In the Name of Allah, the Most Gracious, the Most Merciful
Grande Strategy

The Financial Deepening and Mitigating Risks of Currency Crises in Indonesia

1/27/2011
Dimas Kusuma
Dimas Bagus Wiranata Kusuma
Resident Economist
Email: d.kusuma@grandestrategy.com 

Download paper in PDF format (highly recommended)


ABSTRACT

Background – Financial liberalization and development has been taking place and major interesting phenomena in many developing as well as developed countries. One of most impetus factor to liberalize and integrate their markets is the presumption that such development in financial sector can push up in promoting growth. Subsequently, the turning of economic growth actually can be transmitted through financial intermediation or bank based financing. The present paper is attempting to explore several problems that may be closely related to above topic and expected to capable of generating accommodated conclusions at the end. The present paper would employ some econometric approaches by running several variables as economic indicators. In general, this paper is going to evolve several possible variables which are envisaged significantly effect in accommodating the financial deepening in regards on crisis, particularly currency crises. Inflation (INF), Real Effective Exchange Rate (REER), ratio money supply towards reserve (M2/Res), and Real Credit Growth (RKG) are correspondingly utilized as predictors in looking on the impact of financial deepening on currency crises occurrence.

The present paper ends with some concluded statements as follows (1) financial deepening would lead to a higher incidence of growth inhibiting currency crises due to it might be excessive. The paper comes up with the suggestion that surely the link between finance and growth is seemingly more complex than the single relationships suggest. Therefore, it would appear that financial deepening should be completed by developing associated legal and regulatory institutions so that policies for financial reform and regulation could be well-designed and no longer positively impact to financial stability in financial system.           

Key Words: Financial Stability, Currency Crises, Financial Deepening, and Mitigating Risks
JEL: E44, F15, G01




INTRODUCTION
“What improves the circumstances of the greater part can never be regarded as inconveniency to the whole. No society can surely be flourishing and happy, of which the far greater part of the members are poor and miserable.”
Adam Smith (The Wealth of Nations, Book I Chapter VIII, p.96, Para. 36.)

The global financial has triggered in every aspect of monetary, financial economics and policy. The current issue now, financial liberalization and development have been taking place and major interesting phenomena in many developing as well as developed countries. One of most impetus factor to liberalize and integrate their markets is the presumption that such development in financial sector can push up in promoting growth. Subsequently, the turning of economic growth actually can be transmitted through financial intermediation or bank based financing. Walter Bagehot (1873) and Joseph A. Schumpeter (1912) emphasize the crux of banking role in economic growth.

Indonesia, which is currently categorized as emerging economies, has been transforming its economy in order to maintain a sound economic growth. Therefore, during stabilization period (1969-1996), Indonesia did several reforms and structural changes. In 1983, credit reform began in which the artificial restrictions on the allocation of bank credit and the state bank loans were eliminated. Moreover, financial reform was taken one step further in October 1988 with what referred to as Pacto 88. Under this regulation, restrictions on the operations of foreign banks were eased; the procedures for establishing branch banks were simplified. Pakto 88 also reduced the special privileges and responsibilities of the state-owned financial institutions and narrowed the differential tax treatment affecting various financial instruments. On that sense, the financial deepening was starting to establish by further controlling bank credit in the light of the surge in capital inflows. In brief, before financial crises in 1997, Indonesia had liberalized its economy to expand financial extension for finally fostering economic growth.


TABLE 1.1
  
Source: Bank Indonesia (Various Sources)

However, since economic was overheating, a credit boom had triggered an excessive depreciation of the rupiah. The economy shrank 13,8% during 1998 and annual rate of inflation reached a very high, 77,6% in 1998 (year on year basis). Generally, speaking the process of financial deepening during stabilization and liberalization had ceased since financial crises toke place. Nevertheless, some economists suggest that financial deepening basically has a positive effect on growth if not done to excess (Rousseau, 2007).

Rapid and excessive deepening, as manifested in a credit boom, can be problematic even in the most developed markets because it can both weaken the banking system and bring inflationary pressures. We had experience in financial crises turmoil in 1997/1998 when Indonesia suffered economic crises. At that time, banking system failed to safeguard economic system and also incapable of providing sufficient liquidity needed to take away from crises. It implies to enhancing capital outflows and reserve was depleted tremendously to keep buffering the economy. The excessive credit growth is the outcome of financial liberalization and officially has been inspired by Washington consensus that may have altered the basic structural relationship between finance and growth. Many countries have competed to apply and accelerate financial deepening as the benefits may be got from it. However, such condition has led many countries, unfortunately, to liberalize its economy as pre-condition attaching with financial deepening without developing the associated legal and regulatory institutions. As a consequence, the impact of financial deepening on growth would become smaller and remain a premature financial development. That condition, on contrary, will spur a higher inflation rate and interest rate and finally vulnerable to financial crises. Figure below clearly shows the secular increase in inflation rate and its impact to trigger exchange market pressure during 1997/1998 in Indonesia.   

In contrast, not all economists prescribe and stand in the view that financial deepening negatively affect to financial crises and financial development. As noted by Singh and Weiss (1998) and Diaz Alejandro (1985) that the de-regulation of repressed financial market can impose risks of financial collapse and lead to economy recession. The liberalization financial markets have been occurring in many developing. Country has gained its goal with deepen and force increase in financial depth. Some experiences suggest that even the best approach to regulation and supervision will not be able to completely avoid incidences of financial crisis. An efficient crisis management strategy requires a lender and fiscal support. In our paper we will explain it, detail about this problem, how to free from financial crisis and is it any effect from economy deepening in Indonesia.
 


TABLE 1.2
 
Source: Bank Indonesia (Various Sources)

According to above description, we can get the idea that the relationship between financial deepening and crises might be unstable and fluctuating. Therefore, we would basically investigate further on the relationship between financial deepening and the onset of currency crises. Later, it also continues to conceive on establishing of what possible variables that significantly influences the existing of financial deepening in Indonesia during period of observation. Lastly, the current paper deliberates on the role of early warning system of currency crises approach in mitigating and specifically detecting in such particular period of observations.

