LeveragedBuyoutsinNorway
CarstenBienz1
September2016
Abstract:
This paper reviews several recent studies on buyouts in Norway that analyse the impact of buyout
investorsonthefirmstheyacquire.Aftercontrollingfortheendogeneityofthebuyoutdecision,firms
acquired by private equity funds seem to improve both their operational and financial performance
relativetocomparablefirms.Thereisnoevidencethattheimprovedfinancialperformanceiscausedby
increasedtaxplanning(againrelativetocomparablefirms).Finally,thereislittleevidencethatbuyout
fundsrelyonformalboardworktoeffectthesechangesinfirms.
Introduction
PrivateEquity(PE)inNorwayisstillaquiterecentphenomenonandourunderstandingaboutthisform
offinancing(andgovernance)stemsmostlyfromstudiesabroad.YetinrecentyearsPEactivityinNorway
hasbeenincreasingalot,bothintermsofdealsbutalsointermsoffundsbeingraised.Aroughcountof
buyoutdealsinNHH’sArgentumCentreforPrivateEquitydatabasesuggeststhataround250buyouts
happenedinNorwayupto2013.2Howarethesebuyoutsmanaged?Doestheoperationalandfinancial
performanceoftheacquiredfirmschange?Whatabouttheeffectsforstakeholders?
1
CarstenBienz,NorwegianSchoolofEconomics.Carsten.bienz@nhh.no.Iwouldliketothankthereferee,
HenningFredriksen,JørilMæland,WidarSalbuvikandFrodeSættemfortheircomments.
2
Ascomparedto250inDenmarkand450Swedenoverthesametimeperiod
-1-
Theempiricalevidenceabouttheeffectofbuyoutfundsoncompaniestheyacquiredsuggeststhaton
averageabuyoutisapositiveexperienceforthesefirmsandthestakeholdersinthesefirms.Thereare
individualexceptionsofcourse,liketheSwedishCaremascandal(Johnson,2014),butinthisstudyIwant
tofocusontheoverallimpactofprivateequityonfirmsandnotonselectedcases.Inparticular,Iwantto
focusonthefollowingthreequestions:
1) DobuyoutinvestmentsinNorwayincreasefirmperformance?
2) Whatarethetaxconsequencesofbuyoutinvestments?
3) Howdoescorporategovernancework?
Iamopportunisticinmychoiceoftopicsastherearemanyotherquestionswecouldtrytoanswer.Iam
able to analyse these three questions as I was given the opportunity to supervise three master thesis
writtenbyNHHstudentsinfall2015.Iwillcomplementthesefindingswithsomenewresults.Allstudies
werewrittenondatafromNHH’sArgentumCentreforPrivateEquity.Aswewillseelaterinmoredetail,
thereis,again,littleevidencethatabuyoutisabadexperiencefortheaverageNorwegianfirmthatgoes
throughit.3
Rathertheoppositeistrue-astudyby(Friedrich,2015)suggeststhatNorwegianPEtargetsbecomemore
efficientintheiroperational(i.e.assetturnoverimprovesby50%overathreeyearhorizon)andfinancial
aspects (i.e. the EBITDA/total assets ratio improves by 23% over a three year horizon). There is also
evidencethatPEownershipreducestheprobabilityoffinancialdistressrelativetocomparableno-buyout
firms.(Roald&Roti,2015)investigatewhethersomeoftheincreasedfinancialperformanceinNorway
comesattheexpenseoftax-payers.Yetitseemsthatthereisnoevidenceforsuchaproblem-thereis
nosignformoretaxplanningrelativetocomparablefirmsandbackoftheenvelopecalculationssuggest
thattaxincreasesthroughefficiencyimprovementsarelargerthanreductionsintaxesthroughleverage.
Finally(Farran&Lâm,2015)investigatethechannelthroughwhichbuyoutfundsaffectthe(Norwegian)
companiestheyhaveinvestedin.Interestingly,itturnsoutthatthischannelisnotthefirms’boards.We
seealmostnoinfluenceofthefundmanager’spresenceonthefirm’sboard.4
3
Iamdroppingthequalifier“relativetocomparablefirms”fromnowonforthesakeofexpositionbutall
statementsmadehere,ifnotmentionedotherwise,aremadewiththisqualifierinmind.
4
Itisnotclearwhattheexactchannelis–wecanonlyspeculatethatisthedirectcontactbetweenthefirm’s
managementandthefundmanagers.
-2-
Thefocusofthisstudywillbeontheinvestmentsdonebybuyoutfundsandwewillignorethequestion
of how funds in Norway are structured. For the impact of the fund’s structure on buyouts see (Bienz,
Thorburn,&Walz,2016).InwhatfollowsIwillcontinuebydiscussingsomestylizedfactsbeforeturning
totherelevanttheory.Iwillexplaintheempiricalmethodologyusedinthestudiesweanalysenextand
finallywilldiscussthestudiesthemselves.
PrivateEquity–ashortintroduction.
