Skip to main content

Taking stock of the financial reporting implications of our new mortality assumptions framework

A new framework for future mortality – part 7 of a 7-part series

By Richard Marshall | December 3, 2019

This seventh and final post discusses the implications of adopting the proposed new framework for an insurer’s regulatory balance sheet and other mandatory financial disclosures.
Insurance Consulting and Technology
Insurer Solutions

Understanding the various sources of profit emerging from a portfolio of life business is a complex affair. Ignoring changes in regulatory capital, if we fix the order of the analysis of surplus arising from the business, then we can think of the surplus arising from the change in mortality trends over the year as comprising:

  • the increase/decrease in liabilities resulting from the difference between actual population improvements over the year and those predicted (i.e. higher/lower survivorship), and
  • the change in the present value of the liabilities (relative to that expected) at the end of the year.

This assumes that the difference between actual mortality experience against the previous basis and that expected under that basis, as well as the change in the present value of liabilities due to the change in the base mortality rates has been determined prior to this step.

Unlock More

About the series

In this series of articles, we discuss the impediments to understanding mortality improvements and the dangers of herding and group-think in this process. In response to these issues, we introduce a new framework for the development of mortality improvement assumptions.

The materiality of mortality improvements in the calculation of the best-estimate value of liabilities means that a small change in these assumptions can result in a highly material emergence of (or obliteration of) surplus.

Volatility of this sort is appreciated neither by insurers nor their investors (where relevant). This means that an insurer has to consider how its improvement assumptions could change year-on-year before it adopts this framework for assumption setting.

In this post we look at:

  • Implications for the Solvency II balance sheet
  • Implications for the IFRS 17 balance sheet
  • Implications for disclosures under UK GAAP.

The content of this post is largely focused on financial reporting within the European Economic Area (EEA), but similar considerations will apply more generally to reporting outside of the EEA.

Solvency II balance sheet

The initial impacts of changing the approach to setting mortality improvements are potentially significant. If there is a marked difference between the resulting improvements surface and that implied by the prior approach, then this could mean a significant change in the best-estimate liabilities and capital requirements.

Ignoring reinsurance, taking an informed and carefully-reasoned view that mortality improvements would be higher for annuitants (or lower on protection business) would lead an insurer to hold higher technical provisions. However, chief actuaries might not feel it appropriate to immediately take full credit for the release of technical provisions upon forming a view implying movements in the opposite direction. Releasing technical provisions prematurely only to have to re-establish them later is something that they would be very keen to avoid.

The Solvency II rules technically require assumptions to be best-estimates; it would be inappropriate for an actuary to deliberately introduce a prudent margin into their genuinely-held best-estimate assumptions (where these were based on properly-understood credible data meeting any criteria set out by the insurer). This means that a deliberate more gradual unwinding of technical provisions (for example, to allow for potential future volatility in short-term improvements, in particular, due to the new approach taking more account of new information received each year) would technically be disallowed.

Moreover, whereas, under the old Solvency I, Pillar 1 Peak 1 reporting, an insurer could set up an additional discretionary reserve to protect against volatility, the Solvency II rules do not allow for discretionary reserving.

How then can insurers avoid introducing additional volatility into their balance sheets and potentially spooking investors when changing their assumption-setting approach?

  • One way is to make supplementary disclosures – an adjusted Solvency II balance sheet with additional reserves set up explicitly to smooth such initial and future volatility. This can usefully be combined with other adjustments which give a more investor-focused view of the regulatory balance sheet.
  • Alternatively, they could rely on Directive 2009/138/EC Article 77, which requires the best-estimate to be “based upon up-to-date and credible information and realistic assumptions”. This would involve using the subjective nature of “credibility” to allow a more steady movement towards the best-estimate assumptions and gathering further data and information to support the updated view before adopting it in full.

The latter treatment would be preferable, allowing the Chief Actuary to get comfortable with the new approach to assumption-setting and to see how reliably it predicts (at least short-term) improvements before relying too heavily upon the assumptions. Growing confidence in the approach would allow the assumptions to gradually (over a period of several years) track closer to the new best-estimate. Retaining a small margin for uncertainty within the short-term improvements could also reduce balance-sheet volatility, though such volatility might well be far less significant than the shock due to the initial change in the assumption-setting approach.

Companies can also derive a useful ORSA scenario from this work. As well as giving senior management a feel for the potential direction of travel over the next few years, showing the impact (on the technical provisions and capital requirements) of moving towards the best-estimate assumptions implied by the new approach over a few years would demonstrate to the regulator both the company’s careful approach to the introduction of new assumptions and how it is taking increasing responsibility for the development of its own assumptions, and improving its understanding and management of longevity risk in the process.

