Financial distress and bankruptcy prediction an appropriate models for listed firms in vietnam

Problemstatement

The liberalization of financial institutions has sparked regional competition and significantly impacted national economies, including Vietnam, which was not immune to the effects of the Global Financial Crisis in late 2008 This crisis led to macroeconomic challenges, such as a sharp increase in unemployment and a decline in economic growth The rise in credit risk resulted in the bankruptcy of numerous businesses, including some deemed "too big to fail." In Vietnam, the Ministry of Planning and Investment reported that 68,350 new businesses were established, while over 7,000 firms ceased operations or went bankrupt Additionally, 47,600 firms faced substantial financial difficulties, prompting temporary shutdowns The trend of rising bankruptcies continued, with 18,916 firms declaring bankruptcy in 2016, marking a 4.2% increase from the previous year.

Somecasesareworthconsideringindetails.TheIDICOJoin- stockcompany(PXL)incurredlossesinthreeconsecutiveyearsandth ecompanywasforcedtobedelistedfromtheHoChiMinhC i t y Sto ckExchange(HOSE)in2016.Inthesamefate,VienDongMedi cineCompany(DVD)alsofiledfor bankruptcyin thesameyear.Intheareaofinternationaltradesector,thePhuPhongJoint-

2 stockCompany(PPG)hadtobe dissolvedbecausetheirproduct sproducedfromtheobsoletet e c h n o l o g y couldnotme etthehigh- qualitydemandfromtheEurop eanmarket.Moreover,various mergerandacquisition (M

&A)activities werealsota king placewith smaller co mpaniesbeing acquiredbya muchlargercorporationopera tingin thesameareaofbusiness.Anot ionalexamplecanb e devoted t o t h e M&Aa c t i v i t i e s w hereThanhC o n g (TTC)acq uireds m a l l e r businessesi n c l u d i n g t h e N i n h H o a

Sugar( N H S ) andG i a LaiN hietDien( S E C ) Likewise, a substantialn u m b e r offo reignfirmswantedtoexpandt hemarketinVietnamaswell byconductingvariousM&A activities.Inparticular,Masa nGroup(MSN)wasmergedb ySinghaAsiaHoldingbyth eendof2015.T h e BigCRetai lSystemandtheCentralGroup

Thailandcameintothesamebr inApril2016ortheTCCHoldingCorporationused711milliondollarstotakeovertheMetrogroup in January2016.

The synergies of M&A activities extend beyond the focus of this study; however, senior management aims to avoid being taken over by larger corporations From a macroeconomic perspective, an excess of M&A activities can lead to market power concentration among a few dominant players, negatively impacting consumers Additionally, the rise in bankruptcy filings among firms in Vietnam over the past six years has highlighted credit risk as a significant concern for both company management and policymakers.

Thisstudyisconductedtoconsider,examine,andquantifythefinancialdistressforlistedfirmsi nVietnam.The choice ofVietnaminthisstudyisinteresting.Vietnamhasbeenachievingt h e secondhighesteconomicgrowt hratesintheregionandalsointheworldinrecentyears.Fort h e lastalmost10yearsorso,mostcountriesinth eAsianregionfacedfinancialdifficultiesarisingfromtheglobalfinancialeconomic(GFC).Vietnamescap edfromtheGFCandachievedthehighlevelo f economicgrowth.Vietnamhasbeeng e n e r a l l y consid eredr i s i n g starsi n t h e s t a b l e economicgrowthand development in theworld.

The Pacific Partnership has effectively become obsolete under the new U.S administration, casting doubt on the future of the Trans-Pacific Partnership (TPP) as Australia, Canada, and other member nations consider moving forward without the U.S Instead, Australia is exploring a new trade partnership with the 10 ASEAN nations, potentially led by China While there are no substantial discussions among nations yet, Vietnam is committed to advancing its national economy and integrating it into the global market, independent of the TPP's establishment In light of rapid regional integration, there is an urgent need to understand and compare the credit risks of listed firms in Vietnam, as industry-level credit risks are crucial to this analysis.

Standardapproachestom e a s u r i n g creditr i s k s f o r l i s t e d firmsi n V i e t n a m ando t h e r countriesincludetheaccountingbasedapproachandthemarket- basedapproach.Inthisregard,t h e accounting-basedmodels(suchastheAltmanZ- scoreandtheEMSmodels)andthemarket- basedmodels(suchastheMertonmodels)areemployedtoestimatingthedefaultprobabilityofl i s t e d firmsin thecountryofinterest.

Thisstudydepartsfromthecurrentpractice.Itistheintentionofthisstudytodevelopanewm o d e l in whichthethreepillarsofcreditrisksareconsidered:(i)factorsfromthemarket-basedm o d e l s ; (ii)factorsderivedfromtheaccounting- basedmodels;and(iii)selectedmacroeconomicfactorswhicharelandedonastrongtheoreticalground.Thisapproachisexpectedtoprovideacomprehensiveevidenceinrelationtofinancialdistressandbankr uptcyoflistedfirmsinVietnam.O u r intensiveliteraturereviewindicatesthatthisstudyisthefirstof thiskindinVietnamandp r o b a b l y oneof thefirstfewstudiesin theregion.

ResearchObjectives

 First,thisstudyisconducted toidentify earlywarningindicatorsofcorporatefinancial distress.

 Third,selectedmacroeconomicfactorsaffectingthefinancialdistressandbankruptcyoffirmsa reconsideredandincludedinthe so- called comprehensivemodel ofbankruptcy predictionforVietnam.

 Fourth,aconsiderationo f difference,i f any,betweendifferentp r e d i c t i o n m o d e l s o f financialdistressandb a n k r u p t c y f o r l i s t e d firmsbeforeandaftert h eG l o b a l FinancialC r i s i s

Researchquestions

Ino r d e r t o achievet h e aboveresearchobjectives,t h e f o l l o w i n g researchquestionshaveb e e n rais ed:

 Whatarethe fundamentalindicatorsusingaccountingdata(theAltmanmodel)whichcanb e usedtomeasure financialdistressand bankruptcyforlistedfirmsinVietnam?

 Whatarethemostrelevantindicatorsusingmarketdata(theDistancetoDefaultmodel)f o r measuringfinancialdistress and bankruptcyforlistedfirmsinVietnam?

 Dothepredictioncapacityofthemultivariatediscriminantanalysis(MDA),theDistancet o D efault(DD)andt h e comprehensivem o d e l changebetweencrisisandn o n - c r i s i s periods?

Structureof thethesis

Ther e s t o f t h i s r e s e a r c h i s structuredasf o l l o w s F o l l o w i n g t h i s Introduction,chap ter2 presentst h e backgroundandliteraturereviewo f t h e relatedissues.D a t a description,selectionm o d e l s , theoreticalandempiricalmodelarediscussedinchapter3.Chapter4detailstheresultsand discussion.Severalmajorconclusionsandpolicyimplicationsareconcluded inchapter5.

Thischapterincludesthreesections.ThefirstsectionpresentsthemajorcauseofchoosingVietna mbyanalyzingVietnam e c o n o m y fromt h e pastt o p r e s e n t Moreover,th e effecto f t he globalfi nancialcrisis(GFC)onVietnameconomywouldbeexaminedthoroughlyinthesecondsection.Thela stsectionrepresentsanimportantconceptwithreferencetothefinancialdistresslikelihoodinvariousap proaches,comparisonamongdefault-riskmodelsaswellassomereviewso f defaultprediction inAsianations.

WhyVietnam?

Vietnam is a rapidly developing nation in Asia, having overcome significant wartime challenges throughout its history On September 2, 1945, Nguyen Ai Quoc proclaimed the establishment of the Democratic Republic of Vietnam after a century of political instability However, lasting peace was short-lived as the French and then the Americans invaded the country The North Vietnamese army, led by General Vo Nguyen Giap, achieved a decisive victory against the French at the Battle of Dien Bien Phu, resulting in the Geneva Agreement that divided Vietnam into two regions North Vietnam became the Democratic Republic of Vietnam under the Viet Nam Worker’s Party, while South Vietnam was known as the Republic of Vietnam, initially led by the French and later by the Americans Ultimately, North Vietnam unified the country on April 30, 1975, leading to the establishment of the Socialist Republic of Vietnam in 1976.

Becauseofthelongwarsandcivilwars,Vietnamsufferedfromtheoutdatedtechnologyands l o w e conomicgrowthcomparedwithdevelopedcountries.Afterthewarfinishedin1975,theC o m m u n i s t Partytook thegovernmentandthentheybuilt theeconomyof both nations followingthesocialistorientation.Thanksto thegreatsupportfromthe So VietUnion,Vietnamhadalarges t e p ofbuildinganddevelopingthenation.TheSovietModelwasapplied successfullyinVietnambytheFive-YearPlans.AsinVo,D.H.(2008)’research,thefirstFive- yearplanwasemployedi n theperiodfrom1976to1980.Duringthepost- warreconstruction,thisplanemphasizedthe

Vietnam underwent significant transformations from small-scale to large-scale production, particularly in the heavy industry and agriculture sectors, achieving impressive growth rates of around two percent per year However, the country's challenging economic scenario persisted post-liberation, with GDP growth stagnating at only 0.4 percent annually, while prices surged by approximately 20 percent and the population increased by about 2.3 percent each year This situation has been referred to as the "failure of command economy." Between 1981 and 1985, Vietnam's economy was characterized by a multi-component structure, with North Vietnam featuring individual, collective, and state-run sectors, while South Vietnam was dominated by private capitalism along with a mix of state-private joint ventures, as noted by Riedel and Turley.

(1999).TherewasalargedifferencebetweentheNorthernandSouthern,itmeanst h a t theunification ambitionsofthetwoeconomiesweremetthegreatchallenge.So,thesecondFive-

YearP l a n w a s approvedbyt h e P a r t y Congresst o dealw i t h e c o n o m i c problemsi n t h e previ ousperiodandsetupthenew“familyeconomy”.F o l l o w i n g Vo,D.H.

The initial improvements in national income during the 1980s were overshadowed by a sharp rise in inflation, peaking at 588 percent in 1985, which ultimately undermined the achievements of the economic plans and failed to enhance living standards In light of these challenging circumstances, the Communist Party of Vietnam (CPV) acknowledged the necessity for comprehensive innovation and initiated extensive economic reforms known as the “Doi Moi” transition, emphasizing the significance of a market economy.

