Related topic Subtopic No. Para. Your question Answer

Transcription

1 25 June 2014 Related topic Subtopic No. Para. Your question Answer Valuation V.2.5. Risk margin TP5.4 Under the risk margin transfer scenario there is an assumption that the receiving entity invests its assets so as to minimise the market risk. The Risk Margin is calculated based on projecting this minimised SCR. For firms applying the matching adjustment this results in a risk margin which is the same as for a firm which is not applying the matching adjustment (as in order to minimise the market risk SCR a matching adjustment firm would have to assume it was fully invested in risk-free assets). There is also a stated assumption that the receiving entity has enough assets to cover its technical provisions and SCR. For a matching adjustment firm, there is no mechanism to capture the fact that the technical provisions would increase without being invested in risky assets it would not come through in the SCR and there is no mechanism to allow for the increase in technical provisions to allow for the loss of matching adjustment. For the purpose of the stress tests, and without preempting further developments, insurance and reinsurance undertakings should assume that the assets eligible to the matching adjustment are transferred to the reference undertaking, giving rise to a material market risk. The SCR to be projected should then be the SCR calculated after application of a matching adjustment to the risk-free interest rates term structure and should include the spread risk capital charge Could you confirm this interpretation is appropriate? i.e. The Risk Margin is the same whether a firm is applying the matching adjustment or not.

2 SCR Spread risk (Matching adjustment) SCR In the Spread Risk module the matching adjustment is recalculated following the stress. In particular, Article 163bis of the latest draft Delegated Acts says the fundamental spread should be increase by an absolute amount calculated [ ] The technical specification do not require to reapply the floors of the fundamental spreads when recalculating technical provisions under the stress. Therefore, the interpretation in (ii) is correct. In the Omnibus 2 text, the Article 77c 2(a) defines the fundamental spread as equal to the Cost of Default plus the Cost of Downgrade, subject to a floor of 30/35% of the long-term average spread over risk free rates depending on the asset. On this basis, there are two interpretations: (i) Adjust the Fundamental Spread under the stress to the original fundamental spread (i.e. Cost of Default plus Cost of Downgrade before the floor) and then reapply the floor test. (ii) Apply the adjustment to the Fundamental Spread to the final Fundamental Spread that fed into the matching adjustment calculations after the floor in the original calculation has been applied. The second approach would mean that the floor was not operating correctly, as the technical provisions assumed in the stress would not reflect the technical provisions an entity would actually calculate if spreads were to actually increase to that level since the rate is above the floor. Our approach is therefore to assume that the floor would operate as originally intended and follow the first approach. Could you confirm whether this is a reasonable interpretation?

3 SCR Non-life premium and reserve risk SCR.9.9 " FP = The expected present value of premiums to be earned by the insurance and reinsurance undertaking for each segment for contracts where the initial recognition date falls in the following 12 months but excluding the premiums to be earned during the 12 months after the initial recognition date" Q - Should this bold section be "during the following 12 months"? No. The para SCR 9.9. of EIOPA's technical specifications is correct. SCR CAT risk SCR "... lines of business 7 and 19 as set out in Annex K in relation to contracts that cover windstorm risk and where the risk is situated in windstorm zone i of region r " "... lines of business 6 and 18 as set out in Annex K in relation to contracts that cover windstorm risk and where the risk is situated in windstorm zone i of region r " Q1 - Lines of business 18 and 19 above refer to life business in Annex K. Should this be referring to a line of business in the reinsurance section? Ditto in SCR.9.56, 9.58, 9.65 etc for flood, earthquake etc. Q2 - Where can I find more information about zone i segmentation? Are insurers allowed to define their own zones? In SCR.9.53 which section is "point (b) of paragraph 5" referring to? Q1: the lines of business do not refer to life lines of business (LoB). The annex K needs to be corrected accordingly, such that the life insurance and reinsurance obligations numbering starts at 29. The LoBs 13 to 24 are proportional reinsurance obligations which relate to the obligations included in lines of business 1 to 12 respectively. The LoBs 25 to 28 are the LoBs now numbers in the TS from 13 to 16. Q2: this information is in the Excel file named CAT Helper tab. The zones in the SF are prescribed, therefore the insurers cannot define their own zones when using the SF. SCR Non-life premium and reserve risk SCR 9.9 (SCR 9.9) (page 252) premium risk : Should premium measures be gross or commission and other acquisition expenses? The premiums definition refers to earned premiums after deduction of amounts recoverable from reinsurance contracts. The commission expenses and other acquisition expenses would already be included in the earned premiums.

