BULLETIN OF MONETARY ECONOMICS AND BANKING

ANALISIS TRIWULANAN: Perkembangan Moneter, Perbankan dan Sistem Pembayaran, Triwulan II - 2007 BULLETIN OF MONETARY ECONOMICS AND BANKING Center for...
Author: Doris Willis
2 downloads 2 Views 4MB Size
ANALISIS TRIWULANAN: Perkembangan Moneter, Perbankan dan Sistem Pembayaran, Triwulan II - 2007

BULLETIN OF MONETARY ECONOMICS AND BANKING Center for Central Banking Research and Education Bank Indonesia Patron Dewan Gubernur Bank Indonesia Board of Editor Prof. Dr. Anwar Nasution Prof. Dr. Miranda S. Goeltom Prof. Dr. Insukindro Prof. Dr. Iwan Jaya Azis Prof. Iftekhar Hasan Prof. Dr. Masaaki Komatsu Dr. M. Syamsuddin Dr. Perry Warjiyo Dr. Iskandar Simorangkir Dr. Solikin M. Juhro Dr. Haris Munandar Dr. Andi M. Alfian Parewangi Dr. M. Edhie Purnawan Dr. Burhanuddin Abdullah Editorial Chairman Dr. Perry Warjiyo Dr. Iskandar Simorangkir Managing Editor Dr. Andi M. Alfian Parewangi Secretariat Rita Krisdiana, Skom., ME Wahyu Yuwana Hidayat, SE., MA Tri Subandoro, SE Aliyah Farwah, SP., MSEI The Bulletin of Monetary Economics and Banking (BEMP) is a quarterly accredited journal published by Center for Central Banking Research and Education,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: [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: Statistic Disemination and Management Intern Division - Department of Statistic, Bank Indonesia, Sjafruddin Prawiranegara Building, 2nd Floor – Jl. M.H. Thamrin No. 2 Central Jakarta, Indonesia, phone (021) 2981-6571/2981-5856, fax. (021) 3501912, email: [email protected].

1

BULLETIN of monetary Economics and banking

Volume 16, Number 4, April 2014

QUARTERLY ANALYSIS Monetary, Banking, and Payment System Developments Quarter I - 2014 The Author Team of Quarterly Report, Bank Indonesia The Impact of Crude Palm Oil Price on Rupiah’s Rate Hilda Aprina

291

295

The Regional Impact Transmission via International Trade: An ASIAN-IO Approach Ibrahim, Tri Winarno, Melva Viva, Yanfitri

315

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia Ndari Surjaningsih, Novi Maryaningsih, Myrnawati Savitri

347

Determinant of Capital Ratio: A Panel Data Analysis on State-Owned Banks in Indonesia Pamuji Gesang Raharjo, Dedi Budiman Hakim, Adler Haymans Manurung, Tubagus Nur Ahmad Maulana

369

QUARTERLY ANALYSIS: Monetary, Banking, and Payment System Developments Quarter I - 2014

291

QUARTERLY ANALYSIS Monetary, Banking, and Payment System Developments Quarter I - 2014 The Author Team of Quarterly Report, Bank Indonesia Indonesia’s economy performed an under-controlled economic stability and was sustained by the economic adjustment in quarter I 2014. During the period, inflation was in the declining trend along with smaller current account deficit. The capital inflow also increased along with the improvement of economic fundamental which in turn contributed to the appreciation of Rupiah’s exchange rate. Accordingly, domestic demand was well-managed, even though the growth performed a sharp decrease and was lower than expected as the impact of real export contraction from mining sector. The development was not apart from the policy consistency taken by Bank Indonesia and the government since the mid 2013 to strengthen the economic stability and managed the growth to run proportionally and sustainably. In the quarter I 2014, the domestic demand was well-managed, even though it experienced a slowing down growth due to real export contraction. The under-controlled domestic demand was sustained by the highly growing household consumptions, driven by consumer confidence, and the impact of legislative general election. The investment started to be sustained by nonbuilding investment that had been positively growing, while building investment had been slowing down. Nevertheless, the manageable domestic demand could not resist the slowing down economic growth, which declined lower than expected in the quarter I 2014. The slowing down economic growth was driven by the real export contraction mainly from mining commodities such as coal and mineral concentrate in the form of decreasing demand primarily from Tiongkok, decreasing prices, and also temporarily impact of banning raw materials export policy. From the regional side, the slowing down economic growth occurred in any regions sustained by mining sector which is in the Eastern Indonesia (Kawasan Indonesia Timur/KTI). The export contraction was relatively huge leading to the decrease of net export, even to some extent, import decreased along with moderating trend of domestic demand. The manageable domestic demand sustained by the under-controlled economic stability drove the improvement of Indonesia’s Balance of Payment (IBP). In the quarter I 2014, IBP recorded surplus by USD 2.07 billion sustained by the decrease of current account deficit and the increase of capital inflow. In the quarter I 2014, the current account deficit was recorded by 2.06% to GDP, decreasing from the previous deficit in the quarter IV 2013 by 2.12% to GDP. Improvement of the current account deficit was primarily driven by the decrease of balance of service deficit, especially freight services, along with the decrease of import activities due to

292

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

moderating trend of domestic demand. Meanwhile, the positive sentiment on the improving Indonesia’s economic fundamental impacted the increase of capital inflow in the form of both direct investment and portfolio investment. The increase of capital inflow, which in turn drove the capital and financial transactions, recorded surplus by USD 7.83 billion. The positive development on IBP’s performance eventually contributed to the increasing foreign exchange reserve. In March 2014, Indonesia’s foreign exchange reserve was recorded USD 102.6 billion, or was equal to 5.7 months of import and the government foreign debt settlement, above the international adequacy standard by 3 months of import. The Rupiah’s exchange rate performed the increasing trend in the quarter I 2014, along with the improving economic fundamental and IBP’s performance. At the end of the quarter I 2014, Rupiah appreciated by 7.13% compared to the final level of 2013. The appreciation had primarily occurred since February 2014 along with the increase of capital inflow. The Rupiah’s appreciation was also accompanied by the decreasing volatility. Meanwhile, the micro structure of foreign exchange market performed a positive improvement. The increasing transaction volume of foreign exchange was sustained by the actively foreign exchange transaction and the smaller margin of Rupiah’s bid-ask so that it performed the more liquid condition of foreign exchange market. In the quarter I 2014, inflation development performed improvement to support the inflation achievement target by 4.5+1% in 2014. Inflation was recorded 7.32% (yoy) in the quarter I 2014, much lower than the previous quarter that was 8.38% (yoy). The decrease of inflation was driven by the volatile food group and the core group. The decrease of inflation of volatile food group was driven by harvest on some commodities, even though it was momentarily disrupted by weather and natural disaster in the beginning of the year leading to the unstable production on some commodities. Meanwhile, the decrease of core inflation was sustained by the moderating economy, the minimizing external pressure, and the improving inflation expectation. Nevertheless, inflation on administered price group slightly increased by the increase of LPG 12 kg price and cigarette custom tariff, as well as the implementation of surcharge policy of aircraft tariff. The manageable Indonesia’s economic adjustment was sustained by the stable financial system. The development was sustained by the banking system resistance and the improving performance of financial market. The resistance of banking industry remained strong with the manageable credit risk, liquidity, and market, as well as the strong capital support.The growth of loan to the private sectors decreased from 21.4% (yoy) on the quarter IV 2013 to 19.1% (yoy) in the quarter I 2014, in accordance with moderating domestic demand. Meanwhile, the capital market performance also improved as reflected by the increasing trend of IHSG and decreasing yield of SBN. The improving performance of capital market was sustained by the increasing investor’s optimism on domestic economy.

QUARTERLY ANALYSIS: Monetary, Banking, and Payment System Developments Quarter I - 2014

293

The development of payment system was slowing down in the quarter I 2014 along with the moderating domestic economy. The average of distributed fiat money grew by 13.2% (yoy) in the quarter I 2014, decreased from the previous quarter by 13.4% (yoy). It was driven by the decrease of money demand along with moderating economic growth. Besides, transaction volume of non-cash payment system was slowing down, even the value tended to be stable. In the upcoming periods, economic stability is expected to be manageable and is sustained by the controlled economic adjustment. The economic growth of 2014 is expected to reach 5.15.5%, much lower than the previous expectation by 5.5-5.9%. The expectation was influenced by the unexpected export performance due to raw mineral export restriction policy as well as weaker economic growth of Tiongkok and global commodity prices than previously expected. In the beginning 2015, the economic growth is expected to improve by 5.4-5.8%, even it is much lower than the previous expectation by 5.8-6.2%. Inflation is expected to be lower than inflation in 2013 and range about 4.5+1% in 2014. In 2015, measurable monetary policy supported by the government policy is expected to be able to decrease inflation by 4.0+1%. Some of the risks potentially increase the pressure on economic stability and disrupt the effort on reducing current account deficit to the healthier level require a serious concern. From the global perspective, risks relate to the potential decrease of commodity prices and the slowing down growth of Tiongkok economy potentially increases the current account deficit. The risk of uncertainty on The Fed policy normalization also got a serious concern as it potentially disrupts the prospect of foreign investment. Domestically, the most concern-required aspect is the potential of price pressure regarding the adjustment pressure of the administered price and the increase of food price as delaying impact of flood and the impact of El Nino leading to dry season in some regions.

294

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

This page intentionally left blank

The Impact Of Crude Palm Oil Price On Rupiah’s Rate

295

THE IMPACT OF CRUDE PALM OIL PRICE ON RUPIAH’S RATE

Hilda Aprina1

Abstract Indonesia is a largest producer of Crude Palm Oil in the world, with increasing production and export from time to time. Since Indonesia now adopts a floating exchange rate regime, the export of such commodity may influence the real exchange rate, and this is the aim of this paper. By applying simultaneous equation model on data from 1984 to 2011, we conclude that the increase in CPO price will lead to an appreciation of Rupiah’s real exchange rate. As a major producer of CPO, the authority should be able to control the world price of crude palm oil to help controlling the stability of Rupiah’s rate.

Keywords: CPO, simultaneous equation, real exchange rate. JEL Classification: E2

1 Hilda Aprina, SST is graduated from Sekolah Tinggi Ilmu Statistik

296

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

I. INTRODUCTION Indonesia has been widely-known as an agriculture country since it depends much on the agriculture sector. The horticulture is one of major contributor in agriculture sector, and performs a rapid growth of GDP by 4.47%2 in 2011. According to the General Directorate of Horticulture, the crude palm oil (CPO) is at the top rank among the nine primary horticulture commodities in terms of export; 53.37% in 2011, or equivalent to USD 17.23 billion. The CPO, which is one of the horticulture commodities, has significantly contributed to Indonesia’s foreign exchange with its high economic value on producing vegetable oil. According to Susila in BPS (2008), the CPO performs strategic contribution on Indonesia’s economy through export, poverty reduction, and a more job creations. Furthermore, the CPO is the primary substance of an alternative energy to substitute fossil fuel;widely recognized as important biodiesel energy. It may eventually affectthe size of world demand on CPO.

