This preview has intentionally blurred parts. Sign up to view the full document

View Full Document

Unformatted Document Excerpt

andForecasting ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 1 Overview Regressionanalysis Chapter5 DemandEstimation Hazardswithuseofregression analysis Subjectsofforecasts Prerequisitesofagoodforecast Forecastingtechniques ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 2 Learningobjectives understandimportanceofforecastingin business describesixdifferentforecastingtechniques knowhowtospecifyandinterpretaregression ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 3 Learningobjectives recognizelimitationsofconsumerdata useseasonalandsmoothingmethods ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 4 Datacollection Dataforstudiespertainingtocountries,regions, orindustriesarereadilyavailable Dataforanalysisofspecificproductcategories maybemoredifficulttoobtain buyfromdataproviders(e.g.ACNielsen,IRI) performaconsumersurvey focusgroups technology:pointofsale,barcodes ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 5 Regressionanalysis Regressionanalysis:aprocedurecommonly usedbyeconomiststoestimateconsumer demandwithavailabledata Twotypesofregression: crosssectional:analyzeseveralvariablesfor asingleperiodoftime timeseriesdata:analyzeasinglevariable overmultipleperiodsoftime ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 6 Regressionanalysis Regressionequation:linear,additive eg:Y=a+b1X1+b2X2+b3X3+b4X4 Y:dependentvariable a:constantvalue,yintercept Xn:independentvariables,usedtoexplainY independentvariables) bn:regressioncoefficients(measureimpactof ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 7 Regressionanalysis Interpretingtheregressionresults: coefficients: negativecoefficientshowsthatastheindependent variable(Xn)changes,thevariable(Y)changesinthe oppositedirection positivecoefficientshowsthatastheindependent variable(Xn)changes,thedependentvariable(Y) changesinthesamedirection magnitudeofregressioncoefficientsisameasureof elasticityofeachvariable ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 8 Regressionanalysis Statisticalevaluationofregressionresults: ttest:testofstatisticalsignificanceofeach estimatedcoefficient b t= SE b b=estimatedcoefficient SEb=standarderrorofestimatedcoefficient ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 9 Regressionanalysis Statisticalevaluationofregressionresults: ruleof2:ifabsolutevalueoftisgreater than2,estimatedcoefficientissignificantat the5%level ifcoefficientpassesttest,thevariable hasatrueimpactondemand ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 10 Regressionanalysis Statisticalevaluationofregressionresults R2(coefficientofdetermination):percentage ofvariationinthevariable(Y)accountedfor byvariationinallexplanatoryvariables(Xn) R2valuerangesfrom0.0to1.0 thecloserto1.0,thegreaterthe explanatorypoweroftheregression ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 11 Regressionanalysis Statisticalevaluationofregressionresults Ftest:measuresstatisticalsignificanceofthe entireregressionasawhole(noteach coefficient) ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 12 Regressionresults Stepsforanalyzingregressionresults checkcoefficientsignsandmagnitudes computeimpliedelasticities determinestatisticalsignificance Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 13 ChapterFive Regressionresults Example:estimatingdemandforpizza demandforpizzaaffectedby 1.priceofpizza 2.priceofcomplement(soda) managerscanexpectpricedecreasestolead tolowerrevenue tuitionandlocationarenotsignificant ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 14 Regressionproblems solution:useadvancedcorrectiontechniques, suchastwostageleastsquaresandindirect leastsquares ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 15 Identificationproblem:theestimationof demandmayproducebiasedresultsdueto simultaneousshiftingofsupplyanddemand curves Regressionproblems Multicollinearityproblem:twoormore independentvariablesarehighlycorrelated,thus itisdifficulttoseparatetheeffecteachhason thedependentvariable solution:astandardremedyistodroponeof thecloselyrelatedindependentvariablesfrom theregression ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 16 Regressionproblems Autocorrelationproblem:alsoknownasserial correlation,occurswhenthedependentvariable relatestotheYvariableaccordingtoacertain pattern Note:possiblecausesincludeomittedvariables, ornonlinearity;DurbinWatsonstatisticisused toidentifyautocorrelation solution:tocorrectautocorrelationconsider transformingthedataintoadifferentorderof