RESEARCH DOSSIER
Annuity Contract Payout Value and Product Loadings
Abstract. An annuity's money's-worth ratio compares the expected discounted value of promised payments with the premium paid. This apparently simple measure is indispensable to studying the economics of annuity contracts, yet its interpretation depends on contract contingencies, prospective mortality, the discount curve, tax treatment, market date, and the position from which value is assessed. This monograph establishes a reproducible framework for comparing immediate and deferred life annuities, contractual guarantees, explicit distribution charges, implied price residuals, and the insurer's prospective costs. It critically synthesizes foundational United States and United Kingdom studies, cross-country research, and updated evidence from United States annuity quotes in May 2024. It separates three questions routinely conflated in discussion of annuity prices: what promised payments are worth to a reference survivor population, what they are worth to annuity purchasers, and what resources an insurer requires to finance them. An independently calculated, deliberately synthetic example illustrates how mortality and discount-rate assumptions alter valuation while contractual payouts remain fixed. Historical empirical results show substantial variation across product types, time periods, and valuation bases; the October 2025 revision of research by Poterba and Solomon reports population-basis values of approximately $0.87 and $0.73 per premium dollar for two specific immediate and deferred annuity designs, compared with approximately one dollar under annuitant mortality. Those observations cannot be generalized into a universal industry load. The resulting synthesis relates contractual value, risk pooling, asset-liability management, household welfare, expense attribution, accounting, and disclosure, and defines the evidence necessary to distinguish legitimate financing costs from poor consumer value in a specific market.
Keywords: Life annuities; money's-worth ratio; expected present discounted value; annuitant mortality; mortality selection; expense loading; acquisition commissions; deferred income annuities; discount curves; longevity risk; insurer valuation; retirement decumulation; contract guarantees; insurance economics; reproducibility.
Introduction
The economic value of an annuity is inseparable from its contingent promise. In a conventional immediate life annuity, a policyholder exchanges an initial premium for a specified stream of payments while the covered individual remains alive. The insurer manages a portfolio of such promises, converts premium assets into long-duration cash flows, and bears contractual exposure to the possibility that policyholders live longer than anticipated. A life annuity is therefore simultaneously a product purchased by a household, a contingent claim valued by an actuary, and a long-duration obligation financed by an insurer. Pricing analysis begins by recognizing these related accounting and economic objects.
Yaari's uncertain-lifetime consumption analysis (1965) established a foundational theoretical reason that survival-contingent income can improve consumption opportunities under specified preferences and market conditions. The subsequent welfare analysis of Davidoff, Brown, and Diamond (2005) generalized the annuitization result under certain conditions and examined departures from complete markets. These works identify the value of longevity pooling, which a present-value ratio cannot by itself measure. Conversely, a positive welfare gain from annuitization does not demonstrate that every offered annuity is competitively priced or suitable for every household.
The historical cost literature, particularly Friedman and Warshawsky (1990), examined how market pricing and adverse selection might help explain limited purchases of retail annuities. The influential money's-worth study by Mitchell and colleagues (1999) made the present value of promised payments relative to the initial cost a central empirical statistic. Their object was a single-premium immediate annuity, and comparisons depended on whether the evaluator resembled the general population or the longer-lived population of annuity purchasers.
This Dossier extends the comparison across pricing dates, deferral, period-certain promises, and institutional settings while preserving the original distinction among price, value, risk, and welfare. It addresses three linked research questions. First, how does contract cash-flow valuation change with the mortality basis and the discount curve? Second, what can observed money's-worth ratios establish about commissions, operating expenses, risk capital, and insurer returns? Third, under what conditions can household and insurer values be compared across countries and product designs without confusing distinct contractual rights?
The method is a critical synthesis of original research, institutional primary materials, and a fully specified arithmetic illustration. Published empirical estimates are attributed to their researchers and sample dates. All numerical examples introduced by this Dossier are identified as synthetic, so they cannot be mistaken for observed annuity quotations. Internal calculations do not constitute a new actuarial model or a new authorial theory. The scientific contribution is a disciplined interpretation of established valuation relationships and the construction of an auditable evidence comparison.
Historical and Institutional Foundations
The economic object and the timing of the promise
A lifetime annuity substitutes contingent income for an asset balance under a contract specifying who is covered, which event triggers payment, when the first installment falls due, whether payments continue after death, and which party bears investment risk. Immediate nominal annuities, deferred income annuities, annuities with minimum guaranteed periods, inflation-adjusted annuities, joint-life policies, and variable contracts produce materially different cash-flow distributions. Their quotations cannot be compared by annual payment alone. Commencement, payment frequency, guarantee provisions, and the precise survival condition determine the schedule to be valued.
In a life-only contract with no refund and no guaranteed minimum, payment ceases on death under the contract's timing rule. In a ten-year certain-and-life arrangement, the designated recipient or estate may receive the balance of the guaranteed period after an early death, while survival beyond that period continues the life income. A deferred income annuity requires survival to the commencement date, absent any separately specified death or refund benefit. These differences change both the expected present value and the amount of financial protection provided. They also change the relevance of annuitant mortality selection to the observed money's-worth ratio.
