(aeru) for asean-5 monetary inteGration: an optimum ...

ANALISIS TRIWULANAN: Perkembangan Moneter, Perbankan dan Sistem Pembayaran, Triwulan II - 2007

BULLETIN OF MONETARY ECONOMICS AND BANKING Department of Economic Research and Monetary Policy Bank Indonesia Patron Dewan Gubernur Bank Indonesia Editorial Board Prof. Dr. Anwar Nasution Prof. Dr. Miranda S. Goeltom Prof. Dr. Insukindro Prof. Dr. Iwan Jaya Azis Prof. Iftekhar Hasan Dr. M. Syamsuddin Dr. Perry Warjiyo Prof. Masaaki Komatsu Dr. Iskandar Simorangkir Dr. Solikin M. Juhro Dr. Haris Munandar Dr. Andi M. Alfian Parewangi M. Edhie Purnawan, SE, MA, PhD Dr. Buhanuddin Abdullah, MA Editorial Chairman Dr. Perry Warjiyo Dr. Iskandar Simorangkir Executive Director Dr. Andi M. Alfian Parewangi Secretariat Arifin M. Suriahaminata, MBA Rita Krisdiana, S.Kom, ME The Bulletin of Monetary Economics and Banking (BEMP) is a quarterly accredited journal published by Department of Economic Research and Monetary Policy Bank Indonesia. The views expressed in this publication are those of the author(s) and do not necessarily reflect those of Bank Indonesia. We invite academician and practitioners to write on this journal. Please submit your paper and send it via mail to: [email protected]. See the writing guidance on the back of this book. This journal is published on; January – April – August – October. The digital version including all back issues are available online; please visit our link: –http://www.bi.go.id/web/id/Publikasi/ Jurnal+Ekonomi/. If you are interested to subscribe for printed version, please contact our distribution department: Publication and Administration Section – Department of Economy and Monetary Statistics, Bank Indonesia, Building Sjafruddin Prawiranegara, 2nd Floor - Jl. M. H. Thamrin No. 2 Central Jakarta, Indonesia, Ph. (021) 2310108 / 2310408 ext. 4119, fax. (021) 3802283, email: [email protected].

1

BULLETIN of monetary Economics and banking

Volume 15, Number 3, January 2013

QUARTERLY ANALYSIS : The Progress of Monetary, Banking and Payment System, Quarter IV - 2012 Author Team of Quarterly Report, Bank Indonesia Risk Taking Behavior of Indonesian Banks : Analysis on the Impact of Deposit Insurance Corporation Establishment Moch Doddy Ariefianto, Soenartomo Soepomo

1

3

Capital Flows in Indonesia : the Behavior, the Role, and Its Optimality Uses for the Economy Fiskara Indawan, Sri Fitriani, Meily Ika Permata dan Indriani Karlina

23

The Role of Ssean EXCHANGE Rate Unit (AERU) for Ssean-5 Monetary Integration: an Optimum Currency Area Criteria Dimas Bagus Wiranata Kusuma, Syed Mohammed Abud Ashif, Ali Musa Harahap, Muhammad Alam Omarsyah

55

The Impact of US Crisis on Trade and Stock Market in Indonesia Mita Nezky

83

ANALISIS TRIWULANAN: Perkembangan Moneter, Perbankan dan Sistem Pembayaran, Triwulan III - 2012

1

QUARTERLY ANALYSIS : The Progress of Monetary, Banking and Payment System Quarter IV – 2012

1

QUARTERLY ANALYSIS: The Progress of Monetary, Banking and Payment System Quarter IV - 2012 Author Team of Quarterly Report, Bank Indonesia Indonesia’s economic growth in the fourth quarter 2012 was still going strong, although it was slower than the previous quarter. Indonesia’s economic growth in the fourth quarter of 2012 reached 6.11%, while for the whole of 2012 it reached 6.23%. Good economic growth was supported by quite strong domestic demand. Consumption and investment performance still grew strong during this quarter, though it was moderate compared with the previous period. Export performance began to show improvement in line with the economic recovery in some major trading partner countries. Imports recorded a high growth along with the strong domestic demand. Looking ahead, for the whole of 2013, economic growth is expected to reach the range of 6.3% - 6.8%. The Indonesia’s balance of payments (BOP) in the fourth quarter of 2012 improved. This is reflected in the surplus of 3.2 billion U.S. dollars during the quarter, which was higher than the previous one. Improved performance was driven by an increase in the surplus balance of payments in the Capital and Financial Transaction (CFT) that was greater than the increase in the Current Account deficit (CA). The surplus in CFT balance was supported by the sustained investor confidence, and by additional liquidity in the global financial markets resulting from monetary expansion in developed countries. Conversely, the current account had a deficit due to the slow global economic recovery amid robust domestic demand. Throughout the year of 2012, the balance of payments recorded a surplus of 0.2 billion U.S. dollars. As a result, the amount of reserves at the end of December 2012 stood at 112.8 billion U.S. dollars, equivalent to 6.1 months of imports and government’s foreign debt payments. During the year 2012, the exchange rate depreciated, yet despite this, its volatility was maintained at a relatively low level. On average, the rupiah depreciated by 6.3% (yoy) to Rp9.358 per dollar from Rp8.768 per U.S. dollar in the previous year. Meanwhile, point-to-point, the rupiah depreciated by 5.91%, and closed at Rp9.638 per U.S. dollar with a volatility maintained at the level of 4.3% (annualized). Controlled rupiah volatility is closely linked to Bank Indonesia’s policy in stabilizing the rupiah exchange rate at low levels of volatility. Inflation remained under control during the year 2012 at a low level with the inflation target in the range of 4.5% + 1%. Controlled inflation is a result of Bank Indonesia policies, supported by improving policy coordination with the Government. Inflation in 2012 was recorded

2

Bulletin of Monetary, Economics and Banking, January 2013

at 4.3% (yoy) mainly driven by stable core inflation, controlled volatility of food inflation, and low inflation administered prices. Core inflation was stable, supported by the implementation of a monetary strategy and macro-prudential policy mix to control inflationary pressures from the demand side, the price of imported commodities, and inflation expectations. In addition, subdued inflation was also supported by the more intensive coordination between Bank Indonesia and the Government through the forums of Inflation Control Team at national and regional level (TPI and TPID), especially in efforts to increase production, coordination, distribution, and food price stabilization strategies. Stability of the financial system and banking intermediation function were properly maintained. Solid industry performance was reflected in the high Capital Adequacy Ratio (CAR), which was well above the minimum 8 percent, and the maintenance of the ratio of NonPerforming Loans (NPL) gross under 5%. Meanwhile, credit growth by the end of December 2012 reached 23.1% (yoy), up from 22.3% (yoy) in the previous month. Working capital loans grew quite substantially by 23.2% (yoy), and investment credit growth was stable at a high level of 27.4% (yoy), which is expected to increase the capacity of the national economy. At the same time, consumer credit grew by 20.0% (yoy). Going forward, Bank Indonesia believes the stability of the financial system will remain intact with banking intermediation that will increase along with an increase in performance of the national economy. Solid economic performance in Indonesia cannot be separated from the support of a reliable payment system. In economic activities, the strategic role of payment systems is to ensure the implementation of various payment transactions of economic activity and other activities undertaken by both the public and private sectors. During the fourth quarter of 2012, the payment system demonstrated positive performance. This was supported by Bank Indonesia’s policy to ensure the implementation of an efficient, fast, secure, and reliable payment system. Notwithstanding the positive performance of the payment system, the circulation of money, i.e. currency outside banks as a means of payment, still played an important role in community transactions. This is reflected in the high growth of currency in circulation during the fourth quarter of 2012 along with the solid development of economic activity.

Risk Taking Behavior of Indonesian Banks: Analysis onthe Impact of Deposit Insurance Corporation Establishment

3

Risk Taking Behavior of Indonesian Banks: Analysis on the Impact of Deposit Insurance Corporation Establishment Moch Doddy Ariefianto Soenartomo Soepomo1

Abstract

This paper studies the risk taking behavior of Indonesian Banking Industry, especially before and after the establishment and the implementation of Deposit Insurance Corporation (IDIC).Using common set of explanatory variables; we test several empirical models to reveal the conduct of risk management by banks. In the spirit of BASEL II Accord, this paper take closer look at three types of risk behaviors namely credit risk, market or interest rate risk and operational risk, prior and post the establishment of IDIC. We tested the hypotheses using panel data set of banks operational in period of2000-2009. The dataset consists of 121 banks with semiannual frequency (2420 observations). Our findings show that these variables explain well the three type bank risk exposures. The implementation of IDIC alters the bank behavior albeit in somewhat different way than initially hypothesized. The risk taking responses also varies across bank types. We found that State Owned Enterprise banks (SOE)behave differently relative to the rest types of the bank. Related to size, SOE banks behave more conservative after the implementation of IDIC. On the other hand its response on conditioned capital post the IDIC implementation is the opposite; they became more aggressive. We view the public pressure on this state banks has influenced the way they manage the risk.

Keywords : Risk taking behavior, BASEL II, Deposit Insurance. JEL Classification: G11, G21, G32, C23

1 Authors are lecturer s at Faculty of Economics and Business, Ma Chung University, Malang. They can be contacted by email at [email protected] or [email protected].

4


I. INTRODUCTION Recent theory of banking includes risk management as core function of banks (Freixas and Rochet, 2008). The function takes a central attention especially in a volatile and fast changing today environment. Ideal management objective not only aims at maximizing return subject to resources constraint butalso reasonablerisk. It is a well-established preposition that in the event of asymmetric information, bank managers and or shareholders preferred higher risk portfolio in the expectation of larger return. The situation is worsened when public guarantee for third party funds exists (explicit or implicitly). The latter introduced the moral hazard problem. To overcome the problem, many authorities create a deposit insurance scheme in place of public guarantee. The mechanism may vary, nevertheless in essence it is the banks themselves that raise the funds to back up possibility of bank rush. This scheme is quite old. US Federal Deposit Insurance Corporation was established in 1933 as a response to Great Depression. Indonesia created Deposit Insurance Corporation (IDIC) in September 2005. Its creation was proposed after 1998 crisis revealed major weakness in banking system. With the absent of deposit insurance institution, the government has to bail out the failing banks (or even illiquid banks). The fiscal cost was enormous in which estimation ranged around IDR. 600 Trillion (More than USD 60 billions). Learning from the grim consequences, it is decided a semi public institution should handle the safety of funds and deal with the problems of bank liquidity and solvency. IDIC charges a flat rate of 0.01% based on monthly average deposits. Deposits coverage insured according to current law is maximum Rp. 2 Billions and must comply with deposits remuneration rule. It remains to see whether the establishment of IDIC is effective in preserving public confidence to banks. Nevertheless it must be admitted that IDIC creation is a pivotal point in history of Indonesian banking industry. The largest challenge to date faced by IDIC is the bail out of Bank Century. In the midst of 2008 global crisis, this small commercial bank failed due to weak financial standing. Under the order of Financial System Stability committee, IDIC carried out the bail out of Bank Century whose cost around Rp. 6.7 Trillion (USD 750 Million). The role of risk management is increasingly important, especially in current volatile business environment. Sub Prime Mortgage lost in US which subsequently followed by global crisis has shown that risk management in financial institution is still not adequate. The refinement known as Basel III proposal looks for stricter risk practices and higher capital as risk buffer. Even though the issue is of paramount important, interestingly there are not many empirical searches exploring the issue especially for the case of emerging countries. In this regards we hope that this study would contribute a significant value not only for scientific purpose but also for policy making and regulation.


5

In this paper we take a closer look on how banks in Indonesia practices risk management. Specifically we focus on practices between pre and post IDIC establishment. Although there are various risks inherent in banking operations but here we investigate 3 major types namely credit risk, market risk and operation risk2. Initially we aim to reveal market risk that stems from foreign exchange and interest rate movement. Nevertheless we find the data to be non-supportive for observation in foreign exchange risk taking. Around 58 banks (48% of the sample) are classified as Non Foreign Exchange Banks and Regional Banks. In these types of banks, foreign exchange business is either trivial or nonexistent. Therefore we are focusing more on appropriate class of market risk, that is interest rate risk but still maintaining the notion (that is market risk and interest rate risk are used interchangeably). We establish empirical scheme in testing relationship between these measures of risk with various factors (shareholders drive, competition, firm size, capital, charter value and macroeconomic condition). The aims of this paper first are to identify factors influencing bank risk taking behavior (credit, market and operational risk), second to reveal the pattern of relationship and possible changes related to establishment of IDIC. Further more we attempt to uncover various form of relationship according to types (BI categories), and third to elicit practical and policy implications based on study results. The next section present a brief account on IDIC as one of key turning points in Indonesian banking and also present the recent theories and empirical works in risk management. The third section discusses the methodology and the data, including the robustness check of the model, while the fourth section presents the result and analysis. Conclusion and policy implicationis presented on the last section.

II. THEORY 2.1. Indonesia Deposit Insurance Corporation Deposits insurance is introduced mostly to evade the disastrous effect of bank panics (Freixas and Rochet, 2008). The scheme gives guarantee to public that deposits could still be withdrawn in the event of bank failure. In an explicit insurance scheme, a premium is usually paid and there are several requirements for deposits withdrawal3. The scheme could be run by either a private company, semipublic or government agency. In an implicit insurance, the public sees the scheme as automatic and as a part of procedure to recover confidence to banking system. The first Deposit Insurance was created in United States as a response to Great Depression. Today, according to International Association of Deposits Insurance (IADI) there are 95 countries 2 See Apostolik et al (2009) and Saunders and Cornett (2003) for an excellent text book on various risk inherent in financial firms and banks. 3 This could be in form of maximum withdrawal, characteristic of coverage (types of product, maximum interest rate, nominal amount, etc) and procedure.

6


that have deposits insurance (explicit or implicit). This is more than 60% of existing countries. There are various features of deposits insurance from coverage, characteristics of risk premium, existence of co-insurance and funding. Kunt et al (2005) studied and documented these features (see table 1). Most deposits insurance corporations (DIC) are jointly funded (63%). Around 36% are privately capitalized and only 1% that is pure public funded. Interestingly even though most DIC’s are jointly financed, most of them are administered by the state (60%). Around 27% are jointly operated and only 12% are pure private administration. It seems that government presence intervention is still desired. ��

�

��

��

��

��

��

��

��

��

��

��

In a theoretical work, Diamond and Dybvig (1983) showed that Deposits Insurance could provide a solution to bank runs or bank panics. Nevertheless its implementation is not without problem. Freixas and Rochet (2008) elicit three important aspects related to Deposits Insurance. They are (1) moral hazard Issue; (2) risk based pricing and (3) incomplete information problems. Moral hazard problem arise from weakening incentive of depositors to monitor the banks and increasing risk taking behavior of the manager and shareholder. Deposits insurance works as a put option in part of depositors, and a call option in the view of shareholders. The problems are especially important if the risk premium is not fairly priced: flat or inadequate risk adjusted premium (Greenbaum and Thakor, 2007). Merton (1977) considers deposits insurance as identical to put option of bank assets at strike price of amount of deposits. He postulates that actuarially fair rate of deposits insurance is an increasing function of deposits to asset ratio and volatility of bank assets. This important


7

work has been expanded to various ways; two important works are Pennachi (1987) which includes resolution of bank failure and Acharya and Dreyfus (1988) which considers possibility of authority takes over the bank before it is really insolvent. Chan et al (1997) shows that within incomplete information environment, the fairly priced deposit insurance may not be feasible. This is mainly due to (1) timing problems of lag in policy implementation and (2) adverse selection: private market for insurance premium is ceased to exist. Exposition above shows that the impact of Deposits Insurance to bank behavior might not be so clear. Flat rate might induce more risk taking behavior, however the impact to equilibrium level of deposits and loans margin might diverge (Suarez, 1993). Gennote and Pyle (1991) show that bank might underinvest in loan when bank capital is raised. More recent study by Matutes and Vives (2000) shows that competition might become fiercer with the introduction of deposit insurance, and lead to higher probability of failure.

2.2. A Review on Risk Taking Behavior There are many risk types to consider in bank management. However under spirit of Basel II, generic classes of bank risk could be categorized as credit, market and operational risk. Credit risk could be defined as probabilities that one or more component of bank portfolio experience a default (Freixas and Rochet, 2008). Credit risk could be further classified into individual risk and portfolio risk (Saunders and Cornett, 2003). Individual risk could be measured by standard credit analysis procedure famous as 5C jargon: Capital, Condition, Capacity, Collateral and Character (Apostolik et al, 2009). On the other hand portfolio risk mainly rises from degree of concentration and correlation (lack of diversification). Theoretically a loan could be depicted as a complete contingent contract that specify in every state of nature and interim date of the following conditions (Freixas and Rochet, 2008): a. The amount of repayment b. The interest rate on the remaining debt c. A possible adjustment in the collateral required by the lender d. The actions to be undertaken by the borrower The basic model of lender-borrower first developed by Wilson (1968) hinges on symmetric information assumption. Under this assumption, the sensitivity of repayment rate as a function of firm operation is high when the borrower is more risk averse than the lender (and vice versa). This model has been improved in two important ways. One development is the work of Townend (1979) and Gale and Hellwig (1985) in which they relaxed the assumption into asymmetric information. Lender bears significant cost to

8


reveal the exact nature of borrower business (called state verification). Under this more realistic assumption, lenders should develop incentive compatible contracts. There could be many contracts that have this property thus the next task for the lender is to pick up the most efficient one. An efficient incentive compatible contract is obtained by maximizing the probability of an audit for a fixed expected repayment amount or maximizing expected repayment for a fixed probability of an audit. If both agents are risk neutral, any efficient incentive compatible debt contract is a standard debt contract. The other is introducing a possibility of moral hazard after the loan is granted. Innes (1990) is the most cited work in the area. Assuming the limited liability of the borrower and individual rationality constraints, a model is developed to shape the optimal repayment function. The model also assume the monotone likelihood ratio property (Holmstrom, 1979), in which business result is an appropriate signal of effort. It is shown that the correct incentive compatible repayment scheme is function of effort. The borrower is imposed penalty if the observed result is lower than a particular threshold (obtained via solution to lender return maximizing problem). The borrower is given reward in form of zero additional repayment if his effort exceeds the threshold. There are several other variants and combination works from these two major strains. Bolton and Scharfstein (1990) construct a model in which borrower’s investment is not verifiable. Jappeli et al (2005) propose a model in which borrower could dispute the lender’s claim in court. Hart and Moore (1994) stress an important fact that the contract cannot impose on the borrower any restriction on the freedom of walk away. Myers and Rajan offer a model under condition of possible asset substitution opportunities. These models show that different result could be obtained in the lender-borrower relationship. Pyle (1971) and Hart and Jaffe (1974) are the first theoretical works that give birth to market risk meaning. They view bank as a portfolio manager which obtained funds in various form and tenor then invest them into assets. They regard loans which are inherently non tradable as tradable securities. These securities are valued using a risk free rate asdiscount factor. At a point of time, bank could be in an open position. They have mismatches in deposits versus loan characteristics (especially in tenor and currency denomination). With regard to interest rate, Hart and Jaffe (1974) show that as long as risk free interest rate remains between deposits and lending rate then bank has position in loan (securities) and deposits are positive. Practically market risk could be measure as return variability of a trading portfolio (Saunders and Cornett, 2003). This variability could result from changes in interest rate and exchange rate. There are three types of calculating market risk that commonly used: Risk Metrics4, historic simulation and Monte Carlo. 4 This instrument was first introduced by JP Morgan, see www.jpmorganchase.com for technical documents. Risk metrics was further developed and currently famous as Value at Risk. In a nutshell, this concept describe what is the extent of the loses if the day turn out to be a bad day.


