Inernaional Business Research; Vol. 8, No. 4; 2015 ISSN 1913-9004 E-ISSN 1913-9012 Published by Canadian Cener of Science and Educaion The Influence of Earnings Qualiy and Liquidiy on he Cos of Equiy Ming-Feng Hsu 1 & Jean Yu 1 1 Deparmen of Banking and Finance, Naional Chiayi Universiy, Chiay Taiwan Correspondence: Jean Yu, Deparmen of Banking and Finance, Naional Chiayi Universiy, Chiayi Ciy, 60054 Taiwan. Tel: 886-5-273-2867. E-mail: jean@mail.ncyu.edu.w Received: February 2, 2015 cceped: February 25, 2015 Online Published: March 25, 2015 doi:10.5539/ibr.v8n4p194 URL: hp://dx.doi.org/10.5539/ibr.v8n4p194 bsrac This sudy uses sample companies lised in Taiwan Sock Exchange and GreTai Securiies Marke during 2000 o 2011 o invesigae he influence of earnings equaliy and liquidiy on he cos of equiy. We define discreionary accruals wih hree measures and real earnings managemen wih hree measures as indicaors of earnings qualiy; rading volume, individual sock liquidiy and marke liquidiy as liquidiy measures and individual sock and marke liquidiy risk as liquidiy risk measures. Panel daa is suggesed for his analysis. Firms manipulaing discreionary accruals increase in he cos of equiy, bu ones operaing real earnings managemen decrease in i when considering ha he earnings qualiy and liquidiy direcly impac on i. The cos of equiy is indirecly influenced by earnings qualiy and liquidiy hrough informaion asymmery measured by bid-ask spreads. The resuls show ha no maer firms engaging in discreionary accruals or real earnings can decrease he cos of equiy under higher levels of informaion asymmery. The higher he rading volume or he individual sock liquidiy risk, he lower he cos of equiy when informaion asymmery is low. Keywords: earnings managemen, liquidiy, cos of equiy, informaion asymmery 1. Inroducion The company cos of capial decided by manager decisions, asse valuaion and financial reporing has become an imporan research opic, in paricular he relaionship beween earnings qualiy and he cos of equiy in he accouning field. Francis e al. (2004) presen he accouning-based aribues significanly impac he cos of equiy, especially on he qualiy of accruals. Bhaacharya e al. (2013) follow Francis e al. (2005) using an accruals-based measure as proxy for earnings qualiy, for he firms lis on he New York Sock Exchange (NYSE) and NSDQ o es he relaionship beween informaion asymmery and earnings qualiy. The resuls show ha poor earnings qualiy is highly correlaed o informaion asymmery, hus making he company coss of capial higher. Numerous sudies discuss he effecs on differen proxies of earnings qualiy on eiher equiy or deb coss (see Boosan, 1997; Boosan & Plumlee, 2002; boody e al., 2005; Francis e al., 2005). However, he resuls are inconclusive. The effecs of associaing facors inerac more complex wih he environmen, measured a boh he company and counry level. Gray e al. (2009) analyze relaions beween accruals qualiy and cos of capial for usralian firms. They find significan posiive relaionship beween earnings qualiy and he cos of equiy. These findings are consisen wih he empirical resuls of Francis e al. (2005). In his sudy, we es he relaionship beween accrual earnings qualiy and he cos of equiy for Taiwan's lised companies and find he poor earnings qualiy is associaed wih higher cos of equiy, consisen wih he empirical resuls of Gray e al. (2009). Informaion asymmery embedding fricion reduces volailiy according o corporae financial managemen heory. Chae (2005) presens an inverse relaionship beween sock liquidiy and informaion asymmery before earnings announcemens, bu a posiive relaionship afer he earnings announcemen. ig e al. (2006) find ha greaer deviaion beween excess conrol and ownership have a higher informaion asymmery componen in heir bid ask spread, suggesing poor disclosure qualiy deerioraes liquidiy. Bu Kyle (1985) presens a model shows how a risk neural and perfecly informed insider sraegically profis from his privae informaion and hus induce he liquidiy. Deng and Ong (2014) es he manipulaion effec of earnings managemen o he cos of equiy for he Real Esae Invesmen Truss (REITs). They find managers manipulae earnings managemen during he company's capial raising period o arac uninformed invesors, hus increasing liquidiy, hereby 194
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 reducing he cos of equiy. Numerous sudies explore he reasons for he sock liquidiy, in paricular he impac of liquidiy on he required reurn of sock, which is he cos of capial. Capial sse Pricing Model based on adjused liquidiy is anoher model o exam he effec of liquidiy risk on he sock price or reurns. Following Pasor and Sambaugh (2003), we calculae marke liquidiy risk, and apply discreionary accruals and real earnings managemen as a measure of earnings qualiy o analyzing he impac on he cos of equiy. Our resuls are similar o wha Deng and Ong (2014) find ha he company managers use real earnings managemen o reduce he cos of equiy. Previous sudies invesigae he impac of earnings qualiy as informaion risk on he cos of equiy, some oher sudies analyze he relaionship among he individual liquidiy, liquidiy risk and capial coss. Mos of he sudies show ha earnings qualiy and individual liquidiy affec he cos of equiy. Bhaacharya e al. (2012) use pah analysis o exam he direc and indirec links beween hree measures of earnings qualiy and he cos of equiy. They specify analyical models ha specify boh a direc link and an indirec link ha is reconciled by informaion asymmery consising selecion componen of he bid-ask spread and PIN (he probabiliy of informed rading). They find significan evidence of boh a direc pah from earnings qualiy and an indirec pah from informaion asymmery. Following hree-facor model of Fama-French (1993), we use abnormal reurns as he cos of equiy capial by collecing daa from 2000 o 2011 in Taiwan o discuss a direc impac on cos of equiy from earnings qualiy and liquidiy of he shares, along wih various levels of informaion asymmery facor. Our conribuion is wofold. Firs, he direc impac on equiy of cos is prominen, he worse qualiy of discreionary accruals becomes and he higher rading volume becomes, he lower cos of equiy is. Second, he higher he rading volume or he individual sock liquidiy risk, he lower he cos of equiy when informaion asymmery is low. The paper is organized as follows. Secion I inroduces he background and moivaion. In secion II, we discuss he hypohesis. Secion III describes models, mehodology and variables. Secion IV presens empirical resuls and concludes. 