Journal of Economic & Financial Studies. Financial crises and volatility spillovers among emerging European equity markets

Transcription

1 Financial crisis and volailiy spillover Ergun and Mahmuović, JEFS (014), 0(04), Journal of Economic & Financial Sudies, 0(04), Vol. 0, No. 04: Augus (014) Journal of Economic & Financial Sudies Open access available a hp://journalofeconomics.org Financial crises and volailiy spillovers among emerging European equiy markes Ugur Ergun a*, Zehra Mahmuović a a Managemen Deparmen, Faculy of Economics, Inernaional Burch Universiy, BIH * Corresponding auhor s uergun@ibu.edu.ba H I G H L I G H T S: 1. We sudy he effec of curren financial crisis on equiy marke reurns of Balkan ransiion economies.. The sudy employed generalized auoregressive condiional heeroscedasiciy model for daily daa spanning from 006 o Resul suggess ha volailiy ransmission ake place from Bosnian and Croaian sock markes o Serbian sock marke Aricle Hisory Received: Acceped: Available online: Keywords: Balkan ransiion economies; GARCH (1.1); Equiy markes; Volailiy spillover. ABSTRACT Financial crisis no only have saisically bu also economically significan impac on global equiy marke reurns. This sudy analyzes wheher curren financial crisis affec he equiy marke reurns of Balkan ransiion economies and wha is he exen of such impac by employing Generalized Auoregressive Condiional Heeroscedasiciy model is employed on daily daa spans from 006 o 01 for hree ransiion economies which are Bosnia-Herzegovina, Croaia and Serbia. Empirical resul indicaes ha volailiy of Serbian sock price is influenced by he volailiy of Bosnian and Croaian sock prices. There is one way volailiy ransmission from Bosnian and Croaian sock markes o Serbian sock marke. High degree of volailiy is observed in he sock markes during laes financial crises. JEL Classificaion: C3; C58; G11; F36. DOI: hp://dx.doi.org/ /jefs.vi The Auhors. This is an open access aricle under he erms of he Creaive Commons Aribuion License 4.0, which allows use, disribuion and reproducion in any medium, provided he original work is properly cied. 1.0 Inroducion The global economic crisis brings drop in economic aciviy, illiquidiy of real secor, he financing problems of curren income, growh and developmen of companies, depressed consumer spending and aggregae demand, and an enormous increase in he unemploymen rae. There is almos no economic secor which was no affeced by he emerging marke crisis. In oday s global world crisis are easily ransferred from one marke o anoher. The inensiy of is ransmission depends on he economic srengh of counries in which i originally emerged. Since he U.S. is sill economically sronges counry in he world, is impac on he world s economy is larges, and he crisis ha emerged in U.S. in 199. ( Grea depression. ), quickly spread o European and Asian markes as well. As ime progresses, he righ size and dimension of his crisis became visible, and his U.S. crisis acually demonsraed how inerconneced markes, in fac, are. So ha he marke crisis is no only reained as a local even, raher i akes he global sae. A financial crisis can hus be defined as an ongoing issue which brings here cession ino many counries, and causes a downward flow ino major sock indexes. In 1997, Asian financial crisis emerged. As a major shock, i had a srong influence on invesors in emerging markes. No only were here significan losses on invesmens, bu also many invesors had come o believe ha rapid and relaively seady economic growh in he region was he norm and foresaw only mild marke sebacks. The Asian crisis was followed by he Russian crisis in 1998, and several ohers. In he period of global financial crisis known as Grea recession," has ook Journal of Economic and Financial Sudies. Page 63