One particularly useful recent source is Poterba and Solomon's revised analysis (2025 revision of NBER Working Paper 28557). It evaluates United States quotes as of May 2, 2024, for immediate annuities, deferred annuities, and products with guaranteed payment periods. The researchers describe guarantees as reducing the degree to which payout value depends on the buyer's date of death. Their working paper is circulated for comment and has not been peer reviewed through the NBER working-paper process; its findings are identified here according to that evidence status.
Money's worth in the historical literature
Friedman and Warshawsky (1990) estimated relatively low yields on the individual life annuities available in the United States in their historical sample covering 1968–1983. The published cost-of-annuities study was originally circulated as a 1985 NBER working paper. Its comparisons concern yields against alternative long-term investments, and must not be mechanically converted to a single percentage commission or interpreted as a current retail price quote. Market interest rates, survivor composition, product specifications, and opportunity costs have changed considerably since that period.
Using United States price evidence from the mid-1990s, Mitchell and colleagues' 1999 article demonstrated that valuation with annuitant mortality yields a different numerator from valuation with population mortality for the identical contract. This difference is an economic consequence of selection into purchase and of the chosen demographic reference group. The resulting residual between premium and discounted expected benefits is an aggregate pricing statistic that includes more than selling expenses.
Comparative inquiry was advanced by James and Song (2001), whose study examined Canada, the United Kingdom, Switzerland, Australia, Israel, Chile, and Singapore alongside United States evidence. Their reported annuitant-basis money's-worth ratios exceeded 95% in most studied countries when future payments were discounted at a risk-free rate, while general-population ratios exceeded 90% in most cases. They also emphasized the importance of national interest-rate structures, mortality assumptions, and insurers' investment intermediation. Those estimates depend on the particular historical markets and valuation dates; they are not a timeless global benchmark.
In the United Kingdom, Cannon and Tonks's historical study assembled annuity price information over 1972–2002 and found values of roughly 90%–100% under differing longevity assumptions. Their later United Kingdom annuity price series documents the historical reconstruction for 1957–2002. The authors' research makes clear that time series of nominal annuity rates require contemporaneous rate and mortality context before being interpreted as trends in insurer profitability or customer value. In Annuity Markets, chapter 6, Cannon and Tonks (2008) reviewed money's-worth calculations and observed that estimated markups were generally modest relative to some other financial-services costs. The book chapter supports historical interpretation rather than a contemporary prediction.
Selection and the insured population
The annuitant pool is not a random draw from the population. Purchase can be related to health, private longevity information, wealth, pension arrangements, and product characteristics. Finkelstein and Poterba (2002) compared voluntary and compulsory individual annuity markets in the United Kingdom. They found higher longevity among annuitants than non-annuitants and reported stronger selection effects in the voluntary market, estimating the compulsory market effect at about half the voluntary-market level under their specification. The empirical result is specific to the institutions, cohorts, and methods examined.
Their subsequent policyholder mortality study (2004) detected systematic relationships between mortality outcomes and particular annuity contract characteristics, including payment timing and potential payments to an estate. It did not find substantial differences in mortality by annuity size. This is consequential for inference: evidence of selection on one contractual dimension can coexist with weak or absent selection on another. Treating all annuity purchases as one homogeneous risk class conceals the selection mechanisms relevant to price and value.
The Society of Actuaries provides actual mortality-table reference material. The 2012 Individual Annuity Mortality Basic Table, alongside the Individual Annuity Reserving Report, demonstrates how a specified annuitant mortality basis and separately specified valuation margins can enter actuarial work. An official reserving table has a defined regulatory purpose and cannot automatically be substituted for an unadjusted best-estimate longevity forecast of a retail purchaser in another year or jurisdiction. Projection to a relevant cohort and calendar year requires additional assumptions and documentation.
Money's-Worth Ratios and Reproducible Valuation
The core expected-value calculation
Let P denote the single premium paid at contract issue, t index possible future payment dates, B_t the contractually due cash payment at t conditional on the relevant contingencies, S_t the indicator that the covered life survives as required, and D_t the discount factor corresponding to the valuation objective. In the elementary life-only case with fixed payments b_t and no guarantees, expected present discounted value is
Here is the probability that a purchaser aged x survives to the relevant payment date; T is the maximum modeled payout horizon. Where payments occur monthly, payment timing, mortality interpolation and the term structure must be adjusted accordingly. D_t may be defined from a matched zero-coupon yield curve; using a single flat rate is a simplification. For multiple lives, guarantees, cash refunds, escalation, variable payouts, or contingent death benefits, the formula must instead sum the expected present value of the applicable state-dependent contractual payments. A guaranteed payment may have a probability of payment equal to one even when the annuitant dies, provided the beneficiary receives it and contract conditions are met.
