Marco Corazza Florence Legros Cira Perna Marilena Sibillo Editors Mathematical and Statistical Methods for Actuarial Sciences and Finance Mathematical and Statistical Methods for Actuarial Sciences and Finance Marco Corazza • Florence Legros (cid:129) Cira Perna (cid:129) Marilena Sibillo Editors Mathematical and Statistical Methods for Actuarial Sciences and Finance MAF 2016 123 Editors MarcoCorazza FlorenceLegros DepartmentofEconomics DépartementFinance,Audit,Comptabilité Ca’FoscariUniversityofVenice etContrôle Venezia,Italy ICNBusinessSchool Nancy,France CiraPerna MarilenaSibillo DepartmentofEconomicsandStatistics DepartmentofEconomicsandStatistics UniversityofSalerno UniversityofSalerno Fisciano(SA),Italy Fisciano(SA),Italy ISBN978-3-319-50233-5 ISBN978-3-319-50234-2 (eBook) https://doi.org/10.1007/978-3-319-50234-2 LibraryofCongressControlNumber:2017962872 ©SpringerInternationalPublishingAG2017 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof thematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation, broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformation storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodology nowknownorhereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthisbook arebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsor theeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinorforany errorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictional claimsinpublishedmapsandinstitutionalaffiliations. Printedonacid-freepaper ThisSpringerimprintispublishedbySpringerNature TheregisteredcompanyisSpringerInternationalPublishingAG Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland Preface This volume is a collection of referred papers from the several tens that were presentedattheinternationalconferenceMAF2016—MathematicalandStatistical MethodsforActuarialSciencesandFinance. The conferencewas heldin Paris (France),fromMarch 30to April1, 2016,at the Université Paris-Dauphine.It was organizedby the Executive MBA CHEA— CentredesHautesEtudesd’AssurancesandbytheDepartmentofEconomicsofthe Ca’ FoscariUniversityofVenice (Italy),with the collaborationof the Department ofEconomicsandStatisticsoftheUniversityofSalerno(Italy). Theconferencewastheseventhofaninternationalbiennialserieswhichbeganin 2004.Itwasbornoutoftheideathatenhancedcooperationbetweenmathematicians andstatisticiansworkinginactuarialsciences,ininsurance,andinfinancecanresult inimprovedresearchinthesefields. Thewideparticipationatalltheconferencesintheseriesconstitutesproofofthe meritsofthisidea. The papers published in this volume present theoretical and methodological contributionsandtheirapplicationstorealcontexts. Of course, the success of MAF 2016 would not have been possible without the valuable help of all members of the Scientific Committee, the Organizing Committee,andtheLocalCommittee. Finally, we are pleased to inform readers that the organizing machine for the nexteditionisalreadyinaction:theconferenceMAF2018willbeheldinMadrid (Spain),fromApril4to6,2018(fordetailsvisitthewebsitehttp://www.est-econ. uc3m.es/maf2018/). Welookforwardtoseeingyouthere. Venezia,Italy MarcoCorazza Nancy,France FlorenceLegros Fisciano(SA),Italy CiraPerna Fisciano(SA),Italy MarilenaSibillo August2017 v Contents TheEffectsofCreditRatingAnnouncementsonBondLiquidity: AnEventStudy................................................................... 1 PilarAbad,AntonioDiaz,AnaEscribano,andM.DoloresRobles TheEffectofCreditRatingEventsontheEmergingCDSMarket ........ 17 LauraBallesterandAnaGonzález-Urteaga A Generalised Linear Model Approachto Predict the Result ofResearchEvaluation.......................................................... 29 AntonellaBassoandGiacomodiTollo ProjectingDynamicLifeTablesUsingDataCloning ........................ 43 Andrés Benchimol, Irene Albarrán, Juan Miguel Marín, andPabloAlonso-González MarkovSwitchingGARCHModels:Filtering,Approximations andDuality........................................................................ 59 MonicaBillioandMaddalenaCavicchioli ANetworkApproachtoRiskTheoryandPortfolioSelection .............. 73 RoyCerquetiandClaudioLupi AnEvolutionaryApproachtoImproveaSimpleTradingSystem ......... 83 MarcoCorazza,FrancescaParpinel,andClaudioPizzi Provisions for Outstanding Claims with Distance-Based GeneralizedLinearModels ..................................................... 