Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Ring-Fenced Funds and Restricted Own Funds under Solvency II

Abstract. Ring-fenced funds constrain the prudential value of insurance capital by restricting the losses that designated own-fund items can absorb and the diversification recognized between segregated portfolios and the remainder of an undertaking. The Solvency II framework gives this issue a precise legal and quantitative form: Commission Delegated Regulation (EU) 2015/35 distinguishes the identification of restricted own funds, the adjustment of the reconciliation reserve, and the calculation of notional Solvency Capital Requirements (SCRs). These tests operate on different objects. The existence of an economically dedicated asset portfolio alone does not settle whether its surplus qualifies as a ring-fenced own-fund item; a fund may carry insurance liabilities and regulatory capital requirements even where its assets exceed those liabilities; and an apparently high accounting surplus need not supply usable solvency resources elsewhere. This monograph interprets the EU legal framework as it stands on November 15, 2026, develops a reproducible illustrative balance-sheet calculation, and evaluates legal, actuarial, supervisory, and institutional sources. It also examines a consequential impending reform: Directive (EU) 2025/2, Delegated Regulation (EU) 2026/269, and the July 2026 revised EIOPA Guidelines change the prudential treatment of matching-adjustment portfolios from January 30, 2027. The analysis distinguishes loss-absorption capacity, own-funds eligibility, recognition of diversification, and group transferability; clarifies the interaction with tier restrictions and deferred tax assets; and identifies which assumptions must be resolved before translating these calculations into a real insurer's solvency position. Its original contribution is an integrated, source-traceable interpretation and numerical sensitivity analysis of existing regulatory mechanisms, without asserting a new theory or new regulatory doctrine.

Keywords: Solvency II; ring-fenced funds; restricted own funds; reconciliation reserve; notional Solvency Capital Requirement; capital fungibility; matching adjustment; own-funds tiers; loss absorption; insurance regulation; insurance capital; diversification; deferred tax assets; insurance groups; Directive (EU) 2025/2; Delegated Regulation (EU) 2026/269.

Introduction

Solvency regulation places economic assets and insurance liabilities within an institutional framework for measuring capital. Financial resources acquire prudential value through their capacity to absorb the losses against which the regulatory requirement is constructed. A claim to assets is therefore insufficient to establish that an equal amount is available to every creditor, policyholder group, or business segment. Contractual commitments, law, supervisory approval, and the purpose of a protected fund may channel the resource toward particular liabilities. The associated capital requirement may then have to be measured separately from risk elsewhere in the undertaking.

The Solvency II Directive (2009) establishes the European Union insurance and reinsurance prudential architecture. The Solvency II Delegated Regulation (2015) operationalizes important aspects of valuation, own funds, and the standard formula. Within that architecture, a ring-fenced fund is best analyzed through the loss-absorption restrictions and notional capital computations prescribed in the applicable rules, rather than solely through its label in financial statements or internal asset-management records. The EIOPA 2015 Guidelines provide supervisory interpretation for identifying funds and determining their associated assets, liabilities, and capital charges.

Three related questions organize the investigation. First, what legal and economic circumstances restrict the usable own funds of an undertaking? Second, how do those restrictions modify both the numerator and denominator of a solvency coverage ratio? Third, how should a researcher interpret the transition from the currently applicable matching-adjustment treatment to the amended rules taking effect January 30, 2027? The inquiry is particularly relevant because an insurer can experience an apparent capital improvement through accounting valuation, favorable dependence estimates, or operational restructuring without obtaining a corresponding improvement in transferable, loss-absorbing financial resources.

The method is documentary and analytical. Statutory and delegated legal instruments are interpreted in their applicable time versions, with EIOPA guidelines and questions and answers examined as supervisory materials of a different legal character. Published actuarial scholarship informs the discussion of correlation, aggregation, and intra-group allocation. A self-contained hypothetical numerical example demonstrates the direction and scale of the mechanisms, and sensitivity cases expose assumptions that alter its result. The work does not estimate the distribution of actual insurance losses, test an empirical population of insurers, or substitute for jurisdiction-specific supervisory determinations.

The governing dates are material. This manuscript describes the rules applicable on November 15, 2026. The 2025 amending Directive and the 2026 amending Delegated Regulation are separately examined as the forthcoming regulatory regime. The EIOPA July 2026 revised report expressly states that its revised guidelines apply from January 30, 2027 and that the 2015 guidelines apply until that date. An enactment already in force but not yet applicable to the relevant capital calculation cannot be silently treated as the current operational rule.

Ring-Fenced Fund Structures

Institutional origin and the prudential object

The term ring-fenced fund refers to an arrangement in which specified assets, insurance obligations, financial surplus, or associated own-fund resources have legally or contractually constrained uses. The relevant restrictions may arise from statute, an enforceable policyholder entitlement, trust arrangements, approved product terms, or another source affecting the undertaking's ability to direct resources to losses elsewhere. An arrangement can take the form of a distinguishable portfolio inside a single legal entity. Its presence should accordingly be tested even where no separate company, trust account, or bank account has been created.

Article 80 of the Delegated Regulation supplies the decisive prudential formulation. It addresses own-fund items within a fund whose ability to absorb going-concern losses is reduced because those items can cover only a specified portion of insurance contracts, specified policyholders or beneficiaries, or particular risks or liabilities. The Article 80 restriction test is functional: it concerns the ability of an item to absorb loss across the undertaking, not the superficial separateness of its assets. Article 80(2) also expressly excludes the value of future transfers attributable to shareholders from the restricted own-fund items under this provision, an important qualification when calculating a with-profits estate.

An asset portfolio used to back insurance liabilities must ordinarily meet several different tests. A manager may hold assets in a distinct strategy for duration matching or collateral management. A statute may impose rules on covered assets for policyholder protection. A product may assign a proportion of investment returns to specified participants. Some arrangements restrict the asset strategy while leaving shareholder capital available for wider loss absorption; others restrict the shareholders' claim to the accumulated estate itself. Prudential identification requires tracing which right or obligation carries the restriction, which losses the money can pay, and how the claim behaves under adverse circumstances.

A ring fence is consequently a relation among assets, liabilities, stakeholders, and enforceable allocation rules. The fund's geographical location and the fact that it appears as a separate reporting line do not alone determine this relation. Nor does a general duty to maintain adequate insurance assets prove that the residual surplus is permanently committed to one contractual group. The supervisory question concerns the usable own-fund item and the specific constraint on its loss-absorbing function.