For a constant annual life-only payment A at the end of each year, a flat effective annual discount rate r, and a discrete annual survival probability, write
This annuity factor shows why a higher expected survival probability raises the present value to the buyer, whereas a higher discount rate reduces it, with the quoted contractual payment held constant. In a fully fair, zero-expense hypothetical contract, P=Aa_x(r) on the same mortality, timing, and valuation basis. In an actual insured product, quoted benefits reflect expenses, acquisition, risk transfer, investments, capital, profit objectives, market competition, and regulations. The distance between P and Aa_x(r) therefore does not reveal any single one of these components.
Interpreting an implied loading
An often-used descriptive residual is L=1−MWR, measured as a share of the premium. If the MWR is 0.90, L=0.10 states that modeled expected discounted payments are 90 cents per premium dollar and the residual is 10 cents under the selected assumptions. It does not establish that ten cents were paid as a commission, retained as profit, or lost by an actual purchaser. Change the mortality basis or discount curve and L changes without an amendment to the policy or to its explicit fees.
In a homogeneous illustrative pricing identity, the gross premium may be represented analytically as the expected present cost of benefits plus present values of acquisition costs, maintenance, capital financing, taxes and a prospective margin, all measured on a coherent insurer pricing basis. The components are forecast costs and required compensation, not identical to an ex post income statement. On the insurer side, a risk-free buyer valuation may be higher than a discounted liability calculation using different, regulation-appropriate inputs, and the latter can incorporate uncertainty adjustment and expense obligations. Two independently computed values must be reconciled by basis before their difference is interpreted.
An independently replicated synthetic illustration
The following transparent calculation is authored solely to test the valuation mechanics. It represents no actual product, company quotation, population life table, or regulatory calibration. Assume an age-65 purchaser pays P=$100,000 for A=$6,500 at each calendar-year end while alive, with no refund, death benefit, escalation, or minimum payment period. For attained age 65+k, define the illustrative annual death probability at the kth year of exposure as for k=0,...,49, and stop payments at age 115. The k-year survival probability is the product of over j=0,...,k−1. The assumed discount rate is a flat effective annual 3%. The assumptions are synthetic and chosen for mathematical transparency, not as fitted mortality observations.
This generates an annuity factor of 14.38440, an expected discounted payment value of about $93,499, and MWR 0.9350. To isolate assumptions, repeat the calculation after scaling each annual death probability by 0.8 or 1.2, leaving the premium and payout unchanged; and repeat it with 2% and 4% annual discount rates. The numerical results are independently calculated from the stated recurrence, with survival truncated at 115 rather than extrapolated indefinitely.
| Effective discount rate | Mortality factor versus illustrative base | Annuity factor | MWR |
|---|---|---|---|
| 2% | 0.8 | 17.40322 | 1.1312 |
| 2% | 1.0 | 16.09105 | 1.0459 |
| 2% | 1.2 | 15.02255 | 0.9765 |
| 3% | 0.8 | 15.43682 | 1.0034 |
| 3% | 1.0 | 14.38440 | 0.9350 |
| 3% | 1.2 | 13.51417 | 0.8784 |
| 4% | 0.8 | 13.80564 | 0.8974 |
| 4% | 1.0 | 12.95277 | 0.8419 |
| 4% | 1.2 | 12.23735 | 0.7954 |
The table is a local comparative-static exercise. Its baseline residual of 6.5% changes to −0.3% at the lower illustrative mortality scaling and to 12.2% at the higher scaling, even though the insurer's quoted $6,500 annual payment and $100,000 premium have not moved. It is consequently indefensible to label the residual an observed sales commission. Holding mortality fixed at the baseline instead and raising the flat discount rate from 2% to 4% changes MWR from about 1.0459 to 0.8419. The example shows why precise disclosure of demographic and financial assumptions is part of the empirical claim, not an optional footnote.
The annual end-of-year timing also matters. A product paying monthly from the purchase date will have a different expected present value, particularly where installments are paid in advance. For a deferred product, survival probabilities must be measured from purchase and the summation starts at the contractual commencement date. For a guaranteed-period product, payments due during the minimum term must reflect the contractual rights of beneficiaries. A comparison that overlooks any of these terms substitutes a fictional instrument for the actual one.
Discount rates and the interpretation of safety
A market-consistent comparison requires attention to promised currency, nominal or inflation-indexed benefits, term structure, and credit risk. The value of a nominal guaranteed payout discounted at government security rates is not equivalent to its value at corporate bond yields or at an insurer's expected asset returns. The former can serve as a benchmark for a comparatively low-risk nominal income stream; the latter can capture credit and investment exposures that change the economic meaning of the ratio. An insurer's ability to earn an investment spread is not evidence that the customer can obtain the same risk-adjusted outcome while preserving lifetime pooling.
The 2025 Poterba–Solomon revision explicitly reports results under both Treasury and BBB corporate bond discount curves and shows lower payout values under the latter. The contrast illustrates why a reported MWR must carry its discount basis. For assets and liabilities subject to credit, capital, liquidity, and reinvestment constraints, comparisons must also distinguish realized yields from required or risk-adjusted returns. A lower corporate-discounted ratio does not mechanically measure inefficiency; it answers a different valuation question.
