IFRS 17 BBA, PAA and VFA: Three Measurement Models Explained
Author
Sai Manikanta Pedamallu
Published
Reading Time
19 min read
Table of Contents
IFRS 17 BBA, PAA and VFA: Three Measurement Models Explained
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
The first thing to establish about IFRS 17's three measurement models is that they are not a menu. An insurer does not survey its portfolio and pick whichever model suits it.
The general measurement model is the default and applies to everything unless one of two things happens. The variable fee approach is mandatory for contracts with direct participation features, and prohibited for everything else. The premium allocation approach is optional, available only for contracts meeting specific eligibility criteria.
So the sequence is: test for direct participation first, because that answer is compulsory in both directions. If the contract does not have direct participation features, test PAA eligibility. If the contract is not PAA-eligible, or the insurer chooses not to elect PAA, the general model applies.
Post 73 covered why IFRS 17 exists and the Indian implementation timeline. This post covers the three models and how to determine which applies. Post 75 covers the contractual service margin and risk adjustment mechanics in depth.
The Decision Sequence
| Question | If yes | If no |
|---|---|---|
| Does the contract have direct participation features? | VFA is mandatory | Continue |
| Is the contract PAA-eligible, and does the entity elect PAA? | PAA applies | GMM applies |
Two features of this sequence are worth emphasising because they are frequently misunderstood.
VFA is not optional. A contract meeting the direct participation criteria must use the variable fee approach. An insurer cannot elect the general model for a participating contract because it finds the general model simpler.
VFA is not available by choice. A contract that does not meet the direct participation criteria cannot use the variable fee approach, however much its economics may resemble participating business. The criteria operate as a gate in both directions.
Reinsurance is restricted. For reinsurance contracts, whether assumed or ceded, only the general model and the premium allocation approach are available. The variable fee approach may not be used for reinsurance under any circumstances. The measurement model applied to a reinsurance contract is determined independently of the model applied to the underlying direct contracts.
The General Measurement Model: Four Building Blocks
The general measurement model, commonly called the building block approach, measures a group of insurance contracts as fulfilment cash flows plus the contractual service margin.
Fulfilment cash flows themselves comprise three blocks, and the CSM is the fourth.
Block 1: Estimates of Future Cash Flows
An explicit, unbiased, probability-weighted estimate of the future cash flows within the contract boundary, comprising expected inflows (premiums) and expected outflows (claims, benefits, and directly attributable expenses).
The estimate must be current, meaning updated at each reporting date, and must reflect all reasonable and supportable information available without undue cost or effort. It is a probability-weighted expected value, not a single best-case or worst-case scenario, and it is entity-specific in the sense that it reflects the entity's own expectations, not a market participant's.
Block 2: Discounting
An adjustment to reflect the time value of money and the financial risks associated with the future cash flows, to the extent those risks are not already included in the cash flow estimates.
The discount rates must reflect the characteristics of the cash flows and the liquidity characteristics of the insurance contracts, and must be consistent with observable current market prices for financial instruments with consistent cash flow characteristics.
Block 3: Risk Adjustment for Non-Financial Risk
The compensation the entity requires for bearing uncertainty about the amount and timing of cash flows arising from non-financial risk: mortality, morbidity, lapse, and expense risk.
Financial risk is captured in Block 2 through discounting. The risk adjustment covers only the non-financial component, which prevents double counting.
IFRS 17 does not prescribe a technique for calculating the risk adjustment, but it does require disclosure of the confidence level to which the risk adjustment corresponds, which allows users to compare across entities using different techniques.
Block 4: The Contractual Service Margin
The unearned profit the entity expects to recognise as it provides insurance contract services.
At initial recognition, the CSM is calibrated so that no gain arises on day one. Where fulfilment cash flows produce a net inflow position, the CSM is set equal to that amount, leaving a total measurement of zero and no day-one profit. The CSM is then released to profit or loss over the coverage period as services are provided.
Where fulfilment cash flows produce a net outflow at initial recognition, meaning the group is onerous, there is no CSM. The loss is recognised immediately in profit or loss, and a loss component is established.
The Discount Rate: Two Permitted Approaches
IFRS 17 permits two methods for determining discount rates, and the choice has real consequences for Indian insurers given the structure of the domestic bond market.
The bottom-up approach starts with a liquid risk-free yield curve and adds a premium to reflect the liquidity characteristics of the insurance liabilities. Insurance liabilities are typically less liquid than risk-free instruments, so the illiquidity premium increases the discount rate and reduces the liability.
The top-down approach works in reverse. It starts with the observed market return on a reference portfolio of assets and strips out factors irrelevant to the insurance contracts, principally credit risk and any market risk premium not relevant to the liability cash flows.
In principle both approaches should converge on a similar rate. In practice they frequently do not, and IFRS 17 does not require reconciliation between them.