9

Basel committee (2001) defined operational risk as potential loss (both direct and indirect) due to failure or inadequate internal system. This class of risk includes among other system failure, reputational risk, fraud and strategic risk. Jeitschko and Jeung (2005) develop an interesting theoretical work on aggregate risk positioning by banks which depends on various conditions. There are two critical aspects in their theorem: (1) strategic interaction among three important parties: deposit insurance, shareholder and manager and (2) four types risk profile which are more realistic than strict mean variance ordering that usually assumed in typical studies5. They show that with mean-variance ordering profile: high risk high return, three important parties in banking industry would have positive preference for risky assets. Shareholders are the highest, followed by management and deposit insurance respectively. Saunders et al (1990) investigate empirically the relationship of ownership and bank risk taking. Using a panel dataset consist of 38 US banks at annual frequency during 1978-1985 period, they tested whether stockholder controlled banks have greater incentives to take risk than managerially controlled bank. They experiment with seven capital market risk sensitivity measures (derived by Capital Asset Pricing Model).In their study they find evidence in support of the hypotheses: stockholder controlled banks exhibit significantly higher risk taking behavior than managerially controlled. Risk taking incentive may change before commencement of a business plan (ex ante) and after (ex-post). This possibilityis studied by Galloway et al (1997) using database of 86 US banks, at daily frequency during 1977-1994 periods. They use a market based risk measure, annual standard deviation of weekly equity return and several explanation variables (charter value, market to book value, capital, and operating leverage among others). They find that charter value correlates negatively with risk taking measures. A more recent study conducted by Marco and Fernandez (2007) on relationship of risk taking behavior on ownership structure and size of entities. They use two risk proxies: risk of failure (a Z score) and level of exposure to insolvency (inspired by Value at Risk paradigm). The data are an unbalanced panel consists of 256 Commercial and Savings banks in Spain from 1993 to 2000 at annual frequency (total 1030 observations). Using control variables like return on equity, lending to asset ratio and dummy of size and dynamic panel data econometrics they find evidence that risk proxies are inversely related to size. They also find that commercial banks are more risk inclined than the Savings banks. Our approach in this study is different in several way from previous works outlined above. First we are comprehensive. We include almost all commercial banks operational in Indonesia in post crisis era. Second, we use simple (accounting) risk measures against common set explanatory 5 Mean and variance are characteristics of distribution (also known as the first and second moments). Like a scalar, distribution could also be compared using concept of stochastic dominance. For an introduction to this topic, please see Davidson (2006).

10


variables to verify the relationship hypotheses. Third we take account on the impact of IDIC implementation on risk taking behavior. Last, we also view possibilities of interactions among different types of banks (using BI categories).

III. METHODOLOGY This study aims to reveal the risk taking behavior of the bank (credit, market and operational risk) against common set of independent variables (bank characteristics, competition and macro economy). We use panel dataset from published financial report. Dataset comprises of commercial banks operating during 2000 to 2009 (121 banks) with semiannual frequency. There are 2420 observations in the dataset6. A linear model was used as empirical scheme to test relationships between risk taking and independent variables. Mathematically it could be expressed in the following form

Sit = α0 + αi Xi + εit

(1)

where Sit is a vector of risk taking variables (credit, market and operational) and Xis vector of independent variables. The scheme comprises of 3 risk taking variables and 11independent variables. Definitions, proxies of variables and expected sign of hypotheses are given in Table 1 below. ��

��

��

��

��

��

��

��

�

��

��

��

�

��

��

��

�

��

��

��

�

��

��

��

�

��

��

��

��

��

��

��

��

6 Risks measures are missing or unavailable in several banks especially in early period. Therefore actually we work on unbalanced � data.�� panel The degree of severity of imbalance is different among regressions (credit, market and operational risk).

�

��

��

��

�� Risk Taking Behavior of Indonesian Banks: �� 11 Analysis onthe Impact of Deposit Insurance Corporation Establishment ��

�

��

�

��

��

��

�

��

��

��

�

��

��

��

�

��

��

��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Examination of pre and post IDIC establishment risk behavior is conducted using dummy event variables (notation: IDIC). These categorical variables are used both to signify level effect and behavior shift (interaction term with independent variables). The behavioral shift is assumed only occur to internal characteristics variables. If this regulatory institution works effectively we could expect that all variables are smaller in absolute size (i.e. the dummy event variables would take an opposite numerical value). The empirical scheme will also explore the possible effect on greater detail. Here we will direct our focus on different impact due to types. We use Bank Indonesia (Central Bank) categories and generate dummy variable called TYPE. The categories are State Owned Enterprise (SOE) Banks, Regional Development Bank, Private Foreign Exchange Bank, Private Non Foreign Exchange Bank and Foreign Owned-Joint Venture Bank. The notations are 1, 2, 3, 4 and 5 respectively. Again we look for possible interaction effect from IDIC implementation and bank types.

12


We use 3 econometric techniques to obtain the estimates7. They are Panel Estimated Generalized Least Squares (EGLS), Fixed Effect (FE) and Random Effect (RE). Apriory we don’t know the exact pattern of error component. They could be pooled residual, fixed or random between observations. Here we only take assumption of possible difference between cross section units not period. This is one way error component that can be either fixed or random. Error type could also not include into these classifications, therefore we still maintain pooled estimation with heteroscedasticity robust property (EGLS).

IV. RESULTS AND ANALYSIS This section presents estimation results on various econometrics specification and notes on their robustness (diagnostic test). First we would convey the regression result applied to all data (without including the effect of IDIC implementation). In the next sub section, we will present the impact of IDIC implementation. First we will see its general impact and then explore in greater detail i.e. affect by bank types.

4.1. Overall Behavior Result Table 3 shows the result of all sample credit risk taking regression. Credit risk as proxied by ratio of non performing loan to allowance is regressed to various explanation variables. Here we have results that some what different with previous studies.

��

��

� � � ��

��

��

��

��

��

��

��

7 There are many variants of panel data estimators. We use somewhat standard techniques that are quite popular in usage. Further exposition on this topic could be obtained from Baltagi (2005) and Cameron and Triverdi (2008).


13

First aligned with Saunders et al (1997), we find (credit) risk taking to be negatively affected by bank size. Larger banks tend to be more conservative. The coefficients are the largest among other estimators, ranging from -9.3 (EGLS) to -33.74 (RE). Since these are semi elasticities, 1% increased of size (ceteris paribus) would decrease credit risk taking position 9.3% to 33.74%. We also find competition to be negatively influence (credit) risk taking behavior (EGLS and FE) as suggested by Boyd and De Nicolo (2005). Last we also find Capital to be in line with existing literature since it could go both ways. Our findings then are closer to those of Keeley and Furlong (1990). Contrary to previous studies and intuition, we find that variables of ROE, personnel cost ratio (HRP), and Growth as non-increasing with risk. Contrary to theoretical result of Jeitschko and Jeung (2005), higher value of ROE does not seem to correlate with aggressive risk taking. We think large portion of government bonds and left over effect of 1997 great crisis has tempered management risk appetite while at the same time provide decent return. Indonesian banking industry is dominated by large banks (with share more than 70%) that are recapitalized after the crisis. These would affect the estimation result. The same argument could be said to explain the result of HRP and Growth. ��

��

��

�

�

� �

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Here we also find the strongest impact of the bank size to theirrisk taking behavior (see Table 4). The magnitude of the coefficients is ranged from -1.09 to -31.71 (the largest of all

14


estimates). An increase in 1% of bank size, on average would reduce the loan to deposit ratio by 1.09% to 31.7%. For this market risk type, the role of capital is closer to the one proposed by Jeitschko and Jeung (2005). More capital would likely to induce greater interest rate exposure (through higher LDR). With regards to competition, our findings are closer to that of Keeley (1990). More competition would induce aggressive (interest rate) risk taking. We find non supportive (and mixed) evidence regarding the role of ROE, HRP and growth to interest rate risk taking. ROE and HRP have a small albeit negative effect to LDR while growth estimates are not significant in all techniques used. Again we suggest the left over impact of crisis and recapitalization program should help explain this occurrence. One interesting additional findings are on DIVER variables. DIVER that measures the extent banks operate beyond their traditional intermediation role prove to be risk reducing. It seems that diversification activities taken by banks are detached to intermediation function. Indeed casual observation shows that in last decade other business unrelated to traditional role is flourishing. These activities include among others bank assurance, electronic-internet banking, wealth management, etc. The business mostly is fee based.

��

��

��

��

��

��

�

�

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Last, the evidences from operational risk are also in line with Saunders et al (1997). Here larger banks are associated with less operational risk taking. The magnitude is considerable


15

convergent and lower than the other two risk types. 1% increase in size (ceteris paribus) would reduce ratio of fixed asset to total asset by 1.71% to 2.29%. The role of capital and competition is small although significant and in line with Jeitschko and Jeung (2005) and Keleey (1990). In addition to estimating parameters, we also perform testing on choosing the appropriate model8. First we conduct Redundant Fixed Effect test (Log likelihood ratio) to test whether the Fixed Effect model (FEM) is suitable. The results strongly conforms the use of this model. The null hypotheses of (jointly) zero fixed cross section effect are strongly rejected. The F statistics are 3.58, 3.05 and 4.41 for credit, interest rate and operational risk respectively. Next, we take the analog test for random effect specification (The Hausman Test). Here the null hypotheses of (jointly) no random effect could not be rejected statistically. Taken together both specification tests would conclude that Fixed Effect Model is more superior in estimating the relationship of risk taking behavior to various factors. Based on this result, we decide to use FE technique in further exploration.

4.2. The Impact of IDIC Overall Bank Result Overall regressions show that implementation of IDIC alters the risk taking behavior of banks (see table 6). First we review the credit risk taking behavior. The constant term is increased by 93 points. It would suggest inherently, banks are more aggressive in taking exposure in credit. ��

��

� � � � � � � � � ��

� ��

��

��

��

��

��

��

��

��

��

��

8 We do not present the statistic calculation in this article. Interest reader could contact the author for the result.

16


This evidence contradicts the maintain hypotheses that implementation of IDIC would reduce risk taking drive. We can use the low loan exposures during the first half of the decade to explain this anomaly. After recapitalization, banks are usually reluctant to extend the credit since it might deteriorate their capital position that has just beingrecovered. However things might change once the IDIC was implemented in 2005. Coupled with low realization, the implementation could be seen as a boost to lending activities. As Suarez (1993) first put it, the deposit insurance work to reduce the cost of deposit, the most important input to lending and in the same time increase the value of loan to banks. Both can increase the risk exposure of loan, which is proxied with the ratio of non-performing loan to bad debt allowance. Further evidence on this hypotheses is given by positive and significant of IDIC*CAP interaction term. A one percent increase in capital would result in 2.1 percent more credit risk exposure post IDIC implementation. The credit risk taking behavior increases as the banks have higher charter value (this is contrast to the work of Marchus, 1984). On the other hand, the implementation of IDIC does not have a direct effect on the interest rate risk taking behavior. The estimated coefficient of IDIC, which should alter the constant, is not significant though the sign is negative as expected. However, the implementation of the IDC has an indirect affect through personnel incentive (HRP) and the capital (CAP). A one percent increase of personnel incentive will increase the interest rate risk taking behavior of 0.455, while the capital works in the opposite effect of -0.748 (See column 4 on Table 6). Operational risk is the only measure that in line with hypotheses. The impact of IDIC implementation is negative and significant (-0.852). Nevertheless most explanatory variables are either very small or statistically insignificant. Aligned with above findings, banks size is negatively correlated with risk taking position. Nevertheless related to IDIC implementation, the impacts are diverged. Larger banks are becoming even more reluctant to extend their credit risk position than the smaller ones after IDIC event. One percent increase in size would result in 4.56 percent reductionin credit risk compared to pre IDIC implementation. The interaction of IDIC and the size impacts are either not significant and/or very small for interest rate risk and operational risk.

Controlling the Bank’s Type The responseson IDIC implementation differ markedly across types of banks. We use the State Owned (SOE) Banks as baseline category. To have the net effect of each category, we subtract the baseline coefficient to interaction term9. The algebraic sign is sufficient to show when

9 For example the net impact of size to credit risk taking for Private Foreign Exchange Bank Category is 11.871 that is obtained from -45.862+57.733.


17

particular type of bank behavesmore conservative or more aggressive relative to our benchmark banks (SOE banks). Table 7 present the estimation result for the risk taking behavior using the credit risk proxy. ROE and SIZE are two variables that make the SOE banks to behave more conservative after the implementation of IDIC. We address the explanation of ROE impacton SOE banks to the benefit of recapitalization program as previously stated. Nevertheless the positive sign on Private Foreign Exchange bank and Foreign-Joint Venture category indicates the managerial drives to a more aggressive risk taking behavior, which is widely observed in practice. As explained earlier from Table 6, larger SOE banks tend to choose less risk. Surprisingly, the other types of bank move on the opposite. The Regional and the Private Foreign Exchange Bank respond differently when their size increases. For every 1 percent increase in size after the IDIC implementation, both will increase their exposure on credit risk in the magnitude (net effect) of 5.427 and 11.871 respectively. ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

On the other hand, the increase of capital after the IDIC implementation intensifies the credit risk taking behavior in SOE banks. For every one percent capital increase, the SOE banks increase their credit risk exposure by 13.73 percent, after the implementation of IDIC. Other types of bank response less aggressive or turn out to be aggressive. On average, due to the capital increase, the other bank types are less aggressive by less than a quarter relative to the SOE responses. The response is even negative for Foreign-Joint Venture banks, showing it becomes conservative.Credit risk position response to both HRP and Charter Value does not significantly differ across types of banks. This is in line with the overall regression.

18


��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

The next proxy for risk is interest rate and is termed as market risk (See Table 8). There is little evidence that the response of market or interest rate risk differs across banks types. From Table 8, the bank market risk response differs on only two variables; first is HRP influence on Private Non Foreign Exchange Bank and the other is Size on Foreign-Joint Venture banks. After the implementation of IDIC, the HRP increase the market risk exposure in Private Non Foreign Exchange Bank more than the other bank types. The same effect applies for variable Size in Foreign-Joint Venture banks. ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��


19

Table 9 shows the estimation result for operational risk. We find different response of operational risk between the SOE banks and other bank types. Related to the size of the bank (Size), SOE banks tend to reduce operational risk by 0.406 for every one percentincrease of size after the implementation of IDIC. Contrary to this, the other types of banks raise their operational risk exposure, where the coefficient of interaction term (IDIC*Size) for non-government bank is larger in absolute than the SOE banks. After the implementation of IDIC, as the size of the bank increase, the most conservative positive response belongs to regional banks, while the most aggressive on is private non foreign exchange. Related to the capital, after the implementation of IDIC the SOE banks tend to increase their operational risk as their capital increase. For every one percent increase in capital, the SOE banks raise their operational risk exposure by 0.282 percent. On the other hand, the other bank types only increase their risk exposure by less than one tenth compared to the SOE response. Their response is quite uniform, where their operational risk position on increase by 0.003 after the implementation of IDIC. The rest of variables: ROE, HRP and CV, do not exhibit different impact across bank types.

Comparison Summary Overall view on credit, market and operational risk proxies shows that the SOE banks behave distinctively compared to other bank types. The SOE bankis the most heavily recapitalized one, therefore is closely monitored by various stakeholders10. The larger the bank, the more conservative they will be, and it explain the negative correlation between the risk measure and the bank size. For non-government bank the situation is different. Though also were being recapitalized but they have been sold during the period of 2000-2003, hence are “free” from public control. Moreover, the regional bank, the private non foreign exchange and the foreign-joint venture bank are either small or mostly was self-sustained (not recapitalized by government). Nevertheless due to recapitalization, most SOE banks tended to be too conservative. They are more reluctant to involve in real business activities, there fore they have too low risk exposure. Observing this condition, the public put a pressure on these banks to be more aggressive and to contribute to the business-real sector development11. The implementation of IDIC enforces this trend, since it is perceived as implicit guarantee, and is in line with the theory. The other bank types work on a more sustainable and stable basis. They are free from public pressure and could adhere to their long term business plan without short term discretionary. A particular case is for the foreign-joint venture bank, which become more conservative on taking the credit risk. 10 We should note that in addition to huge economic restructuring, Indonesia is also undergoing a transition to democratic country. Public monitor and participation are rising significantly and various groups in the society are putting high interest on how government spend the money (and also the public corporations performance). 11 Indeed, popular anecdotes are people sees SOE banks management as living on public money. They earn revenue from government bonds that share large portion in the banks book. These bonds many giving coupon that significantly exceed deposits interest rate.

20


V. CONCLUSION For Indonesia, the pivotal point in banking regulation is the implementation of Deposit Insurance Corporation (IDIC). The majority of literature both theoretical and empirical suggests that the IDIC event should reduce the incentive to take the risk. The empirical test on this paper provides us several findings. First, most cases show that risk taking behavior is negatively correlated with the size of the bank. Larger bank tends to be more risk averse than the smaller bank. Second, the implementation of IDIC alter the risk taking behavior, but some what different from initial hypothesis. Excluding the interaction terms, the implementation of IDIC tends to raise the credit risk taking. Third, the implementation of IDIC tends to reduce the operational risk taking, which is aligned with the hypotheses. Fourth, the influence of size is significantly less pronounced post IDIC implementation. By controlling the bank types, the result shows that the government bank (SOE) tends to reduce risk as its size increases; and this is the fifth finding. Sixth, the SOE bank also tends to raise risk exposure as its capital increases, while for the other bank types is the opposite. These findings have several policy implications. First, the IDIC implementation tends to increase the risk taking behavior; and this is a sign of moral hazard as suspected by Freixas and Rochet (2008). The moral hazard in this situation most likely is due to the flat rate insurance premium (Greenbaum and Thakor, 2007). To evade the problems, IDIC should consider using a fairer premium, which based on the risk (risk-based premium). Second, larger banks seem to have self-control mechanism relating to risk since they reduce risk position as their size increase. This finding highlights the importance of bank consolidation through Merger and Acquisition (M&A), and the authority should provide proper incentive for this. To preserve the competition and to maintain the contes table banking market, Bank Indonesia can open the door for new comer to take over the existing banks. Third, the bank excluding SOE banks tend to be more conservative to take risk as capital increase. Since this real phenomenon may arise as the form of capital preservation, the authority should emphasize the importance of capital along with the undergoing implementation of Basel III.


21

References

Apostolik. R., Donohue, C. And P. Went, 2009, Foundations of Banking Risk, Wiley Finance, New Jersey. Boyd. J.H., and G. De Nicolo, 2005, “The Theory of Bank Risk Taking and Competition”, The Journal of Finance, 60, No. 3, page 1239-1343. Chan, Y.S, Greenbaum S.I., and A.V. Thakor, 1992, “Is Fairly priced deposits insurance possible?“, Journal of Finance, 47, page 227-245. Davidson, Russel, 2006, “Stochastic Dominance”, Mc Gill University, Department of Economics Discussion Paper. Freixas. X and Rochet J.C., 2008, Microeconomics of Banking, 2nd Edition, MIT Press. Gale, D. and M. Hellwig, 1985, “Incentive compatible debt contracts: the one period problem, Review of Economic Studies, Vol. 52, page 647-663. Galloway, T.M., Lee W.D., and D.M. Roden, 1997, “Banks changing incentives and opportunities“, Journal of Banking and Finance, 119, page 929-970. Gennote, G and D. Pyle., 1991, “Capital control and bank risk”, Journal Of Banking and Finance, 15, page 805-824. Gorton, G., 1985, “Banks suspension of convertibility”, Journal Of Monetary Economics, 15, page 177-193. Greenbaum, S.I and A. Thakor, 2007, Contemporary Financial Intermediation, Academic Press, San Diego California. Indonesia Deposit Insurance Corporation, Annual Report, 2009. Innes, R.D., 1990, “Limited liability and incentive contracting with ex-ante action choices”, Journal of Economic Theory, Vol. 52, page 45-67. Hart, O., and Moore, 1994, “A theory of debt based on the inalienability of human capital”, Quarterly Journal of Economics, Vol. 109, page 841-879. Holstrom, B., 1979,”Moral hazard and Observability”, Bell Journal of Economics, Vol. 10, page 74-91.

