《毕马威(KPMG):2022技术实践调研报告(2022)(英文版)(79页).pdf》由会员分享,可在线阅读,更多相关《毕马威(KPMG):2022技术实践调研报告(2022)(英文版)(79页).pdf(79页珍藏版)》请在三个皮匠报告上搜索。
1、Technical Practices Survey 2022 Solvency IIFinancial ServicesNovember to the 2022 report It is with the greatest pleasure that we present to you the 2022 edition of our annual Technical Practices Survey.As ever,the focus of this survey is to enable UK life insurance firms to identify the key technic
2、al issues within the industry,and the range of methodologies and approaches that have been adopted by their peers.We are incredibly pleased to see ongoing support for our survey,with 21 participants submitting responses this year,including full submissions from 10 IM firms.We aim to continuously evo
3、lve the survey so that participants find it insightful and relevant to the issues faced within the industry today.The executive summary dashboard overleaf provides an overview of how the key stresses and indicators of risk appetite compare to the median responses provided in this and the previous ye
4、ars survey.We observe that the core stresses such as those for equity and interest rate risk,as well as those relating to underwriting risk have remained relatively stable compared with the prior period.This maintains the theme we have seen in previous years.For credit risk calibrations,there have b
5、een limited changes in corporate credit calibrations.We note that the key focus areas are to develop bespoke credit calibration for illiquid assets to support firms strategy,and the implementation of credit risk,including Matching Adjustment under stress,into SCR calculation processes.Under capital
6、management,target solvency cover ratios have reduced slightly since the previous year.Around a quarter of respondents have reduced capital buffers this year following a general trend for increases last year.These are not wholesale changes,each case is a small refinement.This underlines that capital
7、buffer remains an area of active review within the industry.In order to further support firms in their assessment of the capital management,the report now includes additional calibration points for selected risks at 1-in-10 and 1-in-20 levels.Model risk continues to contribute significantly to opera
8、tional risk capital.Tightening the control environment around actuarial models is therefore a key area of focus at the moment.Each year,in response to market developments and participant feedback,we select thematic areas to explore in more detail in our report.This year those areas include:Historic
9、impact of the Covid-19 pandemic on YE21 assumption setting,where the responses indicate thatthe majority of participants excluded 2020 experience data from their base longevity assumption setting process however most did not make adjustments forsetting mortality improvement assumptions.Very few firm
10、s made differences to the assumption setting process for lapses or partial withdrawals.Only a third offirms continued to hold COVID-19 related provisions atYE21 and most of these are expected to be released over 2022.Forward looking impact of Covid-19 on firms risk calibrations and correlations,wher
11、e a lot of uncertainty remains.There is no real consensus on the expectation for how long COVID-19 will continue to impact experience data.There are also varying approaches as to how the data will be treated in future assumption setting.We trust that you will find the report insightful.Please contac
12、t a member of the team if you would like more information on any of the content.How To Read The Report Throughout the report we have included tables which show the median result from the 2021 report(YE20 medians)for comparison against the responses for this year.In the spirit of being transparent,pa
13、rticularly where firms can provide multiple responses to the same question,we have indicated the number of respondents included in a specific chart with a grey box,thus xThe box and whisker plots,shown illustratively below,has been used extensively within the report.This is read as:the minimum and m
14、aximum data points are shown by the outer grey vertical lines (whiskers).;the inter-quartile range is shown by the box where the lower quartile is shown by the dark blue section and upper quartile is shown by the light blue section;andThe top left hand corner of each page also indicates whether the
15、charts on that page include answers submitted by SF,IM/PIM firms,or both.James Isden Director 2022 KPMG LLP,a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,
16、a 2private English company limited by guarantee.All rights reserved.Technical Practices Survey 2022 Executive Summary The executive summary below provides an overview at a glance of how the median responses for the key stresses and indicators of risk appetite compare to the median responses provided
17、 in the previous years survey.The core stresses such as those on equity,interest rate and underwriting risk have remained relatively stable,as seen in previous years.For the risk appetite,there is a downward shift in the range of Solvency Coverage Ratios from the prior year at both amber and red lev
18、els,with around one-quarter of the respondents that participated in both the years reporting a decrease in capital buffers.For Longevity Risk we have not shown a comparable YE20 Median given the change in survey format this year to include IM01 submissions to the PRA.Median Response(YE21)Median Resp
19、onse(YE20)Matching Adjustment Overall Matching Adjustment (bps)-Average 91 89 Market Risk(99.5%stress)-45%-46%UK Equity Stress Equity Implied Volatility Stress (10 years)14%12%Currency Stress EUR-21%-24%Currency Stress USD-27%-26%Commercial Property Stress-31%-31%Residential Property Stress-27%-30%I
20、nterest Rate Risk (10 years,99.5%stress)Interest Rate Total Up Stress 187 200 Interest Rate Total Down Stress-150-150Interest Rate Volatility Stress (5 X 15 ATM swaption)(bps)19 17 Credit Risk Average Credit Spread Stress (10 years,99.5%stress)Financials A 404 433 Financial BBB 584 591 Non-Financial
21、s -A 258 286 Non-Financials -BBB 406 418 Longevity Risk (99.5%stress)Female (Age 65)Stress (increase in EOL,years)3.0 N/A Male (Age 65)Stress (increase in EOL,years)3.0 N/A Operational Risk Contribution of top six scenarios to risk capital 25%22%21%14%10%9%Information security Model risk Other Produ
22、ct flaws/mis-sellling Failed or ina ppropriate pricing/UW Failure of third party Diversification within scenarios 39%2022 KPMG LLP,a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG
23、 International Limited,a 3private English company limited by guarantee.All rights reserved.44%Other Insurance Risks (99.5%stress)21%21%Expenses Level Stress as%of Best estimate 30%30%Mass Lapse Stress Solvency Cover Ratio Risk Appetite 123%130%Red (Immediate action taken)138%142%Amber(Triggers warni
24、ng)Balance Sheet PreparationSF/IM This section considers some of the key areas in the preparation of a companys base balance sheet.The use of Long Term Guarantee Measures(LTGM)continues to be widespread,with only five out of 17 firms reporting that they do not make use of any LTGM.Transitional Measu
25、res as a%of Technical Provisions remained consistent to previous year as shown in the chart and table below.Key areas of model development remain broadly similar to previous years,with a greater focus now on methodology improvements.1.1 Which of the following Long Term Guarantee Measures do you use
26、in your balance sheet?642Matching adjustment Transitionals Volatility adjustment None Grandfathering sub debt 8 8 8 5 1.2 What are the Transitional Measures as a%of your TechnicalProvisions?(IM firms only)Power BI Desktop17 1.3 What are the key developments or model changes that you will focus on in
27、 2022&2023?Methodology improvements Speed of reporting Other Risk Calibrations -MA Risk Calibrations -Credit Tax Risk Calibrations -Longevity Dependency calibrations Model Validation Adding new entities to the PIM Changes due to M&A activity Modelling alternative assets IMAP Risk Calibrations -Inter
28、est Rates Adding risk types to PIM 7 6 6 4 4 3 3 2 2 2 2 1 1 1 1 18 Responses to Other include model controls,changes to granularity of model outputs,and initiatives to increase automation.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independe
29、nt member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 1 YE20 Median3.2YE21 Median3.3%Power BI DesktopManagement ActionsSF/IM We have observed that firms have well-established management action
30、s for non-profit business,and there have been no significant changes compared to previous years results.As expected,for with-profits business,most companies use some combination of bonus setting,market value reductions and changes to the equity backing ratio.1.4a For non-profit business,which manage
31、ment actions are assumed in the capital measures listed at 31st December 2021?No management actions Lower expenses under mass lapse Other Day-to-day ALM decisions Higher charges for insurance benefits Changes to pension schemes Increases in administration charges Change in reinsurance agreements 6 6
32、 Responses to Other include actions to restore MA compliance,planned cost saving initiatives,and changes to backing portfolios for guaranteed funds.1.4b For with-profit business,which management actions are assumed in the capital measures listed at 31 December 2021?Change in final bonus rates Change
33、 in regular bonus rates Market value reductions Changes to equity backing ratio RemovalRemoval of misc.surplus/planned of misc.surplus/planned e enhancements/past estate distributionsChanges/Introduction of CoG chargesChanges/introduction of CoG chargesOther No management actions 12 9 7 6 6 4 4 1 Re
34、sponses to Other include changes to smoothing limits,rebalancing of dynamic hedges,reduction in level of corporate bonds held,and changes in respect of future discretionary benefits.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent memb
35、er firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 18 13 4 2 2 2 1 1 Risk MarginSF/IM We have not observed any changes to Risk Margin projection methodology,as noted in previous years.We asked fir
36、ms whether any non-insurance risks have been considered within the risk margin calculations,and the only continuing observation is the inclusion of counterparty default risk by most respondents.1.5 What is the Risk Margin as a%of your Technical Provisions?6420IM/PIMSF Power BI Desktop Risk drivers b
37、y risk module The whole capital measure is projected using a single The whole capital measure is projected using a singl risk driver(e.g.assumed to run off in line with BEL)Different approaches for each block of business Actuarial model is able to perform stresses at future Actuarial model performs
38、stresses at future dates by dates for each risk,and capital is then aggregatedProjection is automatically provided by our aggregation Projectionmodel(i.e.Algorithmics,RiskAgility or equivalent)is automatically provided by our aggregati13 3 2 1 2022 KPMG LLP a UK limited liability partnership and a m
39、ember firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 19%IM/PIMSFYE20 Median2.41.0YE21 Median2.80.91.6 How do you project your capital re
40、quirements for the calculation of the Risk Margin?1 Power BI DesktopARisk drivers by risk module risk driver approach is uDifferent approaches for each Different approaches for eblock of businessActuarial model perform stresses at future dates for each risk,andActuarial model performs capital is the
41、n aggregated OtherOtherPillar 2 and ORSASF/IM The difference in Pillar 1 and Pillar 2 balance sheet and capital methodologies for firms continue to demonstrate similar trends to previous years.Only one of the IM firms commented that there was no difference in treatment between Pillar 1 vs 2.As we ha
42、ve seen in previous years,the most common differences relate to the Risk Margin,discount rates,and contract boundaries.Changes in the capital methodology are primarily driven by additional risks in scope for SF firms and a more tailored view of operational risks within the business for the IM firms.
