Instructions regarding reporting templates for the. Non-Life underwriting risk Comparative Study (NLCS) on Internal Models (IMs)

Transcription

1 Instructions regarding repting templates f the Non-Life underwriting risk Comparative Study (NLCS) on Internal Models (IMs) General comments: Last update: 27/04/2018 This document comprises of instructions that each selected solo undertaking participating in the NLCS is invited to follow when completing the questionnaire (put in an Excel wkbook) related to this study. It aims to provide definitions f terms used and guidance in following instructions so that consistency across the NLCS questionnaires can be achieved as much as practicable. F the purpose of these instructions, under the agreed scope of this study, the following terms referred to in the questionnaire are defined: - Undertaking: The term undertaking means full IM undertaking, partial IM undertaking, and IM group in relation to solo results strictly. - Scope: o Undertakings: Only selected solo undertakings with IMs comprising a nonlife underwriting risk module are considered. o Risks: All requested figures are understood to exclude catastrophe risks. o Business: If unmodelled business means business out of scope of the IM then only the figures related to the modelled business should be included. However if unmodelled business encompasses business in the scope of your IM then the related figures should be repted in this questionnaire. - Time hizons: o Ultimate time hizon refers to the non-life model outputs that relate to risk over the time hizon of the run-off of the undertaking s obligations to its policyholders, and f premium risk, including obligations relating to business planned to be written in the 12 months following the reference date. o One-year time hizon is the basis on which the solvency capital requirement (SCR) is calculated under article 101(3) of the Solvency 2 (S2) Directive. - Currency and display: All monetary amounts are to be repted in EUR. If the iginal values are based on a feign currency, please use the daily feign exchange rate at the reference date as published by the ECB 1 and convert it to EUR. Unless otherwise explicitly stated do NOT display monetary figures as 1 dex.en.html Page 1 of 26

2 multiples of EUR (like keur MEUR) and NOT as percentages but as absolute amounts. - Discounting: Where relevant, every amount should be discounted at the applicable risk-free rate. The market risk is out of scope of this study, i.e. the discount rates used should be deterministic and the ones prevailing at t=0. Should another type of discount curve be used please provide explanations in the comments and describe your approach. - Reinsurance: unless stated otherwise, all requested figures are gross of reinsurance. - Premium provisions: Premium provisions changes (see appendix) are to be treated as part of premium risk f this questionnaire unless statedotherwise. - Risk margin: No requested figure pertains to the risk margin. - Expenses: All types of expenses (including fixed costs) related to premium and reserve risks are expected to be covered as in the Solvency 2 balance sheet. o If some expenses are not captured under the IM, please comment on it. o If the IM provides an allocation of expenses at Solvency 2 Lines of Business level, please make sure that internal LoBs results reconcile with the Solvency 2 LoBs results. - Quantiles: High quantiles represent adverse results f the undertakings: the underlying implicit distribution is indeed a loss distribution (i.e. claims - [some positive amounts]). - Reconciliation: Where relevant, please make sure the given figures reconcile with the regulaty national / college-wide IM QRTs. All the repted figures should also be consistent with the EIOPA QRT S Availability of figures: In general it is expected that the requested figures in this questionnaire are available at at least one of the requested granularities (internal Solvency 2 LoBs) and consistently repted f each of these 2 granularities to the extent possible (means add up, etc.) and with proxies 3 if not directly available. The figures should also be consistent with your national regulaty repting. Any deviation from these assumptions should be explicitly mentioned and justified. If adjustments are needed, they should be described and justified in the relevant cells (e.g. comments). When no figure at all is available (even with adjustments), please let the related cells empty and provide the reasons of this unavailability in the comments cells. Where relevant the figures shall take into consideration potential diversification effects. - Solvency 2 segmentations: In both Solvency 2 tabs ( 3.4 QUANT Prem risk - S2 LoBs and 3.6 QUANT Res risk - S2 LoBs ), the 4 Solvency 2 so-called segments (combined LoBs between direct and proptional reinsured business) are the ones from the Annex II of the Delegated Regulation. This segmentation was chosen because the standard fmula coefficients of variation are defined accding to this segmentation and f comparison purposes. - Mapping between internal and Solvency 2 LoBs: In general it is expected that while annuities in payment stemming from long-term bodily injuries claims (e.g. Periodic Payment Orders PPOs in the UK) are allocated to a separate Solvency 2 LoB, the risk associated with such future annuities within the premium 3 The solution making the most sense to you should be chosen and description of the chosen method should be provided and justified. Page 2 of 26

