Please note that it is unavoidable, given the specific industry, that we will use insurance-specific terms. We refer to the glossary at the end of this article (Appendix 1: Table A1) for the explanation of some of these terms. We will also make references to IFRS 4 and the Solvency II framework in this article. Appendix 1: Table A3 provides a comparison between IFRS 17, IFRS 4 and Solvency II on main performance indicators or metrics identified.

It is important to note that the European Union based insurers in our sample have to apply IFRS as endorsed by the European Union (EU) in their consolidated financial statements based on Regulation (EC)1606/2002 of the EU. On 23 November 2021 the EU has published a Commission Regulation endorsing IFRS 17 ‘Insurance Contracts’, albeit with a possibility to ‘carve-out’ the annual cohort requirement of IFRS 17.22 if at least certain conditions are met. As a consequence, it is possible by making use of the exemption (carve-out) to not group contracts within a 12 month period boundary (as required by IFRS 17) but to extend this period based on the specifics of the identified insurance portfolio. This could affect the timing of the recognition of the contractual service margin over the coverage period and the assessment whether or not a group of insurance contracts is (at inception) onerous or has (subsequently) become onerous. We will explain further details, including the actual use of the EU-carve-out in our sample, in section 5.2.3. Our sample also consists of three UK based insurance companies. In the UK IFRS 17 has been endorsed without any exemption.

To close this section on transitioning, it should also be noted that the insurers in our sample decided to use the temporary exemption in IFRS 4 that allows insurers to change from IAS 39 ‘Financial Instruments and Measurement’ to IFRS 9 ‘Financial Instruments’ in the first year’s application of IFRS 17. This temporary exemption is only allowed for entities whose activities are predominantly connected with insurance and is introduced in IFRS 4 to address concerns about the additional accounting mismatches and volatility in profit or loss that might arise when IFRS 9 rather than IAS 39 is applied in conjunction with IFRS 4. This means that the insurers in our sample as per reporting year 2023 change not only from IFRS 4 to IFRS 17 but also from IAS 39 to IFRS 9. In the remainder of this article we will focus on the consequences of the transition towards IFRS 17. Especially in section 6 we will also pay attention to IFRS 9 in the context of reducing ‘accounting mismatches’.

Insurers have been required to restate the balance sheet as of the beginning of the annual period immediately preceding the date of initial application according to the new standard. For the insurance companies in our sample this means that this transition date is January 1, 2022, being the start of the comparative annual period in the 2023 financial statements. This implies relevant challenges, since IFRS 17 requires to evaluate the insurance contracts at this date of transition. For long-standing insurance contracts, data availability is a real challenge. However, IFRS 17 sets out three approaches for determining the CSM at the transition date:

• the full retrospective approach (FRA) in case the full data set is available;
• the modified retrospective approach (MRA) that permits a set of simplification in case the FRA is not applicable;
• the fair value approach (FVA), based on the fair value of a group of insurance contracts.

The three different transition options are likely to lead to a different estimate of the CSM at the transition date. Therefore, the impact on shareholders’ equity upon transition also depends upon the transition approach adopted by the insurer. Given the degree of flexibility permitted by IFRS 17 (in case FRA is not applicable), insurance company’s choices play an important role in the approach chosen (FVA versus MRA) and in the related methodology adopted. The choice could be driven by a preference to show a large CSM at transition date since it is representative of the company embedded profits not yet earned. On the other side, the insurer may prefer to reduce the negative impact on equity arising from a large CSM as per transition date which in normal conditions can be achieved by applying as much as possible the FRA.

Furthermore, we also observed that some companies have used FRA for insurance contracts initially recognised from 2016 onwards (which links directly to the introduction of Solvency II) and either MRA or FVA for insurance contracts entered into prior to 2016. Non-life insurers disclose the majority of groups of insurance contracts under the FRA due to the short-term nature of their business and the availability of historical data. Table 1 gives a summary of the approaches used by the insurers of our sample.

As reported in Table 1, several large insurers adopted different approaches for different types of insurance contracts. In conclusion, given the material impact of the discretional choices involved, a proper comparison of different insurers is not straightforward and may require in-depth analyses of approaches used for different types of insurance contracts.

