Determinants of Capital Structure of Firms in Pre-Post Financial Crisis: Evidence from China

Transcription

1 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) Deerminans of Capial Srucure of Firms in PrePos Financial Crisis: Evidence from China XianZhi Zhang Professor, School of Accouning, Dongbei Universiy of Finance and Economics, Dalian, China Sulan Sikandar Mirza Ph.D suden, School of Accouning, Dongbei Universiy of Finance and Economics, Dalian, China Absrac: This paper examines he impac of he global financial crisis of 278 on 897 Chinese lised nonfinancial firms by examining changes in heir capial srucure from 23 o 212. Panel daa echnique has been used and i is found ha here is noiceable change on boh firm level and macroeconomic level deerminans of he capial srucure afer he financial crisis. Regression analysis has provided very significan resuls. In full period regression, liquidiy has shown no change in boh prepos financial crisis ime periods, while ax, nondeb ax shield, angibiliy, economic developmen and inflaion have shown a very significan and disinc change afer crisis. Bu a he same ime volailiy has shown a very significan change in long erm and in oal leverage while profiabiliy and size has shown significan change jus in shor erm leverage afer crisis. Growh poenial has shown significan change only in oal leverage afer crisis. Analysis shows ha pecking order heory has more explanaions han saic radeoff heory and marke iming heory afer he financial crisis for Chinese lised nonfinancial firms. KEY WORDS: leverage; financial crisis; pecking order heory; capial srucure; nonfinancial Chinese firms 1. Inroducion Every firm's main objecive is firm's value maximizaion. In order o achieve his main objecive, usually every firm has o make vial invesmen decisions and capial srucure choices. Firms have o make very careful decision regarding heir capial financing opions, wheher hey should go public by issuing equiy or seek loans from financial insiuions. The firm's capial srucure in urn deermines he firm s confidence o pursue poenial invesmen opporuniies. For insance, deb is a more convenien and cheaper source of accumulaing funds han equiy because of ease of availabiliy from esablished crediors and ax shields from he governmen. A lo of research work has been done on aaining an opimal level of capial srucure and on he basis of ha various capial srucure heories have been formulaed. There are hree main heories: radeoff heory, pecking order heory, and marke iming heory. Researchers have esed hese heories bu have found ambiguous resuls. For example, in general, he radeoff heory claims ha here is a posiive correlaion beween profiabiliy and leverage raios Harrison & Widjaja (213). However, some researchers, like Rajan and Zingales (1995), have found a negaive correlaion beween profiabiliy and leverage raio. So, hese heories are no reliable for differen ime periods or differen economic condiions o validae i. The global financial crisis of ha became visible in Augus 27 was caused by he subprime morgage crisis in he USA, leading o he bankrupcy of several old and big financial and nonfinancial insiuions. Because of globalizaion, he crisis quickly spread o oher major developed and developing economies, riggering bankrupcy in many forms. Researchers are rying o find ou he impac of he financial crisis on he capial srucure of firms bu hey have mainly focused European and American economies. A sudy by Iqbal and Kume (214) found ha he recen financial crisis had an adverse impac on he capial srucure decisions of UK, French and German firms. The findings show ha boh equiy and deb levels changed significanly in he wake of financial crisis. Harrison & Widjaja (213) conduced a sudy in he USA and found ha he financial crisis of 28 significanly impaced he capial srucure decisions of firms. They found evidence ha in he American economy which is a mosly marke based economy poscrisis capial srucure decisions of firms are more explainable by pecking order heory, which is was no he case prior o he crisis. Researchers found ha conservaive firms were more likely o be ready for rare financial crisis like financial crisis of 278 as hese firms are significanly slow in adjusing he speed of capial srucure in financial disress (Bhamra, Kuehn, and Srebulaev, 211; Ariff, Taufiq, and Shamsher, 28). Bu various sudies have found, as menioned earlier, ha here is no well developed capial srucure model ha could comprehensively define all opimal capial srucure opions. The financial crisis of 278 affeced he whole world in one way or anoher. Is impac on Asian counries 33