2. Hypohesis Design Previous sudies show ha he firms are less likely o do managemen earnings wih higher ransparency of financial reporing when accouning informaion are fully disclosure. Therefore, manager will selecively reveal financial informaion wih asymmeric informaion as hey raionally expec profi from engaging in earnings managemen. Francis e al. (2004) examine seven earnings aribues which have disinguishable effecs on firm-specific risk premium. Boosan and Plumlee (2002) find a posiive associaion beween cos of capial and volunary imely disclosure. They show ha relaion beween disclosure and cos of equiy capial changes from a negaive o a posiive relaion. Bhaacharya e al. (2012) presen poor qualiy of earnings managemen ha induces higher adverse selecion risk and lower capial marke liquidiy. Therefore, invesor requires higher risk premium and hus increase he cos of equiy. Former chairman of he US Securiies and Exchange Commission Levi,. said before ha he mos imporan benefis of he high-qualiy accouning sandards are o improve liquidiy and reduce he cos of capial for sock marke. We apply Fama and French (1993) hree-facor model and follow modified Jones Model in Dechow e al. (1995) o esimaing discreionary accruals as proxies in evaluaing he qualiy of earnings managemen. s he corporae finance lieraure empirically saes ha informaion asymmeries reduce he liquidiy of he company s securiies and hus induce marke fricions, Deng and Ong (2014) find firms wih less liquid are very likely o manipulae earnings prior heir equiy offerings. Therefore, higher uninformed rading follows he real earnings managemen. Firms se he offer price a a smaller discoun afer engaging in real earnings managemen and sock reurns decline in he long run. We propose he following hree hypoheses: Hypohesis 1: When applying discreionary accruals as proxies in evaluaing he qualiy of earnings managemen, poor qualiy managemen increases he cos of equiy, and vice versa. Few sudies apply real earnings managemen as proxy o evaluaing he company's earnings qualiy on he cos of equiy capial by using REITs sample. When engaging in real earnings managemen, firms will increase liquidiy, and reduce he cos of equiy on increasing capial wih cash. In his sudy, we define real earning managemen wih hree measures and impac on cos of equiy. Hypohesis 2: Uninformed raders do no have sufficien informaion when he company involves in real 195
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 earnings managemen, so even wih poor earnings qualiy, cos of equiy sill decreases. The effec of liquidiy flucuaions on asse prices increases he risk of individual invesors wih insufficien informaion. Invesors will require a higher risk premium when liquidiy risks increase, so as sock reurns. In recen years, many researchers have begun o explore he relaionship beween sock marke reurns and liquidiy wih various proxies. Chordia e al. (2001) show resuls of significanly negaive cross-secional relaionship beween sock reurns and he variabiliy of dollar rading volume, which is conrary o radiional hypohesis ha risk-averse invesors who will demand a higher risk premium, and hen increase he company's coss of equiy. On conrary, charya and Pedersen (2005) propose a sock s required reurn depends on is expeced liquidiy as well as on he covariances of is own reurn and liquidiy wih he marke reurn. Moreover, low conemporaneous reurns and high prediced fuure reurns conribue o persisen negaive shock. Therefore, he impac of liquidiy on sock reurns is inconclusive. Lee (2011) uses he liquidiy-adjused capial asse pricing model (LCPM) of charya and Pedersen (2005) on invesigaing from 48 developed and emerging counries around he world. He finds ha various liquidiy risks arising from he covariances of asse reurn (liquidiy) wih local and global marke liquidiy (reurns) are priced. However, liquidiy risk also affeced by he geographic, economic and poliical condiions, especially for he developing counries. Davivongs and Pavadur (2012) also use he liquidiy adjused capial asse pricing model o invesigae he liquidiy risk of socks in China and Taiwan. They find he evidence ha sysemaic liquidiy risk is more imporan han marke risk in Taiwan. Therefore, we form hypohesis 3 as follows: Hypohesis 3: The higher he liquidiy, he lower he cos of equiy. Therefore, he higher he liquidiy risk, he higher he cos of equiy. 3. Research Model In order o examine he effec of earnings qualiy and liquidiy on he cos of equiy capial, a sample of 943 Taiwan-lised companies which excluding financial and insurance-relaed indusries is colleced covering year 2000 o 2011. The disribuion of sample indusry is shown in Table 1. There are 506 companies from elecronics indusry, abou 54% of all samples; followed by he building maerials and chemical indusries, 63 and 53 respecively, each accouning for abou 6% of all samples. The rubber, elecriciy, gas and oil, and ourism indusry, which have he leas companies, are abou 1% of he oal number of samples. Daily sock rading volume and monhly financial informaion are colleced from Taiwan Economic Journal (TEJ) daabase. The final daa decrease o 135,428 afer deducing incomplee daa of 11,334 from 146,762 monhly daa. Table 1. The sample disribuion of lised companies Indusry Number Percen of sample Indusry Number Percen of sample Food 22 2.33 Shipping & Transporaion 23 2.44 Plasic 25 2.65 Trading & Consumers Goods 16 1.70 Texile 48 5.09 Building Maerial & Consrucion 63 6.68 Elecric Machinery 48 5.09 Tourism 10 1.06 Elecronic Indusry 506 53.66 Elecrical & Cable 15 1.59 Bioechnology & Medical Care 10 1.06 Oil, Gas & Elecriciy Indusry 10 1.06 Chemical Indusry 53 5.62 Ohers 50 5.30 Iron & Seel 35 3.71 Toal number 943 100 Rubber 9 0.95 The paper uses individual sock abnormal reurns measured by hree-facor model which is proposed by Fama and French (1993) as he company s cos of capial. We specify he following regression model verificaion: Ri, 0 1EM 2Volume 3LiRisk 4Liqui 5MLiRisk 6MLiqui 7MKT 8SMB 9HML 10Marki jindusryi, j j (1) 196
