EXTERNAL BALANCE AND BUDGET IN MALAYSIA

Transcription

1 ASIAN ACADEMY of MANAGEMENT JOURNAL of ACCOUNTING and FINANCE AAMJAF, Vol. 0, No. 2, 37 54, 204 EXTERNAL BALANCE AND BUDGET IN MALAYSIA ABSTRACT Wong Hock Tsen Faculty of Business, Economics and Accountancy, Universiti Malaysia Sabah, Malaysia This study examines external balance and budget in Malaysia. The unit root test results show that the variables examined are a mixture of stationary and non-stationary variables. The bounds testing results show that there is a long-run relationshi between external balance and its determinants, including budget, and also between budget and its determinants, including external balance. Moreover, the results of the causality analysis show that the current account targeting hyothesis and the Ricardian equivalence hyothesis dominate the relationshi between external balance and budget. Short-run and long-run measures are needed to address external imbalances. A healthy external balance is imortant for sustainable economic growth. Keywords: external balance, budget, Malaysia, causality INTRODUCTION Malaysia is a small oen economy, and its economic growth is strongly influenced by the erformance of the global economy. Malaysia had been achieving raid economic growth, esecially during the eriod from 988 to 996. However, its economic growth slowed down after the Asian Financial crisis in and during the global economic slowdown in The economic growth rate of Malaysia is unimressive comared to the economic growth rate of Malaysia before the Asian financial crisis, that is, during the eriod from 988 to 996. The average economic growth rate (2000 = 00) of Malaysia during was 3.5%, while the average economic growth rate (2000 = 00) from 988 to 996 was 9.% (International Monetary Fund [IMF], 202). Sluggish global economic growth, articularly in the develoed countries, has led to a series of budget deficits in Malaysia with the aim being to stimulate economic growth. Consolidated ublic sector finance in Malaysia roduced deficits of RM 4,685 million, RM 5,52 million, RM 5,80 million and RM 9,555 million for the years 2008, 2009, 200 and 20, resectively (Ministry of Finance Malaysia [MOF], 202). In 2008, the balance of ayments deficit in Malaysia was RM 8,250 million. However, in 2009 and 200, Malaysia exerienced balance of ayments surluses of RM 3,83 and Asian Academy of Management and Penerbit Universiti Sains Malaysia, 204

2 Wong Hock Tsen RM 2,628 million, resectively. In 20, Malaysia exerienced a balance of ayments deficit of RM 94,682 million (MOF, 200, 20, 202). External deficit can be bad or good. External deficit is bad, for examle, because financial regulation fails to fuel credit booms and misbehaviour of government leads to reduced national saving. External deficit is good, for examle, because international borrowing and lending allow for inter-temoral trade, which increases welfare (Blanchard & Milesi-Ferretti, 202,. 40). External imbalances and budget deficits have led to renewed interest in the twin deficits hyothesis (Theofilakou & Stournaras, 202, ). This issue gained much attention in the 980s, esecially in the United States (US) when the US exerienced significant external and budget deficits. The twin deficits hyothesis states that a budget deficit causes an external deficit. The Mundell and Fleming theory ostulates that an increase in the budget deficit will lead to an increase in the real interest rate in the domestic country, which will attract caital inflows. This will cause exchange rates to rise, which will make exorts less cometitive and lead to more imorts. Thus, there will be a current account deficit under a flexible exchange rate regime. Alternatively, the Keynesian absortion theory asserts that an increase in the budget deficit will lead to an increase in the domestic absortion and therefore will induce more imorts in the economy, leading to a current account deficit (Algieri, 203,. 3). In ractice, studies of the twin deficits hyothesis have roduced different results for different countries (Algieri, 203,. ; Chihi & Normandin, 203). This study investigates the imact of budget on external balance in Malaysia over the eriod from 980 to 20. More secifically, this study examines the imacts of consolidated ublic sector finance and federal government finance on balance of trade, balance of services, balance of current account and balance of ayments in Malaysia. The imact of budget could be different for different sub-categories of balance of ayments due to different goods and services elasticities. The revious studies mainly focused on the imact of budget on balance of trade or balance of current account (Kalou & Paleologou, 202). There have been very few studies that examine the imact of budget deficit on goods and services balances under the structural break and also investigate the relationshi between balance of ayments and budget deficit, esecially for Malaysia. The autoregressive distributed lag (ARDL) aroach of Pesaran, Shin, and Smith (200) is used. The aroach is suitable regardless whether all regressors examined are integrated of the same order, that is integrated of order one (I[]), integrated of order zero (I[0]) or a mixture of I() and I(0) variables. Moreover, this study considers the structural break in the examination of causality. 38

