Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Policyholder Optionality and Asset-Liability Feedback

Abstract. Life insurance and annuity contracts often convey choices to policyholders whose exercise changes the amount and timing of insurer liabilities. Surrender, partial withdrawal, premium discontinuance, policy loans, guaranteed crediting arrangements, annuitization and participation rights interact with interest rates, household liquidity needs, distribution incentives, asset valuations and the insurer's capacity to provide cash. These interactions create an asset-liability feedback problem: economic states change policyholder decisions, those decisions change liability cash flows, and the resulting investment, hedging, funding and crediting actions alter the insurer's subsequent exposures. The fundamental research question is how to measure this two-sided adjustment without collapsing heterogeneous contracts into a single lapse assumption or mistaking a favorable present-value remeasurement for immediately available liquidity. This monograph synthesizes contingent-claims valuation, competing-risk and survival models, dynamic asset-liability management, liquidity stress testing and supervisory evidence. It derives the accounting and cash-flow identities that distinguish option value from exercise frequency, economic capital from settlement cash, and portfolio liquidity from aggregate market depth. Illustrative calculations show that an upward interest-rate shift can improve a static duration-gap estimate while triggering additional surrenders and transaction losses; a simple exercise example separately identifies the value of a contractual surrender floor. European and U.S. supervisory findings are compared with the different requirements of Solvency II, IFRS 17, international insurance risk management standards and U.S. own-risk assessment. The analysis concludes that policyholder optionality should be evaluated as a state-dependent joint balance-sheet exposure, with explicit contract rules, nonfinancial behavior, settlement lags, asset-sale constraints, hedge collateral and management responses. This conclusion is an analytical synthesis of established scholarship and supervisory practice. No newly named theory or proprietary model is asserted.

Keywords: Policyholder optionality; surrender option; dynamic lapse; insurance asset-liability management; embedded guarantees; contingent claims; life insurance; annuities; liquidity stress testing; duration gap; behavioral risk; market-consistent valuation; insurance solvency; financial stability; model validation.

Introduction

A policyholder's right to alter, withdraw from or terminate a contract is part of the insurer's financial exposure because the policyholder controls an element of the timing and quantity of future cash flows. The exercise of that right is also a household decision. It responds to market opportunity, personal liquidity, taxation, advice, risk perception, product design and the value of continued insurance protection. The insurer, in turn, responds to aggregate exercise by managing assets, adjusting crediting rates, realizing gains or losses, obtaining funding, rebalancing hedges and, where available, invoking contractually authorized settlement provisions. Each response affects the subsequent asset-liability position. The object of this dossier is the connected sequence of decisions and balance-sheet consequences.

The distinction matters most for long-duration savings and protection business. A traditional participating life contract may give the holder a minimum interest guarantee, a declared bonus, a death benefit and a surrender value. A deferred annuity may have guaranteed minimum benefits and a withdrawal schedule. A fixed annuity with a surrender-charge period offers different incentives from a unit-linked account whose policyholder bears much of the asset-market risk. Even within one product class, jurisdictional contract language, tax treatment and distribution practice change the probability and cost of exercise. Policy behavior is therefore neither a single market factor nor an invariant actuarial decrement.

The Grosen and Jørgensen (2000) valuation identified the economic interaction of interest guarantees, bonuses and surrender rights in participating insurance. The Bacinello (2003) decomposition treated the surrender privilege as an American-style contingent claim alongside a basic contract and participation component. These studies establish a financial valuation perspective. The Kuo, Tsai and Chen (2003) evidence and Eling and Kiesenbauer (2014) analysis establish why observed policy termination requires an empirical behavior perspective as well. The two perspectives address different questions. A contractual option has a state-contingent economic value even when a particular holder never exercises it; an observed surrender may occur for reasons unrelated to maximizing a financial option's value.

The asset side makes this distinction operational. A rise in yields can lower the market value of fixed-income assets and raise the return available on new investments. For some portfolios, the simultaneous reduction in liability present values initially exceeds the asset decline. Yet the same state may make older contractual returns less competitive and prompt withdrawals that require the sale of depreciated bonds. The NAIC discussion of interest-rate exposure and the BIS analysis of rising-rate relief identify these interacting channels. Their coexistence is central to an integrated ALM assessment.

The investigation uses four methods. First, it reconstructs contractual cash-flow rights and separates legally exercisable actions from empirically estimated behavior. Second, it derives present-value, duration and funding identities under explicit timing assumptions. Third, it evaluates empirical and supervisory evidence against the scope of the studied populations, dates and stress-test designs. Fourth, it compares model output under alternative economic, behavioral, liquidity and management-action assumptions. The numerical illustrations are transparent hypothetical cases; they are neither estimates from an insurer portfolio nor fitted forecasts. Evidence about the United States, Germany and European supervisory samples is attributed to those populations rather than extrapolated to every market.

The principal Science connection is Insurance Finance and Asset-Liability Management. Related research intersections include Life Insurance, Annuities and Retirement Insurance, Solvency and Capital, and Enterprise Risk Management in Insurance. The independent boundary of this dossier is the feedback from policyholder options through asset response, liquidity and capital. Standalone estimation of contract-level lapse transition probabilities belongs to the adjacent Policyholder Lapse and Surrender Models dossier; this study uses those probabilities as one part of the joint asset-liability process.

Financial Options in Insurance Policies — Concepts, History and Scientific Object

Contractual rights and financial exposure

A financial option is a right to make a choice contingent on future circumstances, subject to specified contractual conditions. An insurance policy can contain several distinct choices. A whole life policy may accumulate cash value and permit surrender; a universal-life policy may allow flexible premiums and crediting arrangements; a deferred annuity may permit partial withdrawals and later election of annuity payments. Some choices transfer interest-rate, mortality or longevity exposure between policyholder and insurer. Others change the policy's funding trajectory or the period for which the insurer remains committed to a guarantee.

The operative research unit is the policy in a state at a particular decision date. Denote contract terms by cc, observed financial and demographic information by XtX_t, and the set of legally available actions by A(c,t)\mathcal A(c,t). The exercise set may include continuation, surrender, partial withdrawal, paid-up conversion, borrowing or annuitization, depending on the contract. At each date the contractual cash flow depends on the chosen action, and the future contract state changes accordingly. The aggregate liability is the sum of these contract-state exposures over the portfolio, subject to dependencies among holders and the valuation assumptions used.

Surrender and lapse need separate definitions. Surrender ordinarily refers to voluntary termination with a cash payment defined by the contract. Lapse is broader and may occur after premium nonpayment; it need not entail the same payment or imply the same financial motivation. Partial withdrawal keeps some coverage or account value in force, while a policy loan can generate an immediate cash demand without extinguishing the entire liability. Annuitization may convert accumulation rights into a stream of longevity-contingent payments. A guaranteed minimum withdrawal benefit can continue after market account exhaustion under contract-specific conditions. These rights produce different cash-flow profiles, exposures to guaranteed values, and recovery opportunities.