For Indian life insurers with liabilities extending decades beyond the longest liquid point on the government securities curve, the extrapolation of the yield curve beyond observable market data is itself a significant judgment area, with the ultimate forward rate assumption directly affecting the measurement of long-duration liabilities.
GMM Subsequent Measurement: What Moves the CSM
The CSM is not static. At each reporting date it is adjusted, and understanding which movements pass through the CSM and which do not is the core of general model mechanics.
Adjustments that unlock the CSM:
Changes in fulfilment cash flows relating to future service. If mortality assumptions improve, reducing expected future claims, the CSM increases and the additional profit is released over the remaining coverage period rather than recognised immediately.
The effect of new contracts added to the group.
Accretion of interest on the CSM at the discount rate determined at initial recognition.
Adjustments that do not touch the CSM:
Changes in fulfilment cash flows relating to past or current service go directly to profit or loss. A claim incurred in the period is a current service event.
Changes in financial assumptions bypass the CSM entirely under the general model. A movement in the discount rate flows to insurance finance income or expenses, either in profit or loss or in OCI depending on the entity's presentation election.
That last point is the single most important distinction between the general model and the variable fee approach, and it is where the two models diverge economically.
The floor. The CSM cannot be negative. Where unfavourable changes relating to future service exceed the remaining CSM, the CSM is reduced to zero and the excess is recognised immediately as a loss, with a loss component established.
The Premium Allocation Approach: A Simplification With Limits
The PAA is an optional simplified measurement approach, conceptually similar to the unearned premium accounting familiar from general insurance practice.
What PAA Actually Simplifies
This is the point most frequently misunderstood. PAA simplifies only the liability for remaining coverage. It does not simplify the liability for incurred claims.
An insurance contract liability under IFRS 17 has two components:
The liability for remaining coverage (LRC), representing the obligation to provide cover for insured events that have not yet occurred.
The liability for incurred claims (LIC), representing the obligation to settle claims for insured events that have already occurred.
Under PAA, the LRC is measured on a basis derived from premium received, allocated over the coverage period, without a CSM and without an explicit risk adjustment. The unearned premium is the starting point.
The LIC, however, is measured on essentially the same basis as under the general model: discounted probability-weighted expected cash flows, with a risk adjustment for non-financial risk. There is one simplification available: an entity need not discount the LIC where claims are expected to be paid within one year of the incurred date.
For a general insurer, this means the claims reserve requires the full IFRS 17 measurement machinery even where PAA has been elected. PAA is a simplification of the unearned premium side, not of the reserving side.
PAA Eligibility
An entity may apply the PAA where either of the following applies:
The coverage period of each contract in the group is one year or less. This is an automatic qualification requiring no further demonstration.
The entity reasonably expects that PAA would produce a measurement of the LRC that would not differ materially from the general model. For contracts with a coverage period longer than one year, this must be demonstrated, and the demonstration is not available where the entity expects significant variability in the fulfilment cash flows that would affect the LRC during the coverage period.
The second limb is where Indian general insurers with multi-year motor and health products need to perform genuine analysis rather than assuming PAA availability.
Acquisition Cash Flows and the Onerous Test
The PAA provides for insurance acquisition cash flows to be deducted from the LRC, effectively an implicit capitalisation. An entity may alternatively elect to expense acquisition cash flows as incurred, but only where the coverage period of each contract in the group is one year or less.
The onerous contract requirement survives the simplification. Where facts and circumstances indicate a group of contracts is onerous, the entity must measure the LRC using fulfilment cash flows and recognise a loss component. PAA removes the CSM; it does not remove the obligation to identify and recognise losses.
The Variable Fee Approach: For Direct Participating Contracts
The VFA applies to insurance contracts with direct participation features, meaning contracts that are, in substance, investment-related service contracts under which the entity promises an investment return based on underlying items.
The Three Eligibility Conditions
A contract qualifies as having direct participation features only if all three of the following conditions are met at inception:
One. The contractual terms specify that the policyholder participates in a share of a clearly identified pool of underlying items.
Two. The entity expects to pay the policyholder an amount equal to a substantial share of the fair value returns on those underlying items.
Three. The entity expects a substantial proportion of any change in the amounts to be paid to the policyholder to vary with the change in fair value of the underlying items.
A contract failing any one of these conditions must be measured under the general measurement model. The conditions perform a gate-keeping function, preventing the VFA from being applied to traditional insurance products with incidental discretionary features.
The assessment is made at inception and is not subsequently reassessed for changes in expectations, though it is reassessed if the contract terms are modified.
The Crucial Difference: Financial Variables Adjust the CSM
Under the VFA, fulfilment cash flows are measured in the same way as under the general model. The building blocks are identical.
