Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Collective Defined Contribution Benefit Adjustment and Risk Sharing

Abstract. Collective defined contribution (CDC) retirement arrangements transform contributions and collective investment performance into retirement benefits whose level can change through actuarially governed adjustment mechanisms. This research dossier examines how those mechanisms transmit investment, interest-rate, inflation, and longevity risks across current workers, deferred members, retirees, and later entrants. It reconstructs the legal and financial foundations of CDC pensions, formalizes the distinction between aggregate funding balance and cohort-specific value, and compares three related but legally distinct institutional settings: collective money purchase schemes in Great Britain; the Dutch solidarity and flexible contribution arrangements introduced under the Future Pensions Act; and shared-risk pension plans in New Brunswick, Canada. Original analytical work includes a transparent asset–liability accounting model, an illustrative proportional benefit recalibration, stress sensitivities, a cohort-allocation example, and an evaluation framework for adjustment governance. The central finding is that collective funding and formula-driven benefits can distribute financial uncertainty over time and among members, while the resulting allocation of value depends on the exact accrual, discounting, reserve, rebalancing, entry, and exit rules. Aggregate solvency, benefit stability, and intergenerational equity are related institutional objectives whose achievement requires separate evidence. The study interprets established scholarship and original legal materials; it introduces no new named theory, principle, or proprietary model. Its quantitative illustrations are synthetic and are not estimates of any existing pension plan.

Keywords: Collective defined contribution; CDC pensions; collective money purchase; pension benefit adjustment; targeted retirement income; pension funding; intergenerational risk sharing; longevity risk; stochastic asset–liability management; collective pension reserves; cohort equity; pension governance; defined contribution; shared-risk pension plan; target-benefit design; actuarial valuation; Netherlands Future Pensions Act; United Kingdom pension law; New Brunswick pensions; retirement income; actuarial communication.

Introduction

A pension arrangement transforms work-related contributions into purchasing power at retirement. A conventional individual defined contribution (DC) account makes the accumulated investment result visible to its holder but often leaves the conversion of assets into dependable lifetime income to a later decision. A defined benefit (DB) pension generally specifies the promised benefit and distributes deficits through a different set of sponsor, funding, regulatory, and insolvency arrangements. Collective defined contribution arrangements organize a third contractual possibility: plan assets and retirement income are managed collectively, while members participate in the financial consequences through rules that can increase or reduce benefits. The label does not determine the plan's legal status in every jurisdiction. Its operative rights are created by the governing statute, trust or pension documents, and applicable employment and tax law.

In Britain, the Pension Schemes Act 2021 makes valuation and adjustment part of the statutory collective money purchase framework. The Pensions Regulator's CDC supervisory code, in force from July 31, 2026, connects scheme authorization and continuing viability to actuarial calculations, systems, governance, and communication. The Netherlands has pursued a different reconfiguration of occupational pensions. De Nederlandsche Bank's collective reserve guidance describes intergenerational loss and gain sharing inside solidarity and, under certain conditions, flexible contribution contracts. New Brunswick's Pension Benefits Act permits contribution and benefit adjustments according to a shared-risk plan's funding policy. These systems illuminate common financial questions without implying identical pension promises.

The methodological importance of comparing them lies in the meaning of collective risk bearing. A pooled plan can absorb idiosyncratic longevity experience more predictably than a single retiree, and it can smooth financial shocks across benefit dates. It can also place a material share of a loss on members who neither selected the investment portfolio nor anticipated the particular adjustment. The plan's governing rules establish which cohorts carry those effects, when they become visible, and how future members are treated. The legal ability to adjust benefits is therefore part of the financial substance of the promise.

The central questions are how contributions and assets support target benefits, which events cause benefit uplifts or reductions, and whether a rule that balances total assets and projected liabilities also produces defensible results across cohorts. This monograph answers those questions through primary law, supervisory texts, official policy materials, and published economic and actuarial research, supported by explicitly hypothetical calculations. Its original contribution is an integrated analysis that puts accounting balance, economic risk incidence, rights, and governance into the same comparative discussion. No empirical causal estimate or plan-specific actuarial opinion is claimed.

Institutional History, Concepts, and Evidence

The economic object and its boundaries

A collective pension plan is a legal and financial mechanism in which contributions support retirement claims subject to jointly administered investment and risk-allocation decisions. Collective defined contribution is used here for arrangements in which the initial employer financing commitment can be specified in contribution terms and members' retirement benefits remain sensitive to the collective scheme's financial experience. The term must always be mapped to the law governing the actual arrangement. Some collective systems operate with individual pension capital and collective risk-sharing reserves; others express members' claims as target pensions that are adjusted through a funding calculation. A shared-risk target-benefit plan may also change contribution rates in ways that a narrowly defined CDC contract does not.

Five distinctions structure the comparison. A contribution rate specifies the transfer of wages or other income into the retirement fund. An accrued benefit is a rule-defined measure of service or contribution entitlement. A target pension is a projected or currently payable amount whose future path depends on the scheme's adjustment rules. A guarantee is an enforceable undertaking of a specified benefit or protection. A pension reserve is an asset allocation mechanism subject to its legally defined permitted uses. These terms refer to different elements of the pension contract and cannot be interchanged merely because a projection has been expressed in annual income.

The underlying financial flows are contributions, investment income, costs, and benefit payments. The OECD Pensions Outlook 2026 places these flows in the wider evolution of asset-backed retirement systems, reporting USD 78 trillion of OECD pension assets at the end of 2025. That aggregate is a measure of the scale of funded pensions, not evidence of a CDC-specific welfare gain. For an individual collective scheme, the relevant accounts remain its own assets, obligations, benefit formulas, member census, and policy restrictions.

The economic distinctions also differ from employer financial reporting. The IAS 19 classification framework determines whether a post-employment benefit arrangement is a defined contribution or defined benefit plan for a reporting entity from the economic substance of its obligations. The name CDC, standing alone, does not establish the employer's accounting outcome. Contingent or constructive support obligations, the presence of multiple employers, and enforceable scheme terms matter.

The evolution of collective pension promises

Occupational pensions historically combined contributions, investment management, retirement income, and institutional promises in varying proportions. Traditional DB plans concentrated a substantial financing responsibility on sponsors, subject to local funding regulation. Individual DC plans made contribution financing more predictable for employers, while leaving members exposed to accumulation returns, conversion rates, and longevity income decisions. CDC designs seek a collective income path in which target benefits can be supported by shared assets without requiring an unconditional employer guarantee of every future benefit amount.

The welfare proposition of intergenerational pooling has a precise theoretical pedigree. Cui, de Jong, and Ponds (2011), in intergenerational risk-sharing research, use an overlapping-generations model to examine funded arrangements with benefits and contributions that react to fund performance. Their reported welfare improvement depends on the specification of the sharing mechanism and the comparator. It is a conditional model result, not a universal result of institutional pooling.

