Informaton content of Islamc Prvate Debt Announcement: Evdence from Malaysa Sahar Modrzadehbam and Gholamreza Mansourfar Abstract Dfferent types of Islamc debts have been ncreasngly utlzed as preferred means of debt fundng by Malaysan prvate frms n recent years. Ths study examnes the mpact of Islamc debts announcement on prvate frms stock returns. Our sample ncludes forty fve lsted companes on Bursa Malaysa nvolved n ssung of Islamc debts durng 2005 to 2008. The abnormal returns and cumulatve average abnormal returns are calculated and tested usng standard event study methodology. The results show that a sgnfcant, negatve abnormal return occurs one day before announcement date. Ths negatve abnormal return s representng market partcpant s adverse attude toward Islamc prvate debt announcement durng the research perod. Keywords Announcement effect, Event study, Islamc debts, Malaysa, Sukuk I. INTRODUCTION I N the hstory of the busness development n the world, factors that affect on the shareholders wealth have been central ssues n the feld of fnance. Fnancng decsons, whch are to determne how much and what type of debts and equy should be ssued to rase requred capal are among the factors can affect on the frm value and eventually shareholders wealth. Asan fnancal crss n 1997-98 heghtened the need for rsk dversfcaton whn the fnancal system. Over relance on bank loans durng crss perod leadng to lqudy problems drove Malaysan government to support the development of corporate bond markets n order to make greater dversfcaton of cred rsk among economc agents [1]. On the other hand, lqudy problems of the bank made Malaysan frms swch ther fnancng methods to bond offerngs nstead of bank loan or equy fnancng that n turn contrbuted a lot n creatng a prosperous domestc prvate debt secures (PDs) market for rasng capal [1]. Over the past years, attempts to develop the Malaysan corporate debt market have been successful n that the corporate bond market has been postng an average annual growth of 7% snce 2000, reachng RM124.2 bllon as at end-december 2008 [2].Ths growng trend of the market and emergng new types of fnancal nstruments day by day call for constant and dynamc market nvestgatons to provde better nsght of the Sahar Modrzadehbam s wh Graduate School of Management, Unvers Putra Malaysa, Malaysa (phone: +989124261056, fax: +982122022905, emal: sahar_modrzadeh@yahoo.com) Gholamreza Mansourfar s wh Faculty of Economcs and Management, Urma Unversy, Iran (emal: g.mansourfar@urma.ac.r) market for frms, nvestors and government. Ths paper ams to address the valuaton effects of one of the modern fnancal nstrument n ths developng market. A unque aspect of the Malaysan market s the co-exstence of an Islamc secures or sukuk market. Islamc secures are structured to comply wh Sharah prncples, whch prohb the chargng of nterest. Islamc secures appeal to two very dfferent sets of nvestors those nvestng strctly for relgous reasons, and conventonal nvestors seekng lqud, attractvely prced nstruments to nvest n and to dversfy ther portfolos. Malaysa s the poneer of the Islamc capal market. Islamc bond was developed n Malaysa n 1990 and spread out to the world rapdly. The corporate sukuk market n Malaysa has grown enormously n recent years, wh an average annual growth rate of 21% between 2001 and 2008. In fact, the outstandng amounts of sukuk have surpassed the amount of conventonal debt secures n the domestc market. In 2008, the amount of corporate sukuk ssued exceeded RM47 bllon [2]. Indeed, Malaysa stll has the largest sukuk market n the world (countres market share n 2008 s depcted n Fg.1). All n all, everyday much and much attenton pad to ths new fnancal product by frms and nvestors has generated a new seres of concerns for researchers to examne dfferent aspects of ths modern tool. Fg. 1 Sukuk market share n 2008(Source: Ernest and Young Presentaton, Islamc Fnance & Investment World, August 26, 2008) Ths paper crcally dscusses the nformaton content of Islamc prvate debts (IPDs) announcements, and tres to answer the questons of: what knd of nformaton convey to the market after Islamc bond offerngs announcement? Wll ths announcement be perceved favorably or otherwse by nvestors? These questons are answered by studyng changes occurrng n company s stock return around the announcement date durng research perod. II. LITEREATURE REVIEW The frst serous dscussons and analyses of relatonshp between frm value and capal structure n a modern sense emerged wh the Modglan and Mller study [3]. They suggested that the value of the frm and capal structure were 564