2 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) was quie enormous because mos economies here are dependen economies, unlike he selfdriven naure of he big economies of he world. So i was a very urbulen ime for Asia. The inensiy of he crisis was more han expeced so i surprised every naion in his region. According o an ADB repor of 29, he overall GDP of Asia (excluding China and India) in 28 showed a 15% decrease. The Asian counries were in a much worse condiion han he counries from where he crisis originaed. The crisis did no spare he Chinese economy eiher. Following he global decline in economic growh, he Chinese economic growh rae fell from 13 percen o 6.8 percen in he fourh quarer of 28 alone. The main reason was ha he Chinese economy is mainly an expororiened economy. The counry s expor declined sharply because of adverse economic condiions of oher big economies, which ulimaely shaered he Chinese economy. In response o his crisis, Chinese governmen ook very impressive measures like a simulus package of 4 rillion Yuan. The cenral bank of china, he People's Bank of China (PBOC), decreased heir ineres raes sharply which helped boos he growh rae of credi. As early as he firs quarer of 29, he simulus package and he revised moneary policy began o show heir posiive impac on he Chinese economy. Bu i is sill early o alk abou he susainabiliy of he presen growh rae of he Chinese economy as a number of big economies are sill under he influence of he crisis, especially European economies. Till now, no significan work has been done on deermining he impac of financial crisis on capial srucure of Chinese firms. This sudy aims o fill his gap in lieraure wih respec o nonfinancial lised firms of China for he ime period of Lieraure Review Many sudies have been done highlighing differen aspecs of he capial srucure opions for firms. Bu his ques sared wih Modigliani and Miller (1958) when hey gave heir heory of irrelevancy: capial srucure is irrelevan o he firm's value in perfec marke wih symmeric informaion if here are no agency coss, bankrupcy coss and axes. Auhors claim ha price of he equiy would rise by aking deb as a subsiue of equiy in he capial srucure of he firm even by reducing he overall proporion of equiy by keeping he cos of capial consan. Bu his heory sared a new round of couner heories and laer i was found by many researchers ha assumpions made by MM are oo resricive. Till now, cerain heories have been defined o explain various variables o deermine he capial srucure. Pecking order heory, rade off heory and marke iming heory are he main capial srucure heories which have idenified variables like angibiliy, profiabiliy, size, liquidiy and marke o book raio as he deerminans of capial srucure (Tong & Green, 25). Several sudies have shown quie ambiguous resuls afer analyzing hese heories. Pecking order heory claims ha if invesors have asymmeric informaion abou he firm's value hen he equiy can be mispriced by he marke. So, because of his asymmeric informaion firms will prefer reained earnings o deb, shorerm deb over longerm deb and deb over equiy (Chen & Chen, 211). I can be risky for he firm if he marke price of he equiy is undervalued. So, in case of issuance of new sock by he firm o finance new projec, i may creae a siuaion of underinvesmen. So, according o Myers (1984), o avoid his kind of siuaions, firms prefer o adop pecking order, ha is, firms prefer inernal finance (reained earnings) over exernal finance like deb, safe deb over risky deb and converibles and even over equiy like issuance of common sock. Bu according o many researchers, here is no opimal debequiy raio in pracice as every firm is independen o make any decision regarding heir capial srucure. Mosly firms prefer o have pecking order as i makes hem finance heir invesmens hrough inernal sources (reained earnings) raher han looking for exernal financing and i also helps hem o have lower level of leverage (Mayers and Majluf, 1984). Tradeoff heory also alks abou ha he capial srucure opions for a firm bu i mainly focuses on wo opions, deb and equiy. This heory mainly discusses he benefis of deb and also alks abou he agency coss and bankrupcy coss. Bu in all, his heory explains and ries o offse he coss wih immediae access o he money in he form of deb and ax shield benefis provided by deb. There are several facors which can affec he capial srucure decisions of firms. Lieraure abou hose variables can be discussed as follows: 2.1 Size Rajan and Zingales (1995) used four variables of angible asses, marke o book raio, size, and profiabiliy in heir sudy o deermine is relaionship wih capial srucure and hey used economies of G7 counries. They go very ambiguous resuls. hey found ha size and angible asses have posiive relaionship wih he level of deb which is a supporing fac for radeoff heory bu hey also found ha marke o book raio and profiabiliy have negaive relaionship wih level of deb which is a supporing fac for he pecking order heory. Myers (21) explained his phenomenon as he absence of an opimal capial srucure heory which can work beer han is rival opions in differen circumsances. A sudy by Wald (1999) on he firms of USA, UK, Germany, France and Japan o check he correlaion of size and leverage has shown ha here is a posiive relaionship for he firms in USA, UK, France and Japan bu a negaive one for German firms. 34

3 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) Bu sudies conduced on Chinese marke have shown resuls as Srange and Chen (26) has shown ha here is a posiive relaionship beween size and leverage (long erm leverage) for he firms in China bu sudies by Tong and Green, (25), Anwar, S., & Sun, S. (213), Zou and Xiao, (26) and Huang and Song (26) have shown ha leverage and size are negaively correlaed. 2.2 Profiabiliy Researchers have shown conradicory resuls regarding profiabiliy and leverage in heir sudies. Pecking order heory suggess ha here is a negaive relaionship beween profiabiliy and leverage because when firms have large profis hey prefer inernal funds for financing heir invesmens bu radeoff heory claims as he firms are profiable so hey would like o keep heir inernal funds and may access o exernal funds o finance heir invesmens. Several sudies by Chinese researchers on China have shown a negaive relaionship beween profiabiliy and leverage (Chen, 24; Tong and Green, 25; Huang and Song, 26; Anwar, S., & Sun, S. (213); Zou and Xiao, 26; DeJong e al., 28), which suppors he pecking order heory. 2.3 Growh Opporuniies Firms have o expand heir business when here is rapid growh in sales. Myers (1977) suggess ha firms wih high growh may hold more choices for fuure invesmen projecs han low growh firms. Tradeoff heory suggess ha here is negaive relaionship beween growh opporuniies and leverage because firms holding fuure growh opporuniies are like inangible asses and i can' be collaeralized. Agency heory also suppors he radeoff heory. Bu pecking order heory suggess ha when firms expec high fuure growh, hen hey should use equiy financing. Deesomsak e al. (24) have shown in heir sudies ha here's a negaive relaionship beween growh opporuniies and leverage. Zou and Xiao (26), Chen (24) and Anwar, S., & Sun, S. (213) also provide evidence consisen wih he above menioned negaive correlaion of growh opporuniies wih leverage. Wald (1999) has shown ha he USA is he only counry where high growh is associaed wih lower deb/equiy raio. 2.4 Nondeb Tax Shield (NDTS) Nondeb ax shield is applicable for he firms when firm's income is consisenly becoming low or i is negaive. According o DeAngelo and Masulis (198), NDTS is an alernaive o ax shield on deb financing. Sudies have shown quie mixed resuls regarding he relaionship beween NDTS and leverage. Bradley e al. (1984) have shown a posiive relaionship beween he NDTS and leverage bu Wald (1999) has shown a negaive correlaion beween NDTS and leverage. Tradeoff heory argues ha firms would prefer deb for because of nondeb ax shields. Tradeoff heory concludes ha firms will borrow more funds when he ax raes are higher bu usually bigger firms are following his heory han small ones. In China, researchers have shown ha here's a negaive relaionship beween NDTS and leverage. Huang and Song (26) provide evidence for deerminans of capial srucure in Chinese lised firms and find ha leverage and firm size has a posiive correlaion bu i has negaive relaionship wih profiabiliy, nondeb ax shields, growh opporuniies and managerial shareholdings. They provide evidence ha ax has very low significance as Chinese firms consider i only in long erm deb financing. Anwar, S., & Sun, S. (213) also suppor he resuls for NDTS and is relaionship wih leverage. 2.5 Tax Tax is one of he more imporan facors for deermining he capial srucure for firms. Modigliani and Miller (1963) have suggesed ha companies should gain more deb financing for heir ongoing projecs or for new invesmens because of ax deducion associaed wih i due o ineres paymens on deb. Empirical sudies in China have shown very ambiguous resuls. Researchers like Huang and Song (26) have confirmed he relaionship and have showed ha ax is negaively relaed o leverage which suppors pecking order heory. Bu Srange and Chen (26) have shown ha here is no significan relaionship of ax wih leverage in China. 2.6 Tangibiliy Tangibiliy is also one of he imporan deerminans of capial srucure for firms. I can be defined as he collaralizable asse which can be used o ge loan. According o Myers and Majluf (1984) issuing deb by his way, helps a firm o avoid associaed coss. So, his finding suggess ha angibiliy has a posiive correlaion wih leverage which suppors he Tradeoff heory. Many researchers have shown in heir sudies ha here is a posiive relaionship beween angibiliy and leverage (Wald, 1999; Viviani, 28; DeJong e al., 28). Bu 35