Later work sharpens the distinction between welfare gains and fair value. Boelaars and Broeders (2019) examine fair pension risk allocation through the relation between discounting and horizon-dependent benefit adjustment. Their research identifies conditions for arbitrage-free allocation of market risk; it does not prescribe one universally optimal governance arrangement. Yi, Sanders, and Bégin (2022) investigate hidden intergenerational transfers and show, within their modeled arrangements, that entry into a pure CDC plan can produce implicit transfers even without a separately advertised stabilization reserve. The literature thus supports two connected propositions: pooling can improve the timing and distribution of risk, and the exact allocation requires explicit valuation.

Legal distinctions among three institutional settings

The British collective money purchase framework is organized around scheme authorization, trustee duties, scheme-specific actuarial rules, and ongoing supervision. The 2022 CDC Regulations implemented the initial single- and connected-employer framework on August 1, 2022. The 2025 extension regulations, made in December 2025 and commenced July 31, 2026, extended the regulatory system to unconnected multi-employer arrangements. The updated TPR code introduction explicitly addresses both scheme categories. Analysis of the British regime must specify the scheme category and operative regulatory date.

The Netherlands' Future Pensions Act, effective from 2023, is reshaping occupational pension accrual and payout structures around contribution arrangements with investment and risk-allocation rules. The Dutch government transition guidance identifies January 1, 2028 as the latest transition date and describes social-partner decisions concerning solidarity versus flexible contribution arrangements, conversion of existing rights, compensation, and reserves. This is a transition between legally defined Dutch pension contracts. Its individual capital allocation and reserve architecture cannot simply be labeled British-style collective money purchase.

New Brunswick's shared-risk structure was established by legislative reform in 2012. Under the provincial statutory scheme, a plan must maintain a funding policy, investment policy, defined risk-management goals, and governance procedures. Section 100.53 permits funding-policy-driven changes to contributions, base benefits, and ancillary benefits. The provincial supervisory explanation describes a 97.5% modeled probability that base benefits will not be reduced over twenty years and at least 75% of targeted ancillary benefit value paid on average over that period. These regulatory tests are conditional modeled targets, not an insurance guarantee against an adverse realized outcome.

The three legal forms therefore have different adjustment levers. Britain primarily recalibrates collective money purchase benefits under statutory valuation and scheme rules. The Dutch arrangements allocate financial outcomes among member capital, returns, and permitted reserves. New Brunswick additionally makes contribution changes an explicit legal element of shared-risk funding policy. A meaningful comparison is made at the level of legally authorized cash-flow and risk-allocation functions.

The empirical and theoretical research record

The peer-reviewed literature offers mechanisms and conditional results rather than a universal ranking of institutional forms. Wang, Wang, and Zhang (2023) study stochastic CDC investment control with stochastic interest rates, portfolio choice, and benefit adjustment, using a continuous-time model. Chen, Kanagawa, and Zhang (2023) analyze intergenerational allocation with optimization in a fully funded collective DC setting with overlapping generations and realistic investment constraints. Both papers advance the analysis of feasible risk-sharing design but rely on stylized environments and stated behavioral objectives.

The evidence base has three levels. The strongest institutional evidence for what a plan is authorized to do comes from law and operative supervisory rules. Statements about the possible economic effects of a mechanism are supported by transparent actuarial accounting and formal models. Generalizations about actual welfare, benefit volatility, or realized intergenerational transfers require plan-specific time series, member-level exposures, contractual terms, and counterfactual identification. Published article abstracts support a summary of the stated research question and reported conclusions, but they cannot establish the validity of every derivation in an inaccessible full paper. This dossier accordingly distinguishes documentary facts, reconstructed mathematical identities, modeled scenarios, and empirically unmeasured outcomes.

Collective Funding Rules

Contributions, assets, and benefit units

Let time be indexed by t, and let A_t denote the market-consistent or prescribed actuarial value of assets at the valuation date. Let C_t denote contributions entering during the period, P_t benefits paid, E_t expenses, and R_t the asset return net of any costs excluded from E_t. A simple discrete-time asset recursion is

A_(t+1) = A_t × (1 + R_t) + C_t - P_t - E_t.

This is a flow identity, not a pension valuation convention. Real schemes determine precisely when contributions and cash flows occur; a daily-weighted or cash-flow-adjusted calculation may be required to represent realized returns correctly. Receivables, illiquid investments, delayed contributions, and the permitted valuation basis also affect A_t. Employer financing responsibilities depend on the governing documents, as emphasized by the distinction in IAS 19 employee benefits.

Let member i have a projected stream of target payments b_(i,s) at future time s. Let p_(i,s) denote the probability-weighted survival or beneficiary entitlement factor, D_(t,s) the discount factor, and a_(i,s) other contractual conditionality. A stylized actuarial present value is

L_t = sum over i and s>t of [b_(i,s) × p_(i,s) × a_(i,s) × D_(t,s)].

The valuation basis must specify nominal versus real payments; survival and dependent-beneficiary conventions; discount rates; benefit escalation; options; fees; and whether the payments are fully earned, conditional, or projected. An active member's target relating to future service should not automatically be recognized in the same way as a pensioner beneficiary's already accrued rights. A plan may therefore maintain separate valuations for accrued benefits, future service, solvency purposes, and contribution adequacy.

The funding ratio F_t = A_t/L_t is informative only when A_t and L_t are measured on mutually consistent bases and L_t is positive. A ratio above one can coexist with poor liquidity, a concentrated risk portfolio, or an inequitable distribution among cohorts. A ratio below one need not have the same legal consequence in every jurisdiction. The funding policy and actuarial basis determine whether adjustment occurs immediately, over time, after exhausting a buffer, or in combination with contribution changes.

The actuarial funding identity

For clarity, first suppose every target payment scales by the same positive factor k, while survival factors and discount rates stay fixed. Then the value of the adjusted targets is kL_t. Setting this value equal to available assets gives k = A_t/L_t. An underfunded plan with F_t = 0.90 can, under these restrictive assumptions, reduce every modeled payment by 10% and restore equality between the valued payment stream and assets. A surplus ratio of 1.10 would permit an analogous 10% uplift if the governing rules called for full immediate distribution. Neither calculation asserts that any actual plan uses this proportional rule.

The derivation reveals the legal and economic issues that a simple funding ratio conceals. Some benefits may have priority or be excluded from adjustment; current payments can differ from deferred accrual; conditional increases may be treated separately; benefits may be capped; smoothing may postpone part of a change; and longevity risk can affect cohorts differently. In each case, the actual adjustment must solve the plan's valuation equation under its operative benefit and funding rules, rather than multiply the entire liability by a single ratio.

