Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

IFRS 17 Portfolios, Groups, and Annual Cohorts

Abstract. The level of aggregation is a constitutive element of insurance financial reporting under IFRS 17. It determines which contracts share a measurement unit, where anticipated losses are recognized, how unearned profit is preserved and released, and how changes in profitability can be observed across generations of insurance business. This monograph investigates the formation of portfolios comprising contracts subject to similar risks and managed together, their division into initial-profitability categories, and the additional constraint prohibiting contracts issued more than one year apart from occupying the same accounting group. It reconstructs the interaction between grouping, initial recognition, fulfilment cash flows, contractual service margin, onerous losses, subsequent measurement, and disclosures. The research distinguishes the IFRS Accounting Standards issued by the International Accounting Standards Board from jurisdiction-specific endorsement, particularly the narrowly defined European Union option to omit the annual cohort requirement for certain participating and matching-adjustment business. The analytical method combines close reading of authoritative accounting requirements, a source-critical comparison of official implementation decisions and European legislation, examination of empirical research, and explicitly hypothetical numerical illustrations. The examples isolate the economic and accounting effect of combining profitable and onerous contracts, opening new annual cohorts, and allocating group-level estimates. They show that a correctly constructed aggregation map protects information about contemporaneous losses and successive underwriting generations while allowing shared estimation systems, appropriate cash-flow allocation, and systematic profit release. The analysis also establishes boundaries: contract grouping is an accounting unit-of-account rule, not an assertion that insurance risks cannot be pooled economically, and the visibility of initial cohorts alone does not establish complete comparability among insurers. The original contribution is an integrated interpretation of the connections among aggregation, profitability information, institutional exceptions, accounting systems, and financial-statement interpretation. The numerical demonstrations are independently constructed teaching examples, not observations of particular insurers or a newly proposed accounting doctrine.

Keywords: IFRS 17; insurance contracts; level of aggregation; insurance portfolios; similar risks; managed together; groups of insurance contracts; onerous contracts; annual cohorts; contractual service margin; fulfilment cash flows; initial recognition; profitability buckets; financial reporting; intergenerational mutualization; European Union endorsement; IFRS Interpretations Committee; accounting comparability.

Introduction

An insurer can manage risks collectively while reporting its performance through distinct accounting units. A participating life-insurance fund may share investment returns among generations; a property insurer can price, reserve, and reinsure a broad class of risks collectively; and a multinational group can combine information from several operating entities. Financial reporting must nevertheless reveal the emergence of economic losses and the release of profit attributable to services. The grouping of insurance contracts under IFRS 17 gives operational form to that objective.

The IFRS 17 accounting standard issued by the International Accounting Standards Board (IASB) took effect for annual reporting periods beginning on or after January 1, 2023, subject to the specified conditions for earlier application. It replaced the extensive diversity of accounting methods permitted during the IFRS 4 period with a model that combines current estimates of fulfilment cash flows, explicit compensation for nonfinancial risk, and the contractual service margin (CSM). At initial recognition, a profitable group normally carries unearned profit in its CSM; losses arising on an onerous group are recognized immediately. Subsequent results reflect insurance services and relevant changes in assumptions.

The unit at which such calculations are recognized matters. Suppose an insurer issues a loss-making product and a separately profitable product. A measurement scheme that permits unrestricted offset could present a positive combined result while concealing the initial loss. A similar problem occurs through time: favorable economics of a newly issued generation could compensate for deteriorating outcomes of a much older generation. The IASB's explanation of annual cohorts identifies information about performance and changes in profitability as central reasons for preserving separate issue-period groups.

The regulatory construct contains three linked filters: contracts must share a portfolio characterized by similar risks and joint management; the portfolio must be divided by initial profitability; and a group cannot contain contracts issued more than one year apart. These criteria operate with recognition rules and group measurement, rather than being merely labels for a financial-statement note. Under the EU-adopted IFRS 17 text, paragraphs 14-24 establish this accounting architecture, while paragraphs 25-28 govern recognition and inclusion of contracts in groups. The statutory text supplies a directly inspectable public version of the relevant paragraphs; the IASB-issued standard remains the international normative reference.

This investigation asks how aggregation criteria alter the reported timing and transparency of profit and loss; how a group is constructed and maintained when the underlying policies, assumptions, issue dates, and currencies differ; and how institutional deviations and empirical evidence qualify claims about comparability. The method is interpretive and analytical, supported by controlled arithmetic demonstrations. The hypothetical numbers serve to expose accounting relationships, and all inferences from them are identified as examples rather than evidence of industry-wide effects.

A distinction guides every chapter. The portfolio is a population of contracts meeting risk and management criteria. The group is the reporting unit within that population after profitability and issuance-period restrictions. The individual contract remains the source of rights, obligations, and many cash-flow estimates. Portfolio, group, and contract are therefore different levels of description that operate together. The IFRS Foundation's key terms provide the official terminology for portfolios, fulfilment cash flows, coverage periods, CSM, liability for remaining coverage, and liability for incurred claims.

Portfolio Aggregation Criteria

Similar risks and joint management

Paragraph 14 of IFRS 17 defines a portfolio by two simultaneous conditions: contracts are exposed to similar risks and are managed together. The meaning of the conjunction is important. Similarity of insured events alone is insufficient where an insurer manages business according to substantively different regimes. Joint administration or a common information system alone does not establish similar risk. Portfolio boundaries are justified by the nature and extent of risk and by an insurer's actual approach to management.

For example, individual term-life and single-premium fixed annuity contracts may both depend on mortality, yet their exposure directions, product economics, and management objectives differ. A death benefit may become payable upon death, while an annuity may require payment for as long as the annuitant lives. The IFRS 17 key-term explanation identifies product-line differences as an expected reason for separate portfolios. Conversely, contracts need not carry identical risks to share a portfolio: a portfolio is a sufficiently coherent management-and-risk population, not a set of perfectly interchangeable policies.

