<%BANNER%>

Consumption and Investment Decision

Permanent Link: http://ufdc.ufl.edu/UFE0024971/00001

Material Information

Title: Consumption and Investment Decision An Analysis of Aggregate and Time-Additive Models
Physical Description: 1 online resource (110 p.)
Language: english
Creator: Fu, Liang
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: consumption, investment
Accounting -- Dissertations, Academic -- UF
Genre: Business Administration thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: In the presence of an investment opportunity, a borrower decides whether to invest. We find that - depending on the borrower's personal wealth and the presence of limited liability - the investment is not always undertaken, which leads to investment distortion. Furthermore, when the borrower has intertemporal consumption preference rather than aggregate consumption preference, the potential investment distortion problem is weakly exacerbated. The change of the borrower's risk preference from constant absolute risk aversion to decreasing absolute risk aversion lowers the occurrence of the potential investment distortion as his personal wealth increases. Furthermore, when the borrower is privately informed about his personal wealth, the Revelation Principle applies and the naive contract which is derived from public information case is optimal.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Liang Fu.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Demski, Joel S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024971:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024971/00001

Material Information

Title: Consumption and Investment Decision An Analysis of Aggregate and Time-Additive Models
Physical Description: 1 online resource (110 p.)
Language: english
Creator: Fu, Liang
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: consumption, investment
Accounting -- Dissertations, Academic -- UF
Genre: Business Administration thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: In the presence of an investment opportunity, a borrower decides whether to invest. We find that - depending on the borrower's personal wealth and the presence of limited liability - the investment is not always undertaken, which leads to investment distortion. Furthermore, when the borrower has intertemporal consumption preference rather than aggregate consumption preference, the potential investment distortion problem is weakly exacerbated. The change of the borrower's risk preference from constant absolute risk aversion to decreasing absolute risk aversion lowers the occurrence of the potential investment distortion as his personal wealth increases. Furthermore, when the borrower is privately informed about his personal wealth, the Revelation Principle applies and the naive contract which is derived from public information case is optimal.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Liang Fu.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Demski, Joel S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024971:00001


This item has the following downloads:


Full Text

PAGE 1

1

PAGE 2

2

PAGE 3

3

PAGE 4

Iamdeeplyindebtedtomyadvisor,ProfessorJoelS.Demski.ItishisconsistentsupportandguidancethatensuredthesuccessfulcompletionofmyPh.D.program.Hispassioninscholarshiphasgreatlyshapedmyphilosophyonacademiccareer.IwishtothankProfessorHaijinLin,ProfessorGaryMcGill,andProfessorDavidSappingtonfortheirvaluablesuggestionsthatsubstantiallyimprovedthisworkandtheirconstantencouragements.Finally,Iamtrulythankfulformyhusband,JingLiu,forhispatienceandtrust. 4

PAGE 5

page ACKNOWLEDGMENTS ................................. 4 LISTOFTABLES ..................................... 6 ABSTRACT ........................................ 7 CHAPTER 1INTRODUCTION .................................. 8 2LITERATUREREVIEW .............................. 14 3CARA-BORROWER ................................. 20 3.1Introduction ................................... 20 3.2ConsumptionPreferenceCase1:Aggregate ................. 22 3.2.1TheModel ................................ 22 3.2.2InvestmentBehaviors:Borrow,Self-Invest,No-Invest ........ 22 3.2.2.1Benchmark:rst-bestcontract ............... 23 3.2.2.2Second-bestenvironment ................... 26 3.2.2.3Investmentdistortion ..................... 35 3.3ConsumptionPreferenceCase2:Time-Additive ............... 41 3.3.1NoInvestmentDistortion ........................ 42 3.3.2Second-BestContractandPotentialInvestmentDistortion ..... 44 3.3.3WIInvestmentandPotentialInvestmentDistortion ....... 53 3.4InvestmentDistortionandConsumptionSmoothing ............. 58 4DARA-BORROWER ................................. 60 4.1First-BestContractandNo-LenderInvestment ................ 62 4.2Second-BestLoanAgreement ......................... 67 4.2.1AggregateConsumptionPreference .................. 67 4.2.2Time-AdditiveConsumptionPreference ................ 70 4.3ConsumptionPreferenceandInvestmentDistortion ............. 74 5PRIVATEPERSONALWEALTH ......................... 77 6CONCLUSION .................................... 85 APPENDIX:PROOFS ................................... 87 REFERENCES ....................................... 108 BIOGRAPHICALSKETCH ................................ 110 5

PAGE 6

Table page 3-1Second-BestContractsfortheBorrowerwithAggregateConsumption ...... 34 3-2InvestmentDistortionfortheBorrowerwithAggregateConsumption ...... 40 3-3Second-BestContractsfortheBorrowerwithTime-AdditiveConsumption ... 51 3-4InvestmentDistortionfortheBorrowerwithTime-AdditiveConsumption .... 57 4-1InvestmentDistortionforAggregateConsumption ................. 70 4-2LocalCertaintyEquivalentConsumptionsandLocalRiskPremium ....... 73 4-3InvestmentDistortionforTime-AdditiveConsumption .............. 74 4-4Second-BestContractsInvestmentDistortionfortheDARA-Borrower ...... 76 6

PAGE 7

Inthepresenceofaninvestmentopportunity,aborrowerdecideswhethertoinvest.Wendthat-dependingontheborrower'spersonalwealthandthepresenceoflimitedliability-theinvestmentisnotalwaysundertaken,whichleadstoinvestmentdistortion.Furthermore,whentheborrowerhasintertemporalconsumptionpreferenceratherthanaggregateconsumptionpreference,thepotentialinvestmentdistortionproblemisweaklyexacerbated.Thechangeoftheborrower'sriskpreferencefromconstantabsoluteriskaversiontodecreasingabsoluteriskaversionlowerstheoccurrenceofthepotentialinvestmentdistortionashispersonalwealthincreases.Furthermore,whentheborrowerisprivatelyinformedabouthispersonalwealth,theRevelationPrincipleappliesandthenaivecontractwhichisderivedfrompublicinformationcaseisoptimal. 7

PAGE 8

Thisdissertationexaminesconsumptionandinvestmentbehaviorassociatedwithaninvestmentproject.Twopartiesareinvolved:arisk-averseborrowerandarisk-neutrallender.Thedissertationstartswithaborrowerwhodisplaysconstantab-soluteriskaversion(CARAhereafter)withcertaininitialwealthandeitheraggregateconsumptionpreferenceortime-additiveconsumptionpreferenceasabenchmark.Itfocusesonaninvestmentdecisioninthepresenceoflimitedliabilityandtheresultingpotentialinvestmentdistortionaswellasitsinteractionwiththeborrower'sconsumptionpreference. Bydesign,theinvestmentprojectisalwaystakenwhentheborrower'seortispubliclyobservableandtheassociatedrst-bestcontractalwaysexists.However,whentheborrower'seortchoicebecomesprivate,hedoesnotalwaysundertaketheproject,whichleadstoinvestmentdistortion.Inthisdissertationinvestmentdistortionistheborrower'spassinguptheprotableinvestmentopportunity.Basedonthisconditionanddependingontheborrower'spersonalwealth,investmentdistortionoccursifanyofthefollowingsituationsarise: A. Thelackofaloanagreementbetweentheborrowerandthelenderwhentheborroweriswealth-constrained. B. Thewealth-constrainedborrowerprefersthestatusquotothebestavailableloanagreement. C. Thewealth-abundantborrowerprefersthestatusquoto(a)borrowingfromthelender(iffeasible)and(b)self-nancing. InSituationAandSituationB,thewealth-constrainedborrowerreferstotheborrowerwhodoesnothavesucientpersonalwealthtonancetheinvestment.SituationAimpliesthatthereisnoloanarrangementthatmotivatestheborrowerandletsthelenderrecoupherinvestmentatthesametime.Theculpritforthenon-existenceofsuchanagreementisthejointforcefromtwofactorsinthedesign.Theborroweristheparty 8

PAGE 9

SituationBsuggeststhatfortheborrowertodecidebetweenborrowingfromthelender(giventhatsuchanarrangementisinplace)andstayingwiththestatusquo,heconsidershisnetpayofromtheinvestment.Inparticular,theborrowerreceivestheentireexpectedNPVfromtheinvestmentbutheincursapersonalcostofeort;therefore,ifhisnetpayoisnotsucientlylarge,hemightendupgivinguptheinvestmentanddecidetosettlewiththestatusquo.Theeconomicimplicationbehindthisformofinvestmentdistortionisthattheloanagreement,givenitsexistence,istooriskyfortheborrowertoengageinnancingwiththelender. InSituationC,thewealth-abundantborrowerreferstotheborrowerwhohassucientpersonalendowmenttonancetheinvestmentprojectonhisownwithoutresortingtothelender.Iftheloanagreementfails,theborrowerisleftwiththechoicebetweenno-lenderinvestmentandno-investment.Sincetheborrowerisriskaverse,theinvestment 9

PAGE 10

ThisdissertationanalyzesSituationsAthroughCfortheborrowerwitheitheraggregateconsumptionpreferenceortime-additiveconsumptionpreference.Thediscussionfortheaggregateborrowerprovidesabenchmarkfortheinterplaybetweentheborrower'sinvestmentandconsumptionbehaviorbyderivingconditionsunderwhichinvestmentdistortionarises.Thebenchmarkisimportantinthefollowingtworespects:(a)theroleoftheborrower'slimitedliabilityconstraintsonhisconsumptionismadeexplicitindeterminingthefeasibilityoftheloanarrangementbetweentheborrowerandthelenderwhentheformerhasprivateinformationabouthiseortsupplyand(b)theroleoftheborrower'spersonalwealthisnontrivialinpotentialinvestmentdistortion. Withthebenchmarkfortheborrowerwithaggregateconsumptionpreference,theanalysisrevealsthattheborrower'sconsumptionpreferencehasaneectonhisinvestmentdecision,especiallyontheoccurrenceofthepotentialinvestmentdistortion.Theinvestmentdistortionunderaggregateconsumptionpreferenceimpliesthedistortionundertime-additivepreference,butnotviceversa.Inparticular,iftheborrower'spersonalwealthisstrictlylessthantheinvestmentrequirement,theborrower'spreferenceshavenoinuenceonhisinvestmentbehavior.Intuitively,theborrowerdisplaysCARA,whichwouldimplythathispersonalwealthhasnoeectonhisattitudetowardsriskateachinstantiftheconsumptionproblemweretoextendtomultiperiods.Therefore,conditionsunderwhichinvestmentdistortionarisesareexactlythesamefortheborrowerwithaggregateandtime-additiveconsumptionpreferences.However,iftheborrowerhassucientlylargepersonalwealthtoengageinself-nancing,theconcernforintertemporalconsumptionexacerbatesthepotentialinvestmentdistortionproblem.Theaggravationispronouncedfortheborrower'sno-lenderinvestment.Thereasonisthatwithno-lenderinvestment,themagnitudeoftheborrower'spersonalwealthisimportantinshapingthe 10

PAGE 11

WiththeCARA-borrowerasabenchmark,thesubsequentanalysisintroducesaborrowerwithdecreasingabsoluteriskaversion(DARAhereafter)witheitheraggregateortime-additiveconsumptionpreferencetoinvestigatethechangeoftheborrower'sriskpreferenceonhisconsumptionandinvestmentbehavior.Ingeneral,fortheDARA-borrower,theuniquedistinctionbetweentheCARA-andDARA-borroweristhat,ceterisparibus,thelatterislesslikelytoencounterinvestmentdistortionproblemsassociatedwiththesecond-bestloancontract.TheattributeofthemitigatedpotentialinvestmentdistortionisthattheDARA-borrowerchangeshisattitudetowardsriskashispersonalwealthchanges.Inparticular,astheDARA-borrowerbecomeswealthier,hebecomeslessriskaverse,whichinturnleadshimtoundertaketheinvestmentprojectthatwouldotherwiseberejectedbytheCARA-borrower. Althoughtheborrower'shierarchicaloptimalinvestmentdecisionisinconclusivebetweenborrowingandself-nancing,theloanagreementwiththerisk-neutrallenderprovidesrisksharing,whichisimportanttotherisk-averseborrower.Therefore,weretheborrower'spersonalwealthtobeprivateinformation,howtheloancontractwould 11

PAGE 12

Investmentactivitiesarevitaltoaprosperousandwell-functioningeconomy;however,weoftenseesoundinvestmentopportunitiesbeingpassedup.Thisdissertationdesignsasimpletractabletwo-partymodeltoinvestigatethepotentialattributiontopotentialinvestmentdistortion.Thendingsarespecicallyvaluabletotheentrepreneur-typeofborrower,whomaybewealth-constrainedorrestrictedbyriskpreferencebutwhoseekswaystoengageintheproject.Theborrowerwithaggregateconsumptionpreferencecanbeconsideredtorepresentatraditionalentrepreneur,whosekeyfocusistheaggregatereturnofaninvestmentratherthantheintertemporalgain.Ontheotherhand,theborrowerwithtime-additiveconsumptionpreferenceismorelikelytorepresentsmall 12

