DYNAMIC TRADE-OFFS IN FINANCIAL PERFORMANCES OF ROMANIAN COMPANIES

Transcription

1 DYNAMIC TRADE-OFFS IN FINANCIAL PERFORMANCES OF ROMANIAN COMPANIES Alexandra HOROBET Academy of Economic Sudies Buchares, Romania Radu LUPU Academy of Economic Sudies Buchares, Romania Sorin DUMITRESCU Academy of Economic Sudies Buchares, Romania Dan Gabriel DUMITRESCU Academy of Economic Sudies Buchares, Romania Iulia TINTEA Academy of Economic Sudies Buchares, Romania Absrac Our research objecives are wofold: (1) we use he Du Pon model and marke raios such as PER (price-earnings-raio) and EPS (earnings-per-share) o explain he causes of financial performances of Romanian companies lised on he Buchares Sock Exchange for and o idenify he dynamic rade-off among efficiency, profiabiliy and leverage across companies; (2) we develop a cross-secional analysis of he links beween financial saemens based and marke based indicaors, aiming a deermining wheher fundamenal analysis is a significan source of informaion for equiy marke invesors. We find ha over he period Romanian companies performances flucuaed significanly, maybe wih he excepion of financial leverage and o some exen TAT, regardless of how one measures performance, and ha companies were no capable of using he rade-off beween profiabiliy and efficiency o boos heir ROA and ROE. Companies performance in he sock marke is no explained by heir operaional and financial performance, wih he excepion of PER and EPS. These resuls indicae a decoupling beween companies performance refleced in heir financial repors and

2 86 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA he performance invesors obain from rading hem in he sock marke. They also poin owards he lile imporance awarded by for companies performance shown in heir financial repors. Keywords: Financial performance, Romania, fundamenal analysis, marke analysis, Du Pon sysem JEL classificaion: G14; G32; L11 1. INTRODUCTION The use of financial raios for analyzing a firm s performance became a sandard pracice and is one of he cenerpieces of finance courses for undergraduae and graduae classes. These raios ypically use he company s financial saemens Saemen of comprehensive income (formerly Profi and loss accoun, according o IAS 1 Presenaion of financial saemens ) and Saemen of financial posiion (formerly Balance shee, according o IAS 1 Presenaion of financial saemens ) and informaion colleced from financial markes in order o idenify he srenghs and he weaknesses associaed o various areas of firm s performance: liquidiy, efficiency, profiabiliy, solvabiliy and performance on he marke. A he same ime, he analysis based on accouning informaion has limiaions, generaed mainly by he differences across companies and counries in erms of accouning procedures and he disoring effec of inflaion on hese records. Alhough he analysis of a specific area is of obvious imporance and i ineress a paricular sakeholder ypically, shareholders are more ineresed in leverage and profiabiliy raios, crediors show more ineres owards liquidiy and leverage raios he company s managemen should show ineres owards all areas of he business, in an inegraed view. This means he analysis of one paricular business area is no sufficien for properly and fully undersand he business in a saic and dynamic approach. One of he bes known inegraed frameworks of he business s performance analysis is he Du Pon model, characerized by simpliciy and an abiliy o precisely poin ou he weak and srong poins of a business, boh dynamically and across indusries. The Du Pon model explains firs he reurn on asses (ROA), by linking i o he company s efficiency and profiabiliy, as follows: ROA Ne profi Toal Ne profi asses Sales Sales Toal Ne asses profi margin Toal asse urnover (1) The firs componen of ROA, ne profi margin (PM), offers insigh ino company s profiabiliy and abiliy o conrol coss, while he second componen, oal asse urnover (TAT), assesses he firm s efficiency in using is asses in order o generae sales. This decomposiion of ROA shows ha as PM increases profiabiliy increases and TAT increases efficiency of using asses increases - he reurn generaed by he use of company s asses increases. Besides his basic inerpreaion of ROA, here is much more insigh one can gain by analyzing he roo causes of he ROA level and changes in ime. The wo componens of ROA TAT and PM are in a rade-off ype of relaion, depending on he indusry, as a specific ROA level may be achieved wih various combinaions of profi margin and asse urnover, wih firms ranging along so-called ROA lines : each line shows he combinaions of he wo, wih high urnover accompanies by low profi margin and inversely. In an economy, one can find companies aligned along a differen ROA level, wih differen efficiency and profiabiliy levels. Also, a company may remain on he same ROA

