RESEARCH DOSSIER
Mandatory Annuitization Rules and Retirement Income Regulation
Abstract. Mandatory annuitization is a regulatory intervention at the boundary between accumulated retirement capital and the distribution of income over an uncertain lifetime. Its legal architecture determines who must convert capital, how much must be converted, which forms of income qualify, when payment begins, and whether exceptions or competing payout options remain available. The intervention redistributes survival, investment, liquidity, inflation, and bequest risks among retirees, insurers, pension funds, households, and public institutions. These allocations cannot be inferred from the word “mandatory” alone: a required contribution is distinct from an annuity-purchase obligation, a required pension income stream may be delivered through an investment-linked living annuity, and a minimum annual account withdrawal creates no lifetime insurance guarantee. This monograph develops a source-critical comparative examination of retirement payout mandates in the United Kingdom, Singapore, Chile, South Africa, Switzerland, Australia, and the United States. It locates contemporary rules within their legal histories, specifies eligibility thresholds and payout alternatives, and interprets welfare effects using established actuarial present-value analysis and the economics of annuity demand. The United Kingdom’s removal of its age-75 annuitization requirement in 2011, followed by broader defined-contribution withdrawal flexibility in 2015, provides a temporal contrast to Singapore’s cohort- and savings-dependent automatic CPF LIFE inclusion and South Africa’s 2024 preservation and retirement-component architecture. Chile offers statutory choice among annuity and programmed-withdrawal modalities; Swiss occupational pensions employ statutory pension conversion terms alongside lump-sum possibilities; Australian minimum pension payments and U.S. required minimum distributions demonstrate distribution regulation without universal annuity purchase. The central finding is that policy evaluation requires disaggregation of five separate outcomes: the capital legally committed to retirement income, the actual lifetime insurance coverage purchased, the adequacy and real value of payments, the retiree’s remaining liquidity, and the distribution of financial and mortality risk. Comparative evidence supports a conditional case for effective longevity pooling and well-governed payout pathways, while revealing material differences in consumer welfare across health status, income, family structure, inflation exposure, and existing public-pension entitlements. The work provides an original integrative interpretation of verified evidence and transparent hypothetical calculations. It introduces no new proprietary theory, empirical dataset, or claim of scientific priority.
Keywords: Mandatory annuitization; retirement income regulation; lifetime annuity; pension payout mandates; compulsory annuitization; CPF LIFE; programmed withdrawal; living annuity; defined contribution; tax-qualified retirement savings; longevity risk; pension decumulation; insurance regulation; retirement income adequacy; social protection; consumer protection; pension freedom; international comparative law.
Introduction
Retirement policy creates claims on future resources through taxes, contributions, public transfers, funded savings, employer promises, and private insurance contracts. A payout requirement governs the final transformation of accumulated entitlements into usable income. This transformation has consequences for the timing and amount of consumption, the financing of unexpected medical and care expenses, the security of surviving family members, the intergenerational transfer of unspent assets, and the capacity of financial intermediaries to bear longevity and investment risks.
The OECD Pensions Outlook 2026 describes a fundamental contemporary tension. Pension assets across OECD countries reached approximately USD 78 trillion at the end of 2025, yet aggregate asset growth does not assure adequate retirement income for individuals. Contribution levels, accumulation returns, access to advice, and the design of the payout phase all affect retirement outcomes (OECD, 2026). A regulatory policy that effectively annuitizes an inadequate balance may ensure regularity of payment while preserving a low lifetime standard of living. Conversely, access to a substantial account balance without structured lifetime provision may leave a household exposed to depletion in extreme old age.
The scientific object examined here is the legally or regulatorily required conversion of pension wealth into an income arrangement, with special attention to the subset that transfers individual survival risk into a pooled lifetime contract. The distinction matters because a payout pathway can be mandatory without requiring a conventional insured life annuity. Governments also regulate contribution obligations, eligibility for social pensions, benefit commencement, tax-qualified withdrawals, and the technical standards of pension income. These institutions interact, but each changes different margins of household choice.
The study adopts an international comparative rather than a purportedly universal legal description. National provisions are treated as historically situated rules applicable to designated workers, pension products, retirement cohorts, account components, and settlement dates. The phrase “the country requires an annuity” is analytically incomplete until its applicable legal population, product definition, exemptions, and effective date are specified. The OECD pension design recommendation requires coherence between accumulation and payout phases and with the wider pension system. It does not enact an identical annuitization mandate across adhering jurisdictions (OECD, 2022).
The research uses official pension administrator statements, government legislation and tax explanations, international institutional material, and peer-reviewed annuity economics. The legal investigation distinguishes a law, supervisory guidance, administrative implementation, and public explanation. Scholarly claims derived from journal records and abstracts are identified at their actual evidentiary scope. Calculations within this monograph are explicitly stylized, reproducible illustrations of mechanisms, never estimates of national policy impact. All sources are individually recorded in the associated evidence ledger.
Questions and evidentiary sequence
Three questions organize the inquiry. First, which legal and administrative records establish the mandatory conversion threshold and define precisely which retirement assets are affected? Second, how do permitted payout alternatives redistribute longevity, investment, liquidity, and bequest risks? Third, what can be inferred about market development and social welfare when national schemes, historical periods, and heterogeneous populations are compared?
An adequate answer proceeds from the instrument to the eligible individual, from the individual to the feasible contract or payout set, and from that set to economic outcomes. It then interrogates identification: legal availability is different from actual take-up, contract choice is different from realized lifetime security, and measured average payouts are different from welfare experienced by a short-lived or very long-lived retiree.
Research Foundations and Institutional Genealogy
The legal object of annuitization
A life annuity promises specified payments while a designated person remains alive. Its defining economic property is survival contingency: the provider’s aggregate liability depends on the survival distribution of covered lives. An annuity may be immediate or deferred, level or escalating, single-life or joint-life, and may contain a guaranteed minimum payment period. These contract terms change its price and the recipient’s exposure to competing risks. A temporary pension, a drawdown account, and a lifetime annuity may all produce monthly transfers, but only some establish an enforceable promise of payments contingent on continued survival for life.
Mandatory annuitization can intervene at different stages. Law may require a fraction of eligible retirement capital to be converted at retirement, require a minimum income before remaining assets can be withdrawn, automatically enroll individuals in a lifetime payout pool, or prescribe a permissible retirement income product that includes both insured and market-linked variants. A supervisory standard may additionally constrain the pricing, disclosure, solvency, fund governance, or administration of these products. The legal command is therefore a compound specification: covered capital, required benefit, qualifying provider, eligibility date, exceptions, and enforcement.