Impact on the matching adjustment

Annuity writers with a matching adjustment will also need to be careful when making liability changes; changing the reserves may result in more surplus being locked into the matching adjustment fund, or alternatively more money needing to be added to this fund. In the former case, companies will need procedures to demonstrate that a surplus has arisen and to allow this to be released.

Furthermore, changes in the duration of the liabilities may mean that the matching adjustment fund needs rebalancing to meet the matching requirements. With the potential for greater annual volatility in the assumptions, reflecting the nature of new information received each year, rebalancing may need to be more frequent even after the initial change in modelling approach.

Impact on internal model calibrations

For those insurers using an internal model, a significant change in the improvements basis (and the approach to setting this) may give rise to a need for major or minor model changes with a knock-on impact on overall capital requirements.

IFRS 17 balance sheet

On a future IFRS 17 balance sheet, an increase in the volatility of the improvements basis may be less of a concern.

Any change in assumptions resulting in a fall in liabilities will not immediately flow through to the profit and loss (P&L) accounts. Instead, it will accumulate within the contractual service margin (CSM) to be released over the remaining term of the relevant contracts.

Assumption changes resulting in an increase in liabilities would reduce the CSM (to the extent that such a CSM was available) or otherwise be recognised as an immediate loss in the P&L accounts.

This treatment leads to reduced balance sheet volatility, but the recognition of the benefit of a (positive) improvement assumption change will be (potentially greatly) deferred.

Arguably, if there were a change in the outlook for improvements with an expectation that (at least short-term) improvements would become more volatile, then there could be an increase in the risk adjustment (the same confidence measure, or Value-at-Risk could imply a wider range of possible cash-flow patterns – amounts and timings – and therefore a larger deviation from the best-estimate value of liabilities). The run-off of this risk adjustment would likely be quite different from that of the CSM.

Onerous contracts

A change in the improvement assumptions could affect whether new contracts are considered ‘onerous’.

At present, guaranteed annuity options are generally heavily in-the-money for policyholders, so annuity contracts arising from such guarantees are likely to remain onerous unless there are quite extreme changes to improvement assumptions.

For open market annuities (and, perhaps to a lesser extent, internally vesting non-guaranteed annuities). If insurers strengthened improvement assumptions sufficiently, then some annuities could become onerous during the lifetime of the policies. However, this would not affect whether the contracts were considered onerous at inception. So long as such annuities were repriced to reflect the updated improvements basis, the classification of the annuities under the onerous contracts test should not be affected.

Over time, if contracts in a cohort were to become onerous, the CSM for the cohort would be eroded or, if this reached zero, a loss would be recognised on the balance sheet.

Disclosures under current IFRS and UK GAAP rules

For companies producing accounts in accordance with UK GAAP, there is greater freedom in the setting of assumptions, though there will also typically be an expectation that frequent and unnecessary changes to the basis will be avoided. To achieve this aim, companies will typically incorporate a small margin for prudence within their reserving assumptions (per the old Solvency I, Pillar 1 Peak 1 valuation rules).

The incorporation of such a margin within the assumptions means that chief actuaries have much greater discretion about how to update these prudent reserving assumptions over time in response to the change in the best-estimate improvements. They can exercise their judgement as to how they allow for an initial change (due to the change in methodology) and set the size of the margin to apply in their reserving assumptions so as to absorb any volatility in the short-term views of improvement rates over time.

It is also possible to set up explicit volatility reserves if these are felt to be necessary in order to deal with the potential year-on-year changes in views on short-term mortality improvements in response to new information becoming available to the insurer.

Gains are worth the pain

Whatever the reporting regime, the implications of a change in the setting of assumptions needs to be properly thought through – impacts on statutory reserves, risk margins/adjustments, prudent margins, capital requirements, buffers over capital requirements and so on – before a change is made. This way, insurers can anticipate the impact of a change and manage expectations; nobody likes a surprise movement.

Overall, we believe that the benefits of a strong assumption setting framework outweigh the potential short-term pain arising from a transition between approaches.

Find out more

We hope that you have found this series of posts to be interesting and informative. Should you wish to discuss the issues considered in the series further and explore how Willis Towers Watson can support you in developing and adopting a more sophisticated approach to setting improvement assumptions, please contact your local Willis Towers Watson representatives, or alternatively contact the author, Richard Marshall, directly.



Richard Marshall is a Director in Willis Towers Watson’s Insurance Consulting and Technology business and leads the development of mortality and demographic risk models for our UK business.

Title File Type File Size
A new framework for future mortality PDF .8 MB

Related Capabilities

Contact Us