Since1986,Vietnamhasshiftedfromacentrally-plannedtoamarket- orientedeconomy.T h i s overallinnovationwasofficiallythroughtheCPV’sSixthNationalCongress.Int hefirststepo f therenovationprocess,the“entrepreneurialpolicy- makers”intheperiodfrom1986to1994,t h e newForeignInvestmentLawpublishedin1987.Afterth atVietnamwelcomedahugeamounto f ForeignDirectInvestments(FDI)enteredalmosteconomicsector

7 sandthewaveofFDIreached1 0 percentofGDPin1994asdiscussedintheWorldBank(1999).Moreover,theCorporateandPrivateEnterpriseLawwasmodifiedtoencouragethedevelopmentofprivatebusines s.Vietnamwasafoodshortagetobecomethethirdriceexporterintheworld.Therefore,thelivingstandar d

1994,thisfingerwasdoublecomparedt o thepreviousperiod.Inordertoinnovatemorestrongly,Vietna mhasignoredthepastconflictw i t h t h e U S A byn o r m a l i z i n g relationsw i t h t h e m T h i s polit icali n n o v a t i o n openedt h e greatchancetocorporatetodevelopednations,internationalorganizatio nsandmultilateraldonorsasADBasw e l l asW o r l d Banki n t h e p e r i o d from1 9 9 5 t o 1 9 9 9 o r “ E c o n o m i c integrationandadoptionofmarketeconomy”.However,theriskoflosingnationalpoliticst ootherpartieswashigh.TheCPVwouldliketodevelopanationaleconomywithoutanypoliticalthreats.To balancet h einterests,C P V hascontinuedi t s “DoiM o i ” programw i t h threem a j o r targets.

(iii)Theinnovationprocessmustbegradualineveryscenario.Alloftheseobjectiveshavebeeni m p l e m e n t e d byt h e C P V aswellasgovernment’ss o c i o - economicandforeignpolicies.ThewaveofFDIplayedthekeyrole ineconomic growthbecausethe FDImightcreateamillionjobsforthelabormarket.In1996,over10billiondollarsofFDIandabill iondollarscamefromtheADBaswellasWorldBankflowedintothenation.Therefore,theVietnam’GDPin creasedconsiderablyin9.5percentand 9.3percentannuallyin 1995and1996 Thisis thehighestraterecordedafterDoiMoiperiod.

Followingthepreviousachievements,Vietnamenteredanewstageofdevelopmentortheecono micboomandemergingculturalvalueintheperiodfrom2000to2006.Thefinancialmarketi s extendedrapi dlyineconomictermsbutthescaleofVietnamstockmarket’scapitalizationwasinsignificant,approxim ately1percentofGDPin2000.However,thisfingerincreasedsharplyin

22.7percentattheendof2006andVietnamIndexwasalsorisen150percent.Thestockmarketwasa“ferti leland”forfinancialinvestors.

AccordingtoGeneralStatisticsOfficeofVietnam,thee c o n o m y wasrankedat58 tht h e largesteconom yintheworldin2006andtheGDPincreasedanaverageof7.5percentinperiod2000to2005.Acombinati onofhigheconomicgrowthrate,lowinflation,privatizationofState-

OwenE n t e r p r i s e s ( S O E s ) anda largea m o u n t o f FDIassistedVietnamtoobtainthecenturygo alorreducingpovertyratefrom28.9percentto15.5percentin2 0 0 6

Inmorerecent,in theperiodfrom2007 topresent,despitehavingsufferedfromtheimpacto f theeconomiccrisisin2008.Vietnamhasintegrate

8 ddeeplywiththeworldandhasacquiredagreatnumberof economicachievements.VietNamwasamemberof manyorganizations such as theWorldTradeOrganization(WTO),ASIANeconomiccommunity(AEC),Asia-

PacificEconomicCooperation( A P E C ) , InternationalM o n e t a r y Fund(IMF)a n d W o r l d Bank (WB).RegardingWordBank(2016),theGDPgrowthofVietnamis6.7percentin2015andthisnumberhelp sVietnamstayinthehighgrowthcountriesintheworld.Furthermore,thelivingstandardsareconstan tlybeingimprovedovertheyearswiththeGDPpercapitaobtained2,107dollarsin2 0 1 5 aswellast h i s i n c o m e i s f i v e t i m e s highert h a n i n 2 0 0 0 Therefore,Vietnami s o n e o f emergingmarket attractingthelargeforeigninvestment in theworldaswell asVietnamhas beenl o o k i n g alittletigereconomyinSoutheastAsia.

Backgroundto theGlobalFinancialCrises

Theliberalizationoffinancialhasbeenrisingdramaticallyinrecentdecades.Theforeigninv estments(FDI),diversifiedrisk,andinternationaltradewerestimulatedstronglya m o n g nations.Seve ralcountrieshaveobtainedthehigheconomicgrowthandlowunemploymentrate.O n thecontrary,so menationshave fallenconsiderableeconomicvolatilityandfinancialcrises.Iti s theabsolutetruththattherearetwolargef inancialcrises.ThefirstcrisisistheAsianFinancialC r i s i s (AFC).ItbeganinThailandin1997,whic hspreadtoother developingregionsinAsiaandEasternEuropeduring1997- 1998.ThesecondcrisistookintoaccountthisstudyistheGlobalFinancialCrisis(GFC)- startedintheUSin2008anditexpandedrapidlyinthewholeworldwitha dominoeffectduring2008–2009.

The United States was the epicenter of the Global Financial Crisis (GFC) in 2008, which originated from the collapse of the real estate bubble in late 2005 As investment and financial institutions struggled to secure funding, many faced insolvency, leading to significant turmoil in the financial sector The crisis peaked in October 2008 with the bankruptcy of Lehman Brothers, triggering a domino effect that threatened major U.S banks like Morgan Stanley, Citigroup, and AIG This financial turmoil extended beyond the U.S., impacting Western nations such as England, Spain, Iceland, and Ireland The stock markets in New York, London, and Paris recorded historic lows, and a year later, the crisis spread to Asia, affecting even Japan, which, despite having a strong financial system, saw the Nikkei Index plummet to its lowest point in thirty years Similarly, China, as the world's second-largest economy, also felt the repercussions of the GFC.

Viet Nam nationwasalsonegativelyimpactedoftheGFCin2009aftertheeconomicgrowthmiracleobtainsalwaystw ofingers.UndertheimpactoftheGFC,thebankingsystemofVietnamalsofacedagreatnumberofth erisks.ThenegativemacroeconomiccomponentsfromtheGFCmayenhances i g n i f i c a n t l y t hecreditriskandbankruptprobability.

8 5 t o 9 percent.D u e t o t h e i n s t a b i l i t y ofmacroeconomic,t h e NationalA s s e m b l y h a d ad justedthegrowthobjectiveto7percentin2008.Unfortunately,theGDPgrowthofVietNami n 2008rea chedonly5.7percentorlowerthan3.3percentcomparedwiththebeginninggrowthtarget.

Aglanceatthegraph1providedrevealstheGDP growthrates ofVietNamduringtheperiodf rom2 0 0 0 t o 2 0 1 5 Iti s evidentt h a t V i e t n a m experiencedconsiderablef l u c t u a t i o n i n econo micgrowthrate.Itfluctuatedatsomewherebetween6.2and7.5percentpriorto2007,beforedroppingsharplyt o5.4percentin2009.VietNamwasheavilyinfluencedbytheGFCin2008suchastheexportdeclin e,thestockmarketsfaceddifficultiesandinvestorsmetdisadvantages

Viet Nam whereastherealestatewasfrozen.Thus,theGFCisamajorcauseofthedramaticdecreaseintheGDPgrowth ratein2008-

2009.Inthishardtime,theVietNamgovernmentemployedfivegroupso f policiesincludingthepolicies tostimulatingproduction,business,andexport;thepolicyfordemand- stimuluso f investmentandconsumption;t h e expandingfiscalandm o n e t a r y policy;i m p r o v i n g theSocialWelfaresEnsuringpolicies;andenhancingthemanagementandorganizationpolicies.These p o l i c i e s p r o m p t l y andp r o p e r l y p u t VietNamo u t o f t h e crisisr a p i d l y T h e f o l l o w i n g ayearsawarapidclimbinGDPgrowthrate,tosomewhereinthevicinityof6.4percenti n 2 0 1 0 W h e n t h e s t i m u l u s packagewasn o longerworks,t h e globalfinancialcrisisaffectedc o n t i n u o u s l y theeconomyandtheGDPgrowthratewasdecreasedexponentiallyinthebottomof

2 0 1 3 beforeh a v i n g t h e considerablegrowthin2015with6.7percent.Therefore,theVietnam economyhasreturnedtopre-crisisgrowthandwasexpectedto bestrongergrowthin thefuture.

Vietnamnotonlymetthedeclineofgrowth,butalsothisnationfacedthehighinflationint h e peri od2008–2012.Increasingfromapproximately-

1.7percentin2000to7.8percentin2004,t h e inflationratet h e n reacheda peako f 2 3 1 percenti n 2 0 0

2 0 0 8 , werea m a j o r reasonf o r t h e highinflation.Aftert h a t , i t falls

11 significantlyin7 1 p e r c e n t i n 2 0 0 9 b e c a u s e of a combinationo f f i s c a l andm o n e t a r y policies. However,thislowrateexistedonlyinayear.Theinflationrateincreasedrapidly18.7percentin2 0 1 1 whent h e consumingstimulationbudgetm a y notaffectp o s i t i v e l y ont h e e c o n o m y T h i s follo wedbytheperiodofexponentialdecline,withtheinflationrateinthecountryplungedtoal o w o f j u s t 0 9 percenti n 2 0 1 5 T h i s figurei s t h e lowesti n aroundfifteenyearsaswellasdemonstratesagoo dsignaltorehabilitatetheeconomyafterimmersingthestagnancyinalongperiod.

2009witht h e positiveeconomicscenariosuchashighgrowth,lowinflation,lowfinancialexposureand lowdefaultprobability.Thesecondstageispost-

Literaturereviewoncreditmodels

Backgroundtocorporatefinancialdistressmodels

Corporate bankruptcy progresses through five distinct stages Initially, a firm enters an incubation phase where financial difficulties begin to surface This leads to a realization of financial embarrassment, followed by a state of financial insolvency, indicating that the company lacks sufficient funds to meet its obligations As the situation worsens, total insolvency occurs when a firm's physical assets are valued lower than its debts The final stage is confirmed insolvency, where the court officially declares the firm bankrupt, necessitating the sale of assets to satisfy creditors, as noted by Poston, Harmon, and Gramlich (1994) It is important to differentiate between financial distress and bankruptcy; financial distress refers to a firm's inability to cover obligations due to operational losses or high fixed costs, whereas bankruptcy signifies a complete cessation of business operations In some instances, financial distress can be identified before insolvency, allowing for potential recovery through effective board policies, thus mitigating bankruptcy risks Ultimately, all businesses must endure a liquidation period or face dissolution A study by Odosiou & Lee (1993) highlights that nearly all U.S businesses experience this challenging process.

A businessesf r e q u e n t l y loset h e a b i l i t y to payloana p p r o x i m a t e l y twoyearsb e f o r e proc eeding bankruptcy.RegardingBaselC o m m i t t e e on BankingSupervision (2016),th e firmf a l l s i n t h e defaultstatewhentheymaynotrepaybankdebtmorethanapproximately90days.Consequently, thedefaulti s a q u a l i t y proxyforbankruptcyorcreditrisk.