Based on the sample selected, the majority of the insurers reported a drop in the equity at transition date, while only few reported an increase in equity. In particular, several insurers reported an equity drop larger than 10%.

For a more detailed overview of the reported changes in equity, we refer to Appendix 1: Table A4. Though not easy to determine within the annual reports, we have analysed the most important factors affecting the impact on equity as a result of the transition at transition date. We refer to Table 3. With the term one-off we mean that the factor is related to a choice that is only available at transition date.

Considering the several aforementioned effects, it is not possible to provide a common pattern for all the insurers in the sample. Even a single factor can impact insurers in the sample differently depending on the IFRS 4 approach used before transition date.

IFRS 17 provides a detailed set of disclosure requirements specific to the transition impact as per transition date. In summary an entity shall provide:

• a reconciliation of the CSM and the amount of insurance revenue separately for insurance contracts evaluated using the MRA, insurance contracts evaluated using the FVA and all other insurance contracts;
• the methodology/approach used to measure the groups of insurance contracts using FRA, MRA or FVA;
• a disaggregation between insurance finance income/expenses presented in profit or loss (P&L) and other comprehensive income (OCI).

Where the insurer has chosen to adopt the MRA and/or FVA, we have observed differences in disclosing the level of aggregation of the portfolios of insurance contracts for which the transition approach has been applied. Some insurers have provided a breakdown of key insurance KPIs (CSM, insurance revenues, etc.) between the different lines of business, with a clear indication as to what transition approach has been used for each portfolio. Others have shown a fairly high-level overview of these KPIs simply split by transition approach, with no further indication of which transition approach has been adopted for the respective portfolios.

Where the FRA has been used in favour of the MRA or FVA, disclosures of the key assumptions and simplifications applied in the valuation are particularly relevant to understand the process used by the insurers to measure their portfolios on a retrospective IFRS 17 basis, especially for contracts that were closed considerable time prior to the transition date of IFRS 17. Long-standing insurance contracts require often expert judgement since data related to retrospective assumptions are not always available. Under the FRA, insurers should perform a retrospective market-consistent evaluation of liabilities. Expert judgements may involve different types of assumptions, such as retrospective interest rate curves (in the former local currency for contracts issued prior to the Euro introduction), retrospective inflation curves, historical profit sharing, acquisition costs, and retrospective underwriting assumptions.

We have observed that not all insurers in our sample provided a clear and complete disclosure about the methodologies applied to set these assumptions used to evaluate contracts under the FRA method. Nevertheless, it is not possible to provide a detailed breakdown given the large number of possible expert judgements and the different situations of the insurers (in terms of contract issue years and product types).

With regard to the impact, we highlight the following insurers’ disclosures that provided a better-than-average information about the transition effects on equity as per transition date.

Figure 1.

Allianz Group Annual Report 2023, p.153.

Allianz Group provided the effect of the initial application of IFRS 17 in the consolidated statement of changes in equity, per component of equity both on 1/1/2022 and on 1/1/2023, allowing the user of the financial statements to observe the consequences of the transition per component and per date¹.

Munich RE and Gjensidige provided a quantitative disclosure of the transition effects separated per nature. Munich RE explained the table presented in Figure 2 also qualitatively and Gjensidige provided and explained a separated quantitative impact for the discount rate curve choice, risk adjustment, onerous contracts and other drivers (Figure 3).

Figure 2.

Munich RE Annual Report 2023, p. 184

Figure 3.

Gjensidige Forsikring Annual report 2023, p. 167.

As indicated in section 2, IFRS 17 includes a disclosure objective and specific requirements to achieve this objective. Part of these requirements relate to the specific amounts recognised in the financial statements for contracts within the scope of IFRS 17. We will concentrate in this section on the amount recognised as insurance liability, also to limit the scope of the empirical research to what we consider as the most important line item.

In order to assess the diversity of the accounting approaches for measuring the liability and other aspects of the disclosure of the recognised insurance liability, we considered the following aspects observed in the annual reports of our research sample:

• Liability Methodological Assumptions;
• Liability Analysis of Change;
• Insurance Contract Reporting Granularity;
• Risk Reporting and Sensitivity disclosure.