The organization of the present paper is as follows. The next section provides some literature review on the articles of financial deepening on financial crisis. Section III describes the data and empirical approach. Section IV presents results from our VAR and logit analysis. We first look at dynamic responses of financial deepening to financial crises; we then estimate a VAR model to investigate what variables affect the process of financial deepening; and finally by logit model we discern the probability of each variable observed in triggering to financial crisis. To end, section V concludes the paper with a summary of the main findings and some concluding remarks. 


LITERATURE REVIEW
“Why is financial deepening a good thing?” this is one of the issues mentioned by the Deputy General Manager of the BIS, Hervé Hannoun, in the middle of 43rd SEACEN Governors’ Conference 2008. He also mentioned that economist have long insisted on the need to develop strong and deep financial markets in order to promote economic growth. This is because of the key importance of having market-based and diversified channels of intermediation between ultimate borrowers and investors.
Walter Bagehot (1873) and Joseph A. Schumpeter (1912) emphasize the crux of banking role in economic growth. They, then, highlight circumstances when banks can actively spur innovation and future growth by identifying and funding productive investments. Thus, the first prerequisite stage imposed to assure economic growth is that the availability services provided by financial markets which commensurate with the need to finance productive and investment projects in economy. Such condition is capturing a situation of the existence of financial stability in the system.
Generally speaking, the widespread liberalization of financial markets that have been occurring in many developing countries has gained its goals to deepen and force increases in financial depth. In other hand, financial liberalization may lead to a rapid and excessive financial deepening implementation as the further outcome, and it is indicated has tendency in leading to the creation of premature financial development. A premature financial development tends to have a higher risk in facing and creating any possible shocks which readily would be transmitted largely to the financial stability if done to excess and it is fundamentally leads to financial crises. Rousseau and Wachtel (2007) further argue that excessive financial deepening spread into economy may be a result of financial liberalization without completing the associated legal and regulatory institutions which were sufficiently well developed. As a consequence, the impact of financial deepening on growth would become smaller, and ultimately create premature financial development which fundamentally leads to financial crises. Singh and Weisse (1998) and Diaz-Alejandro (1985) also noted that the de-regulation of repressed financial markets can impose risks of financial collapse and consequently, lead to economic recessions.
James (2009) tries to relate the impact of financial liberalization with the innovation and knowledge creation. He notes that the literature suggest that financial reforms may negatively influence innovations in several different theoretical settings. It may do so by weakening the incentives to save and this reducing the domestic resources available for facilitating invention, by producing more instability and triggering financial crises that exacerbate economic fluctuations and dampen knowledge creation, or by enabling the financial sector to expand disproportionately and become more profitable and thus leading to more talent being attracted from the technology sector. He also noted from his findings that financial liberalization relocates talent to the financial system, thus hurting the technology sector. There is also clear evidence supporting the view that financial liberalization retards technological development through inducing financial instability.
In brief, financial deepening will promote growth as long as it is not excessive and those fundamental variables are expected would advocate spelling out precisely the route of financial deepening incurred in the economy whether deteriorate the system or share advantages to financial stability. James (2009) mentioned that whilst financial liberalization appears to have a harmful impact on ideas accumulation, this negative effect can potentially be mitigated through improving the quality of institutions with more effective regulation and supervision.       




EMPIRICAL MODEL AND ESTIMATION PROCEDURES
Data
The analysis is conducted using quarterly data series for the period 1990 quarter 1 to 2010 quarter 3. We are employing several variables to capture the behavior of crisis and financial deepening. Ratio broad money (M2) over Reserve (M2/Res), domestic real credit (RKG), Inflation (INF), and Real effective exchange Rate (REER) are used to further investigate theirs relationships to explain the existing crises and the process of financial deepening. All data are expressed in logarithmic forms, with the exception of inflation, and ratio M2 over reserve (data used is official reserve minus gold). All data are from International Financial Statistic and Bank Indonesia data base in various sources. In addition, to measure crisis, we use the exchange market pressure (EMP) which are comprised of exchange rate variation and reserve variation. And financial deepening is measure from the ratio broad money supply over GDP nominal.
Estimation Procedures   
It is the purpose of the authors to further investigate the directional relation between the existing financial deepening and its possibility leads to crises. The definition of crises here is that the occurrence of currency crises. Our aim is to outline whether in Indonesia the process of financial deepening is likely leads to a currency crises. Thus, our empirical model is specified as follows:
Crisest
Where, crises is the currency crises and financial deepening is the ratio of money supply to GDP. Subsequently, Kaminsky and Reinhart, 1996 used some leading indicator to monitor the possible movement in the sense of measuring unusual behavior in the period preceding a crisis. For Indonesia, the study could only highlight four possible leading indicators to be incorporated into analysis. We propose such variables, namely ratio money supply to reserve, Domestic Real Credit Growth, Real Effective Exchange Rate, and Inflation as our leading indicators. Those selected variables are utilized as departing from study conducted by Susatyo (2002) which had examined several macroeconomic variables and finally obtained keenly 4 out of 20 indicators correspondingly related with the occurrence of currency crises.
Definition of Currency Crises
According to the study conducted by Kaminsky et. Al, “A crisis is defined as a situation in which an attack on the currency leads to a sharp depreciation of the currency, large decline in International reserve of a combination of the two. A crisis so defined includes both successful and unsuccessful attacks on the currency.” (Kaminsky et al, 1997, page 15)
Thus, the currency crisis is more specific defined as the unusual behavior of Exchange Market Pressure (EMP). In addition, Goldstein, Kaminsky and Reinhart (2000) further define EMP as weighted average of exchange rate changes ( ) and rate of change of the reserve (SRt). The weighted which are selected encompass two same of component index with sample volatility. Therefore, EMP is formularized as follows.
EMP=
Where,     And 
The formula above shows the exchanges the reserve and the exchange rate; each of them is positively associated with Exchange Market Pressure (EMP). Thus, an economy justified as crises if EMP exceeds the average plus deviation standard, let’s say we set the threshold as m. if EMP entails the average index of EMP and  Shows deviation standard from EMP index, hence formally we may say that currency crises will occur in a couple period of time. Therefore, the crises can be defined as
Crisest  
Hypothesis and Research Model
In order to elaborate the effect of financial deepening to mitigate crises, the present paper is attempting to present the hypothesis that ultimately can answer to the question whether financial deepening is harm for economy as a whole, in particular in Indonesia which is currently developing its economy through financial based development.
Hypothesis I:
Financial deepening has a significant impact to the occurrence of crises, in particular currency crises.
Hypothesis II
The selected macroeconomic variables or dealing indicators such as ratio M2/Reserve. Real effective ratio rate, inflation and the real credit growth are considered to be significantly effects the process financial deepening.
Hypothesis III
The selected macroeconomic variable or leading Indicators are considered and identified on the main cause in creating a higher probability of currency crises occurrence in Indonesia over period of observation.
Model Analysis
To investigate further toward the mitigation of risk, in particular over crises and in discerning the impact may be observed as the process of financial deepening. The present paper would apply two kind different approaches where each of them is used by looking at previous study as follow:
A.   VAR approach
The analysis used on the present paper is VAR as all variables which are explaining the financial deepening for sure will involve the causal nexus relationship among non-structured variable so that all variable are treated as endogen variable. According to the Sim (1980), VAR approach is enabled to do need to differentiate between exogenous and endogenous variables, and its ability to capture dynamic movement as one or more variable are reached in respond a movement from other variable.
As the variable conceived are considered as endogenous variable. Hence we may write as below expressions:
Or being expressed in the form of matrices as follows:
   =  +    +
Or: 
In addition, if there variable have unit roots, then we can exploit that there may exist co-movement in their behavior and possibilities that they will trend together toward a long run equilibrium state. Then, using the greater representation theorem, we may posit the following testing relationships that constitute a VEC model for financial deepening, as follows:
Where ��Zt contains the growth rate of the variable. The ��’s are estimable parameters. �� is a difference operator, �� is a vector of impulses which represent the unanticipated movements in Zt and is the long run parameter matrix. Then using the saved residuals from the estimation of the long run equilibrium relationship, we can estimate the error correcting as:

Above equation are representing VAR in first differences which are constituted by lt-1. And lt-1 is the value of the residual which estimates the deviation from long run equilibrium in period (t-1). a11…a55 are constituting the parameter and restriction that all ajk(i)=0 can be checked using F-test. In addition aFD, aRKG, aREER, aM2/RES, aINF are the speed of adjustment coefficients and particular interest in that they have important implication for the dynamic of the system. Their restriction can be conducted by using t-test. Asymptotic theory indicator aFD and so on converge to a t-distribution as sample size increases.

B. Logit approach
To evaluate the probability of crises as the process of financial deepening takes place, we attempt to employ logit model which p independent variables and the further study will refer to completely study conducted by Matthieu Bussiere and Marcel Fraztzscher (2002). It is then expressed by:
Where : P0 = Probability Normal regime, P1= Probability Crises Regime, Ln(P1/P0) = log odd, C1,2 = Constanta, β1 = marginal effect of the changes in independent variables Xi,t-1 for the probability of crises in the crises period relative to probability crises in tranquil/normal regime (Matthieu Bussiere and Marcel Fraztzscher, 2002)
Hence, the model specification is obtained as follows:
Crisest = Ln (P1/P0) =Z = C1 + a1M2/ReSt +a2 GKREDt +a3REERt +a4INFt +Et
Where: Ln (P1/P0) = log odd ratio which constitutes currency crises, a0 = Constanta of the model, a1,a2,a3,a4 = parameters of coefficient, Et = error term, M2/RES = ratio M2 over reserve, GKRED = the growth of real credit domestic, REER = real effective exchange rate, and INF = inflation.
Technique of Analysis
Unit Root Test and Co-integration Test
The paper will apply the commonly used tool which is called for Augmented Dickey Fuller (ADF) and Phillips-Perron unit root tests. The estimation procedure takes the following form:
(ADF-Test) :  
(PP-Test)    :  
Where  denotes lag difference of the variable under consideration. M is the number of lags and  is the error term. Based on the critical values of respective statistics, if null hypothesis cannot be rejected, then the time series are non-stationary at the level and therefore need to go through higher order differentiating process to achieve stationarity and to find the order of integration.
To test for co-integration, we employ aVAR-based approach of Johansen (1988) and Johansen and Juselius (JJ, 1990). The latter develops two tests statistic to determine the number of co-integrating vector- the trace and the maximal Eigenvalue (M.E) statistic.
A VAR-based approach is classified as a dynamic model since it can generate variance decomposition and impulse response to further examine short run dynamic interactions among the variables. Generally, time series data get stationary in first difference I (1) order of co-integration. Hence, VECM is utilized to identify the long run behavior of the variables and their short run relations and the therefore can better reflect the relationship among the variables. In addition, in a traditional VAR analysis, Luthepohl and Reimers (1992) show that impulse responses and variance decomposition analysis can be used to obtain information concerning the interactions among variables. The impulse response functions permit inferences on the direction of response of a variable of interest to be a one standard deviation shock in other variables. Meanwhile, variance decompositions indicate the percentage of a variable’s forecast error variance attributable to innovations in all variables considered in the system.      
Logit Model
Logit model is categorized as cumulative discrete function which is appropriate to explain dependent variable which constitutes qualitative response and characterized as dichotomy. Thus, to further comprehend the logit model, below expression explains as cumulative logistic probabilistic:
Overall, logit model is nonlinear model, both in parameter and in variable. Thus, we cannot use OLS to estimate logit parameter. Logit, basically, is specified as probability function. Hence, mathematically, logit is used to solve odds ratio. In addition, odd is defined as ratio of events that have happened to events that have not happened. This can be expressed as follows:
Basically, if the value of P is small, (1-P) is going to close to 1, as consequence the odd entails to zero. In other words, odd represents an indicator which explains whether a country is detected to crises or not. In short, the more odd closes to zero; it means that a country will be small in creating a probability in occurrence a crisis. 