The British viability-certificate guidance calls for annual actuarial valuation to determine benefit adjustments needed to align the values of assets and liabilities. Its rule-based framework is therefore analytically close to this asset–liability logic, although the hypothetical proportional multiplier is deliberately simpler than real scheme valuations. A statutory or scheme-governed benefit change is a financial redistribution across payment dates, not merely an accounting presentation.

Discount rates, survival, and payout duration

A reduction in discount rates generally raises the present value of a fixed stream of positive future nominal payments, all else equal. A positive longevity revision also tends to increase expected life-contingent payment duration. Investment losses directly reduce assets; changes to inflation expectations alter projected index-linked payments; and cohort-specific mortality revisions affect groups unequally. A single collective adjustment rate can therefore transmit fundamentally different underlying shocks.

Duration is a useful first approximation. For a small parallel discount-rate change Δy, the percentage change in a nominal liability can be approximated by -D_mod × Δy, where D_mod is modified duration on the chosen valuation basis. Convexity, nonparallel yield-curve movements, optional benefits, expected indexation and correlated longevity shifts can make this approximation inaccurate. A collective scheme with long-duration nominal pension obligations and comparatively short-duration assets will be sensitive to falling yields. A hedge may reduce this sensitivity while changing the allocation of investment upside and exposure to liquidity requirements.

Longevity pooling is also conditional. Idiosyncratic deaths diversify as group size and independence assumptions improve, while a persistent cohort-wide improvement in survival requires the scheme to finance longer payment streams. Members may value lifetime income protection even when expected benefit amounts are adjustable. The actuarial significance of that distinction is the difference between uncertainty about the timing of death and uncertainty about the amount of payment while alive.

Reserve and smoothing design

A reserve separates the timing of recognized financial gains and losses from the timing of payouts. Suppose the plan earmarks part of its total assets in a stabilization reserve. When investment returns are strong, an admissible portion of the gain may be retained; in a weak year, a transfer from that reserve may soften a reduction in ordinary pension capital or target payments. Because the reserve belongs within total scheme assets, moving money from it to another benefit account does not create new net wealth. It changes which members and payment dates receive the existing wealth, while preserving or consuming future stabilization capacity.

The DNB reserve factsheet describes the Dutch solidarity reserve as a collective asset reserve embedded in the solidarity contribution arrangement. Its stated permissible purposes include intergenerational sharing and moderating pension fluctuations. The text describes constraints on funding from contributions and positive excess returns, rules for distributions, long-term consistency, and cohort-level assessments of benefits and burdens. Those limits are jurisdiction-specific and depend on the particular reserve and contract type.

Smoothing without a separately identified reserve similarly allocates shocks across time. If a plan amortizes an adverse return over several future years, current beneficiaries experience a smaller initial adjustment, while later payments or incoming cohorts may bear residual risk. The resulting intergenerational transfer can be favorable or unfavorable depending on market performance, participant turnover, and the valuation rule. The fair-pensions analysis demonstrates why the discount rate and adjustment process must be analyzed together when judging economic fairness.

Entry, exit, transfers, and plan maturity

The size and age composition of a collective pension population change the economic meaning of its funding rules. Early in a plan's life, contributions may be large relative to pension payments; at maturity, benefit outflows can become dominant. The OECD's pension flow analysis documents how contributions, investment income, benefits and plan maturity jointly drive asset accumulation. It does not provide a single CDC-specific maturity threshold.

New members introduce additional contributions and future claims. If the value of accrued benefits granted to them differs from their economic contribution under the initial conditions, an implicit transfer may occur. Members leaving with a transfer value can carry away more or less than their share of collective economic value depending on the transfer basis and how future adjustment risk is priced. Retirees may be more sensitive to immediate reductions than active contributors because a given nominal cut changes current consumption and their ability to replace lost retirement income can be limited.

An assessment of collective funding must therefore examine four distinct margins: the consistency of aggregate present values; the fairness of accrual and entry pricing; the financing of promised or targeted benefit changes; and the portability of benefit rights at exit. The measures are connected, but passing one does not logically establish the others.

Benefit Uplift and Reduction Triggers

Annual valuation and the adjustment decision

An adjustment decision begins with a valuation date, a defined benefit population, and an agreed projection basis. The administrator must establish the assets available for benefits, the present value of the particular benefits covered by the adjustment calculation, and the actuarial method that translates any difference into an amended pension path. These are analytically separate decisions. Choosing a valuation basis affects the reported surplus or deficit; choosing an adjustment rule determines the immediate distribution of that valuation result among participants.

In Great Britain the statutory foundation is explicit. Section 20 of the Pension Schemes Act 2021 requires scheme trustees to obtain an actuarial valuation showing available assets, required amount, and whether—and by how much—benefits require adjustment. Section 22 addresses the subsequent adjustment of benefits. The 2022 collective scheme regulations require the actuarial methodology stated in scheme rules to determine benefit adjustment and provide for its application across members. Legislative amendments and authorization category must be read together when applying those provisions to an actual 2026 plan.

The Pensions Regulator's viability certificate requirements make the practical distinction between a financially credible design and an actuarial projection that merely balances at a chosen date. Scheme design assessments consider whether the proposed methodology can provide target benefits, whether adjustments can be implemented, and whether the financial assumptions are justified. Valuation accuracy and administration accuracy jointly determine whether the designed adjustment reaches members' accounts and pay slips as intended.

The sign of an adjustment needs an explicit reference point. An increase relative to the previous nominal pension may still be a decline in real purchasing power. A nominal pension maintained at its previous level can represent a real cut when prices rise. A benefit forecast can show an expected increase next year that is smaller than the increase projected last year; this is a downward revision to expectations even if the cash pension actually grows. Public communication must distinguish the current payment, the projected target, the adjustment from last year's target, and purchasing-power performance.

Immediate proportional adjustment: a transparent illustration

Consider a hypothetical plan with actuarially valued target payments of 100 monetary units and available assets of 90 on the same basis. The starting funding ratio is 0.90. Suppose every payment is equally adjustable and the regulator and scheme rules permit immediate, uniform proportional adjustment. The benefit multiplier is 90/100 = 0.90. Each valued target is reduced by 10%, restoring the simplified measured equality. This calculation assumes a fixed valuation basis, no omitted benefit priorities, no transaction costs, and no further stochastic developments.

Now separate causes of the deficit. If the assets remain 90 while a change in market discount rates raises modeled liabilities from 100 to 106, the valuation ratio becomes 90/106, approximately 0.8491. The hypothetical reduction needed for immediate equality increases to approximately 15.09%. If the same assets face mortality changes that raise liability value only to 103, the corresponding multiplier is approximately 0.8738, a reduction of 12.62%. In a different scenario with assets 96 and liabilities 100, the deficit is 4%. The numbers are counterfactual exercises, not estimates for any British, Dutch, or Canadian fund.