Commissions, Expense Loads, and Pricing Residuals
Mapping the costs behind a quoted payout
A quoted annuity payment can reflect at least six channels. Acquisition costs include distribution commissions, application processing, underwriting or eligibility verification, and new-business administration. Maintenance costs include payment operations, customer communication, records, tracing and verification of life status, complaints, and ongoing control systems. Asset-management costs finance the investment portfolio and hedging operations. Longevity and other insurance risks require pooling and reserves, with possible reinsurance. Financing and capital costs compensate providers of risk-bearing capacity. Taxes, applicable assessments, and profit expectations complete the commercial pricing exercise.
These are analytically separable costs, but they need not appear as distinct line items on a consumer's contract. A fixed-immediate-annuity premium may contain implicit loading through the price-to-payment conversion rather than a separately debited fee. Variable annuities, by contrast, commonly specify ongoing account-value charges and explicit riders that operate differently from the income conversion terms of a simple life annuity. Treating the advertised fee of a variable annuity and the money's-worth residual of a fixed immediate annuity as equivalent percentages would combine different denominators, timing, and benefits.
The United States Securities and Exchange Commission describes variable-annuity charges and rider costs, including mortality-and-expense risk charges, administrative charges, underlying investment expenses, and optional benefit charges. Its examples explain charges assessed as percentages of account value, not as a uniform discount to every future guaranteed life payout. The United States National Association of Insurance Commissioners' fixed-deferred-annuity buyer's guide describes charges that reduce contract value, including surrender or withdrawal charges subject to product-specific provisions. The two consumer documents cover different contract classes; neither furnishes a universal expense schedule for immediate lifetime-income annuities.
Distinguishing explicit, implicit, and inferred charges
An explicit charge is specified by the contract or accompanying disclosure and can ordinarily be scheduled or simulated from that instrument. An implicit loading is embedded in the payout conversion terms or interest crediting method and is inferred by comparison to an independently defined benchmark. An ex post expense allocation is obtained from the insurer's accounts after costs have actually been incurred. An estimated money's-worth residual is a valuation difference created by model assumptions. Each is meaningful, and none can be substituted for the others without evidence of equivalence.
For a single fixed payout product, a charge cannot be attributed to distribution simply because an insurance intermediary was involved. A product with substantial fixed administration costs may have a larger cost per dollar of premium for small purchases. Conversely, scale economies, distribution technology, annuitant selection, and market concentration can alter quoted rates without mechanically identifying a commission. An institution that reports acquisition expenses in aggregate may sell several product categories, creating a further allocation problem. A scientifically defensible commission study therefore requires contract or distributor schedules, segment-level accounting, and treatment of rebates, trail compensation, and expense deferrals where they exist.
Insurer profitability must also be evaluated ex post. The premium may finance risky investments and long-duration liabilities while realized mortality, expense inflation, asset returns, surrenders, capital charges, and taxes depart from assumptions. A favorable retrospective outcome can reflect experience variance rather than an excessive ex ante price. An unfavorable outcome can arise even if the original money's-worth ratio was below one. The historical distinction in James and Song's intermediation analysis remains instructive: the price paid to the annuitant is connected to the investment and risk-transformation business of the issuer, so a ratio evaluated at a government security rate does not equal an insurer's net margin.
Decomposing a residual responsibly
A useful empirical reconciliation begins with a precise observation unit: contract i offered on date d for age x, benefit design z, premium P_i, and payout schedule B_i. Estimate the present value under a stated household basis V^H_i and an insurer basis V^I_i. Then collect independently evidenced acquisition cost C^A_i, maintenance cost C^M_i, investment and hedge cost C^I_i, required cost of capital K_i, and pricing allowance or margin M_i. The relation between price and these terms should be tested as an accounting and actuarial reconciliation under compatible assumptions, not imposed as an identity by assigning the unexplained balance to the last variable.
Suppose an analyst reports V^H/P=0.90. The ten-percentage-point residual could combine differences in discount rates, mortality selection, contractual costs, and compensation for bearing risk. If a broker contract independently establishes an upfront commission equal to two percent of premium, this establishes one observed cost item of two percent, but it does not establish that the remaining eight percentage points are corporate profit. In particular, differences between V^H and V^I must be addressed before attempting to allocate the remainder. In this sense, source provenance is as important as the arithmetic of the ratio.
Regulatory discipline and the boundaries of interpretation
Product disclosures, sales conduct, and accounting have different objectives from a single money's-worth measure. The NAIC's Annuity Disclosure Model Regulation listing identifies a model disclosure framework; adoption and effective requirements must be checked state by state rather than presented as one federal legal rule. The NAIC's annuity suitability and best-interest framework concerns recommendations and consumers' objectives. Suitability or best-interest compliance cannot be inferred from a ratio alone because guarantees, liquidity needs, risk tolerance, compensation arrangements, and alternatives remain relevant.
