European Defence Bonds: A Blueprint

The previous article established why the dependency exists and why it is structural. This piece is about the instrument that resolves it — the financial mechanics, the governance design, and who builds it and why they say yes.

The previous article in this series established the diagnosis: over a trillion euros in German retirement capital — funded occupational pension institutions managing approximately €700 billion in directly investable assets, and a life insurance sector with €1,015 billion in total investment assets — is structurally exposed to a US political risk it cannot hedge. Within their sovereign bond portfolios, US Treasuries have become the single largest foreign holding, at 11.6% of sovereign bond allocations in institutional Spezialfonds. European sovereign alternatives do not yet exist at the scale or credit quality these institutions require. That argument stands on its own. This piece does not re-argue it. What it does instead is answer the question serious readers will have asked immediately: how, exactly, would this instrument work?

The answer requires being precise about three things that the broader policy debate consistently conflates: the vehicle through which the bonds would be issued, the governance architecture that determines their credit quality, and the regulatory pathway that allows German pension managers to hold them in the first place. Get any one of these wrong and the instrument either doesn't exist, doesn't price competitively, or doesn't reach the institutional investors it must reach.

The Instrument: Why the ESM Route Is Dead

The European Stability Mechanism is the obvious vehicle. It already has legal personality, carries ECB repo eligibility, and has a decade of market history behind its bonds. The problem is its mandate: the ESM was created for crisis management, not investment financing. Expanding it to cover defence requires amending the ESM Treaty — which requires ratification by all 20 eurozone member states. Any related EU Council legislation enabling the expansion would additionally require unanimity among all 27 member states, where Hungary has exercised veto threats on exactly this kind of initiative before. The ESM route is not merely slow; in the current political environment it is realistically unavailable.

The viable alternative is a Special Purpose Vehicle — a coalition of willing states acting outside the ESM framework, structured from the outset as a capital markets instrument rather than a crisis mechanism. The Warsaw-Berlin-Paris axis is the political core, with the Netherlands, Austria, and Finland as credible additional founding members. This is not a European Union instrument in the formal sense; it does not require unanimous consent from 27 member states. It requires willing counterparties and a governance design rigorous enough to earn investment-grade treatment from rating agencies and ECB repo eligibility from the outset.

That governance design is not optional. It is constitutionally load-bearing.

The Rating Problem

The German Constitutional Court's record on European financial instruments is consistent: unlimited joint liability dies in Karlsruhe. The ESM, the OMT programme, the NGEU — in each case the BVerfG allowed the instrument to survive on the condition that German exposure is capped, ring-fenced, and subject to Bundestag approval per tranche. That constraint is not a political obstacle to be overcome by better negotiation. It is a structural feature of German participation in any joint European financing instrument, and any governance design that ignores it will be struck down.

The Karlsruhe constraint matters directly for the rating, which is where the instrument's credibility with investors is actually determined. The question is what guarantor structure produces a rating sufficient for Solvency II classification — the threshold German pension managers and insurers require to hold it at scale. Under Solvency II, AAA carries zero spread risk capital charge. AA carries materially lower capital requirements than unrated alternatives. The first-issuance rating is a starting point, not a ceiling.

Here the France downgrade is directly relevant. Fitch downgraded France from AA- to A+ in September 2025. S&P followed in October 2025, citing political instability and a deficit trajectory heading toward 121% of GDP by 2028. Moody's placed France on negative outlook. In a guarantor pool, each member's credit quality contributes to the collective yield investors demand on the instrument — France's A+ raises the spread over what a Germany-only instrument would achieve. The governance design has to account for this.

The solution is a structure built around what each partner actually brings, rather than a flat pro-rata guarantee that blends the credits indiscriminately. Germany's AAA — affirmed by S&P, Moody's, Fitch, and Scope as of 2025, stable outlook — provides the financial credibility the instrument's rating depends on. The axis section below explains what France and Poland contribute in return. France's downgrade raises the blended spread the instrument pays — it does not affect the instrument's own credit classification, which flows from the financial governance structure. That distinction matters for pricing; it does not determine whether the deal is possible.