PAGE 13

13

PAGE 14

Thestylizedmodelinthisdissertationpartlyresemblestheconventionalagencyproblem(e.g., GrossmanandHart ( 1983 ))inthatthelender-borrowerrelationshipandthecorrespondingoptimalconsumptionsarecounterpartsofthelabormarketprincipal-agentwagecontract.Sincetheborrowerhasthepotentialtonancetheinvestmentfromthelender,theloancontractbetweentheborrowerandthelenderfeaturesmoralhazardinthedebtmarket.Forexample, GaleandHellwig ( 1985 )focusonacontractbetweeninvestorsandentrepreneurs,bothofwhomoperateinacompetitivenancialmarket.Entrepreneursarejustborrowerswhowishtoundertakeriskyventuresbutlackthenecessaryresourcessotheyturntotheinvestorsforexternalnance.Investorsarebanksorothernancialinstitutions.Theyndthattheoptimal,incentive-compatibledebtcontractinamodelofborrowingandlendingwithasymmetricinformationisthestandarddebtcontract.Inthisdissertation,engaginginborrowingisendogenous,especiallyforthewealth-constrainedborrower;thelenderhasrationalexpectationsoftheborrower'sincentivestoshirk(givenhisprivateeortsupply)andmisreport(givenhisprivateinformationaboutwealth);theinteractionbetweentheborrowerandthelenderunavoidablyraisespotentialinvestmentdistortion,inecientrisksharing,andtheborrower'sconcurrentconcernforintertemporalconsumptionwhenhehastime-additiveconsumptionpreference. ThebenchmarkconsumptionandinvestmentbehaviorsfortheCARA-borrowerinthisdissertationarecloselyrelatedto Tirole ( 2006 )'sbasicmodel,wherehemodelsarisk-neutrallenderandarisk-neutralborrowerandinvestigatescreditrationinginthepresenceofawell-functioningnancialmarket.Tirolendsthatthedrivingforceofpotentialcreditrationingwhentheborrowerhasprivateinformationabouthisactionchoiceisthelevelofhiswealthathand.Thisdissertationextends Tirole ( 2006 )byintroducingtherisk-sharingaspectoftheloanagreement.Therisk-averseborrower, 14

PAGE 15

Tirole ( 2006 )byconsideringtheroletheborrower'sconsumptionpreferenceplaysinpotentialinvestmentdistortion.Thendingsofthisdissertationaremorecompellingintwoways.First,itisperhapsmorerealistictoassumearisk-averseborrowerandnon-negativeconsumptionwhentheproblemathandisinvestmentandconsumptionbehavior.Second,sinceconsumptionbehaviorisnotindependentoftheinvestmentdecision,howtheformeraectsthelatterisitselfofparticularinterest. Thisdissertationfocusesontheunder-investmentproblemthatariseswhenaprotableprojectisnotundertaken.Thenanceliteratureexplainsunder-investmentfromacapitalstructureperspective,whichimpliesthatunder-investmentisaresultofinformationasymmetrybetweentheborrowerandthelender(e.g., MyersandMajluf ( 1984 ))andtheprojectisnotundertakenduetocostlyexternalnancing.Thisdissertationisdistinguishedfromthenanceliteraturebyfocusingonthemoralhazardaspectratherthantheadverseselectionaspectoftheinformationproblem.Inparticular,theborrower'smoral-hazard-associatedincentiveproblem,togetherwiththe ( 2006 )alsoassumesthenon-negativityoftheborrower'sincome.Thelimitedliabilitycontractintheliteraturetakesdierentforms. Sappington ( 1983 )showsthattheprincipaldeliberatelyinducesless-than-ecientoutputinastatethatisnotthemostlyproductivewhencompelledtorespecttheagent'sexpostlimitedliability. LewisandSappington ( 2000 )derivetheoptimalcontractwhentheagenthasprivateinformationabouthiswealth. CheandGale ( 2000 )and LewisandSappington ( 2001 )analyzemodelswheretheagentisprivatelyinformedbothofhisabilityandofhiswealth.Amorerecentpaperby Lu ( 2009 )ndsthatrmsexperienceoverinvestmentratherthanunderinvestmentwhenbankruptcyrisk(calculatedasAltman'sZ-score, Altman ( 1968 ))ishigher. 15

PAGE 16

Theaccountingliteraturefocusesontheuseandthe"quality"ofaccountinginformationinmitigatinginvestmentinecienciesdirectly.Forexample, BensandMonahan ( 2004 )and Bushmanetal. ( 2006 )studyhowarm'sdisclosurepolicyandqualitymitigatetheunder-investmentproblem. Biddleetal. ( 2008 )and Lu ( 2009 )exploretheeectsofaccrualqualityandthedisclosureofnon-nancial-statementinformationonover-investment,respectively.Anotherstreamofaccountingliteratureinvestigatestheconnectionbetweenaccountinginformationandinvestmentbehaviorindirectlyfromtheconnectionbetweenaccountingandthecostofcapitalingeneral. Francisetal. ( 2000 )and Aboodyetal. ( 2005 )emphasizetheroleofaccrual(ormoregenerally,earnings)qualityonthecostofcapital. Bertomeuetal. ( 2008 )furtherendogenouslyconnectdisclosurepolicytoitscapitalstructure.However,theroleoftheborrowerassociatedwiththeinvestmentprojectislargelyoverlookedintheaccountingliterature.Asaneconomicagent,theborrowerisimportantintheinvestmentdecisionandthecorrespondingnancingarrangement.Inparticular,hisunobservableeortchoiceandhispersonalwealthendowmentbothsignicantlyaecttheexistenceoftheloanagreement,whichinturn,inuencestheoccurrenceofinvestmentineciency. Investmentdecisionandconsumptionbehaviorarehardlytwoisolatedevents. Breeden ( 1979 )isthepioneerofcontemporaryresearchonconsumptionandnancialmarketinvestmentinwhathasbecomeknownasthe"consumptioncapitalassetpricingmodel(CAPM)".TheintertemporalCAPMisdevelopedby Merton ( 1973 ). Merton ( 1973 )concludesthatwhensecuritieshavestochasticreturns,anindividual'sportfolioholdingsarefoundintermsofhisindirectutilityfunctionforwealthandequilibriumexpectedassetreturnsarecorrespondinglyfoundintermsofaggregatewealthandthereturnsonassetsthatareperfectlycorrelatedwithchangesinthevariousstatevariables. 16

PAGE 17

( 1979 )extends Merton ( 1973 )andintroducestheconsumption-beta2andarguesthatinarationalexpectationsequilibriumassetpricesintheeconomyarefunctionsoftheconsumptionpreferencesofindividualsandtime,whicharenon-stochastic.Theindividual'sproblemistochooseanoptimalrateofconsumptionandanoptimalportfolioofriskyassetstomaximizetheexpectedvalueateachinstant. TheriskyinvestmentprojectisquitesimilartotheriskyportfoliosinBreeden'swork.However,whileBreedenfocusesmoreontheinvestmentofsecuritiesinthenancialmarket,thisdissertationemphasizesthepotentialaspectofinvestmentdistortionwhentheinvestmentdecisionismoreeconomicandinvolvesrealtechnologyandlabor.Moreover,thisdissertationnotonlydiscussestheconnectionbetweenconsumptionbehaviorandinvestmentbutalsoexplorestheimpactofparticularconsumptionpreferencesonthepotentialinvestmentdistortionproblem.Consumptionspending,asdiscussedby DanthineandDonaldson ( 2005 ),denesa"standardofliving,"andmostindividuals,ifnotall,preferastablestandardoflivingfromtimetotime.3Therefore,intertemporalconsumptionpreferenceisnotanunrealisticfocus.Althoughtheborrowerpreferssmoothconsumption,whentheinvestmentdecisionisconsideredatthesametime,consumptionsmoothnessisnotwithoutacatch.Inparticular,intertemporalconsumptionpreferencemakestheborrowermoresusceptibletoinvestmentineciency.Intuitively,toavoidinvestmentdistortiontheaggregateborrowerfacesafewhurdles.First,whetherthepersonalwealthissucientlylargetonancetheprojectwithoutborrowing.Secondisthefeasibilityoftheloanagreement.However,ceterisparibus,fortheborrowerwith Sharpe ( 1964 )and Lintner ( 1965 )CAPMbeta;theformerdenesassetbetastobemeasuredrelativetochangesintheaggregateconsumptionrate,ratherthanrelativetothemarket,whichisthedenitionofthelatter.3Indeed,consumptionhaslongbeenfoundtobesurprisinglysmoothovertime. Hall ( 1988 )ndsthatariseintheinterestrateisnotaccompaniedbyanincreaseinconsumption.Similarly, Deaton ( 1987 )alsondsthatconsumptionissmoothrelativetoincomeandprices. 17

PAGE 18

Inreality,thelendermaynothaveaccesstothepreciseinformationabouthowmuchpersonalwealththeborrowerpossessesmainlyfortworeasons:rst,theborrowercansimplyconcealsuchinformationanditmaybecomealmostimpossibleorprohibitivelyexpensiveforthelendertoconsultathirdpartytorevealsuchinformation;second,eventhoughthelendercanobservetheborrower'sprivatepersonalwealthinformationwithnoadditionalcost,itcanbeproblematicforthelendertodeterminehisexactnetworth.Forexample,theborrower'spersonalwealthcantaketheformsofcash,realassets,oramixofboth.Whilethevalueofcashiseasilydetermined,theremaynotbeamarketvaluefortheassetsortheborrowermayhavebetterinformationofthemarketabilityoftheassets,bothofwhichmakeitdicultforthelendertohavethesameinformationabouttheborrower'spersonalwealthastheborrower. Intheeconomicsliteraturedealingwithinvestmentbehaviorandprivatelyinformedpersonalwealthisnotuncommon. LewisandSappington ( 2001 )investigateanoptimalcontractbetweenaprincipalwhoistheownerofaprojectandanagentwhohastheskillsrequiredtooperatetheproject.Theagentisprivatelyinformedabouthisabilitytooperatetheproject,hiswealth,andhiseortsupplies. LewisandSappington ( 2001 )ndthatthepoweroftheincentiveschemedoesnotalwaysincreaseashiswealthorabilityaloneincreases.Inotherwords,abilityandwealthactasperfectcomplementsindeterminingthepoweroftheincentiveschemeandanagentrequiresthehigherlevel 18

PAGE 19

LewisandSappington ( 2001 ).Theowneroftheinvestmentprojectistheborrower,notthelender.Giventheborrower'sownershipoftheprojectandthefactthathemaybewealth-constrained,amoralhazardproblemintheborrowingsphereisofprimaryconcern;therefore,theborrower'sabilityisnotconsideredasachoicevariable.Thisdissertationismorecloselyrelatedto LewisandSappington ( 2000 ),wheretheyndthattheprincipalwhoisalsotheowneroftheprojectcaninducetheagent,whoisalsotheprojectoperator,totruthfullyrevealhisprivatelyinformedconstrainedpersonalwealthbypromisingahigherprobabilityofoperationand/oragreatershareofrealizedprotthelargerthebondthatapotentialoperatorpostswhenhiseortisessentiallynon-contractible.Inbothstudies,theasymmetricinformationaboutthepersonalwealthleadstoanadverseselectionproblemandthenon-contractiblepersonaleortsupplyrepresentsamoralhazardproblem.Furthermore,theRevelationPrinciple( Myerson ( 1981 ))appliesandatruthfuldirectmechanisminducestruth-telling.Theultimategoalfor LewisandSappington ( 2000 )'sagentandtheborrowerinChapter5istomaximizetheexpectedprotfromtheoperation/project;however, LewisandSappington ( 2000 )ndthattheagenttruthfullyrevealshiswealthbecauseheispromisedahigherprobabilityofoperationand/oragreatershareofrealizedprotthelargerthebondtheagentposts.InChapter5,theborrowerchoosestruthfulrevelationofhiswealthbecausehedoesnotgainbasedonthelimitedreportingstrategyandtheformofhisutilityfunction. 19

PAGE 20

Toimplementtheproject,theborrowermayapproachthelendertoborrowD00.Uponassessingtheprojectproposedbytheborrower,thelendereitherrejectstheproposaloragreestonance,wherethecorrespondingnancingagreementspeciestheewpaymentsfDH;DLgtothelender.Tobeconsistentwith Tirole ( 2006 ),thelenderoperatesinacompetitivenancialmarket.1Afternancingissecured,theborrowerexertsaneortwhichleadstoabinarycashowX2fXH;XLg,thevalueofwhichisrealizedineitheragoodstateorabadstateandXH>XLimpliesthatthecashowfromthegoodstateisstrictlyhigherthanthatfromthebadstate.2TheborrowercaneitherbehavebyexertingthedesiredeortaH>0,whichyieldsaprobabilitypH2(:5;1)ofXH,3ormisbehavebytakingashirkingactionaL=0withanormalizedzeropersonal 20