3 Dynamic Trade-Offs in Financial Performances of Romanian Companies 87 line, bu wih changing componens he firm migh have been forced o undergo hrough such changes due o he business environmen characerisics -, or move on anoher ROA line, by changing one and/or anoher of ROA componens. The second equaion of he Du Pon model links reurn on asses (ROA) o reurn on equiy (ROE), o explain he causes of he firm residual profiabiliy, as follows: Ne profi Ne profi Toal asses ROE ROA Financial leverage muliplier Equiy Toal asses Equiy (2) This decomposiion of ROE shows ha he residual profiabiliy of he firm, given by he reurns available o common shareholders, afer paying for all company s obligaions, as compared o he amoun hey invesed in he company, depends boh on he reurn he company generaes from using is asses and he degree of leverage used. Basically, a higher leverage more deb in he company s capial srucure will magnify he profiabiliy for shareholders, boh in a posiive or negaive manner. The decomposiion of ROE may be linked o he business environmen and consrains firms face bu i also includes he impac of financial leverage on a company s overall performance. Besides he informaion on company s performance based on accouning daa, marke based measures are criical for undersanding he relaionship beween decisions a company level and invesors assessmen of hese decisions. A company s performance on he marke may be observed wih he aid of various measures, saring wih sock reurn, risk indicaors (such as sandard deviaion of reurns or bea) and moving owards more sophisicaed measures ha link reurns o risks. In he end, a company s performance on he marke is of mos ineres for invesors and, from his perspecive, one can reasonably wonder wheher he informaion provided by companies financial saemens are acknowledged by marke invesors and refleced in marke performance. Our paper aims a invesigaing he dynamic rade-offs in Romanian companies performances from boh a financial and marke perspecive over a ime span ranging beween 2002 and The paper is srucured as follows: Secion 2 offers a brief lieraure review, Secion 3 describes our daa and research mehodology, Secion 4 presens he main resuls and Secion 5 concludes. 2. LITERATURE REVIEW The differences in he profi margin asse urnover mix are no accidenal, bu he resuls of specific condiions ha firms face in heir business environmens. Selling and Sickney (1989) discuss his mix in erms of wo consrains depending on he indusry ha operae in: capaciy consrains and compeiive consrains. On one hand, we have firms operaing under heavy fixed capaciy coss, which canno use heir asse urnover for boosing ROA. Consequenly, hese firms are forced o arac enough capial o finance heir large producion capaciies or, in order o draw invesors owards hem, mus achieve higher profi margins, ypically hrough high enry barriers and/or economies of scale. On he oher hand, here are firms ha operae in inensely compeiive indusries, whose businesses are subjecive o a compeiive consrain. These firms are faced wih an upper limi on he profi margin hey can obain, so he only variable hey can work on is he asse urnover. A he same ime, here are firms ha do no face eiher compeiive or capaciy consrains ha enjoy far more flexibiliy as compared o he ohers for wha concerns changes among he ROA componens, on one hand, bu also changes in ROA hrough ime. The flexibiliy ha

