Legal Analysis: Repara­tions Loan Ukraine

The EU is struggling to find a way to mobilize Russian assets frozen in the EU for Ukraine. The Belgian government in parti­cular is citing legal and financial risks against the proposed “Ukraine repara­tions loan.” The expert opinion shows that such a loan does not entail any substan­tially new risks compared to the EU’s previous sanctions decisions; the advan­tages outweigh the disad­van­tages. We would like to thank attorney Patrick Heinemann for providing us with the expert opinion of the inter­na­tional team of authors.

Legal Analysis

UKRAINE REPARATIONS LOAN: EUROPE AND UK LITIGATION RISK ANALYSIS*

1. KEY CONCLUSIONS

On 3 December 2025, the EU Commission unveiled its most ambitious proposal to date to address Ukraine’s long-term funding needs: a two-part plan to immobilise the Central Bank of Russia’s (CBR) assets and reserves (until Russia ends its war of aggression and pays repara­tions to Ukraine) coupled with a “Repara­tions Loan” of up to EUR 210 billion to be disbursed in tranches over four years. The Repara­tions Loan proposes to use cash balances which accumu­lated at EU financial institutions—most notably Belgium-based Euroclear—following the immobi­li­sation of CBR assets in 2022.

The first part of the plan, the indefinite immobi­lization of CBR reserves was enacted three days ago, on 12 December 2025 as Council Regulation 2025/​2600 (the “Immobi­li­sation Regulation”).  The Repara­tions Loan proposal is to be voted on this week on 18–19 December 2025.

The United Kingdom has announced plans to enact a similar approach for appro­xi­m­ately £8 billion in immobi­lised Russian state money held in the UK.

Opponents question the proposal’s legality, citing the risk of legal claims by Russia or the CBR alleging asset “expro­priation.” The Commission’s view is that the proposal is lawful and does not amount to confis­cation. This has been affirmed by a number of distin­gu­ished academics and practi­tioners.[1]

This paper addresses (i) the litigation risks of the Immobi­li­sation Regulation and (ii) the possible additional litigation risks arising out of enacting the Repara­tions Loan proposal. Its conclusion is that while certain theore­tical litigation risks may have been created on 12 December 2025 under the Immobi­li­sation Regulation, enacting the Repara­tions Loan part of the proposal will add little additional litigation risk as no further loss would be suffered by the CBR.

Any alleged loss stemming from CBR’s inability to use its assets had already arisen under the initial and all subse­quent decisions blocking transac­tions with its assets culmi­nating in the Immobi­li­sation Regulation, which effec­tively froze CBR assets until such time as Russia ends its war and repays Ukraine.  The remainder of the Repara­tions Loan programme is essen­tially an internal matter of EU banking and finance. Any additional litigation risk that may be created by enacting the Repara­tions Loan proposal will be negligible.

In summary, possible litigation risks have arisen out of existing sanctions and the approval of the Immobi­li­sation Regulation. These risks are low and enacting the rest of the EU package that consti­tutes the Repara­tions Loan does not appear to add much, if any, additional risk for the following reasons:

1. Russian courts. Any judgement of a Russian court would not be recog­nised or enforced in the EU or the UK on public policy grounds, alter­na­tively, for lack of juris­diction, provided the relevant respondents do not volun­t­arily consent to Russian court’s juris­diction (including by defending any claim on the merits, as opposed to merely disputing jurisdiction).
2. National courts. Russia or the CBR are unlikely to bring claims in the EU or UK national courts, as doing so will risk waiving sovereign immunity.
3. Court of Justice of the European Union (CJEU). Any damages claim by Russia or the CBR, as non-Member States, would face an excep­tio­nally high bar, requiring proof of a “suffi­ci­ently serious breach” of EU law. That threshold is unlikely to be met by the Immobi­li­sation Regulation, let alone the remainder of the Repara­tions Loan package given that it does not entail the confis­cation of CBR assets. In any event, any claim would be against the EU and not any individual member state. Any challenge to the UK’s parallel scheme would face compa­rable obstacles. UK courts afford the government broad discretion in sanctions and foreign policy, and the scheme’s non-confis­catory design makes it difficult to show unlawfulness or compen­sable loss.
4. Inter­na­tional courts. No inter­na­tional court has juris­diction to hear such claims by Russia or the CBR, as Russia does not accept the juris­diction of either the European Court of Human Rights (ECtHR) or the Inter­na­tional Court of Justice (ICJ).
5. Investor-state arbitration. Claims under bilateral investment treaties (BITs) between Russia and EU Member States /​ UK would likely fall outside tribunal’s juris­diction, as BITs do not protect sovereign assets. In any case, such claims would be weak on the merits (especially on proof of loss) given the Repara­tions Loan’s design, and enforcement of any award in states supporting the asset immobi­li­sation would likely be refused on public policy grounds.
6. State-to-state arbitration. Such procee­dings are likely confined to disputes over BIT inter­pre­tation rather than substantive compen­sation claims. In any event, any compen­sation claim would face strong juris­dic­tional and merits-based defences based on absence of actionable loss.

In addition to not creating any additional legal risk, the Repara­tions Loan mechanism does not appear to create signi­ficant financial risks and this is supported by Standard & Poor and Fitch, the world’s top credit-rating agencies providing independent assess­ments of the credit­wort­hiness of countries, companies and financial instru­ments. Ahead of the 3 December 2025 Repara­tions Loan publi­cation, Fitch stated that Euroclear’s premium “AA/​Stable” debt rating and Belgium’s sovereign (A+/Stable) ratings would remain unaffected by the Repara­tions Loan proposal, provided legal and liquidity risks were shared among EU Member States (as is confirmed in the Commission’s draft legislation).

