RESEARCH DOSSIER
Matching Adjustment and Illiquidity Premium in Insurance Valuation
Abstract. The matching adjustment (MA) combines an actuarial valuation technique with a conditional supervisory recognition of investment spreads. Its economic justification depends on the insurer's ability to meet predictable long-term claims from appropriately matched asset cash flows while bearing the risks that remain in those investments. This monograph develops the financial interpretation of the MA, distinguishes asset illiquidity from the liquidity characteristics of insurance liabilities, reconstructs the role of the fundamental spread, and evaluates eligibility, stress behavior, and institutional governance. It compares operative Solvency II and Solvency UK requirements, the European Union's reforms applicable from January 30, 2027, IFRS 17 discount-rate measurement, and the IAIS Insurance Capital Standard. Empirical bond-market research, primary legislation, supervisory materials, and the PRA's 2025 life insurance stress test supply the evidence. An explicitly hypothetical ten-year cash-flow experiment demonstrates how a 70-basis-point adjustment changes the present value of a 100-unit liability from 82.0348 to 76.6118, and shows why this valuation result cannot be interpreted as economically generated capital without jointly revaluing assets, retained risks, and capital requirements. The original contribution is an integrated analytical synthesis and reproducible illustration, with clear limits on causal inference and jurisdictional generalization. No new proprietary theory or external peer review is claimed.
Keywords: Matching adjustment; illiquidity premium; Solvency II; Solvency UK; fundamental spread; insurance valuation; asset-liability management; annuities; credit risk; liquidity risk; IFRS 17; Insurance Capital Standard; insurance capital; cash-flow matching.
Introduction
A long-term insurer receives premiums and undertakes obligations to pay future claims. Its investments trade in markets that assign prices to uncertain future cash flows, while insurance liabilities have their own timing, guarantees, and contractual options. An asset yield observed today can incorporate the time value of money, expected default losses, compensation for uncertain credit losses, market liquidity effects, embedded options, and other factors. Only an economically and regulatorily justified part of that yield can appropriately affect the valuation of a policyholder promise.
Article 77b of Solvency II restricts the matching adjustment to eligible insurance or reinsurance portfolios with assigned backing assets and supervisory permission. Article 77c's calculation connects effective asset and liability discount rates and requires the exclusion of the fundamental spread for retained risk. The MA is therefore a conditional accounting of asset-liability economics within a prudential measurement system. It does not produce an extra coupon or remove the risk of issuer default.
The investigation employs documentary comparison of dated primary rules, actuarial present-value derivations, qualitative appraisal of empirical financial research, and transparent numerical sensitivity scenarios. Its geographic scope is international, with the legal analysis concentrated on the EU and UK, and accounting and insurance-group comparisons drawn from IFRS 17 and the IAIS standard. All numerical scenarios prepared here are hypothetical rather than observations of an insurer. Results about portfolio eligibility and capital relief are consequently conditional on actual contracts, market data, capital rules, and supervisor permissions.
The study separates three questions: how much of a traded asset's spread represents compensation for illiquidity, how much illiquidity an insurance obligation genuinely possesses, and how much spread a particular supervisor permits in liability valuation. These quantities may be related without being numerically identical.
Liability Illiquidity Economics
What is illiquid?
Asset-market liquidity concerns the ability to exchange an instrument at a known price, size, and speed at limited transaction cost. Bid–ask spreads, trade frequency, market depth, price impact, and the speed of recovery after trades measure different aspects. Insurance-liability liquidity concerns when the insurer can be required to deliver cash and whether the policyholder or beneficiary can accelerate payment. A contract with a distant expected payment is not necessarily illiquid if surrender, commutation, or guarantee options permit an earlier demand.
A long-duration insurer can be a natural holder of less marketable investments when its liabilities generate stable, forecastable cash requirements. Such an investor may be better able to withstand temporary changes in transaction prices than an investor facing daily withdrawals. This economic capacity is never established by the investment's classification alone. Claims, expenses, hedging collateral, reinsurance recapture, funding conditions, and legal options jointly determine the actual ability to hold the position to maturity.
The empirical case for a priced liquidity component is substantial. Chen, Lesmond, and Wei (2007) find that corporate bond liquidity measures help explain yield spreads after selected controls. Bao, Pan, and Wang (2011) identify bond illiquidity from transaction-price evidence. During the financial crisis, Dick-Nielsen, Feldhütter, and Lando (2012) show that liquidity contributions to credit spreads changed sharply and differently across ratings. These studies establish that marketability affects prices; they do not calibrate a specific insurer's allowable MA or guarantee future access to the measured compensation.
The benchmark also matters. Longstaff's Treasury liquidity study finds a liquidity premium in Treasury bond prices compared with government-supported agency instruments. An estimated spread to a Treasury benchmark can therefore reflect liquidity properties of both securities. Bond-market estimates are sensitive to sample selection, tax differences, collateral terms, trading stress, and identification assumptions.
Credit and liquidity in the same spread
For conceptual analysis, write the observed spread as a sum of expected credit loss, compensation for unexpected credit risk, liquidity compensation, and remaining price components. This is a conceptual decomposition rather than a set of directly observable additive factors. Risk and liquidity interact: bonds often become harder to sell when creditworthiness deteriorates, so a yield premium measured during a crisis can mix compensation for multiple correlated exposures.
The insurer still bears credit default even if it has no intention of selling. It can also bear downgrade risk without ultimate default: downgrade can change capital needs, collateral treatment, eligibility, or the cost of finding replacement securities. Embedded prepayment and call options change the scheduled investment payments and may create reinvestment risk. A plausible liquidity benefit must be considered net of these cash-flow and risk-transfer characteristics.