For those aspects we analysed:

• The diversity of the approaches adopted by the insurers in the sample;
• The compliance of the approaches and of the disclosure against IFRS 17 requirements;
• A best practice providing better than average information in the disclosure (if identified).

IFRS 17 gives an entity a choice to present insurance finance income or expenses either in profit or loss (P&L) or to disaggregate the income and expenses in profit or loss and OCI. This choice should be made on a portfolio-by-portfolio basis considering for each portfolio the assets that the entity holds and how it accounts for those assets. As rationale for the option the IASB reasoned that users of financial statements may find that, for some contracts, the presentation of insurance finance income or expenses based on a systematic allocation in profit or loss would be more useful than the presentation of total insurance finance income or expenses in profit or loss. During the drafting process of IFRS 17, concerns were expressed by analysts that this option could impair comparability between entities. These concerns however have not led to further amendments. It is therefore interesting to see whether or not the ‘OCI insurance finance presentation option’ (the disaggregation in P&L and OCI) has been used.

In Table 8 the outcomes are presented and in Appendix 1: Table A2 the choice can also be seen on a company by company basis.

The choice to not use the option relates to the classification of insurance contracts backing financial assets within IFRS 9. Investments in equity instruments should be measured at fair value through P&L (FVTPL) but (only) at initial recognition IFRS 9 gives companies the option to measure these instruments at fair value through OCI (FVOCI). If this IFRS 9 option is used, mismatches will be avoided if the OCI option in IFRS 17 is also used. For the same reason IFRS 9.4.1.5 allows an entity, at initial recognition, to designate financial assets at FVTPL if doing so reduces an accounting mismatch. A portfolio of bonds to be measured at FVOCI according to the classification and measurement rules of IFRS 9, could therefore be measured at FVTPL if doing so reduces an accounting mismatch. And this reduction could be realised if a company does not use the FVOCI option for presenting insurance finance income and expenses. An accounting match is then created by presenting value changes of both insurance liabilities and backing financial assets in P&L. An accounting match does not mean that amounts can be offset in the P&L but only that changes are presented in the same statement. This can cause large volatility with opposite signs in the different subtotals of P&L. An example of a company that uses the P&L matching options in IFRS 9 and IFRS 17 (hence, by not making use of the OCI presentation) is Athora. In Figure 9 a snip of their consolidated statement of profit or loss is presented.

It can clearly be observed in Figure 9 that an all-inclusive P&L approach causes volatility with opposite effects in the investment and insurance finance result. For that reason insurance companies might use other key performance indicators than the IFRS-subtotals to inform the users of financial statements. This topic will be discussed in the next section.

Figure 9.

Athora, snip of Consolidated Statement of Profit or Loss, Annual Report 2023, p. 83.

We observed that especially the larger insurance companies use an adjusted insurance performance measure as indicator of the performance of their insurance business. We refer also to Appendix 1: Table A2 where on company by company basis the results are presented.

Most of the alternative insurance performance measures are described as operating result or profit (8 times). We also see business operating profit, adjusted operating profit and underlying earnings. A reason for presenting an alternative performance measure is explained by a.s.r (see Figure 10) and considered as a best practice in line with the ESMA statements on alternative performance measures (ESMA 2015, 2022).

All ‘APM-using’ companies give reconciliations from an IFRS-subtotal to the alternative performance measure. We have seen many different reconciling items that often seems to portray in different terms similar phenomena. Generally speaking we have observed that the remeasurements of assets and liabilities caused by market price movements are excluded. But to provide the reader an impression of other excluded or included items, we present the following list: deviations from expected returns on assets, discount rate adjustments, income from derivatives, acquisition costs, integration costs, restructuring costs, losses from contracts becoming onerous, amortisation of intangibles, impairments of goodwill, release of claim provisions, foreign exchange impacts, hyperinflation accounting effects, discontinued operations, proportionally consolidation of joint ventures and associates, capital flows including funding costs. We also observed more generic terms like: ‘other one-time items or gains/losses related to corporate transactions’ or ‘accounting mismatches that are dependent on market volatility or relate to events that are considered outside the normal course of business’.