At the international supervisory level, the IAIS Insurance Core Principles and ComFrame establish a framework for policyholder protection and supervision. Their conduct-of-business dimension supports accurate, clear communication about benefits, risks, and costs; implementation varies across jurisdictions. In financial reporting, IFRS 17 Insurance Contracts measures groups of applicable insurance contracts through current fulfillment cash flows and, where relevant, a contractual service margin. The IFRS 17 key-terms explanation includes expected benefits and expenses within fulfillment cash flows and describes the risk adjustment. This accounting structure is not a retail annuity price list, and its measures cannot be called the money's worth of an individual contract without additional reconciliation.
Customer and Insurer Value Across Products and Markets
Three distinct questions of value
The consumer present-value question asks what promised cash flows are worth under the purchaser's survival probabilities and matched opportunity-cost assumptions. The utility or welfare question asks how the contingent income affects consumption security, precautionary holdings, estate objectives, and exposure to catastrophic longevity. The insurer economics question asks whether the asset portfolio, reserves, capital, and expected future cash flows support a sustainable return at the transaction price. Each question concerns the same contract from a different decision position, and each produces useful information without collapsing into the others.
The classical welfare results of Davidoff, Brown, and Diamond demonstrate why lifetime income can have value beyond the expected present value of scheduled dollars when lifetime consumption risk is material. A risk-averse purchaser can rationally value insurance despite an MWR below one; another purchaser with strong bequest or emergency-liquidity needs may prefer a different income arrangement. Household welfare therefore calls for an explicit utility or consumption-risk model in addition to a price study.
Joint-life income introduces another dimension. Brown and Poterba's joint-life study shows that couples' potential gains from annuitization must account for joint-and-survivor payment structures and second-to-die longevity. The payment to the surviving spouse may be a percentage of the original benefit, which changes both household welfare and insurer costs. Inferences based on the survival of a single reference life are insufficient for those products.
Immediate and deferred annuities: contemporary United States evidence
The most directly applicable recent empirical comparison is the October 2025 revision of Poterba and Solomon's money's-worth analysis. Its data consist of United States retail market quotes in May 2024 for immediate, guaranteed-income, and deferred annuity structures. Under the researchers' central population mortality assumptions, the expected discounted value of a simple immediate life annuity purchased at age 65 was approximately $0.87 for each $1 premium. For a deferred annuity purchased at 65 with payments beginning at 75, it was approximately $0.73. Under their annuitant survival assumptions, both approached one dollar per premium dollar. These figures are deliberately not repurposed as current 2026 price observations.
Several scientific implications follow. First, a population buyer and an annuitant with different expected survival can evaluate the same offered contract differently; part of the measured gap is attributable to the reference mortality distribution. Second, long deferral magnifies the sensitivity of a life-contingent contract to whether the purchaser reaches commencement, while a guaranteed period makes more of the payout independent of the date of death. Third, changing the discount curve modifies estimated value even while contractual benefits and purchase prices stay fixed. Fourth, the gap between population and annuitant valuation cannot by itself distinguish unobserved commission, risk pooling cost, selection, and the issuer's balance sheet requirements.
The revised working paper additionally investigates the duration since purchase and reports stronger apparent private-health-information effects soon after purchase than at longer durations, with comparatively small effects on their money's-worth estimates. That result qualifies an otherwise tempting claim that deferred annuities simply eliminate adverse selection. For one thing, mortality composition, purchase timing, and actual contract guarantees remain endogenous; for another, the study is limited to its observational setting and model assumptions.
The United Kingdom and the role of pension institutions
The United Kingdom provides evidence on the relationship between purchasing arrangements, adverse selection, and quoted retirement-income conversion. The comparison of compulsory and voluntary annuity purchasers in Finkelstein and Poterba (2002) shows why institutionally induced participation can alter the pool of insured lives. The later policyholder analysis (2004) adds information on choice across contract characteristics. Together, these results imply that market structure matters for both annuitant mortality and the kinds of products offered. A quotation from a historically mandatory pension-annuity regime cannot be treated as a counterfactual voluntary quote without adjustment.
The price reconstruction by Cannon and Tonks (2004) likewise requires alignment of observed annuity quotes with interest rates, pension account values, and mortality assumptions. A declining annual payout per premium may accompany falling yields or improved longevity rather than an increasing net loading. A cross-year measure of money's worth is accordingly more informative than a series of raw quoted income rates, but only when its own modeling assumptions are kept comparable.
Comparing national markets
Cross-country differences arise from public pension protection, pension conversion regulations, reserve and solvency rules, distribution, mortality, taxes, nominal versus indexed offerings, currency yield curves, and the structure of household demand. The international sample in James and Song (2001) illustrates how an apparently similar contract can be financed and valued under different asset markets. Their estimated high annuitant-basis money's-worth ratios were compatible with a hypothesized insurer investment spread; the claim is an economic explanation supported by their historical evidence, not a direct observation of every insurer's profit account.
