Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Universal Life Crediting and Flexible Premium Architecture

Abstract. Universal life insurance combines mortality protection with a dynamically maintained policy value to which interest credits and contractual deductions are applied. In flexible-premium designs, policyholders may alter contributions and sometimes the amount of insurance; the resulting contract is a system of interacting cash flows, insurance charges, guarantees, behavioral decisions, and legal constraints. This monograph examines universal life crediting and flexible-premium architecture through statutory and supervisory documents, actual insurance contracts, actuarial professional standards, experience studies, and scholarly literature. It reconstructs the account-value mechanism, distinguishes fixed-account crediting, interest-indexed crediting, and separate-account exposure, and examines current versus guaranteed scales. It separates account value, surrender value, death benefit, and independent no-lapse guarantees; identifies why planned premiums cannot in themselves establish long-term coverage sufficiency; and demonstrates the sensitivity of policy sustainability to premium behavior, crediting paths, mortality deductions, withdrawals, and loans. Comparative examination focuses on U.S. state insurance regulation, U.S. federal tax qualification, international insurance-contract accounting, European distribution protections, and selected evidence from China, without equating their legal or product categories. The original contribution is a source-grounded analytical synthesis and reproducible, explicitly illustrative cash-flow sensitivity analysis. The findings support contract-level, path-dependent evaluation of policy sustainability, transparent communication of guarantees and nonguarantees, and empirical models that distinguish premium persistency from termination behavior. This work proposes no new theory or doctrine and makes no claim of newly identified causal effects.

Keywords: Universal life insurance; flexible premium; credited interest; cost of insurance; policy value; cash surrender value; mortality deduction; indexed universal life; variable universal life; no-lapse guarantee; policy persistency; life insurance illustration; secondary guarantee; surrender option; insurance contract law; actuarial assumptions; consumer protection; insurance risk; asset-liability management.

Introduction

The defining financial feature of universal life insurance is a periodically updated contractual value. Premiums or net contributions increase the value; cost-of-insurance, expense, rider, and other deductions reduce it; credited interest or specified investment returns alter its trajectory. The death benefit, surrender benefit, and duration of coverage depend on contractual definitions and can be governed by calculations different from the account-value display. The Universal Life Insurance Model Regulation describes this separation with exceptional clarity: the policy value is an intermediate accounting quantity, and the amount immediately available to the owner may differ. The regulation is an NAIC model rather than itself an automatically effective rule in every U.S. state. Its analytical definitions are used here with this jurisdictional limitation explicit.

Flexible premiums increase the number of decisions facing both the policyholder and insurer. A policyholder can change payment amounts or timing when contractual constraints permit; the insurer remains exposed to mortality and guarantees while managing assets, expenses, crediting strategy, and policy behavior. The 1994 study by Robert Hoyt examined the effect of credited rates, loans, premium flows, and surrender behavior on universal life cash flows. Those mechanisms remain a useful starting point because an aggregate premium forecast cannot be inferred solely from issue-date planned premiums.

The research object is an individual life insurance contract or a closely specified class of contracts possessing a separately identified crediting-and-deduction structure and flexible payment authority. It includes traditional fixed-account universal life, indexed universal life, and variable universal life only where their common cash-flow features can be compared without overlooking differences in investment risk, regulation, and disclosure. A fixed premium life insurance policy, an annuity, a participating whole life contract, and a pension savings account are neighboring objects; their treatment is restricted to comparisons necessary to identify the universal life mechanism.

The dossier has three research questions. First, what exactly is promised and what is contingent in the premium, crediting, charges, and benefit provisions? Second, how do those provisions jointly determine values, policyholder options, lapse exposure, and insurer obligations across time? Third, how much can observational experience, product documents, and regulatory standards establish about sustainability, consumer understanding, and comparative governance?

The method proceeds from legal and product definitions to a reproducible accounting identity, then to scenario analysis and source-critical interpretation. Documentary sources have different evidentiary functions. Model regulations establish model terms; an insurer's filed policy form establishes the terms of that form; a professional actuarial standard addresses work by practitioners within its scope; an industry experience study describes its contributing population; an empirical paper supports claims at the level of its design. The 2015–2021 SOA–LIMRA experience study and its predecessor are therefore treated as evidence about reported U.S. participating experience, not as universal behavioral laws. Every deterministic example is labeled illustrative; no proprietary policy values or original insured-level data are claimed.

The analysis uses a research cutoff of November 16, 2026. The documentary record includes historical policy forms and regulations whose age is disclosed where material. Statements about U.S. federal income tax concern only the United States. Statements about the European Insurance Distribution Directive address the European legal category of insurance-based investment products where it applies. International accounting under IFRS 17 concerns entities within that reporting standard's scope, rather than redefining the consumer contract.

Definitions, Historical Development, and Documentary Foundations

The legal definition of universal life and premium flexibility

The NAIC Universal Life Insurance Model Regulation defines universal life by separately identified interest credits and mortality and expense charges. It defines flexible-premium universal life by the owner's ability to vary independently the amount or timing of one or more premium payments or the amount of insurance. The definition is important for two reasons. It locates the product's identity in the contract's crediting-and-charging structure, and it treats payment flexibility as a contractual power whose content must be specified. In practice, the extent of flexibility is constrained by minimum initial payments, taxation, changes in face amount, underwriting requirements, cash-value sufficiency, and guarantees.

There is an important difference between three quantities. Policy value, often called account or accumulation value in commercial material, is the computational balance to which separately specified transactions are applied. Cash surrender value is defined by the applicable contract and law and can include the effect of surrender charges and loans. Net cash surrender value is the amount available after the contract's relevant adjustments. A fourth quantity, the death benefit, is determined by the death-benefit option, the face amount, and applicable corridor or benefit provisions. These figures should never be silently treated as equal.