Functional identification and enforceable constraints

The 2015 Guidelines on ring-fenced funds guide undertakings through identification, asset and liability boundaries, notional SCR measurement, and interactions with the overall capital calculation. They direct attention to differences among Member State laws and insurance product regimes. Their importance is methodological: comparability of quantitative capital ratios depends on recognizing substantively similar restrictions despite different national legal techniques.

A sound assessment begins with the instrument creating the restriction. Where policy conditions allocate profits to a cohort, one must examine the timing and enforceability of that entitlement, the legal status of unallocated amounts, the treatment of adverse experience, and the scope for transfers to shareholders. A purported restriction that can be unilaterally waived is analytically different from a statutory obligation that remains binding even under stress. Supervisory acceptance of an interpretation may matter where the law or regulatory process makes the classification contingent on approval or documented evidence.

Identification then proceeds through an accounting and actuarial inventory. The researcher associates eligible assets, technical provisions, other liabilities, any expected transfers, and the resulting own-fund items with the protected business. If a single asset supports more than one segment, its allocation requires an explicit rule. If a loss scenario affects different accounts through taxes, options, or guarantees, the assumed attribution must be capable of reconciliation to the undertaking-wide balance sheet. Discretionary benefits and management actions require separate attention: an insurer's formal ability to reduce future bonuses does not automatically release already accumulated surplus for losses in unrelated policyholder groups.

A practical distinction concerns the economic meaning of transfers. The ability to sell a security inside a fund provides internal liquidity, but sale proceeds may remain legally committed to the fund. Conversely, a regulator-approved payment from a surplus account to shareholders may create an actual resource outside the fence only when the legal and valuation conditions are met. Liquidity, transferability, and loss absorption describe different properties. Their combined assessment, rather than reliance on a single indicator, determines the prudential conclusion.

With-profits, surplus funds, trusts, and excluded arrangements

Participating and with-profits life insurance can accumulate amounts whose ultimate distribution depends on investment returns, policyholder participation rules, and supervisory protections. Its accounting presentation may divide technical provisions, future discretionary benefits, shareholders' interest, and surplus funds. The legal constraints governing these components determine whether an excess value is capable of supporting losses beyond the relevant cohort. The EIOPA Guidelines for fund identification supply the starting point; the documents of the actual fund supply the case-specific evidence.

Article 91 of the Directive distinguishes surplus funds made up of accumulated profits that have not been made available for distribution to policyholders and beneficiaries, subject to the directive's treatment. This surplus-funds provision is related to, but conceptually distinct from, the ring-fencing test. Surplus funds are not automatically identified with restricted own-fund items. A surplus may be available under specified rules, may contain amounts already counted elsewhere in the own-funds computation, and may be partly protected for one class of policyholders. The relevant statutory and product structure decides the answer.

Trusts and similar legal structures can strengthen an allocation by identifying beneficiaries and restricting a trustee's powers. Yet the word trust cannot substitute for analysis of enforceability, beneficial ownership, insolvency ranking, and the available management actions. A legally segregated asset pool might chiefly protect policyholders in winding-up, whereas Article 80 concerns limitations on absorbing losses while the undertaking continues operating. The same pool may perform both functions, but the reasoning must identify both.

Likewise, unit-linked liabilities commonly follow designated assets and investment performance; the mere practice of matching those assets with corresponding obligations does not by itself demonstrate that residual own funds are restricted. The appropriate test asks whether there are own-fund items whose loss absorption is constrained in a manner contemplated by Article 80. The distinction is especially important in mixed insurers that combine unit-linked contracts, guaranteed annuities, and participating contracts. The number of product categories is not the number of prudential ring fences: a single contract class could give rise to several enforceable protected funds, while multiple product classes could share one unrestricted capital estate.

The supervisory interpretation of nuanced participating-fund arrangements remains fact dependent. A June 2026 question to EIOPA about national allocation rules for with-profits business was marked rejected, with reference to existing guidance rather than a new universal ruling; its Q&A 3591 record should therefore be read as a documented interpretive question, not as independent binding clarification. It illustrates the need to distinguish allocated individual bonus entitlements in technical provisions from associated surplus or own-fund restrictions. The classification of an insurer's fund cannot be inferred from the procedural status of one Q&A.

Scope, allocation, and documentation

A rigorous fund register records the source of the restriction, beneficiaries, assets, insurance and reinsurance obligations, other liabilities, possible shareholder transfers, asset encumbrances, capital instruments, and the conditions for release. Each item has a valuation date and a link to the relevant legal or contractual evidence. The register is reconciled to the insurer's full Solvency II balance sheet so that a resource or obligation is not lost or duplicated merely because its internal reporting location has changed.

For each fund, the researcher should also specify the applicable calculation basis: standard formula or internal model, scenario set, risk mitigants, treatment of future discretionary benefits, tax assumptions, operational risk, and rules for aggregation with the remainder. These are part of the inference from observed data to regulatory capital; a fund boundary without an agreed modeling boundary leaves the capital computation underdetermined. Modifications of the boundary may change both the estimated solvency requirement and the eligible capital that can cover it.

Documentation should preserve evidence of the actual legal status of constraints through time. Where a rule changes, a transfer is authorized, or a product portfolio closes to new business, the valuation must be evaluated on the new basis at the appropriate date. A year-end diagram of balance-sheet segmentation is consequently less informative than a versioned schedule of restrictions and approved transfers. The central outcome of this chapter is that identifying a ring-fenced fund is a legal-and-actuarial determination of the reach of own-fund items.

Restrictions on Loss Absorption

Transferability and going-concern capital

The Solvency II prudential objective is to ensure that resources credited toward regulatory capital can perform their intended loss-absorption function. Article 93 of the Directive uses permanent availability and subordination, alongside further characteristics, to distinguish classes of own funds. The own-funds classification criteria are relevant because an item that appears as surplus in a valuation can fail a different test governing its capacity, duration, or availability.

A legally restricted resource may be economically valuable and fully sufficient to protect the policyholders of its own fund. Its contribution to general solvency differs because losses in the rest of the firm cannot necessarily be met by that resource. Ring-fenced own funds can therefore have real protective value inside the boundary even when part of their surplus is deducted from the undertaking's freely usable own funds. The adjustment changes the regulatory recognition of where capital can respond to losses.

Suppose a participating fund reports an excess of assets over related liabilities of 150 monetary units. Its fund-specific notional SCR is 100. If the 150 consists of restricted own-fund items that cannot absorb losses outside the fund, crediting all 150 toward enterprise-wide losses would treat the excess 50 as accessible beyond the level needed to cover its own quantified requirement. The ring-fencing rules address this problem through a reduction of the reconciliation reserve. The precise outcome requires the legal identification, valuation, and classification assumptions stated in the next section.