Liability measurement and accounting distinctions
IFRS 17 paragraphs 36 and B79–B81 require discount rates that reflect the time value of money, cash-flow and liquidity characteristics of the insurance contracts, with consistency to relevant observable financial-instrument prices. A bottom-up method adjusts a liquid risk-free curve for the contract's liquidity; a top-down method starts from a reference-asset yield and removes factors irrelevant to the liabilities. These measurement alternatives recognize liability liquidity economics while serving the distinct objective of financial reporting. They do not automatically confer EU or UK matching-adjustment eligibility.
The European Commission's Solvency II explanation places the MA alongside the volatility adjustment, risk margin, and risk-free extrapolation as different long-term-guarantee mechanisms. Each acts through a different regulatory channel. Conclusions about illiquidity should identify the exact mechanism rather than call all discount-rate adjustments a matching adjustment.
Measuring the liquidity component: empirical identification and boundaries
A reliable estimate of bond illiquidity needs a research design that differentiates variation in trading conditions from variation in the expected contractual payoff. Transaction-based measures, including deviations of trade prices from an efficient-price benchmark, are sensitive to quote quality, reporting thresholds, turnover, issue size, dealer inventory and the distribution of trade sizes. Measures based on an observed bid–ask spread may be biased where trades do not occur continuously, particularly when instruments have bespoke covenants or nonstandard collateral. A sample that excludes securities with no trades in the observation window can systematically underrepresent the most difficult-to-sell holdings. The empirical estimate should therefore be interpreted as conditional on the observation process.
The methods in Bao, Pan, and Wang (2011) use the behavior of transaction prices to estimate the importance of illiquidity in the corporate bond market. This methodological choice illustrates a general lesson: the premium cannot be directly identified by labeling an entire yield spread 'illiquid.' Statistical models must control for credit quality, maturity, embedded options and, where possible, taxes and dealer conditions. The 2007 yield-spread study and the crisis-period transaction evidence each provide complementary approaches, but their estimates are population- and period-dependent. Neither supplies a universal regulatory deduction.
Observed yield is also affected by the reference curve. A premium calculated over government securities differs from the spread over overnight indexed swaps and from the premium embedded in a reference portfolio of other corporate securities. Differences in duration, collateral, currency, tax treatment, quotation time and expected recovery can bias such comparisons. A suitable actuarial application therefore identifies the price date, market convention, discount factor curve, asset valuation, and credit-risk adjustment before estimating an additional liquidity component.
Insurance-liability liquidity needs its own evidence. Contract data should identify guaranteed payments, benefit options, lapse or surrender terms, annual expenses, collateral, reinsurance triggers, inflation indexation, and claim settlement flexibility. Empirical policyholder behavior may be conditional on interest rates and market stress; a fixed unconditional lapse percentage can obscure precisely the state in which assets must be sold. If liabilities are fully non-surrenderable but the insurer must meet other group-level cash needs, the available investment horizon can be shorter than the contractual maturity. The relevant unit of analysis is an operating insurance balance sheet with ring-fenced commitments and credible sources of contingency liquidity.
Why the market spread and the permitted adjustment differ
An analytical decomposition of a spread is necessarily model-dependent. Expected loss depends on probability of default, exposure at default, the timing of default and recovery, and the timing of recoveries. An unexpected-loss premium also reflects the market price of bearing tail risk. The liquidity component includes expected transaction costs and the cost of uncertain future trading, whose sensitivity to credit stress may be high. A policyholder's claim has no direct obligation to absorb each of these components merely because the insurer selected the asset.
Regulatory MA and financial-economic illiquidity estimates consequently ask different questions. A financial estimate seeks to explain variation in bond prices; a prudential adjustment seeks to determine which part of an assigned portfolio's return can be incorporated without compromising a conservative valuation of promised insurance benefits. The first requires identification of a price component, and the second additionally requires legal eligibility, cash-flow matching and a retained-risk deduction. A prudent analyst can find evidence for an illiquidity premium and still conclude that no additional MA is permitted for a particular contract or portfolio.
Matching Adjustment Rationale
Present value and discount rates
Let denote the expected liability cash outflow in year and the applicable risk-free discount factor. The risk-free best estimate can be represented as
For an illustration with a flat annual effective risk-free rate and additive adjustment , the simplified form is
Actual Solvency II calculations use a currency-specific term structure and a regulatory formula; the scalar model above is a teaching specialization. For positive future payments, raising lowers the discounted value. The effect is larger for payments farther in the future, other things equal, because the discount factor compounds over more periods.
The conceptual structure of Solvency II Article 77c is an effective asset-liability yield difference less the fundamental spread:
Here is the effective yield obtained by equating the relevant liability cash flows to the assigned asset portfolio's value, is the corresponding risk-free effective liability rate, and is the retained-risk deduction. This shorthand suppresses detailed term-structure and cash-flow conventions. Delegated Regulation Article 53 explicitly addresses default-adjusted expected asset cash flows and the component of the deduction still to be subtracted. Without a consistent treatment of this adjustment, a model can double-count credit loss.
The fundamental spread
The fundamental spread is the principal control over the part of market yield attributed to retained risks. Delegated Regulation Article 54 specifies a prudent, reliable, transparent and objective calculation and relevant regulatory conventions for recovery and historic spread observations. A prescribed recovery assumption is an ingredient in the supervisory measure, not evidence that an individual issuer's actual recovery will equal it.
A market-based credit-risk premium and a regulatory fundamental spread need not coincide. The regulator can apply standardized inputs and prudent conservatism, while a financial economist estimates risk prices with different benchmarks and methods. Comparison requires stating each estimate's purpose, horizon, data, and treatment of correlated tail loss. The correct question is whether retained risks have been adequately recognized when the insurer claims an MA benefit.
