本稿では、この不確実な環境が、保険会社の経済シナリオ、資本要件、ストレステストおよびシナリオテストに与える潜在的意味について調査します。
In this paper, we explore potential implications of this uncertain environment on economic scenarios, capital requirements and stress and scenario testing for insurers.
Background
Since the global financial crisis (“GFC”) of 2007 and 2008, we have seen an almost continuous decline in interest rates, causing financial institutions to focus on the low interest rate environment and its implications for areas such as valuation, capital management and risk management. In 2022 this all changed. Rising inflation forced central banks around the world to increase their rates, materially impacting insurers’ balance sheets and risk profiles.
The first two papers focussed on how increased interest rates impact the level of liability discount rates,1 and what the implications are for interest rate hedging.2 In this third and final briefing note we look at the impact of increased interest rates on the calibration of economic scenarios, the assessment of the appropriateness of the Solvency Capital Requirement (SCR), scenario and sensitivity testing and policyholder behaviour.
What happened to the interest rates?
In our first briefing note we analysed the development of interest rate movements for various maturities, from 2003 onwards. In short, we observed the following:
- Over 2022 there were substantial increases in interest rates. These increases were similar in size to those in the years preceding the GFC but this time the increases took place at a significantly more rapid pace.
- There was a flattening of the yield curve, extending into a significant inversion, which amplified the inversion that had already started halfway through 2019.
- A cluster of volatility occurred that was very similar to the GFC and its aftermath in 2009 but this time driven by rate increases rather than decreases.
Overall, the increased and inverted curve has many implications that insurers should be aware of, several of which we outline in more detail in this paper.
Economic scenarios
Economic scenarios used for risk-neutral valuations, to value the liabilities for products with financial guarantees, depend on the level of the risk-free rates (e.g., the central economic scenario) and (implied) market volatility. The low interest rates and low volatility in the years preceding 2022 caused the returns in these economic scenarios to remain low as well.
As a result, the time value of options and guarantees (TVOG) was very low over the past number of years for many insurance companies offering products with financial guarantees. In the base scenario, options and guarantees were far out-of-the-money for many insurers as the base risk-free curve was materially below the guaranteed rates (or ”strikes”) provided.3
The same held for most economic scenarios underlying the TVOG valuation. Returns in scenarios calibrated were too low for options and guarantees to bite at the guaranteed rates and, as such, the probability of these products returning to being in-the-money (i.e., the time value) was also close to zero for many insurers.
This all changed in 2022. Increasing risk-free rates and interest rate volatility materially impacted Economic Scenario Generator (ESG) calibrations, resulting in a wider, and higher, range of economic scenarios. In the base scenario, for many insurers options and guarantees are still out-of-the-money or at-the-money. Several scenarios now however contain returns sufficiently high for options and guarantees to return to being in-the-money, leading to more material time value outcomes.
Figure 1: EIOPA risk-free curves and ESG 99.5 percentiles
This change in economic scenarios, is also shown in Figure 1. The 99.5-percentile corresponding to a set of economic scenarios per fiscal year (FY) 2021 exceeds the risk-free rate in that year by around two percentage points. One year later, the 99.5-percentile now exceeds the risk-free rate by almost four percentage points—a material impact, especially given that the underlying risk-free rates already increased significantly.
It is likely that, for many insurers with options and guarantees in their books, the TVOG increased materially between FY2021 and FY2022. Where some insurers previously did not have to focus a lot of attention on the output from stochastic models due to materiality, they may not have this ”luxury” anymore. In the new interest rate environment, the methodology, assumptions, results and analyses surrounding the TVOG, as well as the economic scenarios feeding into the calculations and their calibration, will require a lot of attention again to ensure they are appropriate.
Policyholder behaviour
Related to the TVOG, the change in the moneyness of options and guarantees might also impact the dynamic policyholder behaviour (DPB) of an insurer’s portfolio.
DPB reflects the fact that a policyholder’s propensity to exercise options available in a life insurance policy can be influenced by external factors, economic conditions being one of them.