4 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) some researchers also have shown a negaive relaionship beween hese variables (Mazur, 27; Karadeniz e al., 29). Booh e al., (21) have given evidence in heir sudy on he firms of Pakisan, India, Brazil and Turkey ha here's a negaive relaionship beween angibiliy and leverage. A few sudies on Chinese marke have also shown he same findings consisen wih heir counerpars ha here's a posiive relaionship beween angibiliy and leverage. Chen (24); Huang and Song (26); Zou and Xiao (26) have concluded hrough heir empirical findings ha angibiliy and leverage have a posiive relaionship in Chinese firms. 2.7 Liquidiy Liquidiy is a very imporan deerminan of capial srucure for firms. Capial srucure heories have differen argumens abou he relaionship beween liquidiy and leverage. The radeoff heory argues ha here is a posiive relaionship beween liquidiy and leverage as firms wih higher liquidiy raios should go for deb or borrowings while, on he oher hand, pecking order heory believes ha here's a negaive relaionship beween liquidiy and leverage because firm wih higher liquidiy raios prefer o use inernal funds (reained earnings) o finance heir new invesmen projecs (Viviani, 28). Some sudies have shown heir findings which are consisen wih he pecking order heory (Mazure, 27). Some sudies have shown ha macroeconomic condiions, legal sysem and poliical environmen migh influence he capial srucure of a firm. Anoniou, Guney, and Paudyal (28) argue ha UK and USA have common law and a marke based governance srucure some oher counries like France and Germany heir law is codified and bank based governance srucures are he norm. Japan has a blended sysem. In heir sudy of five differen counries, hey found ha here are similariies for he deerminans of capial srucure bu heir imporance varies among counries. So hey suggesed ha counry specific facors are also needed while deermining he capial srucure of firms as only firm specific facors canno explain he capial srucure a all. 2.8 Volailiy Volailiy or business risk is he measuremen of probabiliy of financial disress. In sudies, is relaionship wih leverage has been found negaive. Modigliani Miller heorem predics ha i would have a posiive relaionship wih leverage as firm's sysemaic risk of equiy decreases wih he sysemaic increase of variance for he value of is asses. Bu Choi, J., & Richardson, M. (212), DeJong e al. (28) and Booh e al. (21), have found ha volailiy has a negaive relaionship wih leverage. Huang and Song (26) and Anwar, S., & Sun, S. (213) have checked he relaionship of volailiy wih leverage in Chinese firms and have found he same resuls. They provide evidence ha here's a significanly negaive relaionship beween volailiy and leverage. 2.9 Economic Developmen Some researchers have ried o find ou he relaionship beween GDP or GDP growh and corporae capial srucure. They have found ha here is a negaive correlaion beween GDP or GDP growh and capial srucure (Dincergok & Yalciner, 211; Camara e al., 214). Researchers believe ha a posiive increase in he GDP or GDP growh rae will increase he profis of he companies so firms would prefer o use inernal sources o finance heir projecs, which is a supporing assumpion for pecking order heory (Gajurel, 26). Rajan and Zingales (1995) have given evidence in heir sudy ha if GDP growh rae and leverage are negaively correlaed hen i would confirm he pecking order and if here's a posiive relaionship hen i would suppor he radeoff heory. Anoher sudy conduced by Dejong, Kabir and Nguyen (28) in a sample of 42 counries from , find ha firm specific facors like asse angibiliy, firm size, profiabiliy and growh opporuniies are posiively correlaed wih corporae srucure of he firms in various counries. They also find ha he imporance of hese facors varies among counries. They also give evidence ha here is a posiive relaionship of ED wih he corporae srucure and hey furher argue ha in counries wih sronger legal and sound legal sysems, firms prefer deb over equiy. In shor, counry specific facors are imporan while deciding capial financing opions for firms. 2.1 Inflaion Rae Anoher mos imporan macroeconomic facor which can influence he capial srucure is inflaion rae. Inflaion rae measures he uncerainy in economy. Many researchers have ried o find ou he relaionship beween inflaion rae and capial srucure bu heir findings differ grealy. Frank & Goyal (29) argue ha 36

