2025全球經(jīng)濟(jì)展望

2025

THEMATICOUTLOOK

AIandgeopoliticsforgenewpaths

Keytakeaways

01

MODESTRATECUTSCOULDSUPPORT

SENSITIVEASSETS

RecentFedratecutscouldstartto

takepressureoffinterestrate

sensitivecompaniesandnon-

cash-flowing1assets.This

presentspotentialtailwindsforbiotechstocks,bitcoin,andgold.

02

REBUILDINGTHEU.S.PHYSICALECONOMY

Themescenteredaround

rebuildingthephysicaleconomy,likeinfrastructure,manufacturing,andhomebuildingmaybebetterpoisedtobenefitinthepost-

electionenvironment,astheysitattheintersectionofpolicy

tailwindsandstructuralchanges.

03

AI’SBUILDPHASEACCELERATES

MassiveinvestmentinAI

infrastructureaswellasevermorepowerfulchipsandmodels,are

layingthegroundworkforincreasedadoption.

“

WebelieveAIandgeopoliticswillremainkeythemesfor2025,yettherearesignificant

shiftsintheunderlyingpolicies,

demographics,andtechdevelopmentsthatwilldrivethemforward.Aheadofthenewyear,investorsshouldconsiderwhat

exposuretheyhavetothesethemesandhowtheymaypositiontheirportfoliosforthesestructuraltrends.”

JayJacobs,CFA

U.S.HeadofThematicandActiveETFs,

atBlackRock

iS

1Non-cash-flowingassetsrefertoassetsthathavenoexpectationofeverpayingadividendorcoupon.

iCRMH1124U/S-4024094-1/14

01

MODESTRATECUTSCOULDSUPPORTRATE-SENSITIVECOMPANIESANDASSETS

FromMarch2022toSeptember2024,themarketsenteredaperiodofquicklyrisingandelevated

interestrates,bringingthefederalfundsratetoitshighestlevelssince20012.Thisenvironment

punishedrate-sensitiveinvestmentsontwofronts:

1.Companieswithlongpathstoprofitabilitysawvaluationscontractgivenhigherdiscountrates,and;

2.Firmsdependentonfloatingratedebtorthat

hadtorollovermaturingdebtwerehurtbyhigherfinancingcosts

Additionally,non-cash-flowingassetslikebitcoinandgoldfacedpressurefromrisingopportunitycosts

comparedtoholdinginterest-payingassetslike

bonds.Nowtheseheadwindscouldabate;The

FederalReservehascutratesby75basispoints3asofNovember2024,andwhileratesmaynotreturntopre-pandemiclevels,apathofadditionalmodestratecutsisanticipatedbythemarketfor2025.4

Medicalinnovationgetsashotinthearm

Higherrateshavehadaparticularlynegativeimpactonbiotechfirmsbyshrinkingvaluationsanddrivingupborrowingcosts,causingmanycompaniesto

reduceR&Dspending.Shouldratescontinuetofall,itcouldreducefinancingcostsandpotentiallygive

biotechfirmsmoreconfidencetoexpandR&D

budgets.Increasedspendingcombinedwiththe

potentialintroductionofAItothediscoveryandtrialprocess,couldleadtoasurgeindrugdevelopment.AIisalreadybeingtestedindrugdiscoverytopredictproteinstructuresandchemicalreactions,aswellastoconductdrugtrialsinautomatedlabsanddigital“in-silico”experiments.By2025,over30%ofnew

drugsareexpectedtobediscoveredusinggenerativeAItechniques,potentiallysavingbiotechcompanies25%to50%intimeandcostsfromdiscoveryto

preclinicalstages.5

“

Structuraldemandinthehealthcaresector,drivenbyanagingpopulation,combinedwithrapid

innovationinmedical

technologyanddrug

development,providesastrongfoundationfor

potentialgrowth.”