The model's treatment of flexible and fixed premium forms reflects the fact that an issue-date schedule of future account values cannot be an unconditional promise when payments are under the policyholder's continuing control. Its sections on valuation, nonforfeiture, mandatory policy provisions, annual reporting, and indexed crediting form an integrated regulatory architecture. Section 7 addresses minimum guaranteed interest credits and maximum mortality and expense charges, and Section 9 addresses periodic account reporting. These clauses specify the financial boundaries that an analyst must recover from each applicable contract and adopting jurisdiction.

Documentary evolution and product differentiation

Universal life became a distinctive product family in the era of interest-rate competition and financial product unbundling. Its operation can be described without reducing its commercial history to a single date of invention: periodically credited interest and itemized charges made the savings and protection components more visible, while flexible payments permitted owners to alter funding patterns. Historical product development then differentiated fixed-account crediting, externally indexed interest crediting, and separate-account investment exposure. The terms 'fixed', 'indexed', and 'variable' here identify distinct crediting or allocation mechanisms, not interchangeable guarantees.

Fixed-account universal life generally credits an interest rate specified by the insurer subject to contractual minimums, and future declared rates may vary under the terms of the contract. Interest-indexed universal life calculates at least part of the credited interest under formulas tied to an external benchmark, with contractual parameters such as participation rates, caps, floors, spreads, reset dates, and segment terms. Variable universal life allows designated values to depend on investment-account performance and has different securities and insurance supervisory treatment in the United States. The filed TIAA-CREF variable universal life contract provides documentary evidence of fixed-account and investment-account allocations within one contract, along with premium flexibility, monthly deductions, and a specified no-lapse mechanism. It is a historical contract specimen, not a template that governs every policy.

The NAIC's model distinguishes interest-indexed universal life from other forms by its external referent. Indexed crediting requires special attention because a floor on one segment's gross indexed interest is not a guarantee against erosion of overall policy value after insurance charges, fees, withdrawals, or loan costs. Likewise, participation in an index formula generally does not mean direct ownership of the index constituents. At the analytical level, every claim that an index-based account 'cannot lose' must be recast into the precise contractual promise: which balance, over which period, before or after which deductions, and subject to which crediting conditions?

Why an actual contract matters

A model regulation identifies intended legal categories; the policy form determines the owner's enforceable benefits subject to governing law. The TIAA-CREF filed policy specimen includes an adjustable death benefit, flexible premiums, premium allocations, withdrawal and loan mechanisms, monthly charges, a grace period, and no-lapse testing. Its descriptive passages explicitly warn that planned premiums may be paid while coverage remains vulnerable to lapse under the relevant value test. The no-lapse guarantee is separately conditioned on a cumulative premium test. This simultaneous existence of a planned premium, an account-value test, and a no-lapse test is direct documentary evidence that these concepts should be modeled separately.

A filed specimen also illustrates the need for version discipline. Riders may alter a death benefit or charges. State endorsements may replace language. Issue year, underwriting class, selected benefit options, and the guaranteed charge schedule affect values. Administrative illustrations present assumptions, not necessarily the exact legal hierarchy of policy provisions. A scientific comparison must record specimen date, available attachments, contract definitions, and the level of completeness; the specific 2011 policy specimen filed in 2012 cannot establish the current terms of another insurer's contract or the current law of every state.

The dual architecture of guarantees and discretion

Universal life guarantees have multiple objects. A minimum interest rate controls a crediting variable; a maximum cost-of-insurance schedule controls specified mortality charges; an expense guarantee controls specified charges; a no-lapse provision may maintain insurance when account value would otherwise fail its sufficiency test; and a death-benefit guarantee describes payment upon covered death subject to policy conditions. These guarantees can coexist, overlap, or apply on different conditions. Their consequences depend on their interaction over time.

Insurer discretion over nonguaranteed elements is governed by contract terms, law, and actuarial governance. The Actuarial Standards Board's ASOP No. 2, effective June 2022, includes credited interest, cost-of-insurance charges, expense charges, and indexed parameters among possible nonguaranteed elements. It concerns the actuarial determination of those elements within its professional scope. It supplies a disciplined distinction between guaranteed contractual boundaries and elements that can be redetermined; it does not itself create policyholder rights beyond the law and contract.

Premium Flexibility and the Contractual Cash-Flow Architecture

Premium authorization, planned premiums, and payment behavior

A planned premium is the intended recurring contribution used in an illustration, billing arrangement, or funding strategy. A minimum initial premium may be required to issue the policy; a minimum periodic amount may affect an independent no-lapse test; a sufficient premium for keeping the account-value test satisfied changes with crediting, mortality charges, and costs. These are economically different payment concepts. A contract allowing flexible payment usually does not promise that the initial planned premium will suffice indefinitely. A premium increase may be needed for the same death benefit after weaker crediting or higher charges, even when payments were previously made as planned.

Flexibility is best understood as an option to alter the premium path within the contract's administrative and legal restrictions. A policyholder may skip payments while existing value can fund deductions, pay more after weak performance, or use a single additional contribution. At the same time, the owner cannot assume that overfunding is unlimited. The IRS explanation of Sections 7702 and 7702A identifies the cash-value accumulation test, the alternative guideline premium and corridor framework, and the separate seven-pay test for modified endowment status in U.S. federal tax law. These provisions impose tax-classification constraints independent of the insurer's mechanical ability to accept funds.

Policy funding is path-dependent. Two contracts with identical cumulative nominal premiums can have different values if payments arrive on different dates or if early contributions are partially absorbed by premium loads, fixed charges, or changing mortality deductions. Their net amounts at risk can also diverge when one has a larger accumulation balance. Skipped premiums are not economically equivalent to an interruption in life insurance risk: unless coverage lapses, the insurer continues to bear insured mortality exposure while its deductions continue under the contract. That explains why premium flexibility is not a mere payment convenience.

