Analysis of the Best Current Practices · 0.6 09-05-2016 A. Chamula (EEA) Added sections...

DeliverableD1.1 DisseminationLevel(PU) ContractN.687895

30-06-2016 H2020-687895©ThePROFITConsortium Page1of119

Horizon2020ICT-INFORMATIONANDCOMMUNICATIONTECHNOLOGIES

PROmotingFInancialAwarenessandStabiliTy

H2020–687895

AnalysisoftheBestCurrentPracticesWorkpackageNo. WP1 WorkpackageTitle Requirementsand

FunctionalDesign

TaskNo. T1.1 TaskTitle BestCurrentPractices

MilestoneNo. -

Organization name of leadcontractorforthisdeliverable

DemocritusUniversityofThrace-DUTH

AuthorName(s) IoannisPraggidis,ΕiriniKarapistoli,TsintzosPanagiotis(DUTH), Andrej Chamula (EEA), Georgios Panos,Konstantinos Gkrimmotsis, Nana Oiza Akubelem(UoGLASGOW), Artem Revenko (SWC), AuroraProspero,Gian-LucaGasparini(FEBEA)

Reviewer(s) AnnaSatsiou(CERTH),AntonisSarigiannidis(DUTH)

Status(S:submitted;D:draft) F

Nature R-Report

DisseminationLevel PU-Public

Projectstartdateandduration January2016,36Months

DueDateofDeliverable:ActualSubmissionDate:

June30,2016June30,2016



RevisionHistoryVersion Date Modifiedby Changes0.1 10-01-2016 E.Karapistoli(DUTH) FirstversionofToCprovided

0.2 14-02-2016 I.Praggidis(DUTH) ToCrefinedandfinalised

0.3 15-02-2016 A.Revenko(SWC) Addedsection5

0.4 19-04-2016 E.Karapistoli(DUTH) Addedsection4.2

0.5 24-04-2016 I.Praggidis(DUTH) Addedsection4.1

0.6 09-05-2016 A.Chamula(EEA) Addedsections2.1.1-2.1.3

0.6.1 13-05-2016 A.Chamula(EEA) Addedsection2.1.4

0.6.2 22-05-2016 A.Chamula(EEA) Addedsection2.1.5

0.6.3 25-05-2016 A.Chamula(EEA) Addedsection2.2

0.7 27-05-2016 I.PraggidisandP.Tsintzos(DUTH)

Addedsections4.3-4.4

0.8 27-05-2016 A.ProsperoandG-L.Gasparini(FEBEA)

Addedsection3.4.1

0.8.1 31-05-2016 G.Panos,K.GkrimmotsisandN.Akubelem(UoGLASGOW)

Addedsections3.1,3.2,3.3,3.4.2-3.4.6

0.9 01-06-2016 E.Karapistoli(DUTH) Firstthoroughrevisionofcontributions.Fixedreferences.

0.9.1 05-06-2016 I.Praggidis(DUTH) FirstofficialpreliminarydraftversionofD1.1.Sendtoreviewers.

0.9.2 17-06-2016 A.Satsiou(CERTH) ReviewcommentssentbyA.Satsiou(CERTH)

0.9.3 20-06-2016 A.Sarigiannidis(DUTH) ReviewcommentssentbyA.Sarigiannidis(DUTH)

0.9.4 21-06-2016 G.Panos(UoGLASGOW) RevisionsaccordingtocommentsfromG.Panos(UoGLASGOW)

0.9.5 24-06-2016 E.Karapistoli(DUTH) Executivesummaryandconclusionsadded.FilledouttheAbbreviationstable.Respondedtoreviewers‘comments.

0.9.6 27-06-2016 E.KarapistoliandP.Tsintzos(DUTH)

Proofingandpolishing.

1.0 30-06-2016 I.Praggidis(DUTH) Finalreview,proofingandqualitycontrol.Readyforsubmission.



ListofAbbreviations

APR AnnualPercentageRateBoW Bag-of-WordsCCI ConsumerConfidenceIndexCEDEFOP CentrefortheDevelopmentofVocationalTrainingD2D Doorways2Dreams DSGE DynamicStochasticGeneralEquilibriumECB EuropeanCentralBankECM ErrorCorrectionModelsECO E-learningCommunicationOpen-DataEMMA EuropeanMultipleMOOCAggregatorESCO EuropeanSkills,Competences,QualificationsandOccupationsETF Exchange-TradedFundFCI FinancialConditionsIndexFIBO FinancialIndustryBusinessOntologyFIOS FinancialInformationObservationSystemFINRA FinancialIndustryRegulatoryAuthorityFLEC FinancialLiteracyandEducationCommissionGDP GrossDomesticProductHESC HigherEducationServicesCorporationLOD LinkedOpenDataMCI MonetaryConditionsIndexMOOCS MassiveOpenOnlineCoursesNLP NaturalLanguageProcessingOBI OpenBudgetIndexOECD OrganisationforEconomicCo-operationandDevelopmentOER EuropeanOpenEducationalResourcesOWL WebOntologyLanguagePCA PrincipleComponentAnalysisPOS Part-Of-SpeechSEC SecuritiesandExchangeCommissionSIC StandardIndustrialClassificationSKOS SimpleKnowledgeOrganizationSystemSVM SupportVectorMachinesVAR VectorAutoregressionWFFT WorkplaceFinancialFitnessToolkit



ListofTablesTable3–1Themainfinancialliteracyquestionsandtheirextensions...................................29Table3–2TheOECD's8financialliteracyquestions..............................................................30Table3–3Internationalcomparisonoffinancialliteracy.......................................................31Table3–4Listofindividualfinancialinstitutionsengagedinfinancialeducation.................45Table3–5Listoffinancialassociationsengagedinfinancialeducation.................................47Table3–6ListofonlinefinancialeducationinitiativesintheEuropeanUnion.....................50Table3–7Listofotheronlineresources................................................................................53Table3–8Listofnewspaperswithpersonalfinancesections................................................55Table3–9Listoffinancialliteracy/personalfinanceopenonlinecourses.............................58Table3–10Listoffinancialliteracygames.............................................................................60Table3–11Listoffinancialliteracyandpersonalfinancebooks...........................................63Table3–12PROFITuser-groups.............................................................................................66Table4–1Anoverviewofthesentimentanalysisapproaches..............................................84



ListofFiguresFigure2–1TheWallStreetJournal........................................................................................11Figure2–2EstimatedGDPforecastsinWSJ...........................................................................12Figure2–3TheinspiresectioninCNBC..................................................................................13Figure2–4TheopinionchannelinForbes.............................................................................14Figure2–5ThebusinesssectioninNewYorkTimes..............................................................15Figure2–6TheCNNeconomysection....................................................................................15Figure2–7Thefear&greedindicatorsinCNNMoney..........................................................15Figure2–8Thee-axeswelcomescreen..................................................................................16Figure2–9TheScientific&AcademicPublishingwelcomescreen........................................17Figure2–10TheFirstEUProjectanditsSentifyPortal..........................................................19Figure2–11TheBörseFranfurktwelcomescreen.................................................................20Figure2–12TheU.S.SecuritiesandExchangeCommissioncompanyfilings.........................21Figure2–13Thechartadvisorarticlesininvestopedia.com..................................................22Figure2–14Themymoney.govborrowsection.....................................................................24Figure3–1Financialliteracyaroundtheworld......................................................................32Figure3–2FinancialliteracyintheEU-28..............................................................................32Figure3–3TherelationshipbetweenfinancialliteracyandGDPpercapita..........................34Figure3–4TherelationshipbetweenfinancialliteracyandGDPpercapitaintheEU-28.....35Figure3–5DistributionoffinancialliteracyinGreatBritainandothercountries,bygender38Figure4–1Thesentimentanalysisworkflow.........................................................................81Figure4–2SentimentanalysisinPROFIT...............................................................................85



TableofContents1 Introduction.................................................................................................................9

1.1 Scopeofthedocument.....................................................................................................91.2 Structure..........................................................................................................................10

2 FinancialPlatforms....................................................................................................112.1 Typesofplatforms...........................................................................................................11

2.1.1 Classicaleconomicmedia...............................................................................................112.1.2 Generalmedia................................................................................................................142.1.3 EconomyEnvironment....................................................................................................162.1.4 StockexchangeandsecuritiesWEBsites.......................................................................192.1.5 Educationalwebsites......................................................................................................22

2.2 WEBInformationServices................................................................................................242.2.1 Education........................................................................................................................242.2.2 User’sportfoliomanagement.........................................................................................252.2.3 Communities...................................................................................................................252.2.4 EconomyEnvironment....................................................................................................26

2.3 ThePROFITapproach.......................................................................................................263 FinancialLiteracy.......................................................................................................27

3.1 Theassessmentoffinancialliteracy(ASSESS)...................................................................273.2 Financialliteracydemographics(DESCRIBE).....................................................................363.3 Theimportanceoffinancialliteracy(ATTRIBUTE).............................................................403.4 Financialliteracytoolsandtoolkits(COMPARE)...............................................................42

3.4.1 Financialinstitutionsengagedinfinancialeducation.....................................................423.4.2 Newspaperswithonlinepersonalfinancesections........................................................543.4.3 Openonlinecourses(MOOCs)........................................................................................563.4.4 Financialgamesandgamifiedtools................................................................................583.4.5 Books..............................................................................................................................623.4.6 Europeaneducationalplatforms....................................................................................62

3.5 AsynthesisofbestpracticesandthespaceforPROFITnovelty(SYNTHESIS/NOVEL)........644 FinancialandEconometricForecasting.......................................................................68

4.1 MarketSentimentIndicators...........................................................................................684.1.1 Theoreticaljustificationsofsentimentaffectingtheeconomy......................................684.1.2 Themainandmostcurrentsentimentindexesusedsofarintheliterature.................714.1.3 Therelativeimportanceofsentimentindexesinpredictingthefinancialeconomy.....764.1.4 Concludingremarks........................................................................................................79

4.2 MarketSentimentAnalysis..............................................................................................794.2.1 TheSentimentAnalysisWorkflow..................................................................................804.2.2 State-of-the-ArtinSentimentAnalysis...........................................................................804.2.3 SentimentanalysisinPROFIT..........................................................................................84

4.3 ForecastingModels..........................................................................................................854.3.1 Theinterlinkbetweenfinancialsectorandtherealeconomy.......................................864.3.2 Forecastingmodels.Whathavewelearned?.................................................................894.3.3 ABriefDescriptionoftheForecastingEstimationMethodsinPROFIT..........................924.3.4 ForecastingEstimates.....................................................................................................93

4.4 FinancialandeconometricforecastinginPROFIT.............................................................945 FinancialLinkedOpenData........................................................................................95

5.1 LinkedOpenData(LOD)...................................................................................................955.2 AvailableVocabularies.....................................................................................................96

5.2.1 Abbreviations..................................................................................................................965.2.2 STWThesaurusforEconomics........................................................................................97



5.2.3 EuropeanSkills,Competences,QualificationsandOccupations(ESCO)........................985.2.4 EuroVoc...........................................................................................................................99

5.3 AvailableFinancialLinkedOpenDataResources............................................................1005.3.1 BudgetsandSpendingData..........................................................................................1015.3.2 CorporateReports........................................................................................................104

5.4 Financial-specificLODandthePROFITadvances............................................................1086 Conclusions..............................................................................................................1097 References...............................................................................................................110



ExecutiveSummaryTheneedsforadvancementoffinancialcapabilityandenhancementoffinancialawarenessamonginformedcitizensandmarketparticipantshavebeenidentifiedasamajortargetforimprovedsocialperformance,clientprotectionand,ultimately,greatersocietalwell-being.Acknowledgingtheseneeds,thePROFITprojectanditsplatformaredesignedasasolutioncatering to the exact need of action enhancement for greater financial awareness andcapability.ThisdeliverabledescribesthebestcurrentpracticesthatarerelevanttothePROFITproject.Themaingoalofthisdeliverableistwofold;(a)toreportonthecurrentadvancesintheareaof financial literacy, financial and econometric forecasting and open linked data forsupporting financial decision-making, and (b) to highlight the advances introducedby thePROFITconcepts.The deliverable is divided into threemain parts. The first part reviews themost popularfinancial awareness platforms that are available online and provides an overview of theservicessupportedbythem.Themainpurposeofthefirstpartistogivethereader(incl.thePROFIT industrialpartners) theunderstandingofwhat is the current stateof theart (i.e.,capabilities,limitations)relatedtothePROFITplatformtechnologies.Thesecondpartof thedocument focuseson the twomainscientificpillarsof thePROFITproject, namely the financial literacy and financial forecasting components. Initially, thesecondpartreviewsthebestcurrentpracticesintheacademicandprofessionalagendaonfinancial literacy as well as the currently available financial literacy tools and toolkits.Subsequently, it describes the main market sentiment indicators used in the empiricalresearchandhowtheseindexesareextractedusingsentimentanalysistechniques.Finally,this part reviews the most prominent models used to forecast well-known econometricindicators.ThemainpurposeofthispartistoacquaintthereaderwiththecurrentstateoftheartinfinancialliteracyandfinancialforecastingandtopresentthenovelinsightsofthemethodsandtoolkitstobedevelopedwithinPROFIT.Thefinalpartofthisdeliverablepresentstechnologiesemployedintheenvisagedfinancialawareness platform, including the financial linked open data. Existing thesauri arepresented, followed by a discussion on semantic resources and evolving ontologies. Themainpurposeof thispart is togivethereaderan ideaofhowthePROFITontologymightlooklikeattheend.



1 Introduction

Therecent financialcrisishasgenerated interest inbetterunderstandinghowtopromotemoreresponsibleandprudent individualsavingandborrowingbehaviour (Klaper,Lusardi,and Panos, 2013). The Eurozone debt crisis, with the resulting fiscal consequencesthroughout Europe, further stimulated the discussion on how to promote the efficientallocationoffinancialresourcesandgreaterfinancialstability(LusardiandMitchell,2014),(Lusardi, Michaud andMitchell, 2015). The ability of citizens to make informed financialdecisionsiscriticaltodevelopingsoundpersonalfinance,whichcancontributetoincreasedsavingrates,moreefficientallocationoffinancialresources,andgreaterfinancialstability.Technologicaldevelopmentshaveenabledandenhancedtheavailabilityoflargevolumesofinformation on themes relevant to financial decision-making. Their potential benefits,though, are hindered by the cognitive limitations by individuals when it comes to theprocessingoflargevolumesofinformation,aswellasthedocumentedwidespreadfinancialilliteracyevenwithindevelopedeconomies,includingthoseoftheEuropeanUnion.Acknowledging such needs, the PROFIT (Promoting Financial Awareness and Stability)projectbringstogetherresearchersandprofessionalswithexpertiseacrossaccountingandfinance, economics, information technology, computer/software engineering andeducation,alongwitharangeofprivate,third-sectorandinstitutionalpartners,todevelopafinancial awareness and capability platform for improved social performance, clientprotectionand,ultimately,greaterfinancialstabilityandsocietalwell-being.PROFIT will develop a platform built on Open Source components with aforementionedfunctionalities that will be pilot-tested via the collaboration with the members of theEuropeanFederationofEthicalandAlternativeBanks(FEBEA),aninstitutioncommittedtotheresponsiblebankingandfinanceagenda.TheoutcomesoftheprojectareexpectedtoenableinferencesandspecificappropriatepracticesthatcanbemadeavailabletothewidepublicintheEuropeanUnion.

1.1 Scopeofthedocument

ThisdeliverableprovidesthebaselinedefinitionforcommencingtheworkofWorkPackage1(WP1-RequirementsandFunctionalDesign).TheoverarchinggoalofWP1istostudyandspecify the requirements of the targeted users with respect to the proposed financialawarenessplatform,analysingandturningthemintospecifications.Basedonthisanalysis,itwilldefineanddeliveranumberofrepresentativeusecasesanduserscenariostoexemplifythenoveltiesofthePROFITplatform.Finally,thisworkpackagewillsettheframeworkforplanningtheassessmentactivitiestowardsvalidatingthePROFITtools.In particular, this deliverable will provide a state-of-the-art analysis on the best currentpracticesthatarerelevanttothePROFITproject.Themaingoalistoreportonthecurrentadvances in theareaof financial literacy, financial andeconometric forecasting andopenlinked data for supporting financial decision-making, and to highlight the advancesintroducedbythePROFITconcepts.



1.2 Structure

Thedeliverableisorganizedasfollows:

• Chapter1introducesthedeliverableanddefinesitsscopeandstructure.

• Chapter 2 reviews the various types of financial awareness platforms that areavailableonlineandoutlinestheirservices.

• Chapter 3 reviews the best current practices in the academic and professionalagendaonfinancial literacy. Itbeginsbyprovidingaconceptualoverviewassessingthe importance of financial literacy for economic behaviour of all sorts. It thenreviews and compares the currently available financial literacy tools and toolkits.Finally,itpresentsthenovelinsightsofthePROFITfinancialeducationtoolkit.

• Chapter 4 describes the main market sentiment indicators used in the empiricalresearchandhowtheseindexesareextractedusingsentimentanalysistechniques.Then, ithighlightsthe interlinkbetweenthefinancialsectorandtherealeconomy.Finally, it reviews themain forecastingmodels used in the empirical research andprovidesinformationabouttheirpredictiveperformance.

• Chapter 5 introduces the emerging concept of linked open data and how thisconceptisappliedinPROFIT.

• Finally,Chapter6drawsthemainfindingsandconclusionsofthisdeliverable.



2 FinancialPlatforms

There is wide range of websites that aspires to the term financial awareness platformranging from yahoo personal finance to classical economic media (i.e., the Wall StreetJournal,theCNBC,Forbes,etc.),generalmedia(i.e.,theNewYorkTimes,CNN,etc.),stockexchangeandsecuritieswebsites(i.e.,TheNewYorkStockExchange,BörseFranfurkt, theLondon Stock Exchange, etc.) as well as encyclopedic websites (i.e., investopedia.com,mymoney.gov,about.com,e-axes,etc.)andspecializedfinancial literacyeducationtoolkits(i.e., the Workplace Financial Fitness Toolkit (WFFT), funded by NYSE and the EuronextFoundation).Forbetterunderstandinghowtheunlimitednumberofthesecommunicationmediaworks,wenextprovideanoverviewofthedifferenttypesofplatformsoutliningtheservicessupportedbythem.

2.1 Typesofplatforms

2.1.1 Classicaleconomicmedia

Thisgroupofportalsaresimilartostandardmediaformats.Theysharetohighextendthestructurerelatedtohomeandworld’snews.Thesitestructurecontainstheusual“Home”,“World”, “U.S.”, and “Politics” items.On theother side, it is notunusual to find columnsrelated to Life and Art, for example, Buddhist Treasures from Gobi Desert, or postsdescribingTVepisode“GameofThrones”.Themainfocus,however,remaintofinancial&economyeventsandanalysis.Next,wetakeabrieflookonselectedsites,particularlyWallStreetJournal,CNBC,andForbes.WSJisfocusedtonewsfromU.S.andWorld’sEntrepreneurenvironmentandfinance.

Figure2–1TheWallStreetJournal

Its Policy contains Blog section, consisting from traditional Washington Wire feature,providingthelookbehindthestoriesandwarningsofwhattowatchinthedaysahead.Thesectioncontainsalso“ThinkTank”,ahomeforoutsideanalysisformpoliticalthinkers.Apartofblogs, the video sectionoffers clipsof selectedpoliticalmoments, likehighlightsfrompresidentspeech,politicalcauses,election,etc.ThelastsectioninPoliticsisdedicatedtoNewsPollsaboutElection.SimilarsectioniscoveredbyEconomy.Againablog,named“RealTimeEconomy”.ItoffersanalysisandcommentaryonU.S.andglobaleconomy,centralbankpolicyandeconomics.GDPgrowth,actualeconomicevents like job reports, resultsofmanufacturing sector, carsales,tradedeficitispresentedwithaccompanyingnumbersandgraphs.



Interesting section in Economy is related to Forecasting survey, displaying the estimatedGDP forecasts and actuals on 3, 5, and 7year basis. The survey is performed byWSJ, incooperationwithmorethan60economistsonmorethan10majoreconomicindicatorsonamonthly basis. The indicator includesOil prices,Unemployment, 10 YearNote, CPI index,FederalFundRate,andRecessionProbability.

Figure2–2EstimatedGDPforecastsinWSJ

And, similarly to Politics, there are also videos in Economy section. Short clips capturevarious news from stamp price, FED comments, to Vietnam Investing in Social GoodChallenges. The core info channels are under site structure items “Business”, “Tech”,“Markets”,and“Opinion”.Ratherthanitemizedviewthepurposeofserviceisprovidedinthenextchapter,inconjunctionwithothergroupsites.“Arts” and “Life” coversobligatory topics aboutbooks, films, health, food&drink, sportsetc.WSJ has close relationship to Barron’sweb site, which is primarily dedicated to financialmarkets, representing investment ideas to various securities, and providing the advisorsservice, service for managing user’s portfolio, and notification of important economicevents.Theadvisorsandinvestingideasarepaidservices.CNBCChannelhas its“sister” inWEBsite.Similarly, thesite isorientedprimarilytoworldandU.S.markets.BesidesobligatoryNewssections,thesiterefersto“FinancialMarkets”,“Investing”,and“Tech”,moreelaboratedinServicesview.Interestingsectionis“Makeit”,astorytellingplatformforthosewhowantsmore–startingownbusiness,advancethecareer,lookingforinspiration.Thesectionbringssmallbusinessnews,headlines,informationaboutstartups,toinspirethereaderstomaketheirownwaytomakeabusiness.The section is furtherdivided into “Inspire”, “Guide”,and“InDepth”



chapters.The“Inspire”containsexamplesofsuccessfulbusiness,usuallybasedoninspirationalideas.Forexample,makingthefashiononselfies,clothesforoutdoorexercisesthatcanhidethesweat,standingdeskforchildrentomakehomeworkfun,etc.The“Guide”chapterfocusesmoreon“howto”themes,likestartingthebusinesswithlessthan$1000,tipsforfreshgraduateswhoarelookingfornewjob,or,advicesfromlegends,likeAppleInc.founderSteveWozniakhowtoapproachthestudy.The “In Depth” contains articles grouped by themes like “How I made it”, “Better yourbusiness”,etc.,whichresemblesabitthe“Inspire”.Thesectionalsocontains“Attend”chapter,with links toConferences related tobusiness,presenting founders, CEOs of successful companies, and other entrepreneurs that sharetheirstoriesandadvises.

Figure2–3TheinspiresectioninCNBC

Forbeswelcomesitsreaderswithdailycitationsoffamouspeople–QuoteoftheDay.Aftercouple of seconds the main page will be shown and, according to Forbes motto “TheCapitalistTool”,theflyingBillionaireTickerinthetopofthepageshowsactualearningsofrichest people in the world. The main page is full of Top list, starting from Billionaires,Celebrities,Colleges,toPowerfulwomen.The “Entrepreneur” section is keeping the “Capitalist spirit” and provides the list of bestsmall U.S. companies, best Cities to launch a business, best Franchises to buy, etc.Moreinterestingarticlesarelinkedunderchapterslike“ExitStrategies”thatexplainwhyStartupsusually fail, “Taxes and Law”, with articles related to regulation topics, or “Sales and



Marketing”and“Startup”withsuccessstoriessimilartoCNBS“Inspire”.

Figure2–4TheopinionchannelinForbes

The“Opinion”representsthenewsgroupedtovariouscategories,similartoNewsinothersites.The“Leadership”(apartofcommerciallinks)informsaboutcareeradvices,corporateresponsibility topic, leadership styles, and sales skills. The C-level leadership articles aregroupedinthe“Networks”,startingfromExecutive,throughFinancialtoMarketingOfficers.Theyoungbusinessmenarecapturedinthe“Under30”Section,and“Lists”containsthetopliststypicalforForbes.Forbes investment information services are hidden under “Newsletter”. The sectioncontains recommendations for people who want to invest into securities. Articles arelocked,couldbeaccessedbysubscribingtheservice.Obligatory “Life”, “Video”, and“Magazine”, “Newsletter”encloses theForbes informationportfolio.

2.1.2 Generalmedia

Financeandeconomyiscommontopicingeneral informationmedia,oftencoveredinthe“Business” section. The selected elaborated here are The New York Times, CNN, andBloomberg.TheNewYorkTimesBusinessstartswitheconomyrelatednews,e.g.U.S.jobreport,eventsaboutissuingorclosingbigfunds,informationaboutbigI.P.O.,strikesinworld’scountries,etc.Thenews isgroupedbyusualcategories–“International”, “Technology”, “Economy”,“Energy”, etc. The “Dealbook” is focused on news about mergers and acquisitions, anddivestitures.“YourMoney”sectionbringsnewsandcommentaryaboutinvestments,loans,andretirement.However, thesearenotspecializedservices,butkindofviews,or filtertoarticles,thatyoucanapproachalsofromothersiteplaces.



Figure2–5ThebusinesssectioninNewYorkTimes

CNNeconomysection is“Money”.The initialpagebrings“TopGlobalNews”, forexamplesituation about Zimbabwe, Tech company product news, big stock moves, etc. The sub-sectionsprovidedetailedinfoabout“Markets”.Namely“Pre-MarketTrading”bringsAgencynewsaboutpre-marketmovers,overviewofthecommodity,bonds,andinterestrate.

Figure2–6TheCNNeconomysection

“MarketMovers”sub-sectionshowstopgainers,losers,andmostactivestocks,the“WorldMarkets”depictstheselectedmarketdatainAmerica,Asia,andEurope.

Figure2–7Thefear&greedindicatorsinCNNMoney



“Investingguide”doesnoteducatehowto invest,asoneshouldget impression fromthename, but focuses on news related to investment. For example, silver and oil pricecommentary,DowJonestrends,theoilproducers’contracts,etc.Interestingfeatureiscoveredby“FearandGreed”index,whichreflectsthecurrentmoodonthemarket.Theindexistrackingsevenindicatorsofmarketsentiments,namelyvolatilityindex, following the upward and downward market trend, price momentum, the actualmovementagainst125-dayaverage,pricestrength,pricebreath,putandcalloptions,junkbonddemand,andsafeheavendemand.

2.1.3 EconomyEnvironment

Sitesinthiscategoryofferthereaderstheviewbasedoneconomycontext,understandingthegeneralprinciples,andaskingthequestions,ratherthanclassicalconsumer’sview.e-axesisdedicatedtoprovidinganalysis,researchinfo,commentary,andbookreviews.The“Home” section contains links to recent articles, the topics include Economy growth,Monetary policy, Eurozone debt crisis, Global inflation, etc. Every article is presented bybriefcommentaryandtheTitleofresearch,underwhichthearticleispublished.“Member’s posting” refers to actual contribution of e-axesmembers, categorized to fourmain groups – Commentary, Working papers, Books, and Conferences. The themes arereflecting wide range of various topics, including corruption in development countries,urbanproblemsinAsia,impactofwaterlevelinreservoirtoSingaporeimport,etc.

Figure2–8Thee-axeswelcomescreen

The “Books” section provides reviews of economy books, including Academy sources likePrincetonUniversity Press,OxfordUniversity Press,MIT Press, etc. The site provides also“Calendar of Conferences” with coordinates on date, place, organizer, and keynotespeakers. Examples of upcoming events are “Impact of uncertainty shocks on the globaleconomy”, “Annual conference on Macroeconomic Analysis”, and “EBES JournalConference”scheduledforMay2016.



Apart of the site sectionswithmembers’ contribution, the site refers to three additionalsections referring to International Blogs, Global Research, and 360o Economy View. The“Blogs” section contains selection of International articles. For readerswantedmore infotheoriginallinkisincludedattheendofarticle,sothatallcomments,authors’credentials,relatedcolumns,etc.,areeasilyaccessed.“Global research” archives the research results, usually in PDF form that could bedownloadedbyreader.DocumentsareintroducedbyshortAbstract,andusualcoordinatesreferring to Author, Date, and Affiliation. For example, a Deutsche Bank Affiliate Authordescribes the innovation in financial sector brought by digital currencies, elaborating thepotentialopportunitiesandrisksarisingfromtheglobaladoptionofdigitalcurrencies.“360oEconomyView” is an interestingway to file the commonarticles. Forexample, thetopic“Newandnoteworthybooks” lists theselectedreviews inonearticle, focusingon6titles. Other theme “On economy Rents and Inequality” appoints to 2 links, which arerelevanttohighindustryconcentrationandtheriseininequality,twootherlinksrelatedtopersistenteconomicrents,andattheendthelinktostatistic.The selection is prepared by e-axes team and the focus is to choose the most relevanttopics,insteadofhugenumbersoflinks,whichsavesthetimeofreaders,whicharelookingforparticularthemes.Scientific and Academic Publishing is the open access publisher covering the academicdisciplines. Its chapter “Business and Management” contains economy related Journals“Journal of Economics”, “Finance and Accounting”, and “Microeconomics andMacroeconomic”.

Figure2–9TheScientific&AcademicPublishingwelcomescreen



TheJournalscontainarticles inHTMLandplaintextformat. It ispossibletodownloadthePDF.OlderstuffisclearlyfiledinArchivesection.The Journal of Economics employs economics to analyze problems in business, consumerbehavior, and public policy. It publishes theoretical and empirical papers in finance andrelated themes, for example consequences of adopting IFRS system, relation betweenfinancingtheeducationandunemploymentrate,roleofCorporateSocialResponsibility inCommunitydevelopment,policyinstrumentsandgrowthsustainability,etc.Articlesdonotavoiddevelopmentcountries,likeGhana,Nigeria,Tanzania,orCongo.Journal of Finance and Accounting is dedicated to publishing the empirical research inFinance and Account theory. Empirical approach includes quantitative, qualitativelaboratory,andcombinationmethods,appliedinAuditing,Accounting,Banking,Financing,Investing,andTaxfields.Featuredtopicsare,forexample,ExploringSustainablestrategiesfor Banks, Assessment of Factors affecting Tax morale, Monetary policy and Inflation,ForecastingStockMarketVolatility,EmpiricalInvestigationoftheDebtMaturity,etc.Microeconomics andMacroeconomic Journal introducesbasicmicroeconomic concepts toanalyze issues in business, and to disseminate papers of practical implication for publicpolicy,businesspolicy,andindividualdecisionmaking.Articlescovertopics, like Impacton InflationonStockPricesbyevaluatingthepositivevs.negative impact of inflation,Mutual Interdependence between Human Action and SocialStructureinEconomy,EffectofSupplyandDemandinControllingGDP,andsoon.FIRST Project Portal is a web site, which addresses challenges of growing amounts ofheterogeneousdataandinformationinfinancialmarkets.Forthatpurpose,FIRSTdevelopsandprovidesanInformationandCommunicationTechnology(ICT)infrastructuretocollectand process data to integrate it into a financial knowledge base for further analysis anddeveloping online event detection and prediction models, visualization models, anddecision-supportmodelsthatwilldeliverpertinentinformationtothedecisionmaker.Themain page of the site describes the FIRST objectives, and the Technical details. Thisincludesthecurrentstateoftechnology,theAdvancements,whichdescribeingreaterdetailimplementationof the technical realization, andScaling strategyofdevelopmentprocess.Othersectionsofmainpagecovers:Project Results refers to integrated financial market information system, and relateddeliverables including Ontology, Information Integration, and Data AcquisitionInfrastructure,SemanticExtractionsystem,andDecisionSupportSystem;CaseStudiesintroducescomplementaryandvisionaryusecasesofmarketsurveillance,riskmanagement,andonlineretailbrokerageandinvestmentmanagement.



Figure2–10TheFirstEUProjectanditsSentifyPortal

The Blog section contains articles about PROFIT – technology related news like spamdetection, sentiment extraction from web documents, ontology learning, as well projectachievements and usage. The section “News” refers to press releases, showcases, andactual project presentation; “Dissemination Material” contains deliverables like FinancialMarket Information Systems GUI, Visualization of Textual Streams, Large-scale ontologyreuse,Qualitativemodels,etc.The“Partners”liststhelistofpartiesinvolvedintheproject,and“Links”showsFIRSTinsocialnetworks,publishinginothersites,andrelatedproject.

2.1.4 StockexchangeandsecuritiesWEBsites

Inthissection,webrieflyoverviewsiteswithdirectaccesstothesecuritiesmarketandtheyenvironment.ThelistofsitescoversU.S.andEuropeanregions.TheNewYorkStockExchangeportal isapartof IntercontinentalExchangecompanywebsite.Thefirstserviceinthelistis“DesignatedMarketMaker”forbrokers,whorepresenttheinterest of financial institutions, pension funds, banks, etc. Other services for publiccompaniesincludesInvestorRelationsprograms,Corporategovernancetomanagerisksandenhanceperformance,Marketingandpublicrelations,andNYSENetwork,anentrypointtocrosscountrydiversifiedcommunity.NYSE further offers services for providing information about Stock Exchange Trades andData(Real-Time,Historical,Summarydata,ReferenceData).The“Insight”Sectioncontainseconomyrelatedarticles,forexamplegrowingdemandforGreenEnergy,asaresultofParisAgreement Climate Change Conference, Briefings from Customer Conferences, thecomments related to technology trends like network security, etc. The last section



“Regulation”providestheinfoaboutcompanies listedinStockExchange,andrules,whichcover access and communication with Floor, Listing & Delisting rules, Arbitration rules,ConductsandDisciplinaryrules,etc.Börse Franfurkt focuses more on providing the online information about market. Thesections are filled by types of securities, e.g. Equities, Bonds, ETF, Funds, Commodities,CurrenciesandSustainableSecurities.

Figure2–11TheBörseFranfurktwelcomescreen

Everysecuritiessectionhassubsections,whichenabletheusertogetspecificinfoaboutthesecurities, for example “Overview”, “Search” service to find specific security based onCountry,typeofsecurities (Equity,ETF,Fund,Commodities,…),“Newissues”ofsecurities(InitialPriceOffering),“Indices”,ifrelevantforgivensecurity.Ontopofsecuritiessections,therearetwoother,“News”,and“Knowhow”.“News”referstoDividend/Interest information, commentaryaboutworld’s Stockexchange,or technicalinfolikenewtools,apps,etc.“Knowhow”sectioncontainsquitecomplexglossaryofStockExchangeandfinanceterms,starting fromdefinitionofAAARating,MarketValue,MicroCap, Sector Index, Securities,etc.etc.U.S.SecuritiesandExchangeCommissionisthewebsiteofU.S.regulator(SEC).Itsmissionistoprotectinvestorsandtoactinconjunctionwithpromotingthecapitalformationthatisnecessarytosustaineconomygrowth. SECisdrivenbythesimpleconcept–all investors,largeinstitutionsorsmallprivateindividualsshouldhaveaccesstoimportantfacts,before



theydecidetobuythesecurities.

Figure2–12TheU.S.SecuritiesandExchangeCommissioncompanyfilings

Thesiteprovidesfourmaintypesofinformation:• SEC organization; the section “About”, “What we do” describes the Mission, the

Divisions are described under “Division” section, describing the Function, Structure,News,Regulatoryactions,Lawsandrules.TheSEChasfollowingdivisions:CorporateFinance, Enforcement, Economic and Risk Analysis, Investment Management, andTradingandMarkets.SECalsoexplainshow itsactivitiesareundertaken,specificallyhowInvestigationswork,withlistofmostcommonviolations.

• SECActivitiesunder“Enforcement”section,providingtheinfoabouto Litigation Releases, from actual date to year 1995. The Release contains the

typicallyFraudsandotherActviolations.o The notices and orders concerning the institutions and settlement of

Administrativeproceedingstoviewthesignificantpleadingsanddecisions.o CommissionOpinionsandAdjudicatoryordersissuedbySECo AuditingEnforcementReleaseso TradingsuspensionsbasedonU.S.Securitieslawsupto10workingdaysifpublic

interestorprotectionofinvestors’interestshastobedefended.• LawsrelatedtoSEC

o listing the Acts that govern the securities industry, starting from the first Act1933aftertheGreatDepression,torecentSarbanes-Oxleyfrom2002andJOBSACTfrom2012,

o InformationaboutRulemakingactivities,proposed/interim/finalrules,o Educationmateriallikefastanswers,howtoreadfinancialstatements,glossary

ofcommonTerms,recommendationhowtochecktheInvestmentadvisersand



Brokeragefirms• The Edgar Database of public trading companies, with regular financial statements,

proxystatements,andotherfillingdocuments

2.1.5 Educationalwebsites

Theoverviewoffinancialsiteswillbeenclosedbyeducationalwebs.Threerepresentativeshavedifferentapproachtoeducation.Investopediacoverswiderangeofterms,uptoveryspecifictopicsinterestingalsoforfinancialprofessionals.About.comisageneraleducationsite.ItsMoneysectionisapproachingthewidepublic.And,inthemiddleofthisapproachisMyMoney.org, supportedbyFederalFinancial LiteracyandEducationCommission,and itsmissionistostrengthenfinancialcapabilityandaccesstofinancialservicesforU.S.citizens.Investopedia.commainpage is similar toclassiceconomywebsites,e.g.brings thenewsand comments related toU.S. andworld economy, commenting the future FEDdecisionsabout interestrates,U.S. imports,specialreportsaboutvarious investmentsecurities,etc.The Investopedia’s sections are more focused on various topics, mostly for potential oractualinvestors:- Markets,bringstheinvestmentcolumns,withsubsections

o ETCCenter isoriented topics likemost traded funds,developmentof commodityETFprices,bestdividend-payingfunds

o Analysis andOpinions evaluates the situation in the financialmarket, brining thenews like shares buyback in certain company, opening new subsidiaries, Traders’trends,stockanalysis,similartoCNBCportal

o Sectorsbringsthesnapshotofactualbestandworstperformingeconomicsectors,withrelatedarticles

o Chart Advisor brings various trends in the selected charts, which try to helppredictingthefuturedevelopmentofsecuritiespricesbythepatterns.

Figure2–13Thechartadvisorarticlesininvestopedia.com



The remaining sections are dedicated to explaining the meaning of various technicalindicators,videosexplainingthetradingstrategies,andsoftwarereviewforperformingtheTechnicalanalysis.

- Retirement section brings articles in categories Investing, Living, Insurance andHealth, and Retirement planning. The content is not too much different frominvestingtopics,withaddedrecommendationtopicslike“Whatcountriesmakesyoufeel like millionaire on $3000/Month”, “Avoid Insurance Policy Pitfalls”, “How topreparetohandleNaturalDisasters”etc.

- InvestingsectionbringstheTutorialaboutinvesting,whiletherestofthecontentisfilledby investingnews.Similarapproachcouldbeseen intheSections“ManagingWealth”, “FinancialAdvisor”, “PersonalFinance”,etc.Themore interestingpart isFilingTaxesSection,withtoolsforcalculatingthetax,taxtips,andnewsrelatedtotaxlimitsanddeadlines.

