THAI BUSINESS CYCLE FROM MACROECONOMIC MODEL USING BVAR AND MS-BVAR METHODS

Transcription

1 PEOPLE: Inernaional Journal of Social Sciences Disorn Siriphirunphong e al. Special Issue Vol.1 Issue 1 pp THAI BUSINESS CYCLE FROM MACROECONOMIC MODEL USING BVAR AND MS-BVAR METHODS Disorn Siriphirunphong Faculy of Economics Chiang Mai Universiy Chiang Mai Thailand disorn.ds@gmail.ccom Chukia Chaiboonsri Faculy of Economics Chiang Mai Universiy Chiang Mai Thailand chukia1973@yahoo.com Anuphak Saosaovaphak Faculy of Economics Chiang Mai Universiy Chiang Mai Thailand anuphak@gmail.com Absrac This sudy aims o deermine he impac of imporan componens of Thai business cycle during prosperiy and depression phases. The BVAR and MS-BVAR models are used o analyze he relaionship of each variable. The variables consis of populaion GDP inflaion balance of paymens governmen cash balance ineres rae and exchange rae. The daa correlaed in his sudy are secondary daa during 1979 o 2014 obained from various sources including World Bank World Developmen Indicaors and he Global Developmen Finance daabase World Resources Insiues (WRI) and Bank of Thailand (BOT). The resuls of his sudy indicae ha each variable in his model has saisical significan relaionship. From he analysis each variable has differen impac on Thai business cycle during prosperiy and depression phases. Keywords Business cycle Macroeconomic model BVAR MS-BVAR 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 452

2 PEOPLE: Inernaional Journal of Social Sciences 1. Inroducion Nowadays i is an unavoidable fac ha business cycle becomes more and more imporan in he field of economics. Business cycle or economic cycle can be shorly defined as he upward and downward movemens of Gross Domesic Produc (GDP) levels. Business cycle is also referred o he period of expansions and conracions in he level of business flucuaions around is rend of long-erm growh (Madhani 2010). Undersanding he characerisics and saisical properies of business cycles is very essenial in forming he basis for he consrucion and validaion of heoreical business cycle models. In addiion undersanding he cyclical paerns in economic aciviy and heir causes is exremely crucial o he decisions of policymakers. As a consequence he design of macroeconomic sabilizaion policies is a criically imporan policy objecive for boh developed and developing counries. Therefore a deailed undersanding of he business cycle from macroeconomic model and he ineracion beween policies is a very imporan objecive in he field of economics (Male 2010). Despie no being a developed counry he business cycle has a grea imporance o he economy of Thailand. Thai economy oscillaes from ime o ime wih differen frequency inensiy and duraion. Therefore he business cycle of Thailand needs o be examined in order o undersand he characerisics of business cycle. The principal objecives of he sudy were as follows. To examine he characerisics of Thai business cycle during prosperiy and depression phases. To deermine he impac of imporan componens of Thai business cycle during prosperiy and depression phases. In his sudy populaion GDP inflaion balance of paymens governmen cash balance ineres rae and exchange rae are boh dependen and independen variables. As a resul he model is based on he VAR approach because each variable has an equaion explaining is evoluion based on is own lags and he lags of he oher model variables. The daa analysis is operaed by Bayesian Vecor Auo regression (BVAR) and Markov Swiching Bayesian Vecor Auo regression (MS-BVAR) mehods The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 453

3 PEOPLE: Inernaional Journal of Social Sciences 2. Lieraure Review Dong (2007) aemps o find ou wheher efficiency wage consideraions can help beer undersanding of real business cycle heory. Exending he work of Uhlig & Xu (1996) a new model is developed o analyze effor and business cycles. The resuls demonsrae ha variabiliy of effor due o efficiency wage consideraions help in explaining he large cyclical movemens in employmen as well as he low cyclical movemens in real wages. In he model echnology flucuaes bigger han ha in Hansen s (1985) model bu only 1.5 imes bigger. I is concluded ha efficiency wage heories help undersanding real business cycles in some exen. Arellano (2008) develops a small open economy model o invesigae defaul risk and is ineracion wih oupu consumpion and foreign deb. Her sudy uilizes a model where ineres raes respond o oupu flucuaions hrough endogenous ime-varying defaul probabiliies. The model makes a predicion ha defaul incenives and ineres raes are higher in recessions based on he daa se. The model is able o mach muliple feaures of he daa including he volailiy of ineres raes he negaive correlaion beween oupu and ineres raes he negaive correlaion beween he rade balance and oupu and he high volailiy of consumpion relaive o income. Male (2009) aemps o make an imporan conribuion o he lieraure of he business cycle by exending and generalizing he developing counry sylized facs. Male examines boh classical and growh cycles for a sample of hiry-wo developing counries. The resuls show ha here is he persisence of oupu flucuaions in developing counries and he srong posiive relaionship beween he magniude of his persisence and he level of economic developmen. One of significan finding of his sudy is he model underesimaes oupu persisence in low inflaion counries and overesimaes oupu persisence in high inflaion counries. Wongpunya (2009) documens he sylized facs of he business cycle in Thailand and analyzes he abiliy of real business cycle models o capure hese facs. Her sudy adops he basic framework of real business cycle heory o es he abiliy o explain he salien business cycle feaures of he Thai economy. The resuls show ha he volailiy of invesmen and governmen spending are higher han ha of aggregae oupu. Besides consumpion in he model is less volaile han ha of oupu. These variables are procyclical and are persisen in general. On he conrary ne expors are highly volaile and counercyclical. The models indicae 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 454

