北美精算師協會SOA-研究分享：2022-2023年頂尖精算技術

TopActuarialTechnologiesof2022-2023

September|2023

2

Copyright?2023SocietyofActuariesResearchInstitute

TopActuarialTechnologiesof2022-2023

AUTHORSMariannePurushotham,FSA,MAAA

DararithLy,MBA

SPONSORActuarialInnovationandTechnology

StrategicResearchProgramSteering

Committee

CaveatandDisclaimer

TheopinionsexpressedandconclusionsreachedbytheauthorsaretheirownanddonotrepresentanyofficialpositionoropinionoftheSocietyof

ActuariesResearchInstitute,theSocietyofActuariesoritsmembers.TheSocietyofActuariesResearchInstitutemakesnorepresentationorwarrantytotheaccuracyoftheinformation.

Copyright?2023bytheSocietyofActuariesResearchInstitute.Allrightsreserved.

3

Copyright?2023SocietyofActuariesResearchInstitute

CONTENTS

KeyFindings 4

Methodology 6

Section1:TopActuarialTechnologies 8

1.1DataVisualization 8

1.2CloudComputingandStorage 9

1.3PredictiveModeling 10

Section2:NewDataSources 12

Section3:EmergingTechnologies 13

3.1ArtificialIntelligence/MachineLearning 13

3.2Chatbots 13

3.3UnstructuredData 14

Section4:ActuarialCapabilities 15

Section5:ThePathForward 16

Section6:Acknowledgments 17

AppendixA:InterviewGuide 18

AppendixB:SurveyQuestionnaire 19

AboutTheSocietyofActuariesResearchInstitute 29

4

Copyright?2023SocietyofActuariesResearchInstitute

TopActuarialTechnologiesof2022-2023

Overthepastdecade,theemergenceandapplicationofnewtechnologiesandthegrowthofdatasciencehave

presentedbothchallengesandopportunitiesforactuariesandtheroleoftheprofession.Inordertokeepactuariesinformedaboutnewtechnologiesandtheirimpactonthefuturedirectionoftheindustryandtheprofession,theSOAcommissionedLIMRAtoconductaresearchstudyontheTopActuarialTechnologiesof2019toexaminetopactuarialtechnologiescurrentlyinuse,aswellasthoseexpectedtogrowinthefuture.

Thisreportrepresentsthesecondinstallmentoftheongoingseries.TopActuarialTechnologiesof2022-2023providesupdatesonthecurrentandplannedusesofvarioustechnologytypesandtools,highlightsthose

technologiesexpectedtogrowthefastestamongactuariesinthenext12months,andassessesthestatusofadoptionversusexpectationsfrom2019.

KeyFindings

PrimaryActuarialTechnologies

.Thethreemaintechnologiesusedbysurveyrespondentsintheiractuarialworkduring2022-23aredata

visualization,predictivemodeling,andcloudcomputing/storage.Thiswasalsothecaseforthe2019studyoftopactuarialtechnologies.

.Othertechnologieslessoftencitedasusedincurrentactuarialworkincludeblockchain/distributedledgertechnology,versioncontrol/sharedcodingplatforms,androboticprocessautomation.

.Therearealsoemergingtechnologiesthatactuariesexpecttoleverageintheirworkbeyond2023.ThosemostcommonlyidentifiedincludeArtificialIntelligence,MachineLearning,ChatBots,andUnstructuredData.

FrequencyofUse

.Althoughthetopthreetechnologiesinusefor2022-23arethesameasthosecitedin2019,significant

increasesinthefrequencyofuseforthethreemainactuarialtechnologieshaveoccurred.Thepercentageofsurveyrespondentsfrequentlyusingdatavisualization,predictivemodeling,andcloudcomputingtechnologieshasgrownsignificantlysince2019-from42%to60%(datavisualization),16%to40%(predictivemodeling),

and31%to60%(cloudstorageandcomputing).