Portfolio identification demands more than assigning a product code. An insurer should be able to explain what risks are relevant, whether their extent is sufficiently similar, and how the business is managed. Useful evidence may include underwriting governance, pricing accountability, claims-management structures, asset-liability management decisions, actual portfolio reports, and the organization of reinsurance or capital oversight. These materials are evidence of management practice; none automatically overrides the standard's two criteria. A portfolio description based only on identical reporting software would fail to demonstrate why the risks themselves are similar.

Financial risk requires particular attention. In its October 2022 multi-currency agenda decision, the IFRS Interpretations Committee concluded that an entity considers all relevant risks, including foreign-exchange risk, when identifying portfolios. The Committee also expressly recognized that similar risks need not be identical. Consequently, differences in the currencies of premiums and claims require analysis; they neither compel separate portfolios in every circumstance nor justify treating all currency differences as irrelevant.

This conclusion is wider than a specialized foreign-exchange example. It exposes the difference between the identity of a risk and the degree to which that risk shapes the contract's overall profile. Two policies denominated in different currencies might remain similar if exposures, management practices, and economic sensitivity justify a common portfolio. Equally, material exchange-rate asymmetries could make their risks dissimilar. The decision requires a reasoned, documented judgment grounded in contractual economics, not a mechanical split by currency symbol.

The portfolio is not the reporting group

A portfolio may contain contracts expected to be profitable and contracts expected to be onerous. The portfolio criterion creates the first boundary; profitability and annual issuance filters create the groups that the entity recognizes and measures. The practical implication is that a single managed product family can give rise to multiple accounting groups. A manager may legitimately monitor the product as one economic franchise while the financial statements carry several cohorts and profitability categories.

This distinction also separates IFRS 17 from actuarial risk pooling. Actuarial valuation might estimate claims jointly over a large credible population, using common mortality tables, catastrophe models, loss triangles, or experience studies. IFRS 17 paragraph 24 permits an entity to estimate fulfilment cash flows at a higher aggregation level than the group when those estimates can be allocated appropriately to the groups. It is therefore possible to share models and assumptions without merging the recognized group-level accounting units. The condition is attribution: information cannot be pooled so broadly that group-specific cash flows, risk adjustments, and financial outcomes become indeterminable.

A useful operating map distinguishes four objects: (1) a contract record, including issue and recognition dates; (2) a management portfolio established by similar risks and common management; (3) a group identifier recording initial profitability and allowed issue period; and (4) the valuation and accounting measurements allocated to that group. Group identifiers should be stable after group formation. Model identifiers can be shared across groups, provided the allocation mechanisms deliver the correct recognized amounts. This arrangement satisfies the conceptual separation between estimating an uncertain liability and selecting the accounting unit on which that liability is reported.

Risk heterogeneity and the boundaries of judgment

Similarity is inherently a comparative assessment. An insurer evaluating two disability products might examine benefit trigger, waiting period, benefit duration, renewal guarantees, demographic exposure, currency, indexation, and underwriting control. Different contract characteristics can matter in different combinations. It follows that the evidence needed to identify a portfolio is specific to the products and their management context. The standard does not supply a universal numerical similarity threshold, a single mandatory segmentation algorithm, or a rule that every contract variation generates a new portfolio.

This discretion is constrained by the economic meaning of the standard. A broad portfolio assembled to conceal bad business could fail the similarity or management test; an arbitrarily fragmented portfolio could undermine consistency without giving users better information. The insurer's contemporaneous description of risk and management is therefore important for governance, audit, and comparability. The analytical test is whether a proposed portfolio can be explained from both risk characteristics and how management actually treats the business.

Portfolio-level conclusions

Portfolio selection is the first stage of information preservation. It constrains the initial population before profitability testing begins. The correct decision is sensitive to product economics, currency risk, and actual management. It also determines the cost of the reporting system: overly coarse boundaries create measurement and evidence risks, while highly granular boundaries increase models, reconciliation demands, and controls. The official implementation materials provide background on application, but the decisive judgment in each insurer remains the documented alignment between risks and joint management.

Onerous and Profitable Groups

Three minimum profitability categories

After identifying a portfolio, paragraph 16 requires at least three possible categories, where applicable: contracts onerous at initial recognition; contracts having no significant possibility of becoming onerous subsequently; and the remaining contracts. A particular portfolio can have fewer than three populated categories if one or more categories are empty. The distinction depends on the insurer's assessment at initial recognition and the required evidence about risk of later onerousness, rather than on a retrospectively observed profit outcome.

The category with no significant possibility of future onerousness is especially easy to misread. Its existence does not mean an insurer guarantees positive results for each contract or that adverse future assumptions are impossible. It reflects the entity's assessment at the initial date under IFRS 17's specified information and methodology. An initially profitable contract may belong to the remaining group if there is a meaningful possibility that its group could become onerous. That differentiation expresses uncertainty about the durability of expected profitability, not merely the sign of an initial point estimate.

Paragraphs 17-19 operationalize this assessment through reasonable and supportable information, including the possibility of assessing sets of contracts when that information permits the necessary conclusion. An entity must not assume that a whole product family is uniformly safe because its mean profitability is positive. Nor does the standard require exhaustive calculations of every possible contract state regardless of relevance. The EU-recognized requirements describe how profitability and the probability of onerousness are determined with evidence appropriate to the assessment.

The required partition is a minimum. Paragraph 21 permits further subdivisions, including separate profitable groups with different levels of profitability or different probabilities of becoming onerous, and further division of onerous groups where internal reporting supports that distinction. An insurer may therefore have more than three groups within a permitted issue period for a portfolio. Greater detail should be motivated by relevant information and supported by reliable processes, because disaggregation also increases estimates and reconciliation complexity.

Onerous recognition and the accounting asymmetry

IFRS 17 places losses and expected future profits on different reporting paths. When a group is onerous at initial recognition, the anticipated loss is recognized immediately in profit or loss. For a profitable group, the initial CSM defers unearned profit until services are provided. The IASB's IFRS 17 overview makes this difference central to the representation of insurance-service performance.

