RESEARCH DOSSIER
Insurance Asset-Liability Matching
Abstract. Insurance asset-liability matching is the deliberate coordination of the amount, timing, currency, conditionality, market sensitivity, and legal availability of investments with an insurer's contractual and actuarial obligations. This monograph investigates matching as a problem in contingent cash-flow finance, with particular attention to liability projection, term-structure risk, duration and convexity, immunization, bond selection, longevity, policyholder options, reinvestment, liquidity, credit, capital, accounting, and international supervisory practice. It combines actuarial derivation, critical comparison of historical immunization research, primary-source analysis of insurance standards, institutional evidence through November 14, 2026, and transparent numerical illustrations. It distinguishes cash-flow replication from local interest-rate immunization and economic surplus stability from immediate settlement capacity. The analysis shows why successful matching is conditional on valuation basis, cash-flow assumptions, admissible assets, and stress horizon. It also examines how surrender behavior, collateral, credit deterioration, and changing asset eligibility can defeat an apparently satisfactory baseline hedge. International comparison establishes a shared supervisory emphasis on liability-informed investments while explaining divergent statutory and prudential structures in the United States, European Union, and United Kingdom. The author's contribution is an integrated, evidence-grounded interpretation and explicitly reproducible analytical exposition, without a claim to a new theoretical doctrine. The manuscript is a research reference, not a recommendation concerning any particular insurer or portfolio.
Keywords: insurance asset-liability matching; insurance asset-liability management; contingent liabilities; duration; modified duration; convexity; key-rate duration; cash-flow matching; immunization; reinvestment risk; longevity; lapse risk; surrender options; annuities; insurance liquidity; asset adequacy; asset-liability scenarios; Solvency II; matching adjustment; Solvency UK; IFRS 17; IFRS 9; IAIS; NAIC; actuarial standards.
Introduction and Research Object
Insurance creates obligations whose settlement dates frequently differ from the dates on which premiums are received and invested. A fixed annuity payment may be scheduled decades ahead, yet its occurrence can depend on survival. A life insurance surrender can accelerate payment when market rates change. Catastrophe losses can generate clustered claims before reinsurance recoveries arrive. The insurer's asset portfolio is therefore a financing instrument for obligations that are simultaneously financial, actuarial, contractual, and institutionally constrained.
Insurance asset-liability matching is the process of selecting, structuring, monitoring, and adjusting asset cash flows and exposures in relation to the insurer's liability cash flows and exposures. The word matching describes an objective with several distinct meanings. In deterministic cash-flow matching, asset receipts are timed to fund specified payments. In economic-value matching, asset and liability values respond similarly to particular market shocks. In liquidity matching, cash or readily mobilized resources are available when settlement occurs. In regulatory matching, portfolios meet the eligibility and valuation conditions of a jurisdiction. These meanings overlap, but a result on one dimension does not establish results on the others.
The conceptual starting point is the IAIS asset-liability management requirement, especially ICP 16.5. The international standard requires an explicit ALM policy and assessment of liability cash flows, asset inflows, and economic values under different scenarios. It also makes clear that the task is effective management of mismatches rather than literal elimination of every mismatch. This distinction matters because insurers must preserve liquidity, diversification, return capacity, regulatory compliance, and the resources needed to support future insurance business.
This dossier concentrates on matching itself. General investment optimization, group finance, corporate capital structure, and insurance accounting receive analysis where they alter the feasibility or interpretation of matching. Its principal intellectual question is how an insurer can preserve the capacity to honor state-dependent promises when asset cash flows, liability cash flows, and market prices all evolve. The answer requires identifying which risks are hedgeable, which are subject to estimation, which can be retained with explicit capital, and which reflect the legal architecture of insurance protection.
Method, Scope, and Evidentiary Hierarchy
The investigation uses four mutually supporting methods. First, it derives local present-value sensitivities and cash-flow constraints from transparent assumptions. Second, it examines the foundational immunization literature, including Redington's 1952 valuation study and Fisher and Weil's 1971 paper. Third, it reads primary insurance rules and actuarial standards directly. Fourth, it interprets institutional reports with explicit attention to reference dates, aggregation, and the difference between a reported statistic and a causal inference.
Sources are classified by what they can establish. Statutes and supervisory standards establish requirements within their own legal or professional scope. International principles establish supervisory expectations and comparative norms rather than immediately enforceable obligations in every jurisdiction. Accounting standard-setter materials establish measurement and presentation under the relevant financial reporting system. Historical studies establish a documented intellectual contribution within their original assumptions. Institutional monitoring data describe sampled populations and measurement conventions; they do not establish universal properties of all insurers. The numerical scenarios below are original stylized calculations, not observed balance sheets.
The research uses a November 14, 2026 information cutoff. An effective rule is distinguished from a published future amendment. In particular, the PRA's SS7/18 version history identifies an October 2025 version effective October 27, 2025, and a July 2026 version scheduled to take effect December 31, 2026. The European Solvency II review also includes measures scheduled for January 30, 2027. None is treated as universally operative before its stated date.
This is a comparative, analytical research monograph. It makes no empirical claims based on an undisclosed original sample, no proprietary industry survey, and no claim of statistical identification where source data support only description. All derived examples disclose their inputs so their arithmetic can be reproduced independently.
Liability Cash-Flow Characteristics
The liability as a state-dependent payment
Let L(t,ω) denote the insurer's net obligation at time t in economic state ω. The payment may incorporate benefits, claims, expenses, surrender values, reinsurance recoveries, and contractual participation. The corresponding expected cash flow is E[L(t,ω)] under an explicitly selected probability measure. For financial valuation, the measure and discount factors must remain consistent with the applicable measurement framework; historical or best-estimate probabilities cannot automatically substitute for market-implied financial prices.
Insurance liabilities have five coordinates central to matching: timing, amount, currency, optionality, and contingency. Two portfolios with equal expected present values may differ in each coordinate. A bond portfolio paying 100 at year 10 can fund an unconditional 100 payment at that date, yet will not replicate an annuity that pays 10 annually for as long as a policyholder survives. Equal expected values are weaker than equal state-contingent payments.
The IFRS 17 measurement overview describes fulfillment cash flows as risk-adjusted present values of estimates of future cash flows, supplemented by a contractual service margin for relevant groups. That financial reporting representation is distinct from the physical cash movements that treasury must actually finance. ALM projections may use different cash-flow granularity and stress assumptions from accounting group measurements.