A useful classification follows the settlement event rather than the word used in marketing. For each action the analyst records the eligibility date, notice and administrative lag, contractual payment formula, surrender charges, market-value adjustment, possible taxes, treatment of remaining guarantees, claim or mortality interaction, optional deferral, and allocation to an insurer general or separate account. This classification prevents an estimated termination frequency from being incorrectly multiplied by one universal cash payment.

Option value and contractual persistence

The economic value of a financial surrender option arises because the holder may choose an exercise date when the cash surrender amount compares favorably with the continuation value of the contract. In the Grosen–Jørgensen participating-contract study, financial guarantees and bonus mechanisms jointly determine the contract's liability value. The Bacinello surrender-option study explicitly separates the baseline benefit, participation feature and surrender right. These decompositions illuminate pricing sensitivities, but the exercise decision of a real holder also depends on information, preferences, protection value and practical barriers.

The insurer's option exposure is asymmetric. When external rates rise, a policy paying a low locked-in crediting rate may be surrendered in favor of an attractive alternative. When rates fall, policyholders may retain valuable legacy guarantees. The insurer may then experience both accelerated cash payments in one market regime and prolonged guarantee obligations in another. The shape of those responses depends on contract design. Surrender penalties and market-value adjustments can dampen immediate withdrawals; guarantee preservation, declining insurability, transaction costs and tax effects can encourage continuation. The contract's age and remaining surrender-charge schedule can therefore matter as much as the current interest rate.

The historical embedded-options literature is economically important because it refutes the assumption that actuarially estimated average duration is an immutable property of a liability. Surrender opportunities, bonus rules and minimum returns introduce nonlinear dependence on asset performance and the rate term structure. A policyholder's option can change which future cash flows actually occur, changing the liability's effective duration and convexity. The asset portfolio constructed to match a static expected cash-flow path can then cease to match the liability that emerges after exercise.

Financial and behavioral optionality

A financially rational option benchmark answers how an informed holder who maximizes a specified monetary objective might act under stipulated information and market completeness. An empirical behavioral model answers how contract holders have actually acted in observed conditions, perhaps modified by scenario stress and forward-looking assumptions. Both are valuable, but their probability measures and interpretations differ.

Market-consistent valuation frequently uses a risk-neutral or otherwise prescribed economic valuation framework for traded risks. Empirical lapse and surrender frequencies are estimated under the real-world distribution of behavior. A risk-neutral optimal exercise rule cannot be substituted automatically for a realistic expected surrender rate. Conversely, a historical frequency cannot by itself determine the market-consistent value of a long-dated guarantee in a changing financial environment. The analyst must state whether each reported quantity represents an economic option value, an expected cash payment, a capital stress, an accounting estimate, or a liquidity forecast.

Policyholder welfare also belongs in the interpretation. Surrendering a protective life policy can forfeit death benefits, underwriting status and valuable guarantees. A nominal improvement in externally available yield may be outweighed by those lost rights. This affects the inference from competitive spread alone to actual behavior. It also limits the analogy to deposit withdrawals. Insurance liabilities differ by surrender charges, liquidity horizons, contract-specific protections and insurance benefits; nevertheless, correlated withdrawals can still strain particular insurers.

Portfolio aggregation and common shocks

Let NN contracts have current surrender values Si(t)S_i(t) and conditional surrender probabilities qi(t)q_i(t) over a defined interval. The simplest expected cash outflow is E[Ot∣Xt]=∑i=1Nqi(t)Si(t)E[O_t\mid X_t]=\sum_{i=1}^N q_i(t)S_i(t), assuming the full cash surrender amount is settled during that interval. A partial-withdrawal term, a policy-loan term and any other contractual option payments must be added separately. This expression is an expectation, not a stress quantile, and it does not account for correlated exercise or settlement lag without additional structure.

The dispersion of aggregate cash demand depends on dependence among contracts. If IiI_i indicates surrender, then Var⁡(∑IiSi)\operatorname{Var}(\sum I_iS_i) includes covariance terms 2∑i<jSiSjCov⁡(Ii,Ij)2\sum_{i<j}S_iS_j\operatorname{Cov}(I_i,I_j). Common interest-rate changes, labor-market disruptions, financial advice campaigns, distribution-platform actions and loss of confidence can increase correlation. An annual portfolio-average surrender rate can conceal that tail behavior, especially when contracts contain materially different surrender rights or large account balances.

The internationally relevant scientific object is therefore contract-dependent exercise under common economic shocks and its endogenous effect on financial resources. That object permits comparative evidence while preserving distinctions between products and jurisdictions.

Rate-Sensitive Surrenders — Formal Financial or Actuarial Mechanisms

Interest-rate incentive and household liquidity

Rate-sensitive surrender occurs when a change in financial opportunities modifies the relative attractiveness of retaining an insurance contract. Let rtaltr^{alt}_t denote a suitably risk-adjusted alternative yield, rtpolr^{pol}_t the contract's attainable crediting rate, and ci(t)c_i(t) the annualized economic cost of surrender, including fees, tax and lost benefit value. A simple incentive indicator is zi(t)=rtalt−rtpol−ci(t)z_i(t)=r^{alt}_t-r^{pol}_t-c_i(t). This is a diagnostic spread, not a sufficient behavioral law. It must be modified for differences in duration, credit risk, guarantee structure, protection and tax characteristics.

When zi(t)z_i(t) increases, the economic incentive to replace certain cash-value policies may strengthen. Yet a household facing unemployment may withdraw funds even with an unfavorable yield comparison. Conversely, a household with deteriorated insurability may retain a contract despite a positive spread. The Kuo, Tsai and Chen (2003) cointegration study examines emergency-fund and interest-rate explanations. The authors find unemployment relevant in both long- and short-run relationships, while interest rates mainly affect the long-run relation; their impulse-response comparison reports a larger overall role for interest rates. This empirical result identifies interacting mechanisms in the studied data, not a universal causal coefficient for every line of business.

A firm's exposure changes when the competing mechanisms act together. A recession can simultaneously increase household cash needs, reduce asset market liquidity, widen credit spreads, impair some asset values, and raise the insurer's own funding cost. A monetary tightening can improve reinvestment yields while reducing the market value of legacy bonds and creating interest-rate arbitrage for some annuity holders. Their effects may coincide or offset. The scenario designer should specify the joint path of labor income, consumer liquidity, market yields, spreads and product-specific guarantees.

Observed policy heterogeneity

The German contract-level study by Eling and Kiesenbauer (2014) uses over one million policies from one insurer and compares proportional hazards with generalized linear models while accounting for market turmoil. The authors report that product type, contract age and policyholder characteristics help explain lapse outcomes. The size of this dataset supports precise empirical analysis within its institutional context; it does not remove dependence on the insurer's product mix, distribution network, calendar period and legal environment.

The Deutsche Bundesbank's 2024 financial-stability assessment offers a valuable qualification. It observes that fixed surrender values for many German contracts can remain unchanged by bond-market losses, creating a financial incentive to terminate and reinvest, but judges the risk of a broad wave of lapses limited in its assessment. Reported behavior was only somewhat above pre-rate-rise levels, and a survey found that substantial return differentials would be required to provoke a marked increase. The survey threshold should not be generalized to other countries or contract cohorts. The lesson is that option exposure can be economically material while behavior remains slower and more selective than a simple arbitrage model predicts.