The educational stuff ismore covered in theGuides and Exam preparation, in Referencesection. The Guides are categorized up to 20 divisions, including Industry handbook, theshort describing the terms like Industry, Sections, Federal Reserve system description,MacroandMicroeconomictopics,etc.The Exam preparation contains list of guides to help passing the FINRA, CFA/CFP/CMTcertifications, and Canadian professional exams. The guides are either self studies onlinetextbooks,orreferencestobooks,whichcouldbeobtainedfromreseller.Investopediacontainsalsosimulatorforstockandforextrader(topicisoutofscopeofthisdocument). On the other hand, Advisor insight contains answers to questions issued bypubliccitizens,manyofthemfacespractical lifesituationsabouttaxesanddivorcedpairs,contributiontopensionfunds,moneywithdrawalsfromfunds,howtoplanRetirementfor40-year-oldindividual,etc.Mymoney.gov is chartered by U.S Federal Financial Literacy and Education Commission(FLEC). Iteducatesthewidepublicbyfiveprinciples-Earn,Save&Invest,Protect,Spend,Borrow. Every principle is explained, suggested actions to be taken, and provide Tips &Hints:The educational web site is mapping the links within three streams for supporting thefinancialknowledge:

- Resources for Youth contains games, fun activities, and videos. For kids, there arelinks likeTheCollectorsClub turns the collecting thepocketmoneyandcoins intohobby. The game where goal is to make the longest burning light bulb refers tocommemorative silver dollar, dedicated to Thomas Alva Edison. The C.h.a.n.g.e.project tries to creating habits and financial awareness by advising on moneyspending and saving, introducing the compound interest concept, explaining howmuchcoststhesliceofpizza–ifsomeonewillpay$2forpizzasliceeveryweek,thecostforitis$5200in50years,e.g.untilretirementage.Savingthesameamountat8%ifinteresthoweverbrings$64678,whichistherealpizzacost.



- Resources for Teachers and Educators refers to federal guides and curricula forteaching the financial capability concepts. The Online Resource center containsvideoscoveringtopics likeSavingsgoalsandAccounts,Understandingthe financialinstitutions,Shoppingeffectively,etc.

- Resource for Researches enables searching the Research Clearinghouse, whichcontains the FLEC Reports and Articles from federally supported research onfinancial capability topics. The resources are categorized around mentioned fiveprinciples.

Figure2–14Themymoney.govborrowsection

About.com is general educational site, covering also the financial topic – About Moneychannel. It contains the articles categorized by topics like Insurance, Retirement, Smallbusiness,Savings,etc.Forexample, theCar Insurancecontainsreportabout insuranceforNon-owners,whichissuitableforpersonsborrowingorrentingacar.Similarly,thereadercanfindFinancialplanningarticlesintheRetirementcategory,advisingtoavoidluxurycars,countthemoney,limitthemovingand/orrenovatingthehouse,etc.

2.2 WEBInformationServices

An overview about services the platforms provide. Instead of analysing view to theplatformsonebyoneasprovidedinpreviouschapter,wewill lookonvaluetheplatformsprovideincommon.

2.2.1 Education

The main criterion for considering the service as an Educational one is based either oncurriculumbasedmaterial,which gradually explains and teach the selected topics, or theglossaryapproach,whentermsaregroupedbypredefinedcategories.Thiscriterionstandsbygenericcommentaryarticlesorthematicguides,whichsurfacethepopulartopics.



Thecommoneducationapproachisexplainingthefinancialterms.Theusefulfeatureistherichlistofterms,whichprovideusuallyStockExchangeplatforms,orSECportal.Somewhatdrawback is mutual reference in explanation that lead to tree-like method of studyingbecauseonetermcanbeexplainedbyotherterms,whichwouldneedalsoanexplanation.MorevisualmethodprovidesInvestopediathatcombinesshortarticlewithvideo.Thevideoisanimatedandveryusesapparentsteps–forexample,thedifferencebetweencommonandpreferredstockishighlightedbyshareholdersthatownthecommonstockandcanvoteinannualmeetings.The similar approach like glossary is covered under “Know how” that can form a set ofarticlesthematicallysimilartoselectedterms.The training material is available in the form of online modules, with provided narratorspeech,quiz,explanatoryslides,tests,etc.,usuallyforadultaudience.Specialtyguidesareavailable as a support material for Trainers, which are focused on training the youth orchildren on Mymoney.gov, where fun, cartoon, and game materials could be found forleadingthechildrentobasicfinancialawareness.

2.2.2 User’sportfoliomanagement

Trackingtheusersecuritiesispopularservice,oftenguardedbyregistrationtoprotecttheprivacyofdata.Theportfolioservicebasicallyallowsto

- searchthesecuritiesbasedonidentification,whichcanbeaTicker(incaseofstock),orISINnumber

- registerthesecuritiesintheportfoliodatabasebyenteringthenumberofunits,theprice, date, and the buy or sell transaction.Deposits andwithdrawals can be alsosupportedinordertohavethewholepictureofinvestedmoney.

- Getthenewsrelatedtosecurities,forexampleincaseofstockitisoftenrelatedtoimportant earning events, shareholder meetings, new product or serviceintroduction,bigmovementofprice.

- Comparingtoolsthatoutlinethedifferencebetweentheownportfolioperformanceand the comparable securities, which can be selected from Indices, other Stocks,bonds,funds,ETFs,orotherkindofsecurities

The various tracking tools differ in terms of usability, interfaces (import export of data),graphicalfeatures,andagencyserviceprovided.

2.2.3 Communities

Themost supported community could be foundon investing and security research firms.There are various opinions on how to invest themoney or on the preferred investmentstrategies,usually accompaniedwith various viewsand ratingsofusers. Forexample, themyfool.com users assign the fool caps as ratings to topics if confirmed by other users’experience.



However, the concept of Community of people with common interests, demand, orconcerns receives rather lower attention in most of analyzed platforms. Stock ExchangeportalsandSECdoesnotevensupportblogswithcomments,atleastinactualversions.EconomymediasupportsBlogswithcommentsfromusuallyregisteredusers,similarlytheFIRSTprojectwebsite.TheobviousfeatureslikeFolloworLikeonsocialnetworksarequitecommon.Activemanagementofusergroups,dedicated forcertaincommunityofpeople,moderatingthetopics,oractivepromotingofthecertainaspectoffinancialawarenesshadratherlimitedsupport.

2.2.4 EconomyEnvironment

Inallreviewedplatforms,theEconomyenvironmentisthemostcommontopicapartfromnews.Theserviceisprovidedinseveralways:

- The commented articles related to world and regional economies. The nature ofcomments often expresses the sentiment of author, analysing impact of industrytrends,technologyinnovations,regulationdecisions,politicalevents,andsoon.Thistopiciscommonlycoveredbytraditionaleconomicmedia,aswellasgeneralmediawitheconomysections;

- Academyresearch,coveredbydeepanalysis,oftenasaresultofstudy,forexampleEffectofsocio-economicfactorsonaggregatesavings,thetopicsoftencoveringthedevelopmentcountries;

- Tools for economy environment, like the FIRST portal, the “Fear and Greed”parameterformeasuringthecurrentmoodofthemarket;

- Conferences for presenting the successful entrepreneurs or business trends fromcommercial sphere, or academic conferences, for example to analysingMacroeconomicdata

2.3 ThePROFITapproach

Despite thewide range of financial awareness platforms available today, PROFIT aims atintroducinganentirelynewapproachtofinancialawarenessbyestablishingauser-centredfinancialawarenessplatformthatobtainsfinance-relatedcrowdsourceddatafromtheweband its users in order to create new knowledge towards offering financial information,education and advanced forecasting tools to help users understand financial data andtrendsandempower them indecisionmaking, catering touserprofilesandspecifics.Theplatformwillbe freelyavailablebasedonOpensoftwarecomponentsand licensedundersuitableOpen licensetobeavailable for thegeneraluseandre-usage.To thebestofourknowledge, suchamulti-functional platform for theuser is not available in theEuropeanenvironment so far.Moreover, the PROFIT platformwill be among the firstOpen Sourceparadigmstoillustrateandanalysethepossibilitiesofexploitingthesemantictechnologyinthefinancialawarenessdomainshowcasingthebenefitsthereinaswellastheproblemsandobstaclestoovercomeinthefuture.



3 FinancialLiteracy

Thissectionreviewsthebestcurrentpracticesintheacademicandprofessionalagendaonfinancial literacy. This agenda can be seen as having two distinct phases. Its first phase,whichcoverstheperiodupto2014,aimedatestablishingfinancialliteracyasanimportantanddistinct field in theeconomicsand finance literature.Thisphasewas completedwiththe publication of the review paper by Lusardi and Mitchell (2014) in the Journal ofEconomicLiterature.Thatpublicationset thebenchmark for financial literacyasadistinctfield.Thesecondphaseoftheagendaaimsatestablishingmethodsfordeliveringfinancialliteracy courses and training programmes, aiming at the maximum possible impact onincreasing the financial literacy of the population and, in particular, the target groups ofprimaryinterest.ThePROFITprojectisamongtheprimaryinnovatorsinthisagendaaimingtosetabenchmarkforthedeliveryoffinancialliteracytraininginanopenonlineplatformmodeformassaudiences.Itisworthnotingthattheextensivereviewofallresourcesisnotthepenultimatepurposeof thisworkpackage,asa thorough reviewanduniquesynthesiswillbeconductedalongthedeliveryofWorkPackage3(WP3-FinancialLiteracyforInformedDecision-Making).Thecurrent review intends to motivate the relevant section of the project and highlight itsimportance,potentialandnovelty.

3.1 Theassessmentoffinancialliteracy(ASSESS)

Therecent financialcrisishasgenerated interest inbetterunderstandinghowtopromotemore responsible and prudent individual saving and borrowing behaviour. The Eurozonedebtcrisis,withtheresultingfiscalconsequencesthroughoutEurope,furtherstimulatedthediscussion on how to promote the efficient allocation of financial resources and greaterfinancial stability. In an era of ageing demographics and increased rates of internal andexternalmigrationinEurope,thechallengesforgovernmentbudgetsarevast.Theabilityofconsumers to make informed financial decisions is critical to developing sound personalfinance,whichcancontributetoincreasedsavingrates,moreefficientallocationoffinancialresources, and greater financial stability. From an institutional perspective, increasedfinancial capability can contribute to client protection and social performance targetassessment by financial institutions. From the EuropeanUnion's viewpoint, it can also beconducive to the enhancement of the European identity, open democracy and activecitizenshipforinformedpoliticalawareness.Ultimately,itcancontributetoincreasedlevelsofoverallwell-being,alongwithhigherlevelsofsatisfactionrelatingtoincome,retirement,housingandfinancialsituation.Financial literacy is the ability to use knowledge and skills to effectivelymanage financialresources at a personal-level. It is more than numeracy—“being good with numbers”. Itincludes for example the understanding of how compound interest rates operate, thedifferencebetweenrealandnominalvaluesandtheroleplayedbyinflation,andprinciplesoffinancialriskdiversification,interalia.TheUSAPresident'sAdvisoryCouncilonFinancialCapabilitydefinespersonalfinancialliteracyasthe:“…abilitytouseknowledgeandskillstomanagefinancial resourceseffectively fora lifetimeof financialwell-being”(PACFL,2008).Lusardi and Mitchell (2014) define financial literacy as “…peoples’ ability to process



economic information and make informed decisions about financial planning, wealthaccumulation,debtandpensions”.The assessment of the levels of financial literacy in population samples in the academicliteraturehas seenanumberof approaches,with thedominant schoolsof thoughtbeingthose of Mandell (2004; 2008) and Lusardi (2008). Both approaches were based on thespecificationofquiz-typequestions, installed incustom-made-and lateronmainstream-populationsurveys.The formerapproach isbasedon long test-typequizzes.Asa result itwasmetwithmore criticism,due topeople'sdislikeof feeling tested in surveys.Anotherimportantcriticismstemmedfromthelongnumberofquestionsinvolved,whichwereoftenlacking conceptual clarity and were often inter-related with each other, resulting inconceptualandstatisticalcaveatswhenitcametotheirutilisation.TheapproachbyLusardidominatedthefield,duetoitssimplicityandconceptualclarity.Itisbasedonsetsof3,5,7,and/or17questions.Thequestionsubsetwhichstoodtheacademicscrutinyofeconomictheoristswas focusedon theunderstandingof the three fundamental notionsof finance,namely, interest, inflation and risk. From parity conditions in international finance, tocorporate and public finance, the discussion is always founded around the clearunderstanding of these three notions. Lusardi and Mitchell (2011b) argue that it isimportanttothinkintermsof“basic”and“sophisticated”financialliteracyquestions.Whatdistinguishes the two is their complexity. In otherwords, the sophisticated questions aremoredifficulttoanswercorrectly.Accordingly,onewouldexpectalargeshareofincorrectanddon’tknowanswersforthesophisticatedquestions.Thethreebasicquestionswereinitiallyincludedinanexperimentalfinancialliteracymoduleforthe2004USHealthandRetirementStudy(HRS)andsincethentheybeenaddedsolelyoralong with other questions tomany surveys, such as the National Longitudinal Survey ofYouth(NLSY)for2007-2008,theRandAmericanLifePanel(ALP)in2008,the2009FinancialCapability Study (FINRA, 2010), etc. The World Bank has introduced the questions in anumberof itsowncustom-made surveysandPanosandWreight (2016a;b; c) introducedthethreequestionsintheBritishElectionsStudy2014,amajorUKsurvey.Foreachquestiona list of possible answers is specified (up to five). There is only one correct answer. Onaverage,ittakesamuchshorterperiodoftimetoanswerquestionsofthistype,comparedtoopen-endedquestion.Respondentsareallowedtochoose“Don’tknow”(andshouldbeencouragedtochoosethisoptionratherthanguess).Inadditiontheyareallowedtochoosethe option “Refuse to answer”. It is important to include this option since it allowsrespondenttoobjecttobetestedorobjectiftheybelievethequestionispryingintotheirfinancial circumstances.More recently, Klapper, Lusardi and vanOudheusden (2015)havetransformed thebasic3 financialquestions into4,distinguishingbetweenmerenumeracyand the understanding of interest rates. The 4 questions are presented in Table 3–1. Inaddition, Table3–2presentsa setof8questions froma recenteffortby theOECD (2015).TheOECDnegotiates the same3-4 fundamental concepts, i.e. interest, inflation, risk, andnumeracy;albeit,witha longer setofquestions, inorder todistinguishbetweendifferentlevelsofunderstandingTable3–3presentsan internationalcomparisonof financial literacybasedoncustom-madesurveysindifferentcountries,usingthe3basicquestionsofPanelAofTable1.1.Thefigures



presentedareadaptedfromLusardiandMitchell(2014)andMontagnoli,Moro,Panosand

Table3–1Themainfinancialliteracyquestionsandtheirextensions

PanelA:The3basicquestions(LusardiandMitchell,2014)1. Numeracy[Interest]

Supposeyouhad$100inasavingsaccountandtheinterestratewas2percent per year. After 5 years, howmuch do you think youwouldhaveintheaccountifyouleftthemoneytogrow?

• Morethan$102;• Exactly$102;• Lessthan$102;• Donotknow;• Refuse;

2. InflationImaginethattheinterestrateonyoursavingsaccountwas1percentperyearandinflationwas2percentperyear.After1year,howmuchwouldyoubeabletobuywiththemoneyinthisaccount?

• Morethantoday;• Exactlythesame;• Lessthantoday;• Donotknow;• Refuse;

3. RiskTrueorfalse?Buyingacompanystockusuallyprovidesasaferreturnthanastockmutualfund.

• True;• False;• Donotknow;• Refuse;

PanelB:The5updatedquestions(Klapper,LusardiandvanOudheusden,2015)1. Numeracy[Interest]

Suppose you need to borrow 100 US dollars. Which is the loweramount to pay back: 105 US dollars or 100 US dollars plus threepercent?

• 105USdollars;• 100USdollarsplusthreepercent;

• Donotknow;• Refuse;

2. InflationSuppose over the next 10 years the prices of the things you buydouble.Ifyourincomealsodoubles,willyoubeabletobuylessthanyoucanbuytoday,thesameasyoucanbuytoday,ormorethanyoucanbuytoday?

• Less;• Thesame;• More;• Donotknow;• Refuse;

3a. Compoundinterest1Suppose you had 100 US dollars in a savings account and the bankadds10percentperyeartotheaccount.Howmuchmoneywouldyouhaveintheaccountafterfiveyearsifyoudidnotremoveanymoneyfromtheaccount?

• Morethan150dollars;• Exactly150dollars;• Lessthan150dollars;• Donotknow;• Refuse;

3b. Compoundinterest2Supposeyouputmoneyinthebankfortwoyearsandthebankagreestoadd15percentperyear toyouraccount.Will thebankaddmoremoney toyouraccount the secondyear than itdid the first year,orwillitaddthesameamountofmoneybothyears?

• More;• Thesame;• Donotknow;• Refuse;

4. RiskdiversificationSupposeyouhavesomemoney.Isitsafertoputyourmoneyintoonebusinessorinvestment,ortoputyourmoneyintomultiplebusinessesorinvestments?

• Onebusinessorinvestment;

• Multiplebusinessesorinvestments;

• Donotknow;• Refuse;



Table3–2TheOECD's8financialliteracyquestions

1. Imaginethatfive<brothers>aregivenagiftof$1,000.Ifthe<brothers>havetosharethemoneyequallyhowmuchdoeseachoneget?<brothers>couldbesubstituted.

Recordnumericalresponse.

Optionsof:• Don'tKnow;• Refused;• Irrelevantanswer

2. Nowimaginethatthe<brothers>havetowaitforoneyeartogettheirshareofthe$1,000andinflationstaysatXpercent.Inoneyear’stimetheywillbeabletobuy?

• Morewiththeirshareofthemoneythantheycouldtoday;

• Thesameamount;• Or,lessthantheycouldbuytoday;• Itdependsonthetypesofthingsthattheywanttobuy;

• Don'tknow;• Refused;• IrrelevantanswerNB.D)canbeconsideredcorrectbutshouldnotbereadout/shown

3. Youlend$25toafriendoneeveningandhegivesyou$25backthenextday.Howmuchinteresthashepaidonthisloan?

Correctanswer:None,nothing,zeroetc

Optionsof:• Don'tknow• Refused• irrelevantanswer

4. Supposeyouput$100intoa<nofee>savingsaccountwithaguaranteedinterestrateof2%peryear.Youdon’tmakeanyfurtherpaymentsintothisaccountandyoudon’twithdrawanymoney.Howmuchwouldbeintheaccountattheendofthefirstyear,oncetheinterestpaymentismade?

Correctresponse:102Optionsof:• Don'tknow• Refused• irrelevantanswer

5. andhowmuchwouldbeintheaccountattheendoffiveyears[addifnecessary:rememberingtherearenofees]?Woulditbe:

• Morethan$110• Exactly$110• Lessthan$110• Orisitimpossibletotellfromtheinformationgiven

• Don’tknow• Refused• Irrelevant

6. AninvestmentwithahighreturnislikelytobehighriskAlternative:Ifsomeoneoffersyouthechancetomakealotofmoneythereisalsoachancethatyouwilllosealotofmoney.

• True• False

7. Highinflationmeansthatthecostoflivingisincreasingrapidly

• True• False

8. Itisusuallypossibletoreducetheriskofinvestinginthestockmarketbybuyingawiderangeofstocksandshares.Alternative:Itislesslikelythatyouwillloseallofyourmoneyifyousaveitinmorethanoneplace.

• True• False



Wreight (2016). The figure establishes low levels of financial literacy, with some 40%respondingcorrectlytoallthreequestionsinGreatBritain,30%intheUnitedStates,45%inthe Netherlands, 53% in Germany, 42% in Australia and figures even lower in othercountries,e.g.Japan(27%),NewZealand(24%).Astylizedfactisthat,inthemajorityofthecountries, more than a third of the sample responds that they do not know the correctansweratleastonce.

Table3–3Internationalcomparisonoffinancialliteracy

Country SurveyyearInterestrate Inflation Risk

All3correct

Atleast1"Don'tknow"

GreatBritain 2014 81.3% 69.1% 48.7% 40.2% 28.3%USA 2009 64.9% 64.3% 51.8% 30.2% 42.4%Netherlands 2010 84.8% 76.9% 51.9% 44.8% 37.6%Germany 2009 82.4% 78.4% 61.8% 53.2% 37.0%Japan 2010 70.5% 58.8% 39.5% 27.0% 61.5%Australia 2012 83.1% 69.3% 54.7% 42.7% 41.3%NewZealand 2009 86.0% 81.0% 49.0% 24.0% 7.0%Switzerland 2011 79.3% 78.4% 73.5%* 50.1%* 16.9%*Italy 2007 40.0%* 59.3%* 52.2%* 24.9%* 44.9%*Sweden 2010 35.2%* 59.5% 68.4% 21.4%* 34.7%*France 2011 48.0%* 61.2% 66.8%* 30.9%* 33.4%*Russia 2009 36.3%* 50.8%* 12.8%* 3.7%* 53.7%*Romania 2011 41.3% 31.8%* 14.7% 3.8%* 75.5%* Notes:Figures forGreatBritainand itscountriesareweightedaverages fromtheBritishElectionStudy.ThefiguresfromtheremainingcountriesarefromLusardiandMitchell(2014).Asterisksdenotedifferencesinthewordingorformatofthequestionsindifferentsurveys.Figure3–1presentsaglobalmapoffinancialliteracyforamorerecentglobalsurveyfundedby Standard & Poor's (2014) and conducted by Klapper, Lusardi and van Oudheusden(2015). The countries with the highest financial literacy rates are Australia, Canada,Denmark, Finland, Germany, Israel, the Netherlands, Norway, Sweden, and the UnitedKingdom,whereabout65%ormoreofadultsare financially literate.Ontheotherendofthe spectrum, South Asia is home to countrieswith some of the lowest financial literacyscores,whereonlyaquarterofadults−orfewer−arefinanciallyliterate.Figure3–2highlightsthesegmentofthemapforthe28membercountriesoftheEuropeanUnion.Financial literacyratesvarywidelyacrosstheEuropeanUnion.Onaverage,52%ofadultsarefinancially literate,andtheunderstandingoffinancialconcepts isthehighest innorthern Europe. Denmark, Germany, the Netherlands, and Sweden have the highestliteracy rates in the European Union: at least 65% of their adults are financially literate.RatesaremuchlowerinsouthernEurope.Forexample,inGreeceandSpain,literacyratesare45%and49%,respectively.ItalyandPortugalhavesomeofthelowestliteracyratesinthesouth.FinancialliteracyratesarealsolowamongthecountriesthatjoinedtheEuropeanUnion after 2004. In Bulgaria and Cyprus, 35% of adults are financially literate. Romania,with22%financialliteracy,hasthelowestrateintheEuropeanUnion.



Figure3–1Financialliteracyaroundtheworld

Source:UoGlasgow'sprocessingof theS&P financial literacy survey (2014), conductedbyKlapper, Lusardi andvanOudheusden(2015).

Figure3–2FinancialliteracyintheEU-28

Source:UoGlasgow'sprocessingoftheS&Pfinancialliteracysurvey(2014),conductedbyKlapper,LusardiandvanOudheusden(2015).

(.7,.75](.65,.7](.6,.65](.55,.6](.5,.55](.45,.5](.4,.45](.35,.4](.3,.35](.2,.3][.1,.2]Missing

(.7,.75] (.65,.7] (.6,.65] (.55,.6] (.5,.55] (.45,.5] (.4,.45] (.35,.4] (.3,.35] (.2,.3] [.1,.2] Missing

(.7,.75] (.65,.7] (.6,.65] (.55,.6] (.5,.55] (.45,.5] (.4,.45] (.35,.4] (.3,.35] (.2,.3] [.1,.2] Missing



Itisevidentthatfinancialliteracyratesdifferenormouslybetweenthemajoradvancedandemerging economies in theworld. Klapper, Lusardi and vanOudheusden (2015) highlightthat, on average, 55% of adults in the major advanced economies − Canada, France,Germany,Italy,Japan,theUnitedKingdom,andtheUnitedStates−arefinanciallyliterate.Butevenacross thesecountries, financial literacy rates rangewidely, from37% in Italy to68% inCanada. In contrast, in theBRICSeconomies (Brazil, theRussianFederation, India,China,andSouthAfrica)−onaverage,28%ofadultsarefinanciallyliterate.Disparitiesexistamongthesecountries,too,withratesrangingfrom24%inIndiato42%inSouthAfrica.One frequent claim is that differences in financial literacy around the world might beexplained by income and the level of economic development. Figure 3–3 examines therelationshipbetweenfinancialliteracyandGDPpercapita(accountingforthecostofliving,intermsofPPP$international).Thefigureindicatesthatwealthiercountries,asproxiedbyGDPpercapita,tendtoexhibithigherratesoffinancial literacy.However,therelationshipmostlyholdswhen lookingat thewealthiest 50%of economies. Klapper, Lusardi and vanOudheusden (2015) suggest that around38%of the variation in financial literacy rates intheseeconomiescanbeexplainedbydifferencesinincomeacrosscountries.Forthepoorerhalf of economies, with a GDP per capita of $12,000 or less, there is no evidence thatincome is associatedwith financial literacy. Our own computations also suggest that therelationshipbetweenfinancial literacyandPPP-dividedGDPpercapita ishighlynon-linearandmildlyV-shaped.Theevidence is interpretedas indicative thatnational-levelpolicies,such as those related to education and consumer protection, shape financial literacy intheseeconomiesmorethananyotherfactor.

examinestherelationshipbetweenfinancialliteracyandPPP-dividedGDPpercapitainthe28countriesoftheEuropeanUnion.Theretherelationshipseemstobeclosertolog-linear,withwealthiercountriesexhibitinghigher levelsof financial literacy.Thisobservation isofgreat interest, as it adheres one novel policy element in the agenda of the Europeanintegrationprocess; thatofdifferences in financialknowledgeand resultingdifferences infinancial planning behaviour. Apart from heterogeneity across EU countries at themacroeconomiclevel,itisevidentthattherealsoexistmicroeconomicdifferencesintermsof knowledge. Such differences are likely to exert significant impact on individual andcountryoutcomes.Recentevidenceshowsthatfinancialknowledgeisakeydeterminantofwealthinequality.Itissuggestedthat30-40%percentofretirementwealthinequalityintheUSisaccountedforbyfinancialknowledge(Lusardi,MichaudandMitchell,2016).Inaseriesofstudies,reviewedinthenextsection,ithasbeensuggestedthatindividualswhoaremorefinancially literate make more economically rational decisions pertaining to real estatepurchases, insurance purchases, investing, saving, tax planning, retirement planning,pensionand insuranceplanning.Moregenerally, peoplewhoaremore financially literatemakebetterfinancialdecisions.Thus, it isevidentthatfinancial literacy is likelytoexertalarge indirect impact on financial stability. Given global economic challenges,which haverecently tested the stability of the European Union, policy initiatives regarding theenhancement of financial knowledge at the microeconomic level are likely to entailsignificantpositivemacroeconomicimplications,conducivetofinancialstability.



Figure3–3TherelationshipbetweenfinancialliteracyandGDPpercapita

Source:UoGlasgow'sprocessingoftheS&Pfinancialliteracysurvey(2014)andWorldDevelopmentIndicators(2014)

AFG ALB

ARE

ARG

ARM

AUS

AUT

AZE

BDI

BEL

BENBFA

BGD

BGR

BHR

BIH

BLR

BLZ

BOL

BRA

BTN BWA

CAN

CHE

CHL

CHN

CIVCMR

COG COLCRI CYP

CZE

DEU

DNK

DOMDZA

ECUEGY

ESP

EST

ETH

FIN

FRA

GAB

GBR

GEOGHAGIN

GRC

GTM

HKG

HND

HRV

HTI

HUN

IDN

IND

IRL

IRN

IRQ

ISR

ITAJAM

JOR

JPNKAZKEN

KGZKHM

KOR

KSV

KWTLBN

LKALTU

LUX

LVA

MDA

MDG

MEX

MKD

MLI

MLTMNE

MNG

MRT

MUSMWI MYS

NAM

NER

NGA

NIC

NLD

NOR

NPL

NZL

PAK PANPERPHL

POL

PRI

PRT

ROM

RUS

RWA

SAU

SDN

SEN

SGP

SLE SLV

SRB

SVKSVN

SWE

TCD

TGO

THA

TJK

TKMTUN

TUR

TZA

UGA

UKR

URY

USA

UZBVENVNMWBG

YEM

ZAF

ZAR

ZMBZWE

0

.2

.4

.6

.8

Fina

ncia

l Lite

racy

: %Co

rrect

3 o

ut o

f 4 (S

&P)

6 8 10 12Logarithm of GDP per capita (PPP current international $ - WDI)

Financial literacy in 2014 (S&P) Fitted line from linear regression

Fitted line from 5th order local polynomial regression

AFG ALB

ARE

ARG

ARM

AUS

AUT

AZE

BDI

BEL

BEN

BFA

BGD

BGR

BHR

BIH

BLR

BLZ

BOL

BRA

BTN BWA

CAN

CHE

CHL

CHN

CIVCMR

COG COLCRI CYP

CZE

DEU

DNK

DOMDZA

ECUEGY

ESP

EST

ETH

FIN

FRA

GAB

GBR

GEOGHAGIN

GRC

GTM

HKG

HND

HRV

HTI

HUN

IDN

IND

IRL

IRN

IRQ

ISR

ITA

JAM

JOR

JPNKAZ

KEN

KGZKHM

KOR

KSV

KWTLBN

LKALTU

LUX

LVA

MDA

MDG

MEX

MKD

MLI

MLT

MNE

MNG

MRT

MUSMWI MYS

NAM

NER

NGA

NIC

NLD

NOR

NPL

NZL

PAK PANPERPHL

POL

PRI

PRT

ROM

RUS

RWA

SAU

SDN

SEN

SGP

SLE SLV

SRB

SVK

SVN

SWE

TCD

TGO

THA

TJK

TKMTUN

TUR

TZA

UGA

UKR

URY

USA

UZB

VENVNMWBG

YEM

ZAF

ZAR

ZMBZWE

.2

.4

.6

.8

Fina

ncia

l Lite

racy

: %Co

rrect

3 o

ut o

f 4 (S

&P)

0 20000 40000 60000 80000 100000GDP per capita (PPP current international $ - WDI)





Figure3–4TherelationshipbetweenfinancialliteracyandGDPpercapitaintheEU-28

Source:UoGlasgow'sprocessingoftheS&Pfinancialliteracysurvey(2014)andWorldDevelopmentIndicators(2014)

AUTBEL

BGR CYP

CZE

DEU

DNK

ESP

EST

FIN

FRA

GBR

GRCHRV

HUN IRL

ITALTU

LUX

LVA

MLT

NLD

POL

PRT

ROM

SVK

SVN

SWE

.2

.4

.6

.8

Fina

ncia

l Lite

racy

: %C

orre

ct 3

out

of 4

(S&

P)

20000 40000 60000 80000 100000GDP per capita (PPP current international $ - WDI)


Fitted line from 2nd order local polynomial regression

AUTBEL

BGR CYP

CZE

DEU

DNK

ESP

EST

FIN

FRA

GBR

GRCHRV

HUN IRL

ITALTU

LUX

LVA

MLT

NLD

POL

PRT

ROM

SVK

SVN

SWE

.2

.3

.4

.5

.6

.7

Fina

ncia

l Lite

racy

: %Co

rrect

3 o

ut o

f 4 (S

&P)

9.5 10 10.5 11 11.5Logarithm of GDP per capita (PPP current international $ - WDI)





3.2 Financialliteracydemographics(DESCRIBE)

The very recent booming literature on financial literacy has identified differences inknowledge across demographic groupswhich are of primary interest for informing policyinitiatives aiming at the enhancement of financial capability. Klapper, Lusardi and Panos(2015)findthatinthefinanciallyliterateindividualsintheUSaremorelikelytostarttheirown business, and also perform better while in business. This view on entrepreneurialfinancialknowledgematches thedescriptionof theentrepreneurial figurebyAdamSmith(1876)1.ThereviewpaperbyLusardiandMitchell (2014)establishesthatapart fromtheobservedlowlevelsoffinancialliteracy;lessdevelopedcountrieshaveevenlowerrates.Importantly,financial illiteracy is more severe for some key demographic groups. These include thegeneration of baby boomers, which has neared or gone in retirement, minorities andmigrants,thelesseducated,lowincomeindividuals,etc.Ademographicgroupofparticularinterestinvolvesthefemales,whohavebeenconsistentlyshowntoexhibit lowerlevelsoffinancial literacy. Significantgenderdifferenceshavebeendocumented formostWesterncountries.Figure3.5presentsfiguresfromtheUK,theUSA,Germany,SwitzerlandandtheNetherlands (Panos and Wreight, 2016a). Males are significantly more likely to respondcorrectlytoall threefinancial literacyquestions intheBritishElectionStudy,withfemalessignificantlymorelikelytorespondthattheydon'tknowtheresponsetoatleastoneofthethree questions. This pattern has been also shown in a number of studies for othercountries. Lusardi and Mitchell (2014) suggest that the observed gender differences arelikelytoalsoreflectagreaterlackofconfidenceinfinancialmattersbyfemales,apartfromonlyknowledgedifferences.UsingaScottishsampleofstudentsagedbetween11-17,PanosandWright (2016b) find that the gender gap in financial literacy starts very early in thestudentlife,andthateconomics/finance/businessclassesnarrowtheobservedgendergapinfinancialliteracyatschoolyears.Recognisingtheincreasingneedforfinancialeducation,bothintheschoolcurriculumandinlifelonglearningamongadults,therehavebeensomestudiesthathaveaimedtoassesstheeffectivenessofvariousfinancialeducationprogrammes,bothamongtheyoungandadults.There has been some debate as towhether financial education in high schools has beensuccessful at improving the financial knowledge of participants. Bernheim, Garrett, Maki(2001)findthatmiddle-agedindividualswhotookapersonalfinancialmanagementcourseinhighschoolweremorelikelytosaveahigherproportionoftheirincomesthanindividualswho did not. This emphasises the fact that financial education can help individualsmakebetterfinancialdecisionswhichwouldbenefitthemthroughoutlife.Howeverotherstudieshavequestionedtheeffectivenessof these financialeducationcourses.MandellandKlein(2009) examined the impact of a personal financial management course taken 1-4 yearspreviouslyon79highschoolstudentsintheU.S.Theydiscoveredthatthosewhotookthe

1[Theentrepreneur]must"...beabletoread,write,andaccount,andmustbetolerablejudgetooof,perhaps,fiftyorsixtydifferentsortsofgoods,theirprices,qualities,andthemarketswheretheyaretobehadcheapest.Hemusthaveall theknowledge, inshort, that isnecessaryforagreatmerchant,whichnothinghindershimfrombecomingbutthewantofasufficientcapital."

AdamSmith(1776:BookI,Ch.X:p.12)



course were not more financially literate and did not appear to have better financialbehaviourthanthosewhohadnottakenthecourse.Theseareonlytwoofmanystudiesthathavebeencarriedoutontheeffectivenessofhighschool financial education programmes; a review of these studies indicates contradictoryresults.Walstad, Rebeck, andMacDonald (2010) undertook a studywith 800 high schoolseniorstudentsacrosstheUnitedStatestoassesswhetherapersonalfinanceprogrammewould increase students’ knowledgeofpersonal finance.The studyhadbotha treatmentand a control group andboth groups showed similar levels of financial knowledge at thestart of the programme. The programme was then delivered to the treatment groupthroughfivedifferentvideosegments,whichwerespreadoveratwo-to-fourweekperiod.At the endof the study the treated group showeda substantial increase (19.7%) in theirfinancialknowledgewhereasthecontrolgrouponlysawaverysmallincrease(1%)intheirfinancial knowledge, which led to the conclusion that a well-designed and properlyimplemented financial education programme could significantly improve the financialknowledgeofhighschoolstudents.Luhrmann,Serra-GarciaandWinter(2015)carriedoutastudy,whichevaluatedtheimpactof financial literacy training on teenagers between 14 and 16 years old in German highschools. Their study consisted of three short training modules, which typically took 90minutesandwere taughtby financeprofessionalswhovolunteered to teach teenagers inhigh schools. The trainingmodules focused on shopping, planning, and saving. The studyusedatreatmentgroupof558studentscomparedto158controlstudentsandtheyfoundthatthoseinthetreatmentgroupshowedanincreaseinfinancialknowledgeandinterestinfinancial matters. This again highlights that even a relatively short financial literacyprogrammecanhelptoimprovethefinancialknowledgeofhighschoolstudents.Bruhn et al (2013) carried out a study into the impact of a comprehensive financialeducation programme in Brazil that spanned six states, 868 schools and approximately20,000highschoolstudentsthrougharandomizedcontroltrial.Theprogrammespanned17monthsanditwasintegratedintotheclassroomcurriculaofMathematics,Science,History,and Portuguese, depending on the teacher teaching the course. This was in place ofintroducing a separate financial education course, as itwasdecided that thiswouldhavehadtoogreatanimpactonteacherandstudentworkloads.Onebigdifferencebetweenthisstudyandotherstudieswasthelengthoftimethattheprogrammewasdeliveredover,withtheprogrammeusingtextbookswithmaterials,whichprovidedbetween74and144hoursof teaching. These materials involved interactive case studies which aimed at placingstudentsinfinancialsituationsthattheywerelikelytofaceinthefuture,forexample,familyfinances, entry into the labour market, and preparing for financial independence. Theprogramme helped to improve the financial knowledge of students taking part and alsoincreased the likelihood of financial planning and participation in household financialdecisions by students. This study again highlighted that financial education programmestargeted at high school students can help to improve both the financial knowledge andbehaviourofstudentstakingpart.



