Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Policy Surrender Runs and Disintermediation Risk

Abstract. A life insurer's contractual promise to pay a surrender value makes a long-term insurance liability potentially payable before the investments supporting it mature. Aggregate exercise of these contractual options may produce an insurer-specific liquidity event, a market-wide reallocation of household savings, or a self-reinforcing run when policyholders believe that delayed redemption may worsen their position. This monograph investigates the economic, actuarial, legal, and institutional conditions connecting these outcomes. It separates ordinary lapses from cash surrenders, disintermediation from distress-driven withdrawals, market-value solvency from the timing of cash settlement, and an insurer's own asset sales from their cross-market feedback. Historical analysis of late-twentieth-century rate shocks and the Eurovita episode is integrated with 2025–2026 supervisory data, recent empirical estimates, international risk standards, and a reproducible contingent-liquidity stress illustration. The central analytical finding is conditional: surrenderability generates a liquidity transformation problem when the speed and contractual value of withdrawals outrun realizable asset liquidity; correlated incentives, adverse information, and price-impact feedback increase that problem. Contractual frictions and disciplined balance-sheet management can attenuate it, while illiquid investments, encumbrance, derivatives collateral and concentrated distribution may intensify it. The argument supports product-specific, scenario-dependent governance and supervisory monitoring, with policyholder treatment evaluated alongside financial resilience.

Keywords: Life insurance; policy surrender; disintermediation; liquidity risk; surrender value; lapse; bank-run theory; asset–liability management; mass lapse; interest rate risk; fire sales; solvency; insurance supervision; Eurovita; stress testing; financial stability.

Introduction

Life insurance transforms household contributions into investment portfolios supporting death benefits, accumulated cash values, deferred annuities, and retirement claims. The temporal design of the balance sheet assumes that many obligations will mature over extended horizons. Yet the policyholder may possess an earlier contractual right to extract value. A change in the interest-rate environment or confidence in the insurer can convert that right from a rarely exercised option into a major liquidity demand. The right's specification, its collective exercise, and the marketability of investments jointly determine the insurer's exposure.

The subject therefore calls for a distinction between three related processes. A lapse records termination of contractual coverage, sometimes without a material cash payment; a surrender normally entails an early value payment under an eligible savings or cash-value contract; and disintermediation occurs when funds move away from an insurer, often toward competing instruments as relative yields change. A surrender run refers to an unusually concentrated acceleration of requests driven at least partly by expectations, social information, or concerns about an insurer's ability to honor claims promptly. These terms describe different observables, and their interchangeability would make historical measurement and stress calibration unreliable.

The research question is how a contractual option on insurer liquidity interacts with portfolio duration, cash realization, capital resources, policyholder beliefs, financial regulation, and the behavior of other institutions. A secondary question is why rising interest rates can improve reinvestment prospects while simultaneously exposing individual insurers to immediate claims on liquid funds and to losses on the sale of existing fixed-income holdings. Earlier work in the Federal Reserve's account of disintermediation and the modern life insurance convexity study demonstrates the historical persistence of this tension under different products and market structures.

The method combines conceptual analysis of contracts, a selective critical review of published empirical and theoretical research, close reading of official primary supervisory material, and internally calculated hypothetical scenarios. Institutional propositions are assigned to the jurisdiction and date of the source. An illustrative scenario is labeled as such and makes no claim to reproduce an insurer's confidential cash flows. When journal full text is subscription restricted, the analysis confines source-specific factual attribution to verified bibliographic records and accessible abstracts. This research advances an integrated interpretation of established mechanisms; it makes no claim to introduce a new theory or priority-displacing mathematical model.

Surrender Options and Incentives

Contractual definitions and heterogeneity

An insurer's liability to a terminating policyholder depends on the precise contract. In traditional participating or nonparticipating cash-value life insurance, withdrawal can trigger payment of a stipulated or formula-based surrender value. For some annuities, early withdrawal is subject to a schedule of surrender charges, market-value adjustments, tax consequences, or restrictions on timing. Unit-linked and variable products generally allocate some investment performance to the policyholder, changing the incentives and asset-liability linkage. Policies without positive surrender value may lapse without an equivalent monetary outflow. Loans secured by policy values can also shift funds out of the insurer while the contractual coverage remains in force.

The probability that a policy is surrendered does not identify the amount of liquidity required. A portfolio with many small contracts, modest guaranteed cash values, and substantial charges can have a higher count of surrenders and a smaller cash shock than a portfolio of few large, interest-sensitive savings contracts. Exposure mapping therefore needs the distribution of eligible surrender values, contract anniversaries, charge schedules, minimum guarantees, loan availability, and policyholder cohorts. The German lapse study by Eling and Kiesenbauer (2014) analyzed more than one million contracts and found material product- and policyholder-level differences; it is evidence for heterogeneity rather than a universal withdrawal parameter.

