IFRS 4 vs IFRS 17: What IFRS 4 Allowed and Why It Was Not Good Enough
Author
Sai Manikanta Pedamallu
Published
Reading Time
17 min read
Table of Contents
IFRS 4 vs IFRS 17: What IFRS 4 Allowed and Why It Was Not Good Enough
By Sai Manikanta Pedamallu (ACCA, CMA US, CSCA US, CGMA, ACMA, Dip IFRS, M.Com, MBA, MA)
Lead Instructor, Global Fin X | www.globalfinx.in/manikanta
IFRS 4 is usually described as a standard that permitted everything. That is not quite right, and the inaccuracy matters, because understanding precisely where IFRS 4 had teeth and where it did not explains exactly what IFRS 17 was designed to fix.
IFRS 4 did prohibit certain practices outright. It did impose a measurement floor of sorts. It did require specific disclosures. What it did not do was prescribe how insurance liabilities should be measured, and that single gap is what made comparability impossible for nineteen years.
Post 73 covered why IFRS 17 exists at a high level. This post examines IFRS 4's actual architecture, why an interim standard survived nearly two decades, and the transition bridge Indian insurers crossed on 1 April 2026.
IFRS 4's Architecture: One Exemption, Four Prohibitions
IFRS 4's central provision was a temporary exemption. It relieved insurers from the requirement to consider the IAS 8 hierarchy when selecting accounting policies for insurance contracts, which meant an insurer could continue applying its existing national practices without demonstrating that those practices produced relevant and reliable information.
That exemption was the source of the diversity. But IFRS 4 surrounded it with four specific constraints, and these are frequently overlooked.
Catastrophe and equalisation provisions were prohibited. IFRS 4 explicitly prohibited provisions for possible claims under contracts that are not in existence at the reporting date. Many national frameworks permitted insurers to build reserves against future catastrophic events that had not yet given rise to any contractual obligation. IFRS 4 stopped that, on the straightforward basis that a provision requires a present obligation arising from a past event, and a hypothetical future catastrophe under contracts not yet written is neither.
This had real consequences internationally. Where such provisions existed, they were reclassified from liability to capital in some jurisdictions, abolished entirely in others, and in at least one case abolished from consolidated financial statements while being retained in solo statutory accounts, with supervisors introducing separate regulatory requirements to replace the prudential function the provisions had served.
A liability adequacy test was required. IFRS 4 required insurers to assess at each reporting date whether recognised insurance liabilities were adequate, using current estimates of future cash flows. Where the assessment showed the carrying amount was inadequate, the entire deficiency had to be recognised in profit or loss.
Insurance liabilities could not be derecognised prematurely. An insurer was required to keep insurance liabilities on its balance sheet until they were discharged, cancelled, or expired.
Offsetting was prohibited. Insurance liabilities could not be offset against related reinsurance assets, and income or expense from reinsurance contracts could not be offset against the expense or income from the related insurance contract.
The Liability Adequacy Test: How Far It Actually Went
The liability adequacy test was the closest IFRS 4 came to imposing a measurement requirement, and it is worth understanding why it was insufficient.
Where an insurer's existing accounting policy already included a liability adequacy test meeting specified minimum requirements, IFRS 4 permitted that existing test to continue. Where it did not, IFRS 4 required a comparison of the carrying amount of the relevant insurance liabilities against the current estimates of all contractual cash flows and related cash flows.
The test therefore operated as a floor, not as a measurement basis. It caught liabilities that were plainly insufficient. It did nothing about liabilities that were adequate but measured on assumptions locked in decades earlier, or liabilities that were substantially over-reserved because of undisclosed prudential margins.
A liability could pass the adequacy test comfortably while remaining entirely uninformative about the insurer's actual expected obligations, and two insurers could each pass while reporting materially different amounts for identical portfolios.
What Grandfathering Actually Permitted
Within the exemption, the range of permitted practice was wide.
Locked-in assumptions. Discount rates, mortality, morbidity, and expense assumptions fixed at contract inception and never updated, producing liabilities that reflected economic conditions from the year the policy was written.
Implicit prudential margins. Margins built into assumptions to cover uncertainty, with neither the amount nor the confidence level identified or disclosed.
Profit recognised in relation to premium. Recognition patterns tied to premium receipt rather than to service delivery, permitting substantial day-one profit on long-duration contracts.
Undiscounted long-duration liabilities. In some frameworks, liabilities extending decades into the future were held without discounting at all.
Non-uniform policies across a group. IFRS 4 permitted, though did not require, a subsidiary's insurance accounting to differ from the parent's, so that a single consolidated group could report the same product on different bases depending on which subsidiary issued it.