Actuarial event and transition models

A discrete-time lapse model usually conditions on survival in force through the prior period. The one-period lapse probability qi,t=P(Ti=t∣Ti≥t,Xi,t)q_{i,t}=P(T_i=t\mid T_i\ge t,X_{i,t}) can be modeled by a generalized linear specification, for example logit⁡qi,t=αd(i)+f(durationi,t)+β⊤Zi,t+γzi,t+ut\operatorname{logit}q_{i,t}=\alpha_{d(i)}+f(\text{duration}_{i,t})+\beta^\top Z_{i,t}+\gamma z_{i,t}+u_t. Here αd(i)\alpha_{d(i)} represents product segment, Zi,tZ_{i,t} includes demographic, distribution and contract variables, and utu_t captures calendar or shared conditions. The logistic form is an established illustrative specification; it is not a fitted equation for the present monograph.

Competing-risk designs are more faithful when surrender, death, annuitization, disability, paid-up conversion and premium discontinuance are mutually distinct transitions. If multiple actions are possible, an event-specific hazard must be linked to the probability of still occupying the corresponding contract state. Care is required when a loan or partial withdrawal leaves the contract in force: treating those events as terminal lapses produces incorrect persistence, value and future option exposure. Multi-state cash-flow models may also need repeated choices and restoration or conversion rules.

Time-varying covariates create identification problems. Rates change in response to macroeconomic conditions that also influence unemployment, asset returns and policyholder sentiment. Repricing crediting rates is a management response to market conditions and anticipated lapse pressure. Estimating the causal effect of the credited-rate spread from observational data without a credible design may confound cause and consequence. Product duration, surrender-charge cliff dates, geographic economic variation and repricing policies can assist identification, provided their own selection mechanisms are examined.

Behavioral calibration and stress

Calibration should distinguish observed behavior, management expectation and stress assumptions. An observed lapse rate is a frequency for a defined population and time horizon. A best-estimate projection requires a forward-looking conditional distribution. A liquidity stress may impose correlated withdrawals beyond that distribution. A valuation of contractual optionality may use rational exercise bounds to test the exposure even if actual exercise is incomplete.

Practical validation uses cohort splits by age, duration, product, guarantee moneyness, policy size, distribution channel, charge schedule and calendar period. The analyst should check exposure-weighted as well as policy-count-weighted rates because larger cash values disproportionately influence cash demand. Out-of-time performance is crucial across rate regimes: a model fitted in low-rate years may underestimate behavior after steep rate increases. Cross-sectional accuracy at the policy level does not establish accuracy of portfolio liquidity tails under a new common shock.

There are also data-quality asymmetries. Contract administration often records the termination date and reason with more precision than the holder's actual motivation. Distributor incentives or adviser's outreach may change observed actions without being measured. Administrative data may register withdrawals after the decision or after processing delays. A sensible audit preserves application date, notice date, effective date, settlement date and cash amount where possible. These time stamps matter because a correctly predicted annual surrender rate may still produce a dangerously wrong 30-day funding forecast.

From behavior to asset-liability effect

The insurer's net outflow at time tt is affected by the amount settled for surrender, not only the count of terminated contracts. Let ni,tn_{i,t} represent a settlement indicator or fractional exercise amount and Si,tS_{i,t} the payable value. Then gross option cash demand is Ot=∑ini,tSi,tO_t=\sum_i n_{i,t}S_{i,t}. Net funding demand is gross outflow plus other obligations and collateral needs less scheduled incoming premiums, coupons, redemptions and realizable cash. Uncertainty about settlement lags, large-policy concentrations and correlated rate exposure should be explicitly modeled.

This synthesis yields a practical test: a lapse model used in ALM is fit for purpose only when its event definitions, payment amounts, timing, scenario dependencies and aggregation rules map to the cash ledger. A strong classifier of annual lapse outcomes may be an inadequate liquidity model, just as a market-consistent option value may be an inadequate forecast of observed household behavior.

Guarantees and Exercise — Quantitative Methods and Assumptions

Valuing a right of exercise

The valuation of a surrender right begins with the choices permitted at each contractual decision time. In a simplified market-consistent setting, let StS_t be the contractual surrender payment at date tt, CtC_t the economic value of continued coverage and future contractual choices, and Ft\mathcal F_t the information available at that time. A financial exercise benchmark chooses the action giving the larger continuation or surrender value. If surrender is permitted at specified dates and the other risks are represented consistently, an idealized optimal-stopping expression is

Vt=max⁡(St,EtQ[D(t,t+1)Vt+1+Bt,t+1]),V_t=\max\left(S_t,\mathbb E_t^{Q}\left[D(t,t+1)V_{t+1}+B_{t,t+1}\right]\right),

where DD is a financial discount factor and BB denotes other benefits payable before the next decision. Actual contracts can include death, partial withdrawal, survival-contingent benefits and state-dependent bonuses; those details belong in the state and cash-flow definitions. The displayed recursion is an established contingent-claims benchmark, not a new valuation system. In incomplete markets, choice of valuation measure and the treatment of nonhedgeable behavior require explicit justification.

The financial option is often represented as a put-like right because the holder can put the continuing contract back to the issuer at a stipulated surrender value. The exact payoff is contract-specific: surrender can extinguish protection, cancel guarantees, trigger charges, or interact with tax rules. The Bacinello (2003) study implements a recursive binomial approach for an Italian participating life contract, and the earlier Grosen–Jørgensen analysis identifies the separate economic contribution of guarantees, bonus distribution and surrender rights. Their central implication for ALM is that policy design changes the duration and market sensitivity of the insurer's liabilities.

A conditional valuation illustrates the distinction between contractual value and behavior. Consider a stylized one-year decision node. In two equally weighted risk-neutral financial states, the continuation value at the exercise date is 108 or 92 monetary units; the contractual surrender payment is 100 in either state. If the holder may surrender at the decision date, the corresponding optimal values at that date are 108 and 100. Under an assumed one-year discount rate of 4%, the time-zero financial value of these state-contingent amounts is (0.5×108+0.5×100)/1.04=100(0.5\times108+0.5\times100)/1.04=100. Without the surrender privilege, the otherwise identical continuation amounts have value (0.5×108+0.5×92)/1.04=96.153846(0.5\times108+0.5\times92)/1.04=96.153846. The incremental surrender-option value is therefore 3.846154 units in this artificial two-state example. The calculation assumes no other cash flows, complete exercise at the designated date and a risk-neutral state weighting stipulated for illustration. It does not imply that observed holders would exercise in precisely one-half of real-world outcomes.

The option's positive value follows from the convexity of the exercise maximum under those assumptions. Actual lapse behavior can be delayed or incomplete, changing the realized cash path and a behavior-adjusted liability estimate. This contrast is a necessary model boundary: optimal-exercise calculations may be useful for product design and upper-exposure sensitivities, while observed-exercise models support cash-flow forecasting. Combining them without aligning probability measure, risk adjustment and economic assumptions can count the same option exposure twice.