2. REPARATIONS LOAN: BACKGROUND

Appro­xi­m­ately €176 billion in cash is managed by Euroclear, out of roughly €193 billion in immobi­lised CBR assets held by it. These assets were origi­nally sovereign bonds—a debt owed by the bond-issuer state to the CBR—but by 2025 the majority had matured and converted into cash deposits, i.e. liabi­lities owed by Euroclear to the CBR. The deposits are denomi­nated in multiple currencies, including euros, pounds sterling, U.S. dollars, Canadian dollars, and Japanese yen and held in their respective national banks. Although sanctions block the transfer of cash deposits from banks to Euroclear, the amounts remain recorded on Euroclear’s balance sheet as assets and corre­sponding liabi­lities to the CBR.

Euroclear has been accumu­lating unpre­ce­dented profits from interest on immobi­lised cash deposits, exceeding €14 billion by the first half of 2025.  In February 2024, Council Decision (CFSP) 2024/​577 required these extra­or­dinary revenues to be segre­gated. Fitch subse­quently reported that Euroclear’s financial ratings remained unaffected by this move which was reported by Fitch as a precursor to using the profits to fund loans to Ukraine.

In June 2024, the G7 leaders agreed to provide $50 billion in Extra­or­dinary Revenue Accele­ration (ERA) loans to Ukraine, to be repaid using profits from the immobi­lised Russian sovereign assets held in Euroclear. The EU imple­mented this arran­gement through Regulation (EU) 2024/​2773 (the “ERA Loans Regulation”), which created the Ukraine Loan Coope­ration Mechanism to channel non-repayable EU support toward servicing the G7 ERA loans.

In September 2025, a more extensive proposal emerged to lend the full value of the frozen assets to Ukraine as a “Repara­tions Loan.” The concept involved investing Euroclear’s cash balances in a AAA-rated, zero-coupon European Commission debt instrument to finance appro­xi­m­ately €140 billion EU loan to Ukraine, repayable only if Russia compen­sates Ukraine for war damages.  On 25 October 2025, Fitch again confirmed that Euroclear’s debt rating and Belgium’s sovereign rating would remain unaffected, provided the legal and liquidity risks were shared across EU member states—a principle reflected in the EU Commission’s draft proposals.

The UK government has stated that appro­xi­m­ately £8 billion in CBR cash is currently immobi­lised in the UK financial insti­tu­tions and has announced plans to adopt an approach broadly analogous to the EU model. Under this proposal, the UK would borrow the frozen CBR cash interest-free, back the borrowing with UK sovereign debt, and on-lend the funds to Ukraine on an interest free basis, repayable only if and when Russia pays reparations.

3. EU REPARATIONS LOAN PROPOSAL

The proposal is struc­tured as a tightly-integrated package of EU regulations:

Part I (enacted on 12 December 2025)

• Council Regulation 2025/​2600 enacted on 12 December 2025 under Article 122(1) TFEU (emergency economic measures) prohibits any transfers of immobi­lised assets or reserves of the CBR and the Russian National Wealth Fund (Article 2) until Russian ends its war of aggression and pays repara­tions to Ukraine (Article 4) (the “Immobi­li­sation Regulation”).

Part II (to be agreed on 18–19 December 2025)

• Regulation 3502 of the European Parliament and of the Council under Article 212 TFEU (assis­tance to third countries) estab­lishes the Reparation Loan to Ukraine (the “Repara­tions Loan Regulation”).
• Council Regulation 3501 under Article 122(1) TFEU requires that EU financial insti­tu­tions invest their cash balances resulting from immobi­lised CBR assets into AAA-rated EU Commission debt instru­ments, which “shall be treated as equivalent to cash under appli­cable accounting rules” (Article 4) (the “Investment of Cash Balances Regulation”).
• Council Regulation 3500 under Article 312 TFEU extends EU budgetary guarantees to cover the Repara­tions Loan (the “EU Budgetary Guarantee Regulation”).

Together, this package provides for:

1. Recognition of repara­tions claims. A statutory recognition of Russia’s inter­na­tional law obligation (under the Articles on the Respon­si­bility of States for Inter­na­tio­nally Wrongful Acts (ARSIWA) and customary inter­na­tional law) to pay repara­tions to Ukraine to compensate for damage caused by its illegal war of aggression (Recital (11) of the Repara­tions Loan Regulation) and the use of such claims as colla­teral for the Repara­tions Loan (Article 20(2)(b) of the Repara­tions Loan Regulation).
2. Lock-in of asset immobi­li­sation now in force. A legal basis for the G7 commitment not to release Russian sovereign assets until Russia pays repara­tions to Ukraine (Article 2 of the Immobi­li­sation Regulation Regulation), reducing reliance on the current six-month, unanimity-based sanctions renewals that risk a full unfreeze.
3. Large-scale financial support. Creation of a loan facility of up to €210 billion to Ukraine, usable for macro-financial assis­tance and defence procu­rement, including compen­sation to indivi­duals harmed by Russia’s aggression, such as through the Council of Europe-backed Claims Commission for Ukraine (Recital 24 of the Repara­tions Loan Regulation).
4. Contingent repayment. Ukraine is required to repay the loan only in limited circum­s­tances: if Russia pays repara­tions (up to the amount received), or if Ukraine breaches specified political and rule-of-law conditions.
5. Risk sharing and guarantees. A three-tier system of financial protection for Belgium, Euroclear, and other holders of CBR assets, consisting of: 
   1. Voluntary Member State guarantees (Article 25 of the Repara­tions Loan Regulation);
   2. EU liquidity mechanism (or EU budget guarantees) under the EU Budgetary Guarantee Regulation (which is dependent on that regulation being passed unani­mously); and
   3. Union debt securities as a backstop of last resort.

