Understanding the IFRS 17 income statement: the profit drivers

The previous post in this series worked through the best estimate liability (BEL) analysis of movement (AoM) — the walk from opening to closing reserve per cash-flow type, with expected closing as a named control point. That review establishes what the liabilities did during the period. The income statement review is the next step: it explains why the period produced the surplus or deficit it did. The two reviews are complementary — the liability movements feed the income statement line by line, and a discrepancy in one almost always surfaces in the other.

The general measurement model (GMM) income statement does not present premiums and a loss ratio in the form most readers are used to. It presents a set of profit drivers, each tracing back to a movement in the insurance contract liabilities. Reading the income statement means reading those drivers — knowing where each one lands, and what a variance in it is telling you.

The two results

IFRS 17 paragraph 80 disaggregates the statement of financial performance into two amounts: the insurance service result (insurance revenue minus insurance service expenses) and insurance finance income or expenses (IFIE). The first is the underwriting result — the profit or loss from providing coverage in the period. The second is the financial measurement effect on the liabilities.

In practice the financial side is read more broadly than IFIE alone. IFIE is only the liability side; set against the IFRS 9 investment return on the assets backing those liabilities, it gives the net investment result. Under matched asset-liability management the two largely offset, and the residual is the mismatch worth explaining. The label "net investment result" is a management and analyst construct rather than an IFRS 17-defined line, but it is how the financial result is actually interrogated.

The drivers below are organised by where they sit and which service period they relate to. The organising rule throughout is service-period attribution (paragraphs B96–B97): future-service changes adjust the contractual service margin (CSM); current- and past-service movements flow to profit or loss.

---

Driver 1 — CSM release

The CSM is the unearned profit in the group of contracts. It is recognised in profit or loss across the coverage period through the coverage-unit method (paragraph B119): the coverage units delivered in the period, as a proportion of the total remaining coverage units, applied to the CSM. Reviewing the release starts with the roll-forward.

Each line warrants a check. New-contract additions should be consistent with new-business volumes and initial-recognition assumptions. Interest accretes at the locked-in rate set at initial recognition of each cohort — not the current rate, which applies to the BEL. Future-service estimate changes (non-financial assumption and cash-flow updates) absorb into the CSM; favourable changes increase it, adverse changes reduce it, and an adverse change exceeding the remaining CSM tips the group into a loss component.

The coverage-unit basis is an explicit choice — often proxied by earned premium or expected exposure — and it controls the pace of profit recognition. A useful period-on-period metric is the CSM release % (CSM released divided by the pre-release CSM balance): a compression or expansion in the ratio without a change in the coverage-unit basis or the contract mix is worth investigating.

---

Driver 2 — Risk adjustment: released from coverage, set up on incurred claims

The risk adjustment (RA) for non-financial risk is carried on both the liability for remaining coverage (LRC) and the liability for incurred claims (LIC). That two-sided nature is what makes the RA a profit driver rather than a single line.

As current-period risk expires, the RA on the LRC releases into insurance revenue (paragraph B124(b)) — a gain. At the same time, as claims are incurred, RA is established on the LIC — on the incurred-but-not-reported (IBNR) reserve and the reported claims — inside insurance service expense. The two are the same risk moving from coverage to incurred; the net of the release and the set-up is the period's RA result. A RA release % (RA released divided by the pre-release RA balance) tracks the pace, and a sustained divergence from prior periods without a methodology change warrants explanation.

Future-service RA changes are separate again: where a non-financial assumption update changes the RA that relates to remaining coverage, that change adjusts the CSM (paragraph B96(d)), not the period's result. The RA method itself is principle-based; IFRS 17 prescribes only that the entity disclose the confidence level implied by its technique (paragraph 119).

---

Driver 3 — Premium and acquisition experience, current service

The insurance revenue build measures the amounts earned for coverage at the amounts expected at the start of the period (paragraph B124). Two experience items sit directly in revenue and are observable without reconstruction.

Premium experience for current and past service — actual less expected premiums — is part of insurance revenue (paragraph B124(d)). Acquisition-cost experience for current service is treated the same way: the actual-less-expected on acquisition costs is observable in insurance revenue alongside the premium experience. The acquisition-cost amortisation itself is not a profit driver — it is recovered through revenue and matched by an equal insurance service expense, so it nets to zero in the result.

The future-service portion of these variances is different: it adjusts the CSM (paragraph B96(a)) rather than the period's result. Only the current- and past-service portion is a driver of this period's income statement.

---

Driver 4 — Claims, current service

Revenue carries the expected claims for the period, measured at start-of-period estimates (paragraph B124(a)). The actual claims incurred hit insurance service expense — the claims paid for current-period events plus the IBNR raised for current service plus the reported-claims reserve raised for current service. The gap between the two is the current-service claims experience, and it flows straight to profit or loss (paragraph B97(c)).