4 PEOPLE: Inernaional Journal of Social Sciences a high level of persisence and co-movemen wih oupu. One of he mos imporan resuls is consumpion flucuaes more han oupu does. Bhaacharya e al. (2013) sudy he exen o which financial inegraion impacs of erms of rade shocks on business cycle volailiy in emerging economies. Based on his empirical evidence on he poenial role for erms of rade flucuaions in propagaing he business cycle of an emerging economy his sudy uses a small open economy real business cycle model wih exogenous erms of rade shocks o invesigae he effec of financial openness in his relaion. Their resuls indicae ha he naure of openness migh influence heir abiliy o absorb exernal shocks. More imporan hey find ha in he presence of erms of rade shocks as financial openness increases business cycle volailiy decreases. By increasing capial accoun openness emerging economies can borrow and lend in inernaional financial markes which may help absorb shocks o he economy and sabilize he business cycle. 3. Theory 3.1 Macroeconomic Models Macroeconomic model is an analyical ool used o explain he operaion of he economy of any region or counry. Macroeconomic model is parable of he world. As in any oher parable or heory he macroeconomic model is no expeced o be exremely close o realiy. However i is cerain ha macroeconomic model helps economiss in undersanding he consequences of economic policies in he real world (Gimenez & Jacobs 2008). 3.2 Economic Models a he Bank of Thailand (BOT) The Moneary Policy Commiee (MPC) uilizes economic models as ools o communicae wih he public abou heir policy decision. The MPC emphasizes ha using differen ypes of economic models during he process of forecasing and policy analysis is very imporan (BOT 2008c). 3.3 Business Cycle Facs According o he real business cycle heory he economic flucuaions are mainly caused by produciviy shocks. These produciviy shocks are measured by he Solow residual or he 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 455

5 PEOPLE: Inernaional Journal of Social Sciences change in oupu ha canno be inerpreed by labor and capial changes (Evans 1992). In he real business cycle heory he marke is always cleared by adjusing prices. When firms canno sell heir producs a curren prices hey have o choose eiher reducing heir rae of oupu or decreasing heir prices. Whaever he choice he firms choose i will lead o more quaniy of demand and hus clear he marke (Lucas ). 3.4 The Neoclassical Growh Model (The Solow-Swan Model) Solow (1956) and Swan (1964) assume consan reurn o scale and diminishing reurns o each inpu. The producion funcion is consised of hree inpus which are capial labor and labor augmening echnology. Anoher assumpion is ha a cerain fracion of oupu is invesed in a capial nex period. Saving rae is an exogenous variable in his model. The model indicaes ha he economy will evenually converge o a seady sae no maer where he saring poin of capial is. In he seady sae various variables grow a a consan rae as a raio of he rae of echnology which is predeermined. 3.5 The Basic Neoclassical Growh Model The neoclassical aggregae producion funcion akes he following form: WhereY is oupu K is capial N is labor and X is labor augmening echnology. Y F K N X ) (1) ( For he Cobb-Douglas producion funcion i is se in he form ofy labor share. The physical capial moves along he firs difference equaion in K : 1 K ( N X ) where s K 1 (1 ) K I (2) Wheres he depreciaion rae of capial and I is invesmen. Moreover he economy can only consume and inves from wha i has produced so he addiional equaion is: Y C I (3) 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 456

