Boom-Bust Housing Price Dynamic: The Case of Malaysia

Transcription

1 Inernaional Journal of Economics and Financial Issues ISSN: available a hp: Inernaional Journal of Economics and Financial Issues, 2017, 7(4), Boom-Bus Housing Price Dynamic: The Case of Malaysia Yip Chee Yin 1 *, Woo Kok Hoong 2, Oon Kam Hoe 3, Nabihah Bini Aminaddin 4, Nurfadhilah Bini Abu Hasan 5 1 Universii Tunku Abdul Rahman, Malaysia, 2 Universii Tunku Abdul Rahman, Malaysia, 3 Universii Tunku Abdul Rahman, Malaysia, 4 Universii Tunku Abdul Rahman, Malaysia, 5 Universii Tunku Abdul Rahman, Malaysia. * cyyip@uar.edu.my ABSTRACT This paper aims o differeniae housing price bubble from a housing price cycle hrough he invesigaion and analysis of he price volailiy driving componens using graphical analysis, coinegraing regression and mean reversion regression. The findings sugges ha Malaysia is no facing any housing bubble a he poin of ime since here is no sharp upsurge and rapid fall of house prices (HP). The recen upswing in HP followed by a gradual coming down raher reflec a severe price cycle which sared in The peak was observed in 2013 and since hen has reversed ino a coninuous bu gradual fall. The cycle is persising and has no boomed ou ye. Our resuls show ha he main reasons for he price booms are speculaive herd insinc and lax in house loan lending policy before Subsequenly he various ani cooling measures by he Malaysian governmen have helped o conrol price expansion. Keywords: Sabiliy Tes, Coinegraing Regression, Mean Reversion Regression, Bubbles, Cycles JEL Classificaions: R3, N2, G1, J10 1. INTRODUCTION Housing besides playing he primary role as sheler has become a marke commodiy ha by speculaive buying and selling generaes appreciable wealh during he boom period. In recen years, many counries like China and Unied Kingdom have experienced prolonged period of housing boom which was followed by relaively sharp and rapid weakening of housing prices, causing massive public debs and recession wih severe damages o he economy. However, i has also been proven ha he housing marke is one of he main drivers for economic growh and governmens enac policies o suppor housing finance o promoe home ownership and ulimaely economic growh. Neverheless, esablished lieraure (Cerui e al., 2015; ECB, 2003; Collyns and Senhadji, 2002) has revealed ha financial facors such as credi expansion, low ineres raes and excessive liquidiy, encourage furher demand and risk aking ha lead o speculaive buying. When housing prices ge disconneced from he economic fundamenals and price expansion coninues unchecked driven by speculaion or over opimism, he boom will be followed by bus ha will see widespread sharp and rapid weakening of house prices (HP) if no suiable inervenion from policy makers. The resul as demonsraed in pas boom-bus episodes when hey assumed he characerisics of housing bubbles, is rampan morgage loan defaul ha will rigger financial insabiliy and severe damage o he real economy. However, he bubble can be moderaed ino a cycle of gradual hike and fall in prices. As such, policymakers are confroned wih he conflic beween he objecive o promoe a more dynamic housing indusry hrough home ownership and he associaed risks of an overheaed housing marke. Housing marke and he price dynamics are among he cenre of aenion of policymakers. Housing price cycles and housing bubbles are now of grea ineres o academic researchers as well as real esae propery players. Research sudies including Agnello and Schuknech (2011), show ha he severiy or magniude of he bus and hus he coss on he economy, is srongly correlaed wih he duraion and magniude of he boom. Hence, as analyzed by Hessel and Peeers (2011), if he facors pushing he boom are no well moniored and addressed, non-fundamenals paricularly speculaion or excessive expecaion on he uprend of prices, will cause he boom o morph ino a bubble. The mild form of housing bubbles is also known as housing cycles. However, here are many divergen views of how o idenify he mild form of housing bubbles. 132 Inernaional Journal of Economics and Financial Issues Vol 7 Issue

