2013年4月9日
亚洲主题：消费（第1部分）
下一个浪潮
确立下一轮消费题材投资观点
证券研究报告
亚洲消费：下一轮趋势如何？我们将消费确立为亚洲重大投资主题已有近10年的时间。
我们预计南亚地区下一轮需求浪潮将从必需品转向可选消费品，北亚地区将从商品转向服务。
金融服务和休闲相关行业是主要的受益板块。
明智选择消费板块：归类与简化对消费题材的投资是一个复杂的过程，因为其涉及板块较多且回报显著分化——在过去10年中最领先板块的回报是最落后板块的10倍。
然而，我们认为一个适当的框架可以简化整个过程。
我们界定了10个市场的19个消费板块，并根据它们的行业相关性和与宏观经济的关联将其划分为3类（宏观经济主导、行业主导和个股表现）。
确立投资观点：系统化的题材投资我们基于三个核心标准的筛选框架确定了7个最看好和7个最不看好的板块。
在我们看来，部分中国和印度的板块已再次显现买入良机；许多東盟国家的板块基本面稳健但估值充分。
收入增长、人口结构的变化和科技变革蕴含着四个投资主题：1)消费数字化、2)医疗保健、3)金融服务以及4)娱乐与旅游。
正确选择消费板块的重要性——在过去10年中最领先板块的回报是最落后板块的10倍 Returnsince2001(%)1000 800600400200 0(200) 鄧啟志+852-2978-0722richard.tang@高盛（亚洲）有限责任公司慕天辉,CFA+852-2978-1328timothy.moe@高盛（亚洲）有限责任公司KetakiGarg+91(80)6637-8601ketaki.garg@高盛（印度）证券私人有限公司刘劲津,CFA+852-2978-1224kinger.lau@高盛（亚洲）有限责任公司CaesarMaasry+852-2978-7213caesar.maasry@高盛（亚洲）有限责任公司SunilKoul+852-2978-0924sunil.koul@高盛（亚洲）有限责任公司 AutosPersonalcare Deptstoreso FoodproductsSupermarkets BreweryPCs/mobilesHealthCare HomecareInsurance ClothingTourismPropertysAirlinesEducationConselec 资料来源：FactSet、MSCI、高盛全球经济、商品和策略研究 高盛与其研究报告所分析的企业存在业务关系，并且继续寻求发展这些关系。
因此，投资者应当考虑到本公司可能存在可能影响本
报告客观性的利益冲突，不应视本报告为作出投资决策的唯一因素。
有关分析师的申明和其他重要信息，见信息披露附录，或参阅。
由非美国附属公司聘用的分析师不是美国FINRA的注册/合格研究分析师。
本报告仅供分发给高盛机构客户。
高盛集团 2013年4月9日 TableofContents Asianconsumption:TheNextWave–ExecutivesummaryThemacrocaseforAsianconsumptionTenyearson...howhasitprogressed?
Now,what’snext?
AroadmapforconsumptionpatternevolutionMappingtheterrain:Define,thencharacterizeAnassessmentframeworkforconsumptionsectorsTrendsthatgiverisetonewstructuralopportunitiesDigitalizationofconsumption:TheglobaltrendofgoingonlineHealthcare/wellness:TwindriversofhighereandlongerlivesFinancialservices:IncreasingdemandasAsiansewealthierTourismandentertainment:MorefocusonleisureLocalvs.foreignbrandsinAsiaIndustryconsolidationAppendixI:Top-downindustryanalysis:ParametersandsourcesAppendixII:DetailedstockdataMSCIdisclosureEquitybasketdisclosureDisclosureAppendix AllpricesinthereportareasofthemarketcloseofApril2,2013,unlessmentionedotherwise. 亚太 35581421293031323335363839414142 高盛全球经济、商品和策略研究
2 2013年4月9日 亚太 Asianconsumption:TheNextWave–Executivesummary ConsumptioninAsiaisapowerful,long-termthemewhichmanyinvestorsarefocusedon.Ourobjectiveinthisreportistoidentifyfreshopportunitiesintheconsumptionspaceandtohelpstructurethethoughtprocessforinvestinginthisimportantarea.Wehavefourmaininsights:
1.MacrotrendsConsumptionpatternschangeaspercapitaerises.Basedonourmacroforecasts,weexpectconsumptiondemandtoshiftfromnecessitiestodiscretionaryitems,andfromgoodstoservices.Inparticular: Clothingandapparel,restaurants,orautosmaynolongerbeanewstoryinurbanChina,butwillremainaleadingthemeinIndiaandASEAN, UrbanChinaandthemoreadvancedmarketsinAsiaarerapidlymovingtothenextstage,withtheirconsumerfocusshiftingfromgoodstoservices.
2.InvestingframeworkWeidentify19sectorsthataredirectlyrelatedtoconsumerdemandandnotethatperformancehasvariedwidely,whichunderscorestheimportanceofsectorandstockselection.Tomanagethearrayof10marketsx19sectors,wefocusontwoapproaches: Simplifyingthemonitoringprocess:Wecategorizethesectorsintothreebucketsofmacroplays,industrydriverfocusandstockpicks,basedontheirsectorcoherenceandmacroaffinity. Simplifyingtheselectionprocess:Wescoresectorsbasedonasetofuniformvariablesfocusingonthethreecorecriteria-downindustrylandscape,bottom-upfundamentals,andvaluations.Forstocks,wealsoincludecash-basedreturnmetrics,whichoursinglestockanalystshaveshowntohelpcapturealphaswiththesector.
3.SectorandstockideasWeranktherelativeattractivenessofponentsectorsoftheconsumptionthemeandidentifysevenmostandsevenleastpreferredsectors.Somehighlevelinvestmentconclusionsincludethefollowing: Valueisre-emergingforselectChinaandIndiasectors,especiallyChina,Chinaairlines,ChinainsuranceandIndiaautos. ASEANconsumptionsectorsarefairlyvaluedfortheirsolidfundamentals.SomesectorsinThailandandPhilippineslookrichlyvalued. AnumberofdomesticsectorsinTaiwanlooktohaveabination-downandbottom-upfundamentals,aswellashighvaluations.
4.FourthemesBesidesthecoredriverofhighere,twoadditionalforcesofdemographicsandtechnologypointtoinvestmentopportunitiesinfourthemes.Inanutshell, Increasinguseofandmobileleadstodigitalizationofconsumption; Ageingpopulationincreasesthedemandforhealthcare; Increasingwealthandlongerlifeexpectancyincreasesneedsforfinancialservices; Theincreasingfocusonleisurebringsdemandforentertainmentandtourism. 高盛全球经济、商品和策略研究
3 2013年4月9日 亚太 Exhibit1:Weseethenextwaveofconsumptiondemandshiftingfromnecessitiestodiscretionaries,andfromgoodstoservices ContributiontoPCEgrowth FoodClothingRestaurants AutosHousing ASEANandIndia Culture/EntertainmentFinancialsvcs Chinaandothermore"advanced"EM PercapitaeSource:GoldmanSachsGlobalECSResearch. Exhibit2:Sectorandstockideasbasedonourscreeningmethodology Market Sector Mostfavoredconsumptionsectors India Property China &media China Insurance India Autos China Airlines Indonesia Property Philippines Super-,
hypermarkets&conveniencestores Leastfavoredconsumptionsectors Taiwan Insurance Thailand Healthcare Singapore s Korea s Korea Super-,hypermarkets&conveniencestores Taiwan DepartmentStores Korea Foodproducts/plantations Note:SeeExhibit28forourscreeningdetails. Source:GoldmanSachsGlobalECSResearch. StockideasMostfavoredstockswithinthesector PhoenixMills,HDIL,OberoiRealtyTencent,VIPShop,SouFun,ChangyouPingAn,ChinaTaiping,NCIBajajAuto,TataMotors,HeroMotoCorpChinaEastern,AirChinaBumiSerpongDamaiPuregoldLeastfavoredstockswithinthesectorCathayFin.,ShinKongFin.BangkokDusitMedicalM1LGUplusE-martFarEasternDeptStoresLotteConfectionery,Nongshim Exhibit3:Weseethefollowingtrendsgivingrisetostructuralinvestmentopportunitiesinfourthemes DigitalizationofConsumption Healthcareandwellness Trendsthatgiverisetoinvestmentthemes Demographics e Technology Ageingpopulation Risinghealthawarenessaserises Increasinguseofandmobile Financialservices Longerlifeexpectancy Increasingwealth Entertainment Morefocusonleisureand andtourism qualityoflifease Note:AIA,CMB(H).KBFinancialandE.SunareonourregionalConvictionLisit. Source:GoldmanSachsGlobalECSResearch,GoldmanSachsResearch. ExamplesofbeneficiariesthatareBuy-rated GlobalLogisticProperties,Tencent,VIPShop Mindray,CMS,Shineway AIA,CMB,KBFinancial,
E.Sun GrandKoreaLeisure,Paradise 高盛全球经济、商品和策略研究
4 2013年4月9日 亚太 ThemacrocaseforAsianconsumption Tenyearson...howhasitprogressed?
ThepotentialofAsian/EMconsumerswasfirsthighlightedbyoureconomistsin2004aftertheintroductionoftheterm“BRICs”.Sincethen,ourglobalcolleagueshavewrittenawidearrayofresearchonthistheme.Theimportantpublicationsinclude: GlobalEconomicsPaperNo:118,TheBRICsandGlobalMarkets:Crude,CarsandCapital,Oct14,2004. TheWorldandtheBRICsDream,Feb14,2006. GlobalEconomicsPaperNo:208,TheBRICs10YearsOn:HalfwayThroughtheGreatTransformation,Dec7,2011. China:PortfolioStrategy:PickingtheChineseconsumptionwinners,Feb26,2010. China:PortfolioStrategy:Massmarketconsumption–China’sstructuralopportunity,May4,2012. China:Retail:ManyfacesoftheChineseconsumer,Feb8,2012. Summarizingthefindingsfromourpreviousresearch,wegenerallybelievedtherewerethreemajormacrotrendsprevailing. Risinges:WeexpectstrongeconomicgrowthintheemergingeconomiesonrapidumulationofcapitalandrisingproductivityandforecastepercapitainEMtoalmostdoubleby2030. Urbanization:Thecontinuousmoveofpeoplefromruraltourbanareaswillcreatedemandforhousing,transportation,municationandalsocapitalgoods. Demographics:Wewereforecastingfastgrowthintheworkingagepopulationinmanycountriesfromtheprogressionofvariousagecohorts(i.e.“demographicdividend”)andincreasinglaborparticipationrates.However,thishasrealizedquicklyoverthepastdecade,andsomeofthecountriesinAsiaarenowexpectingageingpopulationsinthenexttenyears. OurconclusionthenwasthatdemandforconsumergoodswouldrapidlyincreaseasAsianspursuehigherlivingstandards,andthegrowthdriverinAsiawouldshiftawayfrominvestmentsandexportstowardsconsumption. Almosttenyearshavepassed.Someoftheconsumptionthemeshaveturnedouttobegoodinvestments,whilesomehavenotplayedout.Atthesametime,therearealsosomenewopportunitiesemerging.Atthebeginningofthisreport,wefirstevaluatetheprogressofAsianconsumptioninthreeaspects: Macro:Themacroevidencehasbeenmixed.Overthepastfewyears,consumptionhasactuallyreduceditsshareofAsianGDP.However,othermacroindicatorssuchasretailsaleshaverecordedstronggrowth. Micro:AgoodexampleisChinaautosales,whichhavenowexceededthoseintheUS.Earningsinconsumersectors(inparticulardiscretionary)havegrownfasterthanthebroaderAsiaindex,andDMconsumersectorstoalargeextent. Investmentreturns:ReturnsfromAsianconsumersectorsalsooutperformedthebroaderregion,althoughtheytrailedUS.Aswehighlightiningsections,however,thesector-widereturnshavemaskedthewidedisparityinperformanceoftheunderlyingindustries. 高盛全球经济、商品和策略研究
5 2013年4月9日 亚太 Exhibit4:WeforecastepercapitainEMtoalmostdoubleby2030Percapitaeformajoreconomies(current,2020and2030) GDPpercapita(USD) 100,000 DM'spercapitaewillrise35% 80,000 inaggregateby2030 60,000 40,000 20,000
0 In18years,EM'sGDPpercapitawillstillbelow(partlyduetoalargepopulation), butitsgrowthratewillbeover3timesthatofDM's Current20202030 EM'spercapitaewillrise175%inaggregateby2030 SwitzerlandAustraliaUSA SingaporeHongKong FranceCanadaGermany IsraelUK ItalyTaiwanJapan SpainKoreaNewZealandRussiaTurkeyBrazilMalaysiaMexicoChinaSouthAfricaThailandIndonesiaPhilippinesIndia Exhibit5:Asiamayhaveitsmiddleclassexpandto44%ofthepopulationby2020 Middleclassas%ofpopulation(definedasesfrom6,000to30,000PPP$) 90%80%70%60%50%40%30%20%10% 0% EuropeLatAmAeJ World
N.AmericaJapan Exhibit6:UrbanizationisoneofthekeydriversforAsia’sconsumptiongrowth Urbanizationratio 90%80%70%60%50%40%30%20%
N.AmericaLatAmEuropeJapan WorldAeJ 198019851990199520002005201020152020198019851990199520002005201020152020 198019851990199520002005201020152020198019851990199520002005201020152020 Exhibit7:Thesignificantchangesindemographicsarehavinganimpactontheconsumptionpattern%ofpopulationinworkingage(left)andelderly(right) 70%65% AeJ LatAmWorldEuropeN.America 30%25%20%15% 60% Japan 10%5% 55% 0% Source:IMF,UnitedNations,GoldmanSachsGlobalECSResearchestimates. Japan EuropeN.America WorldLatAmAeJ 高盛全球经济、商品和策略研究
6 2013年4月9日 亚太 Exhibit8:ConsumptionasapercentageofGDPinAsiahasactuallydroppedinrecentyearsConsumptionas%ofGDP Consumptionas%ofGDP75%70%65%60%55%50%45%40%35%30% TaiwanIndiaKorea AeJ China Exhibit9:RetailsaleshasseenrapidriseinAsiaRetailsalesgrowthforAsiaandothereconomies Retailsales(%change)20%15%10% 5%0%-5%-10% AeJUS JapanEurope(EU-27) 199019952000200520102000200120022003200420052006200720082009201020112012 Exhibit10:China’sautosaleshavesurpassedthoseintheUS...CarsalesinChinaandUS Carsales('000units,monthly) 2,0001,8001,600 USChina 1,400 1,200 1,000 800 600 400 200
0 Exhibit11:...andweseeasimilarcaseputerproductsPCsalesinAsiaandUS PCSalesUnits(mnunits,quarterly)35 US 30 AsiaPacificexJapan 25 20 15 10
5 0 Jan-05Jul-05Jan-06Jul-06Jan-07Jul-07Jan-08Jul-08Jan-09Jul-09Jan-10Jul-10Jan-11Jul-11Jan-12Jul-12Jan-131Q951Q961Q971Q981Q991Q001Q011Q021Q031Q041Q051Q061Q071Q081Q091Q101Q111Q12 Exhibit12:EarningsgrowthinAsia’sconsumersectorshavebeenamongthestrongest EPSCAGR(5y)forglobalconsumersectors Exhibit13:ReturnsfromAsianconsumersectorshaveoutperformedthebroaderregion,buttheytrailedUS ReturnCAGR(5y)forglobalconsumersectors 16% 8% 5yEPSCAGR 14%6% 12% 4%10% 8% 2% 6%0% 4%-2% 2% 0% -4% MXAPJ MXAPJ MXAPJ S&
P SXXP consdiscconsstap consumerconsumer Source:CEIC,FactSet,Haver,IDC,Worldscope,GoldmanSachsGlobalECSResearch. MXAPJconsdisc MXAPJconsstap MXAPJ 5yreturnCAGR S&Pconsumer SXXPconsumer 高盛全球经济、商品和策略研究
7 2013年4月9日 亚太 Now,what’snext?
