Post on 07-Jun-2020
The next frontierThe future of automated financial advice in the UK
Table of contents
Executive summary 01
Part 1: Automated financial advice market overview 05
1.Whatisautomatedfinancialadvice? 06
2. The context 08
3. Theregulatoryperspective 15
4.MachinelearningandArtificialIntelligence 22
Part 2: Automated financial advice – a view of individual markets 23
A. Simplefinancialplanning 24
B. Investing 26
C.Definedcontributionpensionsaving 30
D. At-retirement 33
E. Mortgages 36
F. Individualprotection 39
Next steps – key questions to consider 42
Conclusion 43
Endnotes 44
Executivesummary
Muchhasbeenwrittenabouthowautomatedadvisers,morecommonlytermed‘roboadvisers’,coulddisruptthemarketforinvestmentadvice.Inthispaper,weseektoanswerabroaderquestion:whatisthepotentialforautomatedfinancialadvicetogrowintheUK,overthenextdecade,acrossthefollowingsixmarkets?
Simple financial planning
Investing Defined contribution
pension saving
At-retirement Mortgages Individual protection
Theviewsweexpressinthepaperhavebeeninformedbydiscussionswithkeyplayersandstart-upsinthebanking,insurance,wealthandinvestmentmanagementmarkets,aswellasbyaDeloittesurveyofover2,000consumers.1
Pool of potential adoptersOurmainconclusionisthattheUKoffersarichopportunityforautomatedadvice.Thereisasignificant‘advicegap’,drivenby thehighcostofadvice,lowfinancialliteracy,lowengagement andalackoftrustfollowingpastinstancesofmis-selling.2
Withindividualsbeingincreasinglytaskedwithmanagingtheirownpensionprovision,andinthecontextofarelativelylowstatepension,automatedadvicecanplayakeyroleingeneratinglow-costsolutions.Oursurveyresearchpointstoasizeablepoolofpotentialadoptersineachofthemarketsweexamined(seeFigure1).
OurmainconclusionisthattheUKoffersarichopportunityforautomatedadvice.Thereisasignificant‘advicegap’,drivenbythehighcostofadvice,lowfinancialliteracy,lowengagementandalackoftrust following past instances of mis-selling.
01
The next frontier | The future of automated financial advice in the UK
Figure 1. Estimated number of future users of automated advice
Source: YouGov 23-24 January 2017, Deloitte analysis of consumer survey data. Samples: simple financial planning (2,046), investing (842), defined contribution pension saving (342), at-retirement (173), mortgages (989), protection (457). Note: respondents indicated that they would be willing to pay for automated advice in more than one market.
Investing
Defined contribution pension saving At-retirement
Mortgages Individual protection
Simple financial planning33m heads of families
and single adults
34% willing to pay
around 11mpotential adopters
35% willing to pay
around 3mpotential adopters
19m heads of families and single adults aged under 50
around 7mpotential adopters
18m adults with > £5,000 savings and investments
40% willing to pay
around 7mpotential adopters
4m workers aged 30+ with >£20,000 pension savings
40% willing to pay
around 1-2 millionpotential adopters
6m heads of families with children aged <18
45% willing to pay
around 3mpotential adopters
9m workers with a defined contribution pension
38% willing to pay
0202
The next frontier | The future of automated financial advice in the UK
Overall,oursurveyanalysissuggeststhatupto15millionGBadultswouldbewillingtopayforautomatedadviceinatleastonemarket.However,demandvariesconsiderablybyconsumersegment. Atamacrolevelitis:
• high among consumers in their early forties, despite the fact that this group is often considered to be less enthusiastic about new technologies than millennials
• high among those with above average income, despite the fact thatthisgroupismoreabletoaffordtraditionalfinancialadvicethanthosewithaverageincomes
• consistent, overall, across wealth levels with the exception of defined contribution (DC) pensions, where appetite for automatedadviceishighestamongthosewiththesmallestDCpots.ThisisaclearopportunityinourviewandmayreflectthelackofpublicallyavailableinformationoninvestingwithinaDCpension.
Pricingtheseservicesaffordably,however,willbekey.Ourresearchfound that although more than a third of consumers would be willingtopayforautomatedadvice,theamountstheyarepreparedtopayaregenerallylow.Thatsaid,webelievelargepotentialcustomerpools,coupledwithhighlyefficientdigitalsolutions,couldcreateeconomiesofscale,makingautomatedadvicebothaffordablefortheconsumerandviablefortheprovider.
Figure2capturesourkeyinsightsonwhichplayersmaybemostadvantagedineachmarketandthemostappropriatechannelstodeliverautomatedadvice.Itdrawsonourconsumerresearch,interviewswithsubjectmatterexpertsandconsiderationof a range of future scenarios.
Upto15millionGBadultswouldbewillingtopayforautomatedadvice...
Simple financial planning
Investing Defined contribution
pension saving
At-retirement Mortgages Individual protection
Figure 2. Future of UK automated advice research framework
Source:Deloitteanalysis.Note:financialplanningreferstosimplefinancialplanning,suchasdecidingwhethertoinvestinanIndividualSavingsAccount(ISA)orapension.
Digital only Digital only
Hybrid
Banks Banks
Large, passive fund managers
Retail investment platforms
Life insurers Life insurers
Start-ups in D2C role
Start-ups in B2B role
Optimal channel
Advantages players
03
The next frontier | The future of automated financial advice in the UK
Intermsofchannels,formorecomplexdecisions,customerswillstillwanttospeakwithanadviserinadditiontousingawebsite–theso-calledhybridmodel-toensuretheyhavemadegoodchoices.Hybridmodelsarealsolikelytobeprevalentincertainmarkets(e.g.mortgages,at-retirement)becausetheycanensuresafeguardstothesuitabilityofadvice.
Intermsofindustryplayers,nosingletypeoffinancialservicesproviderisadvantagedacrosstheentireopportunityset.However,incumbentsgenerallyenjoyanadvantageoverstart-upsin direct-to-consumer(D2C)opportunitiesduetotheirexisting scale and brands.
What does the future look like?Todate,automatedadvicehasbeenmostprevalentinwealthmanagement,whichwebelievewillremainfertilegroundforinnovation,giventheincreasedneedforhighernetreturnsfromsavingsinalowinterestrateenvironment.However,thelowcost,highconvenienceandconsistencyofautomatedadvicewilldriveadoptionintootherlargemarkets,especiallythosewherethereisasignificantunmetneedforadvice,suchasretirementproducts,andthosewhichareheavilyintermediatedbutoftenhaveinefficientcustomer-facingprocesses,suchasmortgages.
Lowconsumerfinancialliteracyandengagementarethekeydemandbarrierspreventingautomatedadvicefromreachingitsfull potential. Supply barriers include high customer acquisition costs,lowfeesdrivingthinmargins,incumbentfearsofcustomerswitchingintolowermarginproducts,andregulatoryrisk.Someofthesechallengescanbeovercomebyusingtechnologytomakeiteasierforconsumerstoengagewiththeirfinances,creatingcostefficienciesthrougheconomiesofscale,targetingservicesatparticularconsumersegments,andeffectivecompliancecontrols.
Providersfrequentlyreportthatregulatoryuncertaintyindeterminingwhichservicesareregulatedisabiginhibitor,suggestingthatmoreclarityfromtheregulatorwouldunlockdevelopment.Otherregulatorychallengesincludeclarityofcustomercommunications,thedesignandoversightofalgorithmsandcyberrisk.TheFinancialConductAuthority(FCA)issupportiveofthedevelopmentofautomatedadvicemodelsandisplanningtopublishregulatoryguidanceandfeedbackthatshouldhelpaddresssomeoftheregulatorybarrierstotheirdevelopment.However,automatedadvicecanalsoleadtoregulatoryadvantagesforfirms,by way of a clearer audit trail and consistent customer outcomes.
IntheUS,certainincumbentsdominateintermsofassetgatheringandaredrivingthepathandpaceofadoption.However,thisisduetotheirdecisiontoproactivelyexploitthisopportunity,disruptingmarketsevenattheriskofsomecustomersswitchingtolowermarginproducts.Webelieveadoptingsuchamindsetwillbenecessary,otherwiseincumbentsriskbeingbystanderstothefuture of their industry.
Start-ups,however,haveavarietyofopportunitiesinbothservicesprovidedtootherbusinessesandD2C.Theiradvantagesintechnology,focusandnimblenessenablethemtobuildmoreengaging customer interfaces and adapt faster to changing markets.Duetothehighhurdleofcustomeracquisitioncosts,webelievetheirkeyopportunityinitiallywillbetoofferwhitelabelsolutions to bigger players.
Thispaperisdividedintotwoparts:
• Part 1exploreswhatismeantby‘automatedadvice’,providesanoverviewoftheUKautomatedadvicemarket,andassessestheregulatory challenges.
• Part 2 discusses the opportunities arising from automated advice,thebarriersandtheimplicationsforindustryplayersacrossthefollowingmarkets:simplefinancialplanning,investing,DCpensionsaving,at-retirement,mortgagesandindividualprotection.Itthendrawsoutourkeyquestionsforfirmsandoverallconclusions.
Todate,automatedadvicehasbeenmostprevalentinwealthmanagement,whichwebelievewillremainfertilegroundforinnovation,giventheincreasedneedforhighernetreturnsfromsavingsinalowinterestrateenvironment.
04
The next frontier | The future of automated financial advice in the UK
Part 1Automatedfinancialadvicemarketoverview
0505
The next frontier | The future of automated financial advice in the UK
1.Whatisautomatedfinancialadvice?
Theterms‘automatedfinancialadvice’,‘roboadvice’and‘digitaladvice’areoftenusedinterchangeably,butwhatexactlydotheymean?Theycanbeusedtodescribeservicesrangingfromregulatedadvice,unregulatedguidanceanddiscretionaryinvestmentmanagement.3Theprovisionoftheservicecanbepurelyautomated,orpartlyautomatedwithsomehumanintervention.Theservicemaybeprovidedbyasinglefirm,orbymorethanonefirmpartneringtogether.Thereisnoconsensusonexactlywhichservicesandchannelsareincludedwithintheseterms.Automatedadviceisnottypicallydefinedinlegislationorregulation.Descriptionsbyregulatorshavecommonelementsbutdifferindetails.4
Figure3showstheservicesandchannelsassociatedwiththeterm‘automatedadvice’.Inthispaper,whenwereferto‘automatedadvice’,wemeanthisinagenericsenseincludingalloftheservicesand channels which are shaded in Figure 3. Where we are only referringtocertainservicesorchannels,wewillusespecificterms,suchas‘regulatedadvice’,‘guidance’and‘fullyautomated’.
Currentlythecomputeralgorithmsusedtoprovidetheseservicestendtobeprogrammedtodeliverpre-determinedoutcomesbasedonthedatainputted.However,lookingaheadthereisanopportunityforartificially-intelligentlearningalgorithmstoenable automated systems to respond to a broader range of more complex scenarios. This is further explained in Section 4.
Lookingaheadthereisanopportunity forartificially-intelligentlearningalgorithms to enable automated systems to respond to a broader range of more complex scenarios.
06
The next frontier | The future of automated financial advice in the UK
Figure 3. What is automated financial advice?
Service and regulatory status in the UK*
Regulated advice **Advicegiventoapersononthemeritsoftakingaspecificaction(e.g.buying,selling)inrelationtoaspecificfinancialproduct
Guidance (unregulated)Theprovisionofinformation,genericadviceonwhattypesofproduct may be suitable and/or a general recommendation supportingcustomersinmakingtheir own decisions which does not (inandofitself)involveapersonalrecommendation
Discretionary investment management (regulated)Aninvestmentmanagerdecideswhatproducts to buy and sell on behalf of thecustomer,basedonamandateagreed with the customer
Chan
nel
Traditional face-to-faceCustomersinteractwithahuman,whogeneratesadvice,investmentdecisionsorinformation(dependingontheserviceprovided)withouttheaid of a computer algorithm
Face-to-face assisted by an algorithmCustomersinteractwithahuman,who uses a computer algorithm togenerateadvice,investmentdecisionsorinformation(dependingontheserviceprovided)butcanoverridethealgorithmifneeded
HybridCustomers interact with a website but may also interact with a human (e.g.viaawebchatorbyphone),forexampleifcustomershavequestionsorthefirmneedstoaskfor additional information
Fully automatedCustomers normally interact with a website only. They may still be abletospeaktoahumaniftheyneedtoresolveanyITissues,makea complaint or clarify terms and conditions
Key
Automatedregulatedadvice Automated guidance Automatedinvestmentmanagement
* ThedefinitionsofregulatedadviceandguidanceinFigure3arebasedontheUKregulatorydefinitionsbutarenotdirectquotationsbecausethedetailsofthedefinitionsvaryacrossfinancialservicesmarkets.
**Suitabilityrequirementsapplywhenregulatedadviceinvolvesa‘personalrecommendation’,forexampleadvicebasedonaconsiderationofthecircumstancesofthe customer.
07
The next frontier | The future of automated financial advice in the UK
2. The context
Todate,automatedfinancialadviceintheUK,asinmostoverseasmarkets,hasprimarilytakentheformofinvestmentportfoliooptimisationbasedoncustomerneeds,usinglargely(butnotexclusively)passiveinvestmentproducts.Innovationhaslargelybeenledbystart-ups,whicharehamperedbyveryhighcustomeracquisitioncostsandlowcustomerfees.StringentUKregulatoryrequirementshavemadesomeinnovatorsmorecautious.
Outsideofwealthmanagement,therehasbeenverylimiteddevelopmentofautomatedfinancialadvicebyeitherincumbentsor start-ups,withplayerssuchasHabito(mortgages)andWealthWizards(retirementadvice)beingfewandfarbetween.
However,theneedforaffordableadviceisgrowing,withindividualsbeingincreasinglytaskedwithmanagingtheirownsavings,asevidencedbyashiftfromdefinedbenefit(DB)toDCpensions,auto-enrolmentand‘pensionsfreedom’.5 This is in the context of generallylowfinancialliteracyandengagement.
The potential for automated adviceConsumersneedfinancialadviceatdifferenttimesintheirlives.Theirneedsvaryconsiderablyasdoesthecomplexityofthoseneeds.TakingoutamortgagehaslittleincommonwithtoppingupastocksandsharesISA.Thefinancialdecisionsfacingindividualsatretirement(whenwealthtypicallypeaks)areverydifferentfromthosefacingindividualsjustjoiningtheworkforceforthefirsttime.Wehaveidentifiedsiximportantadvicescenarios(seeFigure4).
Figure 4. Illustrative customer life cycle, wealth by age, UK
Wealth
16-24 25-34 35-44 45-54 55-64 65-74 75-84 85 and over
Retirement
Retirement saving
Finding a mortgage
Buying lifeinsurance
Investment
Financial planning
08
The next frontier | The future of automated financial advice in the UK
Figure 5. Usage of advice on product purchases among GB adults, 2014-16
0% 5% 10% 15% 20% 25% 30% 35% 40%
Non-professional advice
Other professional advice
An accountant or solicitor
Don't know
Mortgage broker
My employer/previous employer
A bank – advice for a fee
A website or online platform
An independent financial adviser
A bank – free advice
Family and/or friends
Purchased without advice
Not purchased product
Main paid-for advice channels
Source: YouGov 23-24 January 2017, Deloitte analysis. Sample: 2,046.
36%
20%
12%
11%
11%
10%
7%
6%
5%
4%
3%
1%
0%
Thehighcostofadviceanddisinclinationtotakeit(regardlessofcost)arekeybarriersacrossthesescenarios.Wenotetheverylowtake-upofanyformofpaidadviceamongoursurveyrespondentswiththerelativelywealthymorelikelythanotherstotakeadvice.Forinstance,amongthosewithsavingsandinvestmentsworthmorethan£30,000,afifthhadtakenadvicefromanindependentfinancialadviserinthepastthreeyearscomparedto11percentamongalladults(seeFigure5).TheincreaseddisclosureofcostsandchargesundertherevisedMarketsinFinancialInstrumentsDirective(MiFIDII)from2018couldfurtherheightenconsumers’costconsciousnessforadviceonretailinvestmentproducts.
Incontrasttoface-to-faceadvice,automatedadvicehasseveraladvantages.Themostcommonlyacceptedare:
• lower costs • highconvenience,givenitsavailability24x7 • consumers may be more willing to disclose details of their financialsituationtoamachineratherthantoahumanadviser
• potentialtoincreaseefficiencyforproviders • adviceisconsistentacrossclientsandcanprovideafullaudittrail.
09
The next frontier | The future of automated financial advice in the UK
Weanalyseconsumers’attitudestowardsautomatedfinancialadviceacrosssixscenariosanddrawoutthekeyimplicationsforindustryplayerslaterinthepaper.However,wesummarisesomeoverallthemesthatapplyinallsixscenariosbelow.
More than a third of respondents in all scenarios showed a willingnesstopayforanautomatedadvicesolution(seeFigure6).Weconsiderthesefigureshighinthecontextoflowtake-upofregulatedadviceandthatautomatedadviceisstillintheearlystagesofdevelopment.
More than a third of respondents in all scenarios showed a willingness to pay for anautomatedfinancialadvicesolution.