5 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) here is no relaionship beween inflaion and capial srucure of a firm. Gajurel (26) argues ha here is a negaive relaionship beween inflaion and capial srucure. Bu some researchers like Se & Sarkhel (21) and Hanousek & Shamshur (211) have found ha here is a posiive relaionship beween inflaion rae and leverage. Camara e al. (214) and Bokpin (29) show ha inflaion rae and leverage have a significanly posiive correlaion. a. Financial Crisis and Capial Srucure Preferences Afer financial crisis, many economiss provided heir views abou he impac of financial crisis on economies. Harrison & Widjaja (213) checked he impac of financial crisis on American nonfinancial firms and found ha size and angibiliy have posiive relaionship wih leverage bu a he same ime, marke o book raio, profiabiliy and liquidiy have negaive correlaion wih leverage, which resuls are conrary o he siuaion ha exised before he crisis. They also found ha hese resuls are in line wih he assumpions of pecking order heory, which offers more explanaions han radeoff heory and marke iming heory during financial crisis. According o Sinan (21) argues ha profiabiliy, nondeb ax shields, volailiy, and liquidiy are negaively correlaed wih leverage based on UK firms afer he financial crisis. He also claims ha he resuls are favoring pecking order heory. Bu a he same ime he also finds ha angibiliy and size are posiively correlaed wih he firm's leverage which is providing suppor for he radeoff heory. Some researchers like Doukas e al., (211) have shown ha capial marke condiions are more favorable afer he crisis. They also believe ha firms usually for deb financing when equiy marke is no favorable. They also find ha firms usually go for long erm deb financing for a leas 5 years. They give evidence ha radeoff heory has conradicions wih capial srucure of such kind of firms even prior o he financial crisis. Afer he financial crisis many researchers believed ha China would no be affeced because of is closed economy and insulaed banking sysem bu acually i hampered he face of Chinese economic growh o a very grea exen. Chinese expor indusry was paricularly affeced by he crisis which in urn impaced he workforce adversely. Wearden & Sanway (28) have shown in heir sudy ha China's GDP was fallen by 9% in he hird quarer of 28 by following 1.1% in he second quarer of 28. Mos of he work has been done in developed economies regarding heory esing and is validiy bu very fewer sudies have been conduced o deermine he applicabiliy of capial srucure heories in developing or ransiional economies especially on china. China economic feaures are differen from oher counries. Till now, no sudy has been presened which is showing any change in he deerminaion of CS for Chinese firms. bu Sudies prior o financial crisis has shown mix resuls in China as Huang and Song (22) sudied he leverage decisions of 799 Chinese lised companies and concluded ha he radeoff model is more explainable han he pecking order hypohesis o deermine he capial srucure of Chinese lised companies. Bu afer ha various sudies have shown ha radeoff heory has found very limied evidence for Chinese lised companies and according o Huang and Song (26) and Lim (212) Chinese lised companies are unlikely o follow he radiional pecking order as in wesern counries bu hey are seem o follow a differen mehod for financing heir firms. They found ha Chinese firms prefer exernal financing han inernal sources. Bu Chen (24) suggesed a new pecking order and claimed ha reained profi is preferred over equiy and equiy is preferred over deb. So, he order is reained profi hen equiy and lasly, deb. 2. Daa and Mehodology In his research aricle, annual daa has been used from he financial saemens of nonfinancial firms of China for he ime period of This ime period exacly show he impac of financial crisis on he capial srucure of firms in china is he period before he financial crisis and is he period afer financial crisis. Boh ime periods equally divided ino 5 years which is represening a very significan amoun of ime for prepos financial crisis analysis of firms. For his research, daa has been aken from very reliable sources of daabase like RESSET, CSMAR and EIU Counry Daa. Firm level daa (Profiabiliy, Size, Tangibiliy, Liquidiy, NonDeb Tax Shield and Volailiy) has been accessed from RESSET and CSMAR while economic daa (Economic Developmen, Tax Rae and Inflaion) has been aken from EIUCounry Daa. Table 1 shows he lis of all variables and is proxies descripions wih expeced relaionships wih leverage. 37