22


Jappelli, M., Pagano, P., and M. Bianco, 2005, “Courts and banks: Effects of judicial enforcement on credit markets”, Journal of Money, Credit and Banking, Vol. 37, page 223-244. Lin, S.L. and Wu, S.J., 2005, “Capital requirements and Risk Taking Behavior in Banks: International Evidence”, ISFA, Working Paper. Marco-Garcia, T., and M.D. Fernandez-Robles, 2008, “Risk-taking behavior and ownership in the banking industry: The Spanish evidence“, Journal of Economics and Business, 60, page 332-354. Marcus, A.J.,1984,”Deregulation and Bank Financial Policy”, Journal of Banking and Finance, Vol. 8., page 557-565. Matutes, C. and X. Vives, 2000, “Imperfect Competition, risk taking and regulation in banking”, European Economic Review, 44, page 1-34. Jeitschko, T.D. and S.D. Jeung, 2005,”Incentives for Risk Taking in Banking-A Unified Approach”, Journal of Banking and Financial, 29, page 759-777. Saunders, A. and M.M. Cornett, 2003, Financial Institutions Management: A Risk Management Approach, McGraw Hill, Singapore. Saunders, A., Strock, E., and N.G. Travlos, 1990,”Ownership Structure, Deregulation and Bank Risk Taking”, The Journal of Finance, Vol. 45, No. 2, page 643-654. Suarez, J., 1993,”Closure rules, market power and risk taking in a dynamic model of bank behavior”, Discussion Paper, Universidad Carlos III, Madrid. Townsend, R, 1979,”Optimal contracts and competitive markets with costly state verification”, Journal of Economic Theory, Vol. 21, page 265-293. Wilson, R, 1968, “On the theory of syndicates”, Econometrica, Vol. 36, page 119-132.

Capital Flows in Indonesia: the Behavior, the Role, and Its Optimality Uses for the Economy

23

Capital Flows in Indonesia: the Behavior, the Role, and Its Optimality Uses for the Economy Fiskara Indawan, Sri Fitriani, Meily Ika Permata dan Indriani Karlina1

Abstract

The abundance of global liquidity post the global crisis resulted in a huge amount of international capital flows to Government Securities (GS) market. Besides useful, the flow of foreign capital potentially give a risk reversal that may leads to instability in domestic financial market. This paper analyzes the determinant of foreign investors including the risk and returns, both from domestic (pull factor) as well as from global (push factor). The result shows that the push factor was instrumentally influence the behavior of foreign investors in the GS (Government Securities) market. For long-term investors, their behavior to place their funds in GS market is influenced by push factor, but not significantly affected by the pull factor. However, for short-term investors, both pull and push factors influence their investment decisions. In addition simulation results indicate that in the future, the prospect of foreign investors in the securities market still faces challenges, particularly from the relatively high volatility as a result of the shock sensitivity of foreign investors on shock that can happen in the uncertainty in the international financial markets due to ongoing debt crisis resolution in developed countries.Concerning these findings, Bank Indonesia and the government needs to maintain and manage the returns and risks of domestic investment on a more competitive and relatively low level by maintaining the strength and resilience of the domestic economy and financial stability.

Keywords : Foreign Exchange, International Lending, Corporate Finance. JEL Classification : F31, F34, G3

1 Economic Researcher in Grup Riset Ekonomi (BRE, Economic Research and Monetary Policy Department (DKM), Bank Indonesia. Ideas in this paper represent the writer’s points of view and do not represent ideas or points of view of either DKM or Bank Indonesia. Email: [email protected], [email protected], [email protected], [email protected].

24


I. INTRODUCTION The abundance of global liquidity post-global crisis that resulted in a flood of international capital flows in the form of investment portfolio into Indonesia will provide a challenge for monetary policy implementation2. Indonesia, like other emerging market countries, has a stronger level of economic growth and higher interest rates, while in the other hand, at the same time developed countries apply extra loosening monetary policy with relatively low interest rates. Both of these factors play an important role in the shifting of international capital flows to emerging markets that have a better rate of return and are supported by economic performance and improving risk (IMF, 2010). In the one hand, the entry of the foreign capital showed increasing international confidence in economic fundamentals reinforced by the increase of Indonesia’s rating to investment grade. Capital flow can increase domestic liquidity and can be used as an alternative investment funding source that is relatively cheaper and can encourage investment activity and encourage the domestic economy. However, besides being useful, in the other hand, international capital flow has high potential risk if it is not managed wisely. Massive capital inflow leads to appreciation of the exchange rate and is able to weaken the competitiveness of exports. Besides, it can lead to higher risks of economic overheating on the economy and increasing pressure on inflation along with the sharp increase in asset prices as well as credit growth and investment that tend to be more expansive. At the same time, the global economic condition that is still vulnerable as well as uncertain international financial markets along with debt crisis in Europe could trigger the fluctuation in international financial markets and lead to high risk of instability in domestic financial markets and the exchange rate in the case of reversal of capital in short time (sudden reversal), especially for the short-term capital flow. Thus, the foreign capital flow is expected to be managed properly in order to give optimal benefits to economic as well as can be minimized the risks. In order to minimize the potential risk in the management of international capital flow, it needs a better understanding about the patterns of capital inflow behavior in financial markets especially in Government Securities (SUN) market along with increasing foreign ownership in the market. In-depth analysis includes several factors, which are the risk factors and the returns that can be derived from domestic (pull factor) or from global (push factor), that affect foreign investors when they decide to make the purchase and sale in Government Securities (SUN) market3. The analysis also needs to be focused to get a good understanding about the characteristics of investor with investment time

2 Some emerging markets facing massive capital inflow has taken several policies beyond interest rate policy in managing capital inflows as macro prudential policy and capital controls (IMF, 2011). Ostry (2010) argued that the policy mix in the face of capital inflows depends on the country’s economic conditions, the level of foreign exchange reserves, quality of prudential rule, strengthening exchange rate and persistence of capital inflows. 3 Analyses using high-frequency (daily data)


25

horizon (long-term and short-term). The results of this study are expected to give appropriate policy recommendations in managing capital inflow, especially in financial markets. The purposes of this study are to identify the foreign investors’ behavior in the Government Securities (SUN) market, both long-term and short-term investors as either in aggregate or individually, especially their net transaction (purchases minus sales) in secondary market of Government Securities (SUN). The analysis of behavior identification includes the factors that affect investors’ motivations which are returns factors and risk factors, whether it is sourced from domestic (pull factors) or from global (push factors); second, to simulate the prospect of foreign investors in the Government Securities (SUN) market by using the estimated model; and third, to recommend the aspects needed to be considered in the management of international capital flow based on the findings in two purposes of this study.

II. THEORY 2.1. Modern Portfolio Theory (MPT) Modern Portfolio Theory (MPT) or Portfolio Theory is a mathematical formulation of the diversification concept in investment, with the aim to investment assets collection selection that gives the most efficient composition either in terms of return or risk. MPT is a financial theory that attempts to establish the composition or proportions of a wide selection of assets in order to maximize the expected return of portfolio for a certain level of risk, or instead to minimize risk for a level of expected return. This theory was first introduced by Harry Markowitz (1952) and developed by James Tobin (1958) by adding an asset that is risk-free into the analysis. If investors, especially foreign investors, have two risky investment portfolio options which is investment in Indonesian financial market that have return RD and variance σ 2D and in international financial market with return RF and variance σ 2F investors can invest their funds with the proportion by ωp for asset in Indonesian financial markets and by 1 - ωp for asset in international financial market, then expected return portfolio and risk of the portfolio are:

ܴ௉ ൌ ߱௉ ܴ஽ ൅ ሺͳ െ ߱௉ ሻܴி

ߪ௉ଶ ൌ ‫ܧ‬ሺܴ௉ െ ‫ܴܧ‬௉ ሻଶ ൌ ߱௉ଶ ߪ஽ଶ ൅ ʹ߱௉ ሺͳ െ ߱௉ ሻߩߪ஽ ߪி ൅ ሺͳ െ ߱௉ ሻଶ ߪிଶ

(1) (2)

where σ 2P is the standard deviation of RD and RF, and r is the correlation between RD and RF. In portfolio theory that used mean-variance model, investor will choose the efficient investment portfolio (efficient portfolio) which has a high return and low risk. In Figure 1, the combinations of all efficient portfolios are in BB curve where the investment risk σ 2P is getting small on each return investment Rp.

26


Figure 1. Efficient Portfolio

In order to know the optimal portfolio allocation between domestic and foreign investment then it is using capital market line MN that is a combination of return and risk from risky and free risk asset. The slope in equilibrium will touch BB curve at the point P, which is portfolio combination that has return RP and level of risk σ 2P . If investors want to gain higher return then they should add their investment portfolio in risky asset as well as having a higher risk so that it moves towards point M. Instead, investors will gain lower return when they have less risky investment so that it moves towards point N. The number of optimal investment w *P is obtained from substitution equation (1) and (2) into the slope σ 2P / (Rp - N) and slope (∂ σ 2P / ∂wp) / (∂RP / ∂wp) as used by Miller (1971), so that obtain:

߱௉‫ כ‬ൌ

Where

‫ܭ‬ൌ

Vଶி

൫ఙಷమ ோା௄൯ ሺ௅ା௄ோሻ

ܴൌ

ൌ ݂ ሺܴǡ ߪ஽ଶ ǡ ߪிଶ ሻ

ሺோವ ିோಷ ሻ ሺோು ିேሻ

െ UV஽ Vி

(3a)

‫ ܮ‬ൌ Vଶ஽ ൅ Vଶி െ ʹUV஽ Vி

(3a)

(3b) (3c)

2.2. Long-term vs. short-term Investors Long-term and short-term investors differ in terms of investment period. Using portfolio theory with mean-variance analysis, Campbell and Viceira (2001) suggested that the short-term investors are facing the following one-period wealth maximization problem:


ଵିఊ

݉ܽ‫ܧݔ‬௧ ܹ௧ାଵ Ȁሺͳ െ ߛሻ

Where

(4a) ଵ

ଵିఊ

ଶ ݈‫ܧ݃݋‬௧ ܹ௧ାଵ ൌ ሺͳ െ ߛሻ‫ܧ‬௧ ‫ݓ‬௧ାଵ ൅ ሺͳ െ ߛሻଶ ߪ௪௧

subject to

27

ଶ

ܹ௧ାଵ ൌ ሺͳ ൅ ܴ௉ǡ௧ାଵ ሻܹ௧

(4b)

(5)

ܴ௉ǡ௧ାଵ ൌ ߙ௧ ܴ௧ାଵ ൅ ሺͳ െ ߙ௧ ሻܴோிǡ௧ାଵ

(6)

where Rp,t+1 is portfolio return, R t+1 is return of risky assets, RRF,t+1 is return of risk-free assets, at is share of the portfolio placed on risky assets, and γ is coefficient of relative risk aversion. Furthermore, substituting equation (6) to (5) and then (4a) and (4), we obtain the maximization problem for the short-term investors as follows: ଵ

ଶ ݉ܽ‫ܧ݃݋݈ݔ‬௧ ൫ͳ ൅ ܴ௉ǡ௧ାଵ ൯ െ ߛߪ௣௧

(7)

ଶ

From equation (7), the short-term investors will achieve maximum wealth by maximizing their portfolio return and minimizing their risk (variance) portfolio. Optimal return and variance portfolio are ଵ

‫ݎ‬௉ǡ௧ାଵ െ ‫ݎ‬௥௙ǡ௧ାଵ ൌ ߙ௧ ሺ‫ݎ‬௧ାଵ െ ‫ݎ‬௥௙ǡ௧ାଵ ሻ ൅ ߙ௧ ሺͳ െ ߙ௧ ሻߪ௧ଶ ଶ

ଶ ߪ௣ǡ௧ ൌ ߙ௧ଶ ߪ௧ଶ

(8) (9)

Meanwhile, the long-term investors face wealth maximization problem for K period ahead, with the following budget constraint: (10)

ܹ௧ା௄ ൌ ൫ͳ ൅ ܴ௉௄ǡ௧ା௄ ൯ܹ௧

The maximization problem for the long-term investor is: ଵ

ଶ ݉ܽ‫ܧ݃݋݈ݔ‬௧ ൫ͳ ൅ ܴ௉ǡ௧ା௄ ൯ െ ߛߪ௣௧ା௄

(11)

ଶ

and the optimal return and variance of the portfolio are: ଵ

‫ܭ‬ሺ‫ݎ‬௉ǡ௧ାଵ െ ‫ݎ‬௥௙ǡ௧ାଵ ሻ ൌ ߙ௧ ‫ܭ‬ሺ‫ݎ‬௧ାଵ െ ‫ݎ‬௥௙ǡ௧ାଵ ሻ ൅ ߙ௧ ሺͳ െ ߙ௧ ሻ‫ߪܭ‬௧ଶ

ଶ ߪ௣ǡ௧ା௄ ൌ ߙ௧ଶ ‫ߪܭ‬௧ଶ

ଶ

(12) (13)

By comparing equations (8) and (9) with (12) and (13), the return and variance portfolio of short-term investors are still optimal for the long-term investors. The mean and the variance of the short-term investors are equal to the long-term investor, multiplied by the factor of K period.

28


2.3. Determinants of Capital Flow The capital inflow to developing countries is driven by several factors. The high degree of financial integration accompanied by the rapid development of technology, especially information and communications technologies play important role in accelerating the increasing mobility of the capital flows. Besides, the development of capital market infrastructure accompanied by the liberalization of capital markets such as the elimination of barriers to repatriate, the reduction of barriers for foreign participation and ownership also contribute to the expansion of the capital flows to developing countries’ markets. There are two major determinants for capital inflow (Agenor, 2004; Calvo et al, 1994): 1. Internal or pull factors, which are linked to domestic policies, such as high productivity levels and growth rates, strong macroeconomic fundamentals, macroeconomic stabilization, structural reforms (for instance the capital liberalization and reduction of fiscal deficit), which would normally be compensated and reflected in the increase of a country’s rating. 2. External or push factors such as (1) the low level of world interest rates, particularly in the U.S. and some other developing countries, which lead to the declining of risk premium, while give higher yield in emerging markets (2) recession or slowing down of level growth in developing countries will lead to low return level and reduce profit opportunity, hence will lead to the transfer of capital from developing countries to emerging markets. Based on type and risk, capital flows can be categorized as follows:

��

��

��

��

��

��

Picture 1. Types of Capital Flow

Larrain et al. (1997) found that long-term flows tend to be influenced by economic fundamentals, while short-term flows are influenced by the interest rate differential. Agung et al. (2011) by using Indonesian data and VAR models, found that the capital inflows into


29

Indonesia are mainly caused by “push factors”, especially from the impact of extra loosening monetary policies adopted by developing countries. In addition, we found that the inflows are particularly vulnerable to the reversal risk.

2.4. Empirical Basis Several studies have analyzed the factors that led the flow of foreign capital flows into developing countries. There are few studies that specifically use VAR to analyze capital flows in emerging countries like Ying and Kim (2001), Vita and Kyaw (2007), Goldfajn and Minella (2005), and Culha (2006). In general, Culha (2006) and Fratzscher (2011) stated that the domestic factor (pull factors) is important factors in attracting capital flows in emerging countries. While Forbes and Warnock (2011) stated that factors from abroad (push factors) to be a driving force of capital flows to emerging countries. Agung et al (2011) using OLS method with monthly data from January 2004 - December 2010 examines factors affecting capital flows in the stock market, SBI and Government Securities (SUN). He found the capital flows are positively influenced by two pull factors namely domestic economic growth (production index) and domestic interest rate changes, and three push factors namely the level of global risk (EMBIG), the global liquidity excess (money supply in the U.S.) and the changes in U.S. interest rates. Meanwhile, capital flows are negatively affected by U.S. economic growth. Furthermore, to examine the effect of capital flows on macroeconomic variables, he used the VAR method on quarterly data for the period of 1994 to 2010. The results were capital flows have a positive influence on foreign exchange reserves, money supply, and stock index, and negative effects on the real exchange rate (appreciation). Nugroho (2010) examined the factors affecting capital flows that are proxied from the foreign exchange transactions with domestic banks from LHBU. Using OLS method and monthly data from January 2002 - March 2010, he found that the capital flows are positively influenced by two factors, which are spread between JIBOR domestic interest rate with LIBOR composite interest rate and economic growth in the U.S. (U.S. consumer confidence and U.S. production index), and was negatively influenced by exchange rate expectation (depreciation led to capital outflow). Cadarajat (2008) by using ARDL method and quarterly data from 1985 to 2007 suggests that the capital flows proxied from FDI, FPI and other investment have positive effect on current account, domestic economic growth and the stock index and have negative effect on country risk and real interest rate. In the financial markets (stocks and Government Securities (GS)) there are positive relationships between the price volatility and the trading volume. The higher the trading volume, the higher will the price volatility in the market. Karpoff (1987) stated that relationship between price volatility with trading volume can give a clearer view of the flow and dissemination of

30


information in the market as well as its structure and its size . Positive relationship between two variables indicates that the market becomes more transparent because there are many investors who can obtain variety of information about market conditions and fundamentals from many sources.

III. METHODOLOGY 3.1. Estimation Technique In a series of financial data, which is usually the high frequencies data, daily or weekly, is often found volatility clustering, where there is a period with high volatility while at different times, there were periods with low volatility. In the period of high volatility, a large shock (residual) tend to be followed by a large shock as well, instead, in periods of low volatility, a small shock will be followed by a little shock too. Ordinary linear regression model emphasizes stable volatility assumptions (homoscedasticity). In the above case, where the homoscedasticity terms cannot be met, one can model the variance of ε_t as a function of the lag of error term. Modeling and forecasting the volatility gives several advantages such as a more efficient estimator if the problem of heteroscedasticity can be solved. Besides, since theforecast confidence interval can vary across time, the variance modeling of error term will help to give more accurate interval. Engle (1982) introduced the concept of Autoregressive Conditional Heteroscedasticity (ARCH). In this model, variance of error term in period t is affected by the square of previous error term (volatility) in several periods. ଶ ߪ௧ଶ ‫ܧ ؠ‬ሼߝ௧ଶ ȁ‫ܫ‬௧ିଵ ሽ ൌ ߱ ൅ ߙߝ௧ିଵ

(14)

With w > 0 and a > 0. The above model is the ARCH (1), with lt-1 is a collection of information that includes e 2t-1 and all previous information. ARCH model (1) states that when 2 large shock occurs in period t - 1, then e t tends to be large, as well as s2 . In other words, t there is a correlation between e 2 and e 2t-1. Unconditional variance from e 2 is: t

ߪ௧ଶ

‫ؠ‬

‫ܧ‬ሼߝ௧ଶ ሽ

ൌ߱൅

ଶ ሽ ߙ‫ܧ‬ሼߝ௧ିଵ

t

(15)

The above equation has a stationary solution which is: s 2 = 1 w , because 0 < a < 1. -a t Keep in mind that unconditional variance does not depend on t. ARCH model (1) can be expanded into ARCH (p) as follows: ଶ ଶ ଶ ଶ ߪ௧ଶ ൌ ߱ ൅ ߙଵ ߝ௧ିଵ ൅ ߙଶ ߝ௧ିଶ ൅ ‫ڮ‬൅ ߙ௣ ߝ௧ି௣ ൌ ߱ ൅ ߙሺ‫ܮ‬ሻߝ௧ିଵ

(16)

The improvement of ARCH model variations that are very useful is introduced by Bollerslev (1986), known as Generalized ARCH or GARCH. GARCH model (q, p) can be written as follows:


௣

௤

ଶ ଶ ߪ௧ଶ ൌ ߱ ൅ σ௝ୀଵ ߙ௝ ߝ௧ି௝ ൅ σ௝ୀଵ ߚ௝ ߪ௧ି௝

Or

31

(17)

ଶ ଶ ൅ ߚሺ‫ܮ‬ሻߪ௧ି௝ ߪ௧ଶ ൌ ߱ ൅ ߙሺ‫ܮ‬ሻߝ௧ି௝

(18)

With w > 0, a > 0 and b > 0. GARCH is a more compatible alternative to model the ARCH with higher order. By using GARCH method the selection of lag et can be minimized.