6 PEOPLE: Inernaional Journal of Social Sciences 3.6 The Opimal Neoclassical Growh Model Based on he Cass (1965) and Koopmans (1965) neoclassical growh model he represenaive household aims o maximize is uiliy subjec o budge consrain: Subjec o K 0 u( C ) 0 Max 1 (1 ) K F( K N X ) C (4) Where s he ime discoun facor and (01). The higher he value of he less willingness for he households o delay heir consumpion is. 3.7 The Real Business Cycle Model Finn Kydland and Edward Presco (1982) use dynamic general equilibrium models in order o explain he business cycle. Their dynamic models overcome he weakness of he saic compuaional general equilibrium ha does no ake capial accumulaion ino accoun. Kydland and Presco use he echnology shock o drive a process of he aggregae flucuaion in he closed economy. 3.8 The Original Real Business Cycle Model Kydland and Presco (1982) creae he iniial se of dynamic general equilibrium models of he 1980s by indicaing ha produciviy shocks are a source of economic flucuaions. Kydland and Presco explain how a well-funcioning marke economy responds o produciviy shocks and displays imporan properies of business cycles in pracical mean. They make an aemp o produce a modern business cycle model which is driven by an aggregae produciviy shock. 3.9 The Basic Real Business Cycle Model The real business cycle model is exended from he neoclassical growh model of he form developed by Solow (1956); Cass (1956); Koopmans (1965). The aggregae producion funcion is he core elemen beween he business cycle and he growh heory. The real business cycle heory sees exogenous sochasic flucuaion in produciviy as he cause of economic variaion. A real business cycle model is exended from he growh model by adding wo 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 457

7 PEOPLE: Inernaional Journal of Social Sciences componens which are variaion in labor inpu and random shocks in produciviy. The model is designed o capure he main feaures of boh growh heory and business cycle based on he assumpion on echnology endowmen and preferences. 4. Mehodology The BVAR and MS-BVAR models used in he sudy of Thai business cycle is based on he properies of Bayes rule (1718). BP POP A1 A2 A7 Are se o be Thailand s key economic indicaors as GCB IR and GDP INF ER respecively. A1 A2 A7 Are evens in a sample space S ha do no happen ogeher ha is U A 7 i 1 i S and A A when i j i j and B is any even ha is he previous value of economic indicaors as POP i GDP i INF BP GCB IR and ER where i is he lag lengh and i 123 n. The previous value of economic indicaors can be expressed as: POP POP POP 1 2 POP n (5) GDP INF BP GCB IR ER GDP GDP 1 2 GDP n (6) INF INF 1 2 INF n (7) BP BP 1 2 BP n (8) GCB GCB 2 GCB 1 n (9) IR IR 1 2 IR n (10) ER ER 1 2 ER n (11) And B S. The condiional probabiliy is wrien as: P( B A ) P( A ) k k ( Ak B) n (12) P i 1 P( B A ) P( A ) i i 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 458

8 PEOPLE: Inernaional Journal of Social Sciences 5. Daa All daa are secondary daa during 1979 o 2014 obained from various sources including World Bank World Developmen Indicaors and he Global Developmen Finance daabase World Resources Insiues (WRI) and Bank of Thailand (BOT). The examined indicaors consis of populaion GDP inflaion balance of paymens governmen cash balance ineres rae and exchange rae. The daa is from annual observaions for 36 years. The daa is convered ino he form of logarihmic reurn. 6. Resuls Table 1: Resuls of MS-BVAR.lag1 phrase 1 POP GDP INF BP GCB IR ER (Inercep) (sd.err) (.sa) POP.lag (sd.err) (.sa) GDP.lag (sd.err) (.sa) INF.lag (sd.err) (.sa) BP.lag (sd.err) (.sa) GCB.lag (sd.err) (.sa) IR.lag (sd.err) (.sa) ER.lag (sd.err) The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 459

9 PEOPLE: Inernaional Journal of Social Sciences (.sa) From able 1 Thai business cycle during prosperiy phrase: For populaion he consan value is and oher variables are se in a fixed equaion sysem. The POP variable of Thai business cycle a ime has a correlaion wih variables POP GDP BP and GCB a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables INF IR and ER a ime 1 in he opposie direcion wih he magniude of and respecively. For GDP he consan value is and oher variables are se in a fixed equaion sysem. The GDP variable of Thai business cycle a ime has a correlaion wih variables POP GDP BP GCB and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables INF and IR a ime 1 in he opposie direcion wih he magniude of and respecively. For inflaion he consan value is and oher variables are se in a fixed equaion sysem. The INF variable of Thai business cycle a ime has a correlaion wih variables POP GDP INF BP and IR a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables GCB and ER a ime 1 in he opposie direcion wih he magniude of and respecively. For balance of paymens he consan value is and oher variables are se in a fixed equaion sysem. The BP variable of Thai business cycle a ime has a correlaion wih variables POP and BP a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables GDP INF GCB IR and ER a ime 1 in he opposie direcion wih he magniude of and respecively. For governmen cash balance he consan value is and oher variables are se in a fixed equaion sysem. The GCB variable of Thai business cycle a ime has a correlaion wih variables POP GDP INF GCB and IR a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables BP and ER a ime 1 in he opposie direcion wih he magniude of and respecively The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 460

