Maintaining a currency has become one of the challenges in the Macroeconomic management of most countries. Among others, Malaysia has a very unique experience in term of currency crises. During East Asian countries crises beginning in late 1997, the currencies of Thailand, Indonesia, The Republic of Korea, Malaysia and Philippines experienced sudden, relatively unexpected depreciation. Currency crisis literature indentified a decline in credit as one of the channels through which crises affect real economic activity. The main objective in this study is to revaluate the causes of currency crises by focusing on the economic variables factors but at the same time also focus on the role played by a broader array of institution factors and crisis episodes. This paper examines the depth of currency crises in Malaysia, for the period of 1969 until 2009 using Econometric equations which are conducted using annual data on real GDP, export, inflation rate, M2, and real exchange rate.
Keyword: Depth of currency crises, real GDP, export, inflation rate, real exchange rate, M2, econometric equations
1.0 Introduction
The decline in the value of a country's currency will brought to a currency crises. This decline in value negatively affects to an economy by creating instabilities in exchange rate which means that one unit of the currency has no longer buys as much as it used to in another. The currency crises also known as a balance of payment crisis which is speculative attack in the foreign exchange market and brought on by a decline in the value of a country's currency. It also can occur when the value of a currency changes quickly. In the other word, it may occur when the value of a currency changes quickly which is undermining its ability to serve as a medium of exchange. We can say that crises develop as an interaction between what those expectation cause to happen and investor expectations. The currency crises not sufficiently stable ones but can be especially destructive to small open economics or bigger. There have different ideas ,thoughts and findings about currency crises. Alesina and Wager (2003) and Calvo and Mishkin (2003) have same idea in two related papers which consider the quality of institution and exchange rate arrangement. They debate and argue about fixed versus over floating exchange rates, deeper institutional features relating to fiscal stability, price and financial stability are more important to avoid crises. Another, john et al. (2000), Block (2002), Gosh and Gosh (2002) and Mulder al. (2002) have the same finding which consider institutional as prime factor that contribute to crises. They consider the number of variables including rule of law, judicial efficiency, corruption, the strength of government, rule of law, the legal regime, contract enforcement, and accounting standards. (Li and Inclan,2001) said the institution that effect to currency crises are through two causal mechanisms which are, first give impact with the health of the national economy that can lead to bad economic fundamental and contribute to currency crises. The effects of institutions are measured in two dimensions. First dimensions are on the probability of currency crises and second dimension is on depth of contraction in output during crisis. Second, informative of institutions. From the financial crises of the East Asian in 1997-1999 has show about unprecedented economic and social distresses on the economic.
1.1 Currency Crises in Malaysia
Since independent, economy in Malaysia has largely been on a sound footing. It has enjoyed full employment in recent years, high growth, and considerable stability in the price, the average rate being 8.7% during 1990-1997. It occupied in the world a high rank at 35 in 1997 with reference to per capita GNP and the size of the aggregate. Furthermore export and import play a major role in a progress of the economy. As a result, the economy is much susceptible to external disturbances and high open. To sustain high investment rates for pushing up growth, foreign capital continually supplements domestic savings. The high performance of the Malaysian economy maintained over the decades. It cannot be slighted without technology or market or even the core capital being local. In July 1997, the financial meltdown starts towards in Malaysia. In the first week of the following month, the major fall in the stock prices occurring sometimes. The stock market fell by 68.58% during the worst patch of about a year and the dollar-ringgit rate plunged by over 37%. In year 2006-2008, Malaysia faced with the global financial crisis with its epicentre in the United States which has brought enormous ramifications for the world economy. Started as an asset bubble caused by an array of financial derivatives and exploded into housing and banking crisis with a cascading affect on consumer in investment demand.(krugman,2009) it quickly grew into a banking crisis with the investment and merchant banks first absorbing the impact before it spread to the commercial banks, from a housing crisis. It sends ripples across export dependent Asian economies with the United States which began to face a contraction as consequences. In Malaysia, although its economy was insulted from the direct effects of financial exposure because of the new derivatives were not allowed into the country. The bitter experience of the Asian financial crisis had already provided the incentive for Bank Negara Malaysia to regulate the financial sector without unduly affecting the stock market.