By definition Private Equity investments are about buying (controlling) equity stakes in privately held
firms.5FamilyofficesorSweden’sInvestorABdothesameyetwewouldnotconsiderthemtobeatypical
example for Private Equity. There are several differences. First, PE typically is organized via limited
durationfunds,withaten-yearcommitment,thatcollectcapitalfrominstitutionalinvestors.Thesefunds
arethenrunbyprofessionalmanagers.Themanagersareusuallycompensatedbyfeesandshareinthe
returntheygenerate(calledcarryintheindustry)(Bienz,Thorburn,&Walz,2016).6
Figure1:PEInvestmentStructure
Thisstructure(showninFigure1: PE Investment Structure)resemblesanidealizedXwherecapitalflows
from the investors (the limited partners or LP) to the fund. The investors are passive and the fund is
managedbythefundmanager,calledGeneralPartner(short:GP).Thefundusesthecapitaltopurchase
firms.Tobemoreprecise,thecapitalfromthefundfinancestheequityportionofthedealbutusually
thereisasubstantialdebtportiontoo.Thisdebtistypicallyraisedfrombanksonacase-by-casebasis.7
5
6
7
Unlike,forexample,activisthedgefundsthatonlybuysmallstakesinpublicfirms.
http://www.dn.no/nyheter/naringsliv/2013/01/04/tok-av-med-herkules
InEuropetypicallybanksarethemainprovidersofcapital.IntheUS,bondmarketsalsooftenplayalargerole
(Axelson,Strömberg,&Weisbach,2009).
-3-
Third,giventhefunds’limitedlife,theinvestmentisalwaysofatemporarynatureandthefunds“exit”
theirinvestments.Anyproceedsfromthefirm(intheformofdividendsorsalesproceeds)arethenpaid
backtotheinvestors,minusthefundmanager’sshareofthereturns(“carry”).
WhattypesoffirmsaretypicaltargetsforPEfunds?ThreereasonsthatmakefirmsinterestingforPE
fundshavebeenproposed:
1. Firmsareunder-levered.
2. Firmsareunder-performing.
3. Firmslackcapitalormanagerialexpertise.
Our perception of buyouts is shaped by the large transaction that make it into the financial press.
ExamplesaretheRJRNabiscobuyoutbyKKRorthepurchaseofISSbyEQTinDenmark,bothtwowellknownbuyoutfunds.Ineachcasealistedfirmwastakenprivate.RJRNabiscowasclearlymismanaged
whileISSwasoperatingwithlowlevelsofleverage.8YetconversationswithNorwegianPEfundmanagers
suggestthatinNorwayoftenthethirdreasonisofmajorimportance.
BothRJRNabiscoandISSwerepubliclytradedbeforetheywereacquiredbyaPEfund.Thisdoesnotseem
tobethecaseinNorwaywhere(Bienz,Thorburn,&Walz,2016)findthatonlythattwooutofmorethan
sixtydealsinNorwayinvolveafirmgoingprivate.MostdealsinNorwayaretransactionswereaprivate
firmissoldtoaPEfund.
ThedebateaboutPrivateEquity
What makes private equity different from other types of (equity) investments? We discussed several
institutionalaspectsthatmakePEdifferent–controllingstakes,theuseoflimiteddurationfunds,theuse
ofprofessionalfundmanagers,thetypeofinvestmentsdone,andthefinancingstructurethatmixesinside
equitywithoutsidedebt.Individuallytakennoneofthesechoicesareveryspecialyetthisparticularsetup
seems to give rise to a relatively special form of financing. Given the complexity of the funding
arrangement,theacademicdebatehasusuallyfocusedonexplainingtheeffectsofoneoftheseaspects
inisolation.
8
(Michel&Shaked,1991)and(ISSA/S:TheBuyout,2013)
-4-
TheacademicdebatetypicallytriestoexplainwhyPEdealsoccurinpracticeandthisexplanationisoften
coupledwiththeimplicitassumptionthatPEownershipofafirmissomethingpositiveforthefirm(and
theshareholdersinthefirm).9
The beginning of the academic debate about PE can usually be traced back to (Jensen, 1986) who
emphasizes the disciplining factor of high leverage or dividend payments. Jensen develops a short
theoreticalmodelwherea(time-inconsistent)managercommitstonotwastingcashbytakingonexternal
leverageorbycommittingtoahighdividend.Suchcommitmentpreventsthemanagerfrominvesting
into negative NPV projects/empire building as the high leverage reduces the firm’s free-cash flow. In
theorysuchcommitmentshouldalsoincentivizemanagersandemployeestoincreaseefficiencyinthe
firm. A famous example of such a strategy is Sealed Air Corporation. (Sealed Air Corp.'s Leveraged
Recapitalization(A),1994)SealedAirCorppaidoutaspecialdividendtwicethevalueofitsbookequity.
Despitetheextremelyhighleverage,thefirmoutperformeditsindustrypeerssignificantlyafterwards.Its
CFOatthetimeattributedthechangetotheincentiveeffectsoftheincreasedleverage.
However, one empirical finding that does not fit the Jensen story is the fact that in some instances
leverage in LBOs is not significantly higher than in normal firms, at least initially. For example, table 1
trackstheleverageratiosfor105Norwegianbuyoutdealsfoundin(Friedrich,2015)fromtheyearofthe
buyouttotheyearandcomparesittosimilarfirms:
LeverageRatio 𝐴𝑠𝑠𝑒𝑡𝑠 − 𝐸𝑞𝑢𝑖𝑡𝑦 Buyout
𝐴𝑠𝑠𝑒𝑡𝑠
Control
Buyout
𝐿𝑇𝐷𝑒𝑏𝑡
𝐿𝑇𝐷𝑒𝑏𝑡 + 𝐸𝑞𝑢𝑖𝑡𝑦 Control
Yearrelativetobuyout
+1
+2
74%
74%
71%
69%
30%
30%
0
73%
72%
32%
26%
25%
23%
+3
71%
70%
28%
24%
Table1:LeverageRatios.Source:Friedrich(2015)
Intheyearofthebuyout,overallleverage(asmeasuredbyassetsminusequitydividedbyequity)isnot
verydifferentbetweencontrolsandbuyouttargetsanditdoesnotchangemuchovertime;yetlongterm
9
Evenifthismightnotbethecaseforallstakeholders.
