2.1 Life insurance risks

2.1.1 Definition

7. Life risk charges are applicable to life business and similar to life health business and consist of the following five sub-risk charges.

• a. Mortality risk: risk of losses due to actual mortality rates higher than expected;
• b. Longevity risk: risk of losses due to actual mortality rates lower than expected;
• c. Morbidity/Disability risk: risk of losses due to unexpected changes in the level, trend or volatility of disability, sickness and morbidity rates;
• d. Lapse risk: risk of losses due to unexpected changes in the level or volatility of rates of policy lapses, terminations, renewals and surrenders;
• e. Expense risk: risk of losses due to unexpected changes in the incidence of expense incurred.

2.1.2 ICS Methodology

8. For each of the five sub-risks listed above, stress factors are applied to the parameters underpinning the calculation of current estimates, in order to generate stressed balance sheets in line with the ICS target criteria – a probability of occurrence of 0.5% over a 1-year time horizon.
9. The calculation of all sub-risk charges are based on the following geographical segmentation, with respect to the location where risks are written:

• a. EEA and Switzerland;
• b. US and Canada;
• c. China;
• d. Japan;
• e. Other developed markets; and
• f. Other emerging markets.

10. The five Life insurance sub-risks charges are aggregated with a correlation matrix; the correlation factors are based on expert judgement whether the corelation between two sub-risks is negative, negligible, low, medium or high.
11. For each of the five sub-risks, the risk charge is calculated both with and without the impact of management actions.
12. The design of Life risks was informed by Field Testing exercises from 2014 to 2019, confidential reporting exercises from 2020 to 2024, as well as public consultations on ICS in 2014, 2016 and 2018. The calibration of Life risks was supported by targeted data collections in 2016, 2018 2019 and 2022 (all risks), as well as 2015 and 2017 (Morbidity/disability risk).

2.1.3 Calibration of risks

13. To inform the calibration of Life risks, supplementary data collections were run in 2018, 2019 and 2022, together with the Field Testing / confidential reporting data collections. In the context of those supplementary Life data collections, Volunteer Groups were requested to provide data covering at least 10 consecutive years (in 2018 and 2019), and 13 consecutive years (in 2022).

2.1.3.1 Mortality risk

14. The calculation of the mortality risk charge is based on a scenario of general permanent increase in mortality rates.
15. Based on a review of existing risk-based supervisory frameworks, and after field testing different factors between 10% and 15%, the stress factor was provisionally set at 12.5% for all regions.
16. In 2018 and 2019, data were collected from IAIGs in order to inform a potential refinement of the calibration. Based on those data and on the assumption that the variable $\dfrac{A}{E}-1$ (where $A$ and $E$ respectively denote actual and expected claims, due to all causes, incurred in each financial year) follows a normal distribution (for which observations are independent and identically distributed), the following 99.5th percentiles  were calculated for mortality rates:

• a. 6% for business written in Japan;
• b. 15% for business written in China; and
• c. 10% for business written in other developed markets.

17. Data for the Japanese market were considered sufficient evidence, with over 10 years of history; therefore, it was decided to set the stress factor for business written in Japan to 10% (including a prudence margin based on expert judgement). The appropriateness of that level was confirmed by data collected in 2022.
18. For business written in China, the 2019 and 2022 data collections provided evidence of a risk factor between 13% and 27%; the risk factor was therefore set to 15%.
19. For other developed markets, data collected in 2022 supported a 12.5% risk factor.
20. For US and Canada and other emerging markets, the lack of historical data provided in all data collections did not allow for an adjustment of the calibration. Therefore, based on expert judgement the stress factor was aligned with other developed markets (12.5%).
21. For EEA and Switzerland, 10 years of historical data collected through the 2019 data collection showed that a 12.5% risk factor was appropriate. Data collected in 2022 continue to support this conclusion.

2.1.3.2 Longevity risk

22. The calculation of the longevity risk charge is based on a scenario of a permanent decrease in mortality rates.
23. Based on results of data provided by IAIGs for the calibration exercise in 2016, a stress factor of 17.5% across all regions was specified  for the 2017 Field Testing. In the absence of any evidence that this level of stress is inappropriate, the 17.5% factor has remained unchanged since then.