Illustrative case Assets Target liability value Funding ratio Immediate uniform benefit change
Starting hypothetical deficit 90 100 0.9000 −10.00%
Same assets, lower discount basis 90 106 0.8491 −15.09%
Same assets, higher longevity value 90 103 0.8738 −12.62%
Smaller deficit 96 100 0.9600 −4.00%
Hypothetical surplus 110 100 1.1000 +10.00%

The proportional calculation is deliberately devoid of legal prioritization. In New Brunswick, for example, base benefits and ancillary benefits have distinct status; some funding policy sequences may preserve base amounts while reducing ancillary improvements first. Dutch reserves can buffer the pensions of eligible cohorts while other allocations follow an age-dependent financial-return policy. Britain's collective scheme methodology may determine an adjustment to annual growth or projected future payments rather than an isolated one-off reduction to every scheduled cash payment. In each jurisdiction, actual law and scheme rules replace the table's simplifying assumptions.

A level adjustment versus a change in future annual indexation

A subtle source of misinterpretation concerns the mathematical form of an adjustment. Let a member's projected annual pension be B_s for each future date s and suppose a plan modifies the currently payable level by multiplier k. The value of an otherwise fixed projection becomes k times its previous value. But if the scheme instead reduces a future annual growth factor by an amount δ, the pension in period s becomes B_s × (1+δ)^(s−t), relative to its unadjusted path. Present values then change differently for different durations, and beneficiaries with earlier and later payment dates may bear different economic fractions of the shock even when every member is subject to the same stated annual rate change.

The same numerical shortfall does not map uniquely to a single annual percentage-point adjustment. A one-off 10% target reduction and a 10% reduction in expected payments in every future year are different from a 10-percentage-point reduction in an annual escalation rate. When communication uses the word “cut,” it should identify which contractual quantity changes: a payment already being made, an initial pension at retirement, a future growth assumption, an ancillary increase, or an actuarially projected present value.

For a purely illustrative five-year accumulation of a target multiplier of 0.90, the constant annual geometric factor is (0.90)^(1/5) ≈ 0.97915, equivalent to approximately −2.085% each year relative to the assumed path. Conversely, a cumulative uplift of 10% corresponds to (1.10)^(1/5) ≈ 1.01924, or approximately +1.924% annually. Neither is a method for setting an actual CDC adjustment, because real pension obligations are paid across heterogeneous remaining lifetimes and their value is calculated with stochastic mortality, discounting, and scheme-specific formulas. The example simply exposes the difference between cumulative level change and repeated annual rate change.

Smoothing, stabilization, and the temporal ownership of shocks

A funding deficit can be recognized immediately, over a specified term, or through a rule that draws upon reserves or temporary future contributions. Financial smoothing changes the date at which the same underlying loss affects benefit amounts. It also influences the value of an option to enter, exit, or retire before a loss is fully recognized. In a closed group with fixed participants, deferral moves risk across the remaining payment schedule. In an open group, it may move risk to future entrants who were absent when the underlying investment loss occurred.

Suppose an illustrative reserve of 6 forms part of a fund's assets, with ordinary benefit capital valued at 94 and total assets of 100. A financial shock reduces total assets by 8 to 92. If a rule transfers 5 from the reserve to ordinary benefit capital, if the full investment loss is assigned to ordinary benefit capital before the reserve transfer, the book allocation becomes 91 ordinary capital and 1 reserve, while total wealth remains 92. It would be erroneous to add the transfer to the total assets as a new economic gain. The simple accounting illustrates the difference between balancing a designated benefit account and providing total solvency. All reserve transfers require reconciliation across sources and destinations.

Under the Netherlands' solidarity and risk-sharing reserve rules, the reserve generally cannot be negative and has an ordinary maximum of 15% of the total relevant pension capital including the reserve. Certain transitional reserves exceeding that ceiling are subject to a special adjustment timetable. Financing from contributions and positive excess returns is subject to stated ceilings. These rules constrain both the size of a buffer and the speed at which it can be replenished. Their effect depends on the fund's actual reserve operating policy rather than the mere presence of the reserve.

The DNB's inflation-risk reserve guidance makes an additional distinction: a permissible reserve mechanism for unexpected inflation must define in advance, using objectively justifiable information, the inflation threshold above which the shock is covered. Expected inflation and unexpected inflation are different actuarial inputs. A rule that covers specified surprise inflation is not automatically a commitment to maintain the entire projected real value of pension payments.

Investment and interest-rate shocks

Consider two members with equal initial accrued pension value but different time until payments. A short-duration retiree's value depends comparatively more on near-term payments; a young contributor's value depends heavily on later cash flows. When discount rates fall, liabilities of long duration tend to rise more. A uniform nominal percentage adjustment may therefore impose a different market-value cost across cohorts. Conversely, a hedge designed to stabilize one generation's liability sensitivity can shift return exposure to another group, depending on the economic rights assigned to each.

Assume a hypothetical market-consistent benefit liability has modified duration 12. A parallel yield reduction of 0.5 percentage point implies a first-order liability change of approximately +6%, holding all other assumptions fixed. For assets with duration 5, the approximate asset valuation increase would be +2.5% if the asset portfolio were a comparable fixed-income instrument. This mismatch lowers funding from the liability perspective, even when both sides' prices rise. The example illustrates the mechanism of interest-rate exposure. Convexity and the actual mix of growth assets, cash flows, and derivative hedges can alter the outcome.

A collective retirement scheme can seek a stable expected income path using asset liability matching, maturity-specific hedges, diversified investment and buffer policies. None of those instruments abolishes the financial cost of income security. An asset portfolio with more growth risk may improve expected long-run return and expected target pensions, yet also raises the probability and magnitude of adverse adjustments. The relevant decision is which return distribution and member-level exposure are authorized, disclosed, and governed.

Longevity shocks and correlation

Pooling member deaths reduces the statistical variability of mortality experience when independent exposures are sufficiently numerous. If many members survive longer than expected because of a sustained population-wide mortality improvement, the fund must finance additional life-contingent payments. This systematic longevity risk is shared across a collective plan only through a specified benefit, reserve, and contribution framework.

A 3% aggregate increase in the present value of target benefits, holding assets constant at 100, would reduce the illustrative funding ratio from 1.00 to 100/103 ≈ 0.9709. It would not imply a universal 2.91% cash pension reduction, because the scheme may differentiate between already payable income, future target escalation, participant rights, and risk-sharing capacity. For groups with markedly different mortality trends, the distribution of an adjustment is also a question of pension design and cohort allocation, not simply actuarial average experience.