������� ���������������������������������������������������������������������� �����

���������������� ����

����

����

����

����

����

����















���������

�����

�����

�����

�����

�����

�����

�����

��������

�����

�����

�����

�����

�����

�����

�����

��������

���

���

����

����

����

����

����

�������

���

���

���

���

���

���

���

��������

���

���

���

���

���

���

���

�������

���

���

���

���

���

���

���

����� ���������

����

����

����

����

����

����

����



���������������������������������������

Since 1984, the output of Indonesia’s CPO has been stable and continuously increasing afterwards. Nevertheless, Malaysia remained at the top rank regarding the market share for export at that moment. In the beginning of 1990, both Indonesia’s and Malaysia’s export share simultaneously increased. However in 1995, Malaysia’s export declined, while Indonesia’s export had been continuously growing until Indonesia declared its position as the major CPO producer in the world, surpassing Malaysia’s position. The CPO production of Indonesia reached 23.9 tons in 2011 which was the highest in the world. 2 BPS (2013). Gross Domestic Product on the Basis of Current Price Based on the Economic Sectors

The Impact Of Crude Palm Oil Price On Rupiah’s Rate

297

With a largest market share for export, Indonesia should have become the standard of world CPO price. For this reason, the world price of CPO may affect the Rupiah’s real exchange rate. Considering since 1977 Indonesia has embraced the floating exchange rate system, thus the role of export commodities becomes urgent onthe volatility of Rupiah. Figure 1 exhibits the relationship between the Rupiah’s real exchange rate and the growth of CPO’s international price.

�������� ��

�������������

���������������������������

������������

�� �� �� � ��� ��� ���� ���� ���� ���� ���� ���� ���� ���� ���� ���� ���� ���� ����� ��������������������������������������������������

Figure 1. Percentage of Rupiah’s Real Exchange Rate Growth, World CPO’s Price, and CPO’s Production of Indonesia Period 2000-2011 (%)

According to Edwards (1986), change in price of leading export commodities usually affects the behavior of exchange rate both directly or through monetary transmissions. Some of the previous literatures focused on the impact of Rupiah’s exchange rate on export of CPO. On this paper, we will deeply focus on the impact of the CPO’s domestic and international price on the Rupiah’s real exchange rate. The paper limits its scope to only focus on the CPO commodity with HS code 1511110000, during the period of 1984 to 2011. This paper will focus on analyzing if the change in CPO’s world price will affect the real exchange rate. The paper attempts to analyze the magnitude of CPO’s world price on the real exchange rate through the change in money supply and the inflations. Moreover, the paper will also provide a comprehensive description on CPO. The next section of this paper outlines the theory and related empirical literature. Section three discusses the data and the methodology, while section four presents the result and its analysis. Section five present the conclusion and the implication of this study.

298

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

II. THEORY The price change of export commodity has significant impact on the dynamics of its real exchange rate (Edward, 2011). In a certain condition such as in the case of commodity boom, a country’s exchange rate will likely appreciate. The change in the price of the export commodity may also affect the monetary sector. Chen and Rogoff (2003) analyzed the real exchange rate of Australia and New Zealand, and argued that the relationship between the exchange rate and the export commodities was driven by the world price of this commodity. This finding is in accordance with Cashin, Cespedes, and Sahay (2004), which found similar evidences in developing countries. In the case of South Africa, Frankel (2007) showed that the mineral was one of important export commodities with significant price impact on the real exchange rate of corresponding country. Ngandy (2005) also argued that the relationship between the export commodity price and the real exchange rate are mostly evident in developing countries. CPO is the leading horticulture commodity and plays an important role with its great contribution for Indonesian economy on foreign exchange, enlarging the market share, and employment3. Trade of CPO and its derivatives is the second largest income source from the non-oil and gas sector, which eventually increases the money supply. According to Boediono (1993), with a surplus balance of payment, there will be more incoming foreign exchange, hence an increase in money supply. Accordingly, an increase of CPO’s world price will raise the national income as well as the money supply. When the money supply increases, the price of goods will also increase. This is in line with the quantity theory of money, which shows a direct relationship between the money supply and the change in price of goods in linear form (Dornbush, 2001, page 373).

MxV=PxY

(1)

Where M is money supply; P is price rate; V money circulation; and Y is output. To measure the average price, economists formulate an index of average prices of distinctive commodities based on their contribution; we recognize it as Consumer Price Index or CPI, (Lipsey, 1995). Inflation is driven by the cost push and demand pull inflation, (Samuelson and Nordhaus, 2004); an increase in money supply will lead to demand pull inflation. Demand pull inflation initially occurs when the aggregate demand increases and creates excess demand within the market. If the full employment condition is already in place, then the subsequent demand will only increase the price (frequently said as pure inflation). The shift up of aggregate demand may be driven by a monetary expansion,including the increaseof government expenditure and

3 BPS (2008), Primary Commodities Discussions, Jakarta: BPS.

The Impact Of Crude Palm Oil Price On Rupiah’s Rate

299

money supply. Meanwhile, the expected inflation depends more on the economic agent’s behavior, which tends to be either adaptive or forward looking. According to Muqrobi (2011), inflation will lower the growth of the output that is going to subsequently decline the import and the demand for foreign currency. This in turn will appreciate the domestic currency. The relationship between the inflation and the exchange rate has been subject to analysis widely. Khodier (2012) argued that there is a strong bidirectional relationship between the inflation and the exchange rate. The statement is in line with Imimol (2011) who argued the inflation depends on the depreciation of nominal exchange rate, the money supply, and the economic growth. The value of the exchange rate can be in nominal and real term. A common nominal exchange rate is bilateral exchange rate between two countries, i.e. Rupiah per US Dollar. Meanwhile the real exchange rate is the nominal exchange rate adjusted to price rate. In detail, the relationship between nominal and real exchange rate can be shown by the following equation.

RER = e x

�∗ �

(2)

where RER is real exchange rate; e is nominal exchange rate; P* is the rate of international price; and P is the rate of domestic price. The above equation implies when real exchange rate is appreciating, the relative domestic prices will be higher, while the price of international goods and services decline. On the other hand, if the real exchange rate depreciates, the price of domestic goods and services will be lower and the price of foreign goods and services will relatively be more expensive. On this regard, the real exchange rate is a benchmark for a country’s competitive advantage of their export commodities in the global market. The following figure shows that the relationship between the net export and the real exchange rate is negative; the lower the real exchange rate, the cheaper the domestic goods and services. The curve showing the savings and investment gap is vertical, since it is independent from the exchange rate (Mankiw, 2007, page 131).The currency exhibited on the figure is US Dollar to foreign currency, so that the direction will be different in this research. If the currency is depreciating, then foreign currency will be appreciating. It drives higher export and lower import. Thus, foreign currency has a linear relationship with export volume while domestic currency has a linear relationship with import volume.

300

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

�������

���

����������� ��

���������

�������������� �������������������� Keterangan: NX

= Net export

RER

= Nilai tukar riil

S

= Tabungan

I

= Investasi

Sumber : Mankiw, 2007

Figure 2. Determinant of Real Currency

where is net export, is real exchange rate; is savings; and is investment.

III. METHODOLOGY 3.1. Simultaneous Equation Model The basic principle of simultaneous equation model is its bidirectional (simultaneous) causality across explanatory variables, and makes difficult to distinguish between the dependent variable and the independent variables (Gujarati, 2004). Thus, it is better to pool a number of variables that can be explained simultaneously by other explanatory variables. The simultaneous model contains more than one equation. In contrast to single equation model, the parameter estimation of certain equation in simultaneous model should consider all information provided by other equation inside the system. In a simultaneous equation model, there are two types of variables: the endogenous, which is determined inside the model; and the predetermined variable, where its value is determined outside the model.Predetermined variables may consist of exogenous variables, both present and lag, and the lag of endogenous variables, (Gujarati, 2004). Equation constructed based on economic model is known as structural equation or behavioral equation as it reflects the structure of economic behavior. The general structure of simultaneous equation model can be expressed as follow, (Gujarati, 2004):

The Impact Of Crude Palm Oil Price On Rupiah’s Rate

301

��� = ������ + ������ + ⋯ + ����� � + ������ + ������ + ⋯ + ������ + ��� ��� = ������ + ������ + ⋯ + ����� � + ������ + ������ + ⋯ + ������ + ��� ��� = ������ + ������ + ⋯ + ����� � + ������ + ������ + ⋯ + ������ + ��� �� � = �� ���� + �� ���� + ⋯ + �� ��� � + �� ���� + �� ���� + ⋯ + �� ���� + �� �

(3)

where: Y1, Y2, ... , YM

= independent or endogenous variables by M units,

X1, X2, ... , XK

= exogenous variables (predetermined) by K units,

U1, U2, ... , UM

= disturbance variables by M units

t = 1, 2, ... , M

= the number of observations,

b

= coefficient of endogenous variables,

g

= coefficient of exogenous variables.

Determining exogenous and endogenous variables may depend on the researcher’s set up but should base on certain theoretical guidelines. The model in this paper refers to Edward (1987). We treat the real exchange rate as endogenous since it depends on the domestic and the international inflation rate,and the world’s price of CPO. Some prior studies also assume the real exchange rate to be endogenous due to its rapidresponse towards shocks.

3.2. Data and Model Specification We use nominal exchange rate of Rupiah against US Dollar (E), money supply (M2), Gross Domestic Product (GDP), Consumer Price Index (CPI), and Producer Price Index (PPI). The data are accessed from the International Financial Statistics (IFS). For the CPO price, the source is accessed from UNCTAD, while the budget deficit is accessed from the Ministry of Finance, Republic of Indonesia. Following Sebastian Edward (1987), we construct an econometric model based on theoretical framework and empirical facts, to investigate the impact of the world price of CPO on money supply, domestic inflation, and the real exchange rate. The structure of the equation system is below.

302

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

Money Supply Growth Function:

��(�� ) = �� + �� ��(��−� ) + �� �� (����) + �� (��(���� ) + ��(��∗ � )) + ���

(4)

Inflation Rate Function:

�� = �� + �� ��(�� ) + �� �� + ��( ��(���� ) + �∗� ) + ���

(5)

Real Exchange Rate Function:

��(���� ) = �� + �� �� + �∗� + �� �� (��∗ � ) + ���

(6)

The expected sign are : a1>0, a2>0, and a3>0; d1>0, d2>0, and d3>0; and μ10 and μ3 � ) + �� �

(1)

The regression threshold method that will be implemented is that used by Hansen (1999) for non–dynamic panel data. The data used is balanced panel with the structure {yit, qit, xit : 1 ≤ i ≤n, 1 ≤ t ≤T}. Dependent variable is shown by yit for individual i and time t which is scalar and xit is an independent variable which is matrix, while qit is variable threshold in the form of scalar and μi shows individual effect. Therefore, the regression threshold method for panel data can be symbolized as follow: 4 Robustness test yang dilakukan : uji Allerano-Bond, Sargan Test untuk menentukan model panel dinamis atau statis, serta Hausman Test untuk menentukan random effect atau fixed effect.