The European Insurance and Occupational Pensions Authority's 2025 costs and performance report studies charges and performance for sampled unit-linked, profit-participating, hybrid, and personal pension products for 2020–2023. These measurements are pertinent to long-term savings-product value, but do not automatically provide payout money's-worth estimates for simple single-premium immediate life annuities. The product taxonomy, observed horizon, investment components, guarantees, and fee denominator must be matched before comparing a reported percentage with an annuity loading.
The OECD Pensions Outlook 2026, published in November 2026, emphasizes converting retirement saving into effective income and the role of choice architecture, advice, governance, and asset-backed pension design. This supplies contemporary policy context. It does not itself supply a standardized, internationally comparable schedule of 2026 retail life-annuity money's-worth ratios. Such a schedule would require a harmonized survey of quotes and assumptions not undertaken here.
Scientific Validity, Sensitivity, and Evidence Limits
Observation design and data provenance
An empirical report should identify the quotation source; collection date; issue date; premium and currency; purchaser age and relevant demographic classification; survivor/beneficiary configuration; payment calendar; escalation; guarantee and refund terms; insurer and intermediary; tax context; whether quotes were binding or indicative; and whether market prices were sampled or selected by the researcher. A repeated observation by the same insurer should be tracked consistently. If a quotation aggregator selects only offers meeting a particular eligibility profile, apparent market averages may be conditional on that platform's coverage.
The 2024 observations used by Poterba and Solomon come from a specific retail quotation source, with product and sample design disclosed in their paper. It would exceed their evidence to attribute their results to every United States insurer, to employer pension buyouts, to all qualified and nonqualified contracts, or to 2026 market conditions. A meaningful replication would obtain licensed contemporaneous quote data, record the exact contract features and underwriting eligibility, use the corresponding discount term structure, and disclose how the survival models and cohort projections were selected.
Mortality modeling requires particular care. A period life table reports mortality rates associated with a given calendar observation framework; the purchaser's lifetime payouts depend on a future cohort path through calendar time. Prospective projections may incorporate improvement, heterogeneity, and uncertainty at advanced ages. Selection can vary with the distribution channel, the option chosen, and time since purchase. The Society of Actuaries payout experience study is based on actual individual payout-annuity experience for 2005–2008; this is valuable historical experience evidence but cannot stand in for a fully projected 2024 or 2026 best-estimate cohort without appropriate transformation.
Sensitivity hierarchy and interpretation
There are several separate layers of uncertainty: contract interpretation; data measurement; demographic estimation; discount-curve choice; and structural economic assumptions about the cost of capital, opportunity costs, and household consumption. An informative sensitivity report varies each layer with others fixed and then tests plausible joint changes. The synthetic sensitivity table above isolates two crucial choices. It does not capture the additional uncertainty from the future path of interest rates, longevity improvements, taxes, and inflation, and its figures must not be treated as empirical intervals.
A robust empirical comparison should report at minimum the population-basis ratio and annuitant-basis ratio, with corresponding mortality source and calendar basis; a risk-free benchmark and a clearly differentiated asset- or credit-sensitive comparator; age, sex where legally relevant and available, and product options; and precision appropriate to original price data. For a guaranteed benefit, the analyst should calculate the portion of EPDV unaffected by survival to distinguish bond-like components from actual longevity protection. This prevents a product with a large minimum guaranteed period from being credited with the same amount of pure survival-risk transfer as a life-only annuity with the same payout present value.
The MWR cannot determine an individual policyholder's utility gain, household income adequacy, liquidity needs, distribution incentives, or insurer solvency. It is also not an internal rate of return realized by a specific buyer. A realized return depends on the lifespan actually lived, contractual survivor benefits, and financing assumptions. An ex ante IRR can be computed using expected benefits and a stated mortality basis, but it should not be confused with a guaranteed investment yield.
Insurance contract measurement and supervisory safeguards
Financial reporting and supervisory valuation impose their own units of account and objectives. Under IFRS 17, a group of contracts may involve estimates of future premiums, benefits, and expenses together with an adjustment for nonfinancial risk and a contractual service margin. The IFRS Foundation's standard overview explains its general measurement and recognition framework; it does not prescribe retail money's-worth ratios for a household's individual offer. Similarly, the NAIC model rules do not become operative uniformly merely by appearing in the NAIC's model-law catalogue. National and state implementation dates must be traced for any jurisdiction-specific legal claim.
The international IAIS supervisory framework provides a benchmark for transparent insurance conduct and the resilience of institutions, but each market's applicable law remains decisive. Evaluating a retailer's stated rate without policy wording, statutory protection, claims priority, or insurer financial soundness can produce a numerically correct but economically incomplete comparison. A higher present-value quote from a riskier issuer is not automatically preferable to a lower nominal quote whose risk characteristics differ.