Capital interpretation
For a simplified balance sheet, surplus , where is asset value, the liability best estimate, and other recognized obligations. Holding all other quantities constant, , so a lower measured best estimate improves the arithmetic surplus. In real stress, asset market values, credit losses, eligible own funds, hedges, capital charges, and other liability components also move.
Economic risk reduction is created through credible matching and risk management; the MA is a regulatory valuation recognition of some consequences of that arrangement. These two ideas should be audited separately. A fall in discounted liabilities alone does not demonstrate newly earned assets, reduction of all risk, or an improvement in the distribution of policyholder payment outcomes.
Matching beyond duration
Equal modified durations do not ensure equal cash flows. Different coupons, maturities, curve exposure, default terms, and options may produce large mismatches under the same interest-rate scenario. Matching analysis should compare expected cash inflows and outflows over time and should test deviations in adverse states. The economic premise is strongest where cash-flow differences can be funded without distressed sales and without dependence on uncertain asset replacement.
Mathematical structure of a matching portfolio
For a deterministic portfolio, define expected asset inflows , liability outflows , initial liquid resources , and net transaction or hedge costs . Ignoring interest earned on intermediate liquidity for the moment, an elementary cumulative cash-flow adequacy condition is
This condition is deliberately stronger than equality of terminal present values. A portfolio can have enough present value to fund all claims and still experience an interim liquidity deficit if its coupons arrive after the liabilities are due. It also differs from the formal eligibility test under each prudential regime; the inequality is an illustrative liquidity feasibility screen. When reinvestment, borrowing, collateral and market states are introduced, each term becomes state-dependent and the condition must hold within defined stress tolerances, rather than automatically in every conceivable state.
The most revealing mismatches are temporal. A ten-year zero-coupon asset is not a perfect match for a ten-year liability portfolio that pays a meaningful part of its claims monthly throughout the decade. A sequence of coupon-paying bonds can produce excess cash early, requiring reinvestment; its effective yield earned on claims depends on reinvestment rates. A duration hedge can reduce the sensitivity to a parallel yield-curve shift while leaving the insurer vulnerable to slope changes, curvature, issuer credit losses and changes in liability experience. Matching must be measured in the units in which cash becomes available.
It is useful to distinguish static matching at a valuation date from dynamic matching under the evolution of liabilities and invested assets. The static projection uses present assumptions and contractual expectations; the dynamic analysis must specify what happens when a bond defaults, a corporate action accelerates payment, an annuitant lives longer than anticipated or a downgrade changes capital requirements. A firm may be capable of maintaining an effective match through rebalancing, but the acquisition cost and financing capacity required to do so are part of the economic assessment.
Return earned through time and the risk of liquidation
A hold-to-maturity strategy can be valid only insofar as payments can be made without an unplanned sale of the assigned asset pool. The insurer must maintain enough cash, high-quality liquid assets or reliable funding to meet claim variations, operating costs and hedge collateral. Market assets can remain nominally unencumbered yet be economically committed to collateral obligations. A supervisory eligibility decision should therefore interact with a documented and tested liquidity plan.
The principal economic test is not whether a matched bond's market price can fluctuate. Prices necessarily fluctuate as rates, spreads and trading conditions change. The test concerns whether those movements create realized losses or prevent the insurer from fulfilling its obligations. If liquidity needs force an asset disposal, the captured yield may differ substantially from the original purchase yield. For private debt, the discount demanded for immediate sale may be larger exactly when other institutions face similar liquidity pressure. The MA must not be treated as a guarantee that portfolio assets can always be retained.
Asset and Liability Eligibility
The EU permission conditions
Under Article 77b, the relevant obligations are specified life insurance and reinsurance portfolios, including qualifying annuities from non-life business. The insurer must assign an asset portfolio intended to cover the best estimate, maintain the assignment, and identify, organize, and manage the relevant assets and obligations separately under prescribed conditions. The regulatory regime also imposes matching and contractual restrictions and prior supervisory approval. Long maturity or high yield alone establishes neither legal eligibility nor economic matching.
Asset selection therefore has two independent dimensions: whether the asset is appropriate to the liability's cash-flow pattern, and whether the asset meets the legal classification and permission requirements. The same applies to obligations. A portfolio of long-term annuities may have stable aggregate expected payments, but longevity uncertainty, indexation, policy options, and contractual expense exposures still require actuarial modeling.
Solvency UK and highly predictable cash flows
The PRA's PS10/24 final policy, effective June 30, 2024, broadened certain investment possibilities and clarified requirements concerning highly predictable asset cash flows, matching tests, fundamental-spread additions, and controls. The policy discusses a 10% limit on the recognized MA benefit attributable to permitted highly predictable cash flows; it is not a blanket 10% ownership cap on all illiquid debt. The amended rules require precise attention to the contractual and supervisory definitions used.
The PRA's SS7/18 version register identifies an October 2025 version in force from October 27, 2025 and a July 2026 version scheduled to become effective December 31, 2026. At the November 15 cutoff, the latter is a future-effective text. The operative supervisory expectations address asset and liability eligibility, matching, fundamental-spread additions, liquidity planning, and continuing portfolio management.
The 2025 Matching Adjustment Investment Accelerator creates a framework for permitted recognition in eligible investment circumstances with accompanying safeguards. Administrative acceleration does not erase the need to appraise asset cash flows, retained risk, concentration, liquidity, and supervisory compliance.