97 TeresaCostaandEvaBoj Profitabilityvs. AttractivenessWithina PerformanceAnalysis ofaLifeAnnuityBusiness ...................................................... 109 EmiliaDiLorenzo,AlbinaOrlando,andMarilenaSibillo vii viii Contents Uncertainty in Historical Value-at-Risk: An Alternative Quantile-BasedRiskMeasure.................................................. 119 DominiqueGuégan,BertrandHassani,andKehanLi ModelingVarianceRiskPremium ............................................. 129 KossiGnameho,JuhoKanniainen,andYeYue CoveredCallWritingandFraming:ACumulativeProspectTheory Approach.......................................................................... 143 MartinaNardonandPaoloPianca OptimalPortfolioSelection for an Investor with Asymmetric AttitudetoGainsandLosses ................................................... 157 SergeiSidorov,AndrewKhomchenko,andSergeiMironov The Effects of Credit Rating Announcements on Bond Liquidity: An Event Study PilarAbad,AntonioDiaz,AnaEscribano,andM.DoloresRobles Abstract ThispaperinvestigatesliquidityshocksontheUScorporatebondmarket aroundcreditrating changeannouncements.These shocksmay be inducedby the informationcontentof the announcementitself, andabnormaltradingactivitycan be triggered by the release of information after any upgrade or downgrade. Our findingsshowthat:(1)themarketanticipatesratingchanges,sincetrendsliquidity proxiespreludethe event,andadditionally,largevolumetransactionsare detected thedaybeforethedowngrade;(2)theconcretematerializationoftheannouncement is not fully anticipated, since we only observe price overreaction immediately after downgrades; (3) a clear asymmetric reaction to positive and negative rating events is observed; (4) different agency-specific and rating-specific features are abletoexplainliquiditybehavioraroundratingevents;(5)financialdistressperiods exacerbateliquidityresponsesderivedfromdowngradesandupgrades. 1 Introduction Informationonratingactionshasbeenapermanentsubjectofdebate.Creditrating agencies (CRAs) state that they consider insider information when assigning and revising ratings, without disclosing specific details to the public at large. The literature examines prices and/or returns responses to rating events. However, the informationaboutthecreditworthinessofissuersdisclosedbyratingactionscannot onlyaffectprices.Besidesthis,itcaninducespecificmarketdynamicsconcerning the liquidity of the re-rated bonds. One important role of ratings is to reduce the informationasymmetry between lenders and borrowers.As this asymmetry is P.Abad UniversidadReyJuanCarlosdeMadrid,Madrid,Spain e-mail:[email protected] A.Diaz(cid:129)A.Escribano((cid:2)) UniversidaddeCastilla-LaMancha,Albacete,Spain e-mail:[email protected];[email protected] M.D.Robles UniversidadComplutensedeMadrid,Madrid,Spain e-mail:[email protected] ©SpringerInternationalPublishingAG2017 1 M.Corazzaetal.(eds.),MathematicalandStatisticalMethodsforActuarial SciencesandFinance,https://doi.org/10.1007/978-3-319-50234-2_1 2 P.Abadetal. inverselyrelatedtoliquidity,ifcreditratingchanges(CRCs) releasespecificnews aboutthefinancialsituationoffirms,theywillaffectfirms’bondliquidity. In order to analyze this question, we go beyond the traditional price analysis by analyzing corporate bond liquidity patterns around CRC announcements. We examine different dimensions of corporate bond liquidity, and compute different adaptationsoftraditionalmicrostructure-basedliquidmeasuresonstockmarketsto bondmarkets,aswellasothertraditionalbondmarketliquiditymeasures.1 Our paper relates to several strands of the literature. First, we contribute to researchthatseekstobetterunderstandtheliquidityofcorporatebondmarkets(e.g. [1, 8] and Chen et al. [3]). Second, to the literature that studies the information contentofratingannouncementsandtheirimpactonbondmarket(e.g.Steinerand Heinke[18]andMay[15])andonstockmarkets(e.g.NordenandWeber[16]). Westudyacomprehensivesampleof2727CRCsinthewholeUScorporatebond market,usingTRACEtransactiondatafrom2002to2010.Wealsostudytheimpact oftherecentglobalfinancialcrisisontheresponseofthedifferentliquidityaspects toCRCannouncements.WeconsiderthedefaultofLehmanBrothersinmid-2008 tobethestartingpointofthefinancialturmoil. Our results indicate three clear patterns