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 where R is he abnormal reurns for he sock i in period measured by Fama-French hree-facor model. EM is earnings qualiy of firm i in period. The earnings qualiy is measured by he modified Jones model proposed by Dechow e al (1995), including abnormal oal discreionary accruals (TD), abnormal discreionary curren accruals (DC), and abnormal discreionary long-erm accruals (DL) as a measure. noher way o measure earnings qualiy is real earnings managemen model proposed by Roychowdhury (2006), which includes abnormal cash flow from operaion(cfo), abnormal producion coss (PC), abnormal discreionary expenses (DE), and inegraed indicaors (CBPM = PC-CFO-DE). Volume (PVolume ) is he oal volume of sock rading (oal ransacion value of sock rading) divided by he oal volume of ousanding shares (oal ransacion value of ousanding shares) of firm i in period. LiRisk is sock liquidiy risk of firm i in period, and LiRisk =(B -m(b ))/s(b ). B is he average monhly bid-ask spread. m(b ) and s(b ) are he mean and sandard deviaion, respecively. Liqui is he liquidiy of he sock of firm i in period, which is equal o he reciprocal of he bid-ask spread. MLiRisk is liquidiy risk of he marke in period, MLiRisk =(MLiqui -m(mliqui ))/s(mliqui ). MLiqui is marke liquidiy and MLiqui = γ /N, i=1 N. m(mliqui ) and (MLiqui ) are he mean and sandard deviaion of MLiqui, respecively. Marke liquidiy is esimaed based on he mehod proposed by Pasor and Sambaugh (2003), he esimaed equaion is as follows. e e R d 1, Ri, d, sign( Ri, d, ) Volume d, (2) Where R d, is reurn rae of sock on day d in period. R e d, is he abnormal reurn on day d in period. sign(r e d,) = -1, when R e d, <0; sign(r e d,) = 0, when R e d, =0; sign(r e d,) =1, when R e d, >0. Volume d, is he firm's sock rading volume on day d in period. KMT are he marke risk facors in period, which equals o he marke porfolio reurn minus he risk-free ineres rae. The risk-free ineres rae is he monh deposi rae of Bank of Taiwan. SMB is he scale risk facor in period, including large-scale and small-scale facor based on he marke value of he sock. HML are he ne marke value raio risk facors and is divided ino hree levels, he highes 30%, he middle 40%, and he lowes 30%. Marke i is equal o 1 when he sample is lised on he Taiwan Sock Exchange, and zero oherwise. Indusry j is equal o 1 if he company is indusry j, and zero oherwise. The descripion of a measure of earnings managemen is depiced as follows. Firs, using he modified Jones model, abnormal oal discreionary accruals (TD ), abnormal discreionary curren accruals (DC ) and abnormal discreionary long-erm accruals (DL ) are esimaed in he following models. T 1 SLES R PPE TD ˆ ˆ ˆ ˆ 0 1 2 3 (3) i, 1 i, 1 i, 1 i, 1 Where T = NI - CFO. NI is he ne profi of firm i in period. he firs. CFO is he cash flow from operaing aciviies of firm i in period. -1 is he oal asse of firm i in he period -1. ΔSLES is he change amoun of sales revenue of firm i beween period and -1. ΔR is he change in he amoun of accouns receivable beween period and -1. PPE is he oal amoun of fixed asses of firm i in period. DC Ci, CSH CL STD 1 SLES R ˆ ˆ 0 1 2 (4) i, 1 i, ˆ i, 1 1 Where ΔC is he change amoun of curren asses of firm i in period. ΔCSH is he change amoun of cash of firm i in period. ΔCL is he change amoun of curren liabiliies of firm i in period. ΔSTD is he change amoun of long-erm deb due wihin one year for he firm i in period. The definiions of oher variables are same as hose of equaions (3). The abnormal long-erm discreionary accruals can be obained using he oal abnormal discreionary accruals measured by formula (3) minus he abnormal curren discreionary accruals measured by formula (4). Essence of real earnings managemen is he use of cash flow from operaion, producion coss, and discreionary expenses as indicaors o manipulae earnings. Manipulaion mehods described in he followings affec he abnormal levels of hree indicaors above. The firs is he excessive price discouns and producion will lead o an abnormally high producion coss, hus reducing he cash flow from operaion. Second, reducing discreionary expendiure resuls in abnormally low discreionary expenses, bu has higher cash flow from operaion. Therefore, under he fixed sales revenue, real earnings manipulaion will resul in abnormally low cash flow from operaion and discreionary expenses, bu will increase abnormal producion coss. Thus come up wih real earnings managemen model is as follows: 197