2 Financial crisis and volailiy spillover Ergun and Mahmuović, JEFS (014), 0(04), place and is considered by many economiss o be he wors financial crisis since he Grea depression in U.S. The financial crisis has caused he emerging and developing economies o replace advanced economies o lead global economic growh. The menioned crisis augh us ha he confidence of financial marke, ones i is shaered, i canno be easily recovered. The laes analysis of he global financial and economic crisis, are including he possibiliy ha he crisis has ouched he boom, and he wors is behind us, bu do no exclude he fac ha he full recovery of he world economy could ake many years. The economy of Bosnia and Herzegovina was in srucural crisis before he economic crisis, due o is lack of planned developmen, incompeen governmen srucures, and poliical inerference in economic decision making. The global crisis has jus pulled back he veils hiding our domesic crisis while opening a large number of oher problems, as in oher counries in ransiion. Despie being one of he world s fases growing counries, Turkey is now facing a number of economic challenges, as a resul, of he global economic downurn. The res of his paper is focusing on he emerging marke crisis which sared in 008 and is sill presen, and how his crisis affecs he equiy markes of developing counries such Bosnia and Herzegovina, and Turkey are. The conribuion of his sudy is wofold. A firs we conduc an analysis covering old Yugoslavian ransiion economies focusing on he Croaia, Serbia and Bosnia-Herzegovina; laer we employ he GARCH (1.1) o examine volailiy spillover and ransmission among sock markes of hose hree counries. A las we analyze he volailiy srucure of he seleced counries sock markes before and afer he laes world financial crises. The remainder of his sudy is srucured as follows: Secion presens a lieraure review. Secion 3 presens he daa and mehodology employed in he empirical analysis. Secion 4 gives he empirical resul, whereas he secion 5 concludes..0 Lieraure review The global economy and deregulaed financial sysem are once again in crisis, and equiy marke is experiencing he lowes level in he pas 70 years. Raing agencies migh add insabiliy o financial markes in emerging economies, is he conclusion drawn from recen research abou emerging marke insabiliy (Kaminsky & Schmukler, 00). Alhough wih he globalizaion effec, in which sock markes are expeced o show uniform responses o shocks, he degree of sock marke reacions o such shocks differs from one region o anoher, depending on he level of inegraion wih he inernaional economy (Nikkinen e al., 008). While financial crises are no a new phenomenon he curren financial crisis differs from many of he previously sudied crises in ha i is boh severe and global. I is well-documened ha inernaional capial markes reac, in erms of reurns and volailiy, quickly and simulaneously o major evens such as he crash of 1987 and he Asian crises in However, he iming and magniude of changes in sock reurns and volailiy differ across markes around he world (Roll, 1988). The cenral message from he findings conduced by Samarakoon (011) is ha emerging markes have large normal sensiiviies o U.S. shocks, and large declines in sock prices in hese markes refleced hese dependencies raher han conagion. The findings underline he imporance of emerging markes as drivers o global asse price developmens in recen years (Sáez, Frazscher, & Thimann, 009). The resuls of IMF sudy conduced in 008 and covering 30 counries indicae ha emerging marke equiy prices are influenced by changes in global financial condiions, such as liquidiy, credi and marke risks, as well as growh and exchange rae expecaions. Therefore, global facors consiue a significan channel for spillovers when he inernaional economic environmen changes. A number of pas financial sudies deal wih financial marke linkages and shock conagion effecs beween maure and emerging equiy markes (e.g. Raanapakorn and Sharma, 00; Bessler and Yang, 003; Chaudhuri and Wu, 003; Wong and Vlaar, 003; Syriopoulos, 004). Summarizing he earlier empirical findings, i can be posulaed ha he medium o long-erm dynamic behavior of he Balkan equiy markes can poenially be affeced by co-movemens and volailiy in leading maure, developed equiy markes. Neverheless, he Balkan equiy markes have been negleced, and empirical research remains hin on his region (Samias e al., 007). Some pas sudies indicae ha emerging sock marke behavior increasingly depends on maure sock marke volailiy swings (e.g. Phylakis and Ravazzolo, 00; Swanson, 003; Yang e al., 006; Syriopoulos, 007). There is general agreemen ha he recen global financial crisis raised he coss and consrains in he financial secor in providing working capial, pre-shipmen expor finance, expor credi insurance and insurance of leers of credi for inernaional rade (Uygur; 010). The global recovery remains fragile and uneven-sluggish in advanced counries, much sronger in emerging and developing counries. The recen crisis has come o Turkey raher lae, bu i hi hard swifly (Ercan, Taymaz & Yeldan; 010). Bosnia and Herzegovina has been facing he difficul challenge of dealing wih he legacies of he crisis and is pas unsusainable policies while no huring he recovery. The mos widely applied models in siuaions where volailiy of reurns is cenral issue are ARCH (Engle, 198) and a generalizaion of Engels s model known as GARCH inroduced by Bollerslev (1986). Journal of Economic and Financial Sudies. Page 64