The asymmetry is an intentional consequence of reporting obligations, not a statement about the insurer's aggregate economic value. An insurance company might rationally operate a short-term loss-making line for strategic reasons, subsidize one policy type with another, or write contracts under regulation that restricts pricing. The consolidated entity may still be financially sound. IFRS 17 nevertheless tests losses using the required groups because positive expected performance elsewhere does not extinguish the unfavorable economics of the onerous group at initial recognition.

The simplified accounting logic can be stated as follows. Let P be the present value of expected premium inflows attributable to a group, O the present value of expected claims and other fulfilment outflows, and RA the compensation for bearing nonfinancial risk. Ignore acquisition cash flows paid before group recognition, taxes, financing effects, and other complications solely for the example. Define the illustrative net pre-margin amount S = P - O - RA. If S is positive, the example has initial unearned profit of S represented by CSM. If S is negative, the magnitude of the deficit is an illustrative initial loss. Actual IFRS 17 measurement uses the full contractual and accounting rules, including relevant cash flows, discounting, and existing balances; the simplified expression is a teaching identity rather than a substitute measurement rule.

Why separate groups? Consider two sets of same-portfolio contracts issued within an allowed annual window. One produces expected net surplus 28 and one produces deficit 11. If measured as one unrestricted accounting population, the net amount would be 17. That combined balance hides the fact that one set is onerous. Group-based recognition separately preserves the 28 of unearned future profit and recognizes the 11 loss immediately. The total arithmetic net is still 17, but the timing and meaning of reported financial performance differ. The distinction concerns information about obligations and service, not the conservation of numerical cash-flow totals.

Classification versus subsequent changes

Groups are formed at initial recognition, and IFRS 17 paragraph 24 prohibits subsequent reassessment of their composition. This is a significant constraint: if a group that was initially profitable later becomes onerous because expected future claims rise, the entity applies the applicable subsequent measurement and loss-recognition requirements to that existing group. It does not reclassify the same constituent contracts into a newly assembled "onerous-at-inception" bucket. The group's historical classification is preserved, and its subsequent measurement reveals the change in economics.

The reverse is equally important. Recovery of estimates after recognition of a loss does not justify rewriting the historical issue population. Reversals and profit adjustments follow the prescribed loss-component and CSM accounting within the group. Holding the unit of account stable supports analysis over time; moving contracts opportunistically among groups would blur whether a reported gain arose from improved experience, new business, or a redefinition of the measurement unit.

Portfolio changes from legitimate operational developments require careful tracing of what actually changed. A new policy issuance, an alteration of contractual terms, or a business combination can have relevant consequences under separate recognition and modification requirements. Each is a distinct event; a general decision to change the managerial view of a product cannot be used as an unrestricted mechanism for retroactively combining or redistributing previously recognized groups.

Regulation-constrained pricing

The three categories operate with an important exception in paragraph 20. Where contracts in one portfolio would fall into different groups solely because law or regulation specifically constrains the entity's practical ability to set different prices or benefit levels for policyholders with different characteristics, the entity may include those contracts in the same group. The scope of this permission is narrow: it depends on the stated cause of the difference and the actual regulatory constraint. It does not permit an entity to bypass profitability grouping whenever cross-subsidization is commercially desirable.

The regulator's control may reflect solidarity, community rating, or mandated benefit terms. Those objectives can affect the prices that risk-sensitive underwriting would otherwise produce. IFRS 17 allows the stated exception because a grouping rule applied without regard to the externally imposed constraint could treat an inseparable regulated business design as artificial loss-making and profitable subdivisions. Even then, applicable portfolio and issuance-period requirements remain relevant. Legal evidence and accounting evidence should be documented together.

What the categories establish

Profitability grouping preserves the distinction between an existing obligation expected to lose money and services expected to generate future profit. The distinction is measured with estimates, so it requires model governance and assumptions review, but it also constrains discretion in reporting. The accounting record should be capable of explaining why particular contracts entered a group and which available information supported the probability-of-onerousness assessment at that date.

Annual Cohort Requirements

The time boundary

Paragraph 22 of IFRS 17 prohibits including contracts issued more than one year apart in the same group. This restriction adds an issuance-time dimension to the portfolio and profitability boundaries. It is commonly implemented with annual issue cohorts, although the standard's literal condition concerns the period separating the contracts included in a group. The boundary should be operationalized consistently, including when writing volumes are seasonal, policy dates differ from recognition dates, and the insurer uses financial years different from calendar years.

The common label "annual cohort" should not be confused with a universal prescription to organize every group by January-to-December calendar years. The governing accounting fact is the permitted issuance interval. Entities use practical windows that consistently satisfy the requirement. The date when a contract was issued, the date it first meets recognition criteria, and the start of coverage may differ; systems should retain each separately rather than substitute one field for all three.

For a long-duration insurance portfolio, new issues may coexist for decades with older contracts whose expected cash flows are still developing. In the absence of a temporal limit, the profitability of new business could be combined with margins on old business and thereby obscure whether the insurer is maintaining underwriting discipline. The IASB annual-cohort rationale connects the separation of generations to the visibility of changes in expected and realized profitability over time.

Annual cohorts should also be distinguished from the annual reporting period. A reporting year is an external calendar for publishing financial statements; a cohort is a restriction on which contracts may occupy a group. Cohort information can feed several subsequent reporting periods as a multi-year contract supplies services and generates claims. A 2024 issue group can remain in the accounts in 2026, and a 2026 cohort can coexist with that earlier group without absorbing its unearned CSM.

A two-generation example

Assume the insurer writes homogeneous long-term contracts that otherwise meet the portfolio and profitability criteria. Its 2025 issuance produces expected surplus of 30 currency units after illustrative risk adjustment, while 2026 issuance is expected to produce a surplus of 12. Both generations may provide services over five years. An unrestricted rolling group would start with profit potential of 30 and later incorporate 12, with results and release patterns influenced by the new issue. Separate cohorts preserve a 2025 initial CSM of 30 and a distinct 2026 initial CSM of 12, subject to subsequent measurement, coverage units, interest, and other applicable adjustments.

