THE RELATIONSHIP BETWEEN FIRM SIZE AND GROWTH OPPORTUNITIES, AND COMPARISON OF CHANGES IN DEBT AND RETAINED EARNINGS

Transcription

1 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle THE RELATIONSHIP BETWEEN FIRM SIZE AND GROWTH OPPORTUNITIES, AND COMPARISON OF CHANGES IN DEBT AND RETAINED EARNINGS * Roya Darabi, Reza Aaeizadeh, Behnam Amin Pour Mamaghan, Seyyed Mohammad Ali Mirmasoumi and Maghsoud Esmail zadeh 3 Deparmen of Accouning, Souh Tehran Branch, Islamic Azad Universiy, Tehran, Iran Deparmen of Researchers and Elie club, Ardabil Branch, Islamic Azad Universiy, Ardabil, Iran 5 Deparmen of Managemen, Germi Branch, Islamic Azad Universiy, Ardabil, Iran *Auhor for Correspondence ABSTRACT The purpose of his research sudy he relaionship beween firm size and growh opporuniies, comparison of changes in deb and reained earnings in he company of he sock exchange has been acceped in Tehran sock during he 006 o 0 is o collec heory mehod of library and o collec financial informaion he financial values of he company and for analysis of mehod panel has been used. The resuls show ha no meaningful relaionship beween he agen and company as much as growh opporuniy Companies wih deb changes and here is profi accumulaed. Keywords: Growh Opporuniies, Firm Size, he Raio of Deb Changes, Raio of Changes in Reained Earnings INTRODUCTION The modern heory of capial srucure began wih he famous proposiion of Modigliani and Miller (958) ha described he condiions of capial srucure irrelevance. Since hen, many economiss have been changing hese condiions o explain facors driving capial srucure decisions. Harris and Raviv (99) synhesized major heoreical lieraure in he field, relaed hese o he known empirical evidence, and suggesed promising avenues for fuure research. They argued ha asymmeric informaion heories of capial srucure are less promising han conrol-based or produc-based heories. The financial crisis of forces o look criically a he modern level of capial srucure heory. The problems of many companies were relaed o heir financing policies. The role of asymmeric informaion and agency problems has o be undersood beer. The marke for morgage-backed securiies which many believe was in he core of financial crisis was full of asymmeric informaion beween invesors and issuers. Numerous scandals including he Bernie Mad off s one illusrae he deph of agency problems in finance. A he same ime, old ideas abou he link beween axes, bankrupcy coss and capial srucure were no receiving enough aenion in heoreical lieraure unil recenly while managers surveys sysemaically reveal heir pracical imporance. (Miglo, 00) Therefore, he deb is one of he main componens of he capial srucure of mos firms, which plays an imporan role in managing financing of companies. Crieria of company growh are considered a crucial ool for managemen decisions. Therefore, idenifying facors affecing growh of he company is one of he mos sensiive and imporan issues. Mos previous researches invesigaed company's growh facor wih a facor, bu in his paper, we focus on menioned issues, we are looking for finding a convenien and effecive relaionship beween several facors influencing he growh of he firm, and is relaion o reained earnings and changes raios in deb. Finally, he main research quesion of research is ha wheher is growh parameers have impac on (growh opporuniies, firm size) on he changes raio in deb and reained earnings of companies or no. The main purpose of his sudy is o find he relaionship beween various facors of he company's growh and susainabiliy of firms' capial srucure. Review of Lieraure Byoun (009) has done a sudy in America eniled "How and when companies change capial srucure o he deermined arges?" The populaion consiss of all companies in America during he period 97 o Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 55

2 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle 003, which following previous sudies, finance companies and uiliy companies were removed from he sample. The resuls showed ha he majoriy of changes in capial srucure, occurs when companies have excess cash and heir deb level is higher han deb capaciy specified in he Company, When companies have a cash defici and heir deb level is lower han he deb capaciy specified in he Company, or when hey have excess cash and heir deb is lower han heir deb capaciy deermined by he company, he speed of hese changes is reduced. Also, he rae of change in he capial srucure of companies ha have surplus cash and heir deb is higher deb capaciy deermined by he company is higher han he rae of changes in companies ha have a cash defici and heir deb is lower han deb capaciy deermined of he company and his may be because changes coss resuling from reduced deb is lower han coss which will only use for increase he deb. Byoun concluded ha companies ha have he deb lower han capaciy