3 External Balance and Budget in Malaysia LITERATURE REVIEW There is a large literature on the relationshi between external balance and budget deficit (Kouassi, Mougoue, & Kymn, 2004; Rafiq, 200; Cama & Gavilan, 20; Jinjarak & Sheffrin, 20; Blanchard & Milesi-Ferretti, 202; Kalou & Paleologou, 202; Mussa, 202; Theofilakou & Stournaras, 202; Nag & Mukherjee, 202; Chihi & Normandin, 203; Hoffmann, 203; Trachanas & Katrakilidis, 203). However, there is no consensus on the relationshi between external balance and budget deficit. Some studies have found that budget deficit causes external deficit. Baharumshah and Lau (2007) investigate the relationshi between current account deficit and budget deficit in Thailand using four variables in a vector autoregressive (VAR) model using quarterly data for the eriod from 976:QI to 200:QIV. The variables emloyed in the study are: current account as a ratio of gross domestic roduct (GDP), budget deficit as a ratio of GDP, the nominal interest rate and the nominal exchange rate, which is the Thai baht against the US dollar. The results show that there is a long-term relationshi among the variables. Budget deficit is found to cause current account deficit and not the reverse. More secifically, an increase in the budget deficit will lead to an increase in the nominal interest rate, and this will areciate the nominal exchange rate and thus roduce a current account deficit. The results of the generalised variance decomosition demonstrate that the nominal exchange rate is the most exogenous variable, and the nominal interest rate is influenced by budget deficit. An increase in the nominal interest rate will dislace rivate investment. In other research, Chihi and Normandin (203) assess the link between external balance and budget deficit balances in 24 develoing countries in Africa, the Americas, Asia and Oceania. The variables studied are the US real treasury bills rate, terms of trade, the real effective exchange rate, GDP divided by consumer rice index (CPI), government exenditure as a ratio of GDP, tax as a ratio of GDP and household exenditure as a ratio of GDP. The results show that in twelve of the countries examined, there is a ositive relationshi between external balance and budget deficit. The domestic resources net of ublic absortions are the most imortant factors exlaining the ositive relationshi between the external deficit and budget deficit for most countries. Budget deficit can influence external deficit and vice versa. Rafiq (200) examines the interaction between budget deficits, current account balances and real exchange rates in the United Kingdom (UK) and US in a time-varying VAR model, which allows for time variation in the stochastic variance and autoregressive arameters, over the eriod from 973:QI to 2008:QIV for the UK and from 973:QI to 2009:QI for the US. The results show that budget deficit reduces the US current account balance. For the UK, budget deficit imroved current account balance. However, the imacts of budget deficits on the UK and the US current account balances have fallen in magnitude 39

4 Wong Hock Tsen over the ast 20 years. The time-varying variance decomosition results reveal that budget deficit shocks layed a key role in driving the UK current account and the real exchange rate fluctuations. However, budget deficit shocks have been a marginal factor in the variation of the US current account and the exchange rate fluctuations. The common finding for the UK and the US is that budget deficit reductions alone cannot eliminate current account imbalances. In contrast to the UK, the findings regarding the US suort the view of using dereciation of the real exchange rate to correct current account imbalance. There are some studies that found no relationshi between external balance and budget. Algieri (203) analyses the relationshis between external balances, namely trade balances, current account balances and budget balances in Greece, Ireland, Italy, Portugal and Sain using quarterly data from the eriod 980:QII to 202:QII. The study uses the Granger causality test and the Toda and Yamamoto causality test. The variables used in the study are government balance as a ratio of GDP and current account balance or trade balance as a ratio of GDP. The results reveal that both tests roduce the same conclusion and suort the Ricardian equivalence hyothesis; more secifically, there is no nexus between current account deficit or trade balance deficit and budget deficit. This imlies that a reduction in budget deficit may not hel reduce external deficits. This is because a budget deficit will lead to inter-temoral reallocation of savings and thus there will be no effect on the interest rate or exchange rate. Therefore there is no effect on the external balance. Rational agents will learn that budget deficit today will lead to increases in taxes in the future. Consequently, the rational agents will save more today to ay more tax in the future (Algieri, 203). The nexus between budget deficit and external deficit is more comlex than the twin deficits hyothesis. There are many factors that can influence external imbalance, such as an internal devaluation olicy to reduce external balance and imrovements in the cometitiveness of tradable goods and services through increases in roductivity and quality (Algieri, 203,. 9). Some studies found that external deficit causes budget deficit. Kalou and Paleologou (202) re-examine the twin deficits hyothesis in a vector error correction model (VECM) including the endogenous determination of structural breaks to determine the causal relationshi between budget deficit and current account deficit in Greece using annual data over the eriod from 960 to The variables analysed are budget deficit as a ratio of GDP, current account balance as a ratio of GDP, the 2-month Treasury bill rate and the nominal effective exchange rate. The two deficits are found to be ositively related and the direction of causality is from current account to budget deficit. In other words, the results suort the current account targeting hyothesis. The hyothesis affirms that current account deficit induces slower economic growth, and subsequently government imlements a budget deficit to stimulate economic 40