Life, annuity, and non-life distinctions
Term life benefits are often concentrated in mortality-contingent payments; the incidence and timing change with age, underwriting, catastrophe mortality, and coverage lapses. Traditional payout annuities generate survival-contingent streams and can lengthen materially when longevity improves. Deferred annuities combine accumulation, election, guarantee, and payout options. Participating policies introduce discretion governed by contractual and regulatory constraints. Unit-linked products pass through parts of market risk while preserving insurer exposures to guaranteed benefits and servicing costs.
Non-life obligations may be shorter in nominal duration but can involve long settlement tails, judicial uncertainty, medical inflation, remediation expenses, currency translation, and catastrophe clustering. Assets matched to the mean non-life payment date may still fail to meet a severe early claims accumulation. The relevant matching objective depends on the contractual cash-flow distribution, not on the label of the insurance line.
Gross, ceded, and net obligations
Reinsurance changes claim amounts, timing, credit exposure, and recoverability. Gross claims must be paid according to policy obligations even if a reinsurer's reimbursement arrives later or becomes disputed. Net present-value matching can conceal interim cash shortfalls. A robust projection therefore preserves gross payments, ceded recoveries, collateral, counterparty conditions, and their date-by-date reconciliation.
Legal availability and segmentation
The same consolidated insurer group may hold assets in legally separate entities or restricted funds. Funds supporting a block of annuities may not be available for another entity's catastrophe claims. Matching calculations should begin at the level at which cash can lawfully be accessed, then aggregate only to the extent that transferability is supported by the legal and supervisory structure. Economically correlated assets are not automatically legally fungible assets.
Actuarial Projection of Contingent Insurance Obligations
Projection design
An effective projection records policy-level or homogeneous cohort data, coverage features, premiums, expected benefits, expenses, taxes where relevant, reinsurance, and management actions. For an annuity issued at age x, a simplified expected payment at time t is B(t) × p(x,t), where p(x,t) is the modeled probability of survival until payment and B(t) is the benefit conditional on survival. If mortality depends on economic or health states, the joint distribution rather than a standalone survival curve may be necessary.
For surrenderable life and annuity products, expected liability flows must be modeled as functions of account value, guaranteed rates, market yields, charges, tax rules where material, and policyholder characteristics. The timing of a surrender can shorten the obligation while increasing immediate cash demand. Historical lapse averages can understate this interaction during a sustained interest-rate shock.
The revised U.S. ASOP No. 7 cash-flow standard, effective June 1, 2026, addresses asset and liability cash-flow projections, scenarios, documentation, and actuarial judgment for life and health cash-flow analysis. This source is particularly relevant to model design. Its revised scope excludes property/casualty actuarial services, which must not be inferred to be covered by the life-and-health standard.
Time buckets and tails
Projection systems commonly use monthly periods during early policy years and longer intervals farther into the future. Such aggregation is a numerical approximation. It can conceal intra-period settlement mismatches and short-notice liquidity needs. A 12-month bucket showing equal annual inflows and outflows may mask claims due in January against coupons received in December. Tail periods require particular attention when very long life benefits extend beyond the observable bond market.
Forecasting should reconcile deterministic expected payments with stochastic percentiles and contingent tail events. The expected annual annuity cash flow is useful for funding a base portfolio, but capital and liquidity resources must address plausible realizations above it. Long-horizon extrapolation of mortality, credit, inflation, and market rates creates structural uncertainty that requires explicit sensitivity analysis.
Dependency and validation
Liability projection is not separable from the investment strategy when contract features respond to financial conditions. Crediting rates, bonus decisions, policyholder behavior, reinvestment income, and reinsurance provisions may depend on the same economic path. Model validation should compare observed and expected lapse and claim experience, examine discontinuities around contractual thresholds, and document judgment used in data-poor tails.
The ASOP No. 22 asset adequacy guidance requires attention to supporting assets and moderately adverse conditions for the particular actuarial opinion within its scope. This illustrates the difference between making a base-case cash-flow forecast and evaluating the adequacy of the financial support for that forecast.
Discounting and the Term Structure
Present value and the valuation convention
For positive, deterministic payments C_i at dates t_i, define their present value under continuously compounded spot rates z(t_i) as
When the payments are uncertain, the valuation must also specify how financial risk, insurance risk, and contractual optionality are reflected. A single constant discount rate can be a useful teaching device but is generally insufficient for a term structure with changing forward rates, spread risk, and contingent insurance payments.
Matching requires consistency of currency, timing, compounding, credit and liquidity assumptions, and the economic meaning of the discount curve. The asset yield-to-maturity of a credit-risky bond is not automatically an appropriate liability discount rate. It includes expected and unexpected credit-related compensation and possibly liquidity and technical market effects that need separate treatment under the applicable framework.
Parallel shifts and their limits
A common first-order approximation assumes an instantaneous small parallel yield-curve shock Δy. Under this assumption, a positive-cash-flow value changes by approximately −D_mod V Δy, where D_mod is modified duration. A term structure shock is rarely perfectly parallel, and the economic difference between assets and liabilities may depend on several key rate points.
Asset and liability spread responses can differ even when their government yield sensitivities appear equal. Corporate bond prices may fall when credit spreads widen while liability discount curves under a given accounting or solvency regime respond differently. Therefore a portfolio with matched interest-rate duration remains exposed to credit spread basis unless another valid hedge offsets it.
Financial reporting versus economic curves
IFRS 17 uses current estimates and relevant discount rates in contract measurement and separates insurance finance effects according to applicable presentation policies. IFRS 9 classification requirements determine whether financial assets fall under amortized cost, fair value through other comprehensive income, or fair value through profit or loss according to the relevant business model and contractual cash-flow characteristics. Changes in financial statement equity or profit may therefore differ from the change in economic asset-liability surplus.
The difference is important for governance. A reported accounting mismatch can reflect a genuine economic exposure, a measurement convention, or both. An ALM committee must understand the mapping rather than optimize exclusively for a reported line item.
Duration, Convexity, and Surplus Sensitivity
Time-weighted sensitivity
For a flat continuously compounded yield y and deterministic cash flows C_i,
Under the specified convention, D is both the value-weighted average payment time and the local modified-duration measure. With annual or semiannual compounding, modified duration and Macaulay duration differ by the appropriate yield-conversion factor. An insurer must state the convention before combining duration figures from different systems.
