Conditional Conservatism and its Effect on Earnings Quality and Stock Prices in Indian Capital Market

Transcription

1 European Journal of Business and Managemen Condiional Conservaism and is Effec on Earnings Qualiy and Sock Prices in Indian Capial Marke Gregory D. Lyimo Assisan Lecurer The Insiue of Finance Managemen (IFM) Shaaban Rober Sree P.O.BOX 98 Dar es salaam Tanzania * of corresponding auhor: Absrac This sudy examines condiional conservaism in Indian capial markes and exen o which condiional conservaism affec earnings qualiy and sock prices. Conservaism principle has been radiionally used in financial reporing for long ime i has been dominan accouning principles (Was 200). In recen years he focus of sandard seers such is IASB have been less conservaive accouning sandard. We used secondary daa for companies lised in Bombay sock exchange from 2006 o 202. We employed pooled panel daa regression wih sandard error robus. We find enough evidence o suppor presence of condiional conservaism during 2006 o 202 in Indian capial marke. We also find ha conservaism do no affec he qualiy of repored earnings however i influence sock prices. Keywords: Conservaism earnings qualiy reurns..0 Inroducion In recen years he qualiy of repored earnings has become key concern for analys and invesors as i influence resources allocaion in capial markes. However he qualiy of repored earnings is influenced by sandards used o measure and managers discreion in he process of measuring. This has lead o he need o undersand he process of measuremen of earnings and facors ha induce bias in he process of measuring he repored earnings in order o avoid poor allocaion of invesmen in financial markes. I is becoming imporan for analyss and invesors o undersand facors ha influence he qualiy of repored earnings so as o avoid losses in rading and maximize heir porfolio. Tradiionally financial reporing is based on conservaism principle which requires anicipaing all possible losses and wries hem off in unfavorable period and underesimae gains in favorable period (Penman& Zhang 2002). This principle has been praciced in he pas and i has been influenial one over ime (Was 200). However in recen years sandard seers for example IASB have developed sandard which are less conservaive in naure in an aemp o reduce conservaism in financial reporing. Sudies have documened mixed findings relaing o he effec of conservaism on earnings qualiy and sock prices. On he one hand Chen e al. (2007) findings sugges ha advanage of conservaism is o decrease he level of earnings managemen. Likewise Fan & Zhang (2007) suggess ha conservaism enhance informaion qualiy. On he oher hand prior sudies have documened negaive impac of conservaism on earnings and reurn. Chen e al. 20 findings reveal ha firms ha are more conservaism are associaed wih poor earnings qualiy. Specifically heir findings sugges ha conservaism reduces he persisence of earnings hus making earnings of poor qualiy. Likewise Hui e al. (2009) find ha more conservaive firms release less earnings forecas. Moreover sudies also have associaed level of conservaism in financial reporing and legal origin. For example Ball e al. (2000) affirms ha companies in common law counries recognize imely bad news han good news as compared o code law counries. Likewise Ball e al. 200 find ha Eas Asian counries Hong Kong Thailand Singapore and Malaysia which are common law counries use condiional conservaism in reporing. Furhermore he level of condiional conservaism has been associaed ownership of companies. For example Ball and Shivakumar (2005) find ha condiional conservaism is more in public companies han in privae companies. Also Risberg (2005) reveals ha independen firms recognize bad news imelier as compared o firms wih pyramid group. So far he debae on he conservaism is far from been resolved one side opposing conservaism and oher side supporing conservaism. Also sudies have shown ha legal origin also influences he level of conservaism accouning. Therefore his sudy is moivaed o examine condiional conservaism in India and he effec of condiional conservaism on earnings qualiy and sock price. This sudy is imporan due o hree reasons. Fis India economy is growing faser and is now shifing from indusrialized o high-ech services oriened. The concern now is he he abiliy of financial repor o reflec he economy changes as his change resuls o he increased research and developmen expenses as well as resrucuring charges. The immediaely wrien hem off by some companies has increased conservaism. Therefore conducing his sudy in India will fill gap in lieraure. Second mos of sudies invesigaing he effec of conservaism on earnings qualiy and sock price have done in 98