4 88 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA a firm has in rading-off asse urnover for profi margin can be also hough in erms of a marginal rae of subsiuion beween efficiency and profiabiliy. For he firms operaing in he ails of he ROA curves, he marginal raes of subsiuion are poor, while for he firms operaing in he area denoed by he cenre of he ROA curve he abiliy o design business sraegies ha emphasize boh efficiency and profiabiliy or one of hem, according o marke condiions, is higher. For wha concerns ROA s behavior in ime Selling and Sickney (1989) advance wo explanaions for i: one is given by he operaing leverage of he firm and he second by he phase of he produc-life cycle of he company s producs. Operaing leverage is ypically defined as he change in earnings before ineres and axes deermined by changes in sales, which depends on he proporion beween fixed coss and variable coss. A higher fixed coss proporion causes a higher variabiliy of ROA, which can have as effec movemens of he firm from one ROA line o anoher. On he oher hand, ROA flucuaes in ime, depending on he produc life cycle: during he inroducion and early growh phases, ROA ends o be negaive, as low sales levels accompanied by high coss linked o produc developmen and markeing effors lead o negaive profi margins. As sales increase in he growh phase, ROA becomes posiive and i increases significanly during he mauriy phase, only o deeriorae during he decline phase and even become negaive. This ROA dependency on he produc-life cycle is suppored empirically by oher sudies, such as Hambrick e al. (1982) or Zeihaml and Fry (1984). In a more recen paper, Tezel and McManus (2003), following he work of Firer (1999), disaggregae he Du Pon ROE equaion as o beer disinguish he impac of boh operaing and financial leverage on i. For a sample of 1,052 US-based non-financial publicly raded firms analyzed for hree consecuive years 1997, 1998 and , he resuls indicae ha smaller firms, ypically riskier, have higher reurn on asses as compared o large firms average reurn on non-operaing asses for 1999 was 13.16% for small firms and 10.01% for large firms -, bu small firms have only slighly higher reurns on equiy compared o large firms for he same year, he average ROE of small firms was 15.24% as compared o 14.01% for large firms. Disaggregaing ROE and ROA, Tezel and McManus find ha small firms have a differenial reurn of 4.30%, lower han he differenial reurn for large firms (5.73%). The reason of his smaller differenial reurn resides, according o he auhors, in a higher financial leverage and operaing leverage of large firms compared o small firms. One imporan issue ackled by research in he field refers o he forecasing abiliy of hisorical values of hese raios. In his respec, i is noeworhy he sudy of Penman (1991) ha reaches he conclusion ha in he shor-erm, given by an approximaion of five years, he curren levels of ROE end o persis in he fuure, bu over he long run, ROE displays a mean revering behavior owards an average economy-wide ROE. Similar resuls were offered by Nissim and Penman (1999) for reurn on ne operaing asses (RNOA), asse urnover and profi margins. On he oher hand, Fairfield and Yohn (2001), using daa for a sample of 14,527 US firms for he period, provide evidence ha disaggregaing reurn on asses ino asse urnover and profi margin does no provide incremenal informaion for forecasing he change in reurn on asses one year ahead, bu ha disaggregaing he change in reurn on asses ino he change in asse urnover and he change in profi margin migh be useful for forecasing he change in reurn on asses one year ahead. The link beween financial indicaors and marke performance has been exensively sudied, paricularly since he adven of CAPM, which assers ha invesors price only sys-

5 Dynamic Trade-Offs in Financial Performances of Romanian Companies 89 emaic risk, measured by bea, and predics a posiive relaionship beween bea and sock reurns (Sharpe, 1964). Empirical sudies conduced on US companies or elsewhere failed o provide a srong link beween bea and sock reurns and idenified fundamenal analysis informaion as offering relevan explanaory facors for sock reurns. Basu (1977) finds ha socks wih low PERs have higher reurns han socks wih high PERs, even afer aking ino accoun he impac of bea, while Bandari (1988) idenified a posiive link beween deb-oequiy raios and sock reurns, in ess ha also included bea and firm size. Maybe one of he bes known ess of he relaionship beween sock reurns and company fundamenals is found in Fama and French (1992) ha invesigaed he relaion beween bea, book-omarke, earnings-o-price raio, financial leverage and company size (measured as marke value of equiy). Their resuls show ha book-o-marke and size capure he explanaory power of he oher facors excep for bea. Laer, Fama and French (1993, 1995) showed ha book-o-marke and marke value of equiy are good proxies for sock reurns sensiiviy o risk facors and ha hese measures are relaed o company s earnings. Moreover, Barbee e al. (1996) idenify sales-o-price raio as a reliable indicaor of firms relaive marke valuaion compared o he radiional price-earnings or book-o-marke indicaors used in previous sudies, due o he fac ha he use of differen accouning mehods for depreciaion and invenory affec companies earnings and book values, bu do no influence sales value. They also find ha sales-o-price raio are able o absorb he role of book-o-marke, marke value and deb raio in explaining U.S. sock reurns during he period. Ouside US, Chan e al. (1991) showed ha sock reurns in Japan are posiively linked o book-o-marke and cash flow yield, while Capaul e al. (1993) evidence ha value socks (wih high booko-marke values), earn higher reurns han growh socks (wih low book-o-marke values) in France, Japan, Swizerland, Japan, Unied Kingdom and Unied Saes. In heir invesigaion of he Korean marke, Mukherji e al. (1997) show ha annual sock reurns during were posiively relaed o firm size, bu no significanly relaed o he earningso-price raio or bea. Also, hey find ha book-o-marke and sales-price raios are more efficien indicaors of value ha he earnings-o-price raio, while bea is a less reliable proxy for risk han he deb-equiy raio. More recenly, Figelman (2007) examines he ineracion beween sock reurn momenum and various earnings measures and finds ha large-capializaion companies wih poor pas reurns and high ROE significanly underperform he marke compared o companies wih poor pas reurns and low ROE. Also, companies wih poor pas reurns and poor earnings qualiy significanly underperform he marke. Based on hese resuls, he auhor concludes ha he marke may no fully recognize companies manipulaion of earnings. Overall, empirical evidences suppor o some exen he link beween marke performance and company fundamenals, bu he conclusions are by no means definie and depend on he manner marke performance is defined. Tradiionally, sudies have focused on sock reurns and have largely ignored oher measures of performance such as Sharpe raio, Treynor raio or Jensen s alpha in erms of heir link o companies financial performance. Our research aemps o complemen he exising research on he Romanian marke wih an analysis ha includes such performance measures: he nex pars describe he daa and mehodology, our resuls and he main conclusions.