Figure 6. Willingness to pay for automated advice by scenario
0% 20% 40% 60% 80% 100%
Simple financial planning
Investing £80 monthlypension contributions
Finding a mortgage
Converting £30,000 pensionsavings into a lump sum and
retirement income
Investing £11,000
Finding life insurance
Yes
Source: YouGov 23-24 January 2017, Deloitte analysis. Samples: Finding life insurance (457), investing £11,000 (842), converting £30,000 pension to lump sum and income (173), finding a mortgage (989), investing £80 monthly pension contributions (382); simple financial planning (2,046).
45%
Don’t know No
15% 40%
40% 8% 53%
40% 16% 44%
38% 16% 46%
35% 20% 45%
34% 12% 54%
10
The next frontier | The future of automated financial advice in the UK
Attitudinal barriers to seeking automated adviceWealsoaskedthoseindividualswhowereunwillingtouseanautomatedsolutiontoexplainthereasonsfortheirreticence,to gauge the attitudinal barriers to adoption. These are remarkablysimilaracrossthesixadvicescenarios,andspeaktothereassuranceconsumersseekaboutthecapabilityoftheseservices,aswellastheneedforanaffordableprice.
Thetwomostcitedreasons(byaroundathirdofrespondents)arealackoftrustinadigitalsolution,aswellasanunwillingnesstopay.Thisisunsurprisinggiventhattheseservicesdonotyetexistatanysubstantialscale.Inaddition,theunwillingnesstopayisalsoevidencedbytheexistenceofan‘advicegap’:peopleunabletogetadviceatapricetheyarewillingtopay.
Asignificantnumberofrespondentssaidtheywouldfinditeasiertospeaktoafinancialadviserthanuseawebsite,implyingthatbuildingeasy-to-usecustomerinterfacesiskeytosuccess.
Withthenotableexceptionsofat-retirementandmortgages,morethan20percentofrespondentsineachoftheotherfourscenariosalsosaidtheywouldnotwanthelpwiththesedecisions.Inourview,thismakesitclearthatprovidermarketingwillneedtoeducateconsumersaboutthevalueofadviceandtacklelowengagementincluding,butnotlimitedto,financialliteracyandinertia.
Despitetheseconcerns,evenamongthoseunwillingtopay forautomatedadvice,veryfewrespondents(atenthorless)ineachscenariobelievetheirfinancialaffairsaretoocomplexfor a website to handle. While some may well be under-estimating the complexityoftheirfinancialaffairs,thissuggeststhatifproviderscandeviseandpriceappropriateservices,themarketforaffordableautomatedadvicecouldpotentiallybeverylarge.
How do attitudes vary by consumer segment? Weanalysedconsumers’willingnesstopayforautomatedadviceacrosstenkeyvariablesincludingage,incomeandwealthtoidentifywheredemandishighest.Figure7outlinesinsightsfromthisanalysis.
Figure 7. Insights from analysis of consumers’ willingness to pay for automated financial advice by demographic
Source: YouGov 23-24 January 2017, Deloitte analysis.
Income
Demand rises with income
Wealth
For financial planning and investing, demand is consistent across
wealth levels
Pension pot
For pension contributions, demand is greatest among
those with pots of less than £10,000
Gender
Men keener for automated advice
than women
Region
Londoners are keenest for automated advice; people in Yorkshire have least interest
Children
Parents want automated advice more than those
without children
Social media
LinkedIn users keenest; Facebook
users less so
Age
Demand highestamong people
aged 18-44
11
The next frontier | The future of automated financial advice in the UK
Automated financial advice has an older target market than many might assumeAutomatedadviceisoftenassociatedwithmillennials(peoplebornbetween1980and2000)becausethisgenerationisconsideredthemosttech-savvyandouranalysisshowsthatdemandisindeedhighestamongtheyoungestagegroupsinthesurvey.However,wealsofoundthatdemandishighacrossfivescenariosamongthoseaged35-44.Inthecaseofpensioncontributions,43percentofrespondentsaged35-44woulduseautomatedadvicecomparedto24percentand21percentofthoseaged45-54and55+.
Theneedforaffordableadviceisgrowing,withindividualsbeingincreasinglytaskedwithmanagingtheirownsavings,asevidencedbyashiftfromdefinedbenefit(DB)toDCpensions,auto-enrolmentand‘pensionsfreedom’.
Figure 8. Willingness to pay for automated advice by age
Finding life insuranceFinding a mortgageInvesting £80 monthlypension contributions
Investing £11,000Simple financial planning
51%
43%46%
32%
20%
63%
55%
63%
37%
23%
36% 35%
43%
24%21%
39% 40%43%
26%
59%
44%
35%
18-24
Source: YouGov 23-24 January 2017, Deloitte analysis. Samples range from 50-793. Excludes at-retirement due to small samples.
25-34 35-44 45-54 55+
12
The next frontier | The future of automated financial advice in the UK
Automated financial advice has the potential to prove popular among people with high incomesDuetoitslowcost,automatedadviceisconsideredapotentialsolutionforthoseonlowerincomeswhostruggletoaffordmainstreamfinancialadvice.However,counterintuitively,wefoundthatacrossfiveadvicescenariosdemandriseswithincome.Forinstance,51percentofthoseonincomesof£45,000to£70,000woulduseanautomatedfinancialplannerbutonly28percentofthoseonincomeslessthan£15,000woulddoso.Forpeopleonlowerincomes,consumerengagementiskeyasthisgroupmaynotperceiveaneedforfinancialadvice.
Asimilarstoryemergedwhenwelookedatdemandforautomatedadviceacrossdifferentwealthbrackets,withdemandsurprisinglyhighamongthosemostabletoaffordfinancialadvice.Forinvesting,forexample,respondentswith£100,000ormoreinsavingsandinvestmentsarealmostaslikely(36percent)asthosewith£1,000ormore(40percent)tooptforautomatedadvice.Inourviewthisreinforcestheconclusionthatautomatedadviceshouldnotbeexclusivelyaimedatlowerwealthsegments.Indeed,incumbents are right to be worried about the potential for wealthy clients to switch into their lower margin products.
Figure 9. Willingness to pay for automated advice by income
Less than £15,000
Source: YouGov 23-24 January 2017; Deloitte analysis. Samples: 55-262. Excludes at-retirement due to small samples.
Finding life insuranceFinding a mortgageInvesting £80 monthlypension contributions
Investing £11,000Simple financial planning
28%
39%
46%
51%
30%
39%
50% 51%
20%
32%
36%
46%
30%
46% 47%
55%
36%
54%52% 53%
£15,000 to £29,999 £30,000 to £44,999 £45,000 to £69,999
13
The next frontier | The future of automated financial advice in the UK
There is a high demand for automated financial advice on pension contributions among the less wealthyForty-fivepercentofthosewithpensionpotsworthlessthan£10,000wouldpayforautomatedadviceonwheretoinvesttheircontributions.Incontrast,amongthosewithpotsof£75,000ormore,thefigureisroughlyathird.WebelievethismayreflectthelackofadviceandinformationthatispublicallyavailableonDCpensionsforthosewithlessthan£50,000toinvest.Incontrast,there is a plethora of websites aimed at small non-pension investors.Wefeeltherelativelyunderservedlow-pension-wealthsegment is a clear opportunity.
Inadditiontothedemographicgroupsdiscussedabove,webelievethereisscopeforproviderstoidentifytargetconsumersegmentsbasedonother,less-obvious,demographic,socio-economicandculturalfactors.InFigure10wehighlightfourpotentialtargetconsumersegmentswithabove-averagepropensitytopayforautomatedadvice.
Figure 10. Characteristics of potential target consumer segments
Consumer segment
Young high earners Tech-savvy over 45s Small DC pot holders Southern home owners
Definition GBadultsaged18-34 with>£25,000income and>£5,000insavings &investments (excludingpensions andpropertywealth)
GBworkersaged45+who useLinkedinwithpension savingsof>£20,000
GBworkerswithaDC pension and total pension savings<£40,000
GBadultsagedunder50 wholiveinLondonor the South East and own property
Number of consumers in segment
1.2 million 1.2 million 2.6 million 3.0million
Willingness to pay for automated advice (peer average)
56%wouldpayfor automatedadviceoninvesting£11,000(40%)
44%wouldpayfor automatedadviceon converting£30,000 pensionsavingsintoa lump sum and retirement income(40%)(35%amongconsumersaged45+)
45%wouldpayfor automatedadviceon investing£80monthly pension contributions (35%)
45%wouldpayforautomatedadviceonfindingamortgage(38%)
Net wealth excluding pensions and properties (peer average)
20%havenetwealthof£10,000-£30,000(11%)
10%havenetwealthof £100,000-£250,000(5%)
28%havenetwealthof<£1,000(15%)
10%havenetwealthof£100,000-£250,000(5%)
Net property wealth (peer average)
• 63%ownproperty(66%) • 6%ownpropertyworth£20,000-£40,000(3%)
• 82%ownproperty(66%) • 19%ownpropertyworth>£500,000(6%)
• 66%ownproperty(75%) • 13%ownpropertyworth£150,000-£200,000(10%)
25%ownpropertyworth>£500,000(6%)
Pension holdings (peer average)
43%holdaDC pension(23%)
33%holdaDCpension(23%) 7%belongtoaDB scheme(27%)
23%holdindividual pensions(12%)
Value of all pensions (peer average)
48%havetotalpensionwealthof£75,000-£100,000(1%)
8%havetotalpensionwealth of£75,000-£100,000(1%)
N/A 9%havetotalpensionwealthof£20,000-£40,000(7%)
Note:FormoreinformationonpeeraveragesseeEndnote7.Source:YouGov23-24January2017,Deloitteanalysis.Samples:younghighearners(54),tech-savvyover45s,smallDCpotholders(112),southernhomeowners(129).
14
The next frontier | The future of automated financial advice in the UK
3.Theregulatoryperspective
Makingfinancialadvicemoreaffordableandattractivetocustomers,includingthosewithsmalleramountsofmoneyormorebasicneeds,isapriorityfortheFCA.Itispartofawiderpublicpolicyobjectiveofensuringindividualssavemoreandarebetterpreparedfortheirretirement.Regulators,bothintheUKandelsewhere,believethatautomatedadvicecouldplayapivotalroleinachievingthisaim.
Intheory,firms’incentivesshouldbealignedwithregulators’inprovidinganefficientandcost-effectivesolutionwhichmeetscustomers’needsandeffectivelymanagesrisks.Automationalsohasthepotentialtooffersometangibleregulatorybenefits.Onefirmplanningtoofferadviceonastandardandlimitedsetofportfoliostoldusthat,providedfirms“getthealgorithmsright”,automatedadvicecouldactuallybe“compliancenirvana”ensuringbothconsistencyofadviceandaclearautomatedaudittrail.
However,inpractice,marketdevelopmentshavebeenslowerthanexpected,withregulatoryuncertaintyandriskconsistentlycitedinourdiscussionswithfirmsasoneofthemainbarrierstotechnologicalinnovationandautomation.CurrentregulationsintheUK,aswellasinothercountriessuchastheUS,AustraliaandGermany,aredesignedtobe‘technologicallyneutral’-whetheryougiveadviceinpersonorviaawebsite,firmsarerequiredtomeet the same set of customer protection rules. So if the rules arethesame,whytheuncertaintyandaddedrisksinrelationtoautomatedadviceandhowcanfirmsaddressthem?
Inthefollowingsectionsweexplorethisquestionfocusingontheadviceboundary,customercommunications,designandsupervisionofalgorithms,andcyber-resilience.Finally,weconcludeonwhetherautomatedadvicecanoffer‘compliancenirvana’, orwhetherit’sactuallya‘compliancenightmare’.
“Automatedadviceisa‘compliancenirvana’.Ifyougetthealgorithmsright,itcan’tgowrong,unlikewithhumanadvisers.Andthereisanaudittrail.” Deloitte interview with wealth manager, 2016
“Thedistinctionbetweenadviceandguidance,oncereasonablyclear,hasbecomemuchgreyerwiththeadvent of platforms and the potential of roboadvice.” John Griffiths Jones, Chair of FCA, February 2017
The advice boundaryAkeyregulatorychallengeforfirmsinprovidingautomatedadviceisunderstandingwhichsideofthe‘adviceboundary’theyfallon–guidanceorregulatedadvice–andhencewhichregulatoryrequirements apply.
Thischallengeexistswhetherafirmprovidestheseservicesonlineor face-to-face. The reason that this is a particularly thorny issue forautomatedmodelsisthatsomeservicesthatareclosertotheboundaryarenotprofitabletoprovideface-to-facebutcouldlendthemselvesverywelltoonlinedistribution,especiallyforretailinvestmentproducts.
Figure11(seeover)illustrateshowthecostofprovidingtheserviceincreasesasfirmsmovefromprovidingguidancetoregulatedadvice,duetotheadditionalconsumerprotectionrules,increasedcomplianceandoversightcosts,andenhancedliabilitythatapplytoregulatedadvice.Therateofreturnthatfirmsreceivealsoincreases,asthegreatervaluetheyprovidetocustomersthroughregulatedadviceratherthanguidancemeanstheycanchargemoreand/or increase customer loyalty.
15
The next frontier | The future of automated financial advice in the UK
Figure 11. Advice boundary8
Guidance Uncertainty Regulated advice
Revenues Costofdeliveringadvice
Notes:
i)Thefigureisdesignedtoillustratethatthecompliance(andrelated)costsofdeliveringadvicerisesteeply,duetothecurrentregulatoryuncertainty,forthoseservicesclosetotheboundarybetweenguidanceandregulatedadvice.Thisisbecausefirms,unabletodetermineexactlywheretheboundaryis,maychoosetocomplywithmorestringentrequirementsthatapplytoregulatedadvicetomakesurethattheyprotectthemselvesagainstanysubsequentclaimthattheservicetheyprovidedwasregulatedadviceratherthanguidance.Therevenuelineindicatesthat,ingeneralterms,revenueswilltendtoriselesssteeplythanthecosts,ascustomerswillnotbewillingtopayverymuchuntilfirmsofferthemactualpersonalisedadvice.Revenuesandcostslinesinthisfigureareforillustrativepurposesonly.
ii)Wherethischartrefersto‘regulatedadvice’,thismeanscaseswherearegulatedpersonalrecommendationhasbeenmadeandsuitabilityrequirementsapply. SeeEndnote8formoreinformation.
Customers enter information, such as on their outgoings and whether
they are saving for retirement, as part of an online financial
‘health check’, which suggests they consider investing more each month in a stocks and shares ISA
A firm sells products via its website. The website has a filtering
tool which allows customers to filter products by objective factors
(e.g. product type, asset class)
Having suggested that the customer invests more in a stocks and shares
ISA, the tool shows the customer a list of 10 ‘high performing funds’
available within an ISA, with a disclaimer saying that this is not a
personal recommendation
In addition to the filtering tool, the website gives general information
about which kinds of products may be appropriate for people with
different goals (e.g. saving for a car, retirement plans)
Customers are asked a more detailed set of questions, including
on their financial circumstances, objectives and attitude to risk, and
they are recommended specific investment products to buy within
the stocks and shares ISA
The filtering tool allows customers to filter products based on factors relating to their life and situation,
such as goals, age, use of tax wrappers, marital status or other
products the customer has
Cost
of d
eliv
erin
g ad
vice
/rev
enue
s
16
The next frontier | The future of automated financial advice in the UK
However,firmscurrentlyviewthecostsandriskofprovidingguidanceclosetotheadviceboundaryastoohigh,asregulatoryuncertaintymakesittoodifficultforthemtounderstandwhereexactlytheboundaryis.Toillustratethis,Figure11providessomeexamplesofadditionalservicesfirmscouldprovidetoclients,butwhereregulatoryuncertaintyactsasabarrier.Ingeneral,muchoftheuncertaintyhingesontheamountofinformationthefirmcanaskthecustomertoprovidebeforeittipstheserviceintopersonalrecommendation,andthusregulatedadvice.
So how can this barrier be overcome? Fromapublicpolicyperspective,thereisatrade-offbetweenenablingconsumerswhocannotaffordadvicetoobtainsomehigh-leveldirectiononwhatproductsmayberelevantforthem,andreducingtheriskthataconsumerissteeredtowardsaproductwhichisnotsuitableinthelightoftheirdetailedfinancialcircumstances.
AspartoftheFinancialAdviceMarketReview(FAMR),HMTreasury(HMT)andtheFCAareseekingtofindtherightbalanceandaretakingforwardtwomainactions.9First,theFCArecentlylaunchedthe‘AdviceUnit’,throughwhichitwillsupportfirmsdevelopingautomatedadvicepropositionsthatcanhelpprovidelowcost,highqualityregulatedadvicetocustomersintheareasofinvestments,pensionsandprotectionmarkets.10Itwilldosobyprovidingthemwithregulatoryfeedback,bothindividually(forfirmsselectedtoreceivedirectsupport)andbypublishingpublicresources,includingfurtherguidanceonstreamlinedadvice,startingfromApril2017.11,12
Second,HMTintendstonarrowtheUKdefinitionofinvestmentadviceforregulatedfirmstobringitintolinewiththeEUdefinitioninMiFIDII,sothatnon-personalrecommendations(whichdonottakeintoaccounttheclient’spersonalcircumstances)providedbyregulatedfirmswillnolongerbeconsideredregulatedadvice.13 Thisshouldmakeiteasierforfirmstoprovidemoredetailedinformationtocustomersaboutthetargetmarketorriskprofileoftheirproducts,withlessriskofinadvertentlystrayingintoregulatedadvice.