3 provisions should be included in the Mot Third-Party Liability Solvency 2 LoB. If it is not the case explanations are expected. F the sake of parsimony and readability, only the cells deemed not self-explanaty are further explained in this document with e.g. the intention and some necessary references. This document is intended to evolve over time to reflect the outcome of the Q&A process. From the template 3.1 QUAL Prem risk - Int LoBs onwards, the data requests only pertain to the following 4 S2 Lines of Business (LoBs): - Mot vehicle liability; - Other mot; - General liability; - Fire and other damage to property. Page 3 of 26

4 Template name Overview 1 ID infmation 2 Segmentations 3.1 QUAL Prem risk - Int LoBs Template description (high level) Overview on the questionnaires content, logic and legend Each submission should include this infmation in der to unambiguously identify the submitting undertaking and the submission date. The segmentations apply to all internal LoBs 4 modelled f premium and reserve risks. It aims to understand the composition of internal LoBs and to map out their relation to the S2 LoBs. Qualitative questions on premium risk at the level of the internal LoBs. These qualitative questions collect infmation on premium risk at the level of the internal LoBs to complement the quantitative questions. They pertain to the internal LoBs mapped to the subset of S2 LoBs. Qualitative questions on reserve risk at the level of the internal LoBs. 3.2 QUAL Res risk - Int LoBs These qualitative questions collect infmation on reserve risk at the level of the internal LoBs to complement the quantitative questions. They pertain to the internal LoBs mapped to the subset of S2 LoBs. Quantitative questions on premium risk at the level of the internal LoBs. 3.3 QUANT Prem risk - Int LoBs These quantitative questions collect infmation on premium risk at the level of the internal LoBs to complement the qualitative questionnaire. They pertain to the internal LoBs mapped to the subset of S2 LoBs. 3.4 QUANT Prem risk - S2 LoBs 3.5 QUANT Res risk - Int LoBs Quantitative questions on premium risk at the level of the S2 LoBs. These quantitative questions collect infmation on premium risk at the level of the selected S2 LoBs in der to compare undertakings IMs outputs on a common basis. Quantitative questions on reserve risk at the level of the internal LoBs. These quantitative questions collect infmation on reserve risk at the level of the internal LoBs to complement the qualitative questionnaire. They pertain to the internal LoBs mapped to the subset of S2 LoBs. 3.6 QUANT Res risk - S2 Quantitative questions on reserve risk at the level of the S2 4 A definition of internal LoB is given in the section dedicated to the Segmentations template. Page 4 of 26

5 LoBs 4. Comments LoBs. These quantitative questions collect infmation on reserve risk at the level of the S2 LoBs in der to compare undertakings IMs outputs on a common basis. The undertaking s comments related to the completion of the above templates. In the first column the undertaking should select from the drop-down list the template name to which the comment in question relates and enter their comment. Comments are especially needed if simplifications were necessary to arrive at some results, to avoid misinterpretations, etc. Each participating undertaking is expected to submit all of the above templates completed in accdance with the outputs available from its IM. If the undertaking is not able to submit a template partly in full, it should provide explanations in the 4. Comments template. F each template the undertaking should enter the infmation required in each greenshaded cell, select infmation required from a drop-down list in each blue-shaded cell, and make no amendments to any other (protected) cell in the template. Instructions: The instructions f each of the above templates are set out in the following tables. The column ITEM identifies the item to be repted by reference to the columns and rows (if applicable) as shown in the template. Page 5 of 26

6 Table of Contents 1. ID INFORMATION SEGMENTATIONS QUAL PREM RISK - INT LOBS QUAL RES RISK - INT LOBS QUANT PREM RISK - INT LOBS QUANT PREM RISK - S2 LOBS QUANT RES RISK - INT LOBS QUANT RES RISK - S2 LOBS COMMENTS APPENDIX NOTATIONS AND RISKS BREAKDOWN FOR PREMIUM & RESERVE RISKS Solvency Capital Requirement and Net Asset Value Stylized Solvency 2 balance sheets...24 Page 6 of 26