Guaranteed returns and persistence

A guaranteed crediting rate binds the insurer's promised return subject to the contractual benefit formula. When market reinvestment rates fall below legacy guarantees, retaining the policy becomes valuable to the holder. Surrenders may decline, extending the duration of the insurer's obligations. When alternative rates exceed credited rates, surrender incentives may strengthen, although the value of protection and contractual penalties still matter. Thus rates influence both liability valuation and how long guarantees persist.

Participation features complicate this relationship. A policyholder may receive bonuses based partly on the insurer's asset performance and smoothing rules. Bonus declarations can be discretionary within contractual, statutory or supervisory bounds; their credibility, path dependence and future management assumptions matter. Contractual guarantees may also interact with investment returns asymmetrically. The Grosen and Jørgensen (2000) analysis shows the sensitivity of participating-policy values to bonus design and guaranteed rates. Surrender and bonus components should therefore be modeled jointly when the contract makes them economically dependent.

Variable and indexed annuities may embed more complex living or death benefit guarantees. Their exercise decisions depend on account values, guarantee bases, withdrawal limits, fees and eligibility conditions. A surrender of a deeply in-the-money guarantee can be uneconomic for the policyholder even when another investment offers a higher headline yield. Modeling only the comparison between credited rate and market rate would erase this protection value. For ALM, product families must be divided according to the source of the guarantee, investment risk allocation and exercise rights.

Interest-rate duration and convexity

For small changes in a parallel discount rate, the first-order market-value change of an asset portfolio is approximately ΔA≃−DAAΔy\Delta A\simeq-D_A A\Delta y, and the corresponding liability value change is ΔL≃−DLLΔy\Delta L\simeq-D_L L\Delta y. Economic surplus E=A−LE=A-L changes by ΔE≃(−DAA+DLL)Δy\Delta E\simeq(-D_A A+D_L L)\Delta y. These are local approximations using consistent discounting and duration measures; they do not incorporate credit spread changes, jumps in surrender rates, asset trading or convexity.

Suppose the hypothetical market-valued assets are 1,000 million units with modified duration 6, and economic liabilities are 900 million with duration 9. A one-percentage-point parallel rise in yields gives an approximate asset loss of 60 million and liability reduction of 81 million; economic surplus therefore rises by 21 million before behavioral feedback and other risks. A one-percentage-point fall gives the opposite first-order result, a 21-million reduction in surplus. The initial economic surplus of 100 million is an accounting identity for this idealized example rather than a regulatory capital determination.

If a yield rise also accelerates surrenders, the actual cash payments and forced-asset-sale costs may offset part of the modeled surplus improvement. If a yield decline induces greater guarantee retention, liability duration may lengthen, magnifying the downward-rate exposure. The second-order terms and endogenous changes in liability behavior can create effective convexity different from that of a static actuarial projection. The BIS discussion of insurer duration gaps highlights why a simple prediction of relief from rising rates can fail once yield-curve shape and liquidity are considered.

Discount curves also move nonparallelly. A short-rate increase may affect reinvestment incentives and surrender opportunities differently from long-term guarantee valuation. Spread widening reduces the value of corporate credit holdings even when the risk-free curve changes little. The sensitivity ledger should therefore distinguish key-rate duration, credit-spread duration, optionality exposure and actual settlement amounts. The sign of an insurer's rate sensitivity is portfolio- and regime-dependent, not universal.

Measurement and financial reporting

Economic ALM, insurance contract accounting and prudential capital are related but different measurement systems. IFRS 17 Insurance Contracts requires current estimates of future fulfillment cash flows, risk adjustment and contractual service margin, with specified separation of certain embedded derivatives and distinct investment components. The accounting standard also separates insurance service results from insurance finance income or expenses. Its effective reporting date is 2023 for annual periods beginning on or after January 1 of that year, subject to the standard's application provisions.

An IFRS 17 liability amount does not equal instantly redeemable cash, and its cash-flow projections are not a standalone liquidity coverage ratio. A prudential Solvency II technical provision and solvency capital requirement use regulatory assumptions and risk measures that need not equal IFRS reporting amounts. A balance-sheet model must retain the mapping between contractual payments, economic valuations, financial-statement entries and regulatory values rather than interpreting their differences as contradictions.

The Solvency II Delegated Regulation includes a lapse-risk submodule in Article 142, defined by the maximum of the permanent increase, permanent decrease and mass-lapse capital shock results under the prescribed framework. The directional symmetry is important: depending on guarantees and future profitability, a decrease in lapses can be adverse because longer persistence retains costly obligations, while an increase can be adverse because surrender realizes embedded options or removes future profits. The regulatory calculation has its own design, eligible options and assumptions. Its result should not be interpreted as a comprehensive liquidity-cash-flow forecast.

The 2025 Solvency II amending directive strengthens several prudential and macroprudential arrangements, with Member State application generally beginning January 30, 2027. This prospective application date must be distinguished from the rules in operation at the November 2026 research date. Stating that the 2027 requirements are already effective would distort the regulatory comparison.

Risk measures and decision horizons

The natural metrics depend on the decision. Pricing may require a state-dependent option value and sensitivities. Asset allocation needs duration and convexity over economic scenarios. Liquidity planning needs gross and net contractual cash flow by date, eligible liquid assets and haircuts. Capital planning needs economic surplus distributions, required capital and management-action assumptions. Product design needs the cost and behavioral consequences of charges and guarantees. One annual lapse-rate statistic cannot substitute for all of them.

Model validation must also respect the time horizon. A one-year reserve sensitivity, a three-month surrender wave and a one-week collateral call are different events. Reporting them as a single capital or liquidity number erases the order in which obligations become payable. A sound actuarial-financial review begins with contract timelines and scenario chronology before aggregating the final metrics.

Asset Response and Funding — Insurance Application and Comparative Evidence

The cash-flow balance sheet

At a given date the insurer holds cash, scheduled coupons and maturities, marketable investments, potentially realizable less-liquid assets, eligible collateral and sources of contingent liquidity. Against these resources are ordinary benefits, expenses, reinsurance payments, withdrawals, surrender settlements, derivatives margin calls, debt service and operational funding needs. The timing and enforceability of each item determines its liquidity relevance.

Let CtC_t be immediately available cash, ItI_t scheduled cash receipts, BtB_t ordinary obligations, OtO_t option-exercise payments, MtM_t hedge and collateral outflows, and FtF_t committed available financing over a specified horizon. The pre-sale funding gap is

Gt=max⁡{0,Bt+Ot+Mt−Ct−It−Ft}.G_t=\max\{0,B_t+O_t+M_t-C_t-I_t-F_t\}.

This is a cash-flow identity under the stated dates and available resources. It does not imply that every line of credit is unconditionally drawable, that every asset is saleable at a quoted price or that collateral is free of encumbrance. A complete implementation specifies haircuts, legal-entity restrictions, funding counterparties, settlement delays and portfolio concentration.