FINDINGS
Before we proceed, it is imperative to perform a priory analysis of the variables temporal properties. We subject each time series to the standard augmented Dickey Fuller (ADF) unit root test. The results indicate that almost all data series under consideration are integrated of order 1, or I (1). That is, they are stationary in their first differences. Accordingly, we implement the Philips-Perron (PP) test for all variables under consideration. The PP test confirms the stationarity in I (1) process.
 Subsequently, we proceed with a cointegration test, as suggested by Johansen (1992) and Johansen and Juselius (1990). Essentially, the test of a VAR- based test, treating all variables as essentially endogenous. In implementing the test, we place emphasis on the pre-condition that the error terms need to be serially uncorrelated. As it can be observed from the table, we find evidence for the presence of a long run relationship among the variables in all systems we estimate. These findings provide an important guide for our dynamic modeling in respect to variables observed shocks in the variation on financial deepening, and creating a higher probability of crises, the issue that we address.
TABLE 4.1
ADF AND PP UNIT ROOT TESTS
Variable
Level
First Difference
ADF
PP
ADF
PP
EMP
-7.981003***
-5.716465***
-9.205674***
-37.86399***
INF
-4.408629***
-4.591088***
-10.58124***
-12.13751***
M2GDP
-1.864321
-1.704157
-4.692668***
-11.80703
M2RES
-1.329521
-1.402575
-6.579466***
-7.275103***
REER
-1.214832
-1.173666
-5.388646***
-6.243173***
RKG
-0.825091
-1.134339
-3.709075***
-6.411127***
*,**,*** denote significance at better that 10%, 5%, and 1%
TABLE 4.2
JOHANSEN-JUSELIUS CO-INTEGRATION TEST
Null Hypothesis
Co-integration Results
Statistical Results/Critical values (5%)
Trace
Maximal Eigen
Trace
Maximal Eigen
r=0
31.30834
31.30693
12.53
11.44
r=1
0.001418
0.001418
  3.84
3.84
The lag order specified for the test is 1, which we find sufficient to render the error term serially uncorrelated. The 5% critical values are based on Osterwald-Lenum (1992). Effective number of observation is 77
TABLE 4.3
LONG RUN CO-INTEGRATING EQUATION (NORMALIZED ON M2GDP)
Constant
EMP
-2.129314
-25.53372***
[-6.62502]
t-Statistic is in parentheses
*** Significant at 1% significance levels
   TABLE 4.4
LONG RUN CO-INTEGRATING EQUATION (NORMALIZED ON M2GDP)
Constant
INF
M2RES
REER
RKG
-64.56817
184.2111***
[ 5.70311]
-1.009500***
[-3.97680]
5.149189***
[ 3.04701]
3.450216***
[ 3.15795]

According to the EMP result, we may say that the financial deepening is affected negatively by the onset of Exchange Market Pressure. The sign is reasonable from a theoretical point of view in which the process of financial deepening would be smaller and insignificant as the financial crisis is taken place. In addition, inflation will cause positively on financial deepening as the higher inflation will encourage a higher on interest rate level. As a higher interest rate established, the marginal propensity to invest might increase and encourage creditor for saving or investing their funds in financial instrument or financial institution. Such situation will force financial institution to transmit in the several investment assets, either categorized as real assets or financial assets. Therefore, the higher of inflation rate will promote the financial depth in the economy. And then, the M2/RES seems to negatively affect the financial deepening in the long run. As noted earlier, the ratio of money supply and reserve indicates the proportion of money supply might be backed by reserve in the sake it will not be oversupply and reflect upon the monetary condition. Generally speaking, we realize that money supply actually as liquidity assets represents the debt of government to economy. Government is trying to maintain it volatility to detach from any instability matter. So that, it must be kept in a proper proportion and precisely commensurate with economic capacity. However, in other side, money supply looks like an engine of growth in the economy since many transactions and financial instrument can be transacted easily by it. In that sense, it is reasonable if it shows negative sign in affecting financial deepening. 
Subsequently, Real Effective exchange rate represents the exchange rate which adjusts for the effect of inflation on the nominal value. In other words, REER could be explained by the theory of purchasing power parity which asserts that in the long run, exchange rate adjust to different inflation rates among countries so that the relative purchasing power of various currencies is equalized. In the context of financial deepening, it would be positively sign as the inflation and the original exchange rate between Indonesia and let say, United States. Usually, the inflation rates in developing countries are higher than in developed countries. Automatically, to keep their real interest rates, the nominal rates would be higher as well. As the nominal rates higher, the capital inflows would flow and trickle down to financial and real sectors. On that reason, the financial markets and financial institutions will take dominant contribution on accelerating the process of financial deepening. Therefore, REER will positively affect in financial deepening process in Indonesia. In terms of Real Credit Growth (RKG) positively affects the financial deepening. It is reasonable since according to Joseph A Schumpeter (1912) emphasize the critical importance of the banking system in supporting financial deepening through fostering economic growth and highlight circumstances when banks can actively spur innovation and future growth by identifying and funding productive investment. In other words, real credit growth may have pushed financial deepening through financial intermediation participating. To some extent, according to Robert G. King and Levine (1993a), empirically show that the level financial intermediation is a good predictor of a long run rate of economic growth, capital accumulation, and productivity improvement. Therefore, we may say that any extensions on credit would spur accelerating of the financial deepening process. According to above explanation, empirically the observed variables all significantly effect on financial deepening process in Indonesia during period of observation.   