A transparent accounting identity

For an illustrative fixed-account monthly policy, let A_t denote the policy value after period t, P_t the gross premium received, l_t the premium load fraction, F_t fixed administrative charges, R_t rider charges, C_t cost of insurance, W_t withdrawals, and i_t the period crediting rate. With a convention that net premium is added, deductions occur, and interest is then credited, an elementary accounting identity is:

A_t = [A_(t−1) + P_t(1 − l_t) − F_t − R_t − C_t − W_t] (1 + i_t).

This is a transparent reconstruction convention, not a universally applicable policy formula. Real contracts can deduct charges before or after credits, use daily interest, split allocations across segments, vary fixed charges, deduct premiums immediately, and apply account-specific loan crediting. The ordering of transactions is financially material. A full replication must follow the filed policy form and administration system rather than substitute a convenient generic formula.

The current period's mortality charge can be represented for analytical purposes as C_t = q_t × NAR_t, where q_t is the applicable contractual monthly cost-of-insurance rate and NAR_t the contractually defined net amount at risk. A simple level-death-benefit approximation is NAR_t = max(B_t − A_(t−1), 0). Particular policies can discount death benefits by monthly interest factors, use different valuation points, include riders, or employ option-specific computations. The TIAA-CREF specimen's net-amount-at-risk provision demonstrates that the actual calculation is more specific than the approximation. The quantitative appendix adopts its own simple definitions and labels them accordingly.

The identity immediately implies that changing P_t affects not only current value but also later interest credits and, under common formulas, later net amount at risk. Lower value can raise future net amount at risk, which can raise future COI deductions, contributing to further value erosion. This is a mechanical feedback under the stated approximation, not an empirical assertion about the size of actual future COI rate changes. Differences between guaranteed and current q_t schedules must be modeled explicitly.

Surrender value and death-benefit options

Suppose a contract shows $30,000 in policy value and imposes a $4,000 surrender charge. If there are no loans or other adjustments, the amount available upon surrender may be $26,000. The example demonstrates a definitional distinction rather than a statement about any specific product's actual charge. Surrender charges can vary by duration, benefit layer, premium, or coverage change; the applicable contract establishes their base and duration. A cash value display is therefore insufficient to infer liquidity.

Death-benefit options alter the relation between mortality protection and accumulated value. One common design offers a level death benefit subject to applicable minimum corridors, while another offers a death benefit that increases with account value; exact names and calculations vary. Under a level benefit and rising account value, net amount at risk can decline. Under a benefit that includes an increment linked to account value, some portion of the insured amount at risk can remain larger. The effect on COI and values depends on age, underwriting class, benefit option, and policy formula. Benefit increases may require evidence of insurability and new charge schedules; reductions may have tax, surrender, or guarantee effects.

Policy loans further separate the relevant balances. Loan principal, accrued loan interest, account collateral, loaned-balance crediting, and unpaid deductions can interact. A loan can reduce net value available on surrender, change the net death benefit, and threaten lapse depending on the contract. The actual filed loan and lapse provisions show why a consumer-facing account value without outstanding indebtedness is an incomplete measure of usable coverage resources.

Consumer flexibility and financial capacity

The contract's flexibility and the owner's economic capacity to exercise it must be distinguished. An owner facing income volatility may value the ability to reduce or postpone premiums. The same owner can find a later catch-up contribution difficult to finance exactly when charges have increased. Conversely, an owner who funds early and conservatively may create a larger margin against adverse crediting. This is an analytical statement about the interaction between contract permissions and a budget constraint; it cannot establish from product design alone which households will benefit.

Research on premium adjustment more broadly offers a caution. Paluszynski and Yu (2023) study price rigidity and commitment in life insurance rather than estimating a universal-life-specific premium choice rule. Their work reinforces the value of separating the power to adjust payments from observed adjustment frequencies and insurer pricing behavior. A scientific account must avoid generalizing a broad life insurance result into a universal life treatment effect.

Crediting, Mortality Deductions, and Guarantees

Declared fixed-account interest

For a traditional fixed-account universal life contract, a periodically declared credited interest rate may be greater than a guaranteed minimum. Insurers earn investment income on their asset portfolios, bear asset-liability and credit risks, and decide current nonguaranteed crediting within the framework of contract provisions, supervisory rules, and actuarial governance. The declared rate is not simply the market risk-free rate. Portfolio yield, reinvestment, expenses, credit spreads, capital, and competitive positioning can influence management decisions. A contract-level minimum is a floor on the designated crediting factor, not a statement that future gross premium deposits will earn that rate without deductions.

The ASOP No. 2 standard calls for relevant considerations in developing nonguaranteed-element scales, including anticipated experience and company practices. Its scope is professional actuarial practice, and any contract-level legal restriction remains controlling. For research purposes, a credible credited-rate forecast needs to distinguish current declared scale, guaranteed minimum, simulated future scales, and any carrier discretion or contractual formulas. Applying a single attractive currently illustrated rate through the entire insured lifetime produces a projection conditioned on an assumption rather than a guaranteed liability for the policyholder.

Interest crediting may occur monthly, daily, or according to segment rules, with compounding determined by policy definitions. The choice between annual effective and nominal monthly rates matters in long horizons. For example, a 4% nominal rate compounded monthly has an effective annual rate slightly greater than 4%; a 4% annual effective rate requires conversion to a lower equivalent monthly periodic rate. Source illustrations should identify which convention is used. When a change in credited rate is delayed relative to market rates, policyholder options may become more or less valuable, an issue connected to surrender behavior and insurer asset-liability management.