352

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

�� � = �� + �1′ �� � � ( �� � ≤ � ) + �2′ �� � � ( �� � > � ) + �� �

(1)

Where I(.) is an indicator function which is valued 1 or 0 depend on its threshold value.

1 jika� ≤ � �

� ( �� � ≤ � )

{0 jika��� �� >

� ( �� � > � )

�� > { 10 jika� jika�� � ≤

� �

So that equation (1) can be written:

�� �

{

�� + �1′ �� � + �� �, � � + �2′� � � + � � �,

�� � ≤ � �� � > �

Or also can be written as follow:

�� �

{ ��

�(�� � ≤ �) � � �(�� � > �) ��

And b = (b’1 b’2) so that equation (1) in above will equall to

�� � = �� + �′ �� �(�) + �� �

(2)

In model (1), sample is grouped into two parts depend on whether those data is located in above or below g threshold. Both of these data groups (regimes) are categorized by the regression slope of b1 and b2. The first group, which is called first regime, contain sample that fulfill qit ≤ g criteria and in the second group is called second regime that contains sample which fulfill qit > g criteria. If b1 ≠ b2 then it is said that there is threshold in regression equation and model (1) is feasible to be used, but if b1 = b2then the ordinary regression model must be used. For that reason, it is needed to test the hypothesis that b1 = b2. The result of this test will determine whether the (1) threshold regression model or the ordinary regression model that will be used. In this threshold model, it assumed that xit and qit are not varied toward time or in other word model used is static fixed effect panel. The error of eit is assumed as independent and identically distributed (i.i.d) with mean 0 (zero) and variance s2 or eit ~i.i.d N(0, s2).

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia

353

Equation (1) if the mean is taken, it will result

�� = �� + � ′ �� (�) + ��

(3)

With 1

�� =



1

σ ��=1 �� �,�� =

1 �� ( �) = �

�� � =

{



σ ��=1 �� � , and



Σ�

� � (�)

�=1



1 �

Σ�

� (�� � ≤ �)

1 �

Σ�

� (�� � > �)

��

�=1 �

��

�=1

The difference between equation (2) and (3) results

�� � − − �� � = [�� � � − − �� �(�)]� ′ + [�� � − − �� � ] Or it can be written

��∗� = � ′ ��∗�(�) + ��∗�

(4)

If the data and individual error are placed in vector with one time period erased, it becomes:

[]

∗ ��2 . ��∗ = .. , ��∗�

[ ] []

∗ ∗ ��2 ��2 (� ) . .. ��∗ (�) = ,��∗ = .. . ��∗� ��∗�(� )

Y*, X*(g) and e* are individual data notation that is stacked then

354

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

[]

[ ] []

�1∗ . � ∗ = .. , ��∗

�1∗ �1∗( �) . . � ∗ (�) = ,� ∗ = .. .. ��∗ ��∗ ( �)

So that the equation (4) can be written

� = � ∗( �) � + � ∗

(5)

For the g value, coefficient or b slope can be estimated with Ordinary Least Square (OLS) method and it will give the same result with equation:

�( �) = ( � ∗( �) ′ � ∗( �))

−1

� ∗ (�)′� ∗

(6)

With residual regression

�^∗ (�) = � ∗ − � ∗ (�)�(�) And the sum of square error

�1 (�) = ^ � ∗ (� )′ � ∗ (�) = � ∗ ′(� − � ∗ (�) ′ (� ∗ (�) ′ � ∗ (�))−1 � ∗ (�) ′)� ∗

(7)

The g estimation is conducted by minimazing the sum of square error in above equation:

� = ���� �� �1( � ) ^

(8)

With the bottom limit and up limit toward threshold:

͟ Γ = ( �, ͟ �) The calculation of value estimation of g threshold is conducted by finding the smallest of S1 (g) value where the most value variation is as many as nT units. For doing minimization, the observation value is ordered continuously then the smallest percentile value is erased, h%, and the biggest percentile value is erased, (1- h)%, for h>0. The residual observation value from the � candidate. This research previous calculation as many as N observations is g value which isas ^ ^ will use simplification method as in Hansen (1999) by finding � candidate in certain quantile, which are {1,00%; 1,25%; 1,50%; 1,75%; …….98,50%; 98,75%; 99,00%} consisted from 400 quantiles.

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia

355

The regime distribution or g determination considers the number of samples in every regime where the number of samples in one of the regimes can not be too low. Thus, it needs to be sured that the minimum of sample percentage is located in the both of these different regimes for instance 1% or 5% of total samples. If ^ � is already gathered then the estimation of b value is the modification from (6): ^

^

^ � = �(� )

(9)

With residual vector ^∗

^ � = �^∗ (�)

(10)

And residual variation ^2 � =

1 1 ^∗′ ^∗ e e = � (�) n(T − 1) n(T − 1) 1

Significancy of the Threshold After getting the threshold value, it is needed to conduct a test whether the influence of this threshold is significant or not with hypothesis as follows:

�0 : �1 = �2 �1 : �1 ≠ �2 The assumption used in this H0 test is that the value of g threshold can not be identified because there is no threshold influence so that the test of its likelihood ratio has no standard distribution. The critical value determination of this model is conducted by bootstrap procedure as shown in Hansen (1996) which aimed to simulate the asymptotic distribution from likelihood ratio test. Below there is no threshold in H0; model (2) can be rewrittenas

�� � = �� + �′ �� � + �� �

(11)

And after passing the process of fixed effect transformation, equation (4) becomes

��∗� = � ′ ��∗� + ��∗�

(12)

b1 parameter is estimated by using least square method then resulting �1 with residual ��∗� and sum of square errors �0 = � ∗′ � ∗. Thereby that likelihood ratio test for H0 can be

calculated based on (12).

356

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

�1 =

^ �0 − �1 (� ) ^2 �

(13)

F1 distribution in the previous explanation is not a standard distribution dominated by ck distribution so the critical value can not be known. The critical value determination from F test is conducted by estimating from F asimptotic distribution through bootstrap procedure. The steps in bootstrap procedure are as follow: 2

1. Treat the independent variable xit and threshold variable qit as given and use the constant value of xit and qit when conducting bootstrap procedure. ∗

^ 2. Take the residual � � � and conduct a grouping based on individual:

^∗

∗ ^∗ �� = (�^�1 , ��2 , … . . , ^ ��∗� )

3. Use sample distribution {�1∗ , �2∗ , … . . , ��∗ } as the empirical distribution for used in bootstrap time. ^

^

^

4. Take random sample (with returning) from empirical distribution on number 3 above and use its error to take bootstrap sample below H0. 5. With this bootstrap sample, estimate models, which are below H0 in equation (12) and H1 in equation (4), and calculate the F1 likelihood ratio according to equation (13). 6. Repeat the steps from number 1 to 5 in many times such as 1.000 times. 7. This bootstrap procedure will result p-value that is asimptotik for F1 below H0. 8. H0 is rejected if the p-value of bootstrap is bigger than the expected critical value.

Consistency of the Threshold After the test of threshold significancy above and proven that the effect of threshold is exist in model (b1 ≠ b2), then the next step is conducting a test whether ^ � is a consistent ^ estimator for g0 (the actual value of g) or not. � is called as a consistent estimator for g0 if fulfilling 2 (two) requirements, first, the likelihood ratio g0 is lower than its critical value, and we use the following null hyphotesis:

�0 : � = �0 �1 : � ≠ �0

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia

357

With likelihood ratio

��1 (�) =

^ �1 (�) − �1 (�) ^2 �

(14)

H0 is rejected if the value of LR1(g0) is bigger than its critical value. For conducting this calculation, used some technical assumptions: if g0 is the actual value of g, q = b2 – b1 and C = naq, where a Є (0, ½). If ft(g) is density function from qit and zit = C’xit, then

Σ �

� ( �) =

�=1

� (��2� | �� � = �)�� (�)

And D = D(g0). Conditional density from qik given qit is symbolized with fk|t (g1 | g2). Some assumptions used are: 1. For every t, (qit, xit, eit) independent and identically distributed (i.i.d) acrossi; 2. For every i, eit i.i.d overt and independent toward { (�� � , �� � )� � =1} and E(eit) = 0; �







3. For every j = 1,….,k, � (��1 = ��2 = ⋯ = �� �) < 1, where �� � is element at-j from xit; 4. For s > 2, E|xit|s< ∞ and E|eit|s< ∞; 5. For C < ∞ and 0 < a < 1/2, q = n-aC ; 6. D(g) continue at g = g0 7. 0 < D < ∞ ; 8. For k> t, fk|t (g0 | g0) < ∞. Under the assumptions from 1 to 8 above and H0 : g = g0,

��1( �) → � � When n

(15)

∞, where x is random variable with distribution function

� (� ≤ � ) = (1 − exp (−�/2))2

(16)

The equation (15) above shows that likelihood ratios are not standard distribution but they are free from disturbance parameter. Moreover, another assumption which is used is (b2 – b1) 0, if n ∞, it means that the slope difference is in between two small regime toward the number of sample. It can be concluded that equation (15) will give a better result in a smaller

358

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

value of (b2 – b1). Nevertherless, if the threshold effect in model is big, then the estimation of this threshold is quite accurate. The distribution function (16) above has an inverse form that is easier to calculate the critical value:

� (�) = −2 log( 1 − √ 1 − � )

(17)

H0 will be rejected in a asimptotik level, if LR1(g0) is more than the value of c(a). The second requirement for the threshold to be valid is the value of ^ � is located in the range confidence interval. According to Hansen (1997), the best procedure to arrange the confidence interval for g was by creating “no-rejection region” using g likelihood ratio statistic. In order to create an asimptotik confidence interval (CI) for g, the confidence level of 1 – a from“no-rejection region” was a group of g values where LR1(g) ≤ c(a). This CI was an output from the calculation of model estimation. To get the least square estimation ^ � , it was conducted by ordering calculation the sum of square error S1(g) in equation (3.17). The sequence of likelihood ratio LR1(g) is re-normalization from the previous calculation number so there is no need to do further calculation.