This anatomy clarifies why the presence of compulsory occupational pension participation is insufficient proof of compulsory private annuity purchase. The OECD pension system architecture differentiates public and occupational schemes, defined benefit and defined contribution structures, and mandatory, quasi-mandatory, and voluntary arrangements (OECD, 2025). A worker may have no choice about participation while retaining choices about retirement timing or payment form; a participant may have an insured lifetime benefit without individually purchasing an insurer-issued retail annuity.
The longevity contingency and international social protection
The public-interest rationale for lifetime income begins with uncertain survival. A person retiring at a specified age knows neither the ultimate lifespan nor the trajectory of future essential expenditure. A fixed stock of resources must therefore finance a random duration of consumption. A collective pool of independent mortality risks can finance survivor-contingent income with mortality credits that an individual self-managed account cannot reproduce without pooling. The achievable credits are affected by cohort selection, expenses, reserves, and interest-rate conditions, and the mechanism does not eliminate inflation, provider credit, or benefit-adequacy risk.
The ILO Convention No. 102 guidance makes the duration of old-age protection a central dimension of social security: an old-age contingency may continue until death, while an exhausted defined-contribution account may cease paying despite the contingency persisting. The ILO also recognizes that legal permission to purchase a life annuity is not sufficient when effective market provision is scarce. This is an institutional observation, not a rule that all signatory states must adopt a particular private-insurance contract. The convention concerns social-protection outcomes, collective financing, and administration as well as payment form (ILO, 1952; ILO guidance).
The OECD recommendation on DC plans similarly places payout design inside a comprehensive pension architecture. Its emphasis on adequate retirement income, stability, governance, and member interests means that annuitization requirements must be evaluated alongside basic pensions, targeted support, contribution careers, disclosure, and the functioning of pension and insurance markets. A formal lifetime payment can satisfy duration but fail on adequacy; an adequate starting payment can lose much of its real value under prolonged inflation; and a favorable projected pension can be unreachable for a worker whose career includes major contribution interruptions.
Historical evolution and policy reversibility
The British experience supplies a rare well-documented reversal of a tax-driven annuitization regime. Historically, tax-privileged retirement saving was expected to finance pension income rather than an unrestricted transfer of wealth into consumption or inheritance. The United Kingdom imposed an effective age-75 annuitization requirement under earlier pension tax arrangements, subject to particular alternatives. HM Treasury’s 2010 consultation on the requirement recorded the announced intention to remove the effective requirement from April 2011 (HM Treasury, 2010).
The HMRC pension manual states the operational chronology precisely: before June 22, 2010, annuitization at 75 was required unless the savings were retained as an alternatively secured pension; the requirement ceased from April 6, 2011; and there has been no general requirement to annuitize at any age since then. From April 6, 2015, the Taxation of Pensions Act 2014 provided further defined-contribution flexibility, including the possibility of withdrawing an entire fund subject to tax. The two dates record distinct reforms. Conflating them would misstate when the legal annuitization mandate ended.
The House of Commons account of drawdown further explains the intermediate period in which drawdown was available but subject to limits, with exceptions for persons possessing sufficient other pension income. The implication is methodological: a binary before-and-after comparison centered only on 2015 cannot identify the effect of removing compulsory annuitization in 2011. Changes to withdrawal limits, investment markets, tax treatment, provider supply, and advice obligations must also be considered.
Retirement income as a regulated sequence of claims
Legislation often governs several different flows bearing the same ordinary-language label of pension. A public old-age transfer may be payable periodically under social legislation; employer-sponsored occupational benefits may be defined by scheme rules; an insurance annuity may constitute a contractual obligation of a regulated provider; and an account-based retirement pension may be a systematic withdrawal from the participant’s own investment assets. These claims differ in enforceability, insolvency exposure, portability, indexation, survivor coverage, and inheritance.
An economically meaningful compulsory conversion rule should therefore be specified at the exact claim level. The first data field is the underlying balance or entitlement. The second is the legally required destination of that balance. The third is the resulting legal claim, including who bears shortfall risk after commencement. The fourth is the permitted amendment, withdrawal, or commutation of the payment. The fifth is the interaction with fiscal support and existing income. This descriptive sequence is an analytic reading procedure drawn from established pension, contract, and public-finance concepts; it is not a newly proposed regulatory doctrine.
Mandatory Conversion Thresholds and Legal Specification
The conversion threshold as a vector of legal conditions
The threshold of mandatory annuitization may be expressed in terms of age, retirement commencement, calendar cohort, contribution history, accumulated balance, fraction of assets, minimum income, or pension-account category. These conditions can be cumulative. A nominally universal scheme may apply in practice only to a defined population above a savings threshold, and the exemption for a pre-existing adequate pension may reduce mandatory purchases among precisely those with other reliable retirement resources.
For comparison, let K denote qualifying retirement capital at the relevant settlement date, let alpha represent the fraction of that capital subject to a conversion requirement, and let E denote the value excluded by explicit grandfathering or exemption. The legally constrained principal can then be represented descriptively as C = alpha × max(K − E, 0), subject to any scheme-specific threshold. This identity does not assert that national laws calculate the base in this way: it exposes the data necessary to reconstruct the actual base in each scheme. Some arrangements apply a retirement-component definition directly; others apply an income sufficiency rule, and others prescribe a plan that generates payments without a private-market purchase.
Singapore: automatic inclusion in national longevity insurance
Singapore’s CPF LIFE is a national lifetime-income annuity arrangement operated within the Central Provident Fund system. The CPF LIFE inclusion rules state that citizens and permanent residents born in 1958 or later who have at least SGD 60,000 of relevant retirement savings when starting monthly payouts are automatically included. Those outside automatic inclusion may join voluntarily, subject to the program’s age conditions. Official guidance further describes exemptions when a qualifying pension or private annuity already supplies at least comparable monthly income (CPF Board, 2026 access).
The rule is therefore cohort-dependent, residency-dependent, balance-dependent, and subject to qualifying alternatives. It cannot accurately be summarized as automatic lifetime annuitization of every CPF member or of every dollar accumulated in CPF. The Ministry of Manpower’s CPF explanation distinguishes CPF LIFE monthly payouts from withdrawable savings above applicable retirement sums and the Retirement Sum Scheme that can pay until allocated savings are exhausted. Retirement income regulation operates jointly with savings set-asides and exceptions.
CPF LIFE also preserves important timing and product choices. According to the CPF payout election instructions, automatically included members may choose to begin payouts from 65 through 70, and those who do not choose a plan before 70 are placed on the Standard Plan with payments commencing at 70. Persons who are not automatically included may opt in within the separately specified age interval. The CPF plan description distinguishes plans with different initial payout and escalation patterns. Automatic membership thus operates inside a structured choice set rather than producing identical contracts for every retiree.