2 European Journal of Business and Managemen developed economies and assume marke efficiency (Ball e al Penman& Zhang 2002 Ball e al. 200 and Kim & Pevzner 200). Therefore conducing his sudy in emerging marke where he level of efficiency of capial marke is less han developed economies will provide new insighs from emerging marke perspecive. Third Ball e al. ( ) affirms ha companies in common law counries recognizes imely bad news han good news as compared o code law counries. India is a common law counry which offers unique seing on legal origin o sudy he rend of condiional conservaism in India over ime. The res of he paper is organized as follows. Secion wo discusses he lieraure review. Secion hree cover daa and mehodology. Secion four presen empirical findings and finally secion five draw he conclusion. 2.0 Lieraure Review Many sudies have invesigaed he relaionship beween condiional conservaism and earnings qualiy. For example Kim & Pevsner (200) invesigae he relaionship beween condiional conservaism and fuure negaive surprise in USA from 97 o Their findings sugges ha condiional conservaism is posiively associaed wih fuure lower likelihood of bad earnings. Their findings also sugges ha conservaism have informaion benefi o users as i influences sock prices. However Abedini e al. (204) examine he level of conservaism and invesors reacion o cash sock in Tehran sock exchange. Their finding shows ha more conservaive firms have negaive effec on reacion of invesor on cash sock. Moreover Kazemi e al. (20) examine he relaionship beween conservaism and earnings qualiy. Their findings sugges ha here is posiive relaionship beween conservaism and earnings qualiy measured by persisence predicabiliy value relevance and imeliness of earnings. However Chen e al. (20) invesigae he relaionship beween conservaism and earnings and find ha firms ha are more conservaism are associaed wih poor earnings qualiy. Their findings sugges ha bias induce in earnings by conservaism accouning makes earnings less persisence earnings. Consequenly Gigler and Hemmer (200) examine he relaionship beween conservaism and volunary disclosure. Their findings show ha more conservaive firms are associaed wih less volunary disclosure. Also Hui e al. (2009) conduced sudy o examine he relaionship beween conservaism and managemen earnings forecas. Their finding show ha firm ha pracice conservaive accouning principle releases less earnings forecas Sudies have invesigaed condiional conservaism and earnings qualiy measured by audior enure. For example Li (200) sudy shows ha here is negaive relaionship beween conservaism and audior s enure for small firms however he sudy reveal posiive associaion beween conservaism and audiors enure for large firms. Likewise Soliman (204) findings show ha here posiive relaionship beween conservaism and audior enure. Using a sample of 2676 firm year observaion from 975 o 2004 Balachandran & Mohanram (20) invesigae he relaionship beween increased accouning conservaism and value relevance of earnings. Their findings reveal ha here is no relaionship beween increased conservaism and decline in value relevance of earnings. Likewise Nahand & Khanlar (202) examine he associaion beween conservaism and informaion conen of earnings. Their finding show non linear relaionship beween conservaism and informaion conen of earnings. Chen e al. (2007) examine relaionship beween earnings qualiy measured by earnings managemen and conservaism. Their findings show ha conservaive accouning improves earnings qualiy by reducing earnings managemen. Likewise Fan & Zhang (202) examines he relaionship beween conservaism and informaion efficiency. Their finding suggess ha conservaive firms are associaed wih improved informaion qualiy. Using conservaism score (CS) as measure of conservaism Penman & Zhang (2002) findings reveal ha poor earnings are associaed wih more conservaism accouning. Also heir finding reveals ha sock prices do no reac o earnings qualiy of firm ha pracice more conservaive accouning. Likewise Lagor (2008) invesigae he relaionship beween conservaism and resaemen announcemen. The findings show ha condiional conservaism increases following resaemen announcemen. Also he findings reveal ha condiional increased more for firm wih deb and compensaion conracing which are significan. The evidence presened so far show mixed findings; some opposing conservaism and oher supporing conservaive as i enhance beer earnings qualiy and influence sock prices. This mixed finding is wha makes his sudy unique o invesigae conservaism rend in India and is effec on earnings qualiy and sock price in India sock marke..0 Hypohesis Mehodology and Daa.Hypohesis We proposed wo research hypohesis saed in alernaive in invesigaing conservaism accouning in India and is effec o earnings qualiy and sock prices. Since Ball e al. (2000) affirms ha companies in common law counries recognize imely bad news han good news as compared o code law counries. We sae our firs 99