PAGE 21

Assumption1suggeststhathigheortisalwayspreferredinequilibriumandtheinvestmentprojectisattractivetotheborrower.Sincethelenderoperatesinacompetitivenancialmarket,alltheexpectedNPVfromtheinvestmentaccruestotheborrower;Assumption1restrictstheborrower'sequilibrium"netmonetarygain"fromtheprojecttobestrictlygreaterthanzero. Assumetheborrowerisanexpectedutilitymaximizerwithconstantabsoluteriskaversion(CARAhereafter).Hisutilityfunctiontakestheformofnegativeexponential,whichisincreasingandconcaveinhisgrossconsumption. 21

PAGE 22

where>0istheArrow-Prattrisk-aversionparameterandaistheborrower'spersonalcostofeort.Theborrower'sobjectiveistomaximizehisexpectedutilityfromhisconsumption. Beforeformallydiscussingtheborrower'sinvestmentoptions,theborrower'sconsumptionbehaviorwhennoinvestmentopportunityispresentisbrieydiscussedforcompleteness. Whenthereisnoinvestmentopportunity,theborrowersimplyconsumeshispersonalwealth,andhisoptimalconsumptionischaracterizedasfollows: ( 2005 )discussconsumptionpreferencesformulti-periodcontractsindetailinChapter25. 22

PAGE 23

Theintuitionforthisconsumptionscheduleisstraightforward.Giventhelackofpersonalbankingaccesstheborrower'soptimalchoiceistoconsumeallthewealthathand.Moreover,sincetheborrowerhasaggregateconsumptionpreferencethetotalconsumptionratherthanperiod-by-periodconsumptionisthefocus.TheborrowercanarbitrarilyallocatehispersonalwealthWbetweentwoconsumptioninstantsoftime;therefore,thereisnouniquesolutionforhisperiod-by-periodconsumption. Therisk-neutrallenderbehavescompetitivelyinthenancialmarketinthesensethatthenancialagreement,ifitexists,makeszeroprot.Therefore,inequilibrium,thelender's(IR)constraintisbinding.6Theborrower'sconsumptionisbinary,i.e.,C2fCH;CLg,whichisboundedabovebythesumofhispersonalwealth(W)andthestochasticcash 23

PAGE 24

Finally,CHandCLaretheborrower'smonetaryholdings,whichareusedforconsumptionpurposes;theyareassumedtobenon-negative.Thesenon-negativityconstraintsareessentiallysimilartothelimitedliabilityconstraintsimposedintheagencyproblems(e.g., Sappington ( 1983 )and Innes ( 1990 ));therefore,theyarereferredtoaslimitedliabilityconstraintshereafter.7 3{2 )andthelimitedliabilityconstraints. 3{2 )allsuggestthatthemagnitudeoftheexpectedNPVHiswhatisimportanttotheanalysis;inotherwords,itisthemagnitudeofXLrelativetothatofXHthatmatters,notthemagnitudeofXLitselfthatplaysaroleintheloanagreement.However,intheno-lenderinvestmentanalysislateron,thesignofXLisimportantindeterminingthefeasibilityoftheno-lenderinvestmentincertaincircumstances. 24

PAGE 25

GrossmanandHart ( 1983 )and Holmstrom ( 1979 ).However,italsohasdistinctcharacteristics.Inequality( 3{2 )suggeststhatweretheinvestmenttobetakentheborrower'sexpectedconsumptionwouldbenogreaterthanthesumofhispersonalwealthendowmentandtheexpectedNPVfromtheinvestment.Foragivenproject,theborrower'spersonalwealthWandtheexpectedNPVHarebothxed,whichimpliesthattheborrower'sconsumptioninexpectationhasanupperbound.Thisisquitedierentfromtheconventionalagencyproblemwheretheborrower'sindividualrationalityconstraintdefactoimposesalowerboundontheagent'scompensation.8Aswilllaterbecomemoreclearerinthesecond-bestcontract,thefactthattheborrower'sequilibriumexpectedconsumptionisboundedabovebyaxedamountplaysanimportantroleinthepotentialinvestmentdistortionproblem. Similartowhatisestablishedintheconventionalagencyproblem,therst-bestcontract[ProgramFB]wouldprovidefullinsurancefortheborrowerweretheborrowertosupplyhigheort.Asaconsequence,therisk-averseborrowerdoesnotbearanyriskassociatedwiththecashowfromtheprojectandhisconsumption. 25

PAGE 26

Assumetheinvestmentprojectisviable,theconsumptionscheduleisriskless.ThisisbecauseJensen'sinequalityimpliesthatriskyconsumptionisnotoptimalintherst-bestcontract.Moreover,theborrower'sutilityfunctionisstrictlyincreasingandconcave;therefore,wehaveU(CaH)=U(W+HaH) whichimplies( 3{3 ). Notethattheborrower'soptiontoself-nancetheinvestmentprojectifheiswealth-abundantisnotdiscussedinthissection.Itturnsoutthatwhentheborrower'seortsupplyispubliclyobservableandcontractible,self-nancingisinferiortoborrowing.Intuitively,theloanagreementprovidesafull-insuredrisklessconsumptionfortheborrower,whichwouldbeinfeasibleweretheborrowertochooseno-lenderinvestmentastheprojectresultsinriskycashow.Therefore,therisk-averseborroweralwaysprefersborrowingtoself-nancingwhenhiseortsupplyispubliclyobservable. 26

PAGE 27

Unliketherst-bestcontract,thesecond-bestenvironmentfeaturespotentialinvestmentdistortionandconsumptiondistortion,whicharedenedasfollows: Thelackofamutuallyagreeablesecond-bestcontractintherstsituationimpliesthatnocontractcanbefoundthatmotivatestheborrowertoworkdiligentlyandletsthelenderrecoupherinvestmentatthesametime.Twopossibilitiescancontributetothistypeofinvestmentdistortion:1)theloanagreementisinfeasiblebecausetheredoesnotexistaconsumptionschedulethatsolvestheborrower'smaximizationproblemwithrespecttotheloancontractand2)giventheexistenceofsuchaconsumptionschedulethatmaximizestheborrower'sexpectedutilityinthesecond-bestloanagreement,the 27

PAGE 28

28

PAGE 29

Theonlydierencebetweenthesecond-bestandtherst-bestloanagreementisthemoralhazardconcernbroughtaboutbytheborrower'sunobservableaction,whichrequiresthecorrespondingnancingarrangementtogranttheborrowersucientstakeintheprojecttoinducehimtoworkdiligently.WithaL=0,inequilibrium,theborrowerbehavesifthefollowingincentivecompatibilityconstraint(IChereafter)holds: Theleft-andright-handsidesof( 3{4 )aretheborrower'sexpectedutilityfromhisconsumptionwhenhechoosestoshirkandtodeliverhigheort,respectively;therefore,( 3{4 )ensuresthattheborrowerismotivatedtoworkdiligently.Rearrangingthetermsin( 3{4 )yields Notethattheright-handsideof( 3{5 )isnegativebecausepH>pLandaH>0ensurethesquare-brackettermisstrictlypositive.Ontheleft-handsideof( 3{5 ),since 29

PAGE 30

whichimpliesthat (1pH)(3{6) Thefollowingassumptionisheldthroughoutthischapter. (1pH)alwaysholds. GivenAssumption2,inequality( 3{5 )issimpliedas where=pHexp(aH)pL 3{7 )constraintisensured.11 3{7 )constraintadded,whichis [ProgramSB]maxCH0;CL0pHexp[(CHaH)](1pH)exp[(CLaH)]s.t.pHCH+(1pH)CLW+HCL+1 3{6 ),iswelldenedforthenaturallogarithmfunctionin( 3{7 ). 30

PAGE 31

3{8 )inProposition 1 ensurestheconstraintsetcanbesatised(i.e.,theconstraintsetisfeasible),which,inturn,determinestheexistenceofthesecond-bestcontract[ProgramSB].Closerinvestigationshowsthatcondition( 3{8 )isalsoanecessaryconditionforasolutiontoexistfor[ProgramSB].12Itisevidentthatgiventheexistenceofthesecond-bestsolutiontheborrower'slimitedliabilityconstraintsarebothsatised.Stateddierently,thelimitedliabilityconstraintsaretrivialiftheconstraintsfor[ProgramSB]arefeasiblebecausenegativeconsumptionneverarise.Proposition 1 impliesthataconsumptiondierentialarisesatthesolutiontotheproblem,i.e.,CH>CL,whichisconsistentwiththeconventionalagencyproblemthattheborrowerbearssomerisktobemotivatedtodeliverhigheort. Undercondition( 3{8 ),theborrower'soptimalconsumptionforthesecond-bestcontractshowsthefollowingcharacteristics:CH>CL0 31

PAGE 32

3{8 )holdsatequality.Theborrower'slimitedliabilityconstraintforCHisobviouslyneverbinding.Thus,thefollowingcorollaryisdevelopedtoaddressthelimitedliabilityconstraintsfor[ProgramSB]: 1 concludesthatthesecond-bestcontracteitherhasaninteriorsolutionornosolution;acornersolutionneverarises.13AssumingCHandCLaretheoptimalsolutionsfromtheconventionalagencyproblem,otherthingsbeingequal,constraint( 3{2 )isreplacedbytheusualindividualrationalityconstraintthattheborrower'sexpectedutilityisnolessthanhisreservationutility.Thelimitedliabilityconstraintsintheconventionalagencyproblemworkinthefollowingway:wereCLtobenegative,thelimitedliabilityconstraintforCLwouldbebindingandthecorrespondingCHwouldgoupbytheamountthatkeepstheconsumptiondierentialconstantatwhatitwouldbeifCLwereallowedtobenegative.However,thispracticeisnotapplicablefor[ProgramSB]andthereasonisinconstraint( 3{2 ),whichimposesanupperboundontheborrower's Thenan"interiorsolution"referstoC0andLC=0anda"cornersolution"referstoC=0andLC<0ThisalsoimpliesthatthelimitedliabilityconstraintsforCarenotbindingwhentheinteriorsolutionarises,whereastheyarebindingforthecornersolution. 32

PAGE 33

3{2 )provides Inequality( 3{9 )impliesthatthereisanupperboundforCH.Therefore,foranyCL<0itisinfeasibletoadjustCHupwardstoinduceCL=0.ThisstatesthatthelimitedliabilityconstraintforCLisneverbindingandCL=0when( 3{8 )issatisedatequalityisstillpartoftheinteriorsolutions.Intuitively,Corollary 1 statesthatnoloanarrangementexiststhatprovidesasolutionofCL<0thatcanmotivatetheborrowerandletthelenderrecoupherinvestmentatthesametime.14 3-1 showswhethertheexistencecondition( 3{8 )forthesecond-bestcontractholds.Thethirdcolumnshowsthesecond-bestcontract(labeledas"SB")describedinProposition 1 .15Thelastcolumnshowstheborrower'sconsumptionfrom[ProgramSB]butwithoutthelimitedliabilityconstraints,i.e.,unrestrictedsecond-bestcontract(labeledas"SBU").Assumptions1and2areallsatisedwiththeseparametervalues:H=pHXH+(1pH)XLI=140L=pLXH+(1pL)XLI=20aH=90<1 (1pH)=109:9 3{9 )alonedoesnotleadtothenon-existenceoftheloancontract,itisajointforcewiththelimitedliabilityconstraintonCLthatpotentiallycausesthefailureoftheloanagreement.WereCLtobeallowedtogonegative,afeasibleCHwouldalwaysbefoundtosatisfy( 3{9 ).15Therst-bestcontractcharacterizedinFact2isnotshownbecauseofitsstraightforwardness. 33

PAGE 34

Second-BestContractsfortheBorrowerwithAggregateConsumption SBU W 3{8 ) N=A N=A 50 29 259 29 200 179 409 179 ConsistentwithProposition 1 ,thesecond-bestcontractisnotalwaysfeasible.ItisevidentfromTable 3-1 thatwhentheborrower'spersonalwealthWis20,thereisnofeasiblesolutionforthesecond-bestcontract.Thisisreectedbythefailuretosatisfycondition( 3{8 )inProposition 1 .16TakingW=20asanexample,wecomparetwosecond-bestcontractscharacterizedinTable 3-1 AtW=20,thecontractSBisobviouslynotfeasible.Thealternativesecond-bestcontractSBU,whichfeaturesthelackoftheborrower'slimitedliabilityconstraints,hassolutionsofCH=229andCL=1.However,thisconsumptionscheduleisnotfeasiblewhenthelimitedliabilityconstraintsareimposed.Weretheborrower'slimitedliabilityconstraintforCLtobebinding,i.e.,CL=0,itwouldfollowfromtheborrower'sICconstraintthatthecorrespondingCHisadjustedupwardsby1tobe230tokeeptheborrowerincentivecompatible.Unfortunately,substitutingCH=230andCL=0intothe( 3{2 )constraintyieldspHCH+(1pH)CL=161W+H=160 whichisaviolationoftheconstraint. 3{8 ),itisnotjustthevariationofWthatmayleadtothenon-existenceofthesecond-bestcontract.Forexpositionalsimplicitytheexample(inTable3-1)focusesontheminimalvariationofoneparameterwhilexingtheotherparameters. 34