A cash payment today changes the insurer's remaining obligation to the departing policyholder. It may also remove future premiums, recurring fees, investment margins, and future death or living benefits. The economic cost of a surrender is thus the sum of immediate liquidity effects, the release or increase of obligations, associated tax and transaction effects, and any consequences for the surviving policyholder pool. A high surrender value relative to the contemporaneous economic value of future benefits can create an exercise incentive, but the relative position depends on guarantees and contract terms, not only on the movement of government bond yields.

Household exercise and opportunity cost

For an individual policyholder, surrender entails a tradeoff among current cash, alternative yields, insurance protection, surrender charges, taxation, health-related re-underwriting, and expected future dividends or bonus allocations. A simple comparison of the declared rate on an old policy with a new deposit rate can overstate the monetary advantage of early withdrawal, because replacing insurance protection may be costly or impossible and the relevant comparison uses net, risk-adjusted, after-tax values. Conversely, low charges and a highly predictable cash value can make interest-sensitive products unusually mobile.

A stylized exercise comparison can be written as E_t = PV_t(alternative investment after costs) + liquidity value of cash − surrender charges − lost coverage value − PV_t(value of continuing policy). A positive E_t indicates a privately attractive exercise under the household's information and preferences. This is an accounting device for organizing incentives, rather than a calibrated behavioral equation or an assertion that households optimize continuously. Many households respond to advice, distribution channels, salience, liquidity needs, or misunderstanding; policyholder advice must therefore be separated from mechanical balance-sheet prediction.

A further asymmetry is that the insurer has selected and valued its investments for expected claim timing. The policyholder retains the legally specified right to choose the timing of a particular withdrawal. The firm's aggregate surrender assumption is consequently a contingent financial input that responds to the economy and the firm itself. The IAIS supervisory principles recognize lapse or withdrawal behavior, options and guarantees, asset marketability and liquidity management as linked enterprise risks.

Disintermediation and the historical record

US experience provides a useful historical antecedent. Curry and Warshawsky (1986), in their Federal Reserve Bulletin study, reported that unusually high inflation and interest rates in the 1970s and early 1980s made alternative investments attractive, accelerated voluntary terminations of ordinary life policies, and encouraged loans against policy values at contractual rates below market rates. The reported episode combines rate-driven substitution with pressure on insurer cash flows; it does not establish that all departures were panic withdrawals.

A distinct crisis mechanism arises when capital constraints alter the prices at which insurers are willing to write new business. Koijen and Yogo (2015) identify unusually steep discounts in the pricing of new US life and annuity products during 2008 and link them to capital and reserve frictions. This establishes the importance of insurer financing conditions without directly measuring a surrender run: product pricing under capital scarcity and withdrawals from outstanding contracts are different margins of adjustment.

The historical evidence has two implications for contemporary interpretation. First, a sharp rise in rates changes the household's outside option before it necessarily changes the return credited by a long-duration insurer. Second, policy loans may create disintermediation-like demands that are missed by a data series limited to full surrenders. A study that observes only lapse count, without associated payments or loans, cannot measure the liquidity burden. The historical products, market access, and regulatory environment also differ from current index-linked annuities, unit-linked contracts, and private-asset-backed liabilities. Quantitative inference across these periods requires explicit adjustment rather than superficial historical analogy.

Interest-Rate Disintermediation

Two-sided valuation consequences

Consider a parallel increase in market yields. The fair value of a fixed-rate bond portfolio declines, while the present value of certain fixed insurance obligations can also decline. Over time, new investments may earn higher returns, improving the economics of reinvestment. These gains and losses arise at different horizons. Immediate surrender payments, however, require cash on dates chosen under the contract and may force asset realization before the insurer can earn the new yield or wait for contractual maturity.

A simplified local duration approximation, ΔV/V ≈ −D_mod Δy + 0.5 C(Δy)^2, describes the market-value sensitivity of assets or fixed liability cash flows to a small yield change, holding other determinants constant. The insurer's economic surplus S = A − L has first-order sensitivity ΔS ≈ −A D_A Δy + L D_L Δy. If liability duration exceeds the asset duration, higher yields may improve economic surplus in a static calculation. Yet if policyholders simultaneously demand cash, the insurer can have a cash shortfall while showing positive modeled economic surplus. These are different risk metrics, and neither should substitute for the other.