Now suppose a revised outlook in 2027 reduces the expected future-service surplus of the 2025 group materially while the 2026 group remains stable. In separate groups, the change is attributed to the 2025 cohort. In a perpetual open group, the aggregation could make the change harder to interpret because the effect of later profitable sales might conceal deterioration of older service obligations. This illustration does not claim that distinct cohorts alone produce complete financial-statement disclosure of all commercial indicators: it demonstrates the information that the measurement unit is designed to preserve.

The effects can be sharper if new business is onerous. Imagine a later 2026 cohort with an initial expected loss of 8 alongside the earlier profitable group. The rules separating initial onerousness and issue periods prevent the newer initial loss from being absorbed by favorable earlier expectations. The reason for separation is not that policyholder experience is independent across generations; indeed, many participating contracts share returns. It is that financial reporting seeks to preserve the economic identity of distinct issuance and profitability groups.

Mutualization and the debate over costs

Intergenerational mutualization introduces a legitimate complication. In some participating products, cash flows to one generation affect the benefits, returns, or obligations associated with another. The insurer may manage the contractual pool jointly, and the allocation of mutualized cash flows among annual groups may require systems and judgments that are costly relative to the value of the information. The IASB retained the requirement after considering concerns of this kind. Its 2020 explanation of the requirement emphasizes the risk that broad aggregation would suppress information about profit patterns.

The decision can be understood as a trade-off between representational granularity and operational complexity. A finer unit of account raises data, valuation, and control costs. It also makes it harder for favorable results from new business to compensate silently for unfavorable results from older generations. The normative balance selected by the IASB does not imply that annual cohorts identify every economic relationship within a participating fund. It requires reporting units to retain at least this degree of vintage differentiation even when contract cash flows are economically linked.

The accounting rules for cash flows that affect other contracts matter here. Mutualized cash-flow relationships can be reflected in estimating fulfilment cash flows under the relevant IFRS 17 provisions. The group boundary and cash-flow allocation problem are therefore logically separate. One determines which contracts are recognized together; the other determines the amounts attributed to each group. Evidence that benefits are shared across generations supports attention to allocation. It does not by itself annul the international annual-cohort requirement.

The European Union exception

European endorsement created a legally significant distinction from IFRS 17 as issued by the IASB. Article 2 of Commission Regulation (EU) 2023/1803 permits an entity, for specifically described groups, not to apply paragraph 22. The first category concerns groups of insurance contracts with direct participation features and investment contracts with discretionary participation features with specified cross-contract cash-flow effects. The second concerns groups managed across generations that meet the Solvency II matching-adjustment conditions and have the required supervisory approval. The exemption is optional and constrained; it is not a general EU-wide deletion of annual cohorts.

The regulation requires disclosure of the use of the exemption as a significant accounting policy, with additional explanatory information such as the portfolios affected. A reader comparing two EU insurers cannot assume that a similar product label implies use of the same grouping method. Nor can the same reader assume that every insurer applying IFRS 17 worldwide has used the EU exception. The scope of adoption and the actual financial reporting framework must be identified before making a cross-company assertion.

The historical justification appears in the original 2021 EU endorsement regulation, which documented debate over intergenerationally mutualized and cash-flow-matched business, the cost-benefit balance, and the need for a narrow optional departure. Later consolidation into Regulation 2023/1803 preserved the substantive option. The distinction is one of legal adoption: the international standard remains the IASB text, while EU-adopted standards include the authorized regional option. The statement that a company "complies with IFRS 17" therefore needs a clear description of the framework under which its statements are prepared.

The exemption is also under institutional review. On September 25, 2026, EFRAG recommended postponing the exemption review because the timing of the IASB's post-implementation review risked leaving insufficient evidence before the statutory EU review point. The current EU regulation calls for Commission review by December 31, 2027. The EFRAG communication is a recommendation, not proof of an adopted amendment or an accomplished postponement. As of the research cutoff of November 14, 2026, the exemption remains part of the EU framework identified here.

Cross-jurisdictional reading

When assessing a multinational insurer, the financial reporting framework matters at least as much as the accounting terminology. Consolidated IFRS statements, EU-adopted IFRS statements, supervisory Solvency II calculations, and local statutory accounts can use different grouping or valuation requirements. The proper research question is: which framework governs the reported number, which specific group rule applies under that framework, and which disclosures identify deviations? Treating the mere use of the acronym IFRS as conclusive would erase distinctions expressly written into endorsement law.

The annual-cohort debate therefore has two analytic levels. At the IASB level it concerns the quality of performance information, the unit of account, and the balance between information and cost. At the EU level it concerns the authorized scope of deviation, supervisory qualifications, and disclosures necessary to interpret the framework. Both levels are integral to comparative research, and neither is reducible to an actuarial calculation alone.

Group Formation and Issuance

Formation is an accounting event sequence

The group construction process is frequently described as a three-step allocation: portfolio, profitability, and cohort. Operationally, it is a controlled sequence of events. The insurer first determines the relevant insurance contracts and identifies portfolios consistent with risk and management. It then applies the prescribed initial-profitability assessment and issuance-period restriction. Finally, it recognizes each group as its constituent contracts meet the recognition rules. A contract's management population can be known before it meets accounting recognition criteria; the final accounting entry nevertheless depends on recognition.

Paragraph 25 requires recognition of an issued group at the earliest of: commencement of the group's coverage period; the date the first policyholder payment becomes due; or, for an onerous group, the date the group becomes onerous. Paragraph 26 addresses absence of a contractual due date and the need to assess evidence of onerousness in advance of ordinary triggers. Paragraph 28 governs additions to a group during a reporting period and the measurement implications of recognized additions. These requirements can be inspected in the EU IFRS 17 standard text.