2 Housing price cycles are common phenomena since HP reflecs he ineracion of supply and demand fundamenals agains he backdrop of shor erm naural inelasiciy of he supply marke (affeced by facors such as land and building permis), resuling in he flucuaions of housing prices. If he self-correcion mechanism on changes of housing prices due o fundamenal facors is overwhelmed by exuberan expecaion of buyers resuling in prolonged price expansion, he cycle could deeriorae o boombus naure. If he episode of volaile housing prices is jus a cycle, he implicaions are generally no so severe. However, i is imperaive ha suiable measures need and should be aken o preven he volaile siuaion from falling ino a housing bubble. However in records, bubbles are idenified on hindsigh base on he emporal and spaial paern of he movemens of housing prices and he magniude of he rail of resuling damages. This scenario herefore underpins he imporance of undersanding he process ha deermines he housing prices so ha he influence of non- fundamenal facors including speculaion, on he rapidly rising HP could be beer assessed. However, economiss have divergen views on he mechanics of he formaion of a bubble. To-dae, here is no consensus wih regard o he srucure of a housing bubble, making i difficul o differeniae a HP cycle wih ha of a housing bubble. In a join criique sudy on a research repor of he U.S. housing marke, Mayer and Shiller (2006), boh discussan auhors give differing views; while Shiller is ouspoken abou he possibiliy of a housing bubble in US, co-auhor Mayer, does no believe ha here is a housing bubble in mos markes in he US hough housing prices are high. Apar from his conroversial argumen, Capozza e al. (2002) employ mean reversion regression o show ha he price cycle in US marke is sill sable while Glindro e al. (2011) conduced a panel daa analysis of nine Asia counries and by using modified mean reversion regression, hey showed ha no sign of bubble is observed. Given he imporance of a healhy growing housing marke o he economy and he damaging effecs of a bus, beer insighs ino he dynamics of housing prices would be helpful o policymakers, developers and buyers/invesors alike. This sudy herefore aemps o propose an approach o differeniae and hus idenify he housing price cycle from he housing bubble, applying in he conex of he Malaysian housing marke. In view of he uncerain naure of differeniaing a housing bubble from ha of a cycle, his paper specifically offers an alernaive approach o invesigae wheher bubble or cycle exiss in he Malaysian housing marke. The proposed sraegy includes differen approaches o validae and corroborae he resuls ha sugges he exisence of bubble in he marke. We conduc a more scienific and rigorous way o compue propery s fundamenal value using prediced value derived from HP regression model. We use a graphical analysis coupled wih qualiaive analysis, uilizing he views and ideas as proposed by well-known economiss, Mayer and Quigley (2003) and Case and Shiller (2004). We define he hreshold value of percenage increase in HP for he formaion of bubble using he resuls from his graphical analysis. The concep is primarily based on Mayer (2011) and Glindro e al. (2011). Nex, we uilize he mean reversion regression model proposed by Capozza e al. (2002) o es for he sabiliy of he HP movemen A Brief Review of he Malaysian Housing Marke The sharp rising of HP in Malaysia over he pas 12 years ( ) was accompanied by he rapidly increasing household deb of which a significan porion was made up of housing loans, over he same period. In is 2012 repor, he Cenral Bank of Malaysia (2013) (Bank Negara Malaysia [BNM]) posed ha propery financing is a significan componen of bank loans, aking a oal of 41% of he oal financing by he banking sysem as a end And of his amoun, 27.4% was on residenial propery financing. The same repor also revealed a finding on housing price rend ha since 2010, curren prices were being deermined by previous prices, a rend reflecing speculaive behavior, an early sign of a housing bubble. Alhough housing price expansion has peaked in 2013 moderaed by he various macro-prudenial and fiscal measures implemened by he governmen, he high rae of household deb and home loans in relaion o gross domesic produc (GDP) have persised, sanding a 89.1% as a end-2015 wih housing loans showing an 11% increase. As poined ou by economiss, in such a scenario of a combinaion of high household debs and housing boom, any unexpeced rapid fall in housing prices and decline in household income will cause financial insabiliy and an ensuing severe economic downurn. Moreover, he consrucion secor is one of he significan conribuors o he Malaysian economic growh (4.4% of GDP in BNM repor 