Define convexity under the same convention as
The second-order approximation is
These expressions are exact local derivatives for the stipulated deterministic cash flows. Their use as finite-shock predictions is approximate. In particular, callable assets and surrenderable liabilities can change cash flows when rates change; a derivative based on frozen cash flows omits this response.
Economic surplus
Let asset value be A, liability value L, and surplus S=A−L. To first order,
The dollar-duration mismatch A D_A−L D_L determines sensitivity of surplus under a common small parallel shock. Matching percentage durations alone does not suffice if A and L differ significantly. If A=120, L=100, D_A=6, and D_L=8, the asset dollar duration is 720 and liability dollar duration is 800; a one-percentage-point rate increase produces approximately +0.8 of surplus improvement in the specified currency units. This is because the liability value falls by more than the asset value. The sign reverses when rates decline.
The second-order surplus term is one half of (A K_A−L K_L)(Δy)^2. Its contribution can matter for larger shocks. When the asset and liability portfolios have embedded options, their effective convexities may change sharply with the rate level, reducing the reliability of a stationary quadratic approximation.
Risk measures and units
DV01 is the local value change associated with a one-basis-point rate move; it must specify both the shock direction and whether it describes an asset, liability, or surplus. For continuous-yield flat shocks, the magnitude of DV01 for a positive-cash-flow position is approximately V D × 0.0001. Dollar duration and DV01 permit coherent aggregation across positions in the same currency and under the same curve convention. Percentage durations do not.
The insurer's central problem is the net distribution of changes in surplus and funding, not an isolated bond-duration target.
Immunization and the Limits of Local Matching
Historical foundations
F. M. Redington's Review of the Principles of Life-Office Valuations developed an influential actuarial approach to protecting the valuation of a liability portfolio against small interest-rate movements. The classical conditions are commonly summarized as equality of present values, equality of first interest-rate sensitivities, and favorable second-order behavior of the asset-liability surplus. These are local conditions around a chosen valuation rate, not a universal claim that all future payments are secured.
Fisher and Weil (1971) examined the risk of interest-rate fluctuations and investment strategies under more general yield-curve assumptions. Their contribution highlights the relevance of discount factors across maturities and the limitations of methods that assume one yield for all cash flows. The historical papers should be read as foundational analyses with stated assumptions, not as empirical demonstrations of contemporary insurance solvency.
The local immunization argument
Suppose S(y)=A(y)−L(y). A stationary local surplus at y_0 satisfies S'(y_0)=0, and a locally convex surplus satisfies S''(y_0)≥0. If S(y_0) is adequate and cash-flow assumptions remain fixed, the first two conditions protect the modeled position against small movements around y_0 to second order. The guarantee disappears when rates move far enough for higher-order terms, policyholder exercise, asset calls, defaults, or rebalancing constraints to dominate.
Immunization also requires maintenance. Coupons are paid, liabilities run off, yields move, and duration changes as cash-flow time horizons shorten. A hedge that met first-order constraints on one valuation date can become misaligned later even without trading. Governance must specify rebalancing thresholds, trading capacity, and the cost of maintaining the hedge.
Cash-flow dedication versus immunization
A dedicated portfolio whose fixed, default-free, same-currency receipts cover each noncontingent liability payment can provide stronger nominal payment-date assurance than duration immunization under its stated assumptions. Even this idealization does not establish protection against inflation, counterparty default, issuer optionality, legal restrictions, or contingent liabilities absent from the target schedule. Actual insurers generally combine dedicated cash-flow layers, duration-based overlays, liquidity reserves, and retained risk capital.
The degree of matching should be set by the liability's structure and the balance-sheet consequences of residual mismatch. A narrow numeric duration gap can coexist with high risk if cash-flow concentration, convexity, optionality, or spread exposure remains unmeasured.
Yield-Curve Shape, Key Rates, and Basis Risk
Nonparallel shocks
A term structure has many independent dimensions. A steepening shock can raise long rates while leaving short rates nearly unchanged; a flattening can do the opposite. A single duration measure collapses exposures at separate maturities into one number, thereby hiding offsetting positions. For insurer liabilities concentrated at 20–30 years, a portfolio of short and intermediate bonds can reproduce a broad duration target while leaving the longest maturities underhedged.
Approximate the present-value change with a vector of key-rate movements:
Here KRD_k is the sensitivity to a defined perturbation of node k, with interpolation rules held fixed. The definition of nodes, curve construction, and interpolation must be identical when comparing assets with liabilities. The sum of key-rate durations need not mechanically equal a separately calculated parallel-shift duration if methodologies or shocks differ.
Basis, spread, and curve construction
A portfolio can hedge Treasury or swap-rate movements while retaining sensitivity to corporate spreads, inflation expectations, currency basis, collateral discounting, and liquidity premia. A liability measured using a prescribed risk-free curve with adjustments may respond differently from the securities held to finance it. The relevant hedge is therefore multidimensional and valuation-basis-specific.
Basis exposures can be significant when hedges use swap curves but valuations depend on government curves, or when asset coupons reference floating rates that reset differently from the insurer's crediting rates. Cross-currency hedges require examination of principal exchange, collateral requirements, and basis costs throughout the contract.
Optionality and effective sensitivities
For an option-bearing contract, duration is usually estimated by repricing the contingent cash flows under positive and negative curve perturbations with a consistent behavior model. The result is an effective sensitivity conditional on that model. A fixed-income bond with an issuer call may shorten when yields fall, just when a long insurance guarantee becomes more expensive. A policyholder's surrender option may alter liability duration as yields rise, making the asset-liability response asymmetric.
Consequently, a credible key-rate hedge must be tested jointly with option exercise and transaction feasibility rather than certified solely on an initial gradient.
Cash-Flow Matching and Asset Selection
The deterministic cash-flow problem
Discretize settlement into periods t=1,…,T. Let c_jt be cash provided at t by one unit of asset j, and let x_j be holdings. A strict gross cash-flow match seeks
subject to budget, eligibility, concentration, credit, currency, and legal constraints. A more realistic formulation incorporates beginning cash balances, coupons reinvested, asset sales, transaction costs, collateral, and contingent liability ranges. Whether an inequality can be satisfied depends on the market's maturity supply and the insurer's existing asset base.