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 PCos 1 CFO 1 1 ˆ ˆ 0 1 1 SLES ˆ 2 1 SLES ˆ 3, 1 1 SLES SLES SLES i 1 ˆ ˆ ˆ ˆ ˆ 0 1 2 3 4, (6) 1 DExpense 1 1 1 ˆ ˆ 0 1 1 1 1 SLSE ˆ 2 Where CFO is cash flow from operaion of firm i in period. PCos is he producion coss of firm i in period. DExpense is discreionary expenses of firm i in period, including adverising, research and developmen, managemen and markeing coss. SLES is sales revenue of firm i in period. ΔSLES is he change amoun of sales revenue beween period and period -1. SLES -1 is sales revenue of firm i in period -1. ΔSLES -1 is he change amoun of sales revenue beween period -1 and period -2. The sandard esimae of real earnings managemen is obained from equaion (5) o equaion (7). The abnormal cash flow from operaion is obained using he acual cash flow from operaion minus he sandard cash flow equaion (5). Similarly, abnormal producion coss can be obained using he acual cos minus producion formula (6); and abnormal discreionary expenses is he acual discreionary expenses minus formula (7). 4. Empirical Resuls 4.1 Summary Saisics Table 2 shows he summary saisics of variables. The average abnormal sock reurns derived from Fama French hree-facor model is -0.31 wih skewed o he righ median of -1.60. In addiion, here are a few companies whose abnormal reurns are less han zero. I may be due o he economic condiion wih he sock index during he sample period in relaively low. Therefore, invesors pick on firms wih good earnings qualiy or high liquidiy o achieve high sock reurn. 1 1, (5) (7) Table 2. Descripive saisics for he measures in he firs regression equaion Mean Median Maximum Minimum Sd. dev. R -0.3039-1.5992 275.45-82.02 16.2655 DC -0.0011-0.0040 3.7654-8.5938 0.2029 DL 0.0006-0.0004 8.3518-4.1949 0.2188 TD -0.0005-0.0018 2.3024-1.8118 0.1309 CFO -0.0035-0.0056 1.2892-3.9988 0.1365 PC 0.0115 0.0163 3.2432-2.0220 0.1708 DE -0.0085-0.0158 0.8805-0.9796 0.0775 LiRisk -0.0004 0.0590 10.5020-10.8038 1.0019 MLiRisk -0.0010-0.0582 11.5654-3.0606 1.0317 Liqui -5.79E+09 1.5060 7.59E+13-8.77E+13 1.13E+12 MLiqui 232.10-58.77 72917-18908 5416 Volume 0.1799 0.0950 2.9105 3.08E-06 0.2303 PVolume 0.1832 0.0972 22.7200 2.98E-06 0.2402 MKT 0.3207-0.0019 51.1862-26.5584 9.0207 SMB 1.5320 1.3328 18.8803-24.9800 4.2383 HML 1.0543 0.9800 25.8576-29.1351 4.9570 Noe. R= abnormal reurn; DC= abnormal discreionary curren accruals; DL = abnormal discreionary long-erm accruals; TD = abnormal oal discreionary accruals; CFO= abnormal cash flow from operaion; PC= abnormal producion coss; DE= abnormal discreionary expenses; Volume= monhly average rading volume; PVolume= monhly average rading amoun; LiRisk= firm liquidiy risk; MLiRisk =marke liquidiy risk; Liqui= firm liquidiy; MLiqui= marke liquidiy; MKT= marke premium facor; SMB=size facor; HML=book-o-marke facor. 198
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 Excep for he abnormal discreionary long-erm accruals, averages and medians of abnormal discreionary curren accruals and abnormal oal discreionary accruals are negaive. Mos of he firms migh experience low earnings in he sample period and ake discreionary wrie-offs o reduce furher he curren period s earnings as he big bah heory of earnings managemen suggess. The saisics presen real earnings managemen by observing averages and medians of cash flow from operaion and discreionary expenses is negaive as well as unusual observaion of posiive averages and medians on producion coss. Individual sock liquidiy sandard deviaion is relaive large wih average skewed o he lef and a median of 1.51. Marke liquidiy is relaively high, so he marke's liquidiy risk is relaively low. Table 3 shows he Pearson and Spearman correlaion coefficien of equiy cos of capial, earnings qualiy and liquidiy wih oher variables. The correlaion coefficiens for earnings qualiy and liquidiy and he cos of equiy capial are consisen. The hree indicaors of discreionary accruals are posiively associaed wih he cos of equiy, bu he ones of real earnings managemen have an opposie relaionship. Tha is, when he company implemening discreionary accruals, he cos of capial will increase, vice versa for he real earnings managemen. Sock rading volume is posiively relaed o he cos of equiy capial. Wheher i is individual sock liquidiy or marke liquidiy risk, he relaionship is in inverse o he cos of equiy. The higher he liquidiy risk, he lower he cos of equiy capial. This is no corresponding o our hypohesis which previous empirical resuls show. Table 3. The correlaion marix on he variables of earnings qualiy, liquidiy and he cos of equiy Correlaion R DC TD DL CFO PC DE Volume PVolume LiRisk MLiRisk Liqui MLiqui R 0.0200 0.0236 0.0048 0.0372-0.0351 0.0128 0.2112 0.1718-0.0961-0.0408 0.0403-0.0448 DC 0.0151 0.2097-0.6314-0.1282-0.0265 0.0208-0.0021-0.0027-0.0061-0.0013-0.0023-0.0006 TD 0.0268 0.1749 0.5226-0.5378 0.0938 0.0010 0.0204 0.0191-0.0260 0.0042 0.0370 0.0045 DL 0.0025-0.8016 0.4485-0.2852 0.1017-0.0081 0.0114 0.0108-0.0081 0.0025 0.0228 0.0022 CFO 0.0139-0.1476-0.6775-0.2774-0.4174 0.0860 0.0209 0.0183-0.0088-0.0042 0.0517-0.0044 PC -0.0162-0.0181 0.1528 0.1093-0.3682-0.5577 0.0709 0.0727 0.0294 0.0071-0.0123 0.0075 DE 0.0193-0.0153-0.0493-0.016 0.1015-0.5102-0.0798-0.0803-0.0032 0.0037-0.0411 0.0031 Volume 0.2775-0.001 0.0046 0.0037 0.0424 0.0283-0.0292 0.9983-0.2786 0.0111 0.0721 0.0075 PVolume 0.1992-0.0022 0.0029 0.0038 0.0396 0.0276-0.0302 0.9311-0.2782 0.0119 0.0691 0.0084 LiRisk -0.0841 0.0015-0.0002-0.0015-0.0132 0.001 0.0101-0.1561-0.1476 0.0155-0.0129 0.0207 MLiRisk -0.0075 0.0009-0.0008-0.0013 0.0009 0.0025-0.0078-0.0225-0.0226 0.014 0.0041 0.9947 Liqui 0.0025 0.0025-0.0038-0.0046-0.0025 0.0045-0.0011-0.0041-0.0043-0.0006 0.0091 0.0034 MLiqui -0.0078 0.002 0.0003-0.0017-0.0003 0.0037-0.009-0.0232-0.0233 0.0137 0.9679 0.0099 Noe. R= abnormal reurn; DC= abnormal discreionary curren accruals; DL = abnormal discreionary long-erm accruals; TD = abnormal oal discreionary accruals; CFO= abnormal cash flow from operaion; PC= abnormal producion coss; DE= abnormal discreionary expenses; Volume= monhly average rading volume; PVolume= monhly average rading amoun; LiRisk= firm liquidiy risk; MLiRisk =marke liquidiy risk; Liqui= firm liquidiy; MLiqui= marke liquidiy. Pearson correlaions are displayed below he main diagonal and Spearman correlaions are displayed above he diagonal in Table. 4.2 Effecs of Earning Managemen and Liquidiy From eiher Lagrange muliplier es or Hausman s es, he resuls of panel daa regression equaion are valid wih fixed effecs. Panel and Panel B in Table 4 presen volume and ransacion value respecively. The resuls find ha hree indicaors of discreionary accruals all hold significan posiive relaionship wih he cos of equiy capial, especially for long-erm and oal accruals. Invesors reques higher cos of equiy capial when a company has poor earnings qualiy of accruals, especially he resuls show ha he long-erm accruals are hree percen han curren accruals (Panel : 0.674 23.77% -0.471 26.22%; Panel B: 0.68 24.18% -0.5178 26.68%). The resuls suppor hypohesis 1. 199