-5-
debtissomewhathigherthanincomparablefirms,thoughitfallsmore.Inaddition,(Bienz,Thorburn,&
Walz,2016)reportthatthechangeinleveragefromtheyearbeforethebuyouttothebuyoutyearisnot
statistically significant.10 Some of this result can be explained by the fact that historically Norwegian
buyoutfundswereconstrainedinusingtheacquiredfirm’sassetsascollateralforleverage,restrictingthe
initialincreaseinleverage.11Yetthisresultalsosuggeststhatincreasedleverageisprobablynottheonly
mechanismthatmattersduringabuyoutanditquestionstheimportanceofJensen’smodel.(Axelson,
Strömberg,&Weisbach,2009)arguethattheuseofoutsidedebtdisciplinestheGPsastheyhavetofind
banks that are willing to lend funds to the GP for the proposed transaction. Theoretically this should
preventGPsfromoverpayingandfromacquiringbadfirms.
(Shleifer & Summers, 1988), on the other hand, explain the gains to the GP and LP not as a result of
increasedefficiencybutastheoutcomeoftherenegotiationofimplicitcontracts.“Implicitcontracts”is
anacademictermforagentlemen’sagreement.Giventhelackofformaldocumentation,suchaninformal
agreementcanbeeasilybrokenwhenownershipchanges.(Shleifer&Summers,1988)andmanyothers
arguethatPEinvestorusetheirpowersasownerstorenegotiate (implicitandexplicit) contractswith
stakeholders,suchasemployeesorbondholders.Thenewownersusethispossibilitytoshiftrentstothe
firms.Thereisoftenanimplicitassumptionthatthisshiftisshortsightedasitdestroysfirmvalueinthe
longrun.
WhilenotexplicitlymentionedinShleiferandSummers,leverageseemstobelargesourceofvaluetoPE
funds(Kaplan,ManagementBuyouts:Evidenceontaxesassourcesofvalue,1989).Thisexplanationrests
ofthefactthatbuyouttargetsinthe1980swereoftennothighlyleveredwithoutthesellersrealizingthis
issue.IfincreasedleverageweretheonlysourceofgainsforLPsandGPs,thenofcourseaPEinvestment
wouldsimplybeataxarbitrageopportunity.
EmpiricallyitseemstheShleiferandSummersconjecturedoesnotgetalotofsupport,possiblywiththe
exceptionofnegativereturnsforsomebondholders.12Agoodexample,andonethatcomesclosesttothe
Norwegianresultsis(Boucly,Sraer,&Thesmar,2011),whofollow839Frenchbuyoutfirmsforthreeyears
10
Themeanleverage(
11
122342567894:
122342
)was78%andfallsto76%inthebuyoutyear.
Thisrestrictionwasusuallydeemedtobeobsoleteoneyearafterthebuyoutandhenceonecanobservelarge
re-capsayearafterthebuyout.
12
InparticularBondholdersnotprotectedbycovenants.See(ISSA/S:TheBuyout,2013)foranexample.
-6-
andcomparethemtoapeergroup.Theyreportthatbuyouttargetsincreasetheirprofitabilityandare
abletogrowfasterthantheirpeers.Thisgrowthseemstobeconcentratedinfirmsthatwereacquired
fromprivateowners(Private-to-private)deals.Thefactthatgrowthseemstobeconcentratedinthese
typesoffirmssuggeststhatPEisabletohelpovercomearestrictedsupplyofcapital.Theyalsoreport
thatgrowthseemstobecomingfromindustriesthataremorefocusedonexternalfinance.13
Theresultsfrom(Boucly,Sraer,&Thesmar,2011)suggestthatarelaxationoffinancingconstraintscauses
firmsacquiredbyprivateequityfundstoincreaseprofitability.However,improvementsinmanagement
qualityarealsooftencitedasareasonforprivateequityinvestments.Theseimprovementsarehardto
measure,possiblywiththeexceptionofboards.Here(Cornelli,Kominek,&Ljungqvist,forthcoming)find
acausallinkbetweenboardmonitoringandfirmperformanceforEasternEuropeanbuyoutdeals.Their
findingspointoutthatanyperformanceincreasebyfirmscanbeexplainedbyacombinationofbetter
oversightandbetteraccesstomanagerialtalent.
Tosumup,existingresearchconjectures(andisabletodocument)anincreaseinfirmperformancefor
companies acquired by PE funds. Both theory and empirical research suggest that this increase in
performancecannotbeexplainedbyasimpletransferfromstakeholderstoshareholders.However,the
exact channel for these improvements is not clear yet. Increased incentives seem to play a role, but
whethertheseincentivescomethroughincreasedleverage,bettercorporategovernance,betteraccess
tocapital,orsimplybecauseaccesstomanagerialtalentincreasesisnotclear.