7 1. ID infmation General comments: This section relates to the identification of each undertaking. ID01 ID02 ITEM Undertaking name Undertaking identification code (preferably Legal Entity Identifier LEI) INSTRUCTIONS Legal name of the undertaking. Needs to be consistent over different submissions. This should be the same as the undertaking name repted in Solvency2 repting template S This should be the same as the undertaking identification code repted in Solvency 2 repting template S ID04 Repting submission date Expected fmat: DD/MM/YYYY. Page 7 of 26

8 2. Segmentations General comments: This section relates to the segmentations used to model the premium and reserve risks f all internal LoBs. An internal LoB should be understood as the most granular level from the IM direct outputs (i.e. no re-runs of the model are needed) at which the probability distribution function of the losses an SCR is available. Infmation on what every internal LoB contains should be given. Furtherme the internal LoBs should be mapped to the S2 ones and necessary adjustments mentioned. S2 LoBs in the drop-down list refer to the non-life LoBs as defined in the Annex I of the Solvency 2 delegated regulation (the so-called level 2). The list of internal LoBs should be complete in contrast to the other templates which focus on a subset of S2 LoBs. This mapping between internal and Solvency 2 LoBs is requested f all internal LoBs in der to check the materiality of the Solvency 2 LoBs left out of this edition of the study and to pave the way towards a me comprehensive analysis of the aggregation. 1-year premium provisions changes Allocation to premium risk As mentioned in the introducty comments, premium provisions changes are to be treated as part of premium risk f this questionnaire unless stated otherwise. This remark applies in particular to the segmentations template: no dedicated premium provisions risk module appears in this template. In case the risk of changes in premium provisions is modelled separately from premium risk, the internal LoBs concerned should be repted with the prefix Premium_Provisions_LoB_(NLCS) in the premium risk table. As mentioned also in the introducty comments in general it is expected that annuities stemming from long-term bodily injuries claims (e.g. Periodic Payment Orders PPOs in the UK) are included in the Mot Third-Party Liability Solvency 2 LoB. If it is not the case explanations are expected. F the sake of parsimony and readability, below are listed the items pertaining to both premium and reserve risks simply separated by a slash ( / ). C01 / 06 ITEM Premium / Reserve risk Internal LoB Name of the internal LoB INSTRUCTIONS This column lists the names of all internal LoBs that are used to calculate the premium / reserve risk. Page 8 of 26

9 C02 / 07 Premium / Reserve risk Description of the internal LoB Please be as descriptive as possible so it is clear what the LoB entered contains. F example, include whether the business is personal commercial, its distribution channel and in which country it is located. Premium risk: In case the risk of changes in premium provisions is modelled separately from premium risk, please rept that internal LoB with the prefix Premium_Provisions_LoB_(NLCS). If applicable, these prefixed internal LoBs will then be filled in in the premium risk-related templates where it makes sense to do so. Premium / Reserve risk: The description of each internal LoB should include the distribution channel as well as any infmation which was relevant f the chosen segmentation. You can also mention these infmation in the relevant comment cells. C03 / 08 Premium / Reserve risk Closest Solvency 2 LoB chosen Choose from the drop-down list Please select the S2 LoB that best fits into each internal LoB. C04 / 09 C05 / 10 Premium / Reserve risk Description of any adjustments perfmed to go from the internal LoB to the S2 one (especially on the SCR) Premium / Reserve risk Expected earned gross premiums over the 12 months after the reference date / Gross BE of outstanding claims as at the reference date If the S2 LoB is composed of me than one internal LoB an internal LoB has to be decomposed in der to be represented in several S2 LoBs, please explain the methods used, especially how the SCR at S2 LoB level was derived from the SCR at the internal LoB level. Measure (weight) to estimate the imptance of each internal LoB. Gross BE of case outstanding claims reserves and IBNR. F consistency reasons with the YE2017 data (expected data), expected data (and not realised data) should also be provided f the YE2016 data. Page 9 of 26