in comparison wih companies which have higher deb han capaciy, use deb for financing. Makouyra and colleagues (0) in a sudy eniled "Capial budgeing, cos of capial and capial srucure", invesigaed he relaionship beween hree variables on he sock exchange in Lain America. They sudied Budgeing 90 acive firms based on he LATAM mehod in connecion wih cos and capial srucure. They compared he resuls of heir analysis wih previous sudies. They have been sudied and compared paricular characerisics and budgeing differences in he wo markes, emerging and SME markes. They observed ha companies wih budgeing srucure use LATAM mehod and sandard capial, bu here is no specific weigh which analyzes liquidiy and capial raioning. Bagheri (009) has invesigaed he impac of financing mehods on he yield and price of shares of lised companies in Tehran Sock Exchange. In his sudy, o find how effec of financing mehods on prices and Sock reurns in lised companies in Tehran Sock Exchange, hree hypoheses have been designed. The firs hypohesis, financing hrough reained earnings more han borrowing on he sock price of lised companies (cemen indusry) is effecive. The second hypohesis, financing hrough reained earnings (reserves) more han borrowing on sock reurns is effecive. The resuls of he firs and second hypohesis sugges ha he impac of financing mehod hrough reained earnings more han borrowing on sock reurns and price is effecive. In he second hypohesis, he average annual oal reurns of companies which have financed hrough reained earnings (reserves) are more han he average annual oal reurns of firms ha have financed hrough borrowing. Ahangar (009) sudied he various mehods of financing of lised companies in Tehran Sock Exchange on he sock reurn. In his sudy, he effec of exernal sources of financing (issuing ordinary shares and receiving loan) on companies lised on sock reurns beween (00-007) was examined. In his sudy, he effec of increasing invesmen and geing a Loan on he annual reurns of 56 firms ha use boh he financial resources (issuing ordinary shares and receiving loan), were compared wih each oher. Saisical analysis showed here is no significan relaionship beween sock issuing and sock reurns, bu here is relaionship beween geing a loan and sock reurn. Meanwhile, here is no significan relaionship beween wo mehods of geing a loan and sock issuing. Theoreical Lieraure of Research According o he heory of sable equilibrium, growing companies in bankrupcy more han oher lose heir value. Several heories which have been proposed in he heory of agency coss, have confirmed his negaive relaionship. In conras, he hierarchy heory of financing opions suggess ha we expec firms wih high invesmen projecs, over ime, have accumulaed a lo of deb (Kimiagari e al., 008). Theoreically, he relaionship beween firm size and financial leverage is no clear. According o ISNA balance model, larger firms have greaer deb capaciy. Also, larger companies usually have more diversiy, and herefore, have more sable cash flows. Sabiliy of cash reduces he risk and heir bankrupcy. They also during he use of deb, have more bargaining power, and can reduce ransacion coss associaed wih he release of long-erm deb. Anoher possibiliy is ha larger companies have more diverse shareholders causing less conrol over he managemen of he company. Thus i is likely ha Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 56

3 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle managers for reducing he risk of personal loss arising from bankrupcies use more deb. However, when he size of he company as a risk defaul variable, is considered, where he coss of financial disress is rivial, should no here is a significan posiive relaionship beween financial leverage and firm size (Kordesani and Najaf 008). The Hypohesis of he Research. There is a significan relaionship beween company's growh opporuniies and deb changes raio of companies.. There is a significan relaionship beween Changes in firm size and he raio of deb changes of companies. 3. There is a significan relaionship beween growh opporuniies of he company and changes raio in reained earnings of companies. 4. There is a significan relaionship firm size and he raio of he changes in reained earnings of he companies. MATERIALS AND METHODS Mehodology Presen sudy in erms of naure and conen is Correlaion sudy, and in erms of ype of work is a research sudy and in erms of purpose in a applicable sudy, and as well as in erms of he mehod for doing research is in he framework of deducive-inducive reasoning. To sudy heoreical foundaions and lieraure review, he library mehod wih using books and papers and heses have been used, and informaion needed is exraced from financial saemens of lised companies on he Sock Exchange. The populaion of he sudy is all lised companies in Tehran Sock Exchange during he period of 006-0, of he 50 companies lised in Tehran Sock Exchange, which mee all of he following crieria: - To March 006 are lised, and heir names unil he end of March 0 from he lis of lised companies are no removed. - During he desired period, heir shares are raded acively on an exchange. 