Indexed universal life

Indexed universal life adds a formula linked to an external index or other defined referent. Participation rates, maximum caps, spreads, floors, terms, and the reference measurement method determine credited amounts. The index credit might be zero for a segment with negative reference-index performance, while monthly mortality and expense deductions continue. The result can be a reduction in overall policy value even where a segment has a zero-interest floor. Policy loans may have distinct interest-crediting and borrowing terms that change net outcomes.

The NAIC's life insurance illustration guidance explains that AG 49 began in 2015, that AG 49-A applies to relevant policies sold on or after December 14, 2020, and that additional changes in 2023 and 2026 addressed illustration limitations and disclosures. A maximum permitted illustrated rate regulates presentation under its adopted legal framework; it is not a prediction of earned future returns. NAIC adoption and state implementation must be distinguished from universal effectiveness. Because product formulas evolve, the applicable guideline edition and policy issue date must be captured when evaluating an illustration.

Interest-indexed products also expose the difference between the economic return of the insurer's supporting assets and the formula credited to the owner. The insurer can manage index exposure through options or other strategies, subject to its particular asset and risk management practices. Crediting rates therefore depend on the formula, the price of optionality, interest-rate conditions, management choices where permitted, and contractual constraints. Research that uses a single historical equity index return as the policy's credited return bypasses these contractual mechanisms.

Variable universal life

Variable universal life introduces separate-account performance into the value path. Allocations, investment charges, fund-level expenses, fixed-account options, transfer conditions, and loan provisions can materially affect results. The SEC-filed variable universal life contract is a useful primary documentary example: its owner may allocate premiums between designated investment options and a fixed account, while monthly charges and coverage tests continue. The contract-specific no-lapse test has a different structure from the investment-account value path. Comparative analysis should therefore state whether the investment loss is borne directly by the owner and which guarantees, if any, constrain the result.

The economic distinction between fixed, indexed, and variable universal life is fundamental to scenario design. A decline in market asset prices does not automatically map one-to-one into a fixed-account policy credit. A stock-index decline does not automatically produce an identical negative indexed interest credit. A variable separate account may record negative investment performance, subject to its terms. All three contract forms can still lose policy value because deductions exceed net premiums and credited additions. These are separate causal channels and should be tested separately.

Cost-of-insurance deductions and the aging exposure

The mortality deduction generally depends on the applicable rate schedule and a measure of insurance at risk. At issue, rates may reflect age, risk class, sex or other legally and actuarially permitted characteristics, coverage amount, product design, and underwriting. In later durations, the applicable per-unit charges can rise with attained age even when mortality assumptions and insurer discretion otherwise remain unchanged. The specific guaranteed maximum COI rates can be substantially higher than current rates, but the width of that distinction is product-specific. Statements of an automatic universal-life rate increase of a fixed percentage every year lack contractual foundation.

Three variables should be separated: the amount of protection the insurer is currently exposed to, the rate charged per unit of exposure, and the owner's payment or cash balance used to finance that charge. When account value grows under a level death benefit, exposure can shrink even if q_t rises. When account value falls, the effect can reverse. Under option structures with an increasing benefit, the exposure path differs. A proper calculation needs the selected option and actual contract formula, not only the sum insured.

Guaranteed maxima provide an important adverse-scenario boundary, but a policy's performance under the guaranteed schedule is not necessarily the same as the insurer's best estimate or actual future scale. The NAIC model's policy guarantees address minimum interest credits and maximum mortality and expense charges. Those boundaries are necessary for a disciplined guaranteed illustration; the economic expectation of future nonguaranteed credits and charges is a separate measurement task.

Illustration governance and the current-versus-guaranteed problem

Life insurance illustrations can be useful because they expose the effect of assumptions across future policy years. Their interpretive danger lies in a visual equivalence between an assumption-driven projected column and a contractual promise. The Life Insurance Illustrations Model Regulation specifies the basic framework for illustrating guaranteed and nonguaranteed elements for policies within its scope. An illustration may involve disciplined current scales, prescribed labels, and restrictions intended to prevent misleading projections; its exact application remains jurisdiction- and product-dependent. A level nonguaranteed credited rate carried forward for decades should be understood as an illustration assumption.

The applicable actuarial professional counterpart is ASOP No. 24, revised in 2024, which addresses actuarial work under the NAIC life illustration model. This is distinguishable from the more general nonguaranteed-element determination work in ASOP No. 2. A review of an illustrated scale should ask who determines it, which experience factors enter it, whether the scale is within the relevant rules, and what consumer-readable warnings accompany guaranteed and nonguaranteed outcomes. Regulator model language, actuarial standards, policy text, and specific issued illustrations each answer different questions.

A robust interpretation compares at least a current illustrated basis, a contractually guaranteed basis, and plausible stressed paths; it also tests actual payment departures. One should not assume that a guaranteed-interest floor paired with guaranteed maximum deductions supplies a lifetime premium guarantee. Any independently promised lifetime or finite-term no-lapse coverage requires separate verification of the relevant premium test, shadow balance, or other guarantee condition.

Policy Sustainability, Behavior, and Adverse Scenarios

Sustainability is a multi-condition question

Policy sustainability means the conditions under which the insurance coverage and intended benefits can continue for a specified horizon. At least three tests matter. The first is account-value sufficiency under the contract's deduction and grace-period terms. The second is the satisfaction of a no-lapse or secondary guarantee, if provided. The third is the policyholder's financial ability and willingness to make future contributions, including any catch-up funding needed when prior assumptions prove optimistic. Tax classification and benefit changes impose additional constraints. A projection satisfying one of these tests can fail another.

Under the simplified balance identity, policy value may become insufficient to cover the next monthly charge. Actual contracts may then allow a grace period and notice procedures, a payment to restore sufficiency, or continued coverage under a secondary guarantee. The filed no-lapse guarantee language demonstrates that an in-force policy can be maintained under a cumulative premium test even when the account-value route alone would be inadequate. It also shows limits and adjustments triggered by loans and withdrawals. The relationship between the no-lapse test and account-value test is thus a contract-specific legal distinction, not an accounting identity.