43、Remove the risk margin Other RFR -change in ILP allowance Contract Boundaries Allow for different risks in RM Pension scheme risk DTA allowance Remove transitionals used in P1 Risk Margin CoC charge RFR -no deduction of CRA 15 1.7a Which of the following areas do you treat differently whenperforming
44、 your Pillar 2 calculations vs Pillar 1 calculations,with regards to Best Estimate Liability /Technical Provisions?6 6 5 5 4 4 3 2 2 10 years 6 years 5 Years 3 years 4 Operational Risk Risk Calibrations Correlations Other Allowance for non-linearity 7 3 3 3 2 Remove Fungibility Constraints Responses
45、 to Other include:-Shareholders interest in the with-profits fund estate recognised under Pillar 2;No tiering of capital under Pillar 2-Longevity risk is excluded from the risk margin calculation-Risk free rates based on gilts and swaps for different blocks of business-Higher operational risk requir
46、ement increases risk margin1.7b Which of the following areas do you treat differently whenperforming your Pillar 2 calculations vs Pillar 1 calculations,with regards to Pillar 2 -Capital?Responses to Other include:-Government bond spread risk not included under Pillar 2(but included in Pillar 1)-Dif
47、fering longevity risk calibration-Differing treatment of volatility adjustment and liquidity premium-Allowance for pipeline major model changes pending formal PRA approval.1.8 For how many years do you project your Pillar 1 BalanceSheet as part of your ORSA?10 3 1 8 2 4 17 2022 KPMG LLP a UK limited
48、 liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 14 17 1.9 How does your company project its future ca
49、pital requirement in the ORSA?Responses to Other include:-A combination of modelling and risk drivers is used for the different capital requirements for each risk-Capital model is used to determine allocated capital requirements at t=0 and this is projected forward using a series of risk drivers and
50、 exposure factors 1 2 1 Power BI Desktop13Volatility Adjustment SF/IM The Volatility Adjustment(VA)continues to be attractive to companies,with 66%of respondents applying VA to with-profit funds and 60%applying it to immediate annuities.There are no firms who plan to apply the VA in the future who d
51、o not currently do so.1.10 For which of the following types of business do you apply a Volatility Adjustment?Non-profit protection Immediate annuities Non-profit savings (e.g.endowment,whole of life)Unit linked with guarantees Currently apply No future plan to apply Unit linked without guarantees 5
52、6 6 4 4 6 2 7 1 8 With-profit business Deferred annuities Bulk purchase annuity buy-ins Bulk purchase annuity buy-outs 6 3 5 3 1 2 1 2 The VA has steadily increased throughout the first six months of the year since YE21,as tabulated below.*Source:PRA Publications 2022 KPMG LLP a UK limited liability
53、 partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 11 Matching Adjustment SF/IM The average base MA rate marginall
54、y increased to 91bps at YE21 from 89bps at YE20 (we have chosen to compare participants using an average for this question to better represent the picture over the year).This is consistent with credit spreads having remained relatively stable over 2021,although we did note some large changes(up and
55、down)for individual firms,in part driven by portfolio optimisation and other factors like ERM restructuring.1.11a Base Matching Adjustment (bps)-Overall 200150100500AAAAAABBBOverall5040 Power BI DesktopAAA AA A BBBOverall1.11b Proportion of base spread realised as MA (%)-Overall 1.11c Bas
56、e Matching Adjustment (bps)-Non-Financial corporates AAAAAABBBOverall2001501005001.11d Base Matching Adjustment (bps)-Financial corporates AAAAAABBBOverall200150100500Charts 1.11c and 1.11d exclude the following categories of assets:Infrastructure Debt,Commercial Real Estate Lending,and Restructured
57、 ERMs.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public Power BI DesktopMatch
58、ing AdjustmentIM Most firms manually allocate assets to the Matching Adjustment Portfolio(MAP),although we are seeing a trend towards more sophisticated approaches through automation.The majority of firms perform their allocation with a target of compliance with PRA Tests 1&3(plus potentially additi
59、onal internal constraints),however,two-thirds of firms also try to optimise the size of the MA through asset allocation.We found that some companies are still planning an extension to the MA coverage showing that this is still an evolving area where firms are keen to optimise the benefit.1.12 As par
60、t of the calculation of the matching adjustment,howare assets hypothecated within the MAP?Manual allocation of assets Automatic allocation of assets Combination of aboveOther5 3 Compliance with PRA test 1 and 3 Maximising MA 8 Cash Callable bonds (using Fisher approach)Inflation linked bonds Suprana
61、tional Non-callable fixed interest Cross currency swaps Inflation swaps Infrastructure debt Commercial mortgages Interest rate swaps Reinsurance asset Private Finance initiative loans Sale and leaseback Social housing loans Ground rent assets (not restructured)Residential mortgage backed securities
62、Secured financing Educational loans Equity release mortgages (not restructured)FX forwards/futures (not restructured)Property rental strips (not restructured)Inflation options Interest rate swaptions Student loans 10 8 9 9 8 6 7 7 6 6 5 4 4 5 2 3 4 3 2 1 2 1 1 1 1 1 1 1 1 2 1 1 1 Currently apply Pla
63、n to apply in future 9 9 10 1 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Publ
64、ic 1.13 What is your objective when allocating assets to the MAP?1.14 Which of the following asset classes (excluding any restructures e.g.equity release mortgage assets)do you have approval to include in your matching adjustment portfolios or do you plan to apply for in the future?6 Power BI Deskto
65、p126 Pension Scheme IM For Pension scheme Pillar I calculations,IAS19 basis emerged as the widespread choice for both Balance sheet and SCR purposes.However,for Pillar 2 it is evenly split between IAS19 and Funding.1.15 What basis do you use to calculate your pension scheme liabilities under Pillar
66、1?Funding basis IAS 19 basis 1.16 What basis do you use to calculate your pension scheme liabilities under Pillar 2?Funding basis IAS 19 basis Pillar 1 -Balance Sheet 6 Pillar 2 -Balance Sheet 3 3 Pillar 1 -SCR 1 5 Pillar 2 -SCR 4 2 6 12 Most respondents do not allow for any fungibility for the pens
67、ion scheme surplus.A couple firms that do allow for it added that they either allow for partial fungibility within the ring-fenced fund which is capped at the SCR level or use it only to offset add-on.1.17 Do you allow capital fungibility for any pension scheme surplus under Pillar 1?No 8%Yes 18%No
68、82%1.18 Are pension scheme risks allowed to diversify with risks on the rest of the business under Pillar 1?Yes 92%11 13 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private
69、 English company limited by guarantee.All rights reserved.Document Classification:KPMG Public Historic Treatment of COVID-19SF/IM Each year,in response to market developments and participant feedback,we select thematic areas to explore in more detail.In line with last year,we have continued to look
70、at the impact of COVID-19.For context,this questionnaire was produced and the responses submitted(for the most part)in Q3 2022.The majority of firms indicated that they excluded 2020 data in their longevity base assumptions setting at YE21 in light of the experience data observed since the start of
71、the COVID-19 pandemic,with 6 of these firms also excluding 2021 data.The most popular approach for lapse bases was to make no change to existing processes and to include all available data,with very few firms indicating they excluded 2020 and 2021 data.There is no real consensus on the approach used
72、 to set mortality assumptions,with responses ranging from firms maintaining their YE20 assumptions to updating their assumptions using all available up to date data.2.1 Did you do anything differently in the assumption setting process at YE21 in light of the experience data observed since the outbre
73、ak of the COVID-19 pandemic?Exclude 2020 data Exclude 2021 data Maintain PY assumption No change to the process Other Use a different model in analysing Use external data e.g.application Longevity -Base l MortalityMorbidityLapse -ULLapse -ProtectionLapse -WPPartial withdrawa Power BI Desktop1818 Res
74、ponses to Other include:Excluding the majority of 2020 data but not all Considering both with and without 2020 and 2021 data when choosing the most appropriate approach per assumption Applying uplifts to morbidity assumptions Increasing the number of years of data included in the experience analysis
75、2.2 Do you include any adjustment to reflect the impact of the COVID-19 pandemic in your longevity improvement assumptions?Yes 11%No 89%19 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International
76、 Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public Use a different model in analysing thedata Use external data e.g.application ofmedical science results 10 6 3 2 1 1 5 3 7 2 2 3 2 5 4 3 1 2 2 2 6 3 1 1 2 6 2 1 1 1 6 3 22 2 4 2 Power BI De
77、sktopHistoric Treatment of COVID-19SF/IM Over two thirds of firms indicated they did not hold any additional provisions in respect of COVID-19 for the purposes of YE21 reporting.Of the firms who did hold such provisions,half indicated that they plan to release these during 2022,whilst the rest indic
78、ated they would maintain these provisions over 2022.2.3 Did you hold an additional provision in respect of COVID-19 for the purpose of YE21 reporting?Yes 30%No 40%2.4 If you held an additional provision in respect of COVID-19 for the purpose of YE21 reporting,do you plan to release any provisions du
79、ring 2022 given the impact of the pandemic observed so far?Yes 60%No 70%Although 6 firms noted in question 2.3 that they held an additional provision,only 5 firms provided further information for question 2.4 20 52.5a In your experience analysis,from the start of the pandemic in 2019,over what perio
80、d will you consider your experience to be impacted by the COVID-19 pandemic?1 5 3 6 Other Exclude entirely No adjustments Apply different weightings Up to YE20 Up to YE21 Up to YE22 Other 8 5 4 Responses to Other include differing impacts dependent on assumption.15 18 2022 KPMG LLP a UK limited liab
81、ility partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 2.5b How will you treat this COVID-19 data within the expe
82、rience investigations going forward?Responses to Other include-including all data just adjusting assumptions for elements of historic experience not expected to impact future experience varying treatment dependant on decrement.1 Power BI DesktopExpenses vary Overheads whiLapse and Expense RiskSF Thi