PanelA:All3correct

PanelB:Atleast1"don'tknow"

Figure3–5DistributionoffinancialliteracyinGreatBritainandothercountries,bygender

Notes:FiguresforGreatBritain,England,ScotlandandWalesareweightedaveragesfromtheBritishElectionStudy,byWreightandPanos(2016a).ThefiguresfromtheremainingcountriesarefromLusardiandMitchell(2014).AnotherrecentstudythatfocusesontheeffectivenessoffinancialeducationinhighschoolsisbyHospido,VillanuevaandZamarro (2015).Their inquirywascompleted inMadridandthe study comprisedof 1,223 students in the 3rd year ofmandatory secondary educationfrom22schools.Forthestudytherewere981studentsinthetreatedgroupand242inthecontrolgroup.Theprogrammeconsistedofa10-hourcourseinfinancialeducationwiththe

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

Great Britain

England Scotland Wales USA Germany Netherlands Switzerland

50.1% 50.1%53.0% 54.6%

38.3%

59.6%55.1%

62.0%

30.6% 30.5% 31.0%34.3%

22.5%

47.5%

35.0%39.3%

Male Female

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

40.0%

45.0%

50.0%

Great Britain

England Scotland Wales USA Germany Netherlands Switzerland

20.3% 20.4%18.1% 18.0%

34.3%

29.9% 29.0%

11.8%

36.0% 36.1%39.5%

24.5%

50.0%

43.3%45.8%

21.6%

Male Female



material for the course drawn from both a Student and Teacher’s Guide and a websitewhich offered online support via practical exercises, activities, games, and additionalresources.Theprogrammeaimedat increasingthefinancialknowledgeoftheparticipantsandprovidingthemwiththetoolstomakebetterfinancialdecisionslaterinlife.Thestudyfound that students of the course performed better in both the arithmetic and non-arithmeticquestionsthanstudentsfromthecontrolgroupinatestoffinancialcompetenceattheendofthecourse2.Thesestudieshaveprimarilyfocusedontheeffectivenessoffinancialeducationasawhole,rather than focusingon themethodof teaching thathasbeenusedduring these coursesand the different studies have providedmixed results as to the effectiveness of financialeducationcoursesinhighschool.Duetotheincreasingimportanceofdevelopingfinancialliteracyinthegeneralpopulationandthegeneralbeliefthatfinancialeducationwillplayakeyroleinimprovingthefinancialliteracyofindividuals,ithasbeenarguedbyLusardietal(2015) that the research in financialeducationshouldmoveaway fromadiscussionas towhethergenericfinancialeducationworks. Instead,theresearchshouldfocusontryingtogainanunderstandingofhowtomakefinancialeducationworkthroughbetterdesignandappropriate delivery methods. By focusing on how to make financial education work todelivermoreeffectivecoursesandproducebetter informed individuals, it is important tolook at the differentways that financial education can be taught andwhichmethodwillresultinthemostinformedparticipants.Collins&Urban(2015)conductafieldstudywhereindividualswithin45creditunionsacrossWisconsin had randomized access to a 10-hour online financial education programmeonsavingsanddebtmanagement.Employeesatthecreditunionsthatofferedtheprograminthe1st half of the study constitute the treatment groupandemployeesat the those thatoffered it in the 2nd half constitute the 'control‘ group. All employees completed a 48-question survey concerning their self-assessed financial knowledge and behaviour at 3points intime, i.e.beforeanyeducation,afterhalfofsiteshadaccesstothetraining,andafter all sites had access. The results show an increase in knowledge regarding interest,loans,creditscores,stocksandbonds,andretirementinvestment.Theyalsoshowthattheeducation is associated with greater retirement plan participation, more frequentemergencysavings,andmoreinstancesofcreatingabudgetorfinancialplan.Thesefindingsarecorroboratedwithactualretirementplancontributionaccounts

2Whiletheabovestudiesindicatetheeffectivesoffinancialeducation,therehavebeensomestudies,whichhavecontradictedthisview.Becchetti,Caiazza,andCoviello (2013)carriedoutastudywith944highschoolstudents from 36 classes in their final year in Italy. The course lasted approximately 3 months and theparticipatingstudentswererequiredtocompleteasurveybothbeforeandafterthecompletionofthecourse.Asthecoursewasbeingtaughtbydifferentteachersindifferentclasses,theteacherswereprovidedwiththesamematerials to teach the course. Thesematerials included: (i) a set of slides; (ii) a short guide for theteacher,whichillustratedtheguidelinestobefollowedintheir lessonsand(iii)amoredetailedguidetotheavailablematerialsspecificallydesignedforthestudents.Thesamesurveywasusedforboththebaselineandthe follow-up survey andwhilst the treated group did perform better in the follow up survey, the controlgroupshowedsimilar improvements,with littlestatisticaldifferencebetweenthetwogroups. It isthereforedifficulttoassesswhetherthecoursehelpedtoimprovethefinancialknowledgeoftheparticipantsorifanyimprovementcamefromstudentslearningtheanswerstothesurveyquestions.



3.3 Theimportanceoffinancialliteracy(ATTRIBUTE)

Theabilityofindividualstomakeinformedfinancialdecisionsiscriticaltodevelopingsoundpersonal finance, which can contribute tomore efficient allocation of financial resourcesandtogreaterfinancialstabilityatboththemicroandmacrolevel(see,e.g.,Lusardi,2008;Lusardi and Tufano, 2009a, b). Recent studies have documented a strong relationshipbetweenfinancialliteracyandasetofeconomicbehaviours.Bernheim(1995,1998)showedthat most households lack basic financial knowledge and cannot perform very simplecalculations,andthatthesavingbehaviourofmanyhouseholdsisdominatedbycruderulesofthumb.Hilgert,Hogarth,andBeverly(2003)findastronglinkbetweenfinancial literacyand day-to-day financial management. Financial literacy has also been linked to a set ofbehavioursrelatedtosaving,wealth,andportfoliochoice.Forexample,severalpapershaveshown that individuals with greater numeracy and financial literacy are more likely toparticipateinfinancialmarketsandtoinvestinstocks(Christelis,Jappelli,andPadula,2010;Yoong, 2011; Almenberg and Dreber, 2011; Christiansen, Joensen, and Rangvid, 2008;Almenberg and Widmark, 2011; Van Rooij, Lusardi, and Alessie, 2011). Moreover, moreliterate individuals aremore likely to choosemutual fundswith lower fees (Hastings andTejeda-Ashton,2008;HastingsandMitchell,2011).Klapper,LusardiandPanos(2013)useuniqueRussianpaneldatafrompriortoandafterthefinancialcrisis,toexaminefinancialliteracyanditseffectsonbehaviour.TheyfindlowlevelsoffinancialliteracyinRussiaandthatfinancialliteracyispositivelyrelatedtoparticipationinfinancial markets and negatively related to the use of informal sources of borrowing.Moreover, individuals with higher financial literacy are significantlymore likely to reporthaving greater availability of unspent income and higher spending capacity. In addition,KlapperandPanos(2011)findthatinacountrywithwidespreadpublicpensionprovisions,financial literacy is significantly and positively related to retirement planning involvingprivatepensionfunds.Similarly,LusardiandMitchell(2007)showthatthosewhodisplayhighlevelsofliteracyaremorelikelytoplanforretirementand,asaresult,accumulatemuchmorewealth,afindingreproduced in many of the countries that are part of an international comparison offinancial literacy (Lusardi andMitchell, 2011b). Moore (2003) reported that respondentswith lower levels of financial literacy are more likely to have costly mortgages. Gerardi,Goette,andMeier(2010)reportthatthosewith lowliteracyaremore likelytodefaultonsub-primemortgagesortohaveproblemswiththem.StangoandZinman(2009) findthatthosewhoarenotabletocorrectlycalculateinterestratesoutofastreamofpaymentsendupborrowingmoreandaccumulatingloweramountsofwealth.Campbell(2006)showsthatindividualswithlowerincomesandlowereducationlevelsarelesslikelytorefinancetheirmortgagesduringaperiodoffallinginterestrates.LusardiandTufano(2009a,b)reportthatindividuals with lower levels of financial literacy tend to transact in high-cost manners.SimilarfindingsarereportedintheUK(DisneyandGathergood,2013).Correlation between financial literacy and behaviour does not mean causation, and it isimportanttoestablishacausal link. Interestingly,Agarwaletal.(2009)showthatfinancialmistakesareprevalentamongtheyoungandtheelderly,whicharethegroupsthatdisplaythelowestlevelsoffinancialknowledge.Calvet,Campbell,andSodini(2007,2009)examine



datafromSwedenandlookatactionsof investorsthatcanbeclassifiedasmistakes.Theyfind that poorer, less educated, and immigrant households - demographic characteristicsthat are strongly associatedwith low financial literacy - aremore likely tomake financialmistakes. These findingshavebeen supported in recent studiesusing randomised controltrialstoexplorethecausalimpactoffinancialliteracyonfinancialoutcomes.Forinstance,inIndonesia,a randomly selectedsetofunbanked individualswereoffered financial literacytraining sessions, whichwere found to increase the demand for banking services amongthose with low initial levels of financial literacy and low levels of education (Cole et al.,2011).Another strategy has been to rely on instrumental variable (IV) estimation. Severalinstruments have been used. For example, Lusardi and Mitchell (2009, 2011) followBernheim andGarrett (2001) in using high school financial literacymandates in differentstates and time periods in theUnited States. Van Rooij, Lusardi, and Alessie (2011) haveusedthefinancialliteracyofothers,suchassiblingsandparents.Behrmanetal.(2010)usedata from the Chilean Social Protection Survey and a set of plausibly exogenousinstrumental variables that satisfy critical diagnostic tests to isolate the causal effects offinancialliteracyonwealth.Thus, individuals who are more financially literate have been shown to make moreeconomically rational decisions pertaining to real estate purchases, insurance purchases,investing,saving,taxplanning,retirementplanning,pensionandinsuranceplanning.Moregenerally, peoplewho aremore financial literatemake better financial decisions (LusardiandMitchell, 2014). Thus, from an institutional perspective, increased financial capabilitycan contribute to client protection and social performance target assessmentby financialinstitutions.

Someveryrecentliteratureestablishessomeparticularlynovelstylizedfactsregardingthelinkbetweenfinancialliteracyandfinancialstability.Inaddition,veryrecentstudiesshowalinkbetweenfinancialliteracyandpublicattitudeswhichcanbethoughtofasconducivetoeconomicstability.RecentevidenceshowsthatfinanciallyliterateindividualsintheUSaremorelikelytostarttheirownbusiness,andalsoperformbetterwhileinbusiness(Klapper,LusardiandPanos,2016a). Inaddition,Klapper, LusardiandPanos (2016b) findapositiverelationshipbetweenfinancialliteracyandtrustininstitutions.At themacro level, and in viewof loomingdebt and retirement crises around theworld,salient political choices − recently involving voting in referendums − are determined byattitudestowardsredistribution,immigration,austerity,aswellastheworkingsofeconomicpartnershipsinmonetaryandfiscalunionsetc.Suchattitudesarelikelytodependupontheunderstanding of the basics of macroeconomic accounts and public finance. Recentevidence shows relationships between financial literacy in the UK and attitudes to theScottish referendum, the EU referendum, and towards immigration (Panos and Wright,2016a; 2016b; 2016c). Moreover, Montagnoli, Moro, Panos, and Wright (2016) find asignificant negative relationship between financial literacy and attitudes favouring stronggovernmentincomeredistributionefforts.



Recentevidenceshowsthatfinancialknowledgeisakeydeterminantofwealthinequality.30-40% percent of retirement wealth inequality in the US is accounted for by financialknowledge(Lusardi,MichaudandMitchell,2016).Thus,recentliteraturestartstoplacefinancialliteracyattheepicentreofpolicydiscussionsregarding financial and macroeconomic stability, going beyond emphasizing its apparentimportance for microeconomic financial stability. Its relationship with the formation ofpublicattitudesisalsoadimensionthatisofprimaryinteresttonationalandinternationalpolicyagendas,withimplicationsregardingthecommunicationofpublicfinance.

3.4 Financialliteracytoolsandtoolkits(COMPARE)

Giventhe importanceoffinancial literacyandthenegativeeffectsof illiteracyonfinancialbehaviourasdocumentedinextantacademicresearch,therehasbeengrowinginterestinaddressing financial illiteracy across countries. In the European Union, the EuropeanCommission has been actively engaged in promoting financial education for its citizens(EuropeanBankingFederation,2009).This section reviews the currently available financial literacy tools and toolkits. Itdistinguishes between: (i) Financial institutions engaged in financial education; (ii)Newspaperswithonlinepersonal finance sections; (iii)Openonline courses (MOOCs); (iv)Books;(v)Financialgamesandgamificationtools,and;(vi)Europeaneducationalplatforms.Thereviewwasconductedusingsearchenginesanddatabasescontainingfinancialliteracyinitiatives such as the Canadian financial literacy database maintained by the FinancialConsumerAgencyofCanadaandtheglobaldatabaseonfinancialeducationmaintainedbytheOECD.ReferenceswerealsomadetopriorreviewsofcurrentbestpracticessuchastheEuropean Banking Federation’s (2009) report on financial literacy, the survey of financialliteracyschemesintheEU-27byHabschicketal.2007,andthereviewfinancialeducationand awarenessby the European insurance and reinsurance federation (Insurance Europe,2011). Additional reviews of existing schemes across European countries were alsoconductedusingtheOECD’sglobaldatabaseonfinancialliteracy.

3.4.1 Financialinstitutionsengagedinfinancialeducation

This subsection discusses financial education initiatives by financial institutions,distinguishingbetween:(i) individualfinancial institutions;(ii)bankingassociations,centralbanks and ministries, and; (iii) financial institutions that are actively engaged in onlinefinancial education using publicly available resources; and (iv) interactive tools byinstitutions.In the first category, twenty-one individual financial institutions are analysed. These arefromEuropeandNorthAmericaandarepresented inTable3–4.This isnotanexhaustivereview, but aims to covering themajor initiatives in this agenda by financial institutions.Moreover, the mix of organisations reviewed is quite diverse, including private banks,ethicalbanks,publicsectorandfederationsoffinancialorganisations.



We start our analysis from the European Union level. TheEuropean BankingFederation(EBF) is the voice of Europe’s banking sector in all regulatory debates atEuropeanandgloballevel.Itrepresents32nationalbankingAssociationsrepresenting4500banksand2.3millionemployees.Whilenotproviding contenton financial literacy, it is a“hub”wherethegeneralpubliccanfindthird-partywebsitesdealingwithfinancialliteracy.A similar “hub approach” is taken by another Europe-wide organisation, the EuropeanBankingandFinancialServicesTrainingAssociation.Deutsche Bank AG is a German global banking and financial services company basedinFrankfurt. The approach adopted by it is direct, with employees going to schools andholdingfinancialliteracylecturesinordertoeducatetheyoung.Thecontentofsuchlessonsisquitestandardandbasic(inaccordancetothetargetaudience),includinginvestmentandretirementfunds,globalfinancialsystemsandfinancialcrises.InDenmarkandNorthernEurope,DanskeBankoffersfinancialliteracymaterialsthattargetthe familyecosystem– fromchildren to youngpeople,parentsand teachers. Thegoalofsuch a strategy is to prepare the next generation of banking customers tomanage theirfinancial affairs. The way in which they deliver the content is bespoke to the targetaudience:advisory services,educationalmaterials,onlinegames,eventsandpresentationmaterialstostimulatefinancialskillsandknowledge.TheBelgianFinancialSectorFederationofferse-learningmaterialstoeveryonewhoneedsabasic trainingand is intended forpeoplewhowish toacquiregeneralknowledgeof theactivities,productsandorganizationofabank.In Italy, theFoundation for Financial Education and Savings,anot-for-profitorganisationfounded by the Italian Banking Association, has the goal to promote financial educationamong citizens ranging from young people to retirees. They try to achieve this goal viaeducational workshops, theatrical shows, multimedia, videos, games, events andcollaborationswithvariousinstitutionsoftheterritory.Itshouldbestressedthatnotonlytraditionalbanks,butalsoethicalbanks(bothexamplesbelow are FEBEA members) offer resources relevant to financial literacy. For example,BancaEticahasaprogrammewhichtargetsyoungpeopleandfeaturesrole-playinggamestothinkaboutmoney, investmentsandinterestandhowtomaketherightdecisionswithrespecttotransparencyandsustainability.Cassa Padana Bcc is another ethical cooperative bank that offers guides to children andimmigrants. Cassa Padana has a training programmededicated to elementary school andhighschoolstudents:throughgames,theyunderstandtheimportanceofproperlyspendingtheir own money. They also deal with more advanced concepts such as saving food,resources,waterand time.The financial inclusionof immigrants is an issueCassaPadanahas been dealing with for years. To improve the relationship between the bank andimmigrants,financialeducationisafundamentalstep.Thankstothecollaborationofseveralemployees of foreign origin, Cassa Padana has produced guides to banking products,translatedintoAlbanian,Arabic,Indian,English,FrenchandSpanish.



In Malta, theMalta Financial Services Centre - a public body which is the regulator offinancialservicesinMalta-isalsoengagedinfinancialliteracyefforts.Itstargetaudienceisthegeneralpublicandhasasection,whichispurelyeducationalexplainingissuesrelatedtobanking, investments, insurance etc. Another public institution that deals with financialliteracyistheMinistryfortheFamilyandSocialSolidarity.ThehighlightistheDraftNationalStrategy document on retirement income and financial literacy 2016-2018. It sets out aprogramme and a number of initiatives that the government will be taking to improvefinancial literacy inMalta.Thisdocument iscurrentlyavailable forpublicconsultationandhasexamplesoffinancialcalculationsandexplanations.In theUK,certain largebanks,suchasBarclaysandRBS,offerspecific financialeducationprogrammestargetedtowardsyoungpeople.Theformereventailorscontentaccordingtothe continent (Africa, Asia-Pacific, Europe and so on) and both take advantage of onlinelearning tomake the content accessible. The goal is tohelp thenext generation improvemoney-managementskills,knowledgeandconfidence.Similarly,Lloyds targetsyoungstersand teachers in order to feed information to the education system. Standard Charteredoffers twoglobal financialeducationprogrammes:FinancialEducation forYouth, teachingyoung people about making a budget and managing their money; and Education forEntrepreneurs, which helps micro and small enterprises to master the basics ofentrepreneurship. Others, like Halifax and HSBC offer websites with generalist content(explanation of jargon, insurance etc.) alongside personal budgeting tools and calculatorsthat integrate with the client’s profile. In this case, the content is delivered via a widespectrumofarticles,videos,appsandsocialmedia.In the United Kingdom, there are also non-bank institutions offering financial educationtools.TheMoneyAdviceServiceisaserviceprovidedbytheConsumerFinancialEducationBody,whichwasestablishedby thenational financial regulator, i.e. theFinancial ServicesAuthority,thatprovidesfree-of-chargeadviceonmoneyandfinancialdecisionstopeopleintheUnitedKingdom. It isan independentorganisation,whosestatutoryobjectivesare: toenhancetheunderstandingandknowledgeofpeopleaboutfinancialmatters,includingtheUKfinancialsystem,andtoenhancetheirabilitytomanagetheirownfinancialaffairs.Themain thematic areas of the educationalmaterial are: debt andborrowing, budgeting andmanagingyourmoney,savingand investing,retirement,workandbenefits, insuranceandmortgages.IntheUnitedStates,theFederalDeposits InsuranceCorporation(FDIC)providesfinancialliteracy training targeted towards low- and moderate-income individuals outside thefinancialmainstream.Interestingly,italsoprovidesmaterialsforinstructors.Thecontentitoffers, in linewith itsmissionanditstargetusers, includesbankservices,credit,chequingaccounts, financial recovery, savings, credit card, borrowing, credit and home ownership.Themodeofdeliveryofthetraininginvolvesslides,CDsandprintedguides.



Table3–4Listofindividualfinancialinstitutionsengagedinfinancialeducation

Financialinstitution Country Link Emphasis/Targetaudience

DeutscheBank Germany https://www.db.com/cr/en/concrete-financial-literacy-initiative.htm

YoungPeople

BankofAmerica USA https://www.bettermoneyhabits.com/index.html

GeneralPublic

Barclays UK https://www.home.barclays/citizenship/supporting-communities/our-programmes-in-uk/barclays-money-skills.html

YoungPeople

BelgianFinancialSectorFederation

Belgium http://www.mijngeldenik.be/ GeneralPublic

CreditSuisse Switzerland https://www.credit-suisse.com/uk/en/about-us/responsibility/economy-society/focus-themes/education.html

GirlsandYoungWomen

Citi Severalcountries

http://www.citigroup.com/citi/citizen/community/curriculum/teens.htm

YoungPeopleandGeneralPublic

DanskeBank Denmark https://danskebank.com/en-uk/CSR/Financialliteracy/Pages/default.aspx

Children,YoungPeople,ParentsandTeachers

FDIC USA https://www.fdic.gov/consumers/consumer/moneysmart/index.html

Low-andmedium-incomeIndividualsOutsidetheFinancialMainstream

Halifax UK http://www.halifax.co.uk/managingyourmoney/

Clients

HSBC UK http://financialplanning.hsbc.co.uk/#/all GeneralPublic,Clients

LloydsBankingGroup UK http://www.lloydsbankinggroup.com/our-group/responsible-business/our-community-programmes/money-for-life/


RBS UK http://www.rbs.com/sustainability/community/young-people.html

YoungPeople

FoundationforFinancialEducation

andSavings

Italy http://www.feduf.it/chi-siamo/i-partecipanti

GeneralPublic,ElderlyandEntrepreneurs

StandardChartered UK https://www.sc.com/en/sustainability/investing-in-communities/financial-education.html

YoungPeopleandEntrepreneurs

MoneyAdviceService UK https://www.moneyadviceservice.org.uk/en

GeneralPublic

EuropeanBankingFederation

EU http://www.ebf-fbe.eu/banks-consumers/financial-education/

GeneralPublic

EuropeanBanking&FinancialServices

TrainingAssociation

EU http://www.ebtn-association.eu/financial-literacy/news

GeneralPublic



Financialinstitution Country Link Emphasis/Targetaudience

Ministry for theFamilyandSocialSolidarity

Malta https://0d2d5d19eb0c0d8cc8c6-a655c0f6dcd98e765a68760c407565ae.ssl.cf3.rackcdn.com/29aad6157295ca5fe74d885d8f69bfab3d3074cc.pdf

GeneralPublic

Malta FinancialServicesCentre

Malta http://mymoneybox.mfsa.com.mt/info/consumer-advice

GeneralPublic

BancaEtica Italy http://www.bancaetica.it/git/firenze/articolo/locandina-scuola-popolare-di-economia-2016


CassaPadanaBcc Italy http://www.popolis.it/linclusione-finanziaria-degli-stranieri/http://www.cassapadana.it/Dettaglio.asp?IdPagina=1&IdNews=193&Id=

YoungPeople,Immigrants

BankofAmerica,oneofthelargestbankholdingcompaniesintheUnitedStatesintermsofassetsoffersfinancialeducationmaterial,comprisingmostlyonlinearticlesonadiversemixof topics, e.g. credit, saving and budgeting, debt, home buying, taxes, personal banking,payingforschool,familiesandmoney,workandincome,andretirementplanning.As a final case from the United States,Citi Group, an Americanmultinationalinvestmentbanking and financial servicescorporation with branches in 140 countries, offers an e-learning trainingprogramme targeting the general public, fromchildrenand teenagers toadults. The main areas covered by the materials are savings, investment planning andbudgeting.OutsideNorthAmericaandtheEU,CreditSuisse,agloballyactiveintegrateduniversalbankproviding solutions to its clients in private banking, investment banking and assetmanagement, is a distinct case in terms of its provision of financial literacy training. Thebankinggroup'sinitiativeonfinancialliteracytraininginvolvesmorepracticalactivities,withlearningprogrammesinBrazil,China,IndiaandRwanda.Furthermore,themaintargetsaregirlsandyoungwomen.AhighlightistheFinancialEducationforGirlsprogramme,aimingtoproviderelevantandtimelyfinancialeducationtogirlsandyoungwomenatkeypoints intheirlives,aspartofaprogrammetoprovidegirlswiththenecessaryknowledge,skillsandattitudestotackledecisionsalongtheirlife’sjourney,whetherthisispersonal,professionalorincontinuingeducation.In addition to the efforts by individual financial institutions for the enhancement of thefinancial skills of their clientele across the EU, there are a number of efforts initiated bybankingassociationsandinstitutionalplayers,suchascentralbanksandministries.Whileanextensivereviewoftheseeffortswillberelevanttoworkpackage3andisbeyondthescopeofthissection,theseeffortsaresummarizedinTable3–5.Theexistinginternet-basedschemesacrossEuropeancountrieshavebeenreviewedandarelisted in Table 3–6. The focus here is identifying tools/channels used to deliver financialeducation. In general, there are multiple channels of delivery of financial literacy. For



Table3–5Listoffinancialassociationsengagedinfinancialeducation

Country Institution Initiative Link

Austria AustrianBankersAssociation

Workshopentitled"debt-freethroughlife",whichwasaimedatprovidingbasicfinancialeducationtailoredtodifferentagegroups

−

Belgium BelgianFinancialSectorFederation

TVseriesonthefinancialworldaimedatthepublicatlarge;creationofaninteractivewebsiteaimedatprovidingfinancialeducation

−

Bulgaria AssociationofbanksinBulgaria

Anumberofactivitiesaimedatenhancingfinancialliteracyincluding:websitesonfinancialeducation,publicationofeducationmaterials,dictionarieswithfinancialtermsandperiodicals.Also,inadedicatedsectiononitswebsiteaimedatinformingthepublicaboutthefinancialproductsandservicesofferedbyfinancialinstitutions.

−

Cyprus AssociationofCypruscommercialbanks

Adhocseminars −

CzechRepublic

CzechBankingAssociation

Creationofaneducationalwebsiteinpartnershipwithotherprofessionalassociationsfromthefinancesector.

http://www.financnivzdelavani.cz/svet-financi

Denmark DanishBankersAssociation

MoneyGuide:awebsite,whichcontainsupdatedpricesforfinancialproductsandservices,aimingtoenableuserstocomparethesebeforemakingaselection.Italsocontainsadictionaryoffinancialterms

http://www.pengepriser.dk


Toolthataidsintheestimationofthecostofborrowing.

http://www.aaop.dk


Entrepreneurs'guide:awebsitewhichprovidesanentrepreneurwithconcreteinformationonthefinancialaspectsofstartingupacompanyOtherinitiativesincludespecialisedseminars

Websitewasnotavailableonthereport.


WebsitesforChildren http://www.pengeby.dk/

Estonia EstonianBankingAssociation

Vikerraadio:Aseriesofradiobroadcastonpersonalfinancetopicssuchassavingsandinvesting

−

Finland FederationofFinnishfinancialservices

Publicationofbooklets(e.g."YouandYourbank");promotionofnationwidecompetitionsaboutfinancialknowledgeinschoolsandshortTV-series

−

France Frenchbankingfederation

Financialliteracyplatform http://www.lesclesdelabanque.com/

France Frenchbankingfederation

Financialliteracyplatform http://www.lafinancepourtous.com/

Germany AssociationofGermanbankers

Financialliteracywebsite http://verbraucher.bankenverband.de/


Schulbank:aninitiativeaimedatpromotingfinancialliteracyinschools

http://schulbank.bankenverband.de/



Country Institution Initiative LinkGermany AssociationofGerman

bankersSchulbankercompetition:sponsoringfinancialliteracycompetitiononfinanceandthewidereconomy

http://schulbanker.de/


Financialliteracyteachingmaterialsonmoneyandconsumptionforschools

http://www.unterrichtshilfe-finanzkompetenz.de/index.htm

Germany Assoc.ofGermanbankers Publicationofbookletsandotherinitiatives −Greece − − −Hungary − − −Iceland Icelandicfinancialservices

associationProvisionofinformationsessionsonfinanceinsecondaryschoolsandcolleges

−

Ireland Irishbankingfederation Financialeducationforschoolagedchildrenandadultsinpartnershipwithotherstakeholders

www.financialeducation.ie

Italy ItalianBankingAssociation

PattiChiariConsortium www.pattichiari.it

Italy BankofItaly Financialeducationwebpageandclassroom-basedfinancialeducationforschoolchildren

http://www.bancaditalia.it

Italy CONSOB Publicationofdocumentsaimedateducatingthepublicaboutinvestments

http://www.consob.it

Italy ItalianTreasury http://www.dt.tesoro.itLuxembourg LuxembourgBankers

AssociationPromotesfinancialeducationeventsandprovidesrelevantinformationindedicatedsectionoftheirwebsite

http://www.abbl.lu/

Malta MalterBankersAssociation

Productionanddisseminationofinformationbooklets/brochuresonawiderangeofpersonalfinancetopics.rganisesseminarsonfinancialtopicsofinterest

−

TheNetherlands

NetherlandsBankersAssociation

Co-founderofCentiQ,aplatformwhichhasfinancialeducationamongitsaims.Co-organisedatelevisionprogrammeaimedateducatingyoungpeople

−

Norway NorwegianFinancialServicesOrganisation

Provisionoffinancialeducationtoschools −

Poland PolishBankingAssociation Severalinitiatives,i.e.:Financialeducationforum,evaluationoffinancialeducationprogrammes,financialeducationsectionofwebsite,supportingfinancialeducationprogrammesonradioandnewspapers,

−

Portugal PortugueseBankAssociation

Publicationofbookletsonfinancialtopics,creationofabankinggamewhichsimulatesthemanagementofabankbranch,provisionofseminarsandonlinecoursesonpersonalfinancetopics,e.g.anopencourseentitled"What'sabank?"

−



Country Institution Initiative LinkSpain SpanishBanking

Association/BankofSpainProvisionofawebportalonfinancialeducation.

http://www.finanzasparatodos.es/

Spain SpanishBankingAssociation/BankofSpain

Provisionofawebportalonfinancialeducation.

http://www.bde.es/clientebanca/home.htm

Sweden SwedishBankingAssociation/FinancialSupervisoryAuthority

Financialeducationprogrammesinschools,creationofafinancialeducationwebsite

http://www.finanskunskap.se/

Switzerland SwissBankersAssociation − https://www.iconomix.ch/

Source:2009EuropeanBankingFederationreportonfinancialliteracyprogrammesbyitsmembersacrosstheEU.This was also supplemented by searching the OECD Database of programmes across the world:http://www.financial-education.org/program.php; http://www.financial-education.org/gdofe.html;http://www.consumerclassroom.eu/resources/theme/financial-literacy#.VxSaJUwrK71instance,Habschicketal.(2007)reportedthat48%offinancial literacyschemesinEuropeusefourormorechannelssuchaswebsites,leafletsandbrochures,printedhandbooksandtrainingcourses.Habschicketal.(2007)providesacomprehensivereviewoffinancialliteracyschemesacrossEurope up until 2007. The report’s findings suggest that there is a wide distribution offinancial literacy schemes across Europe. However, only a few countries are activelyengagedintheprovisionoffinancialeducationsuchastheUK,GermanyandAustria,withEasternEuropeancountriesbeinglessactiveinthisarea.Furthermore,theyfoundthatmostschemeswere targeted at children and young adults in classroomsor universities,with afocuspredominantlyonmoneymanagementandfinancialplanning.About25%ofschemeswere targetedat low-incomeand low-educationgroupsandweremostlyofferedbynon-profit organisations and consumer-protection agencies. A significant number of financialliteracyschemesareinternetbasedandoftenincludeinteractiveelementssuchasgames.All national authorities, non-for-profit organisations and private financial institutions areactivelyinvolvedintheprovisionoffinancialeducation.Additionally, Error! Reference source not found. reviews the more interactive websites,hich provide free financial education resources to enhance financial literacy acrosscountries. These are offered by non-for profit organisations, government authorities orprofessional bodies. For instance, in the US, the institute of certified public accountantsprovides a free website aimed at educating the general public on personal finance(http://www.360financialliteracy.org/); the Australian securities and investmentscommissionalsoprovidesasimilarwebsite(https://www.moneysmart.gov.au/). IntheUK,one of such websites is the “money advice service” (www.moneyadviceservice.org.uk),created as part of the government’s initiative to improve financial literacy across theUK.Similar websites are also available across other EU3countries. In general, these websitesprovide an interactive platform for educating the public on personal finance matters.Content varies across these sites but is mostly centred

3OtherreviewshavementionedEuropeanwebsitewhichprovidesfinancialeducation(Dolceta.eu).However,itappearstonolongerexistorwasplannedbutnotyetimplemented.



Table3–6ListofonlinefinancialeducationinitiativesintheEuropeanUnion

Country Organisation Website Content Availability

Europe-wide

− http://www.consumerclassroom.eu/resources/theme/financial-literacy#.VxSaJUwrK71

- Public

Europe-wide

− http://threecoins.org Aimstodevelopfinancialliteracyamongyoungadults

Public

Austria AustrianNationalBank

https://www.oenb.at/Ueber-Uns/finanzbildung.htmlandhttps://www.eurologisch.at/

Financialeducationmaterialsincludinginformationonmonetarypolicy.Contentisaratherlimitedasseveralkeypersonalfinancetopicsarenotincluded,e.g.insurance

Public

Austria AustrianNationalBank

www.schuldner-hilfe.at FinancialDiversLicense-Aprogrammewhichisaimedatdeliveringfinancialeducationtoyoungpeople.

Public-Non-webbased-Youngpeople

Austria CivilSociety-Klartext

www.schuldenkoffer.at Afinancialeducationprogrammeaimedatteachingyoungadultsaboutmoneymanagement

Public-Non-webbased-Youngpeople

Belgium Belgianfinancialsectorfederation

http://www.monargentetmoi.be/fr

Payments,loans,savings,investmentandbanking

Public

Croatia Charity-PersonalFinanceCounsellingAssociation

http://www.zivotuplusu.info/ Insurance,banking,leasing,savingsandpensionfunds

Public

CzechRepublic

CzechBankingAssociation

http://www.financnivzdelavani.cz/

Banking,financialintermediation,investing,financialmarkets,leasing,payments,insuranceandpensions

Public

Denmark DanishShareholdersAssociation

http://www.shareholders.dk/investorskolen

Bonds,mutualfund,stockinvestmentandanalysis,pensions

Non-webbasedandtargetedatinvestors

Denmark MinistryofEconomicandBusinessAffairs

www.pengeogpensionspanelet.dk

Financialeducation

Denmark DanskeBankGroup http://www.danskebank.com/en-uk/CSR/Financialliteracy/Pages/Moneyville.aspx

Financialliteracy

Denmark DanskeBankGroup http://www.controlyourmoney.dk/

Financialliteracy Public-audience:10-15

Estonia EstonianFinancialSupervisionAuthority

http://kool.minuraha.ee/ Financialliteracy Public-childrenupto12yearsold

Finland BankofFinland http://www.euro.fi/ Financialliteracy




France FrenchBankingFederation

http://www.lesclesdelabanque.com/

Financialeducation

France FrenchBankingFederation

http://www.lafinancepourtous.com/

Financialeducation

France SocieteGenerale https://www.abcbanque.fr/# Costofgoodsandmanagingpocketmoney

Public-children6-10

France BankofFrance www.citedeleconomie.fr/-Games-

Financialliteracygames Public

Germany AssociationofGermanBankers

http://verbraucher.bankenverband.de/

Financialliteracy Public


http://schulbank.bankenverband.de/

Financialliteracy Public-youth


http://www.unterrichtshilfe-finanzkompetenz.de/index.htm

Financialliteracyresourceforteachers

Public

Greece Civilsociety-Internationalandregionalinitiatives

http://www.interfima.org/project-dignity/

Avarietyofpersonalfinancetopicssuchasbanking,insurance,savingsetc

Non-webbased

Hungary CentralBankofHungary

http://www.penziskola.hu/ Financialeducation Public-Secondaryschoolstudentsandteachers

Iceland Instituteforfinancialliteracy-ReykjavikUniversity

http://www.fe.is/ Financialeducation

Ireland CompetitionandConsumerProtectionCommission

http://www.consumerhelp.ie/ Budgeting,banking,savingsandinvestments,insurance,mortgage,loansandcredit,pensions,lifestagesandfinancialfraud

Public

Ireland Moneymanagementandbudgetingservice

http://www.yo-yos.ie/ Alimitedrangeoffinancialeducationtopicssuitedforyoungpeople

Public-youth

Ireland Moneymanagementandbudgetingservice

http://www.moneycounts.ie/ Earningandspendingmoney,budgeting,banking,saving,borrowing,inheritance

Italy ItalianBankingAssociation

www.pattichiari.it Offeringfinancialeducationtohighschoolsandthewiderpublicincitysquares,

Public-youth

Italy BankofItaly http://www.bancaditalia.it Financialeducationwebpageandclassroom-basedfinancialeducationforschoolchildren

Public-youth

Italy CONSOB http://www.consob.it Financialplanning,rightsanddutiesofinvestors,bonds,shares,derivatives

Investors

Italy ItalianTreasury http://www.dt.tesoro.it Italy Insuranceregulator

(IVASS)http://www.educazioneassicurativa.it/guide-pratiche/

Insurance Public




Italy Consumercreditresearchcentre-UniversityofRome-"TorVergata"

www.monitorata.it Consumercredit Public

Lithuania EconomicEducationDevelopmentCentre

http://www.essc.lt/ Financialeducation Public-Childrenuptograde12

Luxembourg Luxembourgbankersassociation

http://www.abbl.lu/en/financialeducation-introductionpage/

Financialeducation Public

Netherlands CivilSociety-NIBUD https://www.nibud.nl/consumenten/themas/kinderen/

Financialeducation Public-youth

Netherlands Ministryoffinance http://www.wijzeringeldzaken.nl/

Personalfinancetopicsincludingsavings,mortgages,pensionetc.

Public

Norwary NorwegianConsumerCouncil

www.finansportalen.no Enablescomparisonofdifferentfinancialproducts

Public

Poland BankPolski http://www.pkobp.pl/sko/lokatka/

Financialeducation Public-Children

Portugal BancodePortugal http://clientebancario.bportugal.pt/pt-PT/DireitosdosClientes/Paginas/default.aspx

Personalfinancetopicsfocusedmostlyonbankingproducts

Public

Slovakia Ministryoffinance http://www.fininfo.sk/sk/ja-a-financie/planovanie-osobnych-a-rodinnych-financii

Awiderangeofpersonalfinancetopicsincludingbanking,savings,investments,loan,leasing,insuranceetc.