3 - Their financial period mus be ended 9 March, and in he course of he sudy, he financial erms have no changed. 4 They are no among financial inermediaion companies (invesmen, holding, leasing, and banking and insurance) because of heir differen performances. 5 - The informaion you need is available. In his sudy, 0 companies as sample are seleced. In his sudy, o evaluae and deermine he appropriae model, panel daa mehod of Chow es using, Hausman es is used and o es he significance of he model, he F saisic and for significance of he coefficiens, T-saisics and for normalizing variables, Kolmogorov-Smirnov es is used and sofware Spss 0, Eviews7 and Miniab6 are used for analysis. Variables of Research Independen Variable Opporuniies for growh: he raio of marke value o book value (Cheng e al., 009) M/B = The value of sock marke Book value of oal asses Firm Size (Size ): is equal o he naural logarihm of he book value of oal asses (Wu and Xu, 005). Dependen Variables The raio of changes in reained earnings (GRE ): This formula is derived from he research of Graham and Harvey (00): GRE = reained earnings a end of year - reained earnings a end of year (-) Book value of oal asses The raio of changes in reained earnings (GRE ): This formula is derived from he research of Graham and Harvey (00): Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 57

4 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle GRE = reained earnings a end of year - reained earnings a end of year (-) Book value of oal asses Conrol Variables Tangible Asses Raio: The ne raio of propery and machines and equipmen over book value of asses (Pasor and verones 003) Ne propery of machines and equipmen Tang = Book value of oal asses Invesmen Policy: According o he following formula is obained: (Wu and e al., 005). INV = Capial expendiure Book value of oal asses Dividend: I is dummy variable ha if he company has paid a cash dividend, i is equal o, and oherwise equal o zero. (Baker and Wurgler, 00) The raio of cash asses: Cash asses divided by book value of oal asses is calculaed by he following formula (Sein, 996). Cash asses CashHolding = Book value of oal asses Sales growh rae of company (SG ): This formula is derived from he research of Cooney e al., (993). S SG = - S I,- S - SG : Sales growh of firm i in year S : Ne sales of firm i in year S - : Ne sales of firm i in year - Descripive Saisics for Research Variables In his secion, using Pearson's correlaion coefficien, he relaionship beween variables of research and he correlaion beween hem is examined. Marix of correlaion coefficiens beween variables is presened in able 4. Based on he resuls of he Pearson saisic, he raio of changes in deb of companies has significan and posiive correlaion wih firm size, and he raio of cash asses has significan and negaive correlaion wih invesmen policy. The raio of changes in reained earnings also has significan and posiive correlaion wih dividend and has significan and negaive correlaion wih he invesmen policy. Growh opporuniies have significan and posiive correlaion wih and he raio of angible asses. Tangible asses raio also has significan and negaive correlaion wih invesmen policy, and he raio of cash asses and he invesmen policy also has significan and posiive correlaion wih he raio of cash asses. Table : Descripive saisics of variables of Research Elongai on Skewne ss The maximu m amoun The Minimu m amoun Sandar d deviaio n Averag e Number of observa ions Variable The raio of changes in he deb The raio of changes in reained earnings Growh opporuniies Company size Tangible asses raio Invesmen policy Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 58

5 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle Dividend According o able, he average raio of changes in deb and he raio of changes in reained earnings of sample companies has been and 0.005, and minimum and maximum values, respecively, has been and Evaluaion of sleekness and elongaion of his variable, which should be 0 and 3, o variable has normal disribuion indicaes ha his variable does no has normal disribuion. Based on he descripive saisics presened in able, he average variables of growh opporuniies and firm size, sample companies during he posiive period, respecively are.0068, and The posiive mean he raio of Tangible asses, invesmen policy, dividend, raio of cash asses and he growh rae of sales respecively are 0.889, , 0.459, 0.79 and Tes for Normal Disribuion of The Dependen Variable Of Research In his sudy, his issue hrough Kolmogorov-Smirnov (KS) is invesigaed. H and null hypohesis in his es