2.1.3.3 Morbidity and Disability risk

24. For the calculation of the morbidity/disability risk charge, the business is split: (resulting in eight segments (4 types of guarantee and for each type, 2 terms of contract).)

• a. By type of guarantee:
  • i. Medical expenses (Category 1)
  • ii. Lump sum in case of a health event (Category 2)
  • iii. Short-term recurring payments (Category 3)
  • iv. Long-term recurring payments (Category 4)
• b. By term of contract:
  • i. Short-term (up to 5 years)
  • ii. Long-term (more than 5 years)

25. For segments in categories 1 to 3, stress factors are applied either to inception / recovery rates (when those are explicitly used for the modelling of claim costs) or otherwise directly to the expected amount of future claims. For segments in category 4, stress factors are applicable to the assumptions underlying the calculation of future claims (inception rates and recovery rates).
26. The distinction between short-term and long-term contracts was introduced to reflect that claim amounts over a short period are likely more volatile than over a longer period.
27. All stresses are assumed to occur simultaneously, therefore risk charges for those four categories are added together.

2.1.3.4 Lapse risk

39. The calculation of the lapse risk charge involves three different scenarios:

• a. A permanent increase in future lapse rates (level and trend, up);
• b. A permanent decrease in future lapse rates (level and trend, down); and
• c. An immediate surrender of a fraction of the in-force policies.

40. For the two scenarios of the level and trend component, an initial placeholder calibration (40% relative increase / decrease in the current estimate lapse rates) was determined based on expert judgement. Based on a Gaussian assumption for the distribution of the variable $\dfrac{A}{E}-1$(ratio of actual to expected number of lapses incurred over the financial year), 10-year historical data submitted by IAIGs in 2018 provided evidence that the 99.5th percentile of lapse rates for business written in Japan should be between 12% and 20%; this evidence was confirmed by the data collected in 2019. Therefore, the stress factor for the level and trend component for business written in Japan was revised to 20%. The data collected in 2019 and 2022 confirmed the appropriateness of the stress factors for Europe, Japan, China and other developed markets. No sufficient data was obtained for other regions.
41. The mass lapse scenario stress factors are 30% (retail policies) and 50% (non-retail). These were determined based on expert judgement after a review of various jurisdictional solvency regimes including those from Solvency II, South Africa, Australia, Canada, Singapore and Japan.
42. Depending on the portfolio structure, either the mass lapse or the level and trend scenario could be the relevant 0.5% probability scenario for a given IAIG. Therefore, the maximum between those two scenarios is retained as the lapse risk charge. For the same reason, the higher of the level and trend up and level and trend down results is retained as the level and trend component.

2.1.3.5 Expense risk

43. The calculation of the expense risk charge involves two simultaneous scenarios:

• a. An increase in the unit expense amount by a factor $x$;
• b. An increase in the level of inflation of expenses by a factor $y$.

44. An initial placeholder calibration of those two scenarios was chosen by expert judgement and a review of this calibration was undertaken in 2017. This took into account:

• a. The responses to the 2016 ICS public consultation;
• b. 9 to 12-year historical data submitted by IAIGs in 2016; and
• c. The Generally Recognised Expense Table (GRET) study published by the Society of Actuaries in the US, based on data from 2008 to 2014.

45. The review concluded that:

• a. A permanent 3% stress on inflation for China and emerging markets is not realistic, and should be graded down over time;
• b. The order of magnitude of the 𝑥 factor is 10%; and
• c. The global level of 99.5% stress (unit expense + inflation) in the US should be approximately 16%.

46. In addition, a review based on Field Testing data and conducted in 2018 showed that a 6% factor for unit expense risk in Japan was appropriate.
47. Based on those considerations, it was decided to introduce a downgrading of the inflation stress after year 10 for Other developed markets, China and Other emerging markets. Other factors were considered globally appropriate. Data collected in 2022 did not provide evidence of calibration inappropriateness.