3.2. Data and Empirical Model In this research, the model which was used was based on the model used by Baggs (2011), which measured the impact of the real exchange rate toward the profitability of a company. The equation inserted the control variable for industry which were the productivity which was measured by TFP, ratio of concentration, and industry sale level, as well as the control variable for the aggregate of macro economy, which were BI Rate such as what had been used in the empirical study on the influence of exchange rate toward the performance of the manufacturing industry in Indonesia (Surjaningsih, N., dkk, 2011). In that research, it was concluded that the appreciation of the real exchange rate did not prove to give pressure toward the performance of the industry sector in Indonesia empirically. However, along the inclined of the content of export industry, the appreciation of the real exchange rate would give pressure toward the performance of the manufacturing industry in Indonesia. Thus, it was indicated that there was a threshold of the real exchange rate which influence negatively toward the performance of the manufacturing industry in Indonesia. Variable indicated to be influenced by the real excahnge rate is the degree of industry sale that on its turn, it will give impact toward the level of company profit. If the real exchange rate starts to give pressure toward the manufacturing sector, then, the level of sale of the industry (measured by the growth) will return to be negative and will decrease the level of the

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia

359

manufacturing industry. Therefore, the variable which interacts with the threshold of the real exchange rate is the SALES with the basic model as follow:

������� � = �� + �1 ������� � + �2 ���� � + �3 ��� �

(18)

+ β1 ������ � I ( ��� � ≤ �) + β2 ������ � I ( ��� � > �) + �� �

This research on the threshold of the real exchange rate in manufacturing industry in Indonesia is conducted in the level of the individual of the company. The threshold of the real exchange rate above is either in the form of level or growth in order to get the sensitivity of the manufacturing industry in Indonesia toward the level of the real exchange rate occurred.The samples used were the individual of company listed in the survey of big-and-medium-scaledindustry published by BPS period 2001-2009. Knowing that the threshold method is design for the data of balanced panel, thus, in this research, there are 225 companies chosen in that period of time as the samples of the research (the Table of Sample Distribution).

������� ������������������� ���� �� �� �� �� �� �� �� �� �� �� �� �� �� ��

������������������� ����������������� ������� ������� ������� ������������������ ������������������������� ���������������� ������������������������������������� ���������������������������� ����������������������������� �������������������� ������������������������������ ������������������������������������������ ����������������������������������������� �����

�������������� �� � �� � � �� � �� �� �� �� � � �� ���

� ����� ���� ���� ���� ���� ���� ���� ���� ���� ����� ���� ���� ���� ����� ������

In this research, the company performance is measured by using profitability or profit level which is that company ability to get profit that is measured relatively toward fixed asset total that can be seen as follows: Profitabilitasit =

outputit- inputit total asetit

360

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

Meanwhile, verify variable used and expectation sign wanted from the econometric test result can be seen in following table: ������� ���������������������������������� ��������

����������

��� ������������� ������������� ������������������

���������������������������������������������������� ����������������������������������������������

�� �������������� ������

�������������������������������������������������� ��������������������������������������������������� �������������������������������������������������� �������

���������

���� �����������

������������ ���������



������������ ���������

������ ������������ ����� ������������� �������

������������ ���������



����������� ���������

� ������ ����������

��������� �����������������������������

����

�������������������������������������������������������� ���������������������������������������������

�����

������������������������������������������������������� ��� ��������� ������� ������������ ��� ���� �������� �� ����������������������������������������������������� �������� ���������������

�������������������������������� ����������������

���������� ��������� ��������������

������

���������������������������������������������������� ���������������������������������������������������� ��������������������������������������������������� ������������������� ��������������������������������������������������������� ������������������������������� ��� ������������������������������������������������� ����������� ��������������������������������������������� �������������������������������������������� ��������������������������������������������� ������ � ��� � �� �� � � � �� ��� ��� � �� ���

��� � ������ ����������

�� ������������������������������������������� ��������������������������������������������� ������������������������������������������������ �������������������������� ������������������������������������������� � ���������� � �� ��������������������������������������������� ������������������������������������������������� ������������������������������������������� �������������������������������������� ���� ������������������������������������������ ����� �����������������������������������������

5 Pemilihan tahun 2007 tersebut alasan kondisi perekonomian Indonesia cukup kondusif untuk melakukan perdagangan �� dengan ��������������������������������������������� ������������������������������ internasional.

����������������������� �� ������������������������������������������ ���������������������������������������������� ������������������������������������

��������������������������������������������� ������������������������������������������������ �������������������������� ������������������������������������������� ����������� Threshold of Real� Exchange Rate and The Performance of Manufacturing Industry in Indonesia 361 �� ��������������������������������������������� ������� ������������������������������������������������� ���������������������������������� ������������������������������������������� �������������������������������������� ���� �������� ���������� ��������� ����������� ���� ������������������������������������������ ������������ � ��� ���������������������������������������������������� ��������� ������������� ���������������������������������������������� ������������� ����� ����������������������������������������� ������������������ ��������� �� ��������������������������������������������� ����������������������������� ������������������������������ ������������ ������ �� �������������������������������������������������� ����������������������� ��������� ������������ �������������� ��������������������������������������������������� ����� ������ �������������������������������������������������� �� ������������������������������������������ ������������� ������� ���������������������������������������������� ������� ������������������������������������ ���� �������������������������������������������������������� ���� ��������������������������������������������� �� �� �� ��

����� ������ ������ ���������� ������������������������������������������������������� ����� ���������������� ��� ��������� ������� ������������ ��� ���� �������� �� ����������������������������������������������������� ���� �������� ��������������������� ��� � �� ��������������� � �� �� � �� � �������������������������������� � � ������ ���������������� �� ��������� ���������� ���������������������������������������������������� ��������� ���������������������������������������������������� ������� �������������� ��������������������������������������������������� �� ������������������������ ������������������� ��� ���������������������������������� ��������������������������������������������������������� �� ������������������ ������������������������������� ��� �������������������� ��� ������������������������������������������������ ��� ������������������������������������������������� ���� ����������� ������������������������������������ �� ��������������������������������������������� ����������������������������������� �� �������������������������������������������� ���������������������������������� ��������������������������������������������� ������� ������������������������������������������������������ ������ ���������������������������������������������� � �� � �� ��� � �������������� � ������ � ������ � �� ��� ��� ������������������������������������������������������������������������������� ��� �� �����

������������ ���������



����������� ���������

� ������ ���������� ��� � ������ ����������

������������ ���� ���������



�������������������������

�� ������������������������������������������� ���������������������������������������������

Economic of scale become an important factor for company in monopolistic competition ������������������������������������������������ �������������������������� market because it will determine the company profit level. Company with a higher technology ������������������������������������������� level will produce an efficient product and give a higher profit level. Control variable which are � ���������� � CR and SALES represents the future prospect in industry sector, where company will have a �� ��������������������������������������������� ������������������������������������������������� higher profit level if operated in more condusive market. If the relation between profit and CR ������������������������������������������� �������������������������������������� is negative, then company will have a better profit level if this company is located inside market that is dominated by 4 big ���� companies (CR which is higher). However, if the the relation between �������� ������� ������� ���������������� profit and CR is positive, it indicates that the theory of SCP was implemented inside the industry. �����will ����������������������������������������� An efficient company capable to sale its product cheaper than its competitor product so �� ��������������������������������������������� that its company market share increase. The increasing of this market share will increase the ������������������������������ �company will have a better profit level if its market concentration ratio in industry. Moreover, ������������ ��������� potency is bigger (SALES is higher). Associated with the influence of real exchange rate �� which ������������������������������������������ ���������������������������������������������� toward company profit level that is doing export, the appreciation of exchange rate will give a ������������������������������������ benefit for companies in overseas because these companies can sell goods in a cheaper price in �







���� �����������������������������������

����� ���������������� ���� ��� �� ����

���������������������

��



362

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

domestic currency money. Therefore domestic company is forced in a high competition either in domestic market or export. In order to maintain its competitiveness, domestic company must do some adjustments and one of these adjustments is by reducing marginal profit so that the company profit level is decline. Meanwhile, BI RATE is a control variable from the side of macro economic which is as measurement approach of external cost to increase the industry sector capacity in order to have a higher economic of scale.

IV. RESULT AND ANALYSIS This research on the threshold of the real exchange rate will focus on the identification of the threshold of the value if the real Rupiah exchange rate in the manufacturing industry in Indonesia both in the form of level of the real exchange rate and the change of the value of exchange rate. As what have been explained, this research used the threshold regression model developed by Hansen using the assumption of panel data of fixed panel. Thus, there was caveat in this research because all models used for KKI 2 were panel data of fixed panel. The basic regression model used is below, with variation of lag addition on its independent variables:

������� � = � + �1 ������� � + �2 ���� � + �3 � �� � + �4 ������ � + �5 � �� � + �� � With the model of threshold regression is:

������� � = �� + �1 ������� � + �2 ���� � + �3 ��� � + β1 ������ � I ( ��� � ≤ �) + β2 ������ � I ( ��� � > �)+ �� � To find out the candidate of the threshold of the real exchange rate in the manufacturing industry sector, it used 400 quantiles which was the threshold was searched in certain quantile {1,00%; 1,25%; 1,50%; 1,75%; ...; 98,50%; 98,75%; 99,00%} along the range level of the real exchange rate on the research period (81,49 ; 101,13) and the changed of the exchange rate (-5,71 ; 20,09). For the test of robustness, a bootsrap was reconducted for about 1000 times. Meanwhile, the regression model separated the exporter company by sorting the companies which answered consistently that they export their output on the questioner of SIBS. For this needs, there had been 16 companies sorted out. The estimation of threshold for the total of the industry was conducted both by level of REER and the growth of REER. To find out the threshold, it was based on the model developed byBaggs (2011), regression also used the lag of independent variable. The estimation of equation both in level and the growth of REER could be seen on table of the Result of the Regression of the Fixed Effect Level REER below. The variable of TFP was positive and significant on influencing the profit level almost in all equation except for the profit

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia

363

using the variable of TFP lag (1). It meant that, the higher technology level would increase the profit level of a company. Thus, the effort to boost the performance of the manufacturing industry sector in a total way could be conducted by using the more advanced technology. While, the positive and siginificant of REER impact toward both on the equation with level of REER and the growth of REER. This findings meant that the level of the real exchange rate (81,49 ; 101,13) and the growth of the exchange rate (-5,71% ; 20,09%) yoy had yet to give pressure toward the performance of the manufacturing industry sector in Indonesia. This result confirmed the previous research which was conducted on the level of sub-sector of industry by using unbalanced panel data.6 Whereas, the variables of BI Rate, CR, and Sales were insignificant for all equation. This insignificancy of the variable of CR meant that the theory of SCP did not applicable for the manufacturing industry sector in Indonesia. In contrast, the significant of TFP strengthtened the theory of Firm Effect Model which stated that the different characteristics in the level of company cause the different in the profitability level. The important of TFP in forming the profitability of company supported the survival of the company which became the samples of the research. Meanwhile, the variable of Sales as the proxy of the market potency was insignificant, this might be because of the structure of the market which tended to be elastic market7 and low profit level. Furthermore, the regression which separated 2 regimes based on the threshold of the exchange rate according to the equation 3.18, it found that there was threshold of the exchange rate on the value level of REER 82,24 and significant on a=5%. The level of the real exchange rate above 82,24until 101,13 was the interval of secure exchange rate for the performance of manufacturing sector. On the other side, the threshold of the growth of the real exchange rate was -5,01% (yoy) and significant on a=3%. The growth of the exchange rate above -5.01% (yoy) until 20,09% (yoy) was the secure growth of the exchange rate for the manufacturing industry in Indonesia. The value of the threshold either by level or by its growth was a consistent threshold value because it was located on the Confidence Interval.