PAGE 35

3{2 ))playanimportantroleindeterminingthefeasibilityofcontractSB. Whenthesecond-bestloanSBisfeasibletheoptimalconsumptionfeaturesconsumptiondierentialCH>CLtoensurethattheborrowerismotivated.Furthermore,theconsumptionspreadalwaysstaysat230(i.e.,CHCL=230),reectingtheborrower'sCARAriskpreference.TheconsumptionbehaviorbetweencontractsSBandSBUareidenticalsincetheborrower'sconsumptionarepositiveandthelimitedliabilityconstraintsarethentrivial. 1 ,thefailuretosatisfycondition( 3{8 )contributestotheinfeasibilityoftheloanagreement.Whenthesecond-bestcontractisfeasible,( 3{2 )suggeststhattheexpectedNPVfromtheinvestmentprojectaccruestotheborrower;therefore,theborrower'snetexpectedgainfromtheinvestmentisHaH.ItturnsoutthatnotanylevelofHaHensurestheinvestmentbehavior.Inparticular,thereisaboundaryconditionfortheborrower'snetgainunderwhichheisbetterowithouttheinvestment.Theconclusionsaresummarizedinthefollowingproposition: 35

PAGE 36

ConditionaofProposition 2 addressestheinfeasibilityofthesecond-bestcontract.Thelackofafeasiblesecond-bestcontractisessentiallyaformofcreditrationingforthewealth-constrainedborrower. Tirole ( 2006 )pointsoutthatinsucientpersonalwealthmaycontributetocreditrationingwhentheborrower'seortsupplyisprivatelyobserved.Theargumentisthatsincetheborrowerenjoystheentirebenetofshirkingbutsharestheinvestmentprojectwiththelender,thecontractbetweentheborrowerandthelenderthenrequirestheborrowertocontributemoreendowmenttoensureheismotivated.ConditionaisconsistentwithTirole'scredit-rationingargument.Basically,fortheborrowerwithpersonalwealthW
PAGE 37

3{8 )inProposition 1 ,wehave Forthewealth-xedborrower,astheright-handsideof( 3{10 )increases,condition( 3{8 )ismorelikelytoholdatequality.Therefore,asthesecond-bestcontractbecomesmorerisky,theborrower'sexpectedutilityfromthesecond-bestcontractisdecreasing;heisthenmorelikelytopreferthestatusquototheloanarrangement.Notethattheincreaseintheright-handsideof( 3{10 )ismorelikelytotriggerConditionbtohold,whichmakestheborrowermorevulnerabletoinvestmentdistortion. Whentheborrowerhassucientpersonalwealth,i.e.,WI,theborrower'sdecisionbetweenthesecond-bestcontract,givenitsexistence,andthestatusquoisnotconclusivetodetermineinvestmentdistortionsincetheborrowerhastheoptiontoself-nancetheinvestmentproject.Whenthesecond-bestcontractisinfeasible,thechoiceisonlybetweenno-lenderinvestmentandthestatusquo.LetCENLdenotetheborrower'scertaintyequivalentconsumptionfromno-lenderinvestment,itthenfollowsfromhisconsumptionbehavior17thatCENL=WI+1

PAGE 38

Giventhattheborrower'scertaintyequivalentconsumptionfromthestatusquoisjusthispersonalwealthW,hisdecisiontodeterminewhetherself-nanceisasoundchoiceboilsdowntocomparingtheborrower'snetcertaintyequivalentconsumptionfromno-lenderinvestmentwithW.Inotherwords,wecomparethestochasticcashowfromtheinvestmentCE(X)withtheincurredinvestmentcost,whichisI+aH.Ifthebenetislessthanthecost,i.e., theprojectistooriskyfortheborrowertoengageinself-nancing;therefore,investmentdistortionarises. Whenthesecond-bestcontractisfeasible,investmentdistortionrequiresthefailureofbothno-lenderinvestmentandloanagreementwiththelender.Weconcludethepotentialinvestmentdistortionwithawealth-abundantborrowerasfollows: whereCE(X)=1 ThesecondconditioninPartbofProposition 3 directlyfollowsfromConditionbofProposition 2 ;itistheconditionunderwhichthestatusquoispreferabletotheloanagreement. Whenthesecond-bestcontractisfeasible,thenextlogicalquestioniswhethertheborroweralwaysprefersborrowingtoself-nancing.Thefollowingcorollaryisdeveloped: 38

PAGE 39

NoInvestSecond-BestLoanNo-LenderInvest(3{12) 3 isonlynecessarybutnotsucientfor( 3{12 ).Furthermore,Corollary 2 impliesthatborrowingisnotalwayspreferabletoself-nancingwhentheborroweriswealth-abundant. LetEUSB(EUNL)andRUdenotetheborrower'sexpectedutilityfromthesecond-bestcontract(no-lenderinvestment)andhisreservationutilityfromno-investment.18Thefollowingexampleisdesignedtoillustrateinvestmentdistortion(orthelackofinvestmentdistortion)discussedinProposition 2 andProposition 3 .TheparametervaluesinTable 3-2 followthoseinTable 3-1 unlessotherwisespecied.Withtheparametervalues,thecertaintyequivalentcashowfromtheinvestmentprojectiscalculatedasCE(X)=243.Furthermore,let"ID"standforwhetherthereisinvestmentdistortionandCHandCLbetheoptimalsecond-bestconsumptionderivedinProposition 1 Theupperexhibitrepresentsthewealth-constrainedborrower.CaseAandCaseBbothsuggestinvestmentdistortion;however,theCaseAscenarioreectsConditionaofProposition 2 thatthesecond-bestcontractisinfeasible(W+HpH 2 thatthesecond-bestcontractexists 39

PAGE 40

InvestmentDistortionfortheBorrowerwithAggregateConsumption aH ID A N=A N=A B 226 6 C 176 123 W=180 aH EUNL D N=A N=A E 450 16 F 341 221 2 holdsandtheborrowerachievesahigherexpectedutilitywithborrowingfromthelenderthanconsuminghispersonalholdings(i.e.,EUSB=:28>RU=:82). Thelowergurefocusesonthewealth-abundantborrowerwithW>I.CaseDandCaseErepresentinvestmentdistortionbutbasedondierentgrounds.CaseDreectsPartaofProposition 3 ,wherethesecond-bestcontractisnotfeasible;therefore,investmentdistortionarisesbecauseno-lenderinvestmentisnotpreferabletothestatusquo(CE(X)
PAGE 41

Second-BestLoanNo-LenderInvest(3{15) ForCaseE,( 3{14 )holdsbut( 3{15 )isviolated.Therefore,asreectedbyTable 3-2 ,theborrower'sexpectedutilityfromno-lenderinvestmentishigherthanthatfromthesecond-bestcontractforCaseE.Intuitively,inthepresenceoftherelativelylargepersonalcostofhigheortsupply(aH=107),althoughtheloanagreementprovidesrisk-sharingitalsoenlargestheconsumptiondierentialtomotivatetheborrower,whichmakestheloanagreementveryriskyandinconsequencereducestheborrower'sexpectedutilitydramaticallygivenhisconcaveutilityfunction.Weretheborrowertochooseself-nancing,hisconsumptionfromno-lenderinvestmentwouldbe380and180whenhighandlowcashowsarerealized,respectively.Thisconsumptionscheduleisnotfeasibleforthesecond-bestloanagreementbecausetheborrowerICconstraintwouldbeviolated.Whentheborrowerweighsthebenetbroughtaboutbyrisk-sharingwiththelenderandthecostofrespectingtheimposed(IC)constraint,theborrowermightbeworseosigningacontractwiththelenderthanjustinvestingonhisown. Onthecontrary,CaseFreectsthesituationwherethesecond-bestcontractistheborrower'soptimalinvestmentdecision.Inparticular,condition( 3{15 )issatisedwithCaseF.Here,themoderatepersonaleortcostoftheborrower(aH=60)makesthebenetfromrisk-sharingoutweighthecostfromapplyingtheconsumptiondierentialtokeeptheborrowermotivated. 41

PAGE 42

Theanalysisfortheborrowerwithtime-additiveconsumptionpreferencesharesquiteafewsimilaritieswiththatfortheborrowerwithaggregateconsumptionpreferenceintheprevioussection.Inparticular,regardlessoftheborrower'sconsumptionpreference,hefacessimilarinvestmentoptionsandhisinvestmentbehaviorfeaturesnoinvestmentdistortionorpotentialinvestmentdistortion. ( 1960 )stationarityaxiomimpliesthatintertemporalpreferencesbetweentwoperiodsremainthesame.Therefore,theborrowerwithtime-additiveconsumptionpreferencehasanaggregateutilityfunctionthatcanbeexpressedasthesumofhisperiod-by-periodutility. 42

PAGE 43

Therst-bestcontractfortheborrowerwithintertemporalconsumptionconcernismodeledsimilarlyto[ProgramFB];however,theborrower'sconsumptionconstraintsaredierent.Att=0,theborrowerhasapersonalwealthlevelofW,thisdoesnotnecessarilyindicatesheconsumestheentireWatt=0,hecanchoosetoconsumeW0WandsaveW1=WW00forthefutureperiodt=1.20Thereforeatt=0,theupperboundfortheborrower'sconsumptionishisinitialpersonalwealthnetofhisinvestmentintheprojectandhisstoredwealth,i.e.,CA0W(ID0)W1 43

PAGE 44

[ProgramFBA]maxCA00;CAH0;CAL0exp(CA0)pHexpCAHaH(1pH)expCALaHs.t.CA0+pHCAH+(1pH)CALW+H whereH=pHXH+(1pH)XLI. Theborrowerwithtime-additiveconsumptionpreferencehasaconsumptionschedulethatresemblesthatoftheborrowerwithaggregateconsumptionpreference,andinparticulartheborrower'sintertemporalnetconsumptionaresmooth,whichsuggestsCA0=CAHaH=CALaH where=pHexp(aH)pL 44

PAGE 45

2CAH=W+HpH 2+1 2 45

PAGE 46

where=pHexp(aH)pL CloserinvestigationrevealsthatinthecornerconsumptionsituationonlytheconstraintforCALisbinding.Thedrivingforceofthispossibilityistheborrower'stime-additiveconsumptionpreference.Recallthatfortheborrowerwithaggregateconsumptionpreference,thereisanupperboundonCH,whichisderivedfrom( 3{2 )constraint.Similarly,whentheborrowerhasatime-additiveconsumptionpreference,inequality( 3{16 )alsosuggeststhatCAHisboundedfromabove Moreover,thesecond-bestoptimalconsumptioninProposition 4 impliesthat Therefore,from( 3{19 )and( 3{20 )itispossibleforCAHtobeadjustedupwardswhenthelimitedliabilityconstraintforCALisbindingbecauseofthepresenceofCA0.Intuitively,ifthecontrolproblemgetssevere,theconsumptiondierentialneedstoincreasetoensuretheborrowerdelivershigheort.Thisdoesnotimposeanyadditionalrestrictionaslongastheoptimalconsumptionisstillinterior;however,whenthelimitedliabilityconstraintofCALisbinding,thequestionofwhetherthereexistsafeasibleconsumptionschedulethatsatisestheborrower'sICAconstraintbecomesimportant.As( 3{19 )shows,CA0providessomerangewhichmakesCAHstillbefeasiblewhenthelimitedliabilityconstraintofCALis 46

PAGE 47

Asindicatedby( 3{20 ),thelimitedliabilityconstraintfortheborrower'st=1consumptionCAHisneverbinding.Furthermore,Proposition 4 alsosuggeststhatthelimitedliabilityconstraintforCA0isnotbindingaswell.Thereasonissimilartothereason,discussedearlier,whythelimitedliabilityconstraintofCLneverbindsfortheborrowerwithaggregateconsumptionpreference.Again,theupperboundin( 3{19 )prohibitstheexistenceofsuchCAHthatcanprovideamutuallyagreeablesecond-bestcontracttorespecttherequiredconsumptiondierentialwhenbothCALandCA0areforcedtobecornersolutions. FromProposition 4 ,itisevidenttheconsumptiondierentialisunavoidableinthesecond-bestcontract;furthermore,consumptionacrossperiodsisnotsmoothing,whichisconsistentwiththeconsumptiondistortiondenition.Deneutilitysmoothingwhenthefollowingequalityholds:exp(CA0)=pHexpCAHaH(1pH)expCALaH 47