Asset-level and liability-level tests
| Feature | Potential role in matching | Residual exposure |
|---|---|---|
| Fixed coupons and principal | Supports contractual payment replication | Default, recovery, rating deterioration |
| Highly predictable but variable payments | May qualify under specific UK permissions | Forecast, timing, and legal-structure risk |
| Thinly traded private loans | May include a market illiquidity premium | Stale valuations and costly disposal |
| Callable or prepayable instruments | Change cash-flow timing | Reinvestment and negative convexity |
| Collateralized lending | May support scheduled payments | Collateral valuation, enforcement, concentration |
| Foreign-currency assets | Broaden portfolio opportunities | Hedge basis, roll and collateral liquidity |
| Structured assets | Can tailor cash-flow profiles | Waterfall complexity, correlation, servicing |
These are financial characteristics, not statements that each asset class satisfies a particular jurisdiction's legal criteria. A sound eligibility review applies the actual rule text, the approved asset and liability pool, and the valuation date.
Operational controls and liquidity
A cash-flow match depends on implementation. Portfolio managers, actuaries, treasury teams, compliance, and risk functions must reconcile contract features, payment projections, internal ratings, prices, and hedges. Rebalancing after downgrade, prepayment, or a change in expected longevity can be expensive. A control environment must measure whether the matched pool survives such events without an unjustified assumption of uninterrupted market liquidity.
The PRA's funded reinsurance supervisory guidance emphasizes collateral quality, counterparty exposure, recapture stress, and costs of rebalancing assets that return to the cedant. A nominally transferred exposure can return to a firm's balance sheet at an adverse moment. Such cases demonstrate why matching is an institution-level liquidity capability as well as a portfolio cash-flow calculation.
Cash-flow eligibility under adverse states
Eligibility assessment should be organized around asset and liability features that change cash-flow timing. For the asset portfolio these include issuer calls, voluntary borrower prepayment, mandatory prepayment, amortization, variable interest provisions, collateral enforcement, waterfall triggers, maturity extensions and currency mismatch. For the liabilities they include mortality, longevity, policyholder options, discretionary benefits, claim expenses and indexation. The same economic scenario can alter both sides. For example, falling interest rates may accelerate mortgage prepayments, reduce reinvestment yields, and increase the value of annuity or guaranteed-benefit obligations.
A well-designed analysis starts with contractual cash-flow schedules and then introduces stochastic overlays. These may include rating-transition matrices, default and recovery distributions, mortality and longevity shocks, collateral requirements, and asset-sale discounts. The choice between stochastic and deterministic treatments must be justified according to materiality. A small assumption error on a long-horizon payment can materially change present value, and a small probability of a large liquidity deficit can be more important than a modest average spread.
A useful quality-control matrix has four columns: source of uncertainty, cash-flow effect, financial measurement effect, and available management response. For an unexpected corporate default, the cash-flow effect is missed coupons or principal and a delayed uncertain recovery; the financial effect is asset impairment and potentially a changed fundamental spread and capital requirement; the response may include recovery management or asset substitution, with costs and timing. For a surge in collateral calls, promised bond cash flows may remain unchanged, while required immediately available cash rises; the response is funding or sale, and its feasibility must be stressed. For longevity deterioration, liability payments may extend; the response is asset purchases or hedges at prevailing conditions. The table's purpose is traceability, not the creation of a new proprietary scoring model.
Fundamental spread calibration and sensitivity
A simplified expected credit-loss approximation can help expose assumptions. For a short interval with default probability , loss given default , and exposure , expected loss is approximately , subject to the timing of default, discounting and contractual terms. For a multi-year bond, survival probabilities, timing of exposure, coupons, recovery and rating migration must be incorporated explicitly. A supervisor's prescribed fundamental spread can include conservative floors or other components beyond an individual's modelled mean loss. The appropriate calculation is the one required by the governing effective legal instrument.
Two sensitivities should always be distinguished. The first is sensitivity to the expected cash flows used to derive effective portfolio yields, including default adjustments. The second is sensitivity to the retained-risk deduction subtracted after those effects have been incorporated. Treating an identical expected-loss quantity as a full asset cash-flow reduction and subtracting it again in full from the adjustment can exaggerate the deduction. Article 53's technical treatment is relevant precisely because the regulatory calculation manages the interaction between these terms.
The risk of a too-small deduction is also important. A model that extrapolates through historically stable markets may be systematically optimistic about a new asset sector or a correlated downturn. Private-lending covenants can produce diverse outcomes, while internal marks and external credit assessments may update at different speeds. Stress testing should therefore vary default frequencies, recoveries, downgrade severity, liquidity discounts, correlation and management-action feasibility jointly. A risk deduction that appears stable under single-variable tests may fail under a coherent adverse economic scenario.
Retained Spread Risk
Defaults and downgrades
A buy-and-hold strategy can reduce the need to trade but cannot eliminate issuer default. Expected nonpayment and tail loss remain part of the asset's economics. Downgrades can raise required capital, worsen marketability, or require portfolio substitution even before default. Risk becomes particularly serious when exposures share economic drivers, collateral markets, or counterparties. The model should consider correlated loss and the distribution of recoveries, not solely expected loss calculated from average historical defaults.
The liquidity value of holding to maturity is state-dependent. An insurer that expected to retain a credit asset may have to sell during an adverse market because of claims, hedge collateral, ratings changes, or recapture of reinsured business. In that state, lower asset prices and higher liquidation costs can occur together. The apparent spread available at purchase is then partly consumed by a forced-sale loss and the costs of maintaining the liability cash-flow match.
Portfolio marks and private credit
Private securities often lack continuous observable transaction prices. Marking models can rely on benchmark curves, comparable loans, valuations of collateral, or stale broker quotes. Apparent price stability can reflect infrequent marks rather than genuinely low economic volatility. An adequate MA analysis should establish the valuation assumptions for private assets independently of the desired liability outcome.