What changes is what happens to the CSM.
Under the general model, changes in financial assumptions bypass the CSM and go directly to insurance finance income or expenses. Under the VFA, the change in the entity's share of the fair value of the underlying items adjusts the CSM.
The economic logic is that for a direct participating contract, the insurer's consideration is effectively a variable fee: a share of the returns on a pool of assets. If those assets perform better, the fee is larger; if they perform worse, the fee is smaller. Because the fee itself is variable and relates to future service, changes in it are appropriately absorbed into the unearned profit balance rather than recognised immediately.
The consequence is that the VFA CSM acts as a buffer against financial market volatility, producing a smoother profit pattern that better reflects the long-term nature of the insurer's management fee.
The floor still applies. If losses from falling asset values or worsening assumptions exceed the remaining CSM, the margin is reduced to zero and cannot go negative, with the excess recognised immediately in profit or loss.
Worked Comparison: The Same Market Movement Under GMM and VFA
The clearest way to see the difference is to run an identical financial assumption change through both models.
A group of contracts has a CSM of Rs. 400 crore at the start of the period. During the period, the fair value of underlying items falls, and the entity's share of that decline is Rs. 60 crore. No other changes occur. Two years of coverage remain, and the CSM is released evenly.
Under the general measurement model:
The Rs. 60 crore effect is a change in financial assumptions. It bypasses the CSM entirely and is recognised as insurance finance expense, in profit or loss or OCI depending on the presentation election.
CSM at period end: Rs. 400 crore less the period's release of Rs. 200 crore = Rs. 200 crore
Insurance finance expense recognised: Rs. 60 crore
Under the variable fee approach:
The Rs. 60 crore effect adjusts the CSM, because it represents a change in the entity's variable fee relating to future service.
CSM before release: Rs. 400 crore less Rs. 60 crore = Rs. 340 crore
CSM released in the period: Rs. 170 crore
CSM at period end: Rs. 170 crore
Insurance finance expense recognised from this movement: nil
The Rs. 60 crore does not disappear. It reduces the profit that will be recognised over the remaining coverage period, rather than hitting profit or loss immediately. Over the full life of the contract, the total profit recognised is the same under both models; the VFA changes when it emerges.
For an Indian insurer with a large unit-linked book, this difference determines whether equity market volatility flows straight into reported earnings or is absorbed into the CSM and released gradually.
Model Comparison
| Feature | GMM (BBA) | PAA | VFA |
|---|---|---|---|
| Status | Default model | Optional simplification | Mandatory for direct participating contracts |
| Availability | All contracts not using VFA or PAA | Contracts with coverage period one year or less, or where PAA approximates GMM | Only contracts meeting all three direct participation conditions |
| CSM for remaining coverage | Yes | No | Yes |
| Explicit risk adjustment in LRC | Yes | No | Yes |
| Liability for incurred claims | Full measurement | Full measurement, with optional non-discounting where settled within a year | Full measurement |
| Changes in financial assumptions | Bypass the CSM, to insurance finance income or expense | Not applicable to LRC | Adjust the CSM |
| Onerous contract loss | Recognised immediately | Recognised immediately where indicated | Recognised immediately |
| Available for reinsurance | Yes | Yes | No |
| Typical products | Term life, non-par savings, annuities, long-term protection | Motor, fire, marine, short-term health, most general insurance | Unit-linked, with-profits, participating savings |
Indian Product Mapping
Variable fee approach. Unit-linked insurance plans are the clearest Indian case, since the policyholder participates directly in a clearly identified pool of underlying assets and the returns vary directly with fund performance. Participating endowment and whole life policies require assessment against the three conditions, and the answer depends on the specific contractual participation mechanics and the bonus declaration framework, which interacts with Section 49 of the Insurance Act as noted in Post 73.
General measurement model. Non-participating term insurance, non-participating guaranteed savings products, annuities, and long-term protection business. Also any participating-looking product that fails one of the three VFA conditions.
Premium allocation approach. Motor, fire, marine, engineering, and most other general insurance lines with annual policies qualify automatically. Short-term health insurance similarly qualifies. Multi-year motor and health policies, which have become more common in the Indian market, require the eligibility demonstration rather than automatic qualification.
Reinsurance. Both inward and outward reinsurance are restricted to the general model or PAA. Given the scale of reinsurance ceded by Indian general insurers, and the obligatory cession to GIC Re, this restriction affects a material portion of the sector's contract population.
One presentational consequence worth noting: because reinsurance is presented separately rather than netted, the familiar gross and net view of premiums and claims does not survive IFRS 17 in its previous form. Reinsurance held is presented as a separate line rather than as a deduction from gross figures.