Regulatory triggers and documented order of actions

The United Kingdom's supervisory approach combines the annual actuarial valuation with authorization, governance, communications and supervisory controls. The CDC systems-and-processes guidance emphasizes operational capacity to administer benefit calculation and change. The member communications code requires communications that can convey reductions as well as increases. The financial computation has therefore to survive translation into reliable member records.

New Brunswick requires the shared-risk plan's funding policy to specify an order for remedial and surplus-utilization actions. Under section 100.53 of its Pension Benefits Act, the administrator may increase, reduce, or suspend contributions and adjust base and ancillary benefits within that policy. The Shared Risk Plans Regulation specifies risk management goals and associated testing. A nominally similar funding shortfall can accordingly produce a change in contributions, suspension of contingent indexing, reduction of an ancillary benefit, reduction of base benefits, or a combination, depending on the binding policy and stress test.

The New Brunswick official consumer explanation characterizes the statutory risk tests as at least a 97.5% modeled probability that past base benefits are not reduced over a twenty-year horizon and a secondary target for ancillary payments averaging at least 75% of their targeted value. The actuarial probabilities concern modeled scenarios under selected assumptions. They are statements about plan-risk design, not evidence that the probability of an actual future cut is known with physical certainty or that a reduction cannot occur.

Cohort Equity and Governance

The distribution of benefits and the incidence of risk

Cohort equity concerns who receives pension value, who contributes it, which financial risks they bear, and how those outcomes change when new members enter or existing members exit. Retirement cohorts differ in age, accrued rights, expected remaining payment duration, future earning capacity, contribution commitments, and alternative retirement resources. A uniform percentage change in a published target has different economic effects depending on these dimensions.

A valuation of the cohort's economic position requires both its expected payments and the market value of its contingent risk exposure. Let V_g denote an actuarially and financially consistent value of the benefit stream allocated to cohort g, including the relevant benefit adjustment policy, and let K_g denote the economic value of past contributions and other transfers allocated to that cohort. A net position N_g = V_g − K_g is a useful accounting description only after all asset ownership, earlier distributions, reserve allocations, risk premia, and discount bases have been defined. A positive N_g is not in itself proof of unlawful redistribution, and a zero N_g is not by itself proof that the arrangement is distributively just.

The analysis in Boelaars and Broeders' Fair Pensions shows why valuation of market-risk-bearing pensions has to integrate the discount rate process and the benefit adjustment mechanism. A mechanical comparison of expected cash payments across generations overlooks the timing and price of risk. The possibility of implicit transfers in collective arrangements without an unconditional sponsor guarantee is investigated by Yi, Sanders, and Bégin through their collective pension redistribution analysis.

A cohort accounting illustration

Consider a hypothetical two-cohort plan with total assets of 100 and target liabilities valued at 100, consisting of 50 for an older cohort and 50 for a younger cohort. If aggregate assets fall to 90 and every target benefit is immediately scaled by 0.90, the revised liability values are 45 and 45. The allocation is symmetric as a percentage of the initial actuarial values, but the present-value losses are also each 5. These equal losses can have unequal welfare effects because the older group is closer to retirement consumption and may have less time to change saving.

Now suppose the plan's rule instead preserves the older group's modeled value at 50 and assigns the remaining 40 of asset backing to the younger group. The valuation still balances at 90; the allocation of the entire loss to the younger cohort is economically different. Or suppose the plan assigns 48 to the older and 42 to the younger cohort, reflecting a distinct adopted intergenerational policy. Aggregate funding equality provides no information about the normative, legal, or contractual acceptability of any allocation. Those issues depend on member rights and precommitted rules.

Hypothetical allocation after the same loss Older cohort Younger cohort Total backed value
Before shock 50 50 100
Uniform 10% reduction 45 45 90
Older cohort fully protected 50 40 90
Partial older-cohort protection 48 42 90

The table isolates one transfer dimension; it does not model market-based differences in duration or the price of risk. Under a fair-value analysis, an older cohort receiving greater current protection might also sacrifice a future gain or pay for the risk transfer through an assigned premium, reserve share, or altered claim. A complete empirical investigation would measure all of these flows rather than infer redistribution from the nominal payment table alone.

Entry, vesting, portability, and retirement timing

The timing of entry can alter the economic exposure of identical contribution amounts. If new members contribute immediately after a loss but receive rights whose initial value is based on an unadjusted target, they may subsidize existing participants. If they enter after a surplus and receive rights priced less favorably than the fund's economic position permits, the converse may occur. The direction is conditional on the contract's rule for opening balances and future claims, not on the mere fact of collective membership.

Retirement timing introduces a different problem. The conversion of accrued target capital or benefit units into a pension may use a mortality basis and a valuation date that differs from the basis used for active members. When financial returns and discount rates change, members retiring just before and just after an adjustment date can experience different initial incomes. Administrative rules on cut-off times, member record corrections, and delayed payment may compound this effect. The fairness assessment should follow legally relevant cohorts—including retirees, deferred beneficiaries, survivors, and members transferring out—not simply broad age bands.

Portability deserves independent attention. A transfer value may be computed from individual pension capital, collective entitlement value, statutory cash equivalent, or other scheme-defined basis. The amount paid to an exiting member changes the remaining asset pool and the value of adjustments borne by those who stay. A policy that treats all exiting participants consistently by formula may still generate a material economic effect on the remaining fund when market conditions change rapidly. Entry and exit should therefore be modeled as explicit events in the asset–liability cash-flow schedule.

Governance of investment and benefit decisions

An effective CDC governance system has to connect actuarial assessment, investment decisions, rule enforcement, member communications and data integrity. The trustee or governing body determines objectives and oversees compliance with applicable law; actuarial specialists calculate liabilities, adjusted benefits and risk projections under adopted methodology; investment specialists manage a portfolio within authorized limits; administrators correctly apply the resulting pension calculations; and supervisors assess compliance with authorization and continuing duties. These functions must be accountable even when they are performed by different vendors.

The OECD defined contribution recommendation situates retirement savings design within governance, risk management and communication principles. In European occupational pension institutions, the IORP II Directive establishes a prudential and governance framework for institutions within its scope, with national implementation and scheme-specific applicability. An institution's status under IORP II, as an insurer, or under distinct occupational pension legislation must be identified before importing any capital, conduct, or reporting requirements.

A governance record should establish which source documents controlled each adjustment, what valuation data and assumptions were used, how the calculation was reviewed independently, which trustee decision was taken, and how the adjustment was entered in payment systems. This audit trail is a practical safeguard against the combination of a correct actuarial method and an incorrect administrative result. It also permits later reconstruction of the impact on individual cohorts when pension outcomes become disputed.

Benefit adjustments and member comprehension

Members reasonably need to understand whether their retirement pension is a guaranteed amount, a target, or a projection subject to annual change. The central content of a comprehensible statement is the amount currently payable, the amount projected under stated assumptions, the circumstances in which each can change, and which party bears the residual investment and longevity risks.