The BIS 2024 analysis of life insurers explains how duration, low-rate profitability pressures, and more complex asset-intensive reinsurance structures affect modern portfolios. The BIS Paper 161 systemic-risk review identifies policy surrenders, derivatives margin, short-term funding and less transparent investments as potentially interacting sources of liquidity exposure. Their presence should be evaluated separately by entity, collateral location, and liquidity access: group-level assets may be legally, operationally, or economically unavailable to meet an immediate local-entity payment.

Optionality, negative convexity, and competing yields

The surrender option changes the effective duration of liabilities as rates change. A high-rate scenario may increase the chance that selected policyholders exercise a guaranteed withdrawal right, causing the duration of the remaining liability pool to shorten or changing its composition. Asset-liability immunization that performs under fixed cash-flow assumptions can then fail under dynamically exercised options. The empirical and calibrated findings of Kubitza, Grochola, and Gründl (2025) connect rate shocks to higher surrender rates, insurer asset sales, and bond-market effects. Their model-specific estimate of annual asset sales and price impact should not be applied as a universal prediction for any other insurer or jurisdiction.

An important refinement concerns contractual pricing. Some guarantees pay fixed cash values while market bond prices fall; these can create a gap between cash requested and the sale value of backing assets. Other designs use market-value adjustments or assign market risk directly to the investor, reducing part of the mismatch while introducing different consumer exposures. Charges and tax treatment may slow exercise, but they are contingent on the contract's age and state law. The timing of rate changes relative to policy anniversaries can produce a pronounced maturity or surrender-charge cliff even where aggregate annual rates look smooth.

Why withdrawal cash differs from reported liability

Surrender cash value, statutory reserve, IFRS or local accounting carrying amount, and economic best estimate are distinct objects. A reserve may be calculated with prescribed assumptions and capital requirements; an accounting liability follows reporting standards and may include different contract boundaries and cash-flow discounting. The cash payable under an elected surrender is governed by contractual and legal rules. Consequently, the reduction in reported liabilities upon surrender need not equal the money paid to the holder. An apparent improvement in a liability metric does not settle the insurer's realized gain or loss.

This distinction also affects regulatory comparison. Solvency ratios measure prescribed asset and liability valuations against capital requirements and can reflect transitional or volatility-related mechanisms; cash is needed to satisfy daily obligations. The ECB's financial stability discussion described the particular vulnerability that rising rates can create when life policies offer relatively low credited returns and insurers have fewer high-quality liquid assets. The problem has persisted conceptually through changing supervisory and market regimes, though its magnitude depends on present contracts and exposures.

Coordination and Run Dynamics

Aggregate lapse versus a surrender run

Policy surrenders can rise in parallel without being coordinated. All households may respond independently to the same outside interest rate, financial need, or product repricing. A systemic liquidity problem can develop from this common exposure alone if the insurer must sell assets on a large scale. A stronger meaning of run additionally entails strategic or informational interaction: each holder's decision is affected by what others have done or are expected to do, the perceived capacity to process redemptions, or a belief that contractual value may become inaccessible. Empirical work must look for such interaction rather than infer it from a steep time series.

The canonical economic reference is Diamond and Dybvig's (1983) bank-run analysis, which shows how demandable liabilities financing less liquid assets can produce a coordination problem. Its institutional detail is deliberately specific to deposits: life policies frequently entail surrender penalties, protection components, administrative processing, tax frictions and different legal remedies. Depositor insurance coverage also differs materially from jurisdiction-specific insurance guaranty arrangements. The appropriate scholarly use is an analogy about liquidity transformation and expectations, not the unqualified transfer of a bank-run equilibrium to every insurer.

Three pathways are distinguishable. The first is a yield-substitution wave, in which higher outside returns trigger a broad but largely fundamental movement of savings. The second is a confidence-driven wave in which emerging concern over the insurer changes withdrawal timing for reasons partly independent of interest rates. The third is a feedback wave: early withdrawals force less favorable sales, reduce financial flexibility or damage reputation, thereby changing incentives of remaining holders. All three can coexist and can also interact with legitimate hardship withdrawals. Evaluating them requires evidence about payments, timing, public announcements, intermediary communications and the nature of counterparties receiving the funds.

The possibility of a run should not obscure the fact that many insurer liabilities have durable funding. Premiums, mortality pools, surrender charges, crediting mechanisms and contract-specific restrictions can stabilize behavior. The 2025 Federal Insurance Office annual analysis emphasizes operating cash flow and the importance of surrender deterrents and liquidity buffers when explaining why bank-like runs are relatively remote for the US life sector as a whole. A statement about the sector cannot be read as a guarantee concerning individual entities or product cohorts.