Aroadmapforconsumptionpatternevolution Asaneconomydevelops,consumerstendtochangetheirconsumptionpatterns–fromnecessitiestoluxuries,fromgoodstoservices,etc.Oneexampleisthegradualdropinproportionofespentonfoodthroughtime.Generallyspeaking,economiesthatarefurtheralongtheirdevelopmentpathstendtohavefoodexpendituretakingalowerportionofthetotal.WecanclassifyAsianmarketsintothreegroupsbasedonthis. Developedmarkets(DM):Regardlessofthestartingpoints,mostdevelopedmarketshave10%-15%oftheirtotalconsumptionexpendituresspentonfood.InAsia,theseincludeHongKong,Korea,SingaporeandTaiwan. Advancedemergingmarkets(EM1):Malaysia,China(urban)andThailandgenerallyhavetheirfoodexpenditurestaking25%-35%oftotal. Emergingmarkets(EM2):ThefoodexpendituresofhouseholdsinChina(rural),India,PhilippinesandIndonesiaarestillclosetohalfoftotalspending. Inourview,thisisbroadlyconsistentwiththestagesofeconomicdevelopmenteachmarketisin.TopredictwherethenextsetofdriversforAsianconsumptiongrowthcouldbe,westudythetypicalevolutionpatternusingUSasanexample,andparetheconsumptionbasketbetweeneachAsianmarketandthatofatypicalDMtoidentify“thegap”. BasedontheexperienceintheUS,webelieveatypicalconsumptionpatternevolutioninvolvesthreestages: oStage1:Atthebeginning,ahousehold’seisheavilyspentonnecessitieswhicharefoodandclothing.Somepaymoreforfoodservices. oStage2:Consumersthengraduallyputmorefocusonthequalityoftheirmodationandalsobegintopurchasetheirowncars.Theproportionofhousingandtransportexpenditurethusincreases. oStage3:Consumptionmovesawayfromgoodstoservices.Peoplespendmoreonleisureandentertainment.Theirdemandforfinancialservicesalsorisesaswealthumulates. oThecontinuousincreaseinspendingonhealthcaremunicationhaslikelybeendrivenbydemographicsandtechnologicaladvancement. OurgapanalysissuggestsmostemergingmarketsinAsiaarestillmeaningfullybehindDMsontheirspendingondiscretionaryitemsandservices,althoughtheydifferinmagnitude.Inourview,theultimatedriverofallthesechangesishouseholde(seeBox#1).Lookingforward,aseconomicdevelopmentprogresses,wethinkthenextstagesofdevelopmentforvariousmarketsandregionswillbeasfollows:
(1)Spendingonculture/entertainmentiscurrentlylowacrosstheboard.
(2)FinancialservicescouldbethenextwaveintheadvancedEMs.Chinahasbeenrapidlydevelopingthisarea,andcouldbearoadmapfortheASEANmarkets. (3)saresaturatedinmostmarketsexceptIndiaandThailand,butthetechnologyadvancement(fromfeaturephonestosmartphones)couldbringaboutaproductupgradestory.
(4)Healthcareisastructuraltrendfromageing,andthisishappeningglobally. Withabroadsenseofwherethenextwaveofconsumptionwillgo,wehaveabetterideaofwhattolookforaswedrilldowntotheopportunitiesofspecificindustrieswithmoregranularanalysesiningsections. 高盛全球经济、商品和策略研究
8 2013年4月9日 亚太 Exhibit14:AsianmarketsbroadlyfallintothreegroupsintermsoftheproportionofesspentonfoodConsumptionpatternevolution–foodas%ofpersonalconsumptionexpenditure Foodas%ofPCE70% ID 60% IN CN-ruralPH 50% 40% 30%US 20% KR HKUK AU GR CN-urban THTWMY JPSG 10% EM2EM1DM 0%1929 1939 1949 1959 Source:CEIC,Haver,GoldmanSachsGlobalECSResearch. 1969 1979 1989 1999 2009 高盛全球经济、商品和策略研究
9 19441954196419741984199420041944195419641974198419942004 2013年4月9日 Exhibit15:Thethreestagesofconsumptionpatternevolution:TakingUSasanexampleContributiontoPCEgrowthinUS ContributiontoPCEgrowth(10yaverage) 30% Restaurants Clothing Food&Beverages 25% 12%10% ContributiontoPCEgrowth(10yaverage) 30% Housing&utilities Transport25% 20% 8% 20% 15% 6% 15% 10% 4% 10% 5% 2% 5% 0% 0% 0% 1944195419641974198419942004 ContributiontoPCEgrowth(10yaverage) 12% Culture/Entertainment Financialservices10% 8% 6% 4% 2% 0% Source:Haver,GoldmanSachsGlobalECSResearch. ContributiontoPCEgrowth FoodClothingRestaurants AutosHousing Time 亚太 12%10%8%6%4%2%0% Culture/EntertainmentFinancialsvcs 高盛全球经济、商品和策略研究 10 2013年4月9日 亚太 Exhibit16:Asiavs.theDMbenchmark:MappingthestructuralopportunitiesBreakdownofpersonalconsumptionexpenditurebycategory ThespendingpatternofanaverageDMconsumer US Food&Beverages 8% Clothing 4% Restaurants 7% Housing&utilities 19% Furniture/Householdfacility4% Transport 10% UK 13%6% 10% Germany 14%5%6% Japan 16%3%7% Australia 14%3%7% DMavg 13%4%7% 24% 25% 25% 23% 23% 5% 6% 4% 5% 5% 14% 13% 11% 10% 12% Consumptionoftenfollowsatypical patternasamarketdevelops Culture/EntertainmentFinancialservices HealthcareCommunicationsEducation Other 9% 11% 9% 10% 10% 10% 8% 5% 7% 4% 9% 6% 20% 2% 5% 4% 6% 8% 2% 2% 3% 3% 2% 3% 3% 2% 1% 2% 4% 2% 7% 6% 6% 10% 7% 7% elerated
byageing,technology, etc. HowdoesthespendingpatternofAsianparetothatinDMs?
Food&BeveragesClothingRestaurantsHousing&utilitiesFurniture/HouseholdfacilityTransport TransportservicesAutosCulture/EntertainmentFinancialservicesLuxurygoodsHealthcareEducationCommunications DMavg13%4%7%23%5%12%2%10%10%5%1%8%2%3% Others 8% Sing9%3%9%19%6%13%3%10%14%5%2%7%3%2% 9% Korea16%5%8%16%3%12%3%9%8%7%0%7%7%4% 6% Taiwan15%5%7%18%5%11%2%10%9%3%1%10%4%4% 8% HK15% 6%10%19% 9%10%7%4% 4%7%2%4%2%2% 8% Malay24%3%10%16%5%13%5%8%5%5%1%2%2%8% 7% China-U27%11%8%10%7%9%3%7%3%2%0%7%9%6% 2% India34% 7%2%13%4%15%8%7%1%6%2%4%1%2% 7% Thai36% 7%9%7%6%15%5%10%5%1%0%6%1%1% 5% China-R41%6%NA19%5%11%NANANANANA7%8%NA 2% Phils43% 1%4%13%4%11%7%4%2%3%1%2%4%3% 8% Source:CEIC,Haver,GoldmanSachsGlobalECSResearch. Indo47%2%2%16%2%12%5%7%1%2%0%3%4%5% 4% 高盛全球经济、商品和策略研究 11 2013年4月9日 亚太 Box#1:Forecastingthepathofconsumptionpatternbypercapitae ForecastingchangeinfoodconsumptionwitheTherelationshipbetweenfoodconsumptionandhouseholdehasbeenwellresearchedineconomicliterature.ThemostwellknownruleisknownasEngel’slaw,whichsaysasacountry’serises,theproportionofespentonfoodtendstofall. Ourobjectiveistwofold.First,wewanttoverifyifmost(orall)majorcountriesinthisstudyfollowthispatternofconsumerbehavior.Second,basedonthiseconomictheory,areweabletoforecasthowthepatternwouldchangeifwehaveasenseofhowquicklyanAsianmarketgrowsitselevel?
Toachievethis,webuildapanelregressionmodeloffoodexpenditureasapercentageofpersonalconsumptionexpenditure(PCE)onpercapitae. esinastherelationshipofpercapitaeandfoodasapercentageofPCEisdefinitelynotlinear.TheEngelcurvefeaturessaturationpropertiesinthatthepaceofdecreaseinproportionoffoodspendingslowsathighelevels.Inotherwords,thereisalimitonhowmuchonecancuttheirfoodexpenditureportionaseincreases.Toincorporatethis,thesimplestwayistoassumeitfollowsalogfunction.Wedefineitasbelow: FoodAs%PCEab*ln(e) WepooldataonfoodasapercentageofPCEandpercapitaeacross15markets.Thelengthoftimeseriesspansfromthelongestof50yearsintheUStotheshortestof11yearsinMalaysia.Werunthedatathroughapanelregression,whichhelpsountforstructuraldifferencesacrossmarkets,andcanmodatedifferentdataavailability. Wefitafixedeffectsmodeltothepaneldata(thenullhypothesisofindividualeffectsbeinguncorrelatedwithotherregressorsisrejectedat1%and5%levelsofsignificancebytheHausmantest).TheresultsshowthatthenaturallogofpercapitaeisasignificantpredictoroffoodexpenditureasapercentageofPCEat1%and5%levelsofsignificanceanditsnegativecoefficientshowsthataspeopleericher,foodexpenditureasashareofPCEdeclinessteadilyuntilitreachesasaturationlevel.TheR-squaredis97%. Withthemodelinplace,weestimatehowfoodspendingasapercentageofPCEwillevolvebyusingIMF’sforecastsofeconomicgrowthandpercapitaelevelsining20years.Thespeedofreductionintheportionofeonfoodvariesmeaningfullyacrossmarkets.Ourmodelsuggeststhatnoneofthe“EM2”marketswillreachtheDMrange(8%-16%)evenafter20years.Withinthe“EM1”markets,MalaysiamayreachlevelssimilartothoseinDMsin10-15years,andaround15-20yearsforChina. Obviouslytheresultsaresolelydependentontheforecastsofeconomicgrowth,whichmayormaynotberealized,andtherearedefinitelyotherfactorsaffectingtheevolutionofconsumptionpattern.Nevertheless,byunderstandingwhatstagesdifferentmarketsareroughlyin,wehaveabettersenseofwhatconsumptionthemes(foodoressories,necessitiesorluxuries,goodsorservices)willbethenextwavesforeachofthem. Engelcurve:AlogfunctionfitthedataneatlyFoodas%ofPCEvs.percapitaeforthe15markets MalaysiaandChinamaybethefirsttoreach“DM”levelsofconsumptionForecastsoffoodas%ofPCEintothenext20years Foodas%ofPCE 70%60%50%40%30%20%10% 0%
0 10,000 20,000 30,000 Percapitae(US$) 40,000 50,000 Foodas%ofPCE70% 60% IN 50% IDCN-ruralPH KR EM2 40% 30% US 20% 10% HKUK AUGR THTW JP SG CN-urbanEM1 MY DM 0%19291939194919591969197919891999200920192029 资料来源：CEIC,Euromonitor,Haver,IMF,GoldmanSachsGlobalECSResearchestimates. 高盛全球经济、商品和策略研究 12 2013年4月9日 亚太 Doesonefactorexplainall?