PAGE 48

DeneCEtasthelocalcertaintyequivalentconsumptionfortheborrowerwithtime-additiveconsumptionpreferenceattimet=0;1,respectively,Corollary 3 thenstatesthatexp(CE0)=exp(CA0)exp(CE1aH)=(1pH)expCALaHpHexpCAHaH Equation( 3{21 )indicatesthattheborrower'snett=1localcertaintyequivalentconsumptionequalshist=0localcertaintyequivalentconsumption.Therefore,eventhoughthereisconsumptiondistortionwithrespecttothesecond-bestconsumptionschedule,theborrower'sdesiretohavesmoothconsumptionispreservedinthesensethatsmoothingutilityisnotdistorted. Recallthatinformulatingthetime-additiveborrower'sconsumptionproblem,theborrowerisassumedtoputW10asideasareserveforhist1consumption.Wearguethatoncethesecond-bestcontract[ProgramSBA]isinplace,itisoptimalfortheborrowertoalwaysstoreadditionalconsumptionW1.Considerthefollowingtwopotentialconsumptionschedules: 48

PAGE 49

B. ItisevidentthatConsumptionBehaviorAstoresnoadditionalconsumptionfort=1whileScheduleBdelayssomeconsumptiontillt1.Fortheborrowerwithtime-additiveconsumptionpreference,theconsumptiondierentialatt=1isdeterminedbyhisattitudetowardsrisk,whichisinvariantwithhispersonalwealthholdings;therefore,whetherW1ispositiveornothasnoeectonhisconsumptiondierentialfort=1. ForConsumptionScheduleA,theborrower'speriod-by-periodexpectedutilityis (3{22)pHexp[(XHDH)](1pH)exp[(XLDL)] (3{23) andforConsumptionScheduleB,hisperiod-by-periodexpectedutilityis (3{24)pHexp[(W1+XHDH)](1pH)exp[(W1+XLDL)] (3{25) Itisevidentthat ( 3{22 )( 3{24 ) 49

PAGE 50

3{23 )( 3{25 ) However,giventheconcavityoftheborrower'sutilityfunction,thefollowingrelationshipholds ( 3{22 )( 3{24 )( 3{25 )( 3{23 ) whichisequivalentto ( 3{22 )+( 3{23 )( 3{24 )+( 3{25 ) wheretheequalityholdsatW1=0.Thisindicatesthattheborrower'sexpectedaggregateutilityisweaklyhigherwhenhehasConsumptionScheduleBratherthanA,whichimpliesthatitisoptimalfortheborrowertostoreadditionalconsumptionatt=0. Beforeproceedingtoinvestmentdistortion,thefollowingexampleisprovidedincomparisonwithTable 3-1 .TheparametervaluesfollowthoseinTable 3-1 unlessotherwisespecied.Otherthanthecolumnfortheborrower'spersonalwealthlevels,Table 3-3 consistsofthreeblocks.Therstandthesecondblocksarethesecond-bestcontractcharacterizedinProposition 4 withinteriorandcornerconsumptionindicated,whicharelabeledas"SBI"and"SBC,"respectively.Forminimalvariation,onlyaHischangedtotriggerthecornerconsumption.21Thelastblockisthesecond-bestcontractifthelimitedliabilityconstraintsareabsentgiventhesameborrower'spersonalcostofeortaHasforthecornerconsumption,whichislabeledas"SBUA."Therst-bestcontractforthetime-additiveborrowerfollowsFact4andissmoothingintheborrower'snetperiod-by-periodconsumption;therefore,itisnottabulated. 50

PAGE 51

Second-BestContractsfortheBorrowerwithTime-AdditiveConsumption SBC SBUA CA0 N=A N=A N=A N=A N=A 225 50 19:1 240 9:7 1:2 269:7 0 5:5 265:4 200 94:1 315 84:7 5:9 477:2 0 8:2 475 Whentheborrower'spersonalwealthholdingsis20,thesecond-bestcontractisnotfeasible.22ThisisbecauseforW=20,condition( 3{18 )isviolated. Whenthesecond-bestcontractexists,itisevidenttheconsumptiondierentialispresentandperiod-by-periodconsumptionisnotsmoothing,bothofwhichindicateconsumptiondistortion.ThesecondblockinTable 3-3 providescornersolutionforthesecond-bestcontract.Intuitively,astheborrower'spersonalcostofequilibriumeortgrows,hisdisutilityfromworkingdiligentlyincreases.Theincentivecontractneedstooerawiderconsumptiondierentialtoensuretheborrowerismotivated,whichmakesitmorelikelytosetothelimitedliabilityconstraintforCAL.ThemagnitudeofCA0providessomecushionforsuchconsumptiondierentialadjustment.Inthepresenceoflimitedliability,oncethelimitedliabilityconstraintforCALisbinding,thesecond-bestcontractreallocatesthedistributionofconsumptionacrossperiodstoensurethefeasibilityofsuchaloanagreement. ThereallocationisnotwithoutitsboundarysincethelimitedliabilityconstraintexistsforCA0aswell.ThelastblockofTable 3-3 forW=20reectssuchacase.WereCALtoberestrictedtobenon-negative,wewouldseeCAH=225+j5:3j=230:3 51

PAGE 52

toensuretheconstraint( 3{16 )holds;however,CA0'slimitedliabilityconstraintisviolated,whichimpliesthereisnosolutionfor[ProgramSBA]. Aftercharacterizingthesecond-bestcontract,wenexttacklethepotentialinvestmentdistortionprobleminthesecond-bestenvironment.Similartotheborrowerwithaggregateconsumptionpreference,forthewealth-constrainedborrower(i.e.,W
PAGE 53

where=pHexp(aH)pL InSection3.2.2.3,theconditionunderwhichthestatusquoispreferabletono-lenderinvestmentfortheborrowerwithaggregateconsumptionpreferenceisderivedas whereCE(X)=1 3{26 )isstillasucientbutnotanecessaryconditionforthestatusquotowinoverno-lenderinvestmentwhentheborrower 53

PAGE 54

investmentdistortioncanstillarise.Thisimpliesthatprojectsthatarenotconsideredriskyfortheborrowerwithaggregateconsumptionpreferencemaybeconsideredriskyfortheborrowerwithtime-additiveconsumptionpreference.Inparticular,theborrowerwhodoesnotachievesmoothingutilitiesacrosstimeismorevulnerabletotheriskinessoftheinvestmentproject.Sincefortherisk-averseborrowertheinvestmentresultinginnon-smoothingintertemporalutilitiesisriskierthantheonewhichprovidesotherwise,itisimportantfortheborrowertosmoothhisutilitiesacrosstime.However,withoutaccesstopersonalbanking,itisnotalwaysfeasiblefortheborrowertodoso. Withno-lenderinvestment,theborrower'sintertemporalconsumptionscheduleisCA0WICAHWCA0I+XHCALWCA0I+XL 3{26 )isasucientandnecessaryconditionforProposition 3 inSection3.2.2.3todeterminetheborrower'spreferenceofstatusquotono-lenderinvestment.24Inequilibrium,theborrower'st=1consumptionCAHorCALarebinding.Furthermore,itispossiblethattheborrower'sintertemporalutilitiesaresmoothedwhenC0=WI 54

PAGE 55

ItturnsoutthatCA0isnotbindingwhenthefollowingrelationshipissatised Otherwise,theborrower'st=0consumptionisequaltoWI.Intuitively,otherthingsbeingequal,( 3{28 )ismorelikelytoberealizedifWisrelativelylarge.ForW>>I,CA0isnotbindingandbyreservingsomeofhispersonalwealthfort=1consumption,theborrowerismorelikelytoachievesmoothingperiod-by-periodutilities,whichisstrictlypreferred.Oncetheborrowerhassmoothingutilities,asdiscussedearlier,anyinvestmentdecisionwithtime-additiveconsumptionpreferenceisnotqualitativelydierentfromthatwithaggregateconsumptionpreference;therefore,( 3{26 )isallthatmatterstodeterminewhethertheborrowerprefersthestatusquotono-lenderinvestment.However,itisinconclusivetodeterminetherelationshipbetweeninequalities( 3{28 )and( 3{26 );therefore,condition( 3{26 )isnolongernecessarytoensurepotentialinvestmentdistortion.Theinvestmentdistortionforthewealth-abundantborrowerwithtime-additiveconsumptionpreferenceissummarizedinthefollowingproposition: whereCE(X)=1

PAGE 56

Thenextnumericalexampleshowsinvestmentdistortionfortheborrowerwithtime-additiveconsumptionpreference.TheparametervaluesmainlyfollowthoseinTable 3-2 exceptfortheborrower'spersonalwealthholdings,whicharespeciedintheTable.Thesuperscript"A"isaddedtothenotationtodistinguishtheborrowerwithtime-additiveconsumptionpreferencefromaggregateconsumptionpreference.Theupperexhibitisforthewealth-constrainedborrowerandthelowerexhibitisforthewealth-abundantborrower;hisconsumptionfromthesecond-bestcontractandno-lenderinvestmentarenottabulated. CaseAinvestmentdistortionoccursbecausethesecond-bestcontractisinfeasible,whichmeansthatcondition( 3{18 )failstohold.Notethatcondition( 3{18 )isequivalenttotheonethatdeterminesthenon-existenceofthesecond-bestcontractfortheborrowerwithaggregateconsumptionpreference.CaseBrepresentsinvestmentdistortionduetotheriskinessofthesecond-bestcontract,whichsuggeststhattheborrowerprefersthestatusquototheinvestment.Closecalculationrevealsthattheborrower's"netgain"(i.e.,HaH)is52,whichisnotsucientlylargefortheborrowertoundertaketheinvestmentproject(therequiredboundaryis56).Thisboundaryconditionisalsoequivalenttotheonethatdeterminespassinguponthesecond-bestcontractbasedonitsriskinessfortheborrowerwithaggregateconsumptionpreference.Therefore,theborrower'sinvestmentdecisionisinvariantwithhisconsumptionpreferenceifheiswealth-constrained.Finally,CaseCshowsnoinvestmentdistortionastheborrower'sexpectedutilityfromthesecond-bestcontractishigherthanhisreservationutility. FortheborrowerwithWI,CaseDandCaseEinTable 3-4 bothreectinvestmentdistortion.Thesecond-bestcontractisinfeasibleforCaseDforXL=130; 56

PAGE 57

InvestmentDistortionfortheBorrowerwithTime-AdditiveConsumption W aH 100 B 88 C 50 Case W aH 109 130 E 90 150 F 90 300 3{26 )(i.e.,CE(X)
PAGE 58

Whentheborroweriswealth-abundant,otherthanseekingaloanagreementandstayingput,hecanalsochoosetoself-nancetheinvestmentproject.Giventheborrower'sconcaveandincreasingutilityfunction,theborrowerwithtime-additiveconsumptionpreferencemaximizeshisexpectedutilitywhenhisperiod-by-periodutilitiesaresmoothing.However,theconcernforconsumptionsmoothingmakestheborrower 58

PAGE 59

59

PAGE 60

TodistinguishtheChapter3borrowerfromtheborrowerdiscussedinthischapter,theborrowerinChapter3islabeledastheCARA-borrower.Chapter3analyzestheCARA-borrower'sconsumptionandinvestmentbehaviorswhilevaryinghisconsumptionpreferences.ThegeneralconclusionisthatpotentialinvestmentdistortionisweaklyaggravatedwhentheCARA-borrowerhastime-additiveconsumptionpreference.Moreover,ifthesecond-bestloanagreementistheoptimalinvestmentdecisionforbothconsumptionpreferences,conditionsunderwhichinvestmentdistortionoccursareidenticalregardlessoftheCARA-borrower'sconsumptionpreference.ThisspecialcaseismainlycausedbythenatureoftheCARA-borrower'sriskpreference.Inparticular,time-additiveconsumptionandinvestmentproblemmaybesimplyconsideredastwoidenticalone-periodconsumptionandinvestmentproblemsfortheCARA-borrower.However,weretheborrower'sriskpreferencetobedierent,suchrepeatedconsumptionandinvestmentdecisionfeaturemaynotbepresent.Inthischaptertheborrowerisassumedtohavedecreasingabsoluteriskaversion(DARAhereafter),whoseutilityfunctiontakestheformofsquare-rootonthenetconsumption.UnlikethenegativeexponentialutilityfunctionwhichistheuniqueutilityformintheCARAclass,thesquare-rootutilityfunctionisonlyoneoftheutilityformsintheDARAclass.Forexpositionalsimplicityandwithalittleabuseofterminology,theborrowerwithsquarerootfunctioninthischapterissometimesreferredtoastheDARA-borrowertoavoidconfusionwiththeCARA-borrower;otherwise,"borrower"inthischapterreferstotheDARA-borrower. TheCARA-borrowerconsumptionandinvestmentanalysesinChapter3showtheimportanceoftheborrower'spersonalwealth.Inparticular,thewealthleveldeterminestheborrower'spotentialinvestmentstrategyandwhethertheloancontractisfeasible.However,oncedierentinvestmentstrategies(i.e.,borrow,no-lenderinvestment, 60