2.0 Literature Review
The causes of the currency crises have been addressed with sequential generation of models developed largely along historical lines. In these models, weak institutions worsen problems associated with uncertainty, risk and contribute to a misallocation of resources thereby setting the stage for currency crises. According to Rajan and Zingales (1998), they consider contract enforcement and opportunities for malfeasance; examine the example of political factors such as divisive and polarized parliaments (Bussiere and Mulder, 1999); Rossi (1999) considers bank supervision, capital account openness, and depositor safety; Johnson et al. (2000) consider a number of variables including judicial efficiency, rule of law, and corruption; Li and Inclan (2001) consider coordinated wage bargaining, stock development, central bank independence, and more; and Mulder et al. (2002) consider among other factors, accounting standards, contract enforcement, and the legal regime.
According to Li and Inclan (2001), there are through two causal mechanisms in institution those effects the currency crises. For the first one is the institution tend to have an impact and correlate with the health of the national economy. For consequence, the institutions that lead to bad economic fundamentals may contribute to currency crises whereas the institutions which help in produce good economic fundamentals remove one reason for currency crises to occur. For the second one is informative of institutions that consequence to correlate with bad economic fundamentals destabilizes market expectations, increase market uncertainty about the likelihood of currency crises, and make currency crises motivated by speculative capital outflows more likely.
The economic variables are included domestic real or credit GDP, export real or credit GDP, foreign real or reserve GDP, M2 or foreign reserves, the inflation rate, the real exchange rate, the trade balance or real GDP, and U.S. interest rate. Rose (1996) note that currency crises will occur when foreign interest rates are high, growth of domestic credit, growth in FDI and output growth are low. According to Furman and Stiglitz (1998) and Chancy and Velasco (1998b) in empirical evidence which is suggested that the growth in the ratio of short term foreign- currency denominated debt to international reserves is an important predictor of currency crises
Export is one of important variables in the rapid industrialization of the economies of East Asia. From newly industrializing economies which is the majority of export of East Asia went to OECD countries in the early 1980's. However, by the late 1990's, export to developed economies (including Japan) is small if want to compare to export to other Asian countries (excluding Japan) where almost large. It is important to incorporate the regional interdependence of the Asian economies, in order to understand the Asia crisis of 1997-1999.
3.0 Data and Methodology
In this section describe the data and methodology used in measure the depth of currency crises. We have collected annual data from 1969 until 2009. Since we are interested not only in explaining currency crises but also in evaluating the predictive power of our model, most of explanatory variables are enter in lagged form. We have a data set of 41 years. Since our dependent variable is dichotomous and we take the value of 1 when there is a crisis and 0 is otherwise. The data we use are M2, inflation, export, balance of payment and real exchange rate.
Model specification:
Depth Currency crisis = β + GDP + M2 + Exchange rate + Inflation rate + Export + e
Source: Basic Econometrics by Gujarati D N. and Porter D C.
The function can also be representing in log-linear econometric format:
LogC = α0 + α1 logGDP + α2 logM2 + α3 logExport + exchange rate + inflation rate + ε
Source: Basic Econometrics by Gujarati D N. and Porter D C.
Where the dependent variable is Depth currency crisis or C and independent variables are GDP, M2, inflation, export, balance of payment and real exchange rate. And 'ε' is the random error term. From the equation above, the positive sign of coefficient for GDP, M2, inflation, export, balance of payment and real exchange rate represent that there are positive relationship between GDP, M2, inflation, export, balance of payment and real exchange rate and dept currency crises in Malaysia. In contrast, if the GDP, M2, inflation, export, balance of payment and real exchange rate have negative correlation to dept of currency crises, it will not help in measurement of currency crises in a country. The hypothesis is stated as below;
H0 : β = 0
H1 : β ≠ 0
The null hypothesis β = 0 (there are no relationship between GDP, M2, inflation, export, balance of payment and real exchange rate and dept currency crises against its alternative β ≠ 0, if less than lower bound critical value (0.05), then we do not reject the null hypothesis. We can conclude that there are significant relationship between independent variables and dependent variable if the t-statistic value greater than 5 percent cortical values then we reject the null hypothesis.
