NAIC经济情景改革：VM-21影响分析模型

NAICeconomic scenario reform:Amodelfor VM-21impact analysis Collin DavidsonZohair MotiwallaParker Henley Executive Summary TheNational Association of Insurance Commissioners (NAIC)continues to evaluate a new economic scenariogenerator(ESG)forU.S.principle-based reserving(PBR). To help companiesnavigate these changes,thispaperpresents an illustrative frameworkfor quantifyingthepotentialimpact of thescenario reformfora prototypicalin-forceblock ofvariable annuities(VAs). Utilizing a few keyeconomicscenario setspresented in the NAIC’s firstfield testing, the analysis looks at VM-21statutoryreserve andC-3 Phase IIcapital impacts across a range ofblock characteristics, along withcapitalmarket sensitivities and measures of scenario dispersion in the tail.Thispaper is intended tolay the groundwork fordeveloping anunderstanding of the complex and potentially materialimpactsthat eventualESG reform is poised tointroduce, so that companies can monitor and plan around thealtered risk profile that emerges. Introduction TheNAIC has been engaged for several yearsinreformingthereal-worldESGused inPBRfor U.S.statutoryrequirements. Thescopeof impact from this change is expected to be significant, leading to reserve and capitalimpacts under theVM-20,1VM-21,2andC-3 PhaseII3frameworks, as well as setting a precursor for eventual VM-224and C-3 Phase I5reform. In 2022, the NAIC rolled out an initial ESG field test that evaluated several test scenariosets produced via the Conning GEMS6model. In early 2023, the NAIC released an update summarizing the initialfindings from the first round ofindustryfield testing and laid out a timeline that included a subsequentindustryfieldtest.7The timingfor the latteris yet to be finalized butisexpected to be rolled out toward the end of 2023. While acknowledgingthatthe effects of ESG reform willeventuallyspan across practically all of life and annuityvaluationin theUnitedStates, the focus of this paper is on the potential impact to VM-21 reserves andrisk-basedcapital (RBC)C-3 capital requirementsforVAs.This is no easy task, as there is substantial uncertainty stillsurroundingwhat thefinal economicscenarios could look like, and variable annuitiesintroducemorecomplexitythanother productsdue tocapital marketexposure,hedge modeling, and dynamicpolicyholderbehavior. Given theimportance ofthe ESGreform efforts, this paper seeks to provide a foundational model to review the VM-21 and C-3impacts of thescenarios used in theinitialround of industryfield testing on some prototypical mixes of variableannuity business.Weanticipate this analysis will beuseful to companiesas they navigate theimpact of theConningGEMS scenariosin bothrounds ofindustry field testing. When the NAIC released the initial report of findings from the initial field test, the results were instructive interms ofbroad observations, but because of the heterogeneity in business mix and strategy across the VA industry itarguablyhad limited utilitywhen trying to understand therelative impact to atypicalVAblock. This paper intentionally seeks tomeet this need bygeneralizingon a few business dimensions to present illustrative impacts across key scenariosfrom the field testing. The goal is not to replicate the entirety of theindustryfield test, but to present some keyfindings when comparing theresults produced by theAmericanAcademyof Actuaries(AAA)Interest Rate Generator,whichcurrently serves as the default ESG,versusthe resultsproduced byConning’s GEMS models. The goals of this paper are to evaluate VA reserves and capital between the AIRG and alternativeeconomicscenarioswith respect tothefollowing: Conditional tail expectation (CTE)dispersion: Evaluatingthe distribution of scenario-level results, particularlyfocusing on the CTE70 and CTE98metricsacross scenario sets.Hedgecredit: Calculatingthe level of hedge costorcredit under different scenarioswhen modeling an implicithedging strategy, which is in line with VM-21 Section 9.B.3 guidance for a future hedging strategy.In thiscontext, hedging is intended to insulate against the market risk introduced by the VA guarantees.Marketsensitivities: Utilizinginstantaneous equity shocks of +/-10% and +/-25%, as well as the implied impactof shocking interest rates by comparingscenarios produced usingtwodifferentstarting yield curvesprovidedinthe field testingsets(the yield curve of December 31,2021,and theyield curve of December 31,2019,plus200basis points[bps]).Mix of business:Reviewingnewversusseasoned business,rider mix—Guaranteed Minimum Death Benefit(GMDB)versusGuaranteed Lifetime Withdrawal Benefit (GLWB)—fund mapping,and equity exposure. The next two sections of this paperprovidesome background on theeconomicscenarios and the model used for theprototypicalVAblock. Followingthat, weanalyze the impact of each ofthe various topicsdescribed above and thenshare some concluding thoughts. Scenarios In this paper, we review scenarios fromthreedifferentreal-worldESGs: 1.TheAAAInterest Rate G