6 90 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA 3. DATA AND RESEARCH METHODOLOGY Our research uses companies lised on he firs and second ier of he Buchares Sock Exchange ha have been coninuously raded on he marke beween 2002 and We included in he sample only he non-financial firms lised on he BSE for hese consecuive eigh years. The reason for he exclusion of financial firms resided on heir much higher financial leverage as compared o he oher firms, which migh have led o misleading resuls. A he same ime, we excluded from our analysis he financial invesmens companies (SIFs), which are a paricular caegory of asses lised and raded on BSE, as hey acually represen invesmen funds socks, wih porfolios formed of a wide range of Romanian companies. Comparing hem wih a ypical non-financial firm would have severely biased he resuls of he research. From he oal number of 35 companies he me our selecion crieria The oal number of firms included in he analysis is 34, wih an approximae oal marke value of equiy of EURO 18.1 billion a he end of 2009 and a share in he BSE s oal marke capializaion ranging from 22.58% in 2009 and 66.02% in For each of hese firms and he eigh years menioned above, we used wo ses of daa in order o invesigae evoluion of heir financial performance and he link beween financial and marke performance. The firs se involves informaion provided in heir income saemens and balance shees ha we used in order o compue he following raios (where designaes he year for which he compuaion was performed): (1) Toal urnover Toal asse urnover (TAT ) Toal asses where oal urnover includes operaing, financial and exraordinary revenues of he firm, while oal asses include curren and ne fixed asses, a heir book values. (2) Ne profi afer ax Profi margin (PM ) Toal urnover where ne profi afer ax is he profi before paying dividends o common shareholders. (3) Ne profi afer ax Reurn on asses ( ROA ) Toal asses ROA was also decomposed ino TAT and PM: ROA TAT PM. (4) Toal asses Financial leverage muliplier (FLM ) Shareholders' equiy where shareholders equiy is compued as he sum of capial provided by shareholders and he accumulaed reained earnings. (5) Ne profi afer ax Reurn on equiy (ROE) Shareholders' equiy ROE was also decomposed ino ROA and FLM: ROE ROA FLM. Besides hese raios we have also used in our analysis wo performance indicaors ha are a mixure beween accouning and marke informaion, as follows: (6) Earnings afer ax Earnings per share (EPS) Shareholders'equiy (7) Marke price of equiy Price earnings raio (PER) EPS PER is one of he mos widely used marke performance indicaors, bu is value should be inerpreed wih precauion. One approach is o consider high values of PER as

7 Dynamic Trade-Offs in Financial Performances of Romanian Companies 91 indicaing good expeced performance from he company, bu one should pay aenion o he fac ha PER is a raion and a high value migh no be necessarily he effec of good expeced performance, bu of a oo small value of EPS. Also, high values of PER are raher common in emerging markes bu in his case we may be faced wih increased values as a resul of a small number of companies available on he marke, and no as an effec on good expeced performance. The oher approach relaed o PER considers is small values as indicaion of good performance, in his case PER being inerpreed as a payback raio i.e. he number of years ha would ake o recover he price paid for company sock from he earnings per share. The second se of daa we used refers o he marke performance of our sample of companies beween 2002 and 2009, based on heir sock marke reurns. We used annualized weekly logarihmic reurns derived from he sock marke prices for he 35 companies, as well as he Buchares Sock Exchange BET Composie Index (BETC), for he period January 2002 December We used weekly reurns in order o diminish he impac of infrequen rading. Daa regarding he daily closing price was colleced from KTD inves, while daa for he BETC index was obained from he Buchares Sock Exchange daabase. The reurns were compued using he closing price for he day wih he minimum average missing observaions, namely he hird day of he week. All prices were denominaed in Romanian currency, RON. Based on weekly reurns we compued sandard deviaion, skewness and kurosis for he eigh years. Also, for all he sample of companies and for each year we compued he following risk-adjused measures of performance: I. R R f Sharpe raio (SR ) σ where R denoes he average annual reurn, R f is he risk free rae and is he average annual sandard deviaion of he weekly reurns. The risk free rae used for compuing he raio is he average ineres rae for deposis wih a one week mauriy, namely he Romanian Inerbank Bid Rae (ROBID), colleced from he Naional Bank s daabase. II. R R f Treynor raio (TreyR ) β where β is he coefficien of sysemaic risk, bea, of each asse. The bea for each year and each company was esimaed by a linear regression of he excess reurn of each company on he excess reurn of he marke index, as follows (3) R R f α β (R m Rf ) ε where R denoes he weekly reurn for each of he 35 companies, R f is he risk free rae, as a weekly ineres rae, while R m is he weekly reurn of he BETC index. The ε erm denoes he residual value of he regression. III. Jensen s index or alpha (Jensen s ), which is anoher risk-adjused performance measure based on he CAPM, being he inercep of he regression esimaed as in (3). This measure offers insigh ino he performance of he company s sock on an yearly basis if posiive, i indicaes a beer performance han he one expeced by he marke and, if negaive, i indicaes a worse performance han he expeced one. Using hese ses of daa we conduced our research in wo main direcions, as follows: 1. We used he Du Pon model and marke raios (PER and EPS) o invesigae he causes of financial performances of Romanian companies lised on he Buchares