Evidence strengths, limitations, and research ethics
The strongest historical propositions in this monograph are attributable to identifiable original studies, published research platforms, mortality table providers, and public regulators. The most recent numerical cross-product money's-worth comparison rests on a revised working paper with transparent sample dates and explicitly disclosed non-peer-reviewed status. We did not acquire a proprietary 2026 quote panel, conduct interviews with insurers, obtain commission schedules or policyholder-level medical records, or re-estimate historical papers' original datasets. The synthetic numerical illustration verifies the mechanics of an annuity valuation, not an empirical pricing fact. These boundaries distinguish supported conclusions from open research questions.
Peer-reviewed research differs in what is observed: studies using actual mortality outcomes identify selection patterns more directly than studies relying only on quoted payouts; quotation studies reveal relative prices but do not disclose every underlying expense; theoretical welfare studies can identify conditions for optimal annuitization but do not measure the prevailing retail charge for a particular contract. A credible synthesis preserves these differences instead of making one source answer all questions.
Integrated Findings and Research Agenda
The accumulated evidence supports the following integrated interpretation. Money's-worth is an ex ante valuation statistic conditioned on a well-defined beneficiary, contract, mortality basis, discount rate, and purchase date. It offers a reproducible way to compare annuity prices when the comparison set is properly harmonized. Selection is a central cause of differences between the population-basis valuation and the valuation relevant to typical purchasers. A product's contingent guarantees and commencement date determine the strength of that effect. These observations are consistent with the evidence from Mitchell and colleagues (1999), Finkelstein and Poterba (2002, 2004), and Poterba and Solomon (2025 revision), while empirical magnitudes remain sample-specific.
The second result concerns expense measurement. A money's-worth residual is a useful indicator for investigating contract pricing. It is not a disaggregated income statement. Direct information about commissions, administration, asset management, capital, taxes, investment spread, and financial performance is needed to explain how premiums support insured benefits. This distinction is especially important when comparing fixed payout annuities with variable contracts that assess separately identified fees on account value. A complete analysis should communicate both explicit customer charges and implied pricing conditions, always with their respective denominators and time horizons.
The third result concerns household and insurer value. A consumer may rationally purchase protection against uncertain longevity even when expected discounted payments are below the purchase premium; the utility literature provides the conditions under which this occurs. An insurer must manage aggregate survival risk, investment and reinvestment uncertainty, expenses, and capital. The ratio is an informative bridge between the parties only when cash-flow rights and valuation bases are reconciled. An inference from the customer ratio directly to an issuer's realized profit omits the balance-sheet economics central to annuity production.
These conclusions suggest a feasible research agenda within established actuarial and insurance-economic methods. First, compile harmonized dated quote panels for immediate, deferred, guaranteed-period, and survivor annuities across representative markets; retain the exact forms and binding terms alongside the stated yields. Second, calculate alternative mortality-basis ratios with documented cohort adjustments and uncertainty ranges. Third, acquire independently verifiable public or confidential expense and commission information and reconstruct cost allocation by segment and distribution method. Fourth, compare the ex ante pricing decomposition with ex post insurer experience and financial reporting under the relevant accounting standards. Fifth, investigate the welfare implications using household circumstances, public pension floors, bequest intentions, and liquidity requirements. Sixth, preserve code, versioned tables, legal effective dates, and source metadata to enable reproducibility. These are open empirical tasks, not claims that such observations were collected or tested in this Dossier.
A practical reporting format follows from the analysis: every comparative table should state the quotation date, premium and currency, payment timing, mortality cohort and projection, discount-curve convention, contract guarantee, calculated EPDV, MWR, and any known explicit fees. The research provenance should indicate whether a number is an observed market quote, a published estimate, a calculated hypothetical result, or a policy statement. That presentation standard is a methodological recommendation derived from existing science, rather than a distinct theory, product model, or doctrinal claim.
Conclusion
Annuity contract payout value is a conditional valuation of contingent payments. Its ratio to premium is an informative and reproducible description of price only when the contractual cash flow, valuation date, discount structure, and mortality population are explicit. The difference between the money's-worth ratio and one is a benchmark residual, within which selection, expenses, capital, asset intermediation, and commercial margins may all be relevant. Isolating any one component requires separate evidence.
Historical studies from the United States, United Kingdom, and other markets demonstrate that the money's worth of annuities varies across eras, institutions, and methods. The revised 2025 analysis of 2024 United States offerings supplies a recent example: simple immediate and deferred life annuities show markedly different present values for a general-population buyer and an annuitant with more favorable survival prospects. The modeled 2024 values are sample findings, and a corresponding 2026 cross-market estimate remains an open empirical undertaking. Household welfare, meanwhile, depends on the benefits of insured lifetime consumption and the opportunity cost of illiquidity, beyond the discounted sum of expected benefits.
A scientifically adequate assessment of product loading therefore joins a contract-level money's-worth calculation to evidence about distribution and maintenance costs, invested assets, risk capital, policyholder selection, legal protections, and insurer financial outcomes. Such a joined analysis permits disciplined comparison of customer value and insurer economics while respecting distinct product forms and institutional settings. The fundamental conclusion is operational: accurate attribution of annuity loads requires identifying the purchaser, the promise, the valuation basis, and the separately evidenced production costs before a pricing residual can be assigned an economic cause.