HowshouldwecomparePEfinancedfirmswithnonPEfinancedones?
In general, we cannot directly compare firms that receive PE financing with non-PE financed firms as
buyouttargetsareselectedbytheirinvestors.Whatdoesthatmeanandhowdoesitaffectouranalysis?
First of all, we need to realize that by being selected by a buyout fund, firms are probably somewhat
differentfromtheaveragefirm.Whatcouldcausethisdifference?Itmightbethatthefirmhashigher
growthpotentialthanacomparablefirm,thatitsmanagementisconsideredtobebetterorthattheGP
hasprivateinformationaboutthefirm.Theproblemforusisthatitisdifficulttofindout.
13
Itisdifficulttodirectlyverifythatprivatetoprivatedealsgrowfasterthanformerlypublicfirmsgiventhat
almostalldealsinNorwayarePrivate-to-privatedealstostartwith,butanecdotalevidence(suchasthebuyoutof
XXL)supportsthisexplanationinNorwaytoo.
-7-
However,whatwecandoistofindfirmsthatdidnotreceivePEfinancingbutthatlooksimilaronthe
outsideandcomparethemtofirmsthatdidreceivefinancing.Wecanthentrytoseeifwefinddifferences
inthefutureperformanceofthesetwotypesoffirms.Inasensetheideabehindthisapproachistocome
ascloseaspossibletoarandomexperimentthatonemightfindinmedicinewhereideallybothtreatment
andcontrolsarerandomlyassigned.Whilewecannotrandomlyassigntreatment,i.e.beingsubjecttoa
buyout,wegotheotherwayaroundandfindcomparablefirmsthatareascloseaspossibletothetreated
firms. In the end we can compute the following difference in difference (another word for a double
difference):14
𝐷𝑖𝑓𝑓 − 𝑖𝑛 − 𝐷𝑖𝑓𝑓 = 𝐸𝐵𝐼𝑇𝐷𝐴
A,4CD
– 𝐸𝐵𝐼𝑇𝐷𝐴
A,4
– 𝐸𝐵𝐼𝑇𝐷𝐴
F,4CD
– 𝐸𝐵𝐼𝑇𝐷𝐴
F,4
TheaboveexamplecomputesthechangeinEBITDAforabuyoutdeal(B)attwopointsintime(tandt+n)
and a control group (C) over the same time period. If this difference is positive, it means that buyout
targetsperformedbetterthantheirpeersandviceversawithrespecttoEBITDA.
Howcanwefindoutwhichfirmsarecomparable?Thereareseveralstatisticaltechniquesthatallcanbe
seenasavariationofamatchingapproachthatcanhelpusfindappropriatecomparablefirms.Thepapers
citedherealluseatechniquecalledprosperityscorematchingthatfindscomparablefirmsbasedonan
indexvalue.Theclosertwofirmsareintheirindexvaluethemorecomparabletheyare.Thissystemhas
theadvantagethatisallowsformanydifferentattributestobeusedbutitworksnotparticularlywellwith
industries(whichcannotbenumericallycompared)andyears(2007iscloseto2008butverydifferent).
ThisprocedureisusedtocapturetheeffectsofthePEfundonthecompany.Wecancallthiseffecta
“treatment” effect. Ideally our procedure has eliminated any “selection” effect, that is that PE fund
managersarejustbetteratselectinggoodfirms.Howrealisticthisassumptionis,isnotentirelyclear.
However,byfocusingonobservablecharacteristicsweshouldattheveryleastminimizeanyselection
effect as much as possible. Of course, if there are any systematic unobservable characteristics – i.e.
buyout funds excel at finding the best management teams or management teams with private
informationproactivelypursuebuyouts–thenourresultscanstillbemisleading.
14
Thetechnicaltermforthisprocedureisdifference-in-differenceaswecomputethedifference
betweentwochanges.
-8-
Datasources
The data comes from several sources. Information about investments is taken from NHH’s Argentum
CentreforPrivateEquity’s(ACPE)database.ThedatabaseiscompiledbyNHHandcontainsinformation
onbuyoutdealsfromthelate1990supto2012.Wealreadymentionedthatthedatabasecontainsaround
250dealsupto2013.Thedatabasealsocontainsinformationonthenameoftheportfoliocompany,the
fundinvestingandthedealyear.ItismatchedtoSNF’saccountingdatabase,maintainedbyAkselMjøsat
NHH,usingorganizationnumbers.Organizationnumbersareindependentlyhand-collectedandverified
bytworesearchassistants.TheSNFaccountingdatabasecoversallNorwegianfirmsfrom1997to2014.
HistoricalownershipinformationwaspurchasedfromBisnodeandcoversallNorwegiandealsupto2012.
Weneedthisinformationtoidentifyholdingcompaniesandtheleverageintheseholdingcompanies.15
DobuyoutinvestmentsinNorwaycreatevalue?
(Friedrich,2015),investigatesthechangesfirmsundergoduringthetimetheyareownedbyabuyout
fundrelativetonon-buyoutfirms.Friedrichstructureshisanalysisalongthreemajordimensions:First,
howdoesthefirmdevelopfinanciallyandintermsofoperatingefficiency?Second,howaredefaultrates
affected?Finally,howarestakeholdersaffected?
Friedrichusespropensityscorematchingtoassignacontrolgrouptoeachfirmgoingthroughabuyout.