Public

Slovakia ChildrenofSlovakiaFoundation

http://www.poznaj.sk/index.html

Personalfinanceforchildren

Public

Slovenia SecuritiesMarketAgency

http://vlagatelj.atvp.si/ENG/ Personalfinancetopicsfocusedmostlyoninvestments

Public

Spain Ministryofeducation http://www.finanzasparatodos.es/

Widerangeofpersonalfinancetopics

Public

Spain Financialstudiesinstitute

http://edufinanciera.com/es/ Widerangeofpersonalfinancetopics

Public

Spain PrivateSectorOrganisation

http://www.jubilaciondefuturo.es/es/

Websitededicatedtocreatingsocialawarenessaboutthepensionssystemsituation

Public-Adult

Sweden Swedishbankingassociation

http://www.finanskunskap.se/

Sweden Publicauthoritiesinassociationwithcivilsocietyandprivatesector

http://gilladinekonomi.se/ Awiderangeofpersonalfinancetopics

Public



Table3–7Listofotheronlineresources

Source Website Content Access

FinancialConsumerAgencyofCanada

http://www.fcac-acfc.gc.ca/Eng/resources/Pages/FLRDSAT-OAEBDRLF.aspx?WT.mc_id=CFLDHPBPERMENG

Databaseandself-assessmentquiz Public

Visa http://www.practicalmoneyskills.com/

Awebsitewhichprovidesfreeeducationalresourcesincludingpersonalfinancearticles,gamesandlessonplans

Public

Jumpstart/Doughmain http://www.irulemoney.com/

Awebsitededicatedtoteenagerswhereexpertsanswertheirquestionsaboutmoneyandmoneymanagement

Public

NorthwesternMutual http://www.themint.org/index.html

Awebsitewhichprovidestoolsandresourcestoteachchildrenandteenstomanagemoneywiselyanddevelopgoodfinancialhabits

Public

AICPA http://www.360financialliteracy.org/

Awebsitewhichoffersgeneralinformationformanagingpersonalfinances.Thewebsiteisstructuredtoprovidetailoredinformationfordifferentcategoriesofusers.

Public

AustralianSecuritiesandInvestmentsCommission

https://www.moneysmart.gov.au/

Awebsitewhichprovidesfinancialeducationandguidancetothepublic.Contentincludes:managingmoney,borrowingandcredit,insurance,retirement,investingandfinancialscams

MutualFundDealersAssociationofCanada

http://www.mfda.ca/investors/Quiz/Quiz-r-01.html

Fraudawarenessquiz Public

MoneyAdviceService(UKGovernment'sinitiativeonaddressingfinancialilliteracy)

https://www.moneyadviceservice.org.uk/en

Contentisstructuredintosectionsasfollows:DebtandBorrowing/Pensionsandretirement/Births,deathsandfamily/Budgetingandmanagingmoney/Workandredundancy/Insurance/SavingandInvestments/Benefits/HomesandMortgagesetc.

Public

FederalFinancialLiteracyandEducationCommission(USA)

http://www.mymoney.gov/Pages/default.aspx

Thewebsiteisorganisedaroundfiveprinciples:Earn/SaveandInvest/Protect/Spend/Borrow.Italsocontainsasectionforentrepreneurswithlinkstorelevantinformationsuchaslinkstosourcesoffinance.

Public

EuropeanCommission http://www.consumerclassroom.eu/

AEurope-basededucationwebsite

PersonalFinanceEducationGroup

http://www.pfeg.org/ Awebsiteaimedatteachingfinancialandbusinessskillstochildrenandyoungadults

Public

on personal finance topics including debt and borrowing, pensions and retirement,budgeting, insurance, savingsand investment,benefits and taxes,mortgagesetc.Contentmaybestructuredbyusergroupsdefinedbyage,profession,maritalstatusor lifestages.There is generally a high level of interaction with the users as most websites featureinteractive elements such as financial calculators,webchats, forums, newsletters, links torelated blogs or other resources, videos, games, quizzes etc. These features enhancelearningandusers’engagement.



3.4.2 Newspaperswithonlinepersonalfinancesections

Information about personal finance is provided by several broadsheet newspapers. Ingeneral,thisiscontainedina“personalfinance”or“money”sectionoftheirwebsites(e.g.,“La Repubblica” in Italy; “The Guardian” in the UK; “DieWelt” in Germany). Some newschannelssuchastheBBC(BritishBroadcastingCorporation)andCNN(CableNewsNetwork)alsoprovidesimilarinformationontheirwebsites.The content of these webpages, reviewed in Table 3–8, comprises mostly news articles,analysisand reportsonavarietyofpersonal finance topics. In somecases, thecontent isclassified into separateheadings relating toanareaofpersonal finance.For instance, theTelegraphnewspaperintheUKclassifiesitspersonalfinance-relatedcontentintoseparatepagesasfollows:investment,property,bills,insurance,pensions,banking,taxandfinancialservices. Content and categories vary greatly across newspapers, with some beingmuchmorecomprehensivethanothers.The focus on news reports about personal finance provides a usefulmeans for updatingknowledge and raising awareness of changes to existing legislation and their impact onfinancialdecisions.However,itmaybearguedthatsuchbenefitislimitedbytheextentofknowledgeonealreadypossessesaboutthesetopics.Forinstance,newsreportsonpensionreforms may have little value for someone with little or no prior understanding of thepensionsystemandtheimportanceofretirementplanning.Furthermore,accesstothecontentofthesewebsitesmayberestrictedtoregistereduserswithpaymentrequired,e.g.theFinancialTimesintheUnitedKingdom.Thetargetaudienceis generally not specified but could be regarded as not suited for children, very youngpeopleandpeoplewithlowlevelsofliteracywhomaystruggletounderstandthecontent.Finally,newspapers’personalfinancesectionsalsofeatureinteractivetoolstoengageusersand enhance financial knowledge, though this varies significantly across newspapers.Examples of such features include: an “Ask the expert” sectionwhere users can interactwith financialexpertsbyemailingpersonal finance-relatedquestions,whichareansweredandcommentedonbyotherusers;interactivetoolstoaidfinancialdecision-makingsuchasmortgage repayment, savings, debt and stamp duty calculators; stock and fund screenertools; glossary of financial terms; newsletters to which users can subscribe to; tools tocompareaccounts,creditcardsandmortgagesacrossdifferentfinancialservicesfirms;andshortvideosandpodcastsonpersonalfinance.Overall, personal finance information offered by newspapers on their websites providepractical and interactive means for individuals to enhance their financial knowledge.However,usersmayrequirepriorfinancialknowledgetomaximisesuchbenefit.



Table3–8Listofnewspaperswithpersonalfinancesections

Newspaper Link Content Public/Licensed

TheGuardian http://www.theguardian.com/uk/money

Contentisclassifiedintosectionsasfollows:Money/Property/Savings/Pensions/Borrowing/Careers

Registrationnotrequired

FinancialTimes

http://www.ft.com/personal-finance

Contentisclassifiedintosectionsasfollows:PropertyandMortgages/Investments/Trading/Pensions/Tax/BankingandSavings

Registrationrequired

Dailyexpress http://www.express.co.uk/finance/personalfinance

Contentisnotclassifiedintosections.Mostlyalistofbriefnewsreportsandanalysisonavarietyofpersonalfinance-relatedtopics.


Telegraph http://www.telegraph.co.uk/money/

Contentisclassifiedintosectionsasfollows:Investing/Property/Bills/Insurance/Pensions/Banking/Tax/Financialservices


USAtoday http://www.usatoday.com/money/personal-finance/

Contentisnotclassifiedintosections.Mostlyalistofnewsreportsandanalysisonavarietyofpersonalfinance-relatedtopics.


WallStreetJournal

http://www.wsj.com/public/page/news-personal-finance.html

Contentisclassifiedintosectionsasfollows:Familyfinance/Wealthadviser/Taxes/Retirementplanning/Realestate/

Registrationrequired

Forbes http://www.forbes.com/finance/#36a00ad31f3a



Bloomberg http://www.bloomberg.com/personal-finance



TheIndependent

http://www.independent.co.uk/money

Contentisnotclassifiedintocategories.Mostlyalistofbriefnewsreportsandanalysisonavarietyofpersonalfinance-relatedtopics.


CNN http://money.cnn.com/pf/



ThisisMoney http://www.thisismoney.co.uk/money/index.html

Contentisclassifiedintosectionsasfollows:Markets/SavingandBanking/Investing/Bills/Cars/Holidays/Cardsandloans/Pensions/Mortgagesandhome

Registrationrequiredonlyforspecificsectionsandfinancialtools

NewYorkTimes

http://www.nytimes.com/pages/your-money/index.html?8qa

Contentisclassifiedintosectionsasfollows:Investments/loans/planning/retirement/credit/insurance


WashingtonPost

https://www.washingtonpost.com/business/get-there/?nid=menu_nav_business-personalfinance

Contentisclassifiedasfollows:Buildit(EarningandInvesting)/Financeit(Borrowing)/Protectit(Insuranceandfinancialfraud)/Saveit(Savingsandretirement)/Spendit.


BBC http://www.bbc.co.uk/news/business/your_money





Newspaper Link Content Public/Licensed

TheTimes http://www.thetimes.co.uk/#section-money

(Paid)Registrationrequiredtoaccesscontent

InvestorsChronicles

http://www.investorschronicle.co.uk/your-money/

Contentisorganisedintocategoriesasfollows:Portfolioclinic(forinvestors),PensionandSIPP(retirement),Property,Paylesstax,savingsandfinancialplanning.

(Paid)Registrationrequiredtoaccesscontent

3.4.3 Openonlinecourses(MOOCs)

The number of financial literacy/personal finance courses provided by online educationalplatforms free of charge and with material available at all times (that is, regardless ofwhetherthecourseiscurrentlyrunning),isquitelimited.ThecoursesarepresentedinTable3–9andashortsummaryfollows:The Open University offers three courses that could be classified as financialliteracy/personal finance courses. The main Open University course is “Managing myMoney”; the other two, namely “Managingmy investments” and “Estimating the cost ofequity”,areextensionstotheformerone,focusingprimarilyonthefieldofinvestmentsandpresentingmorecomplexandsophisticatedconcepts.However,thecoursesoftenoverlap;sometimesevenpresentingtheexactsamematerialintheexactsameway.SincetheOpenUniversity is aUK institution, a quite large portion of thematerial is primarily focused ininformationandconceptsthataremorecountry-specificratherthanuniversallyappliedorrelevant(mostsectionsofthe"Pensions"chapterbeinganexample).Thematerialismostlypresentedinaformakintoabook,astextisthemainelementoftheOpenUniversitycourses.Thisisnotanidealformofpresentationforanonlinecourse,asitmakesitratherunattractive.Manyofthesectionsalsoincludevideoand/oraudiosegmentsin certain sections, aswell as graphs, charts and tables tohelpexplain some conceptsorgive historical data/information, which occasionally improve the presentation of thematerial. The contentof the videoandaudio segmentsmainly includes: talksor lectures,interviews with experts, news segments, street interviews and animated explanations offinancialconcepts.Severalofthesesegmentsareinterestingandusefuladditions,butotherare too long or lack in relevance or specificity. It is suggested that 3 hours perweek bedevotedwhentakingthiscourse,whichmightbealargeamountoftimeforcertaingroupsofpeople(e.g.employeesandstudents).Therearefewinteractiveelementsinthiscourse;specifically, insomeofthesectionstherearesomeactivitiesrequiringthecoursetakertogivearesponseorchooseananswerfromalist,generatinganautomaticresponsewiththecorrectanswerandanexplanation.“Personal and Family Financial Planning” by the University of Florida is similar to“Managing my Money” in that they are both built around a model on how to build apersonalor/andhouseholdfinancialplan.Inaddition,thiscourseisalsofocusedinaspecificcountry, namely the US, meaning that at least a part of it is not universally relevant.Moreover,theycoverapproximatelythesameeconomicandfinancialconcepts.



Thematerial is presented in a form that resembles aUniversity lecture. Prof.Gutter, thecourse’screator,presents thematerial invideos,usingslides,whicharealsoavailable fordownload. This method of presentation might be tiring for some people and make thecoursemoredifficult to follow. Thismight be exacerbatedby the lack in structureof theslides.Moreover, thevideopresentationsareoftenvery long,with theirduration rangingfrom16to32minutes.Anotherproblemmightbethetimethatissuggestedtobedevotedtothiscourse:5-7hours foreachmodule,meaning40-56hours intotal.Thiscoursedoesnotfeatureanyinteractiveelements.Inthe“FinancialAwarenessandConsumerTrainingforStudents–FACTS”programmebythe Higher Education Services Corporation (HESC) of New York State, the material isprimarily focusedonUShighschoolandcollegestudents.Becauseofthat, itcoversmuchfewer financial and economic concepts and subjects than the other courses. Thisprogramme could be interesting with regard to the way the material is differentiatedaccording to the targeted age group. However, it seems that the last module, which issupposedtobetargetedtowardshighschoolstudents,ismostlyasummaryofthepreviousthree, with slight additions or omissions so as to make it more relevant to high schoolstudents.Thematerialisalsopresentedinslides.Inaddition,althoughitistargetedathighschoolandcollegestudents, itdoesnotuseanyof the features thatwouldmake itmoreattractivetoyoungpeople(e.g.videos,graphs,infographics,etc.).“Financial Literacy” byAdvance Learning is a course offered in the platform ALISON, inpartnershipwith Three RiversWorkforce Investment Board. Like all the aforementionedcourses,thisoneisalsoprimarilyfocusedonthecontextoftheUnitedStates.Thiscoursehasseveral interestingfeatures, includingtheuseofanimationforthepresentationofthematerial(althoughthequalityoftheanimationisquiteunimpressive),therecommendationonwhichtopicstopaymoreattentiontowhenstudyingthefollowingmodulebaseontheperformanceinthepre-assessmentquizzesbeforeeachmodule,theembeddedglossary.Italsofeaturessomeinteractiveelements,includingthepre-assessmentquizzesbeforeeachmodule, short assessmentswithin someof the sections, a savingsbar and the right-handsideofthescreen,whichincreasesordecreasesdependingontheanswerstothecourse’squizzesandassessments.Regardingcontent, several subjects (suchas is inflation)arenotanalysedvery thoroughlyandthelevelofdetailsprovidedisquitelow.Hence,theamountoftimeforthecompletionofthiscourseisshorterincomparisontotheothercourses,estimatedtobe45-90minutesforeachofthe7modules,or6-10hoursintotal.“Financial Literacy”, offered in Allversity, is a course is built for the country of Uganda.Hence,abigportionofthematerialisnotrelevantformoredevelopedwesterncountries.SincetheAllversitywebsitewasshutdown, thematerialof thecourse isonlyavailable inWordformat,andthereisnoinformationregardinghowthematerialusedtobepresentedintheplatform.



Table3–9Listoffinancialliteracy/personalfinanceopenonlinecourses

rse Institution Platform Link

Managingmymoney

TheOpenUniversity

OpenLearn http://www.open.edu/openlearn/ocw/course/view.php?id=1166

Managingmyinvestments

TheOpenUniversity

OpenLearn http://www.open.edu/openlearn/ocw/course/view.php?id=1269

Estimatingthecostofequity

TheOpenUniversity

OpenLearn http://www.open.edu/openlearn/money-management/estimating-the-cost-equity/content-section-0

PersonalandFamilyFinancialPlanning

UniversityofFlorida

Coursera https://www.coursera.org/course/uffinancialplanning

FACTS NewYorkState

- https://www.hesc.ny.gov/partner-access/high-school-counselors/financial-awareness-and-consumer-training-for-students-facts.html

3.4.4 Financialgamesandgamifiedtools

Videogamesarefastbecomingapopulartoolforenhancingfinancialcapability.Videogameson personal finance are provided by a variety of organisations, including non-profitorganisations(e.g.D2Dfund),governmentagenciesandauthorities(e.g.FINRA–FinancialIndustryRegulatoryAuthorityintheUSandtheECB–EuropeanCentralBankinEurope)andprivatefinancialinstitutions(suchasVisaandDanskebank).

Thecontentofvideogamesvariessignificantly:someareaimedatassessingusers'financialliteracythroughtestsandquizzeswhicharepresentedusingagame’ssettingsratherthanthe traditional quiz format; others require users to make a series of financial decisions(usuallysimulatingreal-lifesituations)inavirtualsetting.Intheprocess,usersareexposedtopersonal financeconceptsandareable tocomprehendhowcurrent financialdecisionsimpactonfutureoutcomes.Videogames provide an interactive tools that enable users to enhance their financialcapabilities through real life simulations of personal financial decision-making. However,manygamesaredesignedforchildrenandteensandaremostlyfocusedonlimitedfinancialtopicssuchasmoneybasics,savings,budgetingandcredit.While technology-based financial education continues to evolve, the following internet-basedvideogamesarepresentedascurrentlyavailablefreeresources.Itshouldbenoted,however, that this list is innowayexhaustive. The relevant videogamesare reviewed inTable3–10.Bite Club, Farm Blitz, Celebrity Calamity, and Groove Nation are financial literacy gamesdevelopedbytheDoorways2DreamsFund(D2D),partofRAND’sFinancialLiteracyCentre.These games are targeted towards low-income, minority adults with a limited



understanding of the English language. These games “have great potential because theyfacilitate repetitive engagement with lessons and create a ‘performative’ learningenvironment”(Tufano,Flacke,&Maynard,2012).BiteClub(http://financialentertainment.org/play/biteclub.html)issetwithinthecontextofaclub forvampiresand isaccompaniedbyadancebeat.Playersare required to face thedialectics of personal finance:meeting current consumption needs, eliminating debt, andsaving for retirement.Byplayingthegame,agreaterunderstandingofannualpercentagerate(APR)anditsimpactondebtrepaymentisdeveloped.FarmBlitz(http://financialentertainment.org/play/farmblitz.html)isset‘downonthefarm’and is completewithmusic. It is reminiscent of the popular game Farmville available onFacebook. On this farm, however, debt exponentially multiplies like bunnies. Playersmanagedebtandsavingsbutwillalsoneedtoprepareforfinancialemergencies,dueto,forexample,catastrophicweatherconditions.Byplayingthegame,agreaterunderstandingofhigh-interest,short-termdebt,compoundinterest,andnetworthisdeveloped.Celebrity Calamity (http://financialentertainment.org/play/celebritycalamity.html) ismarketed towardswomenages18-35.Players takeon the roleof financialmanager foracelebritywhohasdifficulty keepingher finances in order. The celebrity asks the financialmanager tomake specific purchases for upcoming trips, to sharewith her friends, or forentertainmentpurposes.Theplayerchooseswhethertopaywithacreditcardordebitcardand as a result gains greater insight into finance charges, credit limits, fees and annualpercentage rates (APR). Beyond managing finances, players must work to ensure thecelebrity’shappiness.Theadditionof thispsychologicalelement is interestingas researchdemonstratesthatincomecanhaveanimpactonhappiness(Kahneman&Deaton,2010).Alsomarketedtowardswomenages18-35,GrooveNationisabout“dancing]yourwaytoahealthy budget” (http://financialentertainment.org/play/groovenation.html). Similar toCelebrity Calamity, players serve as financialmanagers. Income, fixed and variable costs,debt, short- and long-term savings, as well as upgrades for dance performances aremanaged. The financial manager learns to set aside earnings from the dancer’s job andperformancesinordertomakeitfromtowntotown.Asthelocationchanges,sodofixedcosts,suchasrent.GrooveNationteachesplayershowtomanagecurrentconsumptionandsaveforfuturegoals.Visa has developed the Financial Football for football enthusiasts. The game simulates asoccer game inwhich users assume the position of a player of a football team.When inpossession of the ball, users are required to answer a question on a variety of personalfinancetopics. If theanswer iscorrect, theball ispassed forwardtowards thegoal; if theanswer is wrong the ball is stolen by the opposing team. Users can choose the level ofdifficultybasedontheiragegroup:Rookie(ages11-14),Pro(ages14-18)andHallofFame(18+).Aspartofthegame,resourcesarealsoprovidedonfinancial topicsassessed inthegame. Educatorsmayuse such teaching resources toeducate studentsprior to thegame(http://www.practicalmoneyskills.com/games/trainingcamp/ff/play/).



Table3–10Listoffinancialliteracygames

Source Link TitleofGame Aimofgame TargetAudience

D2DFund http://financialentertainment.org/

BiteClub Save for retirement whilerunningavampirenightclub

General Public butaimed at low incomeandminorityadults


Celebritycalamity Managecelebritycreditcardandspending



FarmBlitz Manage farm resources tobuild savings and survivefinancialemergencies



GrooveNation Dance and budget on theroadtoLA



RefundRush Helpclientssplittaxrefundsandsaveduringtaxtime.


VISA http://www.financialsoccer.com/play/

FinancialSoccer A basic game to assess thelevel of financial literacyusingasoccertheme

Public with differentlevelofdifficultybasedonplayers'age.

VISA http://www.practicalmoneyskills.com/games/trainingcamp/ff/

FinancialFootball A basic game to assess thelevel of financial literacyusing a (American) footballtheme

Publicwithdifferentlevelofdifficultybasedonplayers'age.

VISA http://www.practicalmoneyskills.com/games/

Countdowntoretirement

A game that enables usersto make financial decisions,incl.savingforretirement.Ametergaugeshowwelltheyareprepared for retirementastheyprogress

Publicbutmostlyrelevantforadultsratherthanyoungchildren


Ed'sbank Agamethatteachesplayersto save for an item in apiggybank

Youngchildren

VISA http://www.practicalmoneyskills.com/games/moneymetropolis/

MoneyMetropolis

Agamethatteachesplayersto set a savings goal, earnmoney and save towardsthatgoal.

Youth/Children


Roadtriptosavings

Agamethatteachesplayerstoearnandsavemoney

Public

SIFMAFoundation

http://www.stockmarketgame.org/expstudent.html

Stockmarketgame

Agamethatteachesplayersaboutinvesting

Youngstudents

MTV http://www.indebted.com/the-game/debtski/

Debtski Teachesplayerstolimitexcessivedebtsandmaximisesavings

Youth



FreedomInstitution

https://demo.moneyisland.com/

MoneyIsland Teachesfinancialresponsibilitythroughthevirtualworld.

Children

Council forEconomicEducation

http://www.genirevolution.org/

GenIRevolution Teachesstudentsavarietyoffinancialconceptsthroughsixteeninteractivemissions

MiddleandHigh-school

Callystro http://www.finlitgame.com/

Journey towardsfinancialliteracy

Financialeducation Children

iGrad http://www.igrad.com/money-smarts-game

Money smartsgame

Financialeducation Youngadults

Jumpstart/Doughmain

http://www.doughmain.com/pub/money-games-for-kids/the-fun-vault/

Thefunvault Financialeducation Children5-8

Jumpstart/Doughmain

http://www.sanddollarcity.com/

Sanddollarcity Financialeducation 8-12yearolds

EuropeanCentralBank

http://www.ecb.europa.eu/ecb/educational/economia/html/index.swf?width=1000&height=750&language=en

Economia Explain how monetarypolicyworks

Public

EuropeanCentralBank

http://www.ecb.europa.eu/ecb/educational/inflationisland/html/index.en.html

Inflationisland Explain the effect ofinflation

Public

DanskeBank

http://moneyville.ie/ For Danish:http://www.pengeby.dk/

Moneyville Financialliteracy Children5-9

DanskeBank

http://www.dreamonthegame.dk/

Dreamon Financialliteracy 15-17yearolds

Age-appropriate lessonson credit, debt, savings, spending, and financial planningare alsoprovided. For those individualswho prefer a different type of football, Financial Soccer isalsoavailable(http://www.financialsoccer.com/).Debt Ski (http://www.indebted.com/the-game/debtski/) is a fairly simple financial literacygamedevelopedbymtvU,MTV’scollegenetwork,andthePeterG.PetersonFoundation.Apiggybankdrivesa jetski throughroughwaters,collectingcoins,wanted items,andbasicnecessities. If sufficient fundsarenotsaved,a financial tsunamicouldset theplayerback.This game drives home a prudentmessage: in order to financially succeed, youmust livewithinyourmeans.Thisincludesmaximizingsavingandminimizingdebt.Ahappinessscore



is also included in this game. It is impacted by debt, savings, and the provision of lifenecessities.Moneyville isagamedevelopedaspartofDanske’sBank'sfinancial literacyprogramme.Itwaslaunchedin2008andtargetedtochildrenaged5-9yearsold.Theaimofthegameistogive children anunderstandingofmoney and teach themabout incomeandexpenditure.Childrenareintroducedtoavirtualtownwheretheycanearnmoneydoingdifferenttaskssuch as sorting parcels at the post office. They are also required to save up in order topurchase items they would like. This enables children to understand the principle ofbudgetingandprioritisingwhenspending.

3.4.5 Books

There isavastrangeofbooksonpersonalfinance.AGoogleBookssearchofthekeyword“personalfinance”returnedoveronemillionhits.“Financialliteracy”returnedover200,000hits.Theseare invariousformats,suchascoursetextbookswithend-of-chapterexercises,guides or self-help books for personal improvement. For instance, there are a number ofpersonal finance titles in thepopular“fordummies”bookseries.Titlesvary,butgenerallyinclude“personalfinance”,“personalfinancialliteracy”or“personalfinancialmanagement”,butmayalsobefocusedonasinglepersonalfinancetopicsuchasinvesting.Theircontent(ascanbededucedfromthetablesofcontents) includesawiderangeofpersonalfinancetopics, suchas savings, investments, credit, insurance,bankingand related concepts,withsomebooksbeingmorecomprehensivethanothers.Anumberofbooksalsoincludetopicsonpublic finance such asmonetary and fiscal policies.Authors range fromeducators andacademicstoexperiencedfinancialplannersandjournalists,etc.Whilemost books are targeted at the general public, a number of books are specificallydirectedtowardscertaingroups.Groupsmaybeagespecific(suchasindividualsinthe20s,30s or post-50 years of age) or by country (some books specifically mention the targetcountry such as Canada or Australia) or by marital status (such as singles, divorcees,newlyweds/youngcouples)orbyoccupation(somebooksaredirectedtoindividualsinthearmedforcesorentrepreneurs)etc.ThesearereviewedinTable3–11.

3.4.6 Europeaneducationalplatforms

This section is dedicated to a short reviewof the online educational platforms in Europe,includingwhethertheyprovideanycoursesonfinancialeducation.Thesecanbedividedintwogroups:theonesfinancedbytheEuropeanUnionandtheones,whichareindependent.

OpenupEd (http://www.openuped.eu/) is essentially a website providing links and shortdescriptions to theMOOCs that its partners provide in their own platforms or websites.These partners include Universities, Open Universities and Technical Universities fromseveralmainlyEuropeancountries(aswellasTurkeyandIsrael).Coursesareofferedinthelocal language inmanycases.ThepersonalfinancecoursesthatareprovidedaretheonesfromtheUKOpenUniversity(seesection3.4.3).



Table3–11Listoffinancialliteracyandpersonalfinancebooks

BookTitle Year Author Publisher Content

Personalfinancialliteracy 2012 JoanRyan Cengage •Howchoicesaffectincome•Income,benefitsandtaxes•Yourpurchasingpower•Financialdecisionsandplanning•Bankingsystem•Personalriskmanagement•Buyingdecisions•Preservingyourcredit•Creditproblemsandlaws•Basicofsavingandinvestment•Savingandinvestingoptions•Buyingandsellinginvestments

Financialliteracyformanagers:FinanceandAccountingforbetterdecision-making

2012 RichardLambert

Whartondigitalpress

Tableofcontentnotincludedinthepreview,butessentiallyaboutAccountingandFinancialManagementforManagers

TheStudent'sGuidetoFinancialLiteracy

2012 RobertLawless

ABC-Clio •Thevalueofaneducation•Theimportanceofsavings•Investments•Principalsofinvesting•Debt•Taxconsiderations•Insurance•TheEntrepreneur•Economics•The10mostcommonmistakestoavoid

TeachingFinancialLiteracyThroughPlay

2015 ChristopherHarris,PatriciaHarris

RosenClassroom

Tableofcontentsnotaccessibleonline

TheCompletePersonalFinanceHandbook

2007 TeriClark Atlanticpublishingcompany

Tableofcontentnotaccessibleonline

TheEverythingPersonalFinanceinYour20sand30sBook

2012 HowardDavidoff

Adamsmedia

Awiderangeofpersonalfinancetopicssuitedforthoseintheir20sand30s

UltimateGuidetoPersonalFinanceforEntrepreneurs

2007 PeterSanderandJeffreyLambert

McGraw-HillCompanies

Tableofcontentnotaccessibleonlinebutthebookaddressespersonalfinancefromtheviewpointofanentrepreneur

Open Education Europa (http://www.openeducationeuropa.eu/en) is a EuropeanCommissionInitiative“toprovideasinglegatewaytoEuropeanOpenEducationalResources(OER)” with the main goal “to offer access to all existing European OER in differentlanguages inorder tobeable topresent themto learners, teachersand researchers”. Forthispurpose,amongotherthings,linkstoMOOCsfromEuropeanInstitutionsareprovidedinthewebsite,similarlytotheOpenupEdprojectabove.AnothersimilaritytoOpenupEdisthat



theonlyavailablefinancialliteracycoursesaretheonesprovidedbytheUKOpenUniversity(see section3.4.3).Asof 17/12/2015, theplatform is “inhibernation”, as it is undergoingredesignandnonewmaterialisuploaded;however,theexistingmaterialisavailable.

ECO (E-learning Communication Open-Data) (http://project.ecolearning.eu/) aims to“extend to a pan-European scale the most successful MOOC experiences in Europe” byimplementing “amergedMOOCplatform integratingdifferentmodulesprovidedby someECOpartners”.CoursesareofferedinmultipleEuropeanplatforms,butcurrentlytherearenofinancialliteracycourses.

EMMA (European Multiple MOOC Aggregator) (http://platform.europeanmoocs.eu/) is amultilingualplatformstillinitsbetaphase,whichaimstooperateintwoways:asahosttotheMOOCsofitspartners,thatisseveralEuropeanUniversities,andalsoas“asystemthatenableslearnerstoconstructtheirowncoursesusingunitsfromMOOCsasbuildingblocks”.Asofnow,nofinancialliteracycoursesareoffered.

Asaconclusion,theEU-financedplatformsseemtohavethesamemaingoal(oratleastoneoftheirmaingoals):tocreateasinglehubfortheMOOCsofferedbyEuropeaninstitutions.All of them offer courses in multiple European languages, but financial literacy/personalfinance is a subject that is addressed only by the Open University among the Europeaninstitutionsinvolvedintheprojects(atleastintheEnglishlanguage).

Most of the independent European platforms offering MOOCs are initiatives either byindividualUniversitiesorprivate companies.Manyof themoffer coursesonly in the locallanguage.Onlytwoofthemofferfinancialliteracycourses:theIrishALISONandtheGermanAllversity(forashortreview,seesection3.4.3).

3.5 AsynthesisofbestpracticesandthespaceforPROFITnovelty(SYNTHESIS/NOVEL)

The provision of financial literacy has become a major policy target. As discussed abovethereisalargegapintheliteratureonfinancialeducationastowhatisthebestmethodofdelivering financialeducationcourses. IthasbeenarguedbyLusardietal. (2015) that theresearch infinancialeducationshouldmoveawayfromadiscussionastowhethergenericfinancial education works. Instead, the research should focus on trying to gain anunderstanding of how to make financial education work through better design andappropriate delivery methods. By focusing on how to make financial education work todelivermore effective courses andproducebetter informed individuals, it is important tolook at the different ways that financial education can be taught and whichmethodwillresultinthemostinformedparticipants.One study thathas lookedat this issue is Lusardi et al (2014),whichdeveloped fourneweducational programmes focusing on risk diversification. The four programmes weredelivered online and comprised of: an informational brochure, a visual interactive tool, awrittennarrative,andavideonarrative.Thestudywasdelivered to892participants fromthe American Life Panel (ALP), which is a representative panel composed of 6,000 UShouseholds,andtheparticipantswereeachallocatedtooneofthefourprogrammesortoa



control group. The study found improvements in the financial knowledgeof individuals ineachofthefourdifferenttreatmentgroups,whichhelpedtoshowthatevenshortdurationinterventions can help to improve financial literacy. The improvements in financialknowledgewasgreatestinthevisualtoolandvideogroupswhichledtheauthorstosuggestthat educational programmes that engage the user emotionally or physically (such aswatchingavideoorusingavisualtool)canbeparticularlyeffectivecomparedtotext-basedor passive and educational programmes (such as reading a narrative) in improvingindividuals knowledge of complex financial matters like risk diversification. This is animportant result as it encourages the use of interactive tools when designing financialeducation courses as using these tools provides the greatest improvement in knowledgeamongstparticipants.Heinberg et al (2014) is another study that aimed to assesswhichmethod of teaching ismosteffectivewhendeliveringafinancialeducationcourse.Forthisstudytheydesignedandtestedafinancialeducationcoursecalled‘FiveSteps’,whichfocusedonfivebasicfinancialplanning concepts that relate to retirement. The five financial conceptsweredelivered toparticipants either through a video or written narrative, with five simple short storiesdescribingthefinancialconceptsverballyandthebenefitsoftakingtheappropriateaction.The study was delivered through the ALP with participants either being exposed to thewrittennarratives,videos,bothmethods,orplacedinthecontrolgroup.Theresultsshowedthat despite only being exposed to the treatment for a short period of time (each of thevideos orwritten narratives only took about 3minutes), individuals showed a substantialincreaseintheirfinancialknowledge.Thetwoabovestudiesarebothveryimportantastheyindicatethatevenashortdurationcoursecanimprovethefinancial literacyof individuals.Thishaspotentialbenefitstoeducatorswhenlookingathowtoimplementfinancialliteracycourses into the school curriculum as these short duration courses should be able to beeasily implemented into the school curriculumwhilst still helping to improve the financialliteracylevelsoftheparticipants.Thus, the PROFIT project comes at a very timely fashion in the policy agenda for theadvancement of financial awareness in the population. Its contribution towards thefinancial-literacyenhancementgoalisintendedviathegenerationofanall-inclusivetoolkit.This will be based on the generation of novel training material from experts in financialeducation. Moreover, it will be complemented by a rich collection of readily availablematerial that is scattered in different locations. Such material involves videos, narrativeswithhistoricalexamples,gamesandinteractivecontentsuchasquizzes.Thegenerationofthematerialisintendedtobeuser-centredandinteractive.Withintheseaims, Task 1.2 (Use Requirements and Application Scenarios) of Work Package 1 (WP1 -Requirements and Functional Design) aims to elicit the views of user groups of primaryinterest, along with bank managers and financial advisors. Based on the review in theprevioussections,theusergroupsthathavebeenidentifiedasofthehighestrelevanceareshowninTable3–12.Theoutcomeofthisexercisewillbediscussedindetailinthenexttaskof this workpackage. It aims to elicit responses regarding the users' needs in terms offinancialtraining,basedon(a)alargesampleofuser-grouprespondentsinaquestionnairesurvey(some300respondents),and(b)interviewswith18cooperativebankexecutivesandfinancialadvisors.



Table3–12PROFITuser-groups

1) Entrepreneurs/latententrepreneurs/socialentrepreneurs/self-employed2) Elderly/retirees/pre-retirees3) Migrants/Membersofanethnicminority4) Children/Parentsofyoungchildren5) Customers:Indebted/Overindebtedhouseholds6) Customers:Investors/Potentialinvestors/Depositors7) Unemployed/trainees8) Activecitizens/taxpayers9) Mortgageowners/homeowners/first-timebuyers10) Professionalsinfinancialservices/financialexperts11) Governmentexecutivesandpolitical-partymembers/localauthorities12) Collectiveinvestors/borrowers/thirdsectororganisationsSource:PROFITpartners'compilationofusergroupsfortask1-2The design of the material in Work Package 3 (WP3 - Financial Literacy for InformedDecision-Making)willbebasedonthefollowingthreeprinciples:(1)generationofprimarytraining material, incorporating novel elements such as notions of public finance andentrepreneurial finance. The aims of the training are to provide users with an enhancedunderstandingofthepivotalfeaturesofpersonalfinancialplanning,e.g.(i)interestrates;(ii)thetimevalueofmoney;(iii)consumerborrowing;(iv)residentialmortgages;(v)savingsandinvestment products; (vi) short and medium-term saving; (vii) saving for retirement I:planning; (viii) inflationand theerosionofpurchasingpower; (ix) personal taxes; (x) basicprobability;(xi)riskversusreturn;(xii)riskdiversification;(xiii)savingforretirementII:riskand retirement planning; (xiv) entrepreneurial finance and planning; (xv)mistakes peoplemake: behavioural and other biases in personal finance; (xvi) public finance for activecitizenship;(b)thecriticalanalysisofexistingpracticesandresources,aimingataselectionofthebestavailableexistingresources;(c)theincorporationofnoveltechniquesaimingatinteractivecontentandincentive-compatibleuserparticipation.The assessment of financial literacy will be based on the following four key principles(LusardiandMitchell,2014):(1)Simplicity:Themeasurementoftheunderstandingofbasicfinancial concepts, is foundedon thenotionsof the very basics of finance; (2) Relevance:Questionsmustrelatetoconceptspertinenttopeople’sday-to-dayfinancialdecisionsoverthe life cycle.Moreover, theymust capture general rather than context-specific ideas. (3)Brevity:Duetotimeandspacelimitationsinmostsurveys,thenumberofquestionsmustbekept to aminimum in order to securewidespread adoption. (4) Capacity to differentiate:Questionsmustdifferentiatebetweenfinancialknowledgelevels,soastocomparepeopleintermsoftheirscoresonacommonsetofquestions.Finally, the design of the material will adapt to the insights from the literature oninformation technology and online education. Insights there suggest that the newgeneration of learners is cradled in technologies, communication, and an abundance ofinformation.Asa result, thedesignof learning technologiesneeds to focuson supportingsociallearningincontext.Alongwithdeliveringcontentthataimsto“teach”thelearnerthebasicprinciplesandtheirextensions,itiswellwithinPROFIT'stargetstoalso(a)supportthelearner in finding the right content (right for the context, the user, the purpose, and



pedagogically); (b) support learners to connect with the right sources of informationprovision,and(c)motivate/incentivizeuserstolearn.Thesetsofprinciplesabove,alongwiththebestpracticesfollowedintheresourcesoftheprevious sub-sections will shape the generation of a very novel and ambitious financialliteracytrainingtoolkitatWorkPackage3(WP3-FinancialLiteracyfor InformedDecision-Making)ofthePROFITproject.



4 FinancialandEconometricForecasting

4.1 MarketSentimentIndicators

In this section it isanalysed the importanceof sentimentproxies in forecasting the futureinnovations of the real economy. In economics, it has been for some decades recognisedthatbusinesscyclefluctuationsareanoutcomeoflesstraditionalsourcessuchassentimentshocks. A long strand of the theoretical literature aims at uncovering the transmissionmechanismsofcausalitybetweenuncertaintyshocksandeconomicgrowth.Theirempiricalcounterparts facea lotofdifficulties inestimatingthiscausality.Thedifficultiesstemfromthe fact that existing measures of sentiment are far from perfect and in fact are bestdescribedasproxiesratherthanrealmeasures.Thereisalsoaneedfordevelopmentofnewproxies of sentiment that take into account the time horizon of sentiment, the type ofsentiment(demandversussupply,policy,etc.)andfinallythenatureofuncertainty.Under this vein, in the subsections that follow we try to set the ground for the nextdevelopmentsneedtotakeplaceinthefieldofsentimentanalysisinthefinancialsector.Forthat reason, we analytically expose how sentiment is treated in theoretical models ineconomics, how it is being estimated and finally what is its overall contribution to theforecast of subsequent economic growth. The best practices described belowwill be thegaugesforthedevelopmentofnewsentimentproxiesandforecastingmodelsforanalysingthecoresegmentsofthefinancialsector.