.Datavisualizationisnotonlythefastestgrowingtechnologyamongactuaries;itisalsousedbymoreactuariescomparedtoothertechnologiesinthesurvey.Intermsofexpectationsregardinggrowthinthetechnologiesusedbeyond2023,actuariesexpecttoseecontinualincreasesintheusageofthethreemaintechnologies,

withnoonesurveyedexpectingtoseeadecrease.

.Factorsdrivingtheincreaseduseofthesetechnologiesincludebothaccountingandregulatorychanges,new

approachestoexperienceanalysisofmortality,morbidity,andbehaviorfactors,andthedrivetowardfasterandmoreeffectiveriskselectiontechniques.

.Dataanalyticsexpertsinterviewedaspartofthisprojectareoptimisticthatactuariesarecapableoflearningnewskills,adapting,andeffectivelyusingnewtechnologiesintheirwork.

5

Copyright?2023SocietyofActuariesResearchInstitute

SoftwareandTools

.PredictivemodelingtoolsmostoftencitedbysurveyedactuariesincludeExcel,R,Python,andSAS.

.DatavisualizationtoolswiththegreatestuseamongsurveyrespondentsincludeMicrosoftOfficeTools(Excel,PowerPoint,Access)(88%ofrespondents),PowerBI(65%ofrespondents)andTableau(47%respondents).

.CloudcomputingandstoragevendorsmostoftencitedbyactuariesincludeMicrosoftAzureCloud(32%)andAmazonWebServices(AWS)(23%).

6

Copyright?2023SocietyofActuariesResearchInstitute

Methodology

Inordertogatherabroadsourceofopinionandinsightregardingnewertechnologiesandtheirapplicationsinactuarialwork,thestudyconsistedofbothqualitativeinterviewsandaquantitativesurvey.AlldatacollectionoccurredbetweenOctober2022andApril2023.

QUALITATIVEINTERVIEWS

Researchersconducted30-minuteinterviewswith18individualswhowereselectedbasedontheirlevelof

involvementwithdataanalyticsandactuarialtechnologyapplicationsfortheirorganization.Ofthe18individuals,14hadalifeandannuityfocus,threehadahealthinsurancefocus,andoneworkedinproperty/casualtyinsurance.Theinterviewquestionsfocusedonthefollowingareas:

.Whattechnologiesareactuariesfrequentlyusingintheircurrentwork?

.Whatdotheyexpecttobeusinginthenext12months?

.Whattechnologiesorareasoftechnologydotheyseeincreasinginthefuture(3-5yearsandbeyond)?

.Howreadyistheactuarialprofessiontousenewtechnologyeffectively?

AninterviewguidewasprovidedtointervieweesinadvanceofthediscussionsandacopyoftheguidecanbefoundinappendixAofthisreport.

QUANTITATIVESURVEY

Followingthe18qualitativeinterviews,ashortquantitativesurveywasdeveloped.Surveyinvitationsweresenttoatargetedsampleofactuariesinthelife,health,retirement,andproperty&casualtyfields,withlineofsightinto

technology,pricing/reserving,andotherfunctions.Thegoalherewastobetterunderstandbroadactuarialprofessionviewsoncurrentandplannedusageofspecifictoolsinvarioustechnologyareas.Thesurveywascompletedby180actuaries.

Basedondiscussionswithinterviewees,thetopactuarialtechnologiesin2022-23werefocusedonthefollowinggeneralareas:

.Datavisualizationisthegraphicalrepresentationofinformationanddatausing

visualelements,suchas

chartsandgraphs,

toprovideameansofidentifyingtrends,outliers,andpatternsindata.

.Predictivemodelingisacommonlyusedsetofstatisticaltechniquestofacilitatethepredictionoffutureoutcomesusinghistoricaldata.

.Cloudstorageisanoffsite,onlinedatastorageandsharingmedium.

.Cloudcomputingisthedeliveryofcomputingservicessuchassoftwareandanalyticstoolsviatheinternet(“thecloud”).Cloudcomputingcanprovidefastercomputingspeedsandeconomiesofscale.