The TPR member communications standards recognize both upward and downward adjustment. New Brunswick's regulatory disclosure requirements expressly call for plain-language statements explaining that benefits may be reduced, along with the funding and transfer-value information needed to interpret the plan. This disclosure principle can be compared across jurisdictions, but the exact legal disclosure obligations differ.

A projection ought to separate expected benefit outcomes from downside stress scenarios, distinguish nominal income from inflation-adjusted purchasing power, and explain reserve policies without describing a conditional buffer as a guarantee. Ex ante explanation matters because an adverse adjustment is most likely to test member understanding precisely when household financing is constrained. A reported average pension improvement does not substitute for disclosure of the distribution of possible individual outcomes.

Protecting governance against selection and conflicts

Risk sharing can create incentives for sponsors, administrators and members that require active supervision. A sponsor may prefer arrangements with bounded contribution costs, while members care about benefit volatility and the treatment of accumulated rights. A plan's selected discount basis could make target pensions appear more affordable even when risk exposure has not changed. A governing board may be tempted to defer visible cuts, thereby transferring financial stress to later participants. A retiree or near-retiree may rationally seek an advantageous exit if transfer values are insensitive to the remaining plan's risk.

Conflict management should therefore include documentation of valuation basis selection, conflicts-of-interest registers, transparent rules for reserves and termination values, assessment of decisions across cohorts, and review of assumptions that materially affect the apparent funding condition. Requiring disclosed principles before a shock occurs reduces the opportunity for opportunistic reallocation afterward.

Relationship to insurers and accounting obligations

A collective occupational retirement plan and an insurance annuity may both make lifetime payments, but their liability counterparties are potentially different. In an insurer-issued annuity, the insurance contract defines the insurer's obligation and the policyholder's rights, subject to insurance valuation and solvency regulation. In a CDC or shared-risk occupational plan, the right to a target benefit and its adjustment is typically governed by a trust or plan statute, potentially without the same insurer-backed guarantee. The plan may purchase insurance for selected liabilities, but that transaction changes the liability boundary and should be analyzed as a distinct risk transfer.

The accounting question is similarly substantive. Under IAS 19 Employee Benefits, a sponsoring employer's classification depends on the nature of its obligations and the applicable accounting rules. A bounded contribution formula does not by itself establish that the employer can never acquire a constructive obligation. Analysts should review plan terms, sponsor support, amendments and actual practices rather than assign an accounting designation from the CDC label.

The solvency consequences can be separated into three accounts: the funded occupational plan's ability to deliver targeted pensions, the participating employer's potential financial obligations, and any insurer's contracted liability and solvency capital. Conflating these accounts would misrepresent which balance sheet absorbs a particular loss.

Methods Validation and Source Criticism

Research design and evidence classification

The research uses documentary legal comparison, actuarial accounting, and critical synthesis of economic and pension research. The object is a mechanism of collectively financed retirement income adjustment. The observational units in legal analysis are enactments, regulations, supervisory standards, and institutional rules. The analytical units in the financial reconstruction are asset and liability flows, pension benefit claims, and cohorts whose rights or outcomes depend on the adjustment method. These units are not interchangeable: an actuarial identity describes possible financial equilibrium, whereas a statutory rule determines how an actual plan must allocate the consequences.

The legal investigation uses original enacted or consolidated texts wherever available. British provisions are read in their post-2025-amendment context and paired with the regulator's July 2026 supervisory code. The Dutch comparison begins with the legal transition under the Future Pensions Act and then reads the DNB's operative explanations of permitted reserves. New Brunswick is examined through the provincial Pension Benefits Act, the Shared Risk Plans Regulation, and the financial supervisor's explanatory materials. Economic propositions are compared against peer-reviewed pension research and public working papers. A statutory text and an academic article are treated as different kinds of evidence: the former establishes a legal provision as of its applicable date; the latter studies potential or observed effects under its specified assumptions.

The source hierarchy is therefore matched to the claim. For a proposition about the legal power to reduce base benefits, the relevant statutory section has priority. For an explanation of what an administrator must communicate, the binding disclosure rule and regulator's current code establish requirements. For a claim that a funding ratio guarantees welfare improvement, neither text is sufficient; one needs a welfare model, stated individual preferences, a feasible comparator, and empirical identification if the claim concerns realized outcomes. The OECD's 2026 pensions assessment provides authoritative contextual information about the wider funded retirement system, but the OECD aggregate is not a sample of CDC plans.

Source accessibility, authorship, and limits of verification

Original legislative and supervisory documents were inspected at the level of accessible official web text. Some cited academic studies were accessible through original journal abstracts, repository descriptions, or working-paper summaries rather than verified full articles. Their use is restricted to identifying the study design, subject, publication details, and conclusions explicitly described in the accessible material. The monograph does not infer unreported estimates, reproduce inaccessible derivations, or extend an author's conclusions beyond the stated model.

Research evidence about intergenerational fairness illustrates this discipline. Cui, de Jong, and Ponds' overlapping-generations pension study reports welfare gains for well-structured funded sharing compared with an individual benchmark in a formal model. This supports a statement of possibility and its assumptions. It does not establish that every actual pension scheme improves members' welfare or that members across all cohorts gain. The Fair Pensions working paper studies arbitrage-free allocation of market risk under specified adjustment and discount-rate mechanisms. The analysis is especially relevant to valuation consistency, not a measurement of any one scheme's realized benefit record.

The Dutch DNB web guidance is regulatory explanation of Dutch rules rather than a complete reproduction of an individual pension contract. The New Brunswick supervisor's 97.5% figure is an objective for model-tested funding security, while the real probability of future benefit reductions depends on model calibration and realized economic experience. The British code describes supervisory expectations and statutory context; actual scheme provisions still have to be obtained from scheme rules, valuation statements, and any relevant formal supervisory decisions.

The study records an evidence cutoff of November 14, 2026. Later amendments, actuarial standard revisions, scheme conversions, or new pension-plan data may change a jurisdictional conclusion. Publication dates, effective dates, and current guidance versions are distinct metadata and should be retained in any future update.

Limitations of the synthetic financial illustrations

The calculations in this dossier are algebraic examples, not calibrated actuarial projections. Their values are deliberately normalized to 100 or similar round amounts. They assume known opening assets and a consistent target liability basis, and then vary specified inputs one at a time. The assumptions omit stochastic employment, wage growth, contribution defaults, changes in member numbers, behavior at retirement, transfer value options, taxes, asset transaction costs, market liquidity constraints, and the correlation of pension risks with members' employment and household finances.