Singapore illustrates a public or quasi-public institutional solution to a problem of lifetime income market participation. Its automatic inclusion rule reduces the need for every covered household to actively select a conventional retail annuity. It also creates administrative responsibilities for cohort identification, savings records, benefit quotation, exceptions, and communication of irreversible or costly payout decisions. Evaluation requires observed outcomes across the actual covered population, including persons below the threshold, not just the performance of the pool among enrolled retirees.
United Kingdom: an ended annuitization requirement and evolving withdrawal rights
The United Kingdom is essential to a comparative account precisely because it formerly maintained and then ended a requirement. The 2010 Treasury consultation addressed an age-75 annuitization obligation associated with pension-tax treatment. The HMRC chronological account identifies April 2011 as the abolition date and April 2015 as a separate liberalization of access to defined-contribution capital. Continuing to describe modern British defined-contribution pension policy as generally requiring annuity purchase would therefore be legally inaccurate.
Removal of a general tax rule does not make every occupational pension promise convertible into any form the individual wishes. Scheme contracts, pension legislation, the form of benefit already accrued, and protections applicable to defined benefit arrangements continue to matter. The HMRC manual expressly distinguishes statutory flexibility from the benefits an individual scheme actually offers. An employer-defined pension can remain a lifetime payment even though a participant in a different defined-contribution arrangement may select drawdown or cash.
The legal comparison thus needs a unit narrower than the state. Consider two British retirees at the same age. One has a defined-contribution account whose plan permits flexible drawdown; the other holds an employer-backed pension defined as a lifelong income claim. Both reside under the same national framework, but their available payout choices differ because the underlying entitlements differ. A national “mandate variable” assigned to both without specifying product and period loses the policy distinction that the research seeks to measure.
Chile: regulated choice of annuity and programmed withdrawal
Chile’s funded pension framework provides a case in which participants encounter regulated retirement payout options rather than universal annuitization. The Superintendencia’s old-age pension guidance describes a balance certification and quotation process using the system for pension amount offers, allowing eligible retirees to evaluate modalities and benefit amounts. The institution identifies multiple pension modalities, including life annuity and programmed withdrawal. Reaching statutory pension age alone does not automatically require immediate retirement (Superintendencia de Pensiones, 2026 access).
In Chile’s programmed-withdrawal rules, the payment depends on accumulated account assets and technical parameters and is recalculated according to regulatory requirements. This form leaves exposure to investment performance and changes in the payout base that differs from the fixed lifetime obligation normally associated with an insurer-issued annuity. It is therefore analytically important to separate the rule that pension resources must be used within the authorized retirement system from any supposed universal duty to choose the life-annuity modality.
The IMF’s 2026 Chile analysis characterizes the system as combining public protection, mandatory contributory capitalization, and voluntary additional savings, with accumulated balances at retirement convertible into an annuity, programmed withdrawal, or a combination. The 2025 pension reform adjusted the institutional framework and contributions. The Chilean authority’s 2025 notice further documents changes to the life-annuity quotation environment, including the elimination of external offers under specified implementation rules. Those reforms affect procurement and competition; they do not establish that programmed withdrawals have been prohibited. Effective dates and deferred amendments must be checked against the relevant pension provisions before any historical dataset is coded.
Chile makes visible the difference between a regulated menu and a mandated product. Multiple payment modalities can coexist with mandatory pension participation and detailed supervision of retirement quotes. Consumer welfare can then depend on mortality information, price transparency, distributional circumstances, insurer competition, and the quality of the individual’s comparison of alternative payment structures.
South Africa: income compulsion with living-annuity alternatives
South Africa’s two-pot reform took effect on September 1, 2024. The government’s enactment announcement records the accompanying Pension Funds Amendment Act and the preservation objectives of the reform. Contributions under the new arrangement are allocated between a savings component, generally one-third, and a retirement component, generally two-thirds, subject to applicable scheme and transitional rules. The retirement component is preserved for retirement income rather than ordinary pre-retirement cash withdrawals (Government of South Africa, 2024).
The detailed SARS lump-sum directive guidance is critical to precise classification. It states that the retirement component is used to provide a pension or purchase a living annuity, guaranteed annuity, or combination, subject to exceptions. It distinguishes vested and nonvested amounts and prescribes a small-balance commutation threshold. In particular, the guidance states that from March 1, 2026 the specified small-annuity commutation threshold increases to ZAR 240,000 from ZAR 165,000 for the preceding period; eligibility depends on the relevant combined component calculation, not the total account balance viewed without classification (SARS, 2026).
This is a substantive annuitization-type mandate with a wider set of permissible income mechanisms than a pure insurance-life-annuity purchase requirement. A living annuity ordinarily leaves investment and drawdown risk with the beneficiary, subject to applicable restrictions, and need not provide an insurer-guaranteed payment for life. Calling both alternatives “annuities” without specifying who bears longevity and investment shortfall risk erases a central welfare difference.
South Africa’s law also embeds historical property and transitional distinctions. Contributions before reform, grandfathered benefits, and certain members have different rights from new flows into the retirement component. An empirical estimate of the proportion of total retirement wealth required to be applied to income cannot simply assign the two-thirds contribution ratio to every accumulated asset. Each individual account must be decomposed into legally classified components, adjusted for the reform commencement date and applicable commutation exceptions.
Switzerland: statutory conversion with elected lump-sum possibilities
Swiss occupational pensions demonstrate a different use of the word conversion. The official 2026 social security overview explains that accrued mandatory occupational pension assets are ordinarily converted into a lifelong pension, subject to recognized lump-sum alternatives, using the statutory minimum conversion rate of 6.8 percent for the mandatory part. The conversion rate translates qualifying accumulated pension assets into annual pension income; it is not an investment yield or a mortality probability. The official overview also notes the relevance of pension-fund regulations and eligible lump-sum payments.
The legal and economic question here is how the statutory conversion formula applies to mandatory versus supplementary occupational assets, how institutional rules permit or constrain lump sums, and how benefit obligations are financed. It would be erroneous to multiply every Swiss pension holding by the statutory rate without determining the statutory base and the fund’s actual plan terms. The Swiss case also illustrates a potential divergence between legislated conversion terms, actuarial cost and financial-market rates, which must be modeled before drawing conclusions about distributional transfers.
Australia and the United States: regulated withdrawals without a general annuity mandate
Australia regulates pension income streams after commencement. The Australian Taxation Office’s minimum payment table lists age-dependent minimum annual pension factors, including 5 percent for an account-based pension holder aged 65–74 under the normal schedule from 2023–24 onward. The ATO description of income streams distinguishes account-based and non-account-based payments, and sets periodic payment standards. An account-based pension with a required minimum payment is not thereby an insured lifetime annuity.