The insurer can close a gap by using cash, reducing discretionary distributions, redirecting incoming cash, selling investments, borrowing against eligible collateral, unwinding positions or obtaining other funding. Each action carries economic, regulatory and operational consequences. A bond held to maturity generates scheduled principal and coupon cash; selling it before maturity realizes the prevailing market price. In a rate-rise scenario that price may be below amortized cost. An insurer that expects to benefit economically from lower present-value liabilities can nevertheless face a cash shortfall before those distant benefits are realized.

The NAIC interest-rate analysis explicitly associates rate increases with potential cash-value withdrawals, falling bond market values and pressured sales. The 2025 U.S. Federal Insurance Office report likewise identifies policyholder behavior, economic conditions, credit downgrades and derivative margin requirements as liquidity drivers; it describes increased surrender activity in the 2024 U.S. life and health sector while observing that the sector continued to meet operating and financial needs. Both findings are portfolio-level observations and should not be converted into the claim that all insurers faced a liquidity crisis.

A transparent funding illustration

Consider a hypothetical portfolio containing 100,000 contracts, each with a payable surrender value of 10,000 monetary units during the selected scenario year. Total current surrender exposure is one billion units. A baseline surrender assumption of 4% produces 40 million units of expected gross payments. In an adverse hypothetical state, a 14% surrender fraction produces 140 million units. The increase of 100 million measures the incremental contractual payment exposure under these simplifying assumptions; it is not yet a measure of loss or regulatory capital.

Assume the adverse-state 140 million is payable during the chosen funding horizon and 70 million of cash plus 20 million of reliably scheduled receipts are available for those same dates. With no other cash demands and no immediately drawable external funding, the funding gap is 50 million. If the insurer must sell securities at a 5% discount to their then-current fair value, the fair value of assets it must dispose of is 50/(1−0.05)=52.63157950/(1-0.05)=52.631579 million. The additional economic transaction loss associated with that discount is 2.631579 million. These values are rounded to six decimal places.

The 5% haircut is applied to the stressed current fair value, not to original book value. If the market rate shock already reduced the securities' fair value, subtracting that full decline again as a transaction cost would double-count the same repricing. Conversely, a valuation gain from lower technical provisions does not supply cash for settlement. The demonstration isolates three quantities: the 140-million cash obligation, the 50-million funding gap after available resources, and the 2.632-million incremental disposal loss. The first is a contract settlement amount, the second a horizon-dependent liquidity gap, and the third a conditional transaction expense.

The result is deliberately sensitive to the assumptions. If settlement is spread over months and scheduled coupons arrive earlier, the funding gap can be smaller. If collateral calls coincide with surrenders, it can be larger. If liquid assets become less marketable or dealer capacity is constrained, the haircut and execution time may worsen. If surrender charges reduce the amount paid or protect existing holders from dilution, the gross outflow can change. An analyst should therefore vary the cash horizon and market-depth assumptions together.

Investment allocation and marketability

An ALM portfolio can match long-term expected cash flows while holding assets that are costly to sell rapidly. Public sovereign bonds, investment-grade credit, private debt, infrastructure, mortgages and structured instruments differ in trading volume, collateral eligibility, quotation quality, contractual lockups and settlement practice. Duration matching alone does not establish liquidity matching. A private loan scheduled to amortize over ten years may economically correspond to long-life liabilities but supply little cash when a surrender event must be paid next week.

The BIS review of changing life-insurer portfolios documents increased exposure to alternative and less-liquid assets and growth in asset-intensive reinsurance. The BIS Paper 161 analysis places the structural shifts in a wider framework of private equity ownership, offshore reinsurance, derivatives and interconnected liquidity risks. These are structural risk channels, not proof that private assets are inappropriate for every insurance balance sheet. Their suitability depends on claim duration, option exercise rights, concentration, asset quality, governance, funding resilience and disclosure.

Liquid-asset inventories should be net of encumbrance. Assets posted as collateral, pledged to secured funding or legally restricted to a separate account may be unavailable for general-account surrenders. Rehypothecation rights and collateral transfer terms add counterparty exposure. Asset-backed liquidity may also fail when credit spreads widen or rating triggers change eligibility. A robust cash waterfall reports immediately available cash separately from contingent sources that require markets or counterparties to remain open.

Derivatives and margin feedback

Derivatives can mitigate interest-rate risk while creating immediate liquidity needs through variation margin. A receiver swap protecting the liability value against declining interest rates can experience cash collateral outflows under rising rates even as a wider ALM measure improves. Collateral settlement may precede asset maturities and policyholder payments. That ordering can require asset sales and create a secondary funding channel.

The relevant test is simultaneous rather than sequential: revalue hedges, compute collateral flows and surrender claims on the same economic path, then allocate unrestricted cash across their due dates. Treating the hedge as a purely offsetting market-value position while ignoring margin can materially understate the required liquidity buffer. The EIOPA 2024 financial-stability assessment identified lapse risk and portfolio liquidity among the major reported drivers of insurers' liquidity concerns, while also flagging derivative-related margin requirements in stressed conditions.

Reinsurance, funds withheld and group funding

Asset-intensive reinsurance can transfer financial risks and supporting assets to reinsurers subject to treaty terms and regulatory treatment. The legal ceding insurer may retain contractual obligations to its policyholders and acquire counterparty credit, collateral, recapture and group concentration risks. A funding shortfall at an assuming reinsurer or a treaty termination can move liabilities and asset-management responsibilities back to the cedant, sometimes during adverse markets. The 2024 BIS analysis of reinsurance structures describes these mechanisms and the role of offshore jurisdictions.

A valid group ALM stress cannot count cash in one legal entity as instantly available to another without demonstrating permitted transfers, dividend capacity and regulatory approvals. Similarly, assets held in funds-withheld arrangements or pledged custody accounts need their actual contractual availability checked. External reinsurance may reduce certain capital needs while increasing the importance of collateral, transparency and cross-border coordination. The same policyholder option can thus create a different financing response depending on the location of assets and risk.

Management actions, incentives and safeguards

Insurers can adjust crediting rates, renewal rates, product terms for new business, hedge positions and investment mix within legal and contract limits. Such actions may improve retention but affect margins and future guarantees. A decision to raise credited returns may reduce withdrawals while extending costly obligations; a decision to sell liquid bonds can meet current claims but leave the remaining portfolio less marketable. Asset sales, repricing and borrowing therefore belong inside the forward-looking scenario rather than being added as free mitigants after the loss calculation.

Governance should specify the order of permitted responses, their lead times, control owners and measurable capacity. Contractholders' legal rights and fair treatment remain constraints on financially attractive management actions. In a solvency or liquidity stress, a planned discretionary measure is credible only if the organization can execute it at the assumed scale and timing without violating the contract, accounting rules or applicable supervision.

The central ALM conclusion of this chapter is a timing result. Financial optionality affects the liabilities' economic amount and expected duration; asset response affects both the insurer's market exposures and its ability to pay at specific dates. The interaction is visible only when the liabilities, investments, collateral and funding sources share one chronological cash-flow ledger.