The word "earliest" matters. The system should test more than a single contractual start field. For example, where the initial premium is due before coverage begins, recognition may occur at the payment-due trigger rather than at the later coverage date. Conversely, where a group becomes onerous before that trigger, paragraph 25(c) can bring forward recognition. A system that records only policy issue month and policy inception date risks missing a material early recognition event.

Contracts added to existing groups

A group is not necessarily fully populated at the first recognition event. Under paragraph 28, the entity includes contracts satisfying recognition criteria in the relevant period and updates the group for permitted additions. The meaning of this permission must be read alongside the cohort constraint. A reporting system cannot leave a group open indefinitely simply because it has a convenient common product identifier. It must assess whether each proposed contract is eligible for that group's portfolio, profitability, and issuance window.

The group becomes stable as a composition of contract identities under paragraph 24. Later economic changes adjust recognized measurements according to IFRS 17 but do not provide a general right to exchange old contracts for newly favorable ones. Both controls are necessary: appropriate addition when a contract first qualifies, and prohibition on later opportunistic regrouping. The audit trail should preserve the date, evidence, and criteria governing each inclusion decision.

A practical group identifier

An adequate data design need not use one universal coding convention, but it must record the underlying dimensions unambiguously. A conceptual group key could include a legal reporting entity, a documented portfolio, an initial-profitability class, and an issuance-window identifier. The group key should be tied to the precise IFRS 17 accounting model and its ledger. For example, "Entity A / Term Life / Remaining Profitability / 2026 Issue Window" identifies a permitted group category more clearly than an arbitrary product code alone. This is a pedagogical data representation, not a new mandatory IFRS coding system.

A robust record includes contract ID, legally relevant issuance date, coverage start, first payment due date, initial onerousness assessment, reporting group ID, effective dates of coverage and contract modifications, allocation of acquisition cash flows where relevant, actuarial data sources, and source-system lineage. These fields serve different tests. Keeping them distinct enables reproducibility when policy administration data change, premiums are adjusted, or valuations are rerun.

Data governance should preserve both decisions and evidence. When two products are classified as similar risk, the record should explain why. When a profitable group is assigned the category with no significant possibility of later onerousness, the record should indicate assumptions and reasonable supporting information at the date. When an entity invokes a regulation-constrained pricing exception, the legal rule and its actual impact on differential pricing should be retained. When the EU annual-cohort option is applied, the exact legal category and disclosure choice should be documented.

General measurement and common estimate engines

Under the general measurement model, a group's carrying amount combines fulfilment cash flows and CSM, with subsequent accounting depending on service, assumptions, finance effects, and whether the group is onerous. The official IFRS 17 terminology guide distinguishes liability for remaining coverage from liability for incurred claims. Although estimates may originate from pooled actuarial engines, financial reporting must attribute the relevant flows and margins to the recognized groups.

The practical problem is one of traceable allocation. A mortality table calibrated to 100,000 policies can legitimately inform multiple groups. The required group-specific cash flows are then estimated from the contracts and assumptions relevant to each group. A catastrophe model might aggregate correlated shocks across contracts, but group allocations must retain the appropriate effects of those risks. A shared engine is a computational resource; its outputs must remain reconcilable to the accounting-unit inventory.

The premium allocation approach (PAA) does not nullify the initial grouping rules. Where PAA eligibility and measurement requirements apply, simplified subsequent measurement can reduce some calculations, particularly for short-duration insurance. Contracts still have to be assigned under the relevant portfolio, profitability, and issue-period rules. Likewise, groups of reinsurance contracts held have modified classification and recognition rules under paragraphs 61-70A, so a direct mapping from underlying issued groups to reinsurance-held groups cannot simply be presumed. These are important neighboring subjects but warrant their own full research treatment.

The multi-currency measurement judgment

Currency is a high-value example of group governance because it connects the first portfolio test to later measurement. In the final IFRIC explanation, the Committee found that all relevant risks, including exchange-rate risk, enter the similar-risk judgment. For multi-currency groups, IFRS 17 and IAS 21 then require the entity to consider currency denomination and translation. The group, including CSM, is treated as a monetary item for the relevant purpose, and an accounting policy determines how its currency or currencies are represented. The decision emphasizes consistency and relevant, reliable information.

That judgment does not permit an accountant to create a separate contractual service margin for each currency as if each were an independent accounting group when the correct unit is a single group. The Committee's discussion confirms the existence of one group-level CSM, with denomination and translation policies controlling the currency measurement. The distinction illustrates a general point: dimensional data may be granular inside a unit of account, while the recognized unit remains the correctly formed group.

Formation control as a reporting safeguard

Grouping errors can propagate from initial recognition across multiple financial reporting periods. An incorrect initial cohort may distort later CSM release. A wrong profitability category may conceal or accelerate a loss. An unrecorded inclusion or modification may produce differences between policy administration counts, actuarial liabilities, and the general ledger. Prevention therefore depends on a full source-to-ledger trace: contract inventory to group assignment, group assignment to cash-flow estimates, estimates to accounting balances, and balances to financial-statement disclosures.

Group-Level Accounting Examples

Method, assumptions, and scope

The numerical examples in this chapter are original, illustrative reconstructions designed to demonstrate effects of grouping. They use arbitrary currency units and independently stipulated estimates. No sample, insurer, jurisdictional financial statement, or market dataset supplied these numbers. Each example deliberately simplifies IFRS 17 to isolate the aggregation issue. Real measurement can include the time value of money, discount-rate changes, acquisition cash-flow assets, multiple services, investment components, finance income or expense, and group-specific contract terms. The examples therefore test conceptual arithmetic rather than prescribe journal entries for every fact pattern.

For the first example, assume all contracts are in one valid portfolio because they are subject to similar risks and managed together. Assume all are issued within the same permitted 2026 annual window. Their classification nevertheless differs at initial recognition. Assume no acquisition cash flows previously recognized, no immediate cash settlement, and no separate financial component in the displayed numbers. All amounts are already stated on a common present-value basis and adjusted for nonfinancial risk where indicated.