83、s section looks at the Standard Formula specific risks.In particular,for the majority of insurers mass lapse is the biting scenario out of the three lapse stresses,although there are a couple of respondents for whom the lapse down stress is the most onerous.For the mass lapse scenario,management act
84、ions provide the main justification for assuming expenses vary with policy numbers,examples of which are considered in chart 3.3 below.These are generally considered over a period of 2-3 years.Other reasons include expense agreements and the ability to recover commissions to offset a larger per-poli
85、cy spread of overheads.3.1 Which of the lapse stresses is the biting scenario for your capital requirement?Mass lapse 3.2 What assumption do you make about expenses in each of the lapse stresses?Expenses vary with policy numbersLapse down 17%83%6 3 2 Overheads which run-Overheads whioff over time Ma
86、ss lapse Lapse down Lapse up Overheads which stay fixed 7 2 11 7 2 11 3.3 Within the mass lapse stress do you assume any further management actions to reduce costs on a permanent basis or while volumes recover?Direct actions on staff (e.g.headcount reductions)Fixed expenditure cuts (e.g.property,equ
87、ipment costs)Less direct actions on staff (e.g.reward changes,headcount freezes)Other actions (e.g.reduced project spend)MiscellaneousOtherOutsourcing cost cuts and change in balance of fixed/variable costs 10 7 5 3 2 2 1 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG
88、 global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public Miscellaneous includes a more ad-hoc approach and a combination of the above.Power BI DesktopExpense and
89、Long Term Equity (LTE)RiskSF As we have seen in previous years,all firms stress overhead and variable expenses and the majority of firms also stress investment expenses.Where investment expenses are not stressed,they are generally defined as a percentage of funds under management.Furthermore,there i
90、s a variety of responses in respect of stressing fixed outsourcing expenses which depends on the contractual agreements in place.For example,some firms do not stress the expenses for inflation while some only stress at the end of the outsourcing term.This year,respondents were asked whether they app
91、ly the less onerous equity stress in respect of strategic long-term equity investments as allowed under the Standard Formula.Only one respondent indicated that this was the case,due to holdings in subsidiaries.3.4 Which of your expenses are subject to the expense stress?Overhead expenses Variable/Ma
92、intenance expenses Investment expenses Other 11 11 9 No 38%Yes 63%11 3.5 Do you stress your fixed outsourced expenses in the expense risk SCR?8 Other includes project expenses.Note that one respondent subjects the investment expenses to the inflation component but not the base component of the expen
93、se stress.3.6 Do you use the Long-Term Equity stress for any assets?No,we have no relevant assets No,despite some qualifying assets Yes 11 9%9%82%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG Intern
94、ational Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 1 Power BI Desktop0(20)(40)(60)(80)EquityIM There have been no significant movements in the equity stress calibrations since the previous year.The median 1-in-200 equity stresses fo
95、r the different currencies have converged slightly,though there has been no material difference year on year in the magnitude of the stresses for any currency.The overall trend is a downward equity stress,though this is very minor.Private equity continues to attract the most onerous stresses with a
96、median of 55%.This is lower than the previous year when we had limited survey data and the median ranged between 65%to 70%.4.1a Equity Stress (%)-Key markets(20)(30)(40)(50)(60)UK Equity 95.0%UK Equity 99.5%US Equity 95.0%US Equity 99.5%EUR Equity 95.0%EUR Equity 99.5%4.1b Equity Stress (%)-OthersEM
97、 Equity 95.0%EM Equity 99.5%Private Equity 95.0%Private Equity 99.5%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights rese
98、rved.Document Classification:KPMG Public UK Equity 99.5%US Equity 99.5%EUR Equity 99.5%EM Equity 99.5%Private Equity 99.5%YE20 Median(46)(47)(44)(45)YE21 Median(45)(45)(43)(45)(55)Power BI DesktopYE21 Median19Equity Volatility and Correlation factorsIM The median for the base equity volatility for a
99、 term of 10 years has seen a small increase from 18%to 19%,in line with the previous trend.The range of responses has contracted somewhat,although this is mostly due to sampling differences.The median 1-in-200 additive stress at the 10 year term has also increased slightly from 12%to 14%.Some firms
100、have decided to strengthen their correlation factors between EUR and other(UK/US)equity,although these have remained largely stable,close to 100%for all currency pairs.4.2 Equity Rate Volatility -Base and additive stress%(10 year)20100Best Estimate90.0%95.0%99.5%Euro&USUK&EuroUK&US100959085 2022 KPM
101、G LLP a UK limited liability partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public%Best estimate 90.0%95.0%99.5%YE20 M
102、edian1830YE21 Median19242633Please note that the medians in the table above are shown as cumulative amounts.4.3 Correlation factors between UK,US and EUR equity assets (%)%Euro&USUK&EuroUK&USYE20 Median909291YE21 Median909390Property and Property volatilityIM The property stresses have weakened slig
103、htly and while many firms apply the same stress to both property types,a few firms differentiate between the two,leading to a higher median for commercial(31%)compared to residential property(27%).4.4 Property Stress (%)(10)(15)(20)(25)(30)(35)(40)Power BI DesktopComm.Property 95.0%Comm.Property 99.
104、5%Res.Property 95.0%Res.Property 99.5%151050Best Estimate90.0%95.0%99.5%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English comp
105、any limited by guarantee.All rights reserved.%Comm.Property 95.0%Comm.Property 99.5%Res.Property 95.0%Res.Property 99.5%YE20 Median(36)(30)YE21 Median(16)(31)(16)(27)Comm.-Commercial,Res.-Residential 4.5 Commercial Property Rate Volatility (%)additive stress (10 year)Power BI DesktopTerm 10 BETerm 1
106、0 90.0%Term 10 95.0%Term 10 99.5%Term 15 BETerm 15 90.0%Term 15 95.0%Term 15 99.5%Inflation and Currency IM For market inflation calibrations,the range of responses has increased with the upper limit increasing from c200 bps to c300 bps while the lower limit remains unchanged.The median stress amoun
107、t is 172 bps.Given current inflation levels,it is no surprise that the median inflation assumption has again increased this year by c.70 bps to c.410 bps for a 10-year term.Interestingly,this has not been mirrored in the 1-in-200 additive stress which has generally been kept level by firms or even d
108、ecreased in some cases.The median currency stresses have remained relatively unchanged for the last few years at the 1-in-200 level for both currencies.For both currencies there is a general consensus around the magnitude of the stresses although there is an outlier for the EUR which has remained co
109、nsistent since the prior year.There is a single respondent who uses the same risk profile for both USD and EUR exchange rates with the remainder using currency specific stresses.4.6a Implied inflation (bps)-Term 10 4003002001004.6b Implied inflation (bps)-Term 15 400300200100BE 90.0%95.0%99.5%BE 90.
110、0%95.0%99.5%*BE stands for Best Estimate4.7 Currency stress (%)-Depreciation with respect to GBP (Percentiles)Change in exchange rate,EUR to GBP Change in exchange rate,USD to GBP 0(10)(20)(30)(40)(50)0(10)(20)(30)(40)(50)0.5%5.0%10.0%0.5%5.0%10.0%2022 KPMG LLP a UK limited liability partnership and
111、 a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Document Classification:KPMG Public 0(50)(100)(150)(200)(250)(300)Interest Rate RiskIM Interest rates rose signific
112、antly over 2021.This resulted in some of the companies responding to the interest rate movements by increasing their interest rate stresses.There is a wide disparity in the interest rate stresses(including interest rate volatility and gilt-swap spread stresses)produced by different companies.This re
113、flects the variety of methodologies adopted in the industry,in particular,whether companies use additive or multiplicative stresses,or a combination of the two.For this years survey we have aligned the terms requested in charts 5.1a and 5.1b to those in IM 01.As such the median boxes for these chart
114、s do not contain equivalent medians for YE20,where terms 2 and 25 year were not requested.5.1a Interest rates 1-in-200 down shocks (bps)GBP300250200150100500Power BI Desktoperm 2 95.0%Term 5 95.0%Term 10 95.TTerm 25 95.Term 2 0.5%Term 5 0.5%Term 10 0.5%TTerm 25 0.5%Term 2 99.5%Term 5 99.5%Term 10 99
115、.TTerm 25 99.Term 2 5.0%Term 5 5.0%Term 10 5.0%Term 15 5.0%Term 25 5.0%(60)(80)(100)(120)(140)(160)08060Term 2 Term 5 Term 10 Term 15 Term 25 Term 2 Term 5 Term 10 Term 25 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Pu
116、blic organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Term 2Term 2Term 5Term 10Term 15Term 25Term 20.5%Term 50.5%Term 100.5%Term 150.5%Term 250.5%YE20 Median(171)(150)(120)YE21 Median(203)(172)(150)
117、(135)(127)5.1b Interest rates 1-in-200 up shocks(bps)GBPTerm 299.5%Term 599.5%Term 1099.5%Term 1599.5%Term 2599.5%YE20 Median217200169YE21 Median2292081871651745.1c Interest rates 1-in-20 down shocks(bps)GBP5.1d Interest rates 1-in-20 up shocks(bps)GBPTerm 5Term 10Term 15Term 25Term 15 Power BI Desk
118、topInterest Rate RiskIM 5.1e Interest rates 1-in-10 down shocks (bps)GBP(60)(80)(100)(120)1201008060Term 5 10.0%Term 10 10.Term 15 10.Term 25 10.Term 2 90.0%Term 5 90.0%Term 10 90.TTerm 25 90.Term 5 95.0%Term 10 95.Term 15 95.Term 25 95.TTerm 5 99.5%Term 10 99.Term 15 99.Term 25 99.5.1f Interest rat
119、es 1-in-10 up shocks (bps)GBPTerm 2 Term 5 Term 10 Term 15 Term 25Term 2 Term 5 Term 10 Term 15 Term 25 5.2a Sovereign/swap spread -1-in-200 stress GBP (bps)5.2b Sovereign/swap spread -1-in-20 stress GBP (bps)Term 2 Term 5 Term 10 Term 15 Term 25 Term 2 Term 5 Term 10 Term 15 T
120、erm 25 The median sovereign/swap spread stresses for YE21 reduced compared to YE20.This reduction is consistent with the fall in sovereign swap spread over 2021.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of i
121、ndependent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopTerm 2 90.0%Term 5 90.0%Term 10 90.Term 15 90.TInterest Rate Risk IM 5.2c Sovereign/swap spread -1-in-10 stress GBP (bps)5.3 Interest rate volatility
122、(bps)5550454035302520020100Best Estimate90.0%95.0%99.5%Term 2 Term 5 Term 10 Term 15 Term 25 5.4 How are shocks applied to interest rates?Additive/Multiplicative with a switching point Multiplicatively (log normal distribution)Other 2 2 6 Other includes additively with a logistic or non-n
123、ormal distribution,multiplicatively with a displacement factor,or by percentiles of another statistical distribution.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG
124、 International Limited,a private English company limited by guarantee.All rights reserved.10 Power BI Desktop8Interest Rate ModellingIM Most respondents allow for negative interest rates in their risk models,with approximately half of those firms applying a lower bound to the interest rates produced
125、 by their interest rate risk model.All respondents indicated that they hold capital for the spread between swaps and gilts.5.5a Does your interest rate model allow for negative interestrates for base and stress purposes?Base Yes,with no lower bound Yes,with lower bound 5 3 Using a displacement facto
126、r(with a Using a displacement factor(withmultiplicative shock model)Using a CEV type model 2Stress 5.5b If your interest rate model allows for negative interest rates,can you set out how this is achieved in the model?Base Stress Yes,with lower bound Yes,with no lower bound 5 1 Using a displacement f
127、actor(with aUsing a displacement factor(with multiplicative shock model)Using a CEV type model Using additive shocks 3 1 8 5 5.5c If your model allows for negative interest rates with a lower bound,how does your firm set the lower bound?Base No responses for Varies by duration Stress Same across all
128、 durations Varies by duration 4 6 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights re
129、served.1 1 2 Same across all durations 2 Power BI DesktopAsset-side Calibration -Total spread riskIM The credit sections cover both SII calibration and modelling methodology adopted by life insurers in the UK.The sections are divided into:Asset side credit risk-risk of change in the market value of
130、credit risky assets Liability side credit risk-risk of change in discount rate used to value annuity liabilities by those firms that have received permission to apply a Matching AdjustmentThe following charts show the change in total bond spreads in bps(incorporating spread volatility and migration
131、risk)for various different ratings at the 99.5th percentile.We also asked for spread only calibrations,however the responses were insufficient for reasonable comparisons to be drawn.6.1a Change in Total Corporate Bond Spreads -Financials 10 years (bps)2,0001,5001,0005000AAA 99.5%AA 99.5%A 99.5%BBB 9
132、9.5%BB 99.5%B 99.5%bps AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%YE20 Median 216 299 433 591 1,031 1,410 YE21 Median 242 279 404 584 1,110 1,476 6.1b Change in Total Corporate Bond Spreads -Financials 15 years (bps)1,5001,0005000AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%2022 KPMG LLP a U
133、K limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.bps AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5
134、%B 99.5%YE20 Median 190 244 417 525 930 1,300 YE21 Median 213 235 349 510 968 1,313 Asset Side Calibration -Total spread riskIM The following charts show the change in total bond spreads in bps(incorporating spread volatility and migration risk)for various different ratings at the 99.5th percentile.
135、6.2a Change in Total Corporate Bond Spreads -Non-Financials 10 years (bps)1,5001,0005000AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%1,5001,0005000AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%Power BI Desktop 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global
136、 Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.bps AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%YE20 Median8882981YE21 Median67351,043
137、6.2b Change in Total Corporate Bond Spreads -Non Financials 15 years (bps)bps AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%YE20 Median8723941YE21 Median8718875Power BI DesktopAsset Side Calibration -Total spread riskIM The following charts show the change in total bond sprea
138、ds in bps(incorporating spread volatility and migration risk)for various different ratings at the 99.5th percentile.6.3a Change in Total Bond Spreads -Commercial Real Estate Lending 10 years (bps)3,0002,5002,0001,5001,0005000AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%3,0002,5002,0001,5001,00050
139、00AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarant
140、ee.All rights reserved.6.3b Change in Total Bond Spreads -Commerical Real Estate Lending 15 years (bps)Power BI DesktopAsset Side Calibration -Total spread riskIM The following charts show the change in total bond spreads in bps(incorporating spread volatility and migration risk)for various differen
141、t ratings at the 99.5th percentile.6.4a Change in Total Bond Spreads -Infrastructure Debt Lending 10 years (bps)1,5001,0005000AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%1,5001,0005000AAA 99.5%AA 99.5%A 99.5%BBB 99.5%BB 99.5%B 99.5%2022 KPMG LLP a UK limited liability partnership and a member fi
142、rm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.6.4b Change in Total Bond Spreads -Infrastructure Debt Lending 15 years (bps)Power BI Deskt
143、opAsset Side Calibration -Credit RiskIM The volatility of credit spread widening shocks(incorporating the effect of spread volatility and migration)has historically been higher for Financial bondsthan for Non-Financial bonds.This is shown below for A and BBB.A full view of other ratings can be found
144、 in the Asset Side Calibration Appendix towards the end of the Credit Risk section.The following charts show the change in total bond spreads(incorporating spread volatility and migration risk)for A and BBB rated bonds at 99.5th and 95th percentiles in bps.6.5a Change in Total Corporate Bond Spreads
145、 -Financials 10 years (bps)8006004002000A 99.5%BBB 99.5%A 95.0%BBB 95.0%8006004002000A 99.5%BBB 99.5%A 95.0%BBB 95.0%6.5b Change in Total Corporate Bond Spreads -Non Financials 10 years (bps)8006004002000A 99.5%BBB 99.5%A 95.0%BBB 95.0%6.5c Change in Total Corporate Bond Spreads -Financials 15 years
146、 (bps)6.5d Change in Total Corporate Bond Spreads -Non Financials 15 years (bps)8006004002000A 99.5%BBB 99.5%A 95.0%BBB 95.0%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated w
147、ith KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopAsset Side Calibration -Credit Stress IM The following charts show the change in total bond spreads in bps(incorporating spread volatility and migration risk)for A and BBB rated bonds at
148、 the 99.5th and 95th percentiles.While we also requested for information on spread only calibrations,we received limited responses so are unable to provide a meaningful comparison.6.5e Change in Total Bond Spreads -Commercial Real Estate Lending 10 years (bps)8006004002000A 99.5%BBB 99.5%A 95.0%BBB
149、95.0%8006004002000A 99.5%BBB 99.5%A 95.0%BBB 95.0%6.5f Change in Total Bond Spreads -Commercial Real Estate Lending 15 years (bps)8006004002000A 99.5%BBB 99.5%A 95.0%BBB 95.0%6.5g Change in Total Bond Spreads -Infrastructure Debt 10 years (bps)6.5h Change in Total Bond Spreads -Infrastructure Debt 1
150、5 years (bps)8006004002000A 99.5%BBB 99.5%A 95.0%BBB 95.0%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by
151、 guarantee.All rights reserved.Power BI DesktopAsset Side Calibration -Transition and Default Stress IM We have introduced new questions in this years survey concerning the probability of downgrade and default.The charts below show each respondents 1-in-200 probability of downgrade and default by cr
152、edit rating for both Financials and Non-Financials,which can be compared against a widely used benchmark which is broadly represented by the red dotted line.Each dot colour in the chart below represents the response of a particular firm.We also asked for transition and default information on Commerc
153、ial Real Estate Lending and Infrastructure Debt however the depth of responses was insufficient to provide meaningful insight.6.6a 1-in-200 probability of downgrade and default for Financial Corporates (%)60 40 20 0 AAA AA A BBB BB B 6.6b 1-in-200 probability of downgrade and default Non-Financial C
154、orporates (%)60 40 20 0 AAA AA A BBB BB B 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All r
155、ights reserved.Power BI DesktopAsset Side Credit Risk -Modelling ApproachIM Within the universe of all credit risky assets,we believe that corporate bonds are the largest single asset class to which life insurers are exposed.Exposure to commercial mortgage real estate and infrastructure investment i
156、s also now increasing.In this section,we discuss the approaches to modelling credit risk that have been adopted across the UK life insurance sector.6.7 What proportion of your portfolio is held in the following currencies (%)?Annuities non-MAP Annuities MAP Other Unit-linked Total spread Spread-only
157、 Default/downgrade-only component 7 3 5 6 3 5 0GBP EURUSDOther4 3 34 2 36.8 Which of the following credit-related stress components do you calibrate?Other refers to other product lines 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classific
158、ation:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.9 Technical PraPocwteri BIcDeess kt oSpurvey2022 6.Credit Risk Asset Side Calibration Credit Risk Drivers 1 2 3 4 5 6 7 8 9 Default
159、 risk drivers Y Y Y Y Y Y Y Y N Matching Adjustment Offset (MAO)Fundamental spread Spread risk driver Transition risk driver Other Common to all in scope asset classes?IM Only one firm reported that they used a specific,non-credit risk driver for an asset class.6.9 What specific risk drivers do you
160、allow for within your capital modelling (e.g.in your proxy model)?Other risk drivers include:3 Other represents a single risk driver that covers downgrades,i.e.transitions,and defaults.4 did not provide information on the Other risk driver9 Other risk driver applicable to Commercial Real Estate Lend
161、ing represents property level,property volatility,net rental income,cost of liquidity,cost of capital 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International
162、Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopIdiosyncraticIdiosyncratic(portfolio/asset specific)(portfolio/asset specifi risks only acroeconomic(market)risk onlyMacroeconomic(market)risks onlyacroeconomic and Idiosyncratic(p Macroeconomic and Idiosyncra
163、tic (portfolio/asset specific)risks only MMAsset Side Calibration -Risk DriversIM Rating and term continue to be the most influential aspects of a credit holding which impact the credit stress calibration.6.10 What is the source of risk drivers for the different credit asset classes listed below in
164、calculating your credit spreads SCR?2 3 2 Corporate Bonds Infrastructure LendingLendingCommercial Real Estate Other alternative credit 1 2 1 2 1 2 8 1 1 8 1 1 6 14 2 Term Security Rating Other industry sectors Financial/Non-financial Duration Domicile of issuer Currency Alternative asset class ABS v
165、s non-ABS Bespoke calibration Multiple of another baseline 4 1 3 2 2 1 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English compa