This variance is not a line that can be read directly off the income statement; it falls out once the expected (in revenue) and the actual incurred (in service expense) are compared. When interrogating it, it is worth distinguishing genuine period experience from model error. A variance that recurs across periods at the same group level points to assumptions that need updating rather than a one-off event.

---

Driver 5 — Expenses, current service

Expense experience works the same way as claims. Revenue carries the attributable expenses expected for the period; the actual attributable expenses incurred hit insurance service expense; the difference is the current-service expense experience, flowing to profit or loss.

Separating variable costs — which should scale with the book — from overheads, and ad-hoc items from recurring costs, isolates what the variance says about the business. New-business acquisition spend allocated into in-force cohort experience without adjustment can inflate the apparent expense variance, so it is worth confirming the allocation before reading the result.

---

Driver 6 — Claims, past service

The LIC carries its own profit driver, separate from the remaining-coverage components, and is reconciled separately (paragraph 100(c)). It is the development of claims incurred in earlier periods.

Two movements make it up. Claims paid on prior claims run against the liability for incurred claims — a charge. Against that, the prior-period IBNR and reported-claim reserves release as those claims report and settle — usually a gain. Together they are the actual-versus-expected on past-service development, and the net flows to profit or loss as a change in the fulfilment cash flows of the LIC (paragraph B97(b)). The figure to interrogate is the change in best estimate versus the assumed development, not simply cash paid versus reserve held: a past-service release can lift the period's result for reasons that have nothing to do with current-period underwriting.

---

Driver 7 — Loss component

A group that is onerous at initial recognition carries a loss component of the LRC instead of a CSM. The combined ratio at initial recognition is the diagnostic; paragraph 47 is the consequence.

The loss component moves through profit or loss in both directions. At recognition the loss posts immediately and the CSM is zero. On subsequent measurement, a favourable future-service change reduces the loss component first, before any CSM is re-established (paragraph 50(b)); an adverse future-service change increases it (paragraph 48). It is systematically allocated to zero by the end of the coverage period (paragraph 52). Reading the loss component means identifying what adverse change created or extended it, and whether the assumption base has been updated.

---

Driver 8 — The net investment result

The financial result sets the investment return on the assets against the finance cost on the liabilities. The asset return is governed by IFRS 9; the liability side is insurance finance income or expense (paragraph 87) — the effect of the time value of money and of financial risk on the insurance contract liabilities.

IFIE has two parts. The unwind reflects the passage of time: the BEL and RA unwind at the current discount rate, while the CSM accretes at the locked-in rate set at initial recognition. The financial-assumption change reflects the discount curve moving: the BEL re-measures at the new rates, and that delta is also IFIE. The income statement review confirms the financial-assumption BEL change matches the figure carried through the analysis of movement.

A reliable sense-check tracks the unwind per cash-flow type — premium present value, claim present value, and expense present value separately — because the net BEL mixes inflows and outflows of different durations, and netting them masks distortions. Read against the IFRS 9 investment return, the net investment result reflects how well the asset strategy is matched to the liabilities: under a well-matched book the two largely offset, and the residual reflects duration mismatch, spread, or reinvestment risk.

---

Reading the release metrics together

Two ratios synthesise the period's profit-recognition mechanics. The CSM release % tracks how fast the unearned profit is being recognised relative to the book's remaining lifetime — a rising ratio means acceleration, from shorter residual coverage, more coverage units per period, or a mix shift toward shorter-duration cohorts. The RA release % tracks how fast risk is being absorbed; because the RA method is principle-based, this ratio is insurer-specific, but a sustained divergence without a methodology or risk-profile change warrants explanation.

Tracking both over time gives a diagnostic for whether the income statement is accelerating or deferring profit recognition, and whether that shift is driven by contract mix, assumption changes, or portfolio growth — rather than reading the income statement lines in isolation.

---

What the review is for

Each line of the GMM income statement traces back to a movement in the liabilities, and each driver is attributable either to service — and flows to the insurance service result — or to the passage of time and financial risk — and flows to the net investment result — or to a re-estimate of future service, which adjusts the CSM or the loss component. A driver that posts in the wrong category distorts both the result and the CSM balance, with compounding effects across future periods. Read alongside the analysis of movement from the prior post, the profit drivers are the bridge between the income statement lines and the decisions they should inform.

---

From review to engagement

If you are building or reviewing an IFRS 17 GMM income statement — in a valuation cycle, a budget process, or a model audit — work on the methodology, the engine configuration, or the review framework sits within the scope of the IFRS 17 work on this site.