is as follows: H0 : Normal Disrbuon H : No Normal Disribuion Of he level of significance of he es saisic is more han 0.05 (Prob> 0.05), H0 hypohesis based on normal variable disribuion will be acceped. In able, he K-S Saisics resuls for facors of he raio of changes in deb and changes in reained earnings raio of companies in he sample are provided. Table : The resuls of normaliy es of dependen variable of research Significance level ( Sig ) Saisics ( KS ) Number (N) Variable The raio of Changes in he deb The raio of changes in reained earnings Given ha, for a variables of raio of changes in deb and he raio of changes in reained earnings, significance level of K-S Saisics is less han 0.05, herefore, H0 hypohesis based on normal variable disribuion is rejeced a he 95% confidence level, indicaing ha he variables of he raio of changes in deb and he raio of changes in reained earnings don' normal disribuion. Being normal of dependen variable is necessary condiion for regression models, so i is necessary ha before hypohesis es, his variable is normalized. In his sudy, for normalizing daa, Johnson's Transformaion funcion is used, and i is analyzed by sofware Miniab6. The resuls of he K-S es, afer normalizing he daa, are provided in able 3 Table 3: The resuls of normaliy es of Dependen variables afer normalizaion process Significance level ( Sig ) Saisics ( KS ) Number (N) Variable The raio of Changes in he deb The raio of changes in reained earnings According o able 3, since afer normalizing daa, significance level (Sig) of saisic Kolmogorov - Smirnov for he dependen variable is greaer han 0.05 (0.967, 0.8), hus he hypohesis a 95% confidence level is confirmed and indicaes ha he variables of raio of changes in deb and raio of changes in reained earnings, afer normalizaion process, have normal disribuion. The Invesigaion of Correlaion Variables In his secion, using Pearson's correlaion coefficien, he relaionship beween variables of research and he correlaion beween hem is examined. Marix of correlaion coefficiens beween variables is presened in able 4. Based on he resuls of he Pearson saisic, he raio of changes in deb of companies has significan and posiive correlaion wih firm size, and he raio of cash asses has Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 59

6 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle significan and negaive correlaion wih invesmen policy. The raio of changes in reained earnings also has significan and posiive correlaion wih dividend and has significan and negaive correlaion wih he invesmen policy. Growh opporuniies have significan and posiive correlaion wih and he raio of angible asses. Tangible asses raio also has significan and negaive correlaion wih invesmen policy, and he raio of cash asses and he invesmen policy also has significan and posiive correlaion wih he raio of cash asses. Table 4: The marix of Pearson correlaion coefficiens beween variables The raio of Rae The The changes The of raio Growh Dividen Invesme raio of Firm in raio of sales of opporunii d n policy Tangibl size reaine changes grow cash es e asses d in deb h asses earning s The raio of changes in deb (P-Value) The raio of changes in reained earnings (0.9) (P-Value) (0.063) (0.55)(0.90) (0.463) 0.00 (0.000) (0.90) (0.899) (0.68) (0.07) (0.00) (0.97) (0.) (0.94) (0.86) (0.988) ((0.000) (0.000) (0.090) 0.00 (0.95) (0.03) (0.64) 0.05 (0.54) (0.85) (0.94) 0.03 (0.753) (0.07) (0.398) (0.000)) (0.000)) (0.000) (0.74) (0.6) (0.03) (0.4) 0.04 (0.7) Growh opporuniies (P-Value) Firm size (P-Value) The raio of Tangible asses (P- Value) Invesmen policy (P-Value) Dividend (P-Value) The raio of cash asses (P-Value) The raio of sales growh P-Value) Resuls of Hypohesis Tesing Tes Resuls of he Firs Research Hypohesis The aim of he es of he firs hypohesis is o invesigae relaionship beween firm growh opporuniies and he raio of changes in deb of companies, and is saisical hypohesis is defined as follows: H : There is no significan relaionship beween growh opporuniies and he raio of changes in deb of 0 companies. H : There is a significan relaionship beween growh opporuniies and he raio of changes in deb of companies. Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 530