166、ny limited by guarantee.All rights reserved.7 8 6.11 In relation to an individual holding in credit,for which of the following factors would a change in the input result in a change in the output total spread stress?1 Power BI DesktopAsset Side Calibration -Credit StressIM 6.12 Have you applied any
167、expert judgements or overlays in respect of credit risk within the calibrations applied for market risks?Yes 44%No 56%6.13 What diversification do you allow for in calculating the credit spreads SCR?For the purpose of this question we consider:-Perfect correlation:+/-100%-Strong correlation:absolute
168、 value of correlation is greater than 70%-Medium correlation:absolute value of correlation between 30 and 70%-Weak correlations:absolute value of correlation is less than 30%.-Uncorrelated:0%correlation3 4 1Credit asset types Credit ratings Duration /duration buckets Spread risk&other credit asset r
169、isk Transition risk&credit asset risk 4 35 12 42 4Perfect Strong Medium The chart above indicates that firms tend to allow for diversification between different risks and credit ratings for the total spreads risk driver.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG g
170、lobal Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.9 8 Power BI DesktopLiability Side Credit Risk -CalibrationIM Credit risk remains a very significant risk a
171、nd focus for a number of firms.We have continued to benchmark the 1-in-200 default and downgrade component of credit stress,as it is indicative of the credit spread stress capital net of offset from changes in MA.The range of 1-in-200 average fundamental spreads stress as a%of total spread stress is
172、 too wide to provide a meaningful benchmark.6.14a Average change in fundamental spreads prior to re-balancing,1-in-200 stress for 10 years (GBP)(bps)250200150100500Financial Corporates(GBP)Non Financial Corporates(GBP)AAAAAABBB6 6.14b Average change in fundamental spreads prior to re-balancing,1-in-
173、200 stress for 15 years (GBP)(bps)250200150100500Financial Corporates(GBP)Non Financial Corporates(GBP)AAA AA A BBB6 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG
174、 International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopLiability Side Calibration -Fundamental Spreads Under StressIM Similar to last year,all respondents adopt a bespoke calibration for components of Fundamental Spreads(FS)under stress.In modelling
175、 FS in stress,cost of downgrade is the assumption that is most commonly varied in comparison to base EIOPA.6.15 Which elements of the fundamental spread calibration do you allow for in your stress calibration?6 PD CoD LTAS LGD/Recovery rate LTA spreads LTA transition matrix RC factorsRisk-free rate
176、5 2 6 4 4 4 2 1 5 5 6.16 Do you allow for a glidepath period for transitions within each of the components respectively?Yes No 67 6.17 To model FS in stress,which of the following assumptions do you vary compared to base EIOPA?4 6 3 3Cost of Downgrade Glidepath Recovery rate 3 5 2 21 1 1Commercial R
177、eal Estate Lending Corporate Bonds Infrastructure Lending Other alternative credit 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private E
178、nglish company limited by guarantee.All rights reserved.7 Risk Correction(RC)factors Power BI DesktopLiability Side Calibration -Fundamental Spreads Under StressIM 6.18 Do you calibrate stressed Fundamental Spreads (FS)at a different granularity as compared to base FS,e.g.more credit rating steps or
179、 asset classes?43%57%Same as base MA Yes 6.19 If you answered yes to question 6.18,is this the same,more or less granular than the base FS?1 2Asset classes beyond financial and non-financial split Credit rating step Term Duration 1 21 22Less Granular More Granular Same 2022 KPMG LLP a UK limited lia
180、bility partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.7 3 Power BI DesktopRMatching Adjustment Under Stress Cal
181、ibrationIM 6.20 Rebalancing strategy -how are defaults treated within the stressed matching adjustment portfolio?Replaced with new bonds as per stressed portfolio Assets replaced with risk-free asset Assume loss by applying stressed LGD Assume loss by applying stressed LGD (ultimate)Remove defaulted
182、 assets from MAP 3 1 1 1 1 6.21 Rebalancing strategy -how are defaults treated within the stressed matching adjustment portfolio?No action assumed as using buy-and-hold No action assumed strategy 3Allow for BBB cliff Adjust CoD Other 4 Replace with assets of eplace withhigher rating assets 32Other 2
183、 Other includes additional capital injected to the MAP to cover cost of increased Fundamental Spreads,this capital dilutes the MA and replacement assets assume same mix as existing(no change to investment strategy)6.22 How do you treat sub-investment grade bonds,e.g.below BBB rating?Other includes b
184、espoke approaches such as replacing bonds with investment grade,and adjust PoD and LTAS.7 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a pr
185、ivate English company limited by guarantee.All rights reserved.6 7 3 Power BI DesktopMatching Adjustment Under Stress CalibrationIM Similar to last year,in demonstrating compliance with the Matching Adjustment regulations under stress,all firms allow for transfer of assets between the non-MA Portfol
186、io and the MA Portfolio,and most of them also allow for changes in liability cashflows.6.23 In demonstrating compliance with the Matching Adjustment regulations under stress,which of the following do you allow for?Transfer assets between non-MAP and the MAP Changes in liability cash flows Reallocati
187、on between Comp A,B and C of the MAP Purchases of corp.bonds at stressed spreads All trading activity to be completed within 2 months Purchases of gilts with zero MA Sale of assets from the MA portfolio Purchases of illiquid assets at stressed spreads 7 6 6 5 3 2 2 1 Restructuring ERM-backed notes t
188、o reflect a change in expected cashflowsRestructuring ERM-backed notes to reflect a change in expected cashfl 1 6.24 When calculating your SCR,how do you validate that the Matching Adjustment under stress passes the PRA tests?Compliance tested in every scenario Compliance tested in every scenario us
189、ed to calibrate proxy model Compliance tested in a sample of the scenarios Off-cycle validation Methodology constructed to ensure tests always Otherpass post-stress if they pass pre-test 2 2 1 1 1 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classifi
190、cation:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.7 6 Power BI DesktopInternal RatingsIM As expected,almost all Internal Model companies maintain an internal credit rating approach
191、.However,we noted a very large reliance on asset managers to provide the rating.6.25 Do you have an internal ratings framework?No 10%Yes 90%6.26 For which of the following asset types do you use either internal ratings supplied by your asset manager,internal ratings derived in-house,or not use inter
192、nal ratings?3 3Commercial Real Estate Lending Equity release mortgages Private placements Infrastructure Lending Social Housing Lending Complex exposures Large exposures Private structured finance 4 21 51 41 1 32 22 24Derived in-house No internal rating Supplied by asset manager 2022 KPMG LLP a UK l
193、imited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.10 9 Power BI DesktopInternal RatingsIM 6.27 Ho
194、w do you validate that your internal rating methodology gives comparable ratings to an external methodology (ECAI consistent)as referenced in SS3/17 (April 2020 update)?Internal validation of Ratings methodology Compare to a sample of externally rated External validation of Ratings methodology Other
195、 Compare method to an ECAI methodology 6 Other includes a comparison of internal and external ratings for assets that have both,and internal rating performance analysis.6.28 If you have an internal ratings framework,which of the following approaches do you use for internal ratings?2 1 3Commercial Re
196、al Estate Lending Equity release mortgages Infrastructure Lending Social Housing Lending Private structured finance 1 1 31 1 31 31 2 A Rating Agency Methodology Scorecard approach Stochastic model Update a Rating Agency Methodology per asset class 2022 KPMG LLP a UK limited liability partnership and
197、 a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.7 6 4 3 3 2 Power BI Desktop1,0008006004002000AAA 95.0%AA 95.0%A 95.0%BBB 95.0%
198、BB 95.0%B 95.0%Asset Side Calibration -AppendixIM The following charts show the change in total bond spreads in bps(incorporating spread volatility and migration risk)for various different ratings at the 95th percentile.6.29a Change in Total Corporate Bond Spreads -Financials 10 years (bps)6.29b Cha
199、nge in Total Corporate Bond Spreads -Financials 15 years (bps)1,0008006004002000AAA 95.0%AA 95.0%A 95.0%BBB 95.0%BB 95.0%B 95.0%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliate
200、d with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopAsset Side Calibration -AppendixIM The following charts show the change in total bond spreads in bps(incorporating spread volatility and migration risk)for various different ratings a
201、t the 95th percentile.6.29c Change in Total Corporate Bond Spreads -Non Financials 10 years (bps)1,0008006004002000AAA 95.0%AA 95.0%A 95.0%BBB 95.0%BB 95.0%B 95.0%6.29d Change in Total Corporate Bond Spreads -Non Financials 15 years (bps)1,0008006004002000AAA 95.0%AA 95.0%A 95.0%BBB 95.0%BB 95.0%B 9
202、5.0%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopAsset S
203、ide Calibration -AppendixIM The following charts show the change in total bond spreads in bps(incorporating spread volatility and migration risk)for various different ratings at the 95th percentile.6.29e Change in Total Bond Spreads -Commercial Real Estate Lending 10 years (bps)1,5001,0005000AAA 95.
204、0%AA 95.0%A 95.0%BBB 95.0%BB 95.0%B 95.0%6.29f Change in Total Bond Spreads -Commerical Real Estate Lending 15 years (bps)1,5001,0005000AAA 95.0%AA 95.0%A 95.0%BBB 95.0%BB 95.0%B 95.0%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG P
205、ublic organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopAsset Side Calibration -AppendixIM The following charts show the change in total bond spreads in bps(incorporating spread volat
206、ility and migration risk)for various different ratings at the 95th percentile.6.29g Change in Total Bond Spreads -Infrastructure Debt Lending 10 years (bps)6004002000AAA 95.0%AA 95.0%A 95.0%BBB 95.0%BB 95.0%B 95.0%6.29h Change in Total Bond Spreads -Infrastructure Debt Lending 15 years (bps)60040020
207、00AAA 95.0%AA 95.0%A 95.0%BBB 95.0%BB 95.0%B 95.0%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarant
208、ee.All rights reserved.Methodology and ApproachIM Given the ongoing focus on Lifetime Mortgages(LTMs)this section captures some of their key calibration details of the IFRS valuation and the treatment under Solvency II.For the IFRS valuation,the majority of firms use a Real World LTM NNEG valuation.