What Big 4 Auditors Focus On
VFA eligibility assessment. Because the VFA is mandatory where the conditions are met and prohibited where they are not, auditors test the assessment of all three conditions at inception, with particular attention to whether the policyholder participates in a clearly identified pool and whether the share of fair value returns is genuinely substantial.
PAA eligibility for contracts longer than one year. Where PAA has been elected for multi-year contracts, auditors test the demonstration that PAA produces a measurement not materially different from the general model, and challenge the assumption where significant variability in fulfilment cash flows is expected.
Completeness of the LIC measurement under PAA. Auditors verify that the liability for incurred claims has been measured on a full IFRS 17 basis, including discounting and risk adjustment, rather than being carried forward on a previous reserving basis on the assumption that PAA simplifies everything.
Discount rate methodology and yield curve extrapolation. Auditors test the bottom-up or top-down approach applied, the illiquidity premium or credit risk deduction, and, for long-duration Indian liabilities, the extrapolation methodology beyond the observable end of the government securities curve.
CSM unlocking classification. Auditors test whether changes have been correctly classified between those relating to future service, which adjust the CSM, and those relating to past or current service, which go to profit or loss, since misclassification directly shifts earnings between periods.
The CSM floor. Auditors verify that where unfavourable changes exceed the remaining CSM, the excess has been recognised immediately as a loss with a loss component established, rather than the CSM being carried at a negative amount.
Dip IFRS Exam Angle
Dip IFRS examines the principles and the selection logic rather than requiring full actuarial computation.
Most tested areas:
Determining which measurement model applies to a described contract, applying the selection sequence: direct participation first, then PAA eligibility, then GMM as default.
Identifying the four building blocks of the general model and what each represents.
Explaining what the PAA simplifies and, critically, what it does not.
Explaining the treatment of changes in financial assumptions under the general model compared with the VFA.
Recognising that the CSM cannot be negative and that the excess is a loss.
Common traps:
Treating the model choice as free. VFA is mandatory where conditions are met and prohibited otherwise; PAA is conditional on eligibility.
Assuming PAA removes the need for full measurement of the liability for incurred claims. It does not.
Applying VFA to a reinsurance contract. VFA is never available for reinsurance.
Passing changes in financial assumptions through the CSM under the general model. They bypass it.
Failing to pass changes in the entity's share of underlying items through the CSM under the VFA. They adjust it.
Recognising a day-one gain on a profitable group. The CSM is calibrated so no gain arises at initial recognition.
Assuming a contract with any participating feature qualifies for VFA. All three conditions must be met.
FAQ
Can an insurer choose the general model for a participating contract because it is simpler to operate?
No. Where a contract meets all three direct participation conditions, the variable fee approach is mandatory. The model is determined by the contract's characteristics, not by operational preference.
Does PAA eliminate the contractual service margin entirely?
It eliminates the CSM for the liability for remaining coverage, which is what the simplification addresses. The liability for incurred claims continues to be measured on a full basis including risk adjustment, and the onerous contract requirement continues to apply.
Why can the variable fee approach not be used for reinsurance?
Reinsurance contracts do not have direct participation features in the sense the VFA contemplates. A reinsurance contract transfers risk between insurers; it does not create a policyholder participation in a clearly identified pool of underlying items. The IASB restricted reinsurance measurement to the general model and PAA accordingly.
What happens when the CSM is exhausted?
The CSM is reduced to zero and cannot go negative. Any further unfavourable change relating to future service is recognised immediately in profit or loss, and a loss component is established. If conditions subsequently improve, the loss component is reversed before the CSM is re-established.
Does the choice between bottom-up and top-down discount rate approaches need to be disclosed?
Yes. IFRS 17 requires disclosure of the approach used to determine discount rates and, where a yield curve is used, the yield curve or range of yield curves applied. The two approaches are not required to be reconciled to each other.
How does the model selection interact with the annual cohort requirement?
They operate together. Contracts are first divided into portfolios of similar risks managed together, then into groups by expected profitability at initial recognition, and no group may contain contracts issued more than a year apart. The measurement model is then applied to each resulting group. A single portfolio can therefore contain multiple groups, each measured under the same model but separately.
Enroll with Global Fin X
The model selection logic is the foundation of IFRS 17: get it wrong and every subsequent measurement is wrong. Post 75 covers the contractual service margin, the risk adjustment, and the disclosure requirements in detail. Our programme covers the full IFRS 17 series with detailed lectures, worked comparisons across the three models, exam-style MCQs, and a dedicated LMS for working professionals.
Enroll Now: Dip IFRS Programme
Faculty profile: www.globalfinx.in/manikanta
This is Post 74 of the Global Fin X IFRS Series. Previous: IFRS 17 Insurance Contracts: Why IFRS 17 Exists and What It Changes. Next: Post 75: IFRS 17 Contractual Service Margin, Risk Adjustment and Disclosure.