Dr.ErinXie

LeadPortfolioManager,HealthSciences,BlackRockFundamentalEquities

Suchinnovationcouldn’tcomeatamorecritical

time;ageingpopulations,particularlyindeveloped

markets,arebolsteringdemandformanydrugsandtreatmentsincategoriesthatdisproportionately

impactseniors.Breakthroughslikepersonalized

cancervaccines,6treatmentsthatcouldeliminatetheneedforinsulintherapyindiabetics,7and

intravenousantibodytreatmentsthatcouldslowtherateofcognitivedeclinewithAlzheimer’spatients8appeartobeonthehorizon.Theserevolutionary

treatmentscouldalsodriveasurgeinmergersandacquisitions.Largepharmaceuticalcompanies

lookingtorefreshtheirproductofferingsamidan

anticipatedmajorpatentcliff9forexisting

blockbusterdrugscouldfurthersupportbiotechvaluations.

2

2.St.LouisFederalReserveBank.“FREDEconomicData-FederalFundsEffectiveRate”Asof10/1/2024.3.Abasispointisonehundredthofonepercent.Exampleonebasispoint=0.01%4.TheWeek,“What’sNextforUSInterestRates.”11/8/20245.Gartner,“BeyondChatGPT:TheFutureofGenerativeAIforEnterprises,”01/26/2023.AndBCG&Wellcome,UnlockingthepotentialofAIinDrugDiscovery,”June2023.Forward-lookingestimatesmaynotcometopass.6.WorldEconomicForum.“12newbreakthroughsinthefightagainstcancer.”O(jiān)ctober21,2024.7.NewsMedical.“Promisingnewtreatmentstrategyfortype2diabetes.”O(jiān)ctober14,2024.8.NewYorkPresbyterian.“WhattoKnowAboutNew

Alzheimer’sDiseaseTreatments.”July2,2024.9.IQVIA“BiopharmaM&Amomentumreturnsbutuncertaintiesremain”July6,2023.

iCRMH1124U/S-4024094-2/14

Modestratecutscouldsupportrate-sensitiveassets

12%

11%

Demographictrendsarelikelytocontinueregardlessoftheeconomicenvironment

Proportionofpopulation65orolder1

12%|198520%|2030e

HealthcarespendingasapercentageofGDP3

19%

13%

12%

10%

9%

6%

4%4%

3%

3x

Spenton

healthcarewhen

+65yearsvs<65years2

20%

OftheU.S.

populationwillbe+65yearsby20301

EmergingmarketsDevelopedmarkets

Source:1.WorldBankdataandestimates,currentasofApril30,2023.2.Deloitte2019OutlookforHealthcare3.TheWorldBank,CurrentasofApril30,2023,basedon2020expenditures.Forillustrativepurposesonly.Thereisnoguaranteethatforecastsmadewillcometopass.

InvestorsinterestedinbiotechnologymayconsidertheiSharesHealthInnovationActive

ETF(BMED

).

Lowerinterestratesmayenhancebitcoin’snear-termattractiveness

Bitcoinisanemergingasset

withuniquedemand

driversstemmingfromitspropertiesasafinite

supply,alternativemonetaryassetthatisdigitallynative,global,anddetachedfromgovernmentsandtheirexistingeconomicandmonetarysystems.

Bitcoinremainsavolatileassetingeneral,butwe

believebitcoin’slong-termfundamentalsare

largelydistinct

fromthetraditionalmacroeconomicindicatorsthatdriveequitiesandother“riskassets,”andforcertainriskfactorsmayevenbeinverted.

However,inthemoreimmediateterm,changesinrealinterestrates(nominalinterestratesminusinflation)dotendtoimpactnon-interestpayingassetslikebitcoinandgoldastheychangethe

opportunitycostoftheseassetscomparedtoincome-payinginvestmentslikebonds.

Forexample,thesharpincreaseinrealinterestratesduring2022wasacontributingfactortothe67%

correctioninbitcointhatyear.Now,amidratecuts,realinterestratescouldfall.Thiscouldcause

investorstoviewbitcoinasincreasinglyattractiverelativetootherassets.

Inadditiontothemacroenvironment,thereisa

renewedsenseofoptimismthatregulatoryclarityforbitcoinanddigitalassetsmorebroadlymayemergefollowingtheU.S.election.President-electDonald

Trumpcampaignedonmaintainingastrategic

bitcoinreserve,whilepro-cryptopoliticiansinthe

HouseandSenateracesfrombothsidesoftheaisleenjoyedelectoralsuccess.Themacroenvironmentcombinedwithsupportivepoliciescouldcombinetoaccelerateandbroadenbitcoin’sadoption.