Theideais,asexplainedabove,tofindoneto(five)comparablefirmsforeachbuyoutdeal.Hematches
againstthepopulationofallfirmsintheBrønnesundregistry.Inordertocompareperformancebetween
buyout firms and controls he tracks the changes in key firm variables. This procedure ensures that a
developmentisnotsomethingthatwouldhavehappenedbecauseofacommontrend.Heusesthedouble
differenceintroducedabove:thechangeovertimeforeachbuyoutfirmiscomparedwithanon-buyout
firmandtherelativechangeiscomputedbylookingatthedifferenceindifference.IncaseofEBITDAthis
is:
𝐷𝑖𝑓𝑓 − 𝑖𝑛 − 𝐷𝑖𝑓𝑓 = 𝐸𝐵𝐼𝑇𝐷𝐴
A,4CD
– 𝐸𝐵𝐼𝑇𝐷𝐴
15
A,4
– 𝐸𝐵𝐼𝑇𝐷𝐴
F,4CD
– 𝐸𝐵𝐼𝑇𝐷𝐴
F,4
Somefundsholdtheirinvestmentsdirectlyandhencewecanmeasureleveragedirectly.Othersuseholding
companies.Inthiscaseweneedtolookupall(historical)holdingcompaniesduringthebuyout.Thisinfoisnot
availableinthestandardaccountingdatawegetfromRAVNorProffasonlythecurrentholdingstructureisgiven.
-9-
PercentageChangesbetweenBuyoutTargetsandControls
FinancialMeasures
0to+1
0to+2
0to+3
EBITDA/totalassets
Netcashflow/totalassets
44.30%**
9.12%
0to+1
0to+2
0to+3
Returnonsales
EBITDA-margin
Assetturnover
25.36%
5.29%
8.95%***
-3.35%
9.30%
19.74%***
21.90%
16.37%
49.56%***
Currentratio
-4.40%
-67.23%***
-48.14%***
Source-Friedrich(2015)
32.47%**
23.60%
26.37%
52.14%
OperationalMeasures
Table2:FinancialandOperationPerformanceofBuyoutTargets:Source:Friedrich(2015).*Denotesstatisticalsignificanceat
the10%level,**denotesstatisticalsignificanceatthe5%level,***denotesstatisticalsignificanceatthe1%level.
Theresultsinthefirstpaneloftable2showthatfinancialperformanceimprovesrelativetocontrolfirms
overathree-yearhorizon,althoughnotallresultsarestatisticallysignificant.Operationalimprovements
alsodisplayapositiveoutcome(Returnonsales,EBITDAmarginandAssetturnover)ascanbeseenfrom
thesecondpaneloftable2.Alsonotehowthefirms’currentratiodeclines.Thishintsatanimproved
liquiditymanagementinthefirm,possiblycausedbystrictercontrolofcapitalinthefirmorbycloser
governance.
Insolvency risk, measured by the firm’s z-score, also decreases more than for comparable firms.16
(Friedrich,2015)usesamodifiedz-scorebelow(wherethemarketvalueofequityisdroppedinfavourof
thebookvalueofequity)andshowsthatinthebuyoutyearthez-scoreisinitiallylowerforbuyoutfirms
butimprovesremarkablyintheyearsafterthebuyout,uptolevelsabovethatofcomparablefirms.
𝑍’ = 0.717 ∗
Z'Score
NOPQ9DRFST94SU
VO4SU122342
Buyout
Control
+ 0.847
Y34.6SPD9DR2
V1
+ 3.107
6A[V
V1
+ 0.42
AOOQ67894:
V1
+ 0.998
^SU32
V1
.
Year
0
1.09
1.16
1
1.34
1.19
2
1.38
1.27
3
1.31
1.27
Table3:ModifiedZ-Score.Source:Friedrich(2015)
16
Az-scoreisanumericalvaluethattriestoevaluateafirm’slikelihoodofdefault.Thehigherthescorethelower
theprobabilityofdefault.Typically,z-scoresarecomputedforpublicfirmsandusethemarketvalueofequity.
-10-
Thereseemtobenoadverseeffectswithrespecttoemploymentalthough(Friedrich,2015)acknowledges
thatthisissueisdifficulttoevaluatewiththedataathand.Buyoutfirmsoftensellandbuyadditional
firmsafterthedealdate,somethingthatisveryhardtokeeptrackoff,andhemayendupmixingorganic
employmentgrowthwithemploymentgrowthfromacquisitions.17
Whatarethetaxconsequencesofbuyoutinvestments?
Thestateandsocietyareaspecialkindofstakeholderinfirmsthroughtheir(tax)claimonfutureprofits.
Yetfirmscanattempttochangethedistributionofcurrentandfutureprofitsthroughmoreaggressive
tax planning or by increasing leverage. Reducing taxes is a simple way of moving wealth from one
stakeholder in the firm (the state) to the equity owners. (Roald & Roti, 2015) attempt to answer this
questionbycomparingPEbackedfirmsandnon-PEbackedfirmswithrespecttotheirbehaviourtowards
measuresoftaxplanning.InordertodosotheyusetaxbalancesheetsobtainedfromtheNorwegiantax
authoritiesandmatchthesetofirms.Thisallowsforacomparisonbetweentaxandfinancialstatements.