3 European Journal of Business and Managemen hypohesis in alernaive form as given below. H : Firms in India use conservaism accouning in financial reporing. Prior sudies have documened ha conservaism accouning is negaively relaed boh earnings qualiy and sock prices (Penman & Zhang 2002 Paek e al and Abedin e al. 204). We form our second hypohesis saed in alernaive as shown below. H 2 : Conservaism has negaive impac on earnings qualiy and sock price..2mehodology We employed quaniaive research mehods o sudy condiional conservaism rend in India and is effec on boh earnings qualiy and sock prices. Our sudy follow model proposed by Basu 997 o invesigae condiional conservaism in India and is effec on sock price as given in equaion one. 0 i 2i i i + ψ () Where; i Earnings per share before exraordinary for firm i for year Price a he end of year for firm i for year P i ( P P ) ) P i Dummy variable equal o if i is negaive and zero oherwise Error erm ψ Similar wih prior sudies he exen of conservaism is measured by he asymmeric imeline coefficien ( ) (Basu 997 Ismail & Elbolok 20). When >0 and i is saisically significance indicaes ha bad news are recognized more quickly ha good news hus firms uses conservaism accouning. We also follow exended Basu (997) model proposed by Ismail & Elbolok (20) o examine he effec of conservaism on earnings qualiy as indicaed on equaion wo CFO N + ψ (2) Where CFO i Cash flow from operaion for firm i in year NP i Ne profi before exra ordinary aciviies for firm i for year. Sample secion The sudy used secondary daa exraced from Prowess daabase ha is mainained by he Cener for Monioring Indian Economy. Our sample size iniial was comprised of 500 companies lised in BSE 500. However we reduced companies wih missing informaion for compuing key variables hus reduced our sample o 7 companies wih 259 firm year observaions. Our sample period is seven spinning from 2006 o Empirical resuls 4. Descripive saisics Table presen he resuls of descripive saisics for our variables. The mean (median) value of earning per scaled by beginnings sock price is (0.0596). The mean (median) value of reurn is 0.7 (0.050) while he mean (median) value of earnings qualiy measured by cash flow from operaion scaled by ne profi is.454 (.0085). 4.2 Pairwise Correlaion Table 2 repors pairwise correlaion among our variables. Reurns ( i ) is significance negaively relaed o negaive reurn ( i ) a 5% level of significance. We also find here posiive significance correlaion beween reurn ( i ) and a 5% level of significance. Moreover here is significance negaive relaionship beween i and a 5% level of significance. 4.Resuls of Heeroskedasiciy es Muli-co lineariy es and Ouliers We repor our resuls of ess and regression esimaion in his secion. Table repor he resuls of mulicollineariy es. We employed variance inflaion facors o deermine wheher here is problem of mulicollineariy among our independen variables. We obained variance inflaion facor values of and 00

4 European Journal of Business and Managemen which are far away from he cu-off value of 0 (Hair e al 995 Gujara 200). Therefore our independen variables are free from problem of muli-collineariy. Table 4 presens he resuls of heeroskedasiciy es. We employed Breusch-Pagan es and obained chi-square value of 7.4 which is far beyond he chi-square value of.84. We herefore rejeced he null hypoheses of consan variance. To alleviae he problem of heeroskedasiciy we employed pooled OLS wih sandard error robus using STATA saisical package (Thompson 20). We also winsorized all our variables a 99% and % o alleviae he problem of exreme observaions ha migh have influence on our sudy similar wih prior sudies (Francis e al 2004). Ou of 259 firm year observaions 24 observaions in his paper are winsorized. 