4 SCR CAT risk SCR 9.46 (SCR 9.46) (page 262). The capital requirement for windstorm is noted as being without deduction of amounts recoverable from reinsurance contracts. This is also the same for other Catastrophe risks. Is this correct? recoveries? Note section 9.49 mentions reinsurance contracts allowing for reinstatements. Please can you clarify whether the capital requirement should include/exclude reinsurance The loss (gross loss) is without deduction of amounts recoverable from reinsurance contracts. However when calculating the capital requirement (change in basic own funds), undertaking should include reinsurance. SCR CAT risk SCR (SCR 9.43) (page 261) Pwindstorm refers to SCR9.59 under earthquake. is this correct? Is this only relevant for non-eu exposures? Windstorm should refer to SCR9.56. SCR CAT risk SCR (SCR 9.53) (page 264) refers to point b) of paragraph 5. Which paragraph is this referring to? SCR CAT risk SCR (SCR 9.54) (page 264) This refers to paragraph 6. Which paragraph is this referring to? SCR CAT risk SCR (SCR 9.57) (page 265) The loss for windstorm other refers to SCR9.59 and SCR9.33 which are Earthquake SCR and NL Lapse Risk. Are these correct? SCR CAT risk SCR (SCR 9.67) (page 268) The definition of Pflood refers to SCR9.83 which is the SCR for hail risk. Is this correct? SCR CAT risk SCR (SCR 9.77) (page 271) This refers to point b) of paragraph 5. Please advise which paragraph this is this referring to? SCR CAT risk SCR (SCR 9.78) (page 271) This refers to paragraph 6. Which section is this in. Is this correct? SCR CAT risk SCR (SCR 9.78) (page 271) The risk weight is set based on property (LoB 7 in Annex K) but does Flood also apply to Motor? should this be included? The reference to paragraph 5 should be replaced by SCR The reference to paragraph 6 should be replaced by SCR.9.51 SCR9.59 should be changed to SCR9.56 which refers to the same risk (windstorm). SCR9.33 should be changed to SCR9.29 which is the formula to compute geographical diversification. The definition of flood should refer to SCR The reference to paragraph 5 should be replaced by SCR The reference to paragraph 6 should be replaced by SCR.9.75 motor LoB was not included in the target calibration of flood risk module, however other motor insurance LoB is included in the flood risk module according to formula in SCR.9.76.

5 SCR CAT risk SCR (SCR 9.79) (page 271) As above, the correlations are set so that the Flood loss equals the loss to property (LoB 7 in Annex K) but does it also apply to Motor as well? SCR CAT risk SCR (SCR 9.109)(page 279) What time period applies to the number of vehicles insured? Should it be a policy count as at the reference date for the SCR calculation? SCR CAT risk SCR (SCR 9.128) (page 284) The correlation matrix for Liability risk includes 5 groups, but only 4 are defined in SCR Please advise the 5th definition? SCR CAT risk Various Is it possible to use the same reinsurance across multiple catastrophe scenarios? For instance if you had a 1m xs 1m catastrophe reinsurance cover with no reinstatements and were subject to both the windstorm and flood catastrophe, both with gross losses in excess of 2m, would you be allowed to take credit for the reinsurance in both the windstorm and flood catastrophe scenarios, or only take credit for it once? motor LoB is not included in the target calibration of flood risk module, however the correlations applies to other motor insurance LoB as well (according to formula in SCR.9.76). The policy count at the reference date for the SCR calculation. Liability risk group 5: Non-proportional reinsurance of obligations relating to insurance obligations included in line of business 8 as set out in Annex K. The same reinsurance can in principle be applied across multiple catastrophe scenarios. But in the concrete case where the cover has no reinstatements, the reinsurance can be applied only once. However, this seems to be quite an unrealistic reinsurance cover. SCR CAT risk Various What policies should the sum insured in the catastrophe module relate to? Is it the sum insured on risk at the valuation date? Or is the total sum insured on risk to expiry allowing for a further year's worth of policies? The policy count at the reference date for the SCR calculation.