two ndependent varables n a perfect capal market wh no corporate ncome taxes and the frm value was determned by s nvestment, not fnancng decsons. However, In 1984, the Peckng order theory proposed by Myers and Majulf [4] mantaned that busnesses adhered to a herarchy of fnancng sources n that they set a hgher prory for nternal fnancng, and then they preferred debt over equy n case of external fnancng. Accordng to ths theory, managers tend to establsh ther capal structure base on the cost of adverse selecton arsng from nformaton asymmetry between better-nformed managers and less-nformed nvestors, whch s much lower for bond than equy. Fndngs by Abhayankar and Dunnng [5] also supported ths theory; ther analyss showed that pure equy fnancng had a relatvely large negatve valuaton effect whle straght debt fnancng had a small non-negatve effect on the frm value. So far, many scholars have argued the effects of dfferent capal structures and fnancng decsons on shareholders wealth [6-9]. Many studes have been exclusvely conducted to examne market partcpant s reacton to bond announcements and s mpact on frm value. Nevertheless, to date there has been ltle agreement on the results. A consderable amount of lerature so far has focused on the group of bonds carryng specal feature of havng both equy and bond components. An early study by Kang and Stulz [10] provded an evdence of a sgnfcantly posve stock prce reacton to the announcement of convertble bonds for the Japanese market. Results of the research by Roon and Veld [11] also showed posve abnormal returns on the announcement date of new ssues of convertble bonds n the Dutch market. In contrast, studes by Arshanapall [12] n the US market and by Cheng et al. [13] n Japanese market dscovered sgnfcant negatve abnormal returns at the announcement tme of the convertble bonds. Ammann et al. [14] measured wealth effects of convertble and exchangeable bonds durng perod of January 1996 to May 2003 for two European markets (Swzerland and Germany) and found sgnfcant negatve abnormal returns on the announcement date for the complete sample whch s consstent wh the results of the study by Dann and Mkkelsonv [15] that had ndcated a sgnfcant negatve stock market reacton to announcement of convertble debt offerng n the US market durng the years 1970 to 1979. Moreover, there are some researches avalable studed announcement effects of purely straght bonds whch carry no specal knd of features. For nstance, Study conducted by Stephen and Davd [16] examned the announcement mpact of hgh-yeld straght debt for a sample consstng of 164 bond offerngs. The result suggested that there was no sgnfcant relatonshp between hgh-yeld straght debt announcement and stock value. Ths fndng contradcted the result of Shawn [17] found a year after, hs result showed that announcement of straght debts were assocated wh sgnfcant negatve abnormal return. The other study by Eckbo [18] examned the valuaton effect of corporate debt offerng on a sample of dfferent types of debts. Ths study reported straght debt offerngs had nonposve prce effects, convertble debt offerngs had sgnfcantly negatve effects and publc utly mortgage bond offerngs had margnally negatve effects, however; ths effect were sgnfcantly negatve when the proceeds were used to fnance the utly s nvestment program. However, there has been ltle dscusson about Islamc bonds. The researches, to date, have tended to focus on other types of debt offerngs. One of the few studes avalable on Islamc bond effects s the nvestgaton by Ashhar [19]. He concentrated on the