Bothoftheseactionswillbewelcomedbyfirms.TheirsuccesswilldependinpartonhowmuchclaritytheFCAisabletoprovideinitsguidance and on the extent to which the FCA is prepared to allow firmstoprovide‘semi-personalised’recommendations,i.e.whereafirmgivesacustomerasteerusingbasicpersonalinformationsuchastheirageandgoalsbutdoesnotaskmoredetailedquestionsabouttheirfinancialcircumstances.
Fornow,toavoidundueregulatoryriskfirmswilleitherchoosetostayclearofthe‘greyarea’andprovideservicesthatareclearlyonlyguidanceorregulatedadvice,orerronthesideofcautionandassumetheyareenteringregulatedadviceterritory.Theywillonlydothelatteriftheybelievethattheeconomicbenefits,directorindirect,intheshortandmediumtermofferanacceptablereturnoninvestment.
Customer communicationsAnotherchallengespecifictoautomatedadvicemodelsisthat,withouthumaninteraction,allinformationprovidedtocustomers,aswellasquestionstocustomersandpossibleanswers,mustbespecifiedinadvance.Thecustomerhaslimitedopportunitytoseekclarification,andtheautomatedadvisercannottestthecustomer’sunderstandingthroughconversationorreadingbodylanguage.
“Informationitselfdoesnotnecessarilyempowertheconsumer.Ourworkonbehaviouraleconomicshasclearlyshownitcanoverwhelm,confuse,distractorevendeterpeoplefrommakingeffectivechoices if presented in a way people struggletoengagewith.” Christopher Woolard, FCA director of strategy and competition, June 2015
Inaddition,anumberofweaknesseshavebeenfoundinfirms’communicationsinautomatedadvice.TheFinancialServicesCustomerPanel(FSCP)recentlyreportedthatfirmsofferingautomatedadviceservicesoftendonotuselanguagethatcustomerscaneasilyunderstand,butinsteadusejargonandpotentially misleading explanations.14 The FSCP also found that thelanguageusedbyfirmsoftenassumesanunrealisticleveloffamiliaritywithconceptssuchas‘funds’and‘ISAs’.
The FCA and FSCP found that most customers do not understand the regulatory distinction between guidance andregulatedadviceorthattheseofferdifferentlevelsofprotection.
17
The next frontier | The future of automated financial advice in the UK
The FCA and FSCP found that most customers do not understand theregulatorydistinctionbetweenguidanceandregulatedadviceorthattheseofferdifferentlevelsofprotection.15 Firms need to communicateclearlywhattheyareoffering,andwhatthismeansforthecustomer.Furthermore,customersmayclickthroughdisclaimerswithoutreadingthemthoroughly,ortheymaynot fullyunderstandthem,andthereforebelievetheyarebeing givenapersonalrecommendation.TheFCAstatesthatif “arecommendationisputforwardinsuchawaythatareasonableobserverwouldviewitasbeingbasedonaconsiderationofa
customer’scircumstancesorpresentedassuitable,thenthisislikelytoamounttoapersonalrecommendation”.16 While the FCA agreesthatacustomer’sperceptionoftheservicereceivedwillnotalwaysbecorrect,firmsareresponsibleforensuringtheytakeallreasonablestepstoavoidanyconfusionormisunderstanding.
InFigure12belowweprovidesomeexamplesonthe‘Dos’and‘Don’ts’forcustomercommunicationsinautomatedadvice and guidance.
Figure 12. ‘Dos’ and ‘Don’ts’ for customer communications for automated advice17
Dos
Considerbehaviouralbiases,e.g.presentbias,lossaversion,overconfidence.Encouragecustomerstothinkabouttheirlong-termfinanceswhenansweringquestionsabouttheirpreferences
Makecommunicationsmorevisual,interactiveandengaging.Thiscanincludetheuseofinteractivevideos
Employbehaviouraltestingtogaugewhethersmallchangesinthewaysinformationispresentedorquestionsareaskedboostsengagement and understanding on the part of the consumer
Testwhethercustomershaveunderstoodtheimplicationsofthechoicestheymakeastheyprogressthroughtheonlineprocess(e.g.byaskingquestions,orspecifyingwhatthefirmisassumingaboutthecustomer)andprovideameansforcustomersto seekfurtherclarifications(e.g.onwebchatorbyprovidingaphonenumber)
Explainclearlythenatureoftheservicebeingofferedandwhatprotectionsareavailabletocustomersifthingsgowrong
Ensure that all important considerations are displayed prominentlywhenprovidinginformationoraskingcustomersquestions
Includelinkstopop-upboxeswherecustomerscanobtainfurther explanations if needed
Don’ts
Usejargon,e.g.‘portfoliomanagement’or‘definedcontributionpensionscheme’withoutaclearexplanationofwhatthe term means
Assumethecustomerknowsthedifferencebetween regulatedadviceandguidance,andtheimplicationsfor consumer protection
Includelongdisclaimersorimportantinformationthatconsumerscan‘tickthrough’withoutreading
Refertodetailedtermsandconditionswithoutprovidingasummaryofkeypoints
Overloadcustomerswithexcessiveinformation
Expressriskonascaleof1to5withoutspecifyingwhatthismeans in practice
Keepriskdisclosurestotheendofthecustomerjourneyoronlyin follow-up sales documentation
18
The next frontier | The future of automated financial advice in the UK
Automatedadviceprovidesanopportunityforfirmstomaketheirriskprofilingprocessesmorevisual,interactiveandengaging.Somefirmsareevenexperimentingwithvirtualrealitytechnologytobringinvestmentdecisionstolife.Forexample,StockCity,byFidelity,isavirtualcitywhichhelpsclientsvisualiseaninvestmentportfolio.18 Eachbuildingintheskylinerepresentsastock;itsheightrepresentsthepriceofthestockintheinvestmentportfolio,whileitswidthrepresentsthenumberofoutstandingshares.Theskylineofthecitychangeswiththefluctuationsinthemarket–adownturnwillcausearainstorm,whileasunnydaymeanssharepricesarerising.
TheFCAisworkingwiththeindustrytoimprovetheclarityofcustomer communications through its Smarter Communications work.19Webelievefirmsshouldproactivelycontributetothedevelopmentandimplementationofindustrybestpractices.Theyshouldalsouseinsightsfrombehaviouraleconomicsandundertaketestinginthedesignoftheircustomerinterface,takingintoconsiderationcustomers’vulnerabilitiesandcognitivebiasesand heuristics which can result in customer responses being influencedbythewaythatquestionsareframed.
Note:Figure13showsbehaviouralbiasesthatcanpreventcustomersfrommakingrationalfinancialdecisions.Behaviouralbiasesarenotalignedwithspecificpotentialsolutions. This is because in certain cases potential solutions relate to more than one bias.
Figure 13. Opportunities to deploy behavioural economics in automated financial advice
Provide customer with comparisons vs. peer groups, e.g. savings rate, accumulated wealth. Make advice communications highly personalised so that customers understand their unique needs.
HerdingSeeking comfort from ‘being with the crowd’.
Use websites to present customers’ wealth in an integrated manner that removes inappropriate segregation between mental accounts.
Mental accountingAllocating wealth to different mental accounts and treating them differently when it may be inappropriate to do so, e.g. an ISA may be more valued than a pension because it is more accessible.
Provide advice on ‘micro-journeys’, i.e. short decision-making processes. Deploy advice through easy-to-use mobile apps.
Financial loss aversionAversion to ‘losing’ leads some people to make sub-optimal investment decisions, e.g. taking too little risk. Proactively contact investors that display behavioural
biases. Educate investors with tools that demonstrate the impact of biases. Tailor advice depending on the propensity of the customer to display biases.
Status quo biasPreferring to stick with the current situation, e.g. not taking advice, not updating life insurance.
Anchor customers’ savings and investment decisions based on prominent and appropriate targets. Use gamification to make chasing goals fun and engaging.
OverconfidenceMaking overly optimistic assumptions about one’s financial future.
Proactively contact customers in advance of when they need to take action, e.g. automated alerts to review finances before life events.
AnchoringMaking decisions based on an arbitrary reference point.
Hyperbolic discountingPreferring smaller payoffs now over larger payoffs later.
Educate customers using online tools that demonstrate realistic assumptions, e.g. projected level of retirement income.
Time loss aversionAversion to ‘losing’ time on administration tasks such as managing finances.
Behavioural biases Potential solutions
19
The next frontier | The future of automated financial advice in the UK
Design and oversight of algorithms Thekeysuccessfactorforanyautomatedadvicemodelisthestrength of the algorithm underpinning it. Flaws in its design and poorgovernancemayresultinwidespreadmis-selling,whichbothfirmsandregulatorsareunderstandablykeentoavoid.
Thereforethewayalgorithmsaredesigned,approvedandreviewedisanewandkeyriskforfirmsdevelopingautomatedadvicemodels.
“Wehaverulesaboutwhatsortofexamsahumanadvisermustpassbeforetheyarequalifiedtoadvise;howwouldweapplythesetoanalgorithm?!” Mary Starks, Director of Competition, FCA, June 2016
AustralianandUSregulatorshavealreadypublishedsomeguidance in this area. Firms can use these sources to inform their thinkingandplans.20,21 Parallels and commonalities can also be foundwiththeMiFIDIIrulesoncontrolsforalgorithmictradingandthesupervisionofinternalmodels.
“Ournumberoneconductriskconcernis the potential for mis-selling en masse.Tomitigatetherisk,thedigitalsolution should recognise when it is no longer appropriate to proceed with the customeralongthejourney.” Deloitte interview with insurer, 2016
1 Aqualifiedadvisershouldalwaysbefullyinvolvedinthedesignprocess,andberesponsibleforunderstandingtheassumptionsandlogicembeddedinthealgorithmineachpotentialscenario,andwhethertheseprovidegoodcustomeroutcomes.
2 Firmsshoulduseafullsetofskillsinthedesignofalgorithms,includingIT,modernportfoliotheory,andbehaviouraleconomics to understand any potential biases in the collection andanalysisofcustomers’data.
3 Exit chutes should be coded into algorithms so that customers whodisplayabnormalbehaviours(potentiallyvulnerablecustomers)orcustomerswithcomplexneedscanberedirectedto human interaction.
4 Governancecommitteesoverseeingautomatedadviceshouldincludeamixofcompliance,technologyandbusinessleadership.Executivemembersshouldbetrainedtounderstandtherisksofusingalgorithms,establishthemetricstobuildintothetestingandqualityassurance(QA)reviewsandregularlyreviewsystemsperformanceandsignoffanychangesrequired.
5 Firmsshouldconduct,anddocument,robusttestingofalgorithms,alwaysinanon-liveenvironment,beforeanyadviceisprovidedtoaclient,andalsosubsequently.(SeeBox1)
6 FirmsneedtoensurethatstaffintheirRisk,ComplianceandInternalAuditteamshavetheskillsneededtounderstandthealgorithms,andsufficientskillstofollowuponinformationprovidedbyautomaticalerts.
7 Firmsshouldhaveclearlinesofaccountability,includingaclearlyidentifiedownerforeachalgorithm,andadequatehumanandtechnologicalresources,throughoutthedevelopment,testing,deployment,monitoring,reviewingandupdatingofalgorithms.
8 Firmsshouldconsiderdeployingalgorithmsgradually,i.e.toalimitednumberofclients,toensuretheyareworkingasexpected.Theirperformanceshouldbecloselymonitored,andthefulldeploymentshouldbesignedoffonlyifQAmetricsarefully met.
9 Thealgorithmownershouldbeinchargeofinitiatingthereviewandupdatingalgorithmswhenevertherearerelevantfactorsthatmayaffectthemand,inturn,thecustomers(e.g.marketorregulatorychanges).
10 Firmsshouldhaveeffectivecontrolsandprocessesinplacetosuspendtheprovisionofadviceifanerrorwithinanalgorithmisdetected.Theyshouldalsoperformamanualreviewofanyadviceprovidedbeforetheerrorwasdetected.
11 Thefirmshouldbeabletoexplaintotheregulatorshowthetoolworksandhowitcomplieswithregulatoryrequirementsandthefirm’sriskappetite.Firmsshouldalsobeabletoprovidedetailsoftheparametersorlimitstowhichthesystemissubjectandthekeycomplianceandriskcontrolsthatareinplace.
Belowisasummaryofsomeoftheessentialprinciplesfirmscanconsiderapplyingwhenestablishingaframeworktodesign,governandsupervisetheiralgorithms.22
20
The next frontier | The future of automated financial advice in the UK
Box 1 – FINRA principles and effective practices for governance and supervision of algorithms
Initial reviews • assessingwhetherthemethodologyatooluses,includinganyrelatedassumptions,iswell-suitedtothetask.
• understanding the data inputs that will be used.
• testingtheoutputtoassesswhetheritconformswithafirm’sexpectations.
Ongoing reviews • assessing whether the models a tool uses remain appropriate asmarketandotherconditionsevolve.
• testing the output of the tool on a regular basis to ensure that it is performing as intended.
• identifyingindividualswhoareresponsibleforsupervising the tool.
“Peopleoftentalkaboutautomatedadvicebeinganinnovationintheadviceprocess,butitalsorequiresinnovationinriskandgovernancepractices.” Deloitte interview with insurer, 2017
Cyber-resilienceFirmslookingtodeveloponlineplatformstoprovideadviceshouldconsidercyberasoneoftheirlargestrisks,bothfromaregulatoryandreputationalperspective.
Fromaregulatoryperspective,firmsshouldensurethattheyhavesufficientlypreparedthemselvesinthreeprincipalareas:23
• risk identification and management–firmsshouldbeabletodemonstrate that they understand the extent of their exposure tocyberandITriskatalllevelsoftheirorganisation
• risk governance–firmsshouldhaverobustinternalgovernancemodelstocopewiththecomplexityandpervasivenessofcyberrisk
• risk resilience–firmsmustdemonstratethattheyhavedevelopedcontingencyplansandcapabilitiestorespondeffectivelytocyberbreachesinawaythatallowsthemtominimise disruption to customers.
Cybersecurityandresiliencearekeyconsiderationsforallinternet-basedcustomerplatforms,notonlyforautomatedadvice.However,firmsbuildingacustomerplatformforthefirsttimewillneedtoensurethatcyberrisk,includingclientdataprotection,isatthetopoftheexecutiveagenda.
Compliance nirvana or nightmare? Automatedadvicefacesdifferentandinsomecasesmorecomplexregulatory challenges than traditional face-to-face channels. Firmswillneedtoadjusttheirriskandcontrolframeworksandmanagementinformationtoreflectthedifferentcustomerjourneysandtoidentify,reviewandmanagethedifferenttypesofrisksautomatedadvicepresents.However,thechallengesinrelationtoconsumercommunications,useofalgorithmsandcyberriskaresurmountableandtherisksmanageable,consistentwiththe analysis set out in this paper.
Atthesametime,automatedadvicealsogivesrisetoregulatorybenefits.Itallowsfirmstomaintainconsistencyofadvice,itprovidesanaudittrailofcustomerinteractionsaspartoftheadviceprocess,aswellasanopportunityforfirmstomakecommunicationsmorevisual,interactiveandengaging,includingthroughexploringtheuseofvirtualrealitytechnology.
OnemajoroutstandingquestioniswhethertheFCAwillbeabletotackletheregulatoryuncertaintyovertheadviceboundarytoanextentwhichmakesfirmswillingtoofferservices‘attheboundary’.WebelievethattheFCA’sworkinthisareawillreduceuncertainty,buteliminatingitentirelywillbechallenging.Ultimately,theFCA’sdecisiononexactlywhatconstitutesregulatedadvicewillbedeterminedbythebalanceitiswillingtostrikebetweenitsconsumerprotectionobjectiveanditsambitiontomakeguidancemorewidelyavailableandaffordableacrossallsegmentsofsociety.
Inconclusion,therefore,whileautomatedadvicepresentsno‘compliancenightmare’,thereisstillsomewaytotravelbeforewereach‘compliancenirvana’.
21
The next frontier | The future of automated financial advice in the UK
Automatedadvicealgorithmscanalsobedesignedwiththeabilityto‘learn’.Inmachinelearning,anapplicationofthebroaderconceptofArtificialIntelligence(AI),computerprogramscanbetrainedbysubjectmatterexperts(SMEs),suchasqualifiedadvisers,torespondtosituationswithoutbeingexplicitlyprogrammed and to change their response when exposed to new data.
OnceSMEscompletethetraining,andlearningalgorithmshavebeenproperlytested,theabilityto‘learn’can(andshould)beswitchedoff,beforetheseAIapplicationsaredeployedtointeractwithconsumers.Atanypointfurtherlearningcanberesumed,in‘offline’mode,byfeedingthecomputerneworadditionaldata.
AsAIapplicationsevolveandincreaseinsophistication(forexample,improvetheirabilitytounderstandandrespondinnaturallanguage),machineswillincreasinglybeabletoreplicate human interaction. Although the regulatory and legal requirementsunderwhichadviceisgivenshouldbecodedintothealgorithmratherthanlearnedovertime,machineswillbeabletohaveprogressivelymorecomplexconversationswithclientsto,forexample,understandtheirpreferencesortesttheirlevelof comprehension. Machine learning is already being applied successfully,particularlyinbusiness-to-businessandfunctionsthatdonotrequirecustomerinput(e.g.portfoliorebalancing).Thefirstinsurancechatbothasalreadybeenlaunched(Spixii),andAmazon’sAlexacanprovideinformationonannuitiesforinstance.24,25,26
AlthoughfirmshaveamuchgreaterabilitytocontrolandtestAIapplicationsthaniscommonlybelieved,regulatorswillbewatchingdevelopmentsclosely.Theprinciplesfordesign,governanceandsupervisionwediscussedinSection3applytobothlearningandnon-learningalgorithms.However,forlearningalgorithms,thereneedstobeastrongfocusonthepoliciesgoverningthequalityofdatafedtothealgorithm,thetype,extent,andlengthofthetraining and the criteria for testing the algorithm.