What the Instrument Should Yield

Any serious financial reader of this piece will want to know whether the instrument can compete on yield — and will not accept an asserted range without derivation. Here is the derivation.

European supranational bonds are priced against mid-swaps — the standard benchmark for non-sovereign fixed income — not directly against German Bunds. This is how the EIB, ESM, and NGEU bonds are all structured, reflecting hedging practice and market convention. The spread over mid-swaps is then converted into a Bund-equivalent yield for comparison. In October 2024, EIB and ESM 10-year bonds traded at approximately 40–50 basis points over equivalent Bunds. NGEU bonds — the EU's pandemic recovery instrument and the closest existing precedent — traded at approximately 50–60 basis points over Bunds, wider due to the instrument's shorter track record and lower secondary market liquidity relative to EIB and ESM paper.

A new defence SPV would price wider than NGEU at launch: the instrument is untested, the secondary market does not yet exist, and investors will demand a new-issue premium for an unfamiliar credit. A realistic first-issuance spread is 60–80 basis points over Bunds. Applied to the prevailing 10-year Bund yield at time of issuance, this gives the instrument's yield for a 10-year tranche. Extending to 20-year maturities — which pension funds structurally prefer for liability matching — adds approximately 30–50 basis points of term premium.

EIB bonds — which carry an explicit German state guarantee — trade tighter than France. There is no structural reason a well-designed defence SPV cannot converge toward EIB pricing over 5–7 years as secondary market liquidity deepens and institutional holders treat it as a permanent asset class. The spread over comparable Bunds is the cost of being new. It is not the cost of being European.

The Demand Side: German Pension Capital as Anchor Investor

German funded occupational pension institutions manage approximately €700 billion in directly investable assets, and a life insurance sector holds a further €1,015 billion in total investment assets. Within those portfolios, US Treasuries already represent 11.6% of sovereign bond holdings in institutional Spezialfonds — the largest foreign sovereign position, ahead of French and Italian paper (Universal Investment, end-2024).

These institutions are already holding dollar-denominated sovereign paper at scale. They are doing so not because US Treasuries are strategically preferred, but because no qualifying European alternative exists at comparable volume and credit quality. A defence SPV that achieves ECB repo eligibility and Solvency II-compatible credit classification — the two gates discussed above — would be the first instrument that structurally competes for that allocation. The question for pension managers is not whether to hold safe sovereign paper. They are required to. The question is whether there is a European alternative worth holding.

The BaFin regulatory pathway is the operational constraint on how fast reallocation can happen. Solvency II investment restrictions, the Anlageverordnung for institutions outside the full Solvency II regime, and internal ALM targets all govern the pace. Reallocation would not happen overnight regardless of instrument quality. But the constraint is pace, not direction. A qualifying instrument creates the option. Regulatory guidance from BaFin confirming asset class treatment would activate reallocation at scale.

The Warsaw-Berlin-Paris Axis

Strip the strategic doctrine and look at the material interests. Every actor in this structure has the same underlying problem: they need capital, industrial capacity, and jobs. A European defence bond instrument delivers all three. The political packaging differs by country. The economic logic is identical.

Germany holds financial governance — not by political choice, but because the structural requirements of the instrument all point to Karlsruhe constraints, AAA-anchor status, and pension fund manager demands. Germany does not need to want governance primacy. The market, the courts, and the regulatory framework require it. In return: a AAA-rated instrument that redirects German pension capital toward European sovereign assets rather than US Treasuries; the industrial scaling of Rheinmetall and the German defence industrial base toward genuine tier-1 status; and the political credibility to lead a European financial architecture rather than merely participate in one. Separately, the March 2025 debt-brake reform creates fiscal space for domestic infrastructure investment — that is German state borrowing, not SPV proceeds, but both work in the same political direction.