10 3.1 QUAL Prem risk - Int LoBs General comments: This template should be filled in befe the template 3.3 QUANT Prem risk - Int LoBs as the internal LoB selection of the fmer will populate the later. With respect to these internal LoBs, the questions below concerning premium risk should be answered. Questions related to the prefixed premium risk internal LoBs ( Premium_Provision_LoB_(NLCS) ) may be filled in where needed and where it makes sense to do so. In der to facilitate the wk of superviss in charge of analyzing and interpreting your answers to these qualitative questions, please use as much as possible the notations of regulaty texts on Solvency 2. F the same reasons, regarding the definition of premium risk and its related SCR, please use the notations defined in appendix and refer to the exposed risks breakdown thereof as much as possible when explaining your approach. QUALPRER0101 QUALPRER0102 QUALPRER0103 QUALPRER0201 ITEM Risk profile Country- and undertakingspecific (from the same country) features Risk profile What is the main type of business written? Please explain if "Other" was chosen. Consistency between the IM and the Solvency 2 balance sheet Please explain how the consistency requirement between premium risk and the INSTRUCTIONS If your undertaking belongs to a group, your group may have this knowledge and you are free to discuss this topic within your group. The ultimate responsibility lies however with you. One aim of this study is to learn about the selfawareness of undertakings with respect to other markets. Entries in the drop-down list: - Retail - Commercial - Industrial - Other Undertakings have to choose one among the possible choices: (generally speaking main means the business with the highest premiums). The consistency requirement can be found in article 121 of the Directive. Page 10 of 26

11 QUALPRER0301 QUALPRER0402 Odd column QUALPRER0503 Odd column Solvency 2 balance sheet has been implemented. Risk and SCR definition Please give the definition chosen f premium risk and its SCR. Risk hizon If premium risk is modelled on an ultimate view first, what method is used to arrive at the one-year view? Please explain if "Other" was chosen. Please provide a brief explanation of the method used. (How is the emergence pattern computed? On what is the emergence pattern applied?) and the reason this modelling approach was chosen. Exposure, claims, premiums and expenses Which methodology is used to model claims generated by this internal LoB? Please explain if Other was chosen. Please use the notations defined in appendix and refer to the exposed risks breakdown thereof as much as possible when explaining your approach. F emergence pattern it is intended a methodology used to get the one-year view from the ultimate view. Please add all comments and description needed about the method used. - Frequency x Individual cost approach: typically used f large claims where the number of claims is calibrated on the yearly number of histical claims and the severity distribution on the yearly individual cost of the histical claims. - Frequency x Average cost approach: typically used f attritional claims where the number of claims is calibrated on the yearly number of histical claims and the severity distribution per claim is calibrated on the yearly average cost of histical claims. - Distribution of loss ratios - Scenario-based approach - Other Page 11 of 26

12 3.2 QUAL Res risk - Int LoBs General comments: This template should be filled in befe the template 3.5 QUANT Res risk - Int LoBs as the internal LoB selection of the fmer will populate the later. With respect to these internal LoBs, the questions below concerning reserve risk should be answered. In der to facilitate the wk of superviss in charge of analyzing and interpreting your answers to these qualitative questions, please use as much as possible the notations of regulaty texts on Solvency 2. F the same reasons, regarding the definition of reserve risk and its related SCR, please use the notations defined in appendix and refer to the exposed risks breakdown thereof as much as possible when explaining your approach. CELL IDENTIFIER QUALRESR0101 QUALRESR0102 QUALRESR0103 Odd column QUALRESR0301 Odd column ITEM Risk profile Country- and undertakingspecific (from the same country) features Risk profile What is the main type of business you have been underwriting? Please explain if "Other" was chosen. Consistency between the IM and the Solvency 2 balance sheet Please explain how the consistency requirement between reserve risk and the Solvency 2 balance sheet has been implemented. INSTRUCTIONS If your undertaking belongs to a group, your group may have this knowledge and you are free to discuss this topic within your group. The ultimate responsibility lies however with you. One aim of this study is to learn about the selfawareness of undertakings with respect to other markets. Entries in the drop-down list: - Retail - Commercial - Industrial - Other Undertakings have to choose one among the possible choices: (generally speaking main means the business with the highest reserves). The consistency requirement can be found in article 121 of the Directive. Page 12 of 26

