Corporate restructuring: empirical evidence on the approval of the reorganization plan

(1)

Availableonlineatwww.sciencedirect.com

RAUSP

Management

Journal

http://rausp.usp.br/ RAUSPManagementJournal53(2018)49–62

Original

Article

Corporate

restructuring:

empirical

evidence

on

the

approval

of

the

reorganization

plan

Reestrutura¸cão

corporativa:

Evidência

empírica

sobre

a

aprova¸cão

do

plano

de

reorganiza¸cão

Vinicius

Augusto

Brunassi

Silva

a,∗

,

Richard

Saito

b

a_Funda¸cão_Escola_de_Comércio_Álvares_Penteado,_São_Paulo,_SP,_Brazil b_Funda¸cão_Getulio_Vargas,_São_Paulo,_SP,_Brazil

Received8July2016;accepted5November2017 Availableonline16January2018 ScientificEditor:MariaSylviaMacchioneSaes

Abstract

Whenacorporationpresentsareorganizationplan,itexpectsitscreditorstoapprovetheplan.Thispaperprovidesempiricalevidenceregarding thelikelihoodofapprovalbasedonreorganizationplansforcreditorsinBrazilthatrequireapprovalbyemployees;andbysecureandunsecure debtholders.Thispaperinvolvesadescriptiveanalysisofthemaincharacteristicsofreorganizationplansbytypeofvote.Usingasampleof120 reorganizationplansproposedbycorporationsfrom2005to2014,wefindthatthelaborclassofcreditorsislikelytoapprovethereorganization planevenwhentheplanisrejected;planswithmoreheterogeneouspaymentforclassesarelesslikelytobeaccepted;plansarelesslikelytobe acceptedwhentherearemoreunsecurecreditors;andplanswithdivestmentproposalsaremorelikelytobeaccepted.Finally,asexpectedgiven thesenioritypositionofsecureddebt,plansarelesslikelytobeacceptedwhentheportionofsecureddebtishigher,andthereverseistruefor unsecureddebt.

©2017DepartamentodeAdministração,FaculdadedeEconomia,AdministraçãoeContabilidadedaUniversidadedeSãoPaulo–FEA/USP. PublishedbyElsevierEditoraLtda.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).

Keywords:Corporaterestructuring;Reorganizationplan;Bankruptcy

Resumo

Quandoumaempresaapresentaumplanoderecuperaçãojudicial,espera-sequeoplanosejaaprovadoporseuscredores.Nesteartigo, apresenta-seaevidênciaempíricasobreavotaçãodoscredorestrabalhistas,comgarantiarealequirografárioeaprobabilidadedeaprovaçãodoplanode recuperaçãojudicialnoBrasil.Oestudoabordaumaanálisedescritivadasprincipaiscaracterísticasdosplanosderecuperaçãojudicialporclasse devoto.Utilizandoumaamostrade120planosderecuperaçãojudicialapresentadosporempresasentre2005e2014,osresultadossugerem que:credorestrabalhistasestãopropensosaaprovaroplanoderecuperaçãomesmoquandooplanoérejeitadopelasdemaisclasses;planoscom propostasdepagamentomaisheterogêneasparaastrêsclassesdecredorespossuemmenorchancedeseremaceitos;achancedeaprovaçãodo planodiminuinoscasosemquemaiscredoresquirografáriosparticipamdavotação;eplanoscompropostadevendadeativospossuemmaior chancedeseremaprovados.Finalmente,maiorconcentraçãodadívidanaclassecomgarantiarealdiminuiachancedeaprovaçãodoplano,eo contrárioocorrenaclassequirografária.

©2017DepartamentodeAdministração,FaculdadedeEconomia,AdministraçãoeContabilidadedaUniversidadedeSãoPaulo–FEA/USP. PublicadoporElsevierEditoraLtda.Este éumartigoOpenAccesssobumalicençaCCBY(http://creativecommons.org/licenses/by/4.0/).

Palavras-chave:Recuperac¸ãodeempresas;Planoderecuperac¸ão;Falência

∗_{Corresponding}_author_at:_Avenida_da_Liberdade,_532,_CEP_01502-001,_São_Paulo,_SP,_Brazil.

E-mail:vinicius.vitio@gmail.com(V.A.Silva).

PeerReviewundertheresponsibilityofDepartamentodeAdministrac¸ão,FaculdadedeEconomia,Administrac¸ãoeContabilidadedaUniversidadedeSão Paulo–FEA/USP.

https://doi.org/10.1016/j.rauspm.2017.12.008

2531-0488/©2017DepartamentodeAdministração,FaculdadedeEconomia,AdministraçãoeContabilidadedaUniversidadedeSãoPaulo–FEA/USP.Published byElsevierEditoraLtda.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).

(2)

Introduction

Whenacompanyfacesfinancialdistressit maychoose to

deviseareorganizationplan.Suchaplanmustbepresentedtoits

creditors,whoultimatelyvotetoapprovethereorganizationplan

ortosubjectthecompanytobankruptcyproceedings.Thispaper

examinesthisdecisionmakingprocess.Wepresentempirical

evidenceregardingtheapprovalofreorganizationplans,aswe

believethat a cleargap existsinhow each class of creditors

decidesto approveor reject theseplansduringthe creditors’

generalmeeting.Collectively,ourresultsshowthatdebtholders’

behaviordependsontheirclaimrights.Moreover,assetdisposal

facilitatestheapprovalofreorganizationplans.

Theliteratureonlawandfinancestatesthatdebt

restructu-ringcanbeconsideredacomplexdecisionprocessinvolvinga

firmanditslenders,andstressesthatdebtreorganizationplans

andclaimholders’relative recoveriesincourtdependonhow

disputesbetweencreditorsareresolved(Gilson,Hotchkiss,&

Ruback, 2000). Empirical evidence demonstrates the

impor-tanceofbankruptcylawtothecreditmarket andenforcement

bycourts.

The interaction between debtors and representative

credi-torsinsituationsoffinancialdistresshasreceivedconsiderable

attention.KordanaandPosner(1999)studybargainingwith

mul-tiplecreditors,incorporatingtheoperationof thevoting rules

for companies filing Chapter 11. Moreover, Winton (1995),

BoltonandScharfstein (1996),Bris andWelch(2005),Hege andMella-Barral(2005),BisinandRampini(2006),Hackbart, Hennessy,andLeland (2007)andvonThadden, Berglöf,and Roland(2010)allpresentmultiple-creditormodelsconsidering

exantecontractingproblemsorexpostanalysisofthose

prob-lems stemmingfromthe individual andcollectiveliquidation

rightsofcreditors.

We provide an analysis based on Kaplan and Stromberg

(2003)andKaplanandStrömberg(2004)usingdatabases

suffer-ingfromsamplebiasnotpresentindatafromquasi-experiments.

Therefore,wedonotaddresscausalityinourstudy;thispaper

providesadescriptiveanalysis.Toourknowledge,thereisalack

ofempiricalexplanationsofhowcreditorsdecidetovoteon

reor-ganizationplans.Moreover,thisisthefirststudytoexaminethe

likelihoodofacceptanceofreorganizationplansbyconsidering

thedecisionprocessofeachclassofclaimholdersandthe

char-acteristicsoftheplansinBrazil.In2005,Law11,101tookeffect

inBrazilwiththegoalofprovidingcreditorswithbetter

condi-tionsforreorganizingorliquidatingcompaniesfacingfinancial

distress.

AsKordanaandPosner(1999)note,littleattentionhasbeen

devotedtotheexaminationofthecorrespondencebetween

vot-ing rules during reorganizations. von Thadden et al. (2010)

suggestthatmanyconflictsofinterestsaresolvedexpostduring

the bankruptcy or reorganization period. Hence, our

descrip-tiveresultsprovideevidenceonthecharacteristicssurrounding

creditor’sdecisionmakingprocesses.

The literature on reorganization and bankruptcy provides

extensivetheoreticalandempiricalanalysis ofdebt

restructu-ring. From an ex ante perspective, studiestry toexplain the

impacts of bankruptcy on firms’ capital structure decisions.

Moreover,theytrytounderstandwhyfirmsborrowfrom

multi-ple creditors, although it couldmake the futureresolution of

distress more complex. The ex post approach aims to show

the bestalternatives forsortingoutclaimsinsituations where

financialdistresshasalreadyoccurred.

HaugenandSenbet(1978)statethatbankruptcyriskimpacts

firms’capitalstructuredecisions.Similarly,thechoiceofdebt

structure influences what occurs in bankruptcy according to

Aghion, Hart, and Moore (1992). From an ex ante

perspec-tive,severalstudiesanalyzetheimportanceofbankruptcywith

respect todebtors’ investments, leverageandincentives prior

tothebankruptcysituation,suchasCornelliandFelli(1997),

Schwartz(1998),BerkovitchandIsrael(1999),andBebchuck (2002)amongothers. Theseresearchers elucidatetheconflict

betweendebtorsandrepresentativecreditors.Unfortunately,in

thispaperwecannotcontrolforexantevariablesbecausethe

majority of our data donot provide financial statements that

allowthis.