2015). A slowdown in he secor will have significan negaive impac on he economy. In addiion, given he slowing naional economy and he many headwinds, boh inernal and exernal, ha he counry is currenly facing, he volaile housing price movemens in he Malaysian marke over he pas decade should be of concern. Housing prices in he Malaysian marke for he period increase by 9.5% per year, comparable o he raes in some OECD counries ha have experienced housing booms before he global financial crisis of which is generally agreed o be ignied by he subprime morgage crisis in U.S.A. Average rise of 8.6% for U.K., 11.3% for U.S.A. and 13.3% for Ireland for he period (Bank of Inernaional Selemen - Propery Price Saisics). In addiion, he sharp increase in housing prices in Malaysia for he period running-up o he 1997 housing bubble bus, was a an annual average of 11%. On he oher hand, he Malaysian economic growh since coming ou from he 1998 financial crisis has range from 3.32% o 9.43% wih he excepion in year 2001 and 2009 when growh dropped o 0.52% and 2.53% (The World Bank - he GlobalEconomy.com). In addiion, oher facors ha fuel housing demand have included significan growh in populaion and heavy rural urban migraion. Afer examining he long run expansion of HP in Malaysia, wo research quesions arises. One, is he recen volaile rend of HP ( ) a housing bubble or jus anoher price cycle in he propery marke? Two, if i is a price cycle, a wha sage is he cycle moving hrough? Is his phase of housing price movemens a housing bubble or a price cycle suppored by fundamenals? Our resuls show ha oher han ciies like Kuala Lumpur and George Town which may be facing an incoming risk of a bubble, Malaysia as a whole, is free of any impending bubble in he immediae fuure. Wha we are experiencing is acually a periodic severe form of HP cycle. However, by providing a new piece of indicaion, i is found ha Malaysia is facing a srong upward Inernaional Journal of Economics and Financial Issues Vol 7 Issue

3 surge on HP especially afer he year Neverheless, he price cycle is sill sable. The res of he paper is organized as follows: Secion 2 reviews seleced conenious housing bubble lieraure. Secion 3 describes he empirical mehodologies. Secion 4 presens he empirical findings and Secion 5 concludes his paper. 2. LITERATURE REVIEW To-dae, here is no consensus on he definiion of housing bubble. For example, i is recognized ha HP in many counries in he OECD have moved in andem wih housing prices in he US during he las years. However, only he supposedly infamous bubble ( ) in he US housing marke has araced considerable aenion and focus. The reason could be ha some economiss have found ha he rampan growh in inernaional housing prices is raional and suppored by fundamenals of supply and demand (Himmelberg e al., 2005; OECD, 2005), while ohers have described hese price rises merely as booms in price cycles. They were careful no o define hem as bubbles (Agnello and Schuknech, 2011; Gallin, 2006). Addiionally influenial housing economiss suggesed differing views abou he exisence of housing bubble in he Unied Saes (Gerardi e al., 2010). In view of his unseled issue abou he exisence of housing bubble, we sar o base our sudy of wheher bubble or cycle on he following definiion by Mayer (2011) and Glindro e al. (2011) and hen suppored by graphical analysis and mean reversion resuls Definiion by Mayer (2011) Housing bubbles represen exreme movemens of HP rising rapidly abou 20%, 30%, or even 40% per year for 2 or 3 years and hen falling jus as rapidly in he following 3 years. This ype of housing bubble happened in Las Vegas, Phoenix and Miami in his decade and in Vancouver, Canada, in he lae 1980s and Japan in he mid-1980s (Mayer, 2011) Definiion by Glindro e al. (2011) Housing bubbles represen exreme movemens of HP rising rapidly abou 20% (Glindro e al., 2011) per year for wo or hree consecuive years and hen falling jus as rapidly in he following 2 or 3 years. Since Glindro e al. (2011) is a sudy of housing price dynamics in nine emergen economies including Malaysia, heir definiion of housing bubble is more appropriae in he Malaysian conex. As such, we se 20% rise in HP as our hreshold value for raising he alarm for inervenion policies and measures o be implemened and o ake effec. Therefore we selec he 20% as he hreshold value o raise he alarm for housing bubble. However, boh definiions are based on he accuraely