3 Financial crisis and volailiy spillover Ergun and Mahmuović, JEFS (014), 0(04), Daa and mehodology 3.01 Daa To invesigae he issue and o analyze he relaionship among equiy reurns and emerging financial crisis his sudy uses daily equiy reurns of SASE (Bosnia and Herzegovina), Serbia (BELEX) and Croaia (CROBEX) for he period of This daa has been applied o esimae he volailiy ransmission among he seleced Balkan counries equiy reurns and he impac of financial crises on he volailiy of seleced equiy reurns. 3.0 Generalized auoregressive condiional heeroscedasiciy (GARCH) Wih he aim of conribuing o he developmen of macroeconomics raional expecaions, Professor Rober F. Engle in 198 in an aricle has inroduced and defined he erm ARCH model. The erm ARCH sands for auoregressive (depending on is pas), condiional, and heeroscedasiciy (non-consan variance). By condiional, i is mean ha likelihood is compued based on an esimaed se of priming values for he squared innovaions (ε ) and he variances (σ ). This means ha ARCH model allows he curren condiional variance o a funcion of he pas squared error erms. Because of his conribuion Engle was considered as he pioneer of he financial economerics. The firs applicaion of ARCH was in a sudy of inflaion raes (Engle, 198), and hereafer a variey of parameric and nonparameric specificaions of auoregressive condiional heeroscedasiciy have been proposed. ARCH models seek o esimae he ime-dependen volailiy as a funcion of observed prior volailiy. The original ARCH model modeled he variance of disurbances as a linear funcion of lagged values and provides a framework of analysis and developmen of ime series models of volailiy. The ARCH model allows he variance of he error o vary over ime, in conras o he sandard ime series regression models where he variance is assumed o be consan. The ARCH model can be wrien as: y 0 q i1 i i where ε is a sequence of independen and idenically disribued random variables wih mean zero and variance 1; α0 >0, α i 0, for i >0. From he srucure of he ARCH model, i can be concluded ha if here is a large shock in previous periods, in oher words if here is a large variance in previous periods, i makes he presen period s variance higher. I means ha large shocks end o follow by anoher large shock. This propery of ARCH sands for he volailiy clusering in financial ime series. As he ARCH models are used o model he volailiy, hey are also used o undersand wheher and o wha exen he volailiy is ransmied across markes. The basic ARCH model has been exended in several ways; in 1986, Bollerslev generalized he auoregressive condiional heeroscedasiciy model, and his GARCH is he mos influenial and widely-used exension of ARCH. Measuring he relaionship beween variables a various poins in ime, raher han using a single correlaion coefficien over he enire sample period, provides informaion on he evoluion of he relaionship over ime. The mehodology applied in his paper is GARCH (1.1) modeling, which is widely applied mehod for analyzing daa a differen. The (1.1) in GARCH (1.1) model in Eq. (3.) refers o he presence of a firs-order GARCH erm and a firs-order ARCH erm Where, y x 1 1 i (3.) (3.) Journal of Economic and Financial Sudies. Page 65

4 Financial crisis and volailiy spillover Ergun and Mahmuović, JEFS (014), 0(04), News abou volailiy from he previous period, measured as he lag of he squared residual from he mean equaion (ARCH erm) las periods forecas variance (GARCH erm) 4.0 Empirical resuls In order o evaluae volailiy spillovers among hree Balkan ransiion counries sock marke reurn we form hree variance equaion model as follows; y βx δσ σ (BELEX) ω α ε σ (SASE) ω ε α σ (CROBEX) ω ε 1 α ε β σ β σ 1 β 1 7 β CROBEX β BELEX......Model 0... (3.4) σ β CROBEX β SASE......Model (3.3) 5 1 β SASE β BELEX......Model 03...(3. 5) Table shows he variance equaion esimaion resuls for hree models. In all hree models, sock marke s volailiy is being influenced by heir pas volailiy. Significan ARCH coefficien in he variance equaion implies ha previous day s sock marke reurn informaion abou volailiy is an imporan facor in defining oday s marke reurn volailiy. I means ha all hree-sock marke s reurn volailiy is influenced by is own shocks which are ARCH and GARCH facors. Furhermore, he resul from he firs model shows ha Serbian sock marke (BELEX) volailiy is being influenced by Bosnian and Croaian sock marke reurns. In he second and hird model ouside shocks are no significan. Bosnian sock marke and Croaian sock marke is no being influenced by he ouside shock from wo neighbor counries. Table GARCH (1.1) Esimaion Resul Variables Model 01 Model 0 Model 03 C 4.50E-07 (003)* (015)* ARCH GARCH CROATIA (04)* (0.5095) BIH (0.0338)* (0.4683) SERBIA (0.7751) 0015 (0.563) Noe: Esimaed coefficien wih p-values in he brackes. * indicaes significanly a he 5% significance level. In he second model, Croaian and Serbian sock marke reurn is no significan meaning ha volailiy in CROBEX and BELEX does no ransmi o SASE. In he hird model also BELEX and SASE is no significan meaning ha here is no volailiy ransmission from hose markes o Croaian sock marke. We can conclude ha unlike Croaian and Bosnian sock markes, Serbian sock marke volailiy is influenced by he Croaian and Bosnian sock marke volailiy. Figure 01: Volailiy of BELEX Journal of Economic and Financial Sudies. Page 66

5 Financial crisis and volailiy spillover Ergun and Mahmuović, JEFS (014), 0(04), Residual Acual Fied Figure 0: Volailiy of COBEX Residual Acual Fied Among he hree sock markes in ransiion, CROBEX has he lowes volailiy feaures during he financial crises free period and moderae volailiy in he laes financial crises as shown in Fig.. Afer crises, all hree sock markes enjoy low and sable volailiy. Figure 03: Volailiy of SASE Residual Acual Fied The las large financial crisis began in 007, reached he boom in 008, and is sill presen. Afer applying he ARCH modeling o he equiy reurns daa colleced for he period of in several markes of ineres for he aricle, i can be saed ha alhough due o he globalizaion effec, in which i is expeced ha sock markes Journal of Economic and Financial Sudies. Page 67