Granger Causality Test
The documented cointegration among the variables suggests only on their long run association and, while it implies causality, does not reveal unfortunately the directions of causation among them. Thus, in order to evaluate and observe the causal nexus among the concerned variables, we can implement the Granger causality test. As noted, with co-integration, the dynamic causal interactions among the variables should be phrased in a vector error correction form.
TABLE 4.5
BIVARIATE GRANGER CAUSALITY RESULTS

Lags: 1
  Null Hypothesis:
Obs
F-Statistic
Probability  
ECM
  M2GDP does not Granger Cause EMP
82
 0.05178
 0.82058
-0.355881*** [3.48637]

  EMP does not Granger Cause M2GDP
 1.59546
0.21026
-0.104943 [-0.30467]







Basically, the present paper would discern the relationship and causality between EMP (exchange market pressure) and M2GDP (financial deepening). Empirically, it shows one cointegration by using Johansen Juselius test, however according to bivariate causality seems not to be significant. Nevertheless, we may note the ECM coefficient suggests insignificant in the long equilibrium. We are expecting that the ECM coefficient results is to be less than zero as it would adjust upward to restore equilibrium. In that sense, with the negative coefficient of the error correction term in the M2GDP equation, it means that the increase on M2GDP (financial deepening) processing tends to increase the EMP in the long run. In addition, the coefficient suggests that deviation from the equilibrium path is adjusted by about 35% the next month through the movements in financial deepening. Indeed, the adjustment toward the long run relationship is quite high. This implies that, after any shocks that forces the financial deepening from their long run values it takes a short quite long for the EMP to return to its equilibrium values.   
TABLE 4.6
BIVARIATE GRANGER CAUSALITY RESULTS
  Null Hypothesis:
Obs
F-Statistic
Probability



  M2RES does not Granger Cause M2GDP
82
 2.99916
 0.08721



  M2GDP does not Granger Cause M2RES
 0.07708
 0.78201

  REER does not Granger Cause M2GDP
82
 2.89021
 0.09305

  M2GDP does not Granger Cause REER
 2.87365
 0.09398

  RKG does not Granger Cause M2GDP
82
 0.20408
 0.65269

  M2GDP does not Granger Cause RKG
 16.5761
 0.00011

  INF does not Granger Cause M2GDP
82
 1.06898
 0.30433

  M2GDP does not Granger Cause INF
 1.18692
 0.27926

  REER does not Granger Cause M2RES
82
 4.72201
 0.03278

  M2RES does not Granger Cause REER
 0.00137
 0.97053

  RKG does not Granger Cause M2RES
82
 1.16151
 0.28443

  M2RES does not Granger Cause RKG
 0.89225
 0.34775

  INF does not Granger Cause M2RES
82
 0.38913
 0.53456

  M2RES does not Granger Cause INF
 0.07933
 0.77894

  RKG does not Granger Cause REER
82
 3.90339
 0.05168

  REER does not Granger Cause RKG
 3.20912
 0.07706

  INF does not Granger Cause REER
82
 0.39690
 0.53051

  REER does not Granger Cause INF
 0.34220
 0.56023

  INF does not Granger Cause RKG
82
 4.25597
 0.04240

  RKG does not Granger Cause INF
 0.18175
 0.67104











According to bivariate granger causality above, we may note that only M2/Res, and REER which have causality relationship to M2/GDP (financial deepening). This means that there is a unidirectional causality from two above variables to M2/GDP and it does seem to confirm on what have mentioned earlier that they are consistent enough in promoting and corroborating the process of financial deepening. And the remaining, inflation (INF) and Real Credit (RKG) are not significant since done to excess. Since according to Peter Rousseau (2007) the financial crises created as rapid and excessive deepening, reflected in a credit boom, has weakened the banking system and brought inflationary pressures so that finally any policies related with them are not extensively stepping up the pace of financial deepening et all.
IRF and VDC
Impulse response functions and variance decompositions capture estimation results of the VAR in the forms that can be easily interpreted. Essentially, impulse response functions trace temporal responses of a variable to its own innovation and innovations in other variables in the system. From the functions, we can observe whether the response financial deepening is significant or persistent. Meanwhile, variance decompositions indicate the fraction of a variable’s forecast error variance attributed to shocks in other variables and, accordingly, provide a natural measure of the of the relative importance of various shocks to the variable of interest. From variance decompositions, we can assess REER, INF, RKG, M2/RES, and EMP are more dominant in accounting for promoting financial deepening (M2/GDP) in the system of VAR.    


FIGURE 4.1
GENERALIZED IMPULSE RESPONSE FUNCTIONS M2/GDP AND THE OTHER CONCERNED VARIABLES

Above figure presents the generalized impulse response functions generated from VAR model for the system consisting all variables observed under consideration. The responses are plotted together with a 2-standard deviation band. If the band does not include the zero line, then the responses are said to be significant. We try to observe the responses of M2/GDP (financial deepening) as any shocks appear on each variable observed. Above figure suggests insignificant responses of any shocks come from REER, INF, M2/RES, and RKG to M2/GDP. In brief, we may testify that no significant Reponses and acerbated financial depth would be occurred in the short run. It confirms the earlier findings that financial deepening process in Indonesia take a quite long time to be in existence. And it might conform to Lucas (1975) which says the financial deepening in some developing countries, in particular, is not well-completed in terms of the associated legal as well as regulatory institutions. Thus, such condition will remain a premature financial development so that it might be vulnerable to financial crises. To that sense, above concerned variables seem to be not significant as in the short run, there would be happened an adjustment process in order to maintain the stability and sustainability in financial deepening process in the long run. However, to gain the further insight on the relative importance of the concerned variables innovations to financial deepening process, we compute variance decompositions up to 30-month horizons.
TABLE 4.7
VARIANCE DECOMPOSITION OF M2GDP
Period
Explained by Innovation in
INF
EMP
M2/RES
RKG
REER
M2/GDP
2
0.521865
0.226545
0.220054
0.035604
0.000826
98.99511
4
0.344953
0.418317
1.065429
0.254723
0.001075
97.91550
8
0.946646
0.631135
2.400091
0.957072
0.079351
94.98571
12
1.560786
1.545328
2.505848
1.543311
0.448535
92.39619
16
2.336302
2.696252
2.398632
1.844612
1.103030
89.62117
20
2.984174
3.819896
2.562311
1.879464
1.834115
86.92004
24
3.353417
4.663126
2.965547
1.835831
2.419526
84.76255
28
3.470407
5.180948
3.380435
1.916042
2.767767
83.28440
30
3.470800
5.341316
3.536517
2.036861
2.862287
82.75222