6 Surjaningsih et al, “Rigiditas Penawaran : Faktor-faktor Penyebab Melemahnya Kinerja Sektor Industri”, Bank Indonesia, 2011. 7 Rata-rata CR sebesar 0,28

364

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

������� ��������������������������������������������������������������������� ��� ��������

��������

��������� ������

������ � ������

��� �������� �����

���������� ����� ���

���

�����

�����

�����

�����

����� ��

�����

������ ����� ��������� ����

��������

�������� �����

�����

�����

������

�����

������

�����

������

�����

�����

������� ��

�������� �����

��������

����� �����

���

����� ���

�����

�����

�����

�����

�����

�����

�����

�����

�����

�����

�����

�����

�����

�����

�����

������� ������������������������������������������������������������������ ��� �������� ������

��������� ������

���

����� �������� �������� ����� ����� �����

������ ��� ����� ���

�����

�����

�����

�����

����� ���

�����

�����

�����

�����

������

�����

�����

�����

�����

������

�����

�����

�����

�����

����� �

�����

�����

�����

�����

�����

������������������������

������

�����

�����

������

������������������������

�����

�����

�����

����� �

���������������

�����

�����

��������������������

� �������

������ ���

�����

�����

�������������������

�����

�����

�����

��� �����

�������� ����� �������� ����� �����

������

������

������������������� ���������

���������

�����

������ ��������������������

�����

�����

������� ��

�������� ����� �������� ����� �����

�����

���������� ���

���������

���

��� �����

��� �����

���������������

���������������

���������������

���� ����

���� ����

���� ����

������������������

������������������

������������������

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia

365

������� ���������������������������������������������������������������������� ��� ��������

��������

��������� ������

��� �������� �����

����� ��� ������

����� �����

����� ���

�����

�����

�����

�����

�����

��������

��� �������� �����

����� ��� ������ ��

���������� ���

������ �����

����� �

����� ���

�����

������ ��

����� �����

�����

�����

������

�����

������

�����

�����

�����

�����

�����

����� �

�����

�����

�����

��������� �����

�����

�������� �����

����� ���

������� ��

����� �����

��������

�����

������� �������������������������������������������������������������������

��������

���������

������ ����������

������ �

���

����� ���

���

���

����� �������� �������� ����� �����

����� �������� �������� ����� �����

�����

���������

���

�����

�����

������

�����

�����

�����

�����

�������������������������

������

������������������������

����� �

�����

�����

�����

���������������������

������ ��

�����

�����

���������������������

�����

�����

�����

������

������������������� ��������� ������������������� � ������� ���������������

������ ���

�����

�����

����� ���

�����

�����

�����

������

�����

�����

�����

�����

������

�����

�����

�����

�����

����� �

�����

������

�����

������� ��

����� �������� �������� ����� �����

����� ������ ���

�����

���������

��� ����� �������������� ����

��� ����� �������������� ����

��� ����� �������������� ����

���� ������������������

���� ������������������

���� ������������������

Translation note: koefisien = coefficient, variabel = variable, test efek threshold = threshold effect test

366

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

The sector of manufacture industry was more sensitive to depreciation than appreciation where the industry structure was still using import component. This statement was confirmed by survey that was taken by Bank Indonesia which was almost 40% respondents that was come from manufacture sector was worried about the sharply depreciation of Rupiah and the portion of manufacture company using import raw material was quite big which was 35,2% of total manufacture industry. The weakening of real exchange rate would increase the production cost because the import raw material was more expensive so that the company profit level would be smaller.

V. CONCLUSION This paper is empirical research, and estimation result confirms the presence of threshold of the real Rupiah exchange rate on the level of 82.24. REER ranging from 82.24 to 101.13 with the growth of -5.01% to 20.09% is secure on supporting the manufacturing company performance. This result asks the attention of monetary authority to keep the value of Rupiah within this range, which also helps to maintain the economic stability. Future research should address some caveats on this paper; first, the use of Hansen method require balanced panel data, which lead to a selection of 225 companies during 20012009. Those samples cover only 1.1 % of the total available data from Survey of Medium and Large Scale Manufacturing Industry of around twenty three thousands per year. Second, future research should leng then historical observations to cover the Asian crisis in 1997/98. This is important to capture the sharp depreciation of the Rupiah exchange rate, which may alter the threshold found on this paper.

Threshold of Real Exchange Rate and The Performance of Manufacturing Industry in Indonesia

367

REFERENCES Bodnar, Gordon M, “Exchange Rate Exposure and Industry Characteristics : Evidence from Canada, Japan, and the USA”, Journal of International Money and Finance (1993), 12, 29-45 Boediono, DR., “Seri Sinopsis Pengantar Ilmu Ekonomi No.1 Ekonomi Mikro”, BPFE, edisi 2, 1999. Depari, Y., dkk, “Perilaku Pembentukan Harga Produk Manufaktur dari sisi Produsen, Distributor dan Pengecer”, Bank Indonesia, 2009. Fung, L., et.al, “Firm Survival, Performance, and the Exchange Rate “, University of Calgary, 2007. Firdaus, Muhammad, “ Ekonometrika untuk Data Panel”, Materi Pelatihan di Bank Indonesia, April 2010. Hansen, Bruce E., “Threshold Effects in Non-Dynamic Panels: Estimation, Testing, and Inference”, Boston College, February 1998. J. Baggs, E. Beaulieu, L. Fung, “Firm Survival, Performance, and the Exchange Rate”, University of Calgary Discussion Paper, 2007. J. Baggs, E., et.al., “Exchange Rate Movements and Firm Dynamics in Canadian Retail Industries”, 2011. Kurniati, Yati, dkk., “Dinamika Industri Manufaktur dan Respon terhadap Siklus Bisnis”, Direktorat Riset dan Kebijakan Moneter, Bank Indonesia, 2010. Mishra, SK, “A Brief History of Production Function” Department of Economic North-Eastern Hill University Shillong (India). Nugroho, Wahyu Agung, dkk., “Struktur Biaya dan Perilaku Pembentukan Harga pada Industri Manufaktur Indonesia”, Bank Indonesia, 2005 Pindyck, Robert S., dan Daniel L. Rubinfeld, “Microeconomics”, edisi ke-4, Prentice-Hall, 1998. Sidek, Noor Zahirah Mohd, “Malaysia : How Much Exchange Rate Misalignment is Detrimental to Export”, Department of Economics, UiTM Malaysia, 2011 Simatupang, P., dkk, “Pengaruh Kenaikan Harga Bahan Bakar Minyak (BBM) Maret 2005 Terhadap Profitabilitas Usaha Jasa Alsintan dan Usaha Tani Padi (Kasus Kab. Sidrap, Sulawesi Selatan dan Kab. Nganjuk, Jawa Timur), Pusat Sosial Ekonomi dan Kebijakan Pertanian, Badan Litbang, Departemen Pertanian, 2005.

368

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

Stierwald, A., “Determinants of Profitability: An Analysis of Large Australian Firms”, Melbourne Institute Working Paper Series No.3/10, April 2010. Surjaningsih, Ndari, dkk., “Rigiditas Penawaran : Faktor-faktor Penyebab Melemahnya Kinerja Industri”, Working Paper No.10/2011, Bank Indonesia. Wooldridge, Jeffrey M. “Introductory Econometrics A Modern Approach”, South-Western Cengace Learning, 2009 Yanuarti, Tri, “Dampak Apresiasi Nilai Tukar terhadap Kinerja Industri Pengolahan”, (2006), Catatan Riset, Bank Indonesia, 2006.

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

369

DETERMINANT OF CAPITAL RATIO: A PANEL DATA ANALYSIS ON STATE-OWNED BANKS IN INDONESIA Pamuji Gesang Raharjo1 Dedi Budiman Hakim Adler Haymans Manurung Tubagus Nur Ahmad Maulana

Abstract

Capital plays important role to support the operational of the banks and to create a sound banking system in aggregate. For this reason, the banks are required to have a sufficient amount of capital, both to support its business expansion as well as a buffer to prevent and to absorb any unexpected losses. This paper analyzes determinants of capital ratio of the state-owned banks in Indonesia. Using panel data regression model, the result shows that the capital ratio of these state-owned banks is affected by the size of the bank, the bank’s leverage, the quality of management, and the interest rate risk. Contrary to the existing literatures, this paper does not support the effect of management capability to generate income on the bank’s capital ratio.

Keywords: Capital structure, state-owned banks, panel estimation. JEL Classification: C23, G21, G32

1 Pamuji Gesang Raharjo is student of Business & Management Program of Postgraduate School, Bogor Agriculture University, Indonesia (corresponding author: [email protected]); Dedi Budiman Hakim is supervising Committee Chairman, Lecture of Postgraduate School, Bogor Agriculture University, Indonesia; Adler Haymans Manurung and Tubagus Nur Ahmad Maulana are supervising Committee Members, Lecture of Postgraduate School, Bogor Agriculture University, Indonesia.

370

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

I. INTRODUCTION Banking is the backbone of the Indonesian economy, since banking is still the main source of financing to drive the wheels of the economy and to create growth. Banks also play an important role in allocation of collected public deposits funds, both in the form of productive investment and lending to productive sectors. Moreover, the banks play important role in facilitating the efficient allocation of investment risk (Diamond and Dybvig, 1983). At macroeconomic level, bank is one of important means to transmit monetary policy (transmission belt), while at microeconomic level, the bank is a major source of financing for businesses and individuals (Koch and Donald, 2000). Banks’ capital structure is fundamentally different from capital structure of non-financial companies, because the characteristics of the banking business and operations are different. The banks also need to have a buffer in accordance with the provisions of the minimum capital requirement set by bank regulators (Saunders, 2008). Bank capital plays a very important role in maintaining safety and creating sound banking system. Banks are required to have a sufficient amount of capital, both to support its business expansion as well as buffer, to prevent and absorb unexpected loss arising from variety of significant risks. Thus the purpose of minimum capital requirement is to ensure the banks to keep enough capital for the risks they take. It is impossible to eliminate altogether the possibility of a bank failure, but the governments can minimize the probability of bank default. By doing this, we may expect to have a stable economic environment where the private individuals and business will have good confidence on the banking systems. Bank Indonesia as the central bank in Indonesia has been endeavored to improve the quality and the quantity of commercial bank’s capital in Indonesia. Improving the quality of bank capital carried out by adjusting the terms of bank capital components and its instruments. On the other hand, increasing the quantity of bank’s capital is done by requiring banks to form additional capital above the minimum capital adequacy requirements, based on bank’s risk profile. Minimum capital requirement based on risk profile is not intended only to anticipate the potential losses arising from risk weighted assets (based on the banks’s credit risk, market risk, and operational risk), but also to anticipate the potensial losses from other risk in the future which has not been incorporated in the weight. The risk profil rating and the minimum capital adequacy requirements as stated on Table 1.