Eventhoughpercapitaeisanimportant(andprobablyadominant)factordrivingthechangeinconsumptionpattern,itisnottheonlyonethatmatters.Wedonotintendtoforecasteveryponentofpersonalconsumptionexpenditurehere,butwethinkitwillbeusefultoponentsthataremoreedrivenfromthosethatarelessdirectlyrelated,orinvolvingotherpotentialdrivers.Toachievethis,wesimplyrunthesamepanelregressionagainfortheponentsparetheirpooledR-squared. Besidesfood,financialservicesandculture/entertainmentcanalsobewellexplainedbythechangeinpercapitae.Asindividualsewealthier,theytendtospendalowerportiononfood,startfocusingoninvestmentsandfinancialservices,andspendmoreonleisure. Statistically,educationalsohasastrongrelationshipwithe,butwethinkitisatwo-wayrelationshipratherthanjustaresultofit. Therelationshipisstilldecentlystrongwithponents,butsignificantlylowerinstrength.Wethinkitsuggeststherearepotentiallysomeotherfactorswhichalsoinfluencethechanges.Forexample,eventhoughweagreethathealthcarespendingtendstoincreasewithebecausepeoplefocusmoreontheirhealthastheyericher,itmayalsobearesultoflivingalongerlife.Demandmunicationshasalsobeendrivenbytechnologicaladvancementandproductlaunches(e.g.andsmartphones).Housingandtransportareoftentheby-productsofurbanizationwhichhasbeenhappeninggloballyoverthepastfewdecades. Engelcurve:alogfunctionfitsthedataneatlyFoodas%ofPCEvs.percapitaeforthe15markets PooledR-sq.frompanelregression1.00.90.80.70.60.5 PercapitaseemstobewellrelatedtoentertainmentexpenseForecastsoffoodas%ofPCEintothenext20years 18%16%14%12%10% 8%6%4%2%0%-2%
0 10,000 20,000 30,000 40,000 Percapitae(US$) 50,000 FoodFinancialsvs. EducationEntertainment OtherHealthCare ClothingFurnitureCommunicationsHousingTransport Entertainmentas%ofPCE Thefitgetsloosermunications...Foodas%ofPCEvs.percapitaeforthe15markets ...andhealthcareexpensesForecastsoffoodas%ofPCEintothenext20years Communicationsas%ofPCEHealthCareas%ofPCE 9%8%7%6%5%4%3%2%1%0%
0 10,000 20,000 30,000 Percapitae(US$) 40,000 50,000 25%20%15%10% 5%0%
0 10,000 20,000 30,000 Percapitae(US$) 40,000 50,000 资料来源：CEIC,Euromonitor,IMF,GoldmanSachsGlobalECSResearchestimates. 高盛全球经济、商品和策略研究 13 2013年4月9日 亚太 Mappingtheterrain:Define,thencharacterize Takingabroaddefinitionofconsumption;weidentify19sectors Consumptionisabroadconcept,asultimatelythedemandformostproductsandservicesislinkeddirectlyorindirectlytoconsumptiondemand.Wethinkitisimportanttodefinetheconsumptionuniverseforthisreportclearlybeforewemoveontothespecificindustryfundamentals. Weidentify19industriesasouruniverse,whicharemoredirectlyrelatedtoconsumerdemandasopposedtoindustrialactivities.EnergyandMaterialsareexcludedforthisreason,andonlyairlinesareincludedinIndustrials.Banks’earningscyclesaredrivenbybothconsumersandcorporatesowedonotincludethemeither.Amongallthepanies,weonlyselectahandfulthatselldirectlytoconsumers–exampleswouldincludeSamsungElectronicsandHTC.Wealsoexcludeutilitiesforitsheavilyregulatednature. TheyformawidedistributioninmarketcapandliquidityThe19industriesdifferwidelyintermsoftheircapandliquidityprofiles.Thelargersectorsareautos,propertyands,buttheremaining60%isalongtailinthedistribution(Exhibit18). MSCIconstituentsrepresentcloseto70%ofthefulluniversebymarketcap,buttheyonlycoverhalfofitintermsofturnover,asthesmallerconsumptionstockstradeproportionatelymore(Exhibit19). Performanceisalsodifferent:Thebesthasoutperformedtheworstby10timesEventhoughmostofthesesectorsareeventuallydrivenbyconsumerdemand,thehomogeneityhasnotappliedtoequityreturns.Disparityinperformancehasbeenmeaningful,withthebestsectorhavingoutperformedtheworstby10timessinceend-2000. Autos,thebestperformer,made915%pricereturn.Theoutperformancewasmostsignificantduring2009and2010,whereearningsgrowthwasaround100%ineachofthetwoyears.Therehasnotbeenmeaningfulre-ratingorde-ratingoverthepastdecade,despitelargeswingsaroundthelong-termaverageP/Eofroughly9X,suggestingthatperformancehaslargelybeenearningsdriven. Thesecondbestispersonalcare,whichhasseldomoutperformedotherssignificantly,butithasbeengoingupsteadilyandconsistentlythusmakingagoodoverallreturn.Thesectorhasre-ratedconsiderably,withforwardP/Erisingfrom5Xin2001to20Xrecently.Earningsgrowthhasbeensteady,ataround10%-20%formostoftheyears. Theworstsectorsarehouseholdappliances/consumerelectronicsandeducationservices,followedbyairlines.Allthreesectorshavehadverylittleearningsgrowthoverthepastdecade.Therehasnotbeenmeaningfulre-ratingorderatinginmultipleseither. Therehasbeenincreasingdivergenceininvestmentreturnsinrecentyears,whichhighlightstheimportanceofbeingmoreselectivewheninvestinginconsumptionthemes.Inthenextsection,weintroduceanassessmentframeworktodistinguishgoodsectorsfrombadones. 高盛全球经济、商品和策略研究 14 2013年4月9日 亚太 Exhibit17:Wedefine19consumptionsectorsPercentagefiguresrepresentsector/industryweightintheconsumptionuniverse,byfloatmarketcap ChinaHongKong IndiaIndonesia KoreaMalaysiaPhilippinesSingaporeTaiwanThailand AeJ GICSsectors GICSindustries/sub-industries Airlines Consumerdiscretionary Airlines 0.20.10.00.00.10.10.00.30.10.0
1 Autos Consumerdiscretionary AutoComponents 0.20.00.40.11.80.20.00.00.60.1
3 Consumerdiscretionary Automobiles 1.10.00.81.12.80.10.00.30.20.0
6 Householdappliances,consumerelectronics Consumerdiscretionary HouseholdDurables 0.30.30.10.00.70.00.00.00.10.0
2 Consumerdiscretionary LeisureEquipment&Products 0.10.10.00.00.00.00.00.00.30.0
0 Clothing/Footwear/Sportswear Consumerdiscretionary TextilesApparel&LuxuryGoods 1.00.20.30.00.30.00.00.00.60.0
2 Consumerdiscretionary SpecialtyRetail 0.70.70.00.10.10.00.00.10.20.1
2 Departmentstores Consumerdiscretionary MultilineRetail 0.30.10.10.20.60.10.40.00.10.1
2 Hotels,tourism,gaming Consumerdiscretionary HotelsRestaurants&Leisure 0.22.90.10.10.40.80.10.70.10.2
6 Educationservices Consumerdiscretionary DiversifiedConsumerServices 0.00.00.00.00.00.00.00.00.00.0
0 Supermarkets,Hypermarkets,conveniencestores Consumerstaples Food&StaplesRetailing 0.60.30.00.10.30.00.10.20.20.7
2 Beers Consumerstaples Beverages 0.20.00.30.00.10.20.10.00.00.2
1 Foodproducts/plantations Consumerstaples FoodProducts 1.90.10.41.00.91.20.30.90.70.4
8 o Consumerstaples o 0.00.00.80.30.50.20.00.00.00.0
2 Homecare Consumerstaples HouseholdProducts 0.10.00.50.20.30.00.00.00.00.0
1 Personalcare Consumerstaples PersonalProducts 0.50.00.40.00.40.00.00.00.00.0
1 Healthcare Healthcare Healthcare 1.20.10.00.21.20.20.00.10.50.3
4 Insurance Financials Insurance 2.53.40.00.01.30.10.00.10.90.2
9 Property Financials RealEstate 4.19.80.40.80.00.70.83.50.60.9 22 &media Consumerdiscretionary Media 0.10.30.40.40.50.10.10.40.00.3
2 Consumerdiscretionary &CatalogRetail 0.10.00.00.00.20.00.00.00.00.0
0 Informationtechnology Software&Services 4.20.00.00.00.90.00.00.00.00.0
5 Informationtechnology HomeEntertainmentSoftware 0.00.00.00.00.30.00.00.00.00.0
0 PCs/mobiles Informationtechnology TechnologyHardware&Equipment0.50.00.00.02.40.00.00.01.00.0
4 s municationservicesmunicationservices 5.20.20.41.00.31.10.31.71.51.1 13 Total 25195 6165
2 8
8 5 100 Source:FactSet,GICS,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 15 2013年4月9日 Exhibit18:AsiaconsumptionsectorsdifferwidelyintermsofcapandliquidityprofilesFreefloatcapandaverageturnoverofeachofthecountrysectors Freefloatcap(US$bn)152 1009080706050403020100 ADVT(US$mn)921 500450400350300250200 150100 500 亚太 Exhibit19:MSCIrepresentsonlyaportionofthefulluniverseBreakingdownourconsumptionuniversetoMSCIandnon-MSCI,byfloatmarketcapandaverageturnover FloatMarketCap AverageTradingVolume NonMSCI31% MSCI69% NonMSCI49% MSCI51% Source:FactSet,MSCI,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 16 2013年4月9日 亚太 Exhibit20:Thebestsectorhasoutperformedtheworstby10timesoverthepastdecadePerformancetimeseriesandannualreturnfigures Priceperformance(Dec2000=100) 1,200 Brewery 1,000 Clothing/Footwear/SportswearDepartmentstores 800 Foodproducts/plantations Supermarkets,Hypermarkets, 600 conveniencestores o 400 Priceperformance(Dec2000=100) 1,200 Airlines 1,000800 Autos Householdappliances,consumerelectronicsHomecare Personalcare 600Property 400 Priceperformance(Dec2000=100)1,200 Hotels,tourism,gaming 1,000 EducationservicesHealthCare 800 Insurance &media 600 PCs/mobiles s 400 200 200 200
0 0
0 Dec-11 Dec-10 Dec-09 Dec-07 2011 Dec-08 Dec-06 Dec-05 Dec-04 Dec-03 Dec-02 2007 Dec-01 Dec-00 2002
Dec-10Dec-11 2003 Dec-09 Dec-08 Dec-07 Dec-06 Dec-05 Dec-04 Dec-03 Dec-02 Dec-01 Dec-00 Dec-00
Dec-01Dec-02Dec-03Dec-04Dec-05Dec-06Dec-07Dec-08Dec-09Dec-10Dec-11 Performance(%) AutosPersonalcareDepartmentstoresoFoodproducts/plantationsSupermarkets,Hypermarkets,conveniencestoresBreweryPCs/mobilesHealthCareHomecareInsuranceClothing/Footwear/SportswearHotels,tourism,gaming&mediaPropertysAirlinesEducationservicesHouseholdappliances,consumerelectronics Max-MinStd.dev./AbsoluteMean 2001 356 45(11) 5(5)295625112515(13)(10)(15)(26)(30)16(33) 883.9 2002 18(11)
(7)(8)2(9)4(22)(20)(15)(24)19
(6)(19)(26)(20)6(37)(12) 561.5 2003 8047613737371144381838422242452225 927 720.5 2004 216 531122312 55(27)141927(4)22135
(7)(3) 571.5 2005 3950563326272159533915 71922 811 36045 570.6 2006 (2)382019392934
(4) 9308932321845442579(17) 2007 15324226493634 725 74434253143524499
3 2008 (52)(24)(51)
(9)(44)(42)(45)(37)(37) 5(45)(61)(56)(48)(58)(39)(59)(53)(60) 2009 19681651183726574712759916895783602493 2010 40182526163830361529(1)1742 77743(17)
(9) 2011 4
(2)(16)25
(7)(5)(16)(19)(14)19(21)(22)(13)(12)(26)5(35)(37)(40) 2012 923131812193517252419 919224214 1(17)(30) 10yearperf93875948145243241039131630830028722317711910658125(72) 105 96 66 193 61 64 72 1011 0.9 0.6 0.4 0.6 0.9 1.4 1.1 0.8 Returndispersion(Std.dev./Absolutemean)2.0 1.6 Increasingdivergenceinsectorreturns 1.2 0.8 0.4 0.0 2012 2010 2009 2008 2006 2005 2004 2001 Note:Thechain-linkedperformanceofHouseholdappliances,consumerelectronicswassignificantlydraggeddownbyanumberofKoreastocksintheearly2000s.Source:FactSet,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 17 2013年4月9日 亚太 Tradingconsumptionsectorsintelligently:Classifyandsimplify Investinginconsumptioncanbeplicateddecisionasitinvolvesalargenumberofdifferentsectors.However,withtheappropriateframeworkwethinktheprocesscanbesimplified.Herewepresentourmethodologytoclassifythesectorsintothreebroadgroupswithdistinctcharacteristics(macroplays,industrydriverfocus,andstockpicks)byanalyzingthemalongtwoaxes:sectorcoherenceandmacroaffinity.RefertoAsiaPacific:PortfolioStrategy:AsiaSectorRadar:Mappingtheterrain,June1,2009andAsiaPacific:PortfolioStrategy:SectorstrategymattersforPMs&stock-pickers,January8,2013forfurtherresearch.
1.Sectorcoherence:Dotheytradeasasectororacollectionofindividualstocks?
Coherenceisanimportantattribute,asitindicateshowimportantitistofocusonidiosyncraticriskswhentradingsectors.Ifasectoris“highlycoherent”,itsuggeststhatmonfactorssharedamongstocks(whichmaybemacroorindustry-specific)dominateindividualrisks,thusthefundamentalanalysescanbeefficientlysimplified.Ontheotherhand,somesectorsaremerelyagroupofstockstradingseparately.Inthiscase,muchmoretimehastobeinvestedinpanyspecifics.
