$200bn of Frozen Russian Assets Just Re-Entered the Debate — Belgium Says 'Door Is Closed', Four EU States Just Pushed It Back Open

On August 27, 2026, the Financial Times reported that Sweden, the Netherlands, Spain and Poland have written to the European Commission asking it to revive a plan to fund Ukraine using more than €200bn of immobilised Russian central bank assets, Reuters confirmed the same day. The push comes less than four months after EU leaders finalised an alternative €90bn Ukraine loan on April 23, 2026 — one that deliberately did not touch the Russian reserves, after Belgian Prime Minister Bart De Wever blocked the so-called reparations loan in December 2025. Belgium's defence minister Theo Francken responded on the record that the door on using the Russian assets is \"non-negotiable\" and \"closed,\" but the four-country letter has now forced the issue back onto the agenda before EU foreign ministers met in Ireland the week of September 1, 2026.

Mechanically, the \"reparations loan\" proposed by the European Commission on December 3, 2025 would have the EU borrow cash balances associated with the immobilised Russian assets — ~€210bn in total — and on-lend to Ukraine as a zero-coupon facility that Kyiv only repays after it receives reparations from Russia. The Commission has framed the structure as debt-neutral for Ukraine and as a way to put the principal at risk rather than only the windfall profit. That is the version Belgium rejected, and it is the version the August 27 letter wants back.

Even without the reparations loan, the EU is already running the second iteration of the same idea. The €90bn loan finalised on April 23, 2026 is funded by EU-Bond issuance on the capital markets; the Commission raised its H1 2026 issuance target from €80bn to €90bn and disclosed €45.4bn issued by mid-June 2026, with a separate €80bn funding target set for H2 2026. The structure used today borrows against the EU's \"headroom\" in the long-term budget. The reparations-loan variant would swap that headroom collateral for the Russian reserves themselves — the same cash balances, lent at zero coupon, that today sit on Euroclear's balance sheet earning interest for the EU.

That is the \"demand floor\" — investors price the bonds knowing that even in a tail scenario where Ukraine defaults and Russia never pays reparations, the principal is anchored to a pool of G7-controlled reserves whose face value already exceeds the issued debt. For pension funds, insurers and reserve managers the structure looks closer to a collateralised EU obligation than a plain sovereign loan, and that is exactly why Belgium's objection is about legal title to the assets, not credit risk.

Euroclear's FY2025 results are the cleanest summary of why Belgium and the company are nervous. Underlying business income excluding Russia rose 6% to nearly €1.9bn, but the headline profit line of €1.1bn still leans on €5bn of interest earned on the immobilised reserves. The company has provisioned €342m of additional direct operating expenses for Russian sanctions and countermeasures — a one-line number that captures the legal, compliance and operational cost of being the place where €195bn of another country's reserves is parked. Move those balances, or change their status, and the €5bn earnings stream becomes a different question entirely: a loan principal rather than windfall profit.

Rheinmetall is the cleanest single read on the transmission chain. Q2 2026 sales came in at €3,289m, up 69% year-on-year, with operating profit of €562m — and order intake of roughly €11bn, up 476% YoY, was driven by an ammunition package for Hungary and 155mm artillery ammunition for Ukraine. Management ended Q1 2026 with €73bn of backlog and the order book blew past €80bn by late July 2026 on Ukraine onshoring announcements. The implication is mechanical: every additional euro the EU channels through the Ukraine loan ultimately lands in Rheinmetall's weapon-and-ammunition book at margins that have lifted gross profit margin to 37.3%.

BNP Paribas and ING Groep are the two European banks most often named in any Russian-asset transmission. Direct Ukraine/Russia exposure is in fact modest — BNP Paribas disclosed combined gross exposure of roughly €3bn in March 2022, equivalent to 0.16% of its risk-weighted assets, and has since wound down its Russian subsidiary business. ING terminated the sale agreement for its Russian business in April 2026 after the proposed buyer could not complete the deal, but offshore exposure has fallen roughly 90% to €0.6bn at year-end 2025. The transmission is therefore not credit risk on Russian counterparties — it is legal and operational exposure as primary dealer and settlement agent for the EU-Bond programme that funds the Ukraine loan.

What changes for the banks is the marginal balance-sheet and fee line. EU-Bond issuance is syndicated through a handful of primary dealers; the planned €170bn of issuance across 2026 generates underwriting fees, market-making P&L and repo exposure for the dealer group. BNP Paribas's 4.97x EV/Sales and 9.0x trailing P/E — among the cheapest large European banks — already reflect a market that is not paying for repricing; if the reparations loan goes through, that multiple gap narrows because BNP is structurally long the EU bond complex. ING Groep sits closer to the Dutch political momentum driving the plan and benefits asymmetrically if the issue is reframed as a Dutch-led initiative.

Step one: the EU issues debt against Russian collateral rather than budget headroom. This converts a frozen-but-earning €195bn securities book into the formal junior-senior structure behind EU-Bond issuance — the legal wrapper around what is effectively a G7-coordinated claim on Russian state assets. Step two: the proceeds are on-lent to Ukraine at zero coupon, debt-neutral for Kyiv, and routed through the €6.1bn-per-tranche EU defense procurement disbursements that have already begun. Step three: those disbursments are bid out by Ukraine's Defence Procurement Agency DOT — Ukraine has contracted more than 22,000 unmanned ground vehicles in 2026 alone — and the bulk of the marginal European kit flows through Rheinmetall for shells, Saab for airframes and missiles, Hensoldt for sensors, BAE Systems for vehicles and light artillery. The US interceptor and avionics line gets pulled through Northrop Grumman and Lockheed Martin under FMS contracts funded by allied contributions from EU member states. The chain is the legal wrapper → EU bond demand → defense procurement → prime order book → component supply chain, and the slack variable in the system is whether Belgium concedes on legal indemnity in the September–October 2026 window.