4.1.1 Theoreticaljustificationsofsentimentaffectingtheeconomy

"Even apart from the instability due to speculation, there is the instability due to thecharacteristicofhumannature thata largeproportionofourpositiveactivitiesdependonspontaneousoptimismratherthanmathematicalexpectations,whethermoralorhedonisticoreconomic.Most,probably,ofourdecisionstodosomethingpositive,thefullconsequencesofwhichwillbedrawnoutovermanydaystocome,canonlybetakenastheresultofanimalspirits—a spontaneous urge to action rather than inaction, and not as the outcome of aweightedaverageofquantitativebenefitsmultipliedbyquantitativeprobabilities."

Keynes,JohnM.(1936),pp.161-162.Therecentglobalfinancialcrisisinbetween2007and2009wasindeedspermaticandleadeconomists,policymakersand investors to realize that the financial sectorplaysa centralrole in themodern economy. Up to today economies still face the deepmacroeconomicimpactoftherecentfinancialcrisisandasDeGrauweandMacchiarelli (2015)state,DSGEmodellers as a recognition of the nexus of the financial sector and the real economy,attempt to introduce financial intermediaries and the banking sector formally in theirmodels, to describe understand and hopefully predict in a positive manner economicdownturns and booms. Scholars, policymakers and investors regard as the origins of the2007-2009 financial crisis the persistent lack of consumer confidence in the 2008 U.S.subprime crisis. Indeed, Akerlof and Shiller (2009) argued for the necessity of an activemacroeconomicpolicytorestoretheconfidenceofconsumersandbusinessesandenhanceeconomic activity as a path out of recession. Most governments and central banks did



undertake measures to restore confidence and economic activities following policiesstemmingfromdifferentschoolsofeconomicthought.ManycentralbanksastheFEDandECB for example undertook measures, both conventional and unconventional, towardmonetaryeasing.Ontheotherhandgovernmentsundertidefinancialconditionsattemptedtoreducedeficitsofevenbalancetheirbudgettorestoremarket,consumerandbusinessesconfidence, under the IMF supporting notion of contractionary expansion and othersfollowedKeynesianexpansionarypolicies.Asscholarsandpolicymakersconvergetowardsaconsensusoverthenexusand impactofthefinancialcycletotherealbusinesscycleandviceversa"investorsentimentcanhaveaprofoundimpactontheeconomy,fuellingboomsandexacerbatingbusts"Lutz(2015).WhileLutz(2015)bringstoourattentionalargeandgrowingliteratureontheeffectsofsentimentonfinancialmarketsandtheeconomy, forexample,Shiller (2000),BrownandCliff (2004),Tetlock(2007),KlingandGao(2008),Kurov(2008),Brunnermeier(2009),Schmeling(2009),Fungetal.(2010),Gençayetal.(2010),Chen(2011),Lux(2011),Singeretal.(2011),Chungetal.(2012),Garcia(2013),andLutz(2015),thestorygoesbackasearlyasKeynes(1936).InspiredfromKeynes,researchersattemptedtoshowwhetherinvestorsentimentcanaffectasset prices as people with high or low sentiment tend to form optimistic or pessimisticjudgments and choices. Keynes (1936) argued that markets can exhibit fluctuations, thatcannot be attributed to commonly modelled exogenous stochastic processes, stemmingfrom the influence of investors “animal spirits”, which tent to move prices in a way notconsistenttofundamentalsfeaturesoftheeconomyandfirm-consumerbehaviour.Indeed,asBakerandWurgler(2007)state,thereareseveralstockmarketeventsthathada"dramatic" level or change in stock prices that seems to defy explanation. Someof theseincidentsBakerandWurgler (2007) refer toasmorecharacteristicare, theGreatCrashof1929,the'TronicsBoomoftheearly1960s,theGo-GoYearsofthelate1960s,theNiftyFiftybubble of the early 1970s, and the Black Monday crash of October 1987. As Baker andWurgler (2007) argue the standard finance model, where unemotional investors alwaysforce capitalmarket prices to equal to the rational present valueof expected future cashflows, has considerably difficulty fitting these patterns. A promising attempt in explainingunconventional innovations inprices isdevelopingbyanapproachofbehavioural finance,wherethestandardmodelfinancialmodelisaugmentedtoincorporatetheassumptionthatinvestorsaresubjecttosentiment.“Animalspirits”was introducedasanotiontodescribethe instincts and emotions that influence humanbehaviour and thus could be proxied byinvestor sentiment, and as Baker andWurgler (2007) state can be defined broadly, as abeliefaboutfuturecashflowsandinvestmentrisksthatisnotjustifiedbythefactsathand. InaseminalpaperbyAzariadis(1981),arguethatpurelysubjectiveagentscansetinmotionbusinesscycles,observingshiftsinanyfactorandthenanimalspiritsorconsumersentimentaffecting price innovations, simply because they are expecting them to change, and pricesignals convey no structural information. Azariadis (1981), show that "even perfectlywellbehaved economies will typically admit rational expectations equilibria in which theexpectations themselves spark fluctuations in the level of business activity. If manyindividualsnaivelybelivethatsunspotsorsomeindexofconfidencearegoodpredictorsoffutureprices,thentheymaytakeactionswhichtendtobearouttheirbeliefs."



De Long, Shleifer, Summers, and Waldmann (1990) in a simple overlapping generationsmodelofanassetmarketinwhichirrationalnoisetraderswitherroneousstochasticbeliefsbothaffectpricesandearnhigherexpectedreturns,maketheroleofinvestorsentimentinfinancial markets formal. They demonstrate that if uninformed noise traders base theirtrading decisions on sentiment and risk-averse arbitrageurs encounter limits to arbitrage,sentiment changes will lead to more noise trading, prices can diverge significantly fromfundamentalvaluesevenintheabsenceoffundamentalrisk,andleadtoexcessvolatility. Zhang(2008)attemptstosummarizehoweconomistshavetraditionallyviewedsentiment.The first strand are asset-pricing theories of classical finance approach asset prices asrationalequilibriumofexpectedfuturepayoffs,reflectingthatchangesinpricesareaffectedbyexternalnewsabout futurecashflowsandthemovementof interestrates,anddonottake into account investor sentiment. The behavioural finance approach assumes thatinvestorsentimentmaysignificantlydistortmarketoutcomestherebyaffectingassetpricesin equilibrium. In the spirit of De Long, Shleifer, Summers, andWaldmann (1990), Zhang(2008)pointsoutthatthenoisetradermodelpositsthatiftherearelimitstoarbitrageandinvestor beliefs are correlated, then noise unrelated to fundamentals, such as sentiment,may lead asset prices to deviate from what is expected from the benchmark of marketefficiency. In amore recent paper Angeletos and La’O (2013) attempt to address the question "Arebusiness cycles and fluctuations in asset markets driven by changes in preferences andtechnologies? Or are the driven by animal spirits, market psychology and self reinforcingwavesofoptimismandpessimism?".Toanswerthisquestiontheyattempttodevelopanewtheory of fluctuations, one that helps accommodate the notions of “animal spirits” and“market sentiment” in unique-equilibrium, rational-expectations, macroeconomic models,developing a neoclassical model economy by allowing trading to be random anddecentralized.Thekeyfeatureoftheirmodelisthedevelopmentofanovelformalizationofextrinsic movements in market expectations, one that requires neither a departure fromrationality nor the introduction ofmultiple equilibria. Theirmain finding is that economicoutcomesandmarketexpectationsmayco-move inresponsetoacertaintypeofextrinsicshocks namely “sentiments”. In this manner, they manage to rationalize random andseemingly inexplicableshifts intheoptimismorpessimismthateconomicagentsmayholdaboutoneanother’schoicesandtherebyfuturemarketconditions. WhileAngeletosandLa’O(2013)proposingthe inclusionoftheirnotionofshocks inDSGEmodels to estimate their contribution to observed business cycles, and study policyimplications inacontextofassetmarkets,DeGrauweandMacchiarelli (2015)highlightingthatthelastfinancialcrisiswasmadepossiblebythefactthatagentsdonotunderstandthecomplexity of the world in which they live, and their cognitive abilities are very limited,argue the need to introduce a banking sector in models that recognize the cognitivelimitationsofagents.InthatspirittheyextendthebehaviouralmacroeconomicmodelofDeGrauwe (2012) to include a banking sector and produces endogenous and self-fulfillingmovements of optimism and pessimism (animal spirits). Their main result is that theexistence of banks intensifies thesemovements, creating a greater scope for booms andbusts. Thus, banks do not create but amplify animal spirits. De Grauwe andMacchiarelli(2015)showthatanincreaseintheequityratiosofbankstendtoreducetheimportanceof



animalspiritsoverthebusinesscycle,highlightingalsotheroleofthecentralbankto“tametheanimalspirits”,bystabilisingoutput.

4.1.2 Themainandmostcurrentsentimentindexesusedsofarintheliterature

"Investorsentiment isnotstraightforwardtomeasure,but there isno fundamental reasonwhyonecannotfindimperfectproxiesthatremainusefulovertime"

BakerandWurgler(2007)As there is an on growing body of literature attempting to shed light on unconventionalasset price movements, not related to fundamentals, and economic performance, manyattempt to found agent "adverse" selection or "irrational" behaviour on the grounds ofbehaviouralfinanceandeconomics.Whatisindisputablychallenginginthisstrata,ishowtoextract from realeconomicand social activity thenecessaryand sufficient information, toconstruct proxies that can be closely related to micro level and macro level economicbehaviorandexplainor forecastpatterns thatwhere insufficientlybeenexplained. to thisdirectiontherearetwomaindistinctpathstoextractsentiment,themoretraditionalwhichisbasedon structureddata, and theamore innovativeandgainingmomentumbasedonunstructureddata.Sentimentindexproxiesbasedonstructureddataresultintheutilizationofexistingdatabasesandusedirectlyasproxies,quantitativedata,ormoresophisticatedindexes constructed on these data. On the other hand, unstructured data, which isqualitativedata,ascanbederivedbymeansoftextualanalysisorsurveymethods,havetoproperlyanalysedandhandled inordertobequantifiedandusesassentimentproxies. Inthissectionwewillattempttopresentandcategorizethemainproxyindexesofsentimentinitsbroadsenseusingthedistinctionofstructured-unstructuredbaseindexes.4.1.2.1 StructuredsentimentindexesInthissubsectionwepresentsomecommonlyusedstructuredsentiment indexes, indirectstructuredsentimentindexesandnewlyproposed.Baker and Wurgler (2007) present and comment on some of the most typical PotentialSentimentProxiesusedintheliterature,whichcanberegardedintheclassificationweuseas structuredsentiment indexes:1) retail investor trades;2)mutual fund flows;3) tradingvolume; 4) premia on dividend-paying stocks; 5) closed-end fund discounts; 6) optionimplied volatility; 7) first-day returns on initial public offerings (IPOs); 8) volume of initialpublicofferings;9)newequityissues;10)andinsidertrading. AsitisseeninBakerandWurgler(2007);GreenwoodandNagel(2006);Barber,Odean,andZhu (2006), the inexperienced retail or individual investor is more likely than theprofessional tobe subject to sentiment. In thisperspectiveKumarandLee (2006) suggestthatsentimentmeasuresforretailinvestorsbasedonwhethersuchinvestorsarebuyingorselling,sincethisisconsistentwithsystemicsentiment. Another sentiment proxy can be easily derived byMutual Fund Flows, since data on theallocationacrossfundcategoriesbyfundinvestorsarewidelyavailable.ForexampleBakerand Wurgler (2007) refer to Brown, Goetzmann, Hiraki, Shiraishi, and Watanabe (2002)



wheretheyproposeanoverallmarketsentimentmeasurebasedonhowfundinvestorsaremoving into and out of, for example, “safe” government bond funds and “risky” growthstockfunds. Regarding Trading volume, ormore generally liquidity, is regarded by Baker andWurgler(2007) as an investor sentiment index based on the argument made by Baker and Stein(2004). The later argue that irrational investors adds to liquiditywhen they areoptimisticand betting on rising stocks rather thanwhen they are pessimistic and betting on fallingstocks. Anotherwaytorevealsentiment istoextract informationbasedonDividendPremium.Asargued by Baker andWurgler (2007) and Baker andWurgler (2004a, b), dividend-payingstocksmaybeinverselyrelatedtosentiment.AsstatedinbyFamaandFrench(2001)whendividendsareatapremium,firmsaremorelikelytopaythem,andarelesssowhentheyarediscounted. As Baker and Wurgler (2007) explain, firms appear to cater to prevailingsentimentfororagainst“safety”whendecidingwhethertopaydividends. Towards a different approach it is argued that that if closed-end funds aredisproportionatelyheldbyretailinvestors,theaveragediscountonclosed-endequityfundsmay be a sentiment index,with the discount increasingwhen retail investors are bearish(seeZweig(1973),Lee,Shleifer,andThaler(1991),andNealandWheatley(1998).AcommonlyusedsentimentproxyistheMarketVolatilityIndex(VIX),whichmeasurestheimpliedvolatilityofoptionsontheStandardandPoor’s100stockindex,isoftencalledthe“investorfeargauge”bypractitioners.AsBakerandWurgler(2007)explain,OptionImpliedVolatility can be derived by options pricing models as a function of options prices. Theimplied causality is when an asset has greater expected volatility options prices rise, andthuscapturingtheinvestorfearasitisexpressedthroughexpectedvolatility. BakerandWurgler(2007)arguesthattheenthusiasmofinvestorsmaycauseinsomecasesremarkable returns on first trading day of Initial public offerings, yielding IPO First-DayReturnsproxy.AtthesametimeIPOvolumedisplayswildfluctuations,duetothefactthatthe underlying demand for initial public offeringsmay be extremely sensitive to investorsentiment. AbroadermeasureofequityfinancingactivitywhichcanberegardedasAnotherpotentialproxy of investor sentiment that usually draws investor's attention is the equity share oftotalequityanddebtissuesbyallcorporations,whichmeasuresallequityofferings,andnotjustIPOs.InanearlierstudyBakerandWurgler(2000),showthathighvaluesoftheequityshareportendlowstockmarketreturnsthatmayimproveforecastsonmarketreturns. Regarding Insider Trading Baker and Wurgler (2007) argue that if sentiment leads tocorrelated mispricings across firms, insider trading patterns may contain a systematicsentiment component, captured by the insiders decisions to form their portfolios withrespecttooutsideinvestors.

4.1.2.1.1Indirectstructuredsentimentindexesmeasures



BrownandCliff (2005) findthatmanycommonlycited indirectmeasuresofsentimentarerelated to direct measures (surveys) of investor sentiment. In their study they utilizefinancialvariablesconsideredbypractitionersasmarketweathervanesandalsoconstructindirect sentiment index measures from structured data. The main consider marketperformance,typeoftradingactivityandderivativevariables. Regarding market performance Brown and Cliff (2005) utilize data for NYSE issues tocalculateamodificationof the ratioof thenumberofadvancing issues todeclining issues(ADV/DEC),thatcanincorporatetheirrespectivevolumes,presentingtheARMSindex. Inorder to capture the typeof tradingactivityBrownandCliff (2005)use thepercentagechange in margin borrowing (DMARGIN) as reported by the Federal Reserve which isregardedasabullishindicator;thepercentagechangeinshortinterest(DSHORTIR)whichisgaragedasabearishindicator;theratioofshortsalestototalsales(SHORTSLS);theratioofspecialists’ shortsales to totalshortsales (SPECIAL) (whenshort-sellingbecomesrelativelylarge,themarketislikelytodecline);andtheratioofodd-lotsalestopurchases(ODDLOT)asbearishmeasure. BrownandCliff(2005)tocapturesentimentindirectlybyderivativestradingactivitytheusethe following variables: the ratio of CBOE equity put to call (PUT/CALL) trading volume(abearishindicator);thechangeinthenetpositioninSPXfuturesbytradertypereportedbythe Commodities Futures Trading Commission (CFTC); the forecasts of commoditymarketreturns collected by Market Vane (MKTVANE) (bullish predictor). They also construct ameasure of expected volatility relative to current volatility using VIX (the S&P 100 Indexoptionvolatility)andSIGistherealizedvolatilitycalculatedfromOpen-High-Low-ClosedataontheS&P100Index.

4.1.2.1.2ProposednewIndirectstructuredsentimentindexes/measures

Taking stock of the innovations on the contraction of new indirect structured sentimentindexesandmeasureswerefertoBrownandCliff(2004)and(2005)asdiscussedpreviously,wheretheyproposeasentimentindex,byemployingPCA(PrincipleComponentAnalysis)tocombineseveralsentimentproxiescommonlyused inthe literature.Also, followingBrownandCliff (2004), Baker andWurgler (2006) construct a sentiment indexbyusing aPCA tocombinesixtypicalproxiesforsentiment. InamorerecentstudyBaker,WurglerandYuan(2011) construct sentiment indices for sixmajor stockmarketsexpanding theapproachofBakerandWurgler's(2006),anddecomposethemintooneglobalandsixlocalindexes. BakerandWurgler(2007)expandtheirpreviousworkandconstructanindexbasedonthesixproxiestheyuseinBakerandWurgler(2006):tradingvolume;thedividendpremium;theclosed-endfunddiscount;thenumberandfirst-dayreturnsonIPOs;andtheequityshareinnewissues,alsotakingintoaccountformacroeconomicindicatorsandtheir impactonthesentimentproxies,asgrowth in industrialproduction; realgrowth indurable;nondurable;and services consumption; growth in employment, and an NBER recession indicator. Theinnovation is that they use the residuals from the regressions of the six basic proxies onmacroeconomicvariablesthenewsentimentproxies.

4.1.2.1.3OtherProxiesofIndirectstructuredsentimentindexes/measuresused



WhilethevastmajorityofstudiesareconsistentwithUSAcases,itisworthnotingastudyofJansenandNahuis(2003)wheretheyusetheEuropeanCommission’sconsumerconfidenceindex (CCI) as a proxy for sentiment and find stock returns and changes in sentiment arepositivelycorrelatedforninecountries,withGermanyasthemainexception.4.1.2.2 UnstructuredsentimentindexesUnstructured sentiment indexes can mainly occur by two types of unstructured datacollection.Thesurveymethodaswellasthetextualanalysisorcontentanalysisorthebag-of-words(BoW)method.

4.1.2.2.1 Unstructuredsentimentindexes:surveys

Inaprevioussectionwedisusedontheindirectstructuredsentimentmeasurescommonlyusedintheliterature.Anapproachtoextractsentimentsimplybyaskinginvestorshowtheyfeel about the market, as argued by Baker and Wurgler (2007) can proxy the marginalirrationalinvestor.ThispathgoesbacktoRobertShillerwhohasconductedinvestorattitudesurveyssince1989.InasimilarapproachUBSandtheGallupOrganizationhaveconductedrandominterviewsofUSinvestorshouseholdsstartingfrom1996to2007,toconstructtheIndex of Investor Optimism. In 2002, UBS extended the Index to France, Germany, Italy,Spain,andtheUnitedKingdom.QiuandWelch(2006)showthatchangesintheUniversityofMichiganConsumerConfidenceIndexcorrelatehighlywithchangesintheUBS/Gallupindex.The Investors Intelligence (II) index is a weekly bull-bear spread by categorizingapproximately 150market newsletters used by Brown andCliff (2005) to forecastmarketreturns.BrownandCliff(2005)alsoutilizedintheirstudyameasurebroadlyunderstoodasof individual investor sentiment, obtained from Surveys conducted by the AmericanAssociationof Individual Investors (AAII).Toconstruct the index, theassociationasksonarandomsampleofitsmemberswheretheythinkthestockmarketwillbein6months:up,down, or the same and then labels these responses as bullish, bearish, or neutral,respectively

4.1.2.2.2Unstructuredsentimentindexes:textualanalysis

Astextualanalysisgainsmomentumsincethe importantworkofTetlock(2007),LoughranandMcDonald(2011),ApelandGrimald(2012)orBakeretal.(2012),themainfocusisnottocreatedifferentquantitativeproxiesbasedon textualanalysis,but rather than improvethe methods and techniques towards improved quality of this proxy by finding theappropriate conversion strategy from qualitative data to quantitative, taking into accountinformationsources,andmethodsofcontendanalysis.Forexampleinarepresentativecase,Tetlock (2007) utilizes the media content of the Wall Street Journal column to measureinvestor sentiment. His main finding is that high media pessimism, which implies lowinvestorsentiment,predictsdownwardpressureonstockpricesfollowedbyareversiontofundamentals. He also relates high market trading volume with an extreme value ofpessimism,consideringthelatterasagoodpredictor. TothisdirectionColmandLiu(2014),inasurveyofmethodsintextualsentimentanalysis,locate three main information sources, corporation disclosures, sentiment that can beextracted by media as in news stories, in-depth commentaries or analyst reports, and



sentimentas itcanbeextractedbythe internet intheformsofposts. Intheysurveytheyrestricttheiranalysiswithrespecttofinance,butthemaininformationsourcescanbeusedforextractingsentimentformoregeneraluse. AsColmand Liu (2014) argue, themost common content analysismethods in the textualsentiment literatureare thedictionary-basedapproachandmachine learning.Thegeneralidea of the dictionary-based approach as Colm and Liu (2014) present it, is that it uses amappingalgorithminwhichacomputerprogramreadstextandclassifiesthewords,phrasesorsentencesintogroupsbasedonpre-defineddictionarycategories,whichisreferredtoasthe‘bag-of-words’. In thedictionary-basedapproachthereare twomain issues,whichwordsshouldbetakeninto account into the dictionary and what is their relative weight. For example in earlierstudiesthemainwordlistscomesfromsocialpsychologyandwheredevelopedby(Stoneetal.,1966), that is theGeneral Inquirer (GI)built-indictionary,whichusedtheHarvard IV-4dictionaries. The GI / Harvard word lists where used for example by Tetlock (2007),Engelberg(2008),Tetlocketal.(2008),Doranetal.(2010),LoughranandMcDonald(2011),Ferrisetal.(2013)andmore. FollowingStone'sapproachHart(2000)developedDICTION5.0,whichcountswordsbasedon33separatedictionaries.ThisdictionarywasusedforexampleinthestudiesofHenryandLeone(2009),DemersandVega(2011),andFerrisetal.(2013). However thisapproachof sentimentextractionwascriticizedandas Li (2010)argues thattheclassificationoftonebasedontheGI/Harvarddoesnotprovidesufficientaccuracy.Toovercomethisdrawbackdictionariesorwordlistsmorecloselyrelatedtofinancehavebeenbuiltbyresearcherssothatmoreaccurateandefficientsentimentscorescanbegenerated,suchaswordlistdevelopedbyHenry(2006,2008). Inmore recent studies,as LiuandMcConnell (2013),andLoughranandMcdonald (2013),thewordlistdevelopedbyLoughranandMcDonald(2011)seemstogainground.TheyuseGI/Harvard negativewords, which have been expanded by inflecting eachword to formsthatretaintheoriginalmeaningoftherootword. Regardingtherelativeimportanceofeachwordfoundinatext,intheearlierstudiestherewhere an equal treatment of words, while Loughran and McDonald (2011a) use twoweighting schemes, a simple proportional weighting and one that weights each wordinverselyproportionaltoitsfrequencyfoundinthetext.InmorerecentstudiesasJegadeeshandWu(2012),theyassignweightsforeachwordbasedonhowthemarkethasreactedtotheminthepast. Ontheotherhandmachine learningreliesonstatistical techniquesto inferthecontentofdocumentsandtoclassifythembasedonstatisticalinference(Li,2010).Thekeydifferenceisthatwhilemachinelearningisalsoregardedasatoolfortextualanalysis,asubsetofthetextisusedasatrainingsetforthealgorithmandmanuallyeachwordinthissetisclassified.



4.1.2.3 LatestinnovationsinsentimentanalysisKontopoulos et al. (2013) argued that sentiment analysis constitutes a rapidly evolvingresearcharea,especiallysincetheemergenceofWeb2.0anditsrelatedtechnologies(socialnetworks,blogs,wikisetc.).Theyhighlight that therecent increaseduseof internet-basedcommunicationviamicro-bloggingandTwittergraspedtheattentionofsentimentanalysisasaprominentrawdatasourcethatcanpotentiallyuncoversentiment.Kontopoulosetal.(2013) criticize machine learning-based approaches that perform sentiment analysis ontweets, and respond to their drawbacks by proposing ontology-based techniques fordetermining the subjectsdiscussed in tweets andbreakingdowneach tweet into a setofaspects relevant to the subject, avoidinguniform scoresof apost, increasing the relevantvalidityoftheposts. AlsoKearneyandLiu(2014)summarizetheimportantandinfluentialfindingsintherelatedliterature about how textual sentiment impacts on individual, firm-level andmarket-levelbehaviourandperformance,andviceversa. Nassirtoussi et al. (2014, 2015) provide an extensive summary on techniques, such asmachine learning on online sentiment analysis, socialmedia textmining, news sentimentanalysis,FOREXmarketpredictionandstockpredictionbasedonnews,highlightingthenewtrend of forecasting economic and financial variables based on the notion of behaviouralfinanceandtheimportanceofsentimentsandmarketperception.Nassirtoussietal.(2015)alsodevelopatechniquetopredictintradaydirectional-movementsofacurrency-pairintheforeignexchangemarketbasedonthetextofbreakingfinancialnews-headlines. Summingup, recentworkhave shown the importanceon theeffects or sentimentor theproxy of sentiments using 1) socialmedia textmining, 2) news, 3) search engine data, 4)tweets.

4.1.3 Therelativeimportanceofsentimentindexesinpredictingthefinancialeconomy

Since the recent financial crisis and the economic crisis succeeded, the theoretical andempiricalmodels seemed inadequate in forecasting the downturn. This fact renewed theattentiononthebehaviouralnatureofeconomicsandfinance,producingagrowingbodyofliterature incorporating traditional sentiment indexes and more modem approaches thatcaptures agents’ sentiment andmarket perception. The empirical literature suggests thattheuseofsentimentmayhaveanimportantrole inpredictingrealeconomyvariablesandfinancialvariables. Carroll et al. (1994) presented evidence that the lagged consumer sentiment has someexplanatorypowerforcurrentchangesinhouseholdspending.McleanandZhao(2014)testwhether recessions and low investor sentiment create financial constraints that limit realinvestmentandemployment. Chen and Chou (2015) show that recession fear does lead to a higher probability ofeconomic downturns; they also provide strong evidence that an increase in marketpessimismmaypushtheeconomyfromanexpansionstatetoarecessionstate.Chenand



Chou (2015) also found some weaker evidence suggesting that a lack of consumerconfidence may trap the economy in a depressed regime longer. They conclude that agreaterlackofconfidenceindeedpushestheeconomyintoadownturn. Hüfner and Schröder (2002) have analysed the forecasting qualities of German leadingindicatorsby focusingon the ifobusiness expectations, the Economic Sentiment Indicatorfor Germany of the European Commission, the PurchasingManagers’ Index and the ZEWIndicatorofEconomicSentiment.HüfnerandSchröder(2002)provideevidencethat,exceptfortheESIN,allotherindicatorsusedintheirstudyareleadingtheyear-on-yeargrowthrateofindustrialproduction. Matsusaka and Sbordone (1995) using vector autoregressions, investigate the link ofconsumer confidence and economic fluctuations. When controlling for fundamentaleconomicvariablestheirmodelsfailtorejectthehypothesisthatconsumersentimentdoesnot cause GNP, highlighting the role of consumer sentiment as an important informationcriterionthatcanimproveGNPforecasting. BatchelorandDua(1998)alsouseconsumerconfidenceandtheyshowthattheinclusionofthisvariablewouldhavebeenhelpfulinpredictingthe1991USrecession.Howevertheyalsoshowthatthisisnotaresultthatcanbegeneralizedinotheryears,andastheyarguethisisdue to the nature of the recession rather than a persistent weakness in the forecastingtechniqueinuse. InasimilarstrandregardingtheusefulnessoftheconsumersentimentindicatorCarrolletal.(1994), Bram and Ludvigson (1998), Howrey (2001), and Ludvigson (2004), argue thatconsumer sentiment can provide important information in explaining consumer spendinginnovations. Howrey (2001) also considers the relationship between consumer confidence and thebusiness cycle, showing that the use of the consumer confidence index increases theaccuracyofone-tofour-quarter-aheadforecastsoftheprobabilityofrecession. LeavingoftheconsumerconfidenceindexasideandusingthePurchasingManager’sIndex(PMI), Dasgupta and Lahiri (1993) show that the index provides additional explanatorypowerinthepredictionofbusinesscyclesturningpoints. FurthermoretherelatedliteraturehasshownthatbusinesssentimentisusefulnotonlyforforecastingGDPand thebusiness cyclebut is also important forGDPnowcasting (see forexample Kauffman 1999; Klein and Moore 1991; Lindsey and Pavur 2005; Banerjee andMarcellino2006;andLahiriandMonokroussos2013). Assentimentindicesseemtohaveincreasedsignificanceforforecasting,thismaybeduetotwo characteristics that theyhave in comparison to traditional leading indicator variables.First, theyareavailable inhigher frequencies and in somecases theyareavailable in realtime and second, by their statistical nature are not subject to revisions (see for exampleKoenig2002).AndasLahiriandMonokroussos(2013)notetherearenoeconomicvariablesof similar importance that are available with the same timeliness. On the other hand,



macroeconomic variables and leading indicators are announcedwith a lag (amonth or aquartermostofthecases)andinmostcasesarerevisedconsiderable.ArecentexampleistheadoptionofESA2010insteadESA1995byEurostatandmemberstates. Christiansen et al. (2014) in an empirical paper for the case of US recessions, provideevidencethatsentimentvariablesplayanimportantroleenhancingthepredictivepowerofrecessions, outperforming classical recession predictors and in excess of common factorsestimatedbasedonlargepanelofeconomicdata.Christiansenetal. (2014)arguethatthebest predictions for future recessions are obtained by the combination of sentimentvariableswith the classical recessionpredictors,orby combining sentiment variableswiththecommonfactors,bothin-sampleandout-of-sample.Christiansenetal.(2014)separatetheirapproachfromtherelatedliteraturebyusingsentimentvariablestoforecastabinaryrecessionindicatorusingprofitmodelsratherthanforecastingcontinuousvariablessuchasGDPgrowth.With this approach they show that the term spread, the short rate, and thestockmarketreturncarryimportantinformationaboutfuturerecessions. Schmeling(2009)usingasampleof18advancedcountriesshowthatsentimentnegativelyforecasts aggregate stock market returns on average across countries. Schmeling (2009)arguesthatwhensentimentishigh,futurestockreturnstendtobelowerandviceversa.Aninteresting finding in this paper is that the predictive power of sentiment is stronger forshortandmedium-termhorizons.OhandSheng(2011)ontheotherhand,showthatstockmicro blog sentiments do have predictive power for simple andmarket adjusted returnsrespectively. They also relate this predictive accuracy with the under reaction hypothesisobservedinbehaviouralfinance. Huangetal.(2014),followingBaker,andWurgler’s(2006)proposeanewinvestorsentimentindexandshowthathasmuchgreaterpredictivepowerthanexistingsentimentindiceshavebothinandoutofsample,andthepredictabilitybecomesbothstatisticallyandeconomicallysignificant. Inaddition, their findingssuggest that itoutperformsmacroeconomicvariablesand can also predict cross-sectional stock returns sorted by industry, size, value, andmomentum.Thedrivingforceofthepredictivepowerintheirmodelsappearstostemfrominvestors’biasedbeliefsaboutfuturecashflows. Nardoetal.(2015)inanextensivesurveyoftherelativeliteratureinvestigatetheinfluenceofonlinenewsonthefinancialmarketandthemagnitudeofthisimpact.InthiseffortNardoet al. (2015) claim that the sequence economic event, investment decisions, stock pricejumps,isfarfrombeingastylisedfact.Theypinpointfromtheliteraturethatmajorityofthelargest movements in the S&P500 and the CRSP Total Market Index cannot be tied tofundamental economic news sufficient to rationalize the size of the observed pricemovements.Fromtherevisionoftheliteraturetheyaretryingtouncoverthepotentiallinkofonline informationwith theaforementioned sequence.Nardoetal. (2015) also studiestheliteraturelinkingchangesinstockreturnsandtradevolumestomeasuresofwebbuzz.Theyobservethatalthoughthewebcantosomeextentanticipatefinancialmovements,thegain seldom exceeds 5%. In addition, they investigate machine learning and sentimentdetectiontoverifythatmoresophisticatedmodelsdonotnecessarilyproducebetterresults.Still,giventheamountofmoneymoveddailybythestockmarkets,evensmallgainscould



be worth exploiting, they also argue that the use of the Internet to predict financialmovementsisapermanentdevelopment. Furthermore the importance of sentiment in commodity markets is also highlighted byBorovkovaandMahakena(2015)whofindsignificantrelationshipsbetweennewssentimentand the dynamic characteristics of natural gas futures returns. Aslo, Feuerriegel andNeumann (2013)provideempiricalevidence thatabnormal returns in commoditymarketscanbeexplained,uptoalargeextent,bynewssentiment.

4.1.4 Concludingremarks

Asscholarsandpolicymakersaremovingtowardsacommonunderstandingoftherelationofthefinancialcycleandtheeconomicbusinesscycle,newevidence,boththeoreticalandempirical,supporttheimportanceofsentimentinbothrealeconomicactivityandfinancialactivity,mainlythroughtheconnectingpathofassetandcommoditypricemovements.Theimportance of sentiment places the behavioural approach of finance and economics in aprominent position in the future of economic research. On the other hand as argued inBaker and Wurgler (2007) "Investor sentiment is not straightforward to measure", so itnatural to observe a vast and on growing literature onmethods and proxies proposed tocapture sentiment. To this strand new information technology as machine learning, textmining, and neural networksmake a promising contribution toward new up to date andmore sophisticated proxies of sentiment that exhibit improved predicative power offorecasting and now casting of fundamental economic indicators, stock market prices,financialindicatorsandcommodities.

4.2 MarketSentimentAnalysis

Therecentyears,therehasbeenasignificantamountofworkonusingsourcesoftextdatafromtheWeb (suchasonlinenewspapers,blogsorTwitter feeds) topredict financialandeconomic variables of interest.Much of this work has relied on some form of sentimentanalysis to represent the text. Sentiment analysis targets open issues in various fields(includingpolitics,psychology,finance,andsociety),becausesentiments’understandingcanlargelyimpactinteractions,policies,anddecision-making.Duetoitsimportance,thisformoftextualinformationprocessinghasbecomeagrowingpartoftheempiricalfinanceresearch.Essentially,theoverarchinggoalinsentimentanalysisistoturntextintodataforanalysisviaapplication of natural language processing (NLP) and analytical methods. There is a widevarietyofvocabularyusedintheliteratureaboutsentimentanalysis.Thesametaskisoftenreferredtoassentimentextraction,sentimentmining,opinionmining,subjectivityanalysis,emotional polarity computation as well as other terms. The same applies for the task ofclassification,i.e.,theidentificationofthepolarityofanopinion.Commonlyusedtermsaresentimentorientation,semanticorientation,opinionpolarity,etc.WefollowTetlock(2007)withregardtotheseterms.Tetlock’s pioneering study (Tetlock, 2007) demonstrated that news stories containinformation relevant to predicting both earnings and stock returns. Within PROFIT, werestrictourfocusonhowtoapplysimilartechniques,butonawidervarietyofdictionaries,datasets,newssources,andmethodologies,inordertoextractinvestorsentimentindexes.



4.2.1 TheSentimentAnalysisWorkflow

An effective sentiment analysis consists of the following steps (see Figure 4-1: (1)informationretrievaloftextual-newsdatawithrespecttoaspecificfinancialinstrument,(2)natural languageprocessing, (3)extractionof themarketsentimentonthesentence level,(4)aggregationofthesentimenttothedocumentlevel.Informationretrievalistheactivityofobtaininginformationresourcesrelevanttoafinancialinstrumentfromacollectionofinformationsources.Searchescanbebasedonpartsoftheoriginaltextsrepresentedindatabasesoronfull-text(orothercontent-based)indexingandcan be restricted to specific websites, blogs or socialmedia resources. Each document isstrippedoff fromnon-textelements, links, tags,comments,andemoticons.Moreover, thetf–idfscheme(anumericalstatisticthatisintendedtoreflecthowimportantawordistoadocumentinacollectionorcorpus)maybeusedasaweightingfactor.In the next step, by using Natural language processing (NLP), we pre-process the text toidentify the various parts of speech, phrases and named entities. NLP of document textsincludes among others part-of-speech (POS) tagging, chunking and named entityrecognition.Wecanalsoapplytokenisation,sentencesplitting,andmorphologicalanalysisinordertofurthercleanthetext(Cunningham,Maynard,Bontcheva,andTablan,2002).After that, we take the pre-processed clean text and extract the market sentiment withrespecttoaspecificfinancialinstrumentorindicator.Theextractionofthemarketsentimenton the sentence level is a knowledge-based approach, which takes into account domainexpertise asmodelled, for example, in a lexiconor in anontology (see section4.2.2.2 fordetails).Afterclassifyingthe identifiedsentences intopositiveandnegativesentences, theaggregation of the sentiment to the document level is done by aggregating the overallsentimentoftheindividualsentencesthatweretaggedpositiveandnegativeinthepreviousstep.Apartfromtheseprocesses,thereisavarietyofothersub-problemsintheareaofsentimentanalysis, including subjectivity classification, opinion spam detection, topic relevance andtopicshift,whichwillbedetailedwithinthedeliverableD4.1(MarketSentimentModelsandIndicators) that comes as part of the Work Package 4 (WP4 - Market Sentiment-basedFinancialForecasting).

4.2.2 State-of-the-ArtinSentimentAnalysis

Inthissection,weperformaliteraturereviewofthestate-of-the-artwithregardtotheareaof sentiment analysis and classification. First, we review the three main approaches totextual content modelling for sentiment analysis, and then we elaborate on the mostpopular sentiment detection approaches that are employed to extract the concealedsentimentsofatextualresource.4.2.2.1 ThesentimentanalysismodelIngeneral,sentimentclassificationcanbedoneatseverallevels:words,phrases,sentences,paragraphs,documentsorevenmultipledocuments.Oneachlevel,theclassificationcanbedoneeitherdirectlyoritcanbebuiltontheresultsofpreviouslevelsbyaggregatingthem.