Using expected asset receipts can produce a misleadingly cheap solution when cash flows depend on default, prepayment, or call decisions. Asset cash flows need a state-dependent representation where risks are material. An expected coupon on a risky bond does not guarantee the liquidity of that coupon on the liability's payment date.
Asset classes and contractual structure
Sovereign and high-grade corporate bonds commonly provide predictable nominal cash flows, subject to credit and market risk. Inflation-linked bonds can address specified indexed benefits but may introduce index lag and basis exposure. Amortizing loans and structured credit can align with annuity cash flows if prepayments, defaults, and embedded options are modeled. Private credit and infrastructure debt may offer long contractual cash-flow horizons yet require particular scrutiny of borrower performance, valuation, collateral, covenants, and realization under stress.
The IAIS investment standards require asset selection appropriate to the nature and duration of liabilities and attention to security, liquidity, diversification, and concentration. An attractive spread is not evidence of matching quality if it comes with the ability of a borrower to defer payment, an unreliable valuation, or a liquidity horizon inconsistent with promised benefits.
Diversification and the limits of dedication
Dedication to specific liabilities can increase certainty for those payments, but excessive concentration in a sector or security weakens resilience if default clusters occur. The insurer must compare concentration risk with the value of more precise timing. A portfolio with near-perfect nominal maturity matching might be less robust than a diversified portfolio with modest deterministic mismatch and sufficient liquid support.
Securities lending, repo, derivative collateral, and asset encumbrance affect the portion of an asset's value actually available to fund the target block. Assets pledged to one obligation cannot be counted in full as immediately available for unrelated claims. Legal-entity and eligibility filters belong inside the cash-flow matching problem rather than being applied only after optimization.
Dynamic portfolio maintenance
Matching changes as assets mature, claims occur, premiums arrive, and new contracts are written. A prudent reinvestment rule specifies the eligible universe, concentration limits, expected settlement needs, and the balance between new asset purchases and hedge instruments. Transaction costs and tax effects where relevant can make exact daily rebalancing economically inefficient. A policy should define tolerable residual risk, not an unattainable condition of permanent zero gap.
Reinvestment and Credit-Performance Uncertainty
Reinvestment risk
Reinvestment risk arises when receipts from assets arrive before the date of liability payment and must earn an unknown future rate. A portfolio that appears to have sufficient present value at inception may fall short if interim receipts can be reinvested only at lower yields. Conversely, rising yields can increase future reinvestment income but reduce the market value of existing securities and raise near-term surrender demand.
The crucial relationship is joint: the insurer cannot assess reinvestment risk without specifying the schedule of receipts, the schedule of obligations, and plausible investment rates. A maturity ladder with numerous near-term maturities financing distant guarantees is structurally exposed to sustained lower rates. Cash-flow dedication reduces this exposure only to the extent that actual receipts align with payments under all relevant states.
Credit, downgrade, and liquidity migration
Credit deterioration changes more than present value. A downgrade can alter statutory capital, asset eligibility, collateral haircuts, and required asset sales. Defaults can delay or eliminate promised receipts; restructurings can change timing. In a stressed market, prices may reflect both expected loss and reduced trading liquidity. If the liability schedule is rigid, the need to realize losses in order to fund benefits can intensify the solvency impact of credit stress.
The 2025 Global Insurance Market Report places private credit, geoeconomic fragmentation, liquidity, and market-risk transmission within international insurance monitoring. Its aggregate observations inform scenario selection but cannot substitute for security-specific default and recovery analysis. A matching portfolio should have issuer-level, group-level, and sector-level concentration limits anchored in the balance sheet's loss-absorption capacity.
Inflation and currency
Nominal bonds provide a poor real hedge for benefits explicitly linked to price inflation unless inflation exposure is separately addressed. Medical claims can grow differently from consumer-price indices. A foreign-currency bond may appear attractive on coupon yield but introduces exchange-rate sensitivity unless benefits are denominated in the same currency or a credible hedge remains effective. Cross-currency derivatives then generate collateral and counterparty risks that must be included in cash projections.
The economic object is the payment in the currency and purchasing-power unit required by the policy, not its reporting-currency equivalent on the day the asset was purchased.
Longevity, Mortality, and Insurance Guarantees
Longevity as a duration transformation
Greater-than-projected survival extends annuity payments and can increase the value of long-term obligations. It can also shift the cash-flow distribution toward later maturities, thereby changing duration even if the interest-rate curve remains fixed. Assets chosen to match an initial expected survival schedule may fail to cover the extended tail. A deterministic longevity stress must therefore reproject the full cash flows before recalculating duration and capital.
Longevity risk cannot generally be eliminated by holding longer bonds alone. A long bond provides cash at specified dates, whereas the magnitude of annuity payments remains survival-contingent. Hedging through reinsurance or market instruments may reduce risk, but basis risk between the insured population and the hedge reference remains.
Mortality and mixed liability portfolios
Mortality risk often creates the reverse timing effect for traditional death benefits, with adverse mortality increasing near-term death claims. A composite life and annuity insurer may obtain some diversification, but the offset depends on business mix, epidemic risk, population differences, and the timing of payments. Correlation assumptions should be stress tested because adverse mortality and market disruption can coincide.
Guarantees as option structures
Minimum-crediting-rate guarantees, guaranteed annuity conversion rates, and minimum withdrawal benefits create financial exposures whose value can increase when market conditions become unfavorable to the insurer. The hedge ratio changes with rates, account values, and time. Portfolio-level effective duration calculated under a central scenario can conceal extreme outcomes when policyholders exercise options collectively.
The insurer should separate the market risk of guarantees, mortality and survival risk, behavioral exercise risk, and operational obligations. Their interaction affects the total funding requirement, but each demands distinct evidence and control mechanisms.
Lapses, Surrenders, and Dynamic Policyholder Behavior
Why lapse assumptions are financial assumptions
Policyholders compare the value of maintaining a contract with alternative uses of funds, subject to surrender charges, tax conditions, personal circumstances, liquidity needs, and product features. The insurer's lapse model must account for this incentive structure. A rise in market yields can simultaneously reduce the value of existing fixed-rate bonds and encourage withdrawals from lower-crediting-rate contracts. The resulting need to sell assets can convert an apparently temporary mark-to-market loss into a realized funding shortfall.