209、Under Solvency II,for stressing the LTMs restructuring,the majority of firms looked through to the underlying assets.As Effective Value Test(EVT)requirements are becoming better understood,more firms are now testing the EVT under all stress scenarios of the proxy model and we expect this will contin
210、ue to increase over time.The Solvency II restructured notes spread over risk-free rate varies significantly with the largest variation seen at the BBB rated level.This variability is part of the rational for the PRA EVT that looks through each firms individual approach to enable the PRA to compare c
211、onsistently between firms.7.1 Do you use risk-neutral or real world parameters,or both,in LTM NNEG valuation used in determining the IFRS Fair value?Look through toLook through tunderlying assets17%Treat like an eTreat like anTreated like a equivalent corporate bond 17%Treat like a corporate bond wi
212、th adjustment 67%Real World Both 80%7.2 Please describe your approach to stressing your LTM restructure 20%56 7.3 What is your calculation approach for PRAs EVT?4003002001000AAAAAABBBPower BI Desktop5 Tested under allTested under ascenarios in the IM Outside of inteTesting a range of scenarios outsi
213、de the IM 40%60%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.7.4 Average
214、 spread over Solvency II risk-free rate by rating Power BI DesktopAssumption CalibrationIM For underlying LTM assumptions the house price growth assumption was between 3.2%and 4.7%(reducing to between 0.7%and 3.3%under stress)and the house price volatility assumption was between 12.5%and 14%(increas
215、ing to between 17.7%and 21%under stress).7.5 House price growth (pre-dilapidation or any other deductions)(%)6420(2)(4)Base AssumptionsAdditive Stress Resulting Stress Assu.esulting Stress Assu.Resulting Stress Assu.6543210Base Assumption7.6 House price growth volatility (%)2520151050Base Assumption
216、sAdditive StressResulting Stress Assumption 7.7 Average base sale cost (%)7.8 Average base voluntary redemption rates for a single life policy,policy year 10 (%)6543210Base AssumptionsAdditive StressResulting Stress Assumption 2022 KPMG LLP a UK limited liability partnership and a member firm of the
217、 KPMG global organisation of independent member firms affiliated with KPMG International Limited,a Document Classification:KPMG Public private English company limited by guarantee.All rights reserved.Resulting Stress Assumption Power BI DesktopAssumption CalibrationIM The median for the additive str
218、ess for average liquidity premium can be compared with the median observed for the stress for non-financial bonds-it lies between than the median additive stress for A-rated and AA-rated bonds.All participants have a property sale delay period between 8-12 months.7.9 Average liquidity premium over t
219、he Solvency II risk-free rate(bps)543210Base AssumptionsAdditive Stress 3.02.52.01.51.00.50.0Base AssumptionsAdditive Stress Resulting SResulting Stress Assu.7.10 Deferment rate used in EVT (%)7.11 What is your property sale delay assumption?12 months 9 months 8 months 2 2 15 2022 KPMG LLP a UK limi
220、ted liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Resulting StressAssumption Resulting Stress Assump
221、tion Power BI Desktop000000.250.0.0.0.0.0.252520200000.2525250.2020200.1515150.1010100.05550.0000Base Mortality Risk IM The charts below use data from firms IM 01 templates.8.1a Change in mortality rate at age 25 (Males&Females)(%)25.2020.1515.1010.05 5.00090.0%95.0%99.5%90.0%95.0%99.5%8.
222、1b Change in mortality rate at age 40 (Males&Females)(%)90.0%95.0%99.5%2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English compa
223、ny limited by guarantee.All rights reserved.8.1c Change in mortality rate at age 55 (Males&Females)(%)Power BI DesktopMortality Catastrophe Risk IM The charts below use data from firms IM 01 templates.Only one firm applied different stresses for different ages at a 90.0%and 95.0%level,and only two f
224、irms applied different stresses for different ages at a 99.5%level,when considering ages 25,40 and 55.8.2a Mortality catastrophe for age 25 (overall)(deaths per 1000)90.0%95.0%99.5%3.0 2.52.01.51.00.50.090.0%95.0%99.5%2015105090.0%95.0%99.5%3.02.52.01.51.00.50.08.2b Mortality catastrophe for age 40
225、(overall)(deaths per 1000)3.02.52.01.51.00.50.090.0%95.0%99.5%8.2c Mortality catastrophe for age 55 (overall)(deaths per 1000)8.2d Mortality catastrophe for age 75 (overall)(deaths per 1000)2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:
226、KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopBase Mortality ModellingSF/IM Most companies set their base mortality assumption by graduating against a standard table,t
227、ypically using tables published by the CMI with the most common tables in use being the 08 series tables.8.3 Which of the following best describes the method used to setyour base mortality assumptions?6%12%76%6%Standard Table Other Internal experience Using a GLM Other includes fitting to internal b
228、ase tables,and using a theoretical model based on individual mortality characteristics calibrated to portfolio experience 17 Other Bulk Scheme Internal Vesting Individual Life Level External Vesting Open Market 7 5 4 3 1 8.4 How many years of historical data do you use to set base mortality assumpti
229、ons?2520151050 8.5 What portfolio segments do you set separate assumptionsfor?00 Series 08 series S3 Series Other 12%24%12%53%Other includes a mixture of bespoke approaches 8.6 Which base mortality tables are your annuitant mortality assumptions based on?Other includes use of the 16 series mortality
230、 tables and England&Wales population tables14 17 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarante
231、e.All rights reserved.YE20 MedianYE21 Median55Years 1 Power BI DesktopBase Mortality Modelling SF/IM A significant proportion of companies also apply adjustments to their base mortality assumptions to reflect risk features relevant to their portfolios,such as anti selection and lifestyle factors,wit
232、h these factors generally applied at individual policy or block of business level.8.7 Which adjustments do you allow for in your base mortality assumptions?Anti-selection Lifestyle factors Other Health factors Late life mortality convergence Smoker status No adjustments 7 7 6 5 5 5 3 Other include a
233、djustments for marital status,temporary selection loadings on the first life,and IBNR adjustments.8.8 At what level are these adjustments applied?Policy level Block /Scheme level Other 14%50%36%Other includes life level and age group level.2022 KPMG LLP a UK limited liability partnership and a membe
234、r firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.15 14 Power BI DesktopLongevity -Annuitant Mortality ImprovementsSF/IM The previous pag
235、es within the mortality section referred to the mortality risks faced by insurers on their protection business.The following pages focus on longevity risks,as faced by firms on their annuity business.The majority of companies plan to adopt the CMI 2021 model for reporting at YE22 with the remainder
236、split broadly evenly between the 2018,2019 and 2020 versions of the model.We note that the Core calibration of the CMI 2021 model does not allow for 2020 or 2021 experience and that the mortality improvements produced by the CMI 2021 Core model are broadly similar to those produced by the 2019 and 2
237、020 models.A similar proportion of companies are planning to adopt the latest available CMI model for reporting at YE22.A similar approach was adopted at YE21.8.9 Which version of the CMI model do you currently use (and plan to use for YE22)for best estimate mortality improvements?2021 2020 2019 201
238、8 2021 2020 2019 2018 5 10 3 10 2022-Plan 8.10a Do you use core,extended or advanced calibration in your longevity improvement basis?8.07.57.06.56.0Core Extended Advanced 8.10b If you use the Extended or Advanced parameterisation of the CMI 2016 model or later,what value of the period smoothing para
239、meter do you use?44%44%11%18 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserve
240、d.18 YE20 MedianYE21 Median8 7Value3 3 2 Power BI DesktopInform judgements in BE assumInform judgements in stress calibLongevity -Annuitant mortality improvementsSF/IM The long-term rates of mortality improvement assumptions are generally higher for males than for females,although the median assumpt
241、ion is 1.5%for both males and females in the survey.The most common adjustments from the CMI Core calibration is to adjust the smoothing parameter.8.11 Long term rate of improvement (%)2.01.81.61.41.21.0FemalesMales8.12 If you use an Extended or Advanced calibration for the CMI model,which calibrati
242、on changes do you make?Smoothing parameter Long-Term Improvement Rates Initial addition to improvements Initial Rates of Improvement Period of age/period convergence Period of cohort convergence Other Speed of convergence 6 5 3 Inform judgements in BE assumptions 63 3 2 1 1 Inform judgements in stre
243、ss calibrationsDirectly set stress calibrations Other includes setting a minimum cohort age and using a set calibration age range 9 9 8.13 If you use cause of death models,what do you use them for?5 2 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Clas
244、sification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.FemalesMalesYE20 MedianYE21 Median1.51.51.51.5Power BI DesktopLongevity -Annuities (Stress)IM The approaches to modelling long
245、evity risk are largely unchanged from YE20,with the majority of firms continuing to use a normal model.As in last years survey,many firms consider multiple models in deriving their longevity trend risk calibrations.Of the companies that allow for data and event risks in their longevity trend risk mo
246、dels,most allow for diversification between these risks,whilst one firm considers event and data risk to form part of a single overall stress continuum.8.14 Which of the following models do you use for modellinglongevity base mortality stresses?Cairns-Blake-Dowd Age-Period-Cohort Cause of death/caus
247、al/Cause of death/caLee-Carter CMI model Other 6 4 4 4 1 1 11%11%78%Normal Cause of death /causal /driver based model Other driver-based model Other includes using a Log-normal distribution combined with expert judgement 8 8.15 Which of the following models do you use for modellinglongevity trend st
248、resses?Other includes using a range of stochastic models to assess the model risk element of stress.8 8.16 How do you consider the Data Risk and Event Risk in your longevity trend risk model?29%43%29%Components of our model Model validation against such risk scenarios Other Other includes as part of
249、 an expert judgement overlay to the outputs of a stochastic model,and covering data risk elsewhere from the longevity trend risk model.20222022 KPMGKPMG LLPLLP a a UKUK limitedlimited liabilityliability partnershippartnership andand a a membermember firmfirm ofof thethe KPMGKPMG globalglobal Documen
250、tDocument Classification:Classification:KPMG KPMG Public Public organisationorganisation ofof independentindependent membermember firmsfirms afaffiliatedfiliated withwith KPMGKPMG InternationalInternational Limited,Limited,a a privateprivate EnglishEnglish companycompany limitedlimited byby guarante
251、e.guarantee.AllAll rightsrights reserved.reserved.7 Power BI DesktopLLongevity -Annuities (Stress)IM 8.17a If you include Data Risk and Event Risk /New Information Risk in your longevity trend risk model,can you please set out how do you allow for diversification between them?25%75%Modelled independ
252、ently allowing for correlation between them Part of a single overall continuum 8.17b Please specify whether the magnitude of stress scenario is deemed to vary for the following heterogeneous groups.Gender Age Other Product Same stress across annuitant portfolio Enhanced /non-enhanced lives evel(base
253、)Level(base)stresses vary by broad product type*stresses vary by broad pruduct type.A s 4 3 3 2 2 1 1*A separate trend stress is applied to deferred BPA members and structured settlements(PPOs)than other annuity.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Do
254、cument Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.4 6 Power BI Desktop50504030BaseBest estimate99.5%50504030BaseBest estimate99.5%Longevity Calibrations -Internal Mo
255、del IM The impact of the longevity improvements is similar to that of the 1-in-200 stress.Expectation of life denoted in years in the charts below.8.18a Expectation of Male Life Aged 50 -IM/PIM only 30252015Base Best estimate99.5%1412108BaseBest estimate99.5%1412108BaseBest estimate99.5%8.18b Expect
256、ation of Female Life Aged 50 -IM/PIM only 8.18c Expectation of Male Life Aged 65 -IM/PIM only 8.18d Expectation of Female Life Aged 65 -IM/PIM only 30252015Base Best estimate99.5%8.18e Expectation of Male Life Aged 80 -IM/PIM only 8.18f Expectation of Female Life Aged 80 -IM/PIM only 2022 KPMG LLP a
257、 UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI Desktop3453453453453455050403090.