7 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle This hypohesis using he model () as he panel daa is esimaed, and if he coefficien is significan a a confidence level of 95%, i will be confirmed. GDi, 0 M / B 4Tang 5Inv 6DivPayer 7CashHolding 8SG () To deermine ha wheher he use of panel daa in esimaion of he model would be effecive or no he Chow es or F-bound and in order o deermine which mehod (fixed effecs or random effecs) o esimae is more appropriae (for deecion Fixed or random differences of secional unis) Housman es is used. The resuls of hese ess are presened in able 5. Table 5: The resuls of Chow and Housman es for model () P-Value Degrees of freedom Saisics value Saisics Number Tes (00.499).8799 F 606 Chow Hausman According o he resuls of he Chow es and is P-Value (0.0000), he hypohesis H 0 is rejeced a he 95% significance level, indicaing ha panel daa mehod can be used. Also according o he resuls of he Hausman es and is P-Value (0.0000) which is less han 0.05, is he hypohesis H 0 is rejeced a he 95% significance level, and he hypohesis H is acceped, so, i is necessary ha model using fixed effecs be esimaed. Jarque-Bera es resuls indicae ha he remaining obained from evaluaion of research model a 95% confidence level, have normal disribuion, so ha he probabiliy of his es (0.854) is greaer han Anoher classical regression assumpion is consisency of variance of he remaining. If, variances are dissimilar, hen esimaor doesn have he leas variance. In his sudy, o invesigae he homogeneiy of variances, es Breusch-Pagan was used. Given he significance level of he es which is less han 0.05 (0.04), H 0 hypohesis based on homogeneiy of variance is rejeced, and can be said ha his model has variance anisoropy problem. In his sudy, o solve his problem, generalized leas squares esimaion (GLS) is used. In his sudy, o es no correlaed residuals which is one of he assumpions of regression analysis, and is called Auocorrelaion, he Durbin-Wason es (D-W) is used. According o preliminary resuls from he model, whereas he Durbin-Wason saisic is.38, and since i is beween.5 and.5, i can be concluded ha he residuals are independen. In addiion, o es wheher he model has linear relaionship and wheher his model in erms of being linear or no, is explained properly, Ramsey es is used. Given ha confidence level of he Ramsey es (0.574) is greaer han 0.05, so he hypohesis H 0 of his es abou model's lineariy is confirmed and he model does no have clear error. Summary resuls of hese ess are presened in able 6. Table 6: The resuls of es relaed o he saisical assumpions of he model () Jarque-Bera saisic Breusch-Pagan es Durbin-Wason saisic Ramsey es P-Value F P-Value D F P-Value According o he resuls of Chow and Hausman es, and also he resuls of es he assumpions of he classical regression, model () using panel daa and as fixed effecs be esimaed. The resuls are presened in able 7. The model is esimaed using Views 7 sofware as follows: GDi, M / Bi, 0.944Tangi,.6609 Invi, DivPayer 0.508CashHolding SGi, Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 53

8 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle Table 7: Resuls of firs hypohesis esing using fixed effecs The dependen variable: he raio of changes in deb Number of views: 606 companies- year Variable Coefficien Saisics P-Value Relaionship Fixed componen Negaive Growh opporuniies Posiive The raio of Tangible asses Posiive Invesmen policy Negaive Dividend Negaive The raio of he cash asses Posiive Rae of sales growh Meaningless The deerminaion coefficien of model SaisicsF ( P-Value ).0308 (0.0000) In invesigaion of being significan of model, given ha he probabiliy of F-saisic is smaller han 0.05 (0.0000), wih 95% confidence, being significance of he model is confirmed. Deermining coefficien of model also indicaes ha, 30.3 percen of raio of changes in deb of companies is explained by he variables enered in he model. In invesigaion of being significan of coefficiens given he resuls presened in able 7, since he possibiliy of -saisic for variable coefficien of growh opporuniies is smaller han 0.05 (0.084), herefore, i is confirmed ha here is significan relaionship beween growh opporuniies and raio of changes in deb a he 95 percen confidence level. A posiive coefficien for his variable (0.047), indicaing ha here is a direc relaionship beween growh opporuniies and he raio of changes in deb of companies. Tes Resuls of he Second Research Hypohesis In he second hypohesis, he relaionship beween firm size and he raio of he changes in deb of companies are also sudied, and is saisical hypohesis is saed as follows: H : There is no significan relaionship beween Firm size and he raio of changes in deb of companies. 0 H : There is a significan relaionship beween Firm size and he raio of changes in deb of companies. This hypohesis using he model () as he panel daa is esimaed, and if he coefficien 3 is significan a a confidence level of 95%, i will be confirmed. GD LnSize Tang Inv DivPayer CashHolding SG () The resuls of he Chow es (o deermine he use of panel daa or hybrid daa mehod) and Hausman (o deermine he use of fixed or random effecs in he mehod of panel daa) for model (), are presened in Table 8. Table 8: The resuls of Chow and Hausman es for model () P-Value Degrees of freedom Saisics value Saisics Number Tes (00.499).058 F 606 Chow Hausman According o he resuls of he Chow es and P-Value, panel daa mehods can be used, and also according o he resuls of Hausman es and P-Value, i is necessary o esimae he model using fixed effecs. In invesigaion of assumpions of he classical regression, he resuls of Jarque-Bera indicae ha, he remaining of he research model a 95% confidence level have normal disribuion, and Breusch- Pagan es resuls based on his issue ha model has problem of non-homogeneiy of variance. In his sudy, o solve his problem, generalized leas squares esimaion (GLS) is used. Saisic of Durbin- Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 53