Underthisscenario,automatedadviserscouldbeabletoprovideadviceagainstarelativelycomplexsetofneedsandgoals,andenhancefirms’abilitytoprofilecustomersand‘test’theirunderstandingofbothproductsandthepotentialrisks.AIwillalsobeabletomonitorandunderstandverycomplexandregularlychangingregulations,probablymuchmoreeffectivelythananysinglehumanadvisercould.Overthenextdecade,therefore,itispossiblethatwewillseeautomatedadvisersentermoresophisticatedadvicemarketsandbeabletodealwithissuesascomplexastaxandholisticfinancialplanning,thusprovidingservicesforwealthiercustomersegments.
Havingsaidthat,thetransitionwillbegradualandprobablyonlypartial,notonlybecauseAItechnologiesarestillrelativelyimmature,butalsobecauseitwillrequireachangeof‘heartsandminds’onthepartofconsumers.Someofthereticenceisgenerational,butsomeisrootedinthehumanneedforformingbondsoffriendship,trust,andempathy.
4.MachinelearningandArtificialIntelligence
22
The next frontier | The future of automated financial advice in the UK
Part 2 Automatedfinancialadvice–aviewofindividualmarketsIn Part 2 we discuss the opportunities arising from automated advice, the barriers and the implications for industry players across the following markets: simple financial planning, investing, DC pension saving, at-retirement, mortgages and individual protection. We then draw out our overall conclusions and the next steps for firms.
As in Part 1, where we refer to ‘automated advice’, this includes regulated advice, guidance and discretionary investment management, which can be delivered via a purely automated system, a hybrid service or face-to-face assisted by an algorithm. Where we refer to specific services or channels we use specific terms such as ‘regulated advice’ or ‘purely automated’.
2323
New horizons | The future of automated financial advice in the UK
The opportunityWebelievethereissignificantunsatisfieddemandforanautomatedfinancialplanningproductortooltohelpwitharangeoffinancialdecisions,especiallyamongyoungermillennials.Consumersneedguidancerangingfrom‘micro-journeys’–bywhichwemeanfinancialdecisionsthatcanbemadebasedonshortadviceprocesses–allthewaythroughtoholisticfinancialplanning,whichtakeslongerandismoreinvolved.Micro-journeysvarybyageandlife-cyclestage.TheyincludedecidingbetweenpayingdowndebtandsavingintoanISA,guidanceonbuildingadepositforahouseandchoosingbetweeninvestmentswithinanISA.
Oursurveyshowsthat34percentofrespondentsarewillingtopayforanautomatedsolutionforfinancialplanning.Weviewthisfigureashigh.Financialengagementandliteracyamongconsumers,whichdrivesinterestinfinancialplanning,isgenerallylow.Moreover,whilesomesuchtoolsareavailableinthemarketplace,awarenessislow.
Pricesensitivityamongtheconsumerswhowoulduseautomatedfinancialplanningishigh.Morethan70percentofrespondentswouldnotpaymorethan£125forthisservice,whichislessthanaquarterofthecostofatypicalfinancialplanningsessionwithanadviser(seeFigure14).27
What are the barriers?Commercialviabilityisthekeyquestionforproviders.Youngermillennialsshowthehighestappetitetouseautomatedfinancialplanning,buttheirsavingsandabilitytopayforadvicearelowrelativetoolderconsumers.Marketingcostswouldneedtobehightoovercomelowconsumerengagementinfinancialmatters,andtheneedtoachievesignificantscaletooffsetlowfees.
Thecombinationoflowfeesandhighmarketingcostsmaketheeconomicschallenging.However,giventhesignificantunsatisfieddemandamongselectgroupssuchasyoungermillennials,offeringafinancialplanningtoolcouldbeapointofdifferentiationforprovidersandameansofgatheringvaluableconsumerdataandincreasingcustomerloyalty.Thiswouldhelpproviderswhoearnconsumers’trusttomarkettotheminfutureyears.Asaresult,somelargerproviderscouldconsiderprovidingfinancialplanningguidance for free or at-cost.
Webelievethereissignificantunsatisfieddemandforanautomatedfinancialplanning product or tool to help with a rangeoffinancialdecisions,especiallyamong younger millennials.
A.Simplefinancialplanning
Key findingThere is strong appetite for automated tools among younger millennials to guide simple‘micro-journeys’suchaschoosingbetweeninvestinginanISAandpayingdowndebt.Thecommercialviabilityoftheseservicesisquestionablegiventhelowabilityofthesecustomerstopay.However,largeinstitutionsmayprovidetheseservicestoengagewithandwintheloyaltyofthesecustomers.
24
The next frontier | The future of automated financial advice in the UK
Figure 14. Willingness to pay for automated advice for a financial review
0% 10% 20% 30% 40% 50% 60% 70% 80%
£500
£475
£450
£400
£250
£125
Less than £125
Source: YouGov 23-24 January 2017, Deloitte analysis. Sample: 697 GB adults who would pay for automated advice for a financial review.
72%
11%
9%
4%
0%
1%
3%
Itiscurrentlydifficulttoprovideaholisticautomatedadvicesolutionforfinancialplanningduetocomplicationssuchastaxandinheritanceplanningwherehumanoverlayisnecessaryatpresent.Inaddition,theshorterattentionspanofmostpeopleinadigitalenvironmentwilllimitthenumberofdetailedquestionstheycanbeaskedrelativetoface-to-faceadvice.However,automatedadviserscanprovidetoolstohelpsimplermicro-journeys,suchaschoosingbetweenISAandpensioncontributions.Adigital-onlysolutionismostlikelytosucceed,withahybridsolutioncombiningdigitalandface-to-faceadviceofferedformorecomplexcases.
IntheUK,adviceisregulatedifitinvolvesrecommendingspecificfinancialproducts.28Whereafirmprovidesregulatedadvicetocustomersbasedontheirindividualcircumstances,thefirmwillbesubjecttosuitabilityrequirements.Iffirmsareshowntohavegivenunsuitableadvicetheyareliabletopayredressandpotentiallyregulatoryfines.However,manymicro-journeys,suchasdecidingwhethertoinvestinanISAorapension,orworkingouthowmuchdepositaconsumerneedstobuyahouse,donotinvolverecommendingspecificfinancialproductsandareonlyregulatedifprovidedalongsideregulatedadvice.Wherefirmsprovideonlyunregulatedservices,theyshouldfacesignificantlyreducedcompliancecosts.Theseservicesmayprovidefirmswithfeweropportunitiestochargecustomersortoselltheirproducts,butcan build customer loyalty.
Wherefirmsprovideregulatedpersonalrecommendationsonretailinvestmentproducts(includinginsurance-basedinvestmentsandpensions),UKrulesrequirethistoberemuneratedthroughadvisercharges.Firmswantingtoprovideregulatedadviceforfreeor below cost will therefore need to consider carefully how this could be funded. More detail on this is set out in Section B.
Implications for industry playersBanksarerelativelywellpositionedbecausetheyhaveunrivalledaccesstomassmarketcustomersandcustomerdata.Theyhavetheabilitytoofferfinancialplanningguidanceasafreeorlow-costsolutiontobuildcustomerloyalty.Moreover,theircustomeraccesssolvesthechallengeofhighcustomeracquisitioncosts.Forbankstoexploitthisopportunityfully,thekeywillbedevelopinganddeployingtherighttoolstoanalysethisdata,whichmayhavetobewith a FinTech partner.
Employerscouldofferonlinefinancialplanningasabenefittotheiremployeesviaafinancialservicesprovider.Massemployerssuchasretailersareuniquelypositionedtoofferthisservicetotheiremployees,againside-steppingthecustomeracquisitioncostproblem.However,theiraccesstoemployeefinancesislimitedtosalaries.Asolutionwouldneedaccesstonon-incomefinancialdatasuch as on borrowings and outgoings. Trust could be a barrier here.
Technology-enabledstart-upsmaybebestpositionedtoprovideengaginginterfacestoencourageadoption,besidesprovidingthetechnologyforanysolution.Ongoingchangesinregulation,suchasPaymentServicesDirectiveII(whichmandatestheopeningupofbankdatatothirdpartieswithcustomerconsent),willhelpthemaccessdata.However,thisstilldoesnotsolvetheproblemofveryhigh customer acquisition costs that any stand-alone start-up will face due to low brand recognition.
25
The next frontier | The future of automated financial advice in the UK
The opportunityWebelievethereisasignificantunsatisfieddemandforefficientwaystoinvestsmallersumsofmoney,bothregularcontributionsaswellaslumpsums.Alltoooften,individualsareeitherriskaverse,preferringtoholdtheirsavingsincash,orlacktheunderstandingtoinvesttheminpotentiallyhighergrowingoryieldingassets.Themarketisverysignificant,with£828billionheldinindividualdepositsathighstreetbanksasofDecember2016.29 CashISAsaloneaccountedfor£269billionoffundsinthe2015-16taxyear,andinflowsintocashISAsamountedtoaround£60billioninboththe2014-15and2015-16taxyears.30Thelevelofinflowsislikelytoincreasefurtherinthe2017-18taxyear,withtheincreaseintheadultISAallowancefrom£15,240to£20,000.
“Automatedadvicewouldallowustoexpandourservicestoawiderrangeofclients,byloweringthesumneededforinvestableassets.” Deloitte interview with UK-based wealth manager, 2016
Fortypercentofrespondentsinoursurveywouldpayforautomatedadvicetohelpwithsuchinvestmentdecisions.Thisisremarkableinthecontextoflowengagementandfinancialliteracy,andintheabsenceofanymajorinitiativesbyUKincumbents.
“Thereisacustomersegmentthatispricesensitiveyetwealthy.Theyunderstandtheimpactoffees,areseekingvalueandarewillingtopayforautomatedadvice.You might see large sums allocated to automatedadvice,representingsmallsharesofwealthyportfolios.” Deloitte interview with UK-based fund manager, 2016
Wealthmanagerstypicallyhaveminimumclientaccountsizesabove£500,000andtheseaccountminimumsarerisingformany players.31Thecostofprocuringhumanadviceat£150perhouronaveragemakesitprohibitivelyexpensiveforsumsbelow£100,000tobeinvestedwithadvice.32However,evenamongwealthierinvestors,costconsciousnessisrising.Webelievethatmanywillallocateaportionoftheirassetstobemanagedpassivelythroughalowcostdigitaladviser.IntheUS,theevidenceshowssizeablevarianceinthedigitaladvicemarket–averageaccountsizesrangefrom$26,410atBettermentto$145,989atPersonalCapital.33Oursurveyresultsbackthisup,showingaclearappetiteamongwealthierinvestorstoinvestrelativelylargesumsofmoneythroughautomatedchannels.Infact,willingnesstopayforautomatedadvicetoinvest£10,000clearlyriseswithwealth(seeFigure15).
B.Investing
Key findingThereissignificantdemandacrossthewealthandincomespectrum,notjustfromthoselookingforalow-costsolutiontoinvestsmallersumsofmoney.Theresultingchallengeforindependentfinancialadvisers(IFAs),wealthandassetmanagerswillbemarginpressurefromthepotentialswitchingintolower-feeautomatedadviceproducts from higher-fee products.
26
The next frontier | The future of automated financial advice in the UK
Thosewillingtouseanautomatedadvicesolutionareclearlypricesensitive(seeFigure16).However,webelievelowcostsolutionsarecommerciallyviable.Morethantwofifthsofrespondentswhowouldpayforautomatedadvicewouldpaylessthan£100.Forlargescaleproviders,thispricepointmaybeachievableasevidencedbyannualmanagementchargesof45basispointsperyearonaccountswithfundsupto£250,000atHargreavesLansdown.34Forexample,acapitalsumof£22,000chargedat45bpswouldgenerateaproviderfeeof£100.
We recognise that a digital asset allocation solution using mainly passiveinvestmentsislikelytobethebestmeansofofferinglow-feeautomatedadvicethatachievesacceptablereturnsforproviders.Forwealthierinvestors,thissolutionmayalsoincorporateactiveinvestments.However,offeringsuchaproductposescleardangersforincumbentwealthandactiveassetmanagers:acorollaryoftherisingpopularityofsuchaplatformisevenmoreattentiontofees.While a substantial element of the business generated will be new totheindustry(i.e.moneythatwasheldincash),therewillbesomeswitchingfromhighermarginaccountsandactiveproducts.
Figure 15. Willingness to invest more than £10,000 using automated financial advice by wealth level
£100,000 or more£30,000 or more£5,000 or more£1,000 or more
13%15%
19%
28%
Source: YouGov 23-24 January 2017, Deloitte analysis. Samples: GB adults with more than £1,000 in savings and investments (excluding pensions and property wealth). Samples: £1,000 or more (1,046), £5,000 or more (842), £30,000 or more (461), £100,000 or more (212).
Figure 16. Willingness to pay for automated financial advice for investing £11,000
0% 10% 20% 30% 40% 50%
£450
£430
£405
£360
£225
£100
Less than £100
Source: YouGov 23-24 January 2017, Deloitte analysis. Sample: GB adults with more than £5,000 in savings and investments (excluding pensions and property wealth) who would pay for automated financial advice for investing £11,000 (335).
42%
28%
16%
6%
3%
4%
2%
27
The next frontier | The future of automated financial advice in the UK
What are the barriers?Wenotethatprovidermarketingwillalsoneedtoconvincepotentialclientsoftheneedtoinvestmoneyinhigherreturnassets as well as educate and reassure them about the capabilities of their automated products.
Lowthresholdswillbekeytoattractingconsumersfromthelesswealthytargetmarketwhodonothaveaccesstotraditional,higher-endwealthmanagers.Inaddition,lowfeesarevitaltobringintheprice-sensitiveconsumerswhoinhabitallsegmentsofthewealthspectrum,asshowninFigure16.However,customeracquisitioncostswillbehigh,especiallyfornewentrantsandthosewithlittle/nobrandawarenessinthismarket.
Asaresultofthesefactors,makingaprofitonstand-aloneautomatedadvicebusinessischallenging,particularlyintheearlyyears.Weestimatethatastand-aloneautomatedadvicebusinessislikelytoneedanywherebetween£4billionto£10billioninassetstobreak-even(seeFigure17)skewedinalllikelihoodtowardsthehigherendofthatrange,alevelsubstantiallyhigherthanwherewearetoday.
Somefirmsdevelopingregulatedautomatedadvicemodelsmaywanttooffertheirservicesbelowcost,orforfree,onanongoingbasistoimprovetheiroveralloffering.FirmswantingtodothiswillneedtoconsidercarefullytheirapproachsincetheFCA’srulesstatethatafirmcan only be remunerated for a personal recommendation on retail investmentproductsbyadvisercharges.35,36
Thisisanareawheremoreregulatoryguidancewouldbebeneficial,andfirmsmayneedtoseekfeedbackontheirproposedapproachfromtheFCA.Inanycase,firmswouldatleastneedtoensurethatcustomers of other business lines did not face higher costs as a result ofautomatedadvicebeingprovidedforfree.Ifaserviceisclearlyguidance,itcanbeprovidedforfree,asguidanceisoutsidethescopeof these rules.
Implications for industry playersAssetmanagerslargelyseethemselvesas‘manufacturers’ofinvestmentproducts.FewhavesizeableD2Cbusinesses,andmost are reluctant to spend the money to build a brand with highrecognitionamongmassmarketconsumers.However,rapidadoptionofdigitaladvicehasthepotentialtoaccelerateanongoingshifttopassiveinvestments.Tomaintaingrowth,assetmanagerswillneedtofocuseitherondifferentiatedproductsand/orfindawayofparticipatinginthischangethrough,forexample,buildingorpartneringtocreateaD2Croute.Thisraiseschannelconflictissuesthattheywillneedtoresolve.Assetmanagerstypicallygothroughintermediaries,suchasplatforms,banksandwealthmanagers,andanymoveintodirectdistributionwouldbringthemintocompetitionwith these established distribution channels.
-20
0
20
40
60
80
100
120
20,00015,00010,0007,5005,0003,0002,5002,0001,5001,000500100
Source: Deloitte analysis
With net fees of 50 basis points and the average account starting at £35,000 break-even is £5-7 billion
With net fees of 75 basis points and the average accountstarting at £35,000, break-even is £4-5 billion
With net fees of 35 basis points and the average accountstarting at £35,000, break-even is £7.5-10 billion
Assets under management (£ million)
Figure 17. Estimated break-even assets under management for a UK stand-alone automated advice platform37
Profi
ts (£
mill
ion)
28
The next frontier | The future of automated financial advice in the UK
Retailinvestmentplatformsandwealthmanagersarethebestpositionedtoaddondigitalinvestmentadviceservicesgiventheir branding and direct access to consumers. They can add this serviceatrelativelylowincrementalcost,althoughtheywillneedtobuild or buy technological expertise.