deermined fundamenal HP. Up-o-dae, here is sill no consensus of wha are he fundamenal variables ha deermine he HP. This sudy aemps o minimize he effec of his uncerainy by using consisency principle approach ha is we invesigae he same phenomenon using a leas more han wo approaches. We review some imporan lieraure abou housing bubbles. Scharfsein and Sein (1990) describe a prominen model wih all raional agens. However, heir model omi considering he fundamenals, and ha raional professional invesmen managers adop a herd behavior o make he invesmen decisions similar o ha made by heir colleagues and by his way hey can proec heir business prospecs based on he fac ha when bubbles burs, hey can share he blame wih he majoriy of managers ha have conribued o inflae he bubble. By his approach, hey have inroduced cerain degree of herd insinc in heir compuaion and analysis. However, heir mehod is unable o explain he remendous volume of ransacions aken place. This sudy uses mean reversion regression o by-pass his problem. Anoher model is presened in Allen and Gale (2000) which explains how a bank-based economy can cause bubbles. In heir model, raional invesors who have only limied liabiliy can sill borrow from banks. So hey have he incenive o coninue invesing even in overpriced asses as long as here is uncerainy abou he duraion of prices remaining above he fundamenal. The raionale behind his behavior is ha in he case of a bursing bubble, he bank has o bear he losses. However, as long as he bubble coninues, he invesor can sill make profis. This sudy agrees wih he imporan role played by banking insiuions. However, we do no agree ha lax in banking regulaions is he sole cause of housing bubble. I is only one of he main causes of housing bubble. This sudy uses qualiaive mehod o analyze his lax in banking regulaion problem. One popular bubble model is o compare observed HP wih fundamenal HP as prediced by using he long-run relaionship beween HP and macroeconomic facors (Abraham and Hendersho, 1996; Kalra e al., 2000; Capozza e al., 2002). However, his approach requires a necessary and sufficien condiion for he presence of a bubble which is defined as he divergence of he acual efficien HP from is fundamenal value and which is raher subjecive. In addiion, correc specificaion of fundamenal variables is crucial o he reliable esimaes of price deviaion from equilibrium since he overvaluaion/bubble par is he residual par ha canno be explained by he lis of fundamenal variables. Glindro e al. (2011) invesigae wha deermines he fundamenal values and shor-erm dynamics of HP in nine Asia-Pacific economies. Ahuja e al. (2010) also use a fundamenal model on asse pricing o invesigae wheher HP are rising oo fas in China. In addiion, Chen e al. (2013) invesigae he exisence of bubbles in he Beijing housing marke from 1998 o 2010 using economic fundamenals. This is he fundamenal model approach o explain formaion of bubbles. However for his fundamenal approach, i is ofen unclear as o how o deermine he fundamenal variables, and o compue he fundamenal value from he fundamenal variables (Yiu e al., 2013). This problem of difficulies in idenifying fundamenal variables moivaes us o conduc deailed analysis of he variables before selecing he so called fundamenal housing variables. To mee his end, we ensure ha he model is robus and consisen, diagnosically accepable, and in line wih convenional wisdom and economic heory. Sudies exclusively on he Malaysian housing marke price movemens are relaively few and among which here is a noable research done by Hussain e al. (2012) who have argued 134 Inernaional Journal of Economics and Financial Issues Vol 7 Issue

4 for he presence of housing bubble in five disrics in Klang Valley, i.e. Ampang, Bau, Kuala Lumpur, Pealing and Seapak during Bubble is measured as he difference beween marke HP and fundamenal HP. However he compuaion of fair (fundamenal) HP and definiion of bubble is no clearly defined in he sudy. Imporanly, he auhors also do no highligh he magniude of bubble or verify is exisence in he sudy. Anoher more recen sudy by Yip e al. (2016) on housing bubble in he Malaysian marke has revealed essenially similar finding as his curren paper. However, Yip e al. (2016) only applied saisical analysis in heir sudy and has no corroboraed heir resuls wih economic and affordabiliy analysis. This presen paper fills up he gaps and pu in deailed economic analysis o back up he saisical findings. 