3

iCRMH1124U/S-4024094-3/14

“

Whilewebelievebitcoin’slong-termadoptiontrajectorywillbeprimarilydrivenbyitsfundamentaluse-caseasaglobalmonetaryalternative,decliningrealinterest

rates–whetherthroughnominalratecutsoranuptickininflation–mayserveasanothercatalyst.”

RobbieMitchnick

HeadofDigitalAssets,BlackRock

Bitcoinhasshownstronginversecorrelationwithrealinterestrates

5YU.S.TreasuriesTIPSYield(InvertedScale)%BitcoinPrice,USD

-3.0$75,000

R:-0.3

-2.0

$60,000

-1.0

$45,000

0.0

$30,000

1.0

$15,000

2.0

$0

3.0

Jan-20Jul-20Jan-21Jul-21Jan-22Jul-22Jan-23Jul-23Jan-24Jul-24

Source:BloombergBitcoinSpotPrice,St.LouisFed,andBlackRockcalculations,asof9/30/24.CorrelationofBitcoinweeklyreturnstotheweeklychangein5YU.S.TreasuriesTIPSyieldoverperioddisplayed(Jan.2020toSeptember2024)is-0.3.Correlationmeasureshowtwosecuritiesmoveinrelationtoeachother.Correlationrangesbetween+1and-1.Acorrelationof+1indicatesreturnsmoveintandem,-1indicatesreturnsmoveinoppositedirections,and0indicatesnocorrelation.Indexperformanceisforillustrativepurposesonly.Indexperformancedoesnotreflectanymanagementfees,transactioncostsorexpenses.Indexesareunmanagedandonecannotinvestdirectlyinanindex.Pastperformancedoesnotguaranteefutureresults.

InvestorsinterestedingainingexposuretobitcoinmayconsidertheiSharesBitcoinTrust

ETF(IBIT

).

TheiSharesBitcoinTrustETFisnotaninvestmentcompanyregisteredundertheInvestmentCompanyActof1940,andthereforeisnotsubjecttothesameregulatoryrequirementsasmutualfundsorETFsregisteredundertheInvestmentCompanyActof1940.

ThisinformationmustbeaccompaniedorprecededbyacurrentiSharesBitcoinTrustETF

prospectus

.Pleasereadtheprospectuscarefullybeforeinvesting.

4

iCRMH1124U/S-4024094-4/14

REBUILDINGTHEU.S.PHYSICAL

02

ECONOMY:INFRASTRUCTURE,

MANUFACTURING,ANDHOUSING

RebuildingthephysicaleconomyintheU.S.-

includingimprovingandrepairinginfrastructure,

expandingmanufacturingcapacity,andacceleratinghomebuilding-hasbecomeatopicofincreasing

consensusacrossboththepublicandprivate

sectors.Webelievespendingandpolicychangeswillcontinuetoacceleratethistheme.

Infrastructureinvestment

toaccelerateintheyearsahead

2021’sbipartisanInfrastructureInvestmentand

JobsAct(IIJA)markedthesinglelargest

infrastructureinvestmentinU.S.history.Theact

allocated$1.2trilliontobuildandrepairthe

nation'sbridges,airports,waterways,publictransitandmore.10Ittakestimetogetshovelsinthe

groundandputtheIIJAinvestmentstowork,asdepictedinthechartbelow,butfederalestimatesfor2025spendingpredictasignificantincreaseoverpriorlevels.

$(Billions)

ForecastofIIJAimplementationsupportsanaccelerationofinfrastructure

BudgetAuthorityEst.Outlays

600

500

400

300

200

100

-

2022202320242025202620272028202920302031

Source:BlackRockusingdatafromCongressionalBudgetOffice.AsofOctober15,2024.Forillustrativepurposesonly.Forward-lookingestimatesmaynotcometopass.

10.U.S.DepartmentofTransportation.“BipartisanInfrastructure(BIL)/InfrastructureInvestmentandJobsAct(IIJA),November15,2021.