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

371

������� ������������������������������������������������������������� ���

�������������������

������������������������ ������������





��





��������������





���������������



������

�������������

����������������������

Commercial banks in Indonesia are also required to have additional capital as a buffer, in accordance with the criteria set by Bank Indonesia. Additional capital consists of capital conservation buffer, countercyclical buffer, and capital surcharge. Capital conservation buffer is the additional capital that serves as a buffer in the event of a loss in the period of crisis. Countercyclical capital buffer is an additional function as a buffer for anticipated losses in the event of excessive credit growth that could potentially disrupt the stability of the financial system. Meanwhile, domestic capital surcharge for systemically important banks is the additional capital for certain bank that serves to reduce the negative impact in the event of a bank failure. Shortly, the domestic capital surcharge is aimed to increase the bank’s ability to absorb losses. Bank Indonesia set the capital conservation buffer at 2.5 % of risk weighted assets, the countercyclical buffer is set in the range of 0% up to 2.5 % of risk weigthed assets, and the capital surcharge is set in the range of 1% up to 2.5 % of risk weighted assets. Commercial banks in Indonesia can be divided into 6 (six) groups; they are State-owned Banks (Bank Persero), National Private Commercial Banks (Bank Umum Swasta Nasional), Regional Development Banks (Bank Pembangunan Daerah), Joint Venture Banks (Bank Campuran), and Foreign Banks (Bank Asing). The group of state-owned Banks consist of Bank Mandiri, BRI, BNI and BTN. The share of state-owned banks’ assets is 36.02 percent by the end of 2012. These state-owned banks enjoy close links with Indonesia’s largest companies, including state-owned enterprises. On the other hand, the state-owned commercial banks are assigned to provide credit to specific sectors, to promote the access to bank services for groups of population or regions not covered by private institutions, to mitigate market failures due to the presence of asymmetric information, to finance socially valuable (but possibly financially unprofitable) projects, and to compete with private institutions to lower the costs of financial intermediation (Yeyati, 2004). Simply stated, the state-owned banks are required to have sufficient capital, both to support its own business operaton and expansion, as well as to stand as agent of development in Indonesia.

372

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

The above illustration underlines the importance of capital for the bank. The purpose of this research is to analyze the determinants of capital adequacy ratio of state-owned banks in Indonesia. Research variables used in this study are divided into two, namely internal factors (bank specific factors) and external factors. Internal factors consist of variables derived from the bank’s performance, as reflected in the balance sheet, income statement, and other financial reports are prepared and published by the bank. On the other hand, the external factors are variables that are not directly related to the bank’s management, but reflect the economic conditions that affect the operation and performance of the bank. Next section of this paper present theory and related empirical studies on the subject. Section three provide methodology and the data we used, while section four explain the result and analysis in details. Section five provide conclusion and will close the presentation of this paper.

II. THEORY 2.1. Capital Structure of the Company Capital structure is concerned with how a firm finances its business by choosing right composition of equity and debt. Several theories are availabel related to the company’s capital structure. David Durand in 1952 developed the theory on capital structure using net income approach. This approach suggests that the use of debt capital by a firm may increase or reduce the cost of capital. Modigliani and Miller (1958) developed a financial theory that became the basic concept of modern capital structure theory; later known as MM Theory. MM theory presumes that capital markets are perfect with no corporate taxes, no bankruptcy costs, no information asymmetry and no agency costs. These assumptions are contradictory to reality, hence cannot be directly applied. Modigliani and Miller (1958) provided the foundational impulse to the study of the capital structure problem by formally proving that, under conditions of complete, perfect and frictionless markets, a firms market value and the welfare of its security holders remain unaffected by financing decisions (Gertler 1988 and Fama 1990). This theoretical proposition carries the implications that: (1) financing and investment policies are independent; (2) internal and external financing are perfect substitutes; and (3) specific type of financing contractual arrangement, either equity or debt, is also irrelevant. The MM theory is widely used by researchers as basic ground on analyzing the capital structure. In 1963, the MM theory was revised by Modligiani and Miller to include the effect of taxes on the value and the cost of capital. With the corporate tax, the value of the firm can vary in accordance with the proportion of debt due to the taxshield of the lending bill (Baral,

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

373

1996). MM theory then known as irrelevancy theory, since this theory states that the proportion of equity and debt has no effect on the firm value (Manurung, 2011). In other words, value of the firm depends on the income generated by its assets, not by how the assets are financed or how the income is divided. Another theory proposed by Donaldson (1961) is pecking order theory, which discusses order financing company. Solomon (1963) developed the theory of optimal capital structure where it is stated that the value of the company will increase up to a certain level, and after that firm value tends to remain constant. With moderate use of debt capital, the value of the company will in turn decrease. Stiglitz (1969), Haugen and the Papas (1971) and Rubenstein have developed a theory of capital structure, known as the trade-off models, which focused on financial distress and taxshield. Jensen and Meckling (1976) put forward the agency theory related to the value of the company due to the conflict between the company’s management (agent) and the shareholders (the principal). Myers and Majluf (1984) suggest that the capital structure can help to mitigate the inefficiencies in a firm’s investment caused by asymmetric information. They show that managers use private information to issue risky securities when they are overpriced. There is a pecking order of corporate financing such that firms prefer internal financing, and if this internal financing is not sufficient, the firms will issue the cheapest security first as external financing source. The theories of the capital structure described above are static and ignore the presences of an optimal capital restructuring in response to the fluctuations of asset value. In other words, the company will always make major adjustments in response to changes in debt the company’s assets. To overcome these weaknesses, Zweibel (1986) and Zechner et al. (1989) developed a dynamic capital structure theory. Zweibel (1986) states that the selection of debt by managers was voluntarily with credible owned limitations. Goldstein, Ju and Leland (2001) also introduced a dynamic structural model with EBIT based models. They argue the firm policy is based on the dynamic capital gains, which depends on the taxshield, the bankruptcy costs, the variability of assets, the interest rate risk, and the size of the recapitalization costs. Baker and Wurger (2002) published a paper on market timing and capital structure stating that the company will issue preferred shares when the stock price is high, and issue bonds when the stock prices is low.

2.2. Bank Capital Structure Bank is simultaneously a firm, a financial intermediary, and a highly regulated business entity. The imposed regulation on bank will determine their capital and their behavior (Marques and Santos, 2004). Given the operational of the bank is different with other industries, and then the business and financial risk of the bank are also different. Therefore, the capital structure of the bank will differ from the non-financial corporation (Buser, 1981).

374

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

Capital regulation affects the capital ratio maintained by the bank. Mishkin (2000) and Ghosh et al. (2003) found that capital requirements affect the bank’s capital structure. Based on previous studies, the bank’s capital ratio is also influenced by other factors, such as asset growth, risk, and profitability. The capital structure taken by management is affected by the owners or shareholders. Ownership structure represents the power to control the management or the company, particularly in deciding important policies. The role of ownership structure on capital structure is inline with the agency cost theory. The agency relationship in banking is quite complex, since it involves the relationship between the shareholders and the management, the relationship between bank and borrowers, and the relationship between bank and regulators. The bank’s capital consists of voluntary and involuntary capital (Besanko and Kanatas, 1996; Cornett and Tehranian, 1992, and Keeley, 1989). Voluntary capital depends on the fundamental of the bank and is voluntarily choosen by the management. On the other hand, involuntary capital is set by the regulator. It is possible for the bank to have excess capital above the minimum capital adequacy ratio for several reasons. One of them is as a hedging strategy when the banks require short term additional capital due to the worsening risk profile or other reasons. If the bank has limited capital, banks can only raise new capital in a short term by selling new shares. The sale of this new share will incur significant transaction costs or may result in share price decline, since investors and public know the bank experiences difficulty on their capital. When the bank’s capital is low, an addition of new capital will also transfer the value to the fixed-income securities holder (including government safety nets). This is similar to the classic conditions whent the debt is too big or debt overhang as proposed by Myers (1977). To that end, the bank tends to maintain higher capital ratios to avoid these problems, and it is also easier for banks to increase the capital when their income is high. Commercial banks have deposits that are insured to protect depositors and to ensure financial stability. In order to mitigate the moral hazard of this insurance, commercial banks must hold minimum amount of capital. Banking regulators use minimum capital requirement to avoid bank failures and to limit the exercised risk taken by the bank. Nevertheless, Kahane (1977), Koehn and Santomero (1980), Kim and Santomero (1988), and Hovikimian and Kane (2000) argues that the capital adequacy ratio is not effective in limiting the bank risk. This is due to strict capital requirements will encourage the bank to maximize the use of their capital, in particular by increasing the risky assets. Other studies explain that factors such as government guarantees (the implicit and explicit deposit insurance, the doctrine of too big - too fail and lender of last resort support), earnings value, and expected bankruptcy costs, will affect the level of capital hold by the bank.

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

375

Assurance given by the government will reduce expected bankruptcy costs as the risk of default is transferred from the bank to the government. This in turn reduces the incentives for depositors to monitor banks closely. At the same time, the bank’s shareholders can take advantage of the slack supervision by increasing the amount of bank loans, which may decrease the ratio of capital and increase the earnings volatility due to inflating risk and the risk transfer to the lender and the guarantor (Hovikimian , Kane and Laeven, 2003). Therefore, the benefits to society arised from government guarantees highly depend on how effectively the regulators on controlling the behavior of banks in shifting risk (Hovikimian , Kane and Laeven 2003). On the other hand, Berger, Herring and Szegö (1995) stated that even the banks are in a business environment that is not regulated and the absence of a government bailout, banks still have to maintain capital due to the demands of the market, which is called the capital adequacy of the market (market capital requirement). Since the failure of a bank could cause instability in the banking system, then the capital of the bank should be regulated. In Indonesia, Yudhistira (2003) examined the impact of capital requirements for the bank using simple model. They found that capital regulation affect the behavior of the banks in Indonesia; hence possibly effect the economy. Tumiwa et al. (2013) found that banking regulations and good risk management practices affect the banks’ capital structure.