The United States generally imposes required minimum distributions on qualifying tax-deferred individual and employer retirement accounts, while allowing other retirement-income arrangements under relevant rules. The IRS required-distribution guidance explains the annual required withdrawal concept and its account- and age-specific treatment, generally beginning at age 73 under its stated rules for affected cohorts, with separate rules for workplace retirement and beneficiaries. The amount must be withdrawn from the relevant account; the rule does not amount to a universal private-annuity purchase obligation.
Together these jurisdictions define a useful negative comparator. The government can compel some decumulation from tax-favored assets without buying out survival risk. Minimum withdrawal law regulates tax timing and the path by which pension capital leaves a privileged account. Whether the withdrawn sum is consumed, reinvested, used for insurance, or retained as cash depends on other rules and decisions. The fact of a mandated distribution cannot, on its own, support an inference of adequate old-age consumption or lifetime protection.
Comparative legal matrix
| Jurisdiction and regime | Covered mandate or default | Qualifying outcome or alternatives | Essential qualification |
|---|---|---|---|
| Singapore, CPF LIFE | Automatic inclusion for specified CPF cohort, residency, and retirement balance | National lifetime payouts with plan and commencement choices | Exemptions and voluntary inclusion outside automatic criteria |
| United Kingdom, modern DC tax regime | General age-75 annuitization requirement ended in 2011 | DC annuity, drawdown, taxable lump-sum access according to applicable schemes | 2015 flexibility is separate from the earlier abolition |
| Chile, funded pension payout | Pension capital directed through regulated pension modalities | Life annuity, programmed withdrawal, combination and other authorized modalities | Choice and quotation rules; no universal annuity-purchase rule |
| South Africa, two-pot retirement component | Retirement-income application of specified component | Pension, guaranteed annuity, living annuity or combination | Vested rights, commutation thresholds and effective dates |
| Switzerland, mandatory occupational pension | Statutory lifelong conversion framework | Pension plus law- and fund-dependent lump-sum possibilities | Statutory conversion rate applies to defined mandatory base |
| Australia, account-based pension | Minimum payment once pension has commenced | Periodic withdrawals from invested account | Payment minimum does not create longevity insurance |
| United States, tax-deferred retirement accounts | Required minimum distributions for covered accounts | Account withdrawals; insurance purchase is a separate choice | Account, cohort, retirement status and beneficiary exceptions |
This matrix encodes the legal logic documented in the primary administrative sources above. It is a comparative research summary, not a substitute for jurisdiction-specific statutory interpretation in an individual case.
Permitted Payout Alternatives and Risk Allocation
Lifetime income and the actuarial present value
The insured lifetime annuity exchanges a premium for survival-contingent claims. Under a simplified annual end-of-year payment convention, let x be purchase age, v(t) the discount factor to payment t, p(x,t) the probability that the covered person survives t years, and B(t) the specified payment if alive. The prospective expected present value of benefits is:
This is an actuarial valuation identity, not a legislated formula or a statement that every annuity is fairly priced. Real contracts may include expenses, profit margins, solvency capital, guarantees, survivor provisions, and tax effects; mortality probabilities can depend on cohort and selection. Payment timing may be monthly rather than annually, and a realistic valuation uses the appropriately matched discount basis. A legal obligation to purchase an annuity changes the demand facing institutions that estimate and price these elements.
If a retiree has eligible capital K, a rule requires fraction alpha to be annuitized, and a stylized insurer offers a constant annual life payment using a purchase factor a, the illustrative initial annual payment is P = alpha K / a. With K = 100,000 units, alpha = 0.60, and a = 15, P = 4,000 units, leaving 40,000 units outside the modeled forced purchase. The factor is assumed, not observed, and no jurisdictional currency or product quotation is implied. The 4,000-unit stream does not establish real-income adequacy without information on taxes, public pensions, housing costs, indexation, and other household resources.
This presentation also shows that a seemingly modest alteration in compulsory fraction changes both payment and liquid wealth. Increasing alpha from 0.60 to 0.80 while retaining the illustrative factor raises the annual payment from 4,000 to approximately 5,333.33 units but reduces residual liquid wealth from 40,000 to 20,000 units. The direction of the trade-off is mechanical; the net welfare effect is conditional on mortality, other insured income, financial needs, and preferences.
Programmed withdrawal and living annuities
An account-based retirement payout operates under a different balance identity. Suppose a retirement account has beginning balance A(t), earns a net investment return r(t), and pays W(t). Under a simple year-end withdrawal convention its balance evolves as A(t+1) = A(t)[1+r(t)] − W(t). Regulatory rules may calculate W(t) from age-based percentages, modeled capital-needs factors, or predetermined income ranges. Unless a separate guarantee or external guarantee fund intervenes, sufficiently adverse returns, long survival, or high withdrawals can exhaust the account.
Chile’s programmed-withdrawal framework and South Africa’s living-annuity allowance establish different institutional instances of a common economic distinction: mandated use for retirement income can leave residual longevity risk with the retiree. The amount of regulatory protection must consequently be evaluated through the distribution of income paths, including extreme-old-age scenarios, rather than through the initial nominal payout alone.
A living annuity can preserve flexibility and potential residual account value. It may also expose household income to return sequence risk: an adverse investment shock early in retirement reduces the capital that can earn subsequent returns, particularly if withdrawals are maintained. The converse is that a guaranteed life annuity may better protect a stable nominal flow, subject to provider security and contract terms, but generally commits capital that would otherwise be available for emergency spending or inheritance.
The effects of inflation and payment escalation
Duration alone is insufficient to establish real lifetime protection. A level nominal annual payment P has purchasing power P/(1+i)^t at year t if the price index grows at a constant illustrative rate i. With hypothetical inflation of 2.5 percent each year over 25 years, purchasing power falls to roughly 53.94 percent of its initial value. The calculation uses only compound arithmetic; it forecasts neither inflation nor any country’s pension benefit.
Index-linked and escalating annuities can address portions of purchasing-power risk, but they require financing through asset returns, lower initial payments, or other pricing terms. A higher initial nominal guaranteed annuity may therefore be less effective at protecting basic consumption in extreme old age than a lower initial but escalating product. Singapore’s CPF LIFE payout options demonstrate why even an automatic national scheme offers distinct payout profiles that ought to be evaluated over time instead of ranked solely by first-month payment.
Survivors, guarantees, and family balance sheets
One-life annuitization can secure payments during the covered individual’s survival while ceasing at death. For a married or dependent household, a joint-life or survivor-continuation benefit may be more important than the nominal amount payable while both partners are alive. Guaranteed minimum payment periods, premium refunds, and survivor percentages transfer different amounts of mortality risk back toward heirs or beneficiaries. Such features can reduce the pure mortality credits available to the primary annuitant and alter price comparisons.