Intotalfourspecificmeasuresareanalysed.Allmeasuresaimtoreflectsystematicdifferencesbetween
profits recorded in the financial accounts and in the tax accounts and are frequently used in the taxavoidanceliterature.18Yet(Roald&Roti,2015)showthattherearenosystematicdifferencesbetween
buyouttargetsandcontrolfirmsonanymeasureoftaxplanning.19Evenmoreso,theyshowthatGPsdo
notseemtotargetfirmswithalargepotentialfortaxsavingswhenselectinginvestments.
Theonlydiscernibledifferenceistheleverageratioofthefirmsgoingthroughabuyout.Heretheyfind
an increase in the leverage ratios of treated firms, consistent with the results in (Friedrich, 2015).
However,boththeoryandempiricalresearchsuggestthatamereincreaseinleverageisnotsufficientto
claimthatbuyoutdealsaredetrimentaltothepublic.Asmentionedbefore,ifincreasesinleveragelead
toanincreaseinefficiency,thentheultimateoutcomeisupintheairinthesensethattheproductivity
gainscouldleadtohigheroveralltaxpaymentsdespitehigherleverageratios.Unfortunately,wedonot
haveadefiniteanswertothisquestion.Betweentheresultsof(Roald&Roti,2015)and(Friedrich,2015)
wedonotknowwhetherhigher(longterm)leverageleadstoloweroveralltaxesorwhethertheefficiency
17
ImaginethefirmAhas10employees,acquiresfirmBwith5butalsolaysofffouremployees.Theneteffect
wouldbeareductioninemployment.Withoutbeingabletofullytrackadd-onacquisitionsthisquestionisdifficult
toanswer.Afullysatisfactorytreatmentwoulduseeitherestablishmentlevelemploymentdataortaxrecords.
18
ExactdefinitionscanbefoundinAppendixAandin(Roald&Roti,2015).
19
Seetable6–theonlyexceptionisthebuyouttargetsleverageratio.
-11-
increasesreportedcompensatefortheincreaseinleverage.
A back of the envelope computation suggests that the increase in productivity clearly dominates the
decreaseintaxes.Let’sconsiderafirmthatinitiallyhasapre-taxreturnonassetsof10%.Togenerate
one kroner of EBIT the firm needs ten kroner of capital. Taking the long-term debt numbers from
(Friedrich,2015),32%forabuyouttarget,a28%taxrate,a5%coupon,andzerodepreciationafirmpays
0.235kronertaxeswhileacomparablefirm(witha26%longtermdebtratio)willpay0.246kronerin
taxes,a1.1ørelossforthestate.Nowlet’sassumethatthebuyouttargetisabletoincreaseEBITDAby
23% after three years relative to comparable firms, as suggested by (Friedrich, 2015), while reducing
leverageto28%.
Assumptions
ROA
TaxRate
ChangeinEBITDA/TA
Interestrate
LeverageRatios
t=0
Buyouts
32%
Controls
26%
0% 10%
28%
23.60%
5%
Depreciation
t=3
28%
24%
Interest
t=0
0.16
0.13
t=3
0.14
0.12
Table1:AssumptionsEfficiencyGains
Taxesinyearone
Taxesinyearthreewith
actualleverageratio
Buyout
target
Taxable
EBIT income
Comparable
Firm
Difference
Taxes
paid
New
EBIT
Taxesinyearthreewith
compleverageratio
Taxable
income
Taxes
paid
Taxable Taxes
income paid Difference
1
0.86
0.235
1.236
1.096
0.307
1.116
0.312
-1.8%
1
0.87
0.244
0.88
0.246
25%
-3%
1
Thisleadstoanincreaseintaxesof0.065kronerto0.306kroner.Ifthebuyouttargetwouldhaveutilized
thecomparablefirm’saverageleverageratioof24%thentaxeswouldhaveincreasedto0.312kroner.
Comparablefirms’taxesmeanwhileincreaseto0.244to0.246kroner.Thischangeissolelydrivenbythe
reductioninthelong-termleverageratiofrom26%to24%.Thenetdifferenceforthestateisa1.1øre
lossinthefirstyearbuta6.5ørenetgainafterthreeyears(ora25%gainrelativetothecomparable
-12-
firm’staxes).AccordingtoFriedrich’snumbers,thestatewouldgainrelativetocomparablefirmsineach
year but the first. This simple calculation illustrates that a major increase in productivity can clearly
outweighanylossfromleverage;however,Ishouldcautionthatitismerelymeantasanillustration.The
problemisthatwecannotbesurethattheincreaseinleverageiswhatcausestheimprovementsinfirm
performance.Otherfactorsmightcausethisimprovement.
Amoreformalanalysisisnotaneasytask.Toproperlyshowthatsucharelationshipexistsonewould
havetoprovethattheincreaseinleveragecausesanincreaseinefficiencyandshowthatthisleadstoa
largertotaltaxpayment.
Internationally,(Badertscher,Katz,&Rego,2010)showthat,similartoNorway,USbuyouttargetsengage
inlesstax-planningthatpubliclylistedfirms.Ontheotherhand,higherleverageratiosintheUSmean
that US buyout targets have lower marginal tax ratios than public firms. Also, (Kaplan, Management
Buyouts:Evidenceontaxesassourcesofvalue,1989)firstreportedthattaxsavingscausedbyincreased
leveragearealargesourceofvalueforUSbuyoutfunds.20
BoardsandPrivateEquity
Our previous discussion centred around the fact that Private Equity firms seem to improve firm
performance.Wealreadydiscussedonepotentialchannel:higherleverage.Often,PEfundsaremajority
ownersinthefirmtheyinvestinto.Thismajorityownershipgivesthemcontroloverthefirmandraises
thequestionifboardsareatoolthatisactivelyusedbyGPs.