6 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) Table 1: Independen variables, heir descripion and expeced relaionship wih leverage Variable name Model Proxy Effec on name leverage (+/) Tax rae TAX i Effecive rae % + Nondeb ax shield NDTS i Depreciaion expenses/oal asses Volailiy VOL i Sandard deviaion of EBT/oal equiy +/ Profiabiliy PROF i Profi before ax/oal equiy +/ Liquidiy Growh Poenial LIQ i GP i Curren asses/curren liabiliies Tobin's Q(raio of marke o book value of asses) +/ +/ Tangibiliy TANG i Ne fixed asses/oal asses + Firm size SIZE i ln(oal asses) +/ Economic Growh ED % change of GDP + Inflaion INF average of consumer price index and producer price index Daa has been carefully seleced for he firms and have excluded firms on he basis of following crieria as menioned by Harrison & Widjaja (213): 1. Financial insiuions such as banks, insurance, leasing, privae equiy and invesmen. 2. Newly lised or delised firms during he period of research Nonavailabiliy of cerain accouns o calculae variables (Profiabiliy, Size, Tangibiliy, Liquidiy, Non Deb Tax Shield, Volailiy, Economic Developmen, Tax Rae and Inflaion) 4. The leverage value is larger han he oal asse value. On he basis of ha here are 897 firms in oal following all he above menioned crieria from various secors of boh Shenzhen and Shanghai sock exchanges. 3.1 Model Developmen STD i = α+ α 1 PROF i + α 2 SIZE i + α 3 TANG i + α 4 LIQ i + α 5 VOL i + α 6 GP i + α 7 NDTS i + α 8 TAX i + α 9 ED i + α 1 INF i +λ+... equ.1 LTD i = α+ α 1 PROF i + α 2 SIZE i + α 3 TANG i + α 4 LIQ i + α 5 VOL i + α 6 GP i + α 7 NDTS i + α 8 TAX i + α 9 ED i + α 1 INF i +λ+... equ.2 TD i = α+ α 1 PROF i + α 2 SIZE i + α 3 TANG i + α 4 LIQ i + α 5 VOL i + α 6 GP i + α 7 NDTS i + α 8 TAX i + α 9 ED i + α 1 INF i +λ+... equ.3 Where, STD i : Shorerm leverage, which can be calculaed by shorerm deb raio. LTD i : Longerm leverage, which can be calculaed by long erm deb raio. TD i : Toal leverage, which can be calculaed by oal deb raio. PROF i : Profiabiliy, calculaed by earnings before ineres and ax (EBT) scaled by oal shareholder's equiy as he index. VOL i :Volailiy, which can be calculaed by Sandard deviaion of EBT/oal equiy. SIZE i : Size, calculaed hrough he logarihm of oal asse as he index. TANG i : Tangibiliy, calculaed by fixed asses divided by oal asses. LIQ i : Liquidiy, by using he Liquidiy raio as he index. GP i : Growh poenial, calculaed as marke o book raio of oal asses. TAX i : Tax, calculaed as he corporae ax rae of he economy. NDTS i : non deb ax shield raio of yearly depreciaion amoun o oal asses. ED i : Economic developmen, calculaed as he average of annual real GDP growh rae. INF i : inflaion, calculaed as he average of consumer price index and producer price index. λ = parameer for ime dummy variables = random error erm i= firms in he same crosssecion (e.g., 1, 2, 3...n) = period of ime (years) Following Harrison & Widjaja (213), daa paneling mehods and random effec model have been used in his model. The RE model applies a differen inercep for each daa uni in boh crosssecion and ime series, hus mainaining he level of degrees of freedom. Daa has been analyzed and run hrough Saa o examine he presence of significan correlaion beween he independen variables (Profiabiliy, Size, Tangibiliy, Liquidiy, NonDeb Tax Shield, Volailiy, GP, ED, Tax Rae and Inflaion) and he dependen variable (shor erm leverage, 38

7 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) long erm leverage and oal leverage). The firs es which is been used prior o regression analysis is he Hausman Specificaion es. The purpose of his es is o idenify ha which specific es (fixed effec or random effec) will explain he regression analysis. The Hausman Specificaion es is conduced wih SAS and able 2 shows ha resuls below. I shows ha fixed effec model will be followed o run oher regression models and i rejecs he null hypohesis. Table2: Hausman specificaion es for overall analysis Models Chisquare saisic Degrees of freedom pvalue Shorerm leverage (STD i ) Longerm leverage (LTD i ) Toal leverage (TD i ) Noes: STD i is defined as curren liabiliies over oal asses; LTD i is defined as noncurren liabiliies over oal asses; TD i is defined as oal liabiliies over oal asses. Variance Inflaion Facor (VIF) es is an imporan es o check he mulicollineariy. If he value is higher i means, he presence of mulicollineariy is sronger. The maximum olerable value o deermine he presence of mulicollineariy is 5 (Harrison & Widjaja, 213). The resuls can be seen (see Table 3) ha in all 3 groups as per periods, he values are ranging from 1.4 o 1.63 and hese resuls imply ha mulicollineariy among he variables is relaively weak. Table3: Variance Inflaion Facor Tes Variables VIF 1/VIF VIF 1/VIF VIF 1/VIF TAX i NDTS i VOL i PROF i LIQ i TANG i SIZE i GP i ED INF Descripive Saisics Descripive saisics has been shown in Table 4, for he dependen and independen variables ha provide some insighs abou he variables. 879 observaions have recorded in he sample of Chinese lised nonfinancial firms. The mean STDi of.614 and a sandard deviaion (SD) of.598 suggess ha oal deb is a dominan mode of financing in China. Whereas minimum value zero of LTDi (mean.17 and SD.236) suggess ha some of he firms do no borrow longerm deb. For firm level variables, we observe low NDTS i wih a mean of.16 and SD of.27. We observe highes variaion of SD 2.95 and 1.765in INF and ED respecively suggesing ha Chinese firms are facing quie a volaile business environmen. 39

8 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) Table4: Descripive saisics Variable Obs. Mean Sd. Dev. Min. Max. STD i LTD i TD i TAX i NDTS i VOL i PROF i LIQ i TANG i SIZE i GP i ED INF In order o es he auocorrelaion, pairwise correlaions mehods have been used. The maximum coefficien of he pairwise correlaion which can be oleraed is.8 (Harrison & Widjaja, 213). Resuls are showing (see ables 57) ha here is no evidence of auocorrelaion in our daa as here are no coefficiens of pairwise correlaion which are larger han. Table5: Pairwise Correlaion Tes for all Daa (23212) STD i LTD i TD i TAX i NDTS i VOL i PROF i LIQ i TANG i SIZE i GP i ED INF STD i 1. LTD i TD i TAX i NDTS i VOL i PROF i LIQ i TANG i SIZE i GP i ED INF Noes: STD i is defined as curren liabiliies over oal asses; LTD i is defined as noncurren liabiliies over oal asses; TD i is defined as oal liabiliies over oal asses; TAX i is defined as % effecive ax rae; NDTS i is defined as depreciaion expense over oal asses; PROF i is defined EBT scaled by oal equiy; VOL i is defined as sandard deviaion of EBT/oal equiy;, LIQ i is defined as curren asses over curren liabiliies; TAN i is defined as ne fixed asses over oal asses; SIZE i is defined as naural logarihm of oal asses; GP i is defined as Tobin's Q(raio of marke o book value of asses); ED i is defined as he yearly % change of GDP and INF i is defined as average of consumer price index and producer price index. 4