A conditional behavioral and balance-sheet representation

Existing actuarial techniques represent exits through contract-level hazard rates conditioned on duration, age, product, charges, economic states and observed history. An illustrative standard logit relationship, log[p/(1−p)] = α + β_1 × (outside yield minus net credited yield) + β_2 × observed concern + β_3 × liquidity need + contract controls, makes the conditioning explicit. This equation is simply a familiar regression specification, not a newly proposed theoretical model. The signs and magnitudes must be empirically estimated; concern is endogenous, measurement is difficult, and observed exits may be influenced by omitted distribution practices. Using one pooled intercept across dissimilar insurance products is usually inadequate.

Aggregate expected payout over a horizon h is the sum across eligible contracts of the expected surrender indicator multiplied by the contractual amount payable upon exercise, with timing aligned to settlement rules. For stress testing, it is often more important to model the distribution of aggregate payout under correlated scenarios than to estimate an average annual lapse rate with minute precision. Correlation can arise from a common policyholder demographic, similar policy vintage, distribution network, or geographically concentrated exposure. Product cohorts with correlated surrender-charge expirations can create synchronized withdrawal opportunities even without an adverse solvency signal.

The insurer's short-horizon cash balance evolves as opening available cash plus actually collectible operating inflows plus available secured funding plus net sale proceeds, less surrendered cash values, benefits, expenses and collateral calls. Each term has a different degree of certainty, settlement delay and legal availability. Encumbered or ring-fenced securities must be excluded from immediate unpledged liquidity unless a lawful and operational release is feasible. Projected premium receipts should be stressed consistently with future policy terminations; otherwise the same surrender shock increases immediate payments while leaving assumed future inflows artificially intact.

Expected liquidity deficits and solvency deterioration are related through possible losses, but their directions and timing can differ. A cash shortfall might be bridged by liquid and creditworthy assets without substantive erosion of economic surplus. Conversely, an insurer can incur significant market-value losses and reduced capital without an immediate withdrawal spike. A joint framework requires both balance-sheet valuation and settlement-timing analysis. The IAIS ICP framework for integrated risk places liquidity assessment alongside market, insurance, concentration and group risks rather than treating it as a stand-alone number.

Endogeneity, identification, and model boundaries

A quantitative paper reporting that rates cause surrender behavior faces a difficult identification problem. Interest rates covary with inflation, household income, financial wealth, debt-service burdens, competitor returns, bond valuation, and company credit conditions. A correlation between higher rates and surrenders may reflect several of these channels. Kubitza, Grochola, and Gründl (2025) use monetary-policy variation in their analysis; the existence of their identification strategy is documented in the publisher abstract, while transferring its numerical coefficients to another population would require a separate external-validity argument.

Estimating a coordination effect requires additional observables. Relevant evidence could include event-time discontinuities around credible announcements, shifts in redemption requests following distributor messages, comparison groups without similar information exposure, and observed changes in withdrawal behavior controlling for outside yields. Insurer administrative delays may themselves both follow and cause increased concern. The resulting simultaneity can distort a naive regression of exits on public confidence. A robust analysis should record submission, processing and actual payment dates separately and should distinguish attempted from completed surrenders.

The distinction between an early-exercise incentive and a destructive run is economically consequential. If an insurer designed sufficiently liquid backing for the contractual option, a high rate of surrender can represent ordinary contractual performance. The label run is justified when concentration and feedback threaten the institution's capacity to perform on fair terms or force behavior with material market spillovers. Such a definition rests on observable mechanisms and does not presume that exercising a lawful contractual right is improper conduct.

Investment Portfolio Liquidation

A reproducible liquidity waterfall

An insurer facing large cash outflows ordinarily attempts to meet obligations through its contractual cash inflows, cash balances, maturing securities, freely available liquid assets and suitable funding lines. Each source has costs and operational constraints. Cash and short sovereign bonds are generally more easily monetized than many private loans, structured products or real estate assets, but market depth, settlement convention, currency and collateral status matter. Selling a bond at its post-rate-shock market price also differs from selling at a liquidity discount below that price. The first is market repricing; the second is additional execution loss or price impact. Their separate identification prevents double counting.

A credible waterfall starts with a contract-specific schedule of payable surrenders and other claims, then deducts realistically available cash and inflows, then measures secured borrowing capacity after eligible collateral haircuts, then considers sale proceeds of unrestricted liquid positions, and finally identifies the remainder that would require illiquid monetization, external support or emergency actions. Funding lines should be tested for covenants, collateral calls, counterparty withdrawal and settlement latency. Reliance on the same corporate bonds as both planned sale inventory and collateral for committed borrowing overstates liquidity.

An empirical observation in the BIS 2025 industry transformation study is that evolving investment structures, derivative use, asset-intensive reinsurance and short-term financing may cause exposures to overlap. The ECB's 2025 analysis of insurer–sovereign links similarly describes less liquid asset allocation and collateral-dependent financial linkages in the euro area. The correct inference is that an asset's presence on a group balance sheet is not sufficient proof of its unrestricted availability within an adverse week.