Example A: separation of three profitability classes

Item at initial recognition Group A: stable profitable Group B: remaining profitable Group C: onerous
Illustrative contracts 100 40 20
Present value of expected premiums 180 90 55
Present value of expected claims 138 70 59
Present value of expected other fulfilment expenses 13 10 6
Risk adjustment for nonfinancial risk 8 7 4
Net expected surplus / (deficit) 21 3 (14)
Initial illustrative CSM 21 3 0
Initial loss recognized in profit or loss 0 0 14

The calculations are reproducible. For Group A, 180 - 138 - 13 - 8 = 21. For Group B, 90 - 70 - 10 - 7 = 3. For Group C, 55 - 59 - 6 - 4 = -14. Total initial illustrative CSM is 24, and total initial loss recognized in profit or loss is 14. Their net is 10. If the three populations were offset in one hypothetical unrestricted bucket, the arithmetic net would also be 10, but the separate loss of 14 would be obscured and the positive CSM would be presented as only 10. The group structure makes both components observable through the accounting model.

This example should be read carefully. The illustration assumes Group A has no significant possibility of future onerousness under the relevant initial assessment and Group B is profitable but cannot meet that stricter assessment. Their different categorization is not inferred merely from the magnitudes 21 and 3. A small initial margin may contribute to risk of onerousness, but the applicable assessment must account for the contract's risk profile, assumptions, and available reasonable information. The classification conditions are stipulated here for demonstration.

The example also shows that recognition and valuation cannot be collapsed into a single "total expected profit" metric. Combined expected net cash surplus can be unchanged across accounting partitions, yet recognized initial losses and deferred future profit differ because the unit-of-account rules constrain offset. The result is logically compatible with the IFRS 17 profit-and-loss framework, which recognizes initial onerous losses and defers expected profits into CSM.

Example B: successive issue cohorts

Assume an entity writes two otherwise comparable profitable groups in successive annual issue windows. Each meets the same portfolio and initial-profitability criteria. Expected initial unearned profits are 30 for the 2025 issuance and 12 for the 2026 issuance. The amounts have been measured at their respective recognition dates. For clarity, assume that subsequent model adjustments and any interest effects have been excluded from a purely illustrative straight-line release exercise. This is not a general IFRS 17 coverage-unit rule, because actual CSM allocation follows the relevant insurance-contract services and coverage-unit requirements.

Initial cohort Illustrative CSM at issue Assumed equal service units Illustrative CSM per equal unit
2025 cohort 30 5 6
2026 cohort 12 4 3

Under these assumptions, the 2025 cohort releases an illustrative six units of initial CSM per equal unit of service, and the 2026 cohort releases three. The service measures are different because the contracts and stipulated service patterns differ. A consolidated presentation can aggregate recognized service results after the calculations, but the group accounting retains the ability to attribute the unearned profit to each original issue period.

Now suppose a deterioration in the expected future cash flows of the older group reduces the 2025 CSM by eight in a subsequent period, with all other effects held constant. The 2025 group's remaining unearned margin absorbs the relevant adjustment to the extent permitted under IFRS 17's subsequent-measurement rules, while the 2026 group remains unchanged. There is no accounting authority for merging the two groups merely because they share a management portfolio. Separate vintage information survives precisely because the first and third grouping filters continue to matter after initial recognition.

Example C: a later onerous cohort

A third issue period illustrates the cost of an unrestricted rolling group. Suppose 2027 issuance has an expected initial deficit of nine units, while the previously issued 2025 and 2026 groups retain positive unearned margins. If an entity improperly joins all vintages and nets their expected economics, it could hide the nine-unit initial loss in an overall positive reserve for future services. The applicable combination of profitability grouping and the one-year issuance restriction instead preserves the 2027 loss on the onerous group.

The existence of economic cross-subsidy would not change the accounting conclusion without a specific applicable exception. A profitable old business book may fund corporate spending or operational investment that supports new business, but that strategic use of resources is not the same as satisfying IFRS 17's conditions for accounting-group membership. If the EU elective exception applies, its legal requirements must be demonstrated before using a different cohort rule; generic reference to mutualization is insufficient.

Example D: an allocation control

Consider a valuation engine that calculates total future claims of 1,000 units across four valid accounting groups, with group-level policy counts and expected claim-cost relativities used to allocate the result. Suppose the documented, calibrated weights are 0.20, 0.25, 0.15, and 0.40. Allocated claims are therefore 200, 250, 150, and 400, and the group total reconciles to 1,000. This is an arithmetic example of higher-level estimation followed by attribution; IFRS 17 paragraph 24 requires appropriate fulfilment cash flows for each group, not a fixed allocation by count or by an arbitrary weight.

A reconciliation test should confirm that weights sum to one, allocated amounts sum to the source model's total, and the chosen weights reflect the relevant economic exposures. More importantly, the insurer must assess whether a single shared estimate is appropriate at all. If the groups differ materially in benefits, duration, or claim severity, a simple count allocation may be misleading. Group-specific estimates or richer allocation logic may be required. The permission to estimate at a higher level is therefore a practical flexibility accompanied by responsibility for group-level accuracy.

The four examples collectively establish that aggregation choice influences loss recognition, unearned profit measurement, generational analysis, and the attribution of actuarial estimates. Those conclusions follow from the accounting structure and the stated arithmetic. They do not establish an empirical estimate of how much insurers' equity, income, or capital changed after IFRS 17 adoption.

Systems and Comparability

From source contract to financial statement

A sound IFRS 17 architecture contains related but distinct layers. Policy systems identify contracts, parties, issuance dates, premium schedules, coverage periods, and claims. Risk and actuarial systems estimate fulfilment cash flows and nonfinancial risk compensation. Grouping controls assign each recognized contract to its permitted accounting population. Accounting engines calculate group-specific CSM, loss components, service results, and finance effects. General ledger and disclosure processes reconcile those results to reported positions and performance.