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 Table 4. The fixed effecs resuls of he influence of earnings qualiy and liquidiy on he cos of equiy Panel R Inercep -2.7042 *** -2.6847 *** -2.6809 *** -2.7073 *** -2.7513 *** -2.7364 *** -2.7519 *** (0.2329) (0.2329) (0.2328) (0.2328) (0.2331) (0.2330) (0.2331) DC 0.4705 * (0.2622) DL 0.6736 *** (0.2377) TD 2.8729 *** (0.3911) CFO 1.0180 *** (0.3544) PC -1.2564 *** (0.2968) DE 2.7603 *** (0.6381) CREM -0.8261 *** (0.1683) Volume 14.3486 *** 14.3501 *** 14.3360 *** 14.3236 *** 14.3577 *** 14.3513 *** 14.3330 *** (0.2155) (0.2155) (0.2154) (0.2157) (0.2155) (0.2155) (0.2155) LiRisk -0.7963 *** -0.7965 *** -0.7969 *** -0.7953 *** -0.7952 *** -0.7982 *** -0.7952 *** (0.0491) (0.0491) (0.0491) (0.0491) (0.0491) (0.0491) (0.0491) MLiRisk 0.4818 ** 0.4776 ** 0.4829 ** 0.4763 ** 0.4759 ** 0.4751 ** 0.4735 ** (0.2140) (0.2140) (0.2139) (0.2140) (0.2140) (0.2140) (0.2140) Liqui 4.84E-14 4.92E-14 4.98E-14 4.90E-14 4.94E-14 4.86E-14 4.94E-14 (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) MLiqui -7.93E-05 ** -7.81E-05 * -7.92E-05 ** -7.81E-05 * -7.79E-05 * -7.76E-05 * -7.73E-05 * (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) MKT 0.4923 *** 0.4897 *** 0.4914 *** 0.4913 *** 0.4928 *** 0.4904 *** 0.4922 *** (0.0313) (0.0313) (0.0313) (0.0313) (0.0313) (0.0313) (0.0313) SMB 0.1703 *** 0.1701 *** 0.1695 *** 0.1709 *** 0.1691 *** 0.1703 *** 0.1700 *** (0.0314) (0.0314) (0.0313) (0.0314) (0.0314) (0.0313) (0.0313) HML 0.1441 *** 0.1442 *** 0.1444 *** 0.1439 *** 0.1436 *** 0.1437 *** 0.1435 *** (0.0229) (0.0229) (0.0229) (0.0229) (0.0229) (0.0229) (0.0229) Marke 0.3206 ** 0.2948 ** 0.3076 ** 0.3062 ** 0.3772 *** 0.3962 *** 0.3764 *** (0.1285) (0.1285) (0.1283) (0.1283) (0.1293) (0.1298) (0.1290) Observaions 67634 67634 67640 67640 67640 67640 67640 dj R 2 0.4082 0.4083 0.4087 0.4083 0.4084 0.4084 0.4084 200
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 Panel B Inercep -2.3311 *** -2.3109 *** -2.3065 *** -2.3370 *** -2.3761 *** -2.3639 *** -2.3837 *** (0.2369) (0.2369) (0.2367) (0.2368) (0.2371) (0.2370) (0.2370) DC 0.5178 * (0.2668) DL 0.6794 *** (0.2418) TD 2.9939 *** (0.3979) CFO 1.4030 *** (0.3605) PC -1.2282 *** (0.3019) DE 2.8472 *** (0.6491) CREM -0.9099 *** (0.1712) PVolume 9.0415 *** 9.0425 *** 9.0326 *** 9.0112 *** 9.0503 *** 9.0462 *** 9.0284 *** (0.2013) (0.2013) (0.2012) (0.2014) (0.2013) (0.2012) (0.2013) LiRisk -0.9492 *** -0.9494 *** -0.9498 *** -0.9476 *** -0.9482 *** -0.9511 *** -0.9479 *** (0.0499) (0.0499) (0.0499) (0.0499) (0.0499) (0.0499) (0.0499) MLiRisk 0.4179 * 0.4134 * 0.4189 * 0.4111 * 0.4117 * 0.4107 * 0.4087 * (0.2177) (0.2177) (0.2176) (0.2177) (0.2176) (0.2176) (0.2176) Liqui 4.36E-14 4.45E-14 4.51E-14 4.44E-14 4.46E-14 4.39E-14 4.48E-14 (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) MLiqui -6.67E-05-6.55E-05-6.67E-05-6.53E-05-6.53E-05-6.49E-05-6.46E-05 (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) MKT 0.5321 *** 0.5294 *** 0.5311 *** 0.5311 *** 0.5325 *** 0.5301 *** 0.5320 *** (0.0318) (0.0318) (0.0318) (0.0318) (0.0318) (0.0318) (0.0318) SMB 0.1993 *** 0.1991 *** 0.1984 *** 0.2001 *** 0.1982 *** 0.1992 *** 0.1989 *** (0.0319) (0.0319) (0.0319) (0.0319) (0.0319) (0.0319) (0.0319) HML 0.1460 *** 0.1462 *** 0.1464 *** 0.1457 *** 0.1456 *** 0.1457 *** 0.1454 *** (0.0233) (0.0233) (0.0233) (0.0233) (0.0233) (0.0233) (0.0233) Marke 0.5404 *** 0.5135 *** 0.5258 *** 0.5229 *** 0.5942 *** 0.6174 *** 0.6014 *** (0.1306) (0.1306) (0.1304) (0.1305) (0.1314) (0.1320) (0.1312) Observaions 67634 67634 67640 67640 67640 67640 67640 dj R 2 0.3877 0.3877 0.3882 0.3878 0.3878 0.3878 0.3879 Noe. R= abnormal reurn; DC= abnormal discreionary curren accruals; DL = abnormal discreionary long-erm accruals; TD = abnormal oal discreionary accruals; CFO= abnormal cash flow from operaion; PC= abnormal producion coss; DE= abnormal discreionary expenses; Volume= monhly average rading volume; PVolume= monhly average rading amoun; LiRisk= firm liquidiy risk; MLiRisk =marke liquidiy risk; Liqui= firm liquidiy; MLiqui= marke liquidiy; MKT= marke premium facor; SMB=size facor; HML=book-o-marke facor; Marke= marke dummy. Sandard errors are indicaed in parenheses. *, **, *** denoe saisical significance a he 10 percen, 5 percen and 1 percen levels (wo-ailed), respecively. 201
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 The hree indicaors of real earnings managemen have significan negaive effecs on equiy coss, wih discreionary expenses he mos prominen. Discreionary expenses are usually he company's adverising, research and oher operaing and adminisraive expenses. If he company did earnings managemen by reducing he above coss in an aim o increase operaing cash flow, hen invesors would no be able o gaher informaion. Consisen wih Deng and Ong (2014) in he sample of REITs, he composie indicaor (CREM) is negaively relaed o he cos of equiy. When he company implemens real earnings managemen, hus reduce he cos of equiy. Therefore, hypohesis 2 holds. The relaionship beween rading volumes and he cos of equiy in boh Panel and Panel B is significanly posiive which indicaing ha he greaer he amoun of sock rading volume, he higher he equiy cos. The cos of equiy for companies lised in he Taiwan Sock Exchange is higher han ha of lised in OTC. Liquidiy in Panel has a significan inverse relaionship wih he cos of equiy capial while marke liquidiy risk is significanly posiive o he cos of equiy indicaing risk premium compensaion. However, here is no significance in Panel B. Marke liquidiy risk in boh Panels is posiively associaed wih he cos of equiy, consisen wih previous empirical resuls of risk premium. Bu individual sock liquidiy risk relaed o he cos is in significanly inverse, implicaing he higher he individual liquidiy risk, he lower he cos of equiy. 