10 PEOPLE: Inernaional Journal of Social Sciences For ineres rae he consan value is and oher variables are se in a fixed equaion sysem. The IR variable of Thai business cycle a ime has a correlaion wih variables POP IR and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables GDP INF BP and GCB a ime 1 in he opposie direcion wih he magniude of and respecively. For exchange rae he consan value is and oher variables are se in a fixed equaion sysem. The ER variable of Thai business cycle a ime has a correlaion wih variables GDP IR and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables POP INF BP and GCB a ime 1 in he opposie direcion wih he magniude of and respecively. Table 2: Resuls of MS-BVAR.lag1 phrase 2 POP GDP INF BP GCB IR ER (Inercep) (sd.err) (.sa) POP.lag (sd.err) (.sa) GDP.lag (sd.err) (.sa) INF.lag (sd.err) (.sa) BP.lag (sd.err) (.sa) GCB.lag (sd.err) (.sa) IR.lag (sd.err) The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 461

11 PEOPLE: Inernaional Journal of Social Sciences (.sa) ER.lag (sd.err) (.sa) From able 2 Thai business cycle during depression phrase: For populaion he consan value is and oher variables are se in a fixed equaion sysem. The POP variable of Thai business cycle a ime has a correlaion wih variables POP INF IR and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables GDP BP and GCB a ime 1 in he opposie direcion wih he magniude of and respecively. For GDP he consan value is and oher variables are se in a fixed equaion sysem. The GDP variable of Thai business cycle a ime has a correlaion wih variables POP GDP and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables INF BP GCB and IR a ime 1 in he opposie direcion wih he magniude of and respecively. For inflaion he consan value is and oher variables are se in a fixed equaion sysem. The INF variable of Thai business cycle a ime has a correlaion wih variables POP GDP INF BP and GCB a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables IR and ER a ime 1 in he opposie direcion wih he magniude of and respecively. For balance of paymens he consan value is and oher variables are se in a fixed equaion sysem. The BP variable of Thai business cycle a ime has a correlaion wih variables BP IR and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables POP GDP INF and GCB a ime 1 in he opposie direcion wih he magniude of and respecively. For governmen cash balance he consan value is and oher variables are se in a fixed equaion sysem. The GCB variable of Thai business cycle a ime has a correlaion 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 462

12 PEOPLE: Inernaional Journal of Social Sciences wih variables POP GCB and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables GDP INF BP and IR a ime 1 in he opposie direcion wih he magniude of and respecively. For ineres rae he consan value is and oher variables are se in a fixed equaion sysem. The IR variable of Thai business cycle a ime has a correlaion wih variables POP GDP BP GCB and IR a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables INF and ER a ime 1 in he opposie direcion wih he magniude of and respecively. For exchange rae he consan value is and oher variables are se in a fixed equaion sysem. The ER variable of Thai business cycle a ime has a correlaion wih variables GDP INF BP GCB and ER a ime 1 in he same direcion wih he magniude of and respecively. I also has a correlaion wih variables POP and IR a ime 1 in he opposie direcion wih he magniude of and respecively. 7. Discussion The sudy of Thai business cycle from macroeconomic model using BVAR and MS- BVAR mehods has found ha during prosperiy phrase populaion GDP inflaion governmen cash balance and exchange rae have posiive correlaion wih GDP and balance of paymens and ineres rae have negaive correlaion wih GDP. This means populaion GDP inflaion governmen cash balance and exchange rae increase he level of GDP and balance of paymens and ineres rae decrease he level of GDP. In addiion during depression phrase GDP inflaion ineres rae and exchange rae have posiive correlaion wih GDP and populaion balance of paymens governmen cash balance have negaive correlaion wih GDP. This means GDP inflaion ineres rae and exchange rae increase he level of GDP and populaion balance of paymens governmen cash balance decrease he level of GDP. For he fuure sudy researcher should ry o sudy oher variables ha have a poenial in affecing he business cycle of Thailand because in his sudy all of sudied variables are merely based on macroeconomic indicaors of he Bank of Thailand which migh no be adequae. The 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 463