Retailinvestmentplatformsareespeciallywellplaced,astheycanattractbusinessawayfromwealthmanagerclientslookingforlower-pricedpassivesolutions.TheUSprovidesevidencethatthisisalreadyhappening.Weestimatethatbetweenthem,twolargeinvestmentplatformshavealreadycapturednearly80percentofthedigitaladvicemarket.38 While select challengers are continuing togrow,incumbentsaregrowingjustasfastbytargetingpotentialentry-levelwealthmanagerclientsattractedbytheirlowerfees.Theimportanceofscalesuggeststhatevenamongincumbentsfewwillbesuccessfulinthischannel:HargreavesLansdownleadsthecurrentD2Cmarketbysomedistanceandhasbeenunrivalledin its success at building a D2C channel with a widely recognised brand.
Wealthmanagerswill,however,needtobemindfulofcustomersswitchingfromhighermargin,activeandadvisedproductranges,tolowermargindigitally-advisedproducts,amidhigherfeetransparency. One option for wealth managers to protect their brandsistolimitaccountminimumstorelativelyhighamountsandofferanadded-valuehybriddigital-humanproduct,suchasadedicatedhumanadviseraboveaminimumaccountsize.
“Investmentexpertiseandsafecustodyare important in a business that depends mainly on trust. This is where incumbentshaveahugeadvantage.” Deloitte interview with UK-based wealth manager, 2016
Forwealthmanagerclients,givenothercomplexitiessuchastaxefficiency,inheritanceandbroaderriskmanagement,adigitalsolutionmay,inanycase,beonlypartofawiderallocationofassets.Overtime,thefocusoftheadvisermayshifttowardsincorporatingnon-investmentmatterssuchastaxplanningintotheinvestmentsolution.Wealthmanagerswhowishtowidentheirreachbeyondtheirexistingclientcharacteristics,giventheirtypicallynicheaudience,willalsohavetospendsignificantamountson customer acquisition costs.
IFAs,inparticular,lookvulnerableintermsofbothfeepoolsaswellasthescopeoftheirtasks.Webelieveretailinvestmentplatformsarewellpositionedtooffermassaffluentclients(withinvestableassetsof£50,000to£250,000)competitiveoffersthatcouldpressureIFAfees.Withtheavailabilityofdigitaltoolstomanageportfolios,theroleofIFAswillshifttowardsareassuchastaxoptimisationandestateplanning,andthegradualnarrowingoftherangeoftaskstheyperformcouldplacefurtherpressureonfees.
Givenbanks’accessandongoingrelationshipswiththeircustomers,targetinginvestorswithsmallsumstoinvestisanopportunitywitharelativelylowacquisitioncost.However,thechallengesherearelikelytobelowfinancialliteracy,highriskaversionandtheneedtokeepthissolutionaffordable.Banksarealsolikelytoneedtobuyintechnologysolutionstoprovideandservicethisoffer,andgrafttheseontotheirlegacysystems.Thatsaid,thiswillbelargelyincrementalbusinesstobanks,whetherbuiltthroughpartnershipsorinternallyandwillbeausefultooltodifferentiatetheiroffer.Theabilitytoviewandmanageinvestmentsonthesameplatformasday-to-daybankingwillappealtomany,andbanksareinpolepositiontoprovidethis.IntheupcomingeraofOpenBankingwherebybankswillberequiredtosharecustomerdatawiththirdparties,addingtheseservicesoffersbanksawayofimprovingcustomerengagementratherthanbeingonthedefensive.
Start-upshavecertainlyledthewayintermsofdeliveringdigitaladvicemodelsthatcandealwithsmallersumsofmoney.Theirfocus has generally been to attract customers through lower fee products,enabledbytheirdigitalplatforms.Whileeveryonewillbeinfavouroflowerfees,engagementwithinvestingisverylowand,inthatcontext,trustintheirfinancialproviderisakeydesiredtrait.Asaresult,customeracquisitioncostsforstart-upsaremagnifiedtouneconomiclevelsbythegenerallyultra-lowawarenessofstart-upbrandsamongallbutthemostdigitallysavvyinvestors.Westruggletofindexamplesofthosethatcanovercomethiskeychallengeandbuildscale.Thus,webelieve,barringaselectfew,amoreviablesolutionislikelytobeeithertopartnerwithortosellincumbentstheirservicesasawhitelabelproduct.
29
The next frontier | The future of automated financial advice in the UK
The opportunityTheUKDCpensionmarketishuge,withmorethan£650billioninestimatedassetsundermanagementandannualprofitsin the region of £1.6 billion.39Inaddition,theUKDCmarketisgrowingquicklyduetoacombinationofpopulationageing,auto-enrolment and the continuing shift from DB to DC pension schemes.WeestimatethattheUKDCprofitpoolwillgrowataround6percenteachyeartoabout£3billionby2025.40
Thedominantworkplacepensionprovidersaretypicallylargeinsurersandassetmanagers.Inourview,automatedadviceonworkplacepensionspresentstwomajoropportunitiesforincumbents to defend their dominant position in DC pensions.
Thefirstopportunityistoincreasecustomerengagement.ManypeoplearedisengagedwiththeirDCpensions.Forinstance,only52percentofDCsavershaveafairlygoodideaofhowtheirfundsareinvested.41 This leads to sub-optimal outcomes. Most obviously,DCsaverssavetoolittleandtaketoolittlerisk,bothof which inhibit the growth of their pension pots. Others fail to maximisetheirtaxbreaks.Nevertheless,theintroductionofthePensionsDashboard,startingwithapilotfromMarch2017,willenablepensionsaverstoseealltheirpensionpotsinoneplace,which could increase engagement.42 We see three main ways in which automation can further increase engagement.
• More proactive.Automatedalertscankeeppeopleupdatedabouttheirbalances,theincometheyareduetoreceiveinretirementandletthemknowwhennewfundsbecomeavailabletoinvestin.Reminderstoreviewsavingsratesand/orinvestmentchoicescanbeissuedperiodically,andchasedupifnoactionistaken.Justunderaquarter(23percent)ofoursurveyrespondentswhohavenotretiredsaidtheywouldlikeonlinehelpunderstandinghowmuchmoneytheyareontracktoreceiveinretirement.
• More accessible.Providerscanmakepensionsavingmoreaccessiblebyusingvisualtoolswithinautomatedadvice.Forexample,visualsthatillustratekeypointssuchasthebuild-upoffundsovertime,theinfluenceofriskonreturnsandprogressagainstgoalswouldlikelyprovemoreaccessibleandengagingthan paper-based communications.
• More powerful.Providerscanuseinsightsfrombehaviouraleconomicstomaketheircustomercommunicationsmorepowerful(seeourdiscussiononcustomercommunicationsinSection3).Researchfoundthathouseholdscuttheirenergyconsumption when they compared it with that of neighbours. The samethinkingcouldbeappliedtopensionswithsaversshownanonymised real life examples of how their peers are doing.
Thesecondopportunityistoimprovepensionsavingdecisionsbymakingadvicemoreaffordable.TheaverageDCpotatretirementis£25,000,thoughmanypeoplehavemultiplepotsandothersources of wealth.43Yetmanyfinancialadviserswillturnawayconsumerswithlessthan£50,000.Fortheseconsumers,thecostofadvicewoulduseuptoomuchoftheirpotstobeworthwhile.
C.Definedcontributionpensionsaving
Key findingDCpensionsareauniqueopportunityforautomatedadviceduetoalargeandfastgrowingmarket,accesstocustomersthroughthetrustedandlow-costworkplacechannelandalackofpublicallyavailableinformationonhowtoinvestcontributions.Thekeychallengewillbedevelopinghighlyengagingmarketingandonlineadviceportals,andensuringthealgorithmsarerobustenoughtomeetregulatory requirements.
30
The next frontier | The future of automated financial advice in the UK
Foremployeeswhocannotaffordtoseeapensionadviser,advicecouldbeprovidedviaaworkplacepensionportalwithminimalhumanintervention.Thewebsitecouldprovidehelpwithkeydecisions,suchaswhichfundtoinvestin.Thiswouldlikelybemuchcheaperthantraditionaladvicefortwomainreasons. One,itwouldsavefinancialadvisers’timeandfees.Two,thecostsof the technology could be spread across many scheme members creatingsignificanteconomiesofscale.Forexample,WealthWizardsprovidesautomatedworkplacepensionadvicefor £100peremployeeperyear.44
Makingpensionadvicemoreaffordableandaccessiblewouldhavealargepositiveimpactonsavers.Accordingtoresearch,UKsaverswithapensionpotof£100,000saveanaverageof£98moreeverymonthandreceiveanadditionalincomeof£3,654everyyearoftheirretirementiftheytakefinancialadvice.45
GivenrelativelylowlevelsofcustomerengagementinDCpensionsaving,oursurveysuggestssurprisinglyhighwillingnesstopayforautomatedadvice:morethanathird(35percent)ofworkerswithaDCpensioninoursurveywouldpayforautomated adviceoninvestingmonthlypensioncontributionsof£80. This represents a population of at least three million people.
What are the barriers?Price sensitivity. Consumers would demand low fees. Roughly two-thirds(68percent)ofworkingDCmemberswhowouldpayforthisservicewouldnotpaymorethan£125,equivalenttoadiscountofmorethan75percentonthetypicalcostofface-to-faceadvicein this scenario.46
Small pots. Akeychallengeforprovidersisthesmallsizeofmany DC pension pots. This necessitates a low-cost solution to keepfeesaffordable.Thisneedisamplifiedbythelowlevelsofengagement and awareness typical of DC plan participants. Thekeytomakingthisviablewillbescale–highervolumescouldcompensateforlowindividualaccountfees.
Acquisition costs. Marketingcampaignstopersuadesaverstotrytheservicecouldprovecostly.Asuccessfulcampaignwouldneedtoovercomelowconsumerengagementandfinancialliteracy,especially for plans with a large number of auto-enrollees.
Despitethesebarriers,alow-costdigitalsolutioncoupledwithalargecustomerbasecanmakethisproductviableonastand-alonebasisinourview.Giventhelowcostimperative,weseealargelyautomatedallocationtooldrivenbypassives,typicallyoperatingonaworkplacewebsite.Weseethisasdigitalbydefault,withhumantouch points as complexity rises.
Figure 18. Willingness to pay for automated financial advice on £80 monthly pension contributions, DC pension members
0% 10% 20% 30% 40% 50% 60% 70% 80%
£500
£475
£450
£400
£250
£125
Less than £125
Source: YouGov 23-24 January 2017, Deloitte analysis. Sample: 134 DC pension members who would pay for automated advice on £80 monthly pension contributions.
68%
14%
7%
6%
2%
1%
2%
31
The next frontier | The future of automated financial advice in the UK
Firmswillneedtothinkaboutwhethertheserviceisregulatedand,ifso,ensuretheyhaverobustprocessestoensurecompliancewithregulatoryrequirements.Ingeneral,theregulatoryconsiderationsrelatingtoadviceonDCpensionschemesaresimilartothoseforotherretailinvestmentproductsasdescribedinSectionB.However,thereisanexemptionfromtherulesonadviserchargingwherefirmsprovideadvicetoanemployerinconnectionwithaworkplacepension scheme.47Nevertheless,anypersonalrecommendationgiventoemployeeswillbecoveredbytherules.
Implications for industry playersWebelievethatthelifeinsurersandassetmanagersthatarethemajorworkplacepensionprovidersarewellpositionedtoautomateadviceonworkplacepensions.Themainreasonisthattheyhaveunrivalledaccesstocustomers.Thisgivesthemagoodchancetoside-step the challenge of high customer acquisition costs that has hamperedpioneersofautomatedadviceinthenon-retirementinvestmentspace.Workplacepensionproviderscommunicatewithaccountholdersandcanusethesecommunicationstoofferthemautomatedadvice.Thechallengewillbetoengagethesecustomers.
“Auto-enrolmentisahugeopportunityforautomatedadvice.Arangeofplayerscouldofferan‘accountforlife’concept. Icouldseebankstyinginpensionadvicewithcurrentaccountsandofferinganaggregatedplatform.”
Deloitte interview with UK-based insurer, 2016
Providinginnovativeautomatedadviceservicescouldalsobeatoolforworkplacepensionproviderstowinbusinessfromemployers.Employers can play an important role in increasing employee engagementwiththeirpension,primarilythroughcommunicatingandreinforcingtheperceivedvalueofworkplacepensionsasthemostimportantcomponentofretirementsavingsformostconsumers.Employersmaybenefitfromsucheffortsiftheyincrease employee satisfaction.
Assetmanagerswillbenefitfromhigherflowsintoinvestmentproductsinwhichsaverswillbeadvisedtoinvest.Weregardtwodistinct categories of asset manager as well-placed to capture thisopportunity.Assetmanagersthatareownedbymajorlifeinsurance groups are well placed because they belong to the samegroupsassomeofthemajorworkplacepensionproviders.Leadingpassivefundmanagersalsoappearlikelywinnersbecausetheirfundsfitbetterwithinanultra-low-costmodel.
32
The next frontier | The future of automated financial advice in the UK
The opportunityConsumersexperiencethegreatestneedforfinancialadvicewhentheyretire.Theyhavetodecidehowtofundretirementwithsavingsaccumulatedoveralifetimeandarefacedwithtoughdecisions–suchashowmuchcashtowithdrawwhileleavingenoughinvestedtoprovideadecentincome–andcomplexproducts.Theneedforat-retirementadvicemarkedlyincreasedwhenPensionsFreedomswereintroducedin2015.Thenewrulesgiveretireesamuchwiderrangeofoptionsonhowtoconverttheirpensionpotintoretirementincomethanpreviouslywhenmanypeoplewere‘forced’intoanannuity.
Howevermanyretireesshunadvice.Currentlyaround33percentofannuitiescustomersand65percentofincomedrawdowncustomersusearegulatedadviser.48Wewouldexpecthighertake-upofadviceforthoseusingincomedrawdownthanforannuities.Incomedrawdowncanberiskierthanannuitiesinthatithasnoin-built insurance against running out of money.
Twomainbarrierspreventanddiscouragemorepeoplefromtakingadvice.Thefirstisdisengagement.Manyretireesaredisengaged with their pensions. This helps explain why only around 20percentofconsumersatretirementusePensionWiseeventhoughitoffersfreeguidance.49Second,formanyadviceistooexpensivetobeworthwhile.Thatsaid,fromApril2017consumerswillbeallowedtoaccess£500tax-freeuptothreetimesfromtheirpensionpottooffsetthecostofregulatedretirementadvice.50 Thiscouldincreasesomeconsumers’abilitytopayforadvice,butconsumerswillstilldemandgoodvalueformoney.Moreover,face-to-faceadviceoftencostsmorethan£500.
Automatedadvicecanhelplowerbothbarriers.Engagement is won in the accumulation phase long before retirement. AsdiscussedinSectionC,automatedadvicecommunicationsthataremoreproactiveandaccessiblethantoday’spaper-basedannualstatementscanbuildcustomer’sinterestandinvolvementinDCsaving.Effortsherewouldlikelygainregulatorysupport.
D. At-retirement
Key findingThereissignificantdemandforlow-costfinancialadviceintheat-retirementmarket.Duetothecomplexityofthedecisionsthatneedtobemade,themosteffectivemodelislikelytobeahybridmodelwheremostoftheprocessisautomatedbutcustomersspeaktoahumanadvisertocompletetheprocess.Thiscansavetimeandcostforadvisersandcustomersalike,allowingbothpartiestofocusonthemost complex decisions. The question is how to design a system sophisticated enoughtodealwithacomplexmarketandtounderstandthedifferentprioritiescustomersmayhavefortheirretirement.
33
The next frontier | The future of automated financial advice in the UK
TheFCAhasbeenconsideringhowfirmscanbetterengage withcustomersaspartofitssmartercommunicationsinitiative.51 Italreadyrequiresfirmstoprovideclientswithasummaryoftheiropenmarketoptionsbeforesellingthemaretirementincomeproduct,andisbehaviourallytestinghowat-retirementcommunicationsfromproviders(knownas‘wake-uppacks’)canbeclarifiedandsimplifiedtohelpconsumersexercisechoiceeffectively.52,53Itisalsoconsultingonrequiringannuityproviderstoshowconsumersthedifferencebetweentheirquoteandthehighestguaranteedquoteavailableontheopenmarket.54 This latterinitiativeshouldencouragecustomerengagementbymakingitclearhowmuchmoneycouldbeatstake.TheFCAhassuggestedthatinfutureitmaydevelopcomparisontoolsforotherretirementincomeproducts,suchasincomedrawdownproducts.55
Inadditiontomakingadvicemoreengaging,automationcanradicallyloweritscost.In-personadviceonconverting£30,000pensionsavingsintoalumpsumandaretirementincomeproduct,whichisatypicaladvicescenario,costsaround£800.56Webelieve thecombinationofawebsiteusingalgorithmsandanadviserspeakingwithcustomersoverthephone(toensureadviceissuitable)candelivergoodqualityadviceatafractionofthiscost. Ourresearchindicatesthatthetimeanadviserwouldneedtospendon automated cases would be less than half that of traditional cases.
Lowcostiskeybecausemanyconsumerswoulddemandadeepdiscountonin-personadvice.Figure19showsthat,amongthosewhowouldpayforit,almosthalf(48percent)wouldpaylessthan£205forautomatedat-retirementadvice.