6 SCR SCR.9.4. Non life CAT risk sub - module SCR Regarding the Credit & Suretyship submodule of the man-made CAT Risk, we note that the definition of P_recession uses the Premiums earned by the insurance or reinsurance undertaking during the last 12 months. However, the definition of L_recession in section SCR uses the definition of premiums earned by the insurance or reinsurance undertaking during the following 12 months to calculate the instantaneous loss. Which premium defintion for the exposure is correct? The correct definition is the definition in SCR Standard_SCR SCR Basis Risk SCR (a) We notice that in paragraph SCR a generic definition of material basis risk is given. Is it the intention that this definition is only supposed to cover insurance risk-mitigation techniques (since it does not appear in chapter 11 on financial risk-mitigation techniques)? (b) We have noticed that the content of paragraph SCR reflects the content of EIOPA s draft GL 3 on basis risk (which are meanwhile also available on EIOPA s homepage). We understand that GL 1 of the paper is supposed to be applied to financial and insurance riskmitigation techniques. Its content however is only reflected in paragraph SCR for financial risk-mitigation techniques. Should the paragraph (with the exception of the first sen-tence) also be applied for insurance risk-mitigation techniques? (a) A comparable generic definition of basis risk in relation to financial risk mitigation techniques is included in SCR on conditions for financial risk mitigation techniques (b) The same conditions and principles on basis risk as in SCR should also apply to assessing material basis risk in relation to insurance risk mitigation techniques

7 Standard_SCR SCR.6.3. Lossgiven-default for risk mitigating contracts SCR (i) Question: We struggle to understand the instruction for the calculation of the term SR for pool-exposures of type A. For the formula under ii. (on page 186) one needs to distinguish between counterparties within the scope of Solvency II and other counterparties. We wonder if that is really the in-tention, given that the formula under ii. uses the term SR (called solvency ratio there) and EOF. That seems to make only sense where the counterparty is not rated (otherwise we use the rating to derive SR) and within the scope of Solvency II. We therefore think the index i should be used for all unrated counterparties within the scope of Solvency II. The in-dex j should be used for all other counterparties. Do you confirm this understanding? (ii) Furthermore, it remains unclear in the Technical Specifications how to derive SR for the counterparties that are not in the scope of Solvency II (whether rated or not). The index i is intended to represent firms within the scope of S2, and the index j the firms that aren t. Contrary to your interpretation, the index is not intended to be linked to whether the firm has a credit rating or not. The correct interpretation should be the following: 1. If a pool member is required to comply with S2 (indexed i), a. If it has a rating, use it to calculate the SR; b. if it doesn t have a rating, use the SR 2. if the firm is not required to comply with S2 (indexed j), a. If it has a rating, use it to calculate the SR; b. if it doesn t have a rating, I. If it is an insurer in a third country which is equivalent, and the firm complies with the local rules, then SR = 1; II. If it is an insurer in a third country which is NOT equivalent, and the firm complies with the local rules, then SR = 0.75.