III. MEHODOLOGY 3.1. Data This study is based on secondary data obtained from the quarterly publication of financial statements of all Indonesia state-owned banks, Indonesian Banking Statistic issued by Bank Indonesia on a monthly basis, and other publications during the period of the first quarter of 2004 to the fourth quarter of 2012. Using panel data provide detailed information on bank behavior across time and space (Baltagi, 2005; Gujarati, 2003). Moreover, panel data is more robust to violations of the Gauss Markov assumptions, namely heteroskedasticity and normality (Wooldridge, 2010). By the end of 2012 there were 120 banks operating in Indonesia and grouped into stateowned commercial banks, foreign exchange national private banks, non foreign exchange national private banks, joint venture banks, and foreign banks. This study analyzes the determinants of state-owned commercial banks capital ratio. Although only consists of four banks, namely Bank BRI, Bank Mandiri, Bank BNI, and Bank BTN, but these four state-owned banks record average market share of 36.02% of total commercial bank assets, savings deposits amounted to 27.25% of total third party fund from all commercial banks, and distributed loan of 35.29% of total commercial banks loan in Indonesia (as December 2012).

376

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

3.2. Empirical Model and Estimation This study examined how the bank characteristic affect their capital adequacy ratio by using a multivariate panel regression model. To see whether the identified bank-specific variables could explain capital adequacy ratio (CAR), we specify the following empirical model:

CARit = β0 + β1LNSIZEit + β2NIMit + β3EQTLit + β4NPLit + β5IRRit + εit Where CARit is capital adequacy ratio of bank i at time t; LNSIZEit is the growth of total assets growth of the bank i at time t (in natural logarithms); NIMit is athe net interest margin of bank i at time t ; EQTLit is total equity to total liabilities ratio of the bank i at time t; NPLit non performing loan ratio of bank i at time t; and IRRit is interest rate risk of bank i at time t. In the above equation b0 is constant and b is coefficient of variables, while εit is the residual. We set the bank’s capital adequacy ratio (CAR) as the dependent variable. CAR is one important indicator on assessing the health of a bank, since the bank’s capital may reflect their ability to develop their business and to manage sufficient buffer to absorb potensial losses. Given the observation period of this study is the first quarter of 2004 to the fourth quarter of 2012, the minimum capital adequacy ratio used in this study still base on Bank Indonesia Regulation No. 10/15/PBI/2008 dated 24 September 2008; where the minimum capital requirement is 8% of risk-weighted assets (RWA). Bank capital is the sum of the core capital (Tier 1 capital), the supplementary capital (Tier 2 capital), and the additional supplementary capital (Tier 3 capital), after taking into account certain deductions in accordance with Bank Indonesia. RWA covers the credit risk, the market risk, and the operational risk. Thus the CAR is calculated as follows: Car =

Tier 1 Capital + Tier 2 Capital + Tier 3 Capital - Deduction Factor of Capital Total Risk Weighted Asset

(1)

In accordance with Bank Indonesia regulations, banks are required to provide core capital (Tier 1 capital) at least 5% of the risk weighted assets, which consist of paid-in capital, additional capital reserves (reserve disclosed), and innovative capital (innovative capital instruments). The additional supplementary capital (Tier 3 capital) can be used for market risk only, but should not exceed 250% of the core capital allocated for market risk. The supplementary capital (Tier 2 capital) and additional supplementary capital (Tier 3 capital) is maximum 100% of the core capital (Tier 1 Capital). Capital adequacy ratios used in this study are stated in the bank financial statements for each period. There are five independent variables in accordance with previous literature; most of them are internal variables for the bank. We get or calculate these variables from the quarterly financial statements published by state-owned banks. The first is natural logarithm of the bank total asset (LNSIZE). The bank’s asset represents the size of the bank as well as the scale of

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

377

economic of the bank, which potentially influence the amount of bank’s capital. In accordance with previous research hypothesis, banks with high income tend to diversify and posses greater investment opportunities. To reduce the cost of capital and to avoid risk, they tend to hold larger equity capital. On this case, the size will positively affect the bank’s capital. On the other hand, easier access to the capital markets and the government guarantee potentially lead the banks to maintain less capital. For this reason, the size of the bank may negatively relate to their capital. Thus, the estimated coefficient from the model can be positive or negative. This study uses logarithm of total assets (LNSIZE) as a proxy of the size of the bank, following Demirguc-Kunt et al. (2004). The formula use in calculating the growth of the bank’s assets as follows:

LNSIZE = Logarithm (Assett+1 / Assett0)

(2)

The second explanatory variabel is the non performing loan (NPL). NPL ratio is one of the key indicators in assessing the performance and the quality of the bank assets. The NPL ratio shows the ability of bank to manage the loans. Higher NPL indicates worse quality of the bank credit, hence bigger credit risk (Nasser, 2003). The non performing loans is classified as substandard, doubtful and loss. NPL is calculated with the following formula:

NPL = (Non Performing Loans / Total Loans) x 100%

(3)

Another determinant of CAR is the quality of bank’s management proxied by the net interest margin (NIM), which becomes the third explanatory variable on our model. NIM is used to measure the ability of management to generate net interest income. NIM may reflect the cost of financial intermediation, and is defined as net interest income divided by average earning assets of the bank. The net interest income is the interest income minus interest expense. Interest income is generated from productive asset. The productive assets of the bank, in accordance with Bank Indonesia’s regulations concerning Commercial Bank Asset Quality Rating, is the provision of bank funds in the form of loans, securities, interbank placements, bill acceptances, bills repurchase agreement, derivative receivables, investments, balance sheet transactions and other forms of similar provision. NIM affects the bank capital positively. High revenue allows the bank to raise additional capital through retained earnings and provide a positive signal to the market (Rime, 2001). On the other hand, high incomes may mean lower probability of failure (Yu, 1995). As a result,

378

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

high income led the management to reduce “capital cushion” given the low risk of failure. Therefore, the NIM may also affect the capital negatively. The fourth independent variable on the model is equity to total liabilities ratio (EQTL). The ratio of total equity to total liabilities (EQTL) is used as a proxy of leverage. A high EQTL signifies low leverage (low debt/liabilities), and accordingly low EQTL reflects high leverage. Banks with high leverage (low EQTL) may find themselves difficult to raise new capital; hence lower equity. We expect the EQTL to positively affect the capital of the bank. The fifth independent variable is interest rate risk (IRR). The main activity of the bank is to collect funds and to distribute it. Interest rate risk is inherent within the bank’s assets and liabilities, ie the risk of losses associated with the different sensitivity of the productive asset and source of bank funds due to interest rate changes. It is also ascociated with the maturity gap between the assets and the liabilities. We calculate the interest rate risk as follows:

IRRit

= Productive Assetit / Liabilitiesit

(4)

where IRR is the interest rate risk; productive assets is calculated as the ratio of bank’s total productive asset equity to total asset; and liabilities is the source of funds for the bank. We estimate the above empirical model using panel data regression. To choose best model variant across Pooled Least Square (PLS), Fixed Effect Model (FEM) and Random Effect Model (REM), we use the Chow-test, Lagrange Multiplier test (LM-test), and the Hausman test. We use the Chow-test to choose between the Pooled Least Squares (PLS) and the Fixed Effect Model. Chow-test assumes the error of the regression is normally distributed with equal variance (s2). If the value of Chow Statistic (F-stat) generated from the test is greater than the F-table, the null hypothesis is rejected so that the model chosen for use is Fixed Effect Model, and vice versa. We use LM-test to chose between the PLS and the REM, while the Hausman-test is used to compare the FEM and the REM.

IV. RESULT AND ANALYSIS Firstly we tested the correlation among Capital Adequacy Ratio (CAR), asset growth (LNSIZE), the ratio of non-performing loans (NPLs), net interest margin (NIM), the ratio of equity the bank’s liabilities (EQTL), and interest rate risk (IRR). The result is presented in Table 2. The NPL had the strongest positive correlation with CAR (0.5862). The other independent variables with significant positive correlation with the CAR are the ratio of equity to total bank

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

379

liabilities (EQTL) and interest rate risk (IRR). On the other hand, net interest margin (NIM) and total equity to total liabilities (EQTL) have a positive correlation with bank’s asset growth (LNSIZE). ������� �������������������� ���

������

���

���

����

���

������

������

�������

������

���

������

�������

������

���

�������

������

�������

������

����

������

������

�������

������

������

���

������

�������

�������

������

������

���

������

����������������������

Tabel 3 provides the summary of descriptive statistics for state-owned banks used in this study. The capital adequacy ratio of state owned banks always above the minimum capital requirement set by Bank Indonesia (8%), for all time of observation (2004Q1 to 2012Q4). The lowest capital adequacy ratio was 12.02% and the highest was 27.81% with an average capital adequacy ratio was 17.50 %. During the same period, the lowest asset growth rate was -12.21% and the highest was 21.65% with average asset growth was 3.42%. The lowest ratio of non-performing loans (NPLs) of state owned banks was 1.74% and the highest was 27.66% with an average of 6.24%, while the net interest margin (NIM) of 3.81% lowest and highest 12.36%. On the other hand, the lowest risk index of state-owned banks is 7.59 and the highest is at 45.64, with an average of 19.68. From the results of the Chow and the Hausman test, the model to use best is FEM. This is in line with Nachrowi and Usman (2006) which states that if the data panels have held a greater amount of time than the number of individuals observed, it is advisable to use the FEM. Table 4 shows the complete results data processing using a data panel fixed effect model to analyze the factors that affected on the capital adequacy ratio of state-owned banks in Indonesia.

380

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

�������� ��������������������� ����������� ��������

��������

��������

�������� ����

�����

������

������

������

�������

������

������

������

������

������

������

������

������

������

������

������

������

��������

������

������

������

������

������

������

������

�������

������

���

������

������

������

������

�������

������

������

������

���

������

������

������

������

������

������

��������

������

����

����

����

����

���

���

������

������

������

������

���

�������

������

���

���

������

���

���

���� ���

���� ����

�����������������������

The coefficient of determination (R-squared) is 53.97%. It states that the variation of explanatory variables (LNSIZE, NPL, NIM, EQTL, and ZRISK) can explain 53.97% the variation of CAR, while 46.03% is explained by other variables. The result indicates other determinant of the state-owned bank’s capital ratio is not included in the model.