Thisisanopportunityforprovidersintwoways.First,itischancetoearnnewadvice-basedrevenues.Manypeoplearewillingtopayforautomatedadviceatretirement.Second,moreimportantly,itis chance to win and retain customers at the crucial at-retirement stage.Morethan300,000peopleretireeachyearwithaDCpot.57 Forprovidersthatselldecumulationproducts,providingeasy-to-useandlow-costadviceisthewaytomarkettheseproductsandtoensurethattheyarewellmatchedtocustomers’needs.
Consumers experience the greatest need forfinancialadvicewhentheyretire.Theyhavetodecidehowtofundretirementwithsavingsaccumulatedoveralifetimeand are faced with tough decisions – such as how much cash to withdraw while leavingenoughinvestedtoprovideadecent income – and complex products.
Figure 19. Willingness to pay for automated advice on converting £30,000 pension savings into a lump sum and a retirement income product
0% 10% 20% 30% 40% 50%
£825
£780
£740
£660
£415
£205
Less than £205
Source: YouGov 23-24 January 2017, Deloitte analysis. Sample: 69 GB workers aged 30+ with pension savings worth more than £20,000 who would pay for automated advice on converting £30,000 pension savings into a lump sum and a retirement income product.
48%
23%
14%
8%
2%
5%
0%
34
The next frontier | The future of automated financial advice in the UK
What are the barriers?Akeyconcernforfirmsisregulatoryrisk.Firmsgivingregulatedadviceonpensionsarelikelytofaceahighdegreeofsupervisoryscrutinygiventhatlargesumsofmoneymaybeinvolvedandconsumerscouldrunoutofmoneyinretirementiftheyaregivenpooradvice.Manyretireeshavecomplexfinancialandpersonalcircumstancesandneedtoweighupcarefullydifferentprioritiesfortheirretirement.Somecustomersmaybevulnerabletobehaviouralbiaseswhenthinkingabouttheirretirement.Examplesincludeprojectionbias,whereconsumersmayassumetheywillbeabletoliveonthesameincomein20years’timewithoutfactoringinchangessuchasthepossibilityofgoingintoacarehome,andoverconfidencebias,whereconsumersmayoverestimatethereturnstheywillearniftheyleavetheirpensioninvestedanddrawitdownovertime.
Toensuretheygivesuitableadvice,firmswillneedtoconsiderhowbehaviouralbiasescouldaffecthowacustomeranswersquestionsandframequestionscarefullytominimisethisrisk.Giventhatbothcustomercircumstancesandretirementoptionsareoftencomplex,itisalsolikelythatfirmswillneedtosupplementtheirautomatedadvicesolutionwithahumaninteraction,sothatahumanadvisercancheckthecustomer’sunderstandingandclarifyanypointsifthefirmneedsadditionalcustomer-specificinformation.Overthelongerterm,wemayseefirmsdevelopingmorefullyautomatedadvicesolutionsthatcanmeetregulatoryrequirements,butintheshortandmediumtermahybridmodelismorelikelytobeeffective.
Formorecomplexcases,itislikelythatfirmswillneedtorefercustomerstoahumanadviserearlierintheprocess.Thisislikelyto be especially true for cases where the FCA has put in place particularlystringentrulestopreventconsumerdetriment.Forexample,wherefirmsgiveregulatedadviceonwhetheracustomershouldtransfermorethan£30,000outofcertaintypesofschemewithguaranteedbenefits,theyarerequiredtostartbyassumingthat a transfer will not be suitable and only recommend it if they can clearly demonstrate that it is suitable.58
Implications for industry playersThelargelifeinsurersandassetmanagersthataremajorworkplacepensionprovidersappearbetterplacedthanotherplayerstowincustomersbyprovidingautomatedat-retirementadvice.Fourreasonsstandout.First,theyhavethegreatestscale,andwithit,abilitytoprovidealow-costservice.Second,theycanusetheirunrivalledcustomerdatafromtheaccumulationphasetoprovidemorepersonalisedadvicethanotherplayers.Third,theycanusethesamedatatopre-populateformsmakingadvicefromthemmoreconvenientthanelsewhere.Finally,forlifeinsurers,onlytheycanprovideannuitieswhich,formanyrisk-aversecustomers,couldbeamoresuitablerecommendationfromadvicethanincomedrawdown.
Foradvisers,automationcanallowthemtospendlesstimeonthemorestraightforwardpartsoftheprocess,suchasdatacapture,andinsteadfocusonthehigh-valueareas.Forinstance,adviserscould automate product recommendation and spend more time advisingontaxconsiderationsinmorecomplexcases.
The large life insurers and asset managers thataremajorworkplacepensionprovidersappear better placed than other players to wincustomersbyprovidingautomatedat-retirementadvice.
35
The next frontier | The future of automated financial advice in the UK
E. Mortgages
Key findingMoreautomationinthemortgageadviceandapplicationprocessescouldunlocksignificantefficiencies.Borrowerswillbenefitfromwiderchoiceandpotentiallylowerfees,whilebothlendersandborrowerswillbenefitfromsmoother,fasterandmoreaccurateprocessingofapplications.Sincecustomersmaybevulnerableand/orhavecomplexfinancialcircumstances,themosteffectiveadvicemodelislikelytobeahybridwheremostoftheprocessisautomatedbutcustomersspeaktoahumanadviserattheend.
The opportunityTheannualmortgagemarketislarge,withgrosslendingvolumesof£246billionin2016.59Mortgagesfinancethelargestlifetimepurchaseandmostvaluableassetofmostindividuals.Moreover,givenvaryingindividualriskappetitesandtimehorizons,thereareseveralcomplexchoicestobemadeincluding,butnotlimitedto,fees,loanduration,interestratetype,andinterest-onlyorcapitalrepaymentoptions.Thiscombinationofahigh-valuedebtproductand the need to choose between multiple product attributes oftencreatesaneedforadvice.IntheUK,theFCArequiresfirmssellingmortgagestogiveadviceinmanycases,underrulesputinplacein2014.60 This has contributed to an increase in the proportionofconsumersreceivingadvicewhentakingoutamortgagefrom67percentin2008to97percentinH22016.61 Meanwhile,intermediatedsalesrosefrom50percentin2009/10to67percentinH22016.62Thelargesizeofthemortgageadvicemarketcreatesanopportunityforbothmortgagelendersandintermediariestoinvestinmoreautomationwhichcanreducecostsforfirmsandimprovethecustomerexperience.
Oursurveyindicatesahighlevelofreceptivitytoadigitalsolutionforthisservice–38percentofallrespondentsarewillingtopayforthisoption.Onaverage,borrowerspayaround£500toamortgagebrokerfortheirservices.63Thesefeescanbehigher,dependingonmortgagevalueandothercomplexities.Brokersarepaid0.3-0.5percentoftheloanvaluebylendersincommissions– an arrangement that is allowed under the UK regulatory regime formortgagesunlikeforinvestments.Somebrokerschoosetoreduceborrowerfees,subsidisingthisfromlendercommissions.Oursurveyresultssuggestthatwhilepotentialusersarepricesensitive,asmanyas47percentofrespondentswillingtopayforanautomatedsolutionwouldpay£125ormoreforthisservice.Whatmakesthispropositionexcitingisthatwebelieveadigitalofferingispotentiallyviableonlenderfeesalone,astheservicecanbeprovidedmorecheaplythanaface-to-faceoffering.
Theannualmortgagemarketislarge,withgrosslendingvolumesof£246billionin2016.Mortgagesfinancethelargestlifetimepurchaseandmostvaluableassetofmostindividuals.
36
The next frontier | The future of automated financial advice in the UK
Webelievethatthishighwillingnesstoadoptadigitalsolutiondespitetheinfancyofthisoffer(withonlyafewstart-upproviders)suggestssignificantinefficienciesinthemortgageapplicationandapprovalsprocesses.
Whilebrokersscanthefullsetofproductsavailableandrelevanttocustomerneedsdependingontheirlevelofserviceoraccesstothemarket,inpractice,theycanonlybefamiliarwiththedetailsoftheunderwritingpracticesofalimitednumberoflenders,knowledgeofwhichiskeytodeliveringaproductsuccessfullytoaborrower.Incontrast,adigitaltoolshouldbecapableofscanningfor products much more widely and automatically assessing the customeragainstthelendingcriteria,benefitingconsumersandpotentiallyloweringrejectionrates.Nevertheless,inpractice,anautomatedsystemislikelytoneedtobesupplementedwithsomeinvolvementfromahuman,particularlyinassessingsuitabilityandcheckingthatthecustomerunderstandstheproduct.
Butequallyasexciting,webelieve,isthepotentialtoincreaseefficiencythroughoutthemortgageapplicationandapprovalprocess,yieldingsizeablesavingsforborrowersandlendersalike.
Mortgagesaredocumentation-heavy–thisrelationshipisdeeplyintermediated,andmanyconsumershaveadifferentmortgageprovidertotheircurrentaccountprovider.Asaresult,borrowersoften face repeated requests for documentation as part of their applications,whichoftenneedstobeverifiedand/orcertified,witha lot of time lost in the processing of this information.
Webelievethatmoreautomationintheadviceandapplicationprocessescouldbenefitborrowersandlendersintwoways:byincreasingefficiencyandbyspeedinguptheprocess.Bydigitallyacceptingandprocessingdocuments,digitalintermediariescanaddsignificantconveniencetocustomers.Iffirmscandeviseaprocesswherebythesedigitalformatsareacceptabletolenders,they can both speed up processing substantially as well as ease pressure points for lenders.
Ultimately,thewiderchoiceadigitalbrokercanofferaborrowerandtheefficiencygainsforlendersandborrowersalikeare hugebenefits.
Oursurveyindicatesahighlevelofreceptivitytoadigitalsolutionforthisservice–38percentofallrespondentsare willing to pay for this option.
Figure 20. Willingness to pay for automated advice to find a mortgage
0% 10% 20% 30% 40% 50% 60%
£500
£475
£450
£400
£250
£125
Less than £125
Source: YouGov 23-24 January 2017, Deloitte analysis. Sample: 378 GB adults aged under 50 who would pay for automated advice to find a mortgage.
42%
28%
16%
6%
3%
4%
2%
37
The next frontier | The future of automated financial advice in the UK
What are the barriers? Anydigitalintermediarywillfaceveryhighcustomeracquisitioncosts.Consumerawarenessofalternativesisverylowandgiventhehighvalueofthisproducttotheborrower,buildingtrustwilltakebothtimeandmoney.
Adigitalintermediarywillneedtobuilddeepknowledgeoftheunderwritingrequirementsacrosslenders.Integratinganydigitaldocument processing capabilities with lender systems is another challenge. The system will also need to be sophisticated enough to considercompetingobjectivesthatthecustomermayhave,suchas price and speed.
Formostpeople,takingoutamortgageisaverymaterialfinancialdecisionandanunsuitablemortgagecouldresultinsignificanthardship,includingpotentialevictionfromtheirhome.Inthiscontext,firmsprovidingadviceonmortgagescanexpectahighdegreeofscrutinyfromtheFCA.TheFCAissupportiveofinnovationintheadvicemarketandiscurrentlyconsideringpotentialregulatorybarrierstofirmsprovidingautomatedadviceformortgagesaspartofitsmortgagesmarketstudy.
However,anyautomatedsolutionwillneedtomeettheFCA’sstrictconsumerprotectionrequirementsformortgageadvice,includingsuitability assessments.64
Inpractice,someregulatoryrequirementsarelikelytoposemorechallengesforanautomatedsolutionthanforahumanadviser.Forexample,anadviserneedstoensurethatthecustomerunderstandstheimplicationsoftakingonasignificantamountofdebt in the face of uncertainties about future interest rates and thecustomer’sfutureincomeandexpenditure.Advisersalsoneedtobeabletoidentifywherecustomersmaybevulnerableormaybeoverlyoptimisticabouttheirabilitytomakerepayments.Humanadvisersmayfindthiseasierastheyaremoreabletoreadbodylanguage,askprobingquestionswhicharespecifictotheinformationprovidedbythecustomer,andgatherbackgroundinformationthroughtalkinginformallytothecustomer.CustomervulnerabilityisakeyfocusfortheFCAacrossallfinancialservicesmarkets,butisespeciallyrelevanttomortgagesastheyarelong-term debt products.65
“Inverycomplexcases,apurelyautomatedservicewouldbedifficult – for example if someone wants to purchase buy-to-let mortgages simultaneously.”
Deloitte interview with UK-based start-up, 2016
Asaresultofthesechallenges,itislikelythatfirmswillneedtosupplementtheirautomatedadvicesolutionwithahumanadviserwhocancheckthecustomer’sunderstandingandclarifyanypointsifthefirmneedsadditionalcustomer-specificinformation.Existingonlinemortgageadvisersrequireallcustomerstospeaktoahumanbyphoneorwebchataspartoftheprocess.Overthelonger-termwemayseefirmsdevelopingmorefullyautomatedadvicesolutionsthatcanmeetregulatoryrequirements,butintheshortandmediumtermahybridmodelismorelikelytobeeffective.
Implications for industry playersTraditionalmortgagebrokerswillbehardesthit.Overtime,feelevelsarelikelytodecline.Webelievethatborrowerfeescouldeventuallydisappear,withintermediariesfundedthroughlenderfees.Withoutaresettotheiroperatingmodel,oramodelthatsignificantlyenhancestheroleofdigitalcapabilities,theyarelikelytoseebothdiminishingshareanddiminishingprofits.
Lenders will see greater product transparency and may face heightenedcompetition.However,thisshouldbeoffsetbysignificantefficiencygainsfromdigitalsolutions.Inaddition,withaccesstotherighttechnology,bankscouldexploittheiradvantagedaccesstocustomerdata(e.g.oncurrentaccounts)anddisintermediatebrokersbydirectlytargetingpotentialmortgagecustomerswithtailoredoffers.Theycouldevensetuptheirownonlinemortgagebrokersthatcouldactasanadditionaldistribution channel.
Webelieveonlineadvisersarewellplaced.However,theyneedtoovercomeseveralchallengesincludinghighcustomeracquisitioncosts,buildingknowledgeoflenderunderwritingcriteriaandmeeting regulatory requirements.
38
The next frontier | The future of automated financial advice in the UK
The opportunityIntheUKthereisalargemarketforindividualprotection.Thistype of product insures policyholders and their dependants from theadversefinancialconsequencesofdeath,illnessordisability.Themainproductlinesaretermassurance,criticalillness,wholelifeandincomeprotection.Weestimatetheprotectionprofitpoolis around £1 billion annually.66
Howeverthereisastubborn‘protectiongap’.Manyconsumersare reluctant to buy enough protection to match their needs. SwissReestimatesthisgapbetweenthelevelofcoverinplaceandthatrequiredtomaintainthelivingstandardsofdependantsis more than £2 trillion for the UK adult population.67
Thekeybarrierstowiderconsumertake-uparealackofperceivedneedand,connectedtothis,perceivedhighcost.Onlyafifth(22percent)ofadultsconsiderlifeinsuranceas‘necessary’andonlytenpercentconsiderit‘affordable’.68 This is despite half (49percent)oftermassurancepolicyholderspayinglessthan£20permonth.69
IntheUKthereisalargemarketforindividualprotection.Thistypeofproduct insures policyholders and their dependantsfromtheadversefinancialconsequencesofdeath,illnessordisability.
Twoissueshaveloweredsales.First,consumersincorrectlyassociate protection with payment protection insurance (PPI).Thisweigheddowndemandfromthelate2000sduetoaccusationsofPPImis-selling.Second,inrecentyearsbanksandbuildingsocietieshavebeenscalingbacktheiradvisoryservicesformassmarketcustomers.Thishasledtofewerbankadvisers,andinturn,fewerrecommendationsofprotection.
Webelievethatautomatedadvicecanaddresstwokeychallengesintheprotectionmarket:thelackofperceivedneedforcoverattherootofthe‘protectiongap’,andtheexpenseattachedtotheadviceprocessthathasledsomeadviserstoretreat.
Responding to life events.Automatedadvicecandemonstratetheneedforprotectioninaproactivewaythatcouldhelpclosethe‘protectiongap’.Morethanhalf(52percent)ofadultsareexpectingachangeintheirpersonalcircumstancesoverthenext12 months.70Theseeventswilltriggerachangeintheneedforprotection.Forinstance,24percentareexpectingtochangejobsandthesemoveswilltriggerachangeintheneedforreplacementincomeintheeventoflong-termillness.71 These people could be targeted with timely communications explaining why they need protection,howtheirneedischangingandwhattodoaboutitwithalinktoanautomatedadviceportal.Wewouldenvisagethisasbeingprovidedbylifeinsurersinpartnershipwithanotherorganisationwithaccesstoawiderpoolofcustomers,suchasabankorretailer.
F.Individualprotection
Key findingAutomatedadvicecanplayakeyroleinmakingtheproductmorerelevantinthedigitalage.In-the-momentadviceandproactivemarketingcanbetterdemonstratethe clear need for protection to customers. A smooth purchasing experience canmakeiteasierforthemtobuyit.Thehurdleswillbeobtainingthenecessarycustomerdataandwinningcustomers’trust.
39
The next frontier | The future of automated financial advice in the UK
Automated sales process.Automatedadvicetoolscanhelpencouragemoreadviserstorecommendprotectionbygivingthemaquicker,light-touchsaleschannel.Twoopportunitiesstandout.First,inthemarketforprotectiontocoveramortgage,application forms could be pre-populated using data captured fromthemortgageapplicationprocess.Thiswouldsavevaluableadvisertimeonthephoneaskingcustomersforbasicinformation.Second,customerscouldbeguidedthroughtherelativelysimplestepsintheadviceprocessonawebsitebeforespeakingwithanadvisertocompletetheprocess,theso-calledhybridmodel.Thiswouldsavecostandallowadviserstofocustheir time on the hardest parts of the process such as choosing a product.Thisrepresentsachanceforadviserstoreducethecostofadviceand,potentially,increaseitsquality.