The NAIC interest-rate analysis explicitly connects rising rates with asset sale pressure and policy lapses or annuity withdrawals. The source supports the mechanism of risk; it does not imply that every life insurer or product experiences the same magnitude of withdrawals.
Path dependency and adverse selection
Lapse behavior may depend on the cumulative rate gap, prior market experience, policyholder communication, distribution channel, and the persistence of the shock. A one-quarter rate increase and a sustained two-year rate increase can produce different surrender patterns even if their final market rate is identical. Products with surrender charges, market-value adjustments, or longer lockups may be less immediately runnable, but contractual and conduct obligations constrain management responses.
Averages across heterogeneous policyholders can obscure behavior of financially sophisticated or highly concentrated cohorts. Insurers should test whether their liabilities are especially sensitive to the actions of a small group of institutional policyholders. The liquidity stress horizon may be shorter than the portfolio's average contractual life.
Feedback from portfolio actions
Management may sell liquid holdings first to meet surrenders, leaving a progressively less liquid and more concentrated asset portfolio. Later withdrawals become costlier, while credit concerns or external news can further influence policyholder actions. This creates feedback between liquidity, portfolio composition, valuation, and behavior. The feedback should be modeled as a scenario mechanism and validated against relevant experience, rather than treated as an asserted universal outcome.
Liquidity, Collateral, and the Settlement Horizon
Solvency and payment are different tests
Economic surplus answers whether modeled assets exceed modeled liabilities on a valuation basis. Liquidity answers whether qualifying cash can be delivered on each due date. A solvent insurer can be forced to realize losses if it has a large obligation before expected asset cash flows arrive. Conversely, an insurer with immediate cash may still have insufficient resources to support distant guarantees.
A practical funding projection tracks beginning liquid balance Q_t, asset receipts C^A_t, premium income P_t, claims and benefits C^L_t, collateral or margin M_t, and accessible financing F_t:
where T_t denotes transaction and other relevant disbursements. This is an accounting identity under the modeled cash categories. A feasible path must maintain specified minimum liquidity and legal eligibility throughout the horizon. F_t must reflect committed and executable funding, not an assumed emergency facility without evidence.
Derivative collateral
Swaps and options can reduce market-value mismatch yet create collateral calls when the hedge moves against the insurer. Those payments may occur before an offsetting benefit is realized in the valuation of long-term insurance liabilities. A derivative strategy therefore requires two linked assessments: economic risk reduction and liquidity capacity under margin stress. Hedge accounting, when applicable, does not supply the cash for collateral calls.
Stress liquidity and asset sales
Liquidity analysis should examine the settlement schedule of securities, market depth, bid-ask costs, repo access, collateral haircuts, and concentration of maturities. Illiquid investments can be appropriate for genuinely illiquid long-term obligations when sufficient liquidity is held elsewhere. The insurer should not infer that life insurance liabilities are universally stable; surrender provisions, derivatives, and reinsurance collateral can create fast-moving cash needs.
The IAIS 2025 liquidity analysis reports liquidity indicators from its monitoring exercise and identifies surrender values, claims, expenses, and funding-related requirements among key liquidity needs. This offers empirical support for modeling liquidity separately from duration. Aggregate indicator strength should not be interpreted as proof that a particular insurer can meet an idiosyncratic run.
Capital, Accounting, and Measurement Regimes
Economic balance sheet and capital
Capital absorbs losses that remain after matching and hedging. An insurer's interest-rate position cannot be evaluated independently of credit, underwriting, operational, counterparty, and liquidity risks. Diversification reduces risk only to the extent supported by the modeled dependence structure and legally available capital. The IAIS ICPs and ComFrame connect investment rules, enterprise risk management, valuation, and solvency capital within the global supervisory architecture.
Economic surplus can be expressed as S=A−L for a selected measurement basis. Under stochastic scenarios S(ω) is a distribution, and capital adequacy depends on the regulatory or internal risk measure, available own funds, and relevant constraints. The same initial duration gap can produce different capital implications under alternative assets, policy types, or regimes.
Financial reporting mismatch
Under the IFRS 9 financial asset framework, qualifying asset holdings may be reported at amortized cost or at fair value through other comprehensive income, or at fair value through profit or loss, depending on contractual cash-flow features and business model. IFRS 17 insurance reporting uses current measurement of fulfillment cash flows and specifies insurance finance presentation. These frameworks may produce accounting sensitivities that differ from the economic change in surplus.
A responsible ALM report includes a reconciliation from economic measures to the relevant financial statement and regulatory measures. Analysts should not label any deviation an error merely because the conventions differ. They should determine the precise mechanism, whether it affects distributable resources, and whether it changes the ability to pay policyholders.
Matching adjustment as valuation regulation
Some solvency frameworks permit a prescribed adjustment to the discount rate for qualifying insurance liabilities supported by a matching portfolio. Such treatment requires careful attention to legal eligibility, residual credit risk, and prescribed valuation rules. It does not establish a universal financial principle that every additional asset spread is available to offset insurance obligations.
The distinction between managing cash flows and qualifying for a matching adjustment is essential: an insurer may engage in sound liability-driven investment without receiving regulatory adjustment benefits, while eligibility for an adjustment does not eliminate residual credit, liquidity, or model risk.
European Union Matching Adjustment
The European Union's Solvency II Directive, Article 77b, provides a legal framework for applying a matching adjustment to the relevant risk-free interest-rate term structure for eligible portfolios of insurance obligations, subject to supervisory approval and qualifying conditions. Its framework requires assigned assets and liabilities, limitations on policyholder options, and cash-flow correspondence in the same currency, among other conditions. The eligibility concept is stricter and more specific than the broad economic notion of a well-managed duration position.
This mechanism influences how insurers construct dedicated portfolios and assess the value of long-dated bonds. The relevant assessment must distinguish statutory best estimates, expected cash-flow replication, residual credit risk, and capital treatment. Conditions differ from a general permission to capitalize all credit spreads as a liability discount.
The amended Solvency II framework has prospective implementation changes. EIOPA's 2026 ring-fenced fund guidance states that revised guidelines are applicable from January 30, 2027; it discusses conditions under which matching-adjustment portfolios may receive diversification treatment without automatically constituting ring-fenced funds. On this monograph's cutoff date, the future applicability is stated explicitly rather than used as current law.