dfference between conventonal bond and Islamc bond announcements and attempted to show that certan types of debts lead to abnormal returns as prevously had clamed by Mkkelson and Partch [8]. The results demonstrated that there was a posve reacton on the Islamc bond ssues announcement; however, no wealth effect was assocated wh the conventonal bond announcement. III. DATA AND SAMPLE Data on Islamc prvate debt ssues were obtaned from the Secury Commsson Malaysa database for the perod of January 1, 2005, to December 31, 2008. The data of the daly closng stock prces for two years before the announcement date of Islamc bond for all the selected companes and for Kuala Lumpur Compose ndex (KLCI) were collected from DataStream. A fnal sample of ssues satsfes the followng selecton crera: 1) The ssung company s lsted n Bursa Malaysa(KLSE) 2) The ssung company s daly common stock prces for the estmaton perod(two years before announcement) are avalable n DataStream 3) Announcement date s dentfable clearly and accurately 4) Matury of the debt must be 1 year or longer. (Ths condon was mposed to avod contamnatng our results wh those that may arse from the sgnalng mplcatons of short-term debt.) 5) Each company s ncluded n the sample only once; even f, had more than one announcement n the research perod After applyng the above screens, we obtaned a sample of 45 Islamc bonds announcement by prvate frms for the perod of 2005 to 2008. In 2005, the number of announcements accounts for 46.6% of the total sample. The number of sample announcements drops largely after year 2005. From 2006 to 2008, there are only 12, 8 and 4 announcements respectvely. However, the market sze has not declned accordngly. IV. METHODOLOGY We use Brown and Warner s [20] standard event study methodology to examne nformaton content of prvate Islamc debts. Before ths, Uday and Nandkumar [21] also used one of the standard event study methodology to test nformaton content of prvate debt placements n the US market. Standard event study s employed to calculate abnormal returns around each announcement of Islamc bond ssue. Accordng to many prevous studes on valuaton effect of company s dfferent announcement, Islamc bond announcement mght have some fnancal mplcatons for the 565
ssung frm. It s hypotheszed that Islamc bond announcement s assocated wh a posve market reacton. An event study s consdered useful because, wh the assumpton of ratonaly n the market place, the share prces wll mmedately reflect the effects of the event. Therefore, the economc mpact of the event can be measured over a shorter perod compared to drect productvy related measures, whch may take many months or years of observatons [22]. Announcement date s a day n whch the bond offerng s frst made known to the publc. Event wndow s consdered -15 to +15 days around the announcement date (22 workng days). There s no standard for event wndow perod; however, the perod of nterest s often expanded to multple days ncludng at least the day of the announcement and the day after the announcement. Ths captures the prce effects of announcements whch occur after the stock market closes on the announcement day [22]. And estmaton perod n ths study ncludes two years (523 workng days) daly nformaton untl ffteen days before (-15) announcement date.(fg.2 depcts event wndow and estmaton perod tmelne) Estmaton perod Event -2 YEARS -15DAYS 0 +15DAYS Fg 2 Estmaton and event wndow tme lne The abnormal return s defned as the dfference between the actual return and the normal return to measure the performance of stock prces of frms on certan days usng (1): AR = R ( α + β R mt ) where; (1) AR : Abnormal returns for frm at tme perod t R : Actual returns for frm at tme perod t R mt : Returns on market portfolo n perod t α β : The constant average return of stock : Beta estmate of stock α and β are estmated over the estmaton perod usng market model whch relates the return of any gven secury to the return of the market portfolo. The return on the Kuala Lumpur Compose Index (KLCI) s used as a proxy of market returns [23]. They are