258、0%95.0%99.5%3455050403090.0%95.0%99.5%Longevity Calibrations -Internal ModelIM Firms who did not respond with a full IM01 submission have been excluded on this page to ensure a like for like comparison between stresses.The ranges and medians are therefore not directly comparable with other Longevity
259、 Calibration insights elsewhere in the report.Expectation of life denoted in years in the charts below.8.18g Expectation of Male Life Aged 50 -IM/PIM only3025201590.0%95.0%99.5%141210890.0%95.0%99.5%141210890.0%95.0%99.5%8.18h Expectation of Female Life Aged 50 -IM/PIM only 8.18i Expectation of Male
260、 Life Aged 65 -IM/PIM only8.18j Expectation of Female Life Aged 65 -IM/PIM only3025201590.0%95.0%99.5%8.18k Expectation of Male Life Aged 80 -IM/PIM only8.18l Expectation of Female Life Aged 80 -IM/PIM only 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Documen
261、t Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopLongevity Calibrations -Internal ModelIM Expectation of life denoted in years in the charts below.8.19a
262、1-in-200 Stress Impact (EoL)-Aged 507654321543212218.19b 1-in-200 Stress Impact (EoL)-Aged 658.19c 1-in-200 Stress Impact (EoL)-Aged 80Expectation of life at male age 80.Expectation of life at female age 6.Expectation of life at male age 65.Expectation of life at female age 5.Expectation of life at
263、male age 50.FemaleMale 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Male
264、 Female Male Female FemaleMale Technical Practices Power BIDesktopSurvey 2022 8.Mortality&Longevity Risk Longevity Calibrations Internal Model IM The tables below sets out,for each age and gender:-Best estimate expectation of life without allowance for mortality improvements-Best estimate allowance
265、for mortality for mortality improvements,as an increase in absolute expectation of life-Overall stress allowance,as an increase absolute expectation of life-Increased in stressed expectation of life,as a percentage of the base without improvements.We note that the stress impact for males is generall
266、y larger than for females.All figures below are shown as market average,with average defined as the mean of the dataset,using data from firms IM 01 templates.Age 50 Female Male Market Average Market Average Base Mortality 36.6 33.3 BE Improvements 2.6 2.5 1-in-200 Stress Impact (EoL)4.4 4.6 1-in-200
267、 Stress Impact (%)*11.1%12.9%Age 65 Female Male 22.8 1.3 2.8 20.3 1.2 2.8 Base Mortality BE Improvements 1-in-200 Stress Impact (EoL)1-in-200 Stress Impact (%)*11.7%13.1%Age 80 Female Male 10.7 0.5 1.4 12.7%9.2 0.4 1.4 14.2%Base Mortality BE Improvements 1-in-200 Stress Impact (EoL)1-in-200 Stress I
268、mpact (%)*Increase in EoL under a 1-in-200 Stress as a%of Base EoL with improvement.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private
269、English company limited by guarantee.All rights reserved.Power BI DesktopOperational Risk CapitalSF/IM This section covers methodologies in respect of Operational Risk Capital.We continue to see a wide range of responses to question 9.1.This reflects differences in the operating models of respondent
270、s and therefore the risks that arise;and differences in the classification of operational risk.It is noticeable that some of the maximums are very high indicating that some respondents have concentrations of risks in certain categories.The median values are more relevant for comparing categories.We
271、have prepared the ranges shown based only on those companies that report non-zero proportions in the category.The responses show that Model Risk is a very serious risk for many companies,this is a similar message to last year.Tightening the control environment around actuarial models is an area of f
272、ocus at the moment which might offer the chance to reduce the capital held for this risk.The high level of capital held for regulatory failures and product flaws reflects the importance of managing the conduct risk agenda.Cyber Risk and Information Security stands out as the responses are very varie
273、d showing variety in both exposure and quantification.It has the highest minimum showing that it is significant for all respondents that categorise it separately and the third highest median showing its overall significance.In relation to model risk,around two-thirds of respondents stated that they
274、had a model risk policy in place.Those respondents that did not have a policy stated that either one was in development,the risk was covered by other policies or there were formal model risk management arrangements in place but not a policy covering them.The majority of respondents have extended the
275、 coverage of the model risk policy beyond actuarial and finance models recognising that a high level of model reliance exists in a number of places within insurers.We found some instances where there were fewer controls mandated for those models outside of actuarial and finance.9.1 What proportion o
276、f your total operational risk capital do each of the following scenarios contribute (%)?Business disruptionChange Management and ProjectsCyber Attack and information securityExternal fraud,financial crime and sanctionsFailed or inappropriate pricing/underwriting processRaw Answer(Numerical)Failure o
277、f third party/outsourcing failureInternal fraudModel risk,including errors in the financial modelsPeople riskProduct flaws and inappropriate sales practices/mis-selllingRegulatory breach0204060 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classificat
278、ion:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Power BI DesktopOperational Risk Capital and CalibrationSF/IM 9.2 Have you implemented a model risk policy to improve model controls,
279、and if so which models are in scope of this policy?Actuarial models Other Financial models Non-financial models 10 9 7 9.3 What type of methodology does your firm use for estimating its operational risk capital requirement?24%6%53%12%6%Statistical frequency/severity models with Statistical models wi
280、th Stat freq/severity models with multiple risk Stat models with cond.dependSimple estimate approach Deterministic approach Othermultiple risk scenarios conditional dependencies Other includes calculating capital deterministically from data from individual risks,making use of deterministic scenario
281、analysis,using a loss data model,and using a hybrid of the scenario options listed above.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a pri
282、vate English company limited by guarantee.All rights reserved.13 16 Power BI DesktopOperational Risk Capital and CalibrationSF/IM It remains common practice to explore a relatively wide number of scenarios to investigate operational risk.Many insurers find the process of holding workshops to explore
283、 operational events to be a useful exercise and therefore insurers are using this as part of their overall risk management as well as to set capital requirements.9.4 For how many different operational risks do you use scenario workshops/expert judgement to set the modelling parameters?821 or above 1
284、1 to 20 10 or below 4 2 9.5 Do you allow for recoveries from corporate insurances onyour operational risk scenarios?Risks from risk assessment Risks from risk assessment procprocesses Historic internal events Control assessments Historic external events Prior years calibration Emerging risks Risk in
285、dicators Forward looking business plans Internal audit findings External events leading to contagion External events leading to contarisk 1413 12 11 10 9 9 7 6 14 15 Yes 46%No 54%9.6 What data do you use in your operational risk calibrationprocess?9.7 How is internal/external data used in your opera
286、tional risk model?To inform the expert judgement Inform expert judgement processprocess To validate operational Validate Op Risk capitalrisk capital Setting distribution parameters Other 13 7 3 1 Other includes applying weightings to data used,and combining scenario analysis using external loss data
287、 with internal data to set model parameters.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All
288、 rights reserved.13 14 3 Power BI DesktopOperational Risk Capital and Calibration SF/IM None of our respondents have made changes to the statistical distributions used to model the frequency or severity of operational risk in the last 12 months.Only one firm stated that it was looking to make change
289、s this year.Given the economic changes seen in recent years,it is not surprising that refining the operational risk model appears to not be an area of focus at the moment.The risk workshops which drive the parameters for use in models is how respondents make sure the operational risk capital takes a
290、ccount of the most recent events and data.The Poisson distribution remains the most common way to model event frequency.For severity,there is a wider variety of distributions used and the use of more than one distribution is also more prevalent.The log-normal distribution remains the most commonly u
291、sed statistical distribution to model severity.9.8 Do you model your frequency and severity distribution separately?No 21%Yes 79%9.9 What statistical distributions are used to model the frequency of your operational risk scenarios?10 Log-normal Generalised Pareto Weibull Gamma Generalised Extreme Va
292、lue Other Poisson Negative Binomial Other 2 3 Other includes the Exponential and Bernoulli distributions,and exposure-based scenario analysis using various distributions.9.10 What statistical distributions are used to model the severity of your operational risk scenarios?Other includes the Normal an
293、d Burr distributions,Cubic Spline,exposure-based scenario analysis using various distributions,and the use of an empirical distribution for each individual risk.15 13 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisatio
294、n of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.14 10 5 2 1 1 5 Power BI DesktopCorrelations,Diversification,and RecoveriesSF/IM Insurers achieve a high level of diversification between operational risks and
295、between operational risks and other risks.Therefore,operational risk contributes less to the overall capital requirement than might appear from the individual scenarios.The correlation parameters that underpin the diversification benefit are relatively subjective and broadly set using pure expert ju
296、dgement.Even the alternative approach of using causal driver analysis is underpinned by expert judgement.The setting of correlation parameters and ensuring that the overall diversification allowance is appropriate will remain an area that insurers need to keep under review.9.11 On what basis are cor