4.3 Effecs of Informaion symmery The findings of his sudy as shown above are differen from previous researches. We follow Bhaacharya e al. (2012) and add he facor of informaion asymmery which has indirec impac on he equiy cos of capial. The bid-ask spread (BS) as a variable of informaion asymmery is calculaed by monhly sock prices as follows: askprice bidprice BS 100%, ( askprice bidprice)/ 2 (8) i Where askprice is daily selling price for socks; bidprice is daily purchase price. We rank absolue value of he bid-ask spread, and he large value indicaes higher degree of informaion asymmery. Companies wih higher degree of informaion asymmery formed in he firs quarile will have dummy variable of HBS se as one, and oherwise 0. The same rule applied for he companies in he boom quarile for dummy variable of LBS. The regression equaions are se as follows: R 0 1EM 2Volume ( PVolume ) 3LiRisk 4HBS 5LBS 6EM HBS 7EM LBS 8 Volume ( PVolume ) HBS 9Volume ( PVolume ) LBS 10LiRisk HBS 11LiRisk LBS 12MKT 13SMB 14HML 15Marke (9) i Excep for informaion asymmery variables, he oher specificaions of variables are he same as hose in regression equaion (1). The panel daa regression resuls are shown in Table 5. ccording o he Lagrange muliplier es or Hausman es, only variable (Volume ) which is monhly average rading volume of he regression equaion is significan a he level of 0.01 wih fixed effecs. However, in regression equaion wih monhly average rading amoun variable (PVolume ), he abnormal cash flow from operaion (CFO ) and abnormal producion coss (PC ) show he bes esimaes under random effecs while he res run he bes under fixed effecs. Mos of he resuls are similar o he prior findings excep for curren accruals. Bu afer adding he ineracion erm of earnings qualiy indicaors and informaion asymmery, when here is high degree of informaion asymmery, abnormal discreionary curren accruals are posiively relaed o he cos of equiy. However, abnormal discreionary long-erm accruals have an inverse relaionship wih he cos of equiy. Thus, company should op for long-erm discreionary accruals as regulaory agency reform may advise. In addiion, when experience higher informaion asymmery, firms benefi from doing real earnings managemen so as o cu cos of equiy. On conrary, companies wih higher ransparency refrain from earnings managemen, oherwise he cos of equiy increases. Under he lower level of informaion asymmery, wih he increase of sock rading volume and sock liquidiy risk will decrease in he cos of equiy capial. This incompleely suppors hypohesis 3. Tha is, firms benefis from he advanage of lower informaion asymmery which reduces he equiy of cos. 202
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 Table 5. The panel resuls from he cos of equiy indirecly influenced by earnings qualiy and liquidiy hrough informaion asymmery Panel R Fixed effecs Inercep -3.3320 *** -3.3149 *** -3.3130 *** -3.3849 *** -3.4091 *** -3.3686 *** -3.4256 *** DC -0.5662 (0.2357) (0.2357) (0.2356) (0.2358) (0.2362) (0.2358) (0.2360) (0.3783) DL 1.3897 *** (0.3458) TD 2.8873 *** (0.6147) CFO 3.9153 *** (0.5762) PC -2.0833 *** (0.4608) DE 2.7745 *** CREM (0.8684) -1.6344 *** (0.2544) Volmue 11.2413 *** 11.2400 *** 11.2285 *** 11.1788 *** 11.2994 *** 11.2635 *** 11.2637 *** (0.3190) (0.3190) (0.3189) (0.3191) (0.3193) (0.3191) (0.3190) LiRisk -0.5747 *** -0.5743 *** -0.5771 *** -0.5823 *** -0.5709 *** -0.5730 *** -0.5727 *** (0.0998) (0.0998) (0.0997) (0.0998) (0.0998) (0.0998) (0.0998) HBS 0.5593 *** 0.5529 *** 0.5901 *** 0.5530 *** 0.6098 *** 0.5602 *** 0.6005 *** (0.1488) (0.1489) (0.1489) (0.1492) (0.1491) (0.1489) (0.1492) LBS 0.8089 *** 0.8026 *** 0.8077 *** 0.7984 *** 0.7883 *** 0.7790 *** 0.7931 *** DC HBS 3.0723 *** (0.1524) (0.1524) (0.1523) (0.1524) (0.1528) (0.1537) (0.1526) (0.6005) DC LBS 0.0209 (0.6514) DL HBS -1.9973 *** (0.5435) DL LBS 0.1917 (0.5930) TD HBS 0.3904 (0.8861) TD LBS 0.0464 (0.9886) CFO HBS -5.0178 *** (0.7942) CFO LBS -2.0931 ** (0.9293) PC HBS 0.1113 (0.6571) 203
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 PC LBS 2.3656 *** (0.7534) DE HBS 2.8062 * (1.5557) DE LBS -2.8857 ** (1.4652) CREM HBS 0.9583 ** (0.3790) CREM LBS 1.3482 *** (0.4190) Volume HBS 15.7296 *** 15.7423 *** 15.7156 *** 15.8291 *** 15.6669 *** 15.7462 *** 15.6596 *** (0.4824) (0.4824) (0.4823) (0.4828) (0.4828) (0.4827) (0.4824) Volume LBS -3.9012 *** -3.8926 *** -3.8886 *** -3.8581 *** -3.9529 *** -3.9189 *** -3.9307 *** (0.4698) (0.4699) (0.4697) (0.4700) (0.4700) (0.4699) (0.4698) LiRisk HBS 0.0148 0.0160 0.0156 0.0234 0.0113 0.0079 0.0149 (0.1136) (0.1136) (0.1135) (0.1135) (0.1136) (0.1136) (0.1135) LiRisk LBS -1.9011 *** -1.8944 *** -1.8720 *** -1.9095 *** -1.9353 *** -1.9247 *** -1.9200 *** (0.2494) (0.2494) (0.2494) (0.2493) (0.2500) (0.2499) (0.2497) MKT 0.4829 *** 0.4795 *** 0.4811 *** 0.4822 *** 0.4837 *** 0.4801 *** 0.4825 *** (0.0304) (0.0304) (0.0303) (0.0303) (0.0304) (0.0304) (0.0304) SMB 0.1583 *** 0.1584 *** 0.1576 *** 0.1589 *** 0.1575 *** 0.1588 *** 0.1580 *** (0.0305) (0.0305) (0.0305) (0.0305) (0.0305) (0.0305) (0.0305) HML 0.1509 *** 0.1502 *** 0.1512 *** 0.1513 *** 0.1507 *** 0.1505 *** 0.1506 *** (0.0224) (0.0224) (0.0224) (0.0224) (0.0224) (0.0224) (0.0224) Marke 0.8570 *** 0.8318 *** 0.8567 *** 0.8774 *** 0.9410 *** 0.9366 *** 0.9585 *** (0.1318) (0.1318) (0.1316) (0.1317) (0.1329) (0.1330) (0.1326) Observaions 67665 67665 67671 67671 67671 67671 67671 dj R 2 0.4294 0.4294 0.4296 0.4296 0.4294 0.4294 0.4295 Hausman es 298.08 *** 300.35 *** 300.72 *** 309.86 *** 301.50 *** 320.93 *** 302.40 *** (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) Panel B Fixed effecs Inercep -3.2439 *** -3.2262 *** -3.2237 *** -3.3014 *** -3.3210 *** -3.2792 *** -3.3435 *** (0.2403) (0.2404) (0.2403) (0.2405) (0.2409) (0.2405) (0.2407) DC -0.5534 (0.3857) DL 1.3951 *** (0.3526) TD 2.9378 *** (0.6267) CFO 4.1508 *** (0.5873) PC -2.0540 *** (0.4699) DE 2.9250 *** (0.8854) 204