in liquidity and prices, depending on the time period around the announcement when we consider the whole sample period. First, we observe trends in prices and liquidity deterioration before the announcement. Additionally, nervousness emerges in the market the day before downgrades.Second, there is price pressure for a few days after the downgrades. This fact could imply transaction prices below fundamental values. Third, we observe that prices converge to the correct value and the level of trading activity clearlyrisesduringthe secondfortnight.Inthe case ofupgrades,thereisno price impact. Asidefromanalyzingimpactsonliquidityderivedfromcreditratingmigrations, we examine the determinants of abnormal liquidity observed before and after the announcements. To find the drivers of abnormal liquidity, we carry out a cross- sectional analysis including as key factors different characteristics of the rating event,theissue,andtheissuer,toexplainliquidityresponsestoratingchanges.Our premiseisthatratingchangesthatprovidemorerelevantinformationtothemarket mustcausestrongerimpactsonliquidity. Our results should enable market participants to manage portfolios, given that theyneedtohaveanunderstandingofthewayinwhichtheliquidityandtheliquidity premiumonpricesbehavearoundCRC. Theremainderofthepaperisarrangedasfollows:Sect.2explainsthehypotheses to be tested. Section 3 presents the data description. Section 4 examines different measuresofabnormalliquidity.ThemainresultsarepresentedinSect.5.Section6 includesthecross-sectionalanalysis.Finally,Sect.7concludes. 1Recentpapersusingsomeofthesemeasurescorroboratetheliquidityeffectsonprices(see,e.g., Baoetal.[1],Dick-Nielsenetal.[8]andFriewaldetal.[10]). TheEffectsofCreditRatingAnnouncementsonBondLiquidity:AnEventStudy 3 2 TheExpected ResponseofLiquidityto RatingActions We consider that the effects of rating changes can be explained by different possible hypotheses. The main hypothesis states that CRAs are supplied with considerablenon-publicinformationaboutfirms,suchasinformationaboutthetotal firmvalueanditsorganizationaleffectiveness.Intheliquidityliterature,themarket microstructure models indicate that trading activity responses to news releases are related to the existence of asymmetric information among informed traders, uninformedtraders,andmarket-makers.KimandVerrecchia[14]statethatthefact thatsometradersareabletomakebetterdecisionsthanothers,basedonthesame information,leadstoinformationasymmetryandpositiveabnormaltradingvolume, despiteareductioninliquidityafterthereleaseofnewinformationaboutthefirm. A rating revision may provide additional information about the firm. Different investors’riskperceptioncaninduceportfoliorebalancingprocesses.Inthiscontext, highertradingactivityafterCRCswillbeexpected. The second theory we analyze is the reputation hypothesis. This hypothesis (Holthausen and Leftwich [11]) states that rating agencies face asymmetric loss functions, and that they allocate more resources to revealing negative credit informationthanpositiveinformation,becausethelossofreputationismoresevere whena false ratingis too highthan whenit is too low.Reputationcosts create an incentive for CRAs to truthfully revealthe investmentquality, since investorscan eventuallylearnandpunishtheagency. Thelasthypothesispointsoutthattheusualliquiditypremiumonpriceswidens afteraCRC.Weexaminewhethertheliquidityimpactonpricesisexacerbatedafter a CRC. If rating announcementsdisclose new and relevant information about the defaultriskofabond,thenpricesshouldimmediatelyincorporatethisinformation. Independently of price adjustment, liquidity may drive additional price changes. The traditional literature considers liquidity to be a key component of corporate bond prices. Recent papers corroboratethis result. For instance, [1] conclude that illiquidity explains a substantial part of the yield spreads of high-rated bonds, overshadowing the credit risk component. Chen et al. [3] and Friewald et al. [10] also observe that the economic impact of liquidity is significantly larger for speculative-gradebonds. 3 Data Description We use two main sources of data in our analysis: the NASD’s Trade Reporting and Compliance Engine (TRACE) transactions data for corporate bonds and the Mergent Fixed Income Securities Database (FISD), with complete information