Options require separate modeling. Borrower prepayment can create early cash receipts that must be reinvested at lower yields; extension provisions can delay expected receipts; covenant breaches can change recovery timing; and structured credit waterfalls may behave differently under stress. These risks alter cash-flow replicability directly. An illiquidity premium estimated from unconditional market spreads cannot automatically compensate for such option exposure.
The UK 2025 empirical stress evidence
The PRA's 2025 Life Insurance Stress Test covered eleven major UK life insurers representing more than 90% of bulk purchase annuity liabilities in the exercise's relevant scope. On positions dated December 31, 2024, the prescribed core financial stress reduced aggregate surplus above capital requirements by £8.6 billion; the sector's aggregate SCR coverage ratio moved from 185% to 154%; and £12.9 billion of assets were downgraded below investment grade. Participating firms still met their capital requirements in this scenario. The result supports a bounded observation of resilience under a specified stress, not an estimate of a universal MA premium or a causal effect of MA reform.
The PRA's accompanying MA analysis describes how the mechanism aims to neutralize some short-term spread volatility on eligible obligations, while defaults, downgrades, and other asset-liability shocks continue to affect capital. These findings must not be mechanically generalized to insurers in other jurisdictions or balance-sheet structures.
Financial frictions can also change insurer behavior. Koijen and Yogo (2015) document crisis-period life insurer pricing anomalies and shadow capital costs. Their empirical outcome concerns insurance pricing under funding and regulatory frictions rather than a causal estimate of MA. Its relevance is that valuation rules and financing conditions can jointly affect incentives, even when promised benefits are unchanged.
Stress-consistent control
The principal scenario families are a widening credit spread with unchanged expected defaults; unexpected defaults and downgrades; a market-liquidity freeze; rising derivative collateral demands; longevity deterioration; early policyholder withdrawals; funded reinsurance recapture; and combinations of these events. Each should include changes to assets, projected liability cash flows, eligible own funds, and required capital.
A stress that assumes the insurer never needs to liquidate assets while simultaneously creating large collateral outflows is inconsistent unless credible alternative funding is documented. Likewise, assuming recovery of full promised principal on a defaulted bond while deducting only a small historic expected loss conceals the actual distribution of credit outcomes. Prudential credibility depends on confronting these state-contingent mismatches.
Constructing a dated regulatory concordance
A legal-economic comparison of MA frameworks should distinguish the date a rule was made, the date it entered into force, the date the insurance undertaking must apply it, and the reporting date on which an observation is measured. The EU's amending Directive (EU) 2025/2 illustrates the distinction: its publication and legal entry precede the January 30, 2027 application of the amended prudential regime. A historical report for December 31, 2026 should not be recalculated using a future provision without explicitly labeling the result as a scenario.
The same discipline is needed for technical material. Draft implementing technical standards, adopted implementing acts, supervisory guidelines and national legislation occupy different positions in the regulatory hierarchy. A document may correctly describe supervisory intent without yet imposing a legally effective obligation. The 2026 EIOPA technical report therefore supports analysis of the envisaged reform, while any assertion that its exact draft provisions are operative requires separate evidence of adoption and entry into application. When a technical document is updated, the referenced version and cutoff date belong in the evidence ledger.
From an actuarial perspective the legal concordance should preserve four independent inputs: the permissible asset and liability population, the valuation rate and fundamental-spread calculation, the capital aggregation rules, and the supervisory approval and governance constraints. A change in one dimension can be mistaken for a change in another if analysts report only a net solvency ratio. For example, SCR diversification may alter required capital while leaving the discounted contractual benefits unchanged. Conversely, a changed MA can alter the best estimate with different effects depending on own-funds eligibility or ring-fencing. These mechanisms deserve separate waterfall reconciliations.
Comparative institutional research design
For comparative work, the insurer's legal entity, reporting currency, product mix and reference portfolio must be normalized before considering differences in regulatory outputs. It is especially misleading to compare an EU annuity portfolio and a UK bulk-purchase annuity insurer with different underwriting dates, investment risk and hedging structures, and then attribute the entire observed solvency difference to the MA regime. A valid comparison would hold an illustrative liability cash-flow vector and portfolio exposures constant, apply each framework's actual eligibility and calculation conventions separately, and explain why some portfolio features are permissible in one regime but excluded in the other.
The PRA's supervisory materials permit a version-controlled UK reference. The EU's 2027 changes require an analogous source hierarchy that prioritizes enacted EU law and binding delegated or implementing measures on the relevant date. These disciplined comparisons also clarify why neither IFRS 17 insurance-contract measurement nor IAIS ICS group solvency rules should be substituted for jurisdiction-specific approvals. Their objectives and institutional scopes are relevant to interpretation, but each sets its own calculation boundaries.
2027 EU Regulatory Transition
Legal dates and interpretation
The currently applicable EU architecture derives from Directive 2009/138/EC and its relevant implementing rules. Directive (EU) 2025/2, adopted on November 27, 2024, entered into force in January 2025 but requires national transposition measures by January 29, 2027, to be applied from January 30, 2027. Entry into force is legally distinct from the scheduled application of amended prudential rules. Any valuation comparison must respect the rules operative on the particular reporting date.
The EIOPA revised matching-adjustment ITS report, published July 15, 2026, describes the treatment of MA portfolios under the revised diversification rules. The review permits full diversification between an MA portfolio and the rest of the undertaking in specified circumstances where the relevant best estimate does not form a ring-fenced fund. The report concerns draft technical standards supporting the incoming framework, with adoption and legal applicability requiring separate verification.