PAGE 61

TheDARA-borrowerfacessimilarinvestmentdecisionsastheCARA-borrower,hehasthechoiceofborrowingfromthelender,engaginginno-lenderinvestment,andconsuminghispersonalwealthandforgoingtheinvestmentproject.Theinvestmentproject,ifengaged,againprovidesabinarycashowX2fXH;XLg.Thelenderisriskneutralandbreaksevenfromherinvolvementintheproject.TheborrowerisexpectedtoworkdiligentlybecauseAssumption1inChapter3appliesforthischapteraswell,whichindicatesthathigheortisalwayspreferredinequilibrium.LetbCandbUdenotetheborrower'saggregatemoneyholdingsacrossperiodsandhisutilityfunction,respectively.Therisk-averseborrower'sutilityisexpressedasbU=bUbCa=q bCa 61

PAGE 62

Thischapterproceedsasfollows.InSection4.1therst-bestcontractandno-lenderinvestmentfortheDARA-borrowerwithaggregateandtime-additiveconsumptionpreferencesarediscussedasbenchmarks.Section4.2characterizesthesecond-bestloanagreementswhentheborrower'seortsupplyisunobservabledependingonhisconsumptionpreferenceandidentiesthesecond-bestloancontractassociatedwithpotentialinvestmentdistortions.Section4.2alsoprovidescomparisonsbetweentheDARA-andtheCARA-borrowerintheirinvestmentbehaviorandtheincidenceofinvestmentdistortion.ThischapterconcludeswithSection4.3,whichprovidesadiscussionoftheeectoftheborrower'sconsumptionpreferenceonhisoptimalinvestmentpolicy.1 62

PAGE 63

bCHaH+(1pH)q bCLaHs.t.pHbCH+(1pH)bCLW+H [ProgramFBDA]maxbCA00;bCAHaH;bCALaHq bCA0+pHq bCAHaH+(1pH)q bCALaHs.t.bCA0+pHbCAH+(1pH)bCALW+H Therst-bestcontractsprovidethefollowingoptimalconsumptionfortheborrower. Intheabsenceofthelender,whentheborrowerhassucientpersonalwealthhehastheoptionofengaginginself-nancing.Theborrowerwhohasaggregateconsumptionpreferencemaximizeshisexpectedutilityfromself-nancing,whichleadstothefollowingprogram: 63

PAGE 64

bCHaH+(1pH)q bCLaHs.t.bCHWI+XHbCLWI+XL However,fromthelimitedliabilityconstraint,itisrequiredthatbCLaH 4{1 )and( 4{2 ),itisrequiredthatpHp 64

PAGE 65

LetCE(X)denotetheborrower'scertaintyequivalentconsumptionwhenW=I,itfollowsthatp 4{3 )inProposition 7 isnotnecessaryfortheborrowertopreferno-lenderinvestmenttonoinvestment.Thereasonisthat( 4{3 )isderivedundertheassumptionthatW=I;therefore,thefailuretosatisfy( 4{3 )doesnotimplytheoccurrenceofinvestmentdistortion. ComparingtheDARA-borrowerwiththeCARA-borrowerintheirno-lenderinvestmentstrategies,itturnsoutthattheconditionwhichleadsbothtypesofborrowertopreferself-nancingisquitesimilar.LetCE(X)betheborrower'scertaintyequivalentconsumptionforW=I,forbothtypesofborrower,thefollowingcorollarycanbedeveloped:

PAGE 66

5 isalsoanecessaryconditionfortheCARA-borrowerbutnotnecessaryfortheDARA-borrower. Next,thepotentialinvestmentdistortioncausedbytheborrower'sunwillingnesstoengageinno-lenderinvestmentisinvestigatedwhentheborrower'sconsumptionpreferencebecomestime-additive. Whentheborrowerhastime-additiveconsumptionpreference,hesolvesthefollowingproblemtojustifyhisno-lenderinvestmentdecision: [ProgramNLDA]maxbCA00;bCAHaH;bCALaHq bCA0+pHq bCAHaH+(1pH)q bCALaHs.t.bCA0WIbCAHWIbCA0+XHbCALWIbCA0+XL 66

PAGE 67

4{4 )with( 4{3 )impliesthattheborrowerrequireshigherCE(X)toengageinno-lenderinvestment(twicetobeexact).Thisisconsistentwiththefactthattheborrowerwithtime-additiveconsumptionpreferencehasintertemporalconsumptionconcern,whichleadstoamoreconstrainedproblem. TheDARA-borrowerinthischapterisassumedtohavedomain-additivepersonalcoststructure.Thiscoststructuremakesitquitediculttoderiveclosed-formsolutionsoftheoptimalconsumptionwithoutimposingfurtherassumptions;however,itisconsistentwiththeCARA-borrowerinChapter3inthecostsphere.Furthermore,thelackofclosed-formexpressionforthesecond-bestloancontractconsumptionhinderstheanalysisofinvestmentdistortionassociatedwiththesecond-bestloanagreement.Therefore,inthissectionnumericalexamplesareappliedtofacilitaterevealingtheinsightsofthesecond-bestloancontractandpotentialinvestmentdistortion. 67

PAGE 68

bCHaH+(1pH)q bCLaHs.t.pHbCH+(1pH)bCLW+H bCH+(1pL)q bCLpHq bCHaH+(1pH)q bCLaH Theborrower's( 4{6 )isbindinginequilibrium,otherwise,therst-bestconsumptionapplywhichleadstodirectlyviolationof( 4{6 )constraintgivenpH>pL.FollowingChapter3,thelenderbreaksevenandallexpectedcashowsfromtheinvestmentprojectaccruetotheborrower;therefore,theborrowerconsumesatmostthesumofhispersonalwealthathandandtheexpectedcashowsfromtheinvestment.Thisimpliesthatthe( 4{5 )constraintisalsobindinginequilibrium.Then,theoptimalconsumptionisderivedfromthefollowingsetofequations:pHbCH+(1pH)bCL=W+HpLq bCH+(1pL)q bCL=pHq bCHaH+(1pH)q bCLaH 68

PAGE 69

Intuitively,personalwealthisofimportanceintheincidenceofinvestmentdistortionassociatedwiththesecond-bestloancontract.TheCARA-borrower'sriskpremiumfortheloanagreementisinvariantwithhispersonalwealthholdings.However,theDARA-borrower'sriskpremiumdecreaseswhenhegetswealthier.Therefore,ceterisparibus,thepotentialinvestmentdistortionislessseverefortheDARA-borrowerashispersonalwealthgrows. LetdRPandRPdenotetheriskpremiumoftheDARA-andtheCARA-borrower'sinthesecond-bestloancontract,respectively.Furthermore,letdEUSB(EUSB),dRU(RU),cID(ID)denotetheDARA(CARA)-borrower'sexpectedutilityfromthesecond-bestloancontract,hisreservationutility,whichisdenedashisutilityfromconsuminghispersonalwealth,andwhetherinvestmentdistortionispresent.TheparametervaluesfollowthoseinChapter3exceptforXH=310.2 4-1 comparestheincidenceofinvestmentdistortionoftheDARA-borrowerwiththoseoftheCARA-borrower.Ingeneral,Table 4-1 indicatesthatinvestmentdistortioncanbemitigatedbychangingtheborrower'sriskpreference.Inparticular,whentheborrower'spersonalwealthisW=50,Case1impliesthatnosolutionexistsfortheDARA-borrower;therefore,investmentdistortionispresent.Furthermore,eventhoughtheloanagreementisfeasiblefortheCARA-borrower,investmentdistortionstilloccursbecauseborrowingistooriskyfortheCARA-borrowertoengagein.Therefore,Case 4-1 69

PAGE 70

InvestmentDistortionforAggregateConsumption W aH EUSB RU ID 50 85 N=A 49:3 135 95 11:5 19:3 11:6 78:8 200 95 14:4 8:7 14:1 78:8 TheDARA-borrowerwithtime-additiveconsumptionpreferencesolvesthefollowingoptimizationproblem: 70

PAGE 71

bCA0+pHq bCAHaH+(1pH)q bCALaHs.t.bCA0+pHbCAH+(1pH)bCALW+HpLq bCAH+(1pL)q bCALpHq bCAHaH+(1pH)q bCALaH dCE1aH=pHq bCAHaH+(1pH)q bCALaHdRP1=EbC1dCE1

PAGE 72

4-2 followthoseinTable 4-1 .Let4dCE(4CE)and4\NCEdenotetheDARA(CARA)-borrower'sdierenceinlocalcertaintyequivalentconsumptionandtheDARA-borrower'sdierenceinnetlocalcertaintyequivalentconsumption,respectively. TheitaliccolumninTable 4-2 indicatesthattheDARA-borrower'st=1localriskpremiumsystematicallydecreaseswiththeincreaseofhispersonalwealthholdings.However,fortheCARA-borrower,hislocalriskpremiumisinvariantwithhispersonalwealthholdings.Specically,forCARA-borrower,anyincreaseofhispersonalwealthisdividedequallyintotwoparts,thersthalfgoestohist=0consumptionC0andthesecondhalfgoestohisexpectedt=1consumption,whichisE(C1).3;4Furthermore,theCARA-borrower'slocalcertaintyconsumptiondierencesareequalandindependentofhispersonalwealthholdings.Thisisreectedbythefactthat However,fortheDARA-borrower,smoothingutilityisnotpresent.Thiscanbeseenfromthefactthat4dCE=dCE1dCE06=aH 3NotethatequallyallocatingwealthacrossperiodsisonlyfeasiblefortheCARA-borrowerwheninteriorsolutionispresent.4Theborrower'sconsumptionatt=0isnotshowninthetable. 72

PAGE 73

LocalCertaintyEquivalentConsumptionsandLocalRiskPremium DARA CARA E(bC1) E(C1) 152:6 137:3 67:9 119:4 102:6 16:8 60 51 152:7 137:5 67:2 119:9 106:1 16:8 60 isdecreasingwiththegrowthofhispersonalwealthholdings.ThisimpliesthattheDARA-borrower'srequiredlocalcertaintyequivalentconsumptionatt=1,wheretheriskpremiumisnecessarytomotivatehimtosupplyhigheort,isdecreasinginmagnitudewithrespecttohislocalcertaintyequivalentconsumptionatt=0,wherenoriskpremiumispresent.ThismeansthattheDARA-borrowerrequireslessriskpremiumwhenhispersonalwealthholdingsgrow,whichcorrespondstothelocalriskpremiumdRP1column. Insum,whentheborrowerhastime-additiveconsumptionpreference,theDARA-borrowermoveshisconsumptionforwardintimetoreducethelocalriskpremiumassociatedwithunobservableeortsupplyatt=1ashispersonalwealthgrows,whiletheCARA-borrowerhasincentivestoequallyallocatetheincreaseofhispersonalwealthacrosseachconsumptionperiodstosmoothintertemporalconsumptionandretainstheunchangedlocalriskpremium. AlthoughtheDARA-borrower'soptimalconsumptionforthesecond-bestloanagreementisdiculttoderivebecauseofthedomain-additivecostpreference,thefollowingnumericalisdesignedtoshowthedierenceinpotentialinvestmentdistortionfortheDARA-andtheCARA-borrower,respectively. LetaH=80andkeeptheotherparametersthesameasinthepreviousnumericalexamples.Table 4-3 comparestheincidentsofinvestmentdistortionfortheDARA-andCARA-borrowerwhenbothhavetime-additiveconsumptionpreference.DenotedEUASB(EUASB),dRPA(RPA),dRUA(RUA),andcIDA(ID)theDARA-(CARA-)borrower's 73

PAGE 74

InvestmentDistortionforTime-AdditiveConsumption W 100 13:1 21 14:1 39:8 150 17:9 15:6 17:3 39:8 InTable 4-3 ,Case1impliesinvestmentdistortionforbothriskpreferenceswhileCase2displaysadierence.First,theCARA-borrowerstillexperiencesinvestmentdistortioneventhoughhispersonalwealthhasgrown,thisisconsistentwiththeconclusioninChapter3thattheCARA-borrower'sriskpreferencedoesnotvarywithhispersonalwealthholdings;therefore,giventhathehasinvestmentdistortiontobeginwith,hewouldencounterthesameinvestmentdistortionregardlessofthechangeinthemagnitudeofhispersonalwealth.However,asthepersonalwealthincreases,theDARA-borrowernolongerexperiencesinvestmentdistortion;thisisbecausehisattitudetowardsriskchangeswiththeincreaseofhispersonalendowment.AstheDARA-borrowerbecomeswealthier,heislessrisk-averse,whichcanbeinferredfromthedecreasingriskpremium;thisfollowsthatheismorelikelytoundertaketheloanarrangementthatwouldotherwisebetooriskywerehetohavelesspersonalwealth. 74

PAGE 75

LettheDARA-borrower'spersonalwealthbexedatW=140.Table 4-4 compareshisinvestmentdecisionsbetweenborrowingandno-investmentwhenhehaseitheraggregateortime-additiveconsumptionpreference. Giventhesamepersonalwealthlevel,Table 4-4 clearlyshowsthatinvestmentdistortionfortheborrowerwithaggregateconsumptionpreferenceimpliesinvestmentdistortionforthesameborrowerbutwithtime-additiveconsumptionpreference.Whentheborrower'spersonalcostofeortisaH=80,noinvestmentdistortionispresentforeitherconsumptionpreference.However,asaHgrowsto98,investmentdistortionoccursforbothconsumptionpreferences.Moreover,thepotentialinvestmentdistortion 75