8 Standard_SCR SCR.6.3. Lossgiven-default for risk mitigating contracts SCR.6.34 SCR.6.34 of the Technical Specification for the Preparatory Phase (Part I) and the "Underlying Assumptions in the standard formula for the Solvency Capital Requirement calculation" paper ("Underlying Assumptions paper", as at 17 April 2014) define the calculation for the LGD for a reinsurance arrangement or securitisation i as follows: LGD(i) = max(0; 50%(Recoverables(i)+50%.RM(re,i))-F.Collateral(i)) The Level 2 text (as at 10 January 2014) defines the LGD for a reinsurance arrangement or securitisation i as follows: This update is to be consistent to the most recent policy developed. LGD(i) = max(50%(recoverables(i)+rm(re,i))-f.collateral(i);0) Please explain the technical reason for introducing another 50% in front of the RM(re,i) component in the Technical Specification for the Preparatory Phase and the Underlying Assumptions paper.

9 SCR.6.36 of the Technical Specification for the Preparatory Phase (Part I) and the Underlying Assumptions paper define the calculation for the LGD for a reinsurance counterparty which has tied up an amount for collateralisation commitments greater than 60% of the assets on its balance sheet, as follows: This update is to be consistent to the most recent policy developed. LGD(i) = max(0; 90%(Recoverables(i)+50%RM(re,i))-F.Collateral(i)) Standard_SCR SCR.6.3. Lossgiven-default for risk mitigating contracts SCR.6.36 The Level 2 text (as at 10 January 2014) defines the LGD for a reinsurance counterparty which has tied up an amount for collateralisation commitments greater than 60% of the assets on its balance sheet, as follows: LGD(i) = max(90%(recoverables(i)+rm(re,i))-f'.collateral(i);0) Please explain the technical reason for introducing another 50% in front of the RM(re,i) component in the Technical Specification for the Preparatory Phase and the Underlying Assumptions paper. Also, please explain why the Level 2 text includes an F' and the Technical Specification for the Preparatory Phase (Part I) and the Underlying Assumptions paper include an F. We are of the view that the same notation should be used consistently across all documents.

10 SCR.6.37 of the Technical Specification for the Preparatory Phase (Part I) defines the calculation for the LGD on a derivative i as follows: LGD(i) = max(0; 90%(Marketvalue(i)+RM(fin,i))-F.Collateral(i)) where F is defined as a Factor to take into account the economic effect of the collateral arrangement in relation to the reinsurance arrangement or securitisation in case of any credit event related to the counterparty i. This is to be consistent to the most recent policy developed. We agree that the F factor for collateral on derivatives should be distinguished from the F factor for collateral in relation to reinsurance. The F factor will be replaced by F. Standard_SCR SCR.6.3. Lossgiven-default for risk mitigating contracts SCR.6.37 In our view, the Technical Specification for the Preparatory Phase (Part I) should distinguish between the economic effect factor for reinsurance arrangements or securitisations (SCR.6.34 and SCR.6.36) and the economic effect factor for a derivative (SCR.6.37). Currently, there is no distinction between both and is denoted by a F throughout. While this distinction is made in the Underlying Assumptions paper as at 17 April 2014,we note that the Level 2 text (as at 10 January 2014) distinguishes between the economic effect factor between non-heavily collateralised reinsurance counterparties (denoted by an F) and heavily collateralised reinsurance counterparties and derivatives (denoted by an F'), rather than between reinsurance counterparties irrespective of whether they are heavily collateralised or not (denoted by an F) and derivatives (denoted by an F'). Please explain.

11 The possibility of using volatility adjustment in the yield curves is mentioned. This adjustment is said to be applied according to LTGA rules and when no matching adjustment is used. Can you please verify exactly in which lines of business the volatility adjustment can be used assuming that no matching adjustment will be used? (i.e. does it apply to: unit linked, unit linked with guarantees, with profits, conventional non profit, term business, health NSLT etc) According to the Directive (Article 77d(1)), the volatility adjustment applies to the relevant risk-free interest rate structure without further limitations. Volatility Adjustment