Financial distress imposes significant costs on debtors, prompting them to take measures to mitigate these expenses (Tinoco & Wilson, 2013) Researchers have developed models that predict and assess the likelihood of financial distress by integrating accounting, market, and macroeconomic variables Gilson (1997) found that high leverage contributes to financial distress due to increased financial expenses and obligations This distress can escalate quickly when cash flow fails to cover financial obligations, potentially resulting in creditors gaining control over the company, as noted by Wruck (1990) When the cost of financial expenses surpasses asset value, the risk of bankruptcy rises sharply Conversely, Foster (1986) argued that financial distress is not insurmountable and can often be resolved through the restructuring of business activities.

Numerous research projects have focused on predicting corporate financial distress, beginning with Beaver's (1966) pioneering accounting-ratio-based model, which has significantly influenced both commercial and academic spheres Beaver utilized a dichotomous classification test to identify financial ratios for bankruptcy prediction, highlighting that the working capital/debt ratio accurately predicts 90% of companies one year before bankruptcy, while the net income/total assets ratio identifies 88% However, this research only considers individual financial ratios and does not incorporate all financial ratios or market indices in failure prediction E.I Altman (1968) advanced the field by developing a multivariate statistical model that distinguishes between failed and non-failed firms, identifying 22 financial ratios categorized into profitability, activity, liquidity, solvency, and leverage Using multivariate discriminant analysis (MDA), Altman's Z-score model accurately predicted bankruptcy 95% of the time one year prior, although its predictive accuracy declines significantly over two to four years, and the model is primarily applicable to publicly traded firms.

E I.Altman,Haldeman,andNarayanan(1977)’sstudyexploredanewfinancialdistressm o d e l (ZETA).T h e d a t a included5 3 failedf i r m s and5 8 non- failedf i r m s concentratingi n manufactureandretailfirmsfrom1969to1975.Throughtheadjustmentth efinancialratiosoftheo l d modelin1968,thesevennewratiosre- estimated successfully thenewmodelbyreplacing bookvaluetomarketvalue.Therefore,ther esultmightcorrectlyidentify95%thefirmsoneyearp r i o r tobankruptcy,andespecially,70%withfi veyearspriortofailure.However,thisresearcho n l y wasappliedinanarrowarea(manufacturerandreta ilerfirm)orbecauseoflackingtheprivatefirm'sdata,hedidnottestthemodelonthesecondarysample.E.I.Altm an(2000)improvedsomedisadvantagesinthemodelin1968and1977byrevealingtheZ’’- scoremodelthatmayonlyhavef o u r financialratiosafterdroppingoutanX5variable(Sales/

Totalassets).Itcouldmeasurethefinancialdistresspredictionbetterthanhistwooldmodels.Alternativ ely,Z’’- scoremodelhastoa considerablen u m b e r o f assumptionsf o r i n d e p e n d e n t variables,s o , t h i s m o d e l i s l i k e t o b e difficulttoapplyinthereality.Ontheotherhand,theAltman’sZ- scorewastesteddirectlyinsomepaperssuchasZmijewski(1984),Holmen(1988),andBegley,Ming,and Watts(1996).Namely,Begley’sresearchi n 1 9 9 6 employedt h e Z - scorem od el t o predicting1365 firmsi n the period1800-

1890with78%accuracyrate.ThefinalversionofZ- scoremodelwastheemergingscoremarketmodel(EMS).EMSmodelincludedthetypicalcharact ersoftheemergingmarket.ESMseemedappropriatetoestimatethedefaultprobabilityfordevelo pingcountriesandrankingthefirmsbythespecificscore.Namely,therewerethreeimportantcompone ntsimpactingstronglyont h e EMSmodel.

(i)Thecurrencydepreciation(ii)itsindustryintegration(iii)thecompetitionint h e industry.EMSmo delwasappliedtotheMexicanfirmsbeforethecrisisin1994byE.I.Altman(2005).Thespecificmodel is that:

The Z-score model, while widely used for bankruptcy prediction, has limitations in accurately reflecting timing issues, as highlighted by Ohlson (1980) He introduced a logit model using data from 105 bankrupt and 2,058 non-bankrupt firms in the U.S between 1970 and 1976 Ohlson's model incorporates various financial metrics, including the logarithm of total assets relative to the GNP price level to indicate company size, total liabilities over total assets to assess financial structure, and current liquidity ratios Additionally, he evaluated performance through net income to total assets and operational funds against total liabilities Zmijewski (1984) further contributed to this field by employing accounting ratios, such as net income to total assets and total debts to total assets, to measure financial distress.

(iii)currentassetstototalliabilities,underrandomexogenoussamplingandtheP r o b i t regressi on.Whereasalmostthebankruptcyassumedthatthemacroeconomicscomponentsm a y notimpactont heiraccounting- basedmodel,thecontroversialhappenedinthereality.Theinterestrateandinflationfactorsaffectstro nglyonthefinancialdistressmodel.Mensah(1984)'ss t u d y reevaluatedthecorporatebankruptcymode l.Hecreatedfourmodelsforfourperiodsoftime1972-1973,1974-1975,1 9 7 6 - 1 9 7 7 and1978-

1979to coverthechangeof theeconomicenvironment.Theoutcomeindicatesthatthenewmodelpredi ctedexactlyoverthefoureconomicstages.

In Greece, Theodossiou (1991) utilized both Logit and Probit models from 1975 to 1986 to identify early warning indicators of financial distress, concluding that the Logit model is likely superior Similarly, Ugurlu and Aksoy (2006) conducted a study in Turkey using multivariate discriminant analysis (MDA) and the Logit model on 27 failed and 27 non-failed firms in the Istanbul stock market from 1996 to 2003, finding that the Logit model outperformed MDA In a more recent study, Stanisic, Mizdrakovic, and Knezevic (2013) developed a bankruptcy prediction model in Serbia, employing Logistic Regression, Decision Trees, and Artificial Neural Networks on a training sample of 130 firms, ultimately revealing that only the Artificial Neural Networks method surpassed Altman’s Z-score model However, the accounting-based approach has faced criticism for not adequately reflecting future trends.

The decline in asset value and liquidity, particularly the reduced ability to raise capital, are primary causes of insolvency Three key components contribute to business defaults: asset value, uncertain risk associated with asset value, and financial leverage When a valuable asset's worth falls below the book value of liabilities, a firm risks entering default Additionally, asset value serves as a strong indicator for bankruptcy predictions, as noted by Black & Cox (1976) and H.E Leland (1994) Consequently, the option-based approach has gained traction in the commercial sector, as highlighted by Black & Scholes (1973) and Merton (1974).

The call option theory is a fundamental concept in market-based approaches, particularly in predicting corporate defaults This theory posits that a firm's equity represents a claim on its assets after settling all financial obligations In this framework, the call option reflects the market value of the assets, while the face value of the debt serves as the strike price At debt maturity, the call option holder will exercise their option if the asset value exceeds the liability's face value Conversely, if the asset value is insufficient to cover the firm's debt, the strike price will not be exercised, potentially leaving the company to its creditors Additionally, the Metron model has become a crucial analytical tool within the Distance-to-Default (DD) model.

In previous studies, Vasicek (1984) analyzed the relationship between asset values and a firm's capital structure liabilities to assess the probability of corporate default Subsequent empirical research by Delianedis & Geske (2003) and H Leland (2002) demonstrated that the theoretical probability derived from structural models is a strong predictor of credit ratings and transitions Various papers highlighted the effectiveness of structural models and the development of option-based models, with Crosbie and Bohn (2003) emphasizing that bankruptcy probability is a critical factor in managing credit portfolios The Distance-to-Default (DD) metric was calculated similarly to option-based models, incorporating asset value estimation and volatility to determine default probability Additionally, the structural model enhanced the reliability of the Merton model by utilizing a global database to measure expected default frequency (EDF) Stein (2005) pointed out the importance of refining the option-based model, suggesting that if failure and non-failure groups were perfectly classified with additional information, the traditional Merton model would not have diminished accuracy.

Insomerecentresearchersemployedthestructuralmodeltomeasurethedefaultriskandt h e n , theexaminationofthecorrelationbetweendefaultriskandothervariables.TheVassalou andX ing(2004)appliedtheMertonmodeltoestimatethedefaultlikelihoodindicator(DLI)fori n d i v i d u a l companiesaswellase x a m t h e relationshipbetweent h e d e f a u l t r i s k andt h e returnequity.Th eFriewald,Wagner,andZechner(2014)illustratedthecloserelationshipbetweenthefirm’sstockreturna ndcreditriskbyusingtheoption- basedmodel.Therefore,thestructuralmodelhadbecometheunderpinningtheoryofagreatnumberbankrupt cypredictionresearchers.Patel&Vlamis(2006)dependedonBSM-

ProbmodelwithcontingentclaimapproachandKVMcorporationframeworkaswellasthedatasetof1 21realestatefirmintheperiod1980-2001tom e a s u r i n g e x a c t l y thedistance-to- defaultandt h e “riskneutral”b a n k r u p t c y p r o b a b i l i t y T h e o u t c o m e dividedthebank ruptcyandnon- bankruptcygroupintotwoerrors.ThetypeIerrorwast h e KVMmodelfailtopredictthebankruptcybutit didoccurwhilethetypeIIerrorwastheKVMm o d e l predictdefaultbutitdidnotoccur.Theresultindicatedt hatthetypeIerrordoesnotappear ino u r estimationwhereastherear e 10o v e r 1 21 firmsfalli n t h e typeI I error.Th e highassetv o l a t i l i t y andthehighleverageweretwodrivingforcesofdefault.

Byström (2006) investigated three assumptions to adapt the Merton Distance-to-Default model for emerging markets and volatile environments, highlighting the significant roles of asset volatility and firm leverage ratios in default risk Bharathan and Shumway (2008) analyzed the accuracy and integration of an option-based model, which was benchmarked against a simpler alternative that did not incorporate default probability Their findings revealed that the alternative model outperformed the hazard model and provided better out-of-sample forecasts compared to the Merton model They concluded that while the structural model lacked sufficient statistical power for predicting default probability, its functional forms were suitable for forecasting defaults.

Filippaki andMamatzakis(2009)calculatedtheMerton- typebankriskandutilizedthepenalVARanalysistoe x a m i n e therelationshipbetweentheefficie ncyandrisk.TheimpactofonestandarddeviationshockstotheDDoninefficiencywasnegativea ndsubstantial.HuangandHe(2010)’researchcalculatedthenewdistance-to- defaultpointofKVMtoestablishanewstructuralmodel.Thenewm o d e l waslikelytoappropriatetoth esevenmajorChinesebanksintheperiodbeforetheglobalfinancialcrisis (2008)2004-2007.

Similarly,D.E.AllenandPowell(2012)incorporatedthemarketvalueofassetstoexaminet h e corp oratedefaultoftheAustraliabanksbytheoption-basedmodelorKVM-

The Merton model indicates that default risk in Australia was likely heightened during the Global Financial Crisis (GFC) due to asset value volatility, with banks exhibiting lower equity levels compared to other sectors Research by Charitou et al (2013) assessed the predictive accuracy of the Black-Scholes-Merton (BSM) bankruptcy model, revealing that incorporating directly market-observable returns on company value enhanced its performance over more complex models In a subsequent study, Agrawal et al (2016) utilized logistic regression and multiple discriminant analysis (MDA) alongside the Merton Distance-to-Default (DD) to effectively distinguish between bankruptcy and non-bankruptcy firms in India, finding the DD variable to be statistically significant in predicting defaults and inversely related to bankruptcy probability, even when the Z-score was included in the model.