5 External Balance and Budget in Malaysia growth in the hoe of reducing the current account deficit. This is true for small oen economies that strongly deend on caital inflows. In a small oen economy, interest rates are exogenous and long-term causality is exected to run from interest rates to current account. In a large economy, interest rates are determined by the budget deficit (Kalou & Paleologou, 202,. 233). In the resence of a growing debt to GDP ratio, domestic develoments will be restrained by the foreign balance. Therefore, tax reforms to curb tax evasion and structural reforms in the financial and labour markets are necessary (Kalou & Paleologou, 202,. 239). DATA AND METHODOLOGY The two budgets used are consolidated ublic sector finance (bd t ) and federal government finance (bd 2t ), which are both exressed as a ratio of GDP. Consolidated ublic sector finance includes federal government finance, state governments' finances, local authorities' finances and statutory bodies' finances (MOF, 202,. 74). The use of the two budgets is to verify the imacts of those budgets on external balances. Balance of trade (bt t ), balance of services (bs t ), balance of current account (bca t ) and balance of ayments (bo t ) are also all exressed as a ratio of GDP. The real interest rate (r t ) is exressed as rt = tb t π t, where tbt is the 3 months treasury bill rate in Malaysia and π t is the inflation in Malaysia. The inflation in Malaysia is calculated as π t = [( cit cit ) / cit ] 00, where ci t is CPI (2000 = 00) in Malaysia. The real exchange rate (rer t ) is exressed by the real effective exchange rate of Malaysia (2000 = 00). The real stock rice return (rs t ) is exressed as rs t = ( st / st ), where st is the stock rice (2000 = 00) in Malaysia divided by CPI in Malaysia. The samle eriod of this study is from 980 to 20. The budget and balance of ayments data were obtained from various issues of Economic Reort, ublished by MOF, and the rest of the data were obtained from International Financial Statistics, available from the IMF. Figure dislays the lots of external balances, and Figure 2 dislays the lots of budgets. Generally, balance of trade, balance of services, balance of current account and balance of ayments fluctuated closely before the Asian financial crisis in However, those series diverged after the crisis. The standard deviations of the series from 980 to 996 were small, namely 5,04.7, 4,339.7, 7,45. and 8,437.7, resectively. The standard deviations increased from 997 to 20, namely 4,355.5, 0,077.6, 39,297.7 and 32,485.3, resectively. Consolidated ublic sector finance and federal government finance fluctuated closely with an uward trend towards zero before the Asian financial 4

6 Wong Hock Tsen crisis. Those series exhibited a high degree of fluctuation after the crisis. Consolidated ublic sector finance and federal government finance were strongly ositively correlated over the eriod from 980 to 20; that is, the coefficient of correlation was 0.7, which is significant at the % level. Figure. The lots of balance of trade, balance of services, balance of current account and balance of ayments in Malaysia, (RM million) Source: MOF, 203 Figure 2. The lots of consolidated ublic sector finance and federal government finance in Malaysia, (RM million) Source: Various issues of Economic Reort, MOF 42