An analysis of the owner's long-term benefit should evaluate actual net death proceeds, loan indebtedness, possible riders, premium outlays, surrender value, and the duration of coverage. A policy might remain in force with low surrender value because a secondary guarantee is satisfied. Another might show positive accumulation value but face changing financing needs for future coverage. Another might have a high illustrative surrender value under assumptions that have not been guaranteed. A single headline account balance cannot rank these contracts reliably.

The cumulative and asymmetric effects of adverse assumptions

A drop in credited interest has a direct effect on the balance in each affected period and an indirect effect through lower future compounding. Under a level death benefit and charge based on net amount at risk, lower policy value can also raise the amount exposed to COI deductions. The combination is more harmful than the lost interest alone under those particular assumptions. Conversely, stronger funding early in the contract can support later value and reduce insured net amount at risk. Such effects are mathematically evident from the stated formula and should be verified for the particular policy.

Premium reductions can be temporary responses to income uncertainty or long-term behavioral changes. A short payment interruption when account value is ample differs from a similar interruption near exhaustion. A catch-up contribution immediately after adverse performance can have a different effect from the same nominal contribution decades earlier. Loans can amplify erosion when borrowing interest exceeds the effective crediting on collateral or when indebtedness reduces the resources supporting the applicable no-lapse test. Policy changes can reset surrender schedules or change mortality charges, and withdrawals can affect death benefits or tax status. The effect of each intervention requires a documented contract rule.

The illustrative appendix uses a single fixed-benefit balance identity to show these relationships over 20 or 35 years. It is deliberately a deterministic sensitivity exercise, not a prediction for any actual policy. The scenarios demonstrate the possibility of a positive balance under some combinations and exhaustion under others. They cannot estimate probability of lapse, insurer solvency, future interest rates, or actual mortality costs without empirical calibration and policy-form replication.

No-lapse guarantees and shadow accounting

A secondary or no-lapse guarantee may be defined by a minimum cumulative premium test, a separate shadow-account balance, or another contract-specific rule. The guarantee can protect coverage for a prescribed period despite weak cash accumulation, provided stipulated conditions remain satisfied. It may include provisions for payment timing, withdrawals, loans, face-amount changes, policy debt, and grace periods. The SEC-filed specimen's guarantee test is concrete evidence of one such mechanism. General statements that universal life 'always' has a guaranteed minimum premium or that a shadow account is the owner-accessible cash account are unsupported.

For an insurer, a secondary guarantee creates an insurance obligation potentially surviving poor account performance. Lapse expectations can be economically important to the valuation of that obligation. If more owners persist when guarantees are valuable, the guarantee's effective cost can exceed a simplistic model that assumes behavior independent of contract value. The same behavior must be analyzed from the consumer side: maintaining a guarantee often requires specific premium discipline, and misunderstanding a small payment deficiency can cause severe coverage consequences. The relevant research question is the distribution of guarantee satisfaction, not simply the contractual presence of the word 'guarantee.'

Surrender, substitution, and liquidity pressure

Life insurance termination may reflect replacement by another contract, competing returns elsewhere, liquidity demand, household shocks, health changes, dissatisfaction, or the cost of maintaining coverage. The literature distinguishes competing explanations. Kuo, Tsai, and Chen (2003) report a macroeconomic cointegration analysis of lapse rates, unemployment, and interest rates; their results distinguish long- and short-run responses and the economic importance of the interest-rate channel. The study is not a randomized experiment, and its aggregate identification cannot determine the lapse hazard of a specific indexed or guaranteed universal life contract.

The Hoyt universal-life cash-flow study directly examined premium payments, loans, and surrenders, showing why an insurer's credited-rate strategy can interact with customer flows. Its historical data and model do not establish magnitudes for current product classes. Both studies nevertheless support the methodological proposition that a credible universal-life projection must model payment, loan, and surrender choices in relation to financial conditions, rather than assume static behavior without testing.

From the insurer perspective, a surrender is not only a statistical count. It changes future claim exposure, asset liquidity, expense recovery, and, where present, guarantee liabilities. Policyholders who are healthier or have less valuable guarantees may be more inclined to terminate in some circumstances, potentially changing the mortality composition of the remaining pool. The direction and scale of such selection depend on measured behavior and underwriting classes. A lapse assumption should be stress-tested around interest rates, policy duration, surrender charge expiration, guarantee value, and economic distress, with the relevant population clearly identified.

The limits of premium persistency as a surrogate for coverage

Premium persistency measures the relationship between actual contributions and a selected premium benchmark; policy persistency concerns whether the contract remains in force. These processes are correlated but not equivalent. A policy can remain active while an owner skips contributions, or terminate after an owner has paid amounts larger than planned. Similarly, an annual ratio of paid to planned premiums does not establish adequate lifetime funding, because it says nothing by itself about interest, COI, loans, expenses, or guarantee tests.

The SOA–LIMRA 2024 study summary describes premium-persistency measures including paid premium relative to planned premium at the beginning of the policy year, a capped paid-to-planned comparison, and current-to-prior paid premium. It also explains conversion from calendar to policy-year presentations and limitations of study population. These measurement choices matter: additional single premiums may inflate an uncapped ratio, while an owner making variable contributions may still be following a rational funding plan. Empirical work should report both behavior measures and the actual policy-state outcomes for which they are proxies.