Obs.
Mean
99
54.1
GPOwnershipinNorwegianPEDeals
Median
Std.Dev.
55.6
30.3
Min
Max
2.18
100
Table4:AverageShareOwnershipbyindividualGPs.
(Bienz,Thorburn,&Walz,2016)showboththeaverageandmedianownershipfor99NorwegianPEdeals
between1997and2009.OnaverageGPsown54%oftheequityintheircompaniesandownthemajority
20
Kaplanalsodocumentslargeincreasesinfirmproductivity,butIamnotawareofanystudythattriesto
determinetheoverallchangeintaxedpaidbycomparingthetwoeffectsforUSbuyouts.
-13-
in57%ofalldeals.21 Hence,inalargefractionofalldealstheGPshavetheabilitytocontroltheboardas
theyseefitto.22
Thequestiontheniswhetherorhowthiscontrolisusedandhowitaffectsportfoliocompanies.DoGPs
directlyinteractwithmanagers;dotheyutilizetheboardorarethereothermechanismsatwork?Board
workcanbedemandingintermsofattentionandtime.PEfundsoftenmanagemultiplefundsandwilltry
to use their time effectively. This begs the question whether boards play a major role in supporting a
portfolio company. On the other hand, boards may matter in cases where the GP owns less than the
absolutemajorityandinsituationsthatinvolveminorityshareholdersotherthanmanagers.
(Farran&Lâm,2015)exploretheboardaspectoftheGP-portfoliocompanyrelationshipandlookatCEO
turnoveraswellasinvestigatingwhetherthepresenceofGPsonthefirm’sboardaffectsthesuccessof
theportfoliocompanies.Inthelatteranalysis,theyfollow(Wintoki,Linck,&Netter,2012)intheirdynamic
panelestimationapproachandaskwhetherspecificaspectsofafirm’sboardcanaccountfordifferences
inoutcomesamongallfirmsthatwereboughtbyaPEfund.Theideabehindadynamicpanelapproachis
thatweshouldbeabletoreducetheendogeneityoffirmperformancebyincludingpastperformanceon
therighthandsideoftheregressionequation.Theyregresscurrentperformanceonpastperformance
andtheextentoftheGP’sboardrepresentation.Otherright-handsidevariablesinclude:CEOchange,an
exitdummyiftheGPhasdivestedhimself,asyndicationdummy,adummyforaNorwegianGP,FirmSize,
anindustryandayeardummy.
Maybethemostinterestingpartoftheiranalysiscanbefoundintable7oftheirthesiswheretheyshow
thatthepresenceoftheGPontheboarddoesnotseemtoaffectfirmperformance.Thereareseveral
different ways to interpret this result. The most straightforward one is simply to say that boards are
superfluousasthereisnorelationshipbetweentheboardandperformance.Thisistheauthor’spreferred
interpretation:“Wefindthatgeneralpartnersdonotprioritizetheboardaslongaseverythingisgoing
according to plan. […] Furthermore our findings suggest that the board is neglected and interaction
21
HerewetreateachGPasanindividualinvestor.HencewecanhavemorethanoneGPasashareholderandwe
donotmeasuretotalPEownershipbutownershipbyindividualGPs.Orputitdifferentlywedonotcontrolfor
syndicationofinvestmentsbetweendifferentGPs.Givencurrentrulesonreportingindividualownershipweoften
donotknowthelevelofmanagerialownership.
22
However,wewouldliketothankourrefereeforpointingoutthatthesenumbersareactuallylowerthanthe
typicalownershipstructureforaprivatelyownedfirm.
-14-
between general partners and management is conducted in alternative ways”. There are two other
possible interpretations. First is the possibility that their analysis simply lacks power to detect this
relationship.Powerisastatisticaltermthatdenotesastatisticalmethods’abilitytodetectasignificant
relationship.Thesecondisthattheboardhasbeenchosenoptimallyandhenceitshouldnothavea
measurable effect. This interpretation is orthogonal to the first explanation as it takes the lack of any
relationshipasanindicationthatthecurrentresultrepresentsanequilibriumoutcomethatcannotbe
improvedupon.
Ifwebelievethefirstinterpretation,namelythatboardsdonotaddvalue,thenthefindingsareinline
withotherpapers,suchas(Wintoki,Linck,&Netter,2012)thatfailtoestablisharelationshipbetween
boardcharacteristicsandfirmperformanceoncepastperformanceisincludedintheanalysis.However
the results are in contrast to (Cornelli, Kominek, & Ljungqvist, forthcoming) who show that boards of
buyouttargetsinEasternEuropeanduringthetransitionperiodsinthe1990sseemedtoimprovefirm
performance.Inaddition,ifwebelievethatboardsdonotaddvalue,thenthisfindingofcourseraisesthe
questionaboutthechannelthatGPsusetoinfluencetheirfirms.InterviewswithGPssuggestthatthey
oftendonotinvolveboardsinthedaytodayworkwithfirmsbutinsteadworkwiththeexecutivedirectors
directly.