Crucially, the proposal does not amount to confis­cation of CBR assets. It tempo­r­arily borrows cash generated by immobi­lised CBR reserves in various European banks without affecting CBR’s balances held at those banks including Euroclear or impairing the CBR’s balance sheet. The accumu­lated cash is replaced with AAA-rated European Commission bonds which, under Article 4 of the Investment of Cash Balances Regulation, are treated as “equivalent to cash” for accounting purposes. In substance, this is effec­tively a cash for bonds swap that releases liquidity for Ukraine while restoring Euroclear and other insti­tu­tions to holding debt securities rather than cash deposits, as was the case when the assets were first immobi­lised in 2022.  Claims that this amounts to a seizure are untenable absent an assertion that EU bonds are worthless.

Given the proposal’s design, the only conceivable scenario in which the CBR could incur a loss is the highly impro­bable event of a default on the EU Commission’s AAA-rated bonds. Multiple layers of guarantees are speci­fi­cally intended to guard against this risk. Once the CBR meets the condi­tions for the return of the blocked assets in Article 6 of the Immobi­li­sation Regulation, namely (i) end of the war of aggression, (ii) repara­tions to Ukraine, and (iii) no threat to the economy of the EU, repay­ments would be made by the EU and, where appli­cable, by Member States or third-party states providing guarantees, up to the amount of frozen funds lent to Ukraine. Accor­dingly, the Repara­tions Loan entails no financial loss for Russia or the CBR.

What if the Repara­tions Loan lacks unanimous support?

Unanimity is not required for the core elements of the proposal. The Immobi­li­sation Regulation has already been passed, while both the Repara­tions Loan Regulation and the Investment of Cash Balances Regulation are subject to Qualified Majority Voting, requiring approval by 55% of Member States repre­senting at least 65% of the EU population. As a result, the Repara­tions Loan proposal can proceed even if Belgium or Hungary withhold consent.

By contrast, the EU Budgetary Guarantee Regulation is subject to “a special legis­lative procedure,” requiring unanimity among all 27 Member States, in addition to the consent of European Parliament.

Failure to pass the EU Budgetary Guarantee Regulation would not invalidate the Repara­tions Loan. It would simply remove the EU-budget backed liquidity cushion, thus elimi­nating the second layer of guarantees.

The practical effect of this would be to limit the loan’s size because the Commission can borrow only amounts fully covered by voluntary Member States guarantees under Article 25 of the Repara­tions Loan Regulation. Because those guarantees are discre­tionary, the total guarantee envelope—and therefore the total funds that could be borrowed from Euroclear and others and on-lent to Ukraine—would remain uncertain.

4. ANALYSIS OF POTENTIAL RUSSIAN LITIGATION RISKS (none of which are increased by adopting the Repara­tions Loan proposal)

The European Commission considers the Reparation Loan proposal to be fully lawful and to carry minimal legal and economic risk. This assessment is supported by an opinion from EU, UK, and US academics,[2] which concludes that “[m]any legal experts have long since dismissed the argument that [the use of CBR assets to fund Ukraine] is illegal under inter­na­tional law,”  as reflected in the Inter­na­tional Law Commission’s Articles on State Respon­si­bility, including the right to take counter­me­a­sures in response to an inter­na­tio­nally wrongful act. This is undoub­tedly correct. Whilst even outright seizure and transfer of Russian assets to Ukraine would be permis­sible under inter­na­tional law;[3] the EU proposal is in any event deliberately struc­tured to avoid confis­cation both in law and in fact.

Critics—most notably Belgium, which hosts Euroclear and the largest share of immobi­lised CBR assets—argue that Belgium or Euroclear could be left bearing repayment risk if a court or arbitral tribunal were to order the return of funds to CBR. However, this risk has already arisen with the passing of the Immobi­li­sation Regulation which Russia may argue amounts to a de facto confis­cation of its reserves by freezing them indefi­nitely. Enacting the remainder of the Repara­tions Loan proposal will not affect the length of time CBR assets will remain immobi­lised and thus cannot be said to cause any additional loss to Russia or the CBR. Accor­dingly, given it will not cause any additional loss, the enactment of the Repara­tions Loan proposal cannot give rise to any additional cause of action by Russia or the CBR which does not already exist under the Immobi­li­sation Regulation.

The analysis below demons­trates that the risk of any enforceable adverse court judgment or arbitral award arising from the enacted Immobi­li­sation Regulation is very low and adopting the Repara­tions Loan program changes little. This conclusion is reinforced by two factors:

1. Despite (i) the immobi­li­sation of CBR assets for almost 4 years; and (ii) the explicit decla­ration in EU law that any profits on the immobi­lised assets do not belong to Russia or the CBR and can be used to finance the ERA loans to Ukraine, the only claim to date was filed by the CBR on 12 December 2025 – in the Moscow City Arbitrazh Court (a first-tier commercial court in Russia); and
2. Credit-rating agencies have confirmed that the Repara­tions Loan would have no impact on either Belgium’s sovereign credit rating or Euroclear’s credit rating.

A. Risk of adverse judgment by Russian courts

This is the sole venue where a claim has been filed illus­t­rating CBR’s reluc­tance to contest the Repara­tions Loan in any independent and neutral forum. The press reports (citing Economic Commis­sioner Valdis Dombrovskis) indicate that Brussels antici­pated such a retali­atory claim.