Wepresent empirical evidenceaboutthe characteristics of

voting on reorganization plans by separately conducting the

analysesaccordingtotheoutcomeofthevoteontheplan.Hence,

thispaperinvestigatesaveryimportantissuefromnotonlya

the-oreticalbutalsoamanagerialperspective.Weaimtounderstand

howdifferentclassesofcreditorsvoteinapprovingorrejecting

recoveryplansbycontrollingfortheconditionsspecifiedinthe

reorganization plans.Weanalyzedatacollectedfrom2005to

2014.Weuse2005asthestartingyearbecausethenewBrazilian

bankruptcylawenteredintoforceinthisyear.

Under the new Brazilianbankruptcy law, creditors playa

more importantrole incompanyrestructuring because of the

newvotingprocedureforreorganizationplans.Afterchoosing

torestructureitsdebtincourt,afirmmustcreateareorganization

plan that presents a solution for its financial distress. Unlike

Chapter11intheUSbankruptcycode,Brazilianbankruptcylaw

doesnotrequireaclaimadministratortoorganizeandprovide

informationonallclaimsandclaimholders.IntheUS,indenture

trusteesactonbehalfofcreditors:therefore,claimholdersdonot

meet tovote onreorganizationplans.Althoughcreditors vote

byfollowingadifferentprocedure,Ponticelli(2012)showsthat

similaritiesexistbetweenBrazilianbankruptcylawandtheUS

bankruptcycodeandAnapolskyandWoods(2013)presentmore

detailsabout thesimilaritiesanddifferencesinreorganization

rulesbetweenthetwocountries.

Whenacreditordoesnotapprovethereorganizationplan

pre-sentedbyaspecificfirm,thedifferentclassesofcreditorsdecide

whether toallowrecover or tosubject thefirmtobankruptcy

togetherinanAssembly,wherethelabor,securedandunsecured

creditorsvoteontheplan.Allthreeclassesofcreditorsmustvote

toapprovetheplan.Withregardtosecuredandunsecured

cred-itors, the planmustbe acceptedby amajorityof creditorsat

themeeting(numbercriteria)andatleasthalfofthetotaldebt

valueforeachclassmustberepresentedduringthevote(value

criteria).

By contrast, for labor creditors, only a majority vote is

required (number criteria). Thesetwo criteria allow firms to

avoidopportunisticbehaviorfromcreditors,wheresome

(3)

special treatment. If the plan is rejected, the firm enters bankruptcy.

WeusedatafromVaradeFalênciaseRecupera¸cãoJudicial

inSãoPauloandfrom firms’website.Basedondata on120

restructuring plansfrom 2005to2014 we find that the labor

classofcreditorsapproves thereorganizationplanevenwhen

theplanisultimatelyrejected.Moreover,wefindthatapproved

planshaveasmallerportionofdebtdiscountedandhighergrace

period on average thanrejected and modified plans. Further,

rejectedreorganizationplanshavehigherdisparitiesinpayment

proposals within thesame class of creditors,and

reorganiza-tionplansthatweremodifiedduringthecreditors’meetinghave

higherdisparities inpayment proposals amongthe classes of

creditors.

To evaluatethe likelihood of acceptanceof reorganization

plans, we run probit regressions. Wefind that asset disposal

increasesthelikelihoodofrestructuringplanapproval.One

pos-sibleinterpretationofthisfindingisthatcollateralisanimportant

determinantofrecoveryplanacceptance.Creditorsseemto

gen-erallypreferthatfirmsliquidateaportionoftheirassetssinceit

facilitatestheirabilitytoreceivecash.Wealsofindthatsecured

debtcreditors havelowerincentivesthanthe otherclasses of

creditorstoacceptreorganization plans;moreover,sincethey

arethelastclassofcreditorstoreceivepaymentafter

liquida-tion,unsecuredcreditorsaremorelikelytoacceptreorganization

plans.Wealsofindthathighdebtvaluesfrombanksinthejunior

classarenegativelyrelatedtoplanacceptanceandthatpayment

disparities among allclasses of creditors seem toreduce the

likelihoodofacceptance.

Thispaperisstructuredasfollows:Thesecondsection

dis-cussestherelatedliterature.Thethirdsectiondescribesourdata.

Thefourthsectiondescribestheempiricalstrategyofanalysis.

Thefifthsectionreportsempiricalresultsandarelated

discus-sion.Thefinalsectionconcludesthepaper.

Relatedliterature

Theimplementationofabankruptcyprocessbylawraises

someconcernsregardingitseffectsonsecurityprices,default

losses,priorityrulesandfinancialreorganization.Accordingto

Gilson(2012),academicresearchonbankruptcyhasbeen

con-centratedinfourmainareas:bankruptcyresolution,bankruptcy

costs (Haugen & Senbet, 1978, 1988), governance changes

in bankruptcy and the effects of bankruptcy on stock prices

(Eberhart,Moore, &Roenfeldt,1990).Thisresearchfocuses

onbankruptcyresolution.

Weareinterestedinunderstandingcreditors’decision

mak-ing about the restructuring process. According to Djankov,

McLiesh,andShleifer (2007) thereis agroupof papers that

connect legal protection for creditors and judicial efficiency.

Forinstance,ClaessensandKlapper(2005)arguethatgreater

judicial efficiency is strongly associated with greater use of

bankruptcy,althoughthecombinationofstrongercreditorrights

andgreaterjudicialefficiencyleadstolessuseofbankruptcy.

Wewouldliketohighlightthestudiesthatseemtoexplain

whytheresolutionoffinancialdistressvariesacrosscountries.

GennaioliandRossi(2010)showthatstrongcreditorprotection

increases the efficiency of the resolution of financial distress

becauseitprovidesjudicialincentives.

Based on a sample of 49 countries,La Porta,

Lopez-De-Silanes, Shleifer, andVishny (1997)find that countries with

poorinvestorprotection(legalrulesandqualityoflaw

enforce-ment)havesmallerandnarrowercapitalmarkets.Toconstruct

ameasureoftheefficiencyofdebtenforcement,Djankov,Hart,

McLiesh, and Shleifer (2008) compare debt enforcementfor

thesamekindof businessin88differentcountries.They find

thatinstitutionsthatregulateinsolvencyusuallyperformpoorly

owingtotheirinefficientbankruptcyprocedures.

PenatiandZingales(1997)revealtheimportanceof

under-standingthelegalenvironmentinwhicharestructuringprocess

occurs because of its effects on parties’ outside options.For

example, the Italianbankruptcy codeincludes two main

pro-cedures (liquidation and reorganization) for addressing an

insolventcompany. Inthe caseofbankruptcy liquidation,the

bankruptcy courtappointsatrusteewho shutsdownthe firm

andsellsitsassets(orevensellsthewholebusiness).The

abso-lutepriorityrulethendetermineshowtheproceedsofthesale

aredividedamongtheclaimants(Franks&Torous,1989,1994).

Sincewefocusonthereorganizationprocessfordistressed

firms, itisimportanttoexamineguidanceprovidedregarding

thepoweroflaw.Inthisregard,PlattandPlatt(2008)examine

factorsthatseemtopredictfinancialdistressintheUS,Europe

andAsia:theyfindthat differencesinaccountingrules,legal

practices,environmentallaws andbusiness practicesbetween

regionsmaylimitthedegreeofconvergenceintheareaof

finan-cialdistress.Indeed,bankruptcylawvariesconsiderablyaround

theworld.

Several studies haveexamined the main characteristics of

restructuringprocessesaroundtheworld.AccordingtoFranks,

Nyborg, andTorous(1996) andHotchkissetal.(2008),who

specificallyfocusontherestructuringofdistressedfirmsinthe

UK,Sweden,France,GermanyandJapan,thedegreetowhich

companies’businessisprotectedfromcreditorsalsovaries

con-siderably.

In studying the primary effects of the new Brazilian

bankruptcylaw,AraujoandFunchal(2009)findthatthenewlaw

hashadarapidandstrongimpactonthenumberof

bankrupt-cies in Brazil. According to these authors, expansion of the

creditmarketisobservable.Moreover,Kadiyala(2011)

investi-gatestheimpactofbankruptcylawreformoncapitalmarketsin

Braziland,basedonanempiricalanalysisoffourdifferentstock

indexes(Bovespa,IBX,IGCXandITAG),showsthataggregate

stockmarket indexes reacted positively bythe timethat new

rules were signed into law. These resultsare consistentwith

thoseofLaPortaetal.(1997),whofindthatbetterbankruptcy

lawsleadtoincreasedequityvalues.

Following Quian and Straham’s (2007) argument that

the quality of the legal environment shapes the

character-istics and terms of bank loans around the world, Araujo,

Ferreira, and Funchal (2012) evaluate the empirical

conse-quencesof the bankruptcyreformoncreditmarketsbyusing

a quasi-experimental approach to compare Brazilian firms

with non-Brazilian firms (companies from Argentina, Chile

(4)

responsibleforbothanincreaseintheamountoflong-termdebt

andareductioninthecostofcapital.