9 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle Wason also indicaes ha he residuals are independen (because Durbin-Wason Saisics is beween.5 -.5). In addiion, Ramsey es resuls indicae ha he model does no have clear error. Summary resuls of hese ess are presened in Table 9. Table 9: The resuls of es relaed o he saisical assumpions of he model () Jarque-Bera saisic Breusch-Pagan es Durbin Wason Ramsey es saisic P-Value F P-Value D F P-Value / According o he resuls of Chow and Hausman ess, and also he resuls of es he assumpions of he classical regression, model () using panel daa and as fixed effecs be esimaed. The resuls are presened in Table 0. Table 0: The resuls of he second hypohesis es using fixed-effecs mehod The dependen variable is he raio Changes in deb Number of views: 606 years - now Variable Coefficien Saisics P-Value Relaion Fixed componen Negaive Growh opporuniies Posiive The raio of Tangible asses Posiive Invesmen policy Negaive Dividend Meaningless The raio of he cash asses Posiive Rae of sales growh Meaningless The deerminaion coefficien of model Saisics F ( P-Value ).509 (0.0000) The model is esimaed using E views 7 sofware as follows: GDi, LnSize 0.535Tangi,.675Inv 0.07 DivPayer 0.390CashHolding SGi, In invesigaion of being significan of model, given ha he probabiliy of F-saisic is smaller han 0.05 (0.0000), wih 95% confidence, being significance of he model is confirmed. Deermining coefficien of model also indicaes ha, percen of raio of changes in deb of companies is explained by he variables enered in he model. In invesigaion of being significan of coefficiens given he resuls presened in Table 0, since he possibiliy of -saisic for variable coefficien of growh opporuniies is smaller han 0.05 (0.0000), herefore, i is confirmed ha here is significan relaionship beween Firm size and raio of changes in deb a he 95 percen confidence level. A posiive coefficien for his variable (.580), indicaing ha here is a direc relaionship beween Firm size and he raio of changes in deb of companies. Tes Resuls of he Third Research Hypohesis In he hird hypohesis, he relaionship beween growh opporuniies and he raio of Changes in reained earnings of companies is sudied, and is saisical hypohesis is saed as follows: H 0 : There is no significan relaionship beween growh opporuniies and he raio of Changes in reained earnings of companies. H : There is a significan relaionship beween growh opporuniies and he raio of Changes in reained earnings of companies. Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 533

10 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle This hypohesis using he model (4) as he panel daa is esimaed, and if he coefficien is significan a a confidence level of 95%, i will be confirmed. GREi, 0 M / Bi, 4Tangi, 5Invi, 6DivPayer 7CashHolding 8SGi, (4) The resuls of he Chow es (o deermine he use of panel daa or hybrid daa mehod) and Hausman (o deermine he use of fixed or random effecs in he mehod of panel daa) for model (4), are presened in Table. Table : The resuls of Chow and Hausman resuls for model (4) P-Value Degrees of freedom Saisics value Saisics Number Tes (00.499).6054 F 606 Chow Hausman According o he resuls of he Chow es and P-Value, panel daa mehods can be used, and also according o he resuls of Hausman es and P-Value, i is necessary o esimae he model using fixed effecs. In invesigaion of assumpions of he classical regression, he resuls of Jarque-Bera indicae ha, he remaining of he research model a 95% confidence level have normal disribuion, and Breusch- Pagan es resuls based on his issue ha model has problem of non-homogeneiy of variance. In his sudy, o solve his problem, generalized leas squares esimaion (GLS) is used. Saisic of Durbin- Wason also indicaes ha he residuals are independen (because Durbin-Wason Saisics is beween.5 -.5). In addiion, Ramsey es resuls indicae ha he model does no have clear error. Summary resuls of hese ess are presened in Table. Table : The resuls of es relaed o he saisical assumpions of he model (4) Durbin Wason Jarque-Bera saisic Breusch-Pagan es Ramsey es saisic P-Value F P-Value D F P-Value According o he resuls of Chow and Hausman ess, and also he resuls of es he assumpions of he classical regression, model (4) using panel daa and as fixed effecs be esimaed. The resuls are presened in Table 3. Table 3: The resuls of he hird hypohesis es using fixed-effecs mehod The dependen variable raio of Changes in reained earnings Number of views: 606 years - now Variable Coefficien Saisics P-Value Relaion Fixed componen Meaningless Growh opporuniies Negaive The raio of Tangible asses Meaningless Invesmen policy Negaive Dividend Posiive The raio of he cash asses Meaningless Rae of sales growh Meaningless The deerminaion coefficien of model Saisics F ( P Value ).8955 (0.0000) Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 534