Oursurveysuggestsautomatedprotectionadvicecouldbeadoptedbymanypeople.Overall,almosthalf(45percent)ofadultswithchildrenaged18andunderwouldpayforautomatedadvicetofindlifeinsurance.Thesepeoplenumberaboutfivemillion.
Almosthalf(45percent)ofadultswithchildrenaged18andunderwouldpayforautomatedadvicetofindlifeinsurance.
What are the barriers?Themainchallengeforanadvicepropositionthattargetscustomersastheirneedsforprotectionchangewouldbewinningcustomers’trust. There are examples from the retail industry where customers objecttopreciselytargetedmarketingbecauseitfeelssomewhatinvasive.Nonetheless,wefeelthatthisbarrierissurmountable.Manymorepeoplerecognisethebenefitofhavinglifeinsurancethanhaveit.Forexample,70percentofparentswithchildrenunder18intheirhouseholdagreelifeinsuranceisworthhavingattherightpricebutonly45percentholdapolicy.72Tous,thissuggeststhatcustomerinertiaisaproblem,i.e.somecustomersareputoffbuyinglifeinsuranceandnevergetroundtoit.Someofthesepeoplewouldlikelybenefitfromanudgetotakeaction.
Connectedtothechallengeofwinningcustomers’trustistheissueofhowtocollect,store,manageandusecustomerdatasecurely,andfirmsneedtoensurethattheyfullytakeintoaccount current and future data protection regulations as theydesigntheirsolutions.ThenewGeneralDataProtectionRegulation(GDPR),whichwillapplytobothregulatedadviceandguidanceandentersintoforcein2018,willmandateorganisationalaccountability,andwillrequirefirmstoimplementrobustprivacygovernanceandingeneraltakeamoreproactiveapproachtoprivacycompliance.73TheGDPRintroducesanewmaximum monetary penalty of 4 per cent of annual global turnoverthatcanbeimposedincasesofseriousnon-compliance.
Figure 21. Willingness to pay for automated financial advice to find life insurance
0% 10% 20% 30% 40% 50%
£150
£140
£130
£120
£75
£40
Less than £40
Source: YouGov 23-24 January 2017, Deloitte analysis. Sample: 208 GB adults with children aged under 18 who would pay for automated advice to find life insurance.
48%
20%
16%
6%
5%
1%
4%
40
The next frontier | The future of automated financial advice in the UK
Whereafirmtargetsitsmarketingcommunicationsbasedonthecustomer’scircumstances,thisdoesnotautomaticallymeanthefirmisprovidingregulatedadvice.However,ifthefirmhighlightstothecustomerthepersonalcircumstanceswhichhaveledtothembeingcontactedandsuggestsspecificproductswhicharerelevantforthem,thisislikelytobeconsideredbytheFCAtobean implicit personal recommendation.74Firmsprovidingpersonalrecommendationsaresubjecttosuitabilityrequirementsandareliabletopayredressandpotentiallyregulatoryfinesiftheyhavenottakenreasonablestepstoensurethesuitabilityoftheiradvice.Therefore,theyneedtoensurethattheycollectallrelevantinformationaboutacustomer’scircumstancesandcandemonstrate suitability.75Inmostcases,thecustomerinformationalreadyheldbythefirmisunlikelytobesufficientforthis.Providerscoulddirectcustomerstoawebsitewheretheyareaskedmorequestionsabouttheircircumstances,althoughthisprocessmaydeter some customers from completing the purchase.
Implications for industry playersLife insurers which distribute their products through partnershipswithbanksandbuildingsocietiesarebestplacedtoofferproactiveautomatedadvice.Thisisbecausebanksandbuildingsocietieshaveaccesstoawidepoolofcustomers,includingthosewhoareuninsured.Theyalsohavedatathatwouldprovideaninsightintothelifeeventsthatchangetheneedforprotection.Wenotethatthelargestprotectionproviderssellthroughthebankchannel.Proactiveautomatedadvicewould,therefore,favourthelargeincumbents.
Adviserswouldbewinnersfromamovetoautomatetheinitialstepsintheadviceprocess.Thiswouldreducetheamountoftime spent on each case and lower costs. As in the mortgage brokingmarket,start-upsmayleadthewayindevelopinghybrid(humanplusdigital)adviceinterfaces.Thekeytosuccessisan easy-to-use interface with a seamless experience when a customermovesfromusingthewebsitetospeakingwithanadviser.Start-upshaveatrackrecordofinnovationbuiltonsmooth customer experience.
Webelievethatautomatedadvicecanaddresstwokeychallengesintheprotectionmarket:thelackofperceivedneedforcoverattherootofthe‘protectiongap’,andtheexpenseattachedtotheadviceprocessthathasledsomeadviserstoretreat.
41
The next frontier | The future of automated financial advice in the UK
Nextsteps–keyquestionstoconsider
Customer analysis • Whichcustomersegmentsshouldyoutarget? • Whatmethodsofin-persontesting(e.g.focus groups)areappropriateforyoutounderstand customerengagement,behaviouralbiasesandpotentialadoptionofnewtechnology?
• Howcanyoudesigndifferentiated,engagingcustomerpropositions,includingacrossabroadrangeofcustomerneeds?
• Whatpotentialchannelconflictswillariseandwhatdistributionmodelsareappropriate?
Regulation • Aretheservicesyouwanttoprovideregulated orunregulated?Ifitisunclear,haveyouengaged withtheFCAtodiscuss?
• Ifregulated,howwillyoucomplywiththerelevantrequirements(e.g.suitability)?Doyouneedahybridsystem(involvingahuman)formorecomplexcases?
• Areyourriskandcontrolframeworksfitforpurposetomanagenewandemergingriskspresentedbyautomatedadvice(e.g.designandoversightofalgorithmsandmachine learning/AIapplications)?
Business model • Where is it feasible for you to maximise profitinthevaluechaine.g.advice,products?
• Howdoyouacquirecustomersprofitably? • Howmuchtocharge(ifanything),andhowshouldyoustructurecharges,e.g.fixedorasapercentageofassets?Considerregulatoryrestrictionsifyouwanttoprovideregulatedadviceoninvestmentproductsforfreeorbelowcost.
• Howcanyouscaleupyourpropositions? • Shouldyougiveadviceonlyonyourownproductsoronthewidermarket?
Technology • HowcanyoumakeuseofAIandmachinelearning, andforwhat?
• Willyoubuildorbuytechnologyinfrastructure? • Howcanyouensurecyberresilienceanddatasecurity? • CanyoueasilyintegratetheautomatedadviceplatformintoexistingITinfrastructure?
• Howdoyouensurelinkagesandconsistencywithotheradvicechannels(e.g.humanorhybridapproaches)?
• Howmuchdoyouwanttoinnovatevs.follow,andwhatdoesthismeanforyourtechnologychoices?
M&A • Inwhichareascanstart-upsand/orthoughtleaderscomplementyouroffering(e.g.technology)?
• How will you structure potential partner relationships: acquisitions,jointventures,strategicstakes?
Talent • What organisational cultural changes are needed toimplementautomatedadvice,e.g.customervs productcentricity,agility?
• Doyouhaveanygapsinrequiredskillsets,e.g.financialadvice,technology,marketing,behaviouraleconomics,riskandcompliance?
• Doyouneedtolooktonon-traditionaltalentpoolsto acquirethese?
42
The next frontier | The future of automated financial advice in the UK
Conclusion
Initialiterationsofautomatedadvicehavecentredonlowcostwealthmanagement.However,webelievethatitskeycharacteristicsofaffordabilityandconveniencewillbeattractivetoawiderangeofconsumersacrossseveraldifferenttypesoffinancialservices.Wefoundevidencefromourconsumersurveyofsignificant,ifprice-sensitive,latentconsumerdemand.Thecostefficiencyofautomatedadvice,itsobjectivityandabilitytomaintainaclearaudittrailarekeypositiveattributesforproviders.Webelievethatincreasedautomationintheprovisionofadviceonretirementproductsandmortgagesisimminent,withahybridmodelinvolvingsomehumaninteractionlikelytodominateinthesemarketsintheshorttomediumterm.
“In5-10yearswewillprobablynotusetheterm‘roboadvice’–digitalwilljustbeanotherchannel.Theprovisionofadvicewillbe‘omni-channel’,althoughwill be underpinned by a consistent underlyingengine.”
Deloitte interview with UK-based insurer, 2016
Firmsprovidingautomatedadvicewillhavetonavigatelowfinancialliteracyandengagement,lowfees,theriskofcustomersswitchingfromhighertolowermarginproducts,aswellasregulatoryriskanduncertainty.However,asexplainedinthispaper,someofthesechallengescanbeovercomethroughinnovativesolutionstomakeiteasierforconsumerstoengagewiththeirfinances,costefficienciesdrivenbyeconomiesofscaleandtailoringservicestoparticularconsumersegments.Fromaregulatoryperspective,firmsmustinvesttoestablishgovernanceandcompliancecontrolsthatarefittodealwiththedifferenttypeandscaleofrisksintroducedbyautomatedadvice.TheFCAplanstopublishadditionalguidancewhichwilllikelyhelpreduceregulatoryuncertaintyinrelationtosomeservices.
Incumbentsarewellplacedbothtoshapeanddrivethepaceofadoptioninmostadvicesituationsastheyhaveaccesstoalargepoolofexistingclients.Iftheydonotlead,otherswill.Thismeansincumbentsmayriskbecomingbystandersastheindustryrapidlyadoptsnewproductsandroutes-to-market,ashashappenedinmanyotherindustries,rangingfromretailtotelecoms.
43
The next frontier | The future of automated financial advice in the UK
Endnotes
1. YouGovplcconductedanonlinesurveyof2,046GBadultson23-24January2017forDeloitte.ThefigureshavebeenweightedandarerepresentativeofallGBadults(aged18+).
2. TheFinancialAdviceMarketReviewdefinedtheadvicegapasasituationinwhichconsumersareunabletogetadviceandguidanceonaneedtheyhaveatapricetheyarewillingtopay.ResponsestotheCallforInputindicatedstronglythatthereisanadvicegap,andthatthisisparticularlysignificantinrelationtopensionsandsavingsand,toalesserextent,protection.The Financial Advice Market Review Final Report,HMTandFCA,March2016https://www.fca.org.uk/publication/corporate/famr-final-report.pdfSeealsoBridging the advice gap Delivering investment products in a post-RDR world,Deloitte,2012 https://www2.deloitte.com/content/dam/Deloitte/uk/Documents/financial-services/deloitte-uk-fs-rdr-bridging-the-advice-gap.pdf
3. Forexplanationsoftheseterms,pleaseseeFigure3.4. Forexample,theFCAdescribesautomatedadviceas‘fullyorpartiallyautomatedonlineservicesandothermodelsthatusetechnologytodeliverlowercostadvice’.In
contrast,descriptionsbyBundesanstaltfür.Finanzdienstleistungsaufsicht(BaFin),theAustralianSecuritiesandInvestmentsCommission(ASIC)andtheUSDepartmentofLaborexcludepartially-automatedmodels,talkingaboutadvicewithoutanyinteractionwithahumanadviser.Theexactservicesdescribedcanalsodiffer.Forexample,whilesomeregulatorstalkonlyaboutadvice,theFCA’sdescriptionalsoincludesdiscretionaryinvestmentmanagementandBaFin’sdescriptionalsoincludesautomated‘socialtrading’(wheretheinvestmentstrategyofasuccessfultraderisautomaticallyreplicatedinthecustomer’sinvestmentportfolio).Descriptionsbyinternationalregulatorybodies,suchastheInternationalOrganizationofSecuritiesCommissions(IOSCO)andtheEuropeanSupervisoryAuthorities(ESAs),arebroad,coveringarangeoffully-orpartially-automatedservices.Advice Unit,FCA,June2016,https://www.fca.org.uk/firms/project-innovate-and-innovation-hub/advice-unit FinTechs: Young IT companies on the financial market,BaFin,January2016https://www.bafin.de/SharedDocs/Veroeffentlichungen/EN/Fachartikel/2016/fa_bj_1601_fintechs_en.html Regulatory Guide 255, Providing digital financial product advice to retail clients,ASIC,August2016,http://download.asic.gov.au/media/3994496/rg255-published-30-august-2016.pdf. BestInterestContractExemption,Federal Register Volume 81, Number 68,April2016,http://webapps.dol.gov/FederalRegister/HtmlDisplay.aspx?DocId=28807&AgencyId=8&DocumentType=2 Joint Committee Discussion Paper on automation in financial advice,EuropeanSupervisoryAuthorities,December2015,https://www.eba.europa.eu/documents/10180/1299866/JC+2015+080+Discussion+Paper+on+automation+in+financial+advice.pdf Report on the IOSCO Social Media and Automation of Advice Tools Surveys,IOSCO,July2014,https://www.iosco.org/library/pubdocs/pdf/IOSCOPD445.pdf
5. In2015,theUKgovernmentchangedthelawtogiveretireesamuchwiderrangeofoptionsonhowtoconverttheirpensionpotintoretirementincomethanpreviously,whenmanypeoplewere‘forced’intoanannuity.
6. Intheconsumersurvey,weaskedaboutconsumers’willingnesstouseandpayforadviceusingabroaddefinitionof‘advice’whichwouldincludeunregulatedguidance.Thisisbecauseconsumersarenotreadilyabletodistinguishbetweenregulatedandunregulatedadvice.
7. Peeraveragesarebasedonthefollowinggroups: •Netwealthexcludingpensionsandproperties–AllGBadults(Sample:2,046) •Netpropertywealth–AllGBadults(Sample:2,046) •Pensionholdings–AllGBadults(Sample:2,046) •Valueofpensions–AllGBadultswithapension(Sample:922) •Willingnesstopayforautomatedadviceoninvesting£11,000–AllGBadultswithmorethan£5,000insavings&investments(excludingpensionsandpropertywealth)(Sample:842) •Willingnesstopayforautomatedadviceonconverting£30,000pensionsavingsintoalumpsumandretirementincome–AllGBworkersaged30+with pensionsavingsworthmorethan£20,000(Sample:173) •Willingnesstopayforautomatedadviceonconverting£30,000pensionsavingsintoalumpsumandretirementincomeamongconsumersaged45+–AllGBworkersaged45+withpensionsavingsworthmorethan£20,000(Sample:117) •Willingnesstopayforautomatedadviceoninvesting£80monthlypensioncontributions–AllGBworkerswithaDCpension(Sample:921) •Willingnesstopayforautomatedadviceonfindingamortgage–AllGBadultsagedunder50(Sample:1,100)
8. HMTintendstoamendtheregulatorydefinitionofadviceforregulatedfirmsgivingadviceoninvestments,DCpensionschemesandinsurancetobringitintolinewiththeMiFIDIIdefinition,whichwillrestrictittocaseswhereapersonalrecommendationismade.By‘regulatedfirms’,HMTmeansauthorisedfirmsexceptthosethatholdonlyoneorbothofthefollowingpermissions:‘advisingoninvestments’or‘agreeingtoadviseoninvestments’.Thischartshowswheretheadviceboundarywilllieforregulatedfirmsafterthischangehasbeenmade.NotethatHMThasnotstatedanyintentiontochangethedefinitionofadviceformortgages.
9. Financial Advice Market Review Final report,March2016,HMTandFCA,https://www.fca.org.uk/publication/corporate/famr-final-report.pdf10. AdviceUnit,FCA,2016,https://www.fca.org.uk/firms/project-innovate-and-innovation-hub/advice-unit11. InApril2017theFCApublishedaconsultationsettingoutproposedguidancetosupportfirmsoffering‘streamlinedadvice’onalimitedrangeofconsumerneeds.In
summer2017theFCAwillpublishanotherconsultationonproposedrevisedguidanceontheamendedadviceperimeterandnon-advisedservices.Insummer2017,theFCAalsoexpectstosetoutfurtherguidanceinformedbytheAdviceUnit’sworkwithfirms.FormoreinformationseeGC17/4,FCA,April2017,https://www.fca.org.uk/publication/guidance-consultation/gc17-04.pdf
12. Streamlinedadviceisacollectivetermusedtodescribeadvisoryservices(suchassimplifiedandfocusedadvice)thatprovideapersonalrecommendationthatislimitedtooneormoreofacustomer’sspecificneeds.Theservicedoesnotinvolveanalysisofthecustomer’scircumstancesthatarenotdirectlyrelevanttothoseneeds.