magnitudeorintroducingleadingorlaggingdata ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 17 Forecasting Examples:commonsubjectsofbusiness forecasts: grossdomesticproduct(GDP) componentsofGDP egconsumptionexpenditure,producerdurable equipmentexpenditure,residential construction industryforecasts egsalesofproductsacrossanindustry salesofaspecificproduct ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 18 Forecasting Agoodforecastshould: beconsistentwithotherpartsofthebusiness bebasedonknowledgeoftherelevantpast considertheeconomicandpoliticalenvironmentas wellaschanges betimely ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 19 Forecastingtechniques Factorsinchoosingtherightforecasting technique: itemtobeforecast interactionofthesituationwiththeforecasting methodology amountofhistoricaldataavailable timeallowedtoprepareforecast Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 20 ChapterFive Forecastingtechniques Approachestoforecasting qualitativeforecastingisbasedonjudgments expressedbyindividualsorgroup quantitativeforecastingutilizessignificant amountsofdataandequations ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 21 Forecastingtechniques Approachestoforecasting naveforecastingprojectspastdatawithout explainingfuturetrends causal(orexplanatory)forecastingattempts toexplainthefunctionalrelationshipsbetween thedependentvariableandtheindependent variables Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 22 ChapterFive Forecastingtechniques Sixforecastingtechniques expertopinion opinionpollsandmarketresearch surveysofspendingplans economicindicators projections econometricmodels ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 23 Forecastingtechniques Expertopiniontechniques Juryofexecutiveopinion:forecasts generatedbyagroupofcorporateexecutives assembledtogetherDrawback:personswith strongpersonalitiesmayexercise disproportionateinfluence ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 24 Forecastingtechniques Expertopiniontechniques TheDelphimethod:aformofexpertopinion forecastingthatusesaseriesofquestions andanswerstoobtainaconsensusforecast, whereexpertsdonotmeet ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 25 Forecastingtechniques Opinionpolls:samplepopulationsaresurveyed todetermineconsumptiontrends mayidentifychangesintrends choiceofsampleisimportant pollingandwillindicatenotonlywhythe questionsmustbesimpleandclear ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 26 Forecastingtechniques Marketresearch:iscloselyrelatedtoopinion consumeris(orisnot)buying,butalso whotheconsumeris howheorsheisusingtheproduct characteristicstheconsumerthinksaremost importantinthepurchasingdecision ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 27 Forecastingtechniques Surveysofspendingplans:yieldsinformation aboutmacrotypedatarelatingtotheeconomy, especially: consumerintentions Examples:SurveyofConsumers(University ofMichigan);ConsumerConfidenceSurvey (ConferenceBoard) ChapterFive inventoriesandsalesexpectations Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 28 Forecastingtechniques Economicindicators:abarometricmethodof forecastingdesignedtoalertbusinessto changesinconditions leading,coincident,andlaggingindicators compositeindex:oneindicatoralonemaynot beveryreliable,butamixofleading indicatorsmaybeeffective Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 29 ChapterFive Forecastingtechniques Leadingindicatorspredictfutureeconomic activity averagehours,manufacturing initialclaimsforunemploymentinsurance manufacturersnewordersforconsumer goodsandmaterials vendorperformance,slowerdeliveries diffusionindex manufacturersneworders,nondefense capitalgoods ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 30 Forecastingtechniques Leadingindicatorspredictfutureeconomic activity buildingpermits,newprivatehousingunits stockprices,500commonstocks moneysupply,M2 interestratespread,10yearTreasurybonds minusfederalfunds indexofconsumerexpectations ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 31 Forecastingtechniques Coincidentindicatorsidentifytrendsincurrent economicactivity employeesonnonagriculturalpayrolls personalincomelesstransferpayments industrialproduction manufacturingandtradesales ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 32 Forecastingtechniques Laggingindicatorsconfirmswingsinpast economicactivity averagedurationofunemployment,weeks ratio,manufacturingandtradeinventoriesto sales changeinlaborcostperunitofoutput, manufacturing(%) ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 33 Forecastingtechniques Laggingindicatorsconfirmswingsinpast