2.Macroaffinity:Dotheytrademoremacroormicro?
“Tradingmacroortradingmicro”pointsustofocusingontherightstrategies.Tradinga“macrosector”isessentiallyexpressingamacroview,sotheimplementationisrelativelystraightforward.“Microsector”ofteninvolvesitsownindustrydrivers,orspecificsupply-demanddynamicsthatbasicmacroindicatorswouldnotfullycapture. AlongthetwoaxeswecanclassifythemintosomebroadcategoriesInourview,thesebucketsdiffervastlyintermsoftradingbehavior.Foraninvestmentmanager,wethinkitmaybebeneficialtorecognizethedifferencesasithelpstooptimizetheallocationoftimeandresourcesamongdifferentsectors. Highcoherence,highmacroaffinity–the“macroplays”:Asthesectorstradecloselywithmacro,andwithinthesectorstockstradecloselywitheachother,investmentdecisionstendtobemoreaboutchoosingbetweenhighandlowbetanames.Alotofautos,insuranceandpropertysectorsfallintothiscategory.Interestinglyenough,mostofthedomesticsectorsinTaiwanexhibitbothhighmacroaffinityandcoherence.WesuspectapossiblereasonisthathistoricallymanagershaveinvestedmoretimeinthestockandsectorspecificsinTaiwantechnames,buthavepaidlessattentiontothenon-techsidethustendtotradethemaltogetherasanexpressionoftheirviewsondomesticconsumersentiment. Highcoherence,lowmacroaffinity–moreindustrydriverfocus:Tradingthesesectorsmeansonehastomonitorafewmoreparameters,whichmayrangefromsamestoresalesgrowthtoARPUtrend.Thegoodnewsisthatstocksinthesesectorsstilltendtotradecloselyenoughwitheachothersopickingtherightsectoralreadytakesushalfwaytooutperformingthebenchmark. Lowcoherence,lowmacroaffinity–it’saboutstockpicking:Inourview,managersneedtoallocatethemosttimeinthese(so-called)sectorsasreturndriversdifferfromstocktostock.Somemayconsistpaniessellingnicheproducts,thusonehastoputmoreeffortintomanagingtheidiosyncraticrisks. Consistentwithourintuition,wefindveryfewsectorsshowinglowcoherencebuthighmacroaffinity. 高盛全球经济、商品和策略研究 18 2013年4月9日 亚太 Exhibit21:WeanalyzesectortradingbehavioralongthetwoaxesofsectorcoherenceandmacroaffinityRankingsofsectorcoherence(left)andmacroaffinity(right)forlargeconsumptionsectors Averagestockcorrelationwithsector Averagecorrelationwithmacroindicators CNInsuranceTWInsurance SGPropertyTWClothing TWAutosMYs TWPropertyINAutos TWFoodproductsCNProperty CNPCs/mobilesTHs KRDeptstoresKRInsuranceSGGamingHKPropertyCNAutosKRAutosMYProperty MYFoodproductsIDProperty CNHypermarketsCNClothingCN IDFoodproductsKRHealthCareIDsCNs HKHotels/gamingINFoodproductsMYHotels/gamingCNFoodproductsSGFoodproducts PHPropertyIDAutos HKClothingKRFoodproductsCNPersonalcare SGsTWsKR&mediaKRElectronicsCNHealthCare INAutosCNPropertyKRInsuranceHKPropertyHKHotels/gaming CNAutosTWProperty TWAutosTWClothingHKClothingCNClothingMYPropertySGPropertyCNPersonalcare KRAutosINFoodproducts PHPropertyTWInsuranceKRElectronicsCNInsuranceSGFoodproductsCNPCs/mobiles SGGamingTWsKRFoodproductsMYFoodproducts KRIDAutos SGsCN CNHypermarketsMYHotels/gaming KRHealthCareKRDeptstoresIDFoodproducts CNsMYsTWFoodproductsCNFoodproductsCNHealthCare IDsTHs IDProperty 0% 20% 40% 60% 80% 0% 10% 20% 30% 40% Noteonmethodology: Weassesssectorcoherencebycalculatingtheaveragestockcorrelationwithsectorreturns(sinceend-2000).Formacroaffinity,weestimateusingthehighestcorrelationamongthatofsectorreturnswithchangesinIP,PMI,retailsales,CPI,exportsandourFinancialConditionsIndex(FCI). Source:CEIC,Haver,FactSet,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 19 2013年4月9日 亚太 Exhibit22:Recognizingthedifferencesintradingbehavioramongsectorshelpstooptimizetheallocationoftimeandresourcesintheinvestmentprocess Classifyingmajorconsumptionsectorsintothreebuckets 40%35% Tradingmacro INAutos MacroPlays Correlationwithmacroindicators 30%25% HKHotels/gamingKRInsuranceCNProperty HKPropertyTWProperty CNAutos TWAutos TWClothing SGProperty 20% CNPCs/mobiles SGGaming TWInsurance 15%indivTirdaudailnsgtocks KR KRDeptstoresMYs CNInsurance Tradingasasector 10% 5% 0%30% CNHealthCareCNFoodproducts IDsIDProperty Stockpicking Tradingmicro TWFoodproducts THs Industrydriverfocus 40% 50% 60% Stockcorrelationwithsector 70% 80% Source:FactSet,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 20 2013年4月9日 亚太 Anassessmentframeworkforconsumptionsectors Screeningforopportunities:theframework Theobjectiveofthisassessmentframeworkistoscreenforinvestmentopportunitiesefficientlyinabroadarrayofcountrysectorsrelatedtoconsumption(10countriesx19sectors),usingsomeuniformvariables.Ourmethodologyfocusesonthreecriteria:-downanalysisoftheindustrylandscape,bottom-upanalysisfocusingongrowthandmargins,aswellasvaluations.Essentiallyweareaskingourselvesthreequestionswhenwegothroughthescreeningprocess:
1.Doesthissectorhaveagoodstory(orabadone)?
2.Isitjustastory,oralsobackedupbycorporateprofits?
3.Hasthemarketpricedinallthesealready?
Top-downindustryanalysisToevaluatetheindustrylandscape,weidentifythreekeyfactors: ration:Alotoftheearly-stagestructuralstoriesarebasedonthepotentialofincreasingration. Fragmentation:Ahighlyfragmentedindustrystructurewilllikelyreducepricingpower,anddampenmargins.Also,inpetitiveindustry,paniesneedtoinvestmuchmoreinbrandbuildingtodifferentiatethemselves.ThiscanrequirealotofR&Dandmarketing,inourview. ASP:Priceorproductupgradecouldbeausefulstrategytodrivefurtherearningsgrowthinanindustrywhichisalreadywellrated.WegenerallyviewlowASPrelativetopeersaspositiveasitindicatespotentialtoraisepricing.Insomecases,however,lowsellingpricescanbetheresultofhighfragmentation,whichsuggeststheindustryneedstogothroughconsolidationbeforeanASPupgradeispossible. Wescorethecountrysectorsbasedontheabovecriteria.Mostofthemcanbebroadlyclassifiedintofourgroups–structuraltrendsfromincreasingration,marketsharestory,price/productupgradestoryandmature,stablemarkets. Wemakesomehigh-levelobservations: IndiaandIndonesiahavelowrationinmanydiscretionaryconsumeritems.WefindsimilartrendsinthePhilippinestoo,buthigherelevelisacriticalelementtodriverationupwards. spendingandhealthcarearethefewsectorswhichwouldlikelybeastructuralstoryacrossmostmarketsinAsia.Wewillstudyeachofthemindetailinaseparatereport. QuiteafewconsumersectorsinChina,inparticularretail,arefragmented.Lackofpricingpowercreatesseriousmarginpressureforindividualplayers,eveniftheunitvolumeisgrowingrapidlyinsomecases.Industryconsolidationislikelyanemergingthemegoingforward. “Tradingup”isthenextearningsdriverforanumberofsectors.AgoodexampleistheessfulproductpremiumizationbyAmorePacificandLGH&HintheKoreacosmeticsindustry. ConsumersectorsinHongKong,Korea,SingaporeandTaiwanaremostlymatureandstablemarkets.Theyhaveestandardcyclicalplays,withonlyafewexceptions. 高盛全球经济、商品和策略研究 21 2013年4月9日 亚太 Bottom-upanalysis:GrowthandmarginsHavinga-downstorygivesinvestorsconfidenceinthelong-termprospectsofanindustry,butthismayormaynotbereflectedpanyprofits.Someoftheminvolveextensivecapexupfront,whichmeanswemayseelimitedpositiveimpacttonear-termearnings.Thisunderpinstheimportanceofhavingbottom-upearningsanalysisaswell.ThetwovariablesweusearetheearningsCAGRandthemarginsinthenextthreeyears(2013-15).Wefindthatanumberofpropertysectorshavebinationofhighgrowthandhighmargins,togetherwithChina.Onthelow-growth-low-marginside,theseincludeChinaautosandTaiwaninsurance. What’spriced:ValuationsTherationaleforvaluationsisprettystraightforward.BesidesP/EandP/B,wealsoincludedividendyieldasoneofourmetricsbecauseourpreviousfindingssuggestedthatyieldmultipleshaveahighpredictivepowerofsubsequentreturnsovera6-24monthperiod(RefertoGOAL–GlobalStrategyPaperNo.3:AsiaPacValuation:Whatworks,andwhen,March12,2012). Exhibit23:Ourmethodologywhichfocusesonthreecriteriaallowsustoscreenforgoodandbadconsumptionsectorsefficiently,inourviewOurscreeningframework Top-downIndustryanalysis Overlappingareas:Sectorsthatweare focusingon Bottom-upEarnings/margins Source:GoldmanSachsGlobalECSResearch. Valuations 高盛全球经济、商品和策略研究 22 2013年4月9日 ChinaHongKong IndiaIndonesia KoreaMalaysiaPhilippinesSingaporeTaiwanThailand 高盛全球经济、商品和策略研究 Exhibit24:India,IndonesiaandPhilippineslooktohave-downinvestmentstories;alotofChinaconsumptionsectorsarehighlyfragmentedScoringconsumptionsectorsbyration,fragmentationandASP AirlinesAutosHouseholdappliances,consumerelectronicsClothing/Footwear/SportswearDepartmentStoresHotels,tourism,gamingSupermarkets,hypermarkets,conveniencestoresBeersFoodproducts/plantationsoHomecarePersonalcareHealthcareInsuranceProperty&mediaPCs/mobiless PF$★★◯ ◯★★ ★◯◯★★◯★★◯★◯◯★ ◯★◯★◯★★★★★◯ ★◯ PF$★ ◯◯◯★◯ ◯◯◯ ★ ◯ ◯◯ ◯◯★◯ ◯◯◯◯◯ PF$★◯★★★★◯★★◯★★★★◯★★★★★★ ★★★ ★★★★ ★★★★★★★★◯★ PF$★◯ ★★◯★★★★★★★★★ ★★★★★ ★★★★ ★★★ ★★ PF$★ ◯★★◯ ◯★◯ ◯◯◯★◯◯◯★ ◯★◯ ◯◯★◯◯ ★◯◯◯ PF$◯ ◯◯ ★★★★★★◯ ◯★◯ ★★◯ ◯★ PF$★ ★★ ★★★★★★ ★★★ ★ ★★★★★★★★★★★ ★ PF$ ◯◯◯◯ ◯◯ ★◯◯ ◯◯◯ ◯★ ★ ◯◯◯◯◯ Notes:P:ration;F:Fragmentation;$:ASP★referstolowerrationvs.peers,lowerindustryfragmentation(i.e.higherindustryconcentration)vs.peers,orlowerASPvs.peers.Wegenerallyviewthisasapositivefactor.□referstohigherrationvs.peers,higherindustryfragmentation(i.e.lowerindustryconcentration)vs.peers,orhigherASPvs.peers.Wegenerallyviewthisasanegativefactor.Dataonthefollowingarenotavailable:Educationservices,andthefragmentationdataonfoodproducts/plantationsandhealthcare. Source:GoldmanSachsGlobalECSResearch,variousindustrysources(refertoAppendixIfordetails). PF$★◯ ◯◯★ ◯★ ◯ ◯◯★◯◯ ◯◯◯ ◯◯ ◯ PF$★ ★★★ ★ ★★★ ◯★★ ◯★ ★ 亚太 23 2013年4月9日 亚太 Exhibit25:Asianconsumptionsectors’storiesbasedonindustrydynamicsCategoriesandsectorexamples STRUCTURALTRENDS (increasingration) AsiaHealthCareChinaAirlines ChinaConsumerelectronicsChinaInsurance China&ASEANspendingIndiaAutos IndiaBreweryIndiaFoodproductsIndia&IndonesiaHome/personalcareIndia,Indonesia&PhilippinesPropertyIndia/ASEANDepartmentstoresPhilippinesHotels&gamingPhilippines&ThailandSupermarkets&conveniencestoresThailandClothing/apparelsThailandDepartmentstores MARKETSHARESTORY (highlyfragmentedindustry,lowpricing) ChinaDepartmentstoresChinaHotels ChinaSupermarkets/hypermarketsChinaBrewery IndiaClothing/apparelsIndias PRICE/PRODUCTUPGRADESTORY (raisingASP) AsiaPersonalcareTaiwanAirlines Taiwan&ThailandHotelsChina&MalaysiaClothing/apparels MalaysiaHomecarePhilipiineso Source:GoldmanSachsGlobalECSResearch. MATUREMARKETS (standardcyclicalplays) HongKong,Korea&TaiwanInsuranceKorea&MalaysiaAutos KoreaConsumerelectronicsKorea&TaiwanDepartmentstores HongKong&SingaporeHotelsKoreaBrewery&oKo
我们预计南亚地区下一轮需求浪潮将从必需品转向可选消费品，北亚地区将从商品转向服务。
金融服务和休闲相关行业是主要的受益板块。
明智选择消费板块：归类与简化对消费题材的投资是一个复杂的过程，因为其涉及板块较多且回报显著分化——在过去10年中最领先板块的回报是最落后板块的10倍。
然而，我们认为一个适当的框架可以简化整个过程。
我们界定了10个市场的19个消费板块，并根据它们的行业相关性和与宏观经济的关联将其划分为3类（宏观经济主导、行业主导和个股表现）。
确立投资观点：系统化的题材投资我们基于三个核心标准的筛选框架确定了7个最看好和7个最不看好的板块。
在我们看来，部分中国和印度的板块已再次显现买入良机；许多東盟国家的板块基本面稳健但估值充分。
收入增长、人口结构的变化和科技变革蕴含着四个投资主题：1)消费数字化、2)医疗保健、3)金融服务以及4)娱乐与旅游。
正确选择消费板块的重要性——在过去10年中最领先板块的回报是最落后板块的10倍 Returnsince2001(%)1000 800600400200 0(200) 鄧啟志+852-2978-0722richard.tang@高盛（亚洲）有限责任公司慕天辉,CFA+852-2978-1328timothy.moe@高盛（亚洲）有限责任公司KetakiGarg+91(80)6637-8601ketaki.garg@高盛（印度）证券私人有限公司刘劲津,CFA+852-2978-1224kinger.lau@高盛（亚洲）有限责任公司CaesarMaasry+852-2978-7213caesar.maasry@高盛（亚洲）有限责任公司SunilKoul+852-2978-0924sunil.koul@高盛（亚洲）有限责任公司 AutosPersonalcare Deptstoreso FoodproductsSupermarkets BreweryPCs/mobilesHealthCare HomecareInsurance ClothingTourismPropertysAirlinesEducationConselec 资料来源：FactSet、MSCI、高盛全球经济、商品和策略研究 高盛与其研究报告所分析的企业存在业务关系，并且继续寻求发展这些关系。
因此，投资者应当考虑到本公司可能存在可能影响本
报告客观性的利益冲突，不应视本报告为作出投资决策的唯一因素。
有关分析师的申明和其他重要信息，见信息披露附录，或参阅。
由非美国附属公司聘用的分析师不是美国FINRA的注册/合格研究分析师。
本报告仅供分发给高盛机构客户。
高盛集团 2013年4月9日 TableofContents Asianconsumption:TheNextWave–ExecutivesummaryThemacrocaseforAsianconsumptionTenyearson...howhasitprogressed?