These prudential constraints must be evaluated under going-concern conditions. The transfer of residual assets following complete run-off or liquidation may be possible while no such transfer is permissible at the reporting date. A theoretical eventual shareholder entitlement is not an unrestricted current claim. This temporal distinction is central to both the Article 80 identification test and the purpose of annual capital-adequacy assessments.

Reconciliation reserve and restricted surplus

The reconciliation reserve is a component of the Solvency II own-funds construction. Its derivation accounts for the excess of assets over liabilities and other items specified in Article 70. The Article 70 reconciliation rules include the ring-fenced adjustment among the deductions. The underlying logic is to avoid allowing the same surplus to appear as freely available capital when its practical ability to meet losses is restricted.

Article 81 requires the undertaking to compare the restricted own-fund items and the fund's notional SCR. In the simplest one-fund case, if all designated items are otherwise eligible and restricted, the conceptual adjustment is the positive excess of those items over the notional SCR. The Article 81 comparison consequently limits the contribution of such an item to the fund-specific capital need. The calculation is subject to the actual composition, classifications, and applicable regulation, and should not be treated as a universal shortcut for every kind of restricted capital instrument.

Let R denote restricted own-fund items in a material fund and K its notional SCR. Under the simplifying assumptions just stated, the illustrative adjustment is

D=max⁡(R−K,0).D = \max(R-K,0).

Let U denote otherwise available unrestricted own funds before tier limits. Then the illustrated pre-tier resource is

F=U+R−D=U+min⁡(R,K).F = U+R-D = U+\min(R,K).

The formula is an analytical rendering of the comparison, not a substitute for the delegated regulation's full valuation, deductions, tiers, or source-allocation rules. In particular, not every accounting asset appears in R, and not every balance-sheet surplus is an unrestricted own-fund item. Where separately classified capital, shareholder transfers, or other adjustments exist, the actual regulatory computation has to preserve their distinct treatment.

An illuminating property follows immediately. Holding U and K fixed, an additional unit of restricted surplus increases F while R is below K; beyond K, the same additional unit produces an equal additional adjustment and no further increase in F. In calculus notation, the simplified marginal contribution changes from one to zero at the threshold K. The result is a direct interpretation of the existing Article 81 rule. It is not a newly proposed capital model: it exposes the consequence of a legal restriction using ordinary arithmetic.

Materiality, exceptions, and possible double counting

For a ring-fenced fund whose assets, liabilities, and risks are immaterial, Article 81(2) permits a different approach: deduct the total restricted own-fund items rather than undertaking a complete notional SCR computation. The immaterial-fund alternative simplifies the task but can be more conservative than the material-fund comparison. Materiality must be assessed according to the actual regulatory context and the fund's economic features. The mere administrative inconvenience of calculating the notional capital requirement is not a materiality criterion.

The central double-counting risk has several sources. A liability valuation may already reflect future discretionary benefits, so treating the same amounts again as a fully available shareholder surplus overstates own funds. A tax asset may be economically unusable outside the ring fence and require restriction in its valuation before being tested for tier eligibility. An asset allocation may assign one investment twice across overlapping administrative portfolios. Finally, a net risk reduction associated with intra-fund diversification must not be counted a second time as unrestricted inter-fund diversification.

These possibilities have different accounting and legal origins. A robust review traces each restriction into the relevant calculation exactly once, with explicit attribution to the article that requires it. An overzealous adjustment can be as misleading as an under-adjustment. The response to overlap is a line-by-line reconciliation between fund-specific resources, the undertaking's balance sheet, notional SCRs, and eligible own funds rather than a blanket additional haircut.

Legal and economic capital availability

Capital availability has multiple dimensions. Legally transferable resources can be moved according to enforceable rules; operationally transferable resources can be mobilized within the relevant time; economically available resources retain value in the adverse scenarios considered; and regulatory eligible resources satisfy the binding classification and quantitative limits. A payment may be permitted by corporate law yet require supervisory consent, may settle too late for a sudden claims event, or may reduce the value of another contract-specific guarantee. Each dimension affects the inference that a particular capital item can absorb the loss of concern.

The ring-fencing adjustment primarily concerns a specific legal and regulatory loss-absorption constraint. Its existence does not, on its own, establish that a fund's assets are illiquid or low quality. A liquid government bond retained within a protected estate may remain unsuitable for losses of a separate estate, while a relatively illiquid but transferable asset may still be counted subject to its proper valuation. Conflating liquidity and fungibility leads to incorrect prudential interpretations.

The same point applies to insurance-group capital. The ability of a subsidiary to satisfy its solo SCR does not guarantee that its surplus can be transferred to a parent or other subsidiaries. The group-level requirements therefore need a separate investigation, developed below. It is methodologically useful to preserve two questions throughout the work: how much capital can support the liabilities of this fund, and how much can legitimately support losses elsewhere in the insurer or group?

Notional Solvency Capital

Portfolio-level and undertaking-level calculations

Capital can be aggregated in different ways depending on which losses and risk pools are permitted to support one another. For the rules applicable on the research cut-off date, Article 216 directs an adjustment to the SCR calculation for ring-fenced funds and approved matching-adjustment portfolios, subject to the relevant occupational-retirement exception. The Article 216 scope must be read together with the detailed Article 217 method.

Under Article 217, the undertaking calculates a notional SCR for each ring-fenced fund, for each matching-adjustment portfolio caught by the current rule, and for the remaining part of the undertaking as if they were separate undertakings. Their notionals are then summed for the regulatory SCR under that method. The Article 217 aggregation rule expressly assumes no diversification between these components. It simultaneously allows the specified intra-component calculations and sets procedures for stress scenarios and the allocation of scenario impacts. The operation is therefore more structured than simply summing arbitrary stand-alone risk-module figures.

In an actual standard-formula calculation, the constituent modules can include market, counterparty-default, life, health, and non-life underwriting risks, as appropriate, together with operational risk and permitted adjustments. The exact construction depends on the regulation's module-specific requirements and the risk exposures allocated to a fund. Internal-model users calculate their fund notionals with the relevant approved model on the basis stipulated by Article 81. The model's representation of restrictions and diversification must be aligned with the legal constraints on own-fund items, rather than merely reproducing the insurer's preferred economic capital decomposition.