Stress illustration and sensitivity

The following numbers are intentionally hypothetical. They describe a generic balance-sheet liquidity allocation in arbitrary currency units, with opening cash of 12, a liquid-securities inventory carried at 28, and an illiquid asset inventory carried at 60. Total assets in this simplified illustration equal 100 before concurrent market valuation changes. Cash inflows during the stress horizon are assumed to be 3. Sale of the liquid inventory realizes 96% of its pre-sale carrying amount; sale of the illiquid inventory realizes 80%. All carrying amounts are assumed to equal the chosen pre-sale reference for the illustration, so reported sale losses are simply the difference between this reference and realized proceeds. The example excludes taxes, regulatory capital ratios, reinsurance recoverables, secured borrowing and new premiums beyond the stated inflow.

Hypothetical scenario Surrender payments Other margin cash calls Available inflow Net immediate cash need Initial cash used Liquid assets sold at carrying value Illiquid assets sold at carrying value Sale loss versus specified carrying values
A: elevated exercise 25 8 3 30 12 18.75 0 0.75
B: concentrated outflow 45 12 3 54 12 28.00 18.90 4.90
C: severe correlated wave 60 15 3 72 12 28.00 41.40 9.40

In A, liquid sales of 18.75 yield 18.00 after the 4% discount and exactly fill the net cash gap of 30 − 12 = 18. In B, the entire liquid portfolio yields 26.88, leaving 15.12 after cash; an illiquid sale of 18.90 realizes 15.12. The corresponding loss is 1.12 + 3.78 = 4.90. In C, the remaining 33.12 requires monetizing 41.40 of illiquid assets at 80%, producing total realized losses of 1.12 + 8.28 = 9.40. These arithmetic results are contingent on the chosen haircut assumptions and do not represent an estimate for a real insurance undertaking.

A further sensitivity exposes the economic importance of market depth. If the illiquid-sale realization rate falls from 80% to 70%, Scenario B requires 15.12/0.70 = 21.60 of illiquid inventory and the total sale loss becomes 1.12 + 6.48 = 7.60. A fixed cash need therefore requires the liquidation of substantially more inventory under thinner markets. If some of the 28 liquid assets are pledged and cannot be released, the residual illiquid requirement rises before changing any assumed policyholder behavior. A proper model would also apply rate- or spread-driven mark-to-market changes to assets and liabilities, then reconcile the timing of loss recognition to the chosen accounting and solvency basis.

The illustrative exercise is consistent with the economics investigated in the 2025 life insurance convexity article: surrender options can create asset sales during rising-rate states. The paper reports a calibrated, context-specific bond-price effect; the illustration above intentionally makes no assertion about equilibrium bond prices. It uses simple accounting to isolate the cash waterfall and shows why an average surrender probability alone is an inadequate stress statistic.

Contagion, funding, and collateral

Liquidation has two possible feedback effects. Selling sufficiently large quantities of the same asset class can depress prices, reducing other insurers' mark-to-market capital and increasing haircuts faced by prospective borrowers. Public sales or supervisory intervention can also affect perceptions of other insurers with similar assets or products. At the system level, the key exposure is the coincidence of portfolio holdings, financing structures and liability exercise incentives, rather than the mere existence of one individual distressed insurer.

There are important limits to these mechanisms. Diversified long-term investors may purchase sales at prices they consider attractive. Insurers with long liabilities and ample cash can sometimes hold investments to maturity while honoring withdrawals; additional capital or committed credit facilities may absorb a shock. It is unsound to assume that every rise in surrender rates causes a fire sale or that every bond sale destabilizes a market. Indeed, the NAIC's August 2026 liquidity-stress findings reported modest aggregate modeled market impacts for 26 large US life insurers under the reviewed scenarios, with most cash deficits absorbed by available cash. These findings are evidence for the specified US supervisory sample, not proof of zero tail risk in all legal entities.

Derivative positions add a different liquidity clock. A hedge may economically offset an asset or liability valuation loss but require variation margin before the underlying cash flows are realized. Repo transactions, securities lending, financing agreements and asset-intensive reinsurance can generate additional collateral demands or restrict assets in stress. A fully effective hedge in present-value terms may therefore coexist with a short-horizon funding problem. The point appears in the IAIS risk-management standards and recent BIS work on systemic linkages. A stress test should join these channels within a consistent timeline rather than add independently calibrated worst cases that cannot occur together.