FunchalandClovis(2009)studyfirms’capitalstructureand

bankruptcylawdesigntoexaminetheeffectofchangesin

pri-oritiesamongcreditorsandfindasignificantimpactonfirm’s

financialpolicyinlinewithlowercostsofcapital.

Exploiting thequality of courtenforcementacross

Brazil-ianjudicialdistricts,Ponticelli(2012)showsthatefficientcourt

enforcementhelps sustainhigher capitalinvestmentand

pro-ductivity for companies. Thus, firms that face better court

enforcement benefitin terms of access toexternal financing,

investmentandproductivity.

Moreover, De Assis (2012) present an interesting study

focusedonanalyzingjudicialrecoveryproceedingsimmediately

followingtheimplementationofthenewbankruptcylaw.This

paperprovidessomeanalysisofrestructuringplansandshows

thattheaveragetimetocompleteallstagesoftheproceedings

exceedsareasonableamountoftime.Ourpapermust

corrob-orate previous work since it intends to present and analyze

informationprovidedinreorganizationplans.

Analysisoftheapprovalofreorganizationplans

UnderthenewBrazilianbankruptcylaw,court-based

restruc-turing permits different means of restructuring, such as a

potentialchangeincorporatecontrol,thestipulationofspecial

termsandconditionsforpaymentsofobligations,andtheright

ofvetofor creditorsregardingrestructuring plans.Theresults

regarding the newBrazilian bankruptcy law show that many

companieshavechosentoadoptarestructuringplaninorderto

addresstheirfinancialproblems.Moreover,thetotalnumberof

restructuringcaseshasincreasedineachyearafterthenewlaw

enteredintoforce. Until2005,bankruptcyinBrazilwasruled

byLaw7661.

Law7661didnotofferconditionsfortherecoveryof

eco-nomicallyviablecompanies that facedfinancial distress.The

oldreorganizationprocedure(knownasconcordata)only

post-ponedcorporate debt.Moreover,asthemainshortcomingsof

theprevioussystem,theliquidationprocesswascharacterized

byextensivebureaucracy,optimalrecoverycouldgenerallynot

beachievedinsituationsofdistress,andfirmsfaceddifficulties

inobtainingnewdebttorestructuretheirbusiness.Moreover,

theinsolvencyprocessdidnoteffectivelyprotect creditrights

afterliquidation.

AccordingtoFunchal(2006),creditorsplayamore

signifi-cantroleintherestructuringprocedureunderthenewBrazilian

bankruptcylawbecausetheyareinvolvedinthenegotiationand

votingonthereorganizationplan.Despitetheimprovementsof

thenewlaw,itcomplicatestheprocessofresolvingfirms’debts

byforcing heterogeneouscreditors tovotetogether (Funchal,

2006).Asnotedearlier,allthreeclassesofcreditorsmustvoteto

approvethefinalplan.However,Brown(1989)findsthat

hetero-geneousgroupsofcreditorsaremoreconcernedwithreceiving

guarantees,whereashomogenouscreditors areprimarily

con-cernedwithparticipatingintherestructuringprocess.

Reorganizationplansmustbeapprovedatthecreditors’

meet-ing,wherecreditorsaredividedintothreeclassesfortheirvote.It

isimportanttohighlightthattaxcreditorsandcreditorsholding

loanssupportedbyfiduciaryalienationofassetsarenotsubject

torecovery:therefore,theydonotvoteonreorganizationplans.

Securedcreditorsvoteasaclassandrepresentanamountup

tothevalueoftheircollateral.Moreover,whencreditorsdemand

morethantheircollateralvalue,theycanvoteasbothsecuredand

unsecuredcreditors,andtheythenrepresentexactlythesame

amount theyownfor each category.Debtors canindicate the

periodthattheybelievetobereasonablebywhichtopaytheir

securedandunsecuredcreditors.However,accordingtoarticle

54ofLaw11.101/2005,debtorscannotstipulateaperiodgreater

thanoneyearforlabordebtintheirrestructuringplan.

Onceacompanyhasdecidedtoundergocourtrestructuring,

the processbywhichcreditors’ acceptanceorrejectionof the

reorganizationplanproceedsthroughthefollowingsteps:

1. Thefirmmustpresentthereorganizationplanincourtwithin

sixtydaysafterdecidingtoundergoarestructuringprocess;

2. The judge communicates that the recovery plan has been

received and sets a deadline for creditors to present any

objection;

3. Labor,securedandunsecuredcreditors(or,potentially,only

oneortwoclasses)voteonthereorganizationplantoaccept,

rejectorpostponeitinordertodemandofadditionalchanges.

All thethreeclasses of creditorsmustvote toapprovethe

plan for it to be approved; otherwise, the firm undergoes

bankruptcy.

Further,thebankruptcylawestablishesthefollowingorder

of debt prioritywhen afirm optsfor bankruptcy or when its

restructuringplanisrejectedinthecreditors’meeting:

1. Labordebtuptothelimitof150minimumwagesperworker;

2. Secureddebtuptothelimitofthecollateral;

3. Taxdebt;

4. Paymentfordebtholderswithspecificandgeneralprivileges;

5. Unsecureddebt.

Datadescription

We use data from different sources to create our sample.

First, wecollectedsomecourtrestructuring plansfrom“Vara

de Falências eRecupera¸cãoJudicial”inSão Paulo,we then

obtained awide variety of restructuring plans from Google®

sincethedataarepublicandusuallyavailableonthewebsites

offirmsandjudicialtrustees.Weconsiderinformationfromboth

privateandpubliccompanies’restructuringplans.

Oursampleincludes120firmsforwhichwehave

informa-tion about labor, securedand unsecured funding from banks

andnonbankcreditors.Since2005,therehavebeenonlyafew

restructuringprocessforpubliccompanies:therefore,themain

part ofoursamplecomprisesprivatefirms. Wecollecteddata

from3differentdocumentsonfirms’reorganizationprocesses,

andweanalyzedthereorganizationplanitself,theminutesfrom

the creditors’ meeting and the relation of each creditor that

(5)

Table1 Sectorrepresentation. Sector #Firms % Industrial 44 36.67 Consumer,noncyclical 38 31.67 Consumer,cyclical 18 15.00 Utilities 14 11.67 Energy 4 3.33 Basicmaterials 2 1.67 Sum 120 100.00

Source:OwnelaborationfollowingtheBloombergsectorclassification.

AsTable1shows,oursampleismoreconcentratedinthe

indus-trialandnoncyclicalsectorsandlessconcentratedinthebasic

materialssector.

Table2summarizesthebasicstatisticsforeachvariable

col-lectedfrom the documents mentionedabove. The definitions

ofthevariablesarealsoprovidedinthenotestothetable.As

shown,firms’averageage(frombirthtotherestructuringdate)

isapproximately31years.

Regarding the number of banks, the statistics show that

approximately seven banks are involved in the restructuring

processperfirm.

Sampleselectionissues

The data in our sample were not provided by a

quasi-experiment. Since we collected our data by searching for

informationprovidedonthewebsitesoffirms,lawyersandjudge

trustees,oursamplemaypresentsomekindofbias.Becausewe

haveanonrandom sample, we canonlyanalyzethe possible

directionof the biasrather thancompletely eliminateit.One

possibletypeofbiasmayberelatedtotheregionas wehave

ahigherconcentrationoffirmsinthesouthandsoutheastand

onlyafewfirmsinthenorthandnortheast.

However,thisdistributionoffirmsisinlinewiththe

popula-tionsofthejudicialdistrictsinBrazil.TheStateofSãoPaulohas

thehighernumberofcivilcourts.Further,inhisanalysisof

judi-cialdistrictsinBrazil,Ponticelli(2012)findsthatthecountry

isdividedinto2738judicialdistricts,withahigher

concentra-tionofjudicialdistrictsandcourtsdealingwithbankruptcyin

southeastandahigherconcentrationofcompaniesinthesouth

andsoutheastofBrazil.Thecharacteristicsofoursamplearein

linewiththesecharacteristics.Fig.1showstheportionoffirms

inoursamplebystate.

Ponticelli(2012)notedthattheStateofSãoPaulohasa

con-centrationofcivilcourts,whichisconsistentwithoursample

becauseagreaternumberof observationswereobtained from

thesoutheast,whichhadmorethan10differentcourtsinSão

Pauloalone.Ponticellialsoshowedthatcourtconcentrationin

thesoutheastisworsethaninotherregionsandindicatedthat

companies in Brazilare extremelyconcentrated in the south

andsoutheast.Wedonotbelievethat suchcharacteristicscan

completelyeliminateabiastowardthisregion;however,weare

confidentthatourdatafollowthecharacteristicsofrecoveryin

Brazilasawhole.