Several interesting results emerge from the variance decompositions. Reaffirming the results from the impulse response functions, variations in the M2/GDP (financial deepening) are accounted mainly by its own shocks and indeed its variations are explained in a negligible portion in regards to the concerned variables. In the short run, after 30-month horizons, variations in the M2/GDP account for only 2-5% of the forecast error variance on variables under consideration. Therefore, we note that there is the relative irresponsiveness the concerned variables to the process of financial deepening; perhaps the economy cannot utilize those instruments effectively in the short run. And also the economy is less capable of spurring financial instruments as explained before Indonesia’s economic system might be less developed and less well supported by the legal associated and the sufficient regulatory institutions. At that reason, the financial deepening is found difficult to expand even though the concerned variables have been run away.
In sum, there are insignificant short run dynamics that we observe tend to answer the question on the role of the concerned variables to affect financial deepening. However, it does not mean that those variables are not important and relevant to be put into our consideration. Since, in the short run, the economy might not be reacted much they could be as being expected in the financial deepening does not reveal too rapid and excessive deepening, and then hopefully can maintain the stability in banking system, and avoid inflationary pressures, and move away from financial crises.       
Application Logit Model in Crises Modeling in Indonesia
Logit regression (logit) analysis is a uni/multivariate technique which allows for estimating the probability that an event occurs or not, by predicting a binary dependent outcome from a set of independent variables. Further, by using logit estimation, we can obtain the specific results on what the concerned variables trigger the currency crises in the form of probability term. The variables used, namely Inflation (INF), Real credit Growth (RKG), ratio M2 over Reserve (M2/Res), and real effective exchange rate (REER) are selected variables which are resulted as empirical research had been completely done by Susatyo (2000) in which he uses 20 leading indicators. Subsequently, by extracting signal approach developed by Kaminsky and Reinhart (1999), he gets those 4 variables. According to signal approach that he did, those 4 variables are able to show 50% probability in triggering a currency crisis in Indonesia. Later, the present paper is aiming to utilize those 4 variables but using the parametric approach, namely logit.
The preceding model involves 83 observations, comprised of 6 observations and 77 observations are grouped as crisis and non-crisis respectively.
TABLE 4.8
REGRESSION RESULTS OF LOGIT MODEL
DEPENDENT VARIABLE: KRISIS
METHOD: ML-BINARY LOGIT
CONVERGENCE ACHIEVED AFTER 7 ITERATIONS

Variable
Coefficient
Std. Error
z-Statistic
Prob. 
C
16.41122
13.71625
1.196480
0.2315
M2RES
0.438312
0.187536
2.337217
0.0194
REER
-2.841761
1.456331
-1.951316
0.0510
RKG
-0.661532
1.282463
-0.515830
0.6060
INF
9.866552
8.481509
1.163302
0.2447
Mean dependent var
0.072289
    S.D. dependent var
0.260540
S.E. of regression
0.257418
    Akaike info criterion
0.551720
Sum squared resid
5.168598
    Schwarz criterion
0.697433
Log likelihood
-17.89637
    Hannan-Quinn criter.
0.610259
Restr. log likelihood
-21.54020
    Avg. log likelihood
-0.215619
LR statistic (4 df)
7.287661
    McFadden R-squared
0.169164
Probability(LR stat)
0.121445



Obs with Dep=0
77
     Total obs
83
Obs with Dep=1
6










Now, let us interpret the regression results given above estimation results. Basically, each slope coefficient in this equation is a partial slope coefficient and measures the change in the estimated logit for unit change in the value of the given regressor (holding other regressor constant). Thus, the M2/Res coefficient of 0,438312 means, with other variables constant, that if M2/Res increases by a unit, on average the estimated logit increases by about 0,44 units, and then suggesting a positive relationship between the two. The INF coefficient also suggests a positive relationship. Conversely, REER and RKG show negative relationships between the two. In terms of significance, M2/Res and REER seem to be significant at 10% level. However, together all the regressors have an insignificant impact on the crisis, as the LR statistic is 7,28, whose p value is about 0,12, which is not significant at 10% level.
Nevertheless, the aim of logit is not to signify whether the regressors are significant or not, but instead of showing for estimating the probability that an event occurs or not. Thus, as noted previously, a more meaningful interpretation is in terms of odds, which are obtained by taking the antilog of the various slope coefficients.
The above results show that intercept coefficient is 16,41. This means
With other variables held constant. Thus, the value of or the probability of occurrence the crisis, held other variables constant is that = 0,99999. In other words, the probability of crisis occurrence with other variables constant is that 99,99%. That seems to be higher and indirectly indicate the crises are dominantly caused by other variables out of the concerned variables above, for example the contagion factor. In addition, the slope of M2/Res is 0,438312. Therefore, the value of probability in occurring crisis if M2/Res tends to fluctuate is that  = 0,607856. Thus, the probability of occurring crises as variables M2/Res is fluctuating is 60,7856%. The further analysis could be, as the present has mentioned in previous chapter, that
P krisis = , P normal                           = .
= . , therefore,                          = 60,7856%.
It obviously shows the risk of crises when M2/Res is fluctuating. In that sense, the M2/Res fluctuation will expound 60,78 times compared as normal condition it does.
Now, we look at the inflation (INF) variable which the slope is 9,866552. That means the probability of crises will be occurred as the inflation is fluctuating is that  or equal with 99.99%. and in terms of REER, we note the slope is -2,841761. Thus, the probability of occurring crises as it is fluctuating, is that,  or equal with 0,05511. This means any shocks or fluctuations in REER variable will force the probability of crises for about 5,511%. Lastly, we compute the probability of crises as real credit (RKG) is fluctuating. Using the antilog of the RKG coefficient, we will get or equal with 0,3404. It means any shocks or fluctuations in credit variable will push up the probability of crises for 34,04%
TABLE 4.9
THE SUMMARY OF LOGIT MODEL ESTIMATIONS
Variable
Parameter
Odd Ratio
Probabilities
Intercept
16.41122 (13.71625)
0,9999
0.2315
RKG
-0.661532 (1.282463)
0,3404
0.6060
INF
9.866552 (8.481509)
0,9999
0.2447
M2/RES
0.438312 (0.187536)
0,60785
0.0194**
REER
-2.841761 (1.456331)
0,05511
0.0510***
Source: Eviews 4
Note: ** and *** are the level of significance at 5% and 10% respectively. The parentheses show the standard errors