The existing 2015 EIOPA Guidelines treat boundary construction and the notional requirement as closely related steps. For a hypothetical insurer with one ring-fenced fund, one might describe the total regulatory capital requirement as K + M where K is its notional fund SCR and M the SCR of the remaining undertaking. That arithmetic assumes the variables already represent the proper final notionals, including any required adjustments and interactions prescribed by the applicable provisions. It is a transparent reduction of the structure to two components; it is not an assertion that every pair of raw module capital requirements can be combined additively.

The special treatment addresses the possibility that favorable dependence between losses in different segments could make total economic losses appear diversifiable even though the resources designated for one segment cannot support the other segment's loss. If an insurer relied on this diversification in its regulatory capital but also counted restricted capital as if it could move among segments, its capital coverage could be doubly optimistic. The legal approach restricts both operations, though in analytically distinct calculations.

Reproducible stylized balance-sheet case

Consider an insurance undertaking with one material fund and a remaining unrestricted business. Monetary amounts are hypothetical units and are independent of any actual insurer. Assume that all fund own-fund items considered are restricted in the sense of Article 80, that fund-specific and remaining notional SCRs have already been determined in conformity with the chosen legal and model bases, and that no further tiers, deferred tax effects, or capital restrictions alter the example. Let:

Symbol Illustrative quantity Amount
R Restricted own-fund items in the fund 150
K Notional SCR of the fund 100
U Other available own-fund items 90
M Notional SCR of the remaining undertaking 80

Using the Article 81 comparison in its simplified form, restricted surplus above the fund's notional capital requirement is 150 − 100 = 50. The illustrated unrestricted-equivalent amount recognized after the fund adjustment is 90 + 150 − 50 = 190. By the simplified two-component aggregation of Article 217, the SCR is 100 + 80 = 180. The corresponding coverage ratio is 190 / 180 = 105.56%. Before adjusting for the restriction, the apparent capital amount would be 90 + 150 = 240 and its comparison against the same 180 SCR would give 133.33%. The 27.78-percentage-point difference is attributable entirely to the restricted-surplus adjustment in this controlled case.

The same insurer with U = 50 would have recognized own funds of 150 and a coverage ratio of 150 / 180 = 83.33%. The change from U = 50 to U = 90 demonstrates that an additional 40 units of genuinely unrestricted own funds raises regulatory coverage by 22.22 percentage points in this setup. Conversely, increasing R from 150 to 190, with K and U fixed, raises the adjustment from 50 to 90 and leaves recognized F at 190. That is the threshold effect derived earlier.

The example has an important counterintuitive consequence: a fund can possess an apparent 50-unit excess of restricted surplus over its fund-specific capital requirement while the insurer's total coverage depends decisively on unrestricted resources elsewhere. Aggregate balance-sheet strength and the ability to resolve a particular distribution of losses are different properties. A portfolio classification that fails to acknowledge the restriction would describe the wrong capacity to transfer resources under stress.

No assertion is made that the firm is legally compliant in a specific jurisdiction based on this single stylized ratio. Its numerator excludes tier-specific complications and its denominator is stipulated rather than calibrated from actual risk data. The analysis demonstrates accounting identities and the direction of regulatory effects, not supervisory approval or a forecast.

Diversification, sensitivity, and limitations

To visualize the separate diversification mechanism, suppose two hypothetical unconstrained risk portfolios have stand-alone capital measures of 100 and 80 under a common quadratic risk aggregation. Let their dependence parameter be 0.25. In that hypothetical statistical model,

Kquad=1002+802+2(0.25)(100)(80)=20,400≈142.83.K_{\mathrm{quad}}=\sqrt{100^2+80^2+2(0.25)(100)(80)}=\sqrt{20{,}400}\approx142.83.

The additive capital figure is 180, so the model suggests a 37.17-unit reduction where diversification is recognized. This is an illustrative quadratic calculation and does not reproduce the overall prescribed Solvency II SCR formula for the insurer. It is used to identify the economic significance of a restriction on cross-fund diversification.

The recognized inter-component diversification advantage in this stylized comparison varies systematically with dependence. At a correlation of −0.25, the quadratic amount is 12,400≈111.36\sqrt{12{,}400}\approx111.36; at 0 it is 16,400≈128.06\sqrt{16{,}400}\approx128.06; at 0.25 it is approximately 142.83; at 0.75 it is 28,400≈168.52\sqrt{28{,}400}\approx168.52; and at 1 it is 180. The difference between 180 and these amounts is the statistical benefit that an aggregation model would report if the components were genuinely poolable. Under the cited Article 217 ring-fencing rule, the regulatory calculation removes diversification between the relevant restricted components even if their economic losses have low correlations.

These observations lead to a precise distinction. Dependence concerns the joint distribution of losses; fungibility concerns whether a resource assigned to one loss can absorb another. Correlation between claim outcomes does not itself create the legal authority to move policyholder-protected money. An insurer that treats these concepts as equivalent may produce an internally consistent statistical model of losses but an institutionally invalid model of available loss-absorbing resources.

The actuarial literature gives a broader foundation to this reasoning. Filipović's multi-level risk aggregation study addresses the consequences of aggregation architecture; Bermúdez, Ferri, and Guillén (2013) analyze correlation sensitivity in underwriting capital. These publications motivate the sensitivity analysis of dependence assumptions. They do not determine the legal classification of a Solvency II fund, and their mathematical statements should not be transplanted into the delegated regulation without identifying the different assumptions.

An insurer's model may also face nonlinear relationships between risk exposures, guaranteed benefits, dynamic management actions, and reinsurance. Diversification can change with stress magnitude, policyholder behavior, and markets. If a fund's liabilities contain options, estimating a stand-alone SCR requires the same governing valuation and stress mechanics under which the options create losses. Allocating a single enterprise-wide stress result in proportion to premiums or asset values would generally require additional justification and could suppress genuine risk differences.

The most useful sensitivity checks therefore run along two axes. First, vary economic assumptions such as correlations, asset shocks, lapse behavior, mortality, and available tax recovery. Second, vary legal and contractual boundary assumptions: which resources are transferable, which shareholder distributions are permitted, which fund is genuinely segregated, and which approvals remain effective. A regulatorily plausible calculation should survive the translation from those assumptions to the actual fund-specific loss-absorption mechanisms.

Own-Funds Eligibility Limits

Tier classification and quantitative ceilings

Available own funds and eligible own funds are related but different quantities. The legal classification of an item reflects its substantive loss-absorbing characteristics, including permanence and subordination, and regulatory limits then determine how much of each class can cover the SCR and the Minimum Capital Requirement (MCR). Article 93's classification characteristics make clear that the economic label equity or debt is insufficient to determine a tier without examining the instrument's terms.