12 Valuation V Methodology for the calculation of the best estimate A direct writer has a practice of purchasing cat cover on an annual basis. Should it assume that cover is purchased in future years (ie allow for risk mitigation in calcualting the SCR at future points in time to calculate the RM and allow for future reinsurance premiums)? When calculating amounts recoverable from reinsurance contracts, insurance and reinsurance undertakings should recognize cash-flows relative to future reinsurance purchasing covering obligations already recognized in the balance sheet to the extent that it is replacing any expiring reinsurance arrangements and if it can be demonstrated that it meets the conditions stated below: - The insurance or reinsurance undertaking has a written policy on the replacement of the reinsurance arrangement; - The replacement of the reinsurance arrangement does not take place more regularly than every 3 months; - The replacement of the reinsurance arrangement is not conditional on any future event which is outside of the control of the insurance or reinsurance undertaking. Where the replacement of the reinsurance arrangement is conditional on any future event that is within the control of the insurance or reinsurance undertaking, then the conditions should be clearly documented in the written policy referred to in point (a); - The replacement of the reinsurance arrangement should be realistic and consistent with the insurance or reinsurance undertaking s current business practice and business strategy. The insurance or reinsurance undertaking should be able to verify that the replacement is realistic through a comparison of the assumed replacement with replacements taken previously by the insurance or reinsurance undertaking; - The risk that the reinsurance arrangement cannot be replaced due to capacity constraints is immaterial;

13 - An appropriate estimate of the future reinsurance premium to be charged is made which reflects the risk that the cost of replacing existing reinsurance arrangements may increase; - The replacement of the reinsurance arrangement is not contrary to the specifications that apply to future management actions set out in TP to TP The question concerns "direct or indirect participations in real estate companies that generate periodic income or which are otherwise intended for investment purposes" and is related to section SCR.5.6.Mktprop property risk of the Technical Specifications for the Preparatory Phase, cf. especially paragraphs SCR.5.57 and SCR.5.58 of that section. As far as we can see, the kind of property investments referred to in the question raised by the undertaking is not explicitly listed in the technical specifications for the preparatory phase and this seems to be a change compared to the specifications used for the LTGA-exercise. Accordingly, our (preliminary) understanding would be that the look through approach should be applied and the investments referred to should consequently be treated within the property risk sub-module. Could you please confirm whether our understanding is correct? Otherwise, a clear instruction is needed regarding the treatment of this kind of investments in the context of the on-going stress test exercise. As regards the calculation of the risk margin and the allowance for reinsurance covers to mitigate SCR risks at future points in time, the same approach should apply, without prejudice to SCR Where undertakings invest in real estate through collective investment undertakings, which are not considered a related undertaking within the meaning of Article 212(1)(b) and 212(2) of Directive 2009/138/EC, or other investments packaged as funds, the look through approach should be applied. For other cases, i.e. equity investments in a company exclusively engaged in facility management, real estate administration, real estate project development or similar activities, undertakings should apply the equity risk sub-module. Direct or indirect participations in real estate companies that generate periodic income or which are otherwise intended for investment purposes should be subject to the equity risk module.

14 A Bulgarian subsidiary has lev and euro contracts (no cross border contracts). Which volatility adjustment (22bps for euro or 20bps for Bulgaria) can be applied for the valuation of the euro contracts? In this context would it make a difference if the currencies had not been pegged to the Euro? The currency of the risk-free interest rates term structure to be used in the calculation of the best estimate should always be the currency of the insurance and reinsurance obligations. Therefore, for contracts whose obligations are denominated and paid in euro, the euro risk-free curve, which the euro volatility adjustment is applied, should be used. Tech.Spec II - LTG LTG Volatility Adjustment Standard_SCR SCR.5. SCR market risk module SCR.5.9 Based on previous submitted question, numbered , ID 1217, p.16 (LTGA exercise) it has been provided that in order to calculate concentration risk, the estimation of the Assets which are in the denominator, would have to exclude assets of this type (i.e. deposits of the undertaking with a bank, which is a parent company, if only this bank operates under Basel II rules). However, it is not clear whether these exposures should have to be excluded also by the nominator for the determination of the E, i.e. the net exposure at default to this specific counterparty. Questions: 1. Deposits with credit institutions and cash at bank are The principle is that if an asset is exluded from the calculation base of the market risk concentration submodule (Assetsi), it should also be excluded from the exposure at default to a single name exposure (Ei). Please see SCR for this interpretation. Please refer to SCR for the list of assets that are excluded from Assetsi and Ei.Cash at bank should be treated in the counterparty default risk module and is excluded from the market risk module. Deposits with credit institutions other than cash at bank should be included in the appropriate market risk sub-modules, especially concentration risk sub-module and spread risk sub-module. For a generic description of the