9 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) Table6: Pairwise Correlaion Tes of all Daa (2327) STD i LTD i TD i TAX i NDTS i VOL i PROF i LIQ i TANG i SIZE i GP i ED INF STD i LTD i TD i TAX i.5 NDTS i VOL i.28 PROF i LIQ i TANG i SIZE i GP i ED INF Noes: STD i is defined as curren liabiliies over oal asses; LTD i is defined as noncurren liabiliies over oal asses; TD i is defined as oal liabiliies over oal asses; TAX i is defined as % effecive ax rae; NDTS i is defined as depreciaion expense over oal asses; PROF i is defined EBT scaled by oal equiy; VOL i is defined as sandard deviaion of EBT/oal equiy;, LIQ i is defined as curren asses over curren liabiliies; TAN i is defined as ne fixed asses over oal asses; SIZE i is defined as naural logarihm of oal asses; GP i is defined as Tobin's Q(raio of marke o book value of asses); ED i is defined as he yearly % change of GDP and INF i is defined as average of consumer price index and producer price index

10 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) Table7: Pairwise Correlaion of all Daa (28212) STD i LTD i TD i TAX i NDTS i VOL i PROF i LIQ i TANG i SIZE i GP i ED INF STD i LTD i TD i TAX i.6 NDTS i VOL i.27 PROF i LIQ i TANG i SIZE i GP i ED INF Noes: STD i is defined as curren liabiliies over oal asses; LTD i is defined as noncurren liabiliies over oal asses; TD i is defined as oal liabiliies over oal asses; TAX i is defined as % effecive ax rae; NDTS i is defined as depreciaion expense over oal asses; PROF i is defined EBT scaled by oal equiy; VOL i is defined as sandard deviaion of EBT/oal equiy;, LIQ i is defined as curren asses over curren liabiliies; TAN i is defined as ne fixed asses over oal asses; SIZE i is defined as naural logarihm of oal asses; GP i is defined as Tobin's Q(raio of marke o book value of asses); ED i is defined as he yearly % change of GDP and INF i is defined as average of consumer price index and producer price index

11 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) 4 Resuls and Discussion In his secion, regression resuls will be discussed. Mainly daa from 23212, has been broken ino 2 caegories, before financial crisis (2327) and afer financial crisis (28212). Regression resuls are very profound. I can be seen (able89111) ha variables are showing heir relaionships differenly in hree se of leverages. All resuls are wihin 95% confidence inerval. Table8: Variables STD i LTD i TD i Coef. P> Coef. P> Coef. P> TAX i.145. *** NDTS i ***.737.*** VOL i.82. ***.2.***.23.13*** PROF i *** LIQ i.112. ***.6.48**.153. *** TANG i.129. ***.72.*** ** SIZE i.12.1 ***.76.***.67. *** GP i.27. ***.9.***.29. *** ED INF Consan.7. *** ***.2.45 **.2.97 * *** *** *** RSq No. of obs No. of firms Noes: STD i is defined as curren liabiliies over oal asses; LTD i is defined as noncurren liabiliies over oal asses; TD i is defined as oal liabiliies over oal asses; TAX i is defined as % effecive ax rae; NDTS i is defined as depreciaion expense over oal asses; PROF i is defined EBT scaled by oal equiy; VOL i is defined as sandard deviaion of EBT/oal equiy;, LIQ i is defined as curren asses over curren liabiliies; TAN i is defined as ne fixed asses over oal asses; SIZE i is defined as naural logarihm of oal asses; GP i is defined as Tobin's Q(raio of marke o book value of asses); ED i is defined as he yearly % change of GDP and INF i is defined as average of consumer price index and producer price index. *** Variable significan a 1%, ** variable significan a 5%, * variable significan a 1% 4.1 Over all period Tax is considered o be an imporan deerminan of capial srucure in Chinese firms. Srange and Chen (26) have provided evidence bu hey have shown ha here is no significan relaionship of ax rae wih leverage in Chinese firms. According o radeoff heory here's a posiive relaionship beween ax savings and leverage bu if ax rae is higher and ax savings are less hen firms will prefer o oher sources of financing like inernal financing or equiy financing. In his sudy he coefficien of ax is.145 which is only significan in shor erm leverage while in long erm and oal deb, i has no significance. I implies ha Chinese firms don' consider ax effec in heir long erm and oal leverage bu in shor erm, i has inverse relaionship wih leverage. So, hese resuls are conrary o he radeoff heory. Nondeb ax shield (NDTS) is significan in only significan in long erm leverage and oal and heir coefficiens are.351and.737, respecively. NDTS is a main deerminan of capial srucure in radeoff heory and i suggess ha when NDTS is higher, firms depends less on debs as heir will be less advanages for firms o use deb and hen ax shields on i. So, his sudy confirms he pecking order heory for Chinese firms. The coefficiens of volailiy (VOL) are.82,.2 and.23 which are very significan resuls. Previous sudies have shown very ambiguous resuls in case of volailiy. Huang and Song (26) have shown ha here is negaive relaionship beween oal leverage and volailiy bu Srange and Chen (26), have given evidence ha here's a significan posiive relaionship beween volailiy and leverage. Bu his sudy also confirms Srange and Chen (26) resuls of and suppor he pecking order heory as Chinese lised firms are suppored by he governmen so i is no easy for banks or oher financial insiuions o close down he operaions of any 43