7 External Balance and Budget in Malaysia Consolidated ublic sector finance and federal government finance were negatively correlated with balance of trade, balance of services, balance of current account and balance of ayments. The correlation coefficients were high between consolidated ublic sector finance and federal government finance with balance of trade ( 0.45, 0.85) and balance of current account ( 0.5, 0.87), but relatively low with balance of services ( 0.42, 0.67) and balance of ayments ( 0.3, 0.34) (Table ). This imlies that budgets have a stronger imact on balance of trade and balance of current account then on balance of services and balance of ayments. Table The coefficients of correlation of consolidated ublic sector finance or federal government finance with balance of trade, balance of services, balance of current account and balance of ayments, resectively, (RM million) bd t bt t bs t bca t bo t bd t.00 bt t bs t bca t bo t bd 2t bt t bs t bca t bo t bd 2t.00 bt t bs t bca t bo t Note: bd t denotes consolidated ublic sector finance while bd 2t denotes federal government finance. Source: Various issues of Economic Reort, MOF. The relationshi between current account deficit and budget deficit can be shown using the national account identity: y t = c t + i t + g t + (x t m t ) () where y t is GDP, c t is rivate consumtion, i t is investment, g t is government consumtion, x t is exorts and m t is imorts. Equation () can be rewritten in terms of the external sector (x t - m t ) as follows: (x t m t ) = y t c t i t g t (2) 43

8 Wong Hock Tsen The national saving (s t ) is equal to y t c t g t and thus equation (2) can be written as follows: (x t m t ) = s t i t (3) The national saving can be divided into government saving (s g ) and rivate saving (s ). Government saving can be defined as (t t g t ), where t t is tax and g t is government consumtion. When (t t g t ) is ositive, government exeriences a budget surlus. When (t t g t ) is negative, government exeriences a budget deficit. Equation (3) can be written as follows: or (x t m t ) = s + (t t g t ) i t (4) (x t m t ) = s i t + bd t (5) where bd t is budget deficit. When (x t m t ) is negative, it imlies a deficit while when (x t m t ) is ositive, it imlies a surlus. When (x t m t ) is negative, a country can finance the external sector through borrowing from abroad. In other words, the country is imorting resent consumtion and exorting future consumtion. The (s i t ) is the rivate saving and investment balance. If the rivate saving and investment are about the same or constant, then external balance (x t m t ) and government balance (bd t ) and will move closely together. If a change in budget deficit is offset by change in saving, that is called the Ricardian equivalence hyothesis, which ostulates that budget and current account are unrelated (Algieri, 203,. 3 5). An inter-temoral shift between taxes and budget deficits does not matter for the real interest rate and investment or current account balance. This means that government deficits are neutral and the twin deficits only haen coincidently. The factors that influence current account are factors such as the resonse of consumtion to various shocks to the economy (Kalou & Paleologou, 202,. 233). Given that rivate saving and investment deend on the interest rate (ir t ), the exchange rate (ex t ) and the stock return (sr t ), equation (5) can be rewritten as follows (Kalou & Paleologou, 202,. 232): (x t m t ) = s (ir t, ex t, sr t ) i t (ir t, ex t, sr t ) + bd t (6) The Zivot and Andrews (992) (ZA) and Lee and Strazicich (2004) (LS) unit root test statistics are used to examine the stationarity of the data. The use of the conventional unit root test statistics can lead to the wrong conclusion about the stationarity of the data if structural breaks exist in the data. The ZA unit root test statistic is an augmented Dickey-Fuller tye endogenous break unit root test. The LS unit root test statistics are an endogenous unit root test for one or two 44

9 External Balance and Budget in Malaysia structural breaks that is unaffected by structural breaks under the null and alternative hyotheses, and thus, surious rejection will not occur. The LS unit root test statistics are based on the Lagrange Multilier test. When the variables in equation (6) are co-integrated, causality will be tested in the VECM as follows: b t = β 0 + βdt + β2i bdt i + β3i rt i + β4i ert i + β5i s t i + 6 i bt i + β7ect + u t β (7) bd t = β 20 + β 2Dt + β 22i bdt i + β 23i rt i + β 24i ert i + β 25i s t i + 26 i bt i + β 27ect + u2t β (8) where is the first difference oerator, b t is balance of trade, balance of services, balance of current account or balance of ayments, D t is the dummy variable to cature the influence of the Asian financial crisis, , bd t is the budget deficit, r t is the real interest rate, er t is the real exchange rate, s t is the real stock rice return and u i,t (i =, 2) is a disturbance term (Kalou & Paleologou, 202,. 232). Baharumshah and Lau (2007) and Kalou and Paleologou (202), amongst others, use the nominal variables in the analysis of the relationshi between external balance and budget. In contrast, Rafiq (200) and others use the real variables. The Mundell and Fleming theory, or the Keynesian absortion theory, imlies that a budget deficit will lead to an external deficit. This hyothesis holds if the coefficient ( β 2i ) in equation (7) is significantly different from zero and the coefficient ( β 26i ) in equation (8) is not significantly different from zero. The Ricardian equivalence hyothesis suggests that there is no nexus between external deficit and budget deficit. This hyothesis holds if the coefficient ( β 2i ) in equation (7) and the coefficient ( β 26i ) in equation (8) are not be significantly different from zero. The current account targeting hyothesis indicates that unidirectional causality runs from external deficit to internal deficit. This hyothesis holds if the coefficient ( β ) in equation (7) is not significantly different from 2i 45