3.1 Unit Root Test
In this paper, unit root test is use to examine the stationary of the time series. It also can be used to determine if trending data should be first differenced or deterministic functions of time to render the data stationary This study use the Augmented Dickey-Fuller (ADF) and Phillips-Perron test statistic (PP) to differentiate the time series data in order to make it stationary. ADF test relies on rejecting a null hypothesis of unit root which is the series are non stationary, in favour of the alternative hypotheses of stationary.
n
Δ Ct = αο + α1C t-1 + ΣαΔCt + e
i= 1
n
Δ Ct = αο + α1 Ct-1 + ΣαΔCt +δ+ et
n= 1
Source: Basic Econometrics by Gujarati D N. and Porter D C.
where C is a time series for currency crises, is a linear time trend, Δ is the difference operator, αο is a constant, n is optimum number of lags in the dependent variable and e is the random error term.
The Phillip-Perron (PP) is equation is thus;
Δ Ct = αο + α Ct-1 + et
Source: Basic Econometrics by Gujarati D N. and Porter D C.
3.2 The Cointegration Test
This test is involved with the presence or otherwise of cointegration between the series of the same order of integration which is through forming a cointegration equation. Basically, in the long run, two or more series move closely together even the series themselves are trended but the difference between them is constant. According to Hall and Henry 1989, it is possible to regard these series as defining a long run equilibrium relationship as the difference between them is stationary. A lack of cointegration will result that variables have no long run relationship which is in principal that they can wander arbitrarily far away from each other (Dickey et. al, 1991).
3.3 The Granger Causality Test
This test are defined as F-test or joint test for the significance of the lagged values of the assumes exogenous variable. From the examination result, it will indicate that we have to reject or accept the null hypothesis and conclude there are exists bidirectional causality between dependent and independent variable
4.0 FINDING
4.1 Stationary Test
According to both the Augmented Dickey Fuller (ADF) and Phillip Perron (PP) test were applied to find the existence of unit root in each of the time series. The result found that dependent variable and independent variables are stationary. This can be seen by comparing the observed values of both ADF and PP test statistic with the critical values of the statistic at 1%, 5% and 10% level of significance. The result from table 1 has provided the evidence from the test. Before using the PP test, ADF test is use to test the variables. If through ADF test the variable is not stationary at level so we test the variable using pp test then we make comparison trough both method which one will result to stationary on variable. If the variables need the test through at 1st or 2nd difference to make it stationary so we examined through it. The comparison through both methods was providing in table 1. For depth of currency crises, there is significance at level through both ADF and PP test statistic but the finding shows that through ADF test, it more significance because the probability result at 0.0000 significance at 1st difference. For M2 there is more significance at 2nd difference through ADF test. For inflation, there is more significance at 1st difference through ADF test. For GDP and export, there is more significance at 1st difference through PP test. Lastly for exchange rate, there is more significance at level through ADF test.
Table 1: ADF test and PP test
Variables
ADF Statistic
PP Statistic
Depth of currency
M2
Inflation
GDP
Export
Exchange rate
(0.0000)**
-6.087489
(0.0000)***
-10.44195
(0.0000)**
-6.708709
(0.9951)**
0.968243
(0.9821)**
0.451160
(0.003)*
-4.826515
(0.002)**
-5.082824
(0.0000)***
-10.75221
(0.0000)**
-6.724699
(0.0060)**
-3.805592
(0.0054)**
-3.845043
(0.0004)*
-4.721409
Note: * significance at level
**significance at 1st difference
***significance at 2nd difference
4.2 Cointegration test
The result from the cointegration test which is the existence of a long term linear relationship is presented on table 3 below. From the cointegration result, both trace and maximum Eigen value statistic indicate there is only three variables have cointegration which is have long run relationship.
Table2: Cointegration Test
Unrestricted Cointegration Rank Test (Trace)
Hypothesized
Trace
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.999042
253.5240
95.75366
0.0000
At most 1 *
0.924083
114.5189
69.81889
0.0000
At most 2 *
0.840569
62.95660
47.85613
0.0010
At most 3
0.559738
26.23368
29.79707
0.1218
At most 4
0.368722
9.825991
15.49471
0.2944
At most 5
0.030805
0.625799
3.841466
0.4289
Trace test indicates 3 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.999042
139.0052
40.07757
0.0001
At most 1 *
0.924083
51.56229
33.87687
0.0002
At most 2 *
0.840569
36.72292
27.58434
0.0026
At most 3
0.559738
16.40769
21.13162
0.2019
At most 4
0.368722
9.200192
14.26460
0.2699
At most 5
0.030805
0.625799
3.841466
0.4289
Max-eigenvalue test indicates 3 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
4.3 Granger Causality Test
Since there was only three variables have existence of cointegration, we try to carry out the granger causality test. The result is reported in table4 below.