Examples of options which can be impacted by DPB on traditional life insurance products include:
- Early guaranteed lapse and surrender
- Guaranteed annuity options
- Option to pay additional premiums on guaranteed terms
- Option to extend the original policy term on guaranteed terms
Assumptions relating to these options might need to be revised and, in cases where some of these options are not modelled explicitly due to materiality, they might need to be reconsidered going forward, especially if DPB is expected to change materially due to the new economic conditions.
Moreover, proposed changes in the Solvency II 2020 review could mean that DPB modelling becomes the default method for most insurers, leading to a considerable expansion of DPB modelling.4
Standard formula appropriateness
Solvency II provides an extensive set of harmonised rules for the assessment of solvency, risk management and reporting of European insurers. The calculation of the Solvency Capital Requirement (SCR) is part of this.
Many European insurers use the standard formula (SF) to calculate the SCR, which aims to capture the risk that an average European insurance company is exposed to. Interest rate risk under the SF is represented by two separate stresses, both representing an instantaneous 1-in-200-year event: an increase in interest rates and a decrease in interest rates.
Per FY2021, the former was represented by a flat one percentage point increase in risk-free rates. With respect to the latter, the SF stress is determined relative to the risk-free rates and zeroised when these rates are negative. With risk-free rates per FY2021 being negative for maturities up to six years and low afterwards, the resulting downward stress was relatively small.
The SF is, however, not perfect and may not be appropriate for all insurance companies. In fact, all insurers using the SF need to assess whether their risk profiles deviate from the assumptions underlying the SF and whether these deviations are significant.5 This assessment of SF appropriateness is performed on an annual basis as part of the annual Own Risk and Solvency Assessment (ORSA) process.
Figure 2: EIOPA base curve (0%-baseline) and EIOPA standard formula up stresses (all per FY2021), and realised increase in rates FY2022
Figure 2 shows the prescribed SF upward stress of one percentage point, the upward stress as proposed in the Solvency II 2020 review, and the actual movement in risk-free rates between FY2021 and FY2022 (the ”realised stress”), in absolute values compared to the FY2021 European Insurance and Occupational Pensions Authority (EIOPA) curve. Herein lie two potential problems:
- Because rates increased more than two percentage points for all maturities, the SF stress (a one percentage point interest rate stress), occurred at least twice over the course of 2022 alone, making it much harder to justify that the SF indeed represents a 1-in-200-year stress.
- In the current framework, calibrations6 underlying the SF interest rate stresses do account for future changes in the shape of the curve, similar to the inversion of the curve observed in 2022. However, simplifications in the SF result in it being largely ignored, and therefore the SF stresses are near-parallel.
The above problems make it increasingly difficult for insurers to demonstrate the appropriateness of the SF interest rate stresses, especially where interest rate risk is a material element of their market risk exposure. Insurers therefore may face challenges when performing their SF appropriateness testing in this year’s ORSA.
A few things to keep in mind when performing this assessment are as follows:
- As a first step, insurers may perform a qualitative analysis and, if that indicates that the deviation is not significant, then a quantitative assessment is not required. This might be the case for insurers with low exposure to interest rate risk.
- A quantitative assessment will likely heavily depend on expert judgement. Interest rate stresses are typically derived using historical market data going back many years. Adding the observed 2022 interest rate movements to this data (e.g., one additional year of data) will likely not lead to massively different stresses. Another way to consider this is to try to back-solve for what level of interest rate stress would result in a materially different SCR (based on the insurer’s own definition of materiality). The insurer could compare this to its expectations of a 1-in-200-year event based on recent market movements and expert judgement to assess whether the SF is still broadly appropriate on the basis of materiality.
- Even if an insurer concludes the SF is still appropriate for use, one might consider allowing for the new interest rate environment when assessing their Own Solvency Needs within the ORSA.
Internal models
SF appropriateness is also important for companies that have an internal model (IM) in place, either full or partial. The starting point of each IM is the appropriateness of its scope. The recent change in interest rate environment could potentially require insurers that use an IM to (re)assess this scope.
Per FY2020, the median interest rate up stress for UK insurance companies using an IM to determine the SCR for interest rate risk was two percentage points. Although the median IM stress is significantly higher than its SF counterpart shown in Figure 2 above, observations similar to the ones made for SF appropriateness in the previous section apply here as well.