5

iCRMH1124U/S-4024094-5/14

RebuildingtheU.S.physicaleconomy

Threeyearsin,theimpactoftheIIJAisalready

visible:over60,000constructionprojectshave

Over

60,000

constructionprojectshaveadvanced

175,000

milesofroadway

advanced,175,000milesofroadway-enoughtocrosstheU.S.60times-arebeingrepaired,and

over10,200bridge-modernizationprojectsare

underway,withmanymoreprogressingacrossthenation.11

Yet,with$720billioninIIJAfundsstilltobe

allocated,thereisasignificantpotentialrunway

aheadforinfrastructurespending,12creatingan

opportunityforfurtherinvestmentasprojectsmovefromconcepttoconstruction.Further,private

spendingininfrastructurecouldhelpaccelerateandsupportthistrend.Overthepastfouryears,privatecompanieshaveannouncednearly$1

trillionincommitmentsacrossU.S.statesandterritories.13Privateinvestmentininfrastructurecouldfurtheraccelerateasgrowinggovernmentindebtednessrequiresnewfinancingmodelstofundcriticalinfrastructureneeds.

U.S.manufacturersbenefitingfromindustrialpolicies

Reshoring

–ortheactofbringingmanufacturingbacktodomesticsoil—gainedprominenceduringthe

pandemicwhenglobaltradewasseverelyimpactedbysupplychainbottlenecks.Byincreasingdomestic

productionandshorteningsupplychains,

governmentsandcompaniescanexertmorecontrolandreduceriskoverpreviouslycomplexandfragilesystems.Severalrecentpiecesoflegislation,liketheIIJA,CHIPS+ScienceAct,andInflationReductionAct,weredesignedto,amongotherthings,accelerate

reshoringbyallocatingbillionsofdollarstoimprovetransportationandsupportsemiconductorandEVmanufacturing,respectively.Semiconductorfirmshavesinceannouncedover80newprojectsintheUS,amountingtonearly$450billioninprivate

investment,14asignthesepolicesarehavinganimpact.

WhilethefutureoftheseBiden-erapoliciesmaybeuncertainunderthenewAdministration,webelieveadditionalpoliciescouldemergefollowingthe2024electionstofurtheracceleratethereshoringtheme.

Thepreviouslymentionedspendingbillstooka

‘carrot’approachatthefederalleveltoincentivizereshoring,butanincreaseintariffsorexportbanscouldintroducea‘stick’approachtoincreasingU.S.manufacturingcompetitivenessversusforeign

exporters.ThenewAdministrationandCongress

couldutilizebothapproachesinthenearterm.

Further,atstateandlocallevels,taxincentivesandprivateinvestmentscouldhelpacceleratethe

buildoutofnewfactories.

U.S.manufacturingremainsaclearareaofpoliticalconsensus,withbothDemocratsandRepublicans

indicating“netfavorable,”meaningalarger

percentageofbothpartiesratedthemanufacturingindustryasfavorablevsunfavorable;plus27%forDemocratsandplus53%amongRepublicans.15Inourview,thepotentialforsupportivepoliciesand

bipartisansupportatmultiplelevelsofgovernmentleaveU.S.manufacturingpoisedforcontinued

growth.

11.U.S.DepartmentofTransportation.“U.S.DoTCelebratesBidenAdministration’sProgressDeliveringontheBipartisanInfrastructureLaw”,09/18/2024.12.U.S.Departmentof

TransportationandConstructionDive.“$720BinIIJAfundsyettobeallocated”,09/19/2024.13.TheWhiteHouse.“InvestingInAmerica.”Accessedon10/30/2024.14.ElectronicDesign.“U.S.FabCapacityPoisedtoExplodeintheNextDecade.”May21,2024.15.VisualCapitalist.“HowDoDemocratsandRepublicansFeelaboutCertainU.S.Industries?”February19,2024.

6

iCRMH1124U/S-4024094-6/14

RebuildingtheU.S.physicaleconomy

Homebuildingbecomesademographicnecessity

Between2012and2023,17.2millionhouseholds

wereformed,yetonly10millionnewsingle-family

homeswerebuilt,resultingina7.2millionhomegap.Housingdemandisoutpacingsupplywith

householdgrowthoutpacingsingle-familypermitsin73ofthe100U.S.metroareas.16Foralmosta

decade,residentialconstructionhaslagged,asshowninthechartbelow.