Empirical Evidence, Comparative Institutions, and Scientific Limits

U.S. industry experience: strengths and denominators

The 2015–2021 SOA–LIMRA study reports 14 contributing companies for its premium-persistency analysis, with approximately half of flexible-premium universal life new-sales market share represented, and describes a separate lapse/surrender component involving 24 contributing companies and a larger estimated market share. The source describes roughly 11.9 million policy-count exposure and approximately 417,000 lapse/surrender terminations in the premium-persistency study summary. Exposure conventions, subset definitions, and the conversion to policy years prevent treating the quotient of these headline totals as a general annual individual lapse probability.

The institutional summary's total-premium figure differs from wording in an indexed PDF highlight: the web summary displays $76.5 billion collected, whereas a PDF preview text presents $76.5 million. Because the apparent unit conflict was not reconciled against a definitive primary data table, the aggregate premium amount is deliberately omitted from quantitative conclusions. This is an editorial source-conflict resolution: other supported observations are retained, while the contested unit is excluded. Any downstream analysis should use the study's final controlled data tables rather than a search-result extract.

The 2018 predecessor study addresses 2009–2013 flexible-premium fixed, indexed, and variable universal life experience. Changes between successive studies must be interpreted with attention to contributing insurers, product mix, issue cohorts, exposure definitions, and market conditions. A difference in a published aggregate is not by itself a causal change in behavior. The later 2022–2025 SOA data solicitation documents the scope of an intended update, including product types and lapse-related factors, but is not a report of completed 2022–2025 experience. As of this dossier's research cutoff, the solicitation is evidence about the research pipeline, not outcome data.

The experience-study distinction suggests useful future work. A reproducible statistical analysis should separately estimate paid-to-planned ratios, partial-payment events, account-value depletion, loan behavior, surrender, lapse without value, reinstatement, and secondary guarantee continuation. Product class and issue cohort should appear alongside duration and interest conditions. Where individual-level observations are available, competing-risk or multistate methods may be useful; this dossier performs no such estimation and makes no empirical hazard estimates.

Direct scholarship and its transfer boundaries

Hoyt (1994) is unusually close to the specific mechanism of this dossier because it studies universal-life premium, loan, and surrender cash flows. Its reported importance of credited-rate policy is consistent with the account-value mechanism, while the historical period limits direct parameter transfer. Kuo and colleagues (2003) provide aggregate statistical evidence about lapse and macroeconomic variables, but the estimated relationship is not automatically the behavior of a guaranteed universal life subset. Paluszynski and Yu (2023) contribute a broader model of commitment and premium rigidity in life insurance; their population and design should not be relabeled as flexible-premium universal life.

A newer market comparison is offered by Jiang, Zeng, and Rui (2025), whose published abstract reports analysis of credited-rate drivers among 38 Chinese life insurers using monthly data from 2015–2021. Its population, product terminology, regulator environment, and interest-rate practices differ from U.S. individual universal life. The paper is useful evidence that credited rates are also studied as management and financial outcomes; its accessible abstract alone does not justify importing coefficients or causal interpretations into this monograph. The full text was not independently available for detailed methods appraisal, which is a stated access limitation.

The strongest synthesis that these sources permit is qualitative and mechanism-specific. Interest crediting, premium decisions, loans, lapses, and insurer constraints are linked; the scale and direction of behavioral responses require empirical identification in the relevant contract population. This conclusion is supported by both direct universal-life work and adjacent life insurance scholarship, with distinctions preserved. It does not imply that every documented relationship has the same size or direction across markets.

U.S. contract regulation, tax status, and consumer disclosure

Three U.S. regulatory layers should be analytically distinguished. State insurance law governs policy filing and many contract, solvency, disclosure, and market conduct matters. The NAIC develops model laws and actuarial guidelines that can influence state rules but do not themselves automatically displace state enactments. Federal tax law establishes life insurance classification and consequences under federal tax provisions. Federal securities law adds another dimension for variable contracts and associated investments.

The NAIC Model Regulation No. 585 addresses universal-life-specific valuation, nonforfeiture, guarantees, and periodic policyowner reporting. The Life Insurance Illustrations Model Regulation addresses presentation of insured benefits and guaranteed versus nonguaranteed elements within its scope. The NAIC's 2026 illustration overview describes the evolution from AG 49 to AG 49-A and later modifications for indexed universal life. A complete compliance conclusion requires the actual adopting state's version, effective date, product form, and illustrations. A model provision is a documentary benchmark rather than proof of uniform state law.

The IRS Chief Counsel letter, dated March 24, 2025, is a concise official explanation of Section 7702 life-insurance qualification tests and Section 7702A modified endowment classification. The key distinction is that meeting a life-insurance definition for federal tax purposes and meeting the separate seven-pay test are different questions. Cash-value accumulation and guideline-premium/corridor tests are alternative Section 7702 routes, subject to their specific requirements. A contract can offer premium flexibility while federal tax limits constrain how that flexibility is exercised. This is a legal measurement issue, not a recommendation to make a particular premium payment.

The 2024 ASOP No. 24 gives actuaries a professional practice standard for illustration work within its scope. The 2022 ASOP No. 2 covers the setting of nonguaranteed elements under its scope and distinguishes that work from illustrations governed by ASOP No. 24. Consumer-facing disclosures should be reconciled against contract language, not treated as a substitute for it. An in-force illustration can reveal changed funding requirements but is still conditional on assumptions.

International financial reporting and European distribution law

International insurance accounting gives a useful cross-sectional perspective on the insurer's economic obligations. IFRS 17 Insurance Contracts, effective for annual periods beginning on or after January 1, 2023, uses current estimates of fulfillment cash flows, a risk adjustment for nonfinancial risk, and, when applicable, a contractual service margin. Its recognition, measurement, and presentation provisions address the insurer's reporting of insurance contracts. A policyholder's illustrated account value is not interchangeable with a group-level IFRS 17 insurance-contract liability or asset. The measurement population, contract boundaries, discount rates, and profit-recognition concepts differ.