8 92 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA Sock Exchange for and o idenify he dynamic rade-off among efficiency, profiabiliy and leverage across companies; 2. We developed a cross-secional analysis of he links beween financial saemens based and marke based indicaors, aiming a deermining wheher fundamenal analysis is a significan source of informaion for equiy marke invesors. The cross-secional analysis was conduced on an yearly basis, based on regressions of he form (4) Y = n + n X n + n, n=1,2,3 where Y is successively represened by EPS, PER, SR, TreyR and Jensen s, and Xs are, successively: ROA; ROE; PM and TAT; PM, TAT and FLM. In (4) n designaes he number of independen variables (1 o 3). We ran 20 cross-secional regressions for each year and 160 for he whole period. Each year we excluded from he regressions he companies where PER was no available, due o he negaive EPS, herefore he sample of companies was homogeneous in he case of each regression a he level of a single year. 4. RESULTS Figure 1 shows he evoluion of he weighed average values of financial repors based indicaors for he enire sample of companies, where he weigh is given by each company s marke capializaion as a proporion of he oal marke capializaion of he sample. We observe he decline in ROA unil 2004 followed by an accenuaed increase in 2005 and anoher increase in 2006 (his is he highes value for all years (11.52%), hen he progressive decline unil 2009, wih a sharp decrease in The main cause of ROA s evoluion was companies profiabiliy, wih he smalles value in 2004 (-7.77%), he peak reached in 2006 (16.58%) and anoher drop unil a value of 6.50% in Ineresingly, profiabiliy wen slighly up in 2009 o 8.89%, which compensaed he effec of TAT drop ha year and overall led o an increase in ROA compared o Romanian companies efficiency was less flucuaing over he years han profiabiliy, wih he highes value recorded in 2002 (0.8572) and he lowes in 2009 (0.5145). In erms of ROE, he evoluion mimics ROA, maybe wih he excepion of when ROE was more oscillaing han ROA, even negaive in Over he years, ROE values ranged beween -6.50% in 2004 and 17.02% in 2006, wih ROA as he main driver (since financial leverage was raher sable, paricularly afer 2003). The evoluion of operaional and financial feaures of Romanian companies was refleced in high volailiy of EPS and PER over he years. I is ineresing o noe ha he indicaors moved closely ogeher beween 2002 and 2008, bu his rend changes in 2009, when EPS declined o is lowes value (1.5%) while PER increased o compared o Oher wo ineresing aspecs o noe are he posiive values of EPS over he years and he raher high values of PER (maybe wih he excepion of 2004, wih an average PER of only 2.71). One should be cauious, hough, o inerpre he high values of PER as indicaed good fuure expeced performance, as i migh be, a leas o some exen, he effec of he small number of companies available for rading in he Buchares Sock Exchange.