297、relations set between operational risks,and between operational risks and other risks?Pure expert judgement Combination Causal driver approach Other 8 3 1 1 Other includes correlation assessments based on qualitative analysis of systemic risk drivers and used in conjunction with a copula.9.12 What d
298、iversification benefit are you able to achieve (%)?806040200Between Operational Risk806040200Between Operational Risk and Other Risks 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms aff
299、iliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.13 YE20 MedianYE21 Median47YE20 MedianYE21 Median52 51%39Power BI DesktopUse of more/different data:cUse of more/different data:calibrationUse of more/different data:vUse of more/different data
300、:validationEnhance calc.of the MAuSEnhance calculation of the MAuSEnhanceEnhance fitting fitting algorithm algorithm without wit changing form of model Direct integration with credit Direct integration with credit modelling Improve granularity of drivers Nothing planned Dependency and Risk Calibrati
301、onIM In general firms continue to perform their risk calibrations either annually or less frequently depending on the materiality of the risk.Continued enhancements to the IT infrastructure(both software and hardware),alongside cloud computing solutions,have meant that we are seeing that firms are i
302、ncreasingly able to calibrate their models on-cycle,however this is not yet universal practice.A majority of firms are looking to increase the amount or improve the quality of the data they are using for calibration and validation,likely in response to the PRAs challenge of proxy models.Matching Adj
303、ustment under Stress(MAuS)continues to be an area in which firms are seeking to make improvements,often driven by enhanced modelling capacity.A number of firms are also investigating more complex ways of fitting proxy models,for example making use of automated fitting routines.10.1 How frequently do
304、 you calibrate the following risks?1 7 1 8 4 3 1 7 1 2 5 1 2 4 1 1 7 4 3 1 1 5 1 3 3 1 3 2 1 1 2 4 Less frequently than annualAnnuallyHalf-yearlyQuarterlyOther Credit risk Correlation matrix Expenses risk Interest rates risk Longevity risk Mass lapse risk Operational risk Persistency risk Equity ris
305、k Mortality risk Mortality Catastrophe risk Proxy models 10 10.2 Do you calibrate your proxy model off cycle?5 5 3 3 2 2 1 98 Other Yes 38%No 63%10.3 Are you planning any other development to your capital model?Other refers to a respondent which does not have a fixed calibration frequency and rather
306、 recalibrates the risks in response to monitoring triggers or to address regulatory or business needs.2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International
307、Limited,a private English company limited by guarantee.All rights reserved.1 Power BI DesktopDiversification LevelSF/IM We havent seen significant re-basing of dependencies between risks(see the correlations section for further detail)and correspondingly,there hasnt been a lot of movement in the div
308、ersification benefits achieved.There is an upward trend in the market risk diversification benefit which is largely down to sampling differences but is also potentially driven by increased granularity of credit risk modelling.10.4 Diversification amongst life risks as a percentage of totalundiversif
309、ied risk (%)50403020100IM/PIM Entity levelIM/PIM MA FundSF Entity level50403020100IM/PIM Entity levelIM/PIM MA FundSF Entity level 10.5 Diversification amongst market risks as a percentage of totalundiversified risk (%)10.6 Diversification between risk modules (%)70605040302010IM/PIMSF10.7 Total Div
310、ersification (%)70605040302010IM/PIMSF 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All righ
311、ts reserved.IM/PIM EntitylevelIM/PIM MAFundSF Entity levelYE20 MedianYE21 Median283913162218IM/PIM EntitylevelIM/PIM MAFundSF Entity levelYE20 Median35371092321%IM/PIMSFYE20 MedianYE21 Median36332322IM/PIMSFYE20 MedianYE21 Median49483936%YE21 Median Power BI DesktopLevel 1Level 2Capital Management S
312、F/IM The capital buffers that respondents use are highly dependent on their risk profile and chosen confidence level.The graphs show a high level of variability overall,but the interquartile range does show more consistency.We have observed a slight reduction in the capital buffer level between the
313、previous year and this years survey.Some of this is caused due to a different set of respondents in each year.Comparing the responses on a like-for-like basis shows that few respondents have made a change to their capital buffer.However,these are not wholesale changes,and they appear to be refinemen
314、ts only.11.1 What coverage ratio for SCR do you set as the Risk Appetite (%)at the operating company level?2000100 Level 2 Level 1 t=0 position only Projection Other 22%17%61%32%68%17 18 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classif
315、ication:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.Level 1Level 2YE20 MedianYE21 Median8%11.2 Do you calibrate your coverage ratio risk appetite using the Level 1 point
316、or the Level 2 point?11.3 Is your coverage ratio risk appetite calibrated using the t=0 position only or do you perform a projection over the first year?Other includes using bespoke approach along with performing projection over a longer period.Level 1-the level below which immediate action would be
317、 taken,for example dividends would not be paid(Amber-Red boundary)Level 2-the level which is considered a warning and below which actions would be planned to recover the position(Green-Amber boundary)Power BI DesktopCapital Management and Recovery and Resolution PlansSF/IM There are fewer respondent
318、s this year that use a 1-in-10 or 1-in-20 confidence level.A number of other firms stated that the method was a more comprehensive approach that considered a range of different scenarios.11.4 If you calibrate your coverage ratio risk appetite using the Risk Appetite level 1 point,what is the underly
319、ing confidence level?1-in-201-in-10Other32%37%32%The Other methods included a number of different confidence levels both stronger than 1-in-20 and weaker than 1-in-10 17 Base Actions that dont need Basespecific approval and aont neere Actions that dd included in the SCR Contingent Actions available
320、and Contingentfeasible but not specifically Actions available Board approved Planned Actions included within Plannedthe approved business plan Actions included witRecovery Actions that would be taken in the calibration stressesRecovery Actions that would OtherOther9 4 3 1 14 11.5 Which of the follow
321、ing features are considered as part of setting your coverage ratio risk appetite?Adjustments for the level of Adjustments forregulatory the level TMTPof tAdjustments for pension scheme Adjustments for pension schrisks emporaryTemporary adjustments adjustments for fora business issuesAdjustmentsAdjus
322、tments for dif forference difference between be notional and regulatory TMTP Adjustments for economic co Other adjustments to the Other adjustments to the regregulatory positionAdjustments for any capital a Adjustments for restrictions from Adjustments for any restrictiothe capital tiering rules Adj
323、ustments for any capital Adjustments for any capital ttargets in sub-funds Adjustments Adjustments forfor fungibilityfungibility 6 5 5 3 2 1 1 11.6 Which Management Actions do you allow for in your coverage ratio Risk Appetite calibration?11.7 How have you defined the point at which your Recovery Pl
324、an is initiated?At Level 1(moving to Risk At LevelAppetite being Red)1(moving to Ris At a defined level below Level At a 1 but above 100%SCRdefined level below L coverage Based on a defined liquidityBased on a defined liquid trigger Based on a defined Based on a defined operoperational trigger At a
325、defined level below 100%At a defined level below 1SCR coverage(but above MCR)No formal triggers set,use No formalBoard/Management triggers set,us discretion 8 5 5 3 2 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Document Classification:KPMG Public organisatio
326、n of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.13 17 3 3 2Adjustments for economic conditions1 2 Adjustments for capital add-ons Power BI Desktop33 months MonthsLiquidity Risk SF/IM Liquidity and short-term
327、cash needs is an area of increasing focus for insurers.Almost all respondents consider more than one time period in their assessment of liquidity risk.There are also a lot of insurers focusing on the very short-term horizons,which is a continuation of a trend seen last year.A few respondents conside
328、r much longer-term horizons as part of the overall liquidity framework.11.8 What is the shortest time horizon you consider for liquidity risk?4 3 2 1 1 day 4 days 1 week 1 month 4 1 4 9 119 20 11.9 How many time horizons are considered in total?11.10 Which time horizons do you use within your liquid
329、ity risk approach?1 day 4 days 1 week 2 2 weeksWeeks 1 month 22 months Months 33 months Months 44 months Months 66 months Months 11 year Year 55 years Years 10 years 20 Years50 Years4 1 4 3 13 1 12 1 2 11 2 1 2022 KPMG LLP a UK limited liability partnership and a member firm of the KPMG global Docum
330、ent Classification:KPMG Public organisation of independent member firms affiliated with KPMG International Limited,a private English company limited by guarantee.All rights reserved.1 19 7 6 5 2 20 years 50 years1 Power BI Desktop(Blank)0%1-5%6-15%16-99%Liquidity Risk SF/IM The understanding of stre
331、ssed positions and reflecting that not all assets are available at the full market price at all times are commonly allowed for in assessing liquidity.However,the approach taken and factors used still differ widely between companies.The chart below shows that a wide range of stresses are applied in o
332、rder to give a full picture of liquidity risk.We have also showed some information about the haircuts typically applied to different asset classes,there is a wide variety of approaches here with cash as well getting differing treatment between respondents.11.11 If your liquidity risk appetite is bas
333、ed on cash assets available in stressed conditions,what stresses do you apply?Changes to asset availablity /haircuts applied Defined scenario impacting outflows%stress applied to asset related inflows Interest rate stresses leading to collateral calls Combined scenario based on 1-in-X confidence level%stress applied to certain outflows only%stress applied to premium inflows%stress applied to all o