In its press release, the CBR states that the claim concerns the “unlawful activities of the Euroclear depository that cause damage to the Bank of Russia and the fact that the European Commission officially considers proposals for direct or indirect use of Bank of Russia assets without autho­ri­sation.” However, the sole “asset” held by CBR with Euroclear is CBR’s claim to repayment of amounts standing to its credit on Euroclear’s balance sheet. This claim remains intact whether or not the cash balances accumu­lated at Euroclear would be invested in the EU bonds under the Investment of Cash Balances Regulation or in any other securities. It is difficult to see what further loss to the CBR, beyond that already caused by the decision to immobilise CBR’s funds, could arise from imple­menting the remainder of the Repara­tions Loan proposal.

Given the Russian courts’ well documented [4] suscep­ti­bility to pressure from the state, parti­cu­larly in cases involving Russian state interests, CBR’s claims in Russian courts are nevert­heless extremely likely to succeed.

However, any such adverse Russian judgment will not be recog­nised or enforced in the EU or  the UK due to (i) the EU ban on enforcement; (ii) the lack of Russian court’s juris­diction (provided Euroclear limits its response to a juris­dic­tional challenge); and (iii) public policy grounds (including lack of fair trial in breach of Article 6 of the European Convention on Human Rights (ECHR)).

          (i) EU ban on enforcement

Article 11 of the Regulation 833/​2014 (adopted in the 15^th sanctions package) bars EU courts from enforcing Russian courts’ judgments made under Article 248 of the Russian Arbitrazh Proce­dural Code, a domestic proce­dural rule that provides Russian courts with exclusive juris­diction where sanctioned parties are involved regardless of the parties’ contractual dispute resolution arrangements.

Article 4 of the Immobi­li­sation Regulation speci­fi­cally prohibits enforcement of any judgment or award arising out of the immobi­li­sation of CBR funds in any EU court.

Thus CBR’s reported claim against Euroclear in the Moscow City Arbitrazh Court would be unenforceable in any EU courts under Article 11 of the Regulation 833/​2014 or Article 4 of the Immobi­li­sation Regulation. 

          (ii) Challenging Russian courts’ juris­diction /​ anti-suit and anti-enforcement injunctions

Any Euroclear-CBR contract would contain standard dispute resolution clauses in favour of Belgian courts or arbitration (not Russian courts). The CBR’s reported filing in Moscow therefore breaches its contractual commit­ments with Euroclear. Any procee­dings brought in a Russian court in violation of the parties’ agreement to resolve their disputes in other forums may be restrained by an anti-suit or an anti-enforcement injunction,[5] and any Russian court judgment obtained in defiance of such an injunction would not be recog­nised or enforced.

Even absent an injunction, EU and UK courts would refuse recognition and enforcement any Russian judgment where the Russian court lacked juris­diction under EU or UK rules (that protect the parties’ agreed-upon contractual dispute resolution choice). Thus Euroclear must ensure that it does not volun­t­arily submit to the Russian court’s juris­diction by taking part in the Russian procee­dings (other than to contest jurisdiction).

          (iii) Public policy grounds including lack of fair trial 

Any Russian court judgment would not be recog­nised or enforced in the UK or EU  where doing so be contrary to public policy.[6]  UK and EU public policy encom­passes both decision to immobilise the CBR funds and the fair trial guarantees enshrined in  Article 6 ECHR. Given the signi­ficant evidence that foreign companies cannot obtain a fair trial in the Russian courts where Russian state interests are involved,[7] an adverse judgement against Euroclear would be refused recognition or enforcement.  The UK or the EU courts are under a duty not to give effect to any foreign judgment that violates fair trial standards[8] or is otherwise contrary to public policy.

B. Risk of adverse judgment by third party national courts

Any claim by Russia against Euroclear or Belgium in any national court (including Belgian courts), is highly unlikely. Russia or the CBR would waive of sovereign immunity by suing in a third party national court, thereby exposing themselves to counter­claims for any seizure of assets in Russia and poten­tially war-related losses caused by Russia’s aggression in Ukraine.

In July 2025, the European Court of Human Rights found Russia respon­sible for losses in Ukraine arising from conduct between 2014 and autumn of 2022. Those losses alone exceed $300 billion, roughly the total amount of immobi­lised Russian sovereign assets. It is implau­sible that Russia would invite further exposure in European domestic courts by litigating amounts disbursed under a Repara­tions Loan that is expli­citly tied to Russia’s obligation to make reparations.

Even if Russia or the CBR brought a claim against Euroclear in the third party national courts, it would likely fail on the merits:

• No loss: Any EU or UK court is unlikely to find in favour of the CBR because the cash-for-bonds structure does not cause any loss. The CBR’s balance at Euroclear and its legal claim to the funds remain unchanged and can be enforced once condi­tions in Article 6 of the Immobi­li­sation Regulation are satisfied. The substitute only affects Euroclear: its claim to cash held at national banks is replaced by a claim against the European Commission under EU bonds. In practical terms, cash is exchanged for instru­ments equivalent to cash, making it difficult for the CBR or Russia to identify any loss. Any residual non-repayment risk lies with Euroclear, which is addressed through robust and layered guarantees.

• Force majeure (poten­tially). Euroclear may also invoke force-majeure protec­tions, depending on its contractual wording with the CBR, where perfor­mance of a contract is prevented by super­vening legal requi­re­ments[9] (which could include legis­lation mandating investment of cash balances into EU bonds).

C. Risk of adverse judgment by the EU Courts (General Court of the European Union and Court of Justice of the European Union (CJEU)).

A non-contractual damages claim may be brought under Article 340 TFEU for harm caused by EU insti­tu­tions in the perfor­mance of their duties. In principle, this means that a claim could be available if an EU legis­lative act, such as the Repara­tions Loan package, causes damage. In practice, successful cases for so-called “public torts” are extremely rare.