The welfare benchmark therefore requires a household rather than an isolated individual when retirement resources are shared. The relevant outcomes include the survivor’s continuing public pension, property tenure, dependency expenses, access to healthcare, and the security of remaining liquid assets. If the compulsory conversion threshold ignores family status, an identical mandated premium may imply different economic exposures and preferences across otherwise similar accounts.
Liquidity, medical expenses, and bequest preferences
Illiquid life income creates an intertemporal commitment. The commitment is valuable when it protects consumption from later market and survival uncertainty; it can impose an opportunity cost when the individual needs a large payment for care, housing repair, family support, or debt settlement. The theoretical annuity welfare analysis by Davidoff, Brown, and Diamond (2005) establishes that substantial positive annuitization is often optimal under market incompleteness, while full annuitization is not generically optimal in the presence of other risks and preferences. Their published analysis is conditional on assumptions rather than a universal command to mandate all retirement capital.
Mandatory income floors and partial-annuitization rules can be interpreted in light of this established distinction. An arrangement that preserves adequate liquid resources while protecting some longevity exposure may accommodate heterogeneous needs more effectively than forcing every retiree to surrender all accessible capital. However, merely permitting partial conversion does not demonstrate that a specific legal percentage is optimal. Calibrating the policy requires distributional information on wealth, health, care expenditures, public pensions, and insurance terms.
Comparing contract economics and consumers' money's worth
Expected discounted payouts relative to the premium are a widely used empirical measure of the money’s worth of annuity products. The 1999 annuity pricing study by Mitchell, Poterba, Warshawsky, and Brown evaluates the expected present value of life-annuity payments under alternative mortality and discount assumptions. This literature establishes why price assessments can change when population rather than annuitant mortality is used, and why simple nominal payout comparisons can mislead.
The money’s-worth ratio is distinct from lifetime utility value. A risk-averse retiree may value insurance against survival to extreme ages even when the expected discounted cash payments are below the premium. Conversely, an annuity that appears attractive under average mortality may yield low expected value for an individual with reliably shortened life expectancy. A coherent welfare study must measure insurance value, financial price, and legal compulsion on different axes.
Market Development, Welfare Effects, and Distribution
Selection and the annuity market
Annuity purchase is affected by private information and observable mortality heterogeneity. People who anticipate longer survival may seek more lifetime-income coverage when purchase is voluntary. The resulting pool can differ systematically from the broader population whose mortality was used in public statistics. A mandate broadening the purchaser pool may alter average longevity risk and the cost of adverse selection, but it can also require persons with shorter expected survival to enter contracts they would otherwise decline.
The British evidence is especially instructive. In Selection Effects in the UK market, Finkelstein and Poterba (2002) compare voluntary and compulsory annuity markets in the United Kingdom and report that annuitants are longer-lived than non-annuitants, with selection differences more pronounced in the voluntary segment. Their reported estimate places adverse selection in the compulsory market at approximately half the level measured in the voluntary market, within the historical markets studied. This empirical result supports the proposition that purchase regime affects pool composition; it does not imply that the same magnitude applies to Singapore CPF LIFE, Chilean products, or South African living annuities in 2026.
The subsequent UK policyholder study by Finkelstein and Poterba (2004) finds relationships between contract attributes and realized mortality even when annuity size itself does not exhibit a substantive mortality association in the analyzed data. Selection may act through timing, estate benefits, or other design terms. A compulsory regime can therefore reduce one form of selection without eliminating differences in the choice among permitted contracts.
Provider solvency and the institutional location of risk
When a private insurer promises lifetime payments, it acquires exposure to lower-than-anticipated mortality, investment returns insufficient to fund long-duration liabilities, changes in credit spreads and interest rates, and operational obligations extending over decades. Prudential regulation and reserves influence the reliability of the promised income. When the payout is supported by a public scheme, sovereign or public-institution commitments and financing become central. When payment comes from a drawdown account, much of the financial risk remains with the member.
These arrangements are not equivalent policy achievements merely because all can produce retirement income. Mandatory demand for insured annuities may expand the duration and scale of insurer liabilities and the need for long-term asset matching. A public pension conversion formula can instead produce intergenerational or cross-cohort financing effects. An account-based pension may generate no new external insurance liability at all. Measuring the market-development effect of a legal reform thus requires classifying the product actually selected and tracing the associated asset and liability.
Income adequacy and distributional inequality
The principal policy objective is the capacity to sustain consumption during retirement. Its measurement cannot be reduced to a count of annuity policies or aggregate pension assets. The OECD’s 2026 pension assessment emphasizes contribution gaps and the difference between large asset pools and individual preparedness. Any conversion rule acts on accumulated wealth; it does not by itself create the contribution history needed to finance a sufficient pension.
Distributional analysis must account for employment, gender, mortality, income, race where lawful and ethically appropriate data exist, access to employer contributions, and differing public transfers. The OECD 2025 old-age poverty chapter reports average relative old-age income poverty for women of 16.9 percent and men of 11.7 percent among the countries in its comparison. Those figures refer to OECD relative household-income measurement and cannot be read as a quantified effect of annuitization mandates. They establish heterogeneity in baseline retirement security, which is necessary context for any claim that compulsory conversion affects all groups similarly.
The selection problem also has a distributional dimension. In a pool with uniform pricing but heterogeneous life expectancy, the expected discounted benefits of long-lived groups are larger, other things equal. Whether that transfer is socially desirable depends on the relevant welfare framework, public support, employment history, financing of contributions, and possible compensating distribution through social pensions. It is not enough to observe that lifetime pooling transfers resources across realized lifespans; that is the insurance mechanism itself.
Existing public pensions and crowding of financial protection
A household already receiving a substantial public lifetime pension has a different marginal need for additional annuity income from a household whose retirement resources are almost entirely in an individual savings account. Mandatory private or occupational annuitization may duplicate some already insured income while reducing discretionary liquidity. Conversely, where social pensions are limited, a well-capitalized lifetime-income product may be particularly consequential for maintaining consumption in extreme old age.
The ILO guidance on social security focuses attention on income continuity throughout the old-age contingency and the possibility that account-depletion benefits end too soon. This identifies a functional test for policy: whether the combined public, occupational and individual system finances a survivable income path, not whether a specific instrument is mandatory in isolation. The amount and adequacy of public transfers must therefore be included in household-level simulations, with benefit taxation and means-testing incorporated where relevant.
Market competition, quotations, and consumer understanding
A mandated purchase pool can produce scale economies and stable demand, potentially supporting more provider offerings and deeper hedging markets. It can also weaken a consumer’s threat to decline annuitization, making quotation transparency and market-conduct supervision especially important. Whether the net result improves prices is an empirical question conditioned on market concentration, expenses, capital rules, distribution incentives, and disclosure.