France gets what it actually needs — and what it would lose in any other arrangement. France does not want to control the financial spreadsheet. France wants to lead what gets built and how it is used. Industrial leadership of FCAS (if the programme survives its current existential crisis), Naval Group, MBDA, and nuclear doctrine primacy. Strategic and operational leadership of the European defence posture. The separation of financial governance (Germany) from strategic and industrial governance (France-led) is not a face-saving formula. It is the only arrangement under which France's core interest — maintaining its position as Europe's leading military power — is preserved while Germany's core interest — constitutional and financial soundness — is also preserved.

FCAS is directly relevant here, and not as a success story. As of March 2026, the programme is at an existential decision point: Phase 2 has been postponed repeatedly, the demonstrator originally expected for 2027 has not been started, and the final decision on whether the programme continues has been kicked to mid-April after a Merz-Macron conversation on 18 March produced no resolution. The Dassault-Airbus workshare dispute — Dassault demanding up to 80% of the fighter jet development, Germany resisting — is not a negotiating posture. It is a structural conflict about whether France controls the most strategically significant asset in the programme. This is precisely why the two-track separation in the bond structure — financial governance separate from strategic/industrial governance, with both tracks explicitly governed and neither subordinated to the other — is a design requirement, not a design preference. The alternative is to reproduce the FCAS dynamic inside a financial instrument, which would be fatal.

Poland has been the most serious conventional military investor in NATO — 4.3% of GDP in 2025, the highest in the alliance, with projections of 4.7–4.8% for 2026. Poland's projected tank fleet by 2030 — over 950 modern main battle tanks across K2, Abrams, and Leopard platforms — exceeds the combined total of Germany, France, the UK, and Italy. Poland's conventional posture is what makes the eastern flank real rather than aspirational. Warsaw has earned the right to be at the table where this instrument is designed — not as the alliance's eastern edge, but as one of its founding pillars.

Poland's non-eurozone status is resolved by precedent, not by workaround. Poland is a shareholder-member of the EIB — the institution that issues AAA-rated euro-denominated bonds with capital guarantees from all 27 EU member states, including non-euro members. Poland already carries contingent liability on EIB paper. Its guarantee contribution to a defence SPV is denominated in euros; currency risk falls on Poland as guarantor, not on bondholders. The EIB is the proof of concept that non-eurozone EU members can underwrite euro-denominated supranational instruments without impairing the instrument's regulatory treatment. Poland participates in both the financial and the strategic governance tracks.

The Transatlantic Statement

This piece does not seek American approval for what Europe is doing. It does not need to.

The structural argument is not about Donald Trump or any specific administration. It is about a dependency that Europe allowed to persist for two decades after it stopped being strategically rational. The window that exists now — created by an American administration that has explicitly de-emphasised European security commitments — is real, but the case for European defence financing capability predates Trump and will remain valid after him.

The weaponisation framing — that European institutions should sell US Treasuries as a geopolitical signal or financial lever — is analytically wrong, and worth naming as such. Serious analysis, including the ECB's own Financial Stability Review, makes clear that this would be self-destructive. European banks and pension funds depend on Treasuries as collateral and high-quality liquid assets; a coordinated sell-off would tighten dollar liquidity globally and hurt Europe first. That is not the argument being made here.

The correct framing, which the ECB itself has articulated, is different: Europe should build an alternative safe asset market, which will naturally reduce the marginal growth in European demand for US Treasuries over time. In November 2025, the ECB's Financial Stability Review explicitly called for "immediate and decisive implementation of policies associated with the savings and investments union and the capital markets union" to foster exactly this alternative. This is not European aggression toward the US financial system. It is rational portfolio management responding to changed political conditions.

For the American foreign policy establishment — the Pentagon, the State Department, serious think tanks — the relevant observation is this: a Europe capable of financing and fielding its own conventional deterrent is a better long-term partner in a world where China is the dominant strategic challenge.

Three realist analysts make the structural case, from different angles. Henry Kissinger's framework of multipolar stability held that capable regional partners reduce the burden on the leading power and anchor regional order. Zbigniew Brzezinski's The Grand Chessboard (1997) assumed a Europe too fragmented to act independently; that assumption is being revised by events, not by European ideology. Barry Posen's Restraint (2014) argues explicitly that European strategic autonomy serves long-term American interests by freeing US capacity for great-power competition elsewhere — precisely the logic that burden-sharing advocates in Washington invoke without following through to its conclusion.