Under the current Article 82 quantitative limits, SCR coverage requires eligible Tier 1 own funds of at least one half of the SCR; Tier 3 eligible own funds must represent less than 15% of the SCR; and combined Tier 2 and Tier 3 eligible own funds may not exceed 50% of the SCR. MCR coverage applies a different set of limits, including at least 80% Tier 1 and no more than 20% Tier 2. The Article 82 eligibility limits also contain additional constraints for specified basic own-fund items within Tier 1. Those further qualifications must be checked for actual capital instruments.

The ring-fenced-fund adjustment should not be assumed to operate as a substitute for tier limits. A capital amount can be restricted in where it is usable and separately limited in how much of its tier can be included in the eligible total. As a result, a simple ratio of total excess assets over liabilities to total SCR overstates solvency coverage if it ignores either test. Conversely, applying an unrestricted tier haircut to an item that has already been excluded from the relevant own-funds basis risks counting the same limitation twice.

Consider a hypothetical undertaking with 100 of otherwise eligible Tier 1 resources and a separate 30 of Tier 3 deferred tax assets before limits, against a 200 SCR. Article 82 prevents the full 30 from being eligible because Tier 3 must remain below 15% of the SCR, even before any special ring-fence restriction is considered. The strict inequality is material: an illustrative amount exactly equal to 30 is not, as written, below 15%. Practical regulatory reporting and the rounding or eligibility procedures applicable to the firm determine the actual admissible number. The example illustrates the separate eligibility dimension rather than prescribing a compliant allocation.

Article 81 compares restricted own funds against a notional SCR rather than granting an unrestricted exception to the hierarchy of own-fund quality. For multi-tier fund resources, one must reconcile which items have been counted and their ability to satisfy the fund-specific requirement. A report that merely supplies R, K, and D without describing the component quality can be insufficient for the final eligible-own-funds calculation.

Deferred tax assets and reporting choices

Deferred tax assets present a special case because their value depends on the realistic future availability of taxable profits or tax liabilities against which they can be used. Ring-fenced insurance business may create legal restrictions on the profits that can support recovery. An asset that is economically meaningful in a consolidated tax assessment might have a different value when assessed for a legally protected segment. The prudential analysis begins with valuation restrictions and then proceeds to own-funds classification.

In its Q&A 2200 clarification, EIOPA notes that Article 15(3) of the delegated regulation requires the valuation of deferred tax assets to reflect legal or regulatory restrictions, including those arising from ring-fenced funds. After such restrictions have been incorporated in valuation, EIOPA describes the permissible entity-level netting of deferred tax assets and liabilities for the resulting Tier 3 position. This sequence matters: legal availability affects the value of the asset itself before the separate eligibility calculation.

The related Q&A 2513 example identifies a reporting problem when a large ring-fenced deferred tax asset increases apparent surplus and the ring-fenced restriction, only to be constrained again by the Tier 3 ceiling. EIOPA permits reporting only the fund's Tier 3 deferred tax assets needed to cover the notional SCR, with an associated adjustment to the reported restricted-own-funds deduction. Its original example concerns a 10 million-euro notional SCR and shows why mechanical recognition of the entire asset can produce an unnecessarily adverse reported result. This interpretation must be applied with the specific reporting template and legal constraints, rather than generalized into a right to recognize unrecoverable taxes.

Two implications follow. A fund-level deferred tax asset should be admitted only to the extent consistent with supported recoverability assumptions. Then its own-funds classification, ring-fenced adjustment, and reporting presentation should be reconciled so that the same constrained amount is not repeatedly penalized through unrelated line items. The calculation is sensitive to the existence of taxable profit sources, timing, group taxation arrangements, and enforceable limitations on using those profits. Each of these factual assumptions must be audited in an actual insurer's valuation.

The treatment also clarifies the boundary of pure algebra. The formula U + min(R,K) does not specify the permissible tax recovery, the tier of the items in R, or the correct template allocation. Treating such an algebraic identity as an executable regulatory calculation without the accounting reconciliation would exceed the evidence supported by the source.

Group transferability and consolidation

Insurance groups can possess abundant solo capital yet insufficient freely available group resources. Legal entity boundaries, minority interests, supervisory restrictions, and national distribution rules can constrain transfers. The Article 330 availability test directs supervisors to consider restrictions on covering losses wherever they arise in the group, legal and regulatory impediments to transfers, and whether resources can be made available within a maximum of nine months. The group assessment therefore adds a time-and-entity dimension to the solo ring-fenced-fund inquiry.

The solo rule asks whether a fund's own-fund item can absorb losses elsewhere in the same undertaking. The group rule asks whether resources attributed to a related undertaking can actually cover losses across the group. Their boundaries need not coincide. An insurer can have a ring-fenced portfolio inside one licensed company and a further transferability restriction at the subsidiary-to-parent level. Equally, a company without a qualifying Article 80 ring fence may still operate in a jurisdiction whose dividend or supervisory regime makes part of its own funds unavailable at group level.

The research literature explores the consequences of allocating capital across group entities. Asimit, Bădescu, Haberman, and Kim (2016) analyze risk allocation within insurance groups under Solvency II. The contribution relevant here is analytical: the allocation of risk and capital can differ from their aggregation. It does not imply that an insurer is legally entitled to implement transfers that its regulator or national law restricts.

In group analysis, the amount of capital that appears in a subsidiary's own-funds calculation is therefore a starting observation, not the conclusion. A robust consolidation process specifies the available fraction, any legal restriction, transfer timing, currency conversion and capital-quality requirements, and the group-level contribution of the subsidiary to regulatory risk. It records restrictions only once and separates legally unavailable resources from statistical diversification assumptions. The direction of an intra-group transfer should also be explicit: upstream dividends and downstream capital injections follow different corporate, contractual, and supervisory paths.

A useful international comparison is the Insurance Capital Standard developed by the International Association of Insurance Supervisors. The IAIS Insurance Capital Standard program addresses globally active insurance groups, and the IAIS Insurance Core Principles and ComFrame articulate broader group-supervision expectations. They do not replace the EU Article 80 or 330 tests; they show that capital resources, their quality, and their transferability are central questions across international supervision. Legal implementation, entity structure, and supervisory mandates must still be treated jurisdiction by jurisdiction.