3.2. Empirical Model To investigate the determinants of foreign investors on Government Securities (SUN), we use econometric models test for equation (24) using GARCH. ݊ ܲ‫ ݃݊݅ݏܣݎ݋ݐݏ݁ݒ݊ܫ݅ݏ݇ܽݏ݊ܽݎܶݐ݁ܰ݅ݏ݅ݏ݋‬ൌ ൫σ݉ ݇ൌͳ ‫ ݇ݎ݋ݐܿܽܨ݄ݏݑܲ ݇ן‬൯ ൅൫σ݇ൌͳ ߚ݇ ܲ‫ ݇ݎ݋ݐܿܽܨ݈݈ݑ‬൯

(19)

The dependent variable is the position or the accumulation of foreign investor transaction (non-residents), both long-term and short-term. Transactions that are analyzed is the position of net inflow (purchases minus sales) for each long-term and short-term investors in aggregate or individually. While the candidate of independent variables are as follows: ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

� ��

� �

��

��

��

��

��

��

��

��

��

��

��

��

��

��

32


We use daily data from 2004 until the end of 2011. The source of foreign investor transactions data was obtained from Government Securities transactions in secondary market, obtained from BI-SSSS system (Bank Indonesia Script less Securities Settlement System). While independent variable data is obtained from Bloomberg. In order to identify foreign investor group behavior, consisting of long-term and short-term investors in Government Securities (SUN) market, we regress the equation (24) with dependent variable of the net transaction (purchase transaction – selling transaction) and the combination from various variables in Table 2 as the independent variables for each group. Each equation contains 4 or 3 independent variables that represent indicators of the global risk and return or the domestic risk and return. Total equation used is 72 equations. From 72 equations, we selected one equation for all foreign investors both aggregately and individually. The selection of the equation is based on the number of significant independent variables.

IV. RESULT AND ANALYSIS 4.1. Stylized Fact of Government Securities (SUN) The transactions of foreign investors in Indonesia tend to increase. In 2011, foreign investors purchase transactions reached 44.7% from the total value of transactions in Indonesia Government Securities (SUN) market (Figure 4). Besides, looking at the net transaction (purchase transactions net minus sales), more foreign investors took selling position during 2006-2008, and recorded the building stock during 2009-2011 (Figure 5). In terms of ownership, as of June 2012, foreign investors have a market share of 27.40% (Figure 6).

��

� ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

� ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 2. Foreign Transaction Value

Figure 3. Net Position of Buy-Sell Foreign Investor

33


� ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 4. Position of Foreign Ownership in SBN

In order to categorize the long-term or the short-term investor, we use the Government Securities (SUN) trading in secondary market. Figure 7 shows that there are 7 most active foreign investors, with the average activity of buying - selling of 70% from the total transaction value of foreign players. As explained before, the difference between the long-term and the short-term investor is on their investment time horizon. In this paper, the long-term investor is defined as foreign institutional investors who have significantly increase their position in Government Securities since 2009, which was the starting period of rapid capital inflows into Government Securities (SUN) market. Instead, the short-term investors are foreign institutional investors that have no significant changes in their Government Securities (SUN) position for the same period. Based on these definitions, then in accordance with Table 4, the investors fall into the category of long-term investors are investors A, B and C and short-term investors are investors D, E, F and G.

� ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

Figure 5. Trading Activity of 7 Largest Foreign Investor

Figure 6. Position of 7 Greatest Foreign Investor (in Billion Rupiahs)

34


��

��

��

��

��

�

�

��

��

��

�

�

��

��

��

�

�

��

��

��

�

�

��

��

��

�

�

��

��

��

�

�

��

��

��

�

�

��

��

��

From table 4, in 2011 when there was rapid of capital inflows, the long-term and shortterm investors have relatively similar portion of 46:34 (Figure 10). This shows that both investors have relatively equal effect on Government Securities (SUN) market.

� ��

��

��

��

��

��

��

��

� ��

��

��

��

��

��

��

Figure 7. Position of Foreign Investor (in Billion Rupiahs)

��

��

��

��

��

��

��

Figure 8. Portion of Long-term and Short-term Investor

Long-term investors tend to be active after the global financial crisis (end of 2008). Meanwhile, the short-term investors have been actively trading in Indonesian Government Securities market, and its movement is very volatile, showing their higher motives for capital gain (indicated from a very high buying and selling activities ).


��

��

��

35

��

��

��

��

��

��

��

��

��

��

Figure 9. Net Activity of Buy-Sell of Long-term Investor

Figure 10. Net Activity of Buy-Sell of Short-term Investor

4.2. Estimation Result Regression results for 72 combinations using GARCH (Equation 19) obtained a general equation that can be used on all foreign investors both groups (long-term and short-term investors) as well as individual of foreign investors. The independent variable of the equation are the 5-year yield as domestic return variables, the 5-year U.S. T Notes yield as foreign return variables, the interbank (PUAB ON) rates as domestic risk variables and the VIX index as foreign risk variable. Next, this general equation will be tested on all investors both group of investors (long-term and short-term).

Long-Term Investor The result of estimated general equation for the long-term investor showed only push factors (global factors), both profit and risk, that significantly affect the behavior of investors in the Government Securities market. The test results remain consistent with the behavior of individual investors. From 3 long term investors tested, two individual investors perform the same behavior (Investor A and B). Meanwhile, the behavior of investor C transaction cannot be well explained by the variable yield5 (proxy for domestic return), USTB (proxy for global return), PUAB (proxy for domestic risk) and the VIX (global risk). Regression results show that considerable improvement of the long-term foreign investor position in the GS market, especially since the beginning of the global financial crisis in 2008 was driven by the push factors which represented from the low foreign interest rates and the relatively high foreign risk. Foreign investors are looking for the alternative placement of investment with relatively high interest rate and low level of risk. The huge push factor shows the vulnerability of the GS market to the risk of large sudden reversal of capital flows when the intensity of risk in the international financial markets increased

36


sharply. Therefore, the government and Bank Indonesia need to continue the awareness of and to monitor the global financial markets developments and prepare the contingency plans to address these risks to minimize their impact on financial stability and domestic economy. The results of impulse response in Figure 12 shows that the shock of a rise in 5 years U.S. T-Notes by 100 basis points would lead to lower position of long-term investors net transaction (or make net sales) by approximately Rp 11 billion at that time (t = 0) with the cumulative impact by Rp 14 billion. While the shock of a rise in VIX index by 100 basis points lead to long-term investors sales by Rp 1.7 billion, with the cumulative impact by Rp 2.4 billion. The result ofindividual impulse response (Investor A and B) shows higher sensitivity. In Figure 13, the increase shock of 100 basis points U.S. T-Notes yields lead to investors A and B to make net sales by Rp 19 billion and similarly, the increase of 100 basis points VIX index lead to net sales by Rp 2 billion at the time t = 0.

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 11. Impulse Response Function Result to Long Term Investor (in Billion Rupiahs)

��

� �

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 12. Impulse Response Function Result to Investor A and B / Long-Term (in Billion Rupiahs)

��


��

�

37

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 12. Impulse Response Function Result to Investor A and B / Long-Term (in Billion Rupiahs) Lanjutan

Short-term Investor In contrast to long-term investors, short-term investor transaction behavior is influenced by both push factors and pull factors (domestic factor) in the Government Securities (GS) market. The regression result suggests that the high frequency of short-term investor transactions in GS market is driven more sensitively on each of changes in the pull factors and the push factors. Both ofthese factors will affect the expected return and the risk tolerance to accept. This is in line with the nature of short-term investor transactions that tend to only look at short-term profits through capital gains. The high frequency of short-term investor transactions that is not accompanied by the increasing position will lead to high volatility in the GS markets, which in turn may affect the stability of the overall financial markets. Therefore, considering short-term investor transaction is very volatile, and then from domestic points of view, the government and Bank Indonesia need to consider the factors that influence the short-term investors by maintaining domestic economic condition such as maintaining domestic competitive interest rate and keeping the level of domestic risk at a fairly low level. Meanwhile, to encounter the risk of capital flows reversal, the government and Bank Indonesia need to continue their awareness on risks in the international financial market and prepare the contingency plans. The results of the individual impulse response in Figure 14 shows that the increase shock Yield5 by 100 basis points would lead the short-term investors to lower net transaction position (or making net sales) by Rp 25.5 billion at that time (t = 0 ) with the cumulative impact by Rp 31.05 billion. On the other hand, the increase shock on 100 basis points UST-5 Year Notes index will lead the short-term investors to net sell position by Rp 80.7 billion, with the cumulative impact by Rp 98.3 billion. Furthermore, the increasing shock on the interbank rate (PUAB ON) by 100 basis points lead the short-term investors to book net sales position by Rp 11.07 billion

38


with cumulative impact by Rp 13.5 billion, while a 100 basis points increase on VIX index lead to net sales by Rp 7.07 billion at the time t = 0 with the cumulative effect by Rp 8.6 billion.

��

��

��

��

��

�

��

��

��

��

��

��

�

��

�

��

��

��

��

��

��

��

��

� � ��

��

��

��

��

��

��

��

��

��

� � ��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 13. Impulse Response Function Result to Position of Short-Term Investor (in Billion Rupiahs)

Although in general the results of the regression for the short term individual investors show similar to the group behavior, the push factor is more consistent than the pull factor. Variable of push factor (U.S. T-Notes and the VIX index) has a significant influence on all tested short-term individual investors (Figure 15 and Figure 16). Meanwhile, the returns of the domestic factors proxied by Yield-5 significantly affect only two over four short term individual investor (Investor F and G) and only one investor (Investor F) that is significantly influenced by the domestic risk proxied from PUAB variable.


��

�

�

��

��

��

��

��

��

�

�

39

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 14. Impulse Response Function Result to Position of Short-Term Investor D and E (in Billion Rp)

��

��

��

��

��

� � ��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

40


��

�

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

� ��

��

��

��

��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 15. Impulse Response Function Result to Position of Short-Term Investor F and G (in Billion Rp)

41


Prospect of Capital Flows In order to know the prospects of the capital flows, we run simulations using the above estimation both for the long-term and the short-term investors. The simulations use three scenarios; mild, moderate, and crises scenario4 with such criteria as outlined in Table 3. Determinations of the criteria are based on the historical patterns in each variable since 2004. ��

��

��

��

Using these criteria, we can determine the daily shock value to use for the following variables. See Table 4. The simulation result is presented in Table 5. ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

The simulation shows the short-term investors are more sensitive to the shock of the four explanatory variables. This is because the four explanatory variables can significantly influence the short-term investor decisions, while long-term investors are affected only by two push factors the U.S. T Notes and the VIX index. This result confirms that in the event of shock that causes changes in the four explanatory variables, the short-term investors react more quickly 4 Mild and Moderate = shock for five consecutive days (1 week); Crisis = shock for 10 consecutive days (2 weeks).

42


to respond the shock. In other words, in case of shock, the market became very volatile as a result of the short-term investors response. Thus, amid the global financial markets that are still vulnerable due to the high uncertainty in the Euro area and the high U.S. government debt, the future prospects of the capital flows on GS market still faces challenges especially the market response on the upcoming shock. The high influence of the global factors in the GS market would lead to high volatility in the GS market. Foreign investor particularly the short-term investors will respond to the shock by making sales that may disrupt the stability of the overall financial markets and the stability in foreign exchange market, which in turn may affect the stability of exchange rate.

V. CONCLUSION This paper provides several important empirical findings. First, the foreign investors in GS market are highly influenced by the global risk factor (push factors). The similar long-term and shortterm investors proportion (46:34) shows that the push factors reflected in the low 5 years yield US T Notes and global risk appetite (VIX), play an important role on their investment decision. Anytime the shock occurs in global financial market, the foreign investors will respond it by massive selling, which potentially disrupts the stability of domestic financial market and the exchange rate. Second, in addition to the push factors, the behavior of the short-term investor is also influenced by pull factors (return and domestic risk), which are reflected with 5 years GS yield and interbank money market (PUAB ON) rate. The increase of GS yield will keep encouraging foreign capital inflow to GS domestic market, while the increase of PUAB ON interest rate will lead to the decrease of foreign capital inflow. The larger number of variables affecting the short-term investor transactions decision relative to the long term investor implies they are more reactive to respond the current shock. The simulation result shows the prospect of foreign investors in GS market still have several challenges in the future, especially the vulnerability of the global financial market due to uncertainty. The strong effect of the push factor on foreign investor transactions shows that the GS market will remain face a relatively high volatility as the impact of foreign investors response on the upcoming shock, especially the short-term investors. These conclusions, lead to some policy implication and recommendations; first, Bank Indonesia and the Government need to continue the effort to keep the domestic return to be more competitive and to manage the investment risk at relatively low, as well as keeping the sustainability of domestic economy, in order to keep the foreign investors place their investment on domestic financial market. Second, Bank Indonesia and the Government should cooperate in formulating a contingency plan to keep the stability of the GS market in the case of excessive volatility due to foreign investor responses, particularly in the worsening of global financial market condition.


43

REFERENCES

Campbel JY and LM Viceira (2001). “Strategic Assets Allocation: Portfolio Choice for Long-term Investors,” http://ciber.fuqua.duke.edu/~charvey/Teaching/BA453_2006/Campbell_Viceira. pdf Chuhan, P, S. Claessens and N. Mamingi (1993). “Equity and Bond Flows to Asia and Latin America; The Role of Global and Country Factors,” Policy Research, Working Papers, WPS 1160. Culha, Ali Askin (2006), “A Structural VAR Analysis of the Determination of Capital Flow into Turkey,” Central Bank Review, Central Bank of the Republic of Turkey. Cuthbertson, K and D. Nitzsche (2005), “Quantitative Financial Economics: Stocks, Bonds and Foreign Exchange,” John Wiley and Sons, Ltd, Second Edition. Forbes, KJ. and FE. Warnock (2011). “Capital Flow Waves: Surges, Stops, Flight, and Retrenchment,” National Bureau of Economic Research, Working Paper 17351, August 2011. Fratzscher, Marcel (2011). “Capital Flows, Push Versul Pull factors and the Global Financial Crisis,” National Bureau of Economic Research, Working Paper 17357, August 2011. Goldfajn, Ilan and A. Minella (2005), “Capital Flows and Controls in Brazil: What have we learned?”, National Bureau of Economic Research, Working Paper 11640, September 2005. Gyntelberg J, M. Loretan, T. Subhani, and E. Chan (2009) “International portfolio rebalancing and exchange rate fuctuations in Thailand,” BIS Working Papers No. 287. Kaminsky, G. and S. Schumukler (2002), “Emerging market Instability: Do Sovereign Ratings Affect Country Risk and Stock Returns?” The World Bank Economic Review, Vol. 16, No. 2, pp 171-195. Miller, NC and N. Whitman (1970). “A Mean-Variance Analysis of United States Long-term Portfolio Foreign Investment,” The Quarterly Journal of Economics, Vol. 84, No. 2 (May, 1970), pp. 175-196 Ostry, JD et al (2011), “Managing Capital Inflows: What Tools to Use? IMF Staff Discussion Note, April 2011.

44


IMF (2011), “Recent Experiences in Managing Capital Inflows-Cross-Cutting Themes and Possible Policy Framework,” Prepared by the Strategy, Policy, and Review Department. Vita, Glauco De and KS Kyaw (2007). “Determinants of Capital Flows to Developing Countries: A Structural VAR Analysis,” Journal of Economic Studies, Vol. 35 No. 4. 2008, pp 304-322. Ying, YH. and Y. Kim (2001). “An Empirical Analysis on Capital Flows: the Case of Korea and Mexico,” Southern Economic Journal, April 2001; 67, 954-967. Agung, Juda., Nugroho, M.N., Yanfitri. 2011. “Arus Modal Jangaka Pendek di Indonesia Pasca Krisis Global: Karakteristik, Prosepek, dan Respon Kebijakan”. Working Paper Bank Indonesia. Akitoby, B. dan T. Stratmann, 2006, “Fiscal policy and Financial Markets”, IMF. Boot, A. et al., 2005, “Credit Rating as Coordination Mechanism”, Oxfords University Press. Cantor, R. dan F. Packer, 1996, “Determinants and Impact of Sovereign Credit Rating”. FRBNY Economic policy Review/October 1996. Cavallo, E., et al., 2008. “Do Credit Rating Agencies Add value? Evidence from the Sovereign Rating Business. Inter-American Development Bank.Edwards, S., 1985. “The Pricing of Bonds and Bank Loans in International Markets: An Empirical Analysis of Developing Countries’ Foreign Borrowing”, National Bureau of Economic Research. Hartilius, K., K. Kashiwase. and L/E. Kodres (2008), “Emerging marketspread Compressions: Is it Real or is it Liquidity? IMF Working Paper 08/10 Jaramillo, L. dan C.M. Tejada, 2011, “Sovereign Credit Ratings and Spread in Emerging market: Does Investment grade Matter?”, IMF. Jeramillo, L., 2010, “Determinants of Investment grade Status in Emerging market”, IMF. Kaminsky, G. dan S. Schmukler, 2001. “Emerging markets Instability: Do Sovereign Rating Affect Country Risk and Stock Return?”, World Bank. Kiff, J. et al., 2012. “Are Rating Agencies Powerful? An Investigation into the Impact and Accuracy of Sovereign Ratings”. IMF. Lee, K., et al. , 2010. “Sovereign debt Rating Changes and Stock Liquidity around the World”, Korea University Business School. Powel, A. dan J.F. Martinez, 2008, “On Emerging Economy Sovereign Spreads and Rating”. Inter-American Development Bank. Rozada, M. G. dan E.L.Yeyati, 2011, “Risk Appetite and Emerging market Spreads”. Universidad Torcuato Di Tella. Rozada, M.G., 2005, “Global Factors and Emerging marketSpreads”, Universidad Torcuato Di Tella.


45

appendix ��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

46


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

47

48


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

49

50


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

51

52


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��


��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

53

54


This page is intentionally left blank

The Role Of Asean Exchange Rate Unit (Aeru) For Asean-5 Monetary Integration : An Optimum Currency Area Criteria

55

the role of asean EXCHANGE rate unit (aeru) for asean-5 monetary integration: an optimum currency area criteria Dimas Bagus Wiranata Kusuma, Syed Mohammed Abud Ashif, Ali Musa Harahap, Muhammad Alam Omarsyah1

Abstract

The idea for regional monetary integration is grounded by the process of convergence theory within the member states. The paper analyses the possibility of monetary union in ASEAN-5 countries, Indonesia, Malaysia, Philippines, Thailand, and Singapore. In terms of volatility,by using nominal deviation indicator assessment, the ASEAN-5 currencies are suggested to peg their national currencies into Yuan since it empirically brings the lowest level of volatility, both during normal and crisis periods.Therefore, Yuan could be proposed as the anchor currency for ASEAN-5 countries. Moreover, valuing the AERU in terms of a weighed average of Yuan is important to determine which countries are considered to be an Optimum Currency Area (OCA). The results statistically suggest that all ASEAN-5 countries could be grouped as OCA according to exchange rate stability criterion.

Keywords : Optimum Currency Area, AERU, ASEAN-5, Exchange Rate Stability JEL Classification : D81, E52, F15, F36

1 Kulliyyah of Economics and Management Sciences, International Islamic University Malaysia; [email protected] (+60102906105), [email protected] (+60-182893070), [email protected] (+60-172905529), [email protected] (+60-173984722).

