摩根斯坦利全新股票盈利预期模型SWEEP

May7,2012USEquityStrategyIntroducingSWEEP:MorganStanley’sSector-WeightedEquityEarningsPredictorThetrajectoryofthestockmarketoverthenextfewmonthswilllikelycomedowntoviewsongrowth.ManyinvestorshavetiredofEuropeannewsanddomesticbanteraboutfiscalcliffsandarelookingtoearningsasabarometer.Assuch,wewantedtomorerigorouslyapproachouroutlookandforecastofS&P500earnings.LaunchingourquantitativeS&Pearningspredictor:SWEEPforecasts13-24month-aheadearningsforeachS&P500sectorandaggregatesthemintoanoverallmarketearningsestimate.Ournewmodelusesarangeofvariables,includingmacroeconomic,financialandtechnicalfactors,toforecastearningsgrowth.Bothstatisticalanalysisandfundamentaljudgmentwereusedinbuildingthemodel.Anewtooltoassistinsettingourpricetarget:Ournew2013S&Pearningsestimateis$98.71,4.3%belowouroldestimateof$103and17.5%belowthe2013bottom-upconsensus,whichisnear$120.Attimes,wemayapplyourjudgmenttoSWEEPresultstoobtainourearningsestimate;inthiscase,however,theresultmatchesourviews.Ournewbullandbear2013earningsare$116.30and$81.13(Exhibit2).For2012,ouryear-endpricetargetfortheS&P500remainsat1167,14.8%belowwherethemarketistradingtoday.Investorscannowlookbehindtheheadlineearningsnumber:SWEEPcontains19factordriversandweshowthecontributionofeachoneto2013earnings(Exhibit19).Interestingly,13of19factorsarecontributingnegativelytoour2013estimate(relativeto2012consensus).Thisgivesusgreaterconfidenceinthisnon-consensusresult.Wealsodecomposeestimatedearningsbysector(Exhibit21).SWEEPforecastsrelativelybetter2013earningsgro[e-mail]toforecastingS&Pearnings,consideringbothmarket-wideearningsforecastsandaggregatedsectorforecasts,andusingdifferentpermutationsofforwardandtrailingearningsasourbase(Exhibit5-Exhibit8).SWEEPissectorbasedandforecastsearningsgrowthoffthe12-monthconsensus.Weintuitivelypreferredthiscombinationbecauseitallowssector-specificfactorstoenter,accountsforsecularshiftsinsectorearningsweightsandutilizesinformationcontainedinanalystforecasts.Testson2-yearreportedS&P500earningsgrowthconfirmedoursupposition:SWEEPhadthesmallesterrorsrelativetoactualgrowthratesamongallsixmodelstested.SWEEPperformswellinnormalmarketconditionsandcapturespost-crisisreboundsinearnings(Exhibit1),butcannotcaptureextremeswings–particularlydownwardmoves.WethereforelayerjudgmentonSWEEPinextremeperiods.Exhibit1SWEEPTracksTwo-YearReportedEPSGrowthinNormalPeriodsSectorModel1(Smoothed4QNTM)vsObserved2-YearGrowth(150%)(100%)(50%)0%50%100%150%200%250%300%899091929394959697989900010203040506070809Smoothed4QNTMModel1Observed2-YearGrowthSource:Factset,MorganStanleyResearchMorganStanleydoesandseekstodobusinesswithcompaniescoveredinMorganStanleyResearch.Asaresult,investorsshouldbeawarethatthefirmmayhaveaconflictofinterestthatcouldaffecttheobjectivityofMorganStanleyResearch.InvestorsshouldconsiderMorganStanleyResearchasonlyasinglefactorinmakingtheirinvestmentdecision.Foranalystcertificationandotherimportantdisclosures,refertotheDisclosureSection,locatedattheendofthisreport.MORGANSTANLEYRESEARCHNORTHAMERICAMorganStanley&Co.LLCAdamS.Parker,[e-mail][phone]BrianT.Hayes,[e-mail]@morganstanley.comAdamJ.Gould,[e-mail]@morganstanley.comRequestyourballot.MorganStanleyappreciatesyoursupportintheInstitutionalInvestor2012All-AmericaEquityResearchTeamSurvey.2MORGANSTANLEYRESEARCHMay7,2012USEquityStrategySWEEP:MorganStanley’sSector-WeightedEquityEarningsPredictorInthisnote,weintroduceSWEEP,ourquantitativemarketearningsforecastingmodel.Thismodel“sweeps”currentlyavailablefinancialandmacroeconomicdatainordertoestimateagrowthrateforS&P500earnings.Bycombiningthisgrowthratewiththecurrentearnings,weobtainanearningsforecastfortheS&P500.HowwillweuseSWEEP?WewilluseSWEEPinformingourEPSestimatesfortheS&P500.Previously,ourS&Pearningsforecasts(pleaseseeUSEquityStrategy:The2012Playbook,January2,2012)incorporatedmacroandfinancialdata,buttherewasnotasystematicorganizationandweightingschemethatallowedustobalancecompetingeffects–suchaspositivesensitivityofenergysectorearningstocrudeoilpricechangesversusnegativesensitivityforconsumerdiscretionaryearnings.WithSWEEP,wecanquantifythesensitivityofearningstomanyvariablesatonceandaccountforvaryingoroppositesectordependence,aswellascorrelationsamongvariables.Periodically,S&Pearningsareinfluencedbyfactorsthataredifficulttoquantify.ExamplesofsucheffectsincludequantitativeeasingbytheFed,majorregulatorychanges,financialcrises,orcrudeoilshocks.Inthesesituations,wewilllayerfundamentaljudgmentontopoftheSWEEPforecasttoobtainourS&Pearningsestimate.Innormalenvironments,orwhenwedonothavefundamentalviewsthatdivergefromSWEEP,wewilldirectlyusethequantitativemodelearningsestimateforouroutlook.Inthisinstance,wehaveadjustedour2013S&PearningsestimatetomatchtheSWEEPforecast.SWEEPisreallyaquarterlymodel,asitcontainsfactorsthatareonlypublishedquarterly(e.g.