Now,what’snext?
AroadmapforconsumptionpatternevolutionMappingtheterrain:Define,thencharacterizeAnassessmentframeworkforconsumptionsectorsTrendsthatgiverisetonewstructuralopportunitiesDigitalizationofconsumption:TheglobaltrendofgoingonlineHealthcare/wellness:TwindriversofhighereandlongerlivesFinancialservices:IncreasingdemandasAsiansewealthierTourismandentertainment:MorefocusonleisureLocalvs.foreignbrandsinAsiaIndustryconsolidationAppendixI:Top-downindustryanalysis:ParametersandsourcesAppendixII:DetailedstockdataMSCIdisclosureEquitybasketdisclosureDisclosureAppendix AllpricesinthereportareasofthemarketcloseofApril2,2013,unlessmentionedotherwise. 亚太 35581421293031323335363839414142 高盛全球经济、商品和策略研究
2 2013年4月9日 亚太 Asianconsumption:TheNextWave–Executivesummary ConsumptioninAsiaisapowerful,long-termthemewhichmanyinvestorsarefocusedon.Ourobjectiveinthisreportistoidentifyfreshopportunitiesintheconsumptionspaceandtohelpstructurethethoughtprocessforinvestinginthisimportantarea.Wehavefourmaininsights:
1.MacrotrendsConsumptionpatternschangeaspercapitaerises.Basedonourmacroforecasts,weexpectconsumptiondemandtoshiftfromnecessitiestodiscretionaryitems,andfromgoodstoservices.Inparticular: Clothingandapparel,restaurants,orautosmaynolongerbeanewstoryinurbanChina,butwillremainaleadingthemeinIndiaandASEAN, UrbanChinaandthemoreadvancedmarketsinAsiaarerapidlymovingtothenextstage,withtheirconsumerfocusshiftingfromgoodstoservices.
2.InvestingframeworkWeidentify19sectorsthataredirectlyrelatedtoconsumerdemandandnotethatperformancehasvariedwidely,whichunderscorestheimportanceofsectorandstockselection.Tomanagethearrayof10marketsx19sectors,wefocusontwoapproaches: Simplifyingthemonitoringprocess:Wecategorizethesectorsintothreebucketsofmacroplays,industrydriverfocusandstockpicks,basedontheirsectorcoherenceandmacroaffinity. Simplifyingtheselectionprocess:Wescoresectorsbasedonasetofuniformvariablesfocusingonthethreecorecriteria-downindustrylandscape,bottom-upfundamentals,andvaluations.Forstocks,wealsoincludecash-basedreturnmetrics,whichoursinglestockanalystshaveshowntohelpcapturealphaswiththesector.
3.SectorandstockideasWeranktherelativeattractivenessofponentsectorsoftheconsumptionthemeandidentifysevenmostandsevenleastpreferredsectors.Somehighlevelinvestmentconclusionsincludethefollowing: Valueisre-emergingforselectChinaandIndiasectors,especiallyChina,Chinaairlines,ChinainsuranceandIndiaautos. ASEANconsumptionsectorsarefairlyvaluedfortheirsolidfundamentals.SomesectorsinThailandandPhilippineslookrichlyvalued. AnumberofdomesticsectorsinTaiwanlooktohaveabination-downandbottom-upfundamentals,aswellashighvaluations.
4.FourthemesBesidesthecoredriverofhighere,twoadditionalforcesofdemographicsandtechnologypointtoinvestmentopportunitiesinfourthemes.Inanutshell, Increasinguseofandmobileleadstodigitalizationofconsumption; Ageingpopulationincreasesthedemandforhealthcare; Increasingwealthandlongerlifeexpectancyincreasesneedsforfinancialservices; Theincreasingfocusonleisurebringsdemandforentertainmentandtourism. 高盛全球经济、商品和策略研究
3 2013年4月9日 亚太 Exhibit1:Weseethenextwaveofconsumptiondemandshiftingfromnecessitiestodiscretionaries,andfromgoodstoservices ContributiontoPCEgrowth FoodClothingRestaurants AutosHousing ASEANandIndia Culture/EntertainmentFinancialsvcs Chinaandothermore"advanced"EM PercapitaeSource:GoldmanSachsGlobalECSResearch. Exhibit2:Sectorandstockideasbasedonourscreeningmethodology Market Sector Mostfavoredconsumptionsectors India Property China &media China Insurance India Autos China Airlines Indonesia Property Philippines Super-,
hypermarkets&conveniencestores Leastfavoredconsumptionsectors Taiwan Insurance Thailand Healthcare Singapore s Korea s Korea Super-,hypermarkets&conveniencestores Taiwan DepartmentStores Korea Foodproducts/plantations Note:SeeExhibit28forourscreeningdetails. Source:GoldmanSachsGlobalECSResearch. StockideasMostfavoredstockswithinthesector PhoenixMills,HDIL,OberoiRealtyTencent,VIPShop,SouFun,ChangyouPingAn,ChinaTaiping,NCIBajajAuto,TataMotors,HeroMotoCorpChinaEastern,AirChinaBumiSerpongDamaiPuregoldLeastfavoredstockswithinthesectorCathayFin.,ShinKongFin.BangkokDusitMedicalM1LGUplusE-martFarEasternDeptStoresLotteConfectionery,Nongshim Exhibit3:Weseethefollowingtrendsgivingrisetostructuralinvestmentopportunitiesinfourthemes DigitalizationofConsumption Healthcareandwellness Trendsthatgiverisetoinvestmentthemes Demographics e Technology Ageingpopulation Risinghealthawarenessaserises Increasinguseofandmobile Financialservices Longerlifeexpectancy Increasingwealth Entertainment Morefocusonleisureand andtourism qualityoflifease Note:AIA,CMB(H).KBFinancialandE.SunareonourregionalConvictionLisit. Source:GoldmanSachsGlobalECSResearch,GoldmanSachsResearch. ExamplesofbeneficiariesthatareBuy-rated GlobalLogisticProperties,Tencent,VIPShop Mindray,CMS,Shineway AIA,CMB,KBFinancial,
E.Sun GrandKoreaLeisure,Paradise 高盛全球经济、商品和策略研究
4 2013年4月9日 亚太 ThemacrocaseforAsianconsumption Tenyearson...howhasitprogressed?
ThepotentialofAsian/EMconsumerswasfirsthighlightedbyoureconomistsin2004aftertheintroductionoftheterm“BRICs”.Sincethen,ourglobalcolleagueshavewrittenawidearrayofresearchonthistheme.Theimportantpublicationsinclude: GlobalEconomicsPaperNo:118,TheBRICsandGlobalMarkets:Crude,CarsandCapital,Oct14,2004. TheWorldandtheBRICsDream,Feb14,2006. GlobalEconomicsPaperNo:208,TheBRICs10YearsOn:HalfwayThroughtheGreatTransformation,Dec7,2011. China:PortfolioStrategy:PickingtheChineseconsumptionwinners,Feb26,2010. China:PortfolioStrategy:Massmarketconsumption–China’sstructuralopportunity,May4,2012. China:Retail:ManyfacesoftheChineseconsumer,Feb8,2012. Summarizingthefindingsfromourpreviousresearch,wegenerallybelievedtherewerethreemajormacrotrendsprevailing. Risinges:WeexpectstrongeconomicgrowthintheemergingeconomiesonrapidumulationofcapitalandrisingproductivityandforecastepercapitainEMtoalmostdoubleby2030. Urbanization:Thecontinuousmoveofpeoplefromruraltourbanareaswillcreatedemandforhousing,transportation,municationandalsocapitalgoods. Demographics:Wewereforecastingfastgrowthintheworkingagepopulationinmanycountriesfromtheprogressionofvariousagecohorts(i.e.“demographicdividend”)andincreasinglaborparticipationrates.However,thishasrealizedquicklyoverthepastdecade,andsomeofthecountriesinAsiaarenowexpectingageingpopulationsinthenexttenyears. OurconclusionthenwasthatdemandforconsumergoodswouldrapidlyincreaseasAsianspursuehigherlivingstandards,andthegrowthdriverinAsiawouldshiftawayfrominvestmentsandexportstowardsconsumption. Almosttenyearshavepassed.Someoftheconsumptionthemeshaveturnedouttobegoodinvestments,whilesomehavenotplayedout.Atthesametime,therearealsosomenewopportunitiesemerging.Atthebeginningofthisreport,wefirstevaluatetheprogressofAsianconsumptioninthreeaspects: Macro:Themacroevidencehasbeenmixed.Overthepastfewyears,consumptionhasactuallyreduceditsshareofAsianGDP.However,othermacroindicatorssuchasretailsaleshaverecordedstronggrowth. Micro:AgoodexampleisChinaautosales,whichhavenowexceededthoseintheUS.Earningsinconsumersectors(inparticulardiscretionary)havegrownfasterthanthebroaderAsiaindex,andDMconsumersectorstoalargeextent. Investmentreturns:ReturnsfromAsianconsumersectorsalsooutperformedthebroaderregion,althoughtheytrailedUS.Aswehighlightiningsections,however,thesector-widereturnshavemaskedthewidedisparityinperformanceoftheunderlyingindustries. 高盛全球经济、商品和策略研究
5 2013年4月9日 亚太 Exhibit4:WeforecastepercapitainEMtoalmostdoubleby2030Percapitaeformajoreconomies(current,2020and2030) GDPpercapita(USD) 100,000 DM'spercapitaewillrise35% 80,000 inaggregateby2030 60,000 40,000 20,000
0 In18years,EM'sGDPpercapitawillstillbelow(partlyduetoalargepopulation), butitsgrowthratewillbeover3timesthatofDM's Current20202030 EM'spercapitaewillrise175%inaggregateby2030 SwitzerlandAustraliaUSA SingaporeHongKong FranceCanadaGermany IsraelUK ItalyTaiwanJapan SpainKoreaNewZealandRussiaTurkeyBrazilMalaysiaMexicoChinaSouthAfricaThailandIndonesiaPhilippinesIndia Exhibit5:Asiamayhaveitsmiddleclassexpandto44%ofthepopulationby2020 Middleclassas%ofpopulation(definedasesfrom6,000to30,000PPP$) 90%80%70%60%50%40%30%20%10% 0% EuropeLatAmAeJ World
N.AmericaJapan Exhibit6:UrbanizationisoneofthekeydriversforAsia’sconsumptiongrowth Urbanizationratio 90%80%70%60%50%40%30%20%
N.AmericaLatAmEuropeJapan WorldAeJ 198019851990199520002005201020152020198019851990199520002005201020152020 198019851990199520002005201020152020198019851990199520002005201020152020 Exhibit7:Thesignificantchangesindemographicsarehavinganimpactontheconsumptionpattern%ofpopulationinworkingage(left)andelderly(right) 70%65% AeJ LatAmWorldEuropeN.America 30%25%20%15% 60% Japan 10%5% 55% 0% Source:IMF,UnitedNations,GoldmanSachsGlobalECSResearchestimates. Japan EuropeN.America WorldLatAmAeJ 高盛全球经济、商品和策略研究
6 2013年4月9日 亚太 Exhibit8:ConsumptionasapercentageofGDPinAsiahasactuallydroppedinrecentyearsConsumptionas%ofGDP Consumptionas%ofGDP75%70%65%60%55%50%45%40%35%30% TaiwanIndiaKorea AeJ China Exhibit9:RetailsaleshasseenrapidriseinAsiaRetailsalesgrowthforAsiaandothereconomies Retailsales(%change)20%15%10% 5%0%-5%-10% AeJUS JapanEurope(EU-27) 199019952000200520102000200120022003200420052006200720082009201020112012 Exhibit10:China’sautosaleshavesurpassedthoseintheUS...CarsalesinChinaandUS Carsales('000units,monthly) 2,0001,8001,600 USChina 1,400 1,200 1,000 800 600 400 200
0 Exhibit11:...andweseeasimilarcaseputerproductsPCsalesinAsiaandUS PCSalesUnits(mnunits,quarterly)35 US 30 AsiaPacificexJapan 25 20 15 10
5 0 Jan-05Jul-05Jan-06Jul-06Jan-07Jul-07Jan-08Jul-08Jan-09Jul-09Jan-10Jul-10Jan-11Jul-11Jan-12Jul-12Jan-131Q951Q961Q971Q981Q991Q001Q011Q021Q031Q041Q051Q061Q071Q081Q091Q101Q111Q12 Exhibit12:EarningsgrowthinAsia’sconsumersectorshavebeenamongthestrongest EPSCAGR(5y)forglobalconsumersectors Exhibit13:ReturnsfromAsianconsumersectorshaveoutperformedthebroaderregion,buttheytrailedUS ReturnCAGR(5y)forglobalconsumersectors 16% 8% 5yEPSCAGR 14%6% 12% 4%10% 8% 2% 6%0% 4%-2% 2% 0% -4% MXAPJ MXAPJ MXAPJ S&
P SXXP consdiscconsstap consumerconsumer Source:CEIC,FactSet,Haver,IDC,Worldscope,GoldmanSachsGlobalECSResearch. MXAPJconsdisc MXAPJconsstap MXAPJ 5yreturnCAGR S&Pconsumer SXXPconsumer 高盛全球经济、商品和策略研究
7 2013年4月9日 亚太 Now,what’snext?