56


I. INTRODUCTION The possibility of creating an ASEAN currency unit (ACU) is a further step of the ASEAN “vision 2020”2. In May 2006, Hyderabad, India, the finance ministers of ASEAN+3 agreed to pursue a study on creating regional monetary units (RMUs) at the ASEAN+3 Finance Ministers’ meeting. They urged to take steps to coordinate their currencies in a way to produce a common regional currency similar to the Euro. The regional financial crises in 1997/1998 eroded the credibility of unilateral fixed exchange rate in ASEAN and then renewed calls for greater monetary integration and then ultimately enhanced regional exchange rate stability3. With respect to exchange rate stability, Euro member countries had adopted the Exchange Rate Mechanism (ERM) or European Monetary System (EMS) over the previous two years prior involvement in Euro. Under such mechanism, participating countries are able to maintain their exchange rate movement within bilateral limits of plus minus 2.25 percent. The table below shows some bilateral nominal ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

2 The ASEAN Heads of States/Governments adopted the Declaration of ASEAN Concord II (Bali Concord II) in 2003, which establishes an ASEAN Community by 2020. ASEAN community will reinforce ASEAN’s centrality and role as the driving force in charting the evolving regional architecture. ASEAN vision 2020 was declared in Kuala Lumpur in December 1997 that decided to transform ASEAN into a stable, prosperous, and highly competitive region with equitable economic development, and reduced poverty and socio-economy disparities. 3 According to World bank (2000), cost recapitulation for crises over GDP were recorded as follows: Indonesia (58% GDP), Malaysia (10%GDP), Thailand (30%GDP), and Korea (10%GDP). 4 Denotes Single European Act which specifically targeted regional issues, recognizing that redistribution of economic resources from richer to poorer areas of the EC was essential in order to achieve harmonious economic integration.


57

exchange rates against ERM currencies. The results suggested over period 1974 to 1990, all joined countries majority are able to keep their currencies volatility less than 2.5 percent, and in average European Countries became less variable against one another, as well as, against US dollar and the Japanese Yen. The question remains is whether ASEAN is ready to establish such monetary union? Such question is relevant when considering the feasibility of ASEAN Currency Unit measurement, which requires the region to satisfy the theory of optimum currency Area (OCA). The definition of this theory is ascribed from the presence of economic convergence criteria of each member state’s economy as an entry condition for union establishment. Later, convergence criteria was ratified by Maastricht Treaty in article 109j, which defines some macroeconomic indicators as convergence measurements, namely; price stability, soundness and sustainability of public finances, exchange rate stability, and convergence in long-term interest rates. Therefore, this study focuses on one parts of above mentioned convergence criteria, namely exchange rate stability achievement. The current crisis in Eurozone,it is not an amazing result nor unprecedented. The basic problem is because of the relaxation on their entrance exams in 1998. Therefore, we may say that the entrance criteria have very little to do with economics, and very much with the politics. During the 1990s, the governments of most EU-countries had made a strong political commitment to go ahead with monetary union. By 1999, large number of countries committed to the monetary union would fail the entrance criteria, and only a few countries would succeed. So politics prevailed and the annoying Maastricht numbers were set aside – which was the right decision and conclusively showed that the Maastricht convergence criteria are irrelevant. Some evidences are denoted by comparing state’s participation criteria fulfillment conditions at the end of 1996 and 1997 when they became subject to assessment, as follows: ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

� �

� �

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�� Banking, �� January�� 58��Bulletin of Monetary, Economics and 2013

��

��

��

�

�

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

� �

� �

��

��

��

��

��

� � ��

��

��

��

��

��

��

��

��

� ��

� ��

��

��

��

��

��

��

��

��

� � � � � � � � � � � � � �

� � � � � � � � � � � � � �

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

��

��

��

��

��

��

��

��

�

�

��

�� the monetary �� union�� processsince � �the For ASEAN �� countries, is theoretically still a long �� shall�� basis. ��Referring �� Bayoumi �� et al�(2000),�the monetary integration be undertaken in a gradual �� development �� and monetary �� obstacle � dissimilarity of the�� level of��economic system �� is the main �� to support economic and monetary integration. The similarity of past macroeconomic policies, �� financial �� systems �� would ��increase �� the integration �� possibility, � stage of economic�� development and �� to thereforethe proposed of ASEAN-5 countries are selected into analysis as the initial step �� form a currency union in ASEAN region. This kind of approach had been tried by Europe when �� they established a union whereas the European Monetary Union (EMU) invited only four of its major members5. They were included as they have comparable population, size, resources, and economies (Day and Herbig, 1995).

In the context of ASEAN-5, these group countries make up over 72 percent in terms of total population inhabitants compared to ASEAN-106. Meanwhile, the degree of economic development in ASEAN-5 is homogeneous and dominant, particularly, if we discern on their trade volume of around 92 percent of the total ASEAN-10 volume of trade. Size of GDP and international reserve posit a tremendous portion for affecting economic policies in the region as almost 96 percent in 2008. In summary, the ability of ASEAN-5 to work together based on above common indicators is apparent and it would sustain the common goal of a successful regional economic cooperation, finally benefiting gradually all participating economies as well as peripheral economies in the region.

5 During the 1950s (initial stage establishment of the European Union), three regional European organizations were form: the European Coal and Steel Community (ECSC), the European Economic Community (EEC), and the European Atomic Energy Community (Euratom). Initially, six states were involved in the formation of these organizations: Belgium, France, Italy, Luxembourg, the Netherlands, and West Germany (the German Federal Republic) where these countries have similar economic stage of development. 6 ASEAN-10 denotes Brunei Darussalam, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Thailand, Vietnam.


59

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�� 60 Bulletin of Monetary, Economics and Banking, January 2013 ��

��

��

��

��

��

��

��

��

��

��

��

��

Several researches have been carried out regarding the reliability and the possibility of ASEAN-5 towards enhanced economic and monetary unification. McAleer and Nam (2005), highlight suitability of establishing a common currency area for ASEAN-5 from the perspective of contagion. They find out (1) contagion was present between all country pairs in ASEAN-5; an indication that the degree of correlation among ASEAN-5 economies had increased during the Asian financial crisis, (2) closer monetary co-operation among ASEAN-5 economies would be feasible. In addition, Ramayandi (2005) discusses on issues and prospects of ASEAN monetary union. In other words, ASEAN-5 is found to be suitable for a monetary co-operation due to their relative symmetrical economic shocks and trade patterns. Finally, monetary union or integration in ASEAN-5 countries is an important step towards economic integration. It could be implemented if ASEAN-5 countries have pursued policy coordination on their exchange rates policies. On this paper, possible research questions are highlighted to address the ASEAN-5 monetary union into existence, first, how to develop the hypothetical currency unit in ASEAN-5 countries, called ASEAN Exchange Rate Unit (AERU) under normal and crisis periods?; second, is AERU stable by pegging its value fix against currency baskets (US$-Euro-Yen-Yuan, US$-Euro-Yen, US$-Euro), and Individual peg currency (US$, Euro, Yen, Yuan, and Singapore Dollar) under normal and crisis periods?; and third, who are the proposed members to form Optimum Currency Area in ASEAN-5 economies? This paper is organized as follows;section two covers the theoretical review. Section three discusses the data and methodology including the calculation steps on AERU, while section four presents the result and analysis. Conclusion and policy recommendation is presented on the last section and will close the presentation.

II. THEORY The Theory of Monetary Union and the Convergence Theory The modern and comprehensive thought regarding OCA theory was initially explained by Robert Mundell in his seminar paper entitled “A Theory of Optimum Currency Area” in 1961. It defines as the optimal geographic domain of a single currency, or of several currencies, whose exchange rates are irrevocably pegged and might be unified. Later in the latest decade, this idea was developed whereas the member states must fulfill the requirements of OCA


61

characteristics. Frankel and Rose (1998) precisely stated that OCA characteristics could be satisfied endogenously. In other words, a group of countries could not meet one or more OCA criteria ex-ante, but ex-post. As OCA closes into the integration idea, Warjiyo (2004) explicitly compares the cost and benefit reaped by joining monetary union. According to Table 4, the main benefit of unification would be the symmetric response towards the onset of shocks. As the economic convergence is achieved, the cost and the threat faced by the country members can be aptly reduced. Meanwhile, the cost is related with the short term adjustment process towards convergence policy. Once the member states can structurally remove the process by fulfilling the stipulated requirement by union, the cost will be gradually eliminated.

��

��

��

��

��

��

��

��

��

The criteria singled out by the “old” OCA theory share a common rationale: by joining a monetary union (MU), a country gives up the possibility of adjusting its nominal exchange rate in response to macroeconomic shocks. The lesser the need for an economy to adjust the nominal exchange rate, the lesser the cost of joining a monetary union. While the “old” OCA theory operates in a “reduction of damages” perspectives, the “new” theory weighs the benefits of OCA membership against its costs. Under the “new” theory, the exchange rates will converge as an outcome, rather than a prerequisite of an OCA membership.

The Rational for Monetary Union The primary aim for establishing the monetary union is exchange rate stability. Yuen (1999) illustrates that the creation of the European Monetary System (EMS) was a response to the external and internal monetary instability of the late 1970s, and the constant feature of the EMS

62


was the quest for stability. In addition, according to Gros and Thygesen (1998), the increasingly tight management of the exchange rate mechanism led to a reduction in the external element of monetary instability to about one-quarter by 1990 in the EU. According to Mundell (1961), the presence of monetary union would bring out economies of scale during the implementation period.It is because, first, when a small country fixes its currency to that of a larger country with an acceptable exchange rate fluctuation, it sets the course for the rest of its macroeconomic policies so that leads to move towards convergence trajectory. Second, the more countries join a currency area, the smaller the proportion to its output of any internal or external disturbance. Third, because money is a unit of account, there are economies of information and convenience in currency unions, and consequently the more countries that join a currency area, the more efficient it will be. Another objective of proposing monetary union is to strengthen monetary policy coordination by setting some member’s currency being pegged into the same basket of currency. Kuroda and Kawai (2002) point out that the creation of Asian Currency Unit (ACU) is likely to act as a statistical indicator summarizing the collective movement of Asian Currencies. This would enable the participating countries to stabilize their exchange rates against the ACU basket and improve the understanding for monetary and exchange rate policy coordination. Kawai, Ogawa, and Ito (2004) suggest that first the monetary Authorities of Asian countries should discuss the exchange rate issue as a part of their surveillance process. The exchange rates of these currencies are linked by terms of trade and competitive prices. Ogawa, Kawasaki and Ito (2002) pointed out possible failure on coordination in choosing an exchange rate system and policy if one country member choosesto peg their currency to USD, since it may has an adverse price effects.

ASEAN-5 Exchange Rate Regimes The Asian currency crisis taught us that the various arrangements of currency regimes would induce coordination failure and finally endanger intra-regional exchange rate stability within the region. Ogawa and Yoshimi (2007, 2008) used the methodology of Frankel and Wei (1994) to investigate actual exchange rate systems and policies conducted by the monetary Authorities of East Asian countries during a period from 1999 to 2007. The study empirically examines what linkage trends each ASEAN-5 currency actually has with three major currencies: the US, the euro, and the Japanese yen. By taking a closer look of above description, we conclude that the ASEAN-5 has a variety of linkages with their major currencies. Under the OCA theory, such condition might reflect different economic interests across the economies.


63

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Classification of ASEAN-5 Economies using Clustering Approach Clustering analysis is a multivariate technique to group research objects based on its characteristics, and provides a homogenous cluster. In the context of monetary union, such approach can be used to estimate which pairs of countries fulfill the OCA properties. Falianty (2005) investigates the feasibility of forming currency union in ASEAN-5 countries by utilizing clustering method. The results show that there are two clusters (groups) that exist, namely group 1: Indonesia and the Philippines, and group 2: Malaysia, Singapore and Thailand. She suggested the currency union is optimum for Singapore, Malaysia, and Thailand. The results strengthen the conclusion that Indonesia and the Philippines are fall behind the OCA cluster in ASEAN-5. In addition, Achsani (2010) was testing the feasibility of ASEAN+3 to form a single currency by undertaking OCA and clustering approach. By using the OCA index, the results present that the Singapore dollar is the most stable currency with the lowest OCA index relative to Malaysian ringgit, Thailand baht, Philippines peso, and Indonesian rupiah. If the ASEAN-5 will establish a single currency, the process should start with integrating Singaporean dollar and Malaysian ringgit and then followed by Thailand baht and Philippines peso. The OCA indices for Indonesia rupiah are extremely higher than the other countries, indicating difficulties to join directly the single currency. On the other hand, the findings of the hierarchical clustering consider two consecutive process; first currency unification includes Singapore and Malaysia, and second, currency unification Thailand and the Philippines. This also highilights the IDR is inappropriate in joining the single currency due to its high dissimilarity with other currencies in the region.

64


III. METHODOLOGY Data and Variables ASEAN-5 countries are seeking the optimal framework in order to reach the ultimate goal of unification, namely the ASEAN-5 monetary union. Three possible basket scenarios are proposed to adopt, (1) the Dollar, Euro, and Yen (DEY)7 and ASEAN Exchange rate Unit (AERU)8, (2) a currency basket composed of ASEAN-5 currencies (AERU), (3) or a strategy of regional monetary integration could make use of both kinds of baskets (Kawai, Ogawa, and Ito 2004). To gain the benefit of AERU in fostering greater intra-regional exchange rate stability, there are five pertinent issues addressed throughout this chapter, particularly associated with the methodology applied to form an AERU hypothetical index and also index for surveillance purposes. First is determining AERU weights; second is calculating the benchmark exchange rate for each ASEAN-5 currency in terms of proposed currency basket; third is calculating the anchor for ASEAN-5 monetary union; fourth is measuring the volatility performance over numerous AERU arrangements; and fifth is assessment of participation in AERU based on ERM II9. This research utilizes the following economic variables in order to form an optimum currency area in ASEAN-5 countries: 1. Trade volume (in million dollars) which consists of the volume of export and import. This variable is based on each trade direction across ASEAN-510 countries. In addition, the trade volume is used to compute the weigh of each currency in the basket for calculating the AERU, both in normal or crisis period. The trade direction also accounts for the trade volume between ASEAN-5 countries with some developed nations, such as US, European Union11, Japan, and China12. The data span from 2004 to 2010 on yearly basis. 2. Nominal GDP (in million dollars) of each ASEAN-5 countries. This variabelis used to calculate the weigh of each observable currency in the basket, both in normal and crisis period. The data spans from 2004 to 2010 on yearly basis. 3. GDP measured at Purchasing power parity (in million dollars) of each ASEAN-5 countries, and used as one of economic criteria to calculate the weigh of each observable currency in the basket. The data spans from 2004 to 2010 on yearly basis.

7 DEY is a common basket based on own trade pattern. 8 AERU grouped as individual-country baskets because it is based on common currency basket weights within the region (Castel et all, 2007). 9 The Exchange Rate Mechanism (ERM II) was set up on 1 January 1999 as a successor to ERM to ensure that exchange rate fluctuations between the euro and other EU currencies do not disrupt economic stability within the single market, and to help non euro-area countries prepare themselves for participation in the euro area. The convergence criterion on exchange rate stability requires participation in ERM II. 10 ASEAN-5 countries denotes for Indonesia, Malaysia, the Philippines, Thailand, and Singapore. 11 The European Union is United Kingdom, Germany, the Netherlands, France, Italy, Belgium, Luxembourg, Denmark, Ireland, Greece, Spain, Portugal, Austria, Finland, Sweden, and other countries in union. 12 Trade direction accounted is also including Hong Kong and China Taipei into RRC.


65

4. International reserve minus gold (in million dollars) as one economic criterion to for the weigh of each observable currency in the basket. The data span from 2004 to 2010 on yearly basis. 5. Baht against US dollar (USD/THB), rupiah against dollar (USD/IDR), ringgit against dollar (USD/MYR), peso against dollar (USD/PHP), and Singapore dollar against dollar (USD/SGD). The data spans from 2004 to 2010 on daily basis. 6. IDR13/AERU, MYR/AERU, PHP/AERU, THB/AERU, SGD/AERU. All data spans from January, 02 2004 to October, 21 2011 on daily basis. 7. US$-Euro/AERU is defined as the value of the AERU in terms of a weighed average of the US dollar and the euro. The data spans from January, 02 2004 to October, 21 2011 on daily basis. 8. US$-Euro-Yen-Yuan/AERU is defined as the value of the AERU in terms of a weighed average of the US dollar, the euro, the yen, and the yuan. The data spans from January, 02 2004 to October, 21 2011 on daily basis. 9. US$-Euro-Yen/AERU is defined as the value of the AERU in terms of a weighed average of the US dollar, the euro, and the yen. The data spans from January, 02 2004 to October, 21 2011 on daily basis. 10. The rate of USD, Euro, Yen, Yuan and Singapore dollar against hypothetical AERU, defined as the value of the AERU in terms of a weighed average of these currencies. The data ranges from January, 02 2004 to October, 21 2011 on daily basis. All data are obtained from International Financial Statistics (IFS), Bank Indonesia (BI), Bank Negara Malaysia (BNM), Bank of Thailand (BOT), Central Bank of the Philippines (CBP), Monetary Authority of Singapore (MAS), and PACIFIC Exchange Rate Services.

The Evaluation in Adopting Monetary Cooperation Proposal in ASEAN-5 Countries When decision is made to seta new currency then one needs to select an anchor, either single or a basket of currencies. The choice depends not only on trade flows but also on the dominance of the dollar in international trade, finance, and in the pricing of commodities. The selected anchor must have high credibility in the presence of high capital mobility;therefore the link is not subject to speculative attacks. There are some issues associated with the use of the AERU as a divergence indicator and an instrument for policy coordination. The first is that the regional countries should establish consensus on technical issues including the assignment of currency weighs, the base year 13 IDR = Indonesian Rupiah; MYR = Malaysian Ringgit; PHP = Philippines Peso; THB = Thailand Bath; SGD = Singapore Dollar.

66


selection, the grouping of currencies, and period of weigh revision. They have to be set correctly in order to resolve the asymmetry problem. For instance, the smaller the currency weigh, the larger the exchange rate fluctuation of a particular country against AERU will be; therefore the higher the intervention burden will be (Moon, et al. 2006). Later Moon, et al (2006) suggest that regional monetary authorities should provide weighs using the combination of GDP based on PPP-value, intraregional trade, and contribution to regional monetary and financial cooperation (e,g., Chiang Mai contribution). In addition, Ogawa and Shimizu (2005) proposed the following basket weighs of AERU, which were adopted from ECU: 1. Trade volume is calculated as a total of export and import volumes from the direction of trade statistics, and central bank of each ASEAN-5 countries. 2. Nominal GDP. 3. GDP measured at Purchasing Power Parity (PPP) is used because the nominal GDP does not always reflect international differences in relative prices. 4. International reserve (minus gold) as indicators of basket weighs from a viewpoint of financial aspect comparison. The proposed currency union is hopefully able to adopt the precedent and un-precedent pressure into a currency union. For such purpose, the coming established currency union has to be able to maintain their stability and volatility over various periods or we can call them as normal and crisis periods. Since the periods of observations range from 2004 to 2010, we divide them into two circumstances based on the standard deviation for normal period as well as for crisis period. Another important technical issue in forming the AERU index is the choice of base year to calculate the exchange rate benchmark14. One of the most popular ways is to choose the year when a fundamental equilibrium of both internal and external sectors is achieved. In other words, the base year is chosen when the total international transaction of the country members countries close to being balanced. On this paper, we use the balance of trade (in US dollars) of ASEAN-5 countries against several combination of the following trading partner: US, EU, Japan, China, and Singapore. The yearly data observed spans from the year 2004 to 2010. The second consideration is the variety of exchange rate systems. Different exchange rate systems among the countries can distort the role of AERU as divergence indicator. Without an appropriate mechanism to reflect this exchange rate systems differences, the divergence indicator may not function properly as a surveillance mechanism. Currently, many ASEAN-5 countries are still using the US dollar as their anchor. The possible suggestion is that the anchor currency should reflect the trade volume of the main trade partner, therefore will consists of hard currencies, such as Singapore dollar. 14 The chosen year based on internal and external equilibrium of trade takes an assumption that a one-year time lag before changes in exchange rates might affect trade volumes. The exchange rate of the AERU in terms of various currency baskets is set unity for the base year.