The One-Way Ratchet on Policy Changes
IFRS 4 did constrain movement between policies, and it constrained it in one direction only.
An insurer could change its accounting policies for insurance contracts only if the change made the financial statements more relevant to users' decision-making needs and no less reliable, or more reliable and no less relevant.
IFRS 4 further specified that an insurer could not introduce certain practices even where its national framework permitted them: excessive prudence, the recognition of future investment margins subject to limited exceptions, and non-uniform accounting policies for subsidiaries where these did not already exist.
The effect was a ratchet. An insurer applying a weak practice could move toward a better one. An insurer applying a strong practice could not move back. Over nineteen years this produced gradual convergence in some jurisdictions, but it produced it unevenly and it could not produce comparability, because the starting points were too far apart and the movement was voluntary.
Unbundling: A Significant Structural Difference
Under IFRS 4, many entities were able to unbundle the insurance elements of contracts and apply the measurement principles of other standards to the separated components, applying IFRS 15 to service elements, IAS 37 to certain obligations, and IFRS 9 to deposit components.
IFRS 17 takes a different position. It generally applies to the whole contract, with separation permitted only in narrow circumstances: distinct investment components, distinct goods and services, and embedded derivatives that meet the separation criteria.
The consequence is that a contract which was previously fragmented across three standards is now measured as a single unit under IFRS 17. For entities that had built their accounting around unbundling, particularly non-insurers issuing contracts with an incidental insurance element, this represents a more fundamental change than the measurement model itself.
Shadow Accounting
IFRS 4 permitted, but did not require, a practice known as shadow accounting.
The problem it addressed: where an insurer measured its insurance liabilities using assumptions that did not reflect current market conditions, but held assets measured at fair value with movements in OCI, an accounting mismatch arose. Asset values moved with markets; liability values did not.
Shadow accounting allowed a recognised but unrealised gain or loss on an asset to affect the measurement of insurance liabilities in the same way that a realised gain or loss would, with the corresponding adjustment recognised in OCI alongside the unrealised asset movement.
It was a mismatch-reduction device layered on top of a measurement basis that created the mismatch in the first place. IFRS 17 addresses the underlying problem directly, through current measurement of liabilities and the OCI presentation option covered in Post 75, making shadow accounting unnecessary.
The IFRS 9 Problem and the Two Reliefs
A genuine practical difficulty emerged as IFRS 9 approached its 2018 effective date while the insurance contracts project remained incomplete.
Insurers would have been required to change the classification and measurement of their financial assets under IFRS 9 in 2018, and then change the measurement of the corresponding insurance liabilities under IFRS 17 several years later. Two disruptive transitions, in opposite halves of the balance sheet, at different times, with a period of amplified accounting mismatch in between.
The IASB responded in September 2016 with amendments to IFRS 4 providing two optional reliefs.
The temporary exemption from IFRS 9. Entities whose activities were predominantly connected with insurance could defer application of IFRS 9 entirely and continue applying IAS 39 until IFRS 17 took effect. This permitted a single simultaneous transition rather than two sequential ones. The exemption had a fixed expiry date, which was extended in June 2020 to align with IFRS 17's 2023 effective date.
The overlay approach. For entities applying IFRS 9 but wanting to reduce the resulting profit or loss volatility, the overlay approach permitted reclassification, from profit or loss to OCI, of the difference between the amounts recognised under IFRS 9 and those that would have been recognised under IAS 39, for designated financial assets backing insurance contracts.
The temporary exemption was the more widely used, and it explains a pattern visible across insurers globally: IFRS 9 and IFRS 17 adopted together on the same date. That is precisely what Indian insurers did on 1 April 2026, adopting Ind AS 109 and Ind AS 117 simultaneously.
Why an Interim Standard Lasted Nineteen Years
IFRS 4 was issued in March 2004 as Phase I of a two-phase project, explicitly intended to be replaced by Phase II. It remained in force until IFRS 17 became effective on 1 January 2023.
The delay was not administrative. The insurance contracts project was among the most technically difficult the IASB has undertaken, and several genuinely hard questions had to be resolved: how to measure liabilities extending decades into the future, how to treat participating features where policyholders share in returns, how to reflect the time value of money for cash flows with no observable market, and how to recognise profit for a service delivered continuously over many years.
Two exposure drafts were issued, in 2007 and 2010, and a revised exposure draft in 2013 following extensive industry response. Each round produced substantial comment and required reconsideration. IFRS 17 itself was then amended before it took effect, with its original 2021 effective date deferred twice.