Figure4–1Thesentimentanalysisworkflow

Sentence-basedanalysis:Typically,atextualresourceismodelledusingthebaselinebag-of-words (BoW)model (Salton andMcGill, 1986). The BoWmodel can be easily applied onsentences,witheachsentencebeingconsideredtocontainasetofwords.Amajorfeatureofthismodelisthattheorderandco-occurrenceofwordsarecompletelyignored.Anotherpopularapproachisatechniquecalledn-grams,whereann-gramrepresentsasequenceofn words. Accordingly, each sentence is modelled with a set of (overlapping) n-grams (N-gramssplitting).Initssimplestform,theso-calledunigrams,then-grammodelisthesameastheBoWmodel. Ingeneral,n-grams(apartfromtheunigrams)considertheorderingofthewords in a sentence. By preserving theword ordering, thismodel enables the betterunderstandingoftheinter-wordsentimentalinfluencesandinteractions.Tofilteroutwordsthat probably do not express sentiment, one can use part-of-speech tagging (N-gramstagging)sothatonlywordswithspecifictags (forexample,adverbs,adjectives,noun,andverbs)areretainedforfurtherprocessing(N-gramsfiltering).FarahBenamaraetal. (2007)suggestthatadjectivesandadverbsaregoodindicatorsregardingthesemanticorientationofasentence.Various features can then be utilised to represent the filtered sentences (modelrepresentation), such as the set of observed n-grams weighted by their frequency ofappearance.Astextualresourcesarequiteoftencharacterizedbycontradictionsordifferentlevelsofsemanticorientation,wemayalsouseadditionalfeatures,suchasintensifiersandnegationwords.Theintensifiersaffectasentimenteitherbyincreasingitsintensity(liketheword“very”or“much”)orbydecreasingitsintensity(liketheword“hardly”or“scarcely”).Negationwordsalsoaffectthesemanticorientationofawordorphrase;forexample,inthesentence “this is good”, thepolarityof theexpressed sentiment is invertedbyadding theword “not”. Accordingly, the sentence “this is not good”, has a negative polarity. In thesentence-based analysis, the final issue to be addressed is how to extract the concealed



sentiments(sentimentdetection).Wediscussthemostpopularapproachesinsection4.2.2.2.Document-basedanalysis:Incasethetextualresourceweaimanalysingisadocument,wecandecompose it intoasetofsentencesorparagraphs(sentencesplitting).Then,wemayuseafilteringapproach(sentencefiltering)toretainonlythosesentencesthatmightbetterconveyinformationrelevanttothedocument’ssentiment.Suchfilteringcouldbebasedondifferent features, such as position, length, and subjectivity. In general, the first and lastsentencesofadocumentarequiteoftenindicativeaboutitssemanticorientation.Becauseadocument is a set of sentences, its representation relies on themodel used to representeachindividualsentence(sentence-basedanalysis).Inordertocapturethesentimentofthedocument as awhole, the sentiments expressed in each individual sentence are typicallyaggregated using appropriate measures, such as an averaging operator (Branavan et. al,2008).Topic-based analysis: Sentiment analysis is without doubt highly topic dependent. Thismeans that usually, we need to identify sentiments expressed by a textual resourcewithrespecttoaspecifictopicoraspecificfinancialinstrument,insteadoftheoverallsentimentexpressed inthetext.Thetopic isthesentence’sactualsubjectforwhichthesentiment isexpressed. In this case, a topic-based sentiment detection approach should be able toidentifythetopic,andthen,capturethesentimentsexpressedwithrespecttoit.Typically,atwo-steps approach is followed in order to proceed: first, the sentences relevant to theconsideredtopicareidentified;andthen,thesentence-filteringprocessisapplied.Finally,amodelrepresentationapproachisusedtocapturethesentimentoftheon-topicsentences(LinandHe,2009).4.2.2.2 SentimentdetectionapproachesTypically,wecapturethesentimentofatextualresourcebyusinglexicon-basedormachinelearningapproaches.Limitedworkhasalsobeenconductedonhybridmethodologiesaswellasonontology-supportedapproaches.Lexicon-BasedApproaches:The lexicon-basedapproach isanunsupervisedtechniquethatextracts the sentiment of a text using lexicons (dictionaries) that may be either domain-specific (such as finance, politics, psychology, etc.) or domain-independent (Turney 2002).Suchdictionariesconsistofwordsthatareannotatedwiththeirsemanticorientationvalue(polarity and strength); usually a positive or negative value is assigned to eachword (forexample,thescorefortheword“good”is0.65andthescorefortheword“disgust”is−0.85).Dictionaries for lexicon-based approaches can be created either manually (Stone4, 1966;Tong,2001;LoughranandMcDonald5,2010),orautomatically,usingseedwordstoexpandthe list of words (Hatzivassiloglou andMcKeown 1997; Turney 2002; Turney and Littman2003). Much of the lexicon-based research has focused on using adjectives as indicatorsregardingthesemanticorientationofthetext(HatzivassiloglouandMcKeown,1997;Wilson,WiebeandHoffmann,2005;HuandLiu,2004;Taboada,Anthony,andVoll2006).Accordingtothis,alistofadjectivesandcorrespondingsemanticorientationvaluesiscompiledintoa

4Availablefromhttp://www.wjh.harvard.edu/~inquirer/5Availableathttp://www.nd.edu/~mcdonald/Word_Lists.html



dictionary.Then,foranygiventext,alladjectivesareextractedandannotatedaccordingtotheirsemanticorientationvalueusingthedictionaryscores.Thesemanticorientationscoresareinturnaggregatedintoasinglescoreforthetext.Machine Learning Approaches: The text classificationapproach is essentially a supervisedclassification task, which involves building classifiers from labelled instances of texts orsentences. The introduction of the machine learning approach in sentiment analysisoriginates fromBoPanget al. (Pang, Lee andVaithyanathan, 2002). This approach is alsoreferredtoasastatisticalormachinelearningapproach.ThemajorityofthestatisticaltextclassificationresearchreliesonSupportVectorMachine(SVM)classifiers,NaïveBayesandMaxentropyclassifiers trainedonaparticulardatasetusing featuressuchasunigramsorbigrams, and with or without part-of-speech (POS) labels (Pang, Lee, and Vaithyanathan,2002;Salvetti,Reichenbach,andLewis,2006).Classifiersbuiltusingthesupervisedmethodsmayachievehighaccuracywhendetectingthepolarityofatext(KennedyandInkpen,2006;Boiyetal.,2007;BartlettandAlbright,2008).However,althoughsuchclassifiersperformverywellinthedomainthattheyaretrainedon,theirperformancedegradeswhenthesameclassifierisusedinadifferentdomain(AueandGamon, 2005). Moreover, utilizing an individual classifier might result in poor sentimentdetection, since the performance of each classifier varies significantly, when for instancesomeone is using different features or weight measures. Therefore, for overcoming thedeficienciesofeachclassifierandtoproceedwithamorerobustandsuccessfulsentimentdetectionprocess,withinPROFITandsimilartoRuiXiaetal.(Xia,Zong,andLi,2011),wewillcreateensembleclassifiers;thatis,acombinationofmultipleclassifiers.HybridApproaches:Itbecameapparentthatthelexicon-basedapproachessufferfromtheirabsolute dependence on lexicons, which are often characterized by word shortage orinappropriate assignment of semantic orientation values. While, machine-learningapproachesovercomethese limitations,theirneedfora largevolumeoftrainingdataalsolessenstheiradvantage(Goncalves,Araújo,Benevenuto,andCha,2013).Accordingly, the hybrid approach targets at solving these limitations by combining theaforementioned approaches. An exemplar use case scenario for a hybrid approach is tofollowa two-stepprocess; i.e., to initially generatea setof trainingdatabyautomaticallyidentifyingthetexts’semanticorientationscore(usingalexicon-basedapproach),andthentoproceedwiththeclassification(usingamachine learningapproach) that is independentfromthelexicons’limitations(Ghiassi,Skinner,andZimbra,2013).AsPrabowoandThelwall(2009) point out, hybrid approaches enhance the stability and accuracy of existingmethodologieswhileexploringthestrongcharacteristicsofeachofthem.Ontology-based approaches: Recently, semanticweb-based sentiment analysis that reliesontheusageofontologieshasstartedtoattractattention.Ontologiesarenotaclassificationmethodinitself,buttheycanbeusedtoenhanceexistingclassificationmethods.ZhouandChaovalit(2008)wereamongthefirsttouseontologiesforthispurpose.Theyincorporatedanontologytoasupervisedmachinelearningtechniqueandabasictermcountingmethod.In theirwork, as inmost otherworks, the ontology contained semantic features andwasusedonlytoidentifythesefeaturesinsidethetext.Cadilhacetal.(Cadilhac,Benamaraand



Aussenac-Gilles,2010)ontheotherhand,theydidnotusetheontologyonlyasataxonomyby taking into account the “is a relation” between concepts. Instead, they presented anapproach that uses other relations between concepts for creating summaries of opinionsandcomputingtheSOscores.Thoughan infancystage,withinPROFIT,wewillexplorethepossibilitiesprovidedbythisclassofsentimentanalysistechniques.Table4–1summarisesthesentimentanalysisapproachesusedintherelevantliterature.Itisworthnotingthattheextensivereviewinsentimentanalysisisnotthepenultimatepurposeofthisworkpackage,asathoroughreviewanduniqueanalysiswillbeconductedalongthedeliveryofthedeliverableD4.1(MarketSentimentModelsandIndicators).

Table4–1Anoverviewofthesentimentanalysisapproaches

Approach Sentimentdictionariesandclassifiers SentimentTypesLexicon-based

GeneralInquirer’sHarvardIV-4psychosocialdictionaryLoughranandMcDonaldfinancialdictionaryLinguisticInquiryandWordCount(LIWC)AFINNdictionaryLexicodersentimentdictionaryWordStatsentimentdictionaryWordNetandWordNet-AffectSentiWordNetSentiStrength

Positive,negative,neutralPositiveandnegativetogetherwithstrengthAffect-relatedsentiments(suchasangry,sad,joyful,fearful,proud,elated,cheerful,gloomy,irritable,listless,depressed,friendly,distant,liking,loving,hating,nervous,anxious,reckless,hostile,jealous,etc.)6

Machinelearning

SupportVectorMachines(SVM)DecisionTreesMultinomialNaïveBayesMaximumentropyclassifierRule-basedclassifier(RBC)GeneralInquirerBasedClassifier(GIBC)Ensembleclassifiers

Hybrid Hybrid classification means applying differentapproachesandclassifiersonasequence(i.e.,GIBCàSVM,RBCàGIBC,RBCàGIBCàSVM)

Ontology-based

An ontology-based classification uses the ontology aspart of the feature extraction process and takes therelationsbetweenconceptsintoaccount

4.2.3 SentimentanalysisinPROFIT

TheapproachofresearchthatweaimaddressingatPROFITcanbespecifiedasfollows(seeFigure4–2).Aleadingcharacteristicoftheobjectofresearchisthefocusontext,soweareonlyinterestedintextclassification.ThesourceofthesetextsislimitedtotheWorldWideWeb (including among others financial newspapers, online reviews, blogs, social media,etc.). Throughout a text there can be subjective or objective statements. We are onlyconcernedwiththeformer.Theapplicationdomainofsentimentanalysisisfinancemeaning

6Scherer,KlausR.EmotionasaMulticomponentProcess:Amodelandsomecross-culturaldata.InP.Shaver,ed.,ReviewofPersonalityandSocialPsychology,vol.5,pp.37-63,1984.



thatsentimentsaboutpersonsorproducts,forexample,areexcluded.Theproblemthatistackledistheextraction,classificationandaggregationofthesesentimentsforthecreationofasentimentindextimeseriesthatreflectsthesentimentofthemarketonagivenday.Theapproach to be followed is automated information extraction. As identified in Section4.2.2.2, various approaches can be used to capture the sentiment of a textual resource.WithinPROFIT,wewillconsideralltheavailableoptionsbeforemakinganyfinaldecision.

Figure4–2SentimentanalysisinPROFIT

4.3 ForecastingModels

Accurate forecasting exercises regarding real economic activity, financial markets andcommodities are the Holy Grail of the theoretical and empirical economic analysis. Bothpublicandprivatesectorrelyonshortor longerhorizon forecasts inorder totakevariouseconomic decisions spanning a wide range of everyday aspects. From the Central Bank’sdecisionofsettingtheeffectiveinterestrates,tothedecisionofaprivateinvestortoholdornot stocks and even to a bet against the realization of a football game. In the US, forexample,priortoeachFederalOpenMarketCommitteemeeting,theFederalReservestaffproduces forecastsof severalmacroeconomicaggregatesathorizonsup to twoyearsasabasis for their discussions. The variables forecasted by the staff include exchange rates,whichinfluencecurrentaccountprojectionsaswellasUSrealGDPgrowth,eventually.The importance of forecasting could be further assessed using some more indicativeexamples.Inpursuingfiscalsustainability,fiscalauthoritiesandcentralbanksitiscrucialtohave good forecasts of the future path of national output. The same holds true for the



effectiveconductofmonetarypolicyinpromotingfinancialstability.Furthermore,long-termlabor contracts ormortgages or fixed deposit accounts are a common feature inmoderneconomies. In such framework it is very important good forecasts of a country’s generalprice index (inflation rate) evolvement for efficient decision-making. Another example iswhether a financial market is in a bear or in a bull sate, which implies very differentinvestment opportunities. For example, during a bear market, stock prices are generallyfallingmakingmoreattractiveriskfreeassets(i.e.,sovereignbonds).Thus,thepredictionofthefuturestateofthemarketcanhelpaninvestortoperformbetterbyshiftingtothemostappropriate state market based investment portfolio. When it comes to the commodityprices,theimportanceofforecastingtoolsareevenmorecrystalclear.Itiscommonnotionthatchangesinthecostofcrudeoilaffecteconomicactivityexertingacountercyclicaleffecton the growth rate of a national economy. This can take place for example through theinvestmentchannel, iftheprivatesectorisreluctanttoproceedtoinvestmentsbecauseoftheir irreversiblenatureandofhighoilpricesorevenbecauseofexpectedhighoilprices.Forthatreason,Centralbanksandinternationalorganizations(InternationalMonetaryFund,WorldBankinteralia)worldwideroutinelyuseshort-horizonforecastsofthequarterlypriceofoil.Inthesamevein,exchangeratesareofgreatimportanceregardingprivateinvestmentdecisions,economicpolicyatnationallevel(especiallyforcountriesthatareheavyimportersand exporters of commodities), as well as risk management. Taking into account all theabove arguments, forecasting remains a fundamental concept of our understanding ofeconomic decisions made by policymakers, firms, investors, and households spanning avoluminousliterature.It is important to note that empirical evidence point to the interconnection betweenfinancial markets, commodities and macroeconomic indicators (Gross Domestic Product,GDP,employment, currentaccount, inflation,etc.). Thismeans forexample thatoilpricesmayconveysomeinformationaspredictivevariablesfortheforecastingofexchangerates(thisofcoursemayholdunderveryspecificassumptions).Macroeconomicvariablescouldalso predict future movements in stock returns or stock market indexes (Dow Jones,Eurostoxx 50 etc.). Finally, many papers find that some financial series have significantpredictivepowerforfutureeconomicactivity(GDP).Someoftheseseriesareinterestratesandspreads,stockprices,monetaryaggregates,surveyforecastsleadingfinancialindicatorsamongothers.Atthecentreofforecastingliesthepredictionforthephaseofthebusinesscycle.AsSiegel(2014)notes,“Ifonecouldpredictinadvancewhenrecessionswilloccur,thegainswouldbesubstantial.Thatisperhapswhybillionsofdollarsarespenttryingtoforecastthebusiness cycle. But the recordof predictingbusiness cycle turningpoints is extremelypoor”. The literature has considered a wide variety of econometric estimation methods,variableselectionandmodelspecification.Despitethischaoticeffort, forecasting isstillanongoingeffort.

4.3.1 Theinterlinkbetweenfinancialsectorandtherealeconomy

ThesevereUSsubprimecrisisof2007thatoriginatedfromthefinancialsectorhasbroughtunder scrutiny the interplay of financial conditions and macroeconomic environment.Financialconditionscanbethoughtofthecurrentstateoffinancialvariablesthatinfluenceeconomicbehaviourand(thereby)thefuturestateoftheeconomy.Generallyspeaking,theuniverseoffinancialvariablesspansthesupplyordemandoffinancialinstrumentsthatarerelated tobroadereconomicactivity. This listmight compriseawidearrayof assetprices



andquantities (both stocks and flows), aswell as indicators of potential asset supply anddemand.Thelattermayrangefromsurveysofcreditavailabilitytothecapitaladequacyoffinancialintermediaries.AFinancialConditionsIndex(FCI)attemptstoextractinformationaboutthefuturestateoftheeconomyconveyedbythesecurrentfinancialvariables. Ideally,anFCIshouldmeasurefinancialshocks–exogenousshiftsinfinancialconditionsthatinfluenceorotherwisepredictfutureeconomicactivity.Truefinancialshocksshouldbedistinguishedfromtheendogenousreflectionorembodimentinfinancialvariablesofpasteconomicactivitythatitselfpredictsfutureactivity.Iftheonlyinformationcontainedinfinancialvariablesaboutfutureeconomicactivitywereofthisendogenousvariety,therewouldbenoreasontoconstructanFCI:Pasteconomicactivityitselfwouldcontainalltherelevantpredictiveinformation.ThegrowingliteratureonthemonetarytransmissionmechanismcouldhostouranalysisforunderstandingFCIs. Searching in these studieswecomeupwithavery influential finding:monetary policy affects the economy by shifting the financial conditions that changeeconomicbehaviour.Thestructureofthefinancialsystemplaysapivotalroleinthevariouschannelsoftransmission.Forexample,thelargecorporatebondmarketintheUnitedStatesand its broadening over time suggest that market prices for credit are more powerfulinfluencesonU.S.economicactivitythanwouldbethecaseinJapanorGermanytoday,orinthe United States decades ago. The state of the economy also matters: For example,financial conditions that influence investment may be less important in periods of largeexcesscapacity.Therecentanalysisofthemonetarytransmissionmechanismbyclassifiesthesechannelsasneoclassicalandnon-neoclassical.Thefirstcategoryiscomprisedoftraditionalinvestment-,consumption- and trade-based channelsof transmission. The investment channel containsboth the impact of long-term interest rates on the user cost of capital and the impact ofassetpriceson thedemand fornewphysicalcapital (Tobin’sq).Theconsumptionchannelcontains both wealth and intertemporal substitution effects. Both the investment- andconsumption-based channels may be influenced by shifts in risk perceptions and risktolerance that in turn affect market risk premia. Finally, the trade channel captures theimpactoftherealexchangerateonnetexports.According to the ICAPM if investment opportunities change over time, then assets’exposures to these changes are important determinants of average returns in addition tothe market beta. The ICAPM dictates that the yield curve is an important part of theinvestmentopportunityset.Twoothervariablesproposedascandidatesforstatevariableswithin the ICAPM are the returns on the HML and SMB portfolios. These factors capturecommonvariationinportfolioreturnsthat is independentofthemarketandthatcarriesadifferent risk premium.Petkova (2006) in a seminal studyprovides a rationale trough theIntertemporalCapitalAssetPricingModelofMerton(1973)forunderstandingthepredictiveability of FF factors for stock returns. She concluded that Fama and French factors areassociatedwithshocksinvariablesthatpredictinvestmentopportunityset.Underanassetpricingmodelcontexttheexpectedreturnsofsecuritiesarerelatedtotheirsensitivitytochangesinthestateofeconomy.Studiesonthisfieldhaverevealedthatsome



statevariablescarryasignificantriskpremiumthatmightchangethroughtime.Inthisvein,Ferson and Harvey (1991) found that the time varying nature of expected risk premiumsaccountsforalargepartofstockreturnspredictability.Thesecondcategory–ornon-neoclassicalset–oftransmissionchannels includesvirtuallyeverything else. Prominent among this category are imperfections in credit supply arisingfrom government intervention, from institutional constraints on intermediaries and frombalancesheetconstraintsofborrowers.Naturally,theimportanceofthesedifferenttransmissioncategoriesmaychangeovertime.For example, a “credit view” – which emphasizes some of the non-neoclassical factors –mighthighlighttheimpactofthedepletionofbankcapitalandthedeclineinborrowernetworthinexplainingtheweakresponseoftheU.S.economytolowpolicyratesintheearly1990s.Aneoclassicalassessmentofthe1998-2002periodmighthighlighttheroleofstockpricesindrivinginvestmentand,toalesserextent,consumption.Notethatbothcategoriesof transmissionchannelsallowfora loose link (orevenfor the lossofa link)betweenthesettingofthepolicytool–typically,therateoninterbanklending–andthebehaviouroftheeconomy. The financial conditions thatmatter for future economic activity are subject toshocksfromsourcesotherthanpolicy,inadditiontopolicyinfluences.Inthetwoexamplesinthepreviousparagraph,theseshockswouldincludechangesinthenetworthoflendersandborrowers,orintherelationshipbetweenassetpricesandeconomicfundamentals.The slope of the yield curve dominated early research on financial conditions indexes.Studiesthatappearedinthelate1980sandearly1990sprovedthattheyieldcurvecarriespredictive power over future economic activity (Estrella and Hardouvelis, 1991). Anotherimportantfinancialvariableisthespreadbetweenthefedfundsrateand10-yearTreasuryyieldthatparticipatesinConferenceBoard’sindexofleadingindicatorssince1996.Amongtheleadingindicatorsofoutputwecanfindcreditrisk,whichisproxiedbythecommercialpaper-Treasury bill spread. In a related study Gilchrist, Yankov, and Zakrajšek (2009)proposed improved credit risk spreads with good forecasting performance over the pastdecade.Thesuperiorityofyieldcurveoverotherfinancialvariablesasrecessionpredictorisbeyondanydoubt,thoughstockmarketperformancehasbeenfoundbysometobeausefulrecessionpredictoraswell (EstrellaandMishkin,1996). Stockmarketvariableshavebeenincludedinindexesofleadingindicatorssincethe1950s.TheBankofCanada (BOC)pioneeredworkonbroader financialconditionmeasures in themid- 1990s, when it introduced its monetary conditions index (MCI). For the BOC, theexchangeratewasthemostimportantadditionalvariable.ItsMCI,therefore,consistedofaweightedaverageofitsrefinancingrateandtheexchangerate.Overthecourseofthelate1990s,MCIsalongthelinesconstructedbyBOCbecameawidelyusedtooltoassessthestanceofmonetarypolicyinmanycountries.Moreover,thescopeofvariablesaugmentingtheeffectsofpolicyrateswasbroadenedtoincludelong-terminterestrates,equityprices,andevenhouseprices(onthegroundsthatrisinghousepricesincreasedtheborrowingcapacityofhouseholds).Thesebroadermeasuresbecameknownasfinancialconditionindexes(FCIs)inordertodistinguishthemfromMCIs.



AvarietyofmethodologiesforconstructingFCIshavebeendevelopedovertime,andtendto fall into two broad categories: a weighted-sum approach and a principal-componentsapproach. In the weighted-sum approach, the weights on each financial variable aregenerallyassignedbasedonestimatesoftherelativeimpactsofchangesinthevariablesonrealGDP.Theseestimatesorweightshavebeengenerated ina varietyofways, includingsimulationswithlarge-scalemacroeconomicmodels,vectorautoregression(VAR)models,orreduced-form demand equations. The second broad approach is a principal componentsmethodology,whichextractsa common factor fromagroupof several financial variables.This common factorcaptures thegreatest commonvariation in thevariablesand iseitherusedas theFCIor is added to thecentralbankpolicy rate tomakeup theFCI (this lattermethod is a combination of the weighted-sum approach and the principal-componentsapproach).

4.3.2 Forecastingmodels.Whathavewelearned?

AsstatedabovethecoreobjectiveofCentralBanks,internationalorganizationsandprivateinvestors is the forecastofGrossDomesticProduct.Although,GDP (or theBusinessCycleanalysis) attracts the most of the interest in the forecasting exercises stock prices andindexes,interestandexchangeratesandcommoditiesarealsointhecentre.Theproposedestimationmethodsforforecastingareinmanycasesthesameregardlessthesubjecttobeforecasted,withthemostofthedifferencestoconcentrateintheselectionofthepredictivevariablesandinthemodelspecification.However,sincethepredictiveperformanceofthemodelsvariesaccordingtotheseriestobeforecastedwewillpresentthemostprominentmodels for each case along with their predictive performance. Our main scope in thissubsection is to expose the evolution of forecasting methods, abstracting from anyunnecessarytechnicalities.Ofcoursethelistisnotexhaustive;howeverwecangraspsomeideasofwhereempiricalresearchstandsandhowwecancopewithit.GrossDomesticProduct(GDP):Manydifferenttypesofmultivariatemodelshavebeenusedin the literature, typically generalizations of the single-equation models. Large scalemacroeconometricmodels that was introduced in ‘70’s performed rather poorly and notbetter than single equation autoregressive models. In the ‘80’s Vector Autoregressivemodels (VAR)were introducedthatrequiredonlyweak identifyingassumptionsrelativetothe largescalemodelspreviouslyused.At thesametime, increasedpopularitygained theDynamicStochasticGeneralEquilibrium(DSGE)models,whicharebasedontheinteractionofmicrofoundeddecisions.ThisleadtotheuseofnewhybridmodelsthatincorporatebothDSGEandVARmodels.Thegeneralfindingfromtheliterature,asitisdescribedinChauvetandPotter(2012)isthatthe DSGE forecasts are slightly superior especially in the medium term (three and fourquarters ahead) to the ones obtained from VARs and BVARs, but still not significantlydifferentfromsimplebenchmarkssuchasunivariateautoregressiveprocesses.Another strandofmodels includes dynamic factormodels,mixed frequencymodels, non-linearmodelsandforecastcombinations.Dynamicfactormodelsprovidetheadvantageofextracting large scale information in a parsimonious framework.Many studies reveal thatdynamic factor models beat univariate and multivariate autoregressive models. Mixedfrequency models explore information in more timely indicators that are available at a



higher frequency. Combined with dynamic factormodels outperform relative to previousmodels. Forecast combinations (Timmerman, 2006) infer about forecasting predictions byestimatingthecorrelationbetweenforecasterrorsandtherelativesizeoftheforecasterrorvariancesofsinglemodels.Resultsshowthatbestresultscanbeobtainedfromcombiningasimpleequal-weightedaverageofsurveyandmodelbasedforecast.On the other hand, empirical research has shown than possible non-linearities in thedynamicsoftheeconomycouldbequite importantfor forecasting.Theclassofnon-linearmodels include an endless variety of models such that dynamic Probit models, Markovswitching models, and threshold multivariate models. Non-linear models may revealadditional information and improve forecasts compared to frameworks that take intoaccountonlytheaveragelineareffectofoneseriesonanother.Combinationoftheabovestated non-linear models show that the recession probabilities perform quite well inestimationwithreal-timedatabasesandaresuccessfultoolinmonitoringthebusinesscycle.Finally, as it was analysed in the previous subsection the forecasting performance of themodels can be further enhancedwith the use of financial series. This result is confirmedacrossanumberofcountries(StockandWatson,2003).Exchangerates:Exchangeratedynamicsareverydifficulttopredictusingeconomicmodelsandarandomwalkwithadriftmodelcanprovideinmanycasesthebestavailableforecast.Reviewingthe literature itseemsthatthere isalmostaconsensusamongresearchersthatlinearspecificationsbasedonTaylorruleandnetforeignassetsfundamentalsmayprovidemoreexplanatorypowerthanmodelsthatrelyontraditionalfundamentalssuchasinterestrate,inflation,output,etc.Rossi(2013),findsthat,althoughsomepredictors(Taylor-rulefundamentalsandnetforeignassets) do exhibit some predictive power at short horizons, and others (monetaryfundamentals, especially in panelmodels) reveal some predictive ability at long horizons,noneof thepredictors,models, or tests systematically find empirical support for superiorexchange rate forecastingabilityofapredictor forallmodels, countriesand timeperiods:typically, when predictability appears, it does so occasionally for some countries and forshortperiodsoftime.Regarding the model specification, in contrast to the Gross Domestic Product case, non-linearmodelsseemtohavetheleastsuccessfulpredictivepower.Amongthelinearmodels,thesingle-equationErrorCorrectionModels(ECM)andthepanelECMmodelsarethemostsuccessful at long horizons, although there is disagreement among researchers about thedegree of robustness of the results. Typically, but not always, for single-equation linearmodels the predictor choice matters more than whether the researcher usescontemporaneous, realized or lagged fundamentals; the linearmonetarymodel does notperformwellatanyhorizons,whereasTaylor-rulefundamentalsandnetforeignassetsaresuccessful predictors at short-horizons.Models with time-varying parameters show somedegreeofsuccess,althoughmostofthefavourableempiricalevidenceisbasedonstudiesinthelate1980s-early1990s,andthemostrecentanalysesreportlesssuccess(althoughwithslightly different specifications). Among the multivariate models, the most successfulspecificationisthepanelECM,althoughthereissomeevidenceinfavourofBVARs,BMAandfactorECMmodels.



In later years it has been proposed an estimation method that is a combination of allavailable macroeconomic variables named “Kitchen sink”. This statistical analysis uses anumber of suggested predictors using penalized least squares. Under this framework theestimatedvarianceislowerthanunderOLSestimationmethod,deliveringahigherdegreeofpredictiveaccuracy(Lietal.,2015).Finally,anewclassofpredictivevariablesiscombinedbyunstructured information from mined news articles. This new dataset increases thepredictivepowerofexistingmodels,providingcorrectinsightsaboutupcomingreversalsofthecurrencymovements.Stock Market Indexes: Literature on stock market predictability is voluminous but stillinconclusive.OnceagainitishighlightedthatsincethecoreobjectiveoftheProfitprojectisto strengthen financial awareness of the users and financial stability, we approach theforecastingexerciseof thestockmarket index fromthescopeofeconomicanalysis ratherthanfromtheviewpointofatrader.Underthisregard,theobjectiveoftheforecastmodelwillbethepredictionofthedirectionofexcessstockreturns.Studies for thepredictabilityof stockmarket returns couch their analysison the so-calledpredictive linear regressions. In the context of predictive linear regressions stock marketreturnsareusuallyregressedagainstthelaggedvaluesofastochasticexplanatoryvariable,forexample, the lagged logdividendyield, thedefault spread, the termspread, the shortterminterestrateassumingconstantcoefficients.However, predictive linear regressions have been in the epicentre of criticism that restsmainly on their poor performance in out-of-sample exercises. The comparison betweenpredictive regressions with historical average returns and shows that historical averagereturnsalmostalwaysoutperformpredictiveregressions intermsofpredictivepower. Inarelatedstudy,Campbelletal.(2008)concludedthattheoutofsampleforecastingabilityofpredictiveregressionscanbesubstantiallyimprovedonceweposerestrictionssuchassignor quantitative on the coefficients and return forecasts of any theoretically motivatedforecastingregression.Modelsof stockmarketpredictabilityhavealso received intensecriticismthat stems fromthequestionofwhetherstockmarket investorscouldhaveexploitedpredictabilitytoearnabnormal returns in real time. Studies attempted to provide an answer to this with themajority reaching evidence against profitability of trading rules built on stock marketpredictability. Generally speaking, studies (see inter alia Marquering and Verbeek, 2004)detectedstrongin-samplestockreturnpredictabilitythatissubstantiallyweakerwhenout-of-sample predictability is under scrutiny. A significant number of studies point tomodelparameter instability or structural breaks which in turn might affect the validity of theforecasting(seeinteraliaPayeandTimmermann,2006).Otherstudies(Henkeletal.,2011)reach the conclusion that predictability is a phenomenon whose strength is distinctivelytime-dependent.A rather prolific strand of literature proposes alternative econometric methods foralleviatingthebiasandperformingreliable inference.Parallel tothestandardstockreturnpredictabilityliteratureanditsrelatedeconometricissuesvariousstudieshaveattemptedto



offeradifferentavenueofresearch.Tothisend,analternativemethodofforecastingstockmarket returns – the sum-of-the-partsmethodwas proposed by Fereirra and Santa-Clara(2011). They decomposed the stockmarket return into three components – the dividendyield, the earnings growth rate, and the growth rate in the price-earnings ratio – andforecast each of these components separately. Their results pointed to the existence ofsignificant predictability in equity returns. Forecasting the future direction of the marketwhetherisabullorbearhasattractedtheinterestofacademicresearch.Forexample,Chen(2009) foundthat theyieldcurvespreadsand inflationwerethemostusefulpredictorsofrecessionsintheUSstockmarketintermsofbothin-sampleandout-of-sampleforecastingperformance.Inarelatedstudy,Nyberg(2013)byemployingaU.S.dataset,theyconcludedthatbearandbullmarketsarepredictableinandoutofsample.Morerecently,Chavaetal.,(2015)analysedstockmarketpredictabilityinUSemployinganoveleconomically-motivatedmeasure of credit standards. Their results revealed that their survey-basedmeasure wasabletopredictaggregatestockmarketreturnsespeciallyinthepost-1990dataperiodandata business cycle frequency. Çakmakli, and van Dijk (2016) employing Factor-augmentedpredictiveregressionmodelsdetectedsignificantstockmarketpredictabilityforUS.Anothernoteworthyfindingofthisstudyisthecollapseofbenchmarkforecastingmodelsduringthe1990s.AninterestingapproachwasproposedbyHongetal.(2007)whoexaminedthepredictabilityofUSstockmarketreturnsemployingpredictiveregressionsandvarious industryportfolioreturns and standard predictive variables as regressors. Their analysis rests on twoassumptions:informationtravelacrossassetmarketsslowlyandmanyinvestorsarenotabletogatherthe informationfromtheassetpricesofmarketsthattheyarenotfamiliarwith.These twocombined togethermight lead to cross-asset returnpredictability. Their resultsrevealed significant forecasting ability of several industry portfolios by up to twomonths.Theiranalysiswasextendedtoothereightinternationalmarkets(UnitedKingdom,Australia,Canada, France, Germany, Japan, Netherlands and Switzerland) and their results werequalitativelythesameasintheUS.

4.3.3 ABriefDescriptionoftheForecastingEstimationMethodsinPROFIT

Inthissubsectionwebrieflypresentthemethodologythatwearegoingtouseinordertoforecast, the phase of the business cycle for the GDP in the Eurozone, the Eurostoxx 50which is Europe’s leading Blue Chip Index for the Eurozone, the Eurodollar exchange rateand the oil price. This selection of predictions corresponds to the most widely used byCentralBanks,Internationalorganizationsandprivatefunds.Ofcoursethepalletofpossibleusefulseriestobeforecastedconvergestoinfinity,butconditionalontheextremeamountofeffort required foreach forecastexercisewe confineourselves to theabovementionedseries.RegardingtheproposedmethodologyfortheforecastingofGDPandtheEurostoxx50,wewillemployasetofdynamicmodelsfollowing.Morespecifically,respondingtothecriticismandaccountingforthenon-linearnatureofstockmarketreturnsmostlikelyinducedbytheinterplay amongstmarket participants, managers, and central banks, we opt for a novel,semi-parametricestimationmethodwhichhelpsdealmoreeffectivelywithnon-linearitiesorstructuralbreaks in thedata (Florackisetal.,2015). In thecaseofoilpricesandexchangeratewewillfollowrecentempiricalliteratureusinglinearmodels.



As shown before existing studies in the literature differ considerably depending on thecharacteristicsof theirdata.Theselectionofpredictorscancruciallyaffect the forecastingpowerofthemodelinuse.Thus,inordertoavoidanyinconsistenciesandbiasesstemmingfromtheselectionofthepredictorserieswewillemploynewmethodsinfeatureselectionand component analysis. Doing so, we will be able to use the most of the informationavailablebythedataavoidinganyproblemsfromoverestimatingthemodel.Furthermore,following recent literaturewewilluseadditional information fromunstructured textdata,enhancing the predictive power of our forecasting models. More technical analysis andintuition for the data and methodology selection process will be provided within thedeliverables D4.1 (Market Sentiment Models and Indicators) and D4.2 (AdvancedEconometric and Machine Learning Forecasting) of Work Package 4 (WP4 - MarketSentiment-basedFinancialForecasting).

4.3.4 ForecastingEstimates

In this subsectionwepresent sources of economic forecasts for themember countries ofEuropeanUnionprovidedby theEuropeanCommission, the InternationalMonetaryFund,the European Central Bank and the Organization of Economic Cooperation andDevelopment.AMECO (European Commission - Economic and Financial Affairs: DG ECFIN's economicforecasts concentrateon theEU, its individualmember states,and theeuroareabutalsoinclude outlooks for some of the world's other major economies, and countries that arecandidates for EUmembership. The forecasts extend over a time horizon of at least twoyears and cover about 180 variables. The forecasts are not based on a centralisedeconometricmodel,but areanalysesmadebyDGECFIN countrydesks,usingmodels andexpertknowledge.Consistency isensuredbyanumberofcross-countryandcross-variablechecks.Forecasts for theEUandtheeuroareaarecreatedbyaggregatingthedataof theindividualmemberstates.DGECFIN'sforecastsarepublishedthreetimesayearinsyncwiththeEU'sannualcycleofeconomicsurveillanceprocedures,knownastheEuropeanSemester.Priortotheadoptionof these procedures in the autumn of 2012, two fully-fledged forecasts and two interimforecastswerepublishedeachyear.IMF (World Economic Outlook Database): The World Economic Outlook (WEO) databasecontains selected macroeconomic data series from the statistical appendix of the WorldEconomic Outlook report, which presents the IMF staff's analysis and projections ofeconomicdevelopmentsatthegloballevel,inmajorcountrygroupsandinmanyindividualcountries.TheWEOisreleasedinAprilandSeptember/Octobereachyear.OECD (Economic outlook, Analysis and Forecasts): The OECD’s forecasts combine expertjudgement with a variety of existing and new information relevant to current andprospectivedevelopments.Theseincluderevisedpolicysettings,recentstatisticaloutturnsand conjunctural indicators, combined with analyses based on specific economic andstatisticalmodelsandanalyticaltechniques,asoutlinedbelow.