Alaginresidentialconstructionhasledtosupplyanddemandimbalance

EquilibriumSurplus/Deficit(K)

1,000K

750K

500K

250K

-

(250K)

(500K)

(750K)

(1,000K)

(1,250K)

(1,500K)

Jan-1964

May-1966

Sep-1968

Jan-1971

May-1973

Sep-1975

Jan-1978

May-1980

Sep-1982

Jan-1985

May-1987

Sep-1989

Jan-1992

May-1994

Sep-1996

Jan-1999

May-2001

Sep-2003

Jan-2006

May-2008

Sep-2010

Jan-2013

May-2015

Sep-2017

Jan-2020

May-2022

Sep-2024

Source:RSMConsultants.“HousingandConstructionIndustryOutlook”withdatafromU.S.Census.Gov-NewResidentialConstructionchartfrom1959–2024.

16.R,“U.S.housingsupplygapgrowsin2023;growthoutpacespermitsinfast-growingsunbeltmetros,”2/27/2024.

7

iCRMH1124U/S-4024094-7/14

RebuildingtheU.S.physicaleconomy

Thislackofhousingsupplyhasbeenmetwith

increaseddemandasmoreandmorepeopleare

lookingtobuyhomes.Demographicsareplayinga

“

majorroleindrivingthishousingdemand,

especiallywithMillennials,whohaveovertaken

BabyBoomersasAmerica’slargestgeneration.17Millennialsarefeelingtheeffectsofhousing

Asreshoringinitiatives

continuetoplayoutduetoacombinationof

legislation,geopolitical

tensionsandafocuson

enhancingresilienceof

supplychains,companies

whospecializeinareassuchasbuildingproductsand

transportationequipmentcouldbenefit.”

shortages,whichiscausingaffordabilitytobeoutofreachformany.Infact,86%ofAmerican

renterssaytheywouldliketobuyahomebutcannotaffordone.18

Whilehousingbecameanationaltopic,discussedbybothmajorpartycandidatesduringtheir

campaigns,increasinghomeconstructionisstillhighlydependentonfactorslikeinterestratesandstate&localgovernments.Arenewedfocusonthepoliciesneededtosupportandacceleratehomeconstruction-alongwithpotentiallylowerinterestratesthatcouldsupportfinancingof

homebuildingaswellasmakemortgagesmoreaffordable-supportourviewthathomebuildingwillaccelerate.

AstheU.S.navigatesaneweraofbuildingoutitsphysicaleconomy,investorsinterestedinthese

value-orientedmanufacturingandhousing

themesmayconsidertheiSharesU.S.

InfrastructureETF(

IFRA

),iSharesU.S.

ManufacturingETF(

MADE

),andtheiShares

U.S.HomeConstructionETF(

ITB

).

TonyDeSpirito

GlobalChiefInvestmentOfficerofBlackRockFundamentalEquities

Investorsinterestedinanactiveapproachthatcanpotentiallycapturetheserebuildingthemes,may

considertheiSharesLargeCapValueActiveETF(

BLCV

),whichisactivelymanagedbyTonyDeSpiritoandtheBlackRockIncome&ValueTeam,whoseektoallocatetothemostcompellingcompaniesinthe

LargeCapValue

space,whichcanincludethemeslikemanufacturingandinfrastructure.

8

17.PewResearchCenter“MillennialsovertakeBabyBoomersasAmerica’slargestgeneration.”April28,2020.PewResearchCenter“MillennialsovertakeBabyBoomersasAmerica’slargestgeneration.”April28,2020.18.CNN.“MorethanhalfofAmericanrenterswhowanttobuyahomefearthey’llneveraffordone."July29,2024.

iCRMH1124U/S-4024094-8/14

03

AI’SBUILDPHASEACCELERATESINFRASTRUCTUREANDMODELS

AI’sinsatiabledemand

Ithasbeenjusttwoyearssincethereleaseof

ChatGPTinNovember2022,whichignitedthe

generativeAIrevolution.SurgingAIoptimism,

sincethen,includingestimationsofupto$15.4trillion19forthetotalannualvalueofAIand

analyticsacrossindustries,hascatalyzedmassiveinvestmentinAIinfrastructureinanarmsrace

amonghyperscalers.ItisourbeliefthatAI’s

continuedinfrastructurebuildout,alongwith

hardwareandmodelupgrades,ashighlighted

below,willdriveevermorepowerfulAItoolsintheyearsahead.SecondaryimpactsofAI’srise,

includingAI-politicsandcybersecurity,maybecomeincreasinglyrelevant.