10 Wong Hock Tsen zero and the coefficient ( β 26i ) in equation (8) is significantly different from zero (Kalou & Paleologou, 202,. 233). RESULTS AND DISCUSSION The results of the ZA and LS unit root test statistics are reorted in Table 2. The lag lengths used to estimate the ZA and LS unit root test statistics are based on the t-statistic, that is, the number of lags for which the last included lag has a marginal significance level less than the cut-off given by the 0% level. The fraction of entries on each end of data to exclude as the breaks and minimum ga between breaks is the 0% level. The results of the ZA and LS unit root test statistics show that all the variables excet balance of ayments and the real stock rice return are mostly non-stationary in their levels but become stationary after taking the first differences. Thus the variables examined are a mixture of I() and I(0) variables. The results of the Elliot, Rothenberg and Stock (ERS) and Phillis and Perron (PP) unit root test statistics are reorted in Table 3. The lag lengths used to comute the ERS unit root test statistics are based on the Schwarz Bayesian Criterion. The lag lengths used to comute the PP unit root test statistics are based on the Newey-West automatic bandwidth selection with the maximum lag length set to seven. Generally, the ERS and the PP unit root test statistics show about the same conclusion as the ZA and LS unit root test statistics. Table 2 The results of the Zivot and Andrews (992) (ZA) and Lee and Strazicich (2004) (LS) unit root test statistics ZA Crash LS Crash ZA Break LS Crash bt t (998) 3.480* (997) (998) (996) bt t *** (996) ** (999) ** (996) * (999) bs t *** (2000) (999) 5.202** (2000) ** (2000) bs t 7.255*** (985) *** 995) *** (2000) *** (997) bca t (998) ** (997) (998) * (996) bca t *** (996) ***(999) ** (996) ** (987) bo t 5.355** (994) *** (996) ** (994) 5.54*** (2006) bo t *** (994) 5.566*** (996) 6.05*** (994) *** (2003) bd t (2008) (2008) (995) 4.966** (994) bd t *** (2004) *** (993) *** (2004) ** (2006) bd 2t (983).2878 (986) (988) * (997) (continued on next age) 46

11 External Balance and Budget in Malaysia Table 2 (continued) ZA Crash LS Crash ZA Break LS Crash bd 2t *** (2003) *** (994) *** (998) *** (988) r t 6.42*** (998).5893 (997) (998) 6.030*** (986) r t *** (986) *** (2009) (989) 6.572*** (2009) er t (986) 3.085* (2009) 6.2*** (998) (998) er t ** (985) *** (2000) *** (989) ** (988) s t *** (987) 5.730*** (2003) *** (995) *** (996) s t 8.42*** (999) *** (2009) *** (990) *** (997) Notes: Crash denotes the ZA or LS unit root test statistic for testing an abrut change in level but no change in the trend rate. Break denotes the ZA or LS unit root test statistic for testing an abrut change in level and a change in the trend rate. Values in arentheses are the breaks. The critical values can be obtained from Zivot and Andrews (992) and Lee and Strazicich (2004). *** (**, *) denotes significance at the % (5%, 0%) level. Table 3 The results of the Elliott, Rothenberg and Stock (ERS) and Phillis and Perron (PP) unit root test statistics ERS No trend ERS Trend PP No trend PP Trend bt t bt t *** *** *** 4.790*** bs t bs t *** *** *** *** bca t bca t 3.905*** 4.650*** 5.066*** *** bo t *** *** *** 4.846*** bo t *** 7.306*** *** 8.267*** bd t bd t *** *** *** bd 2t bd 2t *** *** *** *** r t ** * 3.502** r t *** *** 7.435*** 8.83*** er t er t 4.848*** *** *** 4.039** s t 6.475*** *** 7.464*** *** s t *** 5.33*** *** Notes: ERS No trend denotes the Elliott, Rothenberg and Stock test statistic estimated with the model included a constant only. ERS Trend denotes the Elliott, Rothenberg and Stock test statistic estimated with the model included a constant and a time trend. PP No trend denotes the Phillis and Perron test statistic estimated with the model included a constant only. PP Trend denotes the Phillis and Perron test statistic estimated with the model included a constant and a time trend. *** (**, *) denotes significance at the % (5%, 0%) level. 47