economicactivity averageprimeratechargedbybanks commercialandindustrialloansoutstanding ratio,consumerinstallmentcreditoutstanding topersonalincome changeinconsumerpriceindexforservices ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 34 Forecastingtechniques Economicindicators:drawbacks leadingindicatorindexhasforecasta recessionwhennoneensued achangeintheindexdoesnotindicatethe precisesizeofthedeclineorincrease thedataaresubjecttorevisionintheensuing months ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 35 Forecastingtechniques Trendprojections:aformofnaveforecasting thatprojectstrendsfrompastdatawithouttaking intoconsiderationreasonsforthechange compoundgrowthrate visualtimeseriesprojections leastsquarestimeseriesprojection ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 36 Forecastingtechniques Compoundgrowthrate:forecastingby projectingtheaveragegrowthrateofthepast intothefuture providesarelativelysimpleandtimely forecast appropriatewhenthevariabletobepredicted increasesataconstant% Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 37 ChapterFive Forecastingtechniques Generalcompoundgrowthrateformula: E=B(1+i)n E=finalvalue n=yearsintheseries B=beginningvalue i=constantgrowthrate ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 38 Forecastingtechniques Visualtimeseriesprojections:plotting observationsonagraphandviewingtheshape ofthedataandanytrends ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 39 Forecastingtechniques Timeseriesanalysis:anavemethodof forecastingfrompastdatabyusingleast squaresstatisticalmethodstoidentifytrends, cycles,seasonalityandirregularmovements ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 40 Forecastingtechniques Timeseriesanalysis: Advantages: easytocalculate doesnotrequiremuchjudgmentoranalytical skill describesthebestpossiblefitforpastdata usuallyreasonablyreliableintheshortrun ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 41 Forecastingtechniques Timeseriesdatacanberepresentedas: Yt=f(Tt,Ct,St,Rt) Yt=actualvalueofthedataattimet Tt=trendcomponentatt Ct=cyclicalcomponentatt St=seasonalcomponentatt Rt=randomcomponentatt ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 42 Forecastingtechniques Timeseriescomponents:seasonality needtoidentifyandremoveseasonalfactors, usingmovingaveragestoisolatethosefactors removeseasonalitybydividingdatabyseasonal factor ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 43 Forecastingtechniques Timeseriescomponents:trend toremovetrendline,useleastsquaresmethod possiblebestfitlinestyles: straightLine:Y=a+b(t) exponentialLine:Y=abt quadraticLine:Y=a+b(t)+c(t)2 chooseonewithbestR2 Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 44 ChapterFive Forecastingtechniques Timeseriescomponents:cycle,noise isolatecyclebysmoothingwithamoving average randomfactorscannotbepredictedand shouldbeignored Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 45 ChapterFive Forecastingtechniques Smoothingtechniques movingaverage exponentialsmoothing workbestwhen: nostrongtrendinseries infrequentchangesindirectionofseries fluctuationsarerandomratherthanseasonal orcyclical ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 46 Forecastingtechniques Movingaverage:averageofactualpastresults usedtoforecastoneperiodahead Et+1=(Xt+Xt1++XtN+1)/N Et+1=forecastfornextperiod Xt,Xt1=actualvaluesattheirrespective times N=numberofobservationsincludedin average ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 47 Forecastingtechniques Exponentialsmoothing:allowsfordecreasing importanceofinformationinthemoredistant past,throughgeometricprogression Et+1=wXt+(1w)Et w=weightassignedtoanactual observationatperiodt ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 48 Forecastingtechniques Econometricmodels:causalorexplanatory modelsofforecasting regressionanalysis multipleequationsystems endogenousvariables:dependentvariables thatmayinfluenceotherdependent variables exogenousvariables:fromoutsidethe system,trulyindependentvariables Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 49 ChapterFive Forecastingtechniques Example:econometricmodel Suits(1958)forecastdemandfornewautomobiles R=a0+a1Y+a2P/M+a3S+a4X R=retailsales Y=realdisposableincome P=realretailpriceofcars M=averagecreditterms S=existingstock X=dummyvariable ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 50 Globalapplication Example:forecastingexchangerates GDP interestrates inflationrates balanceofpayments ChapterFive Copyright2009PearsonEducation,Inc.PublishingasPrenticeHall. 51 ... View Full Document

End of Preview

Sign up now to access the rest of the document