67

Watanabe and Ogura (2006) have studied on the Regional Monetary unit (RMU). However, given that a currency union takes long process to be realized, they propose to create AERU. Eichengreen (2006) calls this a parallel currency approach. The present paper considers the method of Ogawa and Shimizu (2005), which follows the same principle of the European Currency Unit under the EMS, which is, computed as the weighed average of each country’s currency within the region. In the same way that the ECU was defined as a basket of currencies of EU member countries, the AERU is defined as a basket currency of the ASEAN-5 (Indonesia, Malaysia, the Philippines, Singapore, and Thailand).

Calculation Steps According to Ogawa and Shimizu (2005), for the benchmark period, the exchange rate of the AERU in terms of various baskets arrangement is set at unity. Setting for unity implies that a weighed trade proportion is set on a hundred percent for each benchmark calculation separately. Then, several steps are incorporated into calculation as follows:

Step 1: Determining the AERU Weights As mentioned previously, that this paper compares four different economic size indicators and then tries to select which two out of four indicators has the highest stability when each of them is applied on every single currency basket observed15, either in normal or crisis period. Generally, the weigh of the basket is supposed to represent the weigh of the country’s economic importance and contribution to economic cooperation in the region. Following Ogawa and Shimizu (2005), we use four different economic size indicators16, then calculate the optimal share weighs17 for the last three or four years. Since the present paper is comparing the best currency peg for ASEAN-5, the normal period and crisis period is set separately by taking three years average of normal period18 (2004-2006) and four years average of crisis period19 (2007-2010). The AERU weigh for each ASEAN-5 currency = Average Benchmark Exchange Rate for each ASEAN-5 country based on the each observable currency basket X economic size indicator20. 15 We observe eights different basket currencies, namely (1) US$-Euro, (2) US$-Euro-Yen-Yuan, (3) US$-Euro-Yen, (4) US$, (5) Euro, (6) Yen, (7) Yuan, (8) Singapore Dollar. 16 They used 1) trade volume; 2) Nominal GDP; 3) GDP measured at Purchasing Power parity; 4) International Reserves (minus Gold). From the stand point of stability vis-à-vis the US$-Euro basket currency, the PPP measured GDP and trade volume were chosen as weights. 17 It is calculated by comparing the standard deviation of each the value of AERU in terms of a weighted average the numerous exchange rate regimes. Then, the two lowest of economic size indicators are incorporated by computing theirs arithmetic shares for normal as well as crisis period separately. 18 Three years for normal are determined by considering one year before and after the year of normal period, which is set in 2005 for ASEAN-5 national currencies against US dollar. 19 Three years for crisis are determined by considering one year before and after the year of crisis period, which is set in 2008 for ASEAN-5 national currencies against US dollar. 20 See Ogawa and Shimizu (2005).

68


For example, the AERU weigh in terms of US$-Euro/AERU for IDR = Average benchmark for IDR in terms of US$-Euro X trade volume. We then can call it as the IDR weigh in terms of US$-Euro.

Step 2: Calculating the benchmark exchange rate for each ASEAN-5 currency in terms of proposed currency baskets21 Benchmark22 exchange rate calculation for each ASEAN-5 currency in terms of (US$Euro/AERU) = average of ((USD/Each ASEAN-5 Currency X a weighed trade proportion23 with US) + (EURO/Each ASEAN-5 Currency X a weighed trade proportion with European Union)). For example: Benchmark of IDR for US$-Euro/AERU = average of ((USD/IDR X 58%) + (Euro/IDR X 42%)).

Step 3: Calculating and selecting the anchor for ASEAN-5 monetary union The value of AERU in terms of a weighed average of the US dollar and the Euro (US$-Euro/ AERU) is calculated as follows: US$-Euro/AERU24= ((IDR weigh)*(US$-Euro/IDR)) + ((MYR weigh)*(US$-Euro/MYR) + ((PHP weigh)*(US$-Euro/PHP)) + ((THB weigh)*(US$-Euro/THB)) + ((SGD weigh)*(US$Euro/SGD))25

Step 4 :Measuring the Volatility Performance over Numerous AERU Arrangements The less volatility of AERU against selected currencies basket is better alternative for providing the best anchor in pursuing the optimum currency area criteria. An OCA criterion is maintaining the exchange rate stability stipulated in the Maastricht Treatyof plus minus 15 percent from central parity. 21 The same formula is applied for other currency arrangements, namely (1) US$-Euro-Yen-Yuan/AERU, (2) US$-Euro-Yen/AERU, (3) US$/AERU, (4) Euro/AERU, (5) Yen/AERU, (6) Yuan/AERU, (7) Singapore dollar/AERU. 22 The benchmark period refers to the year which the total international transaction of the members countries are as close to being balanced as possible and their balances with the rest of the world are also small as possible. The base year calculation uses the balance of Trade (exports volume minus imports volume in US dollars) of ASEAN-5 countries (1) within ASEAN-5 countries, (2) with US, EU, Japan, and China, (3) with US, EU, and Japan, (4) with US and EU, (5) with US, (6) with EU, (7) with Japan, (8) with China, and (9) with Singapore. The result is 2007 and 2008. 23 It is accounted from the proportion of balance of trade of ASEAN-5 over the several countries trade partners in percentage point, namely the proportion of ASEAN-5 balanced trade (1) with US, EU, Japan, and China, (2) with US, EU, and Japan, (3) with US and EU. 24 AERU represents Indonesian Rupiah (IDR), Malaysian Ringgit (MYR), the Philippines Peso (PHP), Thailand Bath (THB), and Singapore Dollar (SGD). 25 The same formula is applied for other currency arrangements, namely (1) US$-Euro-Yen-Yuan/AERU, (2) US$-Euro-Yen/AERU, (3) US$/AERU, (4) Euro/AERU, (5) Yen/AERU, (6) Yuan/AERU, (7) Singapore dollar/AERU.


69

In the literature, different approaches for measuring exchange rate volatility have been applied, however there is no consensus on which measure is the most appropriate. Usually, the average absolute difference between the previous period forward rate and the current spot rate is considered to be the best indicator of the exchange rate volatility. The present paper follows Ogawa and Shimizu (2006) to measure on exchange rate volatility, namely nominal deviation indicator (NDI). It indicates how far each ASEAN-5 currency deviates from the benchmark exchange rate in terms of the AERU. The nominal AERU deviation indicator is calculated as follows:

ܰ‫ݎ݋ݐܽܿ݅݀݊ܫ݊݋݅ݐܽ݅ݒ݁ܦ݈ܽ݊݅݉݋‬ሺΨሻ ൌ ܾ݄݁݊ܿ݉ܽ‫ܷܴܧܣ݂݋݁ݐܽݎ݄݁݃݊ܽܿݔ݁݇ݎ‬ ‫ܷܴܧܣ݂݋݁ݐܽݎ݄݁݃݊ܽܿݔ݈݁ܽݑݐܿܣ‬ ൰ െ൬ ൰ ൬ ܽܿ‫ݕܿ݊݁ݎݎݑ‬ ܽܿ‫ݕܿ݊݁ݎݎݑ‬ ܾ݄݁݊ܿ݉ܽ‫ܷܴܧܣ݂݋݁ݐܽݎ݄݁݃݊ܽܿݔ݁݇ݎ‬ ൰ ൬ ܽܿ‫ݕܿ݊݁ݎݎݑ‬

Furthermore, this paper proposes the criteria as the threshold to categorize the degree of volatility into low, medium and high volatility. The optimal threshold is set to resolve the asymmetry problem and to strengthen symmetry policy response among member countries. Moon (2006) suggests that the smaller currency weigh of a country is, the larger the exchange rate fluctuation of the country against the AERU will be, vice versa. He later suggests that every country weigh should be not more than 0.33 to maintain national currency stability against currency union. Therefore, the upper limit to be agreed upon is 0.33 in order to limit the overwhelming fluctuation in a currency union. The measurement of volatility is covering AERU against basket currencies as well as AERU against ASEAN-5 currencies, respectively. In addition, to categorize the degree of volatility among ASEAN-5 currencies and common basket currencies against AERU, the rules below are set as follows: 1. If the number of nominal deviation exceeds the allowed bands26 and the average number of nominal deviation is within 0.67 - 1, it is categorized as having high level of volatility. 2. If the number of nominal deviation exceeds the allowed bands and the average number of nominal deviation is between 0.34 - 0.66, it reflects medium level of volatility. 3. If the number of nominal deviation exceeds the allowed bands and the average number of nominal deviation is within 0 - 0.33, it indicates as low level of volatility.

Step 5 :Assesment of participation in AERU based on ERM II Assesment of participation is set to provide a more published picture about exchange rate volatility and more serious data base for assessment as stated in Maastrcht Treaty. As stipulated 26 The threshold is calculated from plus minus 15 percent (ERM II) of nominal deviation indicator for normal period and crisis period.

70


in the Maastricht, exchange rate stability would have to be fulfilled by member state for at least recent two years and has not undergone devaluation. For such reason, the simulation participation which is formatted based on the most stable basket currency against AERU for ASEAN-5 countries. The upper and lower band are set by multiplying plus and minus 15 percent towards each benchmark currencies of each ASEAN-5 currency against AERU, both in normal and crisis periods. For normal period, the range is set from January 2004 to December 2005 on daily basis, while for the crisis period, the range spans from January 2008 to December 2009 on daily basis. Technically, to assess the participation criteria among ASEAN-5 countries under AERU arrangement, the paper utilizes average exchange rate movement (AERM), average number fluctuation within bands (ANFB), average number volatility (ANV), standard deviation exchange rate movement (SERM), and level of volatility (LV). All of those measurements would generate a conclusion on the exchange rate stability decision. In addition, the exchange rate stability decision is based on the rule set by nominal deviation in step 4.

V. RESULT AND ANALYSIS Descriptive Assesment for Monetary Union in ASEAN-5 Countries A descriptive assessment on asymmetric shock shows the possibility of ASEAN-5 to form a union. Overall, either supply and demand shocks seem similar, even though for Indonesia, they are relatively lower compared to the rest of the countries. In other words, the degree of symmetric shocks in AEAN-5 is compatible to proceed, considering the Euro zone shocks which appears to be only slightly lower27.

��

��

��

��

��

��

�

��

Figure 1. Symmetry of Demand and Supply Shocks in ASEAN-5 and EURO Zone

27 Indonesia is an outlier whose demand and supply shocks do not seem to be well synchronized with the rest of ASEAN countries.


71

In addition, this assessment is in line with many literatures that ASEAN-5 countries do not experience many asymmetric shocks. See for instance, Xu (2004) who computed the percentage of the variation in demand and supply shocks that can be attributed to common shocks. Following Blanchard-Quah, he applied structural VAR procedure and factor analysis to identify the supply and demand shocks, and then estimates the common component of these shocks28. The share of the total variation captured by this common component can be interpreted as expressing the degree of symmetry in the shocks. Therefore, ASEAN-5 are eligible to create an union based on the degree of similarity, and integration in terms of their economic structure which provide symmetric shocks.

The Weigh and Benchmark Exchange Rate Calculation This part covers the firstand the second step as outlined before, namely (i) determining the AERU weights, and (ii) calculating the benchmark exchange ratefor each ASEAN-5 currencies in terms of proposed currency basket. On addressing the weighs on each currency basket, the value of AERU would have been shared into several currencies. The shared average of particularly common basket currency is based on important trading partners with ASEAN-5 countries. As the common basket currency is quoted in terms of a shared average of the US$-Euro, the US$-Euro-Yen-Yuan, the US$Euro-Yen, and some individual basket currencies, the shares on those basket currencies are set at unity. ��

��

��

��

��

��

��

��

��

�

�

�

��

��

��

��

��

�

��

��

��

��

�

�

��

��

�

�

�

�

��

�

��

�

�

� �

��

�

�

��

�

��

�

�

�

��

�

��

�

�

�

�

��

��

On calculating the countries’ shares over various economic size indicators, we use the four distinct economic size indicators to determine the best weighs. This weigh will influence the amount of convertion rate for each national currency. 28 This approach is widely used, but is subject to an important criticism. This is that the shocks identified as demand shocks are in fact temporary shocks, while the shocks identified as supply shocks are in fact permanent shocks.

72


��

��

��

��

��

��

�

�

�

�

�

�

�

�

�

��

��

��

��

��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Table 7 shows that Indonesia has the highest shares in terms of nominal GDP and GDP measured at purchasing power parity (PPP). In contrast, Singapore records the highest shares in terms of trade volume and international reserve (minus Gold), either in normal or crisis period. Among the rest, the Philippines is accounted for the lowest shares over the four various economic size indicators. The benchmark period is chosen in order to calculate the benchmark exchange rate. The benchmark period is defined when the total balance of trade of ASEAN-5 is relatively close to zerowith the following trading partner: within ASEAN-5; with US, EU, Japan, and China; with US, EU, and Japan; with US and EU; with US; with EU; with Japan; with China; with Singapore. The table 8 shows the balance of trade of the ASEAN-5 from 2004 to 2010. It shows that 2008 is the year which majority of the balance of trade is close to zero. The purpose under such rule mentioned by De Grauwe (2007) is that the countries under monetary union will be able to pay their debts without creating surprise inflation and or devaluation, which reduces the real value of the debts, but increase its nominal value and downgrade government credibility. The absence of trade deficits is the indicator to show their commitments to preserve the union from defaults. ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��


73

Assuming a one-year time lag before changes in exchange rates affect trade volumes, we should select 2007 and 2008 as the benchmark periods. The benchmark exchange rate is calculated separately for each ASEAN-5 country in terms of the AERU. The benchmark for normal and crisis period is not different. The use of the benchmark is to calculate the conversion of each ASEAN-5 national currency against AERU (with various proposed basket currency, including for commodity peg currency). The calculation result is presented below. ��

��

�

��

�

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

��

��

��

��

��

�

��

��

��

��

��

��

�

��

��

��

��

��

��

�

��

��

��

��

��

��

�

��

��

��

��

��

��

��

According to table 9, benchmark of IDR is the smallest against AERU for every currency basket. On the other hand, SGD shows the biggest value of benchmark against AERU. These imply that the value of IDR in every basket currency must be the highest, and SGD is viceversa. Selecting the most stable among the four types of indicators is based on the lowest standard deviation in terms of rate of change (%). Therefore, the use of statistical measurement in rates of change (%) can captures the stability of ASEAN-5 currency against AERU (Ogawa Shimizu, 2005). Following Ogawa and Shimizu work (2005), we will take two most stable types of AERU and use them to calculate the weighs of currency basket in AERU for ASEAN-5 currencies. Table 10 provides the summary of selected best indicators over various currency baskets against AERU. The results shows that the nominal GDP and GDP measured at PPP are the most stable indicators across proposed currency baskets, either under normal and crisis periods. GDP, either nominal or real, is the best indicators as weigh since in a monetary union a country needs strong economic structures, accompanied by a credible macroeconomic policies, which lead to structural convergence.

74


��

��

�

��

�

��

�

��

�

��

�

��

�

��

�

��

�

��

��

��

��

Choosing the anchorover numerous AERU arrangements This part covers the third and the fourth of calculation step outlined before, namely the calculation of ancor currencies for ASEAN-5 monetary union, and measuring the volatility performance over numerous AERU arrangements. The nominal deviation indicator (NDI) is exercised to measure the degree of volatility of currency basket alternatives. The results is presented in Table 11, both for normal and crisis periods. ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��


75

Overall, the selected currencies are fluctuating during the two separated condition. It is evident that Yuan is the most stable currencies throughout the observation, both in crisis and normal period, with average volatility of 0.01. This leads us to choose Chinese Yuan as the currency anchor for AERU arrangement within ASEAN-5 countries. Beside NDI indicator, we also consider the average and the standard deviation of band fluctuation29. Empirically, table 12 shows the summary of currency movement over various currency baskets, either normal and crisis periods. In general, all ASEAN-5 currencies are moving or fluctuating against selected exchange rate arrangements under AERU. During normal period, interestingly all ASEAN-5 currencies tend to depreciate against selected currency arrangements under AERU calculation. This depreciation occurs under normal period in which during this period, all ASEAN-5 countries were promoting their export and benefited from existing depreciation. Differently, during crisis period, depreciation occurred when the ASEAN-5 countries pegged to US$-EuroYen, Yen, Yuan, and Singapore Dollar (SGD). Those currencies are not source of current global financial crisis and so much relies upon their economy on trade. Such that, once ASEAN-5 pegged to those currencies, ASEAN-5 economies are attempting to boost their exports and reduce their volume of imports which are needed to underpin the trade balance policy in the mid of crisis. In addition, under AERU framework, ASEAN-5 economies are yet buffering their economy by compounding their money supply, even though could elevate inflation rate. ��

��

�

��

�

��

�

��

�

��

�

��

�

��

�

��

�

��

��

��

��

29 The former and latter is calculated by taking average and standard deviation on exchange rate deviation from benchmark exchange rate over period of observations.

76


In brief, under monetary union or AERU framework, ASEAN-5 economies remain fragile for instability unless they keep strengthening coordination on their exchange rate management. Meanwhile, ASEAN-5 currencies tends to appreciate against US$-Euro, US$-Euro-Yen-Yuan, US$, and Euro. This is in line with the current situation in Europe and US economies. Under AERU, ASEAN-5 countries realized that holding risky currencies may eventually transmit their crisis into ASEAN-5 economies. Moreover, during crisis, US and European countries called back their currencies in order to reduce their money supply all over the globe and reduced their inflation rates. In other words, under AERU, member states should pegged the currencies to credible and economically sound currency to avoid the union being trapped into the same crisis or financial instability.

Assessment of Participation in AERU Based on ERM II This section covers the fifth step where we will asses the participation of each ASEAN-5 member in AERU. From the previous section, we have acknowledged Chinese Yuan as the currency anchor for AERU, since it is the most stable currency peg. Now we can proceed to establish assessment criteria which allowus to propose the country member of the OCA. The assesment is developed for a period of 2 years and the margin set is plus and minus 15 percent from the AERU conversion rate (as benchmark) in terms of Chinese Yuan for five currencies, both in normal and crisis periods. The test for a 2 years period refers to participation criteria fulfillment condition in the EU, which is extended on the present study into two different condition; normal and crisis. As noted earlier, normal period is 2004-2006, and crisis period is 2007-2010. Thus, the assessment on exchange rate stability is conducted 2 years before the currency union implementation. For normal period, its implementation will be in 2006, so that the assessment criterion becomes 2004-2005. Similarly for crisis period, 2008-2009 becomes subject to assessment since policy makers agree to implement the currency union in 2010. Table 13 shows the statistical evaluation of simulation of ERM II for ASEAN-5 currencies in terms of Yuan. According to AERM statistical calculation, ASEAN-5 currencies against AERU depreciate during crisis period compared with normal period, except for Malaysian ringgit and Singapore dollar. Meanwhile, based on SERM, during crisis period, the standard deviation for ASEAN-5 currencies has bigger dispersion than in normal period, except for Malaysian ringgit and Singapore dollar which remain constant. Then, the ANF results outline that all ASEAN-5 currencies are having low volatility, either in normal or crisis period, except for Indonesian rupiah in crisis period. However, all in all, ASEAN-5 currencies against AERU have shown that under Yuan currency basket peg, the exchange rate stability will have to be in place once ASEAN-5 countries agree to join a currency union.