The architecture does not require one technology vendor, database, or actuarial projection method. It requires explicit evidence that the flow of data can preserve the IFRS 17 unit of account. The same contract should not migrate between groups as models are recalibrated. A valuation system should not erase prior cohort membership when a portfolio manager changes organizational responsibility. A financial statement number should be reconcilable to the underlying group population and the assumptions used in calculating it.

Material controls include completeness of contract capture, stable group keys, date consistency, dual evidence for portfolio judgment, profitability classification, exception eligibility, and audit trails for recalculations. Group assignment must be reconciled at each reporting close with contract movements and the accounting rollforward. A useful reconciliation separates beginning group carrying values, new recognition, cash flows, service release, changes in future-service estimates, incurred-claims movements, and applicable finance effects. The categories must follow IFRS 17's detailed measurement requirements; they should not be conflated into an undifferentiated movement total.

Comparability within the same standard

Uniform recognition of a minimum group structure makes different insurance businesses more interpretable, but it cannot eliminate all judgment. The assessment of similar risks and management practices, probabilities of future onerousness, fulfilment assumptions, risk adjustments, and contract services can vary among issuers. The international standard constrains these decisions but often depends on information about contracts and management that is not identical across firms. A statement that all IFRS 17 companies are perfectly comparable at a portfolio or cohort level would exceed what the rules themselves guarantee.

An especially relevant empirical source is ter Hoeven, Borelli, and van Vlijmen's (2024) study of European insurers' first-year reporting. It examined the 2023 annual reports of 24 large European insurance companies and found areas where more disaggregation and clearer analyses of movements in insurance liabilities and sensitivity information would improve understanding. The study also identifies impediments to comparability from accounting choices and the EU endorsement option. Its sample and disclosure focus support claims about observable reporting practices, not a causal estimate of portfolio grouping's economic effect on the entire global insurance industry.

Empirical findings should also be treated as time-specific. First-year adoption is a transition context, in which historical IFRS 4 variation, transition approaches, accounting policy selections, and initial systems configuration may strongly influence numbers. A comparison of 2023 statements does not establish that deficiencies remained unchanged in 2025 or 2026. Later reports can improve without implying a change in the underlying rule; conversely, a stable rule can still yield differing disclosures. It is therefore appropriate to use the study to motivate disclosure questions rather than convert it into a universal claim of IFRS 17 failure or success.

The broader valuation discussion by Vazov and Hristozov (2025), in their IFRS 17 valuation and profit analysis, explores how recognition, measurement approaches, and profit management interact after adoption. This study is useful for contextualizing operational and reporting effects. Its comparative, example-based approach should not be treated as a controlled causal test that annual cohorts alone caused a specified improvement in global accounting quality. The analytical evidence needed for that stronger inference would require carefully matched observations and explicit separation of concurrent regulatory and market changes.

Comparability across frameworks

The European cohort option is a concrete example of framework-dependent comparability. A company applying the IASB-issued standard with annual cohorts and a qualifying EU company electing the exception may represent similar economic participating business using different grouping boundaries. Their totals may still be reconcilable for some purposes, yet the pattern of CSM and recognized service results can reflect differences in the unit of account. Disclosures describing the option are therefore necessary to interpret the statements.

As of September 2026, EFRAG's public recommendation on review timing indicates that the exemption remains a live policy issue. The recommendation seeks further evidence from the IASB's post-implementation process before reconsideration. A scheduled IASB meeting for November 17-19, 2026 includes an IFRS 17 review initiation item. Because that meeting follows this dossier's November 14 research cutoff, its future decisions are not presented as accomplished facts. Likewise, any proposal to postpone the EU review deadline should be distinguished from an adopted legal amendment.

The same methodological caution applies to national adoption dates and regulatory overlays. A jurisdiction may adopt IFRS 17 at a different time, attach different financial statement requirements, or use separate solvency reporting. The correct comparison should record the reporting period, applied framework, actual grouping policies, and treatment of optional departures. A broad statement about "IFRS 17 internationally" is analytically incomplete when the sample combines distinct adoption regimes.

Financial-statement disclosure and its limits

IFRS 17 requires disclosures sufficient to help users assess the effect of insurance contracts on financial position, performance, and cash flows. The required reconciliations and judgments offer insight into different measurement components. Yet the underlying group-level unit of account does not mean that each individual group must necessarily be presented as a separate line in the public financial statements. Accounting measurement granularity and external disclosure granularity are related but distinct. Paragraph 95, for example, warns against obscuring useful information through excessive aggregation or insignificant detail.

Consequently, an analyst should not infer the absence of annual cohorts from a financial statement note that discloses information by broader product line. The question is whether the company constructed and measured its groups correctly and whether the published disclosures comply with the relevant disclosure requirements. A company could maintain thousands of internal accounting groups while presenting coherent aggregates externally; the aggregate disclosures still must avoid obscuring information about materially different characteristics.

Several focused questions improve interpretation. Does the insurer identify and explain significant portfolio groupings? Are initial onerous losses and loss components intelligible? Are movement analyses sufficient to understand liability changes and CSM release? Which key assumptions drive variation between current and prior periods? Has an EU option been applied and to which populations? These are analytically grounded reading tests. They do not amount to a proposal for new mandatory disclosures beyond the governing standards.

Reporting integrity and model risk

The most important operational failure is often a mismatch across data layers. A policy system may classify contracts by marketing product, an actuarial engine by homogeneous claim category, and a ledger by legal entity. None alone necessarily equals the IFRS 17 portfolio or group. A reconciliation that uses one label across all systems without mapping their different meanings can silently create inaccurate allocations.

For group QA, the system should test uniqueness of contract-to-group attribution, enforce the one-year issue interval unless a valid adopted exception is documented, require an initial-profitability decision when recognition occurs, and preserve immutable historical group membership for subsequent valuation. It should also test whether allocated estimates reproduce upstream totals and whether actual group-level calculations reconcile to the reported financial statements. Such tests improve traceability; they cannot replace professional assessment of whether the underlying group criteria were applied appropriately.