Any such claim would be against the EU (not Belgium or any Member State), and the Commission appears to have assessed the risk as close to nil. The prospect of success is remote because:

• Russia/​CBR unlikely to have any standing. Article 340 TFEU is designed for claims by indivi­duals or corporate entities; it is doubtful that a non-EU sovereign state or its organ (Russia/​CBR) has standing to bring such a claim. 

• Merits threshold is extremely high. The claimant must show a “suffi­ci­ently serious breach” of EU law and a direct causal link to actual loss. The Repara­tions Loan package is struc­tured to avoid confis­cation and, as designed, does not deprive the CBR of its legal claim against Euroclear. Absent loss and an identi­fiable serious breach, liability is unlikely.

Russia could also theore­ti­cally bring a claim under Article 263 TFEU challenging the legality of any Repara­tions Loan legis­lation.[10] However, any such judicial review challenge will not be against Belgium or Euroclear and likely to fail for the same reasons as Venezuela’s challenge to EU restrictive measures in Venezuela v Council of the European Union (T‑65/​18 RENV).[11]

D. Risk of adverse judgment by inter­na­tional courts

No inter­na­tional court has juris­diction to hear any potential claims by Russia or the CBR for alleged losses.

European Court of Human Rights (ECtHR)

The Court’s juris­diction extends only to Council of Europe member states that have ratified the ECHR and its Protocols. Russia ceased to be a member of the Council of Europe in March 2022 and automa­ti­cally withdrew from the ECHR in September 2022 and therefore cannot bring new appli­ca­tions, including through its state organs (such as the CBR), to the ECtHR.

Even if juris­diction existed, Article 1 of Protocol No. 1 to the ECHR protects private property rights, not sovereign assets, and would not apply to claims concerning central bank reserves.

Moreover, any appli­cation by Russia risks being joined with the pending cases brought by Ukraine and Nether­lands against Russia for human rights viola­tions arising from its illegal invasion of, and aggression against, Ukraine since 2014 and the downing of MH-17. In light of the adverse findings made against Russia in July 2025, it is highly unlikely that Russia would invite such a joinder by initiating new procee­dings before the ECtHR.

Inter­na­tional Court of Justice (ICJ)

The ICJ can exercise juris­diction only in limited circum­s­tances, none of which would apply in the Repara­tions Loan scenario:

1. Compulsory juris­diction by consent. Russia has not accepted the ICJ’s compulsory juris­diction and is highly unlikely to do so, given its ongoing breaches of inter­na­tional law.
2. Special agreement via written consent. Juris­diction may arise from written consent by both parties to a dispute, but the EU, Belgium, the UK, or any other parti­ci­pating state in the Repara­tions Loan would not provide such consent.
3. Juris­diction under a treaty. The only conceivable treaty Russia could invoke is the 2004 United Nations Convention on Juris­dic­tional Immunities of States and their Property. But this convention applies only to judicial, as opposed to legis­lative or executive actions (Article 1). In any event, that avenue is forec­losed because the Convention is not yet in force and has not been ratified by Russia. 

E. Risk of an adverse award under investor-state dispute settlement (ISDS)

Arbitration claims under bilateral investment treaties (BITs) between Russia and EU Member States also pose a potential litigation risk, including under the Belgium/​Luxembourg – Soviet Union  Russian Federation Bilateral Investment Treaty 1989 (the “BL-RF BIT”), which is the most plausible treaty route for a claim impli­cating Belgium. The analysis below generally applies to other Russia-EU Member State BITs with similar clauses.

For comple­teness, the 1994 Partnership and Coope­ration Agreement  between the EU and Russian Federation is unlikely to generate awards: it contains no investor-state arbitration clause, and its state-to-state settlement process contem­plates consul­ta­tions and diplomacy rather than arbitration.

Overall ISDS risk remains low due to:

(1) The likely lack of tribunal juris­diction over a Russia/​CBR claim;

(2) An extremely low likelihood of liability on the merits; and

(3) Limited enforcea­bility in juris­dic­tions supporting immobi­li­sation, on public policy grounds. 

(1) Lack of jurisdiction

The BL-RF BIT was not designed to cover central bank invest­ments, so it is unlikely to apply to the CBR assets. That said, it cannot be entirely ruled out that a tribunal (depending on its compo­sition) might nevert­heless assert juris­diction to consider such a claim.

Strong juris­dic­tional objec­tions include:

(i) Investment and terri­torial nexus. The BIT only covers assets invested “in accordance with” host-State legis­lation and “in the territory” of the host State (Art. 1.1.1 of the BL-RF BIT). Cash balances accumu­lated at Euroclear as a result of immobi­li­sation do not cleanly satisfy those requi­re­ments: the under­lying cash is held at national banks in the currencies’ issuing juris­dic­tions, and Euroclear’s role as a securities depository (record-keeping and settlement) hardly consti­tutes “investment” under the BIT.

(ii) Narrow consent to arbitration. The consent clause (Article 9(1)) is arguably limited to the amount and method of compen­sation payable under Article 5, not disputes over whether expro­priation occurred or concerning the legality of immobi­li­sation, swaps, or transfers as envisaged under the Repara­tions Loan proposal. This reading is supported by Russia’s own arguments in Berschader v Russian Federation, where a tribunal concluded that the parties to the BL-RF BIT intended such disputes “to be submitted to dispute resolution proce­dures provided for under the appli­cable contract or alter­na­tively to the domestic courts of the Contracting Party in which the investment is made.”[12] This means that the CBR would first need a Belgian court judgment estab­li­shing an expro­priation before arbitration over compen­sation could proceed. Notably the decision of the Russian court will not suffice since the judgment must be of the court of the country where the investment is made, ie Belgium, or any other contrac­tually designated forum (which is not a Russian court).