3. EMPIRICAL METHODOLOGIES 3.1. Fundamenal HP Mos bubble models need fundamenal HP for compuaion. Fundamenal HP are aken o be he fied HP which are obained by regressing acual HP on housing deerminans like morgage rae (MTR), morgage credi o GDP, ineres rae and unemploymen. The difference beween acual and fundamenal prices is defined as overvaluaion of prices (OVP) which can be due o normal price adjusmen or speculaive aciviy. Mos sudies consider speculaive aciviy as he bubble componen. However, in his sudy, we classify OVP as he bubble componen as illusraed by equaion 1: P = f +B (1) Where, P, f, B respecively represens HP, fundamenal HP and bubble componen. We furher assume ha bubble componen consiss of adjusmen of prices due o shor run price dynamics and prices due o speculaion. Basically, his assumpion is reasonably valid only if he speculaive sub-componen is very much larger han he shor run dynamic price adjusmen sub-componen. This is because shor run dynamic price adjusmen can adjus he price level back o normal price level and hus canno cause he formaion of bubble. Fundamenal HP can be aken as: The long erm average prices, or he long run equilibrium fied HP. We choose he laer because i is more plausible. We compue his long run equilibrium HP using a hree sep approach. Firsly, we idenify he housing deerminans ha have significan predicive informaion abou HP. We do his idenificaion by going hrough lieraure review, and hen compile a lis of he deerminans from which we selec by meri of significance base on convenional wisdom and sepwise regression. Secondly we run he coinegraion regression and hirdly compue he bubble componen and hen analyze wih he aid of graphs and convenional wisdom. The seleced significan deerminans for his sudy are GDP, MTR, morgage credi o GDP raio (MGD), exchange rae (Japanese Yen/Ringgi) (EXJ) and exchange rae (Hong Kong Dollar/Ringgi) (EXH). Our HP are in fac HP indexes which has he advanage of compensaing he difficuly of observing rens for he houses (Mayer and Shiller, 2006) Coinegraing Regression We deermine wheher each deerminan is an inegraed series of order 1 ha is I(1). If all he series are I(1) we can conduc coinegraing (long run) regression using boh fully modified ordinary leas square (FMOLS) and Canonical Coinegraing Regression (CCR) models. The fied values from he regression are our fundamenal HP Compuaion of Bubble Componen We compue he bubble componen values (B) by subracing he fied HP from he acual HP. I is given by equaion 2 in erm of percenage: B =lnp lnp ˆ p The crierion for likely o be bubble is given by equaion 3: B p >20% (3) 3.4. Housing Bubble Invesigaion Mehodologies Any housing bubble should consis of a price booming period and followed by a rapid fall in prices for cerain lengh of ime. Thus he firs sep we should do is o show ha here is a rapid price booming for cerain period of ime. This we can do by hree differen approaches: One, fixing he hreshold for booming price surge; wo, analyzing he uprend graph of prices using common psychology; hird, esing he sabiliy of he price movemen. The followings are he hree ypes of mehodology Graphical Analysis We analyze price behavior from line graphs by using common logical deducion. In addiion, we also analyze HP-income graph. Besides using normal line graph, we also use % incremen yearon-year line graph for he analysis. Nex by using hisorical daa rend, we define he hreshold value for he formaion of bubble Price Sabiliy Model (PSM) We use PSM o es he sabiliy of he price cycle which can be used as one of he measures o idenify wheher i is a cycle or bubble. If he price cycle is unsable, we can conclude enaively ha i is indeed a bubble. PSM is based on mean reversion regression, a shor run regression. Mean reversion regression is popularized by Capozza e al. (2002) who proposes ha HP changes in he shor run which are governed by reversion o fundamenal values and also by serial correlaion. In anoher words, deviaion from he fundamenal HP are mainly due o economic shocks and ha his produces he shor-erm dynamics of HP. This means reversion process is summed up by he formula given in equaion 4: p = α p + β (p -p ) + γ p * * Where, p is log of (observed) real HP and is he difference operaor. he long run HP fundamenals which is obained by FMOLS or CCR coinegraed regression analysis. In efficien housing markes, prices will adjus insananeously so as o mainain γ = 1 and α = 0. However, since housing is a slowclearing durable asse, i is reasonable o expec ha he curren price changes are parly due o previous changes in own price (2) (4) Inernaional Journal of Economics and Financial Issues Vol 7 Issue