Aroadmapforconsumptionpatternevolution Asaneconomydevelops,consumerstendtochangetheirconsumptionpatterns–fromnecessitiestoluxuries,fromgoodstoservices,etc.Oneexampleisthegradualdropinproportionofespentonfoodthroughtime.Generallyspeaking,economiesthatarefurtheralongtheirdevelopmentpathstendtohavefoodexpendituretakingalowerportionofthetotal.WecanclassifyAsianmarketsintothreegroupsbasedonthis. Developedmarkets(DM):Regardlessofthestartingpoints,mostdevelopedmarketshave10%-15%oftheirtotalconsumptionexpendituresspentonfood.InAsia,theseincludeHongKong,Korea,SingaporeandTaiwan. Advancedemergingmarkets(EM1):Malaysia,China(urban)andThailandgenerallyhavetheirfoodexpenditurestaking25%-35%oftotal. Emergingmarkets(EM2):ThefoodexpendituresofhouseholdsinChina(rural),India,PhilippinesandIndonesiaarestillclosetohalfoftotalspending. Inourview,thisisbroadlyconsistentwiththestagesofeconomicdevelopmenteachmarketisin.TopredictwherethenextsetofdriversforAsianconsumptiongrowthcouldbe,westudythetypicalevolutionpatternusingUSasanexample,andparetheconsumptionbasketbetweeneachAsianmarketandthatofatypicalDMtoidentify“thegap”. BasedontheexperienceintheUS,webelieveatypicalconsumptionpatternevolutioninvolvesthreestages: oStage1:Atthebeginning,ahousehold’seisheavilyspentonnecessitieswhicharefoodandclothing.Somepaymoreforfoodservices. oStage2:Consumersthengraduallyputmorefocusonthequalityoftheirmodationandalsobegintopurchasetheirowncars.Theproportionofhousingandtransportexpenditurethusincreases. oStage3:Consumptionmovesawayfromgoodstoservices.Peoplespendmoreonleisureandentertainment.Theirdemandforfinancialservicesalsorisesaswealthumulates. oThecontinuousincreaseinspendingonhealthcaremunicationhaslikelybeendrivenbydemographicsandtechnologicaladvancement. OurgapanalysissuggestsmostemergingmarketsinAsiaarestillmeaningfullybehindDMsontheirspendingondiscretionaryitemsandservices,althoughtheydifferinmagnitude.Inourview,theultimatedriverofallthesechangesishouseholde(seeBox#1).Lookingforward,aseconomicdevelopmentprogresses,wethinkthenextstagesofdevelopmentforvariousmarketsandregionswillbeasfollows:
(1)Spendingonculture/entertainmentiscurrentlylowacrosstheboard.
(2)FinancialservicescouldbethenextwaveintheadvancedEMs.Chinahasbeenrapidlydevelopingthisarea,andcouldbearoadmapfortheASEANmarkets. (3)saresaturatedinmostmarketsexceptIndiaandThailand,butthetechnologyadvancement(fromfeaturephonestosmartphones)couldbringaboutaproductupgradestory.
(4)Healthcareisastructuraltrendfromageing,andthisishappeningglobally. Withabroadsenseofwherethenextwaveofconsumptionwillgo,wehaveabetterideaofwhattolookforaswedrilldowntotheopportunitiesofspecificindustrieswithmoregranularanalysesiningsections. 高盛全球经济、商品和策略研究
8 2013年4月9日 亚太 Exhibit14:AsianmarketsbroadlyfallintothreegroupsintermsoftheproportionofesspentonfoodConsumptionpatternevolution–foodas%ofpersonalconsumptionexpenditure Foodas%ofPCE70% ID 60% IN CN-ruralPH 50% 40% 30%US 20% KR HKUK AU GR CN-urban THTWMY JPSG 10% EM2EM1DM 0%1929 1939 1949 1959 Source:CEIC,Haver,GoldmanSachsGlobalECSResearch. 1969 1979 1989 1999 2009 高盛全球经济、商品和策略研究
9 19441954196419741984199420041944195419641974198419942004 2013年4月9日 Exhibit15:Thethreestagesofconsumptionpatternevolution:TakingUSasanexampleContributiontoPCEgrowthinUS ContributiontoPCEgrowth(10yaverage) 30% Restaurants Clothing Food&Beverages 25% 12%10% ContributiontoPCEgrowth(10yaverage) 30% Housing&utilities Transport25% 20% 8% 20% 15% 6% 15% 10% 4% 10% 5% 2% 5% 0% 0% 0% 1944195419641974198419942004 ContributiontoPCEgrowth(10yaverage) 12% Culture/Entertainment Financialservices10% 8% 6% 4% 2% 0% Source:Haver,GoldmanSachsGlobalECSResearch. ContributiontoPCEgrowth FoodClothingRestaurants AutosHousing Time 亚太 12%10%8%6%4%2%0% Culture/EntertainmentFinancialsvcs 高盛全球经济、商品和策略研究 10 2013年4月9日 亚太 Exhibit16:Asiavs.theDMbenchmark:MappingthestructuralopportunitiesBreakdownofpersonalconsumptionexpenditurebycategory ThespendingpatternofanaverageDMconsumer US Food&Beverages 8% Clothing 4% Restaurants 7% Housing&utilities 19% Furniture/Householdfacility4% Transport 10% UK 13%6% 10% Germany 14%5%6% Japan 16%3%7% Australia 14%3%7% DMavg 13%4%7% 24% 25% 25% 23% 23% 5% 6% 4% 5% 5% 14% 13% 11% 10% 12% Consumptionoftenfollowsatypical patternasamarketdevelops Culture/EntertainmentFinancialservices HealthcareCommunicationsEducation Other 9% 11% 9% 10% 10% 10% 8% 5% 7% 4% 9% 6% 20% 2% 5% 4% 6% 8% 2% 2% 3% 3% 2% 3% 3% 2% 1% 2% 4% 2% 7% 6% 6% 10% 7% 7% elerated
byageing,technology, etc. HowdoesthespendingpatternofAsianparetothatinDMs?
Food&BeveragesClothingRestaurantsHousing&utilitiesFurniture/HouseholdfacilityTransport TransportservicesAutosCulture/EntertainmentFinancialservicesLuxurygoodsHealthcareEducationCommunications DMavg13%4%7%23%5%12%2%10%10%5%1%8%2%3% Others 8% Sing9%3%9%19%6%13%3%10%14%5%2%7%3%2% 9% Korea16%5%8%16%3%12%3%9%8%7%0%7%7%4% 6% Taiwan15%5%7%18%5%11%2%10%9%3%1%10%4%4% 8% HK15% 6%10%19% 9%10%7%4% 4%7%2%4%2%2% 8% Malay24%3%10%16%5%13%5%8%5%5%1%2%2%8% 7% China-U27%11%8%10%7%9%3%7%3%2%0%7%9%6% 2% India34% 7%2%13%4%15%8%7%1%6%2%4%1%2% 7% Thai36% 7%9%7%6%15%5%10%5%1%0%6%1%1% 5% China-R41%6%NA19%5%11%NANANANANA7%8%NA 2% Phils43% 1%4%13%4%11%7%4%2%3%1%2%4%3% 8% Source:CEIC,Haver,GoldmanSachsGlobalECSResearch. Indo47%2%2%16%2%12%5%7%1%2%0%3%4%5% 4% 高盛全球经济、商品和策略研究 11 2013年4月9日 亚太 Box#1:Forecastingthepathofconsumptionpatternbypercapitae ForecastingchangeinfoodconsumptionwitheTherelationshipbetweenfoodconsumptionandhouseholdehasbeenwellresearchedineconomicliterature.ThemostwellknownruleisknownasEngel’slaw,whichsaysasacountry’serises,theproportionofespentonfoodtendstofall. Ourobjectiveistwofold.First,wewanttoverifyifmost(orall)majorcountriesinthisstudyfollowthispatternofconsumerbehavior.Second,basedonthiseconomictheory,areweabletoforecasthowthepatternwouldchangeifwehaveasenseofhowquicklyanAsianmarketgrowsitselevel?
Toachievethis,webuildapanelregressionmodeloffoodexpenditureasapercentageofpersonalconsumptionexpenditure(PCE)onpercapitae. esinastherelationshipofpercapitaeandfoodasapercentageofPCEisdefinitelynotlinear.TheEngelcurvefeaturessaturationpropertiesinthatthepaceofdecreaseinproportionoffoodspendingslowsathighelevels.Inotherwords,thereisalimitonhowmuchonecancuttheirfoodexpenditureportionaseincreases.Toincorporatethis,thesimplestwayistoassumeitfollowsalogfunction.Wedefineitasbelow: FoodAs%PCEab*ln(e) WepooldataonfoodasapercentageofPCEandpercapitaeacross15markets.Thelengthoftimeseriesspansfromthelongestof50yearsintheUStotheshortestof11yearsinMalaysia.Werunthedatathroughapanelregression,whichhelpsountforstructuraldifferencesacrossmarkets,andcanmodatedifferentdataavailability. Wefitafixedeffectsmodeltothepaneldata(thenullhypothesisofindividualeffectsbeinguncorrelatedwithotherregressorsisrejectedat1%and5%levelsofsignificancebytheHausmantest).TheresultsshowthatthenaturallogofpercapitaeisasignificantpredictoroffoodexpenditureasapercentageofPCEat1%and5%levelsofsignificanceanditsnegativecoefficientshowsthataspeopleericher,foodexpenditureasashareofPCEdeclinessteadilyuntilitreachesasaturationlevel.TheR-squaredis97%. Withthemodelinplace,weestimatehowfoodspendingasapercentageofPCEwillevolvebyusingIMF’sforecastsofeconomicgrowthandpercapitaelevelsining20years.Thespeedofreductionintheportionofeonfoodvariesmeaningfullyacrossmarkets.Ourmodelsuggeststhatnoneofthe“EM2”marketswillreachtheDMrange(8%-16%)evenafter20years.Withinthe“EM1”markets,MalaysiamayreachlevelssimilartothoseinDMsin10-15years,andaround15-20yearsforChina. Obviouslytheresultsaresolelydependentontheforecastsofeconomicgrowth,whichmayormaynotberealized,andtherearedefinitelyotherfactorsaffectingtheevolutionofconsumptionpattern.Nevertheless,byunderstandingwhatstagesdifferentmarketsareroughlyin,wehaveabettersenseofwhatconsumptionthemes(foodoressories,necessitiesorluxuries,goodsorservices)willbethenextwavesforeachofthem. Engelcurve:AlogfunctionfitthedataneatlyFoodas%ofPCEvs.percapitaeforthe15markets MalaysiaandChinamaybethefirsttoreach“DM”levelsofconsumptionForecastsoffoodas%ofPCEintothenext20years Foodas%ofPCE 70%60%50%40%30%20%10% 0%
0 10,000 20,000 30,000 Percapitae(US$) 40,000 50,000 Foodas%ofPCE70% 60% IN 50% IDCN-ruralPH KR EM2 40% 30% US 20% 10% HKUK AUGR THTW JP SG CN-urbanEM1 MY DM 0%19291939194919591969197919891999200920192029 资料来源：CEIC,Euromonitor,Haver,IMF,GoldmanSachsGlobalECSResearchestimates. 高盛全球经济、商品和策略研究 12 2013年4月9日 亚太 Doesonefactorexplainall?