The European Insurance Distribution Directive Article 29 addresses customer information, risks, costs, and charges for insurance-based investment products within its scope. Article 30 of the Directive addresses suitability and appropriateness, including knowledge, experience, financial situation, risk tolerance, and ability to bear losses in specified advice and sales contexts. Those rules illustrate the importance of assessing an owner's capacity to sustain a complex long-duration product. Their legal category is not automatically coextensive with every U.S. universal life form. Equating U.S. illustration compliance with European investment-product suitability would erase a substantive regulatory distinction.

Cross-border comparison requires a common analytic vocabulary and separate local legal mappings. An account-value mechanism can be compared economically even when products have different domestic names, tax treatments, guarantees, or investment risks. For each jurisdiction, the dossier's approach requires identifying the policy contract, crediting method, policyholder options, regulatory perimeter, income-tax treatment where material, accounting regime, and empirical market population. Only then is a claim about policyholder protection or sustainability comparable.

Evidence design and measurement discipline

A credible empirical study must define its unit of observation. Contract-month data would support cash-flow reconstruction more directly than annual aggregate premium totals. At minimum, a research dataset would distinguish issue age, underwriting class, duration, product chassis, face amount, death-benefit option, crediting scale, guaranteed scale, premium plan, actual cash contributions, withdrawals, indebtedness, monthly COI, expense charges, cash and account values, guarantee status, and termination reason. Any statutory privacy and contractual restrictions on data must be observed. The present dossier uses no proprietary individual-policy dataset.

The primary identification problem is endogenous behavior. Owners select funding amounts partly in response to past investment outcomes, household finances, health information, sales advice, and expectations. Insurers may adjust current crediting rates in response to asset yields, competition, profitability, and solvency considerations. Observed associations among paid premium, credited rate, and lapse can therefore arise through several channels. Estimating a causal effect of crediting on lapse needs a credible exogenous source of variation, a carefully justified structural assumption, or a documented experimental design. None is established merely by juxtaposing two industry summary statistics.

Measurement error can be severe. 'Lapse' may combine surrender with value and forfeiture without value; an administrative 'planned premium' may be outdated; terminations may differ by product rider; a paid-premium ratio may embed large single additions; and an annual snapshot can miss month-by-month guarantee-test failures. Study authors' definitions and exclusion rules are therefore part of the evidence, not footnotes to be ignored. The SOA–LIMRA study documentation is explicit about population and measurement choices, providing a better foundation for research than unsourced industry-wide anecdotes.

Integrated Scientific Synthesis

The object of evaluation is a conditional path

A universal life contract cannot be adequately represented by an issue-date planned premium and an assumed long-run credited rate alone. Its economically relevant object is a contingent trajectory of policy balances, deductions, guarantee tests, benefits, and owner actions under specified contract terms. This conclusion follows directly from the account-value identity, the variable contribution authority, and the separate conditions under which a policy remains in force. It is an integrated analytical inference from established contractual mechanisms, rather than a new conceptual system.

Under a level death benefit with an account-dependent net amount at risk, the crediting and mortality deduction processes interact. An optimistic projected crediting rate can understate future expected mortality deductions if it produces a larger simulated account value than a plausible stressed path. A lower contribution path can likewise increase net amount at risk in later periods. A conventional projection that varies only the investment return misses a relevant interaction. The effect is bounded by the actual policy's death-benefit and charge formulas, so no universal numeric multiplier can be asserted.

At least four records should travel together in scholarly evaluation: the contract and applicable endorsements; a transparent calculation of account value, available surrender value, and death benefit; explicit guaranteed and scenario-specific nonguaranteed scales; and the independent guarantee/lapse tests. If one of those records is missing, the project can identify a mechanism but cannot make a source-verified statement about that policy's future coverage duration.

Behavior is jointly relevant to welfare and liability

Premium flexibility can be useful for households with variable income while making future funding less predictable for insurers. The same feature alters the timing of mortality risk, savings accumulation, and expense recovery. Whether that produces a better individual outcome depends on circumstances the contract text alone cannot determine. Empirical research should estimate how actual funding behavior changes with income shocks, credited rates, loans, advice, and guarantee value, then evaluate coverage continuation and household welfare separately.

A guarantee can be particularly valuable when poor account performance threatens coverage. Its economic value to the owner and cost to the insurer depend on the likelihood that guarantee conditions remain satisfied. Policyholders who understand and finance those conditions may behave differently from owners who misunderstand them. Consequently, assuming behavior independent of guarantee value is a substantive model choice requiring sensitivity analysis. The SOA experience research program treats product structure and behavior factors as relevant objects of study, while its future data collection cannot predetermine findings.

Disclosure quality is a causal mechanism only when tested

Clearer presentation of guaranteed versus nonguaranteed values, the funding required for no-lapse conditions, and the difference between policy and surrender value is logically relevant to informed choice. It is a normative and mechanism-based inference, not an experimentally proven effect of a particular disclosure intervention. A practical evaluation might compare comprehension of a single optimistic illustration against a display containing three scenarios, explicit premium shortfalls, and guarantee-status flags. The test would require observed comprehension and later behavior, rather than merely counting regulatory disclosure fields.

The NAIC illustration framework and the European distribution requirements approach consumer understanding through different legal instruments. Their comparative value is the shared concern for realistic communication of complex financial protection, with jurisdiction-specific implementation. Neither authorizes a claim that all future interest credits are guaranteed or that the same consumer test governs all products.

Limits of scientific novelty and remaining questions

This dossier's contribution is the explicit integration of contract-state accounting, supervisory guarantees, crediting strategies, policyholder behavior evidence, tax constraints, and international financial reporting categories in one source-mapped study. The cash-flow identity and the scenario calculations employ familiar actuarial and financial accounting principles. The analytical distinctions are derived from already documented contract terms and existing professional research. No new theorem, proprietary economic model, empirical causal estimate, or priority claim is asserted.