AIscalinglawssuggestthatAIperformancescaleswiththesizeofthemodelandthedataonwhichittrains.Thisscalingrequiressignificantlyhigher

computationalpower,whichisfundamentaltotheAIrevolution.Someestimatessuggestthatby

2030,futureAImodelscouldbetrainedwithupto10,000timesmorecomputationalpowerthan

modelslikeGPT-4.20

Whiletherehasbeenskepticismaround“Ifbiggerisbetter”whenitcomestoAI21,weareenteringthe

thirdyearofAI’s“BuildPhase”,whichhaswitnessedasignificantsurgeofinvestmentinAIdatacenters,withthedatacenterGPUandAIASICS(custom-builtchips)marketsexpectedtoreachacombined$156billionby2025and$233billionby2029.22

“

Webelievemostofthe

growthisstillahead,with

globalITspending

predictedtojumpupwardsof9%in2025,markingoneofthelargestpercent

increasesinthiscentury.23

WEAREHERE

2.

ADOPTIONPHASE

ThisiswhenAIsystemswillbeintegratedintobothconsumerand

enterpriseapplications.

3.

TRANSFORMATIONPHASE

ThefinalphasewillseeAIfundamentallyreshapetheglobaleconomy.New

industriesandbusinessmodelswillemerge.

1.

BUILDPHASE

CharacterizedbysignificantinvestmentinAIstructure,

ParticularlyinAIdatacenter,computinghardware

andcloud.

Source:BlackRockasofOctober2024.Forillustrativepurposesonly.Viewsaresubjecttochange..

19.McKinsey.“theExecutivesAIplaybook.”Accessedon11/13/2024.20.SingularityHub,“AIModelsScaledUp10,000xArePossibleby2030,ReportSays,”August29,2024.Forward-lookingestimatesmaynotcometopass.21.Reuters“OpenAIandothersseeknewpathtosmarterAIascurrentmethodshitlimitations.November11,2024.22.YoleGroup“DatacenterGPUandAIASCIrevenuecouldreach$156billionby2025and233billionby2029.April12,2024.Forward-lookingestimatesmaynotcometopass.23.Gartner.“GartnerprojectsmajorITspendingincreasesfor2025.”O(jiān)ctober24,2024.Forward-lookingestimatesmaynotcometopass.9

iCRMH1124U/S-4024094-9/14

AI’sBuildPhaseaccelerates

TheGPUsattheheartofthesedatacentersare

gettingasignificantupgradein2025.Nvidia’s

BlackwellGB200GPUsareexpectedtoshipatscalein2025.Leadingtechfirmshavealreadyputin

orderswithforthissuperchip,whichis25timesmorepowerefficientand30timesfasterthanits

predecessorH100GPU.24Itisimportantto

underscorethatthisleapinbothspeedand

efficiencymayallowmorebusinessesbeyondjustmega-captechfirmstointegrateadvancedAI

capabilitiesinreal-timeapplicationsonanunprecedentedscale.

AImodelscouldevolvein2025,too.OpenAIhas

announcedtheupcomingGPT-5largelanguage

modeliscurrentlyintraining,whichcouldpush

multi-modelAIevenfurther25.GPT-5couldintroduceadvancedreasoning,improvedreliability,and

#ofParameters/Intelligence

autonomousAIagentscapableofmanagingreal-worldtaskswithouthumansupervision.These

featuresaimtomakeAImoreeffectiveincomplex,nuancedenvironments,fromdynamicproblem-

solvingtoseamlessprocessautomation.OpenAI

CEOSamAltmanhintedattheleapforward,saying,"GPT-4isthedumbestmodelanyofyouwillever

havetouseagainbyalot.“26

Modelsaren’tjustgettingsmarterandmorecomplex.Insomeinstances,theyaregettingsmaller.More

efficientSmallLanguageModels(SLMs)arecheaperandrequirelesstrainingdataandenergythanlargelanguagemodels(LLMs).SLMs,whichfocuson

specificdatasetsandusefewerparameters,canbe

betweenfiveand29timeslessexpensive.27These

modelsarerunningonthe"edge"—deviceslike

smartphonesorsensorsthatprocessdatalocally,

ratherthanrelyingondistantservers.ManySLMsarenowpoweringnewAIapplicationsdirectlyonthese

edgedevices,andthenumberisgrowingfast.