9 Dynamic Trade-Offs in Financial Performances of Romanian Companies 93 Noe: lhs ROA, ROE, TAT, EPS; rhs FLM, PM, PER Figure no. 1 Financial indicaors for Romanian companies, Figure 2 presens he weighed average values for companies sock reurns, sandard deviaions, skewness and kurosis, based on series of weekly reurns. I is easily observable he flucuaing evoluion of reurns and sandard deviaions wih a close link beween he wo he increased risk associaed wih negaive reurns in 2008 is also eviden. Skewness is flucuaing in a small range: we may observe years of average posiive values (2002, 2004, 2006, 2007) and years of average negaive values (2003, 2005, 2008 and 2009 wih he lowes value recorded in 2008, of ). Kurosis is highly posiive for all years, indicaing he presence of fa ails in reurns disribuions, bu he peak is recorded in 2003 (7.0988). Noe: lhs - Re, SD; rhs - Skew, Kur Figure no. 2 Weighed average values of reurns, sandard deviaions, skewness and kurosis for Romanian companies, The risk-adjused performance measures for Romanian companies, presened in Figure 3, drives our aenion owards some noeworhy observaions: firs, he average bea of he sample ranges beween in 2004 and in 2002 overall, bea is close o he

10 94 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA marke value of one, which indicaes ha he sample does no have on an aggregaed basis a significanly differen level of risk han he marke; second, he Sharpe raio (SR) is flucuaing over he years, ranging beween in 2008 (no surprise here!) and in 2004, wih anoher negaive value in 2007 ( ). This shows ha overall Romanian companies were no able o offer invesors significanly beer risk-adjused reurns as compared o he risk-free rae and ha in some years an invesmen in risk-free asses would have provided invesors wih superior performance. When we analyze he Treynor raio, which shows he bea risk-ajused reurn, he values are highly flucuaing from one year o he oher, wih no wo consecuive years recording posiive values, which may be inerpreed as an impossibiliy of Romanian companies o consanly offer good risk-adjused performance for invesors. Jensen s alpha is also flucuaing: we observe five years wih negaive alphas, indicaing a worse performance of companies agains he marke expecaions, and hree years wih posiive values, indicaing beer han expeced performances. Noe: lhs - SR, Bea, TreyR; rhs - Jensen s alpha Figure no. 3. Risk-adjused performance measures, Profiabiliy represens he main driver for company reurns on asses over , while efficiency plays a smaller role, and financial leverage does no add much o he company performance as measured by reurn on equiy. Table 1 shows basic descripive saisics of PM, TAT and FLM and Figure 4 plos he sample means (simple arihmeic means) agains sandard deviaions across companies over he years. We observe a more homogeneous disribuion of profiabiliy during , followed by wider disribuions, paricularly during In he case of efficiency he disribuion is wider in 2002 and 2003 hen again in 2008 and Overall, he sandard deviaions for TAT are higher han he ones for PM, which is no uncommon for our sample ha includes companies from diverse indusries, each wih differen efficiency levels.

11 Dynamic Trade-Offs in Financial Performances of Romanian Companies 95 Table no. 1 Descripive saisics for PM, TAT and FLM, Profi margin (PM) Asses urnover (TAT) Financial leverage (FLM) Mean Minimum Maximum Sd.Dev. Mean Minimum Maximum Sd.Dev. Mean Minimum Maximum Sd.Dev Financial leverage displays a more heerogeneous paern over he years: 2002 is a differen year compared o he res of he period, wih an average leverage of 7.46, while aferwards leverage flucuaes beween 1.71 in 2003 and 2.73 in The high value of leverage in 2002 is due o one company only, Mechel Targovise, ha recorded a financial leverage of , correced aferwards. Figure no. 4 Box & whisker plos for PM, TAT and FLM, In erms of he flexibiliy ha Romanian companies may have regarding heir rade-off beween profiabiliy and efficiency o boos ROA, Figure 5 shows he combinaions beween PM and TAT and he correlaions beween he wo indicaors across companies.

12 96 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA Figure no. 5 PM-TAT correlaions, Typically, he correlaion level indicaes he degree of flexibiliy available for companies regarding he use of one of he wo wih he aim of increasing ROA: a negaive correlaion suggess ha companies in he sample operae a he ends of ROA curves, which indicaes less flexibiliy in boosing ROA (or, differenly pu, hey may use only one of he wo drivers of performance, profiabiliy or efficiency, in order o increase ROA level); a close o zero correlaion may signify ha companies dispose of enough flexibiliy o improve ROA by simulaneously adjusing heir profiabiliy and efficiency levels; a posiive correlaion indicaes ha he presupposed link beween efficiency and profiabiliy is no presen across he sample used in he analysis. For our sample of companies we observe negaive correlaions in four of he eigh years we invesigae (2002 o 2004 and 2006), slighly posiive correlaions in 2005 and 2007 and posiive correlaions in 2008 and 2009, alhough none of correlaions values are significan a he 95% level. Overall, Romanian