12 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) lised company in China. So, in case of any financial disress, Chinese lised firms can have access o exernal funds or debs. The coefficien of profiabiliy is.7,.92 and.11 in STD, LTD and TD, respecively. Bu profiabiliy is only significan in long erm leverage. I means profiabiliy has inverse relaionship wih leverage wih Chinese firms bu in only in long erm in he period of So, i implies ha more profiable a firm, he lower is leverage and i suppors o pecking order heory which suggess ha firms prefer heir inernal funds o finance heir projecs or invesmens. So, pecking order heory claims ha firms wih higher profiabiliy will have lower level of leverage in heir capial srucure. Liquidiy (LIQ) is overall significan bu has shown posiive relaionship in LTD and negaive in boh STD and TD. Pecking order heory suggess ha as firms wih more liquidiy raio are more able o finance heir projecs so firms will no go for exernal financing or debs. So, pecking order heory claims ha here is negaive relaionship beween liquidiy and leverage. So, i in his sudy, Chinese firms follows pecking order heory for heir long erm leverage (.112) and oal leverage (.153). Bu conrary o his, radeoff heory shows a direc relaionship beween liquidiy and leverage as firms have more liquidiy raio, hey will hold i and can use i in he ime of disress and hey will prefer o use deb. So, radeoff heory suggess ha a highly liquidae firm will go for more leverage. So, his sudy shows ha Chinese firms follow rade off heory in case of long erm leverage (.6). Tangibiliy (TANG) is a very imporan variable o deermine he capial srucure of Chinese firms as i serve as he collaeral for debs and i helps o moderae he conflics beween shareholder and bond holder. In his sudy, angibiliy has significan coefficiens in all leverages. Bu angibiliy has shown negaive relaionship in shor erm leverage (.129) which suppors he pecking order heory. Bu radeoff heory claims ha angibiliy and leverage have posiive relaionship as firms can uilize heir angible asses o heir advanage and can ge more exernal financing (deb) agains heir angible asses. So, resuls show ha Chinese firms from 23212, follow radeoff heory for long erm leverage (.72) and oal leverage (.84). The coefficien of size (SIZE) is.12,.76 and.67 in shor erm leverage, long erm leverage and oal leverage, respecively. Size is significan in all leverages bu i has shown very conradicory resuls. In his sudy resuls have shown ha Chinese firms have posiive relaionship beween size and leverage in shor erm leverage and long erm leverage. I means bigger firms will prefer o have more deb. I follows he radeoff heory. While conrary o his, pecking order heory claims ha firms wih bigger size will prefer heir inernal resources o finance heir operaions. So, size will have a negaive relaionship wih leverage. So, in his sudy, Chinese firms have shown a negaive relaionship beween oal leverage and size. Growh poenial (GP) has.27,.9 and.29 significan coefficien values for shor erm leverage, long erm leverage and oal leverage, respecively. I shows ha Chinese firms are end o have inverse relaionship wih leverage as hey have abundance of funds o inves heir operaions smoohly bu in shor erm and long erm. I means Chineselised firms wih good growh opporuniies in he fuure prefer o be less dependen on exernal financing in he form of deb. Bu afer analyzing he resuls of SIZE and GP, i can be saed ha smaller firms end o use heir own funds o growh bu bigger firms prefer o use deb for heir growh. I is conrary o pecking order heory in case of shor erm and long erm leverage. Bu for oal leverage, Chinese firms prefer o use exernal sources of funds o run heir operaions. Possible explanaion can be ha as growh opporuniy is an inangible asse for a firm which is likely o be smashed up in case of any crisis (Huang and Song, 26). Researchers have ried o find ou he relaionship beween macroeconomic variables and leverage. Rajan and Zingales (1995), Booh e al. (21), Chen (24) have found ha here's a posiive relaionship beween growh and leverage. Joeveer (213) have showed ha GDP growh is very imporan facor o deermine he capial srucures of firms in hose counries where he economic condiions are very sable and progressive. He suggess ha GDP growh rae may affec he leverage as i is indirecly relaed o growh opporuniies. If GDP growh is good hen firms would be having more growh opporuniies o expand heir business so for ha purpose hey need more finances o suppor heir operaions and companies will prefer o have deb o have exra funds o mee heir requiremens for expansion. This explanaion favors he pecking order heory. POT suggess ha here's a posiive relaionship beween GDP growh and leverage. The pecking order hypohesis predics ha firms wih higher growh opporuniies indicae he greaer demand of capial, hus exernal fund is preferred hrough deb financing (Lim, 212). Bu Jong., Kabir, and Nguyen, (28) in heir sudy of 42 counries have eviden ha here's negaive relaionship beween growh and leverage in hose counries where economic 44