The same structural argument appears in John Mearsheimer's work, and it is correct. It is not cited here because Mearsheimer's pre-war arguments on NATO enlargement as a cause of the Ukraine war make him a liability among this article's primary audience — Central and Eastern European policymakers and German conservatives for whom that position is neither academic nor settled. The structural logic stands without him.

Europe is building what it should have built twenty years ago. The alliance continues — as a partnership between capable equals, not as a protection arrangement. That is not a threat. It is the description of a more durable relationship.

The Obstacles That Require Steamrolling

The ECB credentialing process is more tractable than often assumed. There are two distinct processes, frequently conflated: ECB repo eligibility (using the instrument as collateral in ECB credit operations) and ECB asset purchase programme inclusion (the ECB buying the instrument outright). Only the first matters for pension fund regulatory treatment. Repo eligibility does not require Governing Council approval for individual instruments. It follows an NCB assessment process against standing criteria in Guideline ECB/2014/60: the instrument must be listed on a regulated market, meet the credit quality threshold (CQS 1–2, meaning AAA/AA equivalent from two recognised agencies), and satisfy structural requirements on form and settlement. Once an NCB confirms eligibility, the instrument is added to the daily-updated list of eligible marketable assets. A well-structured SPV meeting the credit criteria from issuance could achieve repo eligibility through this process within weeks or months — not subject to a Governing Council political veto. The Governing Council sets the framework; it does not vote on individual instruments. What does require Governing Council approval is inclusion in asset purchase programmes — a higher bar, and not required for the pension fund argument.

Karlsruhe is a constraint, not a veto. The BVerfG's track record on European financial instruments is entirely manageable as long as the instrument is designed with the constitutional requirements built in from the start rather than retrofitted after political negotiation. Capped guarantees, per-tranche Bundestag approval, ring-fenced liability — these are not concessions to German politics. They are the governance architecture that makes the instrument sound.

German coalition politics has shifted. The March 2025 debt-brake amendment — which carved out defence spending and infrastructure from the constitutional deficit limits — was not a one-off. It represents a genuine reorientation of the CDU/CSU position on European defence financing. The Merz government's fiscal loosening for defence, averaging 3.6% of GDP deficit for 2026–2030 per Scope Ratings, creates the political space that was structurally unavailable under Scholz. The obstacle is no longer the political will — it is the institutional design.

Rating agencies are obstacles of the solvable kind — and the governance template they require already exists.

The NGEU's Recovery and Resilience Facility established the conditionality model: ring-fenced eligible uses (national recovery plans), milestone-triggered disbursements, and Commission oversight authority to withhold tranches. Rating agencies awarded the EU its AAA rating on the basis of that structure. The defence SPV needs an equivalent. It now has one.

The Security Action for Europe — SAFE, adopted by the Council in May 2025 and already disbursing to 16 member states — is operationally that equivalent. Participating states submit National Defence Investment Plans defining eligible procurement. Pre-financing of up to 15% is released after Commission assessment. Subsequent tranches require biannual milestone reporting, and the Commission holds authority to withhold or recover funds in cases of budget substitution. The eligible categories — air and missile defence, ground combat systems, strategic enablers, drones, cyber — are defined in the regulation, not left to political discretion.

SAFE is a loan instrument: the EU borrows on capital markets and passes the funds to member states. The defence SPV is a capital markets instrument: pension funds hold the bonds directly. The governance architecture is portable between the two. The SPV raises capital via bond issuance; disburses proceeds to participating states against their Commission-assessed NDIPs; releases subsequent tranches against verified procurement milestones — contracts signed, equipment ordered, deliveries confirmed. A supervisory board, structured analogously to the EIB's board of governors — which includes all 27 EU member states, including non-euro members — holds tranche suspension authority and reports to participating state parliaments.