13 Dynamic Trade-Offs in Financial Performances of Romanian Companies 97 companies have somehow increased heir flexibiliy regarding he use of profiabiliy and efficiency beween 2002 and 2007, bu he las wo years have probably led o divergen behavior of companies in heir aemp o deal wih he crisis effecs. I remains o be seen wheher he relaionship beween PM and TAT will reurn o a negaive one afer he crisis will be over. The las par of our research aims a idenifying significan relaions beween Romanian companies performance measured wih he help of financial repors daa and he one measured in he capial marke. The resuls of he cross-secional regressions conduced on an annual basis of he form specified in (4) above are presened in Table 2. A firs glance we observe ha financial repors based performance has explanaory power for EPS and PER, bu he link beween hem varies from one year o he oher. Ineresingly, he direcion of he link, given by he sign of he saisically significan coefficiens, does no change over he years. EPS is posiively linked o all financial raios, wih he excepion of 2007, a year when none of he independen variables explains EPS disribuion across companies in he sample. For he remaining years, ROA is a significan explanaory variable in 2005, 2008 and 2009; ROE explains EPS in 2003, 2005, 2006 and 2007; PM resuls in being significan for EPS only in 2005 and 2006; TAT is a powerful explanaion for EPS in 2002, 2003, 2004 and 2008, and FLM explains EPS in 2005 and Looking over he sample and he years, we may inerpre hese resuls as normal: companies wih higher profiabiliy and higher efficiency in using asses have higher earnings-per-share han he ohers, and companies have also used he financial leverage as a ool for increasing EPS. In he case of PER, he influence of hese variable is also presen over he years, wih he excepion of 2003, and he signs of he regression coefficien do no change from one year o he oher. PER is negaively linked o ROA (in 2002, 2004, 2005 and 2008), ROE (in 2002, 2004, 2005, 2007 and 2008), PM (in 2004, 2005, 2006, 2007 and 2008), TAT (in 2007 and 2008) and FLM (in 2007 only). Again, hese resuls are no surprising, as PER is inversely relaed o EPS, herefore he higher he values of hese raios he higher he EPS and he lower he PER and he oher way around. The link beween marke-based performance and financial repors informaion is, as one may observe from Table 2, weaker han for EPS and PER. ROA does no seem o represen a driver for companies marke performance, as i has no significan explanaory power in any of he eigh years. ROE, PM and TAT are posiively linked o Sharpe raio or Jensen s alpha bu only in 2006 and 2007, which were economic growh years, while financial leverage negaively influenced marke performance measured hrough Jensen s alpha in Also, companies wih higher efficiency in using heir asses had lower Treynor raios. We may inerpre hese resuls as indicaing a clear decoupling beween companies performance refleced in heir financial repors and he performance invesors obain from rading hem in he sock marke. They may also poin owards he lile imporance awarded by invesors in he Romanian sock exchange for companies performance shown in heir financial repors and, consequenly, may make us conclude ha fundamenal analysis does no seem o be used as a ool for making decisions in he capial marke. Table no. 2 Cross-secional regressions resuls, Inercep ROA ROE PM TAT FLM Adj. R EPS ** EPS * PER * ** 0.123

14 98 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA Inercep ROA ROE PM TAT FLM Adj. R 2 PER * ** EPS * EPS * * EPS * * SR 0.175*** ** Jensen s ** EPS * * EPS * * PER * ** PER * * PER * ** PER ** ** EPS ** EPS * * EPS * ** * PER * * PER * * PER * * PER * * EPS *** EPS * 6.444* * PER *** PER ** *** ** SR *** SR *** Treynor 5.277** * Treynor * PER * * PER * * ** PER * * ** SR * ** SR ** SR ** Jensen s *** 0.053* α Jensen s * 0.008*** α 2008 EPS * EPS * EPS ** EPS ** PER * ** 0.117