12 Wong Hock Tsen The results of the bounds testing aroach are reorted in Table 4. The Wald-statistics are found to be statistically significant at the % level. Hence there is a long-term relationshi between balance of trade, balance of service, balance of current account and balance of ayments and determinants of balance trade, balance of service, balance of current account and balance of ayments, resectively including consolidated ublic sector finance or federal government finance. Additionally, there is a long-term relationshi between consolidated ublic sector finance or federal government finance and its determinants including balance of trade, balance of service, balance of current account or balance of ayments. In other words, those variables are moving together and would not move far from each other in the long term. The results of the Johansen likelihood ratio test statistics, namely, the maximum eigenvalue statistic (λ Max ) and the trace statistic (λ Trace ), are comuted with unrestricted intercets and no trends in the VAR are reorted in Table 5. The lag lengths used to comute the VAR are based on the Schwarz Bayesian Criterion. The results show that the null hyothesis of no co-integration among balance of trade, balance of service, balance of current account or balance of ayments and its determinants is rejected. There is at least one co-integrating vector among those variables. Thus, testing of the Granger causality should be tested in the VECM. Table 4 The results of the bounds testing aroach for co-integration bd t bd 2t bt t 63.2*** *** bd t *** *** bs t *** *** bd t *** *** bca t 7.277*** *** bd t *** *** bo t *** *** bd t *** *** Note: Values are Wald-statistics, *** denotes significance at the % level. The results of the co-integration test show that there is long-term relationshi between external balance and its determinants, including budget, and also between budget and its determinants, including external balance. Chihi and Normandin (203), among others, reort a ositive relationshi between the external deficit and budget deficit. Consolidated ublic sector finance is found to Granger cause balance of trade and balance of current account. However, federal government finance is found to Granger cause balance of ayments. Hence these 48

13 External Balance and Budget in Malaysia findings suort the current account targeting hyothesis. Kalou and Paleologou (202), among others, document that the current account targeting hyothesis haens in Greece. In other words, fiscal olicy can be used to address external imbalance. Federal government finance is found to Granger cause balance of services and not the reverse. This suorts the Mundell and Fleming theory or the Keynesian absortion theory. Baharumshah and Lau (2007) and others arrive at the same conclusion for Thailand. However, there is no Granger causality found between consolidated ublic sector finance and balance of services. Thus the imacts of consolidated ublic sector finance and federal government finance on external balances are not the same. Additionally, consolidated ublic sector finance is found to have marginally more imact on external balances than federal government finance has. Budget may reduce imbalance of some goods and services but not others due to different elasticities of goods and services. To imlement an effective olicy, the government should determine the link between external balances and budgets and also the causality between them. Table 5 The results of the Johansen Likelihood Ratio Test statistics λ Max Test statistic H 0 : r = 0 r r 2 r 3 r 4 H a : r = r = 2 r = 3 r = 4 r = 5 bd t bt t 35.94* bs t 35.8* bca t * bo t 45.68* bd 2t bt t 55.43* 34.28* 30.30* 20.50* 0.04 bs t 43.79* 3.93* bca t 48.23* 4.52* 24.60* 5.06* 0.02 bo t 5.87* 29.90* 2.29* λ Trace Test statistic H 0 : r = 0 r r 2 r 3 r 4 H a : r r 2 r 3 r 4 r = 5 bd 2t bt t 95.92* 59.99* 33.* bs t 80.9* (continued on next age) 49