Institutional Backstops

Eurovita: a documented jurisdiction-specific intervention

Eurovita supplies a concrete case in which financial strain, rising-rate disintermediation and legal intervention overlapped. The Italian supervisory authority IVASS reported that Eurovita and its holding company fell below solvency capital requirements in September 2022 after deficiencies in governance and risk management, and that an attempted recovery plan was inadequate. In the official IVASS 2023 supervisory account, a rapid increase in surrenders from segregated-account contracts contributed to the decision to suspend the right to exercise surrenders on February 6, 2023. The contemporaneous IVASS suspension order identifies the legal instrument and affected contractual category.

The chronology matters. The authority appointed a provisional administrator on January 31, 2023; suspension followed on February 6, rather than being the initial recognition of trouble. The measure was later extended while the authorities and market participants designed a portfolio transfer. In June 2023, IVASS documented the transfer agreement involving major insurance companies and distributing banks and confirmed an extension through October 31. The October 2023 revocation and compulsory liquidation were followed by authorization to transfer the insurance business to Cronos Vita. The IVASS annual review describes the transfer as protective of policyholder interests. The episode should be understood through its formal decisions and financing arrangements rather than condensed into an unsourced story of instantaneous failure.

Eurovita does not provide a numerical estimate of the probability of a run for all European insurers. It establishes that surrender acceleration can become a material supervisory concern and that a jurisdiction may use temporary restrictions while pursuing a continuity solution. The case also exposes a consumer tradeoff: a restriction intended to protect collective recovery can impose immediate costs on holders with pressing liquidity needs. The merits of such intervention depend on statutory authority, proportionality, communication and the credible prospect of an orderly outcome. Historical actions in Italy do not themselves define the law applicable in another country.

The IMF's 2026 Italy FSAP technical note revisits Italian insurers' lapse sensitivity, notes strengthened liquidity monitoring and contingency planning, and explains the interaction between guaranteed contracts, rate changes and market value effects. Its findings concern the institutional and regulatory circumstances examined in the 2026 mission. They corroborate the need for scenario-specific governance while leaving the particular consequences of future rate shocks uncertain.

International standards, the EU transition, and US practice

International insurance supervision has long addressed liquidity through enterprise risk management. The IAIS Insurance Core Principles require an insurer's enterprise risk management framework to address liquidity in normal and stressed conditions. The guidance considers asset marketability, policyholder lapse and withdrawal, collateral demands, contingent financing and concentrations. These are internationally recognized supervisory principles, applied through jurisdictional regimes; they are not a direct substitute for enforceable local law.

The European Union adopted a revised Solvency II framework in Directive (EU) 2025/2. Among its liquidity and macroprudential provisions, Article 144b addresses supervisory powers in exceptional circumstances and provides for suspension of life policyholder redemption rights as a last resort subject to specified conditions, public explanation and time limits. The distinction between adoption, national transposition and application is decisive: EIOPA's July 2026 final liquidity guidelines state an application date of January 30, 2027. The amended framework and its detailed guidance must therefore be treated as forward-looking for a November 2026 assessment, rather than silently projected backward onto the Eurovita intervention.

In the United States, insurer regulation remains state-centered and differs from deposit-taking bank regulation. The NAIC's ORSA account describes an ongoing enterprise assessment of relevant risks, including liquidity, and their implications for solvency. The August 2026 NAIC macroprudential materials describe the completed liquidity stress-testing framework and its use in identifying potential securities sales during modeled stressed cash-flow deficits. Specific supervisory practices, filing populations, guaranty associations, security arrangements and emergency powers require state- and product-level verification before any legal or consumer conclusion is drawn.

The United Kingdom offers a separate resilience benchmark. The Prudential Regulation Authority's 2025 Life Insurance Stress Test results reported that participating major life insurers retained substantial aggregate capital coverage under a severe financial-market scenario. That exercise emphasized market and credit shocks; its findings are not a direct measurement of an insurer-wide policyholder surrender panic or an unconditional assurance of daily liquidity. The proper comparison is between the scenario definitions and tested exposures, not between headline capital ratios and an assumed number of redemptions.

Contract design, governance, and supervisory escalation

Long-horizon protection depends on explicit liquidity engineering at product origination. A contract with a guaranteed withdrawal value should be priced and backed by assets under a range of exercise times, not solely under central mortality and lapse assumptions. Surrender charges or market-value adjustments can reduce opportunistic interest-rate exercise when lawful and comprehensible, though they shift costs and risks toward the departing policyholder. Exclusion of high-cash-value products from an analysis can produce a misleadingly comfortable liquidity ratio even if most total policy count appears stable.