11 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle The model is esimaed using E views 7 sofware as follows: In invesigaion of being significan of model, given ha he probabiliy of F-saisic is smaller han 0.05 (0.0000), wih 95% confidence, being significance of he model is confirmed. Deermining coefficien of model also indicaes ha, 8.70 percen of raio of Changes in reained earnings is explained by he variables enered in he model. In invesigaion of being significan of coefficiens given he resuls presened in Table 3, since he possibiliy of -saisic for variable coefficien of growh opporuniies is smaller han 0.05 (0.0038), herefore, i is confirmed ha here is significan relaionship beween Growh opporuniies and raio of Changes in reained earnings a he 95 percen confidence level. A negaive coefficien for his variable (-.0006), indicaing ha here is an inverse relaionship beween Growh opporuniies and raio of Changes in reained earnings. Tes Resuls of he Fourh Research Hypohesis The purpose of he fourh hypohesis is o examine asymmeric relaionship beween cash flows due o Firm size and he raio of changes in reained earnings of he companies, and is saisical hypohesis is saed as follows: H 0 : There is no significan relaionship beween Firm size and he raio of changes in reained earnings of he companies. H : There is a significan relaionship beween Firm size and he raio of changes in reained earnings of he companies. This hypohesis using he model (6) as he panel daa is esimaed, and if he coefficien 3 is significan a a confidence level of 95%, i will be confirmed. GRE LnSize Tang Inv DivPayer CashHolding SG (6) The resuls of he Chow es (o deermine he use of panel daa or hybrid daa mehod) and Hausman (o deermine he use of fixed or random effecs in he mehod of panel daa) for model (6), are presened in Table 4. Table 4: The resuls of Chow and Hausman es for he model (6) P-Value Degrees of freedom Saisics value Saisics Number Tes (00.499).6749 F 606 Chow Hausman According o he resuls of he Chow es and P-Value, panel daa mehods can be used, and also according o he resuls of Hausman es and P-Value, i is necessary o esimae he model using fixed effecs. In invesigaion of assumpions of he classical regression, he resuls of Jarque-Bera indicae ha, he remaining of he research model a 95% confidence level have normal disribuion, and Breusch- Pagan es resuls based on his issue ha model has problem of non-homogeneiy of variance. In his sudy, o solve his problem, generalized leas squares esimaion (GLS) is used. Saisic of Durbin-Wason also indicaes ha he residuals are independen (because Durbin-Wason Saisics is beween.5 -.5). In addiion, Ramsey es resuls indicae ha he model does no have clear error. Summary resuls of hese ess are presened in Table 5. Table 5: The resuls of es relaed o he saisical assumpions of he model (6) Jarque-Bera saisic Breusch-Pagan es Durbin Wason Ramsey es saisic P-Value F P-Value D F P-Value Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 535

12 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle According o he resuls of Chow and Hausman ess, and also he resuls of es he assumpions of he classical regression, model (6) using panel daa and as fixed effecs be esimaed. The resuls are presened in Table 6. Table 6: The resuls of he fourh hypohesis es using fixed-effecs mehod The dependen variable raio of Changes in reained earnings Number of views: 606 years - now Variable Coefficien Saisics P-Value Relaion Fixed componen Negaive Growh opporuniies Posiive The raio of Tangible asses Meaningless Invesmen policy Negaive Dividend Posiive The raio of he cash asses Meaningless Rae of sales growh Meaningless The deerminaion coefficien of model Saisics F ( P-Value ) 9 (0.0000) The model is esimaed using Eviews 7 sofware as follows: GRE 0 SG In LnSize 0.03Tangi, Invi, DivPayer 0.078CashHolding invesigaion of being significan of model, given ha he probabiliy of F-saisic is smaller han 0.05 (0.0000), wih 95% confidence, being significance of he model is confirmed. Deermining coefficien of model also indicaes ha, 9.64 percen of raio of Changes in reained earnings is explained by he variables enered in he model. In invesigaion of being significan of coefficiens given he resuls presened in Table 6, since he possibiliy of -saisic for variable coefficien of growh opporuniies is smaller han 