15 differentiated on the basis of the tenor? i.e. fixed tenor => Deposits with credit institutions and not fixed tenor => cash at bank? 2. Time deposits to a parent undertaking, which is a bank under Basell II are treated, with regards to the calculation of SCR, in the spread risk and concentration risk modules? 3. Time deposits to a parent undertaking, which is a bank under Basell II, bear concentration risk? Are they included in the calculation of E but excluded from the calculation of Assets? distinction between cash at bank and other deposits please refer to answer to previous question: Cash at bank" should be understood as meaning cash on a current account or cash at hand that is immediately available to be drawn up by the undertaking under all circumstances. Deposits should be understood as including deposits with credit institutions, deposits with ceding undertakings or deposits received from reinsurance undertakings as specified in in Article 6 of Directive 91/674/EEC.

16 Annex E to the Technical Specification for the Preparartory Phase Part I We have some questions about Annex E in which discount factors in the derivation of the risk margin are discussed. Annex E shows that the risk margin is just sufficient in the best estimate scenario to release an amount of (rf+coc)*scr0 in the first year when invested in the risk free asset. SCR0 in this case is the capital buffer commensurate with the 99.5% confidence level over the first year of projection, determined prior to adding the risk margin itself to the balance sheet. The risk margin subsequently provides an additional buffer that is available to fund benefits to policyholders in a worst case scenario. Hence by adding the risk margin to the liability, the likelihood of meeting policyholder liabilities increases beyond the 99.5% confidence level, and also the capital provider is less exposed to risk than without the addition of the risk margin. When the risk margin is small compared to the SCR, the impact of adding the risk margin to the balance sheet may be negligible. However if it is large compared to the SCR, the degree of protection for policyholders may increase far above the targeted 99.5% confidence level over a one year period and also, the loss incurred by the capital provider in an adverse scenario is significantly reduced. The latter implies that also the risk exposure of the capital provider is materially reduced by the addition of the risk margin to the liability. Nevertheless, the investor is compensated through the release of the risk margin as if the SCR serves as the only buffer against adverse deviations of the liability. Level I legislation states that the risk margin must reflect the cost of providing the SCR. Assuming that a reduction in risk exposure leads to a Annex E is supplementary technical background that describes some technical aspects regarding the discount factors to be used in the calculation of the risk margin. The risk margin should be calculated in accordance with section v2.5 of technical specification part I. We would note that the risk margin must reflect the cost of holding capital equal to the SCR until the liabilities have gone off the books. For particularly long tail business, this may result in a risk margin which is close in magnitude to SCR0.

17 lower cost of capital for the investor, the derivation in annex E shows that the risk margin is at least at the level prescribed by Level I legislation. The addition of the risk margin to the balance sheet reduces the risk to the investor, but this reduction is not taken into account when determining his cost of capital. Furthermore, when the duration of liabilities is very long, the risk margin may even be higher than SCR0, the capital initially provided. The level I text stipulates that the risk margin must reflect the cost of providing the SCR, so that the cost of providing the SCR as reflected in the risk margin, is then in excess of the actual amount of SCR provided. We have the following questions about this approach: * In case of a risk margin that is in the same magnitude as the SCR, is the increase of available funds beyond the one year 99.5% confidence level by addition of the risk margin to the balance sheet, and the corresponding reduction of risk exposure for the capital provider, intentional? * Alternatively, should the SCR be interpreted as the capital required to meet obligations at a 99,5% level of confidence over a single year after addition of the risk margin? * Is a risk margin in excess of SCR0 intended and considered reflective of the cost of providing capital in the amount of SCR0, as stipulated in the level I legislation?