OECD (Composite Leading Indicator): Leading indicators comprise the composite leadingindicator(CLI)andstandardisedbusinessandconsumerconfidenceindicators.Theyprovidequalitative information useful formonitoring the current economic situation and advancewarningofturningpointsineconomicactivity.EuropeanCentralBank(MacroeconomicProjections):Themacroeconomicprojectionscovertheoutlookfortheeuroarea,inparticularwithregardtoGDPandinflation.Theprojectionsarepublishedfourtimesayear:inMarchandSeptember,whentheyareproducedbyECBstaffmembers;inJuneandDecember,whentheyareproducedjointlybyeuroareanationalcentralbanksandECBstaffmembers.

4.4 FinancialandeconometricforecastinginPROFIT

Predictionofeconomicandfinancialvariablesisaverydifficultcourseasitincludesvariouscomplications.Theresearcherneedstoanswertomanyquestionsregardingthepredictorsthatshoulduse,theforecasthorizon,themodeltobeemployed,thedatafrequencyandsoforth.Manyauthorsconjecturedthatsamplingerror,modelmisspecificationandinstabilitiescouldpotentiallyexplainthepoorforecastingperformanceoftheeconomicmodels.Thus, the core objective of this deliverable andmostly ofWork Package 4 (WP4 -WP4 -Market Sentiment-based Financial Forecasting) is to provide a reliable overview ofestablishedfindingsthatcanbehelpfultobuildmodelswiththebestavailableforecastingability.Itseemsfromthediscussionintheprevioussectionsthatrecentdevelopmentinthewaythatweanalysebiddatasetscanprovideanadvantageoftheforecastingperformanceofnewlyconstructedmodelsrelativetotherecentpast.Tothisend,withinPROFITweaimatextractinginformationfirstbynewunstructureddatasetsandsecondlybyemployingalargersetofpredictivevariables.Bythesemeans,wewillbeabletosolveproblemsofselectionbias.Ontopofthat,wewillmakeuseofnewmethodsinnon-parametric analysis in order to avoid to theextent that it is feasible anyproblemsarisingfromspecificationerrors.Thenewcapturedinformationwillrefertotheprevailingsentimentonthefinancialmarket.It is well understood by now that the predictive power of sentiment is pronounced inforecastingeconomic and financial variables.However as sentiment is still regardedas anamorphousconceptwithouta reliablemeasureof it there isplentyof room indevelopingnew proxies. In this vein, during the projectwewill develop newmeasures of sentimentanalysisbothintermsofconceptandalso intermsofmethodology.Thefirstreferstothedevelopmentofanewcomposite indexof sentimentanalysiscombining information fromvariousfinancialmarkets.All this information isvaluabletothe investors,households,andpolicymakersinformingexpectationsforupcomingrealizations,andsubsequentlymakethemoreappropriateeconomicdecisions.



5 FinancialLinkedOpenData

5.1 LinkedOpenData(LOD)

Linked(Open)Data(Aueret.al,2013), (HeathandBizer,2011), (BorrowerandKaltenböck,2011)isanemergingconceptforpublishingstructureddataontheWebasinterlinkedRDFgraphs. It isbasedonbasicprinciplessuchas: (i)useHTTPURIstonamethings, (ii) returnRDFaboutthosethingswhentheirURIsarelookedup,and(iii)includelinkstorelatedRDFdocuments elsewhere on theWeb. The differences betweenOpenData and LinkedOpenDataarebestdescribedby5StarsModelintroducedbySirTimBerners-Leein20107whereLODcomesintoplayfor4and5starsofthemodel:

1star InformationisavailableontheWeb(anyformat)underanopenlicense

2stars Informationisavailableasstructureddata(e.g.Excelinsteadofanimagescanofatable)

3stars Non-proprietaryformatsareused(e.g.CSVinsteadofExcel)

4stars URIidentificationisusedsothatpeoplecanpointatindividualdata

5stars Dataislinkedtootherdatatoprovidecontext

OnecanbenefitfromLinkedDatatechnologyasapublisherofdataandasaconsumer.Forexample, as a consumer of 5 stars open data one would be able to link and bookmarkparticular pieces of data as well as discover new data of interest while consuming otherinformation.Asapublisherwith5starsdataoneincreasesthevisibilityandthevalueofthedataandatthesametimeenablesfine-granularcontroloverthedata8.In addition to providing guidelines for publishing and consuming data on the Web andinterlinking data just like Web pages, Linked Data is also emerging as a promising DataIntegration and Data Interchange paradigm as well as this Semantic Web providescomprehensivemechanismsforcomplexqueryingandreasoning.The growing adoption of these principles, aswell as the increasingly common practice ofembeddingmetadataintoHTMLpages(suchaspromotedbyschema.org),haveledtoavastvolumeofRDFdatabeingmadeopenlyavailableonline,contributingtowhatisoftencalledthe “LOD Cloud"9. Huge amounts of data containing millions of individual instances andrelations between them are now openly available on the web. The English version ofDBpedia10,whichisthemachine-readablesemanticwebversionofWikipedia,containsover4.5millionofdifferentthings.

7dvcs.w3.org/hg/gld/raw-file/default/glossary/index.html#linked-open-data85stardata.info/en/9lod-cloud.net10wiki.dbpedia.org



5.2 AvailableVocabularies

Controlledvocabulariesofferawaytostandardizethedatarepresentationand, therefore,easytheconsumptionofthedata.Inthecontrolledvocabulariesmainclassesandpredicatesare defined. Among themost widely known and used controlled vocabularies are SimpleKnowledgeOrganization System (SKOS)11andWebOntology Language (OWL)12. The SKOSdescribes theconstructs toconstitute thesauriandOWL is thesemanticweb language fordescribing ontologies. Besides such general controlled vocabularies, domain specificvocabularies areoftendeveloped. In this chapter some thesauri andontologies related tofinancialLODarelisted.First it isnecessary tounderstandhowcomplex theneededontologies (the complexityofthesauri is fixed) could be. The experts from Tecnología, Información y Finanzas (TIF)(Castellisetal.,2004)haveidentifiedthattheResourceDescriptionFramework/RDF(withits model of representing data as triples) should be enough. The transitive closure ofsubPropertyOf and subClassOf relations fromOWL, the domain and range entailments ofproperties,andthe implicationsofsubPropertyOfandsubClassOf,weretheonly inferencemechanisms used by TIF. The project team has created an ontology, however, all theelements of the ontology are described in Spanish. Nomultilingualism support has beenconsidered,asthebusinessactivitiesofTIFaremainlyfocusedontheSpanishmarket.

5.2.1 Abbreviations

FIBO13isamodularizedformalmodeloftheconceptsrepresentedbyfinanceindustrytermsas used in official financial organization documents such as contracts, product/servicespecificationsandgovernanceandregulatorycompliancedocuments.ThisisreferredtoasaBusiness Conceptual Model as distinct from models or descriptions of data or ITimplementations.Thescopeoffinanceindustryencompassesabroadrangeoforganizationsthat manage money, including credit unions, banks, credit card companies, insurancecompanies, consumer finance companies, stock brokerages, investment funds and somegovernment sponsored enterprises. This particular specification defines the Foundationsmodule of FIBO: a set of business concepts which are intended to support the financialindustry terms semantics presented in other FIBO specifications. Foundations is itselfsegmented into a number of models or ontologies. The FIBO Foundationsmodels definegeneralconceptsthatarenotuniquetothefinancialindustry,butneededtohelpdefinethefinancialconcepts.FIBOFoundationsthereforeincludesanumberofbasiclegal,contractualandorganizationalconcepts,amongothers.Conceptswhichareavailable inother industrystandardsarenot included,but insomecasesa“Proxy”concept is includedforreference,forexampleforaddressandcountryconcepts.Therationaleforincludingtheseistwo-fold:● Concepts in the financial industry are generally specializations ofmore general, non-

financial concepts suchas contracts, commitments, transactions,organizationsand soon, These are included in FIBO Foundations so that specializations of them may bedefinedinotherFIBOspecifications;

11https://en.wikipedia.org/wiki/Simple_Knowledge_Organization_System12https://en.wikipedia.org/wiki/Web_Ontology_Language13edmcouncil.org/financialbusiness



● Properties of financial industry concepts frequently need to be framed in terms ofrelationships to non-financial concepts such as countries, jurisdictions, addresses andthe like. These are included in FIBO Foundations so that properties in other FIBOspecificationsmaymakereferencetothem.

OneofthekeybenefitsofFIBOwithrespecttodata,messageorreasoningmeta-modelsisthatitcanprovideasemanticanchorfirmlyrootedintheconceptsasunderstoodandusedbypeopleinthefinanceindustry.FIBOenablesthecreationoflogicaldatamodelssuchthatthoselogicalmodelsderivetheirformalsemanticsfromFIBO.FIBOsupportsthederivationofontologies to support semantic reasoningandqueryingapplications. SinceFIBO itself isframedusingtheformalconstructsoftheOWLlanguage,suchoperationalontologiesmaybe derived directly from the FIBO conceptual ontologies,with adaptation as necessary tosupportanyapplicationspecificconstraints.FIBOallowsdisambiguationofnewandexistingregulation. To the extent that regulatory requirements reference the formal concepts inFIBO,termsreferredtointheseregulatoryrequirements,or inreportsthataremandated,wouldbesemanticallyunambiguous.One important goal of FIBO is for the formal business definitions to be used in legaldocuments such as contracts, terms and conditions of sales and payment, IP protection,compliancereports;andtounderpinlessformallanguageusedinadvertisingandcustomer-facingwebsites. Thebusiness termsanddefinitions in this specificationmaybeusedas areferencemodeltowhichfirmswouldtietheirownproprietarymodels(semanticmodelsorontologies);andalsoasacatalogueforalloftherelevantdatamodels.FIBO is managed as a collaborative effort among financial institutions and vendors toprecisely define the terms and characteristics of financial instruments, business entities,pricingandfinancialprocesses.FIBOismanagedbythemembersoftheEDMCouncilviatheFIBO Content team (FCT) process. Content teams are formed to perform an in-depthevaluationof a FIBOdomain. FCTs are operational for Equities, Loans, Indices/Indicators,FIBO-Foundations,BusinessEntities,FIBOprocessesandvendortesting.

5.2.2 STWThesaurusforEconomics

STWThesaurus14for Economics is a controlled, structured vocabulary for subject indexingand retrieval of economics literature. The vocabulary covers all economics-related subjectareasand,onabroader level,themost importantrelatedsubjects(e.g.socialsciences). Intotal, STWcontains about6,000descriptors (keywords) andabout20,000non-descriptors(searchterms).Alldescriptorsarebi-lingual,GermanandEnglish.STWhelpsuserstosearchtheZBWcatalogueECONISwiththemostsuitableterms,becausetheholdingsareindexedwith its descriptors. Having been developed as a general indexing and retrieval tool foreconomics publications, STW can be used by other organisations for indexing their ownmaterials.DetailsforusearegivenontheSTWonlinehomepage.TheSTWthesaurusconsistsofsevenmainsubjectgroups:A GeneralDescriptorsB BusinessEconomics

14zbw.eu/stw/version/latest/about



G GeographicNamesN RelatedSubjectAreasP CommoditiesV EconomicsW EconomicSectorsandIndustryStudies[A]containsalistofgeneralterms.Thesetermsdonotbelongtoanyspecialsubjectareasandthereforepossessonlylittleinformationvaluewhenusedalone.Thesubjectgroupingsofparts[B]und[V]followtheusualsubjectfieldsinnationalandinternationalresearch.Thedescriptorgroupingsofparts[P]und[W]aremostlytakenfromthecurrentcommodityandsectorclassificationsofStatistischesBundesamt.Thegeographicsectionsofpart[G]reflectthecontinentalsubdivisions,startingwithEurope.ExclusiveresponsibilityforthecontinuousmaintenanceoftheSTWlieswithZBW.Aregulareditorial team verifies and decides on changes and updates concerning vocabulary andthesaurusstructure.Proposalsbyindexersandthecurrentprofessionaldiscussionaretakenas sources for changes to be made. The STW master file has been implemented into athesaurus maintenance tool based on an in-house development. The consistentmaintenance and revision is assured by carrying out plausibility checks and removingduplicate terms. Each descriptor is given an ID-number, which serves as a unique andinalterablekey.

5.2.3 EuropeanSkills,Competences,QualificationsandOccupations(ESCO)

ESCO15is themultilingualclassificationofEuropeanSkills,Competences,QualificationsandOccupations.ESCO ispartof theEurope2020strategy.TheCommissionservices launchedtheproject in2010withanopen stakeholder consultation.DGEmployment, SocialAffairsand Inclusion – supported by the European Centre for the Development of VocationalTrainingCedefop–coordinatethedevelopmentofESCO.Stakeholdersareclosely involvedinthedevelopmentanddisseminationofESCO.The ESCO classification identifies and categorises skills, competences, qualifications andoccupationsrelevantfortheEUlabourmarketandeducationandtraining.Itsystematicallyshows the relationships between the different concepts. ESCO has been developed in anopenITformat,isavailableforusefreeofchargebyeveryoneandcanbeaccessedviatheESCOportal.The first version of ESCO was published on 23 October 2013. This release marks thebeginningofthepilotandtestingphase,includingtheESCOmappingpilot.Untilendof2016theclassificationwillbecompletelyrevised.ThefinalproductwillbelaunchedasESCOv1.ESCO is structured in three pillars: occupations; skills and competences; qualifications. Allthree pillars are structured hierarchically and interrelated with each other. The ESCOoccupationspillarcontainsoccupationgroupsandoccupations.Theskillsandcompetencespillar contains skills, competences, knowledge as well as skills and competence groupconcepts. The qualifications pillar contains qualification groups and qualifications. The15ec.europa.eu/esco/portal/escopedia



hierarchicallystructuredqualificationgroupsarebasedontheISCEDFieldsofEducationandTraining201316.DG Employment, Social Affairs and Inclusion – supported by the European Centre for theDevelopmentofVocational Training (Cedefop) andbyexternal stakeholders – coordinatesthedevelopmentofESCO.ShapingESCOintoanup-to-date,practicaltoolcanonlybedonefrom the bottom up, through the active involvement of people from the education andtraining sector as well as from the labour market. Stakeholders contributing to thedevelopmentoftheclassificationinclude:● employers'organisations● tradeunions● employmentservices● educationinstitutions● trainingorganisations● statisticalorganisations● Europeanandnationalsectorskillscouncilsandnetworks

5.2.4 EuroVoc

EuroVoc17isamultilingualthesaurus,whichwasoriginallybuiltupspecificallyforprocessingthe documentary information of the EU institutions. It is a multi-disciplinary thesauruscovering fields, which are sufficiently wide-ranging to encompass both Community andnational points of view,with a certain emphasis on parliamentary activities. EuroVoc is acontrolledsetofvocabulary,whichcanbeusedoutsidetheEU institutions,particularlybyparliaments.The aim of the thesaurus is to provide the information management and disseminationserviceswithacoherent indexingtool fortheeffectivemanagementof theirdocumentaryresources and to enable users to carry out documentary searches using controlledvocabulary.TheEuroVocthesauruscoversalltheactivityfieldsoftheEuropeanUnioninstitutions:● Politics● Internationalrelations● Europeancommunities● Law● Economics● Trade● Finance● Socialquestions● Educationandcommunications● Science

16 https://ec.europa.eu/esco/portal/escopedia/International_Standard_Classification_of_Education%253A_Fields_of_Education_and_Training_2013_%2528ISCED-F%252917http://eurovoc.europa.eu/drupal/?q=abouteurovoc&cl=en



● Businessandcompetition● Employmentandworkingconditions● Transport● Environment● Agriculture,forestryandfisheries● Agri-foodstuffs● Production,technologyandresearch● Energy● Industry● Geography● InternationalorganisationsSomefieldsaremorehighlydevelopedthanothersbecausetheyaremorecloselyrelatedtotheEU’scentresofinterest.Thus,forexample,thenamesoftheregionsofeachEUMemberStateareinEuroVocbutnotthoseofthirdcountries.Thegroupingofdescriptorsintofieldsis to a certain extent arbitrary. One of EuroVoc’s distinctive features is the limitation ofpolyhierarchy.Descriptorswhichcouldfitintotwoormoresubjectfieldsarethusgenerallyassignedonlytothefieldwhichseemsthemostnaturalforusers, inordertofacilitatethemanagementofthethesaurusandlimititsvolume.TheEuroVocthesaurusisadaptedcontinuallytotakeaccountofdevelopmentsinthefieldsin which the European Union institutions are active and of changes in its languagearrangements.Themaintenance teamatthePublicationsOfficecollectsandexaminestheproposals, put forward by the national parliaments, the European institutions or agenciesand the private users. A maintenance system has been established to deal with theadministrationandmaintenanceofthethesaurus.Themaintenanceteamcoordinatestheworkof theMaintenanceCommitteeand is responsible for ITandtechnicaldevelopmentsandformonitoringtranslations;TheMaintenanceCommitteemadeupofrepresentativesofthevariousEUinstitutions,votesonthevariousproposalsanddecidesontheamendmentstobemadetothethesaurus;TheSteeringCommitteesupervisestheprojectandofficiallyadoptseachnewversion.

5.3 AvailableFinancialLinkedOpenDataResources

There is a high public demand to increase transparency in government spending. Openspending data has the power to reduce corruption by increasing accountability andstrengthens democracy because voters canmake better informed decisions. An informedand trusting public also strengthens the government itself because it is more likely tocommit to largeprojects (Höffner,Martin and Lehmann, 2015). Following these ideas theinitiativesofEuropeanCommission18haveencouragedtheemergenceof largeamountsofpublicsectordatatobecomefreelyavailableforuseandredistributionwithoutrestriction.TheEUhasalsomandated that collected financial, economicand legaldata setsbemadeavailable as Open Data for integration and innovative reuse within new products andservices19.

18EuropeanCommission.EuropeanInformationSociety,PublicSectorInformation–RawDataforNewServicesandProducts.http://ec.europa.eu/information_society/policy/psi/index_en.htm2003.19EuropeanCommission.MESPIR.ec.europa.eu/information_society/policy/psi/mepsir/index_en.htm2006.



Theopen financial data, like thedatapublishedbyEurostat20(EU), theWorldBank21, andInternationalMonetary Fund22, could potentially be exploited in order to build advancedinformation portals for the financial sector (O'Riain, Curry and Harth, 2012). Furtherexamplesinclude,forexample,OpenTED23,aninitiativethatprovidesdatadumpsoftendersfrom the joint European procurement system more easily accessible to journalists andresearchers.However,sincethebeginnings,nearlyalloftheopendatainitiativesaroundtheglobesufferfromrelativelylowusageoftheirproduceddata.Existingprojectsarebasicallyactase-catalogsofdatathatarefragmentedintopic,placeandtimesincetheydonotsharecommonstandardsandmethodologies.Incaseswhereopendataexist,thebasicobstacleisthe fact that there are not even commonpractices for representing themain actors (e.g.payers and payees) and the type of payments. Therefore, it is impossible to interlink theavailabledatainmeaningfulwaysandsupportservicesanddecision-making(Vafopoulosetal.,2016)Financial data is hardly valuable in pieces, without having clear relations to the overallpicture.AsHatsuKim,VPGlobalFundamentalsThomsonReuters,noted“whenitcomestovaluingcompanies,quarterlyandyearlyfilingsareinsufficientininformationtermsandhaveto be considered together with information on markets and exchange rates”24. When itcomes to corporate data, analysts and investors constructing detailed insight into anorganisation develop their understanding through examination of a diverse range ofbusinessand financial information.Performingananalysis can require informationvaryingfrom operational figures, new product announcements, risk exposure, sector spend,independent analysis and customer sentiment. The source of this information includesinternalreports,socialplatforms,marketingbriefs,regulatoryfilings,analystreports,pressreleases, government statistics and third party information providers. Informationprofessionalsfacethedifficultyofhowtoachievefasterandmoreaccurateanalysisacrossthese disparate financial information sources that present and behave as islands ofinformation. It is clearly necessary to have a more holistic view on the financial data.However, low quality andmultilevel data fragmentation raise barriers in interconnecting,analysing and inferencing frommultiple data sources. For instance, budget and spendingclassificationsareincompatibleevenwithinasingleorganizationorcountryandthereisnotastandardmethodforrepresentingcompanynamesandactivities(Vafopoulosetal.,2016).

5.3.1 BudgetsandSpendingData

Severalcountriespublishtheirfinancialdata,budgetdocuments,andreportsinopenaccess.Notable examples include USA25, Brazil26, Korea27, France28. Clearspending29is a Sunlight

20http://epp.eurostat.ec.europa.eu/portal/page/portal/eurostat/home/21http://data.worldbank.org/data-catalog22http://www.imf.org/external/np/fin/tad/exfin1.aspx23http://ted.openspending.org/24W3C.ReportfortheWorkshoponImprovingAccesstoFinancialDataontheWeb,Co-organizedbyW3CandXBRL International, and hosted by FDIC, Arlington, Virginia USA, 5-6 October; 2009http://www.w3.org/2009/03/xbrl/report.html.25usaspending.gov26portaldatransparencia.gov.br27eng.openfiscaldata.go.kr



Foundation’s project that analyses the spendingdatauploaded in theofficial portal ofUSgovernment.Insomecountriestherearecivilinitiativesthatservethepurposeofcollectingspending data, e.g. in Russia30. In the field of open budgets, the International BudgetPartnership (IBP) advocates for public access to accountable budget systems. One of itsprojects,theOpenBudgetSurveyTracker(OBSTracker)31allowscitizenstomonitorwhethercentral governments are releasing the requisite information on how the government ismanaging public finances. Although the information about reports is linked to country’sprofilesandcanbetrackedoverdifferentyears,theOBSTrackeronlymonitorsthefactofpublishingthereports,nottheinformationcontainedinthereports.ComplementedbytheOpen Budget Index (OBI) issued every two years, and the Open Budget Survey thepartnership offer a comprehensive stack of tools and reports tomonitor the openness ofstatebudgets32.Moreover,IBPofferaninteractiveinflationguide.Theguidemaybeusedtotoexplainthebasicsofinflation,differenttypesofinflation,pricelevels,nominalversusrealprices,andhowtodovariouscalculationsusinginflation.However,thedataaboutspendingevenifprovidedbythegovernmentisusuallynotlinked.Starting from2015,EUfundstheOpenBudgetsproject33toprovideascalableplatformforpublic administrations to publish open budget data that is easy-to-use, flexible, andattractiveforall.ThegoalofOpenBudgetsistoprovideitsuserswitheasytousetoolsforcomparative analysis and data mining. Different tools will be built to perform temporal,spatial, and administrative comparative analysis and data mining operations. Thevisualisationservicecontributestotheanalysisandpresentationofthedatabyprovidingofftheshelfvisualizationtoolsontheplatform34.The intentionoftheproject istocollectthe“flat”OpenData for the governmental portals and to link it using LOD cloud. These newsemanticresourcesarethenprocessedfurther.Cleaningupofdatainresourcesisrequiredsincetheuploadedbudgetdatausuallycontains inconsistenciesand/ornon-optimalstates(e.g.duplicationsandcontradictions).Moreover,thedatawillbeenrichedbylinkingthemtootherrelatedsemanticresources(inthecreatedknowledgebaseandLODCloud).Asofnow(April2016)noresultsfromOpenBudgetsareavailableyet.NoreportsareavailableinWorkPackage2(WP2-LinkedDataLifeCycle)thatdealswithdatalinking.Onlysomeresultsonthelinkingofthevocabulariesarereported35.OpenSpendingwebsite36offersaneasysystemtoupload,exploreandsharepublicfinancedata (e.g.budgetsorexpendituredatabases). Itaims to trackandanalysepublic spendingworldwideand,atApril2016,containsmorethan28millionfinancialentries inmorethan1100datasets.Datasetscanbesubmittedandmodifiedbyanyonebuttheyhavetopassasanity check from the OpenSpending Data Team, which also cleans the data beforepublishing. OpenSpending hosts transactional as well as budgetary data with a focus ongovernment finance. Itcontainsthisdata instructuredformstored indatabasetablesand

28data.gouv.fr29sunlightfoundation.com/clearspending/30clearspending.ru31obstracker.org32OBSTrackerlaunchreportinternationalbudget.org/wp-content/uploads/Sept-2014-OBSTracker-Report.pdf33openbudgets.eu34openbudgets.eu/about/technical-structure35openbudgets.eu/assets/deliverables/D1.8.pdf36openspending.org



providessearchingandfilteringaswellasvisualizationsandaJSONRESTinterface.Thedatasets differ in granularity and type of accompanying information, but they share the samemeta model. The system offers functionality for explorative analysis as well as forvisualization of the data. The data is open for use for any purposes (non-profit and for-profit)37.Newdatasetsare frequentlyadded.Therearepropertieswith thesamename indifferentdatasetswhere it isunknown if theyspecifythesameproperty (Höffner,MartinandLehmann,2015).In order to correlate the new datasets and the properties LinkedSpending38link the datafrom OpenSpending using semantic web principles. LinkedSpending follows the RDF DataCube39recommendationtomakeheavyuseoftheSDMXmodelformeasures,attributesanddimensions.Thedatasetsareveryheterogeneousbuttherearesomeproperties,whicharecommonlyspecifiedandthusmodelledwithestablishedvocabularies.Theyearanddate,adata set and an observation refers to, respectively, is expressed by sdmx-dimension:refPeriod and XSD. Currencies are taken from DBpedia and countries arerepresentedusing the vocabularyof LinkedGeoData40, ahub for spatial linkeddata. SomeamountofdataisimportedfromLinkedGeoDatacountriesandDBpediacurrencies.Becauseof the limited number of countries and currencies, and properties values imported percountry and currency, theamountofdata is too small to consider federatedquerying.Asmostcountriesandcurrenciesarestableinthemediumterm,thisdataneedstobeupdatedonlyinfrequently.Thelinkingofentities,properties,andattributesisdonemostlyautomaticallybasedonthestringcomparison,i.e.ifthenamesoftwoentitiesarethesameanditisnotexplicitlystatedthat the entities are different then it is assumed that they are the same. In otherwords,propertieswiththesamenameindifferentdatasetsnothavingamappingentrythatstatesotherwise are assumed to represent the same concept and thus given the same URL. Asnoted in thepaper, there is a possibility ofmismatch, however, such amismatchhasnotbeen spotted yet. If necessary, improving the automatic matching is part of future work(Höffner,MartinandLehmann,2015)Interlinkingofentitiesisbasicallydoneusingthesameentity and not using owl:sameAs. The data conversion process is controlled by a webapplication,whichconstantlychecksforaddedandmodifieddatasetsfromOpenSpending,whichareautomaticallyqueuedforconversionbutcanalsobemanuallymanaged.Updatesdon’tinterrupttheaccessibilityoftheSPARQLendpointandtheservicesbuildingonit.ThePublicSpending41projectwas the first attempt toharvest, align, interlink, analyze anddistributeasLinkedOpenDatamassiveamountsofpublicspendingdatafromseven(localandnational)governments.Intheprojectconsortiumproceededasfollows.Eachdatasetissubject to preprocessing anddata preparation. Even though the data is open, there is nohomogeneityintheinformationthedatasetscarry,notonlyacrossdomains(e.g.Greecevs.Australia),butalsowithinthesamedomain(e.g.datapublishedbydifferentdepartmentsintheUK).Becauseoftheheterogeneityofthedescriptionsacrossdatasets,minimumsetsof

37community.openspending.org/about/38linkedspending.aksw.org39https://www.w3.org/TR/vocab-data-cube/40linkedgeodata.org41www.publicspending.net



characteristicshavebeen identifiedanduniversallyapplied inorder toprovidea commondenominationofthedatamodel.Thebasicconceptsinthepublicspendingdomainarethepaymentanditsparticipantsalongwiththeirmetadata.PaymentsaregivenURIsbasedontheiruniqueidentifiers,buttheURIsthemselvesindicatethedomainoforigin.TheEuropeanUnionhasestablishedtheuseoftheCommon Procurement Vocabulary 2008 (CPV) as mandatory for all public procurementnoticesaccordingtotheRegulation(EC)Nº2195/2002oftheEuropeanParliament.Inordertocreateacomparisonofthecontractsobjectpublishedbydifferentpublicadministrationsahubclassificationmustbeusedtounifythosedescriptions.TheCPVhasbeenselectedasahubclassification.SomelinksfrominternationalproductmappingstoCPVwereavailableasLOD, the rest of them have been created applying custom NLP techniques to link adescriptionofaproduct/servicetoaCPVconcept.Payers and payees are described using some form of unique identification in order togenerateURIs,dependingonthedataorigin,aswellastheirnames.Avarietyoferrorscanbefoundsuchasmisspellingerrors,samecompanyunderdifferentnames,useofdifferentkindofacronyms,etc.Forexampleacompanysuchas“Oracle”,“Accenture”or“Capgemini”canbefoundunderdifferentnameconventions:“OracleAust.”,“OracleUniversity”,“OracleCorpora tion (Aust) Pty Ltd”, “Oracle Corp Aust P/L”, “Accentrure”, “Accenture Australia”and“CAPGemini”amongothers.IntheSemanticWebareaandmorespecificallyintheLODinitiative one of the principles lies in providing a real unique identification of resourcesthrough URIs. Thus reconciliation techniques coming from the ontology mapping andalignmentareasoralgorithmsbasedonNaturalLanguageProcessinghavebeendesignedtolink similar resources already available in different vocabularies, datasets or databases. Acontext-aware method based on NLP techniques has been designed, customized andimplemented trying to exploit the naming convention of the dataset. The technique togenerateauniquenamebeforeperformingthereconciliationprocessisastepwisemethod,inwhicheachstepperformsafilteroverthestringliteraltryingtoremoveallunnecessarywords inthenametofinallyusean iterativeprocessofstringcomparisonandgroupingtogenerateauniqueandrelevantnamefortheinputdataset.Afterpreparingthedata,RDFmodelswerecreatedusingthepublicspending.grontologyaswell as other widely used vocabularies such as Dublin Core and FOAF. The resulting RDFmodelswerethenserializedinRDF/XMLformanduploadedtotheVirtuosoQuadStoreheldinpublicspending.gr. Foreachdomainof interest thedata is stored inadedicatednamedgraph.ThisprocedurecreatedsevennamedgraphsthatcoverspendingdataforGreece,theUS,theUK,Australia,thecityofChicagoandthestatesofAlaskaandMassachusetts.Intotalthere were identified 87,629,622 triples and 1,466,520,742,920 Euros of payments. Thedifferentnamedgraphsare interlinked intwodifferent levels: (i)ontheconceptualmodellevel they use the same vocabularies, and (ii) on the datamodel through the use of CPVcodesandaggregationsofcompaniesfromdifferentdatasets.Thismakesthesupersetopentocomparisonsandanalysisacrossdomains.

5.3.2 CorporateReports

So far most of the financial open data is related to companies’ reports. The eXtensibleBusiness Reporting Language (XBRL) standardises financial reporting with a machine-



interpretable format that makes corporate reports easier to consume and integrate. TheXBRL standard isanXML formatand is thereforemoreamenable toautomaticprocessingthan traditional financial information representations such as PDF, HTML and textdocuments.However,XBRL-basedinformationsourcesonlyprovidepartofthepicture,andmanyotherdatasourcesareusedinconjunctionwithXBRL.Withinthewiderglobalfinancialand business information ecosystem of corporate press releases, government statistics,market press coverage and third party information providers, XBRL data repositoriesrepresents yet another data silo in a global landscape of disconnected silos. In order toconsumethedatatheuserscouldfindusefulseveralanalyses:● Background information analysis, e.g., looking at company information from different

sourcessuchastheaddress,thefoundingdateandtheindustry.● Multi-companyKPIanalysis,e.g.,comparingKPIsovertimeforseveralcompaniessuch

asthestockmarketpriceforcompaniesfromthesameindustry.● Cross-data-sources KPI analysis, e.g., comparing values from heterogeneous datasets

suchas theEarningsperShare fromyearlybalance sheetswithpricesper share fromelectronic stock quotes as well as Total Assets published using the US-GAAP version2009andversion2011.

Combining XBRL with the plethora of non-XBRL financial information remains at an earlystage of realisation. Significant efforts are required on the part of financial informationconsumers to integrate XBRL data with other data expressed in a wide variety of dataformats.• Automatically deriving information from XBRL is difficult since formal semantics are

limited (M. Spies, 2010). Relationships between financial concepts, such as“SalesRevenueNet”and“Revenues”intheU.S.GenerallyAcceptedAccountingPrinciples(US-GAAP),areonlytextuallydescribed.

• Financial information fromdifferent XBRLdocuments often cannot be compared sinceaccounting and regulatory organisations do not align their taxonomies of financialconcepts;newversions,e.g.,ofUS-GAAP,lackbackwardcompatibility;andXBRLallowspublisherstodefinetheirownconcepts.

• Gathering informationaboutacompany isdifficultsincetherearenouniquecompanyidentifiersacrossdifferentreportingsources5andrelationshipsbetweencompaniesareobscure.

• Otherfinance-relatedOpenDatasuchasstockquotesandbackgroundinformationarepublishedusingdifferentdatamodels(Kämpgenetal.,2014).

Next,wereviewseveralprojectsthattacklesomeoftheabovementionedissues.

In FLORA project (Razmiski et al., 2012) authors consider the problem of fostering thetransparent access to financial data through its dataset developed from extraction andtriplificationof thedata contained in the financial statementsof companies registeredonU.S. Securities and Exchange Commission (SEC) 42 through filings and forms stored on

42sec.gov



EDGAR43 System and furthermore allowing an easy combination of many informationsources. Moreover authors aim at increasing the FLORA functionality through the use offrameworks fordiscovering relationshipsbetweencompanies containedonFLORAdatasetand data itemswithin different LinkedData sources, i.e. linking of entities. Hence, in theproject a complex, unified process of transforming unstructured financial data into aninterlinked, navigable knowledge base for financial information management andinformationdiscoverywithintheWebofdatawasdesignedandimplemented.The transformation and generation process of FLORA dataset consists in investigatingappropriatedatasources thatare rich in financial information:balancesheets,cash flows,andincomestatementsfromcompanies.Asof2012authorshavegeneratedadatasetthatstoresRDFtriplesgeneratedfromtheextractionof409.374financialstatements,whichhavebeen published in XBRL format by different U.S. companies through the EDGAR systemfillings.As external dataset authorshave chosenDBpedia.Authors approached the taskof linking8915companiestotheirDBpediaentrieswithdifferent linguistic techniques.Thetaskwasapproachedusingthetechniquesofstringcomparison.Thebestresultswereachievedaftereliminating stopwords and allowing Levenshtein distance to vary between 0 and 1.Withthesesettingsitwaspossibletodiscover3652links.Theexperiment,however,showsthatthecoverageofthefinancialdomainisrathersmall.Thelackofdatathatcouldbeusedtounivocallyidentifycompanyconcepts(suchasCIKortickersymbol)makesdatasetinterlinkingstilladifficulttaskrequiringvarioustechniquesfordisambiguationandmanualverification.In the Financial Information Observation System (FIOS) (Kämpgen et al., 2014) authorsmodeltheXBRLdatausingtheRDFDataCubeVocabulary.Thisisdoneinordertoachievethe requirement identified in the paper, namely, to enable the user to browse the data(“overviewfirst,zoom-in,detailsondemand”)andcreatehisownanalysesifpossiblewithExcel-likefunctionality.Moreover,thesystemshouldbeabletoeasyprocessandstorenewdata.ThedatawastakenfromdifferentsourcessuchasDBpedia,EDGAR,andYahoo!Finance44.Allpiecesofdatawasuniquelyidentified.EveryXBRLinstanceandtaxonomywasmodelledasamultidimensionaldataset,i.e.,collectionoffactswithindependentdimensionvariablesand dependentmeasure variables, using awell-adopted Linked Data vocabulary, the RDFDataCubeVocabulary.Similarly, stockquotes fromYahoo!FinancecanbemodelledusingData Cube Vocabulary: every daily collection of values is a dataset, every stock quotecontainsasdimensions thecompany, thedate thevalue isvalidand thestockquote typesuchaspriceatstockmarketopening(Open).Onthenextstepitisnecessarytolinkthedifferententities.Whereastimeperiodscaneasilybematchedbycomparingcanonical representationsof time, for linkingbetweendifferent

43sec.gov/edgar/searchedgar/companysearch.html44finance.yahoo.com



URIforcompaniesandfinancialconceptsacrossdatasourcesmappingsneedtobeavailable.EntitiesorpropertiesinRDFcanexplicitlybestatedasequivalentviarelationshipsbetweentheirURIs.Tomodel finance relatedmetadata, e.g., about companies and industries,weusewidely-adoptedLinkedDatavocabularies,e.g.,FOAFandtheDBpediaontology.The industryofacompany can be represented using SKOS classification in FIOS hierarchies, e.g., the SECprovides for companies the Standard Industrial Classification (SIC) hierarchy, e.g., SICconcept“SERVICES-BUSINESSSERVICES,NEC”skos:narrower“MASTERCARDINC”.The Semantic XBRL project45aims at triplification of the XBRL data. Semantic XBRL is arepositorythatcontainssemanticdatageneratedfromthedatasubmittedbyparticipantsintheXBRLProgramEDGARpromotedbytheU.SSecuritiesandExchangeCommission(SEC).Semantic data is generated by translating XBRL XML to RDF following the XML SemanticsReuse Methodology, which is implemented by the ReDeFer tools. First, there is theXSD2OWLtranslation,whichgeneratesXBRLOWLontologiesfromtheXBRLXMLSchemas.Then XML2RDF,which translate XBRL XML instances to RDF taking into account the XBRLOWLontologies.ThedifferencebetweenFIOSandtheSemanticXBRListhatthelattertiesRDFrepresentationsofXBRLclosetotheoriginalXMLdatawhichmakemixingwithotherdatasourcesdifficult.In the field of company data, OpenCorporates46aggregates company information fromdifferentcountriesand jurisdictions.Theopencorporates.comteam isworkingoncreatingLinked Data representations out of their databases, by mapping company metadata tocertified ontologies such as the Core Business Vocabulary and linking them to other datahubs,suchasDBpedia.organdGeonames.Financialstatementsandaccountingreportsthatarepublishedbycompaniescontainimportantdata.Intheareaofproducts,OpenProductData47hasbeeninitiatedbyPhilippePlagnol(nowishostedbyOKFasacommunityproject)asapublicdatabaseofproductdataconnectedtobarcodes in order to empower consumers with useful and machine-readable productinformation(e.g.price).The Thomson Reuters Permanent Identifier or PermID48is a machine readable, 64 bitnumberusedtocreateauniquereference-whichwillneverchangeovertime-toapieceofinformation. Thomson Reuters has been using the PermID company-wide to improveinternal data federation. A subset of PermIDs is available through an open license on thepermidsite.Thisprovidessignificantandno-costaccesstoavailableThomsonReuters-linkeddata.ThomsonReuterspromisetocontinuetoupdate,buildandadministerourPermIDs.