4.. Pooled Regression Resuls and Discussion This par repors he resuls of regression for whole sample and year basis regression resuls on asymmeric imelines and also we repor he effec of condiional conservaism on earnings qualiy and sock prices. Table 5 repors he resuls of esimaion of condiional conservaism. We employed Basu 997 model and find ha asymmeric imelines coefficien ( p<0.05). This evidence suppors our firs hypohesis ha firms in India uses conservaism accouning. We also find ha earnings are sensiive o bad news a.97 imes ( as compared o sensiiviy o good news. ( 2) Table 6 repor condiional conservaism values of our full sample from 2006 o 202 in year basis. We find ha in each year he asymmeric imeliness coefficien is greaer han zero ( >0) ha indicaes he presence of conservaism yearly in India. In year 2006 he asymmeric imeliness is p <0.05 likewise in 2007 he asymmeric imeliness coefficien is p <0.05. We also find ha he coefficien of asymmeric imeliness is p <0.0 in Moreover in 202 we find he asymmeric imeliness coefficien of p<0.05. However we find insignifican asymmeric imeliness coefficien in Therefore on average our year basis resuls from 2006 o 202 sugges presence of condiional conservaism in India. Also our resuls sugges ha here is negaive relaionship beween sock price and condiional conservaism. As condiional conservaism increases sock price ends o fall. Table 7 presens he resul of effec of condiional conservaism on earnings qualiy. We employed model developed by Ismail & Elbolok 20 o examine he effec of conservaism on earnings. We find ha here is no p>0.). Therefore presence significan relaionship beween conservaism and earnings qualiy ( 4 of conservaism does no influence earnings qualiy in India. The overall findings show prevalence of conservaism in India. The findings also reveal ha he level of sensiiviy of earnings o bad news in India is.97 imes as compared o posiive reurn. However his level of sensiiviy o bad news is low as compared o sensiiviy of 5.69 repored in Egyp capial markes by Ismail & Elbolok (20). We also collecively find ha condiional conservaism do no influence earnings qualiy however i influence sock prices in Indian sock marke. 5.0 Conclusion We examine condiional conservaism in conex of Indian sock marke and exen o which condiional conservaism affec boh sock prices and earnings qualiy. Conservaism principle has been radiionally been used in financial reporing and i has been influenial principle over ime (Was 200). We specifically invesigaed wheher financial markes in India reflec more bad news imely han good news and exen o his affec he qualiy of earnings. Our evidence suggess ha financial markes reflec more bad news han good news ( p<0.05). We find ha earnings are sensiive o bad news a.97 ( as compared o sensiiviy o good news. Our evidence suppor Ball e al findings ha ( 2) firms in common law counries recognize imely bad news han good news as compared o code law counries. We also find ha condiional conservaism do no affec he qualiy of repored earnings however i influence sock prices. Our findings have implicaion o regulaors o develop or adop accouning sandard ha reduces or eliminae conservaism hence improve sock prices. This paper suggess ha fuure research should focus on sudying he effec of condiional conservaism on earnings qualiy. Specifically fuure sudies should examine effecs of condiional conservaism on earning qualiy proxies such accrual qualiy smoohness persisence and predicabiliy. Reference Abedin B. Ranjabar M. & Mozaffar A. (204) Invesing effec of accouning conservaism and earnings on reacion on invesor o cash sock of companies acceped in Tehran sock exchange Inernaional Journal of Academic Research in Accouning Finance and Managemen Science 4 () -9. 0

5 European Journal of Business and Managemen Balachandran s & Mohanram P (20) Is he decline in he value relevance of accouning driven by increased conservaism? Review of Accouning Sudies 6 (2) Ball R. Robin A. and Wu J. S (200) Incenive versus sandards: Properies of accouning income in four Asian counries and implicaion for accepance of IAS Journal of Accouning and Economics 6 (-) Ball R and Shivakumar L. (2005) Earnings qualiy in UK privae firms: Comparaive loss recogniion imeliness Journal of Accouning and Economics 9 () Ball R. Kohar P and Robin A. (2000) The effec of inernaional insiuion facors on properies of accouning earnings Journal of Accouning and Economics vol. 29 () pp.-5. Basu S. (997) The conservaism principle and he asymmeric imeliness of earnings Journal of Accouning and Economics vol. 24 () pp. -7. Chen Q. Hemmer and Zhang. Y. (2007) On he relaion beween conservaism in accouning sandards and incenives for earnings managemen Journal of Accouning Research Vol.45 () pp Chen L. Folson D Paek W & Sam H (20) Accouning conservaism earnings persisence and pricing muliples on earnings Accouning Horizon (doi: 0.208acch-50664) Fan Q. & Zhang Z. 202 Accouning conservaism informaion aggregaion and he qualiy of financial reporing Conemporary Accouning Research vol.29 () pp Francis J Lafond R. Olsson M P & Schipper (2004) Cos of capial and Earnings Aribue The Accouning Review vol.79 No. 4 pp Gigler F and Hemmer T. 200 Conservaism opimal disclosure policy and he imeliness of financial repors The Accouning Review vol. 76 (4) pp Gujara