Moreover, Brazilianjudicial trustees are appointed by the

court,butaspecificlawyermayreceivemorecomplicatedcases

basedonhisreputationor knowledgeaboutbankruptcy

situa-tions.Fortunately,wecanchecktheminutesoftheAssemblyto

determinewhetheralawyerrepresentedtherecoveringfirmora

judicialtrusteewasinchargeofthecase.Withsuchinformation,

wecandeterminewhetherthereisapatternrelatedtolawyers

inoursampledata.Our casesare spreadoutamongdifferent

lawyers,whichreducesthepossibilitythataspecificlawyeris

Table2

Descriptivestatistics.

Totalsample

Variable NOBS Measurement Mean Std.Dev. Min Max

Labordebt% 120 % 0.033 0.0685 0 0.5547

Secureddebt% 120 % 0.178 0.223 0 0.9081

Unsecureddebt% 120 % 0.789 0.2314 0.085 1

Securedbankloan% 114 % 0.361 0.4138 0 1

Unsecuredbankloan% 114 % 0.408 0.321 0 1

Labordebt# 112 Numerical 349.5 1064 0 7278

Secureddebt# 115 Numerical 2.852 4.3774 0 31

Unsecureddebt# 115 Numerical 306.7 472.73 2 2754 Concentrationoftop10creditors 109 % 0.682 0.1891 0.173 0.994 Concentrationoftop10creditors(noBanks) 109 % 0.329 0.2211 0.003 0.924

NumberofBanks 110 Numerical 6.51 3.6086 0 18

FirmAge 111 Years 31.33 23.22 4 120

Paymentyears 114 Years 11 4.14 3 22

Thetablereportsdescriptivestatisticsfromfirms’restructuringplans.ThevariableLabordebt(%)istheportionofthefirstclassoffirmdebt.ThevariableSecured debt(%)representstheportionofthesecondclassoffirmdebt.ThevariableUnsecuredistheportionofthethirdclassoffirmdebt.Securedbankloan%showsthe portionofbankloansthatconstitutesecureddebt.Unsecuredbankloan%showstheportionofbankloansthatconstituteunsecureddebt.Labor#,Secured#and Unsecured#arethenumberoflabor,securedandunsecureddebtholders,respectively.Top10(%)istheproportionofthedebtheldbythe10debtholderswiththe highestamountofdebt,andTop10(%,nobanks)istheproportionofthedebtheldbythe10debtholderswiththehighestamountofdebt,excludingbanks.The variableNumberofBankspresentsthenumberofbanksoperatinginthelistofcreditors.ThevariablePaymentyearsistheperiodoftime,asstatedbythefirms, neededtosettletheirdebt.ThevariableFirmAgeindicatestheperiodoftimefrombirthtotherestructuringyear.

(6)

Firms by state 60 50 40 30 20 10 0 SP RS GO SC RJ MT MG PE RO PA State Brazilian states: MG = Minas gerais GO = Goias PE = Pemambuco RS = Rio grande do sul

RO = Rondônia SC = Santa catarina MT = Mato grosso RJ = Rio de janeiro SP = Sao paulo PA = Para

Fig.1.Portionoffirmsinthesamplebystate. Source:Ownelaboration.

drivingourresult.Althoughwehaveshownthecharacteristics

ofourdatatoanalyzethepotentialforbiasinourdatawecannot

makedefinitiveconclusionssincewedonothavearandom

sam-ple.Wearequiteconfidentthatanybiasinoursampleisrelated

theconcentrationofcaseshandledbylawyersandbelongingto

particularregions.

Empiricalstrategyofanalysis

Thispaperinvolvesbothadescriptive andan econometric

analysisofthemaincharacteristics ofreorganizationplansby

typeofvote.Itisimportanttohighlightthatwedonotintendto

identifyacausalrelationbetweenthevariablessincewearenot

conductingacontrolledexperimentprovidedbyanexogenous

shock.Weprovideaninitialanalysisofreorganizationplansby

typeofvote,andwethenconductaneconometricanalysisto

calculatecontrolledcorrelationsbetweentheindependent

vari-ablesandthelikelihoodthatcreditorsinitiallyaccepttheplan

(i.e.,withoutmodification).

Descriptiveanalysis

Thispartofthepaperaimstoprovideempiricalevidenceof

thecharacteristicsofthereorganizationplansbyseparatingthe

reorganizationplansaccordingtotheAssemblyresults.

Specif-ically, sinceplanscanbe approved, modifiedor rejected, we

decidedtocapturethecharacteristicsofeachplanandcompare

themaccordingtothepossibleresultsoftheAssembly.We

fur-therconductadescriptiveanalysisontheproposalthatdebtors

presentedtoclaimholdersregardingpayment.Thispartofthe

analysisexaminestheportionofdebtdiscountedfromthe

origi-naldebtvalue,thegraceperiodsuggestedbydebtorstopostpone

thefirstpaymentandthecorrectionformofthedebtpayment

providedduringthereorganizationperiod.

We also perform a descriptive analysis of disparities in

paymentproposalsamongcreditors.

Econometricregressions

To identify which kind of outcome one can expect for a

restructuringplan,wedecidedtoadoptaprobitregression.Since

the readers of thispaper likely come from different areas of

knowledge,weprovideabriefdescriptionrelatedtoour

empir-icalmodel.

Ourvariableofinterestisadichotomousdependentvariable

(creditorsmustapproveorrejectthereorganizationplan).

There-fore,ourdependentvariableisanominalscalevariable,taking

valuesof 1 forapproved plansand0 otherwise.Aresearcher

canchooseamongseveraltechniquestorunaregressionwhen

thedependentvariableisbinary,suchastheLinearProbability

Model(LPM),theLogitModelortheProbitModel.

TheLPMassumesthattheprobabilityofthedependent

vari-able(toapprovetheplan,inthiscase)moveslinearlywiththe

valueoftheexplanatoryvariables.Moreover,thereisno

guaran-teethattheestimatedvalueintheLPMmodelwillliebetween0

and1.Hence,logitandprobitmodelsarethepreferredchoicefor

modelingourdecisionmakingprocesssincetheyfitanonlinear

function to the data andguarantee that the estimated

proba-bilitieswill liebetween0and1.Thesetwomodelsgenerally

presentsimilarresults.Thelogitmodelconsidersthecumulative

distributionfunctionofthelogisticdistribution,whilethe

pro-bitmodelconsidersthecumulativedistributionfunctionofthe

standardnormaldistribution.Logitmodelsareusuallyapplied

whenresearcherswanttointerpretcoefficientsoftheregression

interms ofodds ratios(which isnotour case).Logit models

presentslightlyfattertailsandtheirconditionalprobabilityof

rejectingorapprovingthereorganizationplanapproaches0or

1 ataslower rate comparedtoprobit models.Therefore, we

(7)

Our model aims to determinethe likelihood that a firms’

restructuring planis accepted when creditors do not demand

additionalchanges. Therefore, we do not separate plans that

wereapproved,modifiedorrejectedinoureconometricanalysis

inordertorunamultinomialprobitregression.Modifiedplans

havemanyofmodificationsfromaspecificgroupofcreditors

thatdefineconditionsforapprovingtheplan.Hence,modified

planswere first rejected by creditors before being modified.

Wedecided tocapture the likelihood that afirms’

restructu-ringplanisacceptedwithoutadditionalchanges.Therefore,our

researchquestionisasfollows:Whatarethemaindeterminants

ofreorganization planacceptancefor eachclass ofcreditors?

At a first glance, we would like to evaluate the factors that

affectcreditors’ decision about the reorganizationplan. How

doheterogeneouscreditorsbehaveinthedecisionprocessfor

reorganizationplans?Doesthedecisiontoacceptreorganization

plansliewithbanks?Doesitconcernspecifyingcollateral?

Giambonaetal.(2013)arguethatbecausetherearedifferent

classesofdebt,reorganizationplansareoftenrejectedin

Chap-ter11proceedings.Previousworkshavealsostressedthatthe

allocationofresourcestodifferentclaimholdersthatisspecified

inarecoveryplanisasimportantasthepotentialvaluethatthe

restructuringwillengender.Further,Brown(1989)arguesthat

problemscanriseinthepresenceofheterogeneouscreditors.

Following these studies, we evaluate the likelihood of

approvalaccordingtothecategoriesofcreditors.Accordingly,

asexplanatoryvariablesinourregressionsweadopttheratioof

eachdebtcategorytothetotaldebtandthenumberofcreditors.

Concerning the restructuring process, Senbet and Wang

(2010) state that creditors generally prefer asset liquidation,

sincesuch aprocedurefacilitatestheirabilitytoreceivecash.

In our empirical model, we examine the effect of asset

dis-posalwhen acollateralassetfor debtpayment isspecified in

thereorganizationplan.

Forthispurpose,werunregressionswithlabor,securedand

unsecureddebtastheexplanatoryvariablesandwemodeleach

regressionwhilecontrollingforagroupofvariablesthateach

categoryofcreditorsshouldconsiderinvotingonthe

reorgani-zationplan.