Matching Adjustment Interactions

Current portfolio treatment

The matching adjustment is a valuation mechanism for approved portfolios of eligible insurance liabilities and assigned assets. It modifies the relevant risk-free interest-rate term structure in the calculation of best-estimate liabilities subject to specified regulatory conditions. The resulting difference in the valuation of liabilities may materially affect the insurer's solvency balance sheet, but that valuation mechanism has a different legal object from the classification of own-fund items as restricted for purposes of Article 80.

Under the current delegated regulation, however, the calculation framework links the mechanisms operationally. Article 216 covers both ring-fenced funds and portfolios for which use of the matching adjustment has been approved, and Article 217's notional SCR method includes each matching-adjustment portfolio as a separate component in addition to each ring-fenced fund and the rest of the undertaking. This method restricts recognition of cross-component diversification regardless of whether an approved matching-adjustment portfolio also meets the separate Article 80 restriction test.

A portfolio can therefore meet matching-adjustment eligibility conditions without its mere status as an approved matching-adjustment portfolio determining the nature of all associated own funds. Distinguishing the mechanisms prevents an analytical error: the formal valuation treatment of long-term asset-liability cash-flow matching should not be used as a stand-in for the enforceability of a policyholder surplus allocation. The present capital-aggregation rule brings the two categories together for a specified computational purpose, but the categories have distinct underlying legal tests.

The question is consequential for retirement and annuity businesses that manage long-duration liabilities through dedicated fixed-income portfolios. Asset dedication can support cash-flow matching and risk control, while the ability to divert a portfolio's surplus to other losses depends on legal and supervisory arrangements. Reinsurance protection, collateral restrictions, and guarantees may change both exposures and the capital quality of related resources. Any calculation must determine whether the matching portfolio is additionally an Article 80 fund and maintain the appropriate notional calculations without assuming two independent sets of capital resources for one underlying asset.

Matching adjustment, ring-fencing, and accounting distinctions

Matching-adjustment approval, prudential restrictions on own-fund transferability, and financial-statement presentation answer separate questions. The first determines whether a particular liability discounting treatment and its associated conditions apply. The second asks whether own-fund resources can cover losses across portfolios. The third describes accounting recognition and measurement under the relevant accounting framework. Although these questions interact in the same economic balance sheet, they have different units of analysis and should be reconciled explicitly.

A matching-adjustment portfolio's lower liability value can increase the apparent excess of assets over liabilities. That does not alone establish that any increase represents freely transferable capital. If legal rules restrict the newly recognized surplus to policyholders or particular risk obligations, the own-funds limitation must still be tested. Conversely, where the obligations and economic assets are assigned to a matching portfolio but the fund does not satisfy Article 80's restriction criteria, the new diversification rules applicable in 2027 create a materially different capital position from the present Article 217 aggregation.

Distinctions in the treatment of adjustment benefits have also featured in the wider review of Solvency II. The 2025 amendment to the Directive is part of a comprehensive review of long-term guarantee measures, proportionality, supervision, and capital requirements. The scope of this monograph is intentionally narrower: its subject is the regulatory status of ring-fenced resources and the recognition of cross-portfolio diversification. Any projection of an insurer's post-review solvency ratio would additionally need to model all other changes to technical provisions, risk margin, and supervisory requirements that apply to that insurer.

The prudential distinction has an operational implication for data systems. Firms should record at least three independent flags for an insurance portfolio: the presence of a binding ring-fence restriction; authorization to apply the matching adjustment; and the accounting or operational segmentation used for valuation. A binary field named restricted portfolio is too ambiguous to drive a calculation that depends on independent legal conditions. Separate records preserve the ability to change one classification without silently modifying the others.

Comparative regulatory context

Differences among national supervisory systems confirm the need to distinguish legal classification from generic economic labels. In the United Kingdom, the Prudential Regulation Authority's with-profits supervisory statement provides an example of a jurisdiction-specific approach to governance, policyholder interests, and management of such insurance business. The UK post-withdrawal prudential framework has its own evolution, so a current EU article number cannot be applied automatically to UK firms by referring to historical EU-derived wording.

Similarly, the IAIS Insurance Capital Standard framework concerns a global group-supervision project. The international question whether capital is capable of supporting losses across participating entities is comparable to the EU's availability concern, but the quantitative implementation, institutional authority, and applicable legal rules must be separately established. The same economic word diversification may be used in all frameworks while denoting differently constrained and differently calculated benefits.

An international scientific comparison should therefore state the jurisdiction, reporting entity, valuation date, regulatory version, definition of own funds, permitted transfer paths, and the risk aggregation model. Without that information, apparent differences in solvency ratios between jurisdictions can reflect incompatible measurement rules rather than differences in underlying insurance risk. The comparative finding is methodological and supports evidence-based interpretation, not an attempt to prescribe a universal ring-fencing mechanism.

2027 EU Rule Changes

Legal chronology and implementation

The European Union has adopted legislation changing the operation of ring-fenced funds and matching-adjustment portfolios, with the relevant substantive application date in 2027. The 2025 Solvency II amending Directive and the 2026 delegated amendment must be distinguished from the consolidated text describing rules currently applied in supervision. The delegated act was adopted October 29, 2025 and published in February 2026, while its relevant regulatory amendments are scheduled for application on January 30, 2027.

EIOPA published the revised ring-fenced-fund Guidelines report on July 15, 2026. Its official explanatory page states that the guidelines applicable until January 30, 2027 are those issued in 2015, and that the revised guidance applies from that date. This statement provides a clear supervisory temporal boundary. The legislative basis and actual effective provisions still require reference to the EU instruments themselves, because guidelines operate within the governing legal framework.

The temporal sequence creates two simultaneous documentary realities in November 2026. The current insurer SCR is calculated under the presently applicable Article 216–217 rules that generally separate matching-adjustment portfolios for purposes of the relevant notional aggregation. Meanwhile, insurers preparing 2027 projections should model the future provisions under which matching-adjustment approval on its own does not trigger the same no-diversification treatment. A scientific analysis must label these as separate measurement regimes and avoid treating a later source as evidence that the earlier rule has already ceased to apply.

Changes to Articles 81, 216, and 217

The 2026 delegated amendment changes the designation and scope of Article 81 so that the reconciliation-reserve adjustment addresses ring-fenced funds. Its amendments to Articles 216 and 217 remove the former automatic pairing of ring-fenced funds with matching-adjustment portfolios in the special notional-SCR aggregation method. For 2027 application, the relevant Article 216 method is activated by a qualifying ring-fenced fund; matching-adjustment authorization by itself no longer supplies an independent trigger.