0.05 (0.005), herefore, i is confirmed ha here is significan relaionship beween Firm size and raio of Changes in reained earnings a he 95 percen confidence level. A posiive coefficien for his variable (.995), indicaing ha here is a direc relaionship beween Firm size and raio of Changes in reained earnings. CONCLUSION In his sudy, we have examined he relaionship beween various measures of company growh and susainabiliy of he capial srucure of lised companies in Tehran Sock Exchange wih a oal of 0 samples in 006 o 0. To gaher informaion, he library mehod, and also for exracing saisical informaion, he audied financial saemens of lised companies in Tehran Sock Exchange are used, and for analysis, panel daa mehod is used. Using he firs hypohesis can be concluded ha here is a direc relaionship beween company's growh opporuniies and he raio of changes in deb. The resul of he firs hypohesis is in accordance wih he significan relaionship beween he independen variable and dependen of research of Ali (006), bu in erms of he ype of relaionship (direc or inverse), i is relaed o he resuls of Cheng e al., (009) and i is inconsisen wih he research of Graham and Harvey (00). In relaion o confirmaion of he second hypohesis, i can be concluded ha here is significan and direc relaionship beween Firm Size he raio of changes in deb of companies. In relaion o confirmaion of he hird hypohesis, i can be concluded ha here is significan and inverse relaionship beween growh opporuniies for he company and he raio of changes in reained earnings of companies. In relaion o confirmaion of he fourh hypohesis, i can be concluded ha here is significan and direc relaionship beween Firm size and he raio of changes in reained earnings of companies. The resuls of fourh hypohesis of our sudy in erms of relaionship are consisen wih he resuls of some sudies such Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 536

13 Indian Journal of Fundamenal and Applied Life Sciences ISSN: (Online) An Open Access, Online Inernaional Journal Available a hp://hp:// 04 Vol. 4 (S), pp /Darabi e al. Research Aricle as Liu (009) and pasor and veronesi (003), and i is inconsisen wih he resuls of oher research such as: Kachik and colleagues (990). REFERENCES Ali A (006). How persisen is he impac of marke iming on capial srucure? Bagheri M (009). The invesigaion of impac of financing mehods on he yield and price of shares of lised companies in Tehran Sock Exchange, MSc Thesis, Faculy of Managemen. Islamic Azad Universiy, Babol. Baker M and Wurgler J (00). Marke iming and capial srucure. Journal of Finance Byoun Soku (009). How and when do firms adjus heir capial srucures oward arges?.forhcoming Journal of Finance. Chang X and Dasgupa S (009). Targe behavior and financing: how conclusive is he evidence? Journal of Finance Cooney J and Kalay A (993). Posiive informaion from equiy issue announcemens. Journal of Financial Economics Farajpour Ahangar M (009). Sudying he effecs of differen mehods of financing of lised companies in Tehran Sock Exchange on sock reurns, MSc Thesis, Faculy of Managemen, Islamic Azad Universiy, Babol. Graham J and Harvey C (00). The heory and pracice of corporae finance: evidence from he field. Journal of Financial Economics Kimiagari Ali Mohamed and Eynali S (0). A comprehensive model of capial srucure (firms lised in Tehran Sock Exchange). Financial Research (5) Korajczyk R, Lucas D and McDonald R (990). Undersanding Sock Price Behavioraround he Time of Equiy Issues. Edied by Hubbard RG, In: Asymmeric InFormaion, Corporae Finance, and Invesmen. (Universiy of Chicago Press), Chicago. Kordesani GH and Mazaher Najafi Omran (008). The invesigaion of he deerminans of capial srucure: An empirical es of ISNA equilibrium heory versus hierarchical heory, Journal of Financial Research 0(5) Liu X (009). Hisorical marke-o-book in a parial adjusmen model of leverage. Journal of Corporae Finance Myers S and Majluf N (008). Corporae financing and invesmen decisions when firms have informaion ha invesors do no have. Journal of Financial Economics Pasor L and Veronesi P (003). Sock valuaion and learning abou profiabiliy. Journal of Finance Sein J (996). Raional capial budgeing in an irraional world. Journal of Business Wu X and Wang Z (005). Equiy financing in a Myers Majluf framework wih privaebenefis of conrol. Journal of Corporae Finance Wu X and Xu L (005). The value informaion of financing decisions and corporaegovernance during and afer he Japanese deregulaion. Journal of Business Wu X, Sercu P and Yao J (009). Does compeiion from new equiy miigae bank renexracion? Insighs from Japanese daa. Journal of Banking and Finance Wu X, Wang Z and Yao J (005). Undersanding he posiive announcemen effecs ofprivae equiy placemens: new insighs from Hong Kong daa. Review of Finance Copyrigh 04 Cenre for Info Bio Technology (CIBTech) 537