.
Chapter 6 turns from risk to response. Section 6.1 asks why regulation is not the enemy of RWAs, Section 6.2 maps the patchwork of rules that exists today, Section 6.3 follows the key jurisdictions already moving, and Section 6.4 explains why the regulatory question is the central question. Where Chapter 5 asked what breaks, Chapter 6 asks who writes the rules that decide what happens next.
- The standard-setters agree on principle but not on implementation. The FSB's October 2025 peer review of 37 jurisdictions' crypto-asset and stablecoin regimes found "significant gaps and inconsistencies" and cross-border cooperation that is "fragmented, inconsistent, and insufficient."
- One token can be a different instrument at each border. The Czech and Slovak authorities marked ESMA's classification guidelines partly "Not applicable" because their securities laws lack a tokenized form, and Robinhood sells stock tokens in the EU "under MiFID II as derivatives" while SEC staff attach registration and exchange-trading conditions to tokenized security-based swaps.
- No one yet agrees which country's law decides who owns a token. UNIDROIT's Principle 5 binds no State until adopted and leaves the token-to-asset link to other law, the Hague Securities Convention has three parties, and in May 2026 the European Commission was still asking whether the EU needs a conflict-of-law rule for tokens.
- Failure stays national. When FTX failed, the Bahamas regulator moved assets the US Chapter 11 debtors claimed, the sides valued them at about $296 million and more than $3.5 billion, and a bespoke settlement took about 13 months. The ESRB found EU authorities split on whether multi-issuer stablecoins are permitted.
- Fragmentation resolves as arbitrage toward the lightest regime or as silos that split one exposure by residence. The IMF's Tobias Adrian calls the choice "a trade-off between preserving sovereign control and enabling global integration," which is the question Chapter 6 takes up.