Governance should require the insurer to identify the legal entity responsible for payments, the country and currency in which cash must be delivered, the seniority and encumbrance of usable assets, and the availability of external funding under simultaneous sector stress. Management information must distinguish cash on hand, liquid market value, saleable value after stress haircuts, collateral already pledged, uncommitted credit lines, and assets held by another group company. Controls should also detect increases in surrender requests before payment completes; processing backlogs and an unusual proportion of partial withdrawals may signal an emerging pressure point.

A proportionate escalation sequence typically begins with more frequent cash-flow reporting, contract-level and distributor-level outflow analysis, enhanced liquidity reserves, reconsideration of new product pricing and investments, and credible contingency funding plans. Later steps depend on statutory authority and circumstances, including restructuring or supervisory intervention. Any power affecting lawful policyholder withdrawals should be assessed against the intended collective benefit, the scope of hardship consequences, and available continuity solutions. Supervisory discretion can preserve value only when the wider intervention is credible and operationally feasible.

Stress and Consumer Outcomes

International observations through November 2026

Contemporary indicators are informative precisely because they do not all tell the same story. The EIOPA June 2026 Financial Stability Report reported that the median life lapse rate in its relevant sample decreased from 3.8% in 2024 to 3.2% in 2025 as interest rates stabilized; its upper-tail statistic also decreased. This observation cautions against assuming a monotonically worsening surrender wave throughout every high-rate period. EIOPA's October 2026 risk dashboard rated liquidity and funding risks as medium at the report's cut-off, with broadly stable cash and liquid asset positions overall. An aggregate category cannot capture the tail exposure of one product issuer.

US materials report large aggregate payment flows alongside positive operating cash generation. The NAIC's 2026 sector assessment stated that 2025 life-industry surrender benefits were $487.6 billion, a 0.7% annual increase, while operating cash flow reached $219.2 billion. A separate 2026 Federal Insurance Office report reported $487.2 billion in surrender payments and $226.1 billion of operating cash flows for the US life-and-health sector under its own data sourcing and definitions. These reported figures should be preserved as separate institutional series rather than treated as exact duplicates or averaged into a synthetic observation. Neither amount describes a run by itself; the ratio to available short-horizon cash and the concentration of payment requests are required.

The NAIC's 2026 review of liquidity stress filings covering 26 large US life insurers found that most modeled cash-flow deficits could be met from cash. In the severe scenarios reported, modeled corporate investment-grade bond sales required to close deficits represented 3.1% of average daily trading volume and 0.04% of outstanding corporate investment-grade bonds; reported Treasury sales were also modest on the specified denominator. This finding challenges any claim of inevitable large-scale market disruption from sector-wide surrender stress. Its limits are the sampled insurers, selected stress assumptions, specified market benchmarks and absence of a full endogenous crisis feedback simulation in the published summary.

The cumulative international evidence thus supports a distributional rather than a binary assessment. Liquid, strongly capitalized insurers may absorb elevated surrender demands with modest trading. A thinly capitalized or heavily encumbered issuer with concentrated surrenderable liabilities may be much more vulnerable, especially when market prices and insurer reputation are changing rapidly. The market and supervisory environment in 2026 contains both genuine resilience indicators and material tail-risk channels. The evidence does not justify either a forecast of imminent universal life-insurer runs or a blanket denial of disintermediation risk.

A comprehensive validation protocol

A rigorous empirical or supervisory evaluation would assemble contract-level exposure tables reconciled to policy administration, statutory or regulatory returns and asset accounting. These should identify original and remaining term, guarantees, current cash value, account type, surrender-charge phase, product distribution channel, permitted partial withdrawals and policy loans. A cohort-level hazard or cash-flow approach then estimates ordinary behavior and plausible stress response, including correlated rate shocks, income shocks, perceived insurer distress, cyber or operational disruption and intermediary-driven accelerations. The uncertainty associated with small samples, structural changes and infrequent run-like episodes must be reported.

The asset side requires position-level duration, currency, credit quality, legal ownership, collateral pledges, market depth, settlement conventions and stress haircuts. Reinsurance recoverables, security trusts, derivatives margining and group transfers should be valued in the time frame when cash is required. The exercise should calculate at least daily/weekly, quarterly and longer-term liquidity outcomes because a firm solvent over a year can fail to settle obligations during a concentrated week. Reverse stress testing should identify combinations of lapse and asset-sale assumptions that exhaust practical funding capacity; it is a way to locate a boundary, not a forecast that the boundary will be reached.

Validation also requires backtesting surrender amounts against historical policyholder experience, out-of-time periods and rate changes. The Eling and Kiesenbauer lapse evidence supports treating policy and holder characteristics explicitly, whereas the Kubitza and coauthors' rate sensitivity analysis motivates sensitivity to economic incentives. Behavioral assumptions should be tested against events for which direct exposure data are actually available, and reported results should distinguish predictive skill in normal times from hypothetical crisis extrapolation.