Thesurveyfocusedontheseareas,collectinggreaterdetailregardingspecifictechnicaltoolsusedineachcategory,andthefrequencyofuseofthesetoolsbyactuaries.

Amongthe180surveyrespondents,therewereasmallnumberofinstanceswheremultipleactuariesfromthesamecompanycompletedaresponse.Giventhesmallnumberofinstances,adecisionwasmadetoincludeallindividualsurveyresponsesintheanalysis,regardlessoftheactuary’semployer.

7

Copyright?2023SocietyofActuariesResearchInstitute

SURVEYRESPONDENTPROFILE

Surveyrespondentsrepresentedvariousareasofindustrypractice,withthelargestproportionofrespondents

focusedonthelife,healthandretirementareasinboth2019and2022-23.Therewasaslightlygreaterdiversityofrespondentsin2022-23withgreaterpercentagesofactuariesfromproperty/casualty,othernon-insuranceand

healthareas(figure1a).

Figure1a

SURVEYRESPONDENTPROFILE–PRIMARYAREAOFPRACTICE

2022-23Study

Life

Health

Retirement

Other,noninsurance

Otherinsurance

Property/Casualty

Investment

Management

10%

3%4%1%

1%

6%2%1%

20%42%

15%

50%

20%

26%

2019Study

Also,thereisagoodvarietyintermsoftenureintheactuarialprofession,withaclusterinthe20yearsandlesscategoryandanotherclusterintheover20–30yearcategory(figure1b).Theaveragetenureofactuarial

respondingtothesurveywasjustover21years.

Figure1b

SURVEYRESPONDENTPROFILE–NUMBEROFYEARSINTHEACTUARIALPROFESSION

16.8%

72.6%

10.6%

Under10years

10-30years

Morethan30years

8

Section1:TopActuarialTechnologies

Forpurposesofthisresearch,thetoptechnologiesaredefinedasthosemostfrequentlyidentifiedbyactuariesasexpectedtoincreaseinusageoverthenext12months.Thisdefinitionincludesbothtechnologiescurrentlyusedbytheactuary,aswellasthosethatactuariesarenotusingtodaybutplantobeginusinginthecomingyear.

Thetopthreetechnologiesidentifiedbysurveyrespondentsfor2022-23arethesameasthoseidentifiedby

respondentstothe2019study–datavisualization,predictivemodeling,andcloudcomputingandstorage.Figure2showsthepercentageofrespondentswhoidentifiedthesetoptechnologiesinthe2019and2022-23surveys.Theremainderofthissectionwillexaminetheseresponsesingreaterdetail.

Figure2

TECHNOLOGYAREASEXPECTEDTOGROWFASTESTINTHENEXT12MONTHS*

Datavisualization

PredictiveModeling

Cloudcomputing&storage

0%10%20%30%40%50%60%70%

5

7%

64%

45%

56

51%

%

60%

2019

2022

*Percentagesrepresentthepercentofactuariessurveyedwhobelievetheirusagewillincreaseinthenext12months.

1.1DATAVISUALIZATION

Datavisualizationcontinuestobethemostcommonlycitedtechnologywheregrowthisexpectedforusein

actuarialwork.Fromthecurrentstudy,60%ofsurveyedactuariesutilizedatavisualizationOftenorFrequently,withalmosttwo-thirdsofcurrentusersplanningtoincreasetheirusage(64%),andtheremainingone-thirdexpectingtouseitaboutthesame(figure3).ThemostwidelyuseddatavisualizationtoolsincludeMSOfficeTools(Excel,

PowerPoint,Access)(88%),PowerBI(65%),Tableau(47%),andR/PythonNativeGraphicalTools(41%).

Typically,thesetoolsareusedfordevelopingvisualsforreportsorpresentations(e.g.,pricingormodelingresults,financialresults),fordecision-making(e.g.,predictivemodelselection,assumptionsettingreviews),andforcreatinginteractivedashboardstobeusedfordatareviewandanalysisorforprovidingKeyPerformanceIndicators(KPIs)tolineofbusinessleads.