Inadditiontotheamountofdebtfromheterogeneous

credi-torsmentionedabove,wealsoconsidertheamountofsecured

andunsecuredbankloanswithclaimsatthe creditors’

meet-ing as a control measure. There is no consensus based on

empiricalevidenceaboutthe roleof banksintheapprovalof

reorganizationplans.AccordingtoGilson(1990),Brown(1989)

andJames(1996),bankshelpreduceholdoutandinformation

problems in private restructuring. However, Helwege (1999)

argues that bank debt is related to a slower debt

restructu-ringprocess.Althoughwecannotspecifytheexpectedsignfor

bankdebtinourempiricalinvestigation,bankdebtnevertheless

seemstobenecessary toincludeas acontrol variableinour

regressions.

Finally,wealsocontrolfortheperiodoftimestatedbythe

firmsbywhichtosettletheirdebt,thetypeoffirm,disparitiesin

paymentproposalsamongcreditorsandmodificationsin

corpo-rate ownership.Since the reorganizationplan ismade before

creditors vote, all explanatory variables are specified for the

periodbeforetheacceptance,requestedmodificationorrejection

oftheplan.

Ourfirstempiricalmodelisdesignedforlaborcreditors.As

specifiedearlier,suchcreditorsarethefirstcategoryofcreditors

toreceiveanyamountofmoneyifaplanisrejected.According

tolaw, this categoryof creditors mustreceive payment from

debtorswithinaone-yearperiod.Hence,wedonotneedtorun

aregressionthat controlsfor theperiod of timestatedby the

firm to payits debtholders for this groupof creditors. Since

weincludemanyofthesamevariablesinsubsequentequations

below, we avoidrepeating the definition of each variablefor

allthe equations.Therefore,for each equation,we repeatthe

definitionforthedependentvariableandprovidethedefinitions

forthecontrolvariablesthatwerenotmentionedfor previous

equations.

Werunaregressionwithboththeportionofdebtandnumber

ofcreditorsasexplanatoryvariablesfortheclassesofcreditors.

Ourfirstequationisspecifiedasfollows:

yt =β0+β1Labor(%)t−1+β2Labor(#)+β3Typet−1

+β4Asset Disposalt−1+β5Total Debt(ln)t−1

+β6Dif Classest−1+β7Ownershipreorg+ut (1)

Yt=Pr(Y=1/X)=dummyvariablethatequals 1if the

reor-ganization plan is accepted without changes and 0 if it is

either acceptedwithmodifications suggestedby thecreditors

orrejected.

Labor (%)=ratio of labordebt tototal debt. The variable

Laboristhefirstclassofthefirm’sdebt.

Labor(#)=thisisthenumber(quantity)oflabordebtholders.

Type=dummyvariablethat equals1 forcorporations(S.A

firms).

Asset Disposal=dummy variablethat equals 1 if the firm

choosesacollateralassetfordebtpaymentintherestructuring

plan.

Total Debt(ln)=thevariableismeasuredasthelogarithmof

thefirm’stotaldebt.

DiffClasses=dummyvariablethatequals1fordisparitiesin

thepaymentproposalsamongthethreeclassesofclaimholders.

OwnershipReorg=dummy variable that equals 1 for

changesincorporatecontrol.

In the first equation, we are interested in the sign of the

variableslabor andasset disposal.It is difficult todetermine

howlaborcreditorsshouldbehavewithrespectto

reorganiza-tionplans,sincetherearenoobservableconditionsforwhichto

controlintheplananalysis,suchasemployment.

Laborcreditorsarethefirstcategoryofcreditorstoreceive

paymentinthecaseofbankruptcy:therefore,theyhave

incen-tivestorejecttheplanwhencreditorsareperceivedtohavean

advantage.Nevertheless,ifworkersbelieveintheirfirmandif

theyfear that they mayfaceproblems when returningto the

labormarket,theyhaveincentivestoapprovetheplan.Inthis

equation,weexpectonlyassetdisposaltohaveapositivesign

owingtotheguaranteeof cash,asnotedinSenbetandWang

(8)

AccordingtoBrown,James,andMooradian(1994),evidence

indicates that asset sales benefit creditors more than

equity-holders incasesof distress. Moreover,Asquith, Gertner,and

Scharfstein (1994) show that asset sales represent an

impor-tantmeansforfirmstoavoidbankruptcy.Thenullhypothesis

ofourequationsisthatnoneofthevariablesmentionedbelow

influencesthe acceptanceof the restructuring plan. Basedon

previousworks,weproposethefollowinghypotheses:

Hypothesis H1. Asset disposal may influence creditors to

acceptthereorganizationplan. Therefore,weexpecttofinda

positivesignforassetdisposalinourregressions(β>0).

Hypothesis H1must holdfor allregressionsinthispaper.

With regard to securedcreditors, we believethat our task is

easier.Eq.(2)isspecifiedasfollows:

yt =β0+β1Secured(%)t−1+β2Secured(#)t−1+β3SBLt−1

+β4Typet−1+β5Asset Disposalt−1

+β6Total Debt(ln)t−1+β7Dif Classest−1

+β8Dif SameClass+β9Ownershipreorg

+β10Pt−1+ut (2)

yt=Pr(Y=1/X)=dummyvariablethatequals1ifthe

reorganiza-tionplanisacceptedwithoutchangesand0ifitiseitheraccepted

withmodificationssuggestedbycreditorsorrejected.

Secured(%)=ratioofsecureddebttototaldebt.

Secured(#)=thisisthenumber(quantity)ofsecure

debthold-ers.

SBL=variablethat specifiestheportionof bankloansthat

constitutesecureddebt.

DifSame Class=dummyvariablethatequals1for

dispari-tiesinpaymentproposalswithinthesameclassofclaimholders.

P=thevariablePayment yearsistheperiodoftimestatedby

thefirmtosettleitsdebt.

Secured creditors own assets as collateral and receive

paymentafterlabordebtorsinthecaseofbankruptcy.Therefore,

webelievethatthiscategoryofcreditorsalsohasincentivesto

reject the reorganizationplan. According toBrouwer (2006),

somecountriesattempttoattenuatetheconflictthatarisesfrom

securedcreditors’righttoclaimtheircollateralbyapplyingan

automaticstay.However,suchameasuremaynotbesufficient

toconvincecreditorstoacceptthereorganizationplan.In

addi-tiontothecontrol variablesincluded inthe firstequation,we

addedtwomorevariablesrelatedtosecuredcreditors’decision

to accept the plan. For this kindof creditor,we believethat

the period of timestated by afirm to settleits debt andthe

amountofsecuredbankloanscaninfluencethelikelihoodthat

therestructuringplanisaccepted.

Therefore,wespecifyoursecondhypothesisasfollows:

HypothesisH2. Securedcreditorshaveincentivestorejectthe

reorganizationplan.Therefore,weexpecttofindanegativesign

forthecoefficientofthisvariable(β<0).

Finally,weanalyzethesamedecisionwithrespectto

unse-curedcreditors.Asthesecreditorsarethelastgroupofcreditors

toreceiveanyvaluefromliquidation,theyhaveincentivesto

Fig.2.Portionoffirmsinthesampleperyear. Source:Ownelaboration.

accept thereorganizationplan. Theincentivesof thisclass of

creditorsareclearlymorealignedwithshareholdersthanthose

of the otherclasses. Junior creditors areout of the moneyin

mostcases,andthedecisiontocontinuethebusiness(evenifit

isinefficient)canprovideanupsideforthisclassaccordingto

GertnerandScharfstein(1991).Focusingonunsecured

credi-tors,Eq.(3)isspecifiedasfollows:

yt =β0+β1Unsecured(%)t−1+β2Unsecured(#)t−1

+β3UBLt−1+β4Typet−1+β5Asset Disposalt−1

+β6Total Debt(ln)t−1+β7Dif Classest−1

+β8Dif SameClass+β9Ownership reorg

+β10Pt−1+υ (3)

yt=Pr(Y=1/X)=dummyvariablethatequals1ifthe

reorganiza-tionplanisacceptedwithoutchangesand0ifitiseitheraccepted

withmodificationssuggestedbycreditorsorrejected.

Unsecured(%)=ratioofunsecureddebttototaldebt.

Unsecured (#)=this is the number (quantity) of unsecure

debtholders.

UBL=unsecured bank loan indicates the portion of bank

loansthatconstituteunsecureddebt.

Weimpose thesamemodificationinEq. (3) regardingthe

separation of the value andnumber criteria specified for the

previousequations.Wepresentourthirdhypothesisasfollows:

Hypothesis H3. Unsecured creditors have incentives to

approvethe reorganizationplan. Therefore,we expecttofind

apositivesignforthecoefficientofthisvariable(β>0).

Empiricalresultsanddiscussion

Thissectionreportstheempiricalresultsfromourdescriptive

andeconometricanalyses.Fig.2showstheportionoffirmsin

thesampleperyear.

We have more reorganization cases in 2012, 2009, 2013

(9)

Table3

Quorumandvoteanalysisforthereorganizations.