The revised Article 216 retains a specifically delimited exception concerning undertakings that obtained supervisory approval for the Directive's Article 304 regime before the relevant January 2027 cut-off. It would be inaccurate to infer that every ring-fenced fund becomes entitled to full cross-fund diversification after the review. In fact, qualifying legal restrictions continue to constrain the recognized diversification of actual ring-fenced funds. The reform changes the link between matching adjustment and ring fencing; it does not abolish the prudential importance of restricted own funds.

EIOPA states the policy purpose directly in its July 2026 reform explanation: undertakings using the matching adjustment can assume full diversification between that portfolio and the rest of the business when the associated best-estimate obligations do not constitute a ring-fenced fund. The phrase do not constitute a ring-fenced fund is the operative condition. It requires a legal and contractual examination, not merely a statement that management would prefer to pool risks.

The distinction can be displayed without projecting any insurer's actual model output:

Question Rules applicable November 15, 2026 Framework applying January 30, 2027
Does a qualifying ring-fenced fund require the restriction analysis? Yes, through Articles 80 and 81 Yes, under the amended provisions
Is a matching-adjustment portfolio independently captured by the special Article 217 aggregation because it is a matching-adjustment portfolio? Generally yes under current Articles 216–217 No; ring-fenced-fund status supplies the relevant trigger
Is full cross-portfolio diversification permitted where the matching-adjustment portfolio is not a ring-fenced fund? Not under the current special method solely because the portfolio qualifies for matching adjustment The amended approach permits full diversification, subject to the other governing SCR rules
Does a ring-fence label determine deferred tax recoverability or tier eligibility? No; separate valuation and eligibility tests remain necessary No; separate valuation and eligibility tests remain necessary

This comparison is a regulatory interpretation of the stated provisions. A firm-specific effect depends on the classification of the matching-adjustment book, model approval, transitional conditions, and whether other insurance-business restrictions independently establish a ring-fenced fund.

Prudential effects and transition controls

The reform creates potential benefits in capital aggregation for insurers whose matching-adjustment portfolios were separated by the current method without an underlying ring fence. In the toy quadratic example developed above, recognizing diversification could move an illustrative capital measure from 180 toward 142.83 under a 0.25-dependence assumption, holding all other features constant. That number is a sensitivity illustration and is not a forecast of the effect of the 2027 EU law. Real capital effects may be smaller, larger, or offset by other regulatory changes, and the legal test of whether the fund truly qualifies for greater diversification remains decisive.

A key risk is classification drift. An undertaking might simply rename a legally restricted participating fund as a matching-adjustment portfolio and attempt to claim the resulting diversification. The legal mechanics preclude such a shortcut: the resources remain subject to Article 80 when their permitted loss-absorption uses are unchanged. The correct implementation tests the legal restriction anew and records why it does or does not exist. It then computes the applicable capital using the correct January 2027 rule version.

The transition is also a governance exercise. Firms need a register of matching-adjustment portfolios and potentially restricted subfunds, a documented comparison of current and revised Articles 81 and 216–217, updated actuarial assumptions, and controlled changes to systems that produce fund-specific and total SCR measures. Discrepancies between policyholder obligations, investment-management systems, actuarial cash flows, capital templates, and financial reporting should be resolved before a firm makes a material claim about newly available diversification.

Supervisors and researchers must distinguish an improvement caused by the recognition of a dependence benefit from an improvement caused by a genuinely increased stock of eligible resources. If the same economic balance sheet produces a lower SCR solely because the special aggregation rule changes, the reported coverage ratio may increase even where the insurer has received no new contributions and its fund distributions remain unchanged. This does not make the new method illegitimate. It means the interpretation of the change must identify its measurement cause and explain how the law conditions recognition on actual restrictions.

Disclosure also requires temporal precision. Historical series built across January 2027 may contain a structural break in measured SCR and eligible own funds due to changes in law and model treatment. An observed rise in a solvency ratio across the boundary is not necessarily evidence of reduced claim volatility, better underwriting, or higher cash holdings. Comparative studies should separate the contribution of changed parameters, changed model structure, changed fund classification, and real balance-sheet movements wherever the data permit.

Research findings and conditions for generalization

The combined legal and mathematical evidence supports five findings. First, a restricted own-fund item acquires its regulatory character through limits on usable loss absorption, not from the physical or administrative location of the assets. Second, the reconciliation-reserve adjustment and the notional SCR aggregation are distinct legal mechanisms that may simultaneously affect a coverage ratio. Third, the presence of a matching-adjustment portfolio is currently relevant to the special capital calculation even where its classification under Article 80 requires an independent inquiry. Fourth, the 2027 amendment narrows that special link, retaining ring-fence protections while permitting recognized diversification for qualifying matching-adjustment portfolios that do not constitute ring-fenced funds. Fifth, own-funds tiers, deferred tax recoverability, and group transferability impose additional tests that cannot be collapsed into the ring-fenced-surplus adjustment.

The research also identifies the principal limitations of extrapolation. The delegated regulation supplies the applicable method and legal criteria; it does not determine, absent the contracts and national legal evidence, whether a particular company's participating fund meets them. EIOPA guidelines help align supervisory interpretation but are different in legal status from the EU delegated regulation. Published actuarial aggregation studies clarify dependence and allocation, but they do not validate the underlying legal right to mobilize capital. A fully quantified real-insurer result requires audited balance-sheet data, fund-level risk-model outputs, enforceable contract evidence, valuation assumptions, and regulatory version control.

These findings offer a research agenda rather than a new named theoretical system. Future empirical work could examine how the 2027 reform changes reported SCR and capital ratios in a documented panel of affected undertakings, controlling for life-insurance business mix, reinsurance, internal-model status, and prevailing market shocks. Such work would require access to sufficiently granular disclosures and explicit rules for distinguishing legal ring fences from managerial portfolio segmentation. The present monograph establishes the documentary and interpretive requirements on which a credible empirical design would depend.

Conclusion

Ring-fenced funds and restricted own funds expose the difference between a resource's recognized economic value and its regulatory capacity to absorb losses across insurance obligations. In Solvency II, the distinction has operational consequences for the reconciliation reserve, eligible own funds, and the Solvency Capital Requirement. Article 80 identifies the relevant transferability restrictions. Article 81 limits the surplus credited through the reconciliation reserve by reference to the fund's notional SCR. Articles 216 and 217 specify a further capital computation that, under rules applicable in November 2026, addresses ring-fenced funds and matching-adjustment portfolios without recognizing diversification between the prescribed components. Separate own-funds quality and group availability rules preserve additional constraints on the reported resource.