5 levels, by he deviaion from he fundamenal value and parly by conemporaneous adjusmen o changes in fundamenals. For he case when α < 1 and β > 0, he price change is saionary because of he fac ha α < 1. Previous changes in own price level has lile effec on presen changes in price level and ha deviaion of presen price from fundamenal values is minimum, indicaing ha he housing cycles are sable. Tha is here is less danger of housing cycle becomes a bubble. Pu i differenly, price increase is due o fundamenal adjusmen of prices which is classified as price cycle phenomenon. However if α 1 i means ha he daa series is nonsaionary, producing he housing cycles which are explosive. This is confirmed by he condiion β 0, meaning ha curren HP are higher han he fundamenal prices producing housing cycles which are no sable. Housing bubbles are likely o form. 4. EMPIRICAL FINDINGS 4.1. Graphical Analysis Figure 1 shows he house price movemen rend wih hree prominen price cycles: , and For he firs cycle, i sars o rise rapidly from 4% o 25% wihin 2 years and hen drops drasically from 25% o 5% wihin 2 years also. This saisfies he bubble characerisics of Mayer and Glindro. Similar price movemen can be observed from However, for he hird cycle, i sared o spiral up from 2% in 2009 reaching a peak of 12% in 2013 and hen urned around and decreased o abou 6% in This hird cycle does no show sharp price movemen. So his emporal paern does no fi he Mayer or Glindro crierion for bubble. This graphical analysis suggess ha we are no in he mids of a bubble bu raher a prolonged price cycle. One new indicaion seems o emerge from analysis of income and HP graph as shown in Figure 2 and which suppors he above conclusion. Beween 2001 and 2008, income moves parallel wih HP which is he norm since wih more income, more people can afford o buy houses and hus pushing up HP. However, despie he drasic drop in income for he period , HP sill spiral upward. Subsequenly income goes up in 2010 bu i drops again in All hese while, HP move up raher exponenially. The wo graphs sugges ha HP moving up faser han income afer Apparenly HP are no sensiive o he flucuaion of income, a possible indicaion ha invesors are convinced ha HP will coninue o rise and hus very cerain of a profiable reurn for heir invesmens. This resul is in line wih he finding noed in BNM 2012 repor of he rend of curren prices are affeced by pas prices. This is a vial sign for he formaion of housing bubble. Sill, his may no saisfy he crierion for he formaion of a bubble. Income may no move up in andem wih he rise of HP. Bu his could be a emporary phenomenon (a shor run economic shock) as once he economy picks up income level will move up Empirical Analysis To idenify cycle or bubble, we compue he difference beween he observed HP and fundamenal HP which are prediced based on he long run relaionship beween HP and macroeconomic and financial facors. However, in order o obain he correc se of housing deerminans o run effecive coinegraing regression, we examine pas lieraure o obain housing deerminans, followed by general o specific rule o run regression and hen analyzing he resuls from wo differen long run coinegraing regression. Prior o performing coinegraing regression, Granger (1969) causaliy es and Johansen s (1994) mulivariae approach o coinegraion es have o be performed iniially. I is necessary o confirm wheher he dependen variable is I(1) and all he explanaory variables are also I(1) and no I(0) explanaory variable. For his purpose, we conduc Augmened Dickey Fuller and Kwiakowski Phillips Schmid Shin uni roo es. Following he confirmaion of he iniial saisical ess, we perform he Johansen mulivariae coinegraion es and Granger causaliy analysis. I is found ha all he variables, HP, GDP, MTR, exchange rae of Japanese Yen o Ringgi (EXJ), exchange rae of Hong Kong dollar o Ringgi (EXH) and household deb o gross domesic produc (MGD) are I(1) and coinegraed, meaning ha here is a long run relaionship among hem. Nex, we run he VECM model and he end resul is ha GDP, MTR, MGD, EXJ and EXH each has long run causaliy wih HP, running from each deerminan o HP. However, he resuls of Granger causaliy es sugges ha Figure 1: Year on year house price changes in % 136 Inernaional Journal of Economics and Financial Issues Vol 7 Issue