Governance extends to documentary evidence. Management assertions about similar risks, legal restrictions, and participation features should be retained with their source periods and responsible decision-makers. Changes in actuarial assumptions should be differentiated from changes in contract populations. Evidence of source-data completeness, model validation, and financial reconciliation should be available for review. The substantive reason is that a wrong initial group can affect many subsequent reporting periods; a one-time classification error can become a persistent profit-recognition error.

Synthesis and evaluation

The studied sources support five linked findings. First, IFRS 17 portfolios are constructed through risk similarity and joint management, rather than through insurer-wide economic pooling. Second, profitability grouping provides immediate visibility of expected losses while deferring unearned positive margins. Third, the issuance-window rule protects the distinct accounting histories of successive generations. Fourth, regulatory exceptions and multi-currency judgments are part of a correct international interpretation of grouping, rather than minor technical footnotes. Fifth, successful application depends on an auditable chain from contracts and actuarial data to groups, liabilities, and disclosures.

These findings are integrative analytical conclusions drawn from the standard and from controlled demonstrations. They do not create a new theoretical system, a new IFRS requirement, or an unverified claim of scientific priority. Their importance is practical and interpretive: unit-of-account choices establish what financial statements can reveal about how insurance obligations and profits evolve.

Conclusion

IFRS 17 uses portfolios, initial-profitability groups, and annual issuance restrictions to give insurance contract measurement a defined unit of account. A portfolio contains contracts exposed to similar risks and managed together. Within it, grouping separates initial onerous contracts, contracts with no significant possibility of subsequently becoming onerous, and other contracts. The further rule against combining issuance dates more than one year apart gives reporting a generational dimension. These rules operate with recognition dates, fulfillment cash flows, CSM, immediate onerous losses, and subsequent measurement, and their combined effect is central to the timing and interpretation of reported insurer performance.

The numerical illustrations demonstrate the informational consequence of these boundaries. Under the stipulated assumptions, expected net amounts can be the same when added together, while immediate loss recognition and deferred profits differ sharply under correct and incorrect aggregation. Separate issuance groups retain distinguishable histories of profit release and estimate changes. Higher-level actuarial estimation can coexist with group-level accounting if the insurer attributes appropriate amounts to each reporting unit and maintains reconciliation.

The international interpretation must identify the exact accounting framework. IFRS as issued by the IASB retains the annual-cohort requirement, while the EU-adopted framework permits a limited elective exemption for specified participating and matching-adjustment business with associated disclosures. The 2022 IFRS Interpretations Committee decision establishes that foreign-exchange risks enter the risk-similarity judgment and clarifies multi-currency group measurement. Current institutional discussion does not itself amend either set of binding requirements.

Available empirical research supports a qualified conclusion about comparability. Grouping rules improve the structural conditions for observing losses and vintage profitability, while differences in assumptions, adoption choices, transition, disclosure quality, and regional endorsements continue to affect cross-company interpretation. The studies cited here document particular first-year reporting experience and analytical contexts; they do not supply a global causal estimate attributable solely to annual cohorts.

The central scholarly conclusion is that aggregation under IFRS 17 is a reporting decision with lasting informational consequences. It translates the economics of insurance contracts into stable accounting units, preserves distinctions essential to loss and profit recognition, and places substantive demands on data, valuation methods, and institutional disclosure. Its value is realized through accurate contract classification, documented judgment, consistent allocation, and auditable reporting practice.

References

  • European Commission. (2021). Commission Regulation (EU) 2021/2036 of November 19, 2021, amending Regulation (EC) No. 1126/2008 as regards International Financial Reporting Standard 17. Official Journal of the European Union. Original endorsement regulation.

  • European Commission. (2023). Commission Regulation (EU) 2023/1803 of September 13, 2023, adopting certain international accounting standards in accordance with Regulation (EC) No. 1606/2002. Official Journal of the European Union. Regulation and IFRS 17 annex.

  • European Commission. (2024). Commission Regulation (EU) 2023/1803, consolidated version of January 9, 2024, Article 2. EUR-Lex. Consolidated adoption and cohort exception.

  • European Financial Reporting Advisory Group. (2026, September 25). EFRAG asks the European Commission to delay the IFRS 17 annual cohort exemption review. EFRAG news and recommendation.

  • IFRS Foundation. (2020, April). IFRS 17 Insurance Contracts: Why annual cohorts? Article concerning the IASB's rationale for the requirement. Annual-cohort explanation.

  • IFRS Foundation. (2022, September and October). IFRIC Update September 2022, including the October addendum: Multi-currency Groups of Insurance Contracts (IFRS 17 and IAS 21). Final agenda decision.

  • IFRS Foundation. (2023, effective January 1; website consulted November 2026). IFRS 17 Insurance Contracts: Issued standards overview and effective date. Issued-standard overview.

  • IFRS Foundation. (n.d.). IFRS 17 Insurance Contracts: Key terms. Educational terminology resource. Key-term glossary.

  • IFRS Foundation. (n.d.). IFRS 17 Insurance Contracts: Supporting materials. Standards implementation and educational resources. Implementation materials.

  • IFRS Foundation. (2026). International Accounting Standards Board meeting, November 17-19, 2026: Agenda item on initiation of the post-implementation review of IFRS 17. Future meeting agenda at cutoff.

  • ter Hoeven, R., Borelli, A., & van Vlijmen, P. (2024). First year's application of IFRS 17 in the financial statements of European insurance companies. Maandblad voor Accountancy en Bedrijfseconomie, 98(6), 403-422. https://doi.org/10.5117/mab.98.139090. Peer-reviewed study.

  • Vazov, R., & Hristozov, Z. (2025). Specific features of the application of IFRS 17: Valuation of insurance contracts and profit and loss management. Journal of Risk and Financial Management, 18(12), Article 706. https://doi.org/10.3390/jrfm18120706. Journal article.