Dynamic Behavioral Feedback — Stress Testing, Model Limits and Verification

The endogeneity of insurance balance-sheet response

A dynamic ALM projection must allow a financial event to change policyholder behavior and then allow that behavior to change the financial state of the insurer. A simplified annual projection that applies the same lapse rate regardless of yield, guarantees and insurer management is useful as a baseline. It cannot fully describe a liquidity episode in which a rate shock changes surrender incentives, withdrawals force asset disposals, those disposals reduce future coupon receipts, and the insurer subsequently raises credited rates or alters its investment strategy.

One can describe the feedback with four existing classes of equations, without claiming a novel model. First, economic factors such as the yield curve YtY_t, credit spreads sts_t, unemployment utu_t and market liquidity mtm_t evolve under a stipulated scenario or stochastic process. Second, conditional policy actions are generated by qi,t=g(ci,Xi,t,Yt,st,ut,managementt)q_{i,t}=g(c_i,X_{i,t},Y_t,s_t,u_t,\text{management}_t), subject to the contract's feasible actions. Third, these actions create settlement cash flows OtO_t and the insurer revalues its assets and liabilities. Fourth, a documented response rule determines sales, collateral postings, borrowing, crediting changes and future portfolio composition. Repeating the sequence for each date creates the state path used to assess future exposures.

The critical point is simultaneity across economic channels, even when operational decisions occur in an ordered daily schedule. A market-rate rise may simultaneously alter the option value of surrender, the market value of an asset, derivative collateral, funding spreads and the price at which securities can be sold. If the simulation computes each effect against a different economic curve or different observation date, the resulting cash and capital numbers are internally inconsistent.

Some feedback enters through external markets. An individual insurer's sales usually need not move aggregate market prices. During a broad industry shock, however, correlated sales may encounter limited dealer capacity and adverse market depth. Whether that mechanism is material depends on the scale of aggregate selling relative to trading volumes, asset concentration and the willingness of other investors to transact. The BIS analysis of sector transformation discusses growing connections between insurer balance sheets and broader funding markets. A systemic liquidity spiral is a scenario to be investigated, not an assumption to impose on every insurer.

Path dependence and policyholder guarantees

A current exercise rate does not fully describe the future liability state. The set of policies remaining after a surrender wave may have different guarantees, ages, risk profiles and cash values from the original portfolio. Financially motivated withdrawals can selectively remove contracts with low guarantees while leaving more valuable legacy promises in force. This selection alters the guarantee burden per remaining policy and can extend liability duration. The insurer's ability to reinvest surrendered funds also depends on whether assets were sold, held or replaced.

For example, suppose a product cohort consists of two equal groups by policy count, one with attractive old guarantees and another with weaker credited rates. If 2% of the first group but 18% of the second group surrenders in a rate shock, the overall count-weighted surrender rate is 10%. Yet the remaining cohort contains more of the expensive old-guarantee business. Forecasting later liability duration using the original product proportions would misstate persistence. This example is a deterministic selection demonstration; its rates are assumed, not empirical.

A surrender charge may fall away at a contract anniversary, creating concentrated future exercise even if current withdrawals are low. Similarly, a guaranteed rate may reset, a bonus may vest, or a retirement-age threshold may open annuitization rights. These scheduled contractual dates are known information and should be preserved in the projection. They can generate cash-flow cliffs that annual average assumptions obscure.

The interactions are consistent with the option decomposition in Bacinello (2003) and the observed influence of contract duration in Eling and Kiesenbauer (2014). The studies support different components of the analysis: contingent-claims sensitivity in one case, observed heterogeneity in the other. Together they justify explicit contract-state transitions without asserting that one empirical coefficient is sufficient for all markets.

A stress matrix with distinct transmission channels

A robust analysis should present multiple paths instead of one mechanical increase in lapse rates. A low-yield persistence stress retains costly guarantees and extends liability duration. An upward-rate replacement stress increases competitive yield incentives and possible surrender settlements while lowering the value of existing fixed-rate assets. A credit-and-liquidity stress widens asset spreads, reduces dealer liquidity, increases collateral requirements and may also affect policyholder confidence. A household-income stress increases emergency withdrawals without requiring a yield arbitrage incentive. A combined stress applies these factors jointly, with correlation assumptions made explicit.

Each path should record the scenario horizon, affected contract population, withdrawal timing, economic state, contractual payment rules, available liquidity, securities sold, funding drawn, realized losses, surviving liability states and residual capital. Its design must avoid treating an extreme value of every individual risk variable as a coherent joint economic event without a causal story or probability interpretation.

The 2024 EIOPA insurance stress test offers a useful institutional precedent for simultaneous capital and liquidity assessment. It examines an adverse geopolitical and economic scenario and distinguishes a fixed-balance-sheet approach from constrained management actions. EIOPA reports that the tested insurers had significant capital buffers and sufficient liquid assets to meet modeled cash needs, even though cash holdings alone could be insufficient. The report also recommends improvements in some participants' liquidity-flow and look-through modeling. The result describes the participating European sample and scenario; it does not certify every insurer or eliminate future liquidity risk.

Simulation design and nonfinancial behavior

For large portfolios, stochastic projection can sample market paths and contract actions jointly. A two-stage procedure simulates financial and macroeconomic states and then conditional insurance events, while a feedback procedure updates the insurer's balance sheet and management choices after settlements. The events can be generated as dependent random outcomes through common risk factors, latent shocks, hierarchical effects or other justified mechanisms. The tail of cash demand may differ materially from a binomial approximation with independent policies when common shocks dominate.

The probability measures must be consistent with each output. Real-world simulation supports expected cash demand, stress distributions and economic-capital assessment under a defined risk appetite. Market-consistent valuation of hedgeable option components may require risk-neutral techniques, while nonmarket behavior and incomplete insurance markets call for transparent assumptions. A model that mixes real-world lapse frequencies with risk-neutral asset returns without reconciling the measure can produce a number that lacks a coherent interpretation.

Calibration can be constrained by history. Severe modern rate regimes, private asset markets, distribution technology and new products may not be represented in long runs of older policy data. Parametric extrapolation should be labeled as such, and stress overrides documented with their origin. A credible high-surrender scenario is not established merely because its chosen percentage is large. Credibility comes from contract eligibility, incentive structure, household capacity to act, observed time lags, peer comparisons and operational settlement capacity.

Behavioral feedback may also be affected by management and conduct. Crediting-rate changes, customer communications and retention offers can alter exercise probability. Since these are choices by the insurer, they should be specified as controlled actions with implementation delays. The same historical data may show that rates were raised in response to potential lapses, producing reverse causation in a regression of lapse against credited rates. The correct scientific statement about the observed association remains conditional until an identification strategy supports a stronger claim.

Independent model validation

Validation begins at the contract level. The analyst verifies legal exercise eligibility, surrender calculations, fees, charges, guarantee persistence and event timing against actual policy documents. Next comes the data audit: reconciliation of opening and closing in-force counts, surrender notices, payments, partial withdrawals, loans, deaths and conversions. This should be followed by statistical tests of predicted actions under out-of-sample periods, with separate measures for the number of actions and the cash value exercised.