6 HP o GDP and HP o EXH are unidirecional. The conclusion is ha here may no be any shor run causaliy from GDP, MGD, MTR, EXJ and EXH o HP. This suggess ha here is no shor run equilibrium among he variables in level. Indirecly his lends suppor o our preference for long run causaliy analysis. The final resul is: We conclude ha HP in he real esae secor and is deerminans in Malaysia are coinegraed and ha here is a meaningful long-run relaionship beween hem. Then from he resuls of coinegraing regression of Table 1, we compue he fied values of HP using he bes fied model which happened o be FMOLS. The fied values or he fundamenal prices are shown in equaion 5: ˆp =E[ p ] Where, ˆp represens he fied HP and p is he HP Long Run (Coinegraing) Regression Resuls When FMOLS is compared wih CCR, boh of hem produce a se of resuls which is no in conflic wih economic heory Figure 2: RHPC = Overall real house price index/gross domesic produc per capia (5) and convenional wisdom. However, FMOLS has a beer se of adjused R square, long run variance and sandard error of regression. Hence, we selec he se of resuls produced using FMOLS o compue he fundamenal HP using equaion 5. Nex, we compue B p by using Equaion 2 and keeping in mind ha B p mus be more han 20% for possible bubble o exis. The resuls as shown in Table 2 indicae ha B p is always <20%. The crierion in equaion 3 is no saisfied and herefore i is expeced ha here is no bubble ye bu price cycle. Values in Table 3 shows ha α < 1 and β > 0, suggesing ha he price change is saionary and ha HP and is fundamenal values are very close o each oher, which indicaes ha he HP cycle is a play and ha i is sable. Capozza e al. (2002) found ha HP increases are due o fundamenals only. In addiion, i is found ha (1 + α-β) 2-4α = > 0. This implies ha he ransiory pah in response o changes in equilibrium HP value suggess a damped flucuaion around he equilibrium level. The propery of oscillaion is deermined by he magniude of α + β. Usually a higher α indicaes a higher ampliude of price oscillaion while a higher β implies a higher frequency of he flucuaion process. As a whole, since he HP changes are sable, i lends suppor o he previous graphical analysis ha he HP are efficien in he sense ha i follows he fundamenal pah and recover he equilibrium price afer absorbing some shor erm economic shocks. 5. CONCLUSION Source: Real house price index/gross domesic produc per capia is obained form inernaional moneary fund Table 1: Coinegraing regression resuls wih house price as dependen variable Model FMOLS CCR Model 1 Model 2 GDP 0.12 (0.01) 0.11 (0.02) MGD (0.00) (0.00) MTR 2.01 (0.01) 2.03 (0.02) EXJ 4.64 (0.00) 4.62 (0.00) EXH 1.34 (0.00) 1.35 (0.00) Adjused R² Long run VAR S.E. regression VAR: Variance, S.E. regression: Sandard error of regression, GDP: Gross domesic produc, FMOLS: Fully modified ordinary leas square, CCR: Canonical coinegraing regression From a policy perspecive, i is imporan o idenify he componens of housing price overvaluaion - wheher due o housing marke fricions or unsusainable overconfidence and high expecaion of capial gains from he housing marke, as each driving facor would require differen conainmen approach. From he afore described sudy, employing graphical analysis and some psychology consideraion, we come o he conclusion ha he Malaysian housing marke is facing a severe price cycle saring from 2009 o 2016 and he cycle is sill persising. This cycle is no classified as bubble because i does no spiral up sharply for wo consecuive years and fall rapidly for he nex wo subsequen years. Moreover, when his cycle is analyzed, i is found o be sable. Basing on hese resuls, we conclude ha i is a price cycle. However, he price behaviour of being disconneced from influence of income suggess overconfidence on he par of house buyers or invesors. Addiionally, his cycle demonsraes ha price expansion has peaked in 2013 and since hen he prices have sared o come down gradually. Furhermore, he HP -income graph suggess ha increase in income canno cach up wih he rise in HP and hus causing housing affordabiliy problem afer Oher han income and HP impariy, land cos, consrucion cos and compliance cos have gone up drasically, roughly in he raio of 20%: 70% and 10% which are due mainly o he worsening of exchange rae wih respec o he USD. Therefore Table 2: Annual OVP in percen Years Toal B p =annual OVP in % OVP: Overvaluaion of prices Inernaional Journal of Economics and Financial Issues Vol 7 Issue