The EU framework demonstrates how accounting and solvency rules can influence asset demand. It also illustrates a governance concern: once an adjustment depends on eligibility, changing asset characteristics or liability behavior can have nonlinear effects on reported own funds. An internal matching analysis should stress both market risk and the potential loss or reduction of favorable regulatory treatment.
United Kingdom Matching Adjustment and Investment Accelerator
The United Kingdom's post-reform matching adjustment operates under the Prudential Regulation Authority's Solvency UK framework. The PRA's current SS7/18 guidance, effective from October 27, 2025 as of this research date, addresses asset and liability eligibility, cash-flow matching, calculation, governance, permissions, and ongoing compliance. It should be read with the specific PRA Rulebook requirements and relevant policy statements, rather than conflated with EU law.
The October 2025 Matching Adjustment Investment Accelerator policy introduced an avenue for eligible firms with permission to include limited self-assessed assets featuring characteristics outside their existing matching-adjustment permission, subject to defined controls and a subsequent application process. Its purpose includes reducing delays in investment while maintaining supervisory safeguards. The instrument adds an eligibility and governance dimension to the investment decision: an asset's economic suitability is only one part of the institution's effective capacity to include it in a specific matching portfolio.
The PRA's SS7/18 publication record separately identifies a July 2026 version scheduled to become effective December 31, 2026. The future version is not presented as already operative on November 14, 2026. This date distinction prevents an otherwise subtle regulatory error when comparing portfolios across time.
From a scientific perspective, the UK case reveals a trade-off between early investment access, the predictability of cash flows, and the risk that an asset's eligibility assessment may require remediation. The relevant cash-flow analysis should incorporate the liquidity and capital consequences of a possible ineligibility finding, not merely the yield expected if the preferred regulatory treatment continues.
United States Asset Adequacy and Cash-Flow Testing
United States insurance regulation is largely state based, with NAIC models, statutory accounting, actuarial opinions, and capital assessment supporting a coordinated but jurisdiction-dependent structure. The model Actuarial Opinion and Memorandum Regulation appears in the NAIC's model-law catalogue, while adopted state law controls its implementation. A model text is not automatically an enforceable national statute.
The NAIC overview of interest-rate risk describes statutory asset adequacy cash-flow testing and explains the exposure to reinvestment risk, surrender acceleration, and forced asset sales. It also notes the use of deterministic and stochastic interest-rate scenarios in life insurance assessment. Those descriptions support the analytical importance of scenario testing but should not be converted into one universal required number of scenarios for every state, product, or reporting date.
U.S. professional actuarial guidance provides a particularly useful modeling comparison. ASOP No. 22 governs specified asset adequacy opinions and related analysis. The revised ASOP No. 7, effective June 1, 2026, governs relevant life and health cash-flow analysis. Together they illustrate the distinction between constructing projections and issuing an adequacy opinion about the reserves and supporting assets under applicable circumstances.
A comparative inference follows: the United States approach often places substantial weight on prospective cash-flow testing and appointed-actuary judgment, while EU and UK matching-adjustment regimes directly connect qualifying portfolio structure to liability discount treatment. Each can be studied through a common economic ALM framework, but statutory outputs and permissions remain jurisdiction specific.
International Empirical Evidence and Comparative Interpretation
The IAIS Global Insurance Market Report 2025 reports that insurers in its monitoring exercise had reduced duration gaps in recent years, while warning that divergence in national monetary policies can alter market exposure. It also identifies geoeconomic fragmentation and private credit as concerns for asset-liability and liquidity risk analysis. The reported pattern is a sample- and measurement-specific result, not proof that all insurers have eliminated interest-rate mismatch.
EIOPA's October 2026 Insurance Risk Dashboard uses Q2 2026 quarterly reporting and year-end 2025 annual indicators. It describes median insurer bond exposure of 49.7% of total assets in Q2 2026, stable duration mismatch at year-end 2025, and elevated market risk. The dashboard reports median holdings and other distributional measures under specified reporting conventions. The figures must not be treated as portfolio weights for every European insurer.
The EIOPA 2024 insurance stress test found sensitivity to simultaneous increases in yields and risk premia under its adverse scenario. It reported important capital effects under a fixed balance sheet and mitigation by reactive management actions, while also assessing liquidity. This observation supports joint modeling of asset prices, liability values, and management responses, with the limitation that stress-test scenarios and participant samples do not directly predict future events.
These sources support a restrained synthesis. Duration matching has improved in some observed samples, but robust protection requires a wider set of measures: curve-shape sensitivities, spread responses, liquid funding, surrender stress, and asset quality. Aggregate changes in duration gaps do not identify the causal effects of a specific regulation or investment strategy without further research.
Stochastic Asset-Liability Modeling
Joint state processes
A coherent stochastic ALM model simulates interest-rate curves, credit spreads, defaults, equity returns where relevant, inflation, exchange rates, mortality and longevity, policyholder behavior, and the availability of liquidity. Asset and liability cash flows then respond to the same economic paths. The dependence structure matters: a rate rise coupled with a widening credit spread and increased surrenders is different from the sum of three independent, single-factor shocks.
Let the state vector be X_t and the control vector u_t describe permissible investment and risk-management actions. A general recursion has the schematic form
This is a generic modeling notation, not a new theory or proposed proprietary model. Its value lies in specifying which variables jointly determine cash flows and which management actions can lawfully and operationally be taken in each scenario.
Scenarios and probability measures
Pricing and capital measurement may require different probability measures and calibration approaches. Market-consistent valuation of financial contingencies must be distinguished from realistic projections of future cash outcomes and from conservative regulatory stress conditions. Mixing them without explanation can produce a numerical result that has no interpretable probability or accounting meaning.
At minimum, analysis should cover rates falling, rates rising, nonparallel curve shifts, credit spread widening, defaults, liquidity stress, surrender shocks, longevity deviation, and combinations. Dynamic hedging paths should include realistic transaction capacity, trading costs, and collateral. A policy that calls for asset sales in a nonexistent market is not an executable risk mitigation strategy.
Model governance
Calibration requires data provenance, reconciliation to observed positions, sensitivity to long-run assumptions, version control, independent review, and clear escalation triggers. Tail events often provide sparse data. Judgment should therefore be documented as an assumption, compared with alternative scenarios, and measured for its impact on the decision. Complexity is justified when it improves the reliability of a material decision; a simpler model may be more transparent for straightforward liabilities.