6 Financial crisis and volailiy spillover Ergun and Mahmuović, JEFS (014), 0(04), show uniform responses o financial crashes, he degree of sock marke volailiy differs from one region o anoher, depending on he level of inegraion wih he inernaional economy. Observing sock markes of he neighbor counries of Bosnia and Herzegovina, Croaia and Serbia, we can see uneven developmens and differen responses o he emerging financial crisis. 5.0 Conclusion This sudy explores he volailiy spillover among he hree sock markes of Balkan counries in ransiion which are Croaia, Bosnia and Herzegovina and Serbia. Given ha he globalizaion has inegraed financial markes, he hypohesis is ha he emerging financial crisis has affeced all sock markes worldwide o some degree, alhough shock sensiiviy is significanly more obvious in emerging counries. Resul from he GARCH (1.1) model implies ha only Serbian sock marke volailiy is being influenced by he Bosnia and Herzegovina and Croaia. Lack of exernal volailiy ransmission o Bosnian and Croaian sock markes form he oher wo former Yugoslavian counries are observed. Addiionally all hree-sock marke s volailiy is being influenced by is own shocks and informaion channels. Bosnian and Croaian sock markes have higher volailiy feaures compared o Serbian sock marke. All hree sock markes show sable low volailiy afer laes financial crises. Our resuls have implicaions for inernaional porfolio invesors which are looking for diversified markes. References Bessler, D. A., and Yang, J The srucure of inerdependence in inernaional sockmarkes. Journal of Inernaional Money and Finance, :61-87 Bollerslev, T Generalized auoregressive condiional heeroskedasiciy. Journal of Economerics, 31(03): Chaudhuri, K., and Wu, Y Random walk versus breaking rend in sockprices: evidence from emerging markes. Journal of Banking & Finance 7(4): Engle, R. F Auoregressive condiional heeroskedasiciy wih esimaes of he variance of U.K. inflaion. Economerica 50(4): Ercan, H., Taymaz, E., &Yeldan, E., 010. Crisis and Turkey: Impac analysis of crisis response measures. Inernaional Labor Organizaion, Kaminsky, G., & Schmukler, S. L. 00. Emerging markes insabiliy: Do sovereign raings affec counry risk and sock reurns. World Bank Economic Review 16(): Nikkinen, J., Omran, M. M., Sahlsröm, P., &Äijö, J Sock Reurns and Volailiy Following he Sepember 11 Aacks: Evidence from 53 Equiy Markes. In. Review of Financial Analysis 17:7-46. Phylakis, K., and Ravazzolo, F. 00. Measuring financial and economic inegraion wih equiy prices in emerging markes. Journal of Inernaional Money and Finance 1(6): Raanapakorn, O., & Sharma, S.C., 00. Inerrelaionships among regional sock indices. Review of Financial Economics 11(): Roll, R., The inernaional crash of Ocober 1987.Financial Analyss Journal 44(5): Saez, L. C., Frazscher, M., &Thimann, C The ransmission of emerging marke shocks o global equiy markes. Journal of Empirical Finance, 16:-17. Samarakoon, L. P., 011. Sock marke inerdependencies, conagion, and he US financial crisis: The case of emerging and froniermarkes. Inernaional Financial Markes, Insiuions and Money 1(5): Samias, A., Kenourgios, D., &Palalidis, N. 007.Inegraion and behavioral paerns in emerging Balkan sock markes.working paper, EFMA Annual Conference, 5-0. Swanson, P.E., 003. The inerrelaedness of global equiy markes, money markes, and foreign exchange markes. Inernaional Review of Financial Analysis 1(): Syriopoulos, T., 004. Inernaional porfolio diversificaion o cenral European sock markes. Applied Financial Economics, Taylor and Francis Journals 14(17): Syriopoulos, T., 007. Dynamic linkages beween emerging European and developed sock markes: Has he EMU any impac? Inernaional Review of Financial Analysis 16(1): Uygur, E., 010. The global crisis and he Turkish economy. Third World Nework, 131: Wong, A. S. K., and Vlaar, P. J. G., 003.Modelling ime-varying correlaions of financial markes. WO Research Memorandum, no. 739: 16. Yang, J., Hsiao, C., Li, Q., and Wang, Z., 006. The emerging marke crises and sock marke linkages: furher evidence. Journal of Applied Economerics 1(6): Journal of Economic and Financial Sudies. Page 68