13 PEOPLE: Inernaional Journal of Social Sciences researcher should collec more daa in order o ge more accurae resuls. Moreover he business cycle of oher counries in Souh Eas Asia region can be ineresing o sudy as well. 8. Conclusion The sudy of Thai business cycle from macroeconomic model using BVAR and MS- BVAR mehods aims o deermine he impac of imporan componens of Thai business cycle during prosperiy and depression phases. The BVAR and MS-BVAR models are used o analyze he relaionship of each variable. The daa correlaed in his sudy are secondary daa during 1979 o 2014 obained from various sources including World Bank World Developmen Indicaors and he Global Developmen Finance daabase World Resources Insiues (WRI) and Bank of Thailand (BOT). The resuls of his sudy indicae ha during prosperiy phrase populaion GDP inflaion governmen cash balance and exchange rae have posiive correlaion wih GDP and balance of paymens and ineres rae have negaive correlaion wih GDP. In addiion during depression phrase GDP inflaion ineres rae and exchange rae have posiive correlaion wih GDP and populaion balance of paymens governmen cash balance have negaive correlaion wih GDP. To sum up during prosperiy phrase he GDP of Thailand rises by he increase of populaion GDP inflaion rae governmen cash balance and exchange rae and he decrease of balance of paymens and ineres rae. Besides during depression phrase he GDP of Thailand rises by he increase of GDP inflaion rae ineres rae and exchange rae and he decrease of populaion balance of paymens governmen cash balance. Consequenly Thai governmen should aim o increase populaion GDP inflaion rae governmen cash balance and exchange rae during prosperiy phrase and increase GDP inflaion rae ineres rae and exchange rae during depression phrase in order o enrich Thai economy. REFERENCES Arellano C. (2008). Defaul risk and income flucuaions in emerging economies. American Economic Review 98(3): Bank of Thailand (2008a April 30). Thailand s macroeconomic indicaors Rerieved from hp:// The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 464

14 PEOPLE: Inernaional Journal of Social Sciences Bank of Thailand.(2008b April 1). Roles and responsibiliies of he bank of Thailand. Rerieved from hp:// And Responsibiliies Aspx. Bank of Thailand (2008c April 1). Economic models a he bank of Thailand. Rerieved from hp:// Model Aspx Bayes T. (1718) Bayes heorem Scoland UK: Universiy of Edinburgh Bhaacharya R. Panaik I. & Pundi M. (2013).Emerging economy business cycles: Financial inegraion and erms of rade shocks.imf Working Paper 13/119. Cass D. (1965) Opimum growh in an aggregaive model of capial accumulaion Review of Economic Sudies 32: 233 Dong C. (2007) Effor unemploymen and business cycle (Maser disseraion of Economics and managemen Science Humbold-Universiäzu Berlin) Evans C. L. (1992). Produciviy shocks and real business cycle Journal of Moneary Economics 29: Gimenez J. D. & Jacobs G. (2008).Macroeconomics for everyone. Unpublished manuscrip. Hansen G. D. (1985). Indivisible labor and he business cycle Journal of Moneary Economics 16: Koopmans T. C. (1965). On he concep of opimal economic growh: The economeric approach o developmen planning. Chicago IL: Rand-McNally. Kydland F. E. & Presco E. C. (1982).Time o build and aggregae flucuaions. Economeric a 50: Lucas R. E. (1972). Expecaions and he neuraliy of money Journal of Economic Theory 4: Lucas R. E. (1973). Some inernaional evidence on oupu-inflaion radeoffs American Economic Review 63: Madhani P. M. (2010). Rebalancing fixed and variable pay in a sales organizaion: a business cycle perspecive. Compension & Benefis Reviews 42: Male R. L. (2009).Developing counry business cycles: Characerizing he cycle and invesigaing he oupu persisence problem. (Docoral disseraion Universiy of York) 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 465

15 PEOPLE: Inernaional Journal of Social Sciences Solow R. M. (1956). A conribuion o he heory of economic growh The Quarerly Journal of Economics 70: Swan T. (1964). Growh models of golded ages and producion funcions. In K Berril (Eds) Economic Developmen wih Special Reference o Eas Asia London: Macmillan Uhlig H. & Xu Y. (1996).Effor and he cycle: Cyclical implicaion of efficiency wages. Neherlands: Tilburg Universiy. Wongpunya N. (2009). Thai business cycles: Theory and pracice. (Docoral Disseraion Universiy of Bah) 2015 The auhor and GRDS Publishing. All righs reserved. Available Online a: hp://grdspublishing.org/people/people.hml 466