Comparisonofaccounting-basedandmarket-basedmodels

Itshould beborne in mindthatthemarket-basedhas been appealingonseveralgrounds: (i)t h e timelinessofthecorporatebankruptcypredictionmayberisenexponentially bytheco mbinationo f t h e m a r k e t - b a s e d v a r i a b l e s ( i i ) T h e v o l a t i l i t y oft h e m a r k e t - b a s e d variablesi s calculateddirectlybythemarketindextoenhancingdramatically thepowerf ulindicatorofthedefaultrisk.Thefluctuationplaysakeyroleinthedefaultprediction.Thehighervolat ilityledtogreatertheprobabilityofthedefaultwasproposedbytheBeaver,McNichols,andRhie(2 005).

Market prices are more effective for predicting corporate defaults because they incorporate forward-looking information about future cash flows, unlike accounting-based metrics, which focus on past performance According to Hillegeist, Keating, Cram, and Lundstedt (2004), the stock market encompasses a wide range of financial information, including data found in accounting statements, suggesting that market information can serve as an alternative source By utilizing forward-looking information, market prices provide a more accurate basis for predicting corporate defaults Relying solely on accounting statements for bankruptcy predictions overlooks critical forward-looking data, limiting the analysis to historical performance and potentially masking relevant financial insights that market-based approaches can reveal.

Anotherstreamofthedefaultpredictionliteraturefocusesonthemarket- basedpredictionm o d e l , asubstantialnumberoftheempiricalstudieshasillustratedtheinferior ityoftheaccounting-basedmodeloverthemarket- basedmodelandviceversa.Unfortunately,theoutcomesacquiredfromthoseresearchershavebeenunob viousbecauseofnumeroussuppositionsaswellaswithoutthesolidbaseofthetheory.Inaprevious paper,theHillegeistetal.(2004)revealedt h a t n o t o n l y theAltman’Z-scoreandO - s c o r e m o d e l m a y notoutperformt h e Black-Scholes-

Mertonmodelabouttheinformationoftheprobabilitybankruptcybutalsothestudyemployingt h e market- basedmodelisthebestrepresentativeoftheprobabilitydefault.Furthermore,Gharghori,Chan,andFaf f(2006)attemptedtocomparethreekindsofthedefaultriskmodelsi n c l u d i n g option- basedmodel,path-dependentbarrieroptionmodel,andZ- scoremodel.Theresultdemonstratedthattheoption- basedmodeloutperformsothermodelsandshouldbeemployedto

19 rankingfirmsbybankruptcyprobability.Likewise,theW.Miller(2009)performedtwocorporatedefaultp redictionmodelsincludingtheAltman’Z-scoreappliesresearcherandtheDistance-to-

Defaultmodelemploysthepractitioners.Themanufacturingfirmsandnon- manufacturingfirmsweretestedintheresearch.TheoutcomeindicatedthattheDistance-to-

Defaultmodeloutperformst h e Z-scoremodelandtheDDis moredurable than theaccounting- basedmodel.

The option-based model relies on several assumptions, such as the normality of stock returns, as highlighted by Allen and Saunders (2002) They noted that this model simplifies reality by assuming a single type of debt and necessitates perfect estimation of unobservable asset values and volatility In contrast, Campbell, Hilscher, and Szilagyi (2006) found that a fully controlled accounting-based model demonstrates strong predictive power Additionally, Reisz and Perlich (2007) suggested that the Z-score model may outperform the Merton model in predictive capability over a year Furthermore, Agarwal and Taffler (2008) argued that the market-based model does not surpass the accounting-based model, which is heavily reliant on financial ratios, indicating a significant disparity in predictive accuracy between the two models.

Eveniftheonemodelmaybesuperiortoothers,itdoesnotmeanthatthesuperiormodelisl i k e l y tob eabandonedaltogether.Therefore,thecombinationoftwokindsofthemodelintothegeneralm o d e l i s p o s s i b l e m o r e powerfulp r e d i c t i o n t h a n o n e approacheithert h e t r a d i t i o n a l acco unting-ratio-basedorthemarket-ratio- basedmodel.Inparticular,Y.Wu,Gaunt,andGray(2010)’s researchcollectedt h e dataseto f t h e c o m p a n y bankruptciesf r o m NewGenerationResearch(www.bankruptcydata.com),Compustata ndCRSPintheperiod1980-

MDAmodelbasedonaccountingvariables,Oh ls on (1980)– logitmodelwith accountingratios,Zmijewski(1984)– p r o b i t modelusingaccountingdata,Shumway(2001)– hazardmodelwithbothaccountingandmarketvariables;andHillegeistetal.(2004)–BSM-

Probmodelbasedonbothaccountingandmarketvariables.According tothetests,thekeyvariableswereclassifiedthoroughlytobuildthecomprehensivemodelthatreflectac countinginformation,marketdata,andfirmcharacteristics.Consequently,thenewcomprehensivemode lcapturingdiversifysidesofthebankruptprobabilityoutperformedothermodelandwasseeminglythemostre liablemodeltopredictthefuturedefault.

Furthermore,LiandM i u (2010)employeda b i n a r y quantileregressionapproacht o b u i l d i n g succ essfulhybridb a n k r u p t c y predictionm o d e l l i n k i n g t h e a c c o u n t i n g - b a s e d and,t h e market-basedmodel.Thedefaultandnon-defaultgroupwere classifiedclearlyfromCompustatdatabasei n theperiod1996-

2006.Theresultdemonstratedthatthez-coredrivenfromaccounting- basedapproachisstatisticallysignificantinexplainingthosecompanieshavingthegoodcreditqualit yw h i l e theDistance-to-Default(DD)variablestakenfromthemarket- basedmodelarestatisticallysignificantininterpretingthosecompanieshavingthepoorcreditquality.The TinocoandWilson(2013)employedthesampleof23218UKfirmsinthe period1980-

2011andNeutralnetwork(MLP)t e c h n i q u e t o e s t a b l i s h t h e defaultpredictionm o d e l Especially,a newm o d e l i s t h e combinationofth eaccounting,market,andmacroeconomicdata.Inmorerecentwork,Trujillo-Ponce,Samaniego- Medina,andC a r d o n e -

R i p o r t e l l a (2014)illustratedt h a t t h e comprehensivem o d e l i n c l u d i n g market- basedanda c c o u n t i n g - b a s e d i s t h e m o s t reliablem o d e l f o r p r e d i c t i n g financialdistresseitherz-coreorKVM-

Studiesonfinancialdistressin thecontextofAsiaandVietnam

Research on bankruptcy in China highlights the significant influence of institutional backgrounds on firms' decisions during financial distress According to Fan, Huang, and Zhu (2008), the behavior of distressed firms is primarily determined by their characteristics rather than external factors such as bankruptcy laws and creditor actions Notably, ownership structure plays a crucial role, with State-Owned Enterprises (SOEs) exhibiting less sensitivity to financial distress compared to non-SOEs This results in SOEs being more likely to maintain higher leverage, a greater proportion of long-term liabilities, and increased levels of external investment during post-distress periods.

21 kgrounddoesnotonlyexplainthecross- sectiondifferencesi n firmb e h a v i o r b u t alsoe x p l a i n h o w firmsa d a p t t h e i r d e c i s i o n s i n distress,t h e researchutilizedthedifferentdefinitionofdistress,differentcriteriafordistress,differe ntspecification,differentregressionmethod.Forthoserobustnesstests,thehypothesisperformancerema insunchanged.

DependingontheoriginalZ- score’Altman,L.Zhang,Altman,andYen(2010)’theresearchdevelopedinto theChinaZ- score.Namely,theform of thenewmodel wassimilarto theZ-

Scoreemergingmarketmodel(EMS).Fromfifteenfinancialratios,theauthorclassifiedintoonlyfourm a j o r variablesbythediscriminantanalysisandtheworkingcapitalovertotalassetvariablewaso n e offourvariablesthatareidenticaltoEMSmodel.ThenewZ- scoremodelwasappliedtoaconsiderablenumberofthechinalistedfirmstakenfromTinysoftFinance AnalysisDatabaseaswellasShenzhenGuoTaiAnInformationtechnologyintheperiod1998–

2008.Surprisingly,whilet h e originalZ- scoremodelonlyforecastsaccuratelyayearinadvance,theChinaZ- scoremodelpredictsthreeyearsinadvancewith80percent accuracy.Moreover,intheW a n g & Campbell(2010)’study,theOhlsonmodelwasre-estimatedtopredicting thefinancialdistresswithnumerousChinapubliclylistedfirmsintheperiod1998-

2008.Theoutcomeindicatedthatifthetotalliabilityishigherthanthetotalassetorthenetincomeisn egativetwoyearscontinuously,t h e firmsmayfallinthebankruptstate.Inmorerecent,PaoloneandRangon e(2015),theemergingmarketmodel(EMS)wasappliedtoforecastingtheeffectoftheglobalcrisisonth ebankruptcyi n China.Inparticular,theaccountingdataofthe3220Chinesepubliclytradedfirmsdur ingtheglobalcrisisperiod(2008-

2014)wascollectedtocalculatingtheEMSscoreforeachstateandint h e wholeperiod.Theresultindicat edthat71.93percentdonothavethebankruptprobabilityando n l y 6.18percenthavebankrupt probabilityin thecrisisperiod.

In a study by Low & Yatim (2001), the efficiency of predicting default probability in Malaysian companies was analyzed, focusing on profitability and liquidity The research involved approximately 176 companies across nine industries, utilizing logistic regression Surprisingly, traditional financial ratios for profitability and liquidity proved ineffective in measuring financial distress for firms with high profitability and liquidity Instead, cash position emerged as the most critical ratio, directly influencing the likelihood of financial distress; a stronger cash position correlated with a lower probability of insolvency, achieving predictive accuracy exceeding 80% In a separate study in Thailand, Meeampol & Noonoi (2014) employed the Z-score and Emerging Market Score (EMS) models to predict financial distress among listed companies on the Thai stock exchange Their findings indicated that both models could accurately forecast default probabilities, although the Z-score model demonstrated a higher accuracy level compared to the EMS model.

There are limited bankruptcy studies in Vietnam, with most researchers utilizing well-known techniques like the Z-score and Merton model to predict financial distress A notable example is Canh's (2013) research, which employed the Z-score Altman model for five state banks and twenty commercial banks from 2002 to 2012 This long dataset enabled an examination of the global financial crisis's impact on the banking system, revealing that post-crisis default risk is significantly higher than pre-crisis levels, and the bankruptcy rate for state banks exceeds that of commercial banks in Vietnam While Canh's study took an accounting-based approach, Chi and Anh (2013) applied the market-based Merton model to predict the default probability of 6,598 customers at Vietcombank during 2008.