CONCLUSIONS
The present paper analyzes three issues in looking at the relations between financial deepening and the probability of occurrence of crises, in particular, currency crises in the case of Indonesia. By also employing several variables under consideration, the present paper is able to identify empirically on what variables importantly effect to financial deepening and theirs relation to crises. Initially, we ensure whether data used are stationer and integrated, and we find that our data stationer in first difference I(1) by using ADF and PP tests. Then, according to Johansen and Juselius test, the data happen to be found evidence for the presence of a long run relationship among the variables. And then by establishing the long run co-integrating equation, we suggest that the financial deepening is affected negatively by the onset of Exchange Market Pressure (EMP). The sign is reasonable from a theoretical point of view in which the process of financial deepening would be smaller and insignificant as the financial crisis is taken place. The other findings related according to the empirical research as follows:
a.      Inflation will cause positively on financial deepening as the higher inflation will encourage a higher on interest rate level
b.      The M2/RES seems to negatively affect the financial deepening in the long run
c.      REER will positively affect in financial deepening process in Indonesia
d.     Real Credit Growth (RKG) positively affects the financial deepening
Therefore, according to long run co-integrating equation, empirically the observed variables all significantly effect on financial deepening process in Indonesia during period of observation.   
The causality test suggests finding evidence that there is independence causality from EMP and M2GDP. However, the M2GDP coefficient shows negatively significant in the long run. It means that the increase on M2GDP (financial deepening) processing tends to increase the EMP in the long run. In addition, the coefficient suggests that deviation from the equilibrium path is adjusted by about 35% the next month through the movements in financial deepening. Indeed, the adjustment toward the long run relationship is quite high. This implies that, after any shocks that forces the financial deepening from their long run values it takes a short quite long for the EMP to return to its equilibrium values. Subsequently, the other variables show in respect to M2/GDP (financial deepening), is that, M2/Res and REER have a unidirectional causality to M2/GDP, and conversely, INF and RKG is not significant in causal directional relationship to financial deepening (M2/GDP).
Meanwhile, the IRF suggests that above concerned variables (REER, INF, RKG, M2/RES) seem to be not significant as in the short run, there would be happened an adjustment process in order to maintain the stability and sustainability in financial deepening process in the long run. The variance decomposition confirms the earlier findings that suggest in the short run, after 30-month horizons, variations in the M2/GDP account for only 2-5% of the forecast error variance on variables under consideration. Therefore, we note that there is the relative irresponsiveness the concerned variables to the process of financial deepening; perhaps the economy cannot utilize those instruments effectively in the short run.
According to logit results, we get empirical evidence that only M2/Res and REER significantly effect on crises. However, the aim of logit is not to signify whether the regressors are significant or not, but instead of showing for estimating the probability that an event occurs or not. Thus, as noted previously, a more meaningful interpretation is in terms of odds, which are obtained by taking the antilog of the various slope coefficients. The odds show the probability given for a crisis occurrence as one of concerned variables is fluctuating. The M2/RES and INF seems to show a significant probability since their probability are above 50%, meanwhile the remaining, RKG and REER are below 50%. However, the concerned variables tend to fluctuate and effect to a higher probability crises occurrence as shocks imposed to them.
Finally, the present paper contains the recommendations in regard to financial deepening process, is that, financial deepening does not reveal an excessive growth and liberalization as pre-condition of fueling financial deepening should be attributed with sufficient on the associated legal and regulatory institutions so that financial development produced is not premature and does not lead to financial crises which exacerbate and deteriorate the economy as well as all aspects of human beings. Therefore, the synergic and harmonious policies need to be well developed to promote financial deepening which can positively support the economic growth and stability within the country as well as international economy.  