(iii) CBR will not qualify as an “investor.” Although the definition appears broad, BITs operate to protect private investors within a three-party structure (host state–investor–home state).  An exception may apply if a state-controlled insti­tution acts in a commercial, rather than a govern­mental capacity.[13] The CBR is manifestly not a private investor as it performs sovereign functions and would ordinarily invoke sovereign immunity to protect its assets. It would be internally contra­dictory for the CBR to claim sovereign status for immunity purposes while claiming private-investor status for BIT protection. In other words, a claim by the CBR would amount to a dispute between two states acting as equals, not a dispute between a state and a protected investor, and therefore falls outside the scope of BIT protection.[14]

(2) No liability on the merits 

Even if juris­diction were found, liability is unlikely:

(i) Public order carve-out. Invest­ments are protected subject to measures necessary to maintain public order[15] (Art. 4.2 of the BL-RF BIT). EU legis­lation aligned with the UN General Assembly resolu­tions recog­nising Ukraine’s right to repara­tions from Russia provides a strong public order justi­fi­cation for any actions by Belgium under the Repara­tions Loan proposal.

(ii) Public inter­na­tional law context. A tribunal would apply public inter­na­tional law, including decisions of the ICJ on protection against expro­priation under BITs. In Certain Iranian Assets,[16] the ICJ indicated that not every seizure of a state bank’s assets consti­tutes a compen­sable expro­priation; liability only arises where the expro­priatory measures is itself tainted by a specific illegality, such as a denial of justice or the appli­cation of executive measures that themselves breach inter­na­tional law. The Repara­tions Loan is designed to avoid confis­cation and is a propor­tionate response to Russia’s grave breaches of inter­na­tional law, including Article 2(4) of the UN Charter. The CBR is insepa­rable from the Russian state and is one of the key instru­mental actors in Russia’s ongoing breaches of inter­na­tional law in occupied terri­tories in Ukraine.[17] Council Regulation EU 2025/​1494 (imple­menting the EU’s 18^th sanctions package) also recog­nised that “the Russian banking and financial sector is key to Russia’s war effort,”  making any compen­sation claim arising from the Repara­tions Loan unlikely to succeed.

(iii) Clean hands. Investment tribunals have applied a clean-hands doctrine to deny relief where the claimant’s unlawful conduct contri­buted materially to the situation. Al Warraq v. Republic of Indonesia illus­trates this approach: the tribunal held that a claimant’s viola­tions of the host state’s laws were “preju­dicial to the public interest” and therefore stripped him of BIT protection, barring his claim despite findings of unfair treatment. Russia’s aggression similarly materially contri­butes to the circum­s­tances giving rise to the measures at issue.

(iv) No loss. Russia’s/CBR’s claim to its funds remains intact and Russia’s assets will one day be unfrozen provided Russia complies with the condi­tions in Article 4 of the Immobi­li­sation Regulation (end of its war of aggression and payment of repara­tions to Ukraine).

(3) No enforcement on public policy grounds

Even if Russia/​the CBR succeeded in obtaining an award against Belgium, enforcement in the EU, UK, U.S., Canada, Japan, and any other states supporting immobi­li­sation would likely be refused as contrary to the public policy of those states—specifically, the policy that Russian sovereign assets should not be released until Russia pays repara­tions to Ukraine.

The 18^th EU sanctions package (imple­mented in Council Regulation EU 2025/​1494) intro­duces protec­tions against sanctions-related BIT procee­dings, including a non-recogni­ti­on/non-enforcement approach. Recital (22) and the operative provi­sions specify that Member States should not recognise or enforce injunc­tions, orders, judgments, or arbitral decisions arising from ISDS procee­dings connected to sanctions measures, and treat effective imple­men­tation of the “no-claims” clause as EU and Member State public policy.

The Reparation Loan Regulation contains analogous language (Recital (57)), urging Member States to contest ISDS procee­dings and resist enforcement on public policy grounds. Similarly, Article 4(1) of the Immobi­li­sation Regulation provides that no arbitral award (or court judgment) obtained by Russia or CBR arising out of the immobi­li­sation of CBR assets will be recog­nised, given effect to or enforced in the EU.

Existing EU burden-sharing (limited relevance)

EU Regulation No 912/​2014 provides a framework for allocating financial respon­si­bility where Member State action required by EU law triggers investment-treaty liability. However, it will not apply to BITs to which the EU is not a party, including the BL-RF BIT. This supports the policy case for termi­nating legacy Member State BITs in favour of an EU-level approach (as per Recital (59) of the Repara­tions Loan Regulation).

Distin­gu­ishing sanctions-related BIT claims by private individuals

Any hypothe­tical CBR claim should be distin­gu­ished from pending claims by sanctioned private indivi­duals (none of which have succeeded to date). The widely reported $16 billion arbitration claim by the sanctioned Mikhail Fridman against Luxem­bourg under the BL-RF BIT illus­trates the diffe­rence from any possible claim by the CBR under the Repara­tions Loan proposal:

(i) Fridman can readily establish “investor” status (unlike the CBR).

(ii) His claim follows the Judgment of the General Court of the EU (First Chamber)[18] annulling earlier sanctions listings for eviden­tiary reasons (notably, he remains sanctioned on the basis of new evidence obtained). Fridman can therefore rely on that annulment in his BIT claim. By contrast, it is extremely unlikely that the EU courts would annul the Repara­tions Loan legis­lation, as the CJEU has never allowed a merits challenge to Common Foreign and Security Policy measures, treating them as excep­tional instru­ments largely insulated from challenge.