Eventhoughpercapitaeisanimportant(andprobablyadominant)factordrivingthechangeinconsumptionpattern,itisnottheonlyonethatmatters.Wedonotintendtoforecasteveryponentofpersonalconsumptionexpenditurehere,butwethinkitwillbeusefultoponentsthataremoreedrivenfromthosethatarelessdirectlyrelated,orinvolvingotherpotentialdrivers.Toachievethis,wesimplyrunthesamepanelregressionagainfortheponentsparetheirpooledR-squared. Besidesfood,financialservicesandculture/entertainmentcanalsobewellexplainedbythechangeinpercapitae.Asindividualsewealthier,theytendtospendalowerportiononfood,startfocusingoninvestmentsandfinancialservices,andspendmoreonleisure. Statistically,educationalsohasastrongrelationshipwithe,butwethinkitisatwo-wayrelationshipratherthanjustaresultofit. Therelationshipisstilldecentlystrongwithponents,butsignificantlylowerinstrength.Wethinkitsuggeststherearepotentiallysomeotherfactorswhichalsoinfluencethechanges.Forexample,eventhoughweagreethathealthcarespendingtendstoincreasewithebecausepeoplefocusmoreontheirhealthastheyericher,itmayalsobearesultoflivingalongerlife.Demandmunicationshasalsobeendrivenbytechnologicaladvancementandproductlaunches(e.g.andsmartphones).Housingandtransportareoftentheby-productsofurbanizationwhichhasbeenhappeninggloballyoverthepastfewdecades. Engelcurve:alogfunctionfitsthedataneatlyFoodas%ofPCEvs.percapitaeforthe15markets PooledR-sq.frompanelregression1.00.90.80.70.60.5 PercapitaseemstobewellrelatedtoentertainmentexpenseForecastsoffoodas%ofPCEintothenext20years 18%16%14%12%10% 8%6%4%2%0%-2%
0 10,000 20,000 30,000 40,000 Percapitae(US$) 50,000 FoodFinancialsvs. EducationEntertainment OtherHealthCare ClothingFurnitureCommunicationsHousingTransport Entertainmentas%ofPCE Thefitgetsloosermunications...Foodas%ofPCEvs.percapitaeforthe15markets ...andhealthcareexpensesForecastsoffoodas%ofPCEintothenext20years Communicationsas%ofPCEHealthCareas%ofPCE 9%8%7%6%5%4%3%2%1%0%
0 10,000 20,000 30,000 Percapitae(US$) 40,000 50,000 25%20%15%10% 5%0%
0 10,000 20,000 30,000 Percapitae(US$) 40,000 50,000 资料来源：CEIC,Euromonitor,IMF,GoldmanSachsGlobalECSResearchestimates. 高盛全球经济、商品和策略研究 13 2013年4月9日 亚太 Mappingtheterrain:Define,thencharacterize Takingabroaddefinitionofconsumption;weidentify19sectors Consumptionisabroadconcept,asultimatelythedemandformostproductsandservicesislinkeddirectlyorindirectlytoconsumptiondemand.Wethinkitisimportanttodefinetheconsumptionuniverseforthisreportclearlybeforewemoveontothespecificindustryfundamentals. Weidentify19industriesasouruniverse,whicharemoredirectlyrelatedtoconsumerdemandasopposedtoindustrialactivities.EnergyandMaterialsareexcludedforthisreason,andonlyairlinesareincludedinIndustrials.Banks’earningscyclesaredrivenbybothconsumersandcorporatesowedonotincludethemeither.Amongallthepanies,weonlyselectahandfulthatselldirectlytoconsumers–exampleswouldincludeSamsungElectronicsandHTC.Wealsoexcludeutilitiesforitsheavilyregulatednature. TheyformawidedistributioninmarketcapandliquidityThe19industriesdifferwidelyintermsoftheircapandliquidityprofiles.Thelargersectorsareautos,propertyands,buttheremaining60%isalongtailinthedistribution(Exhibit18). MSCIconstituentsrepresentcloseto70%ofthefulluniversebymarketcap,buttheyonlycoverhalfofitintermsofturnover,asthesmallerconsumptionstockstradeproportionatelymore(Exhibit19). Performanceisalsodifferent:Thebesthasoutperformedtheworstby10timesEventhoughmostofthesesectorsareeventuallydrivenbyconsumerdemand,thehomogeneityhasnotappliedtoequityreturns.Disparityinperformancehasbeenmeaningful,withthebestsectorhavingoutperformedtheworstby10timessinceend-2000. Autos,thebestperformer,made915%pricereturn.Theoutperformancewasmostsignificantduring2009and2010,whereearningsgrowthwasaround100%ineachofthetwoyears.Therehasnotbeenmeaningfulre-ratingorde-ratingoverthepastdecade,despitelargeswingsaroundthelong-termaverageP/Eofroughly9X,suggestingthatperformancehaslargelybeenearningsdriven. Thesecondbestispersonalcare,whichhasseldomoutperformedotherssignificantly,butithasbeengoingupsteadilyandconsistentlythusmakingagoodoverallreturn.Thesectorhasre-ratedconsiderably,withforwardP/Erisingfrom5Xin2001to20Xrecently.Earningsgrowthhasbeensteady,ataround10%-20%formostoftheyears. Theworstsectorsarehouseholdappliances/consumerelectronicsandeducationservices,followedbyairlines.Allthreesectorshavehadverylittleearningsgrowthoverthepastdecade.Therehasnotbeenmeaningfulre-ratingorderatinginmultipleseither. Therehasbeenincreasingdivergenceininvestmentreturnsinrecentyears,whichhighlightstheimportanceofbeingmoreselectivewheninvestinginconsumptionthemes.Inthenextsection,weintroduceanassessmentframeworktodistinguishgoodsectorsfrombadones. 高盛全球经济、商品和策略研究 14 2013年4月9日 亚太 Exhibit17:Wedefine19consumptionsectorsPercentagefiguresrepresentsector/industryweightintheconsumptionuniverse,byfloatmarketcap ChinaHongKong IndiaIndonesia KoreaMalaysiaPhilippinesSingaporeTaiwanThailand AeJ GICSsectors GICSindustries/sub-industries Airlines Consumerdiscretionary Airlines 0.20.10.00.00.10.10.00.30.10.0
1 Autos Consumerdiscretionary AutoComponents 0.20.00.40.11.80.20.00.00.60.1
3 Consumerdiscretionary Automobiles 1.10.00.81.12.80.10.00.30.20.0
6 Householdappliances,consumerelectronics Consumerdiscretionary HouseholdDurables 0.30.30.10.00.70.00.00.00.10.0
2 Consumerdiscretionary LeisureEquipment&Products 0.10.10.00.00.00.00.00.00.30.0
0 Clothing/Footwear/Sportswear Consumerdiscretionary TextilesApparel&LuxuryGoods 1.00.20.30.00.30.00.00.00.60.0
2 Consumerdiscretionary SpecialtyRetail 0.70.70.00.10.10.00.00.10.20.1
2 Departmentstores Consumerdiscretionary MultilineRetail 0.30.10.10.20.60.10.40.00.10.1
2 Hotels,tourism,gaming Consumerdiscretionary HotelsRestaurants&Leisure 0.22.90.10.10.40.80.10.70.10.2
6 Educationservices Consumerdiscretionary DiversifiedConsumerServices 0.00.00.00.00.00.00.00.00.00.0
0 Supermarkets,Hypermarkets,conveniencestores Consumerstaples Food&StaplesRetailing 0.60.30.00.10.30.00.10.20.20.7
2 Beers Consumerstaples Beverages 0.20.00.30.00.10.20.10.00.00.2
1 Foodproducts/plantations Consumerstaples FoodProducts 1.90.10.41.00.91.20.30.90.70.4
8 o Consumerstaples o 0.00.00.80.30.50.20.00.00.00.0
2 Homecare Consumerstaples HouseholdProducts 0.10.00.50.20.30.00.00.00.00.0
1 Personalcare Consumerstaples PersonalProducts 0.50.00.40.00.40.00.00.00.00.0
1 Healthcare Healthcare Healthcare 1.20.10.00.21.20.20.00.10.50.3
4 Insurance Financials Insurance 2.53.40.00.01.30.10.00.10.90.2
9 Property Financials RealEstate 4.19.80.40.80.00.70.83.50.60.9 22 &media Consumerdiscretionary Media 0.10.30.40.40.50.10.10.40.00.3
2 Consumerdiscretionary &CatalogRetail 0.10.00.00.00.20.00.00.00.00.0
0 Informationtechnology Software&Services 4.20.00.00.00.90.00.00.00.00.0
5 Informationtechnology HomeEntertainmentSoftware 0.00.00.00.00.30.00.00.00.00.0
0 PCs/mobiles Informationtechnology TechnologyHardware&Equipment0.50.00.00.02.40.00.00.01.00.0
4 s municationservicesmunicationservices 5.20.20.41.00.31.10.31.71.51.1 13 Total 25195 6165
2 8
8 5 100 Source:FactSet,GICS,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 15 2013年4月9日 Exhibit18:AsiaconsumptionsectorsdifferwidelyintermsofcapandliquidityprofilesFreefloatcapandaverageturnoverofeachofthecountrysectors Freefloatcap(US$bn)152 1009080706050403020100 ADVT(US$mn)921 500450400350300250200 150100 500 亚太 Exhibit19:MSCIrepresentsonlyaportionofthefulluniverseBreakingdownourconsumptionuniversetoMSCIandnon-MSCI,byfloatmarketcapandaverageturnover FloatMarketCap AverageTradingVolume NonMSCI31% MSCI69% NonMSCI49% MSCI51% Source:FactSet,MSCI,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 16 2013年4月9日 亚太 Exhibit20:Thebestsectorhasoutperformedtheworstby10timesoverthepastdecadePerformancetimeseriesandannualreturnfigures Priceperformance(Dec2000=100) 1,200 Brewery 1,000 Clothing/Footwear/SportswearDepartmentstores 800 Foodproducts/plantations Supermarkets,Hypermarkets, 600 conveniencestores o 400 Priceperformance(Dec2000=100) 1,200 Airlines 1,000800 Autos Householdappliances,consumerelectronicsHomecare Personalcare 600Property 400 Priceperformance(Dec2000=100)1,200 Hotels,tourism,gaming 1,000 EducationservicesHealthCare 800 Insurance &media 600 PCs/mobiles s 400 200 200 200
0 0
0 Dec-11 Dec-10 Dec-09 Dec-07 2011 Dec-08 Dec-06 Dec-05 Dec-04 Dec-03 Dec-02 2007 Dec-01 Dec-00 2002
Dec-10Dec-11 2003 Dec-09 Dec-08 Dec-07 Dec-06 Dec-05 Dec-04 Dec-03 Dec-02 Dec-01 Dec-00 Dec-00
Dec-01Dec-02Dec-03Dec-04Dec-05Dec-06Dec-07Dec-08Dec-09Dec-10Dec-11 Performance(%) AutosPersonalcareDepartmentstoresoFoodproducts/plantationsSupermarkets,Hypermarkets,conveniencestoresBreweryPCs/mobilesHealthCareHomecareInsuranceClothing/Footwear/SportswearHotels,tourism,gaming&mediaPropertysAirlinesEducationservicesHouseholdappliances,consumerelectronics Max-MinStd.dev./AbsoluteMean 2001 356 45(11) 5(5)295625112515(13)(10)(15)(26)(30)16(33) 883.9 2002 18(11)
(7)(8)2(9)4(22)(20)(15)(24)19
(6)(19)(26)(20)6(37)(12) 561.5 2003 8047613737371144381838422242452225 927 720.5 2004 216 531122312 55(27)141927(4)22135
(7)(3) 571.5 2005 3950563326272159533915 71922 811 36045 570.6 2006 (2)382019392934
(4) 9308932321845442579(17) 2007 15324226493634 725 74434253143524499
3 2008 (52)(24)(51)
(9)(44)(42)(45)(37)(37) 5(45)(61)(56)(48)(58)(39)(59)(53)(60) 2009 19681651183726574712759916895783602493 2010 40182526163830361529(1)1742 77743(17)
(9) 2011 4
(2)(16)25
(7)(5)(16)(19)(14)19(21)(22)(13)(12)(26)5(35)(37)(40) 2012 923131812193517252419 919224214 1(17)(30) 10yearperf93875948145243241039131630830028722317711910658125(72) 105 96 66 193 61 64 72 1011 0.9 0.6 0.4 0.6 0.9 1.4 1.1 0.8 Returndispersion(Std.dev./Absolutemean)2.0 1.6 Increasingdivergenceinsectorreturns 1.2 0.8 0.4 0.0 2012 2010 2009 2008 2006 2005 2004 2001 Note:Thechain-linkedperformanceofHouseholdappliances,consumerelectronicswassignificantlydraggeddownbyanumberofKoreastocksintheearly2000s.Source:FactSet,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 17 2013年4月9日 亚太 Tradingconsumptionsectorsintelligently:Classifyandsimplify Investinginconsumptioncanbeplicateddecisionasitinvolvesalargenumberofdifferentsectors.However,withtheappropriateframeworkwethinktheprocesscanbesimplified.Herewepresentourmethodologytoclassifythesectorsintothreebroadgroupswithdistinctcharacteristics(macroplays,industrydriverfocus,andstockpicks)byanalyzingthemalongtwoaxes:sectorcoherenceandmacroaffinity.RefertoAsiaPacific:PortfolioStrategy:AsiaSectorRadar:Mappingtheterrain,June1,2009andAsiaPacific:PortfolioStrategy:SectorstrategymattersforPMs&stock-pickers,January8,2013forfurtherresearch.
1.Sectorcoherence:Dotheytradeasasectororacollectionofindividualstocks?
Coherenceisanimportantattribute,asitindicateshowimportantitistofocusonidiosyncraticriskswhentradingsectors.Ifasectoris“highlycoherent”,itsuggeststhatmonfactorssharedamongstocks(whichmaybemacroorindustry-specific)dominateindividualrisks,thusthefundamentalanalysescanbeefficientlysimplified.Ontheotherhand,somesectorsaremerelyagroupofstockstradingseparately.Inthiscase,muchmoretimehastobeinvestedinpanyspecifics.
2.Macroaffinity:Dotheytrademoremacroormicro?