REFERENCES
Adiningsih, S., D.N. Setiawati, dan Sholihah, 2002, Early Warning Systems For Macroeconomic Vulnerability  in Indonesia, Final Report, EADN Regional Project.
Aizenman, Joshua. 2002. Predicting Financial Crises : Early Warning Systems. Paper Presented Univ. of sao Paolo Brazil by World Bank Institute
Andreou, Irene, Gilles Dufrenot, Alain Sand-Zantman, and Aleksandra Zdzienicka-Durand. 2006. A Forewarning Indicator System For Financial Crises : The Case of Six Central and Eastern European Countries. Journal of Economic
Ardıc Oya Pınar and H. Evren Damar (2006), “Financial Sector Deepening and Economic Growth: Evidence From Turkey”, November.
Asian Development Bank (ADB). 2005. Early Warning System For Financial Crises : Application to East Asia. New York : Palgrave Macmillian
Bank Indonesia. 2007. IMF dan Stabilitas Keuangan Internasional : Suatu Tinjauan Kritis. Jakarta : PT Elex Media Komputindo
Bill and Melinda Gates Foundation (2005), “Financial Deepening Challenge Fund Strategic      Project Review”, December.
Budsayaplakorn, Saksit, Sel Diboooglu, and Ike Mathur. Can macroeconomic Indicators Predict a Currency Crisis? Evidence from Selected Southeast Asian Countries. Journal of Economic Department of Finance Southeast Illinois University
Bussiere Matthieu, and Marcel Fratzscher. 2002. Toward A New Early Warning System of Financial Crises. European Central Bank Working Paper No. 145
Dornbusch, Rudiger, Stanley Fischer, dan Richard Startz. MAKROEKONOMI. New York
Fritz, Richard G., “Time Series Evidence of the Causal Relationship Between Financial Deepening and Economic Development”, Journal of Economic Development.
Gie, Kwik Kian. Gonjang-Ganjing EKONOMI INDONESIA. PT Gramedia Pustaka Utama : Jakarta
Ginebri, Sergio et all, “Financial Deepening, Trade Openness and Growth: a Multivariate Cointegrated Analysis of the Complementary Effects”.
Gujarati, Damodar. 1978. EKONOMETRIKA DASAR. Terjemahan. Jakarta : Erlangga
Herrera, Santiago dan Conrado Garcia, 1999, User’s Guide to an Early Warning System for Macroeconomics Vulnerability in Latin American Countries, Paper prsented in the XVII Latin American Meeting of the Econometric Society, August, Cancun.
Husman, A. Jardine. 2006. Dampak Fluktuasi Nilai Tukar terhadap Output dan Harga : Perbandingan Dua Rezim Nilai Tukar. Working Paper No 17, Bank Indonesia
Jacobs, Jan P.A.N., Gerard H. Kuper, and Lestano. 2004. Financial Crisis Identification : A Survey. Working Paper, Department Of Economics, University Of Groningen, Netherland
Jeong, Hyeok and Robert M. Townsend (2007), “Sources of TFP Growth: Occupational Choice and Financial Deepening”, Economic Theory (2007) 32: 179–221 DOI 10.1007/s00199-006-0201-8, April.
Kaminsky, L., Graciela. 2003. Varieties Of Currency Crises. National Bureau Of Economic Research Working Paper
Kaminsky, Graciela, Saul Lizondo, dan C.M. Reinhart, 1997, Leading Indicators Of Currency Crises, Maret, IMF Staff Papers Vol 5. No. 1, pp.1-48, Washington DC: IMF
Khan, Mohsin S. (2002), “Inflation, Financial Deepening, and Economic Growth”, Banco de Mexico Conference on Macroeconomic Stability, Financial Markets and Economic Development, Mexico City, November.
Kardoyo, Hadi dan Mudrajad Kuncoro. 2001. Analisis Kurs Valas dengan Pendekatan Box Jenkins : Studi Empiris Rp/US$ dan Rp/Yen, 1983.2-2000.3. Department of Economic of Gajahmada University
Kiyotaki,Nobuhiro and John Moore (2005), “Financial Deepening”, Journal of the European Economic Association, May.
Krznar, Ivo. 2004. Currency Crisis : Theory and Practice with Application to Croatia. Working Paper 12, Croatian National Bank
Lestano, Jan Jacobs, dan Gerard H. Kuper, 2003, Indicators Of Financial Crises do Work!: An Early Warning System for Six Asian Countries, CCSO Working Paper 13, Department Of Economics, University Of Groningen, Netherland
Lestano, Jan Jacobs, dan Gerard H. Kuper, 2004. Currency Crises in Asia : A Multivariate Logit Approach, CCSO Working Paper 13, Department Of Economics, University Of Groningen, Netherland
Lynch, David (1996), “Measuring Financial Sector Development: A Study of Selected Asia-Pacific Countries”, The Developing Economies, XXXIV-1, March.
Mankiw, N. Gregory. 2003. Teori Makroekonomi. Edisi Kelima. New York and Basingstoke : World Publishers McGraw-Hill Companies
Nachrowi, Djalal Nachrowi, dan Hardius Usman. 2002. PENGGUNAAN TEKNIK EKONOMETRI. Jakarta : PT RajaGrafindo Persada
National Bank of Rumania. 2006. Early Warning System : The Case Of Rumania. Occasional Paper No. 5
Nzotta, Samuel Mbadike and Emeka .J. Okereke (2009), “Financial Deepening and Economic Development of Nigeria: An Empirical Investigation”, African Journal of Accounting, Economics, Finance and Banking Research Vol. 5. No. 5.
Nwezeaku, N.C et all (2010), “The Effects of Financial Deepening on Stock Market Returns and Volatility: Evidence from Nigeria”, International Research Journal of Finance and Economics - Issue 40.
Oliver, Burkart, and Virginie Coudert. 2002.  Leading Indicators of Currency Crises for Emerging Countries. Published in Emerging Markets Review, Volume 3
Pohan, Aulia. 2008. Potret Kebijakan Moneter Indonesia. Edisi Pertama. Jakarta : PT RajaGrafindo Persada
-----------------.2008. Kerangka Kebijakan Moneter dan Implementasinya di Indonesia. Jakarta : PT RajaGrafindo Persada
Rivai, Veithzal, Andria Permata Veithzal, dan Verry., N.,Idroes. 2007. Bank and Financial Institution Management : Conventional & Sharia System. Jakarta : PT RajaGrafindo Persada
Sabirin, Syahril. 2002. Peran Bank Indonesia Dalam Financial Stability. Disampaikan pada Seminar mengenai Otoritas Jasa Keuangan, di Jakarta 27 Februari 2002
Singh, Raju Jan et all (2009), “Financial Deepening in the CFA Franc Zone: The Role of Institutions”, IMF Working Paper 09/113.
Sumodiningrat, Gunawan. EKONOMETRIKA PENGANTAR. Edisi Kedua. Yogyakarta : BPFE
Tambunan, Tulus. 1998. Krisis Ekonomi dan Masa Depan Reformasi, Jakarta : PT  RajaGrafindo Persada
Tambunan Tulus. 2002. Building “ An Early Warning System” For Indonesia With The Signal Approach. Final Report, EADN Regional Project
Widarjono, Agus. 2005. EKONOMETRIKA : TEORI DAN APLIKASI. Edisi Pertama. Yogyakarta : Ekononisia
Zhuang, Juzhong, and Malcolm Dowling. 2002. Causes of The 1997 Asian Financial Crises : What Can An Early Warning System Model Tell Us? ERD Working Paper Series, Economics and Research Departement, Asian Development Bank. Forthcoming
Zina, Naceur Ben & Borhen Trigui, “Financial Deepening in economic development: theory and lessons from Tunisia”.




Vision Without Glasses

0 comments:

Post a Comment