Interest rate risk modules need to be reassessed and recalibrated, potentially using additional expert judgement. Because increased interest rates and volatility might also have impacted correlations with other risk exposures (especially market risk and lapse risk), the same holds for the aggregation sub-module. This all might lead to an increased overall SCR, impacting the solvency coverage ratio and risk management activity of an insurer.
IM companies that do not yet determine their SCR for interest rate risk using an IM are also likely to be affected by the new interest rate environment. The (re)assessment of the SF appropriateness for interest rate risk might even lead to the inclusion of both the market risk and aggregation modules in the scope of the IM. This will have material consequences on the capital requirements, risk management and operations of an insurer.
Stress and scenario testing
Interest rate risk is integral to stress and scenario testing in risk reports such as the Solvency and Financial Condition Report (SFCR) and ORSA.
The low interest rate environment gave insurers little reason to reconsider their interest rate stresses on a regular basis. As a result, many insurers were using a fixed set of interest rate scenarios over the past few years. This allowed for comparability of outcomes between years, whilst maintaining an appropriate representation of their interest rate risk profiles.
The new interest rate environment will likely force insurance companies to reconsider their interest rate scenarios. There are several areas insurers could consider when determining their new scenarios. Examples are:
- Level: What is an appropriate increase or decrease in rates to represent the insurer’s risk profile?
- Shape: Are there changes in the shape of the curve that need to be considered?
- Central bank policy: Are there policy changes, such as rate hikes anticipated, that should be reflected in the scenarios?
- Liquidity: Does the new interest rate environment impact the future liquidity of the fixed-income market, impacting the way the curve is constructed?
- Ultimate Forward Rate (UFR): Can the UFR be considered fixed over the next few years or is there reason to make an alternative assumption?
- Solvency II 2020 review: Should the scenarios already account for future changes due to the Solvency II 2020 review?
- Management actions: Are there any new management actions as a result of the new interest rate environment that need to be considered?
Conclusions
The new interest rate environment creates a lot of uncertainty and insurers will need to refocus their attention to interest rate risk as a result. This could have implications on insurers’ balance sheets and risk profiles, especially for those with material interest rate risk exposures.
Not only do insurers need to be able to explain movements in reported results to stakeholders, they also need to evaluate their balance sheet management and how to best mitigate interest rate risk.
In this and the previous papers we discussed several areas that are impacted by the recent changes in interest rates. A key conclusion from these papers is that insurers now need to take a step back, analyse the impact of the new interest rate environment on their balance sheets and risk profiles and move forward with a better understanding of the potential risks associated with more volatile interest rates, now that a low interest rate environment is not the default anymore.
1 Ruissaard, M., Broens, J., Zandbergen, F. et al. (27 January 2023). The new interest rate environment: Back to normal? – Part 1. Milliman Briefing Note. Retrieved 3 March 2023 from https://www.milliman.com/en/insight/new-interest-rate-environment-back-to-normal.
2 Ruissaard, M., Broens, J., Zandbergen, F. et al. (21 February 2023). The new interest rate environment: Back to normal? – Part 2. Milliman Briefing Note. Retrieved 3 March 2023 from https://www.milliman.com/en/insight/the-new-interest-rate-environment-back-to-normal-part-2.
3 In using the term “moneyness” we view the options and guarantees provided by insurers as financial products containing a call option.
4 European Commission. Insurance & reinsurance firms – review of prudential rules (Solvency II Directive). Retrieved 3 March 2023 from https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12461-Insurance-reinsurance-firms-review-of-prudential-rules-Solvency-II-Directive-_en. .
5 Hooghwerff, S., van der Kamp, R., & Clarke, S. (June 2017). Judging the appropriateness of the Standard Formula under Solvency II. Milliman White Paper. Retrieved 3 March 2023 from https://www.milliman.com/en/insight/judging-the-appropriateness-of-the-standard-formula-under-solvency-ii.
6 EIOPA (15 April 2010). Solvency II Calibration Paper. Retrieved 3 March 2023 from https://www.eiopa.europa.eu/sites/default/files/publications/submissions/ceiops-calibration-paper-solvency-ii.pdf.