14 Wong Hock Tsen Table 5 (continued) bca t 92.90* 62.88* 34.54* bo t 0.2* 55.53* bd 2t bt t 40.56* 85.2* 50.84* bs t 98.30* 54.50* bca t 29.44* 8.2* 39.68* bo t 4.26* 62.39* 32.49* Notes: The VAR = 2 is used in the estimations. The critical value is based on MacKinnon et al. (999). * denotes significance at the 5% level. The results of the Granger causality test are reorted in Table 6. The lag length used to comute the Granger causality test statistic is based on the Akaike Information Criterion. Generally, the estimated models fulfil the conditions of normality of disturbance terms, no autocorrelation, homoscedasticity of disturbance terms and no functional form of the model. However, the Granger causality test statistics are based on the ordinary least squares with Newey-West corrected standard errors when autocorrelation or autocorrelation and heteroscedasticity of disturbance terms are found to be statistically significant. The short-term Granger causality, or weak Granger causality, which shows the short-term influence of the indeendent variables on the deendent variable, can be tested through the F-statistic on the coefficients of the lagged differences. For consolidated ublic sector finance, there are long-term relationshis between balance of trade, balance of current account or balance of ayments and budget which are imlied by the significance of the error correction terms, excet balance of services when it is a deendent variable. On the whole, there is no causality from consolidated ublic sector finance to balance of trade, balance of service, balance of current account or balance of ayments. Nonetheless, there are causalities from balance of trade and balance of current account balance to consolidated ublic sector finance. Thus this suorts the current account targeting hyothesis. For federal government finance, long-term relationshis are found between external balances and budget, excet balance of services when it is deendent variable and balance of trade and balance of account when they are indeendent variables. There is no causality from federal government finance to external balances, excet federal government finance is found to influence balance of services. Additionally, causality is found from balance of ayments to federal government finance. There is no evidence of causality between balance of trade and balance of ayments and federal government finance. This suorts the Ricardian equivalence hyothesis. On the whole, this study finds evidence that the current account targeting hyothesis and the Ricardian equivalence hyothesis dominate the relationshi between external balances and budgets. 50

15 External Balance and Budget in Malaysia The finding that budgets cause external balances imlies that the government can reduce the budget deficit to imrove the external deficit. In addition, the government should consider other measures such as controlling inflation and maintaining the cometitiveness of exorts. In the long run, a small oen economy such as Malaysia should reform its tax to curve tax evasion and continue to transform its economy successfully. Cutting ublic sending to reduce the budget deficit is also imortant (Kalou & Paleologou, 202,. 239). In the long run, the focus should be on imroving roductivity and quality through technological advancement to enhance the exort cometitiveness of Malaysia. Malaysia aims to achieve a high-income economy by the year In the 0th Malaysia Plan (20 205), strategies have been roosed for transforming the Malaysian economy, maintaining full emloyment, ursuing roductivity-led growth, increasing the dynamism of the rivate sector, romoting growth in rivate consumtion, diversifying exort markets, sustaining balance of ayments surlus and imroving the efficiency of fiscal olicy. Also in the 0th Malaysia Plan, several measures have been imlemented to ensure the sustainability of ublic finance in the long run. A two-year rolling lan has been imlemented to rovide flexibility in exenditure management in tandem with fiscal resources and olicy riorities. Value management is required for develoment rojects above RM 50 million, and outcome-based budgeting will be introduced. The efficiency and effectiveness of government sending shall be assessed (MOF, 202,. 23). However, the imrovement of external balances deends largely on how successfully and effectively those strategies are imlemented. Table 6 The results of the Granger Causality Test Deendent variable bd t i ec t Normal LM Hetero Reset bt t *** *** ** bs t *** bca t ** *** bo t *** Deendent variable b t i ec t Normal LM Hetero Reset bd t ( bt t ) * *** bd t ( bs t ) *** bd t ( bca t ) * 3.560*** bd t ( bo t ) ** (continued on next age) 5

16 Wong Hock Tsen Table 6 (continued) Deendent variable bd 2t i ec t Normal LM Hetero Reset bt t * *** * ** bs t 4.590** *** bca t ** bo t *** * Deendent variable b t i ec t Normal LM Hetero Reset bd 2t ( bt t ) ** bd 2t ( bs t ) * bd 2t ( bca t ) * bd 2t ( bo t ) * ** Notes: b t i denotes bt t i, bs t i, bca t i or bo t i. The values under the columns bd t i, bd 2t i and b t i are the F-statistics whilst the values under the column ec t are the t statistics. Normal denotes Jargue Bera test of the residual normality. LM denotes Breusch Godfrey serial correlation test. Hetero denotes the residual heteroskedasticity test. Reset denotes Ramsey regression equation secification error test (RESET). *** (**, *) denotes significance at the % (5%, 0%) level. CONCLUSION This study has examined the link between external balance and budget in Malaysia. It is imortant to determine the link between external balances and budgets, and also the causality between them, for imlementing an effective budget olicy. There is a long-term relationshi between external balances and budgets. This study finds evidence that the current account targeting hyothesis and the Ricardian equivalence hyothesis dominate the relationshi between external balances and budgets. Consolidated ublic sector finance is found to have marginally more imact on external balances than federal government finance has. A budget may reduce the imbalance of certain goods and services, or one comonent of balance of ayments but not others due to different elasticities of goods and services. One way to address the external imbalance, esecially in the short term, is through the use of the budget. In the short term, it is crucial for the government to control inflation in the country. Cutting ublic sending and the effective use of the ublic sending are imortant to address external balance. In the long-run, the focus should be on imroving roductivity and the quality of roducts and services through technological advancement to enhance the exort cometitiveness of Malaysia. Successful and effective transformation of the Malaysian economy to a higher level is crucial. 52