Even the simple funding equation contains an important boundary. The formula A/L can indicate how much a uniform adjustment of modeled liabilities would be needed under a linear benefit rule. It becomes inappropriate when projected benefits are nonlinear, when contingent increases have asymmetric conditions, when different claims have priority, or when the discount basis changes as adjustments are applied. In those cases, the adjustment should be solved iteratively or by a full actuarial simulation from the actual plan terms.

The discount-rate illustration uses a first-order duration approximation. A one-time 50-basis-point movement can change both asset prices and expected returns, and different maturities may move independently. A realistic sensitivity study requires term-structure stresses and an asset-liability matching analysis. Mortality shocks require exposure-weighted cohort projection, and fiscal and social transfers require a broader household model to understand welfare effects. None of these elements is inferred from the small illustrative tables.

Risks of comparing legally different schemes

The principal comparative risk is false equivalence. British collective money purchase, Dutch contribution arrangements with collective reserves, and New Brunswick shared-risk target benefits are related by risk-sharing mechanisms, while differences in individual capital accounting, benefit guarantees, employer obligations, ancillary entitlements and contribution levers are legally material. In one system benefits may be adjusted through a common future target escalation rate; in another collective reserve transfers change the allocation of individual capital; in a third the funding policy may prioritize higher contributions or contingent indexing adjustments. The measured rate and timing of a member's benefit change are therefore not automatically comparable.

A second risk is a misleading time comparison. The British 2022 framework and the July 2026 extension operate over different time periods and scheme categories. The Netherlands is in a transition process scheduled to conclude by January 2028. New Brunswick arrangements have plan-specific conversion histories and funding policies. A time-series study spanning changes in statutory rights must identify which version of the law, actuarial basis, and benefit formula applied in each observation.

A third risk is survivor and selection bias. Long-lived retirees receive more payments and remain in an annuity population longer; members may transfer or retire at economically favorable dates; and employers joining a collective arrangement may differ from those remaining in conventional DB or individual DC plans. Comparing average current pension payments across plans without accounting for age, salary, contribution histories, selection, and the present value of accrued rights can exaggerate or understate the value of pooling.

Validation of numerical and legal inferences

Internal numerical validation consists of recalculating asset–liability identities, confirming that the rows of each allocation table sum to total modeled assets, checking percentage adjustments against their denominators, and identifying the assumptions necessary for the formula. For the immediate adjustment example, 90/106 = 0.8490566 and 90/103 = 0.8737864; the corresponding reductions are 15.09434% and 12.62136%. The five-year annualization gives a growth factor of approximately 0.97915 for a cumulative 10% reduction and 1.01924 for a cumulative 10% increase. The apparent precision of these figures represents arithmetic only; no confidence intervals attach to them because they are synthetic inputs rather than sampled observations.

Legal validation requires four fields for each proposition: jurisdiction, operative legal source, version or effective date, and the specific action it permits or requires. For example, New Brunswick section 100.53 is a source for a plan administrator's policy-constrained ability to change base and ancillary benefits or contributions; it is not a legal source for a British trustee's authority. The Dutch reserve maximum belongs to the specified Netherlands statutory regime; it is not a generic cap for CDC pensions in every country.

The publication's scientific originality lies in the integrated critical interpretation of these documented mechanisms and in the explicit demonstration of the gap between collective funding balance and cohort allocation. Its simplified equations reconstruct established financial logic. The analysis makes no claim to a newly discovered pension law, a new theory of intergenerational justice, or a novel stochastic optimization model.

Reproducible evaluation for a plan-level extension

A future empirical investigation could be conducted with a standardized data dictionary containing membership start and end dates, contributions by employer and employee, time-indexed assets, historical investment cash flows, actuarial bases by valuation date, benefit rules and amendments, accrued and projected pension values, retirement dates, reserve balances, reserve transfers, benefit reductions and uplifts, mortality exposures, member communications, and transfer values. Data would be anonymized and linked across time, with distinct fields for legal obligations and actuarial targets.

The first empirical task would be to reconstruct the annual asset reconciliation from audited statements. The second would reproduce the scheme actuary's adjustment using the member census and operative rules in force at each valuation. The third would allocate realized and projected changes by cohort and calculate distributional measures under both actuarial and market-consistent bases. The fourth would compare alternatives under a transparent counterfactual method, such as individual DC with equivalent contribution rates, a separately purchased annuity, or a plausible alternate collective adjustment policy. The fifth would evaluate sensitivity to mortality, investment return, interest rates, inflation, mobility, and assumptions about member welfare.

An outcome assessment could report the incidence, frequency, and size of actual cash-benefit reductions; probability distributions for future pensions; downside consumption adequacy; variation across cohorts; transfers at entry and exit; and the amount of employer funding support. These measures should be reported alongside scheme maturity and liability duration. No statement that one pension architecture is empirically superior is warranted solely from the normalized examples in this manuscript.

Integrated Comparative Synthesis

The common financial problem

Each studied system must reconcile the desire for reliable retirement income with uncertain investment returns and the longevity of its members. Collective financing permits pooling of idiosyncratic demographic experience and joint investment management. Benefit adjustment and reserve policies are instruments for aligning the value of target claims with resources under stress. A plan that permits changes to benefits can remain financially balanced without forcing an unrestricted sponsor guarantee, but the quality of its retirement outcome depends on how it distributes those changes and whether it communicates the uncertainty.

The OECD's good-design pension recommendation gives a broadly relevant institutional setting: the design of payout options and the governance of pension assets need to support retirement income objectives. The recommendation is not a substitute for local law. The relevant question is which legal and financial allocation each country's system actually makes.

International comparison of adjustment and governance

Dimension Great Britain: collective money purchase Netherlands: contribution arrangements with collective reserves New Brunswick: shared-risk plans
Core legal architecture Authorized collective money purchase scheme, governed by the 2021 Act and regulations Solidarity or flexible contribution contracts under Dutch pension legislation and transition rules Registered shared-risk pension plan with funding and investment policies
Main benefit mechanism Scheme-rule actuarial valuation and annual benefit adjustment Individual pension capital and permitted collective return/risk allocations, including reserves Funding-policy sequencing for contributions, base benefits, and ancillary benefits
Risk-bearing and smoothing Collective valuation of benefit targets and rule-governed adjustment Cohort-specific return allocation and solidarity/risk-sharing reserve mechanisms Modeled risk targets, funding restoration, contingent benefits, and potential contribution changes
Employer commitment Depends on the statutory collective money purchase and scheme financing terms Depends on the occupational contract and employer/pension-fund arrangements Defined by plan documents and statutory shared-risk funding policy
Key analytical warning Uniform statutory adjustment does not imply uniform welfare effect Collective reserve is part of total assets and cannot finance arbitrary transfers Modeled target security is not a guarantee of realized benefits
Primary supervisory/legal evidence Pensions Act, collective scheme regulations and TPR's 2026 code Dutch pension law and DNB reserve guidance Pension Benefits Act, Shared Risk Plans Regulation and FCNB guidance

The table compares mechanisms, not performance league tables. It provides neither comparable countrywide CDC adoption counts nor scheme-level estimates of actual cuts and uplifts. Those observations would require matched data, a common benefit metric, and adjustment for institutional differences.