The scale of the eventual implementation confirmed the difficulty. Industry commentary has been consistent that implementation was hugely complex, took longer, and cost more than originally anticipated.
IFRS 4 and IFRS 17 Compared
| Area | IFRS 4 | IFRS 17 |
|---|---|---|
| Measurement basis | Grandfathered national practices; no prescribed model | Single model: fulfilment cash flows plus CSM, with PAA and VFA variants |
| Assumptions | Frequently locked in at inception | Current at each reporting date |
| Discounting | Not required in all frameworks | Required, with prescribed principles for rate determination |
| Prudence | Implicit, undisclosed margins | Explicit risk adjustment with disclosed confidence level |
| Profit recognition | Often tied to premium receipt; day-one gains possible | CSM calibrated to prevent day-one gain; released as service is provided |
| Onerous contracts | Liability adequacy test as a floor | Immediate loss recognition with a loss component |
| Grouping | Not prescribed | Portfolios, profitability groups, and annual cohorts, fixed at inception |
| Unbundling | Widely permitted | Generally the whole contract; narrow separation only |
| Catastrophe and equalisation provisions | Prohibited | Prohibited |
| Offsetting insurance and reinsurance | Prohibited | Prohibited; reinsurance presented separately |
| Shadow accounting | Permitted | Not needed; addressed by current measurement and the OCI option |
| Non-uniform group policies | Permitted where already existing | Not permitted |
| Interaction with IFRS 9 | Temporary exemption and overlay approach available | Applied alongside IFRS 9 |
The Transition Bridge: Three Approaches
Moving from IFRS 4 to IFRS 17 requires establishing an opening balance for contracts already in force, and this is where the transition becomes genuinely consequential.
The full retrospective approach is the default. The entity applies IFRS 17 as if it had always been applied, reconstructing groups, cash flow estimates, discount rates, risk adjustments, and the CSM from the date each group of contracts was issued.
The modified retrospective approach is available only where full retrospective application is impracticable. It permits specified simplifications while requiring the entity to achieve the closest possible outcome to full retrospective application using reasonable and supportable information available without undue cost or effort.
The fair value approach is also available only where full retrospective application is impracticable. The CSM at transition is determined as the difference between the fair value of the group of contracts and the fulfilment cash flows measured at that date.
Why the Choice Matters So Much
The transition approach determines the opening CSM, and the opening CSM determines all future profit emergence for contracts in force at transition.
A higher opening CSM means more unearned profit to release in future periods. A lower opening CSM means less. For a life insurer with decades of in-force business, the opening CSM can represent the majority of the profit the existing book will ever report, and the transition approach selected effectively fixes it.
This is why the impracticability assessment receives such scrutiny. The full retrospective approach is not optional where it is practicable, and an entity that adopts a simplified approach without genuinely establishing impracticability has made a choice with direct and lasting earnings consequences.
For Indian insurers with long-duration participating and non-participating life business written over many decades, historical data at the granularity IFRS 17 requires frequently does not exist, and the modified retrospective or fair value approaches will apply to substantial portions of the in-force book.
The Indian Transition: Ind AS 104 to Ind AS 117
Indian insurers were applying Ind AS 104, India's equivalent of IFRS 4, and moved to Ind AS 117 with a transition date of 1 April 2026 under the IRDAI roadmap, as covered in Post 73.
Three features of the Indian position are worth drawing out.
Simultaneous adoption of Ind AS 109. Indian insurers adopted the financial instruments standard alongside the insurance contracts standard, exactly the outcome the IFRS 4 temporary exemption was designed to permit. The asset and liability sides of the balance sheet changed measurement basis on the same date.
Ind AS 104 continued for consolidation during the interim. The MCA's September 2024 transitional relief permitted insurers to continue preparing financial statements under Ind AS 104 for the purpose of consolidation by a parent, investor, or venturer until IRDAI notified implementation. This created a period during which an insurance subsidiary's contribution to a listed parent's consolidated accounts was still on the old basis.
Dual reporting during transition. As covered in Post 73, IRDAI required insurers to furnish Ind AS financial statements under Schedule IIA while separately publishing Schedule II Indian GAAP information, meaning the old and new bases run in parallel and are directly comparable for the first two years.
That parallel running is unusual and genuinely useful. Most jurisdictions transitioned with a restated comparative and nothing more. Indian users will be able to see the same insurer's results on both bases simultaneously, which makes the effect of the change unusually visible.
What Big 4 Auditors Focus On
The impracticability assessment for transition. Where a modified retrospective or fair value approach has been applied, auditors test whether full retrospective application was genuinely impracticable rather than merely difficult or expensive, and whether the entity has documented the specific information that was unavailable.
