IFRS 17 revisited

…which made comparison and interpretations of numbers between companies problematic. IFRS 17 has delivered consistency, and delivered on the following items:

·      Current assumptions: where under IFRS 4 the use of locked-in or past pricing assumptions was allowed, IFRS 17 uses current assumptions.

·      Stochastic valuation/valuation of guarantees: under IFRS 4, requirements relating to the valuation of guarantees or profit sharing were limited to a liability adequacy test. This did not consider the full scope of probable future scenarios. IFRS 17 prescribes sensitivity analysis, providing a more accurate picture of insurance liabilities.

·      Risk-based thinking: IFRS 17 stimulates risk-based thinking due to the valuation and disclosure of sensitivity analysis, the requirement to aggregate products with similar risks into portfolios, and the explicit component of the risk adjustment.

·      Cooperation between actuaries, accountants, and other finance functions: IFRS 17 has clear links with other accounting standards like IFRS 9 and 15, and IAS19 and therefore requires interaction between finance functions. This is especially true for contracts under the Variable Fee Approach (VFA), where there is a clear link between assets and liabilities.

·      Contractual Service Margin (CSM): IFRS 17 introduces the concept of CSM, which reflects the unearned profit on the contract, released over the lifetime of the contract. The CSM thus prevents profit to be recognized at the moment of sale and profits will be released when services are provided which, in my opinion, better reflects the profit during the lifetime of policies.

The Bad

A framework can never be perfect, and with the benefits above come some less beneficial implications:

–               Copy paste Solvency II?: as IFRS 17 is a market-consistent framework, the option to use existing Solvency II assumptions, methodologies and processes is tempting. Such an approach reduces the additional benefits and insights IFRS 17 can provide. Could IFRS 17 become ‘Solvency II plus P&L’, and become just another market- consistent metric?

–               Are the standard approaches under IFRS 17 really standard in practice?

•               Transition: The standards prescribe the full retrospective approach to be the standard transition method. However, this requires past information such as assumptions on expenses, lapse rates, and mortality and morbidity rates. For most companies, applying the full retrospective approach is not feasible, which begs the question whether it makes sense to have chosen this approach as the standards’ starting point.

•               Measurement model: although the General Measurement Model (GMM) is the standard model for valuing contracts, in certain cases where the interpretation is not clear-cut (for example life products with discretionary profit sharing), companies can choose the Variable Fee Approach (VFA) over the GMM, which allows for more stable CSM and Profit and Loss (P&L) patterns. It therefore might be tempting for companies to apply VFA thereby increasing their hold on future P&L.

–               Asymmetric treatment gains and losses: due to the asymmetric treatment of profitable contracts (profit recognition deferred through CSM) and loss-making contracts (loss recognized immediately), there might be a tendency for firms to modify portfolios, allocate assets, or apply measurement models to avoid a loss component. The most immediate impacts of these actions will be seen at transition, but the asymmetry will also impact future new business at inception and subsequent measurement and may cause ‘accounting-driven’ management actions further down the line.

–               Lack of quantitative steering: like other IFRS standards, IFRS 17 is a principle based framework. As a result, the classification of contracts, application of measurement models and assessment of materiality are based on qualitative and not quantitative assessments. This approach forces users to think about the framework and, perhaps, better understand its intentions, but it can also result in different interpretations which might result in inconsistency, making it harder to compare results.

•               Subadditivity not always guaranteed: similar to Solvency II, classifying products into homogeneous groups might result in different results for an identical underlying risk profile. Consider the example of the Non-Distinct Investment Component (NDIC) for an endowment product. Economically, the product consists of a pure endowment plus a term assurance product for a single policyholder. The NDIC is the amount that has to be paid to the policyholder under different circumstances or insurance events.

A common method to calculate the NDIC is to consider the minimum payment at surrender, death, or maturity. However, this would result in a different NDIC value depending on whether the endowment is considered as a whole, or as the sum of its two separate components:

Download Table 1 at the end of the article

In other words, although the total risk profile is identical, the total NDIC is not.

Asymmetric treatment of gains and losses: similarly, the classification of policies into similar risks managed together may also result in different outcomes for identical risks. In the case of endowment, it is common for insurance companies to profit from the mortality component, and incur losses on the longevity component:

Download Table 2 at the end of the article

As, under IFRS 17, losses have to be taken into account immediately, while profits are spread over the lifetime of the policy, the two separate policies (for a single policyholder) would results in a loss at the start of the contract, while for the combined endowment product, there is no P&L impact at inception.

The Ugly

Finally, we highlight a few causes for concern:

Firstly, dual reporting (comparative figures) required during 2022 has put companies under additional pressure – and all to produce parallel results in two bases (IFRS 4 and 17) which are very difficult to compare. An impact study approach more like Solvency II, where different studies were performed before implementing, would have resulted in a more gradual implementation, and provided an opportunity for companies, investors and auditors to get on the same page.

Secondly, the guidance has clear shortcomings with regards to expenses and inflation. In the disclosures, expense results are included in the operating result line, which also includes results from demographic actuarial assumptions, making the two hard to disentangle. In recent years, lower asset returns, increasing longevity, and automation have placed more focus on expenses and operational efficiency, especially for life insurers. A more in-depth section on expense result would therefore have been a valuable source of information for investors and other stakeholders.

The standard gives limited guidance on inflation. For example, for interest rates, there is guidance on deriving the interest curve either via bottom-up or top-down approach, while for inflation no specific guidance is provided. In hindsight, when the standard was drafted, inflation was not a big topic, and hence providing guidance on nominal vs real rates and the use of inflation indices as market assumptions was sufficient. However, as the inflation environment has changed dramatically in the last year, additional guidance on longer term inflation would have been helpful to ensure industry-consistent application. ■