45rhizomik.net/semanticxbrl/html/46opencorporates.com47product.okfn.org48permid.org



5.4 Financial-specificLODandthePROFITadvances

Inthissection,wehaveanalysedtheadvantagesoflinked(open)dataoverthedatathatisnot linked.Besidesgeneraladvantagessuchasextensibilityandreusabilitytherearesomefeatures that are specific to the financial field, the main being that the financial andeconomicdataonlymakessensewhenputintotheglobalcontext.Several available controlled vocabularies were identified and described during this studywherebysomeofthemarealsoavailableasLinkedOpenDataandcanbeused.Theycoverpretty large variety of different topics and constitute a solid baseline for furtherdevelopment. On the next phases of the project the described vocabularies may becomplemented by further ontologies and thesauri that cover other areas related to theproject(forexample,descriptionofeducationalmaterials).Thealreadyidentifiedcontrolledvocabularies will be extended themselves if the need for this will be identified. Thecontrolled vocabularies lay a grounding for linking and integrating the data into the datacloudoftheproject.Moreover,theyaddasemanticlayerforthedatafacilitatingthebetterunderstandingofthedataandmatchingthedatawiththeuserprofiles.Anoverviewof theexistingapproaches to linking the financial data reveals that themainobstacle to the consumption of the published financial data is the linking of the entities.WithinPROFIT,wewilladdressthisobstaclebyperformingthehigherlevellinkingtogetherwiththeanalysisofthetrendsandthemarketsentimentinthefinancialfield.



6 Conclusions

Thisdeliverable isregardedasan initialsketchfortheroadmaptobefollowedbyPROFIT.Overall, it discussed the current advances in the area of financial literacy, financial andeconometricforecastingandopenlinkeddataforsupportingfinancialdecision-making,andattemptedtoshedlighttotheadvancestobeintroducedbythePROFITplatformintheareaoffinancialdecisionmakingandawareness,namely

a) Novelfinancialeducationtoolkits;b) Novel financial forecasting models incorporating notions of market sentiment to

identifymarkettrendsandthreats;c) Advancedcrowd-sourcingtoolsbasedonsemantictechnologiestoprocessfinancial

data,extractandpresentcollectiveknowledge;d) Incorporation of linked data principles towards developing and establishing a

methodologyforadatalifecycle.Thestate-of-the-artsectionsofthisreport includedoverviewsofmethods,toolsandfieldsof science relevant to the PROFIT project.More, elaborative reviews and analyseswill beperformedintherespectivedeliverablesofeachuniqueresearchanddevelopmentpillarofPROFIT.Regarding the structure, this deliverablewas divided into threemain parts. The first partreviewed the most popular financial awareness platforms that are available online andprovidedanoverviewoftheservicessupportedbythem.Thesecondpartof thedocument focusedon the twomainscientificpillarsof thePROFITproject, namely the financial literacy and financial forecasting components. Initially, thesecondpartreviewedthebestcurrentpracticesintheacademicandprofessionalagendaonfinancial literacy as well as the currently available financial literacy tools and toolkits.Subsequently, it described the main market sentiment indicators used in the empiricalresearchandhowtheseindexesareextractedusingsentimentanalysistechniques.Finally,itreviewedthemostprominentmodelsusedtoforecastwell-knowneconomicindicators.Thefinalpartofthedeliverablepresentedtechnologiesemployedintheenvisagedfinancialawareness platform, including the financial linked open data. Existing thesauri arepresented,followedbyadiscussiononsemanticresourcesandevolvingontologies.



7 References

Agarwal,S.,J.Driscoll,X.Gabaix,andD.Laibson(2009).Theageofreason:Financialdecisionsoverthelifecyclewithimplicationsforregulation.BrookingsPapersonEconomicActivity,Fall2009,pp.51–101.

Angeletos,G.M.,andJ.La'O(2013).Sentiments,Econometrica,81(2),pp.739-780.

Akerlof,G.A.,andR.J.Shiller(2010).Animalspirits:Howhumanpsychologydrivestheeconomy,andwhyitmattersforglobalcapitalism.PrincetonUniversityPress.

Almenberg,J.,andA.Dreber(2011).Gender,financialliteracyandstockmarketparticipation.WorkingPaper,StockholmSchoolofEconomics.

Almenberg,J.,andJ.Widmark(2011).Numeracy,financial literacyandparticipationinassetmarkets.Mimeo,SwedishMinistryofFinance.

Apel,M.,&Grimaldi,M.(2012).Theinformationcontentofcentralbankminutes.RiksbankResearchPaperSeries,(92).

Aue,AnthonyandMichaelGamon.2005.Customizingsentimentclassifierstonewdomains:Acasestudy.InProceedingsoftheInternationalConferenceonRecentAdvancesinNaturalLanguageProcessing,Borovets,Bulgaria.

Auer, S., J. Lehmann, A.-C. N. Ngomo, and A. Zaveri (2013). Introduction to Linked Data and itsLifecycleontheWeb.InReasoningWeb.Springer.

Azariadis,C.(1981).Self-fulfillingprophecies.JournalofEconomicTheory,25(3),pp.380-396.

Baker, M., & Stein, J. C. (2004). Market liquidity as a sentiment indicator. Journal of FinancialMarkets,7(3),271-299.

Baker,M.,&Wurgler,J.(2004).Acateringtheoryofdividends.TheJournalofFinance,59(3),1125-1165.

Baker, M., & Wurgler, J. (2004). Appearing and disappearing dividends: The link to cateringincentives.JournalofFinancialEconomics,73(2),271-288.

Baker,M., &Wurgler, J. (2000). The equity share in new issues and aggregate stock returns. theJournalofFinance,55(5),2219-2257.

Baker, M., & Wurgler, J. (2006). Investor sentiment and the cross-section of stock returns. TheJournalofFinance,61(4),1645-1680.

Baker, M., & Wurgler, J. (2007). Investor Sentiment in the Stock Market. Journal of EconomicPerspectives,21(2),129-152.

Baker,M.,Wurgler,J.,&Yuan,Y.(2012).Global,local,andcontagiousinvestorsentiment.JournalofFinancialEconomics,104(2),272-287.

Barber,B.M.,Odean,T.,&Zhu,N.(2009).Systematicnoise.JournalofFinancialMarkets,12(4),547-569.

Banerjee,A.,&Marcellino,M. (2006).Arethereanyreliable leading indicators forUS inflationandGDPgrowth?.InternationalJournalofForecasting,22(1),137-151.

Batchelor, R., & Dua, P. (1998). Improving macro-economic forecasts: The role of consumerconfidence.InternationalJournalofForecasting,14(1),71-81.

Bram, J.,& Ludvigson, S. C. (1998).Does consumer confidence forecast household expenditure?Asentimentindexhorserace.EconomicPolicyReview,4(2).



S.R.K.Branavan,HarrChen, JacobEisenstein,andReginaBarzilay.2008. Learningdocument-levelsemantic properties from free-text annotations. In Proceedings of the AnnualMeeting of theAssocia-tionforComputationalLinguistics(ACL-2008)

Bauer,F.,andM.Kaltenböck(2011).Linkedopendata:Theessentials.Edition.

Becchetti,L.,Caiazza,S.,andD.Coviello (2013).Financialeducationand investmentattitudes inhighschools: evidence froma randomized experiment.Applied Financial Economics, 23(10), pp. 817-836.

Behrman, J., Mitchell, O. S., Soo, C., and D. Bravo (2010). Financial literacy, schooling, and wealthaccumulation.NationalBureauofEconomicResearchWorkingpaperNo.16452.

Benamara,F.,C.Cesarano,A.Picariello,D.ReforgiatoandV.-S.Subrahmanian,“SentimentAnalysis:AdjectivesandAdverbsAreBetterthanAdjectivesAlone,”Proc. Int’lConf.WeblogsandSocialMedia,2007.

Bernheim,D. (1995).Dohouseholds appreciate their financial vulnerabilities?An analysis of actions,perceptions,andpublicpolicy, in:TaxPolicyandEconomicGrowth,AmericanCouncil forCapitalFormation,Washington,DC,pp.1–30.

Bernheim,D.(1998).Financialilliteracy,educationandretirementsaving.InO.MitchellandS.Schieber(eds.), LivingwithDefined Contribution Pensions, University of Pennsylvania Press, Philadelphia,pp.38–68.

Bernheim,D.,andD.M.Garrett (2001).Theeffectsof financialeducation in theworkplace:evidencefromasurveyofhouseholds.JournalofPublicEconomics,87(7-8),pp.1487–1519.

Bernheim,B.D.,Garrett,D.M.,andD.M.Maki,(2001).Educationandsaving:Thelong-termeffectsofhighschoolfinancialcurriculummandates.JournalofPublicEconomics,80(3),pp.435-465.

Boiy, Erik, Pieter Hens, Koen Deschacht, and Marie-Francine Moens. 2007. Automatic sentimentanalysis of on-line text. In Proceedings of the 11th International Conference on ElectronicPublishing,pages349–360,Vienna.

Borovkova, S.,&Mahakena,D. (2015).News, volatility and jumps: the caseofnatural gas futures.QuantitativeFinance,15(7),1217-1242.

Brown,G.W.,&Cliff,M.T. (2004). Investor sentimentand thenear-termstockmarket. JournalofEmpiricalFinance,11(1),pp.1-27.

Brown,G.W.,&Cliff,M.T.(2005).InvestorSentimentandAssetValuation.TheJournalofBusiness,78(2),pp.405-440.

Brown,S.J.,Goetzmann,W.N.,Hiraki,T.,Shirishi,N.,&Watanabe,M.(2003).InvestorsentimentinJapaneseandUSdailymutualfundflows(No.w9470).NationalBureauofEconomicResearch.

Bruhn,M.,deSouzaLeão,L.,Legovini,A.,Marchetti,R.,andB.Zia(2013).Theimpactofhighschoolfinancial education: Experimental evidence from Brazil.World Bank Policy Research WorkingPaperNo.6723.

Brunnermeier, M. K. (2009). Deciphering the Liquidity and Credit Crunch 2007-2008. Journal ofEconomicPerspectives,23(1),pp.77-100.

Cadilhac, A., Benamara, F., & Aussenac-Gilles, N. (2010). Ontolexical resources for feature basedopinion mining : a case-study. Proceedings of the 6th Workshop on Ontologies and LexicalResources,23rdInternationalConferenceonComputationalLinguistics,pp.77-86.

Çakmaklı, C., vanDijk,D..2016. Getting themost out ofmacroeconomic information for predictingexcessstockreturns.InternationalJournalofForecasting,32,3,pp.650-668.



Calvet, L., Campbell, J. and P. Sodini (2007). Downor out: Assessing thewelfare costs of householdinvestmentmistakes.JournalofPoliticalEconomy,115,pp.707–747.

Calvet, L., Campbell, J., and P. Sodini (2009). Measuring the financial sophistication of households.AmericanEconomicReview.99(2),pp.393–398.

Campbell,J.(2006).HouseholdFinance.JournalofFinance.61,pp.1553–1604.

Campbell,J.Y.andS.B.Thompson.2008.Predictingexcessstockreturnsoutofsample:Cananythingbeatthehistoricalaverage?TheReviewofFinancialStudies21(4),pp.1509-1531.

Carroll, C. D., Fuhrer, J. C., &Wilcox, D.W. (1994). Does consumer sentiment forecast householdspending?Ifso,why?.TheAmericanEconomicReview,84(5),1397-1408.

Castells, P., Foncillas, B., Lara, R., Rico,M., & Alonso, J. L. (2004). Semantic web technologies foreconomic and financial information management. In The semantic web: Research andapplications.SpringerBerlinHeidelberg,pp.473-487.

Chava, S, Gallmeyer,M , Park,H..2015. Credit conditions and stock return predictability. Journal ofMonetaryEconomics,74,pp.117-132.

Chauvet,M.andS.Potter(2012),“ForecastingOutputduringtheGreatRecession,”WorkingPaper,UniversityofCaliforniaRiverside.

Chen,S-S.,2009.Predicting thebear stockmarket:Macroeconomicvariablesas leading indicators.JournalofBanking&Finance33,2,pp.211–223.

Chen,S.S.,&Chou,Y.H.Doesfearleadtorecessions?.MacroeconomicDynamics,1-17.

Chen, S. S. (2011). Lack of consumer confidence and stock returns. Journal of Empirical Finance,18(2),225-236.

Christelis, D., Jappelli, T., and M. Padula (2010). Cognitive abilities and portfolio choice. EuropeanEconomicReview.54,pp.18–38.

Christiansen,C.,Joensen,J.,andJ.Rangvid(2008).Areeconomistsmorelikelytoholdstocks?ReviewofFinance.12,pp.465–496.

Christiansen, C., Eriksen, J. N., & Møller, S. V. (2014). Forecasting US recessions: The role ofsentiment.JournalofBanking&Finance,49,459-468.

Chung,S.L.,Hung,C.H.,&Yeh,C.Y.(2012).Whendoesinvestorsentimentpredictstockreturns?.JournalofEmpiricalFinance,19(2),217-240.

Cole,S.,Sampson,T.,andB.Zia(2011).Pricesorknowledge?Whatdrivesdemandforfinancialservicesinemergingmarkets?JournalofFinance.66(6),pp.1933–1967.

Collins, C., and C. Urban (2015).Workplace financial education and retirement. Economic Inquiry,Forthcoming.

Cunningham,H.,Maynard,D.,Bontcheva,K.,&Tablan,V.(2002).GATE:AFrameworkandGraphicalDevelopment Environment for Robust NLP Tools and Applications. Proceedings of the 40thAnniversaryMeetingoftheAssociationforComputationalLinguistics(ACL’02).Philadelphia.

Dasgupta,S.,&Lahiri,K.(1993).OntheuseofdispersionmeasuresfromNAPMsurveysinbusinesscycleforecasting.JournalofForecasting,12(3-4),239-253.

DeGrauwe,P.,(2012).LecturesonBehavioralMacroeconomics.PrincetonUniversityPress.

De Grauwe, P., & Macchiarelli, C. (2015). Animal spirits and credit cycles. Journal of EconomicDynamicsandControl,59,pp.95-117.



DeLong, J.B., Shleifer,A., Summers, L.H.,&Waldmann,R. J. (1990).Noise trader risk in financialmarkets.JournalofpoliticalEconomy,703-738.

Demers,E.A.,&Vega,C.(2011).Linguistictoneinearningsannouncements:NewsorNoise?.SSRNWorkingPaperSeries.

Disney,R.,andJ.Gathergood,2014.FinancialLiteracyandConsumerCreditPortfolios.JournalofBankingandFinance,37,pp.2246-2254.

Doran,J.S.,Peterson,D.R.,&Price,S.M.(2012).Earningsconferencecallcontentandstockprice:thecaseofREITs.TheJournalofRealEstateFinanceandEconomics,45(2),402-434.

EuropeanBankingFederation (2009). Financial literacy -empoweringconsumers tomake the rightchoices.Report,Brussels:May2009.

Engelberg, J. (2008). Costly informationprocessing: evidence from information announcements. InAFA2009SanFranciscoMeetingsPaper.

Estrella,A.andHardouvelis,G.A.(1991)Thetermstructureasapredictorofrealeconomicactivity,TheJournalofFinance,46,pp.555–76.

Estrella, A. and Mishkin, F.S. (1998). Predicting U.S. Recessions: Financial Variables as LeadingIndicators.TheReviewofEconomicsandStatistics,80(1),pp.45-61.

Ferris, S. P. (2013). The Effect of Issuer Conservatism on IPO Pricing and Performance. Review ofFinance,17(3),993-1027.

Ferreira,M.A,Santa-Clara,P..2011. Forecastingstockmarket returns: thesumof theparts ismorethanthewhole.JournalofFinancialEconomics,100,pp.514–537.

Fama, E. F.,& French, K. R. (2001). Disappearing dividends: changing firm characteristics or lowerpropensitytopay?.JournalofFinancialeconomics,60(1),3-43.

Ferson,W.E. andHarvey, C.R., 1991. The variation of economic risk premiums. Journal of PoliticalEconomy,pp.385-415.

Feuerriegel,S.,&Neumann,D.(2013).Newsornoise?Hownewsdrivescommodityprices.

Florackis, C., Kanas, A., & Kostakis, A. 2015. Dividend policy, managerial ownership and debtfinancing:Anon-parametricperspective.EuropeanJournalofOperationalResearch241(3),pp.783-795.

Fung,A.K.W.,Lam,K.,&Lam,K.M.(2010).DothepricesofstockindexfuturesinAsiaoverreacttoUSmarketreturns?.JournalofEmpiricalFinance,17(3),428-440.

Garcia,D.(2013).Sentimentduringrecessions.TheJournalofFinance,68(3),1267-1300.

Gilchrist, S., Yankov, V. and Zakrajšek, E., 2009. Credit market shocks and economic fluctuations:Evidence from corporate bond and stock markets. Journal of Monetary Economics, 56(4),pp.471-493.

Gençay, R., & Gradojevic, N. (2010). Crash of'87—Was it expected?: Aggregate market fears andlong-rangedependence.JournalofEmpiricalFinance,17(2),270-282.

GerardSaltonandMichael J.McGill.1986. Introduction toModern InformationRetrieval.McGraw-Hill,Inc.,NewYork,NY,USA.

Gerardi,K.S.,Goette,L.,andS.Meier(2010).Financialliteracyandsubprimemortgagedelinquency:Evidencefromasurveymatchedtoadministrativedata.FRBAtlantaworkingpaper2010-10.

Ghiassi,M.,Skinner,J.,&Zimbra,D.(2013).Twitterbrandsentimentanalysis:Ahybridsystemusingn-gramanalysisanddynamicartificialneuralnetwork.ExpertSystemswithapplications,40(16),pp.6266-6282.



Goncalves,P., M. Araújo, F. Benevenuto, and M. Cha, “Comparing and Combining SentimentAnalysisMethods,”Proc.1stACMConf.OnlineSocialNetworks,2013,pp.27–38.

Greenwood, R., & Nagel, S. (2009). Inexperienced investors and bubbles. Journal of FinancialEconomics,93(2),239-258.

Habschick,M.,Seidl,B.,andJ.Evers,withD.KloseandY.Parsian(2007).SurveyoffinancialliteracyschemesintheEU27.VTMarkt/2006/26H-FinalReport.

Hart,R.P.(2000).DICTION5.0.Austin,TX:Digitex.

Hastings,J.,andO.S.Mitchell(2011).Howfinancial literacyandimpatienceshaperetirementwealthandinvestmentbehaviours.NBERWorkingPaper16740.

Hastings,J.,andL.Tejeda-Ashton(2008).FinancialLiteracy,Information,andDemandElasticity:SurveyandExperimental Evidence fromMexico.NationalBureauof EconomicResearchWorkingPaper14538

Hatzivassiloglou,V.,andK.McKeown(1997).PredictingtheSemanticOrientationofAdjectives.pp.174-181.

Heath,T.,andC.Bizer(2011).LinkedData:EvolvingtheWebintoaGlobalDataSpace(1stEdition),volume 1 of Synthesis Lectures on the Semantic Web: Theory and Technology. Morgan &Claypool.

Henry,E.(2006).Marketreactiontoverbalcomponentsofearningspressreleases:Eventstudyusingapredictivealgorithm.JournalofEmergingTechnologiesinAccounting,3(1),1-19.

Henry, E. (2008). Are investors influenced by how earnings press releases arewritten?. Journal ofBusinessCommunication,45(4),363-407.

Henry,E.,&Leone,A.J.(2009).Measuringqualitativeinformationincapitalmarketsresearch.

Hu,MinqingandBing Liu. 2004.Miningand summarizing customer reviews. InProceedingsof theACMSIGKDD International Conference on KnowledgeDiscovery andDataMining (KDD-2004),pages168–177,Seattle,WA.

Hüfner, F. P., & Schröder, M. (2002). Forecasting economic activity in Germany-how useful aresentimentindicators?.

Lusardi, A., Samek, A., Kapteyn, A., Glinert, L., Hung, A., and A. Heinberg, 2015. Visual tools andnarratives: newways to improve financial literacy. Journal of Pension Economics and Finance.pp.1-27.

Henkel, S.J., J. Spencer Martin, Federico Nardari.2011. The time-varying short horizon returnpredictability.JournalofFinancialEconomics,99(3),pp.560-580.

Hilgert, M., J. Hogarth, and S. Beverly (2003). Household Financial Management: The ConnectionbetweenKnowledgeandBehavior,FederalReserveBulletin,pp.309-332.

Höffner,K.,Martin,M.,&Lehmann,J.(2015).LinkedSpending:OpenSpendingbecomesLinkedOpenData.SemanticWeb,7(1),pp.95-104.

Hong,H, Torous, W.,Valkanov, R. 207. Do industries lead stock markets? Journal of FinancialEconomics,83,pp.367-396.

Hospido, L., Villanueva, E., & Zamarro, G. (2015). Finance for All: The Impact of Financial LiteracyTraining in Compulsory Secondary Education in Spain. Banco de Espana Working Paper No.1502.

Howrey,E.P.(2001).Thepredictivepoweroftheindexofconsumersentiment.Brookingspapersoneconomicactivity,2001(1),175-216.



Huang,D., Jiang, F., Tu, J.,& Zhou,G. (2014). Investor sentiment aligned: A powerful predictor ofstockreturns.ReviewofFinancialStudies,hhu080.

Jansen,W.J.,&Nahuis,N.J.(2003).Thestockmarketandconsumerconfidence:Europeanevidence.EconomicsLetters,79(1),89-98.

Jegadeesh, N., & Wu, D. (2013). Word power: A new approach for content analysis. Journal ofFinancialEconomics,110(3),712-729.

KahnemanD.,andA.Deaton(2010).Highincomeimprovesevaluationoflifebutnotemotionalwell-being.ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica,107(38),pp.16489-16493.

Li, J., Tsiakas, I. and Wang, W., 2015. Predicting exchange rates out of sample: Can economicfundamentalsbeattherandomwalk?.JournalofFinancialEconometrics,13(2),pp.293-341.

Lin,C.,andHe,Y."Jointsentiment/topicmodelforsentimentanalysis."Proceedingsofthe18thACMconferenceonInformationandknowledgemanagement.ACM,2009.

Lührmann,M., Serra-Garcia,M., and J.Winter (2015). Teaching teenagers in finance: Does itwork?JournalofBanking&Finance,54(C),pp.160-174.

Kauffman,R.G.(1999).IndicatorqualitiesoftheNAPMreportonbusiness.JournalofSupplyChainManagement,35(2),29.

Kämpgen,B.,Weller,T.,O’Riain,S.,Weber,C.,&Harth,A.(2014).Acceptingthexbrlchallengewithlinkeddataforfinancialdataintegration.InTheSemanticWeb:TrendsandChallenges(pp.595-610).SpringerInternationalPublishing.

Keynes,J.M.(1936).Thegeneraltheoryofinterest,employmentandmoney.

Kearney, C., & Liu, S. (2014). Textual sentiment in finance: A survey of methods and models.InternationalReviewofFinancialAnalysis,33,171-185.

Kennedy, Alistair and Diana Inkpen. 2006. Sentiment classification of movie and product reviewsusingcontextualvalenceshifters.ComputationalIntelligence,22(2):110–125.

KlapperL.F.,LusardiA.,andG.A.Panos(2013).FinancialLiteracyanditsconsequences:EvidencefromRussiaduringthefinancialcrisis.JournalofBankingandFinance,37(10),pp.3904-3923

KlapperL.,LusardiA.,andG.Panos(2016a).Financialliteracyandentrepreneurship.GlobalFinancialLiteracyExcellenceCenterworkingpaper2015-3.

Klapper L., Lusardi A., and G. Panos (2016b). Financial literacy and trust in financial institutions.Workingpaper.

Klapper L., Lusardi A. and P. vanOudheusden (2015). Financial literacy around theworld: InsightsfromtheStandard&Poor'sratingsservicesglobalfinancialliteracysurvey.S&PReport.

Klapper L. F., and G. A. Panos (2011). Financial literacy and retirement planning: The Russian case.JournalofPensionEconomicsandFinance,10(4),pp.599-618.

Kling, G., & Gao, L. (2008). Chinese institutional investors’ sentiment. Journal of InternationalFinancialMarkets,InstitutionsandMoney,18(4),pp.374-387.

Koenig,E.F.(2002).Usingthepurchasingmanagers’indextoassesstheeconomy’sstrengthandthelikelydirectionofmonetarypolicy.FederalReserveBankofDallasEconomicandFinancialPolicyReview,1(6),1-14.

Kontopoulos, E., Berberidis, C., Dergiades, T., & Bassiliades, N. (2013). Ontology-based sentimentanalysisoftwitterposts.Expertsystemswithapplications,40(10),4065-4074.



Kurov,A. (2008). Investorsentiment, tradingbehaviorand informationalefficiency in index futuresmarkets.FinancialReview,43(1),pp.107-127.

Kumar, A., & Lee, C. (2006). Retail investor sentiment and return comovements. The Journal ofFinance,61(5),2451-2486.

Lahiri, K., & Moore, G. H. (1992). Leading economic indicators: new approaches and forecastingrecords.CambridgeUniversityPress.

Lahiri, K., & Monokroussos, G. (2013). Nowcasting US GDP: The role of ISM business surveys.InternationalJournalofForecasting,29(4),644-658.

Lee,C.,Shleifer,A.,&Thaler,R.H. (1991). Investorsentimentandtheclosed-endfundpuzzle.TheJournalofFinance,46(1),75-109.

Li, F. (2010). The information content of forward-looking statements in corporate filings—A naïveBayesianmachinelearningapproach.JournalofAccountingResearch,48(5),1049-1102.

Lindsey,M.D.,&Pavur,R.J.(2005).AsthePMIturns:atoolforsupplychainmanagers.JournalofSupplyChainManagement,41(1),30-39.

Liu,B.(2010).SentimentAnalysisandSubjectivity.HandbookofNaturalLanguageProcessing,2,pp.627-666.

Liu, B., & McConnell, J. J. (2013). The role of the media in corporate governance: Do the mediainfluencemanagers'capitalallocationdecisions?.JournalofFinancialEconomics,110(1),1-17.

Loughran,T.,&McDonald,B.(2010).WhenisaLiabilitynotaLiability?TextualAnalysis,Dictionaries,and10-Ks.

Loughran, T., & McDonald, B. (2011b). Barron's Red Flags: Do They Actually Work?. Journal ofBehavioralFinance,12(2),90-97.

Loughran,T.,&McDonald,B.(2013).IPOfirst-dayreturns,offerpricerevisions,volatility,andformS-1language.JournalofFinancialEconomics,109(2),307-326.

Ludvigson, S. C. (2004). Consumer confidence and consumer spending. The Journal of EconomicPerspectives,18(2),29-50.

Lusardi, A. (2008). Overcoming the saving slump: How to increase the effectiveness of financialeducationandsavingprograms.UniversityofChicagoPress:Chicago.

Lusardi, A., Michaud, P. C., & Mitchell, O. S. (2015).Optimal financial knowledge and wealthinequality.JournalofPoliticalEconomy,forthcoming.

Lusardi,A.andO.S.Mitchell(2007).Babyboomerretirementsecurity:Theroleofplanning,financialliteracy,andhousingwealth.JournalofMonetaryEconomics,54,pp.205–224

Lusardi, A., and O. S. Mitchell (2009). How ordinary consumers make complex economic decisions:FinancialliteracyandretirementreadinessNBERWorkingPaper15350.

Lusardi, A., and O. S. Mitchell (2011a). Financial Literacy and Planning: Implications for RetirementWellbeing, forthcoming in A. Lusardi andO.Mitchell (eds.), Financial Literacy: Implications forRetirementSecurityandtheFinancialMarketplace,OxfordUniversityPress.

Lusardi, A., and O. S.Mitchell (2011b). Financial literacy around the world: an overview. Journal ofPensionEconomicsandFinance,10(4),pp.497–508.

Lusardi,Annamaria,andMitchellS.Olivia(2014).“TheEconomicImportanceofFinancialLiteracy:TheoryandEvidence”,JournalofEconomicLiterature.52(1),pp.5-44.

Lusardi,A.,andP.Tufano(2009a).DebtLiteracy,FinancialExperiencesandOverindebtedness.NationalBureauofEconomicResearchWorkingPaper14808.



Lusardi,A.andP.Tufano(2009b).TeachWorkersaboutthePerilsofDebt.HarvardBusinessReview.

Lutz, C. (2015). The impact of conventional and unconventional monetary policy on investorsentiment.JournalofBanking&Finance,61,pp.89-105.

Lutz,C.(2015).TheAsymmetricEffectsofInvestorSentiment.MacroeconomicDynamics,pp.1-27.

Lux, T. (2011). Sentiment dynamics and stock returns: the case of the German stock market.Empiricaleconomics,41(3),663-679.

Mandell, L. (2004). “Financial Literacy: Are We Improving?”, Washington, D.C.: Jump$tartCoalitionforPersonalFinancialLiteracy.

Mandell,L.(2008).“TheFinancialLiteracyofYoungAmericanAdults:Resultsofthe2008NationalJump$tartCoalitionSurveyofHighSchoolSeniorsandCollegeStudents”,Washington,D.C.:JumpstartCoalition.

Mandell,L.,andL.S.Klein(2009).Theimpactoffinancialliteracyeducationonsubsequentfinancialbehavior.JournalofFinancialCounselingandPlanning,20(1),pp.15-24.

Marquering,W.,& Verbeek,M. (2004). The economic value of predicting stock index returns andvolatility.JournalofFinancialandQuantitativeAnalysis,39(2),pp.407-429.

Matsusaka, J. G., & Sbordone, A. M. (1995). Consumer confidence and economic fluctuations.EconomicInquiry,33(2),296-318.

McLean,R.D.,&Zhao,M.(2014).Thebusinesscycle,investorsentiment,andcostlyexternalfinance.TheJournalofFinance,69(3),1377-1409.

Montagnoli A., MoroM., Panos G. A. and R. E. Wright (2016). Financial literacy and attitudes toredistribution.Workingpaper.

Moore, D. (2003). Survey of Financial Literacy inWashington State: Knowledge, Behavior, Attitudes,and Experiences. Technical Report n. 03-39. Social and Economic Sciences Research Center,WashingtonStateUniversity.

Nardo,M.,Petracco-Giudici,M.,&Naltsidis,M.(2015).WalkingDownWallStreetWithATablet:ASurveyofStockMarketPredictionsUsingTheWeb.JournalofEconomicSurveys.

Nassirtoussi,A.K.,Aghabozorgi,S.,Wah,T.Y.,&Ngo,D.C.L.(2015).Textminingofnews-headlinesforFOREXmarketprediction:AMulti-layerDimensionReductionAlgorithmwithsemanticsandsentiment.ExpertSystemswithApplications,42(1),306-324.

Nassirtoussi, A. K., Aghabozorgi, S., Wah, T. Y., & Ngo, D. C. L. (2014). Text mining for marketprediction:Asystematicreview.ExpertSystemswithApplications,41(16),7653-7670.

Neal,R.,&Wheatley,S.M. (1998).Domeasuresof investorsentimentpredict returns?. JournalofFinancialandQuantitativeAnalysis,33(04),523-547.

Nyberg,H. (2013).Predictingbearandbull stockmarketswithdynamicbinary timeseriesModels.JournalofBankingandFinance,37,pp.3351-3363.

Oh, C., & Sheng, O. (2011). Investigating predictive power of stock micro blog sentiment inforecastingfuturestockpricedirectionalmovement.

OrganisationforEconomicCooperationandDevelopment-OECD(2015).2015OECD/INFEtoolkitformeasuringfinancialliteracyandfinancialinclusion.Report.

O'Riain, S., Curry, E., and A. Harth (2012). XBRL and open data for global financial ecosystems: Alinkeddataapproach. International JournalofAccounting InformationSystems,13(2),pp.141-162.



Pang, B., and L. Lee (2008). Opinion Mining and Sentiment Analysis. Foundations and Trends inInformationRetrieval,2(1–2),pp.1–135.

Pang,B., L. Lee,andS.Vaithyanathan (2002). ThumbsUp?SentimentClassificationUsingMachineLearningTechniques.Proc.Conf.EmpiricalMethodsinNaturalLanguageProcessing,pp.79–86.

Panos G. A., and R. EWright (2016a). Financial literacy and attitudes to the Scottish referendum.Workingpaper.

Panos G. A., and R. E.Wright (2016b). Financial literacy and attitudes to the EU-exit referendum.Workingpaper.

PanosG.A.andR.EWright(2016c).Financialliteracyandattitudestoimmigration.Workingpaper.

Paye,B.S. andA. Timmermann (2006). Instability of returnpredictionmodels. Journal of EmpiricalFinance,13(3),pp.274-315.

Petkova,R. (2006).Do theFama–French factorsproxy for innovations inpredictivevariables?.TheJournalofFinance,61(2),pp.581-612.

Prabowo, R., and M. Thelwall (2009). Sentiment analysis: A combined approach. Journal ofInformetrics,3(2),pp.143-157.

Qiu,L.X.,&Welch,I.,(2006).InvestorSentimentMeasures.WorkingPaperBrownUniversityandNBER.

Radzimski, M., Sánchez-Cervantes, J. L., Rodríguez-González, A., Gómez-Berbís, J. M., & García-Crespo, Á. (2012). FLORA–Publishing Unstructured Financial Information in the Linked OpenData Cloud. In First InternationalWorkshop on Finance and Economics on the SemanticWeb(FEOSW).

Rossi,B.,2013.ExchangeRatePredictability.JournalofEconomicLiterature,51(4),pp.1063-1119.

Salvetti,Franco,ChristophReichenbach,andStephenLewis.2006.Opinionpolarityidentificationofmovie reviews. In Janyce Wiebe, editor, Computing Attitude and Affect in Text: Theory andApplications.Springer,Dordrecht,pages303–316.

Schmeling,M.(2009).Investorsentimentandstockreturns:Someinternationalevidence.JournalofEmpiricalFinance,16(3),pp.394-408.

Schmeling,M.(2009).Investorsentimentandstockreturns:Someinternationalevidence.JournalofEmpiricalFinance,16(3),394-408.

Shiller,R.J.(2015).Irrationalexuberance.Princetonuniversitypress.

Singer,N.,Laser,S.,&Dreher,F.(2013).Publishedstockrecommendationsasinvestorsentimentinthenear-termstockmarket.EmpiricalEconomics,45(3),1233-1249.

Smith,A.(1776).TheWealthofNations.Peteresfield,UK:HarrimanHouseLtd.,2007Edition.

Spies, M. (2010). An ontology modelling perspective on business reporting. Information Systems,35(4),pp.404-416.

Stango,V.andJ.Zinman(2009).Exponentialgrowthbiasandhouseholdfinance.JournalofFinance64,pp.2807–2849.

Stock, J.H. and M.W. Watson (2003). Forecasting Output and Inflation: The Role of Asset Prices.JournalofEconomicLiterature,41(3),pp.788–829.

Stone,P.J.(1966).TheGeneralInquirer:AComputerApproachtoContentAnalysis.TheMITPress.

Stone, P. J., Dunphy,D. C.,& Smith,M. S. (1966). TheGeneral Inquirer: A Computer Approach toContentAnalysis.



Taboada, Maite, Caroline Anthony, and Kimberly Voll. 2006. Creating semantic orientationdictionaries. In Proceedings of 5th International Conference on Language Resources andEvaluation(LREC),pages427–432,Genoa.

Tetlock,P.C.(2007).Givingcontenttoinvestorsentiment:Theroleofmediainthestockmarket.TheJournalofFinance,62(3),1139-1168.

Tetlock,P.C., Saar-Tsechansky,&Macskassy,S. (2008).More thanwords:Quantifying language tomeasurefirms'fundamentals.TheJournalofFinance,63(3),1437-1467.

Timmermann,A.(2006).Forecastcombinations.HandbookofEconomicForecasting,1,pp.135-196.

Tong, Richard M. 2001. An operational system for detecting and tracking opinions in on-linediscussions. In Working Notes of the ACM SIGIR 2001 Workshop on Operational TextClassification,pages1–6,NewYork,NY.

Tufano,P.,Flacke,T.,andN.W.Maynard(2012).Raisingawarenessofgovernmentinitiativesandtaxtimesavingsopportunities.FinancialLiteracyCenterworkingpaper925-5SA.

Turney, Peter. 2002. Thumbs up or thumbs down? Semantic orientation applied to unsupervisedclassificationof reviews. InProceedingsof40thMeetingof theAssociation forComputationalLinguistics,pages417–424,Philadelphia,PA.

Turney, Peter and Michael Littman. 2003. Measuring praise and criticism: Inference of semanticorientationfromassociation.ACMTransactionsonInformationSystems,21(4):315–346.

Vafopoulos,M.N., Vafeiadis,G., Razis,G., Anagnostopoulos, I.,Negkas,D., and L.Galanos (2016).LinkedOpenEconomy:TakeFullAdvantageofEconomicData.

Van Rooij, M., A. Lusardi, and R. Alessie (2011). Financial literacy and stock market participation.JournalofFinancialEconomics,101(2),pp.449-472.

WalstadW. B., Rebeck, K., and E. A.MacDonald (2010). The effects of financial education on thefinancialknowledgeofschoolstudents.JournalofConsumerAffairs,44(2),pp.336-357.

Wilson, T., Wiebe, J., and P. Hoffmann (2005). Recognizing Contextual Polarity in Phrase-LevelSentimentAnalysis.pp.347-354.

Xia, R., C. Zong, and S. Li (2011). “Ensemble of Feature Sets and Classification Algorithms forSentimentClassification”.InformationSciences,181(6),pp.138–1152.

Yoong,J.(2011).Financialilliteracyandstockmarketparticipation:EvidencefromtheRANDAmericanLife Panel. In Olivia S. Mitchell and Annamaria Lusardi. Financial Literacy: Implications forRetirementSecurityandtheFinancialMarketplace.Forthcoming,OxfordUniversityPress.

Zhang, C. (2008). Defining, modeling, and measuring investor sentiment. University of California,Berkeley,DepartmentofEconomics.

Zhou,L.,andP.Chaovalit(2008).Ontology-SupportedPolarityMining.59(1),pp.98-110.

Zweig,M. E. (1973). An investor expectations stock price predictivemodel using closed-end fundpremiums.TheJournalofFinance,28(1),67-78.

Analysis of the Best Current Practices · 0.6 09-05-2016 A. Chamula (EEA) Added sections...

Documents

Transcript of Analysis of the Best Current Practices · 0.6 09-05-2016 A. Chamula (EEA) Added sections...