PAGE 76

Second-BestContractsInvestmentDistortionfortheDARA-Borrower 12:8 11:8 16:7 12:2 11:8 16:7 11:5 11:8 16:7 Iftheborroweriswealth-abundantandself-nancestheinvestmentproject,similarlogicfollows.Inotherwords,sincetheborrowerwithtime-additiveconsumptionpreferencefacesmorehurdleswithrespecttotheinvestmentbehavior,thepotentialinvestmentdistortionproblemisworsenedastheconsumptionpreferenceemphasizesintertemporalconsumption. 76

PAGE 77

BothChapter3andChapter4discusstheborrower'sconsumptionandinvestmentbehaviorundertheassumptionthattheborrower'spersonalwealthispubliclyobservablebuteortsupplyisunobservable.Thischapterrelaxesthisassumptionandfocusesonthescenariowheretheborrower'spersonalwealthisprivateinformationwhilehiseortsupplyisalsoprivate.Thentheproblembecomesanentanglementofmoralhazardandadverseselection. TheanalysisisfocusedontheCARA-borrowerwithaggregateconsumptionpreference.Thetimelineisrevisedasfollows:attimet=0,amenuofcontractswhichspeciestheCARA-borrower'sconsumptionandthelender'sinvestmentandpayofromtheprojectisoered.TheCARA-borrowersubsequentlyobserveshispersonalwealthW2fWH;WLgwhilethelenderremainsuninformed.However,thelenderrequirestheCARA-borrowerreporthispersonalwealthfW2nfWH;fWLoandpostanup-frontbondintheamountofhisreport.LetW 77

PAGE 78

WeusethefollowingnotationtodenotethechoicevariablesintheCARA-borrower'soptimizingprogram. wherefW2nfWH;fWLo. ( 1985 )and LaontandMartimort ( 2001 )havedetaileddiscussesofRevelationprinciple. 78

PAGE 79

wherefW2nfWH;fWLo. OncetheCARA-borrowerknowshiswealthtype,hemakesreportingandeortsupplychoices.LetEUW;fW;adenotetheCARA-borrower'sexpectedutilitywhenheobserveshispersonalwealthtobeW2fWH;WLg,butchoosestoreportfW2nfWH;fWLoandsupplyaneortofa2faH;aLg.Foreachofhiswealthtypes,hehasthefollowingfeasiblestrategiesandtheassociatedexpectedutilities: Type Report Eort 79

PAGE 80

InChapter3,theCARA-borrower'sconsumptionisupper-boundedbythesumofhisremainedwealthafterputtinghisshareofinvestmentintheprojectandthecashowfromtheinvestmentnetthelender'spayo.Theseconsumptionconstraintsalsoapplyinthischapter: (5{6)CLfWHfWHhID0fWHi+XLDL(fWH) (5{7)CHfWLfWLhID0fWLi+XHDH(fWL) (5{8)CLfWLfWLhID0fWLi+XLDL(fWL) (5{9) TheCARA-borrowerwithprivateinformationthensolvesthefollowingproblemwithrespecttotheloancontract: [ProgramgSB]maxC(fW)0;D0(fW);D(fW)EUWRPjaHs.t.( 5{1 )( 5{2 )-( 5{5 )( 5{6 )-( 5{9 ) 80

PAGE 81

Denethe"naivecontract"astheoptimalloancontracttheCARA-borrowerwouldachieveifhispersonalwealthwerepublic;i.e.,theoptimalsecond-bestloanagreementinChapter3isthenaivecontract.Thefollowingconclusionisdevelopedfor[ProgramgSB]: 3{8 inChapter3).WhentheCARA-borrowerhasprivateinformationabouthispersonalwealth,hisreporting 81

PAGE 82

LettheCARA-borrower'spersonalcostofeortbeaH=30andhisbinarypersonalwealthtakethevalueofeitherWL=140orWH=300.TherestoftheparametervaluesfollowthoseinChapter3.2TheprobabilityoftheCARA-borrowerbeingwealthy,i.e.,hehasapersonalwealthof300isassumedtobe=:7.Theoptimalcontractfor[ProgramgSB]showsasolution:CHfWH=456andCLfWH=403CHfWL=296andCLfWL=243 withanexpectedutilityof:0374.ThisconsumptionschedulecoincideswiththenaivecontractinChapter3whentheCARA-borrower'spublicpersonalwealthis300and140,respectively.Furthermore,notethattheconsumptiondierencesbetweenreportingtruthfully(i.e.,300)andmisreporting(i.e.,140)isthedierencebetweenthereportedwealth(i.e.,300140=160).Therefore,byignoringtheCARA-borrower'sprivatepersonalwealthsourceandoeringhimtheconsumptionschedulethathewouldachieveifhiswealthwerepublic,theCARA-borrowerhasnoincentivetomisreporthiswealth. Chapter3demonstratesinvestmentdistortionariseswhen1)nosolutionexistsfortheloancontractand2)theloanagreementistooriskyandtheCARA-borrowerprefersthestatusquo.SincetheCARA-borrower'sconsumptionbehaviorfollowsthatinChapter3,thecorrespondingpotentialinvestmentdistortionproblemisinvarianttowhetherprivateinformationispresent. 82

PAGE 83

InChapter4,wendthatitisdiculttoderiveclosed-formconsumptionfortheDARA-borrowerbecauseofthedomain-additivecoststructure,thesamedicultyapplieshere.UnliketheCARA-borrowerwhoseriskpreferencestaysthesameregardlessofhispersonalwealth,theDARA-borrowerchangeshisattitudetowardsriskbasedonhiswealth.Inparticular,themorewealthiertheDARA-borroweris,thelessrisk-averseheis,whichimpliesthattheriskpremiumassociatedwiththeloancontractisdecreasingashispersonalwealthincreases.Therefore,thelesspersonalwealththeDARA-borrowerreports,themoreriskyloanagreementhefaces.Thisimpliesthatitisnotinthe 83

PAGE 84

Theborrower'smisreportingstrategyisone-directionalinthischapter.However,over-reportingmaybevaluabletotheborrower.Inparticular,whentheborrowerobserveshispersonalwealthtobelessthanW 84

PAGE 85

Thisdissertationcharacterizesaborrower'sinvestmentandconsumptionbehaviorswithrespecttoaninvestmentprojectandinvestigateswhethertheborrower'sconsumptionpreferenceaectstheinvestmentdecision.Itisconcludedthattheprojectmaybeinecientlyabandoned,whichresultsininvestmentdistortion,andtheinvestmentdistortionproblemisworsenedwhentheborrowerfocusesmoreontheintertemporalconsumptionthanonaggregateconsumption.Theborrower'sinitialwealthendowmentisimportantindeterminingtheformofpotentialinvestmentdistortion.Forthewealth-constrainedborrower,thejointforceofupper-boundedexpectedprotfromtheinvestmentandthelimitedliabilityconstraintsdeterminesthefeasibilityoftheloanarrangementbetweentheborrowerandthelender,which,inturn,isconclusiveaboutthepossibilityofinvestmentdistortion.Theborrower'sconsumptionpreferencehasnoeectontheoccurrenceofinvestmentdistortion.However,thewealth-abundantborrowernotonlyfocusesontheforcenotedabovebutalsoweighsthecertaintyequivalentcashowoftheinvestmentandtheentirepersonalcostofimplementingsuchinvestment,whichistheriskinessoftheinvestmentitself,todeterminewhethertheprojectisundertaken.Theborrower'sintertemporalconsumptionpreferenceisnolongerirrelevantintheinvestmentdecision.Theconditionsunderwhichinvestmentdistortionarisesfortheborrowerwithaggregateconsumptionpreferenceimplythosefortheborrowerwithtime-additiveconsumptionpreferencebutnotviceversa,whichsuggeststhattheinvestmentdistortionproblemisexacerbatedwhenthewealth-abundantborrowerhasintertemporalconsumptionstructure. However,astheborrower'sriskpreferencechanges,theassociatedpotentialinvestmentdistortionproblemdiers.TheDARA-borrowerexperienceslesssevereinvestmentdistortionthantheCARA-borrowerbecausetheincreaseinhispersonalwealthmakeshimlessriskaverse;therefore,otherthingsbeingequal,theloancontractbetween 85

PAGE 86

WhentheCARA-borrowerhasprivateinformationabouthispersonalwealth,theRevelationPrincipleapplieswhichmotivatestheCARA-borrowertotruthfullyreporthisprivateinformation.AnaivecontractwhichignorestheCARA-borrower'sprivatesourceofinformationandtreatsitaspublicinformationisafeasibleandoptimalloanagreement.TheprivateinformationcaseoftheDARA-borrowerremainsanopenissue,thoughthenaivecontractretainsattractivetruthincentives. Inclosing,afewissuesareraised.Itisassumedthattheborrowerhasallthebargainingpowerwhiletheroleofthelenderisquitepassive.Inreal-worldsituations,theborrower,especiallythewealth-constrainedborrower,tendstobethepartythatdoesnothavemarketpowerinthenancialmarket.Therefore,whetherandhowtheresultsofthisdissertationchangeunderalternativebargainingstructuresisstillanopenbutinterestingquestion.Second,thepossibilityofover-reportingwhentheborrower'spersonalwealthisprivatelyobservedwarrantsinvestigation.Over-reportingstrategyinthepresenceofprivateinformationisparticularlyvaluabletothewealth-constrainedborrower,byborrowingtoposttherequiredup-frontbond,theborrowerisabletosecuretheloanagreementandimplementtheinvestmentproject. 86

PAGE 87

1 andCorollary 1 where=pHexp(aH)pL RearrangingtermsoftheconstraintsprovidesCHCL+1 (A-3)CHW+H(1pH)CL IfW+HpH (A-5) GiventhelimitedliabilityconstraintthatCL0,if( A-5 )holds,theconstraintsetof( A-3 )and( A-4 )isneversatised;whichbycontradiction,suggeststhatthereexistsnofeasiblesolutionifW+HpH Therefore,[ProgramSB]hasfeasiblesolutionif (A-6) Assumecondition( A-6 )holds,thefollowinganalysisshowstheexistenceofthesolutionanditscharacterizations. Notethattheobjectivefunctionin[ProgramSB]isconcaveandcontinuous,givencondition( A-6 ),theconstraintsetiscompact,thenbasedontheWeierstrassTheorem,thereexistsasolutionfor[ProgramSB]. 87

PAGE 88

A-1 )isnotbinding,thelenderisgettingrentfromherrepayment,whichcontradictsacompetitivenancialmarketassumption.If( A-2 )isnotbinding,therst-bestcontractconsumptionscheduleCH=CLapplies,whichleadstoaviolationof( A-2 )foranyaH>0.Thenthesolutionfor[ProgramSB]issolvedfromthefollowingsetofequationspHCH+(1pH)CL=W+HCL+1 (A-7)CL=W+HpH (A-8) Withsolution( A-7 )and( A-8 ),therst-orderconditionswithrespecttoCHandCLarebothzero,whichimpliesthat( A-7 )and( A-8 )areinteriorsolutions.QED. 2 1 andCorollary 1 Partb.LetCESBandCERUdenotetheborrower'scertaintyequivalentconsumptionfromthesecond-bestcontractandfromconsuminghispersonalwealthathand,respectively.GivenProposition 1 'soptimalsecond-bestcontract,CESBandCERUarecharacterizedasCESB=W+HpH (A-9)CERU=W where=pHpL 88

PAGE 89

IfHaH
PAGE 90

(A-11) whereCE(X)=1 GivenCERU=W 2 ,wecompareno-lenderinvestmentwithno-investment.Notethatthepersonalcostofeortisconsideredfortheformer.Then,ifCE(X)CESBaH>CENLaH 4 andCorollary 3

PAGE 91

where=pHexp(aH)pL RearrangingtermsoftheconstraintsprovidesCAHCAL+1 (A-14)CAHW+HCA0(1pH)CAL IfW+HpH (A-16) GiventhelimitedliabilityconstraintthatCA0;CAL0,if( A-16 )holds,theconstraintset( A-14 )and( A-15 )isneversatised,whichbycontradiction,suggeststhatthereexistsnofeasiblesolutionifW+HpH Therefore,[ProgramSBA]hasfeasiblesolutionif (A-17) Assumecondition( A-17 )holds,thefollowinganalysisshowstheexistenceofthesolutionanditscharacterizations. Notethattheobjectivefunctionin[ProgramSBA]isconcaveandcontinuous,givencondition( A-17 ),theconstraintsetiscompact,thenbasedontheWeierstrassTheorem,thereexistsasolutionfor[ProgramSBA]. Notethat>1;itthenfollowsfromtheborrower'sICAconstraintthatCAHCAL+1 91