13 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) condiions like capial formaion and GDP growh rae along wih legal enforcemen and credior/shareholder righ proecion are beer. This sudy has found ha coefficiens of economic developmen (ED) are.7 and.12 and significan for shor erm leverage and oal leverage for he period of for Chinese lised nonfinancial firms. These resuls show ha capial srucure of Chinese firms is posiively affeced by GDP growh in shor erm which suppors ha argumen of pecking order heory bu in case of oal leverage, GDP growh has negaive relaionship wih leverage which is conrary o pecking order heory and suppors he radeoff heory (Joeveer, 213). Inflaion is a very imporan facor o deermine counry's sable growh. If here's any increase in inflaion rae, i may brings uncerainy in he economy which ulimaely affec he firm's capial srucure. High inflaion may lessen he imporance of leverage because of higher bankrupcy coss associaed wih deb for firms (Balacı & Ayaydın, 214). Higher inflaion increases he cos of deb so firms reduce heir deb raios. So inflaion has a negaive relaionship wih leverage. Joeveer (213) has provided he evidence ha here is a negaive relaionship beween inflaion and leverage. This sudy shows ha Chinese firms have significan negaive relaionship wih leverage in shor erm (.2) and long erm (.2) which is conrary o oher sudies. The possible explanaion can be as menioned by Srange and Chen (26) Chinese lised firms are suppored by Chinese governmen. So deb financing is easy for hem o finance heir projecs and macroeconomic variables are very much under conrol so i's no creaing any uncerain siuaions for Chinese firms. Table9: 2327 Variables STD i LTD i TD i Coef. P> Coef. P> Coef. P> TAX i NDTS i VOL i.51.*** PROF i.5.***.119.*** LIQ i.94.***.8.77*.121.*** TANG i ***.147.5*** SIZE i ***.251.*** GP i.38.***.12.*** ED INF Consan.14.***.7.1***.42.***.1.11*** *** *** *** RSq No. of obs No. of firms Noes: STD i is defined as curren liabiliies over oal asses; LTD i is defined as noncurren liabiliies over oal asses; TD i is defined as oal liabiliies over oal asses; TAX i is defined as % effecive ax rae; NDTS i is defined as depreciaion expense over oal asses; PROF i is defined EBT scaled by oal equiy; VOL i is defined as sandard deviaion of EBT/oal equiy;, LIQ i is defined as curren asses over curren liabiliies; TAN i is defined as ne fixed asses over oal asses; SIZE i is defined as naural logarihm of oal asses; GP i is defined as Tobin's Q(raio of marke o book value of asses); ED i is defined as he yearly % change of GDP and INF i is defined as average of consumer price index and producer price index. *** variable significan a 1% ** variable significan a 5% * variable significan a 1% 4.2 PreFinancial Crisis (2327) Afer analyzing he daa (see able9) for 2327(prefinancial crisis), i is showing Tax, NDTS, Tangibiliy and size are insignifican in shor erm while volailiy, profiabiliy, liquidiy growh poenial, economic 45

14 Research Journal of Finance and Accouning ISSN (Paper) ISSN (Online) developmen and inflaion significan before financial crisis. In case of long erm many variables like ax, NDTS, volailiy and inflaion are insignifican bu profiabiliy, liquidiy, angibiliy, size, growh poenial and economic developmen are significan. Tax, NDTS, volailiy, profiabiliy, growh poenial and inflaion are insignifican in oal leverage bu liquidiy, angibiliy size and economic developmen are significan. Tax and NDTS has shown no significance for all kinds of leverages. This may explain he fac ha during he ime of economic expansion Chinese firms didn' consider much abou ax and ax shields for heir capial srucure. As shown in able. 9, from 23212, ax is only significan in shor erm leverage bu in he period of 2327, i's no significan a all. This proves he evidence provided by Srange and Chen (26). Volailiy is only significan posiively in shor erm. I means Chinese firms are dependen on exernal source of financing o mee heir shor erm needs in case of any disress bu for long erm and oal leverage here's no significan relaionship has been found beween leverage and volailiy. Profiabiliy has given ambiguous resuls in his sudy prior o financial crisis. Profiabiliy is significan in boh STD and LTD bu is relaionship wih leverage is posiive and negaive, respecively. I shows ha Chinese firms prefer for debs or loans in heir shor erm o finance heir operaions bu for shor erm hey depend on hemselves. So, in case of profiabiliy, Chinese firms follow radeoff heory in heir shor erm bu conrary o his, hey follow pecking order heory o mee heir long erm financial obligaions. Liquidiy is overall significan bu negaively correlaed in STD and TD. These resuls show ha Chinese lised firms have no or less endency o go for exernal sources of finances if he liquidiy is high. Even liquidiy is posiively correlaed o long erm deb raio bu i is very less significan. These resuls are inline wih he oal sample (23212) analysis. So, hese resuls sugges ha Chinese firms follow he pecking order heory for heir capial srucure. Tangibiliy is significan in LTD and TD bu negaively correlaed only o STD. unlike o oal sample (23 212) where, i is significan in all leverages, bu prior o financial crisis i is no significan in shor erm leverage. These resuls confirm he previous sudies especially by Huang and Song (26) in which hey have shown ha angibiliy has significan posiive relaionship wih longerm deb raio and oal deb raio. Size is also significan in LTD and TD bu only has negaive relaionship wih STD. hese resuls show ha in he period of economic growh, Chineselised firms prefer o rely on heir own resources o cope up wih heir financial needs. As inerpreed by Huang and Song (26), size can be considered as he proxy of bankrupcy cos as for Chinese lised firms i has less or no affec on heir leverage because sae is mosly, he main shareholder so here are very rare chances o go bankrup. Prior o financial crisis period (23212), growh poenial (GP) is significan negaively correlaed o leverage in shor erm and long erm leverages. I shows ha firms wih more growh opporuniies are less likely o depend on exernal financing. These resuls are conrary o pecking order heory. Economic developmen (ED) is significan in all kinds of leverages bu i is negaively correlaed only in LTD which explains ha GDP growh developmen is very imporan for Chinese firms in economic developmen period (2327). Resuls have shown ha Chinese firms have less dependence on long erm deb raio which also confirms he previous sudies (Jong e al., 27). Bu i is posiively correlaed wih shor erm leverage and oal leverage. These resuls confirm he pecking order heory. Inflaion is only significan in STD bu posiively correlaed o all kinds of leverages which show ha any increase in inflaion is posiively correlaed wih shor erm leverage. Inflaion is a very imporan measure of economic sabiliy in any counry. For Chinese lised firms, if here is any change in inflaion, i can affec he financial decisions of capial srucure. Bu for long erm and oal leverage, inflaion has no affec on Chinese lised firm's capaciy o mee heir financial obligaion as mosly firms are suppored by he sae (Huang and Song, 26) Pos Financial Crisis Analysis and Is Comparison wih PreFinancial Crisis Period From able 111, i can be seen ha here is a significan amoun of change in daa afer financial crisis. Tax is significan and negaively correlaed o all kinds of leverages. Table 11, is showing a very clear picure of he impac on he deerminans of capial srucure for Chinese lised firms before and afer crisis. Before 46