2013 Theresultindicatedthatthedefaultprobabilityofthewholeportfolioissomewhereinthevicinit yof2 6 % orVND6319billion.Surprisingly,almostlarge- sizefirmshadhigherthedefaultprobabilityt h a n thesmall- sizecompaniesandtheclassificationamongthee c o n o m i c sectors.Namely,thehighestbankruptpro babilitywastheproduction,seafoodprocessingindustryand,electricitywhilet h e lowestdefaultprobabilitywas the road transportandinlandwaterways.

Thischapterconcentratesinthedatasourceandresearchmethod.Thefinancialdistresslikelih oodiscarefullyestimatedbytheaccounting-basedandmarket- basedapproach.Notonlyalargenumberofvariablesiscollectedtobuildthenewmodelincludingt heaccounting,marketandmacroeconomicvariables,butalsothetheoryoflogistic regressionandaspecialtechniqueofcomparingamongmodels would bepresentedcompletelyin twolastsections.

Data

Thisstudyisconductedonadatasetwhichcoversmorethan800listedfirmsinHanoistockexchange( HNX)andHoChiMinhstockexchange(HOSE)fortheperiodfrom2003to2016.Alldataarecollectedfrom Bloombergforall11differentsectors.Almosttheaccounting datacomesfromt h e financials t a t e m e n t w h i l e t h e s t o c k priceplaysa m a j o r parto f marketd a t a MacroeconomicdataiscollectedfromtheWorldBankwebsite.Especially,themodelwillpredictcorporat ebankruptcyandcompareeachotherthroughtheGFCstageincludingthepre-andpost-t h e GFC.

Comprehensive model (EMS + DD + Macro)

Analyticalframework

Comparisonamongmodels:themodelfitstandardasthePseudoR 2and ReceiverOperatingCharacteristicsar ea(ROC)

Eachmodelwouldestimatedefaultprobabilityinpre-crisis,duringcrisis,post- crisisandthew h o l e period.

Int e r m s o f v a l u a b l e informationfrompreviousstudies,t h e m e t h o d o l o g y i s a d a p t e d t o analyticalframeworkabove.Inparticular,thefinancialriskisestimatedbyaccounting- basedorE M S modelandmarket- basedapproachesorDDmodel.Inordertocoveralmostfactorsaffectingt o businesses,t h e comprehens ivem o d e l i n c l u d i n g t h e a c c o u n t i n g , marketasw e l l asmacroeconomicvariablesisest ablishedcompletelyandlogisticregressionisemployedtom e a s u r i n g thefinancialdistress.No tonlythePseudoR 2and ReceiverOperatingCharacteristics area(ROC)areusedforcomparingamongmodelsbutalsothelikelihoodofinsolvencyisexameds e r i o u s l y indifferentscenariosaspreandpost-crisisstages.

Estimatingfinancialdistress

Emergingmarketscoringmodel(EMS)

Noonecandenythatthemultiplediscriminantanalysis(MDA)isoneofthebestwaystopredict t h e defaultp r o b a b i l i t y whent h e d e p e n d e n t variablei s qualitativeform.T h i s statisticaltechniqu eisacquaintedwithclassifyingobservationsintooneofsome priorgroupshavingtheprivatecha racteristic.Inthisstudy,thebankruptandnon- bankruptgroupisdeterminedefficientlybytheMDAtechnique.Thescorethatiscalculated fromth ediscriminant functionisusedforclassifyingthe default andnon-defaultgroup.

NotonlytheE.I.Altman(1968)’researchexploresthediscriminanttechniquebutalso,he constructsa m o d e l t o p r e d i c t financialdistressw i t h t h e financialand,economicratios.T h e fu ndamentalf u n c t i o n i s b u i l t byt h e essentialtraditionaldiscriminantm o d e l andt h e n , i t i s developedt h e m u l t i p l e discriminantanalysis( M D A ) Aftert e s t i n g a considerablen u m b e r o f financialvariables,thefivefollowing variables:( i ) working capitalsover totalassets(X 1 ), (ii) retainedearningsovertotalassets(X2),

(iv)marketvalueofequityoverbookvalueoftotaldebt(X3)and(v)salesovertotalassets(X4).Eachindicator measureliquidity,profitability,thep r o d u c t i v i t y ofassets,s o l v e n c y prevailedandsalesgen eratingabilityofassetsrespectively.

DependingontheZ- score’Altmanin1968thatonlyestimatemanufacturingpubliclytradedfirms,E.Altman(1983)modifiedth eZ-scoremodeltoZ’- scoremodelthatisusedtotheprivatemanufacturingfirmso r unpublishedcompanies.Inparticular,t h e marketvalueofe q u i t y wasreplacedbythebookvalueofequity.Forhewouldliketoestimatemorety peoffirms,theZ’’- scoremodelisinventedbyhimselfE I.Altman(2000).TheSalesoverthetotalasset(X5)isdroppe dout themodelbecauseof minimizingthepotential industryeffect.

This research utilizes the Emerging Markets Scoring Model (EMS) developed by E.I Altman in 2005 to estimate the default probability of emerging market credits The EMS model is chosen for its reliance on fundamental financial reviews from qualitative risk assessments and serves as a modified rating that reflects specific credit risks It is particularly well-suited for developing nations like Vietnam, capturing the advantages and addressing the disadvantages of previous models, including the Z, Z’, and Z’’-score models The coefficients of the four variables, X1 to X4, align closely with the Z’’-score model, with the EMS score incorporating a constant term of +3.25.

� 3 : Earningsbeforeinterestand taxes(operatingprofit)overtotalassets(EBIT/TA)

- 4.15≤EMS−Score≤5.85:Greyzone,warningzone.Thefinancialexposureina lowlevelor thepotentialbankruptcy

- EMS−Scorez 95%Conf Interval

ItisgenerallyacceptedthattheEmergingMarketScore(EMS)andDistancetodefaultaret w o ke yproxiesforthebankruptcyoffirms.Table7presentsthatEMSisnegativelysignificantat1percentw hiletheDistancetodefault(DD)isat5percentlevelofsignificance.Thisresultalsoindicatesthatthelargerth eEMSandDistancetodefault,the lower theprobabilityofdefault.

2016.81.2percentoffirmbelongstothesafezonewithnoprobabilityofdefaultandgrayzoneorthelowle velofdefaultwhereasonly18.8percentfallsinthehighprobabilityofb a n k r u p t c y zone.Moreover,t heZ- scoremodeldemonstratesthesimilarresulttoEMSmodelatt a b l e 9withisrelatively80percentofthec ompanyliesinthefinancialhealth.Consequently,mostVietnamese firmshaveagoodandstable economicoperation.

48Table10alsorevealsDistancetodefaultdiscriminationtabletodistinguishbetweenhealthyandfina nciallydistressed firmsappliedthemappingofS&PcreditingtoKVMEstimatedDefault

Frequency(EDF)values.Afirmisclassifiedasnon- bankruptcyifitsEDFislowerthan0.52%andbankruptcyifitsEDFishigherthan6.94%.Itcanbeconclu dedthatthismethodindicatestheimpressivediscriminationaccuracywithregardtoDDmodel.Surprisingly ,thesafeandgrayzonei s considerablylowerthantheEMScasewithisonlyapproximately50percentortheB ankruptcyz o n e ofDDmodelistwicethatoftheEMSmodel.ItmeansthatahalfofVietnamesefirmsfallint h e financialdistressandthedifferentresultbetweentheaccounting-based (EMS)andmarket- basedapproach(DD)couldbeinterpretedeasilybytheaccuracyoffinancialdata.Thisresultisp r e t t y suitabilityto thestatisticsfromtheMinistryofPlanningandInvestment.

Table10:Creditrating ofDistanceto default(DD)model

Many Vietnamese companies often manipulate their financial statements to present a favorable image to investors and the State Securities Commission of Vietnam, which can facilitate capital raising through stock market listings Consequently, accounting outcomes frequently reflect an overly optimistic view of Vietnam's financial landscape In contrast, a market-based approach provides a direct observation of stock market performance, helping to eliminate adjustments to financial data and offering a more honest depiction of a firm's financial health This method also reveals the volatility of companies at different market stages As a result, the positive portrayal of Vietnamese firms often exists only on paper Financial investors should focus on market dynamics, while the government needs to implement appropriate policies to support businesses rather than relying solely on financial statements or balance sheets.

Factorsaffectthefinancialdistress

Table11presentstheresultofaxtlogitregressionoffinancialdistress.Variousmodelshavebeenused t o c o n s i d e r t h e separateeffectsfromaccountingfactors;m a r k e t - b a s e d factors;andmacroeconomicfactorsonfinancialdistressandbankruptcyforVietnam’slistedfir ms.Inaddition,becauseofthecorrelationsbetweentheinflationandtheShort- termTreasurybill,variousm o d e l s arerunconcurrentlytoovercomethecorrelationamongvariables. Insummary,the followingmodels areconsideredinthisstudy.

- Model4includesallaccountingvariablesplustheshort-termTreasurybill in oneyear;

- Model6includesthemarketvariablesplustheshort-termtreasurybill in oneyear;

Variable Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

Note: WCTA:T h e worki ng c a p i t al over t ot al as se t ;RETA:Re t a i n ed e a r n i n g s over t ot a l asset ;EBITTA:

E a r n i n g s beforeinterestandtaxes(operatingprofit)toTotalassets;BVETA:Bookvalueofequityto

T o t a l liabilities;MVE:Marketvalueofequity;PRICE:Stockprice;VOL_MVE:Volatilityofmarket v a l u e o f e q u i t y ;L E V E R A G E :L e v e r a g e r a t i o ;I N F L A T I O N :I n f l a t i o n ;S H T

Theresultingmeasurementfromxtlogitregressionofthedefaultindicators ontheindepend entvariablesisalsoillustratedinTable11.Alloftheaccountingvariablesinmodel1ares t a t i s t i c a l l y significantat1-

10%.Itsuggeststhattheaccountingmodelisapowerfulpredictorofdefaultprobability.Moreover,t h e c oefficientmeasurementexpressest h e predictedsign.T h e negativesignoft h e w o r k i n g capitalovert h e t o t a l asset(WCTA)p r e s e n t i n g t h e netfinancial liquidityoftheassets,showsthatthehighertheworkingcapital,thehigherperformanceofthec o m p a n y aswellasthen,itislowerthedefaultprobability.Likewise,thenegativesignoftheretainede arningsoverthetotalasset(RETA)indicatessuggestthehigherprofitability,thelowerp r o b a b i l i t y ofenteringthebankruptcy.Theearningsbeforeinterestandtaxesoverthetotalasset(EBIT/

TA)hasalsothenegativesign,whichrepresentstheproductivityofthecompany’sassetse x c l u d i n g thetaxandleveragecomponentsortheearningpoweroftheasset,illustratesthatthehigherlevel ofEBIT,thehigheritsperformanceandthereforethelowerthedefaultlikelihood.T h e finalaccounti ngvariableisthebookvalueofequityoverthetotalliability(BVE/

TL)orthec a p a b i l i t y ofcovering thefinancialdebt bythefirm’s asset.The negativesignin dicatesthata higherleverageoffirmmayexpressthelowerbankruptprobability.Thesignofallacc ountingvariablesisappropriatetotheexpectationthatisforecastedin theprevioussection.