17 External Balance and Budget in Malaysia ACKNOWLEDGEMENT The author would like to thank the Ministry of Education Malaysia for roviding the funding for this study through the Fundamental Research Grant Scheme (FRGS//203/SS07/UMS/02/2). Additionally, the author would like to thank the reviewers of the journal for their comments. An early version of the aer was resented at the 0th Asian Academy of Management International Conference, August 203, Hotel Bayview Beach Resort Batu Feringgi, Penang, Malaysia. All remaining errors are the author's. REFERENCES Algieri, B. (203). An emirical analysis of the nexus between external balance and government budget balance: The case of the GIIPS countries. Economic Systems, 37(2), Baharumshah, A. Z. & Lau, E. (2007). Dynamics of fiscal and current account deficits in Thailand: An emirical investigation. Journal of Economic Studies, 34(6), Blanchard, O. & Milesi-Ferretti, G. M. (202). (Why) Should current account balances be reduced? IMF Economic Review, 60(), Cama, J. M. & Gavilan, A. (20). Current accounts in the euro area: An intertemoral aroach. Journal of International Money and Finance, 30(), Chihi, F. & Normandin, M. (203). External and budget deficits in some develoing countries. Journal of International Money and Finance, 32, Hoffmann, M. (203). What drives China's current account? Journal of International Money and Finance, 32(C), International Monetary Fund (IMF). (202). International financial statistics yearbook. Washington D.C., USA: IMF. Jinjarak, Y. & Sheffrin, S. M. (20). Causality, real estate rices, and the current account. Journal of Macroeconomics, 33(2), Kalou, S. & Paleologou, S. M. (202). The twin deficits hyothesis: Revisiting an EMU country. Journal of Policy Modeling, 34(2), Kouassi, E., Mougoue, M. & Kymn, K. O. (2004). Causality tests of the relationshi between the twin deficits. Emirical Economics, 29(3), Lee, J. & Strazicich, M. C. (2004). Minimum LM unit root test with one structural break (Working Paer). NC, USA: Deartment of Economics, Aalachian State University. MacKinnon, J. G., Haug, A. & Michelis, L. (999). Numerical distribution functions of likelihood ratio tests for cointegration. Journal of Alied Econometrics, 4(5), Ministry of Finance Malaysia (MOF). (200). Economic reort 200/20. Kuala Lumur, Malaysia: Percetakan National Malaysia Berhad. MOF. (20). Economic reort 20/202. Kuala Lumur, Malaysia: Percetakan National Malaysia Berhad. 53

18 Wong Hock Tsen MOF. (202). Economic reort 202/203. Kuala Lumur, Malaysia: Percetakan National Malaysia Berhad. Mussa, M. (202). Rerint of: The dollar and the current account deficit: How much should we worry? Journal of Policy Modeling, 34(4), Nag, B. & Mukherjee, J. (202). The sustainability of trade deficits in the resence of endogenous structural breaks: Evidence from the Indian economy. Journal of Asian Economics, 23(5), Pesaran, M. H., Shin, Y. & Smith, R. J. (200). Bounds testing aroaches to the analysis of level relationshis. Journal of Alied Econometrics, 6(3), Rafiq, S. (200). Fiscal stance, the current account and the real exchange rate: Some emirical estimates from a time-varying framework. Structural Change and Economic Dynamics, 2(4), Theofilakou, N. & Stournaras, Y. (202). Current account adjustments in OECD countries revisited: The role of the fiscal stance. Journal of Policy Modeling, 34(5), Trachanas, E. & Katrakilidis, C. (203). The dynamic linkages of fiscal and current account deficits: New evidence from five highly indebted Euroean countries accounting for regime shifts and asymmetries. Economic Modelling, 3, Zivot, E. & Andrews, D. (992). Further evidence of the great crash, the oil-rice shock and the unit root hyothesis. Journal of Business and Economic Statistics, 0(3),