7 Table 3: Shor run house price dynamics Variable Coefficiens saisical P value C p * p p * p Adjused R² F (P value) S.E. regression Persisence parameer α= 0.322, mean reversion parameer β=0.0046, conemporaneous adjusmen parameer γ= he scenario poins o he soluion ha o overcome he curren housing problem is o hasen economic growh o creae higher income for he people. 6. ACKNOWLEDGMENTS We wish o hank Higher Educaion Minisry for financially supporing his sudy hrough is FRGS research gran (FRGS/1/2016/SS08/UTAR/02/2). REFERENCES Abraham, J., Hendersho, P. (1996), Bubbles in meropolian housing markes. Journal of Housing Research, 7, Agnello, L., Schuknech, L. (2011), Booms and Buss in Housing Markes- Deerminans and Implicaions. European Cenral Bank Working Paper Series No. 1071/July, Ahuja, A., Cheung, L., Genberg, H., Porer, N., Zhang, W. (2010), Are house prices rising oo fas in China? HKMA Working Paper No. 08/2010. Allen, F., Gale, D. (2000), Bubbles and crises. The Economic Journal, 110, Bank Negara Malaysia. (2013), Developmens in he Housing Marke and Implicaions on Financial Sabiliy. Financial Sabiliy and Sysem Paymens Repor, Capozza, D., Hendersho, P., Mack, C., Mayer, C. (2002), Deerminans of Real House Price Dynamics. NBER Working Paper No. 9262, Ocober. Case, K.E., Shiller, R.J. (2004), Is here a Bubble in he Housing Marke. Cowles Foundaion Paper No. 1089, Yale Universiy, Cerui, E., Dagher, J., Dell Ariccia, G. (2015), Housing Finance and Real Esae Booms: A Cross-Counry Perspecive. IMF Discussion Noe SDN 15/ Chen, R.D., Gan, C., Hu, B., Cohen, D.A. (2013), An empirical analysis of house price bubble: A case sudy of Beijing housing marke. Research in Applied Economics, 5(1), Collyns, C., Senhadji, A. (2002), Lending Boom, Real Esae Bubbles and he Asian Crisis. IMF Working Paper WP/02/20. ECB. (2003), Srucural Facors in he EU Housing Markes. European Cenral Bank, March, Gallin, J. (2006), The long-run relaionship beween house prices and income: Evidence from local housing markes. Real Esae Economics, 34, Gerardi, K.S., Fooe, C.L., Willen, P.S. (2010), Reasonable People Did Disagree: Opimism and Pessimism abou he U.S. Housing Marke before he Crash. Public Policy Discussion Papers No. 10-5, Federal Reserve Bank of Boson. Glindro, E.L., Subhanij, T., Szeo, J., Zhu, H. (2011), Deerminans of house prices in nine Asia pacific economies. Inernaional Journal of Cenral Banking, 7(3), Granger, C. (1969), Invesigaing causal relaions by economeric models and cross-specral mehods. Economerica, 37, Hessel, J., Peeers, J. (2011), Housing bubbles, he leverage cycle and he role of he cenral bank. De Nederlandsche Bank Occasional Sudies, 9(5), Himmelberg, C., Mayer, C., Sinai, T. (2005), Assessing high house prices: Bubbles, fundamenals and mispercepions. Journal of Economic Perspecives, 19(4), Hussain, M.Y., Rahman, R.A., Hussain, F.N.M., Lyndon, N., Ibrahim, N.M. (2012), Housing bubbles assessmen : Experiences in Klang Valley, Malaysia. Advances in Naural and Applied Sciences, 6(1), Johansen, S. (1994), The role of he consan and linear erms in coinegraion analysis of non-saionary variables. Economeric Review, 13, Kalra, S., Mihaljek, D., Duenwald, C. (2000), Propery Prices and Speculaive Bubbles: Evidence from Hong Kong SAR. Working Paper No. WP/00/2, Inernaional Moneary Fund. Mayer, C. (2011), Housing bubbles: A survey. Annual Review of Economics, 3, Mayer, C., Quigley, J.M. (2003), Commens and Discussion: Is here a Bubble in he Housing Marke? Brookings Paper on Economic Aciviy. Vol. 2003, Paper No. 2, The Brookings Insiuion. Mayer, C., Shiller, R.J. (2006), Bubble, bubble, where is he housing bubble-commens and discussion. Brookings Papers on Economic Aciviy, 2006, The Brookings Insiuion. OECD. (2005), Recen house price developmens: The role of fundamenals, OECD Economic Oulook, 78, Scharfsein, D.S., Sein, J.C. (1990), Herd behaviour and invesmen. The American Economic Review, 80, Yip, C.Y., Wong, W.C., Woo, K.H. (2016), Deecing Malaysian housing bubbles. American Journal of Applied Sciences, 13, Yiu, M.S., Yu, J., Jin, L. (2013), Deecing bubbles in Hong Kong residenial propery marke. Journal of Asian Economics, 28, Inernaional Journal of Economics and Financial Issues Vol 7 Issue