(iii) His alleged loss relates to the loss of value of his Alfa Group, which assets are held by a company regis­tered in Luxem­bourg (ultim­ately a question of expert company valuation); a CBR claim with respect to the Repara­tions Loan (if conceivable at all) would be capped in practical terms by the value of immobi­lised assets.

F. Risk of adverse award in the inter-state arbitration

Article 9 of the BL-RF BIT provides a three-stage state-to-state dispute resolution mechanism: (1) diplo­matic efforts; (2) referral to a joint commission; and (3) if unresolved within six months,  ad hoc arbitration.

There is a debate whether such inter-state dispute resolution clauses cover the full range of treaty disputes or are limited to questions of treaty inter­pre­tation. The prevailing view is that BITs deliberately bifurcate: inter­pre­tation disputes go to state-to-state arbitration, while alleged breaches of substantive investor protec­tions go through investor-state arbitration route (Article 10).[19]

Even if a tribunal admitted any substantive (as opposed to inter­pre­tative) claims by Russia under Article 9, they would likely fail on:

(1) Juris­diction: CBR’s custodial relati­onship with Euroclear does not constitute an “investment” made “in the territory” of Belgium; Euroclear’s depository role is primarily record-keeping and settlement facilitation.

(2) Merits: generally recog­nised principles of inter­na­tional law, especially counter­me­a­sures, would provide strong defences and preclude liability. Criti­cally, Article 9(6) requires the tribunal to decide on the basis of the Treaty and “generally recognized norms and principles of inter­na­tional law,” effec­tively inviting counter­me­a­sures or essential-security-interest defences where the measures respond to gross breaches of inter­na­tional law and are anchored in EU legis­lation and UN General Assembly-level recognition of Ukraine’s reparation claim.

REPUTATIONAL RISKS

It has been argued that even unsuc­cessful litigation challenging the Repara­tions Loan could generate reputa­tional harm to Belgium or Euroclear. That argument is weak. As explained by a number of prominent Belgian and inter­na­tional jurists in their opinion: “[i]nvoking reputa­tional risk in this context would be tanta­mount to arguing that one should refrain from seizing the bank accounts of drug traffi­ckers or the proceeds of corruption, on the grounds that this might worry other clients about the security of their funds. It is precisely the opposite: what builds the reputation of a financial system is its ability to distin­guish between legitimate and illicit assets and to handle the latter in accordance with the law.”

Moreover, the proposal is designed as a non-confis­catory mechanism that repur­poses an aggressor’s immobi­lised liquidity to mitigate the damage it inflicts on Ukraine and its neighbours.

CONCLUSION

The EU Repara­tions Loan proposal provides a non-confis­catory financing mechanism for Ukraine that is consistent with EU and inter­na­tional law. To address legal and financial concerns from Belgium and others, it embeds a robust two- to three-layer guarantee (depending on whether the EU budgetary guarantee layer is adopted by unanimity) to ensure that any amounts ultim­ately payable to the CBR are satisfied.

The proposal also provides that any residual BIT-award risk should be shared among Member States through guarantee agree­ments (Recital 58), with Germany alone calcu­lated to bear appro­xi­m­ately 25 percent of the risk.

Euroclear has already reported holding appro­xi­m­ately €5 billion as a “buffer” for sanctions-related legal risk and may also indefi­nitely retain 3 percent of proceeds from immobi­lised assets for adminis­trative fees and an additional 10 per cent. as a temporary legal buffer.

In short, any possible risks faced by Belgium—described in Prime Minister De Wever’s 27 November 2025 letter to President von der Leyen as the only country with “skin in the game”—have already been created by sanctions and on 12 December 2025 by the adoption of the Immobi­li­sation Regulation. No material new risks will be created by adopting the full Repara­tions Loan plan and any such negli­gible risks are materially outweighed by the proposal’s benefits for European peace, security, stability, and the long-term viability of Ukraine.

*AUTHORS AND SIGNATORIES

Tetyana Nesterchuk (Barrister, Fountain Court Chambers; Non-resident Senior Legal Fellow at KSE Institute)

Anna Vlasyuk (Head of Inter­na­tional Law and Policy Research, KSE Institute)

Dr Patrick Heinemann (Partner, Bender Harrer Krevet)

Dr Anton Moiseienko (Senior Lecturer, Research Director, School of Law, The Australian National University)

Philip Zelikow (Botha-Chan Senior Fellow, Hoover Insti­tution, Stanford University)

Yuliya Ziskina (Senior Legal Fellow, Razom for Ukraine)

[1] See the opinion by Francis Biesmans (economist and statis­tician, professor emeritus, University of Lorraine); Samuel Cogolati (doctor of inter­na­tional law, KU Leuven); Paul De Grauwe (professor, John Paulson Chair in European Political Economy, London School of Economics); Pierre Klein (professor, Centre for Inter­na­tional Law, ULB); André Lange (member of the Board of Directors of the association For Ukraine, for their freedom and ours); Gerard Roland (former E. Morris Cox Professor of Economics and Professor of Political Science at UC Berkeley and ULB).

[2] See fn. 1 above.