9

Figure3

CURRENTANDEXPECTEDFUTUREUSEOFDATAVISUALIZATIONBYACTUARIES

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

0.6

0.64

0.57

0.40.42

0.29

0.18

0.11

Rarely/NoneOccasionalFrequentPlantouseMore

CURRENTUSE

2019Study

2022-23Study

Basedoninterviewsandopen-endedresponses,thistrendisdrivenbyincreasedrecognitionofthevalueofdata

visualizationtoolsinunderstandingandexplainingdata,andtheaccessibilityofmorecentralizeddatasourcesto

actuaries(e.g.,centralizeddatawarehouses).Inaddition,pointandclicktools,likeTableauandPowerBI,aremuchmoreaccessibletotheaverageuserwithamuchquickerlearningperiod.

1.2CLOUDCOMPUTINGANDSTORAGE

Usageofcloudcomputingand/orstoragehasincreasedsignificantlysincethe2019study,fromabout31%usingitfrequentlyoroftento69%in2022-23.Whileoneinthreeexpecttousethistechnologyaboutthesameamount,sixintenanticipateincreasedusage.Thisseemstoverifytheshiftfromstoringdataincentralizedwarehouseson

premisestothird-partystorage,oftencloud-basedproviders.Cloudcomputingismainlybeingusedtofacilitatesharingofdata,allowformoreeffectivecodingcollaboration,andincreasethespeedofcomputationformassivemodelsthatarebeingusedbytheindustryintoday’senvironment.

Basedoninterviewsandopen-endedsurveyresponses,thesechangesarelargelytheresultofmajordata-focusedeffortstorespondtorecentaccountingandregulatorychanges.ImplementationsofIFRS17,LTDI,principles-basedreserving,stochasticmodeling,andpredictivemodelingapplicationswereallcitedasprojectswithcurrentfocusbycompaniesthathavefast-trackedtheneedforgreaterstoragespaceandfasterprocessingspeeds.

Amongthosesurveyed,twoplatformsweremostprevalent–amajority(56%)ofrespondentorganizationsuse

eitherMicrosoftAzureorAmazonWebServices.FarfewermentionedotherservicessuchasSnowflake(18%)andDataBricks(13%).

10

Figure4

CURRENTANDEXPECTEDFUTUREUSEOFCLOUDCOMPUTINGANDSTORAGEAMONGACTUARIES

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

0.69

0.6

0.51

0.36

0.31

0.18

0.12

0.33

Rarely/NoneOccasionalFrequentPlantouseMore

CURRENTUSE

2019Study

2022-23Study

1.3PREDICTIVEMODELING

Propertyandcasualtyactuarieshavebeenapplyingthesetechniquesformanyyears,andtherestoftheprofessionisincreasinglyfindingnewapplications.Inreviewingtheresponsesinthissection,itisimportanttokeepinmindthatmostoftheintervieweesandsurveyrespondentsarecurrentlyworkinginthelife,healthandretirement

industriesratherthaninproperty/casualty,investmentmanagementorotherinsuranceornon-insuranceareas.

Theuseofpredictivemodelingandresearcharoundpotentialapplicationshasincreasedsubstantiallywithintheactuarialcommunity.In2019,only16%ofsurveyrespondentsindicatedtheywereusingpredictivemodelingoftenorfrequently,whilethefigurehasmorethandoubledforthecurrentreportto40%(figure5).Thiswascoupled

withadecreaseinthosewhoexpecttousepredictivemodelingmoreinthefuture(from56%in2019to45%in2022-23).However,thepotentialincreaseinthenextseveralyearsisstillquitelargeandthedecreaseshouldbeexpectedgiventhesignificantincreaseinfrequentusers.

Figure5

CURRENTANDEXPECTEDFUTUREUSEOFPREDICTIVEANALYTICSAMONGACTUARIES

0.6

0.5

0.4

0.3

0.2

0.1

0

0.56

45

2019Study

2022-23Study

0.45

0.39

0.4

0.