This conditionality design is stronger than the RRF in one key respect. The RRF tied disbursements to policy reforms — inherently political, contested, subject to Commission-member state negotiation. Procurement milestones tie disbursements to verifiable physical outcomes against tangible assets. Rating agencies can model that. Karlsruhe can accept it: liability is capped per tranche, conditioned on milestone delivery, and subject to Bundestag approval at each tranche decision. The governance question is not unanswered. The template is operational.

The Window — And What Has Already Started

The standard framing treats this as a future scenario. It is already beginning.

The Euronext European Defence Bond label launched in July 2025. BPCE inaugurated the market in August 2025 with a €750 million issuance that was nearly four times oversubscribed. Bpifrance issued €1 billion in November 2025, with an order book above €3.8 billion. Luxembourg published its sovereign Defence Bond Framework in October 2025 and announced inaugural sovereign issuance in January 2026. The SAFE instrument — €150 billion in EU loans to member states for defence — is operational. National escape clauses under the Stability and Growth Pact have been activated for 15 member states, with Germany's own clause activated in October 2025, specifically to accommodate defence and infrastructure spending at scale.

This is not a future scenario. The early market infrastructure exists.

What does not yet exist is a sovereign-quality, supranational SPV instrument — ECB repo-eligible and Solvency II-compatible — that German Pensionskassen and life insurers can hold at scale. The Euronext label covers corporate and financial institution paper — useful, but not a substitute for the sovereign-quality instrument. That is the gap this instrument would fill.

The realistic sequencing from here:

2026–2027: SPV structure negotiated within the Warsaw-Berlin-Paris axis. Conditionality and governance framework agreed. Inaugural supranational issuance — small by design, to establish the yield curve. NCB eligibility assessment completed; instrument achieves ECB repo eligibility. FCAS impasse either resolved (defining the strategic governance track) or formally restructured (Franco-Spanish fighter, German-led Combat Cloud — separating the tracks that should be separated in any case).

2027–2028: BaFin regulatory guidance confirming asset class treatment for German pension purposes. Secondary market develops. Spread tightens from first-issuance 60–80bps over Bunds toward EIB-equivalent levels as the instrument proves its governance stability. First major pension fund reallocation into the instrument, visible in Universal Investment Spezialfonds data.

2028–2030: Industrial programmes funded. First capabilities delivered that would not have existed under the previous national procurement model. The instrument achieves sufficient scale to be a meaningful allocation across German and broader European institutional portfolios. European Defence Readiness Roadmap 2030 targets — 40% joint procurement, 55% from European suppliers — become financially credible rather than aspirational.

2030–2033: Yield curve established across multiple maturities. Instrument treated as a standard European safe asset alongside Bunds and EIB paper. The original question — where is retirement capital safe? — has a European answer that didn't exist five years earlier.

The window is not closing. But it is not staying open indefinitely either. The political conditions that make this possible — German defence spending reform, French willingness to accept financial governance separation, Polish conventional credibility, an American administration that has forced the question — are specific to this period. A future US administration with more traditional alliance instincts would reduce the urgency without eliminating the structural logic. Build it now, while the logic and the political conditions coincide.

What to watch for — three leading indicators that Europe is actually moving, not just talking:

1. First issuance of a sovereign-quality, supranational defence SPV bond by any European coalition. The Euronext-label corporate and financial institution paper already exists — that is not this. The signal is an instrument with the governance architecture, credit rating, and ECB repo eligibility that institutional investors require. That is what does not yet exist.
2. BaFin regulatory guidance acknowledging a new defence bond asset class. Without this, pension managers cannot reallocate at scale regardless of instrument quality.
3. A Franco-German joint procurement announcement that excludes US prime contractors. This is the industrial governance track made concrete. It is also the indicator that the FCAS impasse has been resolved in a direction that preserves European industrial leadership rather than outsourcing it to Lockheed Martin.

This is the answer to the trillion-euro question. The instrument is buildable, the template exists, and the window is open. Whether it stays open long enough depends on decisions being made now — in Berlin, Paris, and Warsaw — about whether Europe is serious or just loud. No permission from Washington required.

---

Published: 2026-03-30