Five analytical findings

First, a collective funding ratio is a necessary financial indicator under its stated basis but an incomplete measure of income security. It captures a relationship between valued resources and target commitments at one date. It omits the probability and timing of a future payment change, the member's own risk exposure, and the rights held against employers and any insurers. A financially balanced pension fund can still have a highly volatile target income stream.

Second, benefit adjustment converts asset and longevity uncertainty into an explicit retirement income mechanism. Under the British statutory structure, adjustment belongs to the scheme's central method and oversight. Under the Dutch structure, the financial allocation also depends on reserves and rules for distributing returns among cohorts. Under the New Brunswick structure, funding policy can modify both contributions and different tiers of benefits. The same word “sharing” conceals these different contractual instruments.

Third, the timing of a benefit change is itself an allocation of pension value. A delayed response can protect a current pensioner from an immediate financial shock while leaving a larger residual adjustment risk for another year or cohort. An immediate reduction can place more loss on current beneficiaries. Reserve usage, transfer value policy, and entry pricing can modify this allocation. The fair-value research in Boelaars and Broeders (2019) explains why the price of market risk cannot be assessed separately from the discount and adjustment processes.

Fourth, the legal existence of a shared-risk system does not empirically prove superior welfare. Cui, de Jong, and Ponds' intergenerational sharing analysis finds potential gains in a modeled setting, while the hidden-transfer study illustrates how particular CDC transactions can redistribute economic value. The welfare gain of pooling has to be assessed alongside its financing, transition and incidence.

Fifth, governance has an economic function because the operative rules determine pension payoffs. Accurate actuarial calculations, credible valuation assumptions, independent review, effective investment controls, and comprehensible communications directly affect the reliability of pension outcomes. Institutional accountability and benefit-rule predictability belong in the financial analysis.

Welfare, stability, and the risk-sharing tradeoff

A member's valuation of collective retirement income depends on consumption prospects, risk aversion, existing pension wealth, horizon, and confidence in administrative performance. Stable projected payments can be highly valuable to a household whose basic spending is fixed. A younger member may value investment upside and the capacity to absorb investment volatility. A common adjustment policy can therefore increase expected welfare for one member while reducing it for another under reasonable preferences.

Intergenerational pooling is economically meaningful because people born in different years cannot trade complete claims on all future pension contingencies. A sufficiently well-governed collective institution may provide a form of long-term risk diversification otherwise costly or unavailable to individuals. It may also create an implicit obligation for later participants if benefits paid earlier are insulated from shocks. The existence and direction of such a transfer must be measured, not presumed.

The practical tradeoff is among the level of initial projected retirement income, the distribution of subsequent benefit changes, the expected returns and risks of underlying assets, and the funding support available from sponsors or participants. Increasing the projected pension by using a more optimistic expected return assumption is a valuation choice with consequences; it does not increase initial actual assets. Increasing the reserve by withholding gains provides more future buffering potential but reduces immediately allocated pension value. Guaranteeing a minimum payment requires another economically credible source of support or capital. These are statements of financing arithmetic and contractual design, not a normative prescription to prefer one scheme over all others.

Evidence priorities for future comparative research

The highest-value missing evidence is longitudinal plan-level member data paired with complete adjustment rules. For Britain, such data would permit assessment of the realized annual benefit path in authorized collective money purchase schemes, controlling for the 2026 regulatory extension. For the Netherlands, analysis could trace how allocations and reserves affect different age cohorts through the transition to the new pension framework. For New Brunswick, the funding policy and historical actuarial valuation series could show how contingent indexing, contributions, and base-benefit protections interacted over market cycles.

An informative comparative study would focus on consistent economic measures rather than equal legal labels. These include the present value and purchasing power of pensions by retirement cohort, annual volatility of actual cash payments, the distribution of downside outcomes, open versus closed group funding, and the actuarial credibility of the investment and mortality assumptions. Governance and communication evidence should be part of the dataset because member choice and trust can affect retirement behavior.

The mathematical cases presented here can serve as deterministic unit tests for a more elaborate actuarial engine: when all claims are proportional, the resulting adjustments must preserve balance by construction; when liability duration increases, rate sensitivity must be reconsidered; and when an internal reserve transfer is made, total fund assets must remain unchanged absent outside cash flows. Such tests help prevent arithmetic and accounting errors, while empirical validation requires actual contractual terms and member histories.

Conclusion

Collective defined contribution benefit adjustment is the contractual and actuarial process by which a jointly financed retirement system adapts targeted pension payments to financial and demographic experience. The process connects contributions, assets, valuation assumptions, promised or targeted payment streams, reserves, and legally authorized changes in retirement income. Its central accounting condition is the relationship between resources and the present value of the relevant benefit claims. Its economic substance lies in the distribution of the resulting financial gain or loss among people who differ in age, contribution history, retirement timing, and exposure to future uncertainty.

The primary-source comparison establishes that the major jurisdictions studied use related financial ideas through different legal forms. British collective money purchase schemes are structured around actuarial valuations and rule-governed benefit adjustments, with a supervisory framework extended to multi-employer designs in July 2026. Dutch solidarity and flexible pension contracts permit specific forms of collective risk sharing and reserves within contribution-based pension contracts. New Brunswick's shared-risk plans can adjust contributions and distinct categories of benefits according to legislatively defined funding policies and modeled risk-management objectives. Financial mechanisms must be analyzed within these original forms.

The analytical reconstruction identifies four consequential relationships. A funding ratio measures collective financial balance on a stated basis but cannot determine cohort fairness. Smoothing changes the timing and incidence of shocks while conserving total wealth unless new funding arrives. Benefit stability is partly a function of legal priority, actuarial basis, and reserve allocation. Finally, market-consistent valuation and transparent member communication are necessary complements to benefit-adjustment mechanics because the price and experience of risk depend on both.

The original contribution is a comparative scientific synthesis of enacted rules, supervisory requirements, pension economics, and internally consistent asset–liability examples. The results justify investigating CDC and shared-risk retirement arrangements as substantive forms of institutional risk allocation, with careful attention to both potential pooling benefits and implicit intergenerational transfers. They do not establish a universal optimal system or a new proprietary pension theory. Further conclusions about actual welfare, observed reduction frequency and cohort-specific transfers require primary scheme records and longitudinal evidence. The decisive object of evaluation is the verifiable relationship among adopted rules, resources, members' legal entitlements, and the pension income those rules ultimately deliver.

References