EIOPA's July 2026 implementation announcement distinguishes final guidelines from draft implementing technical standards and identifies January 30, 2027 as the transition date. A future capital analysis must separate any change in liability discounting from the different change in risk diversification and SCR aggregation. Combining them into one unlabeled 'MA improvement' would prevent correct attribution of the result.
EU versus UK arrangements
Solvency UK already operates under its 2024 matching-adjustment reforms and the October 2025 MAIA regime. Those provisions are separate from the EU's 2027 amendments. Both address the challenge of long-term guarantees and retained credit risk, but eligibility, deduction rules, governance duties, diversification, and effective dates differ. Regulatory comparisons need instrument-level citations and the relevant date rather than merely a shared name for the adjustment.
Accounting and international supervisory comparison
The IAIS Insurance Capital Standard describes discounting through an adjusted risk-free yield curve, with different asset-based spread adjustments for specified liability buckets. It is an international group-capital framework whose methodology cannot be equated by default to the EU MA. Similarly, IFRS 17 discounting guidance bases measurement on contract cash-flow characteristics and market consistency, allowing bottom-up and top-down approaches. A prudential permission does not determine the financial-reporting liability without applying the separate accounting standard.
| Regime | Main purpose | Liquidity-related mechanism | Principal qualification |
|---|---|---|---|
| Current EU Solvency II | Prudential valuation and capital | Approved MA with fundamental-spread deduction | Eligible assigned portfolios and matching conditions |
| EU rules applying from January 30, 2027 | Revised prudential regime | Amended MA and associated SCR treatment | Effective implementing rules and ring-fencing conditions |
| Solvency UK at the cutoff | UK prudential solvency | MA, highly predictable cash-flow permissions, MAIA | PRA approvals, attestation, risk and liquidity management |
| IFRS 17 | Financial statement measurement | Liability-specific liquidity in discount rates | Market consistency and exclusion of irrelevant factors |
| IAIS ICS | International group-capital measurement | Adjusted yield curves and liability buckets | Specified asset eligibility, risk correction and bucket rules |
The purpose of this comparison is to establish methodological distinctions. It does not assert numerical equivalence between statutory measures or uniform applicability across legal entities.
Stress design with economically coherent variables
The base numerical case is deliberately constructed from a certain terminal payment, a flat risk-free rate, and a scalar deduction. A full model replaces that simplicity with a timeline. At each reporting date the analyst should identify the risk-free term structure, the market value of assigned assets, projected asset cash flows after specified credit assumptions, liability best-estimate cash flows, portfolio-specific MA, eligible own funds, and SCR. Each scenario is then applied consistently across all variables.
One meaningful pair of scenarios separates a pure liquidity disruption from a credit-loss event. In the first, bid–ask spreads widen and observed market prices fall, while modeled contractual survival and expected recovery remain unchanged. An investor capable of retaining the assets may avoid realizing the full trading discount, although fair-value movements and collateral needs persist. In the second, spreads widen because default and recovery prospects deteriorate; cash-flow shortfall and capital impairment can be real even if the firm retains the assets. The first scenario is closest to the stabilizing rationale of the MA; the second tests the reliability of the retained-risk deduction. Real crises often mix the two.
Further scenarios change the liability itself. A mortality or longevity shock changes the expected annuity-payment vector, while a policyholder option can accelerate claims and thereby shorten the time available for asset income. An interest-rate shock may change the market values of investments and obligations, collateral needs, prepayment speeds, and the effectiveness of hedges. A funded-reinsurance recapture event can return assets and liabilities to a cedant under unfavorable market conditions. The PRA's funded reinsurance guidance gives a concrete supervisory setting in which those collateral and recapture relationships must be tested.
These scenarios can be organized as a financial reconciliation: opening eligible own funds plus asset revaluation plus liability revaluation plus credit loss and other adjustments equals closing eligible own funds, subject to accounting and regulatory definitions. The SCR must then be recalculated using the scenario's asset and liability risk concentrations and diversification rules. This explicit accounting prevents a favorable liability-side MA movement from being interpreted without its corresponding investment and capital effects.
A multi-payment extension
Consider two deterministic payments, 40 units after five years and 60 units after ten years. Under the same 2.00% flat risk-free rate, their combined present value is
For an illustrative constant 70-basis-point adjustment, the corresponding value is
The present-value difference is the sum of the separate maturity effects, not a single reduction inferred from the final year ten payment. This model generalizes straightforwardly to monthly annuity cash flows. Its economic use depends on correctly specifying future mortality, expenses, credit, investment cash flows and the applicable currency-specific curve. A common duration statistic summarizes part of the sensitivity but should not replace this time-by-time cash-flow analysis.
Model governance and scientific reproducibility
For every numerical output, the research record should preserve the original input vector, the valuation date, source URL and version, compounding convention, interpolation and extrapolation methods, default and recovery assumptions, cash-flow model, shock design, and the version of any supervisory permission. Rounding belongs only in presentation: calculations should be performed at unrounded precision so that changing the display convention does not silently alter the result. Alternative scenarios should be computed from the same underlying cash-flow data rather than changing the asset portfolio and the liability model without attribution.
Independent reproduction should verify that the adjustment is applied only to eligible cash flows, that retained-risk costs have not been subtracted twice, that the proper curve is used for each currency and maturity, and that stress movements on assets and liabilities share the same economic state. If the exercise is based on confidential insurer data, the public report can still disclose methods, model structure, relevant aggregation and sensitivity conditions while protecting protected customer information. Reproducibility is principally the traceability of assumptions, transformations and calculations.
The boundaries of the analysis are substantive. A deterministic example can demonstrate the mathematical relationship between an interest-rate adjustment and a present value. A sector stress test can document a contingent result for participating firms under a designed shock. Neither establishes a new universal relationship between MA use and default-free economic capital. The purpose of an integrated study is to state precisely what each piece of evidence warrants and how the different pieces constrain the final interpretation.