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 CREM -1.6850 *** (0.2593) PVolmue 8.1202 *** 8.1202 *** 8.1105 *** 8.0615 *** 8.1795 *** 8.1435 *** 8.1492 *** (0.3202) (0.3202) (0.3201) (0.3203) (0.3205) (0.3203) (0.3202) Lirisk -0.6673 *** -0.6668 *** -0.6697 *** -0.6747 *** -0.6635 *** -0.6654 *** -0.6650 *** (0.1017) (0.1018) (0.1017) (0.1017) (0.1017) (0.1018) (0.1017) HBS 1.0735 *** 1.0683 *** 1.1076 *** 1.0803 *** 1.1254 *** 1.0749 *** 1.1221 *** (0.1524) (0.1525) (0.1525) (0.1528) (0.1527) (0.1524) (0.1528) LBS 1.2112 *** 1.2056 *** 1.2092 *** 1.1991 *** 1.1944 *** 1.1893 *** 1.1976 *** (0.1482) (0.1482) (0.1480) (0.1482) (0.1486) (0.1495) (0.1484) DC HBS 3.1717 *** (0.6122) DC LBS 0.0776 (0.6641) DL HBS -2.0010 *** (0.5542) DL LBS 0.1481 (0.6046) TD HBS 0.5531 (0.9033) TD LBS 0.0455 (1.0079) CFO HBS -4.7408 *** (0.8096) CFO LBS -2.1765 ** (0.9473) PC HBS 0.0879 (0.6700) PC LBS 2.1681 *** (0.7681) DE HBS 2.1533 (1.5863) DE LBS -2.4694 * (1.4938) CREM HBS 0.8983 ** (0.3864) CREM LBS 1.2690 *** (0.4272) PVolume HBS 13.0023 *** 13.0087 *** 12.9802 *** 13.0725 *** 12.9339 *** 13.0082 *** 12.9157 *** (0.4929) (0.4929) (0.4928) (0.4932) (0.4933) (0.4932) (0.4929) PVolume LBS -4.8356 *** -4.8309 *** -4.8234 *** -4.7940 *** -4.8893 *** -4.8572 *** -4.8697 *** (0.4284) (0.4284) (0.4283) (0.4285) (0.4285) (0.4284) (0.4283) LiRisk HBS -0.0480-0.0469-0.0474-0.0393-0.0515-0.0543-0.0484 (0.1158) (0.1158) (0.1158) (0.1158) (0.1158) (0.1158) (0.1158) LiRisk LBS -1.8220 *** -1.8160 *** -1.7936 *** -1.8329 *** -1.8520 *** -1.8392 *** -1.8385 *** (0.2543) (0.2543) (0.2543) (0.2542) (0.2549) (0.2548) (0.2546) 205
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 MKT 0.5192 *** 0.5156 *** 0.5173 *** 0.5183 *** 0.5198 *** 0.5162 *** 0.5186 *** (0.0309) (0.0309) (0.0309) (0.0309) (0.0309) (0.0309) (0.0309) SMB 0.1813 *** 0.1813 *** 0.1805 *** 0.1820 *** 0.1804 *** 0.1816 *** 0.1809 *** (0.0311) (0.0311) (0.0311) (0.0311) (0.0311) (0.0311) (0.0311) HML 0.1526 *** 0.1519 *** 0.1529 *** 0.1528 *** 0.1523 *** 0.1522 *** 0.1521 *** (0.0229) (0.0229) (0.0229) (0.0229) (0.0229) (0.0229) (0.0229) Marke 1.0282 *** 1.0022 *** 1.0270 *** 1.0473 *** 1.1121 *** 1.1090 *** 1.1354 *** (0.1344) (0.1343) (0.1342) (0.1342) (0.1355) (0.1356) (0.1352) Observaions 67665 67665 67671 67671 67671 67671 67671 dj R 2 0.4069 0.4069 0.4072 0.4071 0.4069 0.4068 0.4071 Hausman es 271.35 *** 273.74 *** 273.89 *** 0.00 0.00 293.77 *** 275.86 *** (0.0000) (0.0000) (0.0000) (1.0000) (1.0000) (0.0000) (0.0000) Panel C Random effecs Inercep -4.3027 *** -4.2918 *** -4.2837 *** -4.3689 *** -4.3817 *** -4.3548 *** -4.4055 *** (0.2908) (0.2910) (0.2902) (0.2891) (0.2904) (0.2867) (0.2901) DC -0.4553 (0.3856) DL 1.3201 *** (0.3525) DT 2.9605 *** (0.6266) CFO 4.1643 *** (0.5873) PC -2.1392 *** (0.4697) DE 3.0073 *** (0.8852) CREM -1.7206 *** (0.2593) PVolmue 7.9909 *** 7.9909 *** 7.9803 *** 7.9286 *** 8.0518 *** 8.0063 *** 8.0196 *** (0.3195) (0.3195) (0.3194) (0.3196) (0.3198) (0.3196) (0.3195) LiRisk -0.6284 *** -0.6277 *** -0.6304 *** -0.6346 *** -0.6242 *** -0.6242 *** -0.6256 *** (0.1015) (0.1015) (0.1015) (0.1015) (0.1015) (0.1015) (0.1015) HBS 1.0573 *** 1.0514 *** 1.0919 *** 1.0633 *** 1.1114 *** 1.0577 *** 1.1069 *** (0.1524) (0.1525) (0.1525) (0.1527) (0.1526) (0.1524) (0.1527) LBS 1.1946 *** 1.1891 *** 1.1927 *** 1.1829 *** 1.1776 *** 1.1717 *** 1.1806 *** (0.1482) (0.1481) (0.1480) (0.1482) (0.1486) (0.1494) (0.1484) DC HBS 3.1878 *** (0.6122) DC LBS 0.0158 (0.6640) DL HBS -2.0038 *** (0.5541) DL LBS 0.1956 (0.6045) 206
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 TD HBS 0.5643 (0.9032) TD LBS 0.0282 (1.0078) CFO HBS -4.7850 *** (0.8096) CFO LBS -2.2027 ** (0.9473) PC HBS 0.1020 (0.6699) PC LBS 2.2401 *** (0.7680) DE HBS 2.0157 (1.5861) DE LBS -2.5820 * (1.4936) CREM HBS 0.9190 ** (0.3864) CREM LBS 1.3057 *** (0.4272) PVolume HBS 12.9388 *** 12.9453 *** 12.9166 *** 13.0097 *** 12.8679 *** 12.9419 *** 12.8505 *** (0.4927) (0.4928) (0.4926) (0.4931) (0.4931) (0.4931) (0.4928) PVolume LBS -4.7343 *** -4.7295 *** -4.7217 *** -4.6900 *** -4.7900 *** -4.7514 *** -4.7687 *** (0.4282) (0.4282) (0.4281) (0.4283) (0.4283) (0.4282) (0.4281) LiRisk HBS -0.0672-0.0661-0.0668-0.0592-0.0711-0.0748-0.0680 (0.1157) (0.1157) (0.1157) (0.1157) (0.1157) (0.1157) (0.1157) LiRisk LBS -1.7733 *** -1.7673 *** -1.7450 *** -1.7832 *** -1.8045 *** -1.7893 *** -1.7908 *** (0.2542) (0.2541) (0.2541) (0.2540) (0.2548) (0.2547) (0.2544) MKT 0.8916 *** 0.8900 *** 0.8917 *** 0.8950 *** 0.8933 *** 0.8988 *** 0.8931 *** (0.0174) (0.0175) (0.0174) (0.0172) (0.0173) (0.0170) (0.0173) SMB 0.4503 *** 0.4514 *** 0.4508 *** 0.4539 *** 0.4502 *** 0.4572 *** 0.4513 *** (0.0230) (0.0230) (0.0230) (0.0229) (0.0230) (0.0227) (0.0229) HML 0.1346 *** 0.1338 *** 0.1347 *** 0.1351 *** 0.1344 *** 0.1348 *** 0.1342 *** (0.0192) (0.0192) (0.0192) (0.0191) (0.0192) (0.0190) (0.0191) Marke 1.5725 *** 1.5486 *** 1.5729 *** 1.5954 *** 1.6602 *** 1.6656 *** 1.6832 *** (0.1279) (0.1279) (0.1277) (0.1277) (0.1290) (0.1290) (0.1287) Observaions 67665 67665 67671 67671 67671 67671 67671 dj R 2 0.1034 0.1032 0.1041 0.1047 0.1039 0.1056 0.1041 Noe. R= abnormal reurn; DC= abnormal discreionary curren accruals; DL = abnormal discreionary long-erm accruals; TD = abnormal oal discreionary accruals; CFO= abnormal cash flow from operaion; PC= abnormal producion coss; DE= abnormal discreionary expenses; Volume= monhly average rading volume; PVolume= monhly average rading amoun; LiRisk= firm liquidiy risk; MLiRisk =marke liquidiy risk; Liqui= firm liquidiy; MLiqui= marke liquidiy; HBS= he high level of informaion asymmery; LBS= he low level of informaion asymmery; MKT= marke premium facor; SMB=size facor; HML=book-o-marke facor; Marke= marke dummy. Sandard errors are indicaed in parenheses. *, **, *** denoe saisical significance a he 10 percen, 5 percen and 1 percen levels (wo-ailed), respecively. 207