calculated by runnng regresson of secury returns aganst the market returns. After estmatng the abnormal returns for each frm, the abnormal return for all of the frms on each day of the event wndow are then aggregated and averaged. By averagng across all frms, the effects of other events would be mnmzed. The cross-sectonal average abnormal returns for each tme perod t, are calculated as (2); where N s equal to the number of frms n the sample: AAR t = 1 N N t = 1 AR The AAR t s the average estmated percentage devaton of the returns of the sampled stock from the normal relatonshp to the market. The standard devaton of the AAR of the estmated wndow calculated as well to test and analyze whether the Average Abnormal Returns (AAR) are statstcally dfferent from zero by(3): t statstc = AAR t / δ (AAR) where, (3) AAR : Average abnormal return of perod t t δ : Standard devaton of average abnormal return over the estmaton wndow A t-value at 5% sgnfcant level means that the partcular one-day resdual s sgnfcantly dfferent from zero at 95% confdence level. To observe the cumulatve effects, the cumulatve abnormal returns (CAAR -t1, +t2 ) are computed by summng up the AAR t over varous tme perods of nterest relatve to the event perod (-t 1 to +t 2 ) usng (4): + t2 ( t1, + t2) = AAR t 1 (2) CAAR (4) CAAR t s a more precse representatve of the longer term effect on share prces from bond offerng announcements. The t-value for the CAAR t s gven as (5): t = CAAR / δ ( CAAR ) (5) t The standard devaton of CAAR s defned as (6); where N s the number of days n the CAAR statstc: δ ( CAAR ) = δ ( AAR ) N (6) V. RESULTS Abnormal returns are calculated for 22 days durng the event wndow. A cumulatve average abnormal return s also computed to test cumulatve effect of nformaton for the market reacton. Table 1 demonstrates a summary of average abnormal returns and ther t-tests on day -1,day 0 and day +1 of bond announcement. The average abnormal returns (AAR) are -0.00132 on day 0, -0.0023on day +1 and -0.00724 for day -1. The negatve abnormal return on day before announcement s hghly sgnfcant at 5% level and t-values for the announcement day and day +1 are. Ths early market reacton to Islamc bond announcements was also dscovered prevously by Ashhar [19]; where was found that the early market reacton was posve. Regardless of the reacton knd (to be posve or negatve), possble reason for early response could be the fact that nformaton of Islamc bond offerng often leaks out to the market before the announcement. It s also well known that event study accurately explaned 80-90 percent of nformaton content reflected n prces before the announcement date 566
[24].Sgnfcant negatve abnormal return on the day before announcement shows that the announcement of Islamc bonds n the market has reflected a bad news durng the years of 2005 to 2008 n Malaysan market. Fg 3 shows AAR durng the 22-day event wndow. The market abnormal return has a sharp downward trend over days -3,-2 and -1 and experence negatve values for the perod from day -2 to day +2 whch would be a result of fallng demand for the stock on those days. After day +2, the negatve trend goes up and adjusts gradually n the followng days. TABLE I SUMMARY OF THE AVERAGE ABNORMAL RETURN (AAR T ) ON DAY -1 TO DAY +1 Measure Day -2 Day -1 0 Day+1 Day+2 AAR -0.00365-0.00724-0.00132-0.0023-0.00232 T-statstcs -1.0480797-2.080332-0.3783392-0.6618723-0.6651651 Result sgnfcant Fg. 3 Market reacton n the form of average abnormal return to the Islamc Prvate Debt 567