77

��

��

��

��

��

�

��

�

�

�

�

�

�

�

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

The flexible judgment in determining ASEAN-5 participation utilizes the exchange rate stability criteria. Empirically, it is based on the value of AERM, ANFB, ANV, SERM, and LV indicate stability path. The results suggest that all ASEAN-5 countries have low level of volatility, except for Indonesian rupiah, but her volatility is yet considered low, around 0.37, during crisis period. Therefore, ASEAN-5 countries under Yuan currency basket peg could be considered into optimum currency area (OCA) based on their exchange rate stability criterion. Based on this, we conclude that all ASEAN-5 currencies are eligible to join a currency union against AERU, both using normal period and crisis periods. The finding is in line with the study conducted by Shirono (2009) which suggested currency unions along with China tend to generate higher average welfare gains for East Asian, including ASEAN countries than currency unions with Japan or the United States. This trend is likely to continue if China’s role continues to rise in the regional trade. However, the current issue arises as the movement of Chinese Yuan is highly pegged to US Dollar and is well-known as a heavily manipulated currency against US dollar. This situation open further research, particularly on how to set a stable currency union in the mid of manipulated exchange rate policies by China’s government.

30 AERM is defined as the degree of which exchange rate is fluctuating over the period of observation. The higher number of AERM shows that the national currencies are under pressure or getting shocks. 31 ANFB denotes for the number of data in daily basis where the fluctuated currencies are no longer exceeding the allowed bands, namely plus and minus 15 percents. Hence, the smaller ANFB indicates the national currency is no longer fluctuating sharply due to moving within tolerable bands. 32 ANV represents the number of data in daily basis where are no longer in the allowed bands. Hence, the higher of the number shows that the national currency is increasingly unstable. 33 SERM covers the distance between the value on the specific date and the mean of data observed. It informs that the higher of SERM would lead to instability position on the observed currency. 34 LV demonstrates the degree in which the volatility is well-categorized according to nominal deviation results of each national currency against AERU. The low level of volatility indicates the country satisfies the exchange rate stability criteria and ultimately is eligible as member of union.

78


In addition, this finding is also relevant with the growing role of China in global economy. In the context of exchange rate policy, this tren provide opportunity for strengthening economic relationship within ASEAN-5. In addition, choosing Yuan as a common anchor currency by ASEAN-5’s is supported by the network effects theory. This theory states that the utility of a consumer on particular good is dependent on the number of other individuals consuming the same good (Katz and Shapiro, 1985). There are two implications for this, first, a minimum level of agents consuming the same good (critical mass) is necessary for the initial adoption of a network good (Farrel and Soloner, 1986); second, the demand for network commodities is associated with a bandwagon effect, i.e. the more individuals use the good, the more incentive for others to also use it. These implications will apply on money as a network good and will led to interesting results in the form of monetary integration. In reality, the network effect does occur in ASEAN-5 countries on their trade relationship with China. The table below demonstrates the growing importance of China within ASEAN-5 trade direction.

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Table 14 shows ASEAN-5 countries are dominantly trading with China, around 44% of the four selected trading partner. Swoboda (1968) argues that if residents of a country can only hold non-interest bearing foreign currency assets, and their revenues or expenditures are at least partly denominated in a foreign currency, and also owing to transaction costs (e.g. brokers’ fees, bookkeeping, psychological inconvenience, etc.), then it is profitable for them to hold foreign currency cash balances. Krugman (1980) develops a formal three-country, three-currency model, where the transaction costs decline as the size of the market increases. He shows that only the currency with dominant economy can serve as a vehicle currency. Moreover, once a currency serve as international medium of exchange, its vehicle role becomes self-reinforcing and may persist even when its economic power diminishes. This theoretical view is in line with high penetration of Cina towards ASEAN-5 market.


79

On the other hand, China’s government keeps trying to postion Yuan as a gold-backed currency35, which help it to become major international currency. Many economist realize if China is trying to position the Yuan as the alternative global reserve currency. Currently, China is the sixth largest holder of gold reserves in the world, and officially has of 1,054.1 tones of gold. This is less than half of Euro debtor nation; France and Italy, who have 2,435.4 and 2,451.8 tons of gold reserve respectively. These arguments support the ASEAN-5 country to peg their currency to Yuan.

V. CONCLUSION The paper measures the degree of volatility among various currency basket arrangements using nominal deviation indicator. By exercising of each ASEAN-5 national currency against AERU which are pegged to various selected currencies, the study reveals that ASEAN-5 currencies are recommended to value the AERU into Yuan to maintain ASEAN-5 currencies’ stability. Furthermore, the results show that using Yuan as the anchor currency will smooth the fluctuation bands of plus and minus 15 percent from each ASEAN-5 currency benchmark in terms of AERU. This finding leads us to the conclusion of this paper that ASEAN-5 can form a currency union, which is most suitable to be pegged to Yuan. This conclusion has several implications for political leadersin this region, first, they should complete the internal market unification in the short run, such as AFTA; second, strengthen the competitiveness on goods and services; third, enhance coordination and surveillance on economic policy; fourth, budgetary adjustments in highly debted or deficit countries; fifth, commit to accelerate the establishment of the ASEAN Monetary Institute (AMI), replicating the European Monetary Institute (EMI), which will be institutionally prepared to be the ASEAN Central Bank (ACB).

35 Turkey and China had signed a trade agreement to only use their currencies in the trade. China has signed up similar agreement with almost all of Asia, Belarus, Argentina, Brazil (The Wall Street Journal).

80


REFERENCES

Achsani, Noer Azam, and Hermanto Siregar, 2010, Classification of the ASEAN+3 Economies Using Fuzzy Clustering Approach.EuroJournals Vol. 39 No. 4, pp. 489-497 Achsani, Noer Azam, and Titis P., 2010, Testing the Feasibility of ASEAN+3 Single Currency Comparing Optimum Currency Area and Clustering Approach. Euro Journals Publishing, Inc Issue 37 Agarwal, Aman, Penm, and Wong, 2004, ASEAN DOLLAR: A common Currency Establishment for Stronger Economic Growth of ASEAN Region. Paper presented at the International Conference on Business, Banking, and Finance: Trinidad and Tobago Azali, Wong, et.al. 2007, The ASEAN-5 Future Currency: Maastricht Criteria, MPRA Paper No. 10272 Bayoumi and Mauro, 2000, On Regional Monetary Arrangements for ASEAN. Journal of Japanese and International Economics 14, 121-148 Eichengreen, Barry, 2006, Global Imbalances: The Blind men and The Elephant. Brookings Policy Brief 1 (January) Falianty, Telisa Aulia, 2006, Feasibility of Forming Currency Union in ASEAN-5 Countries. Research Laboratory University of Indonesia: Indonesia Farrel J., Saloner G., 1986, Installed Base and Compatibility: Innovation, Product Preannouncements, and Predation. American Economic Review;76;940-955 Frankel, Jeffrey, 1999, No Single Currency Regime is right for all countries or at all times. Working paper NBER WP No. 7338. Frankel and Rose, 1998, The Endogeneity of the Optimum Currency Area Criteria. Economic Journal, Royal Economic Society, Vol.108 (449), pages 1009-25, July Frankel and Wei, 1994, APEC and other regional economic arrangement in the Pacific. Pacific Basin Working Paper Series 94-04, Federal Reserve bank of San Francisco. Gross and Thygesen, 1998, European Monetary Integration, from the European Monetary System to Economic and Monetary Union. Harlow Essex/New York, Longman. Ito, Takatoshi, 2006, On Determinants of the Yen Weigh in the Implicit Basket System in East Asia. RIETI Discussion Paper Series 06-E-020


81

Katz M.L, Shapiro C., 1985, Network Externality, Competition, and Comparability. American Economic Review; 75; 424-440 Krugman P., 1980, Vehicle Currencies and the Structure of International Exchange. Journal of money, credit, and banking;12 (3); 513-526 Kuroda and Kawai, 2002, Strengthening Regional Financial Cooperation. Pacific Economic Papers 322, pp. 21-35. Kusuma, Dimas Bagus, and Arief D.P., 2010, Analysis of Implementation Optimum Currency Area and Its Volatility: Case Study ASEAN-5+3. Bulletin of Monetary Economics and Banking Vol. 13, No. 2 McAleer, M.J. and Nam, J.C.W, 2005, Testing for Cointegration in ASEAN Exchange Rates. Mathematics and Computers in Simulation, 68: pp. 519-527 Moon, et al., 2006, Regional Currency Unit in Asia: Property and Perspective. KIEP Working Paper, 06-03. Mundell, 1961, ” A theory of Optimum Currency Areas”, American Economic Review, Vol. 51, p657-665 Ogawa, Eiji, 2006, Adopting a common currency basket arrangement into the “ASEAN plus three”. RIETI Discussion paper series 06-E-028 Ogawa, Eiji, and Junko Shimizu, 2006, “A Deviation Measurement for Coordinated Exchange Rate Policies in East Asia,” RIETI Discussion paper 2006/01 06-E-002 Ogawa and Ito, 2002, On the Desirability of a regional basket currency arrangement.Journal of the Japanese and International Economies 16:317-34. Ogawa and Shimizu, 2005a, A deviation measurement for coordinated policies in East Asia. RIETI Discussion Paper Series, 05-E-017. Ogawa and Shimizu, 2005b, Risk Properties of AMU Denominated Asian Bond. Journal of Asian Economics, Vol. 16, issue 4, 590-611 Ogawa and Yoshimi, 2008, Widening Deviation Among East Asian Currencies. RIETI Discussion paper series 08-E-90. Swoboda A., 1968, The Euro-Dollar Market: An Interpretation. Essays in International Finance; p.64 Warjiyo, Perry, 2004, Materi Kuliah Ekonomi Keuangan Internasional. Post-Graduate Program, Economic Science, University of Indonesia Watanabe and Ogura, 2006, How Far a Part are two ACUs from each other?: Asian Currency Unit and Asian Currency Union. Bank of Japan Working Paper Series

82


Williamson, John, 2005, A currency Basket for East Asia, not Just China. Institute for International Economics, No. PB05-1 Yuen, Hazel, 1999, Globalization and Single Currency the Prospects of Monetary Integration in East Asia. Paper prepared for conference on ”The Challenges of globalization”, Bangkok.

The Impact of US Crisis on Trade and Stock Market in Indonesia

83


Mita Nezky1

Abstract

This paper analyzes the impact of the 2008 financial crisis in United States on the Indonesian economy, by using the Structural Vector Autoregression (SVAR) model of 5 variables; the Dow Jones Industrial Average, the exchange rate, the Jakarta Composite Index (JCI), the production index, and the international trade tax income. The results showed that the US crisis affected the capital market in Indonesia, where the Dow Jones Industrial Average plays a greater role in explaining the JCIcompared to Rupiah exchange rate, production index and the trade income tax. In addition, the US crisis affected the volume and the trade income tax in Indonesia. These empirical results have policy implications for the Capital Market and Financial Institution Supervisory Board (BAPPEPAM-LK) as the stock market regulator to intervene or to suspend the trade in stock market when volatility exceeds the psychological threshold. The results also emphasized the necessity to diversify the country’s export destinations and to increase the quality and the value-added of Indonesian exports.

Keywords : US Crisis, stock market, trade, SVAR. JEL Classification : G18

1 Mita Nezky, ME is graduated from Master of Economic in Public Policy, University of Indonesia; [email protected].

84


I. INTRODUCTION The U.S. economy is the largest economy in the world with a GDP of $ 15.5 trillion2 by the end of 2011 (one quarter of world GDP). Before the financial crisis in 2008, steady economic growth resulted in low unemployment and inflation in the United States. In early 2007, the unemployment rate in the United States was 4.4% with an inflation rate of 2.1%. During the crisis of 2008, the unemployment rate in the United States increased to 6.8% with an inflation rate of 5.6%3. In mid-2007, the U.S. experienced a subprime mortgage crisis that peaked in September 2008, and was subsequently marked by the announcement of the bankruptcy of several financial institutions. The beginning of the problems occurred in 2000-2001, when the dotcom4 stocks in the United States collapsed, and the companies that issued shares could not repay the loans to the bank. To overcome this, the Federal Reserve (U.S. Central Bank) cut interest rates. Low interest rates were used by developers and housing finance companies. The houses built by developers and financed by the housing finance companies were cheap homes, sold to low-income home-buyers who had inadequate financial guarantees. With the collapse of the stock value of these companies, the banks were facing defaults of the debtor (developers and housing financing companies). According to Crockett (1997), financial stability is closely related to the health of an economy. The more healthy a country’s financial sector, the more healthy the economy, and vice versa. Thus, the development of the financial sector, including capital markets, is a key indicator to maintain a healthy or stable economy. Price movements of stocks, bonds, and other securities in the stock market of a country are a reflection of investors’ perception of the condition of the capital markets. This perception will ultimately affect investment funds coming into the country, and likewise affect the state of the economy concerned. This is not exclusive to the United States, as it is also prevalent in Europe and Asia, including Indonesia. The Rupiah exchange rate against the U.S. dollar began to decline from mid-2008 and continued to depreciate until it reached its lowest level at the beginning of 2009 at the amount of Rp. 11,900 per 1 USD. Changes in exchange rates, either appreciation or depreciation will affect import-export activities in the country, because the USD is still the dominant currency in global trade payments. The increase or decrease in exports and imports will affect state revenues derived from international trade taxes. Depreciation of the Rupiah mid-2008 led to an increase in exports that affected the acceptance of duties and taxes, in particular international trade in general. Changes in the value of exports and imports also affect the Gross Domestic Product (GDP) of 2 International Financial Statistics (IMF, 2011). 3 Ibid. 4 Dotcom companies are the companiesthat run most of its business on the internet, for example, www.amazon.com ; www.amcy. com ; www.e-loft.com ; www.flooz.com; etc.


85

Indonesia. The production index is an economic indicator that is often used to replace the GDP as its data publication is done every month. Given the background above, this paper analyzes how the U.S. financial crisis 2007-2011 affected capital markets in Indonesia and the Indonesian government revenue derived from international trade taxes. Explicitly, the first purpose of this paper is to examine whether there was any effect of the U.S. financial crisis on Indonesia’s capital market and revenues derived from international trade taxes; and second, to provide policy recommendations in order to maintain the stability of the Indonesian capital markets and fixed increase government revenue from international trade taxes (ITT). The second part of this paper will examine theoretical basis, and the third part will review the data and methodology used for the analysis. Results and analysis are presented in the fourth section, while the conclusions and recommendations are presented in the last section to complete this paper.

II. THEORY The financial crisis can be attributed to a few key things; first, the failure of the financial market, second, a situation in which an institution or a financial institution lost most of its assets, third, banking panic, credit defaults and the recession, and fourth, the collapse of the stock exchange and falling currency values (see among others, Mishkin (1992), Allen (2001), Eichengreen and Portes (1987), and Jickling (2008)). Many researchers have investigated financial crises and these crises are generally divided into three types according to the background and the characteristics of the crisis. The firstgenerationof crisis-related fiscal and monetary problems commenced with a crisis in Mexico 1973-1982 (Kaminsky, 2003). Flood and Garber (1984) and Krugman (2007) further stated that in addition to fiscal and monetary issues, first generation crisis were caused by macroeconomic instability. Alongside this, a currency crisis can also be caused by a government’s budget deficit and a system of fixed exchange rates. Second generation crisis was first presented by Obstfeld (1994) and Cole and Keho (1996). One example of this event is the crisis that hit the European financial system in 1992 and 1993. According to Obstfeld, second-generation crisis is a crisis caused by the application of the conflict of fixed exchange rates against the government’s desire for monetary expansion. Third generation crisis is a combination of the first and second generation crises, as such it is also known as the twin crises. According to Krugman (2001), Cartapanis and Gilles (2002), twin crises are caused by a worsening of the banking system and a drop in the exchange rate. One example of the twin crises is crises that hit Asia in 1997. According to Kaminsky and Reinhart (1999), twin crises are caused by weak economic fundamentals of the country. In 2003, Kaminsky added that the cause of the third generation crises is also a crisis of moral

86


hazard and asymmetric information. This type of crisis is characterized by anincrease total credit and a sudden rise in asset prices. Various empirical studies have been conducted of the financial crises, some of which examined the relationship between the financial crisis with the stock indices, exchange rates, trading volumes, and revenues from international trade. However, in the author’s opinion, little research has been done on the relationship of a financial crisis with international trade taxes. Therefore, the author’s interest is to explore the effect of the financial crisis on international trade taxes. This paper refers to research conducted by Fang, Lai, and Thompson (2007) and Zhang brothers and Han (2010), but research for this thesis puts more emphasis on international trade taxes. Fang, Lai, and Thompson (2007) to did more research on the exchange rate and exports. According to Fang, Lai, and Thompson (2007), there is an (positive) influence between the exchange rate depreciation on export revenues in eight countries in Asia. However, there is not always a positive relationship between exchange rate depreciation with export revenues because the declining value of the currency is not always accompanied by increased export demand. In addition, exchange rate risks encourage exporters to hedge. Research on the relationship between financial crises with international trade was also carried out by the Zhang brothers and Han (2010). They assumed that the U.S. financial crisis affected the state of the economy in the Asia Pacific region through the trade channel. More specifically, their study showed that the financial crisis of the United States economy affected the Asia Pacific region through three channels, namely the banking sector, flight-to-quality, and the stock market.

III. METHODOLOGY 3.1. Data and Variables Monthly time series data from January 2007 to December 2011 were used in preparing this paper. The period was determined by the movement of the financial crisis in the United States. Monthly data was used to capture the movement and to provide a more accurate analysis of changes in variables such as the Dow Jones Industrial Average (DJI), Jakarta Composite Index (JCI), and exchange rates (Euro against the USD) that occurred within a relatively short time. The data used was quite diverse in measurement unit. Data from the DJI, JCI, and EXCRATE have units inthe thousandths, the hundredths for PI, while the data units for ITT are in the trillions. To make the data more uniform to facilitate interpretation, it was converted usinga natural logarithm. According Nachrowi and Usman (2006), using a logarithmic transformation of the data is intended to minimize the scale between the independent variables. If the range of values observed become increasinglynarrow, the expected variation of the error will not differ across groups of observations.