At the ALM level, validation becomes an accounting identity and scenario-reconciliation exercise. For each date, opening cash plus receipts and asset-sale proceeds less benefits, surrenders, expenses, collateral and financing service must equal closing cash. Opening market assets plus realized investment returns and trades must reconcile to ending assets, allowing for market revaluation and losses. Liability values must be recomputed from the surviving contract states; they cannot be updated solely by subtracting surrendered cash from a stale reserve. Capital results should be computed under the same economic scenarios and regulatory definitions.

Independent benchmark calculations help expose numerical errors. The duration calculation of a one-percentage-point rate shock, the two-state surrender-value exercise and the 140-million settlement stress are sufficiently simple to be checked with exact or high-precision arithmetic. Complex Monte Carlo systems should reproduce these limits under constrained inputs. A monotonicity assertion, however, must be tied to controlled assumptions: increasing the surrender rate does not always decrease insurer value because the profit contribution, guarantee cost and surrender payout differ among products.

Model governance and decision use

An integrated model requires documented ownership of the interest-rate process, behavior assumptions, asset valuations, legal contract mapping, reinsurance recoveries, collateral rules and management actions. Each component needs its data lineage and version. A model change that alters surrender probabilities should trigger checks of liquidity and capital results, and a change to investment liquidity haircuts should trigger review of funding policy. The point is to preserve the causal and computational connections, not simply to maintain separate validated spreadsheets.

The IAIS Insurance Core Principles, particularly ICP 16, require an integrated enterprise risk management framework that covers ALM, liquidity, investments, risk interdependencies and own risk and solvency assessment. The NAIC ORSA framework likewise places current and prospective risk, including liquidity and market risk, inside management's assessment of risk and capital. These frameworks provide supervisory foundations for the integrated model-governance conclusion.

The output should distinguish limits that can be implemented from metrics that merely look precise. A forward-looking liquidity buffer, concentration ceiling, collateral reserve or contingent-funding plan requires a named entity, quantity, time horizon and enforcement mechanism. An extreme but statistically uncertain cash-demand distribution can still support prudent governance when the uncertainty is communicated and measured through sensitivities.

Liquidity and Solvency Cases — Integrated Scientific Interpretation and Implications

The European stress-test evidence

The EIOPA 2024 stress-test results establish three linked observations. The participant sample began with substantial solvency resources; a severe macro-financial shock could still cause a large decline in capital; and liquid asset holdings could cover projected outflows despite temporary insufficiency of cash balances. The exercise's aggregate initial solvency ratio was reported as 221.8%, and EIOPA's scenario involved more than €270 billion of capital decline before the mitigating effect of allowed management responses. These are reported outputs of that exercise, not forecasts for November 2026 and not data about a randomly selected European insurer.

The distinction between immediate cash and saleable assets makes the stress informative for policyholder optionality. A surrender shock can be payable before ordinary asset maturities, forcing asset conversion without necessarily implying insolvency. A reduction in regulatory own funds can occur without a corresponding cash outflow; conversely, a liquidity shortage may arise even while a portfolio has positive economic surplus. Capital adequacy and liquidity sufficiency therefore require separate metrics and a consistent common scenario.

The EIOPA June 2024 financial-stability report stated that lapse risk was a leading perceived liquidity driver in its insurer survey and noted that portfolios had remained sufficient to meet the additional surrenders then reported. It also identified derivative margin calls as a potential source of strain. This combination supports the need to model risk channels jointly while retaining the evidence that supervisors found resilience in the observed period.

The October 2026 EIOPA Risk Dashboard reported liquidity and funding risk at a medium level, with Q2 2026 median cash holdings of about 0.7% of total assets and median liquid assets of about 46%. These median measures describe different liquidity concepts and are not an immediately available cash balance for any single firm. The dashboard's contemporary assessment helps date the sector context; the historical stress-test results remain tied to their own 2024 calibration.

The U.S. experience and the limits of aggregate inference

The NAIC discussion of interest-rate transmission identifies the competitive-yield mechanism: rising rates can encourage withdrawals or replacement of cash-value policies and annuities, while also lowering the value of existing bond investments. The Federal Insurance Office's 2025 annual report observed that elevated rates contributed to higher surrender activity in the U.S. life-and-health sector during 2024, but the sector maintained its ability to meet obligations, supported by strong premium growth and operating cash flow. The report also noted increasing investment illiquidity and pressure on a current liquidity measure.

Aggregate resilience does not rule out firm-specific fragility. An insurer with highly surrenderable contracts, short-term collateral obligations and a concentrated illiquid asset portfolio may face a materially different profile from a sector average. Conversely, a diversified insurer with durable guarantee value and a ladder of readily marketable assets may experience modest net withdrawals even in a higher-rate environment. Market depth and legal-entity funding access matter as much as headline lapse rates.

At the NAIC's August 2026 summer meeting, the macroprudential liquidity materials presented 2025 U.S. life-sector surrender benefits of approximately $487.6 billion, a 0.7% increase over 2024, and an average surrender-to-premiums ratio of 61.1%. The same materials reported that modeled extreme liquidity stress for the 26 large life insurers subject to the discussed stress test entailed relatively small aggregate market sales compared with benchmark daily bond-market turnover. These statements refer to an NAIC meeting presentation and its modeled sample; they are not independently verified transaction-level results and do not imply that every insurer could liquidate its portfolio without cost.

Together these materials show why a credible assessment should combine insurer-specific contract terms, liquidity horizon and portfolio marketability with supervisory sector evidence. A high volume of reported surrenders can coexist with positive operating inflows and a manageable liquidity position. The analytical task is to reconcile both sides of the cash-flow account rather than interpret the gross outflow alone as failure.

German life-insurance persistence and contractual design

The Deutsche Bundesbank's 2024 financial-stability review provides a different institutional setting. Fixed surrender values can make reinvestment attractive when bond yields rise and asset values fall. Yet the Bundesbank assessed the risk of a broad surge in policy lapses as limited and observed relatively moderate responses. This evidence warns against converting the existence of a contractual exercise incentive into an automatic mass-withdrawal assumption.

The result is consistent with the empirical finding that policyholder response varies by contract type and characteristics in Eling and Kiesenbauer (2014). In some portfolios, retaining insurance guarantees or protection dominates a simple market-yield difference. In others, large annuity account holders may react more directly to available reinvestment opportunities. Jurisdictional rules, product ages and distribution channels can produce different behavior even under similar yield movements.

The Eurovita case and the joint liquidity-solvency channel

Italy's Eurovita episode in 2022–2023 illustrates the conjunction of rate exposure, policyholder optionality and institutional response in an actual insurer. The BIS Paper 161 case study describes how rising yields reduced the market value of a fixed-income portfolio backing guaranteed contracts, while higher available rates increased incentives to surrender. The paper reports a decline in the insurer's solvency capital ratio from approximately 230% to approximately 130% by the end of 2022, followed by intensifying redemption demands and the need for additional capital. The episode culminated in special administration and a temporary suspension of surrenders in February 2023 while an industry-backed transfer solution was organized. The BIS synthesis draws on supervisory developments and reported contemporary information, including media accounts; its illustrative ratio and financing figures retain that source provenance.