44
The next frontier | The future of automated financial advice in the UK
13. Amending the definition of financial advice: consultation response,HMT,March2017,https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/594790/pu2041_amending_definiton_financial_advice_response_2017_FINAL.pdf. By‘regulatedfirms’,HMTmeansauthorisedfirmsexceptthosethatholdonlyoneorbothofthefollowingpermissions:‘advisingoninvestments’or‘agreeingtoadviseoninvestments’.UKfirmscurrentlyfacetwodefinitionsoffinancialadvice:oneisdefinedinMiFID,andanotherintheRegulatedActivitiesOrder(RAO).Thelatterpre-datestheMiFIDdefinitionandiswiderinscope.Theregulatedactivityof‘advisingoninvestments’underArticle53oftheRAOiswiderinscopethan‘investmentadvice’underMiFID.ThisisbecauseMiFIDrequiresadvicetobeofapersonalnaturewhereastheRAOdoesnot.ForadvicetoberegulatedunderArticle53oftheRAO,itmust: •relatetoarelevantinvestment,whichincludescontractsofinsurance •begiventoapersonintheircapacityasaninvestororpotentialinvestor(orintheircapacityasagentforaninvestororpotentialinvestor) •relatetothemeritsofthembuying,selling,subscribingfororunderwritingtheinvestment(orexercisingrightstobuy,sell,subscribefororunderwritesuchan investment) MiFIDinvestmentadviceinvolvestheprovisionofpersonalrecommendationstoacustomer,eitheruponthecustomer’srequestoronthefirm’sinitiative.Itcomprises three main elements: •theremustbearecommendationthatismadetoapersonintheircapacityasaninvestororpotentialinvestor(orintheircapacityasanagentforaninvestoror personalinvestor) •therecommendationmustbepresentedassuitableforthepersontowhomitismadeorbasedontheinvestor’scircumstances •therecommendationmustrelatetotakingcertainstepsinrespectofaparticularinvestmentwhichisaMiFIDfinancialinstrument,namelytobuy,sell,subscribe for,exchange,redeem,holdorunderwriteaparticularfinancialinstrument(orexercisearighttobuy,sell,subscribefor,exchange,orredeemafinancial instrument)FAMRrecommendedamendingthedefinitionofadviceintheRAOtobringitinlinewithMiFIDtohelpremoveuncertainty.
14. Online investment and advice services – the consumer experience,ConsumerPanel,December2016,https://www.fs-cp.org.uk/sites/default/files/final_panel_position_paper_online_investment_and_advice_services.pdf
15. Online investment and advice services – the consumer experience,ConsumerPanel,December2016,https://www.fs-cp.org.uk/sites/default/files/final_panel_position_paper_online_investment_and_advice_services.pdfandFG15/1 Retail investment advice: clarifying the boundaries and exploring the barriers to market development, FCA,January2015,https://www.fca.org.uk/publication/finalised-guidance/fg15-01.pdf
16. FG15/1 Retail investment advice: clarifying the boundaries and exploring the barriers to market development,FCA,January2015,https://www.fca.org.uk/publication/finalised-guidance/fg15-01.pdf https://www.fca.org.uk/publication/finalised-guidance/fg15-01.pdf
17. Compiled from our discussions with clients and from Feedback Statement on Smarter Consumer Communications,FCA,October2016,https://www.fca.org.uk/publications/discussion-papers/smarter-consumer-communications-further-step-journey
18. FidelityLabsCityVisualisation,http://new.fidelitylabs.com/city/19. Feedback Statement on Smarter Consumer Communications,FCA,October2016,https://www.fca.org.uk/publications/discussion-papers/smarter-consumer-
communications-further-step-journey20. Report on digital investment advice,FinancialIndustryRegulationAuthority(FINRA),March2016,https://www.finra.org/sites/default/files/digital-investment-advice-
report.pdf21. Regulatory guide 155 – Providing digital financial product advice to retail clients,ASIC,August2016,http://download.asic.gov.au/media/3994496/rg255-published-30-
august-2016.pdf22. The Promise of FinTech – Something New Under the Sun?MarkCarney,BankofEngland,January2017,http://www.bankofengland.co.uk/publications/Pages/
speeches/2017/956.aspx.Iftheuseofautomatedplatformsbecamewidespread,andtheunderpinningalgorithmsusedverysimilarlogic,theremaybeariskthatasignificantnumberofcustomersmayenduptransactinginthesamewaysinrelationtocertainfinancialproducts.TheESAsJointCommitteehasalsohighlightedthis‘herdingrisk’and,likeMrCarney,believesitcouldleadtoincreasedvolatilityinthemarketorincreasepro-cyclicality,whichinturnwouldresultincustomerdetrimentintheeventofamarket‘shock’event.
23. Regulating cyber-resilience,Deloitte,December2016http://blogs.deloitte.co.uk/financialservices/2016/12/regulating-cyber-resilience.html24. Spixii,https://spixii.ai/25. “Alexa, what’s an annuity?” Insurance company Aviva is using Amazon Echo,CityAM,January2017,http://www.cityam.com/256375/alexa-whats-annuity-insurance-
company-aviva-using-amazon26. Achatbotisacomputerprogramdesignedtosimulateconversationwithhumanusers,especiallyovertheinternet.27. Cost of Advice Guide,Unbiased,https://www.unbiased.co.uk/cost-of-financial-advice28. SeeSection1formoredetailontheregulatorydefinitionofadvice.29. BritishBankers’AssociationHighStreetBanking,December2016,https://www.bba.org.uk/news/statistics/high-street-banking/30. HerMajesty’sRevenueandCustomsIndividualSavingsAccountStatistics,August2016,https://www.gov.uk/government/uploads/system/uploads/attachment_data/
file/547217/Full_Statistics_Release_August_2016.pdf31. Private client wealth managers shun moderately rich,FinancialTimes,June2015https://www.ft.com/content/a5c19f24-15c4-11e5-be54-00144feabdc032. See for example Cost of Advice Guide,Unbiased,https://www.unbiased.co.uk/cost-of-financial-advice33. DeloitteanalysisbasedonFormADVdated31January2017forBettermenthttps://www.adviserinfo.sec.gov/IAPD/content/ViewForm/crd_iapd_stream_pdf.
aspx?ORG_PK=149117andFormADVdated2August2016forPersonalCapitalhttps://www.adviserinfo.sec.gov/IAPD/content/ViewForm/crd_iapd_stream_pdf.aspx?ORG_PK=155172
34. Charges and interest rates,HargreavesLansdown,http://www.hl.co.uk/investment-services/fund-and-share-account/charges-and-interest-rates35. SeeCOBS6.1AintheFCAHandbook,https://www.handbook.fca.org.uk/handbook/COBS/6/1A.html36. TheFCAhasproposedguidanceonadviserchargesforverticallyintegratedfirms(VIFs)inrelationtotheinterpretationof‘longterm’andtheflexibilityallowedinthe
cross-subsidisationrules.FormoreinformationseeQuarterlyConsultationNo15,FCA,December2016,https://www.fca.org.uk/publication/consultation/cp16-39.pdf
37. TheDeloitteUKInsightteamdevelopedamodeltoestimatetheprofitabilityofstand-aloneroboadvisersatdifferentlevelsofAuMunderthreefeescenarios.Thisincludedassumptionsonadministration,compensationandmarketingcosts.Theunderlyingdatawasobtainedfrompubliclyavailablesourcesincludingthereportsand disclosures made by a range of UK and US asset and wealth managers as well as our own proprietary data sources.
38. DeloitteanalysisbasedonproviderSecuritiesandExchangeCommissionfilings
45
The next frontier | The future of automated financial advice in the UK
39. TheDeloitteUKInsuranceInsightteamdevelopedamodeltoestimateUKlifeinsuranceprofitsbyproduct2015-25.Profitswerederivedfromestimatedassetsandprofitmarginscombinedwithgrowthassumptions.Theunderlyingdatawasobtainedfrompublicandprivatesourcesincludingannualreports,regulatoryreturnsandanalystreports.FormoreinformationseeUKlifeinsurancefutures,Deloitte,January2016,https://www2.deloitte.com/uk/en/pages/financial-services/articles/uk-life-insurance-futures.html
40. UK life insurance futures,Deloitte,January2016,https://www2.deloitte.com/uk/en/pages/financial-services/articles/uk-life-insurance-futures.html41. Workplace Pensions, UK, June 2016,Mintel,June201642. Pensions Dashboard prototype to be ready by spring 2017,HMT,September2016,https://www.gov.uk/government/news/pensions-dashboard-prototype-to-be-ready-
by-spring-201743. DC trust: a presentation of scheme return data 2013 – 2014,ThePensionsRegulator,http://www.thepensionsregulator.gov.uk/doc-library/dc-trust-a-presentation-of-
scheme-return-data-2014.aspx44. Receiving pension savings advice for £100 a year,WealthWizards,https://www.wealthwizards.com/blog/receiving-pension-savings-advice-for-100-a-year/45. The facts and figures behind the value of advice,Unbiased,https://www.unbiased.co.uk/value-of-advice/report-stats46. Cost of Advice Guide, Unbiased,https://www.unbiased.co.uk/cost-of-financial-advice47. SeeCOBS6.1A.1RintheFCAHandbook,https://www.handbook.fca.org.uk/handbook/COBS/6/1A.html48. Data Bulletin Issue 8,FCA,February2017,https://www.fca.org.uk/publication/data/data-bulletin-issue-8.pdf.ThedatacoverstheperiodApriltoSeptember2016.By
‘drawdown’wemeananarrangementwherebythecustomercanwithdrawmoneyfromtheirpensionpotovertimetocreateanincome.49. TheFCAfoundthatinQ3201520%ofconsumerstoldtheirproviderstheyhadusedPensionWise.Providersareonlyrequiredtorecordwhetheraconsumersaid
theyusedPensionWisewhenconsumersarenotusingaregulatedadviser.Data Bulletin Supplement Retirement income market data,FCA,April2016,https://www.fca.org.uk/publication/data/data%20bulletin%20suppl%20apr%2016.pdf
50. Introducing a Pensions Advice Allowance: response to the consultation,HMT,February2017,https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/589163/Pensions_Advice_Allowance_response_to_the_consultation_web.pdf
51. FS16/10 Smarter Consumer Communications,FCA,October2016,https://www.fca.org.uk/publication/feedback/fs16-10.pdf52. SeeCOBS19.4intheFCAHandbook,https://www.handbook.fca.org.uk/handbook/COBS/19/4.html.53. Update on our work around pensions and retirement income,FCA,July2016,https://www.fca.org.uk/publications/market-studies/update-our-work-around-pensions-
and-retirement-income54. CP16/37 Implementing information prompts in the annuity market,FCA,November2016,https://www.fca.org.uk/publication/consultation/cp16-37.pdf55. Update on our work around pensions and retirement income,FCA,July2016,https://www.fca.org.uk/publications/market-studies/update-our-work-around-pensions-
and-retirement-income56. CostofAdviceGuide,Unbiased,https://www.unbiased.co.uk/cost-of-financial-advice57. Pension freedom for 400,000 hardworking people from today,HMT,March2014,https://www.gov.uk/government/news/pensions-freedom-for-400000-hardworking-
people-from-today 58. ThisincludestransferringfromaDBpensionschemeintoaDCscheme,orgivingupaDCpensionschemewithaguaranteedannuityrate.59. Market commentary January 2017,CouncilofMortgageLenders,https://www.cml.org.uk/news/news-and-views/market-commentary-january-2017/60. SeeMCOBS4intheFCAHandbook,https://www.handbook.fca.org.uk/handbook/MCOB/4/?view=chapter.Inparticular:
•Forthemajorityofcustomers,firmsarerequiredtoprovideregulatedadvicetocustomersifthereisanyspokenorotherinteractivedialoguebetweenthefirm and the customer during the sale. •Forcertainvulnerablecustomers,salesmustalwaysbeadvised. •Incaseswhereacustomerisahighnetworthorprofessionalcustomer,orwheretheloanissolelyforabusinesspurpose,thecustomercanoptnottoreceive adviceevenifthereisaspokenorinteractivedialoguebetweenthefirmandthecustomerduringthesale.
61. Mortgages Market Study Terms of Reference,FinancialConductAuthority,December2016,https://www.fca.org.uk/publication/market-studies/ms16-02-1.pdf62. Mortgages Market Study Terms of Reference,FinancialConductAuthority,December2016,https://www.fca.org.uk/publication/market-studies/ms16-02-1.pdf63. A guide to mortgage fees and costs,MoneyAdviceService,https://www.moneyadviceservice.org.uk/en/articles/mortgage-related-fees-and-costs-at-a-glance64. Mortgages Market Study Terms of Reference,FinancialConductAuthority,December2016,https://www.fca.org.uk/publication/market-studies/ms16-02-1.pdf65. Our Future Mission,FCA,October2016,https://www.fca.org.uk/publication/corporate/our-future-mission.pdfandConsumer Vulnerability,FCA,February2015,
https://www.fca.org.uk/publication/occasional-papers/occasional-paper-8.pdf66. AccordingtodatafromtheAssociationofBritishInsurers,2014grosswrittenpremiumstotalled£8,920millionacrosspurelifeinsurance,incomeprotection,
standalonecriticalillness,long-termcareandotherprotectionbusiness.TheDeloitteUKInsuranceInsightteamdevelopedamodeltoestimateUKlifeinsuranceprofitsbyproduct2015-25.Profitswerederivedfromestimatedassetsandprofitmarginscombinedwithgrowthassumptions.Theunderlyingdatawasobtainedfrompublicandprivatesourcesincludingannualreports,regulatoryreturnsandanalystreports.FormoreinformationseeUK life insurance futures,Deloitte,January2016,https://www2.deloitte.com/uk/en/pages/financial-services/articles/uk-life-insurance-futures.html
67. Swiss Re’s Term and Health Watch 2012 reports significant increases in UK Life Assurance Protection Gap, overall UK long-term protection market remains resilient in 2011, SwissRe,June2012,http://www.swissre.com/media/news_releases/nr_20120611_Term_Health_Watch.html
68. Term Assurance, UK, June 2015,Mintel,June201569. Term Assurance, UK, June 2015,Mintel,June2015.Note:thepremiumfigureisbasedonconsumerresearchnotindustrypremiumdata70. Term Assurance, UK, June 2015,Mintel,June201571. Term Assurance, UK, June 2015,Mintel,June201572. Term Assurance, UK, June 2015,Mintel,June201573. TheEUGeneralDataProtectionRegulation,Official Journal of the European Union,May2016,http://eur-lex.europa.eu/legal-content/EN/TXT/
PDF/?uri=CELEX:32016R0679&from=EN74. ThisscenarioissimilartoanexampletheFCAgivesinitsfinalisedguidanceonthedefinitionofretailinvestmentadvice.Thisguidancedoesnotapplytoprotection
productswithoutaninvestmentelement.However,thedefinitionofapersonalrecommendationintheFCAHandbookisthesameforprotectionproductsasforinvestmentproducts,sosimilarprinciplesarelikelytoapply.FG15/1 Finalised guidance on retail investment advice,FCA,January2015,https://www.fca.org.uk/publication/finalised-guidance/fg15-01.pdf
75. ICOBS5.3oftheFCAHandbooksetsoutrulesforadviceonprotectionproducts.Thesuitabilityrulesarenotasstringentasforinvestmentproducts,butafirmmusttakereasonablecaretoensurethesuitabilityofitsadviceforanycustomerwhoisentitledtorelyuponitsjudgment. https://www.handbook.fca.org.uk/handbook/ICOBS/5/3.html
46
The next frontier | The future of automated financial advice in the UK
Gavin NorwoodPartner Deloitte Digital+441179841154gnorwood@deloitte.co.uk
Margaret Doyle PartnerHead of Financial Services Insight+442070076311madoyle@deloitte.co.uk
Tony GaughanPartnerHead of Investment Management+442073032790tgaughan@deloitte.co.uk
Neil TomlinsonPartnerHead of Banking+442073032333ntomlinson@deloitte.co.uk
David StrachanPartnerHead of the EMEA Centre for Regulatory Strategy+442073034791dastrachan@deloitte.co.uk
Cindy ChanPartnerRetail Conduct Risk Lead+442073035836cichan@deloitte.co.uk
David RushPartnerHead of Insurance+442073036302drush@deloitte.co.uk
Rahul SharmaDirectorFinancial Services Insight+442070075557rahulsharma@deloitte.co.uk
Joy KershawManagerEMEA Centre for Regulatory Strategy+442073036575jkershaw@deloitte.co.uk
Peter EvansManagerFinancial Services Insight+442073030010peevans@deloitte.co.uk
Rosalind FergussonSenior ManagerEMEA Centre for Regulatory Strategy+442073037997rfergusson@deloitte.co.uk
Valeria GalloManagerEMEA Centre for Regulatory Strategy+442073036745vgallo@deloitte.co.uk
Contacts
Authors
47
The next frontier | The future of automated financial advice in the UK
Notes
48
The next frontier | The future of automated financial advice in the UK
Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited (“DTTL”), a UK private company limited by guarantee, and its network of member firms, each of which is a legally separate and independent entity. Please see www.deloitte.co.uk/about for a detailed description of the legal structure of DTTL and its member firms.
Deloitte LLP is the United Kingdom member firm of DTTL.
This publication has been written in general terms and therefore cannot be relied on to cover specific situations; application of the principles set out will depend upon the particular circumstances involved and we recommend that you obtain professional advice before acting or refraining from acting on any of the contents of this publication. Deloitte LLP would be pleased to advise readers on how to apply the principles set out in this publication to their specific circumstances. Deloitte LLP accepts no duty of care or liability for any loss occasioned to any person acting or refraining from action as a result of any material in this publication.
© 2017 Deloitte LLP. All rights reserved.
Deloitte LLP is a limited liability partnership registered in England and Wales with registered number OC303675 and its registered office at 2 New Street Square, London EC4A 3BZ, United Kingdom. Tel: +44 (0) 20 7936 3000 Fax: +44 (0) 20 7583 1198.
Designed and produced by The Creative Studio at Deloitte, London. J11582