Figure 10.

a.s.r. Annual report, p. 322.

Despite the fact that reconciliations are given, it is difficult to compare insurance companies’ alternative performance measures due to the large variation in reconciling items. Thereby acknowledging the fact that reconciliations sometimes differ because the IFRS subtotal is also subject to important accounting policy choices, like the use of the OCI option, the options in IFRS 9 and the EU-carve-out option but also the choices that have been made at transition date to IFRS 17. In the ESMA statements, the ESMA requires that companies separately identify and explain the material reconciling items. We see especially in the specific explanations room for further improvement.

As a best practice of a reconciliation statement we have selected the reconciliation matrix of Zurich Insurance (see Figure 11). The APM measure is divided over the main segments and comparatives are included. In the footnotes (not presented in the figure due to readability) further explanations are given.

With regard to the qualitative explanation of the reconciling items (including comparisons and visuals) a.s.r gives a good example in a separate section (7.10) of their 2023 annual report. The combination of these practices could lead in our view to a ‘best of both worlds’.

We have also examined another important ESMA requirement: whether or not the APMs and the related IFRS performance measures are presented with equal prominence in the key figures of the annual report. We note that almost all companies (N = 12) apply equal prominence in these sections. In two cases we observed that in the ‘at a glance’ page (all important metrics in one visual on one page in the front of the report) missed the IFRS subtotal but included the related APM.⁵

In the period preceding IFRS 4 almost all the insurers used to present the operating result with the purpose to provide an adjusted result. That this practice continues under IFRS 17 in half of our sample is interesting given the fact that IFRS 17 introduced several accounting items with the objective to provide faithfully the financial performance of the insurance activities, like:

Figure 11.

Zurich Insurance, Annual Report 2023, p. 398.

• The introduction of an insurance service margin and an insurance finance margin to better reflect the profitability of a portfolio over the contractual service period and to distinguish between service and financial performance;
• The CSM that smooths profits over time and that avoids that market (under VFA model) or underwriting fluctuations affects the operating profit;
• The OCI options in IFRS 17 (and IFRS 9) that allows the company to present gains and losses due to interest rate and other market fluctuations in other comprehensive income (OCI) without impacting profit or loss;
• The allowance for an illiquidity premium on the discount curve that, for many insurers, increases the asset-liability matching thereby reducing the impact on equity/profit or loss arising from short-term market movements.

We observe that out of the 17 insurance companies that use the OCI option for insurance finance expenses or income, 6 (35%) are using an alternative performance measure related to the insurance result. Gjensidige was the only insurance company not making use of an APM while also not making use of the OCI option. This company uses the subtotals of the income statement as performance metrics to explain the insurance service and finance result. We might reluctantly conclude that not using the OCI option might be related to the fact that these companies focus primarily on the alternative performance measure and are not willing to create another performance metric through the disaggregation in P&L and OCI.

In one case we could also clearly observe that the annual cohort requirement of IFRS 17 led to an adjusted operating result, as can be seen in Figure 12.

Figure 12.

Prudence Plc Annual Report 2023, p. 250.

Users of insurer’s financial statements should be aware that the annual cohort requirement is sometimes used, sometimes not used (by making use of the EU carve-out) and sometimes used but excluded from the adjusted operating profit. We refer again for the use of this EU carve-out option to Appendix 1: Table A2.

In this section we have focussed on the insurance result performance measure. But the ESMA definition of an alternative performance is broad and contains also balance sheet metrics, operating cash flows and capital related metrics. Often insurers present several APMs together. Examples are combined ratio, operating/organic capital creation and operating return on equity. Of course, Solvency II ratios are also presented but these are not APMs because they are required by the regulator and have to be measured according to their applicable rules.

In conclusion, the market is not presenting yet a common framework of performance measures based on IFRS 17 and users of the financial statements should consider carefully how the APM is determined. We encourage insurance companies to apply the ESMA Statement on Alternative Performance Measures and corresponding guidance (ESMA 2015, 2022) on the several APMs they use.