The IAIS explicit ALM policy standard supports integrating investment, liability strategy, and scenario valuation within the insurer's enterprise risk management framework. The model should provide decision evidence for governance rather than substitute for an accountable decision-making process.
Integrated Numerical Case Studies
Case study: maturity mismatch hidden by initial present value
Consider an insurer owing 100 currency units at year 10. At a flat continuously compounded 4% yield, the present value is 100e^(−0.4)=67.0320. Suppose it buys a default-free zero-coupon security maturing at year 5 with face amount F selected to equal the liability's initial present value. Since the bond value is F e^(−0.2), F=100e^(−0.2)=81.8731. At inception, the asset and liability values are both 67.0320; initial economic surplus is zero. Yet the bond pays 81.8731 at year 5, and those proceeds must be reinvested for another five years.
If the five-year reinvestment rate equals 4% continuously compounded, the proceeds grow to 81.8731e^(0.2)=100 and fund the benefit. If the reinvestment rate falls immediately to 1% for years 5–10, the proceeds grow to 81.8731e^(0.05)=86.0708, leaving a nominal shortfall of 13.9292. If the insurer instead held a zero-coupon claim of 100 maturing at year 10 with equivalent credit and legal characteristics, that reinvestment exposure would be absent. The calculation isolates cash-flow timing and reinvestment; it does not model interim changes in market value, taxes, defaults, or regulatory reserve rules.
Case study: equal duration with surplus sensitivity to valuation amounts
Suppose an insurer has assets worth 120 with modified duration 6 and liabilities worth 100 with modified duration 8. Their percentage duration gap is −2 years, but the relevant dollar durations are 720 and 800. For a +100 basis-point parallel yield move, first-order changes are −7.20 in asset value and −8.00 in liability value. Surplus therefore rises by approximately 0.80, from 20.00 to 20.80. A −100 basis-point move approximately reduces surplus by 0.80 before convexity. This reveals why duration assessment must weight sensitivities by present value.
If asset duration were adjusted to 800/120=6.6667, first-order dollar durations would match. This would neutralize local surplus sensitivity to a small parallel shock under fixed cash flows. It would leave nonparallel curve, spread, reinvestment, convexity, liquidity, and behavioral risks unaddressed.
Case study: same aggregate duration, different curve exposure
Consider two hypothetical portfolios with identical initial values and reported flat-curve duration of eight years. Portfolio A has all payments near eight years; Portfolio B combines large near-term and distant payments to produce the same average. A localized upward shock to yields at the long end can affect Portfolio B substantially more than Portfolio A. The exact magnitude requires the cash-flow amounts and curve interpolation; none is fabricated here. The analytic result is structural: one scalar first moment cannot identify the full distribution of payments across maturities.
For any proposed hedge, a practitioner should therefore compare the vectors of time-bucket cash flows and key-rate sensitivities before claiming that a shared duration means equivalent protection.
Case study: surrender shock and economic liquidity
Assume assets of 110 and liabilities of 100 under a stated economic basis, yielding surplus of 10. If 30 of immediately payable surrenders arise while only 10 of unencumbered cash is available, a liquidity gap of 20 exists before asset sales or funding. The existence of positive accounting or economic surplus does not close that gap on the required day. If securities must be sold at a 5% execution discount to their pre-stress marked value to raise 20 in cash, 21.0526 of pre-discount asset value must be liquidated, generating an execution loss of about 1.0526. This arithmetic is independent of whether the underlying assets were economically solvent over a longer horizon. It demonstrates the importance of settlement timing and executable liquidity.
The example assumes immediate full surrender payment, a fixed discount for the required asset sales, and no alternative credit facilities. It is illustrative and must not be interpreted as a forecast of surrender incidence or market haircuts.
Sensitivity and reproducibility
The numerical cases use deterministic inputs and elementary formulas, with amounts rounded to four decimals where appropriate. The first case can be verified by substituting the stated rates into exponential discounting. The second follows directly from dollar-duration derivatives. The third is a qualitative counterexample regarding distributions of maturity; its absence of invented cash-flow numbers is intentional. The fourth follows from dividing required cash by the sale-price fraction 0.95.
No scenario is inferred from actual insurer portfolios. Their role is to test conceptual distinctions, disclose assumptions, and prevent a false equivalence between value matching, payment matching, and liquidity matching.
Comparative Asset-Liability Governance
Accountable mandates
The insurer's board and senior management must specify the objective of matching, the permissible risk budget, legal-entity constraints, and escalation rules. The investment committee, actuarial function, risk function, and treasury should use mutually reconcilable liability and asset data. Responsibilities can differ across firms, but the governance record should make clear who authorizes investment, who measures risk, and who can challenge modeled assumptions.
An effective mandate addresses the allowed duration and key-rate gaps, eligible asset types, concentrations, liquidity reserves, collateral obligations, credit migration, and permissible hedges. It also specifies what happens after a threshold breach. A limit that cannot trigger action in a fast market is an incomplete control.
Performance and incentives
An asset manager rewarded solely for yield over a benchmark may be encouraged to accumulate spread and liquidity risk inconsistent with long-term policy promises. Liability-aware evaluation requires comparing investment outcomes with the behavior and valuation of the funded liabilities. The result must be interpreted across market states and should include the cost of maintaining hedges and capital.
The IAIS guidance on investment risk emphasizes liability characteristics and risk management. Its framework supports a decision process in which portfolio return is evaluated in relation to policyholder obligations rather than as an isolated objective.
Comparative institutional synthesis
Across IAIS principles, European prudential requirements, UK matching-adjustment permissions, and U.S. asset adequacy practice, a common pattern emerges: the analysis starts with promised cash flows and tests whether supporting assets remain adequate under relevant uncertainty. Differences concern the legal status of portfolios, valuation adjustments, regulatory permissions, and professional actuarial duties. Such differences should be documented as institutional characteristics rather than collapsed into one global rule.
The central governance conclusion is operational. Insurance matching is a continuous process of projection, valuation, funding, capital assessment, and accountable response. It cannot be delegated entirely to a duration statistic or an optimization engine.