Chile’s pension modality quotation process is a concrete institutional response to difficult retirement contract comparisons. The 2025 supervisory reform notice illustrates how reforms can target the quotation process and external offer channels without eliminating all payout choices. A comparison of prices before and after such reform should isolate changes in quote quality, take-up, insurer entry, and annuity terms rather than assuming that every change reflects the compulsory status of conversion.
The OECD’s pension-plan recommendation emphasizes well-designed communication, governance, and decisions in the interests of members. For pension decumulation, the choice environment is particularly consequential because the consumer faces mortality uncertainty, irreversible or difficult-to-reverse elections, and a long horizon for investment and inflation outcomes. Disclosure should therefore distinguish guaranteed from contingent payments, show survivor and death-benefit features, state fees and indexation rules, and make the costs of payment flexibility visible.
Assessing reforms without overstating causality
A before-and-after change in annuity sales does not establish the welfare impact of a mandate. Product pricing, interest rates, demographic cohort size, tax rules, employment participation, pension wealth, and distribution channels may all change simultaneously. In the United Kingdom, 2011 and 2015 represent different interventions; South Africa’s September 2024 reform combined new access to a savings component with strengthened preservation of retirement contributions; Chile’s 2025 changes modified a wider pension architecture. A simple national time trend would confound these multiple moving components.
Policy evaluation should specify the outcome of interest in advance. Suitable outcomes include the proportion of retirees with guaranteed income lasting until death, household real consumption at advanced ages, the number and severity of asset-depletion events, premium money’s worth, complaint and mis-selling rates, voluntary versus mandatory annuity participation, and changes in net public expenditure. Reliable cross-jurisdiction inference requires harmonized definitions, not merely common labels.
Robustness, Independent Validation, and Source Criticism
Evidence hierarchy and legal effective dates
The study distinguishes four levels of proof. Binding law and directly operative statutory rules establish legal obligations. Regulatory instructions and official pension administrator notices establish applied procedures within their documented authority. International institutional reports provide comparative definitions and measurements. Peer-reviewed economics supports mechanisms and empirical findings for the studied populations, with portability to other jurisdictions subject to investigation. Each category answers a different scientific question.
For the United Kingdom, the temporal sequence is decisive. The HMRC account of 2011 abolition must take precedence over retrospective summaries that compress changes into the phrase “pension freedoms in 2015.” For South Africa, the SARS directive details identify the 2024 component structure and the 2026 change in the small-annuity threshold. For Singapore, the CPF eligibility record identifies the particular birth cohort, minimum savings and exemption conditions. Removing any of these qualifiers would produce a materially different legal description.
Chile has further implementation-date complexity. The 2025 reform contains amendments with distinct commencement rules; some provisions may take effect later than the statute’s enactment. The current programmed-withdrawal compendium and official explanatory material therefore need to be compared against the legal state on the observation date, particularly if a later statistical dataset spans the reform period. An assertion about “the 2025 rule” is incomplete without the specific amendment and effective date.
Reproducible arithmetic checks
Three numerical illustrations in this monograph use explicit assumptions. First, for K = 100,000, alpha = 0.60 and purchase factor a = 15, premium commitment is 60,000 and stylized annual income is 4,000. Second, for alpha = 0.80, the premium commitment rises to 80,000 and annual income to 5,333.33 when other assumptions are held constant. Third, 25 years of 2.5 percent annual inflation yield a real-value factor of 1/(1.025)^25 = approximately 0.53939. These are ordinary mathematical identities computed on stated illustrative inputs.
Such calculations do not simulate mortality distributions, estimate annuity demand elasticity, price actual insurance contracts, or predict public budgets. A fuller model would require age-by-cohort survival probabilities, discount curves, benefit expenses, tax schedules, public-pension entitlements, withdrawal constraints, risk preferences, household composition, and correlated expenditure risks. Claims beyond the illustrative scope would require independently sourced empirical parameters and sensitivity analysis.
Sensitivity of annuity valuation to mortality and rates
Expected present value depends jointly on survival and discounting. A longevity improvement assumption raises future survival probabilities and can increase the actuarial cost of a fixed lifelong benefit. A rise in discount rates, under otherwise consistent measurement assumptions, decreases the present value of deferred nominal cash flows. The size of each response depends on duration and product design. Indexation, return guarantees, and joint-life provisions alter the relevant sensitivity.
The Mitchell and colleagues pricing study and Finkelstein–Poterba contract evidence support treating market prices, participant mortality and product characteristics as separate variables. Money’s worth is particularly sensitive to which mortality table corresponds to the population whose expected payouts are evaluated. Using a general-population table for a selected annuitant portfolio or using an annuitant table to evaluate the welfare of mandated short-lived participants would change the question being answered.
Behavioral and institutional selection
Voluntary purchasers are not a random sample of all retirees. Health information, financial literacy, access to advice, inherited wealth, marital status, tax position and preferences can influence decisions. When a mandate expands coverage, its marginal participants may differ materially from pre-reform voluntary buyers. As a result, the welfare consequences cannot be obtained by multiplying pre-reform mean voluntary annuity returns by the new number of mandatory purchasers.
The historical compulsory-versus-voluntary comparison provides evidence that the degree of adverse selection can differ across purchase regimes. Its data are historical and specific to the markets studied. An independent contemporary evaluation should test whether cohort-specific mortality, socioeconomic composition, risk premiums and contract choice have changed rather than transplanting its quantitative selection estimate into new legal settings.
How to identify a policy effect
A credible causal design might compare cohorts just above and below a rule’s age or savings threshold, provided the assignment mechanism cannot be strategically manipulated and other determinants change smoothly. Singapore’s balance-based inclusion threshold motivates such a design, but threshold-crossing behavior, eligibility exemptions, and retirement timing must be investigated before a discontinuity interpretation is credible. A differences-in-differences design may instead exploit changes in rules among comparable pension products or cohorts, but requires defensible parallel pre-policy trends.
South Africa’s two-pot regime could be analyzed using administrative records on component composition, observed withdrawals, later annuity choices and retirement outcomes, while accounting for grandfathering and changes in tax behavior. A British event study would distinguish 2011 abolition, 2015 flexibility, shifts in interest rates and subsequent regulatory interventions. Chilean analysis would separate pension benefit modality choices from concurrent changes to public transfers and contribution rules. These are prospective research designs, not completed causal estimates within this manuscript.
External validity, missing data, and uncertainty
Observed annuity take-up may understate lifetime insurance if a person also has a defined benefit pension, and may overstate it if the recorded product is a living annuity without a guaranteed minimum lifetime payment. Nominal spending surveys may miss imputed housing services, informal support or unrecorded medical costs. Insurer data may omit pension fund liabilities; pension regulator statistics may aggregate products with different guarantees. Death records may be delayed or incomplete, with resulting mortality selection bias.