PAGE 92

Inequilibrium,bothconstraintsin[ProgramSBA]arebinding.If( A-12 )isnotbinding,thelenderisgettingrentfromherrepayment,whichcontradictsacompetitivenancialmarketassumption.If( A-13 )isnotbinding,therst-bestcontractconsumptionscheduleCA0=CAHaH=CALaLapplies,whichleadstoaviolationof( A-13 )foranyaH>0.Therefore,wehaveCAL+1 SubstitutingCAHin( A-18 )into( A-19 )andsolvingforCA0 Given( A-18 )and( A-20 ),theborrower'speriodt=0expectedutilityis Substituting,theborrower'sproblembecomes [ProgramSBA0]maxCAL0expW+HpH

PAGE 93

(A-23)CALL0CAL=0 (A-24) where( A-24 )isthecomplementaryslacknesscondition.Solving( A-23 )and( A-24 )yieldstheoptimalsolutionofCAL: IfW+HpH 2;L0CAL=0 (A-25) Otherwise (A-26) where=pHpL Therefore,( A-25 )and( A-26 )aretheinteriorandcornersolutionsforCAL,respectively.SubstitutingCALin( A-25 )into( A-21 )and( A-22 ),( )=( ) Moreover,substitutingCALin( A-25 )into( A-19 )and( A-20 ),optimalconsumptionfor[ProgramSBA]isCA0=W+HpH 2CAH=W+HpH 2+1 2 NotethataH+1 A-26 )into 93

PAGE 94

A-19 )and( A-20 ),theoptimalcornerconsumptionfor[ProgramSBA]isCA0=W+HpH Notethatfortheaboveconsumptionschedule,given( A-17 ),theborrower'slimitedliabilityconstraintsareallsatised.Moreover,therst-orderconditionwithrespecttoCA0for[ProgramSBA]is0;whichimpliesthattheoptimalCA0isaninteriorsolution.QED. 4 a. b. HaH
PAGE 95

Whenthesecond-bestloancontractprovidescornerconsumption,giventheconcavityoftheborrower'sutilityfunction,hisexpectedutilityisstrictlylowerthanthatunderinteriorconsumption.Therefore,ceterisparibus,ifHaH1and<1,thenpH (A-32) However,given( A-32 ),inequalities( A-30 )and( A-31 )cannotcoexist.ThisimpliesthatgivenHaHpH Forthesecond-bestinteriorconsumption,investmentdistortionisnotpresentwhenexp(CA0)exp(W thenbasedonCorollary 3 and( A-29 ),exp(CA0)pHexpCAHaH(1pH)expCALaH2exp(W whichindicatesaabsenceofinvestmentdistortion.QED. 95

PAGE 96

5 4 Thefollowingderivestheconditionsunderwhichtheborrowerprefersno-investmenttono-lenderinvestment.Theborrower'sno-lenderinvestmentprogramwhenhehastime-additiveconsumptionpreferenceis [ProgramNLA]maxCA00;CAH0;CAL0exp(CA0)pHexpCAHaH(1pH)expCALaHs.t.CA0WICAHWCA0I+XHCALWCA0I+XL (A-33)LCAL=(1pH)expCALaH3+6=0 (A-34)Rearrangingtermsfor( A-33 )and( A-34 ),givennon-negativemultipliers,2;3>0 Thentheborrower'sconsumptionconstraintsatt1arebothbinding,i.e.,CAH=WCA0I+XH 96

PAGE 97

SubstitutingthebindingCAHandCALintotheobjectivefunction,theprogrambecomesmaxCA00exp(CA0)+expCA0exp[(WIaH)]EU(X)s.t.CA0WI whereEU(X)=(1pH)exp(XL)pHexp(XH).TakingthederivativeandsolvingforCA0, 2 (A-38) whereCE(X)=1 Given( A-37 ),( A-38 )onlyappliesif Otherwise,for itfollowsthatCA0=WI A-39 )holds,given( A-35 ),( A-36 ),and( A-38 )theborrower'soptimalconsumptionisderivedasCA0=WIaH+CE(X) 2CAH=WI+aHCE(X) 2+XHCAL=WI+aHCE(X) 2+XL

PAGE 98

2RUA=2expW 2
PAGE 99

A-40 )hastohold,i.e.,CE(X)+IaH>W.Summarizingcase2impliesthatforXL0,investmentdistortionarisesif 2WI+aH,nosolutionexistsfortheborrower'sno-lenderinvestmentbehavior.WhenCE(X)WI+aH,investmentdistortionanalysisfollowstheanalysisofCase1above. Second,ifXL
PAGE 100

NotethatatW=I,theborrowerconsumesthestochasticcashowsfromhisinvestment;therefore,letCE(X)denotetheborrower'scertaintyequivalentconsumptionatW=I,thenitfollowsthatp ForanyWI,giventheborrower'sdecreasingabsoluteriskaversionpreference,itisalwaysthecasethatif( A-41 )holds,thefollowingrelationshipistrue:EUDNL(WI)p

PAGE 101

A-42 )isthesucientconditionfortheborrowertoengageinno-lenderinvestmentovernoinvestmentintheabsenceofthelenderforanyWI.QED. 5 :DirectlyfollowstheProofofProposition 3 inChapter3andtheProofofProposition 7 above.QED. 8 : whichcanbesimpliedasp whereCE(X)followsthedenitionintheProofofProposition 7 anddenotestheborrower'scertaintyequivalentconsumptionatt=1.Therefore,atW=I,investmentdistortionarisesif 101

PAGE 102

A-43 )issucienttoavoidinvestmentdistortionintheabsenceofthelender.QED. 6 : [ProgramSBD]maxbCHaH;bCLaHpHq bCHaH+(1pH)q bCLaHs.t.pHbCH+(1pH)bCL=W+HpLq bCH+(1pL)q bCL=pHq bCHaH+(1pH)q bCLaH bCHaH+(1pH)s TakingthetotalderivativewithrespecttoaHyieldsdobj daH=1 224pH1 bCHaH+(1pH)1 2pH0@1 bCHaH1 daH<0 102

PAGE 103

A-44 )withrespecttoWimpliesthatdobj dW>0 QED. 9 AndtheCARA-borrower'sexpectedutilitiesfromeachreportingandeortsupplystrategyare: 103

PAGE 104

(A-47)CLfWHfWHhID0fWHi+XLDL(fWH) (A-48)CHfWLfWLhID0fWLi+XHDH(fWL) (A-49)CLfWLfWLhID0fWLi+XLDL(fWL) (A-50) Notethatfrominequalities( A-47 )and( A-48 ),thefollowingrelationshipholds: 104

PAGE 105

A-45 ),( A-51 )becomes Similarly,given( A-49 )and( A-50 ),togetherwith( A-46 ),wehave Therefore,theCARA-borrower'sconsumptionconstraints( A-47 )to( A-50 )andthelender'sindividualrationalityconstraints( A-45 )and( A-46 )becometheCARA-borrower'sexpectedconsumptionconstraints( A-52 )and( A-53 ). IgnoringtheCARA-borrower'slimitedliabilityconstraints,let160betheLagrangianmultipliersfortheCARA-borrower'sfourincentivecompatibilityconstraintsandtheexpectedconsumptionconstraints( A-52 )and( A-53 ).TakingthederivativesoftheLagrangianfunctionwithrespecttotheCARA-borrower'sconsumptionwhenhighwealthisreportedCfWH,wehave (1pH)=5 (1pH) However,thisiscontrarytotheassumptionthatpH>:5;therefore,5>0.Giventhat5>0,if2=0,wehaveexphCHfWHiexp(aH)(+1+3)=5exphCLfWHiexp(aH)(+1+3)=5

PAGE 106

where=pHexp(aH)pL Similarly,takingthederivativesoftheLagrangianfunctionwithrespecttotheCARA-borrower'sconsumptionwhenlowwealthisreportedCfWL,wederivethat4;6>0,whichimpliesthatthefourthincentivecompatibilityconstraintand( A-53 )arebinding:EUWL;fWL;aH=EUWL;fWL;aLpHCHfWL+(1pH)CLfWL=fWL+H 106

PAGE 107

GiventhatfWH>fWLW Notethattheoptimalsolutionsfor[ProgramgSB]areexactlythesameastheoptimalconsumptionfromthenaivecontractwhentheCARA-borrower'spublicobservablepersonalwealthareWHandWL,respectively.Therefore,thenaivecontractisoptimalfor[ProgramgSB].QED. 107

PAGE 108

Aboody,D.,Hughes,J.,Liu,J.,2005.Earningsquality,insidertrading,andcostofcapital.JournalofAccountingResearch43(5),651{73. Altman,E.,1968.Financialratios,discriminantanalysis,andthepredictionofcorporatebankruptcy.JournalofFinance23(4),589{609. Bens,D.,Monahan,S.,2004.Disclosurequalityandtheexcessvalueofdiversication.JournalofAccountingResearch42(4),691{730. Bertomeu,J.,Beyer,A.,Dye,R.,2008.Capitalstructure,costofcapital,andvoluntarydisclosure.WorkingPaper. Biddle,G.,Hilary,G.,Verdi,R.,2008.Howdoesnancialreportingqualityimproveinvestmentecientcy?WorkingPaper. Breeden,D.,1979.Anintertemporalassetpricingmodelwithstochasticconsumptionandinvestmentopportunities.JournalofFinancialEconomics7(3),265{96. Bushman,R.,Piotroski,J.,Smith,A.,2006.Capitalallocationandtimelyaccountingrecognitionofeconomiclosses.WorkingPaper. Che,Y.,Gale,I.,2000.Theoptimalmechanismforsellingtoabudget-constrainedbuyer.JournalofEconomicTheory92(2),198{233. Christensen,P.,Feltham,G.,2005.EconomicsofAccounting.PerformanceEvaluation.SpringerScienceBusinessMedia,Inc. Danthine,J.,Donaldson,J.,2005.IntermediateFinancialTheory.ElsevierAcademicPress. Deaton,A.,1987.Life-cycleModelsofConsumption:IstheEvidenceConsistentwiththeTheory? Francis,J.,LaFond,R.,Olsson,P.,Schipper,K.,2000.Themarketpricingofaccrualsquality.JournalofAccountingandEconomics39(2),295{327. Gale,D.,Hellwig,M.,1985.Incentive-compatibledebtcontracts:Theone-periodproblem.ReviewofEconomicStudies52(4),647{63. Grossman,S.,Hart,O.,1983.Ananalysisoftheprincipal-agentproblem.Econometrica51(1),7{45. Hall,R.,1988.Intertemporalsubstitutioninconsumption.JournalofPoliticalEconomy96(2),339{57. Harris,M.,Townsend,R.,1985.AllocationMechanism,AsymmetricInformationandthe"RevelationPrinciple". 108

PAGE 109

Innes,R.,1990.Limitedliabilityandincentivecontractingwithex-anteactionchoices.JournalofEconomicTheory52(1),45{67. Koopmans,T.,1960.Stationaryordinalutilityandimpatience.Econometrica28(2),287{309. Laont,J.,Martimort,D.,2001.TheTheoryofIncentives:ThePrincipal-AgentModel.PrincetonUniversityPress. Lewis,T.,Sappington,D.,2000.Contractingwithwealth-constrainedagents.InternationalEconomicReview41(3),743{67. Lewis,T.,Sappington,D.,2001.Optimalcontractingwithprivateknowledgeofwealthandability.ReviewofEconomicStudies68(1),21{44. Lintner,J.,1965.Thevaluationofriskyassetsandtheselectionofriskyinvestmentinstockportfoliosandcapitalbudgets.ReviewofEconomicStudies47(1),13{37. Lu,R.,2009.Doesdisclosureofnon-nancialstatementinformationintheannualreportsrelaxrms'nancingconstraints?WorkingPaper. Merton,R.,1973.Anintertemporalcapitalassetpricingmodel.Econometrica41(5),867{87. Myers,S.,Majluf,N.,1984.Corporatenancingandinvestmentdecisionswhenrmshaveinformationthatinvestorsdonot.JournalofFinancialEconomics13(2),187{221. Myerson,R.,1981.Optimalauctiondesign.MathematicsofOperationsResearch6(1),58{73. Sappington,D.,1983.Limitedliabilitycontractsbetweenprincipalandagent.JournalofEconomicTheory29(1),1{21. Sharpe,W.,1964.Capitalassetprices-atheoryofmarketequilibriumunderconditionsofrisk.JournalofFinance19(3),425{42. Tirole,J.,2006.TheTheoryofCorporateFinance.PrincetonUniversityPress/PrincetonandOxford. 109

PAGE 110

LiangFuearnedherBachelorofArtsinInternationaleconomicsfromSouthwesternUniversityofFinanceandEconomicsin1998.SheearnedherMasterofArtsineconomicsfromKentStateUniversityin2002.ShejoinedthePh.D.programofFisherSchoolofAccountingattheUniversityofFloridaandshereceivedherdegreeofDoctorofPhilosophyinAccountingin2009. 110