The final audit must test the same policyholder group under every balance-sheet schedule: if surrender removes a policy, future premium inflows, investment crediting obligations and future benefits cannot remain unchanged. Sale proceeds and valuation haircuts must reconcile to position-level reference prices; derivatives gains and collateral outflows cannot be counted inconsistently; capital charges and cash positions must be tracked without equating statutory surplus to immediately transferable funds. The most consequential limitation in any global comparison is the scarcity of harmonized, insurer-specific daily surrender and funding data. This limits the empirical strength of claims about international run frequency, even while official cases demonstrate possible episodes.

Distributional consequences and consumer protection

Policyholders have different reasons to withdraw. A household facing unemployment, illness or retirement expenses may depend on cash value as a contractual contingency. Another may surrender for yield arbitrage after receiving a new offer from a distributor. Restricting access to funds affects these groups differently. A fair response requires clarity about cash values and charges, accurate and timely information, lawful prioritization rules, protection of beneficiaries and an effective complaints route. These considerations are integral to the economic meaning of a liquidity measure: protecting aggregate asset value while denying timely cash to vulnerable holders can impose a large welfare cost even if modeled solvency improves.

There is also an intertemporal distribution problem. Early surrenders may extract contractual cash amounts while remaining policyholders absorb the effects of asset sales, changed expenses or weaker future participation returns. Conversely, broadly imposed gates can cause households with legitimate liquidity needs to bear costs in order to preserve value for those who remain. The balance depends on contract terms, law, institutional alternatives, and the credibility of eventual payment. The Eurovita supervisory record illustrates why portfolio continuity, orderly transfer and policyholder protection must be analyzed together when restrictions are adopted.

Distribution channels deserve separate scrutiny because they can affect exit timing. A bank distributor or broker may explain a higher-yield alternative accurately, encourage replacement while omitting lost guarantees, or communicate solvency concerns. These possibilities must be established case by case and cannot be assumed from the existence of a sales network. Appropriate conduct analysis asks whether replacement advice fairly accounts for charges, tax, mortality coverage and the customer's own financial needs. A rise in completed surrenders may reflect informed exercise of options as well as potentially harmful switching practices; the distinction requires documented evidence.

The principal institutional objective is the continued and fair performance of enforceable insurance obligations under adverse conditions. That objective is served by sound product pricing, transparent withdrawal terms, realistic asset-liability matching, genuinely usable liquidity, robust monitoring and credible resolution planning. The latest BIS systemic-risk review and the IMF's Italian supervisory analysis both support monitoring complexity and liquidity interdependence, while the design of the consumer response belongs to the relevant law and the facts of the case.

Conclusion

Policy surrender runs and disintermediation risk arise from the interaction of contractual withdrawal rights, policyholder incentives, asset and liability duration, market liquidity, and expectations about financial performance. A higher outside yield can accelerate ordinary exits. A decline in confidence can concentrate payment requests. Once sales or funding constraints affect financial resources and perceptions, feedback may intensify the original withdrawal pressure. The causal pathways are empirically distinguishable, and their analytical separation improves actuarial measurement and crisis diagnosis.

The decisive balance-sheet variable is the ability to turn legally available assets and realistic inflows into cash on the actual dates when surrender payments and collateral calls fall due. Solvency and duration measures remain essential, but neither independently establishes that short-horizon obligations can be met without severe costs. Equally, a rise in lapse statistics establishes no direct amount of surrender cash and no evidence of strategic run behavior until the contract, cash-flow, and information channels are examined.

Historical disintermediation, the documented Eurovita intervention, and 2025–2026 international supervisory evidence support this conditional conclusion. They also reveal substantial heterogeneity: some institutions show resilience under published stress scenarios, while specific products, concentrated distributors, encumbered collateral and illiquid investment strategies can expose a firm to a sharper liquidity event. Institutional responses require legal precision, especially where redemption suspension is an exceptional measure subject to future regulatory application dates. Policyholder welfare and contractual fairness are constitutive elements of the risk assessment.

The original contribution of this dossier is an evidence-connected comparative synthesis of established economic, contractual, actuarial and supervisory mechanisms, supported by a transparent hypothetical liquidity waterfall and explicit source limitations. No new theoretical system, empirical priority claim, universal numerical forecast or privately collected dataset is asserted. Further progress requires harmonized contract-level and payment-timing data, credible identification of confidence interactions and scenario-linked market liquidity evidence. The scientific conclusion is that sound surrender-risk management must reconcile law, household choice, short-horizon cash, economic value and institutional confidence within one auditable assessment.

References