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 5. Conclusions In his sudy, we invesigae he impac of earnings qualiy and liquidiy on cos of equiy for lised firms in Taiwan from 2000 o 2011. The proxies of discreionary accruals are calculaed by using abnormal discreionary curren accruals, abnormal discreionary long-erm accruals and abnormal oal discreionary accruals. Under real earnings managemen, we apply cash flow from operaion, producion coss and discreionary expense as he indicaors. We also apply individual liquidiy risk and marke liquidiy risk as liquidiy risk proxy. While liquidiy indicaors are esimaed using sock rading volume, individual sock liquidiy and marke liquidiy. The proxy of cos of equiy uses he esimae of Fama-French hree-facor model abnormal reurns of socks. The resuls show ha hree indicaors of discreionary accruals are posiively relaed o he cos of equiy capial while he hree real earnings indicaors have he opposie effecs on he cos of equiy. Conrary o prior sudies, rading volume and he cos of equiy show a significan posiive relaionship while he negaive relaionship presened beween individual sock liquidiy risk and he cos of equiy. fer aking he number of op and boom quarile as he indicaors of informaion asymmery and applying earnings qualiy, rading volume and liquidiy risk as ineracion erms, he panel daa regression analysis of long-erm discreionary accruals or real earnings managemen show ha he higher informaion asymmery, he lower company's cos of equiy when earnings qualiy deerioraes. Under higher asymmeric informaion, uninformed raders may fail o respond o he earnings managemen. In erms of rading volume and liquidiy risk, if he liquidiy and liquidiy risk are high, he cos of equiy is low when he degree of informaion asymmery is low. This resul confirms ha managers need o fully disclose informaion o reduce he exernal cos of equiy capial. References boody, D., Hughes, J., & Liu, J. (2005). Earnings qualiy, insider rading, and cos of capial. Journal of ccouning Research, 43(5), 651-673. hp://dx.doi.org/10.1111/j.1475-679x.2005.00185.x charya, V. V., & Pedersen, L. H. (2005). sse pricing wih liquidiy risk. Journal of Financial Economics, 77, 375-410. hp://dx.doi.org/10.1016/j.jfineco.2004.06.007 ig, N., Fong, W. M., Gadhoum, Y., & Lang, L. H. P. (2006). Effecs of large shareholding on informaion asymmery and sock liquidiy. Journal of Banking & Finance, 30, 2875-2892. hp://dx.doi.org/10.1016/j.jbankfin.2005.12.002 Bhaacharya, N., Desa H., & Venkaaraman, K. (2013). Does earnings qualiy affec informaion asymmery? Evidence from rading coss. Conemporary ccouning Research, 30(2), 482-516. hp://dx.doi.org/10.1111/j.1911-3846.2012.01161.x Bhaacharya, N., Ecker, F., Olsson, P. M., & Schipper, K. (2012). Direc and mediaed associaions among earnings qualiy, informaion asymmery, and he cos of equiy. The ccouning Review, 87(2), 449-482. hp://dx.doi.org/10.2308/accr-10200 Boosan, C. (1997). Disclosure level and he cos of equiy capial. The ccouning Review, 72(3), 323-349. Boosan, C., & Plumlee, M. (2002). re-examinaion of disclosure level and he expeced cos of equiy capial. Journal of ccouning Research, 40(1), 21-40. hp://dx.doi.org/10.1111/1475-679x.00037 Chae, J. (2005). Trading volume, informaion asymmery, and iming informaion. The Journal of Finance, 60(1), 413-442. hp://dx.doi.org/10.1111/j.1540-6261.2005.00734.x Chordia, T., Subrahmanyam,., & nshuman, V. R. (2001). Trading aciviy and expeced sock reurns. Journal of Financial Economics, 59(1), 3-32. hp://dx.doi.org/10.1016/s0304-405x(00)00080-5 Davivongs, K., & Pavabur, P. (2012). Pricing of liquidiy risk in emerging markes: Evidence from Greaer China. Inernaional Review of Business Research Papers, 8(1), 20-32. Dechow, P. M., Sloan, P., & Sweeney,. P. (1995). Deecing earnings managemen. The ccouning Review, 70(2), 193-225. Deng, X., & Ong, S. E. (2014). Real earnings managemen, liquidiy and reis SEO dynamics. REUE-SS annual meeing, Philadelphia, P, 3-5 January. Fama, E. F., & French, K. R. (1993). Common risk facors in he reurns on socks and bonds. Journal of Financial Economics, 33, 3-56. hp://dx.doi.org/10.1016/0304-405x(93)90023-5 Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2004). Cos of equiy and earnings aribues. The ccouning Review, 79(4), 967-1010. hp://dx.doi.org/10.2308/accr.2004.79.4.967 208
www.ccsene.org/ibr Inernaional Business Research Vol. 8, No. 4; 2015 Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2005). The marke pricing of accruals qualiy. Journal of ccouning and Economics, 39, 295-327. hp://dx.doi.org/10.1016/j.jacceco.2004.06.003 Gray, P., Koh, P. S., & Tong, Y. H. (2009). ccruals qualiy, informaion risk and cos of capial: Evidence from usralia. Journal of Business Finance & ccouning, 36(1)&(2), 51-72. Kyle,. S. (1985). Coninuous aucions and insider rading. Economerica, 53(6), 1315-1336. hp://dx.doi.org/10.2307/1913210 Lee, K. H. (2011). The world price of liquidiy risk. Journal of Financial Economics, 99(1), 136-161. hp://dx.doi.org/10.1016/j.jfineco.2010.08.003 Pasor, L., & Sambaugh, R. F. (2003). Liquidiy risk and expeced sock reurns. Journal of Poliical Economy, 111, 642-685. hp://dx.doi.org/10.1086/374184 Roychowdhury, S. (2006). Earnings managemen hrough real aciviies manipulaion. Journal of ccouning and Economics, 42, 335-370. hp://dx.doi.org/10.1016/j.jacceco.2006.01.002 Copyrighs Copyrigh for his aricle is reained by he auhor(s), wih firs publicaion righs graned o he journal. This is an open-access aricle disribued under he erms and condiions of he Creaive Commons ribuion license (hp://creaivecommons.org/licenses/by/3.0/). 209