13 QUALPRER0201 QUALRESR0302 Odd column QUALRESR0701 Odd column QUALRESR0901 Risk and SCR definition Please give the definition chosen f reserve risk and its SCR. Risk Hizon If reserve risk is modelled on an ultimate view first, what method is used to arrive at the one-year view? Please explain if "Other" was chosen. Please provide a brief explanation of the method used. (How is the emergence pattern computed? On what is the emergence pattern applied?) and the reason this modelling approach was chosen. Computations of volatilities If an ultimate view is modelled first, what methodology is used to model it? Please provide a brief description of the approach and explain if "Other" was chosen. Reinsurance How do you take into account the risk mitigation effects of reinsurance f this internal LoB? Please use the notations defined in appendix and refer to the exposed risks breakdown thereof as much as possible when explaining your approach. F emergence pattern it is intended a methodology used to get the one year view from the ultimate view. Please add all comments and description needed about the method used. Entries in the drop-down list: - Closed-fmula approach (e.g. Mack) - Simulation-based approach (e.g. bootstrap variants) - Internal benchmark (e.g. parameters from an internal LoBs are used on another internal LoB) - External benchmark (e.g. parameters from an external source are used on an internal LoB) - Other To our knowledge the reinsurance effects in reserve risk are usually modelled in a much less granular way than in premium risk (e.g. gross-tonet ratio). Please describe your approach in the free-text cell. Page 13 of 26

14 3.3 QUANT Prem risk - Int LoBs General comments: This template should be filled in after the template 3.1 QUAL Prem risk - Int LoBs as the internal LoB selection of the later will populate the fmer. With respect to these internal LoBs, the questions below concerning premium risk should be answered. Prefixed internal LoBs ( Premium_Provision_LoB_(NLCS) ) may be filled in needed and where it makes sense to do so. where As mentioned in the introducty comments, below are slightly me technical definitions related to both time hizons: - Ultimate time hizon refers to the non-life model outputs that relate to risk over the time hizon of the run-off of the undertaking s obligations to its policyholders, and f premium risk, including obligations relating to business planned to be written in the 12 months following the reference date. The ultimate view (the same applies mutatis mutandis to reserve risk) is the risk of an adverse development between the reference date and the time hizon of the obligations of the estimated ultimate cost of future claims. - One-year time hizon is the basis on which the solvency capital requirement (SCR) is calculated under article 101(3) of the Solvency 2 Directive. The one-year view (the same applies mutatis mutandis to reserve risk) is the risk of an adverse development between the reference date and one year after this reference date of the estimated ultimate cost of future claims. Regarding parameter uncertainty the following definition 6 was chosen: the prediction uncertainty ( err) of e.g. the claim development result in reserve risk is generally split into a process err and an estimation err, the latter being also called parameter uncertainty parameter err. Parameter uncertainty is the possibility that the parameters used to define the model are increct given that the model is crect. This possibility occurs, firstly, because there is only a (sometimes very) limited amount of data to estimate the parameters. Secondly, the parameters themselves will evolve through time, so those applicable f future events are always unknown at the present. If a CAT risk is included in the LoB Fire (f example Windstms), it is not required to make an artificial run to exclude it. This specificity should however be mentioned in the comments cells. 6 F references, see f instance: Cairns (A. J. G.) (2000) A Discussion of Parameter and Model Uncertainty in Insurance Walker (S.) (1996) Wkshop on Stochastic Err, Parameter Err & Model Err Page 14 of 26

15 ITEM INSTRUCTIONS PRERG(N)101/102 Odd and even columns (merged cells) PRERG104 Odd and even columns (merged cells) PRERG201 Odd Even column PRERG301 to 306 Odd even column PRERG401 to 403 Odd even column Gross (net) premium risk data Expected written/earned premiums over the 12 months after the reference date Expected combined ratio used in calculating the BE of premium provisions in the Solvency 2 balance sheet as at the reference date (t = 0) Gross premium risk data One-year Ultimate Mean of claims and expenses Gross premium risk data One-year Ultimate 0.1%, 0.5%, 5%, 95%, 99.5%, 99.9% percentiles of premium risk Gross premium risk data One-year Ultimate Effects of parameter uncertainty (if applicable and F consistency reasons with the YE2017 data (expected data), expected data (and not realised data) should also be provided f the YE2016 data. See general comments on time hizons. The mean of claims and expenses is the sum of mean of claim and mean of expenses. See general comments on time hizons. Depending on the definition chosen f the premium risk and its SCR (incl. confidence level and risk measure), the two quantities 99.5%thquantile and SCR of premium risk might differ. See general comments on time hizons. If your final SCR figures include parameter uncertainty, then please give here figures without parameter uncertainty. Vice-versa, if your final SCR figures don't include parameter uncertainty, then please give here figures with parameter uncertainty. If not available, please let the cell empty. Page 15 of 26