AImodelevolution

10T?100T

GPT-5

?

GPT-41T

GPT-3175B

GPT-2

GPT1.5B

120M

Transformer

201720202023Next5years2030s

Source:BlackRock,October21,2024.Forillustrativepurposesonly.Thereisnoguaranteethatanyforecastsmadewillcometopass.Viewsaresubjecttochange.

FrontierModels

ChatGPT

InvestorsinterestedintheAIthemecanlearnmoreaboutthefullAIvaluechainwiththeiSharesFutureAI&TechETF(

ARTY

),orwithanactivestrategylikethe

iSharesA.I.InnovationandTechActiveETF

(BAI

)whichseekstomaximizetotalreturn.

24.Nvidia,“GB200NVL72,”AccessedNovember6,2024.25.OpenAI,“OpenAIBoardFormsSafetyandSecurityCommittee,”May2024.26.Medium,”DumbestModel‘InOpenAI’sAIEvolution.”May2024.27.ATrade-offAnalysisofReplacingProprietaryLLMswithOpenSourceSLMsinProduction.arXiv,2023.

10

iCRMH1124U/S-4024094-10/14

AI’sBuildPhaseaccelerates

ImplicationsoftheAIrevolution

AI'seconomicpotentialhastwonotableconsequences:

1.theimportanceofthistechnologymaymean

inconsistentglobalregulationandprotectionistpolicies.

2.AIcouldmakeproprietarydatamorevaluable,requiringcommensurateinvestmentin

cybersecuritytoprotectwhatmaybecomecompanies'bestasset.

TheU.S.techsectorishighlyglobal,deriving60%ofitsrevenuesfromoverseas.28WithAIbecomingsuchanimportantpotentialeconomicgrowthengine,

variousjurisdictionsareweighingmoreprotectionistpoliciesthatbalancesupportingtheirnational

champions,withsecuringdataandintellectual

property,andensuringnationalsecurity.U.S.tech

firmsmayfindthemselvesparticularlyvulnerabletotheseregulations,creatingadivergencebetween

firmswithamoredomesticfocusonhiringandsalesversusthosethataremoregloballydependent.

InvestorsinterestedinthethemeofU.S.TechIndependencemayconsider

iSharesU.S.TechIndependenceFocusedETF

(IETC

).

InvestorsinterestedinthethemeofcybersecuritymayconsidertheiSharesCybersecurityand

Tech

ETF(IHAK

).

Cybersecurityisanotherareathatmaystandto

benefitfromAI’srise.AttheheartofAIisdata,whichcanbeusedtotrainlanguagemodelsordrive

insightsformachinelearningtools.Likethe

combustionengine’simpactonoil,theriseofAImaymakeproprietarydataanincreasinglyvaluable

resource.

Yethavingamorevaluableresourcerequiresmore

investmentinprotectingthatresource.Cyberattacksarerisingalarmingly,withsignificantglobalattacksexpectedtodoubleby2024comparedto2020.29Thenumberofattacksrose28%from4Q2023tothefirstthreemonthsof2024.30Thefinancialimpactisalsoescalating,withglobalcybercrimecostsprojectedtoreach$10.5trillionannuallyby2025,morethan

triplingfrom$3trillionin2015.31Inresponseto

risingthreats,cybersecurityinvestmentshave

surged.Globalspendingisprojectedtoreach$215billionin2024,a14.3%increasefrom2023.32AstheAImegaforceevolves,investmentsincybersecuritycouldgrowcommensurately.

“

AI’srisehasgeopolitical

implicationsasfragmentedregulationanddesireto

controlthetechnologymayleadtodivergencebetweenUS-focusedAInames

andthosewithmoreglobalexposure.”