�������� ���������������������� ��������

�����

��

�������

�����



������

������

������

������

������

�������

������

�������

������

���

������

������

������

������

���

������

������

������

������

����

������

������

������

������

���

�������

������

�������

������

��������

������

������������������

������������������

������

������������������

������

������ ������ �������� ������� ������

�������������������� ����������������� ������������������� ������������������

������� ������� ������� ������

������������������ ����������������� �������������� ����������� �����������������

������

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

381

The estimation shows that the growth of the bank total assets negatively affects the state-owned bank capital ratio and is statistically significant at α = 0.05. This means a 1% addition of the bank’s assets (in logarithm) led to a reduction of the bank’s capital adequacy ratios by 0.08%. The growth of the bank’s assets is mainly due to the increae of productive assets, both in loans and investments in other risky assets. The increase of growth in loans and risky financial instruments will raise the bank’s potential losses from bad debts and losses from declining of financial instruments price hold by the banks. In accordance with the regulation on bank capital, the raise of bank’s risk weighted assets will lower the bank’s capital adequacy ratio. Our result is inline with research conducted by Kane (2000), Mishkin (2006), and Rime (2001), who obtained a negative relationship between the size and the capital ratio. From a safety net perspective (systemic risk), this findings confirm the common believe that larger banks are ‘too-big-to-fail’ or “too-big-to-discipline-adequately”. High non performing loans (NPL) are commonly associated with high risk and poor management (Barrios and Blanco, 2003). The estimation result shows that the non-performing loan positively and sifnificantly effects the bank capital ratio. This indicates the bank anticipate any increase in potential losses from bad loan by raising their capital. This resulst is inline with Ahmad et al. (2009). The net interest margin (NIM) has a positive coefficient but not statistically significant. The state owned banks in Indonesia have been very profitable, retained a lot of earnings, however, this finding indicates that it does not affect their capital ratios. Generally, the interest margin (NIM) positively affects the bank capital, since high revenues allow the banks to raise additional capital through retained earnings and to give positive signal to the value of the company, (Rime, 2001). A high earnings or franchise value provides bank managers an easier access to equity capital and a self-regulatory incentive to minimize risk taking (Cebenoyan et al. 1999; Saunders and Wilson 2001; Ahmad et al., 2009). But again, the insignificant of the estimated NIM coefficient from our model, does not support those literatures. The ratio of equity to total bank liabilities (EQTL) is statistically significant at the 0.01 level and has a positive effect on state-owned bank capital ratio. The state owned banks in Indonesia tend to hold high capital and low leverage. The positive sign of EQTL coefficient indicates a negative relationship between the bank leverage and the risk-weighted capital adequacy ratio (Ahmad et al. 2009). The last explanantory variables, the variable interest rate risk (IRR), is statitically significant on affecting the state-owned banks capital ratio. The coefficient of IRR shows that a one unit increase in interest rate risk will reduce the bank’s capital adequacy ratio by 0.07 units.

382

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

V. CONCLUSIONS This paper analyzes the determinant of capital ratio of the state-owned bank in Indonesia. In line with other economies experiences and existing literatures, the capital ratio of the state owned banks in Indonesia is determined by the asset growth (LNSIZE), the amount of nonperforming loans (NPL), interest rate risk (IRR), the equity to total liabilities ratio (EQTL), and the net interest margin (NIM). Except for the net interest margin (NIM), this paper did not find significant effect of the NIM to the bank’s capital adequacy ratio. This study covers only state-owned banks in Indonesia. With possible differences in business characteristics, incentives and organization structure, then future study should incorporate other types of bank, particularly those with different interest rate sensitivity.

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

383

REFERENCES

Ahmad, R., Ariff, M., and Skully, M.J. (2009). The Determinants of Bank Capital Ratios in a Developing Economy. Asia-Pacific Finan Markets 15:255-272. Baker, Malcom, and Jeffrey Wurgler. (2002), Market Timing and Capital Structure. Journal of Finance 57, 1-32. Baltagi, B. H. (2005). Econometric Analysis of Panel Data, Chichester: John Wiley and Sons. Baral, Keshar Jung. (1996), Capital Structure and Cost of Capital in Public Sector Enterprises in Nepal. Ph.D thesis. Delhi University. Barrios, V.E.J. and Blanco, J. (2003). The Effectiveness of Bank Capital Adequacy Regulations: An Empirical Approach. Journal of Banking and Finance, 27, pp.1935-1958. Berger, A., DeYoung, R., Flannery, R., Lee, D., and Oztekin, O. (2008). How Do Large Banking Organizations Manage Their Capital Ratios?. The Federal Reserve Bank of Kansas City, Research Working Paper 08-01. Berger, A.N, Herring, H.J., and Szego G.P. (1995). The Role of Capital in Financial Institutions. Wharton School Center for Financial Institutions,Working Paper 95-01. Besanko, D. and Kanatas, G. (1996). The Regulation of Bank Capital: Do Capital Standards PromoteBank Safety?. Journal of Financial Intermediation. 5, 160-183. Buser, Stephen A., Andrew H. Chen, and Edward J. Kane. (1981). Federal Deposit Insurance, Regulatory Policy, and Optimal Bank Capital. The journal of Finance, 36 (1), 51-60. Cebenoyan S., Cooperman ES., and Register CA. (1999). Ownership Structure, Charter Value, and Risk-taking Behavior of Thrifts. Financial Management. 26, 43-60. doi: 10.2307/3666116 Cornett, M. M. and H. Tehranian. (1992). Changes in Corporate Performance Associated with Bank Acquisitions. Journal of Financial Economics. 31, 211-234. Demerguc-Kunt A., L. Leaven and R. Levine (2003), The Impact of Bank Regulation, Concentration and Institution on Bank Margin, National Bureau of Economic Research. Working Paper 9890. Diamond and Dybvig (1983). Bank Runs, Deposit Insurance, and Liquidity. The Journal of Political Economy, Vol. 91, No. 3, pp. 401-419. Donaldson, G. (1961), Corporate Debt Capacity. Harvard University Press, Boston.

384

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

Durand, David. (1952), Cost of Debt and Equity Funds for Business: Trends and Problem Measurement. Conference on Research in Business finance. The National Bureau of Economic Research. p. 215-262. Furlong, F. T. and Keeley, M.C. (1989). Capital Regulation and Bank Risk-Taking: A Note. Journal of Banking and Finance 13 (6): 883–91. Goldstein, Ju and Leland H. (2001). An EBIT-Based Model of Dynamic Capital Structure. Journal of Business, vol. 74, no. 4, pp.483-512. Gujarati, D. N. (2003). Basic econometrics. New York, USA, McGrawHill. Hovikimian, A., and E. Kane. (2000). Effectiveness of Capital Regulation at U.S. Commercial Banks, 1985 to 1994. Journal of Finance 55: 451-469. Hovikimian, A., E. Kane, and L. Laeven. (2003). How Country and Safety-Net Characteristics Affect Bank Risk-Shifting. Journal of Financial Services Research 23: 177-204. Jensen, Michael C. and William H. Meckling. 1976. Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics 3: 305-360. Kane, E.J. (2000). Incentives for banking megamergers: What motives might regulators infer from event study evidence?. Journal of Money, Credit and Banking 32, 671-701. Kahane, Y. (1977). Capital Adequacy and the Regulation of Financial Intermediaries. Journal of Banking and Finance 1, 207-218. Kim, D., and A. M. Santomero. (1988). Risk in Banking and Capital Regulation. Journal of Finance 43 (5): 1219–33 Koch TW and Mac Donald SS. (2000), Bank Management. Fourth Edition. Orlando. The Dryden Press. Harcourt Brace College Publishers. Koehn, M. and A. M. Santomero (1980). Regulation of Bank Capital and Portfolio Risk, Journal of Finance 35, 1235—1244. Manurung, Adler H. (2011), Metode Penelitian: Keuangan, Investasi, dan Akuntansi Empiris. PT. Adler Manurung Press. Mishkin, F.S. (2006). How Big A Problem Is Too Big To Fail?. Journal of Economic Literature 44, 988-1004. Modigliani, F, and Miller, M.H. (1958). The Cost of Capital, Corporation Finance, and the Theory of Investment. American Economic Review Vol.XLVIII, No.3, (June 1958), pp.26197, reprinted in The Theory of Business Finance (2e), Stephen H. Archer and Charles A. D’Ambrosio, eds., New York, Macmillan, 1976 (page references to reprint). Modigliani, Franco and Merton Miller (1963). Corporate Income Taxes and the Cost of Capital: A Correction. American Economic Review, vol.53, pp. 433-443

Determinant Of Capital Ratio: A Panel Data Analysis On State-Owned Banks In Indonesia

385

Myers, S. C. (1977). Determinants of Corporate Borrowing. Journal of Financial Economics, 5, 147–175. Myers, S.C. and Majluf, N.S. (1984). Corporate Financing and Investment Decisions When Firm Have Information That Investors Do Not Have. Journal of Financial Economics 13, pp. 187-221. Nasser, Etty M. (2003). Perbandingan Kinerja Bank Pemerintah dan Bank Swasta Dengan Rasio CAMEL Serta Pengaruhnya Terhadap Harga Saham. Media Riset Akuntansi, Auditing dan Informasi, Vol. 3, No. 3. Nachrowi ND, Usman H. (2006). Pendekatan Populer dan Praktis Ekonometrika Untuk Analisis Ekonomi dan Keuangan. Lembaga Penerbit Fakultas Ekonomi Universitas Indonesia. Rime, B. (2001). Capital Requirements and Bank Behaviour: Empirical evidence for Switzerland. Journal of Banking and Finance 25: 789-805. Saunders A., and Wilson B. (2001). An Analysis of Bank Charter Value and Its Riskconstraining Incentives. Journal of Financial Sevices Research, 19, 185-195. doi: 10.1023/ A:1011163522271 Solomon, Ezra. (1963). The Theory of Financial Management. New York: Columbia University Press. Stiglitz, J.E. (1969). A Re-Examination of the Modigliani-Miller Theorem. American Economic Review, 59:5, pp. 784-793. Tumiwa RAF, Sudarma M, Salim U, and Djumahir. 2013. Banking Regulation, Risk Management, and Capital Structure Decisions: A Study on Rural Banks in Indonesia. Research Journal of Finance and Accounting, 4(15): 27-35. Wooldridge, J.M. (2010). Econometric Analysis of Cross Section and Panel Data. 2nd Edition. MIT Press. Yu, H. (1995). The Determinants of Interest Rate Margins: Empirical Evidence on the Canadian Banking Industry. International Journal of Finance, 7: 33-45 Yudhistira, D. (2003). The impact of Bank Capital Requirement in Indonesia. Available: http://129.3.20.41/eps/fin/papers/0212/02122002.pdf Zweibel, J. (1996). Dynamic Capital Structure under Managerial Entrenchment. American Economic Review 86 (5), 1197–1215.

386

Bulletin of Monetary, Economics and Banking, Volume 16, Number 4, April 2014

Halaman ini sengaja dikosongkan

387

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:

388

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.

Suggest Documents