[3] See “On proposed counter­me­a­sures against Russia to compensate injured states for losses caused by Russia’s war of aggression against Ukraine” by Professor Dapo Akande (Chichele Professor of Public Inter­na­tional Law Oxford University Essex Court Chambers, London), Professor Olivier Corten (Center for Inter­na­tional Law Université libre de Bruxelles), Professor Shotaro Hamamoto (School of Government/​Graduate School of Law,  Kyoto University) Professor Pierre Klein (Center for Inter­na­tional Law, Université libre de Bruxelles), Harold Hongju Koh (Sterling Professor of Inter­na­tional Law, Yale Law School), Paul Reichler (Public Inter­na­tional Law Practi­tioner, 11 King’s Bench Walk Chambers, London), Professor Hélène Ruiz Fabri (Sorbonne Law School, Université Paris), Professor Philippe Sands (University College London, 11 King’s Bench Walk Chambers, London), Professor Emeritus Nico Schrijver (Grotius Centre for Inter­na­tional Legal Studies, Leiden University, the Nether­lands), Professor Christian J. Tams (University of Glasgow, 11 King’s Bench Walk Chambers, London),  Philip Zelikow (Senior Fellow, Hoover Insti­tution Stanford University)

[4] The evidence of Russian courts’ lack of impar­tiality and indepen­dence from the Russian state was accepted by the English courts on a number of occasions. See, for instance, Russian Aircraft Operator Policy Claims (Juris­diction Appli­ca­tions) [2024] EWHC 734 (Comm) where the English court found that foreign companies were unlikely to receive a fair trial in Russia (relying on agreed expert evidence from both parties that at least “in cases which are of suffi­cient interest to the Russian State, it is capable of affecting the outcome of judicial decisions” (at [259]-[263])).

[5] See Google LLC v NAO Tsargrad Meda [2025] EWHC 94 (Comm) where the English Commercial court granted Google an anti-enforcement injunction to prevent Russian companies from enforcing Russian court orders—specifically, compounding fines of extra­or­dinary value—outside Russia, as these orders stemmed from procee­dings brought in breach of exclusive juris­diction and arbitration agree­ments in favour of England and Wales or London arbitration. The court also granted an anti-anti-suit injunction to pre-empt the Russian court’s anti-suit injunction. The court found it just and conve­nient to grant such injunc­tions, noting that the Russian procee­dings relied on Article 248 of the Russian Arbitrazh Proce­dural Code, which under­mined the parties’ contractual choice of dispute resolution venue.

[6] See Dicey, Morris & Collins on The Conflict of Laws, 16^th edn : „RULE 5 – English courts will not enforce or recognise a right … or legal relati­onship arising under the law of a foreign country, if the enforcement or recognition of such right … or legal relati­onship would be incon­sistent with the funda­mental public policy of English law.“

[7] See fn 4 above.

[8] See, for instance, OJSC Bank of Moscow v Chern­yakov [2016] EWHC 2583 (Comm) at [10].

[9] See Lewison, „The Inter­pre­tation of Contracts” 8th Ed. at [13.06]; RTI v MUR Shipping BV [2024] UKSC 18 at [2] under UK law; Reichs­ge­richt, judgement of 7 April 1927 – IV 745/​26 –, RGZ 117, 12 (13); Bundes­ge­richtshof, judgment of 16 May 2017 – X ZR 142/​15 –, BGHZ 215, 81 margin no. 8 under German law: Cour de cassation, Arrêt du 14 avril 2006, Pourvoi n°02–11.168 under French law.

[10] See Venezuela v Council of the European Union (C‑872/​19 P) where it was found that Venezuela had standing to bring such a challenge.

[11] In summary, the court found that the state’s right to be heard prior to the imposition of sanctions does not exist (§§36–45); the obligation to state reasons will have been fulfilled (§§36–45); there will have been no manifest error in the assessment of the facts (§§36–45); and whether or not sanctions constitute lawful counter­me­a­sures is irrelevant (§§36–45).

[12] Berschader v Russian Federation, SCC Case No. 080/​2004, Award (21 April 2006) at §153 which reflected Russia’s arguments that: “the Soviet Union proceeded on the basis that the question of the presence or absence of an act of expro­priation must in every parti­cular case be decided by the national court of the state in the territory of which the expro­priation was alleged to have taken place. The Respondent [Russia] maintains that this was a point of principle of the Soviet Union and relies upon the dispute settlement provi­sions of all the treaties concluded by the Soviet Union in support of this contention.” (see § 63) This argument would therefore apply to all historic BITs inherited by Russia from the Soviet Union.

[13] Rudolf Dolzer, Ursula Kriebaum and Christoph Schreuer, Principles of Inter­na­tional Investment Law (3rd edn, Oxford University Press 2022) 172.

[14] See Dr Patrick Heinemann’s paper “Make Russia Pay: on the confis­cation of assets of the Russian Central Bank in Germany”.

[15] Art. 4.2 of the BL-RF BIT: “sous réserve des mesures néces­saires au maintien de l’ordre public”

[16] Islamic Rep. of Iran v. United States, Judgment, 2023 I.C.J. Rep. Mar. 30, 2023 at § 184: “A specific element of illegality related to that decision is required to turn it into a compen­sable expro­priation. Such an element of illegality is present, in certain situa­tions, when a depri­vation of property results from a denial of justice, or when a judicial organ applies legis­lative or executive measures that infringe inter­na­tional law and thereby causes a depri­vation of property.“

[17] See, for instance, a brief filed by a group of NGOs with the Inter­na­tional Criminal Court (ICC) on CBR’s role in war crimes and crimes against humanity on temporary occupied terri­tories of Ukraine.

[18] This judgment has been appealed to the Court of Justice of the European Union (CJEU) but CJEU is likely to uphold it based on early indication of the Advocate General’s opinion.

[19] The view that, in the two-track (investor-state and inter-state) regime, inter-state disputes should not infringe on investor-state disputes is  expressed in the Expert Opinion with Respect to Juris­diction, Prof. W. Michael Reisman (English) (Republic of Ecuador v. United States of America (PCA Case No. 2012–5)