Initiallyapprovedplans–quorumofvotes

Measurement Percentiles Mean Std.Dev.

Smallest 25% 50% 75% Largest

Labor (%) 0 0.64 0.94 1 1 0.7684 0.3145

Secured (%) 0.75 1 1 1 1 0.971 0.078

Unsecured (%) 0.1 0.46 0.575 0.66 1 0.588 0.217 Initiallyapprovedplans–votesinfavorofplanacceptance

Labor (%) 0.96 1 1 1 1 0.998 0.0085

Secured (%) 0.53 0.68 1 1 1 0.8548 0.1875

Unsecured (%) 0.52 0.63 0.77 0.95 1 0.7647 0.1694 Modifiedplans–quorumofvotes

Labor (%) 0 0.51 0.725 0.97 1 0.6553 0.3111

Secured (%) 0.12 0.9 1 1 1 0.8954 0.2236

Unsecured (%) 0.01 0.535 0.75 0.855 1 0.675 0.26 Modifiedplans–votesinfavorofacceptance(modifiedplan)

Measurement Percentiles Std.Dev.

Smallest 25% 50% 75% Largest Mean

Labor (%) 0.51 1 1 1 1 0.96 0.1047

Secured (%) 0.5 0.81 1 1 1 0.8956 0.1623

Unsecured (%) 0.52 0.65 0.8 0.91 1 0.7846 0.1591 Rejectedplans–qorumofvotes

Labor (%) 0 0.16 0.57 1 1 0.5475 0.4093

Secured (%) 0.21 0.89 1 1 1 0.8981 0.2371

Unsecured (%) 0.23 0.42 0.53 0.84 1 0.6 0.2379

Rejectedplans–votesinfavorofrejectingtheplan

Measurement Percentiles Std.Dev.

Smallest 25% 50% 75% Largest Mean

Labor (%) 0 0 0 0 1 0.1042 0.2854

Secured (%) 0.02 0.99 1 1 1 0.8984 0.2717

Unsecured (%) 0 0.46 0.68 0.85 1 0.6081 0.3085

Source:Ownelaboration.

Thistablepresentsthepercentageofaquorumandtheoutcomeofthevotesduringthecreditors’generalmeetingforeachpossibleresult.

2014becausewestopped collectingdatainthemiddleofthe

year.

Table3 presentsresults regardingthe quorum of creditors

whovotedonthe planduring thecreditors’ generalmeeting.

Forapprovedplanswithoutmodifications,theresultsrevealthat

moresecuredcreditors werepresent tovoteon theplan than

laborandunsecuredcreditors.However,therateofacceptance

ishigheramonglaborcreditors thanamongtheotherclasses.

Theresultsforthemodifiedplansaresimilar.

Furthermore, the number of no shows at the vote is

higher among unsecured creditors for approved plans, yet

(10)

Table4

Proposalforpaymenttoclaimholders.

A.Debtor’sproposaltoclaimholders Approved

Debtdiscount (%ofDebt) 0 0 0.2 0.375 0.6 0.195 0.2092 Graceperiod (Months) 0 12 24 24 48 20.4 11.357

Modified

Debtdiscount (%ofDebt) 0 0 0.3 0.5 0.7 0.27 0.2516

Graceperiod (Months) 0 9 12 24 60 15.92 13.11

Rejected

Debtdiscount (%ofDebt) 0 0 0.4 0.5 0.8 0.313 0.2739 Graceperiod (Months) 12 12 22 24 36 19.23 7.72

B.Correctionformofthedebtpaymentduringthereorganizationperiod Correctionformforthedebtpayment

Correctionform Approved Modified Rejected

%ofthetotal %ofthetotal %ofthetotal

Inflation(only) 23.08 10.26 41.18

Inflation+fixedinterestrate 3.85 2.56 0.00

Fixedinterestrate 19.23 17.95 5.88

Floatinginterestrate 23.08 43.59 35.29

N.I. 30.77 25.64 17.65

Total 100.00 100.00 100.00

ThistableshowsthedescriptiveresultsforallpaymentproposalsaccordingtotheresultsoftheAssembly.Debtdiscountistheportionofdebtdiscountedfromthe originaldebtvalue.Graceperiodistheperiodfromtheplanvotetothefirstcreditor’spayment.ThevariableN.I.indicatestheportionofreorganizationplansthat didnotpresentacorrectionformforthedebtpayment.

among labor creditors is similar to that among unsecured

creditors.

Theanalysisofrejectedplansisinteresting.Secured

credi-torsshowthehighestportionofrejections,whilelaborcreditors

rejected the planin only 10.42% (mean) of the cases inthe

sample.Infact,whentheplanwasinitiallyacceptedthelowest

portionoflaborcreditorstoaccepttheplanis96%.Thus,labor

creditorsapprovedtheplaninmostofthecases,evenwhenthe

otherclassesdecidedtorejectit.

Table 4 shows the proposal of payment to claimholders

accordingtotheoutcomeofthevoteintheAssembly.The

aver-ageportionofdebtdiscountedishigherforrejectedplans,while

theaveragegraceperiod(13.11months)ishigherformodified

plans.

Table4,partB,alsorevealsthatinflationindexeswereused

asthemainstrategyforcorrectingdebtpaymentsforrejected

plans,whilefloatinginterestrateswerethemainstrategyused

intheothercases.

Regarding the disparities in payment proposals, Table 5

shows that plans approved with no modifications are more

homogenous among claimholders. Moreover, modified plans

show the greatest disparities in payment proposals among

classes,whereasrejectedplansshowthegreatestdisparitiesin

paymentproposalswithinthesameclassofcreditors.

As shown for our first probit, regression in Table 6, the

coefficient for the variable AssetDisposal is positive and

significant. The interpretation of this result is quite simple,

sincehaving agreateramountof collateralfordebtincreases

the likelihood that the restructuring plan will be accepted.

Claimholders vote on the plan within an environment of

uncertainty: therefore, it benefits all creditors to associate

an asset with collateral. This result thus seems to confirm

that creditors generally prefer to liquidate a portion of the

firm’sassets,sincesuchaprocedurefacilitatestheirabilityto

receivecash. Thisfinding supportsthefirst hypothesisofour

(11)

Table5

Paymentdisparities.

Debtpaymentspecification Approved Modified Rejected

Sameclass Amongclasses Sameclass Amongclasses Sameclass Amongclasses %ofallcasesinthecategory

Debtdiscount 0 36 11.11 40.74 14.29 25

Graceperiod 8 0 9.26 25.93 10.71 21.43

Debtpaymentcorrection 0 12 1.85 16.67 3.57 3.57

Thistableshowsthedisparitiesinpaymentproposalsinthereorganizationforthesameclassofcreditorsandforthreedifferentclassesofcreditors.Disparitiesmean theportionofdifferentcasesforeachpaymentspecificationdividedbythetotalnumberofcases.

Thedifferenceamongclassesissignificantwhenyearfixed

effectsaretakenintoaccount.However,wedidnotfind

signifi-cantcoefficientsfortheremainingvariables.

Theempiricalresultsforthesecondequationarepresented

inTable7.Asshown,therelationshipbetweentheportionof

securedcreditdebtandthelikelihoodofacceptanceissignificant

andnegative.ThisresultisinagreementwithSenbetandWang

(2010)andGiambonaetal.(2013).

Asmentionedearlier,securedcreditorsseemtohave

incen-tives to liquidate a company and to get their money back.

In addition tosecured debt, asset disposal and disparities in

payment proposals among the classes are significant in the

regression.Thesignforassetdisposalremainspositivesignand

disparitiesinpaymentproposalsamongclassesshowsanegative

Table6

Probitresultsforlabordebt.

PlanAcceptance–Eq.(1)–labordecision Labor–valueandnumbercriteria Variables %and# Labordebt(%) −2.6129 −3.3078 (3.023) (3.0049) Labordebt(#) −00011 −0.0004 (0.00013) (0.0001) Type 0.4516 0.5273 (0.2953) (0.4167) AssetDisposal 0.4674 0.5253** (0.2953) (0.2642) LnTotalDebt 0.1182 0.0833 (0.1183) (0.1103) DifClasses −0.5173* ₋_0.3218 (0.3078) (0.2849) Ownershipreorg 0.2927 0.4297 (0.3452) (0.3179) Constant ₋2.2241 −1.998 (2.1135) (1.7945)

YearFE? Yes No

Observations 108 112 Pseudo-R2 0.1904 0.1506 Thistableshowstheresultsforlabordebt.PlanAcceptanceisadummyvariable thattakesthevalueof1ifthereorganizationplanisacceptedwithoutchanges and0ifthereorganizationplaniseitheracceptedwithmodificationssuggested bythecreditorsorrejected.Standarderrorsareshowninparentheses.

* _Denote_significance_at_the_10%_level. ** _Denote_significance_at_the_5%_level. ***_Denote_significance_at_the_1%_level.

sign.Regardingpaymentyears,onecouldexpectanegativesign

intheregressionbecauseitrepresentstheperiodoftimestated

bythe firmstosettletheirdebt: therefore,ashorterperiod is

betterforclaimholders.However,thecoefficientisnot

statisti-callysignificant.