0.32

0.28

0.16

Rarely/NoneOccasionalFrequentPlantouseMore

CURRENTUSE

11

Copyright?2023SocietyofActuariesResearchInstitute

ExcelandRwerethemostcommonlycitedtechnologytoolsusedforpredictivemodeling,with44%and45%ofrespondents,respectively,indicatingthattheyuseoneofbothofthesetoolsintheirwork.ExcelandRwere

followedcloselybyPython,at39%,andSAS(Baseand/orEnterpriseGuideversions)at15%.

Themostcommonlymentionedusecaseforpredictivemodelinginactuarialworkcontinuestobeanalysisand

modeling/forecastingofproductexperience.Thisincludesmortalityandmorbidityexperience,aswellas

policyholderbehaviorfocusedfactors,includinglapseandsurrender,aswellastheutilizationofguaranteedandnon-guaranteedelectedbenefits.Somerespondentstalkedaboutexperimentspairingpredictivemodelingtoolswithbehavioraleconomicsprincipalsinmoreadvancedapplications.Predictivemodelsallowtheactuarytobetterunderstandthekeydriversofproductexperience,allowingformoreaccuratepricingandforecasting.

12

Copyright?2023SocietyofActuariesResearchInstitute

Section2:NewDataSources

Basedoninterviewswithactuariesandsurveyresponsesfromboththe2019and2022-23studies,theutilizationofnewdatasourceshasemergedasanotherfocusforapplicationsofnewactuarialtechnology.Thegrowthinthis

areaseemstobedrivenbyinsurerseffortstounderwritebusiness(lifeandhealthproductswerebothmentioned)

morequickly,butwithoutsignificantlossofriskassessmentaccuracy.Acceleratedunderwritingprogramsarenowprevalentintheindustryandmanyoftheexternaldatasourcescitedbysurveyrespondentsarefocusedinthis

area.

Almosthalfoftheactuariessurveyeduseexternaldata,includingprescriptiondata(48%)andElectronicHealthRecords(EHRs)(48%),asadditionaldatasources.Othercommonexternaldatasourcesincludewearabletrackers(12%)andtelematics(11%)(figure6).SomeInsurTechcompaniesarealsoexamininglinksbetweenmortalityandgeneticinformationorevendentaldatarecords.

Figure6

USEOFNEWSOURCESOFDATA

Unstructureddata TelematicsGenetictesting

Facialrecognition

Wearabletrackers(Fitbit,AppleWatch,etc)PrescriptionData

ElectronicHealthRecords

YesNo

0%10%20%30%40%50%60%70%80%90%100%

Notapplicabletomytypeofinsurancecompany/Unknown

Whiletechnologythatallowscompaniestoaccessprescriptioninformationhasbeeninuseforadecadeormore,theuseofelectronichealthrecordsisnewtomanycompanies.TheabilityofEHRtoallowcompaniestodirectly

accesshealthrecordscouldreplacethelargelymanualandexpensiveprocessofobtainingphysicianhealthrecords,allowforamorecomprehensivehealthdataprofile,andultimatelyleadtoabetterandlesscostlyriskselection

process.

Despitesignificantlylowercurrentadoption,somenewdatasourceswarrantamention,includinginternet-of-things(IoTs),dronetechnology,facialrecognition,andgenetictesting.Companiesareexploringdatasourcedthrough

fitnesstrackersandtelematics.Theseeffortshaveproducedimmenseamountsofdatacausingchallengesinstorage,understanding,anddefiningapplicableuses.

13

Copyright?2023SocietyofActuariesResearchInstitute

Section3:EmergingTechnologies

Inthethreeyearssincethe2019reportwaspublished,theadoptionandutilizationoftopactuarialtechnologies

increasedatafastpace.Atthesametime,newtechnologiescontinuedtoemerge.Thevelocityoftechnological

advancementisvirtuallycertainthateveryyearwillbringnewtoolsforactuariestounderstandandassessfortheirabilitytomakecurrentprocessesmoreefficient,increasethequalityofwork,andmaintainrelevanceinthe

industry.