The cumulatve average abnormal returns (CAAR) for three sub-perods are shown n Table 2. Besdes AAR, CAAR s a useful statstcal analyss whch helps to get a sense of the aggregate effect of the abnormal returns. Partcularly, the CAAR can prove very useful, f the nfluence of the event durng the event wndow s not lmed only to the event date. As llustrated n the table, CAAR for the perods of day-1 to day +1, day -1 to day0 and day0 to day+1 are -0.01087, -0.00856 and -0.00362 respectvely. As s seen n the table, although t-statstcs for the perods (-1,+1) and (-1,0) show values almost sgnfcant, none of the t-values are absolutely sgnfcant at 5% level. Accordng to Fg 4 depctng Graph CAAR on the event wndow, CAAR values are entrely posve before day -1 and after +4. Negatve CAAR values for the days -1 to day +4 represent the negatve nterpretaton of Islamc bond announcement by market partcpants. Moreover, a downward trend startng from day -3 to day +2 of CAAR can be consdered as a sgn of anomaly n the market when there s a general upward trend for the whole perod. TABLE I I SUMMARY OF THE CUMULATIVE AVERAGE ABNORMAL RETURN (CAAR) Measure Day(-1,1) Day(-1,0) Day(0,+1) CAAR -0.01087-0.00856-0.00362 T-test -1.80165-1.73854-0.73554 Result Fg.4 overall market reactons n the form of cumulatve abnormal return to the specal dvdend announcement 568
The result of negatve market reacton to the Islamc Prvate debt of ths study s n contrast wh the fndngs of the study by Ashhar [19] that found a wealth effect on the Islamc bond ssues n Malaysa for the perod 2001 to 2006. Ths contradcton can be explaned by the dfferent tme perods of two studes. Ths study covers the perod of fnancal crss tme startng n 2007 that probably s a reason of nvestor s averseness. The other reason mght explan ths conflct of results would be nvestor s msconcepton about the real nature of the Islamc bonds. They attrbuted ther features to bond and faled to recognze ther smlares to eques features n prmary years. However, after a whle, nvestors became more nformed about the true characterstcs of Islamc bonds and learned about s more common features wh equy than bond. As a result, they started to treat Islamc bonds as eques rather than bonds. Islamc bond carres specfc features that dfferentate from conventonal bonds. Characterstc of Islamc bond ssues have to comply wh Shar ah that approved by Sharah Advsory Councl (SAC) of the Secures Commsson Malaysa. Base on Sharah, Islamc bonds transform blateral rsk-reward sharng between borrowers and lenders nto the market-based refnancng of shar ah-complant lendng or trust-based nvestment and Investors own the underlyng asset va SPV that funds (un)secured payments to nvestors from drect nvestment n real, relgously-sanctoned economc actvy. Islamc bonds do not pay nterest, but generate returns through commodzaton of capal gans. Islamc bond cannot be classfed exclusvely n debt category because also shares some stock features. Thus, consderng the smlar characterstcs of Islamc bonds and equy, we are able to explan negatve market reacton to Islamc bond offerng by relyng on the results of pror studes on the negatve mpact of equy announcement on the stock return. The emprcal evdence suggests that pure equy offers have a relatvely large negatve effect whle ssues of straght debt have a small non-negatve effect on the value of the ssung frm. Offers for sale of convertble secures, whch combne characterstcs of both debt and equy, have negatve wealth effects that le between those observed for pure equy and straght debt [5]. VI. CONCLUSION Ths study set out to determne the nformaton content of Islamc prvate debt offerngs. For ths purpose, we examned stock prce reacton to the announcement of IPD s ssuance durng the perod 2005 through 2008 n Malaysan market for a sample consstng of forty fve lsted companes n Bursa Malaysa. The average abnormal return and cumulatve average abnormal returns around announcement date were examned usng standard event study methodology. The average abnormal return was sgnfcantly negatve on day -1 but not on day 0 and day +1. The day -1 average abnormal return was also the lowest durng the 22 days event wndows. The possble reason s the leakage of nformaton to the market before IPDs announcements. The cumulatve abnormal return for three sub-perods of (-1,+1),(-1,0) and (0,+1) were also analyzed; the results revealed negatve CAAR. Ths fndng s neher consstent wh the results of the prevous study by Ashhar [19] usng the IPD s data durng years 2001 through 2006 n Malaysa nor does support the peckng order theory whch expects debts to create value for a frm. Thus, the result rejects our hypothess of a posve market reacton on Islamc bond announcement. Ths dverse market reacton would be attrbuted to the nvestors ncreased awareness regardng Islamc bonds features havng n common wh equy. Moreover, fnancal crss of 2007 can be consdered as a major contrbutng factor n the nvestor s reluctance. REFERENCES [1] I. Muhammad bn, and W. Adran, The corporate bond market n Malaysa Developng corporate bond markets n Asa. Bank Negara Malaysa, 2006. [2] Bank Negara Malaysa and Secures Commsson Malaysa, Malaysan Debt Secures and Sukuk Market. 