Thecoefficientofthevariablesecuredbankloanisnot

statis-ticallysignificant.Asexplainedearlier,thisvariablecouldshow

eitherapositiveoranegativesign.

Table8presentsthesameanalysisforEq.(3),wherewefocus

ontheroleofunsecuredcreditors.

Table7

Probitresultsforsecureddebt.

PlanAcceptance–Eq.(2)–secureddecision Secured–valueandnumbercriteria

Variables %and# Secureddebt(%) ₋1.6489* −1.2868 (0.9109) (0.8255) Secureddebt(#) 0.2671 0.0153 (0.421) (0.0377) SBL 0.2911 0.0245 (0.4044) (0.3587) Type 0.3615 0.4368 (0.4563) (0.4239) AssetDisposal 0.6389* _0.7346** (0.3306) (0.3062) LnTotalDebt 0.1543 0.1347 (0.1167) (0.1085) DifClasses −0.5834* ₋_0.3559 (0.3164) (0.2853) DifSameClass −0.3957 −0.3741 (0.3509) (0.3365) Ownershipreorg 0.1657 0.3228

(0.3552) (0.3238) Paymenttimeyears ₋0.0212 −0.02441

(0.036) (0.338) Constant ₋2.7797 −2.449 (1.9251) (1.7047)

YearFE? Yes No

Observations 105 108 Pseudo-R2 0.1962 0.1558 Thistableshowstheresultsforsecureddebt.PlanAcceptanceisadummy variablethattakesthevalueof1ifthereorganizationplanisacceptedwithout changesand0ifthereorganizationplaniseitheracceptedwithmodifications suggestedbythecreditorsorrejected.Standarderrorsareshowninparentheses.

* _Denote_significance_at_the_10%_level. **_Denote_significance_at_the_5%_level.

(12)

Table8

Probitresultsforunsecureddebt.

Plan Acceptance–Eq.(3)–unsecureddecision Unsecured–valueandnumbercriteria

Variables %and# Unsecureddebt(%) 1.7764** _1.5624** (0.8433) (0.7017) Unsecureddebt(#) −0.0015*** ₋_0.0011*** (0.0005) (0.00044) UBL ₋1.2891*** ₋_1.2354** (0.5548) (0.5386) Type 0.7974 0.8046 (0.5167) (0.4732) AssetDisposal 0.8123** 0.9548*** (0.3591) (0.3318) Ln Total Debt 0.3729** 0.2716** (0.1601) (0.4733) Dif Classes −0.9174** ₋_0.5668* (0.3746) (0.3132) DifSameClass −0.6023 −0.4493 (0.3891) (0.3548) Ownershipreorg 0.2293 0.3628

(0.3972) (0.3572) Paymenttimeyears −0.1793 −0.0306 (0.3799) (0.0357) Constant −7.5088*** ₋_5.348**

(2.8419) (2.2522)

YearFE? Yes No

Observations 105 108 Pseudo-R2 0.312 0.2588 Thistableshowstheresultsforunsecureddebt.PlanAcceptanceisadummy variablethattakesthevalueof1ifthereorganizationplanisacceptedwithout changesand0ifthereorganizationplaniseitheracceptedwithmodifications suggestedbythecreditorsorrejected.Standarderrorsareshowninparentheses.

* _Denote_significance_at_the_10%_level. **_Denote_significance_at_the_5%_level. ***_Denote_significance_at_the_1%_level.

Asshown,thevariableunsecured,measuredasaportionof

unsecureddebttototaldebt,ispositivelyrelatedtothelikelihood

thatthereorganizationplanisacceptedwithoutmodifications.

Thisresultsupportsthethirdhypothesisofthispaper.Onecan

alsoseethatthecoefficientforthenumberofunsecuredcreditors

hasanegativesign.

Increasingthenumberofcreditorsinvolvedinthevoteona

reorganizationplancanengendermorecoordinationproblems,

andourresultssupportsuchastatement.Asexplainedearlier,

unsecuredcreditors haveincentives to approvethe

restructu-ringplan,sincetheyreceivepaymentafterothercreditors.The

coefficientforthevariableassetdisposalremainspositiveand

significant. Further, the coefficient for the difference among

classesremainssignificantinthesamedirection.Sincehigher

levelsofdebtdistancecreditorsfromat-the-moneypositionsin

thecaseofliquidation,webelievethatthedirectionofthesign

isaccurate.

InTables6–8,weinvestigatethelikelihoodthat

reorganiza-tionplansareacceptedwhenfirmspresentthemtoclaimholders

during the creditors’ general meeting. Wefind evidence that

moreheterogeneousproposals amongclasses ofclaimholders

reducesthe likelihood of plan acceptance. Several

complica-tions mayappear in situations wheremultiple creditors have

different interests and receive different contracts from the

firm. These issueshave received considerable attention from

theoreticalstudies,suchasKordanaandPosner(1999),Brisand

Welch(2005)andvonThaddenetal.(2010).Differentinterests

andheterogeneousproposalscanleadtoarejectionofthe

reorga-nizationplanevenwhenitappearstobegoodforthecompanyas

awhole.Thus,ourresultsareinlinewiththepreviousliterature.

A higher portion of unsecuredclaims increases the

likeli-hoodofplanacceptance.However,reorganizationplansareless

likelytobeapprovedincasesofagreaternumberofunsecured

creditors.Ahigherconcentrationofdebtinthehandsofsecured

creditors reducesthechancesofplanacceptance.Gertnerand

Scharfstein (1991) explain that the incentives of unsecured

creditorsaremoreclearlyalignedwithshareholdersthanthose

of the otherclasses. Junior creditors areout of the moneyin

mostcases,andthedecisiontocontinuethebusiness(evenifit

isinefficient)canprovideanupsideforthisclass.Therefore,they

haveincentivestoapprovethereorganizationplan.Nevertheless,

securedcreditorshaveincentivestoreceivecashinaccordanceto

theircollaterallevelasquicklyaspossible.Theymightsearch

for better investmentopportunities instead of waiting for the

distressresolution.

In Brazil,securedcreditorsarebanksinmostcases,which

hasanimportantimplication.Resolution2.682/1999fromCVM

(TheSecuritiesandExchangeCommissionofBrazil)specifies

that banksmustclassifycreditsfromapprovedreorganization

plansascategory H.Therefore,itrequiresbankstoprovision

100% of the borrowed amount as a guarantee of operation;

thisrequirement reducesbank interestbecause theyearnless

inthissituationcomparedtoanothermarketalternatives.This

implication is line with our results regarding secured

credi-tors(banks,inmostofourcases)andunsecuredbankloansin

Tables7and8.

Inaddition,divestmentproposalsincreasethelikelihoodof

plan acceptance.Thisresult isin linewithSenbet andWang

(2010)andGiambonaetal.(2013),whoarguethathigherlevels

ofassetliquidationfacilitatethereorganizationprocess.Brown

etal.(1994)alsoshow thatasset salesbenefitcreditorsmore

thanequityholdersincasesofdistress.

Concludingremarks

Inthispaper,westudythelikelihoodofacceptanceof

reor-ganization plans based on a sample of 120 Brazilian firms

for the periodfrom 2005 to2014.To ourknowledge, thisis

the firstpaper toevaluatethemaindrivers ofthe approvalof

reorganizationplansduringthecreditors’meeting.Some

impor-tant resultsarenotableinthispaper.First,weshowthatasset

disposalfacilitatestheapprovalofreorganizationplansas

col-lateralisanimportantdeterminantofrecoveryplanacceptance.

Indeed,collateralhelpsreduceacreditor’slossexpectations

dur-ingthesetofchallengesthatafirmundergoingrestructuringis

facing.

Second,weconfirm thatsecureddebtcreditorshavelower

incentivestoacceptreorganizationplans.Sincethesecreditors

haveaspecificamountincollateral,theymayprefertoliquidate

(13)

portion of unsecuredcreditors seemsto be relatedto higher

likelihoodof acceptance. Sincethesecreditors are more

sim-ilartoequityholdersandareusuallyoutofthemoney,theymay

haveincentivestoapprovetheplanevenifdoingsoisnotbest

decisionforcreditorsasawhole.

Thisstudyalsohassomelimitations.Unfortunately,wedo

notexaminecausalityinthispaper.However,sinceourpurpose

inthisstudy wastoshowthecharacteristicsofreorganization

plansbythetypeofvote,oureconometricresultsmerely

corrob-orateourdescriptiveanalysisbycontrollingforcertainvariables

showonlytherelation.

Futureresearchcanbeconductedtoanalyzewhetherfirms

whose reorganization plan was approved achieved success

throughrecovery.Wecouldnotanalyzethisissueinthispaper

becausemostofthefirmsarestillinrecovery.

Conﬂictsofinterest

Theauthorsdeclarenoconflictsofinterest.