Intervieweesandsurveyrespondentsidentifiedanumberoftechnologiesthatarenotcurrentlyinwidespreaduse,butareexpectedtogrowinuseamongactuariesinthenearandlongerterm.

3.1ARTIFICIALINTELLIGENCE/MACHINELEARNING

ArtificialIntelligence(AI)wascertainlytopofmindformany,asevidencedbytheinterviewsandopen-endedsurveyresponses.ActuariesinterviewedforthisstudydiscussedthepotentialforassistivetechnologieslikeAIand

machinelearningtorevolutionizetheverynatureofactuarialwork.AIcanbedeployedtoautomateprocesses,

enhanceriskmodeling,andimprovedecision-making.SomeorganizationsarealreadyusingAIforclaimsprocessingandfrauddetection,aswellasacceleratingtheirunderwritingprocesses.Machinelearningalgorithmscananalyzehistoricaldatatomakepredictionsandenhanceriskmanagementstrategies.

TherewerealsoconcernsregardingAIandmachinelearningvoicedbysomeactuaries.Someseethepotentialforautomationtoreplacelower-levelactuarialrolesinthenearfutureandpossiblyevenhigher-leveljobsovertime.

Therewasalsoconcernthattheskillsetrequiredtousemachinelearningmodelseffectivelyandresponsiblymaybemarkedlydifferentfromwhatactuarieshaveusedinthepast.

“Ifeelasthoughthenatureofactuarialtechnologyandworkhasthepotentialtochangedramaticallyoverthenext

fewyearsasactuariesbecomemorefamiliarwithassistivetechnologysuchasAIandmachinelearning.Theskillset

neededtoproperlyusethistechnologymaybequitedifferentfromwhatactuariesareusedto.”

3.2CHATBOTS

Whileonly19%ofrespondentsarecurrentlyusingChatBots(ChatGPT,etc.),another9%areplanningtousethistechnologyinthenext12months.

Intheinsurancesector,ChatBotsarelargelydeployedinsupportareastohandlemoreroutinecustomerserviceorclaimsfunctions.However,ChatBotscollectimportantcustomerdataduringtheseinteractions,whichwouldallowactuariesanddatascientiststoconductdata-drivenanalysisifthatdatacanbecapturedandorganized.

Someactuariesinterviewedforthisstudyarejustbeginningtothinkaboutpotentialusesoftechnologieslike

ChatGPTinactuarialwork.Oneofthecommonlycitedareastobeginexploringistheabilitytoprovidecode

snippetsinlanguageslikeR,Pythonorothercommonlyusedcodinglanguagesandreceiveaquickresponseonhowtocorrectcodelogic.

Wewouldliketopointoutthatutilizationoftechnologieslikethesecouldbelimitedbyinternalgovernance.SomecompaniesarestillevaluatingtheuseofChatGPTand,inthemeantime,haveprohibiteditsuseduetorisk

concerns.Ifapproved,weexpectthepercentageofuserswouldincreasesignificantly.

14

Copyright?2023SocietyofActuariesResearchInstitute

3.3UNSTRUCTUREDDATA

Currently,44%ofsurveyedactuariesuseunstructureddataandanother7%plantouseitinthenext12months.Typically,unstructureddatadoesnothaveapredefineddatamodelororganizationalstructure.Examples

mentionedincluderecordingsortranscriptsofcustomerserviceinteractions,competitorwebsiteinformation,earningscalltranscripts,andinformationfromsocialmediafeedsorothersimilarsources.

Someactuariesseeanopportunitytobetterunderstandkeydriversoflapseandsurrenderbyminingdata

generatedthroughuseofNaturalLanguageProcessingwithcallcenterinteractionsandOpticalContentReadersconvertingpreviouslyanalogdataintoelectronicformats.Todate,actuarieshavetrackedresultsofkeyexperiencefactorsimpactingprofitability,buttheadditionaldatacollectedthroughthesetoolsholdsthepromiseofprovidingexplanationsfortheresultsastheyemerge.