The analysis reveals four market variables, with only three—price (PRICE), market value of equity (MVE), and leverage ratio (LEVERAGE)—showing significant relationships at the 1% level Notably, market price has a significantly negative correlation with the probability of financial distress, indicating that higher market prices are associated with better future returns for both investors and firms Companies can raise substantial capital when prices are high, while low prices can lead to financial losses MVE also exhibits a negative relationship with default probability, as a lower asset value compared to the liability's face value increases the risk of financial distress; thus, a higher market value of equity correlates with a lower default probability Additionally, the leverage ratio presents a negative relationship, suggesting that higher leverage levels reduce the likelihood of default, as the tax shield from debt can enhance equity returns and lower default risk However, there is no evidence that the remaining volatility of equity variables correlates with financial distress.

Models 3 and 4 incorporate accounting variables along with two key macroeconomic indicators: inflation and the Treasury bill rate, both significant at 1% The negative coefficient for inflation indicates that higher inflation correlates with an increased likelihood of financial distress, holding other factors constant Conversely, a positive relationship exists between the Treasury bill rate and financial distress, suggesting that a higher Treasury bill interest rate leads to increased financial costs for firms, making it challenging for them to mobilize capital for unexpected payments The findings from these macroeconomic indicators align with prior expectations, and models 5 and 6 yield similar results to the earlier models.

Variable Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

The article discusses key financial metrics essential for evaluating a company's performance and stability Working capital over total assets (WCTA) indicates liquidity, while retained earnings to total assets (RETA) reflect a firm's ability to reinvest profits Earnings before interest and taxes to total assets (EBITTA) assess operational efficiency, and the book value of equity to total liabilities (BVETA) measures financial leverage Market value of equity (MVE) and stock price (PRICE) provide insights into market perception, while volatility of market value of equity (VOL_MVE) highlights investment risk Additionally, the leverage ratio (LEVERAGE) and inflation (INFLATION) are critical for understanding financial health and economic impact.

In logistic regression, unlike linear models, the parameters are not easily interpretable, making the marginal effect essential for understanding the influence of regressors on the response variable Table 12 presents data from Vietnam between 2003 and 2016, highlighting that earnings before interest and taxes over total assets (EBITTA) has the most significant impact on financial distress, while market value volatility (Vol_MVE) has the least The model outcomes indicate that four accounting variables and two macroeconomic variables significantly affect the likelihood of financial distress, whereas only two market variables and two macroeconomic variables show substantial influence in model 5 Thus, accounting and macroeconomic variables play a crucial role in predicting default probability.

Measure Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

Table 13 summarizes the performance of eight statistically significant models used to measure financial distress for listed firms in Vietnam The area under the ROC curve (AUC) is utilized to estimate the predictive accuracy of these models A good model exhibits an ROC curve that rises directly from (0,1) to (1,1), while a model with no discriminatory power will have an AUC of 0 or an ROC of 0.5; a perfect model will achieve an ROC of 1 Models 3 and 4, which incorporate accounting and macroeconomic indicators, are identified as the best performers with AUCs of 0.9354 and 0.9352, respectively In contrast, Models 2, 5, and 6, which include market variables, demonstrate lower predictability regarding default probability Both Nagelkerke's R² and Cox and Snell's R² tests yield similar results, indicating that Model 3 is the highest performer with values of 0.395 and 0.596, respectively The presence of accounting and macroeconomic factors in the model contributes significantly to its effectiveness, and the marginal effect analysis reveals that the impact of market variables on default probability diminishes when included in the model.

Financialdistressin variousscenarios

Inordertounderstandthefinancialdistressindifferentcircumstances,theglobalfinancialcrisis( GFC)ischosenasamilestonedividedintodistinctphasesbecauseofthegreateffectonVietnameco nomy.Therefore,thedatasetisdividedintotwosub-periods:pre-crisis(2003-2009)andpost- crisis(2010-2016).

Table14presentstheresultofafinancialdistressinpre- crisisperiods.WhiletheEarningsbeforeinterestandt a x e s t o Totalassets( E B I T T A ) i s n e g a t i v e l y s i g n i f i c a n t at1 percent,t w o macroeconomicvariablesinvolvingtheinflationandTreasurybil larepositivelysignificantat5percentwith thefinancialdistress.Inaddition,therearealsotwomarketvariablesaresignificant.T h e priceandlever agehaveanegativeimpacton thecompanies’probabilityoffinancialdistress.Thisresultrevealsthatthetwomarketfactorsincluding themarketvalueofequity(MVE)andv o l a t i l i t y of MVEdo notaffectthedefaultprobabilityoffirmsinthepre-crisisperiod.

Table14:FinancialdistressofVietnam’slistedfirms:variousmodelsinpre-crisisperiod(2003-

Variable Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

WCTA:T he working capi t al overt otalasset ;RETA:Retained e ar ni ng s over totalasset;EBIT

TA:E a r n i n g s beforeinterestandtaxes(operatingprofit)toTotalassets;BVETA:Bookvalueofequit ytoT o t a l liabilities;MVE:Marketvalueofequity;PRICE:Stockprice;VOL_MVE :Volatilityofmar ketv a l u e o f e q u i t y ;L E V E R A G E :L e v e r a g e r a t i o ;I N F L A T I O N :I n f l a t i o n ;S H

Furthermore,theROCscoreinTable15indicatesalsothebestmodelbelongtothemodel3 and4i ncludingtheaccountingandmarketvariables.ItprovidesevidencetoconfirmthatModels2 , 5,6includingmai nlymarketvariablesappeartobetheworstperformingmodelsincomparisonw i t h othermodelsincludin gaccountingfactorsand/ormacroeconomicfactors.

Measure Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

5percentandthreemarketvariablesincludingtheprice,marketvalueofequityandleverage aresignificant.Inaddition,theROCscoreinTable17alsoillustratesthatModels 3,

4includingaccountingandmacroeconomichaveahighestROCscore(0.9394forboth).Therefore,the yarethebestmodelsinpost- crisiscase.TheworstmodelsstillbelongtoModels2,5 and6includingthemarketaswellasmacroeconomicvar iableswiththelowestROCscore(0.682and0.681relatively).

Table16:FinancialdistressofVietnam’slistedfirms:variousmodelsinpost-crisisperiod(2010-

Variable Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

WCTA:T he working capi t al overt otalasset ;RETA:Ret ai ned ea rn i n gs overtotal asset;EBITT

A:E a r n i n g s beforeinterestandtaxes(operatingprofit)toTotalassets;BVETA:Bookvalueofequity toT o t a l liabilities;MVE:Marketvalueofequity;PRICE:Stockprice;VOL_MVE:Volatilityofmark etv a l u e o f e q u i t y ;L E V E R A G E :L e v e r a g e r a t i o ;I N F L A T I O N :I n f l a t i o n ;S H

Measure Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

Sector No.ofObservations Safe Gray Bankruptcy

Inordertoundertakethefinancialriskforeachfield,allofthefirmsaredividedintotensecto rsbythe GlobalIndustryClassification Standard(GICS)i n t a b l e 1 8 T h e highestr i s k i s Consu merStaplessectorincludingtheFood&StaplesRetailing;Household&PersonalProductsandFoodBever age&Tobaccowith is

The Vietnamese Consumer Staples sector faces significant financial distress, with a 53.2% decline reported Despite typically experiencing double-digit growth, companies like Masan Group (MSN) are encountering high risks, highlighted by a $1.1 billion M&A deal with Thailand's Singha Group and a subsequent over 10% drop in stock value due to a scandal involving fish sauce products Similarly, Kinh Do Group (KDC) is restructuring its investment portfolio, having sold its candy segment to foreign investors to secure capital for new market ventures, while grappling with intense competition The Utilities, Consumer Discretionary, and Energy sectors also present high risks, with exposure rates around 50%, in contrast to the Health and Education sectors, which have the lowest exposure at 33.8% Traphaco Joint Stock Company (TRA) leads the pharmaceutical industry by leveraging advanced technology in business management and production, resulting in significant revenue and earnings growth With widespread distribution channels across Vietnam, Traphaco is poised to maintain its leadership in the pharmaceutical sector.

Thefinalchapterwillsumupallofthemainfindingsderivedfromtheempiricalresults.Intermsofthe seoutcomes,policyimplicationsareprovidedappropriatelytopractitioners,investors,andgovernment.So medirectionsoffurtherstudyandseveralminorlimitationswillbeexploredo b v i o u s l y bytheen dofthischapter.

Conclusions

The economic shock has significantly impacted the global economy, pushing companies into challenging situations and exposing them to potential bankruptcy risks The 2008 global financial crisis serves as a stark example, as many nations faced macroeconomic challenges such as high inflation, rising interest rates, increased unemployment, and declining output growth, leading to a surge in bankruptcies and default risks Despite Vietnam's impressive economic growth rate, it too is grappling with financial risks, with the number of newly established companies nearly matching those that go bankrupt This research aims to identify early warning indicators of corporate financial distress through accounting-based (EMS) and market-based (DD) models Additionally, it seeks to develop a bankruptcy prediction model using market data for listed firms in Vietnam, focusing on relevant indicators for measuring financial distress The study will also consider key macroeconomic factors influencing financial distress and bankruptcy, culminating in a comprehensive bankruptcy prediction model for Vietnam Finally, it will explore differences in financial distress and bankruptcy prediction models for listed firms before and after the Global Financial Crisis, assessing the predictive capacity of accounting and market variables during crisis and non-crisis periods.

Mostofthepreviousdefaultpredictionmodelsfocususuallyononesideoffinancialdistress.T h e ideal modelofaccountingsideistheZ-scoremodelwhereasDistance-to-

Defaultmodelistheo p t i m a l modelofmarketside.Eachmodelhasitsstrengthsandweaknessandth econcentrationo n o n e aspectcouldn o t b e consideredt h e o p t i m a l choice.Moreover,a l m o s t pre viousdefaultm o d e l onlyappropriatedtodevelopedcountrieswhilethesuitablemodelfordevelopingc ountriesi s beingdiscussedonvariousaspectsaswellasthereisalsonostudyoncomparisonamongmodelsf o r emergingmarkets.

In this research, a total of 6,736 observations from 2016 are analyzed using the Emerging Market Score Model (EMS) and the Distance to Default (DD) model to identify early signals of financial distress, specifically within the context of Vietnam The study employs logistic regression on a comprehensive model that incorporates key aspects of financial distress, including accounting factors, market factors, and two macroeconomic indicators To evaluate the effectiveness of different default prediction models, the Area Under the ROC Curve (AUC) is utilized Additionally, the research examines the potential effects of financial distress across two distinct sub-periods: the pre-GFC period (2003-2009) and the post-GFC period (2010-2016).