“Tradingmacroortradingmicro”pointsustofocusingontherightstrategies.Tradinga“macrosector”isessentiallyexpressingamacroview,sotheimplementationisrelativelystraightforward.“Microsector”ofteninvolvesitsownindustrydrivers,orspecificsupply-demanddynamicsthatbasicmacroindicatorswouldnotfullycapture. AlongthetwoaxeswecanclassifythemintosomebroadcategoriesInourview,thesebucketsdiffervastlyintermsoftradingbehavior.Foraninvestmentmanager,wethinkitmaybebeneficialtorecognizethedifferencesasithelpstooptimizetheallocationoftimeandresourcesamongdifferentsectors. Highcoherence,highmacroaffinity–the“macroplays”:Asthesectorstradecloselywithmacro,andwithinthesectorstockstradecloselywitheachother,investmentdecisionstendtobemoreaboutchoosingbetweenhighandlowbetanames.Alotofautos,insuranceandpropertysectorsfallintothiscategory.Interestinglyenough,mostofthedomesticsectorsinTaiwanexhibitbothhighmacroaffinityandcoherence.WesuspectapossiblereasonisthathistoricallymanagershaveinvestedmoretimeinthestockandsectorspecificsinTaiwantechnames,buthavepaidlessattentiontothenon-techsidethustendtotradethemaltogetherasanexpressionoftheirviewsondomesticconsumersentiment. Highcoherence,lowmacroaffinity–moreindustrydriverfocus:Tradingthesesectorsmeansonehastomonitorafewmoreparameters,whichmayrangefromsamestoresalesgrowthtoARPUtrend.Thegoodnewsisthatstocksinthesesectorsstilltendtotradecloselyenoughwitheachothersopickingtherightsectoralreadytakesushalfwaytooutperformingthebenchmark. Lowcoherence,lowmacroaffinity–it’saboutstockpicking:Inourview,managersneedtoallocatethemosttimeinthese(so-called)sectorsasreturndriversdifferfromstocktostock.Somemayconsistpaniessellingnicheproducts,thusonehastoputmoreeffortintomanagingtheidiosyncraticrisks. Consistentwithourintuition,wefindveryfewsectorsshowinglowcoherencebuthighmacroaffinity. 高盛全球经济、商品和策略研究 18 2013年4月9日 亚太 Exhibit21:WeanalyzesectortradingbehavioralongthetwoaxesofsectorcoherenceandmacroaffinityRankingsofsectorcoherence(left)andmacroaffinity(right)forlargeconsumptionsectors Averagestockcorrelationwithsector Averagecorrelationwithmacroindicators CNInsuranceTWInsurance SGPropertyTWClothing TWAutosMYs TWPropertyINAutos TWFoodproductsCNProperty CNPCs/mobilesTHs KRDeptstoresKRInsuranceSGGamingHKPropertyCNAutosKRAutosMYProperty MYFoodproductsIDProperty CNHypermarketsCNClothingCN IDFoodproductsKRHealthCareIDsCNs HKHotels/gamingINFoodproductsMYHotels/gamingCNFoodproductsSGFoodproducts PHPropertyIDAutos HKClothingKRFoodproductsCNPersonalcare SGsTWsKR&mediaKRElectronicsCNHealthCare INAutosCNPropertyKRInsuranceHKPropertyHKHotels/gaming CNAutosTWProperty TWAutosTWClothingHKClothingCNClothingMYPropertySGPropertyCNPersonalcare KRAutosINFoodproducts PHPropertyTWInsuranceKRElectronicsCNInsuranceSGFoodproductsCNPCs/mobiles SGGamingTWsKRFoodproductsMYFoodproducts KRIDAutos SGsCN CNHypermarketsMYHotels/gaming KRHealthCareKRDeptstoresIDFoodproducts CNsMYsTWFoodproductsCNFoodproductsCNHealthCare IDsTHs IDProperty 0% 20% 40% 60% 80% 0% 10% 20% 30% 40% Noteonmethodology: Weassesssectorcoherencebycalculatingtheaveragestockcorrelationwithsectorreturns(sinceend-2000).Formacroaffinity,weestimateusingthehighestcorrelationamongthatofsectorreturnswithchangesinIP,PMI,retailsales,CPI,exportsandourFinancialConditionsIndex(FCI). Source:CEIC,Haver,FactSet,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 19 2013年4月9日 亚太 Exhibit22:Recognizingthedifferencesintradingbehavioramongsectorshelpstooptimizetheallocationoftimeandresourcesintheinvestmentprocess Classifyingmajorconsumptionsectorsintothreebuckets 40%35% Tradingmacro INAutos MacroPlays Correlationwithmacroindicators 30%25% HKHotels/gamingKRInsuranceCNProperty HKPropertyTWProperty CNAutos TWAutos TWClothing SGProperty 20% CNPCs/mobiles SGGaming TWInsurance 15%indivTirdaudailnsgtocks KR KRDeptstoresMYs CNInsurance Tradingasasector 10% 5% 0%30% CNHealthCareCNFoodproducts IDsIDProperty Stockpicking Tradingmicro TWFoodproducts THs Industrydriverfocus 40% 50% 60% Stockcorrelationwithsector 70% 80% Source:FactSet,GoldmanSachsGlobalECSResearch. 高盛全球经济、商品和策略研究 20 2013年4月9日 亚太 Anassessmentframeworkforconsumptionsectors Screeningforopportunities:theframework Theobjectiveofthisassessmentframeworkistoscreenforinvestmentopportunitiesefficientlyinabroadarrayofcountrysectorsrelatedtoconsumption(10countriesx19sectors),usingsomeuniformvariables.Ourmethodologyfocusesonthreecriteria:-downanalysisoftheindustrylandscape,bottom-upanalysisfocusingongrowthandmargins,aswellasvaluations.Essentiallyweareaskingourselvesthreequestionswhenwegothroughthescreeningprocess:
1.Doesthissectorhaveagoodstory(orabadone)?
2.Isitjustastory,oralsobackedupbycorporateprofits?
3.Hasthemarketpricedinallthesealready?
Top-downindustryanalysisToevaluatetheindustrylandscape,weidentifythreekeyfactors: ration:Alotoftheearly-stagestructuralstoriesarebasedonthepotentialofincreasingration. Fragmentation:Ahighlyfragmentedindustrystructurewilllikelyreducepricingpower,anddampenmargins.Also,inpetitiveindustry,paniesneedtoinvestmuchmoreinbrandbuildingtodifferentiatethemselves.ThiscanrequirealotofR&Dandmarketing,inourview. ASP:Priceorproductupgradecouldbeausefulstrategytodrivefurtherearningsgrowthinanindustrywhichisalreadywellrated.WegenerallyviewlowASPrelativetopeersaspositiveasitindicatespotentialtoraisepricing.Insomecases,however,lowsellingpricescanbetheresultofhighfragmentation,whichsuggeststheindustryneedstogothroughconsolidationbeforeanASPupgradeispossible. Wescorethecountrysectorsbasedontheabovecriteria.Mostofthemcanbebroadlyclassifiedintofourgroups–structuraltrendsfromincreasingration,marketsharestory,price/productupgradestoryandmature,stablemarkets. Wemakesomehigh-levelobservations: IndiaandIndonesiahavelowrationinmanydiscretionaryconsumeritems.WefindsimilartrendsinthePhilippinestoo,buthigherelevelisacriticalelementtodriverationupwards. spendingandhealthcarearethefewsectorswhichwouldlikelybeastructuralstoryacrossmostmarketsinAsia.Wewillstudyeachofthemindetailinaseparatereport. QuiteafewconsumersectorsinChina,inparticularretail,arefragmented.Lackofpricingpowercreatesseriousmarginpressureforindividualplayers,eveniftheunitvolumeisgrowingrapidlyinsomecases.Industryconsolidationislikelyanemergingthemegoingforward. “Tradingup”isthenextearningsdriverforanumberofsectors.AgoodexampleistheessfulproductpremiumizationbyAmorePacificandLGH&HintheKoreacosmeticsindustry. ConsumersectorsinHongKong,Korea,SingaporeandTaiwanaremostlymatureandstablemarkets.Theyhaveestandardcyclicalplays,withonlyafewexceptions. 高盛全球经济、商品和策略研究 21 2013年4月9日 亚太 Bottom-upanalysis:GrowthandmarginsHavinga-downstorygivesinvestorsconfidenceinthelong-termprospectsofanindustry,butthismayormaynotbereflectedpanyprofits.Someoftheminvolveextensivecapexupfront,whichmeanswemayseelimitedpositiveimpacttonear-termearnings.Thisunderpinstheimportanceofhavingbottom-upearningsanalysisaswell.ThetwovariablesweusearetheearningsCAGRandthemarginsinthenextthreeyears(2013-15).Wefindthatanumberofpropertysectorshavebinationofhighgrowthandhighmargins,togetherwithChina.Onthelow-growth-low-marginside,theseincludeChinaautosandTaiwaninsurance. What’spriced:ValuationsTherationaleforvaluationsisprettystraightforward.BesidesP/EandP/B,wealsoincludedividendyieldasoneofourmetricsbecauseourpreviousfindingssuggestedthatyieldmultipleshaveahighpredictivepowerofsubsequentreturnsovera6-24monthperiod(RefertoGOAL–GlobalStrategyPaperNo.3:AsiaPacValuation:Whatworks,andwhen,March12,2012). Exhibit23:Ourmethodologywhichfocusesonthreecriteriaallowsustoscreenforgoodandbadconsumptionsectorsefficiently,inourviewOurscreeningframework Top-downIndustryanalysis Overlappingareas:Sectorsthatweare focusingon Bottom-upEarnings/margins Source:GoldmanSachsGlobalECSResearch. Valuations 高盛全球经济、商品和策略研究 22 2013年4月9日 ChinaHongKong IndiaIndonesia KoreaMalaysiaPhilippinesSingaporeTaiwanThailand 高盛全球经济、商品和策略研究 Exhibit24:India,IndonesiaandPhilippineslooktohave-downinvestmentstories;alotofChinaconsumptionsectorsarehighlyfragmentedScoringconsumptionsectorsbyration,fragmentationandASP AirlinesAutosHouseholdappliances,consumerelectronicsClothing/Footwear/SportswearDepartmentStoresHotels,tourism,gamingSupermarkets,hypermarkets,conveniencestoresBeersFoodproducts/plantationsoHomecarePersonalcareHealthcareInsuranceProperty&mediaPCs/mobiless PF$★★◯ ◯★★ ★◯◯★★◯★★◯★◯◯★ ◯★◯★◯★★★★★◯ ★◯ PF$★ ◯◯◯★◯ ◯◯◯ ★ ◯ ◯◯ ◯◯★◯ ◯◯◯◯◯ PF$★◯★★★★◯★★◯★★★★◯★★★★★★ ★★★ ★★★★ ★★★★★★★★◯★ PF$★◯ ★★◯★★★★★★★★★ ★★★★★ ★★★★ ★★★ ★★ PF$★ ◯★★◯ ◯★◯ ◯◯◯★◯◯◯★ ◯★◯ ◯◯★◯◯ ★◯◯◯ PF$◯ ◯◯ ★★★★★★◯ ◯★◯ ★★◯ ◯★ PF$★ ★★ ★★★★★★ ★★★ ★ ★★★★★★★★★★★ ★ PF$ ◯◯◯◯ ◯◯ ★◯◯ ◯◯◯ ◯★ ★ ◯◯◯◯◯ Notes:P:ration;F:Fragmentation;$:ASP★referstolowerrationvs.peers,lowerindustryfragmentation(i.e.higherindustryconcentration)vs.peers,orlowerASPvs.peers.Wegenerallyviewthisasapositivefactor.□referstohigherrationvs.peers,higherindustryfragmentation(i.e.lowerindustryconcentration)vs.peers,orhigherASPvs.peers.Wegenerallyviewthisasanegativefactor.Dataonthefollowingarenotavailable:Educationservices,andthefragmentationdataonfoodproducts/plantationsandhealthcare. Source:GoldmanSachsGlobalECSResearch,variousindustrysources(refertoAppendixIfordetails). PF$★◯ ◯◯★ ◯★ ◯ ◯◯★◯◯ ◯◯◯ ◯◯ ◯ PF$★ ★★★ ★ ★★★ ◯★★ ◯★ ★ 亚太 23 2013年4月9日 亚太 Exhibit25:Asianconsumptionsectors’storiesbasedonindustrydynamicsCategoriesandsectorexamples STRUCTURALTRENDS (increasingration) AsiaHealthCareChinaAirlines ChinaConsumerelectronicsChinaInsurance China&ASEANspendingIndiaAutos IndiaBreweryIndiaFoodproductsIndia&IndonesiaHome/personalcareIndia,Indonesia&PhilippinesPropertyIndia/ASEANDepartmentstoresPhilippinesHotels&gamingPhilippines&ThailandSupermarkets&conveniencestoresThailandClothing/apparelsThailandDepartmentstores MARKETSHARESTORY (highlyfragmentedindustry,lowpricing) ChinaDepartmentstoresChinaHotels ChinaSupermarkets/hypermarketsChinaBrewery IndiaClothing/apparelsIndias PRICE/PRODUCTUPGRADESTORY (raisingASP) AsiaPersonalcareTaiwanAirlines Taiwan&ThailandHotelsChina&MalaysiaClothing/apparels MalaysiaHomecarePhilipiineso Source:GoldmanSachsGlobalECSResearch. MATUREMARKETS (standardcyclicalplays) HongKong,Korea&TaiwanInsuranceKorea&MalaysiaAutos KoreaConsumerelectronicsKorea&TaiwanDepartmentstores HongKong&SingaporeHotelsKoreaBrewery&oKo
声明：
该资讯来自于互联网网友发布，如有侵犯您的权益请联系我们。