Valuation and Capital Effects
Explicit assumptions
A hypothetical insurer owes a single fixed payment of 100 currency units at the end of year ten. Assume a flat 2.00% annual effective risk-free rate, a 3.50% equivalent yield for an assigned asset portfolio, and an illustrative 0.80% retained-risk deduction. Ignore taxes, expenses, mortality, options, market-value changes, risk margin, capital requirements, and timing uncertainty in this initial calculation. These artificial values are chosen to make the arithmetic reproducible, not to assert a current regulatory fundamental-spread calibration or establish actual permission.
In the simplified expression, . Thus
The liability value falls by 5.4230 units, or approximately 6.61% of the risk-free figure. The hypothetical 100-unit asset payment discounted at 3.50% has value . That asset value is not identical to the MA-adjusted liability value because the adjustment retains a credit-risk deduction. A valuation comparison should use full asset and liability cash-flow vectors and clarify every discounting convention.
Sensitivity table
Each row below changes one or more explicitly stated assumptions while retaining the ten-year payment and 2.00% risk-free rate.
| Illustration | Asset yield | Fundamental spread | MA | Liability present value | Reduction versus risk-free |
|---|---|---|---|---|---|
| Risk-free baseline | — | — | 0 bp | 82.0348 | 0.0000 |
| Base assumptions | 3.50% | 0.80% | 70 bp | 76.6118 | 5.4230 |
| Larger retained-risk deduction | 3.50% | 1.40% | 10 bp | 81.2349 | 0.7999 |
| Market-spread widening | 4.50% | 1.40% | 110 bp | 73.6908 | 8.3440 |
The final row reveals a critical interpretive hazard. Increasing asset yield by 100 basis points while increasing the fundamental spread by 60 basis points raises the illustrative MA, thereby lowering the liability present value. Yet the market value of the same ten-year 100-unit asset payment falls from 70.8919 at 3.50% to 64.3928 at 4.50%. A higher MA can therefore coincide with a falling asset value and credit deterioration. The change in measured liability value is not a demonstration that the firm has acquired additional wealth.
The third row isolates the deduction effect: the 60-basis-point increase reduces the adjustment from 70 to 10 basis points and the present-value benefit from 5.4230 to 0.7999 units. The result quantifies the dependence of measured benefit on retained-risk attribution.
Payment timing and interest-rate sensitivity
If the same 100-unit payment were due in five years, risk-free value would be , and the value at 2.70% would be , for a difference of 3.0449. Shortening cash-flow duration reduces the adjustment's monetary effect in this example. An unexpected early payment also makes the investment strategy harder to execute if the backing asset is scheduled to pay much later.
For strictly positive future cash flows discounted at a flat adjusted rate, differentiation gives
The sensitivity depends on timing and rate level; convexity means finite rate increases and decreases need not have equal absolute value effects. This mathematical property is a valuation statement, independent of claims about how default, liquidity, or solvency requirements will behave under the same scenario.
Solvency reconstruction
Define a simplified solvency coverage ratio . To infer the effect of MA on , one must recalculate the same-scenario changes in eligible own funds and SCR. The reconciliation should include current asset prices, actual or expected defaults, migration of credit ratings, recovery values, liability cash-flow revisions, the revised MA and fundamental spread, hedge collateral, funding costs, management actions, and any risk-margin or diversification effects.
The PRA's empirical stress-test report demonstrates that joint balance-sheet modeling is essential. Its aggregate 185%-to-154% coverage movement belongs to the specific 2025 UK exercise and is not a universal sensitivity coefficient for MA. Likewise, the numerical table above is a reproducible accounting experiment, not evidence of an insurer's realized capital ratio.
Research interpretation and scientific contribution
The evidence supports five integrated findings. Liability illiquidity must be established from payment and option characteristics rather than a distant contractual maturity alone. An illiquid asset's spread is heterogeneous and may conceal credit, optionality and market risk. The fundamental spread is a prudential retained-risk control rather than an observable pure liquidity price. Cash-flow matching and sustainable liquidity management jointly support the capacity to earn return through maturity. Finally, a higher liability discount rate changes measured best estimates while a valid capital conclusion requires simultaneous asset and capital revaluation.
The scientific contribution is the systematic reconciliation of these financial, actuarial, accounting, and supervisory distinctions with a fully replicable numerical example and a dated EU/UK regulatory comparison. The sources do not support an assertion of priority for a new theory, model family, principle, or doctrine, and none is claimed.
Conclusion
The matching adjustment is a conditional mechanism for reflecting part of the yield from qualifying assigned assets in the valuation of specific long-term insurance liabilities. Its economic foundation is cash-flow matching and the insurer's ability to retain assets through market volatility while meeting policyholder commitments. Its retained-risk deduction and supervisory restrictions seek to prevent risky asset yields from being treated as unconditional resources for policyholder payments.
An asset-market illiquidity premium, the liquidity characteristics of an insurance obligation, and the legally permitted matching adjustment are separate quantities. The ten-year example establishes that modest discount-rate changes materially affect measured liabilities, while its spread-widening case establishes why the resulting liability change cannot stand alone as evidence of stronger solvency. Default, downgrade, forced-sale, collateral, option, concentration, and reinsurance risks remain relevant to the institution's actual capacity to pay.
The EU's reforms applicable from January 30, 2027 must be distinguished from the current EU regime and from the UK's operative 2024–2025 framework. IFRS 17 and the IAIS ICS provide additional perspectives through their own measurement objectives. Robust analysis ties the adjustment to a dated legal framework, eligible cash-flow portfolios, coherent retained-risk assumptions, and stress-consistent reconstruction of both sides of the insurance balance sheet. These are the conditions under which a valuation adjustment acquires credible economic interpretation.