87

3.2. Estimation Technique This paper uses a Structural Vector Autoregression (SVAR) model for making data estimates. The estimation technique is preceded by some standard steps including stationarity test data or stationary stochastic process (Ajija et.al, 2011), and the determination of the optimal number of lag with lag order selection criteria. In this study, the amount of lag will be determined based on the criteria with longest lag. The SVAR model is a development of the VAR model that was first introduced by Christopher Sims (1980). According to Sims, if there is a simultaneous relationship between observed variables then these variables should be treated equally; there would be no endogenous and exogenous variables. Development within the SVAR model is done by setting restrictions on cross-variable correlation in the system of equations. Limitations or restrictionsare intended to separate the movement of the endogenous variables into several components with reference to the underlying shock. According to Enders (2003), Structural VAR is used to prove an economic theory or to seek theoretical basis of a shock (Bilmeier and Bonatot, 2002). The Structural VAR model used in this paper consists of five variables that make up the five equations, where the Dow Jones Industrial Average (DJI) is used as a proxy of a crisis in America, the Jakarta Composite Index (JCI) as a proxy of the state of capital markets in Indonesia, the value of the Rupiah against the USD, the production index as a proxy of the state of the Indonesian economy (GDP), and international trade taxes as a proxy of state revenue. The specific SVAR model in reduced form is, (1) where Xt is a vector for five variables used (DJI, CJI, EXCRATE, PI, and ITT); A0 iscontemporaneous relations between variables; A(L) is the finite-order matrix polynomial with the lag operator L; εt is the vector structural disturbance; and B is a non-zero diagonal matrix. Basically, there are several ways to set restrictions on the SVAR model, among others are the long run restriction, impact, and the sign restriction. These restrictions help in the identification of the model and also functions in applying theory to the model. The Dow Jones Industrial Average variable is considered an independent variable so that the variable that can influenceit is the shock of Dow Jones Industrial Average itself. Shock of Jakarta Composite Index (JCI), exchange rate, production index, and international trade taxes cannot be considered to affect the shock to the Dow Jones Industrial Average. Thus, the first equation in the SVAR system is as follows: (2)

88


Shock to the Jakarta Composite Index (JCI) is affected by the shock of the Dow Jones Industrial Average since the event of shock on the Dow Jones Industrial Average; Jakarta Composite Index (JCI) is the first variables that respond to shock. It follows that shock to the Composite Stock Index is influenced by the Dow Jones Industrial Average shock and shock on Jakarta Composite Index (JCI) itself. (3) If there is a shock in the Dow Jones Industrial Average which is followed by a shock in the Jakarta Composite Index (JCI), and then a shock in the exchange rate, these events would be associated with production activities for exports and imports, where the production index is assumed to respond to shocks in the first place. There fore, shock on the production index is affected by the shock on the Dow Jones Industrial Average, shock on Jakarta Composite Index (JCI), the exchange rate shock, and a shock to the production index itself. (4) The exchange rate is assumed to be influenced by shock to the Dow Jones Industrial Average (DJI), shock to the Jakarta Composite Index (JCI), and shock to the exchange rate itself. On the other hand, the shock on the international trade tax is assumed to be influenced by shock to the Dow Jones Industrial Average, shock on the Jakarta Composite Index (JCI), the exchange rate shock, the shock on the production index, and a shock to the international trade tax itself. (5) (6) Basically, the restriction used refers to Sims (1980) where C0εt = et, then C0 is restricted as a triangular matrix, providing a system that is just identified. Matrix C0 is a measure of impact of structural shock from the endogenous variables thus it is classified as impact restriction, with a number of restrictions as n x (n-1)/2 or in this case as much as 10 (ten) restrictions. In addition, the author also provides an additional restriction to the diagonal elements b11 =b22 =...b55 =1, so the equation system becomes over-identified. This additional restriction is a normalization model for ease of interpretation where the reduced form disturbance (e) will correspond with the structural shock (ε). Normalization is merely scaling that does not change the essence and interpretation of the estimated results. Thus, with a total of 15 restrictions, the SVAR system specifications used are as follows:


89

(7)

To ensure the validity of the final model used, testing was done for the stability of the model on the condition that the entire root lies inside the unit circle (absolute value less than one unit root). If the condition is stable, the next step is the Impulse Response Function (IRF) and Variance Decomposition (VD).

IV. ANALYSIS AND RESULTS From the stationarity test, the results obtained indicate that all the variables were stationary in first differences. Based on the testing of lag length, the recommended amount of lag is lag 1 and lag 2. The selected optimal lag is Lag 1 (one) on the basis that the longer the lag, the more observations that are ‘missing’ (Nachrowi and Usman, 2006). By using Inverse Roots of AR Characteristics polynomial, the estimated empirical model is proved to be stable,hence can be used to analyze the Structural Impulse Response Function and Variance Decomposition.

4.1. Impulse Response Function Figure1 shows the impulse response function of the variables studied. When there is a decline in the Dow Jones Industrial (DJI), The Jakarta Composite Index (JCI) will drop until month 12 and then increase to neutral after 40 months. The impulse response function above indicates that The Jakarta Composite Index (JCI) will immediately respond to a reduction in the Dow Jones. The DJI decline triggered a liquidity crisis among American and European investors that also affected the Jakarta Composite Index (JCI). While in Indonesia, domestic investors experienced panic and uncertainty over the state of the economy that caused The Jakarta Composite Index (JCI) to plummet. The positive influence of the shock will occur within a period of 12 months. After that period, Jakarta Composite Index (JCI) would respond opposite (negative) of the DJI. This is caused by the re-entry of funds characterized by the growth of investor confidence in the country so as to attract foreign investors to invest their funds in the Jakarta Composite Index (JCI). According to the IRF, the effect of the DJI shock becomes neutral (no effect) within 40 months. In that time frame it is expected that investors both within and outside the country would have received confirmation that the economy is no longer influenced the DJI shock.

90


��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 1. Impulse Response Function

In terms of exchange rate movements of the Rupiah against the U.S. dollar, the IRF results show that when the Dow Jones (DJI) average index increase, then the Rupiah would weaken against the USD. A DJI decline will reduce liquidity in USD in the domestic market of various countries so that the USD will be weakened. Weakening of the USD does not directly strengthen the Rupiah. According to the Chief Economist at PT. Bank Mandiri, Destry Damayanti, speculation of homeland expectations of economic conditions make investors hold on to their USD which


91

weakens the Rupiah. In addition, the current account deficit caused by the increasing number of imports (due to a weakening USD makes the price of foreign goods cheaper than domestic goods), which worsens the strength of the Rupiah. Based on the IRF graph, the condition will last about 40 months and then became neutral (no effect). The Jakarta Composite Index (JCI) has an impact on dollar value that can be seen from the Figure 1. The Jakarta Composite Index (JCI) decline will put pressure on the Rupiah resulting in depreciation. The Jakarta Composite Index (JCI) will also cause a decline in dollar liquidity in the domestic market so that the minimum value of the rupiah would weaken. Based on the Figure 1, the condition will last about 40 months and then become neutral (with no effect). This same period of turbulence and stabilization is seen for the stock index in response to a DJI shock which explains why the exchange rate (EXCRATE) will stabilize within a period of 40 months from the effects of shock to the JCI. If the dollar is depreciating, the value of the goods in the country (Indonesia) is relatively less expensive compared to the price of foreign goods which would lead to increased exports. However, the declining purchasing power in the United States caused by the financial crisis, has affected trade activities in US and other countries that made the United States an export destination. This will directly and indirectly affect the level of Indonesia’s exports. According to the Figure 5, the EXCRATE shock effect would be neutral in a period of 36 months. In aggregate, the growth of Indonesia’s exports decreased from 9.5% in 2008 to 5.9 in 2009%. A slowing export growth due to reduction on global demand, which is triggered by the world recession, caused a slowdown in manufacturing and agricultural exports, falling commodity prices (mined goods wereone-fifth the price in 2007 compared to the price in 2006), and rising global unemployment rate affected the production level of export commodity especially processing and craft commodity. The Dow Jones was also affected by the international trade taxes (ITT). A decrease in the DJI will reduce the liquidity of the USD in the United States domestic market and will affect Indonesia’s purchasing power and exports to this country. Based on Statistic Indonesia data obtained from the website, the value of Indonesia’s exports to the United States decreased by USD 2 trillion during 2008 to 2009. This decline in export value depressed the trade tax income (ITT). In the 19th month ITT started to rise, and exporters began eyeing export destinations other than the United States and beyond. In addition, the increase in ITT was also caused by increased imports. The effect of the shock became neutral at month 40 as the import/export market conditions is expected to be stable by that time.

92


��

��

�

��

�

��

�

��

��

��

��

��

��

��

��

�

��

��

��

�

��

��

�

��

�

��

��

��

��

��

��

��

��

��

��

��

��

��

�

��

� ��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Based on data obtained in 2012 from CBS, the United States is Indonesia’s third largest export destination after Japan and China, then followed by other Asian countries such as Singapore, the Republic of Korea, and India. A decrease in purchasing power of the United States led to decreased demand for imports which in turn affected the value of Indonesia’s exports to the United States. The composition Indonesia exports from 2007 to 2011 consisted of 20.38% in oil and gas exports, and 79.62% in non-oil exports. Although the value of Indonesia’s exports ��

��

�

��

�

��

�

��

�

��

�

��

�

��

�

��

�

��

��

��

��

��

��

��

��

��


93

to the United States was only 9% (on average in the period 2007-2011), the United States is the main destination of Indonesia’s non-oil industries exports (Table 2), hence changes in demand for U.S. exports would greatly affect the value of exports from Indonesia as a whole.

4.2. Structural Variance Decomposition Variance decomposition of the forecast error equation system gives information of how large the role of each variable in explaining variation in a particular variable in the SVAR system. Changes in the DJI (Dow Jones Industrial Average) play a greater role in explaining changes of Jakarta Composite Index (JCI), compared to EXCRATE (the exchange rate), PI (Production Index) and ITT (International Trade Tax). The bankruptcy of major financial firms like Lehman Brothers in 2008 led to panic as investors pulled their money out of the stock and pushed flight-to-quality. This causeda drop in the Jakarta Composite Index to its lowest level. A number of companies in the United States and Europe were experiencing a liquidity crisis and panic that encouraged domestic investors’ impairment to theJakarta Composite Index (JCI). On the other hand, the estimated results indicate that changes in the Jakarta Composite Index (JCI) has larger role in explaining changes in EXCRATE (exchange rate) compared to changes in the DJI (Dow Jones Industrial Average), PI (Production Index) and ITT (International Trade tax). The Jakarta Composite Index (JCI) is an index of major stocks that describes the movement of stock prices in the Indonesia stock exchange with a market capitalization of Rp. 3.4 Trillion. If there is a decrease The Jakarta Composite Index (JCI), investors will pull their money out of the stock, causing liquidity in the Indonesian domestic market to increase. This will cause the value of dollar to decline or depreciate. JCI changes also play a role in explaining changes in PI (Production Index) compared changes in DJI (Dow Jones Industrial Average), EXCRATE (exchange rate), and the ITT (International Trade Tax). Jakarta Composite Index (JCI) is a stock index comprised of large companies that have gone public such as the PT. Astra Agro Lestari (plantations), PT. Bumi Resources Tbk (mining), Elnusa Tbk (mining), Dynaplast Tbk (industrial), Betonjaya Manunggal Tbk (industry), Barito Pacific Tbk (industry), Gajah Tunggal Tbk (industry). Changes in The Jakarta Composite Index (JCI) would cause an increase or decrease in the value of the industrial companies incorporated in The Jakarta Composite Index (JCI) that will also affect the value of PI (Production Index). If the value of Jakarta Composite Index (JCI) declined the book value of these companies will decline and will affect their assets and capital. It will affect the activities of production, production output and production indices in the end. Although associated with international trade taxes (ITT), changes in the DJI (Dow Jones Industrial Average) plays a greater role in explaining changes in ITT compared JCI, EXCRATE (exchange rate), and PI (Production Index). The financial crisis that hit the United States to led to a global recessiondue to the significant role of the United States in the world economy.

94


The U.S. economy is one of the world’s largest economies. This is reflected in its GDP that reached USD 15.5 trillion at the end of 2011, or a quarter of total world GDP. The U.S. trade balance is always in deficit. This is due to the high number of residents of the United States of 306 million people so the demand for clothing, food, goods and shelter is high. On the other hand, three-quarters of the working population is in the tertiary sector (service industry) and not in the secondary sector (manufacturing industry) or the primary sector (mining, fisheries, and agriculture), so it cannot meet the high primary needs in the country. Although Indonesia is not its main country of origin of imports, Indonesia is one of the main countries of origin for U.S. imports of natural rubber commodity, crude oil, bauxite and aluminum, industrial organics chemical, tin and aluminum processed products, processed rubber products, processed products palm oil, fishery products, and textile. Conversely, U.S. export commodities to Indonesia are wheat, passenger aircraft, machinery, and cars. With such a relationship, the crisis in the U.S. would have a major impact on trade volumes to Indonesia. On the financial side, the crisis in the U.S. raised the issue of liquidity of U.S. financial firms as the Lehman Brothers (the fourth-largest financial company in the U.S.) went bankrupt. This caused the decline of the Dow Jones. The Dow Jones Industrial Average (DJI) is an indicator of the state of the financial sector (capital markets), so changes in the value of the DJI would affect investors’ the decisions and confidence in investing. The decline of the Dow Jones, liquidity problems and high levels of debt caused the U.S. recession affecting the purchasing power of the U.S. In terms of international trade, decline in purchasing power would also affect U.S. import demand which would in turn affect world export demand. Countries that were directly affected are the countries that make the U.S. its largest export market, namely China, Japan, Germany, and Indonesia. This would reduce the level of Indonesian exports and directly impact on international tradetax income.

V. CONCLUSIONS Referring to the empirical analysis this paper provides two (2) conclusions; first, the crisis in the United States has a significant effect on the capital market in Indonesia. Movement of the Jakarta Composite Index (JCI) delivers a proven direct response to Dow Jones Industrial Average (DJI). Changes in the Dow Jones Industrial Average (DJI) have a larger role in explaining movements of the Jakarta Composite Index (JCI) than the exchange rate, the Production Index (PI), and the International Trade Tax (ITT). The first conclusion is consistent with the fact that Indonesia’s capital market is still strongly influenced by foreign capital markets, so that the event of a large shock on foreign stock indices will easily cause panic among domestic investors. In connection with trade, the shock on the Dow Jones Industrial Average (DJI) will be positive response on the Trade Tax International (ITT), which means that if there is an increase in the Dow Jones Industrial Average (DJI), the International Trade Tax (ITT) will go up, and vice versa. Changes in the Dow Jones Industrial Average (DJI) play a greater role in explaining changes


95

in International Trade Tax (ITT) than the other three variables, namely the Jakarta Composite Index (JCI), exchange rate (EXCRATE), and Production Index (PI). This means that the Jakarta Composite Index (JCI) effecton the International Trade Tax (ITT) is more significant than other variables. Shock on the Dow Jones Industrial Average would affect liquidity of the USD in the United States domestic market that will affect the purchasing power and Indonesian exports to the United States. Changes in the purchasing power will also affect U.S. import demand which impact on the world export demand. Countries that were directly affected are the countries that make the U.S. their largest export market, namely China, Japan, Germany, and Indonesia. This will reduce the level of Indonesian exports and directly impact on reception duties (International Trade Tax). These results lead to the conclusion that the crisis in the United States both affects the volume of trade and the taxation of international income. It should be emphasized the needs to internalize other important variables into the model. In addition, the determination of restrictions on the estimated SVAR models can be studied further to provide more accurate results. Both of these opportunities would further research. The conclusion of this paper has policy implications in the event of a crisis for example a marked significant decline in the Dow Jones index. The Capital Market and Financial Institution Supervisory Board (BAPPEPAM-LK), as a regulator of stock trading on the stock exchange, may intervene when there is a drop in stocks that go beyond the psychological threshold and/or halt trading on the Stock Exchange for a period of time. This is in accordance with Article 5 of Act 8 of 1995 regarding Capital Market. BAPPEPAM-LK can urge member companies to limit the exchange of securities using a margin threshold to protect the interests of investors from the use of debt. Besides that BAPPEPAM-LK can support a socialization program of investment in the capital market for the community, there by increasing the domestic (local) investor base. Related to the effect of an external crisis concerning the decline in International Trade Tax, the Ministry of Commerce should encourage diversification of export destinations. This is in line with efforts to reduce dependence on exports of Indonesia to the United States. For the Ministry of Industry, the conclusions in this paper imply the need to improve the quality and value-addedof export commodities. Attempts by the government to improve the quality of these goods could include provisions in education, training, and mentoring for entrepreneurs (exporters), both large and small. Providing additional capital for small entrepreneurs or SME’s is necessary for small businesses with bright prospects to grow in the international market. The government can also work with national banks or local banksin providing credit with low interest charged. In addition, the Ministry of Industry could encourage the promotion of value-added products or value-added exports.

96


REFERENCES

Ajija, Shochrul, 2011, et.al. ´Cara Cerdas Menguasai Eviews´. Salemba Empat Allen, Franklin, 2001, ‘Financial Structure and Financial Crisis’. Wharton School, University of Pennsylvania, International Review of Finance. Bilmeier and Bonatot, 2002, ‘Exchange Rate Pass-Through and Monetary Policy in Croatia’ Bappenas, 2009, ‘Kebijakan Nasional Dalam Mencegah dan Mengantisipasi Dampak Krisis Keuangan Global’, Buku Pegangan, 2009. Crockett, Andrew, 1997, ‘Why is Financial Stability a Goal of Public Policy’, In Maintaining Financial Stability in a Goal Economy, A Symposium Sponsored by The Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming. Eichengreen, Barry and Richard Portes, 1987, ‘The Antomy of Financial Crises’, Working Paper No. 2126. National Bureau of Economics Research, Cambridge. Enders, Walter, 2003, ‘Applied Econometric Time Series’, Iowa State University. Fang, Wenshwo, Yihao Lai, and Henry Thompson, 2007, ‘Exchange Rate, Exchange Risk, and Asian Export Revenue’, International Review of Economics and Finance16 : 237-254. Fang, Wenshwo, Yihao Lai, and Stephen M Miller, 2009, ‘Does Exchange Rate Risk Affect Exports Asymmetrically?’, Journal of International Money and Finance 28: 215-239. Nachrowi, D Nachrowi., and Hardius Usman, 2006, ‘Pendekatan Populer dan Praktis Ekonometrika Untuk Analisis Ekonomi dan Keuangan’, Depok, LP FE UI. Sims, C. A., 1980, Macroeconomics and reality, Econometrica, 48(1):1–48. Zhang, Wenlang, Zhiwei Zhang, and Gaofeng Han, 2010, ‘How Does the US Credit Crisis Affect The Asia Pasific Economies? Analysis Based on A General Equilibrium Model’, Journal of Asian Economics 21 : 280-292.

97

WRITING GUIDANCE

1. The paper should be original and should not violate any copyrights. The submitted paper should have never been published or not being submitted to other publisher. The copyright of the published paper is retained to the author. 2. The Bulletin of Monetary Economics and Banking provide a financial incentive between IDR 5,000,000. 3. The paper should be submitted in two formats (i) Microsoft Word (*.doc) and (ii) portable digital file (*.PDF). These files should be sent to the following mail address: [email protected] and Cc to: [email protected]

You may save your files in CD and send it to the following Editorial Address: Bulletin of Monetary Economics and Banking Department of Economic Research and Monetary Policy-Bank Indonesia Building Syafruddin Prawiranegara, 20th Floor JI. M. H. Thamrin No. 2 Central Jakarta, Indonesia Ph. +62-21-2310108 / 2310408 ext. 4119, Fax: 62-21-3800394

4. To avoid missing fonts or other compatibility issues, any special characters or mathematical expression (equations, symbols, matrix, etc.) must be written using Microsoft Equation. 5. The submitted paper should contain (i) an abstract of maximum one page A4, (ii) keywords and (iii) JEL classification code. See the JEL code at http://www.aeaweb.org/journal/jel_class_ system.html. 6. The paper must contain the followings:

The background, the aim of the paper and its distinction to previous study

Theory and review of literatures

Methodology (quantitative methodology is preferred)

Result and analysis

Policy and further study implication

7. The citation should be in footnote and not in endnote. 8. The references must obey the following rule:

98

Bulletin of Monetary Economics and Banking, October 2010

a. Book:

Hanke John E. and Arthur G. Reitsch, (1940), Business Forecasting, Prentice-Hall, New Jersey.

b. Article in journal:

Rangazas, Peter. (2000) “Schooling and Economic Growth: A King-Rebelo Experiment with Human Capital”, Journal of Monetary Economics, October, 46(2), page. 397-416.

c. Article in book edited by other people:

Frankel, Jeffrey A. and Andrew K., Rose. (1995) “Empirical Research on Nominal Exchange Rates”, in Gene Grossman and Kenneth Rogoff, eds.,”Handbook of International Economics. Amsterdam: North-Holland, page. 397-416.

d. Working papers:

Kremer, Michael and Daniel, Chen. (2000) “Income Distribution Dynamics with Endogenous Fertility”. National Bureau of Economic Research (Cambridge, MA) Working Paper No.7530.

e. Mimeo or unpublished work:

Knowles, John. “Can Parental Decision Explain U.S. Income Inequality?”, Mimeo, University of Pennsylvania, 1999.

f. Article from web or other electronic form:

Summers, Robert and Alan W., Heston. (1997) “Penn World Table, Version 5.6” http://pwt.econ.unpenn.edu/

g. Article in newspaper, magazine or equal periodicals:

Begley, Sharon. (1993) “Killed by Kindness”, Newsweek, April 12, page. 50-56.

9. The paper should be submitted along with curriculum vitae complete with mail address and phone number.