Research with original individual-policy data could estimate the distribution of policy sustainability horizons, quantify owner responses to changes in declared versus market rates, compare guarantee satisfaction by product form, or measure the consequences of more intelligible funding illustrations. International comparisons could test whether changes in investment-based distribution standards affect comprehension or lapse behavior. These are researchable questions; their answers remain open until appropriate data, design, and legal mappings are supplied.

Conclusion

Universal life crediting and flexible-premium architecture is a long-duration contractual arrangement in which premiums, interest credits, mortality deductions, charges, benefit options, and owner behavior jointly determine a policy's evolving values. The mechanisms are documented in the NAIC universal-life model, actual filed insurance policy language, actuarial standards, and experience-study research. Account value, surrender value, death benefit, and no-lapse eligibility require separate analysis. Planned premiums and current illustrated credits are conditional inputs; guarantees require precise identification by object, duration, and prerequisites.

The financial analysis establishes a path-dependent sustainability problem. Premium timing changes later investment accumulation; policy-value variation can change net amount at risk and subsequent mortality deductions; borrowing or withdrawals can affect liquidity and separate guarantee tests. Deterministic stress testing makes these relationships transparent while leaving empirical probabilities and individual contractual outcomes to appropriately specified data and policy conditions.

The empirical literature establishes the relevance of credited-rate strategy, competing lapse motivations, actual premium-persistency behavior, and institutional heterogeneity. It does not justify one universal lapse elasticity, one future credited rate, or one long-term premium sufficiency judgment across fixed, indexed, and variable universal life. Regulators and professional standards address distinct facets of contract design, presentation, actuarial discretion, and insurer reporting. Responsible scholarship preserves those boundaries while comparing common mechanisms.

The resulting research position is clear: a sustainable universal life analysis requires an explicit contract, a defined funding and crediting path, a complete charge schedule, an independently evaluated guarantee test, and empirically justified behavior assumptions. The research dossier is an original critical synthesis of documented knowledge and reproducible illustrative calculations; its scientific claims remain tied to the supplied primary sources and their recorded limitations.

Appendix: Reproducible Illustrative Calculations

Assumptions and calculation convention

The examples use a hypothetical fixed-account universal life arrangement with an initial policy value of $12,000, level death benefit of $200,000, monthly gross premium P, a 5% load on each premium, a monthly administrative deduction of $15, and zero policy loans, withdrawals, and riders. The initial monthly COI rate per dollar of net amount at risk is q_0; the rate grows at a constant 3% per year (or 4% in the longer lapse illustration). The gross annual nominal credited interest rate r is compounded monthly at r/12. Net amount at risk before each monthly deduction is max(200,000 − preceding policy value, 0). Crediting follows all monthly additions and deductions. No surrender charge, tax limit, guaranteed scale, benefit corridor, insurer default, or independent no-lapse guarantee is modeled. This is an arithmetic demonstration, not a marketed policy or a prediction.

For month t beginning at t = 1, the computational rules are:

q_t = q_0 × (1 + g)^((t−1)/12);
C_t = q_t × max(200,000 − A_(t−1), 0);
A_t = [A_(t−1) + 0.95 P − 15 − C_t] × (1 + r/12).

The formula is applied successively; if A_t becomes negative, the illustrative account-value coverage test is deemed exhausted in that month. Results assume full specified premium receipt at each monthly transaction, no grace period, and no policy guarantee overriding exhaustion. A reproduction can calculate the rows directly with the stated recurrence and ordinary double-precision arithmetic.

Twenty-year crediting, premium, and COI scenarios

Scenario Monthly premium Nominal annual credited rate Initial monthly COI rate Yearly COI escalation Policy value after 240 months
Baseline $300 3.5% 0.00030 3.0% $97,304.70
Lower crediting $300 1.5% 0.00030 3.0% $74,119.14
Lower premium $180 3.5% 0.00030 3.0% $55,211.11
Higher initial COI $300 3.5% 0.00050 3.0% $82,015.73
Joint adverse case $180 1.5% 0.00050 3.0% $26,107.49
Higher crediting $300 5.0% 0.00030 3.0% $120,445.59

These values follow the recurrence specified above, using no intermediate rounding. They are account values under the stipulated hypothetical transaction convention, not death benefits, surrender values, insurer reserves, or guarantees. Their differences demonstrate conditional sensitivity rather than a probability distribution. The lower-premium illustration is not claimed to cause an imminent lapse because the assumed starting value and term still permit a positive balance after 20 years.

Illustrative exhaustion horizons

For a separate long-horizon demonstration, set initial account value $12,000, initial monthly COI rate 0.00050, annual COI escalation 4%, nominal annual credited rate 1.5%, the same $15 monthly administration charge and 5% premium load, the same $200,000 level benefit, and no guarantee override. With gross monthly premiums of $0, $50, $100, and $150, the first month of negative calculated account value occurs respectively at months 100, 146, 230, and 354. These are arithmetic thresholds under the supplied assumptions, not insurer-issued lapse determinations. Real contracts have additional grace and guarantee provisions; the sign of A_t in this simplified exercise cannot establish their legal result.

Replication and epistemic boundary

The accounting exercise can be reproduced with any spreadsheet or script that loops the formulas from t = 1 to the selected horizon. Scenario choices are deliberately synthetic; no mortality experience table is represented by q_0 or g, and there is no claim that these rates are typical for a specified insured age. More elaborate scientific work would use the contract's actual monthly deductions, applicable mortality rates, premium allocation timing, surrender schedule, separate-account crediting, guarantee ledger, and legally relevant age and tax tests. The present calculations test mechanisms and conditional sensitivity, not real-world calibration.

References