Eurovita shows why a solvency issue and a liquidity issue can reinforce one another. Asset losses weaken the capital position. The same rate environment gives some policyholders a reason to withdraw. Paying those withdrawals can entail asset disposals and a further reduction in resources or future earnings capacity. The need for recapitalization, a supervisory intervention and a portfolio transfer then alters the feasible actions of the insurer and the rights available to holders during the intervention. The sequence is economically important because it contains several distinct stages, with different timing and legal mechanisms, rather than one unconditional claim about surrender rates causing failure.

The case supports a disciplined interpretation. Supervisory intervention and the subsequent preservation of contracts show that the realized policyholder outcome depends on the resolution framework and the participation of other insurers. The existence of one distressed life insurer cannot be generalized to the European market as a whole, particularly in light of the stronger aggregate resilience found in EIOPA's 2024 sample. The lesson for ALM is to model contract exercise, asset liquidation, capital erosion and supervisory constraints jointly under plausible scenarios. It is an observational case with multiple contributing causes; it does not provide a separately identified causal coefficient for surrender behavior.

Regulatory interpretation across systems

European and U.S. approaches share a concern with policyholder obligations, market exposures and prospective risk management but use different institutional instruments. In the EU, the Solvency II lapse-risk framework specifies capital stresses for lapse direction and mass lapse under defined conditions. EIOPA's stress-testing work supplements that framework by assessing liquidity alongside capital under broader scenarios. In the United States, NAIC ORSA promotes insurer-level prospective risk and capital assessment, while liquidity stress testing and state financial supervision supply additional perspectives. Internationally, IAIS ICP 16 emphasizes risk interdependencies and integrated enterprise risk management.

Financial reporting adds a separate perspective. IFRS 17 describes the measurement and presentation of insurance contract groups, while national statutory accounting systems may treat asset valuation and capital recognition differently. The prudential, accounting, economic and liquidity measures should be reconciled, yet each answers a distinct question. Cross-country comparisons must identify which regime, reporting period and legal entity the stated measure belongs to.

The EU 2025/2 amending directive introduces changes whose applicable national implementation generally starts on January 30, 2027. The EU insurance recovery and resolution framework also provides a prospective framework applying from January 30, 2027. Both must be described as future implementation at the November 2026 date of this monograph. They help frame future recovery and macroprudential analysis but cannot be represented as already binding in every Member State.

A comparative decision table

Decision question Primary economic object Necessary evidence Typical failure of interpretation
What is the surrender option worth? State-contingent contractual choice Contract terms, market curves, valuation assumptions Treating the value as observed exercise frequency
How many policies surrender? Conditional behavior over a defined interval Contract-level events, exposures, cohort variables Treating rates from one product or country as universal
How much cash must be paid? Settlement cash flow by date Exercise amounts, charges, notices, settlement lags Applying annual lapse counts to an immediate funding horizon
How do assets respond? Revaluation, income, sale capacity and collateral Portfolio holdings, maturity ladders, liquidity haircuts Equating fair value with freely spendable cash
How much capital changes? Economic or regulatory loss distribution Liability valuation, asset shocks, applicable regime Equating a solvency ratio with a liquidity coverage ratio
Which actions are credible? Legally and operationally executable response Authority, contracts, funding commitments, lead times Assuming unconstrained sales or unlimited group transfers

The table is an analytical synthesis of the reviewed insurance-finance literature and supervisory materials. It is a decision-audit aid rather than a newly proposed theoretical framework.

Integrated interpretation

The evidence supports a conditional assessment of amplification. Optionality becomes more consequential where accessible surrender rights coincide with increased alternative yields, reduced market values of backing assets, concentrated cash-settlement dates and limited liquidity. Amplification may be muted where protection value, surrender charges, policy tenure and liquid holdings moderate exercise and settlement. A change in interest rates can also improve economic surplus under one duration configuration while weakening the liquid asset position in the same scenario.

Scientific interpretation must preserve scale. A contract-level option value is a property of terms and valuation assumptions. A portfolio-level withdrawal estimate aggregates heterogeneous behavior. A firm-level funding gap adds liquidity resources and restrictions. A sector-level financial-stability assessment also needs market depth, correlated behavior and cross-institutional connections. Movement between these scales requires a documented aggregation and identification procedure.

The policyholder's choice and the insurer's financial response are linked by contracts, dates and markets. The central empirical and modeling task is to preserve those links long enough to produce an intelligible answer to how much can be paid, when it is payable, what must be sold, what risk remains and whether the insurer can still honor its surviving promises.

Conclusion

Policyholder optionality is a material component of insurance finance whenever contractual exercise changes the timing, amount or persistence of insured obligations. Its effects on asset-liability management are determined by the interaction of contractual guarantees, state-dependent behavior, investment performance, liquidity and the financing choices of the insurer. A full analysis must therefore follow the life of the contract through the cash settlement and subsequent balance-sheet response.

The reviewed contingent-claims literature establishes that surrender rights and guarantees can have separate, economically significant values. Empirical lapse research establishes that actual exercise varies with market conditions, household circumstances, product design and contract duration. Supervisory evidence from the United States and Europe shows that rising rates, lapses, less-liquid assets and derivative collateral can be connected sources of strain, while the studied sectors and stress-test participants have also demonstrated substantial liquidity and capital resilience. These findings are compatible because economic option exposure, observed surrender behavior and institutional solvency are different measurements.

Four conclusions follow from the synthesis. First, policy actions must be classified by their contractual state transitions, payment values and effective settlement dates. Second, rate and spread scenarios must jointly revalue assets, insurance guarantees and policyholder choices while preserving the distinction between financial valuation and real-world behavior. Third, liquidity planning must measure unrestricted cash, scheduled inflows, collateral needs, sale capacity and funding arrangements by legal entity and horizon. Fourth, validation must reconcile modeled behavior, asset trading, cash movements and economic or regulatory capital under the same scenarios.

The hypothetical calculations provide transparent benchmarks: a two-state surrender privilege adds 3.846154 units of contingent-claim value under the stipulated assumptions; a static one-percentage-point rate rise improves surplus by 21 million in the illustrative negative-duration-gap portfolio; and a 14% surrender event can create a 50-million funding shortfall and 2.631579 million additional sale-cost loss in the separate liquidity illustration. These figures are independent pedagogical examples with specified assumptions, not empirical predictions or a claim that the examples describe one common insurer.

A disciplined research program combines legally defined contract options, empirically tested exercise behavior, actuarial cash-flow models, dynamic investment and collateral accounting, and model governance. Such a program can distinguish solvency resilience from settlement capacity, quantify sensitivity without mistaking a benchmark for a forecast, and identify where portfolio-specific feedback creates material risk. The analytical synthesis extends the application of established insurance finance and risk science. It asserts no new named theory, doctrine, principle or proprietary model.

References