References
Bao, Jack, Jun Pan, and Jiang Wang. 2011. “The Illiquidity of Corporate Bonds.” The Journal of Finance 66 (3): 911–946. https://doi.org/10.1111/j.1540-6261.2011.01655.x
Bank of England, Prudential Regulation Authority. 2018, updated 2025. “SS7/18 — Solvency II: Matching Adjustment.” October 2025 current version, effective October 27, 2025; future version listed for December 31, 2026. https://www.bankofengland.co.uk/prudential-regulation/publication/2018/solvency-2-matching-adjustment-ss
Bank of England, Prudential Regulation Authority. 2024. “PS10/24 — Review of Solvency II: Reform of the Matching Adjustment.” June 6, 2024. https://www.bankofengland.co.uk/prudential-regulation/publication/2024/june/review-of-solvency-ii-reform-of-the-matching-adjustment-policy-statement
Bank of England, Prudential Regulation Authority. 2024, updated 2025. “SS5/24 — Funded Reinsurance.” July 26, 2024; current version effective October 27, 2025. https://www.bankofengland.co.uk/prudential-regulation/publication/2024/july/funded-reinsurance-implementation-approach
Bank of England, Prudential Regulation Authority. 2025. “PS17/25 — Matching Adjustment Investment Accelerator.” October 23, 2025. https://www.bankofengland.co.uk/prudential-regulation/publication/2025/october/matching-adjustment-investment-accelerator
Bank of England, Prudential Regulation Authority. 2025. “Life Insurance Stress Test: 2025 Results.” November 17, 2025. https://www.bankofengland.co.uk/prudential-regulation/publication/2025/november/list-2025-results-report
Chen, Long, David A. Lesmond, and Jason Wei. 2007. “Corporate Yield Spreads and Bond Liquidity.” The Journal of Finance 62 (1): 119–149. https://doi.org/10.1111/j.1540-6261.2007.01203.x
Dick-Nielsen, Jens, Peter Feldhütter, and David Lando. 2012. “Corporate Bond Liquidity Before and After the Onset of the Subprime Crisis.” Journal of Financial Economics 103 (3): 471–492. https://research-api.cbs.dk/ws/files/58847171/corporate_bon_liquidity_david_lando.pdf
European Commission. 2025. “Questions and Answers on the Solvency II Delegated Regulation.” October 29, 2025. https://finance.ec.europa.eu/news/questions-and-answers-solvency-ii-delegated-regulation-2025-10-29_en
European Insurance and Occupational Pensions Authority (EIOPA). n.d. “Matching Adjustment to the Relevant Risk-Free Interest Rate Term Structure: Article 77b.” Solvency II Single Rulebook. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2162_en
European Insurance and Occupational Pensions Authority (EIOPA). n.d. “Calculation of the Matching Adjustment: Article 77c.” Solvency II Single Rulebook. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2163_en
European Insurance and Occupational Pensions Authority (EIOPA). n.d. “Calculation of the Matching Adjustment: Article 53.” Solvency II Single Rulebook, Delegated Regulation (EU) 2015/35. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2466_en
European Insurance and Occupational Pensions Authority (EIOPA). n.d. “Calculation of the Fundamental Spread: Article 54.” Solvency II Single Rulebook, Delegated Regulation (EU) 2015/35. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2467_en
European Insurance and Occupational Pensions Authority (EIOPA). 2026. “Final Report on Revised ITS on the Matching Adjustments — Solvency II Review.” July 15, 2026. https://www.eiopa.europa.eu/publications/final-report-revised-its-matching-adjustments-solvency-ii-review_en
European Insurance and Occupational Pensions Authority (EIOPA). 2026. “EIOPA Completes Solvency II Review Mandate with Final Guidelines and Draft Technical Standards.” July 15, 2026. https://www.eiopa.europa.eu/eiopa-completes-solvency-ii-review-mandate-final-guidelines-and-draft-technical-standards-revised-2026-07-15_en
European Parliament and Council of the European Union. 2009, consolidated January 17, 2025. Directive 2009/138/EC on the Taking-Up and Pursuit of the Business of Insurance and Reinsurance (Solvency II). https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A02009L0138-20250117
European Parliament and Council of the European Union. 2025. Directive (EU) 2025/2 of November 27, 2024 Amending Solvency II. Official Journal of the European Union, January 8, 2025. https://eur-lex.europa.eu/eli/dir/2025/2/oj/eng
IFRS Foundation, International Accounting Standards Board. 2017, amended 2020. IFRS 17 Insurance Contracts, including paragraphs 36 and B72–B85. IFRS Foundation text incorporating June 2020 amendments. https://www.ifrs.org/content/dam/ifrs/project/amendments-to-ifrs-17/ifrs-17-incorporating-the-june-2020-amendments.pdf
International Association of Insurance Supervisors (IAIS). 2024. Insurance Capital Standard, Level 1 and Level 2 Text, Section 3.2.5, “Discounting.” IAIS online standard tool. https://www.iais.org/ics-online-tool/
Koijen, Ralph S. J., and Motohiro Yogo. 2015. “The Cost of Financial Frictions for Life Insurers.” American Economic Review 105 (1): 445–475. https://www.aeaweb.org/articles?id=10.1257%2Faer.20121036
Longstaff, Francis A. 2004. “The Flight-to-Liquidity Premium in U.S. Treasury Bond Prices.” The Journal of Business 77 (3): 511–526. https://www.jstor.org/stable/10.1086/386528