16 available) Page 16 of 26

17 3.4 QUANT Prem risk - S2 LoBs General comments: Prefixed internal LoBs ( Premium_Provision_LoB_(NLCS) ) must be included. Individual cell references of this template are the same as the ones mentioned in the section 3.3 QUANT Prem risk - Int LoBs. The only additional cell in this template is the one below: CELL IDENTIFIER PRERG104 Odd and even columns (merged cells) ITEM Gross premium risk data Volume measure of the premium risk in the Standard Fmula INSTRUCTIONS Please insert the volume measure asked at SF level (i.e. as defined in the SF). Page 17 of 26

18 3.5 QUANT Res risk - Int LoBs General comments: This template should be filled in after the template 3.2 QUAL Res risk - Int LoBs as the internal LoB selection of the later will populate the fmer. As mentioned in the introducty comments, below are slightly me technical definitions related to both time hizons: - Ultimate time hizon refers to the non-life model outputs that relate to risk over the time hizon of the run-off of the undertaking s obligations to its policyholders, and f premium risk, including obligations relating to business planned to be written in the 12 months following the reference date. The ultimate view (the same applies mutatis mutandis to premium risk) is the risk of an adverse development between the reference date and the time hizon of the obligations of the estimated ultimate cost of claims already occurred. - One-year time hizon is the basis on which the solvency capital requirement (SCR) is calculated under article 101(3) of the Solvency 2 Directive. The one-year view (the same applies mutatis mutandis to premium risk) is the risk of an adverse development between the reference date and one year after this reference date of the estimated ultimate cost of claims already occurred. Regarding parameter uncertainty the following definition 7 was chosen: the prediction uncertainty ( err) of e.g. the claim development result in reserve risk is generally split into a process err and an estimation err, the latter being also called parameter uncertainty parameter err. Parameter uncertainty is the possibility that the parameters used to define the model are increct given that the model is crect. This possibility occurs, firstly, because there is only a (sometimes very) limited amount of data to estimate the parameters. Secondly, the parameters themselves will evolve through time, so those applicable f future events are always unknown at the present. RESR002 Odd and even columns (merged cells) RESRG101 Odd and even columns (merged cells) ITEM Data-histy Number of calendar years used to calibrate attritional claims Gross reserve risk data Claims provisions derived from the IM befe any Solvency 2 INSTRUCTIONS What is looked f here is the data histy and not the development of claims. Some accident years may have incomplete developments. Claims provisions as of the reference date and derived from the internal model. Give this value befe any adjustment (e.g. scaling f consistency reasons). In case there is no difference to the standard fmula claim provisions, please explain the 7 F references, see f instance: Cairns (A. J. G.) (2000) A Discussion of Parameter and Model Uncertainty in Insurance Walker (S.) (1996) Wkshop on Stochastic Err, Parameter Err & Model Err Page 18 of 26

19 balance sheet reconciliation adjustments rationale and approach. RESRG102 Odd and even columns (merged cells) RESRG201 to 206 Odd even column Gross reserve risk data BE of claims provisions Gross reserve risk data One-year Ultimate 0.1%, 0.5%, 5%, 95%, 99.5%, 99.9% percentiles of reserve risk Gross reserve risk data Provisions as of the reference date See general comments on time hizons. Depending on the definition chosen f the reserve risk and its SCR (incl. confidence level and risk measure), the two quantities 99.5%th-quantile and SCR of reserve risk might differ. RESRG301 Odd even column RESRG302 Odd even column One-year Ultimate Effects of parameter uncertainty (if applicable and available) Standard deviation of reserve risk without parameter uncertainty Gross reserve risk data One-year Ultimate Effects of parameter uncertainty (if applicable and available) 99.5% percentile of reserve risk without parameter See general comments on time hizons. Figure to be provided on a best-efft basis. Please give results without parameter uncertainty. If the results had to be acquired via approximations, please describe them in the comments template. If not available, please let the cell empty. See general comments on time hizons. Figure to be provided on a best-efft basis. Please give results without parameter uncertainty. If the results had to be acquired via approximations, please describe them in the comments template. If not available, please let the cell empty. Page 19 of 26