A cross-national synthesis should prioritize definitions that can be audited at the contract and account level. Useful disaggregations include initial and real monthly income, account value retained after conversion, realized age at death, surviving household members, administrative fees, provider credit strength, and the presence of public income floors. An unobserved selection mechanism must be presented as an uncertainty rather than silently interpreted as evidence of success or failure.
Verification boundaries of this study
The source record includes official material available publicly as of the research cut-off, including legislative and administrator descriptions, institutional comparative reports, and publisher bibliographic records and abstracts for scholarly papers. The study does not reproduce individual statutes article by article for every jurisdiction; it does not claim access to private insurer quotations or confidential administrative data; it does not perform a quantitative meta-analysis or field experiment. The academic sources are used at the level directly supported by the publicly inspected journal record. These boundaries determine the strength of each conclusion.
Documentary consistency checks were conducted between public authorities and economic categories, including independent confirmation that the British 2011 reform and 2015 reform are distinct; that South Africa permits living as well as guaranteed annuities; that Singapore’s automatic entry has explicit conditions; and that the Australian and U.S. minimum-withdrawal rules do not themselves purchase survival insurance. The linked references preserve the original institutional provenance of these claims.
Integrated Analytical Synthesis and Scientific Agenda
The central finding: compulsion acts on different economic margins
The evidence establishes that mandatory annuitization is a family of legal interventions with different economic content. Singapore’s program automatically includes qualifying participants in national lifetime insurance. South Africa binds specified retirement contributions to an income purpose but allows different instruments, some exposing the member to continuing investment and longevity risk. Switzerland uses a statutory pension conversion framework while preserving defined lump-sum rights. Chile permits a regulated choice among payout modalities. The United Kingdom removed a general age-75 purchase requirement. Australia and the United States regulate minimum withdrawals without generally compelling insurance annuity purchase.
The distinction can be expressed through a set of measurable outcomes. A legally constrained share of capital records how much retirement wealth must enter specified payout channels. A lifelong guaranteed share records how much is converted into enforceable payments for the recipient’s remaining life. A real-income floor records how much purchasing power is protected across inflation and survival states. A liquidity share records accessible wealth after settlement. Finally, risk-bearing location identifies the party responsible for adverse investment, longevity and solvency outcomes. These are descriptive dimensions of established financial and legal arrangements, not a new proprietary theory or normative formula.
Different mandates can increase one indicator and leave another unchanged. Increasing the portion that must be used for a living annuity can raise required retirement-income allocation without raising the portion protected by insurer-guaranteed lifetime payments. Increasing a fixed nominal annuity purchase can reduce outliving risk while leaving real expenditure risk exposed to inflation. Permitting cash withdrawals can raise liquidity while increasing the need for future public protection if savings are depleted. The relevant policy question is which combination serves the specified social objective for the eligible population.
The welfare balance is heterogeneous
The 2005 annuity welfare analysis supports the value of substantial longevity pooling under many familiar conditions while rejecting a universal inference that full annuitization must be optimal under all incomplete-market environments. The British selection evidence shows that the pool created by compulsion differs from one generated through voluntary purchase. These conclusions are mutually consistent. Pooling can generate real insurance value while an indiscriminate mandate imposes higher opportunity costs on particular groups.
An individual with short expected survival, high immediate care expenditure, or a strong need to support dependents may place a different value on irreversible annuity conversion from an otherwise comparable individual with long expected survival and limited other lifetime income. A couple has different survivor-risk exposure from an individual living alone. Workers with interrupted contribution histories may enter retirement with too little capital for meaningful guaranteed benefits even when every eligible unit is annuitized. Regulatory design must therefore consider benefit adequacy, health heterogeneity, survivor rights and access to residual liquid savings together.
Fiscal and competitive effects
The expected fiscal effect of an annuitization mandate depends on existing guarantees, social assistance rules, taxation, investment and mortality risk distribution, and the way additional lifetime income changes later transfer claims. A policy could reduce some instances of very late-life account exhaustion while generating public costs through guarantees or social transfers for persons with inadequate retirement capital. Without linked longitudinal household and administrative data, the sign and scale of the net fiscal effect remain empirical questions.
Market design also matters. When purchase is mandatory, competition for insurance provision must be evaluated against transparent quotation, disclosure, distribution incentives, risk-based pricing and financial security. The Chilean pension quotation guidance and OECD pension-plan recommendation point to different institutional approaches to these objectives. Neither the existence of a quotation system nor automatic assignment alone establishes that consumers receive the best feasible lifetime value.
Research priorities
A substantive international research program should construct a legal-event dataset that identifies statutory clause, policy commencement, eligible cohort, account component, conversion fraction, exemptions, permitted payout instruments, and the corresponding definition of guarantee. The dataset should not assign a single undifferentiated “mandatory annuitization” indicator to a country-year when simultaneously operating schemes impose different rules.
It should then link administrative pension histories to contract and benefit outcomes, mortality or survival data, and household consumption where permitted under privacy and research governance. Evaluations should measure both average and tail outcomes, including extreme longevity, early death, bereavement, disability, household care expenditure and inflation scenarios. Source provenance must distinguish enacted law from supervisory interpretation and current operational practice.
Annuity market studies should separately measure premium money’s worth, insurance value to heterogeneously situated retirees, insurer financial resilience, and the distribution of the cost of guarantees. Mitchell and colleagues’ valuation approach and Finkelstein and Poterba’s contract selection analysis provide methodological antecedents for such work. Historical estimation must use jurisdiction- and period-specific mortality and discount curves, not generalized market averages assumed to be universal.
Finally, pension policy should be evaluated as a full lifecycle system. Saving adequacy, fund returns, retirement timing, pension access, public benefits, income conversion, and later-care contingencies interact. The 2026 OECD assessment is valuable precisely because it situates payout choices within the accumulation and institutional conditions that determine how much retirement wealth exists in the first place. A technically successful mandate applied to insufficient capital cannot, by itself, secure sufficient retirement income.
Conclusion
Mandatory annuitization rules govern the legal transition from retirement capital to future income. Their scientific analysis requires exact identification of the covered assets, applicable persons, commencement dates, permissible payout products, exceptions, and institutions bearing the economic risks. The comparative record demonstrates several independent legal architectures: automatic inclusion in pooled lifetime insurance, compulsory application of certain balances to retirement income, statutory occupational pension conversion, regulated choice of pension modality, and mandatory minimum account distributions. They produce distinct protections and obligations.
The United Kingdom’s 2011 abolition and 2015 expansion of withdrawal freedoms must be treated as separate events. Singapore CPF LIFE requires a cohort-, residency- and balance-specific analysis. Chile combines mandatory saving with legally regulated payout alternatives. South Africa’s retirement component constrains payout use while preserving a living-annuity alternative and defined exemptions. Swiss occupational pension conversion rules must be read together with the statutory asset base and lump-sum rights. Australian and U.S. withdrawal minima do not establish a compulsory life-insurance annuity.