Consistency of the transition approach across groups. The approach is determined group by group. Auditors test whether the entity has applied full retrospective where it was practicable and a simplified approach only where it was not, rather than applying a single approach across the entire in-force book for convenience.
Opening CSM derivation. Because the opening CSM drives future profit emergence, auditors test its calculation in detail under whichever approach was applied, including the reasonableness of the assumptions used to reconstruct historical positions.
Completeness of contracts brought into scope. IFRS 17 applies based on the contract rather than the entity, and applies to the whole contract rather than permitting the unbundling that IFRS 4 allowed. Auditors test whether contracts previously accounted for under other standards through unbundling have been correctly brought within IFRS 17.
Removal of grandfathered practices. Auditors verify that practices permitted under IFRS 4 but not under IFRS 17, including undisclosed prudential margins, locked-in assumptions, and non-uniform group policies, have been genuinely eliminated rather than carried forward through the transition.
Dip IFRS Exam Angle
This comparison is examined conceptually, testing understanding of why the replacement was necessary rather than requiring computation.
Most tested areas:
Explaining what IFRS 4 permitted and why that prevented comparability.
Identifying the specific practices IFRS 4 prohibited, including catastrophe and equalisation provisions and offsetting of insurance against reinsurance.
Explaining the liability adequacy test and why a floor is not equivalent to a measurement basis.
Describing the three transition approaches and the condition attaching to the two simplified ones.
Explaining why the opening CSM is so consequential.
Common traps:
Describing IFRS 4 as permitting anything. It prohibited catastrophe provisions, required a liability adequacy test, prohibited premature derecognition, and prohibited offsetting.
Treating the transition approaches as a free choice. Full retrospective is the default; the other two require impracticability.
Assuming IFRS 17 permits the same unbundling IFRS 4 allowed. It generally applies to the whole contract.
Forgetting that IFRS 4's policy change constraint operated in one direction only, permitting improvement but not regression.
Assuming shadow accounting continues under IFRS 17. It is unnecessary because the underlying mismatch is addressed directly.
FAQ
Did IFRS 4 really allow insurers to do whatever they wanted?
No. It permitted continuation of existing measurement practices, which was the source of the diversity, but it prohibited catastrophe and equalisation provisions, required a liability adequacy test, prohibited derecognition of liabilities before they were discharged, and prohibited offsetting insurance liabilities against reinsurance assets.
What was the liability adequacy test and why was it not enough?
It required insurers to assess whether recognised liabilities were adequate against current estimates of future cash flows, recognising any deficiency immediately. It functioned as a floor, catching clearly insufficient liabilities, but it said nothing about liabilities that were adequate yet measured on decades-old assumptions or inflated by undisclosed margins.
Why did insurers get an exemption from IFRS 9?
Because applying IFRS 9 in 2018 and IFRS 17 several years later would have meant two disruptive transitions on opposite sides of the balance sheet, with amplified accounting mismatch in between. The temporary exemption allowed entities predominantly engaged in insurance to defer IFRS 9 and adopt both standards together.
Why does the transition approach matter more than most transition provisions?
Because it determines the opening CSM, which represents the unearned profit on all contracts already in force. For a long-duration life book, that opening balance can represent most of the profit the existing business will ever report, and the approach selected effectively fixes it.
Can an insurer choose the fair value approach because it produces a preferable opening CSM?
No. The fair value and modified retrospective approaches are available only where full retrospective application is impracticable. Impracticability must be genuinely established, not asserted, and auditors test the specific information that was unavailable.
What happens to shadow accounting under IFRS 17?
It becomes unnecessary. Shadow accounting existed to reduce mismatches arising from liabilities measured on outdated assumptions while assets were measured at fair value. IFRS 17 measures liabilities currently and provides the OCI presentation option, addressing the mismatch at source.
Enroll with Global Fin X
Understanding what IFRS 4 permitted, and what it did not, is what makes IFRS 17's design intelligible rather than arbitrary. Each of IFRS 17's demanding requirements exists because a specific IFRS 4 practice made comparison impossible. Our programme covers the full IFRS 17 series with detailed lectures, transition scenarios, exam-style MCQs, and a dedicated LMS for working professionals.
Enroll Now: Dip IFRS Programme
Faculty profile: www.globalfinx.in/manikanta
This is Post 76 of the Global Fin X IFRS Series. Previous: IFRS 17 Contractual Service Margin, Risk Adjustment and Disclosure. Next: Post 77: IAS 21 Effects of Changes in Foreign Exchange Rates: Functional Currency, Translation and Monetary Items.