DYNAMIC SOLVENCY TESTING (AGN7) UPDATE PREPARED BY: KEN TANG, HSBC INSURANCE

Transcription

1 DYNAMIC SOLVENCY TESTING (AGN7) UPDATE PREPARED BY: KEN TANG, HSBC INSURANCE

2 BACKGROUND & PURPOSE OF THIS UPDATE DST is an important tool for Appointed Actuaries to ensure the Insurer s financial condition continues to be healthy, and to identify and mitigate any threats to solvency of the business. The AGN7 requirement first came into effect from The last revision of the guidance was 4 years ago (2011). There is a need to refresh the guidance in order to keep the requirements up-to-date. For instance there is increasing emphasis on new/emerging risks (e.g. operational risk) so the guidance is expanded to cover the testing of such additional scenarios. Coincidentally there was an IMF review earlier in 2015 on the regulatory environment in HK, as part of the Financial Services Assessment Program (FSAP). With reference to DST their specific comments were (i) whether the solvency requirement for ILAS products is sufficient (ii) mortality stress appears to be too light. We are largely keeping most of the guidance unchanged (only minor revisions). The revised guidance has run through ASHK Life Committee and ASHK Professional Matters Committee. Next step will be circulation to ASHK members for feedback.

3 SUMMARY OF EXISTING REQUIREMENTS DST examines the effect of various plausible adverse scenarios on the Insurer s capital adequacy, over a projected period of 3 years. It includes 1 base scenario (the best estimate) 6 simple prescribed scenarios (independent shocks in mortality/morbidity, lapse, interest rates up and rates down, High and low growth in sales) 3 additional plausible adverse scenarios 1-year short term shock scenario associated with epidemic outbreak Medium term inflationary scenario Medium term deflationary scenario Financial condition is considered satisfactory if Under base scenario, minimum capital requirement is met (i.e. solvency > 100%) Under all other adverse scenarios, insurer s assets exceed liabilities Frequency is to conduct the DST annually Ripple effects (e.g. mgt actions, p/h actions) are recommended to consider in the projections

4 The testing scenarios A base scenario (the best estimate) (No change) 6 simple prescribed scenarios (independent shocks in mortality/morbidity, lapse, interest rates up and rates down, High and low growth in sales) (No change) 5 additional plausible adverse scenarios 1-year short term shock scenario associated with epidemic outbreak (Slight Revision in stress parameters) Medium term inflationary scenario (No change) Medium term deflationary scenario (No change) Scenario considering occurrence of severe operational incident (New Scenario) Scenario considering occurrence of severe counterparty defaults (New Scenario) Frequency is to conduct the DST annually (No change) Ripple effects (e.g. mgt actions, p/h actions) are recommended to consider in the projections (Added a paragraph to elaborate the importance of factoring in p/h actions)

5 Change #1-1-year short term shock scenario associated with epidemic outbreak (Slight Revision in stress parameters) 15% deterioration in mortality rates in first year Changed to Uniform excess death of 0.75 per thousand at all ages 25% shocks to equities and real estate investments 80% of projected plan sales volumes 5% addition / reduction in lapse rates

6 Change #2 New scenario considering occurrence of a severe operational incident The Actuary should consider the resilience of the business to operational incidents, particularly for classes of business which have a relatively low minimum solvency requirement under the ICO, in particular Class C linked business. If Class C business is a significant product type for the company, then at least one unit linked operational incident scenario should be included in the DST report, with an aggregate loss amount not less than 1% of funds under management. Some of the most common operational incidents giving rise to significant financial loss for insurers are as follows:- For all classes of business mis-selling resulting either in a regulatory fine or a need to compensate affected policyholders For all classes of business business operation failure which may result from a number of causes including system failure, inappropriate system input, internal or external fraud, etc. For linked business only - unit pricing errors which require policyholders who have lost out financially to be compensated without recovering funds from those that have gained from the error. For linked business only investment allocation error which causes the fund s assets to no longer be reflective of the investment objectives in the fund brochure, with a result that the fund underperforms against customer expectations.

7 Change #2 New scenario considering occurrence of a severe operational incident The range of operational incidents is quite wide and by nature hard to predict. Judgment is expected for actuaries to determine the most relevant scenario considering the particular circumstances of the company. Examples to be tested can include - Unit pricing errors of inforce unit-linked policies leading to a loss of 1% of AUM Operational failures causing business interruptions, leading to a loss of revenue from a key business line without significant reduction in costs for a quarter Mis-selling identified on a major product leading to customer redress payments on entire block of policies (e.g. premium refund plus interests to all inforce customers)

8 Change #3 New scenario considering occurrence of severe counterparty defaults The Actuary should consider the resilience of the business to counterparty default events. An example would be an economic recession triggering defaults of major corporate bond issuers, reinsurers and derivative counterparties. Some guidance and considerations that the Actuary may consider for bond defaults include: The Actuary may consider the impact of defaults on the company's fixed income portfolio equivalent to 0.5% market value of its investment grade bonds (excluding AAA rated bonds and highly rated sovereign bonds) and 3% market value of its non-investment grade bonds, or default on the largest bond issuer of the company's fixed income portfolio, whichever impact is higher. These bond default events typically occur during economic downturn, where there will be shocks in other asset prices as well (e.g. equities and other derivatives). Where appropriate the Actuary should consider testing the compound impacts of these to the entire portfolio of the business. When a bond defaults, one can expect a portion of the principal and accrued interest can be recovered. The Actuary should apply judgment on the assumed recovery rate, but typically it should not be greater than 50% if the event is assumed to happen as a result of an economic recession. On the other hand, some considerations that the Actuary may consider for reinsurer defaults include: The Actuary should test the impact of the default of a reinsurer which the company has the most significant business with. The impact analysis should take into account any collateralization or other guarantee mechanisms provided by the reinsurer under specific reinsurance agreements. An example would be coinsurance on RMB-denominated products.

9 Change #4 a paragraph added to further elaborate the importance of policyholder actions Assumed policyholder action Under certain adverse situations, the Insurer may suffer further losses due to action from policyholders. In particular, in high interest rate scenarios, products with higher cash values, e.g. universal life products, are more likely to experience higher policyholder disintermediation. This could be due to the fact that many policyholders' financial situation will be adversely affected by higher interest rates because they may have mortgage loans and/or other liabilities or because the health of the economy may be adversely affected with impacts on unemployment, etc. Furthermore, when interest rates rise, some policyholders may surrender their policies in search of higher returns elsewhere. This situation may be exacerbated by the fact that in some cases the cash used to purchase the insurance product has been acquired through premium financing. The losses that the Insurer may suffer as a result of the above should be considered within the projections. For products that have a material savings element, it is common to model an add-on to the rate of lapsation which is a function of market interest rates less the rate of return to policyholders expected under the contract. It may be helpful under adverse scenarios to report results with and without assumed policyholder action.

10 THANK YOU R E S T R I C T E D