Model Limitations, Validation, and Research Priorities
Sources of model error
Mortality improvement, future lapse behavior, expense inflation, callable asset behavior, private-asset valuations, and tail dependence are imperfectly observed. Alternative reasonable assumptions may yield materially different asset allocations. Scenario selection can also create a false sense of security if it omits severe but plausible combinations. Model uncertainty must be incorporated into governance through sensitivity bounds, expert challenge, and capital or liquidity buffers appropriate to the decision.
Market observability and very long obligations
Observable liquid yield curves rarely extend indefinitely. Insurers with obligations beyond traded maturities must extrapolate or use synthetic hedges. Such techniques depend on assumptions about forward rates, basis behavior, and market accessibility. Claims of very-long-horizon duration precision should disclose the role of extrapolation. The apparent availability of a theoretical hedge is not evidence that sufficient instruments can be traded at acceptable prices.
Limits of the empirical comparison
The IAIS monitoring evidence and EIOPA's latest risk indicators draw on different populations, regulatory definitions, periods, and data aggregation. A cross-report ranking of insurer matching quality would require harmonized measurements and an identified sample; such a ranking is not offered here. Generalizations are limited to mechanisms supported by standards and observations reported for the relevant populations.
Research priorities
Future empirical research can compare the effectiveness of key-rate and cash-flow matching under joint surrender and spread shocks, test the liquidity consequences of derivative overlays, study realized outcomes of matching-adjustment eligibility changes, and investigate the sensitivity of very long annuity portfolios to joint longevity and yield-curve shifts. These are research questions, not newly declared theories. Robust evaluation would require matched insurer data, clear measurement baselines, and appropriate causal or descriptive methods.
Conclusion
Insurance asset-liability matching is a liability-informed financial discipline grounded in the exact shape of future policyholder obligations. Its elementary measures are cash-flow schedules, present values, dollar duration, convexity, key-rate exposures, and liquidity resources. Each contributes information that the others cannot replace. Matching of initial values is insufficient when payments occur at different dates. Matching of local duration is insufficient when the yield curve changes shape, options are exercised, or credit spreads move. A favorable economic surplus is insufficient when claims or collateral payments must be delivered before funding can be mobilized.
The analytical results show that duration immunization is a local sensitivity condition, cash-flow dedication is a stronger but assumption-dependent nominal funding technique, and stochastic projection is necessary when liability behavior reacts to the financial environment. Longevity, surrender decisions, defaults, reinvestment, currency and inflation introduce residual risks that must be explicitly measured and financed. The comparative review confirms that the IAIS, EU, UK, and U.S. frameworks share a concern for liability-appropriate investment while applying different legal tests and measurement conventions.
The author's synthesis therefore establishes a practical interpretive conclusion: the credibility of long-term insurance protection is assessed by combining solvency value, dated settlement capacity, contractual optionality, legally usable assets, and accountable governance. The durable object of matching is the insurer's capacity to fulfill contingent promises across plausible states of the world. This is a research conclusion drawn from established actuarial finance, supervisory requirements, and transparent analysis; it creates no new named doctrine.
References
Actuarial Standards Board. (2021). Statements of Actuarial Opinion Based on Asset Adequacy Analysis for Life Insurance, Annuity, or Health Insurance Reserves and Other Liabilities (Actuarial Standard of Practice No. 22; effective June 1, 2022). Official ASOP No. 22 text.
Actuarial Standards Board. (2025). Life or Health Cash Flow Analysis (Actuarial Standard of Practice No. 7; effective June 1, 2026). Official ASOP No. 7 text.
Bank of England, Prudential Regulation Authority. (2018–2026). SS7/18: Solvency II: Matching Adjustment. Publication and version-history record; current version effective October 27, 2025; future revision effective December 31, 2026. PRA supervisory statement.
Bank of England, Prudential Regulation Authority. (2025). PS17/25: Matching Adjustment Investment Accelerator. Published October 23, 2025; effective October 27, 2025, subject to noted reporting implementation dates. PRA policy statement.
European Insurance and Occupational Pensions Authority. (2024). Insurance Stress Test 2024. Results released December 17, 2024. EIOPA stress-test materials.
European Insurance and Occupational Pensions Authority. (2026). Insurance Risk Dashboard: October 2026 (EIOPA-BoS-26-495). Quarterly data reference Q2 2026; annual indicators year-end 2025. Official risk dashboard.
European Insurance and Occupational Pensions Authority. (2026). Final Report on Revised Guidelines on Ring-Fenced Funds—Solvency II Review. Published July 15, 2026; guidelines scheduled to apply January 30, 2027. EIOPA revised guidelines report.
European Parliament and Council. (2009, as consolidated January 17, 2025). Directive 2009/138/EC on the Taking-up and Pursuit of the Business of Insurance and Reinsurance (Solvency II), Article 77b. Official EUR-Lex consolidated text.
Fisher, L., & Weil, R. L. (1971). Coping with the risk of interest-rate fluctuations: Returns to bondholders from naive and optimal strategies. The Journal of Business, 44(4), 408–431. https://doi.org/10.1086/295402. Bibliographic publisher-index record.
IFRS Foundation. (2017, with subsequent amendments and implementation materials). IFRS 17 Insurance Contracts. Effective for annual periods beginning January 1, 2023. IFRS 17 standard overview.
IFRS Foundation. (2014, with subsequent amendments). IFRS 9 Financial Instruments. IFRS 9 standard overview.
International Association of Insurance Supervisors. (2024). Insurance Core Principles and Common Framework for the Supervision of Internationally Active Insurance Groups (December 2024). IAIS official standards text.
International Association of Insurance Supervisors. (2024, continuously maintained). ICP and ComFrame Online Tool, including ICP 15 (Investments) and ICP 16.5 (Asset-Liability Management). IAIS online principles.
International Association of Insurance Supervisors. (2025). Global Insurance Market Report 2025. December 2025. Official GIMAR report.
National Association of Insurance Commissioners. (n.d., accessed November 14, 2026). Interest Rates and Insurance. NAIC research overview.
National Association of Insurance Commissioners. (2026). Actuarial Guidelines. Institutional overview, updated February 11, 2026. NAIC actuarial guidelines.
National Association of Insurance Commissioners. (n.d., accessed November 14, 2026). Model Laws. Includes Actuarial Opinion and Memorandum Regulation (MO-822). NAIC model-law catalogue.
Redington, F. M. (1952). Review of the principles of life-office valuations. Journal of the Institute of Actuaries, 78(3), 286–340. Cambridge journal record.