20 uncertainty RESRG303 Odd even column RESRN101 Odd column RESRN102 Odd column RESRN201 to 206 Odd even column RESRN207 Odd even column Gross reserve risk data One-year Ultimate Effects of parameter uncertainty (if applicable and available) SCR of reserve risk without parameter uncertainty Net reserve risk data Claims provisions derived from the IM befe any Solvency 2 balance sheet reconciliation adjustments Net reserve risk data BE of claims provisions Net reserve risk data One-year Ultimate 0.1%, 0.5%, 5%, 95%, 99.5%, 99.9% percentiles of reserve risk Net reserve risk data One-year Ultimate Mean of reserve See general comments on time hizons. Figure to be provided on a best-efft basis. Please give results without parameter uncertainty. If the results had to be acquired via approximations, please describe them in the comments template. If not available, please let the cell empty. Claims provisions as of the reference date and derived from the internal model. Give this value befe any adjustment (e.g. scaling f consistency reasons). In case there is no difference to the standard fmula claim provisions, please explain the rationale and approach. Best Estimate as of the reference date. See general comments on time hizons. See general comments on time hizons. Figure to be provided on a best-efft basis only if readily available. Page 20 of 26

21 risk RESRN208 Odd even column RESRN209 Odd even column Net reserve risk data One-year Ultimate Standard deviation of reserve risk Net reserve risk data One-year Ultimate SCR of reserve risk See general comments on time hizons. Figure to be provided on a best-efft basis only if readily available. See general comments on time hizons. Figure to be provided on a best-efft basis only if readily available. Page 21 of 26

22 3.6 QUANT Res risk - S2 LoBs General comments: Individual cell references of this template are the same as the ones mentioned in the section 3.5 QUANT Res risk - Int LoBs. The only additional cell in this template is the one below: CELL IDENTIFIER RESRG102 Odd and even columns (merged cells) ITEM Gross reserve risk data Volume measure of the reserve risk in the Standard Fmula INSTRUCTIONS Please insert the volume measure asked at SF level (i.e. as defined in the SF). Page 22 of 26

23 4. Comments General comments: This section relates to any additional comment which might be necessary to contextualize the infmation provided in any of the previous sections. This includes e.g. clarifications, suggestions f future versions of this template, etc. Please add lines as appropriate. CELL IDENTIFIER ITEM INSTRUCTIONS C01 Tab name Options: - Overview ID infmation Segmentations QUAL Prem risk - Int LoBs QUAL Res risk - Int LoBs QUANT Prem risk - Int LoBs QUANT Prem risk - S2 LoBs QUANT Res risk - Int LoBs QUANT Res risk - S2 LoBs C02 Row reference Row reference to be used may be found in red coloured cells (e.g. R001 in Template 2.0 Segmentations ) If an entire row is affected please leave the column reference empty. C03 Column reference If applicable Column reference to be used may be found in red coloured cells (e.g. C01 in Template 2.0 Segmentations ) If an entire column is affected please leave row reference empty. C04 Comment number f row and column reference In case of several comments related to one cell C05 Comment Description of issue. Please keep your explanation sht. Page 23 of 26

24 5. Appendix Notations and risks breakdown f premium & reserve risks The intention of this appendix is to provide undertakings with common notations and risks breakdowns f premium and reserve risks. The expected benefits of providing such a common risk alphabet are twofold: 1) to help undertakings to understand what is expected in the qualitative questions on Risk and SCR definition and give them a reference point to answer these questions and 2) to help superviss with the analysis of the answers to these questions. The aim of this appendix is to define common notations as well as theetical (i.e. pure Solvency 2-compliant) premium and reserve risks derived from an annual variation of basic own funds computed on a Solvency 2 balance sheet, as specified in the Directive (e.g. article 101). To that purpose a basic definition of the general SCR (i.e. regardless of the underlying risk) is first given and then two simplistic 8 Solvency 2 balance sheets seen at two different dates separated by one year from each other, namely year N and year N + 1, are set up in der to look me closely at the sole non-life underwriting risk SCR Solvency Capital Requirement and Net Asset Value The Solvency 2 Directive defines the SCR as the amount to which the basic own funds at the end of year N could be reduced 9 so as to ensure that the probability of economic insolvency in a one-year time equals 0.5%. Economic insolvency in a one-year time is defined as the point in time where the net asset value (NAV basic own funds) at the end of year N + 1 becomes negative, i.e. fmally 10 : SCR = NAV(N) such as P[NAV(N + 1) < 0] = P[NAV(N) + NAV < 0] = 0.5% where NAV(N + 1) NAV(N) = NAV. It then comes that SCR = q 0.5% ( NAV) = q 99.5% ( NAV) Stylized Solvency 2 balance sheets Below is a simplistic Solvency 2 balance sheet of an undertaking only subject to the nonlife underwriting risk between two consecutive years from which the NAV will be computed: 8 In particular the financial / investment income is neglected. 9 In other terms, it means that the SCR at the end of year N is the amount of NAV(N) which saturates the inequality P[NAV(N) + NAV < 0] 0.5%. 10 To be fully accurate, a discount fact should be added, i.e. the random variable that we look at should be NAV(N) + δ NAV, with δ the discount fact between the end of year N and the end of year N + 1. Nevertheless, f the sake of simplicity and readability, this fact will be simply omitted here. Page 24 of 26