D. (200) Basic economerics New York: Mc-Graw-Hill?Irwin Hair J Anderson R Taham R & Black W (995) Mulivariae daa analysis: New Jersey: Prenice-Hall. Hu K. Masunaga S. and Morse D (2009) The impac of conservaism on managemen earnings forecass Journal of Accouning and Economics vol. 47 () pp Ismail T H & Elbolok R M; (20) Do condiional and uncondiional conservaism impac earnings qualiy and sock price in Egyp? Research journal of Finance and Accouning vol. 2(2) Kazem H Hamma H and Faridvand R (20) Invesigaing he relaionship beween conservaism and earnings aribue World Applied Science Journal vol. 2 (9) Kim B and PevznerM. (200) Condiional conservaism and fuure negaive surprise: Empirical invesigaion Journal of Accouning and Public Policy vol.29 (4) pp.-29. Lagor W.D (2008) Condiional and uncondiional conservaism following reporing failure: An empirical sudy PhD hesis Florida sae universiy assessed hrough ProQues no L D. (200) Does audior enure effec accouning conservaism? Journal of Accouning and Public Policy vol.29 () pp Nahand Y. & Khanlar M. (202) Invesigaion on informaion conen of conservaive and non-conservaive accouning earnings. Inernaional Journal of Finance and Accouning vol. (4) pp Penman. S & Zhang X J. (2002) Accouning conservaism he qualiy of earnings and sock reurns The Accouning Review vol. 77(2) Risberg M. (2005) Corporae governance and Accouning qualiy: Is bad news good news? Working paper Universiy of Goeborg (accessed 25 h March 204). Thompson S B (20) Simple formula for sandard errors ha cluser by boh firm and ime Journal of Financial Economics vol Was R L (200) Conservaism in Accouning par I: Explanaions and implicaions Accouning Horizon vol. 7() Soliman M (204) Auding qualiy characerisics and accouning conservaism: An empirical sudy of he lised companies in Egyp Corporae Ownership & Conrol vol. (2)

6 European Journal of Business and Managemen Table : Descripive saisics Name µ δ ρ50 Kurosis Skewness Min Max N i i CF NP i Noes: Noes: Descripive saisics of he variables included in his sudy. Nnumber of observaion ρ50median µmean δsandard deviaion. Earnings per share before exraordinary i dummy variable equal o if i is negaive and zero oherwise. ( P P ) ) P NP sand for ne profi before exra ordinary aciviies CF sand for cash flow from operaion. Table 2: Pairwise correlaion i P i i i i * Noes: Noes: pair wise correlaion coefficien of he variables. ** depics up o 5% level of significance. * * 0.80* i Table : Muli-collineariy es resuls using Variance inflaion facor Variable VIF VIF i i Mean VIF.8 Noes: Muli-collineariy es using Variance inflaion facor Table 4: Heeroskedasiciy es resuls Breusch-Pagan Cook-Weisberg es for heeroskedasiciy Ho: Consan variance Variables: dr06 re drro6 chi2() 7.4 Prob > chi Noes: This able presens resuls of heeroskedasiciy es using Breusch-Pagan es. 0

7 European Journal of Business and Managemen Table 5: Esimaion condiional conservaism P Independen Variable i + ψ 0 2 i i 0 Coefficien and probabiliy values aached o independen variables: i 2 i 0.08 *** (0.00) ** (0.08) ** (0.00) No of observaion Overall R- square (0.02) Noes: This represens resuls of pooled OLS regression using sandard error (clusered by firm and year). The values presened in parenheses represen he probabiliy values aached o he coefficien. *** ** and * represen levels of significance a % 5% and 0% respecively according o wo ailed hypoheses Table 6: Year wise regression for measuring condiional conservaism P Independen Variable i + ψ Coefficien and probabiliy values aached o independen variables: i 2 i No of observaion R-square *** 0.0** * ** (0.00) (0.004) (0.069) (0.00) *** -0.05** (0.00) (0.002) (0.02) (0.047) *** (0.00) (0.74) (0.29) (0.552) ** E-05*** (7.84) (0.02) () (0.00) *** -0.05* -7.4E (0.00) (0.06) (0.99) (0.79) (6.7) (-2.0) (-0.78) (0.5) ** (0.026) (0.287) -.4E-05 (.0) 0.240** (0.027) Noes: This represens resuls of pooled OLS regression using sandard error (clusered by firm and year). The values presened in parenheses represen he probabiliy values aached o he coefficien. *** ** and * represen levels of significance a % 5% and 0% respecively according o wo ailed hypoheses Table 7: Impac of conservaism on Earnings qualiy P 0 2 4CFO N Independen Variable P i 0 + ψ Coefficien and probabiliy values aached o independen variables: No of observaion i 2 i 4 CFO i N Overall R- square *** (0.00) * (0.082) ** (0.00) (0.29) (0.522) Noes: This represens resuls of pooled OLS regression using sandard error (clusered by firm and year). The values presened in parenheses represen he probabiliy values aached o he coefficien. *** ** and * represen levels of significance a % 5% and 0% respecively according o wo ailed hypoheses. 04