Finally,figure7summarizessomeadditionaltechnologiesmentionedbybothintervieweesandsurveyrespondents,includingroboticprocessautomation,versioncontrolandsharedcodingplatforms(suchasGit),aswellas

distributedledger/blockchaintechnology.Althoughasignificantnumberofrespondentsindicatedtheyworkwithroboticprocessautomation(33%)orversioncontrol/sharedcodingplatforms(43%),blockchainanddistributedledgertechnologieshavenotmadesignificantheadwaywiththissamplepopulation.

Figure7

OTHEREMERGINGTECHNOLOGIES

RoboticProcessAutomation

Versioncontrol/Sharedcodingplatforms

Blockchain

0%10%20%30%40%50%60%70%80%90%100%

YesNoNotapplicabletomytypeofinsurancecompany/Unknown

Inthecontextofthisstudy,roboticprocessautomationreferstotheautomationoftasksthatarecurrently

performedbyactuaries,buthavepotentialtobecompletelyorlargelyautomated,e.g.,automationofstandard

actuarialfinancialreportsthatdrawfrominternaldatasources.Severalintervieweestalkedaboutthecombinationofcentralizedwarehousesmakingdatamuchmoreaccessibletoactuariesandtheemergenceofnewtools,

allowingthemtoeasilypulldataandpopulatestandardreporttemplates.

Expertsbelievethepaceofdevelopmentandintroductionofnewtechnologieswillcontinuefortheforeseeablefuturemeaningthatactuariesneedtobepreparedtoidentify,evaluate,andadoptnewapplicationsinanefficientmannerinordertoremainrelevantasaprofession.

15

Copyright?2023SocietyofActuariesResearchInstitute

Section4:ActuarialCapabilities

Expertsinterviewedaspartofthisprojectareoptimisticthatactuarieswillcontinuetoevolvebylearningnewskillsandeffectivelyusingnewtechnologies.

Whileneweractuariesaregettingmoreexposuretopredictivemodelinganddatasciencetechniquesintheir

training,itismorechallengingforolderactuariestoacquiretheseskills.Manyoftheexpertsinterviewedbelievedtheactuarialprofessionaldevelopmentofferingsshouldintegratemorecodingandpredictivemodelingtechniquesspecificallyinsupportofactuarialfunctionssuchaspricing,reserving,experienceanalysisandassumptionsetting,andmortalitymodeling.

TheSocietyofActuarieshasimplementedchangestobothitsbasiceducationanditsprofessionaldevelopment

programs,includingapredictiveanalyticsexam,anadvancedtopicsinpredictiveanalyticsexam,andanadditionalcertificationinpredictivemodelingtoexpandtheactuarialskillset.

Severalofthoseinterviewedforthisreportfeltthatactuariesshouldnotbetryingtocompetewithdatascientists,butshouldworktomoreeffectivelycomplementeachother.Therequiredskillsetsareoverlappingbut,whiledatascientistsfocusonunderstandingdataandthestatisticalmodelsavailabletomodelthatdata,actuariesare

insuranceandfinancialservicestechnicalexpertswhounderstandthebusiness,theproducts,andtheaccountingandregulatoryfunctionsoftheindustry.Asactuaries,weareuniquelyqualifiedtoworkwithdatascientiststohelpselectandexplainthestatisticalmodelssothattheycanbeleveragedtothegreatestbenefitoftheindustry.

16

Copyright?2023SocietyofActuariesResearchInstitute

Section5:ThePathForward

Thelastquestioninthesurveyallowedactuariestoaddanyadditionalcommentstheyhadregardingthedirectionofactuarialtechnology.

Afewthemesemergehereincluding:

.AutomationandthespeedwithwhichmanybelieveAItechnologywillallowactuariestoeliminatetime-consumingandmanualprocesseswascitedasabenefit.Thiswasalsomentionedasaconcernintermsofhowfutureentry-levelactuarieswillgaintheunderstandingthatcomeswithdoingthework.

.Collaborationwillbecomemorecriticalovertime.Severalexpressedtheimportance