References

Aghion,P.,Hart,O.D.,&Moore,J.(1992).Theeconomicsofbankruptcy reform.JournalofLaw,Economics,&Organization,8,523–546.

Anapolsky,J.M.,&Woods,J.F.(2013).PitfallsinBrazilianbankruptcylaw forinternationalbondinvestors.JournalofBusiness&TechnologyLaw,8, 397–450.

Araujo,A.P.,Ferreira,R.V.X.,&Funchal,B.(2012).TheBrazilianbankruptcy lawexperience.JournalofCorporateFinance,18,994–1004.

Araujo,A.P.,&Funchal,B.(2009).LeideFalênciasBrasileira,A.N.Primeiros Impactos.JournalofPoliticalEconomy,29,191–212.

Asquith,P.,Gertner,R.,&Scharfstein,D.(1994).Anatomyoffinancialdistress: Anexaminationofjunk-bondissuers.QuarterlyJournalofEconomics,109, 625–658.

Berkovitch,E.,&Israel,R.(1999).Optimalbankruptcylawsacrossdifferent economicsystems.ReviewofFinancialStudies,12,347–377.

Bebchuck,L.(2002).Exantecostsofviolatingabsolutepriorityinbankruptcy. JournalofFinance,57,445–460.

Bisin,A.,&Rampini, A.(2006).Exclusivecontractsandtheinstitutionof bankruptcy.EconomicTheory,27,277–304.

Bolton,P.,&Scharfstein,D.(1996).Optimaldebtstructurewithmultiple cred-itors.JournalofPoliticalEconomy,104,1–25.

Bris,A.,&Welch,I.(2005).Theoptimalconcentrationofcreditors.Journalof Finance,60,2193–2212.

Brouwer,M.(2006).ReorganizationinUSandEuropeanbankruptcylaw. Euro-peanJournalofLawandEconomics,22,5–20.

Brown,D.T.(1989).Claimholderincentiveconflictsinreorganization:Therole ofbankruptcylaw.ReviewofFinancialStudies,2,109–123.

Brown,D.T.,James,C.M.,&Mooradian,R.M.(1994).Assetsalesby finan-ciallydistressedfirms.JournalofCorporateFinance,1,233–257.

Claessens,S.,&Klapper, L.(2005).Bankruptcyaroundtheworld: Expla-nations of its relative use. American Law and Economics Review, 7, 253–283.

Cornelli,F.,&Felli,L.(1997).Exanteefficiencyofbankruptcyprocedures. EuropeanEconomicReview,41,475–485.

DeAssis,D.V.X.D.(2012).UmaAnáliseEmpíricasobreoProcessode Recu-pera¸cãoEconômicapelaviajudicialadotadopelassociedadesempresárias. Dissertac¸ãodeMestrado–Fundac¸ãoGetulioVargas,Direito/Rio,Riode Janeiro.

Djankov,S.,Hart,O.,McLiesh,C.,&Shleifer,A.(2008).Debtenforcement aroundtheworld.JournalofPoliticalEconomy,116,1105–1149.

Djankov,S.,McLiesh,C.,&Shleifer,A.(2007).Privatecreditin129countries. JournalofFinancialEconomics,84,299–329.

Eberhart,A.C.,Moore,W.T.,&Roenfeldt,R.L.(1990).Securitypricingand deviationsfromtheabsolutepriorityruleinbankruptcyproceedings.Journal ofFinance,45,1457–1469.

Franks,J.R.,Nyborg,K.G.,&Torous,W.N.(1996).AcomparisonofUS,UK, andGermaninsolvencycodes.FinancialManagement,25,86–101.

Franks,J.R.,&Torous,W.N.(1989).AnempiricalinvestigationofU.S.firms inreorganization.JournalofFinance,44,747–770.

Franks,J.R.,&Torous,W.N.(1994).Acomparisonoffinancialrecontracting indistressedexchangesandchapter11reorganizations.JournalofFinancial Economics,35,349–370.

Funchal,B.(2006).EssaysonCreditandBankruptcyLaw.InTese(Doutorado). RiodeJaneiro,FGV:EPGE–Fundac¸ãoGetulioVargas.

Funchal,B.,&Clovis,M.(2009).Firms’capitalstructureandbankruptcylaw design.JournalofFinancialEconomicPolicy,3,264–275.

Gennaioli,N.,&Rossi,S.(2010).Judicialdiscretionincorporatebankruptcy. ReviewofFinancialStudies,23,4078–4114.

Gertner,R.,&Scharfstein,D.(1991).Atheoryofworkoutsandtheeffectsof reorganizationlaw.JournalofFinance,46,1189–1222.

Giambona, E., Lopez-de-Silanes, F., & Matta, R. (2013). Improved creditor protection and verifiability in the U.S. In Working paper..

https://editorialexpress.com/cgi-bin/conference/download.cgi?dbname= sbe35&paperid=22(Accessed23.08.14)

Gilson,S.(1990).Bankruptcy,boards,banks,andblockholders:evidenceon changesincorporateownershipandcontrolwhenfirmsdefault.Journalof FinancialEconomics,27,315–353.

Gilson,S.C.(2012).Creatingvaluethrough corporaterestructuring:Case studiesinbankruptcies,buyouts,andbreakups(2nded.).Hoboken,NJ: Wiley.http://dx.doi.org/10.2139/ssrn.278580

Gilson,S.C.,Hotchkiss,E.S.,&Ruback,R.S.(2000).Valuationofbankrupt firms.ReviewofFinancialStudies,13,43–74.

Hackbart,D.,Hennessy,C.,&Leland,H.(2007).Canthetradeofftheoryexplain debtstructure?ReviewofFinancialStudies,20,1389–1428.

Haugen, R.A.,& Senbet,L.W. (1978).Theinsignificance ofbankruptcy coststothetheoryofoptimalcapitalstructure.JournalofFinance,33, 383–393.

Haugen,R.A.,&Senbet,L.W.(1988).Bankruptcyandagencycosts:Their significancetothetheoryofoptimalcapitalstructure.JournalofFinancial andQuantitativeAnalysis,23,27–38.

Hege,U.,&Mella-Barral,P.(2005).Repeateddilutionofdiffuselyhelddebt. JournalofBusiness,78,737–786.

Helwege,J.(1999).Howlong dojunkbondsspendindefault?Journalof Finance,54,341–357.

Hotchkiss,E.S.,John,K.,Mooradian,R.,&Thorburn,K.S.(2008).Bankruptcy andtheresolutionoffinancial distress.In Workingpaper..Availableat

http://mba.tuck.dartmouth.edu/Pages/Faculty/Karin.Thorburn/publications/ Ch14-N53090.pdf

James,C.(1996).Bankdebtrestructuringsandthecompositionofexchange offersinfinancialdistress.JournalofFinance,51,711–727.

Kadiyala,P.(2011).Impactofbankruptcylawreformoncapitalmarketsin Brazil.InvestmentManagement&FinancialInnovations,8,31–41.

Kaplan,S.N.,&Stromberg,P.(2003).Financialcontractingtheorymeetsthe realworld:Anempiricalanalysisofventurecapitalcontracts.Reviewof EconomicStudies,2003,281–315.

Kaplan, S. N., & Strömberg, P. (2004). Characteristics, contracts, and actions:Evidencefromventurecapitalistanalyses.JournalofFinance,59, 2177–2210.

Kordana,K.A.,&Posner,E.A.(1999).ApositivetheoryofChapter11.NYU LawReview,74,161–234.

LaPorta,R.,Lopez-De-Silanes,F.,Shleifer,A.,&Vishny,R.W.(1997).Legal determinantsofexternalfinance.JournalofFinance,52,1131–1150.

Penati, A., & Zingales, L. (1997). Efficiency and distribution in financial restructuring: The Case of Ferruzzi Group. In Working paper..http://faculty.chicagobooth.edu/finance/papers/ferfin.pdf(accessed 08.09.13)

Platt,H.,&Platt,M.(2008).Financialdistresscomparisonacrossthreeglobal regions.JournalofRiskandFinancialManagement,1,129–162.

(14)

Ponticelli,J.(2012).Courtenforcementandfirmproductivity:Evidencefrom abankruptcyreforminBrazil.InJobmarketpaper.Barcelona:University PompeuFabra.

Qian,J.,&Strahan,P.E.(2007).Howlawsandinstitutionsshapefinancial contracts:Thecaseofbankloans.JournalofFinance,62,2803–2834.

Schwartz,A.(1998).Acontracttheoryapproachtobusinessbankruptcy.Yale LawJournal,107,1806–1851.

Senbet,L.W.,&Wang,T.Y.(2010).Corporatefinancialdistressandbankruptcy: Asurvey.Finance,5,243–335.

vonThadden,E.,Berglöf,E.,&Roland,G.(2010).Thedesignofcorporatedebt structureandbankruptcy.ReviewofFinancialStudies,23,2648–2679.

Winton,A.(1995).Costlystateverificationandmultipleinvestors:Theroleof seniority.ReviewofFinancialStudies,8,91–123.