The established actuarial and economic literature supports substantial value from pooling uncertain survival, while showing that health, family structure, available public benefits, liquidity needs, product design, pricing, and real-income exposure condition household welfare. The original contribution of this monograph is an integrated, source-specific comparison of these legal and economic margins and a documented agenda for empirical validation. It makes no claim to have produced a new annuitization theory or a causal estimate unavailable in the verified evidence.
An effective international evaluation will measure lifetime protection, real adequacy, retained liquidity, and distributional effects together, using observed legal and contract conditions. Scientific closure of the question of a particular mandate’s net welfare impact requires the national data and causal tests specified above. The documentary and conceptual conclusions established here stand independently of those future empirical investigations.
References
Organisation for Economic Co-operation and Development. (2026, November 4). OECD Pensions Outlook 2026: Are We Saving Enough for Retirement? OECD Publishing. https://www.oecd.org/en/publications/oecd-pensions-outlook-2026_a51aa775-en/full-report.html
Organisation for Economic Co-operation and Development. (2022, February 23). Recommendation of the Council for the Good Design of Defined Contribution Pension Plans (OECD/LEGAL/0467). https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0467
International Labour Organization. (2024 access; Convention adopted 1952). Frequently asked questions: Social Security (Minimum Standards) Convention, 1952 (No. 102). https://www.ilo.org/frequently-asked-questions-social-security-minimum-standards-convention
HM Revenue & Customs. (2026 access). Inheritance Tax Manual IHTM17021: Pensions—Types of pension scheme: Registered pension schemes. https://www.gov.uk/hmrc-internal-manuals/inheritance-tax-manual/ihtm17021
HM Treasury. (2010, July 15). Removing the requirement to annuitise by age 75: Consultation and government response. https://www.gov.uk/government/consultations/removing-the-requirement-to-annuitise-by-age-75
House of Commons Library. (2014). Pensions: Income drawdown. Standard Note SN00712. https://commonslibrary.parliament.uk/research-briefings/sn00712/
Central Provident Fund Board. (2026 access). CPF LIFE: Inclusion, benefits, options, and exemptions. https://www.cpf.gov.sg/member/retirement-income/monthly-payouts/cpf-life
Central Provident Fund Board. (2026 access). When do I have to choose my CPF LIFE plan? https://www.cpf.gov.sg/service/article/when-do-i-have-to-choose-my-cpf-life-plan
Ministry of Manpower, Singapore. (2026 access). How you can use your CPF: Monthly payouts and CPF LIFE. https://www.mom.gov.sg/employment-practices/central-provident-fund/how-you-can-use-your-cpf
Superintendencia de Pensiones, Chile. (2026 access). Pensión de vejez: Pension initiation and benefit modality choice. https://www.spensiones.cl/portal/institucional/594/w3-propertyvalue-9921.html
Superintendencia de Pensiones, Chile. (updated 2025). Compendio de Normas del Sistema de Pensiones, Chapter III: Retiro Programado. https://www.spensiones.cl/portal/compendio/596/w3-propertyvalue-3223.html
International Monetary Fund. (2026). Chile: Selected Issues. IMF Country Report 2026/170, discussion of pension design and the 2025 reform. https://www.elibrary.imf.org/view/journals/002/2026/170/article-A001-en.xml
Subsecretaría de Previsión Social, Chile. (2025). Superintendencia de Pensiones y CMF: Normas conjuntas sobre eliminación de ofertas externas de rentas vitalicias. https://previsionsocial.gob.cl/superintendencia-de-pensiones-y-cmf-emiten-norma-conjunta-que-regula-la-eliminacion-de-la-posibilidad-de-comercializar-ofertas-externas-de-rentas-vitalicias-otro-cambio-de-la-reforma-previsional/
South African Revenue Service. (2026). Guide to Complete the Lump Sum Tax Directive Application Forms: Retirement and two-pot annuitization provisions. https://www.sars.gov.za/guide-to-complete-the-lump-sum-tax-directive-application-forms/
Government of South Africa. (2024, July 21). President Cyril Ramaphosa signs Pension Funds Amendment Bill into law. https://www.gov.za/news/media-statements/president-cyril-ramaphosa-signs-pension-funds-amendment-bill-law-21-jul-2024
Swiss Social Insurance / AHV-IV Information Centre. (2026). Social security in Switzerland: Occupational benefit plan (PP). Federal overview valid January 1, 2026. https://www.ahv-iv.ch/p/890.e
Australian Taxation Office. (2026 access). Payments from super: Minimum annual payments for super income streams. https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/payments-from-super
Australian Taxation Office. (2025, December 1). Income streams: Information for APRA-regulated funds on paying benefits. https://www.ato.gov.au/tax-and-super-professionals/for-superannuation-professionals/apra-regulated-funds/paying-benefits/income-streams
Internal Revenue Service. (2026 access). Retirement plan and IRA required minimum distributions FAQs. https://www.irs.gov/retirement-plans/retirement-plan-and-ira-required-minimum-distributions-faqs
Davidoff, T., Brown, J. R., & Diamond, P. A. (2005). Annuities and individual welfare. American Economic Review, 95(5), 1573–1590. https://doi.org/10.1257/000282805775014281. https://www.aeaweb.org/articles?id=10.1257%2F000282805775014281
Finkelstein, A., & Poterba, J. (2002). Selection effects in the United Kingdom individual annuities market. Economic Journal, 112(476), 28–50. https://doi.org/10.1111/1468-0297.0j672. https://academic.oup.com/ej/article-abstract/112/476/28/5060949
Finkelstein, A., & Poterba, J. (2004). Adverse selection in insurance markets: Policyholder evidence from the U.K. annuity market. Journal of Political Economy, 112(1), 183–208. https://doi.org/10.1086/379936. https://www.journals.uchicago.edu/doi/10.1086/379936
Mitchell, O. S., Poterba, J. M., Warshawsky, M. J., & Brown, J. R. (1999). New evidence on the money’s worth of individual annuities. American Economic Review, 89(5), 1299–1318. https://doi.org/10.1257/aer.89.5.1299. https://www.aeaweb.org/articles?id=10.1257%2Faer.89.5.1299
Organisation for Economic Co-operation and Development. (2025). Pensions at a Glance 2025: Old-age income poverty. OECD Publishing. https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/old-age-income-poverty_3cd946d7.html
Organisation for Economic Co-operation and Development. (2025). Pensions at a Glance 2025: Architecture of national pension systems. OECD Publishing. https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en/full-report/architecture-of-national-pension-systems_2850d0fe.html