References
Brown, Jeffrey R., and James M. Poterba. 2000. “Joint Life Annuities and Annuity Demand by Married Couples.” Journal of Risk and Insurance 67 (4): 527–553. NBER Working Paper 7199. Original research source.
Cannon, Edmund, and Ian Tonks. 2002. “Annuity Prices, Money's Worth and Replacement Ratios: UK Experience 1972–2002.” University of Bristol, Centre for Market and Public Organisation, Working Paper 02/051. Original research source.
Cannon, Edmund, and Ian Tonks. 2004. “UK Annuity Price Series, 1957–2002.” Financial History Review 11 (2): 165–196. DOI: 10.1017/S0968565004000125.
Cannon, Edmund, and Ian Tonks. 2008. “Money's Worth Calculations.” In Annuity Markets, 117–139. Oxford: Oxford University Press. DOI: 10.1093/acprof:oso/9780199216994.003.0006.
Davidoff, Thomas, Jeffrey R. Brown, and Peter A. Diamond. 2005. “Annuities and Individual Welfare.” American Economic Review 95 (5): 1573–1590. DOI: 10.1257/000282805775014281. Publisher record: AEA article.
European Insurance and Occupational Pensions Authority (EIOPA). 2025. Costs and Past Performance Report 2025. Published April 15. Official source.
Finkelstein, Amy, and James Poterba. 2002. “Selection Effects in the United Kingdom Individual Annuities Market.” The Economic Journal 112 (476): 28–50. DOI: 10.1111/1468-0297.0j672. Publisher record: Economic Journal article.
Finkelstein, Amy, and James Poterba. 2004. “Adverse Selection in Insurance Markets: Policyholder Evidence from the U.K. Annuity Market.” Journal of Political Economy 112 (1): 183–208. DOI: 10.1086/379936.
Friedman, Benjamin M., and Mark J. Warshawsky. 1985. The Cost of Annuities: Implications for Saving Behavior and Bequests. NBER Working Paper 1682; published in The Quarterly Journal of Economics 105 (1) (1990): 135–154. Original research source.
Friedman, Benjamin M., and Mark J. Warshawsky. 1990. “The Cost of Annuities: Implications for Saving Behavior and Bequests.” The Quarterly Journal of Economics 105 (1): 135–154. DOI: 10.2307/2937822.
International Association of Insurance Supervisors (IAIS). 2024. Insurance Core Principles and Common Framework for the Supervision of Internationally Active Insurance Groups (ICPs and ComFrame). Updated December 2024. Official source.
International Financial Reporting Standards (IFRS) Foundation. 2023. IFRS 17 Insurance Contracts. Effective for annual reporting periods beginning on or after January 1, 2023. Official source.
International Financial Reporting Standards (IFRS) Foundation. n.d. IFRS 17 Insurance Contracts—Key Terms. Official source.
James, Estelle, and Xue Song. 2001. Annuities Markets Around the World: Money's Worth and Risk Intermediation. Working paper, SSRN. Original research source. DOI: DOI: 10.2139/ssrn.287375.
Mitchell, Olivia S., James M. Poterba, Mark J. Warshawsky, and Jeffrey R. Brown. 1999. “New Evidence on the Money's Worth of Individual Annuities.” American Economic Review 89 (5): 1299–1318. DOI: 10.1257/aer.89.5.1299. Publisher version: Journal article.
National Association of Insurance Commissioners (NAIC). 2022. Buyer's Guide to Fixed Deferred Annuities. Official source.
National Association of Insurance Commissioners (NAIC). n.d. Model Laws: Annuities/Variable Contracts, including Annuity Disclosure Model Regulation No. 245. Official source.
National Association of Insurance Commissioners (NAIC). 2025. Annuity Suitability and Best Interest Standard. Last updated February 10. Official source.
Organisation for Economic Co-operation and Development (OECD). 2026. OECD Pensions Outlook 2026. Paris: OECD Publishing. November. Official source.
Poterba, James M., and Adam Solomon. 2021, revised October 2025. New Evidence on the Money's Worth of Immediate and Deferred Individual Annuities. NBER Working Paper 28557. Empirical quotes dated May 2, 2024. Original research source. Revised text: Original research source.
Society of Actuaries (SOA). 2012. 2012 Individual Annuity Mortality Basic Table—Male, SOA Table Identity 2581. Official source.
Society of Actuaries (SOA). 2012. 2012 Individual Annuity Reserving Report and Table. Official source.
Society of Actuaries (SOA). 2014. 2005–08 Individual Payout Annuity Experience Report and Excel Files. Official source.
United States Securities and Exchange Commission (SEC). n.d. Variable Annuities: What You Should Know. Investor publication. Official source.
Yaari, Menahem E. 1965. “Uncertain Lifetime, Life Insurance, and the Theory of the Consumer.” Review of Economic Studies 32 (2): 137–150. DOI: 10.2307/2296058.