25 N N + 1 Assets NAV(N) Assets NAV(N + 1) Economic value of non-life liabilities(n) + Written premiums(n + 1) - Claims settlements(n + 1) - Expenses(N + 1) Economic value of nonlife liabilities (N + 1) It then comes: NAV = Written premiums(n + 1) Claims settlements(n + 1) Expenses(N + 1) 11 + Economic value of non-life liabilities(n) Economic value of non-life liabilities(n + 1). Article 77 of the Directive defines the economic value of liabilities as The value of technical provisions shall be equal to the sum of a best estimate and a risk margin [ ]. In particular f the non-life underwriting risk: Economic value of non-life liabilities(n) = BE PCO (N) + BE Premium provisions (N) + RiskMargin(N) where BE stands f Best Estimate and PCO f Provisions f Claims Outstanding. The claims settled during the year N + 1 are then split accding to their accident years, i.e. between claims occurred until the end of year N (index N ) and those occurred only in year N + 1 (index N + 1 ): Claims settlements(n + 1) = Claims settlements N (N + 1) + Claims settlements N+1 (N + 1). The same split is perfmed and the same notations adopted f the Best Estimate of PCO computed at the end of year N + 1: It finally comes: NAV BE PCO (N + 1) = BE PCO N (N + 1) + BE PCO N+1 (N + 1). = Written premiums(n + 1) Claims settlements(n + 1) Expenses(N + 1) + BE Premium provisions (N) BE Premium provisions (N + 1) + BE PCO (N) BE PCO (N + 1) + RiskMargin(N) RiskMargin(N + 1) = BE Premium provisions (N) + Written premiums(n + 1) BE Premium provisions (N + 1) Claims settlements N+1 (N + 1) BE PCO N+1 (N + 1) Expenses(N + 1) + BE PCO (N) (Claims settlements N (N + 1) + BE PCO N (N + 1)) + RiskMargin(N) RiskMargin(N + 1) 11 These expenses refer to the so-called Unallocated Loss Adjustment Expenses (ULAE contrary to ALAE, the Allocated ones, which can be tied to a specific claim f lawyer expenses, legal procedures costs, etc. and are already embedded in the claims costs), the acquisition and the administrative costs. Page 25 of 26

26 = Technical result of the new accident year + Claims development result 12 due to fmer accident years + Risk margin variation. Article 13 (30) of the Directive defines the non-life underwriting risks as the risk of loss of adverse change in the value of insurance liabilities, due to inadequate pricing and provisioning assumptions. A premium ( pricing) risk can then be defined as the adverse (i.e. negative) evolution over one year of the following random variable: A theetical metric of the premium risk Technical result of the new accident year = BE Premium provisions (N) + Written premiums(n + 1) BE Premium provisions (N + 1) Claims settlements N+1 (N + 1) BE PCO N+1 (N + 1) Expenses(N + 1). Note that the random quantity Claims settlements N+1 (N + 1) + BE PCO N+1 (N + 1) may be modelled as a whole (e.g. as the estimated ultimate cost of the claims occurred during the accident year N + 1). As f the reserve risk, it can be defined as the adverse (i.e. negative) evolution over one year of the following random variable: A theetical metric of the reserve risk Claims development result due to fmer accident years = BE PCO (N) (Claims settlements N (N + 1) + BE PCO N (N + 1)). This random variable should be centred (i.e. with conditional expectation until N equal to 0). As a side and final remark, the random variable of the ultimate counterpart of the (oneyear) reserve risk defined above can be written: Ultimate reserve risk random variable = BE PCO (N) i 1 Claims settlements N (N + i). This random variable should also be centred (i.e. with conditional expectation until N equal to 0). 12 Accding to Merz & Wüthrich s terminology. Alternative terms could be used, e.g. Annual run-off result, Annual bonus-malus result Annual reserves release result. Page 26 of 26