Table 3: Granger Causality test
Null hypothesis
F-Statistic
Prob.
LNM2 does not Granger Cause LNO
7.96823
0.0044
LNO does not Granger Cause LNM2
0.11412
0.8929
LNGDP does not Granger Cause LNO
9.35021
0.0023
LNO does not Granger Cause LNGDP
0.93350
0.4149
LNEXPORT does not Granger Cause LNO
10.6758
0.0013
LNO does not Granger Cause LNEXPORT
1.27523
0.3080
INFLATION_RATE does not Granger Cause LNO
1.50408
0.2539
LNO does not Granger Cause INFLATION_RATE
0.37942
0.6906
EXCHANGE_RATE does not Granger Cause LNO LNO does not Granger Cause EXCHANGE_RATE
0.38474 1.67916
0.6872 0.2198
LNGDP does not Granger Cause LNM2
LNM2 does not Granger Cause LNGDP
2.40497
4.79738
0.1055
0.0146
LNEXPORT does not Granger Cause LNM2
LNM2 does not Granger Cause LNEXPORT
2.24840
2.44960
0.1210
0.1014
INFLATION_RATE does not Granger Cause LNM2
LNM2 does not Granger Cause INFLATION_RATE
0.87620
7.69642
0.4256
0.0017
EXCHANGE_RATE does not Granger Cause LNM2
LNM2 does not Granger Cause EXCHANGE_RATE
0.47237
5.62952
0.6276
0.0077
LNEXPORT does not Granger Cause LNGDP
LNGDP does not Granger Cause LNEXPORT
0.94100
0.26969
0.4002
0.7652
INFLATION_RATE does not Granger Cause LNGDP
LNGDP does not Granger Cause INFLATION_RATE
0.35717
4.68142
0.7022
0.0160
EXCHANGE_RATE does not Granger Cause LNGDP
LNGDP does not Granger Cause EXCHANGE_RATE
1.45569
5.12881
0.2474
0.0113
INFLATION_RATE does not Granger Cause LNEXPORT
LNEXPORT does not Granger Cause INFLATION_RATE
0.60222
1.64690
0.5533
0.2076
EXCHANGE_RATE does not Granger Cause LNEXPORT
LNEXPORT does not Granger Cause EXCHANGE_RATE
1.15607
5.14944
0.3268
0.0111
EXCHANGE_RATE does not Granger Cause INFLATION_RATE
INFLATION_RATE does not Granger Cause EXCHANGE_RATE
9.39794
1.73255
0.0006
0.1921
There are relationship found to exist between, first, LNM2, LNGDP and LNExport with LNO (dept currency crises). Second, LNGDP with LNM2 and LNM2 with LNGDP. Third, LNexport with LNM2 and LNM2 with LNExport. Forth, LNM2 with inflation rate. Fifth, LNGDP with inflation rate. Next is LNGDP with exchange rate. The next one is, LNexport with exchange rate and the last one is inflation rate with exchange rate.
The estimation result indicated that we reject the null hypothesis for both LNGDP and LNM2 and LNexport and LNM2 then we conclude that there exist bidirectional causality between GDP with M2 and export with M2 at 5% level of significant. There was statistical insignificant relationship found to exist between inflation rate with LNO(dept currency crises), exchange rate with LN0, LNexport with LNGDP and LNinflation rate with LNexport
5.0 Conclusion
In this paper we try to explain the currency crises by using some variables that we assume as crucial indicators. The stationary property of the data and the order of integration of the data were tested using both Augmented Dickey-Fuller (ADF) test and the Phillip-Perron (PP) test. Since we used single education model(s), the application of Johansen multivariate approach to cointegration was necessary test for the long run relationship among the variables. The result showed there are three variables existence of cointegration among the variables tested. The granger causality test shoes a bidirectional relationship between GDP (LNGDP) and M2 (LNM2) and also a bidirectional relationship between export (LNexport) and M2(LNM2). The result also shows that the causalation between inflation rate and depth of currency, exchange rate and depth of currency, export and GDP, and also inflation rate and export were statically insignificant.
Appendix
Notes on Variables:
M2= Money Supply
GDP= Gross Domestic Product