Anotherveryeffectivewaytochangeafirmisthereplacementofthefirm’sCEO,anotherpartof(Farran
&Lâm,2015)analysis.WhileCEOturnoverishigh(about50%ofallCEOsendupbeingreplaced)they
reportthatinitialreplacementsarenotexplainedbyobservablefactors,suchaspastperformanceorfirm
characteristics.Badpastperformanceseemstoplayarole,butonlyinlaterstagesoftherelationship.
MostlikelyunobservedCEOcharacteristicsplayaroleintheinitialreplacementdecision,soitispossible
thattheseinitialreplacementsaredrivenbyaperceivedlackofmanagerialexperienceinexistingCEOs
butwithoutbetterdatathisquestionishardtoevaluate.
Conclusion
WehavediscussedsomeaspectsPEownershipofNorwegianfirmsinthispaper.Inparticular,thefinancial
and operational performance of these firms was analysed. On most measures (both financial and
operational)usedfirmsseemtoimproverelativetoacarefullyselectedgroupofcontrols.Thereislittle
indicationthattheseimprovementsaretothedetrimentofotherstakeholders.Thereisalsoevidence
thatPEfirmsdonotsystematicallyengageintaxplanningtoalargerextentthancomparablefirms.Finally,
-15-
thereseemstobeanindicationthatPEfundsdonotuseboardsinordertoimprovethefirmstheyown
butthatotherchannelsarebeingused,suchasdirectcontactbetweentheGPandtheCEO.
There are also many questions that should still be answered. For example, the interaction between
leverageandperformanceimprovementsneedsbeanalysedinmoredetail.Wealsoknowverylittleabout
theeffectofPEownershiponcompetitorsortheeffectontheindustriesinwhichthebuyouthappens.
Bibliography
Axelson,U.,Strömberg,P.,&Weisbach,M.(2009).WhyareBuyoutsLevered?TheFinancial
StructureofPrivateEquityFirms.TheJournalofFinance,64,1549-1582.
Badertscher,B.,Katz,S.,&Rego,S.(2010).TheImpactofPrivateEquityOwnershipOnPorfolio
Firms'CorporateTaxPlanning.Workingpaper.
Bienz,C.,Thorburn,K.,&Walz,U.(2016).Coinvestmentandrisktakinginprivateequityfunds.
WorkingPaper.
Boucly,Q.,Sraer,D.,&Thesmar,D.(2011).GrowthLBOs.JournalofFinancialEconomics,102,
432-453.
Cornelli,F.,Kominek,Z.,&Ljungqvist,A.(forthcoming).MonitoringManagers:DoesitMatter?
JournalofFinance.
Farran,H.,&Lâm,K.-A.(2015).CorporateGoveranceinPrivateEquity:DoBoardsReally
Matter?MasterThesis,NHH.
Friedrich,T.(2015).TheImpactofBuyoutsonCompanyPerformance:EvidencefromNorway.
MasterThesis,NHH.
ISSA/S:TheBuyout,N9-214-027(HarvardBusinessSchool2013).
Jensen,M.(1986).Agencycostoffreecashflow,copratefinanceandtakeovers.American
EconomicReview,76(2).
Johnson,S.(2014,April4).Reuters.RetrievedFebruary9,2016,from
http://www.reuters.com/article/sweden-privateequity-idUSL5N0MP3A020140404
Kaplan,S.(1989).ManagementBuyouts:Evidenceontaxesassourcesofvalue.TheJournalof
Finance,44,611-632.
Kaplan,S.,&Strömberg,P.(2009).LeveragedBuyoutsandPrivateEquity.JournalofEconomic
Perspectives,22,121-146.
Lichtenberg,F.,&Siegel,D.(1990).Theeffectsofleveragedbuyoutsonproductivityand
relatedaspectsoffirmbehaviour.JournalofFinancialEconomics,27,165-194.
Michel,A.,&Shaked,I.(1991).RJRNabisco-ACaseStudyofaComplexLeveragedBuyout.
FinancialAnalystsJournal,47,15-27.
Roald,A.K.,&Roti,T.(2015).TaxPlanninginNorwegianPrivayeEquity-BackedCompanies.
MasterThesis,NHH.
SealedAirCorp.'sLeveragedRecapitalization(A),HBS9-294-122(1994).
Shleifer,A.,&Summers,L.(1988).BreachofTrustinHostileTakeovers.InA.J.Auerbach(Ed.),
CorporateTakeovers:CausesandConsequences.Chicago:UniversityofChicagoPress.
Wintoki,B.,Linck,J.,&Netter,J.(2012).Endogeneityandthedynamicsofinternalcorporate
governance.JournalofFinancialEconomics,105,581–606.
-16-
AppendixA–MeasuresofTaxPlanning
1.
Thedifferencebetweenprofitsbetweenthetaxandfinancialaccounts(TotalBookTaxDifference)
2.
Discretionarydifferencesbetweenfinancialandtaxaccounts
3.
𝐶𝑎𝑠ℎ𝑒𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒𝑡𝑎𝑥𝑟𝑎𝑡𝑒 = 4.
𝑀𝑎𝑟𝑔𝑖𝑛𝑎𝑙𝑇𝑎𝑥𝑅𝑎𝑡𝑒 =
FS2f4Sg32TS9h
(jP354Sg9DkOl3–^T3k9SU943l2)
FS2fVSg32jS9h
qT3PS49DRjPOr942
-17-