Empirical findings from this study indicate that accounting, market, and macroeconomic variables significantly influence the financial distress likelihood of Vietnamese firms during the research period Specifically, four accounting proxies from the EMS model show a negative relationship with default probability, suggesting that higher financial liquidity, asset productivity, solvency, and profitability correlate with lower financial distress Additionally, market-based variables reveal a negative relationship between the market value of equity (MVE) and financial distress likelihood, indicating that larger firms tend to have a lower default probability Conversely, a higher leverage ratio is associated with increased financial distress, although substantial leverage can enhance return on equity due to tax shields Furthermore, the market price of a firm's equity negatively impacts the probability of financial distress, as higher prices enable companies to raise significant capital, while lower prices can lead to financial losses Notably, inflation and short-term Treasury bill interest rates exhibit a positive relationship with financial distress.

Theintentionofthisstudyistodevelopacomprehensivemodel,whichisthefirstofitskindi n Vietnam, t o i n c l u d e variousfactorsderivedf r o m t h e s t r o n g groundo f a c c o u n t i n g m o d e l s ; market- basedmodels;andmacroeconomicfactorstoconsidertheeffectofthesefactorsonthefinancialdist resslikelihoodoftheVietnamesefirms.Theentireresearchperiodfrom2003-2016,whichisthensub- dividedintothepre-GFCperiod(2003-2009)andthepost-GFCperiod(2010-

Onbalance,p r o x i e s f r o m accountingm o d e l , t h e market- basedm o d e l , andtypicalmacroeconomicfactorshaveallcontributedeffecttothefinancialdistresso fVietnameselistedfirmsf o r t h e researchperiodwhentheyareconsideredi n isolation.However,whe na comprehensivemodelisdeveloped,theeffectfromaccountingfactorsappeartobestrongerincomp arisonwiththemarketfactors.Findingsfromthisstudyalsoconfirmthatthemodelofdefaultpredictionincludi ngaccountingandmacroeconomicfactorsappeartobebetterthanthemarketfactorswithmacroeco nomicfundamentals.Furthermore,almostsectorsofVietnameconomyhavea highrateoffinancialrisk Thehighestexposure istheConsumerStaplessector with is53.2percentfallinfinancialdistres swhereasthelowestistheHealth&Educationsectorwithonly

Policyimplications

Foracademics

This study offers valuable insights for academics by developing a reliable bankruptcy prediction model tailored to early-stage corporate financial distress in listed firms in Vietnam Unlike previous research that often relied on a singular approach—either accounting-based or market-based—this paper integrates diverse factors from multiple methodologies to enhance predictive accuracy Many earlier studies utilized models designed for developed nations, which may not be suitable for emerging markets like Vietnam, leading to significant limitations in their findings Additionally, this research employs a unique method of model comparison using ROC curves to analyze the predictive power of each model through specific graphs and figures As a result, this study holds substantial significance in the academic landscape.

FortheVietnameseGovernment

The independent variables in this study are derived from two reliable models: the Emerging Score Model (EMS) and the Distance-to-Default Model (DD) These variables have been validated in previous research, significantly enhancing the model's accuracy The empirical results can assist policymakers in identifying signals of financial distress, potentially preventing large-scale bankruptcies across industries Awareness of financial distress is crucial for government intervention, allowing for a reassessment of industry operations before crises occur This study emphasizes the need for practical application rather than theoretical reports to uncover the primary causes of bankruptcy Additionally, it introduces a new default model aimed at mitigating financial risk Given the limited resources of governments, including Vietnam's, the model highlights key financial ratios that influence distress, enabling a targeted approach to reduce default probabilities and optimize resource allocation.

Moreover,thisstudyalsoexampredictionmodelsoffinancialdistressforlistedfirmsindiffe rentperiodsasb e f o r e andaftert h e G l o b a l FinancialC r i s i s Forestimatingt h e d e f a u l t probabilitythroughmultiformscenariosandthecomparisonoftheaccuracyandpowerfulpredictionofd istressmodel,thefinancialexposuremaybepreventedinoneofthefirststagesoft h e distressbytheeffe ctivesolutionsaffectingtheheartofthefinancialdistressoffirmsinthew h o l e economy.Theg overnmentcouldpreparethesolutionsforvariouscircumstanceswhenthefinancialriskcomesto theeconomy.Therefore,thisresultis alsousefulforthepolicymakers,theo u t c o m e o f t h e p r e d i c t i v e m o d e l assistt h e m i n k e e p i n g t h e s t a b l e e c o n o m y i n t h e d i f f i c u l t circumstance.

Forpractitionersandinvestors

Infact,thisresearchidentifiesearlywarningindicatorsofcorporatefinancialdistressissoessent ialtomanypractitioners.Theymayminimizetheirexpenseforavoidingtheprobabilityofdefault.The practitionerscouldalsoanalyzethe variousfactorsoffinancialdistresswith thehighr e l i a b i l i t y insteadofdependingonalargenumberofindicatorsfromthefinancialsta tementandbalancesheet.Thisquantitativeanalysisisusedtomakinginvestmentstrategytothemselvesa ndt h e i r customers.T h e financialemployeesm a k e m a n y s o l u t i o n s t o reducingfinancialr i s k bymodifyingfinancialr a t i o formt h e m o d e l i n c l u d i n g accounting,m a r k e t andmacroe conomicelements.Correctingtimelyfinancialfactorscandecreasetheconsequencesofbankruptcy.

Professional investors prioritize maintaining reserve funds to mitigate financial risks When default risks are accurately predicted, they diversify their portfolios into safer investments, sometimes reallocating these funds into their investment budgets This strategy helps them avoid significant expenses associated with financial distress during business bankruptcies Additionally, recognizing bankruptcy signals can guide investors in making informed investment decisions, as empirical evidence highlights default probabilities across various industries and firms Investors typically prefer sectors with lower bankruptcy risks and may select specific firms based on their financial stability In the banking sector, early identification of default risks is crucial for making sound lending decisions, as lenders must assess the financial health of their borrowers to avoid potential bad debts.

Limitationandfurtherresearch

Despite its significant contributions, this study has some minor limitations Data was collected from the Hanoi Stock Exchange (HNX) and Ho Chi Minh Stock Exchange (HOSE) for the period from 2003 to 2016 to cover the entire economy However, the two exchanges have different trading rules and market scales, and merging them may slightly reduce data accuracy Additionally, the stock market does not fully represent the overall economy, as many state-owned enterprises remain unlisted The Vietnamese stock exchange is still in a developmental stage, characterized by unstable trading rules and market size, which can change based on investor demand Consequently, it is not yet a fully professional or entirely trustworthy market.

Ins o m e furtheri m p r o v e m e n t s , t h i s researchexaminest h e financialdistressi n variou sscenariosaspre-andpost- crisisperiods.Itwouldbedeeperanalysistoconsiderseparatelythed u r i n g thecrisisperiod.The policymakersandpractitionersneedtobeawareofthelargeimpacto f theglobalfinancialcrisisontheec onomy.Moreover,thevariousfactorsaffectingtheb a n k r u p t c y s h o u l d b e a l s o examinedt o m a k e thedevelopments t r a t e g y f o r eachsector.T h e m e t h o d o l o g y ofthisstudymaybee xpandedtoChina,ASEANcommunityandotheremergingmarketsin theworld.

Agarwal,V.,&Taffler,R.(2008).Comparingtheperformanceofmarket-basedandaccounting- basedbankruptcypredictionmodels.Journalof banking&finance,32(8),1541-1551.

( 2 0 1 6 ) P r e d i c t i n g f i n a n c i a l distress:revisitingthe option basedmodel.South

(2009).Transitionalcreditmodelinganditsrelationshiptomarketvalueatrisk:anAustraliansectora lperspective.Accounting&Finance,49(3), 425-444.

(1983).CorporateFinancialDistress:ACompleteGuidetoPredicting,avoidinganddealingwith bankruptcy.NewYork.

(1968).Financialratios,d i s c r i m i n a n t analysisandt h e p r e d i c t i o n o f corporatebankr uptcy.TheJournalofFinance,23(4),589-609.

Altman,E.I.(2000).Predictingfinancialdistressofcompanies:RevisitingtheZ- scoreandZETAm o d e l s SternSchoolof Business,NewYorkUniversity,9-12.

Altman,E.I.,Haldeman,R G., &Narayanan,P.(1977).Zetatoanalysisanewmodelto identifyb a n k r u p t c y risk ofcorporations.Journalof banking&finance,1(1),29-54.

Andrade,G.,&Kaplan,S.N.(1998).Howcostlyisfinancial(noteconomic)distress?

Asquith,P.,Gertner,R., &Scharfstein,D. (1994).Anatomyoffinancialdistress:Anexaminationo f junk- bondissuers.Thequarterlyjournal ofeconomics,109(3),625-658.

(2013).Impactofmacroeconomicforcesonnonperformingloans:Anempiricalstudyo f comme rcialb a n k s i n Pakistan.wseasT r a n s a c t i o n s o n B u s i n e s s a n d Economics,10(1),40-48.

(1996).B a n k r u p t c y classificatione r r o r s i n t h e 1 9 8 0 s : A n empiricalanalysisofAlt man'sandOhlson'smodels.Reviewof AccountingStudies,1(4),267-284.

Campbell,J.Y.,Hilscher,J.,&Szilagyi,J.(2006).Insearchofdistressrisk:NationalBureauofEconomic Research.

Canh,N P (2013)BankRisk PreandPostGlobalFinancialCrisis inVietNam:ASurvey

Castro,Vítor."Macroeconomic determinantsof thecreditrisk in thebankingsystem:Thecaseoft h e GIPSI."EconomicModelling31(2013):672-683.

Cox, D R., &Snell,E.J.(1989).Analysisof binarydata(Vol.32) CRCPress.

Chaibi,Hasna,andZiedFtiti."Creditrisk determinants:E vi de nc e fromacross- country study."Researchin international business andfinance33(2015):1-16.

(2013).Alternativebankruptcy predictionmodelsusingoption- pricingtheory.Journalofbanking&finance,37(7),2329-2 3 4 1

Chí,L.Đ.,&Anh,L.T.(2013).KếthợpphươngphápCVaRvàmôhìnhMerton/KMVđểđolường rủirovỡnợ-BằngchứngthựcnghiệmởViệtNam.TạpchíPhát triểnvàHộinhập(5(15)),10-15. Crosbie,P., &Bohn,J.(2003).Modellingdefaultrisk KMVmodellingmethodology. Ćurak,M.,Pepur,S.,&Poposki,K.(2013).Determinantsofnon-performingloans– evidencefromSoutheasternbankingsystems.Banks&BankSystem,8(1),45-54.

Delianedis,G.,&Geske,R.L.(2003).Creditriskandrisk- neutraldefaultprobabilities:i n f o r m a t i o n aboutratingmigrationsanddefaults.Paper presentedattheEFA2003annualconferencepaper.