The controlled numerical example demonstrates why a fund with 150 units of restricted resources and a 100-unit notional SCR can contribute less to the total usable capital of the insurer than its full reported surplus suggests. The separate quadratic sensitivity example demonstrates how recognizing or disallowing inter-component diversification changes the denominator of a ratio even while the underlying distributions of risk remain unaltered. These two mechanisms must be measured and interpreted individually. Statistical dependence cannot establish legal capital fungibility; legal fungibility, in turn, does not imply perfect statistical correlation or universal liquidity.

The 2027 Solvency II reforms create a clear temporal transition. From January 30, 2027, the amended treatment permits matching-adjustment portfolios that do not constitute ring-fenced funds to benefit from cross-portfolio diversification under the applicable calculation framework. Genuine ring-fencing restrictions remain prudentially significant. The transition accordingly raises the importance of evidence about the legal status of surplus, the contractual allocation of profits, the treatment of policyholder rights, and the actual boundary of fund-specific liabilities.

A reliable analysis of insurer solvency retains four independently documented dimensions: the legal availability of own funds; their valuation and tier eligibility; the applicable diversification and notional-SCR method; and their transferability across group entities and time. The integrative result is a more precise interpretation of the existing regulatory framework and a traceable analytical basis for supervisory, actuarial, and economic research. Scientific claims made in this manuscript extend to the identified legal sources and stated hypothetical assumptions. Application to individual insurers requires original institutional evidence and the regulation applicable at the valuation date.

References

  • European Parliament and Council. (2009). Directive 2009/138/EC of 25 November 2009 on the taking-up and pursuit of the business of Insurance and Reinsurance (Solvency II). Official Journal of the European Union. Official source or publication.

  • European Commission. (2014/2015). Commission Delegated Regulation (EU) 2015/35 of 10 October 2014 supplementing Directive 2009/138/EC. Official Journal of the European Union. Official source or publication.

  • European Parliament and Council. (2025). Directive (EU) 2025/2 amending Directive 2009/138/EC as regards proportionality, quality of supervision, reporting, long-term guarantee measures, macro-prudential tools, sustainability risks, and group and cross-border supervision. Official Journal of the European Union. Official source or publication.

  • European Commission. (2026). Commission Delegated Regulation (EU) 2026/269 of 29 October 2025 amending Delegated Regulation (EU) 2015/35 as regards technical provisions, long-term guarantee measures, own funds, risk requirements, reporting and disclosure, proportionality, and group solvency. Official Journal of the European Union. Official source or publication.

  • European Insurance and Occupational Pensions Authority (EIOPA). (n.d.). Delegated Regulation (EU) 2015/35, Article 70: Reconciliation reserve. Solvency II Single Rulebook. Official source or publication.

  • EIOPA. (n.d.). Delegated Regulation (EU) 2015/35, Article 80: Ring-fenced funds requiring adjustments. Solvency II Single Rulebook. Official source or publication.

  • EIOPA. (n.d.). Delegated Regulation (EU) 2015/35, Article 81: Adjustment for ring-fenced funds and matching adjustment portfolios. Solvency II Single Rulebook; pre-January 2027 text. Official source or publication.

  • EIOPA. (n.d.). Delegated Regulation (EU) 2015/35, Article 82: Eligibility and limits applicable to Tiers 1, 2 and 3. Solvency II Single Rulebook. Official source or publication.

  • EIOPA. (n.d.). Delegated Regulation (EU) 2015/35, Article 216: Calculation of the SCR in the case of ring-fenced funds and matching adjustment portfolios. Solvency II Single Rulebook; pre-January 2027 text. Official source or publication.

  • EIOPA. (n.d.). Delegated Regulation (EU) 2015/35, Article 217: SCR calculation method for ring-fenced funds and matching adjustment portfolios. Solvency II Single Rulebook; pre-January 2027 text. Official source or publication.

  • EIOPA. (n.d.). Delegated Regulation (EU) 2015/35, Article 330: Availability at group level of the eligible own funds of related undertakings. Solvency II Single Rulebook. Official source or publication.

  • EIOPA. (n.d.). Directive 2009/138/EC, Article 91: Surplus funds. Solvency II Single Rulebook. Official source or publication.

  • EIOPA. (n.d.). Directive 2009/138/EC, Article 93: Characteristics and features used to classify own funds into tiers. Solvency II Single Rulebook. Official source or publication.

  • EIOPA. (2015). Guidelines on ring-fenced funds (EIOPA-BoS-14/169), accompanied by supervisory publication and subsequent language files. Official source or publication.

  • EIOPA. (2026, July 15). Final report on revised Guidelines on ring-fenced funds—Solvency II Review; revised Guidelines EIOPA-BoS-26/284 applying from January 30, 2027. Official source or publication.

  • EIOPA. (n.d.). Q&A 2200: Deferred tax assets and ring-fenced funds; valuation restrictions and netting. EIOPA Q&A database. Official source or publication.

  • EIOPA. (n.d.). Q&A 2513: Ring-fenced deferred tax assets and own-funds reporting template S.23.01. EIOPA Q&A database. Official source or publication.

  • EIOPA. (2026). Q&A 3591: National contribution-based allocation rules for traditional with-profits business; rejected question record. EIOPA Q&A database. Official source or publication.

  • International Association of Insurance Supervisors (IAIS). (2024–2026). Insurance Capital Standard (ICS): international group solvency capital framework and implementation materials. IAIS. Official source or publication.

  • IAIS. (2024). Insurance Core Principles and Common Framework for the Supervision of Internationally Active Insurance Groups (ComFrame), December 2024 consolidated text. Official source or publication.

  • Filipović, D. (2009). Multi-level risk aggregation. ASTIN Bulletin, 39(2), 565–575. https://doi.org/10.2143/AST.39.2.2044648. Official source or publication.

  • Bermúdez, L., Ferri, A., & Guillén, M. (2013). A correlation sensitivity analysis of non-life underwriting risk in solvency capital requirement estimation. ASTIN Bulletin, 43(1), 21–37. https://doi.org/10.1017/asb.2012.1. Official source or publication.

  • Asimit, A. V., Bădescu, A. M., Haberman, S., & Kim, E.-S. (2016). Efficient risk allocation within a non-life insurance group under Solvency II regime. Insurance: Mathematics and Economics, 66, 69–76. https://doi.org/10.1016/j.insmatheco.2015.10.008. Official source or publication.

  • Prudential Regulation Authority (PRA). (2015; subsequent updates). Supervisory Statement SS14/15: With-profits. Bank of England. Official source or publication.