2009 [3] F. Modglan, and M.H. Mller, The Cost of Capal, Corporaton Fnance and The Theory of Investment. Amercan Economc Revew, vol. XLVIII(3), pp. 261-297, 1958. [4] S.C. Myers, and N.S. Majluf, Corporate fnancng and nvestment decsons when frms have nformaton that nvestors do not have. Journal of Fnancal Economcs, vol. 13, pp. 187-221, 1984. [5] A. Abhayankar, and A. Dunnng, Wealth effects of convertble bond and convertble preference share ssues: An emprcal analyss of the UK market. Journal of Bankng & Fnance, vol. 23,pp. 1043-1065, 1999. [6] M.T. Bradshaw, S.A. Rchardson, and R.G. Sloan, The relaton between corporate fnancng actves, analysts' forecasts and stock returns. Journal of Accountng and Economcs,vol 42, pp. 53-85, 2006. [7] D. Magenns, E. Watts, and S. Wrght, Convertble notes: the debt versus equy classfcaton problem. Journal of Multnatonal Fnancal Management, vol.8, pp.303-315,1998. [8] W.H. Mkkelson, and M.M. Partch, Valuaton effects of secury offerngs and the ssuance process. Journal of Fnancal Economcs, vol. 15, pp. 31-60, 1986. [9] C.R. Harvey, K.V. Lns, and A.H. Roper, The effect of capal structure when expected agency costs are extreme." Journal of Fnancal Economcs, vol.74, pp. 3-30, 2004 [10] J.K. Kang, and R.M. Stulz, How dfferent s Japanese corporate fnance? an nvestgaton of the nformaton content of new secury ssues. The Revew of Fnancal Studes, vol. 9, pp. 109-139, 1996. [11] De. F. Roon, and C. Veld, Announcement effects of convertble bond loans and warrant-bond loans: An emprcal analyss for the Dutch market. Journal of Bankng & Fnance, vol. 22, pp. 1481-1506, 1998 [12] B. Arshanapall, F. Fabozz, L. N. Swzer, and G. Gosseln, New evdence on the market mpact of Convertble Bond Issues n U.S. Paper presented at annual meetng of the Calforna,Oct 2004. [13] W. Cheng, N. Vsaltanachot, and P. Kesayan, A Stock Market Reacton Followng Convertble Bond Issuance: Evdence from Japan. Internatonal Journal of Busness,10(4), 2005 [14] M. Ammann, M. Fehr, and R. Sez, New evdence on the announcement effect of convertble and exchangeable bonds. Journal of Multnatonal Fnancal Management, vol. 16,pp. 43-63,2006. [15] L.Y. Dann, and W.H. Mkkelson, Convertble debt ssuance, capal structure change and fnancng-related nformaton - Some new evdence. Journal of Fnancal Economcs, vol. 13, pp. 157-186, 1984 [16] P. Stephen, and J. Davd, Stock prce effect of hgh yeld debt ssues. The Revew of Fnancal Studes, vol. 71, pp. 393-410, 1990. [17] D.H. Shawn, E.H. Shelly, and B.H. Steven, The market reacton to straght debt ssues: the effect of free cash flow. The Journal of Fnancal Research, vol. XXI, pp. 219-228, 1998. [18] B.E. Eckbo, Valuaton effects of corporate debt offerngs. Journal of Fnancal Economcs, vol.15, pp.119-151, 1986. 569
[19] Z.M. Ashhar, Conventonal vs Islamc Bond Announcements: The Effects on Shareholders Wealth. Internatonal Journal of Busness and Management, vol 4,pp. 6,2009 [20] S.J Brown, and J.B. Warner, Usng daly stock returns: the case of event studes. Journal of Fnancal Economcs, vol. 14, pp. 3 31, 1985. [21] C. Uday, and N. Nandkumar, The Informaton Content of Prvate Debt Placements. Journal of Busness Fnance & Accountng, vol. 35, pp. 1164-1195, 2008. [22] A.C. MacKnlay, Event Studes n Economcs and Fnance. Journal of Economc Lerature, vol. 35, pp. 13-39, 1997. [23] M.N. Annuar, and M. Shamsher, Capal structure. Capal Market Revew, vol. 2,pp. 171-177, 1993. [24] M.a. Shamsher, and H. Taufq, Polcal Events and Share Prces of Polcally-Connected Frms. Workng Paper, 2009. 570