Dobbs,R.,Lund,S.,Koller,T.,&Shwayder,A.(2013).QEandultra- lowinterestrates:Distributionaleffectsandrisks.McKinseyGlobalInstitute.Availableat:http ://www McKinsey. com/insights/economic_studies/qe_and_ultra_low_interest_rates_distributional_effects_an d_risks.

Fan,J.P.,Huang,J.,&Zhu,N.(2008).FinancialDistresswithoutBankruptcy:TheCaseofChina.

Fofack,Hippolyte."NonperformingloansinSub-SaharanAfrica:causalanalysisandmacroeconomic implications."(2005).

(2009).Capitalstructured e c i s i o n s : whichfactorsarer e l i a b l y important?.Financialmana gement,38(1),1-37.

Gharghori,P.,Chan,H.,&Faff,R.(2006).Investigatingtheperformanceofalternativedefault- r i s k models:Option-basedversusaccounting- basedapproaches.AustralianJournalofManagement,31(2), 207-234.

(1997).Transactionscostsandcapitalstructurechoice:Evidencefromfinanciallydistressedfir ms.TheJournal ofFinance,52(1),161-196.

(1988).Usingfinancialratiostopredictbankruptcy:anevaluationofclassicmodelsu s i n g recent evidence.AkronBusinessandEconomic Review,19(1), 52-63.

(2010).EnactmentofDefaultPointinKMVModelonCMBC,SPDB,CMB,H u a x i a BankandS DB.International Journal ofFinancialResearch,1(1),30.

Jimenez,Gabriel,VicenteSalas,andJ e s u s Saurina."Determinantso f collateral."Journalo f financi aleconomics81,no 2(2006):255-281

(2010).Studiesofdiscriminantanalysisandlogisticregressionmodelapplicationi n creditr i s k f o r China'sl i s t e d companies.M a n a g e m e n t Sciencea n d Engineering,4(4),24.

Koutsomanoli-Filippaki,A.,&Mamatzakis,E.(2009).Performance andMerton- typedefaultrisko f listedbanksintheEU:apanelVARapproach.Journalofbanking&financ e,33(11),2050-2061.

Leland,H.(2002).Predictions ofexpecteddefaultfrequenciesinstructuralmodels ofdebt.PaperpresentedattheVeniceConferenceonCreditRisk.

(1994).Corporatedebtvalue,bondcovenants,andoptimalcapitalstructure.T he JournalofFi nance,49(4),1213-1252.

(2010).Ahybridbankruptcypredictionmodelwithdynamicloadingsonaccounting-ratio- basedandmarket- basedi n f o r m a t i o n : A b i n a r y q u a n t i l e regressionapproach.Journalof

(2004).T h e empiricalr e l a t i o n s h i p betweenaveragea s s e t correlation,t h e firmp r o b a b i l i t y ofdefault,andassetsize.JournalofFinancialIntermediation,13(2),265-283.

Low,S.W.,Nor,F.M.,&Yatim,P.(2001).Predictingcorporatefinancial distressusingthelogitm o d e l :

(2015).Thecontributionofmacroeconomicfactorstotheprediction ofsmallbankfailures. Journal ofInternationalFinancialMarkets,Institutions,andMoney, 39,25-39.

Meeampol,S.,Lerskullawat,P.,Wongsorntham,A.,Srinammuang,P.,Rodpetch,V.,&Noonoi,

S c o r e ModelT o PredictBankruptcy:A CaseStudyOfListedCompaniesInTheStockExcha ngeOfThailand(Set).Knowledgelearningforqualitylife.InInternationalConference(pp.

(2015).T he Ap pl ic a ti on o f t he EmergingMarketS c o r e Modeli n C h i n a DuringtheGlo balCrisisPeriod:ACountertrend.ChineseBusinessReview,14(10),484-498.

Poudel,RaviPrakashSharma."MacroeconomicDeterminantsofcreditriskinNepalesebankingindustr y."InProceedingsof21stInternationalBusinessResearchConference,pp.10-11.2 0 1 3

Reisz, A S., &Perlich,C.(2007).Amarket- basedframeworkforbankruptcyprediction.Journalo f FinancialStability,3(2),85-131.

(2001).Forecastingb a n k r u p t c y m o r e accurately:A simplehazardm o d e l * T h e Journal ofBusiness,74(1),101-124.

(2005).Therelationshipbetweendefaultpredictionandlendingprofits:IntegratingR O C anal ysisandloanpricing.Journalof banking&finance,29(5), 1213-1236.

(1991).Alternativemodelsforassessingthefinancialcondition ofbusinessinGreece.Journ alof BusinessFinance&Accounting,18(5),697-720.

(1993).Meanandv o l a t i l i t y spilloversa c r o s s m a j o r nationals t o c k markets:Furtherempiric alevidence.Journalof FinancialResearch,16(4),337-350.

(2011).Creditriskdeterminantsofpublicandprivates e c t o r b a n k s i n India.EuropeanJo urnalo f Economics,Financea n d Administrative Sciences,34,147-154

(2015).VietnamBankingIndustryReport.Footnote:BaoTranT r a n , BerniceOng&ScottWeld on,VietnamBanking IndustryReport,7.

Trujillo-Ponce,A.,Samaniego-Medina,R.,&Cardone-Riportella,C.

(2014).Examiningwhatbeste x p l a i n s corporatecreditr i s k : accounting- basedversusmarket- basedm o d e l s Journalo f B u s i n e s s EconomicsandManagement,15(2), 253-276.

(2006).Predictionofcorporatefinancialdistressinanemergingmarket:t h e caseofTurkey.Cro ss-CulturalManagement:AnInternationalJournal,13(4),277-295.

TheEconomicsofMeasuring FiscalDecentralisation:PartIV:FiscalDecentralisationi n V i e t n a m , C h i n a , andSelectedAsianN a t i o n s B u s i n e s s School,Economics,UniversityofWe sternAustralia.

Washington,GitongaKariuki."EffectsofmacroeconomicvariablesoncreditriskintheKenyanbankin gsystem."International Journal ofBusinessandCommerce3, no 9(2014):1.

GDP(currentUS$).Retrievedfromhttp://data.worldbank.o rg/indicator/NY.GDP.MKTP.CD

(1990).Financialdistress,reorganization,andorganizationalefficiency.Journaloffinancialeco nomics,27(2),419-444.

(2008).Analyzingt h e financialdistresso f C h i n e s e p u b l i c companiesusingprobabilisticneura lnetworksandmultivariatediscriminateanalysis.Socio-EconomicPlanningSciences,42(3),206- 220.

Xiaogang,C.,Skully,M.,&Brown,K.(2005).BankingefficiencyinChina:ApplicationofDEAt o pre- andpost-deregulationeras:1993–2000.ChinaEconomic Review,16(3),229-245.

Zhang,B.Y.,Zhou,H., &Zhu,H.(2009).Explainingcreditdefaultswapspreadswith the equityv o l a t i l i t y andj u m p r i s k s o f in di vi du al firms.Reviewo f FinancialStudies,2 2(1 2),5 0 9 9 - 5 1 3 1

(2010).Corporatef i n a n c i a l distressdiagnosism o d e l andapplicationi n creditr a t i n g f o r l i s t i n g firmsi n C h i n a Frontierso f ComputerSciencei n C h i n a , 4(2),220-236.

( 1 9 8 4 ) Methodologicalissuesrelatedt o t h e e s t i m a t i o n o f financialdistresspredictionmo dels.Journal ofaccountingresearch,59-82.

Variable Obs Mean Std.Dev Min Max

Note:T h i s tablereportsthedescriptivestatisticsofindependentvariablesasWCTA:Theworkingcapitalover totalasset;RETA:Retainedearningsovertotalasset;EBITTA:Earningsbeforeinterestandtaxes(operatin gp r o f i t ) toTotalassets;BVETA:BookvalueofequitytoTotalliabilities;MVE:Marketvalueofequity;PRICE:

Variable Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

Note:Thistablepresentsthemarginaleffectforaccounting,marketandmacroeconomicvariables inpre-crisis.Model

1andModel2showonlyaccountingvariablesaswellasmarketvariablesrespectively.Model3andModel4revealtheaccountingvar iablesplustwomacroeconomicvariableswhileModel5andModel6indicatethemarketvariablesplusmacroeconomicv a r i a b l e s M o d e l 7 andM o d e l 8 presentcomprehensivemodelincludinga l l o f thev a r i a b l e s a r e accounting,marketa ndmacroeconomicvariables.Themarginaleffectisemployedtointerpretdirectlytheeffectof theregressorsontheresponsevariable.

Variable Obs Mean Std.Dev Min Max

Note:ThistablereportsthedescriptivestatisticsofindependentvariablesasWCTA:Theworkingcapitalovertotal asset;RETA:Retainedearnings overtotalasset;EBITTA:Earningsbeforeinterestand taxes(operatingprofit)t o Totalassets;BVETA:BookvalueofequitytoTotalliabilities;MVE:Marketvalueofequity;

PRICE:Stockp r i c e ;VOL_MVE:Volatilityofmarketvalueofequity;LEVERAGE:Leverageratio;INFLATION:I nflation;S H T B R D E F :Short-termtreasurybillinoneyear

Variable Model1 Model2 Model3 Model4 Model5 Model6 Model7 Model8

Note:Thistablepresentsthe marginaleffectforaccounting,marketandmacroeconomicvariablesinpost-crisis.Model

1andModel2showonlyaccountingvariablesaswellasmarketvariablesrespectively.Model3andModel4revealtheaccountingvaria blesplustwomacroeconomicvariableswhileModel5andModel6indicatethemarketvariablesplusmacroeconomicv a r i a b l e s M o d e l 7 andM o d e l 8 presentcomprehensivem o d e l i n c l u d i n g a l l ofthev a r i a b l e s a r e accounting,marketand macroeconomicvariables.Themarginaleffectisemployedtointerpretdirectlytheeffectoftheregressorsontheresponsevariable.

Model 3 ROC area: 0.9354 Model 7 ROC area: 0.9341

Note: This figurerevealstheAreasunderthereceiveroperatingcharacteristiccurveofthemodel3,model5andmodel7.T hegoodmodelmayhavethe

ROCcurvestraightupfrom(0,1)to(1,1)afterwardacross(1,1).Moreover,theaccurateratio(AR)ofthemodelhasdefinedthear eaundertheROCcurve(AUC).TheperfectmodelmayhaveA R =1ortheROCscore=1whilethemodelthathasnodiscriminato rypowerhasAR=0orROC=0.TheModel3includingtheaccountingandmacroeconomicindicatorsarethebestmodelpos sessingARis0.9354unitandthismodelalsohastheROCcurvegofurtheronthetopleft.

Model 3 ROC area: 0.9091 Model 7 ROC area: 0.9089

Note: This figurereveals theAreasunderthereceiveroperatingcharacteristic curveofmodel3,model5andmodel7 inthepre- crisisperiod.ThegoodmodelmayhavetheROCcurvestraightupfrom(0,1)to(1,1)afterwardacross( 1 , 1 ) Moreover,the accurateratio(AR)ofthemodelhasdefinedtheareaundertheROCcurve(AUC).Theperfectmodelmay haveAR