15 Dynamic Trade-Offs in Financial Performances of Romanian Companies 99 Inercep ROA ROE PM TAT FLM Adj. R 2 PER * ** PER * ** ** PER ** ** Treynor *** * Treynor ** EPS ** Noe: *, ** and *** denoe significance a 1%, 5% and 10% level, respecively. 5. CONCLUSIONS Our paper invesigaed he dynamic rade-offs in Romanian companies performances from boh a financial and marke perspecive over a ime span ranging beween 2002 and We used a sample of 34 non-financial companies raded on he Buchares Sock Exchange and developed he analysis in wo main direcions: he use of Du Pon model and marke raios (PER and EPS) o observe he causes of financial performances in he sample, and he developmen of a cross-secional analysis on he links beween financial saemens based and marke performance indicaors. We find ha over he period Romanian companies performances flucuaed significanly, maybe wih he excepion of financial leverage and o some exen TAT, regardless of how one measures performance. Companies were no capable of using he rade-off beween profiabiliy and efficiency o boos heir ROA and ROE, wih profiabiliy displaying high volailiy during We also find ha companies performance in he sock marke is no explained by heir operaional and financial performance, wih he excepion of PER and EPS. We may inerpre hese resuls as indicaing a clear decoupling beween companies performance refleced in heir financial repors and he performance invesors obain from rading hem in he sock marke. They may also poin owards he lile imporance awarded by invesors in he Romanian sock exchange for companies performance shown in heir financial repors and, consequenly, may make us conclude ha fundamenal analysis does no seem o be used as a ool for making decisions in he capial marke. Acknowledgmens This paper presens resuls achieved wihin he research projec Modeling he ineracion beween he capial marke and he foreign exchange marke. Implicaions for financial sabiliy in emerging markes, Projec code IDEI_1782, Projec s financer: CNCSIS, PNII/IDEI References [1] Barbeee, W.C., Sandip, M., Gary, A.R., Do Sales-Price and Deb-Equiy Explain Sock Reurns Beer Than Book-Marke and Firm Size?, Financial Analyss Journal, March/April 1996, pp [2] Basu, S., The Invesmen Performance of Common Socks in relaion o Their Price-Earnings Raios: A Tes of he Efficien Marke Hypohesis, Journal of Finance, July 1977, pp [3] Bhandari, L.C., Deb/Equiy Raio and Expeced Common Sock Reurns: Empirical Evidence, Journal of Finance, June 1988, pp [4] Capaul, C., Rowley, I., Sharpe, W.F., Inernaional Value and Growh Sock Reurns, Financial Analyss Journal, January/February 1993, pp

16 100 Alexandra HOROBET, Radu LUPU, Sorin DUMITRESCU, Dan DUMITRESCU, Iulia TINTEA [5] Chan, L.K.C., Hamao, Y., Lakonishok, J., Fundamenals and Sock Reurns in Japan, Journal of Finance, December 1991, pp [6] Fairfield, P.M., Yohn, T.L., Using Asse Turnover and Profi Margin o Forecas Changes in Profiabiliy, Review of Accouning Sudies, December 2001, p [7] Fama, E.F., French, K.R., The Cross-Secion of Expeced Sock Reurns, Journal of Finance, June 1992, pp [8] Fama, E.F., French, K.R., Common Risk Facors in he Reurns on Socks and Bonds, Journal of Financial Economics, February 1993, pp [9] Fama, E.F., French, K.R., Size and Book-o-Marke Facors in Earnings and Reurns, Journal of Finance, March 1995, pp [10] Figleman, I., Ineracion of Sock Reurn Momenum wih Earnings Measures, Financial Analyss Journal, May/June 2007, pp [11] Firer, C., Driving Financial Performance hrough he Du Pon Ideniy: A Sraegic Use of Financial Analysis and Planning, Financial Pracice and Educaion, Spring/Summer 1999, p [12] Hambrick, D.C., Mac Millan, I.C., Day, D.L., Sraegic Aribues and Performance in he BCG Marix: A PIMS-based Analysis of Indusrial Produc Businesses, Academy of Managemen Journal, Sepember 1982, p [13] Mukherji, S., Dha, M.S., Kim, Y.H., A Fundamenal Analysis of Korean Sock Reurns, Financial Analyss Journal, May/June 1997, pp [14] Nissim, D., Penman S.H., Raio Analysis and Equiy Valuaion, Working Paper, hp://ssrn.com/absrac= [15] Penman, S., An Evaluaion of he Accouning Rae of Reurn, Journal of Accouning, Audiing, and Finance, 6, p [16] Sharpe, W.F., Capial Asse Prices: A Theory of Marke Equilibrium under Condiions of Risk, Journal of Finance, July 1964, pp [17] Tezel, A., McManus, G.M., Disaggregaing he Reurn on Equiy: An Expanded Leverage Approach, Journal of Applied Finance, Spring 2003, p [18] Zeihaml, C.P., Fry, L.W., Conexual and Sraegic Differences Among Maure Businesses in Four Dynamic Performance Siuaions, Academy of Managemen Journal, December 1984, p