BULLETIN OF MONETARY ECONOMICS AND BANKING · The Bulletin of Monetary Economics and Banking ... Tumpak Silalahi, ... The author thanks to Dr. Noer Azzam,

1ANALISIS TRIWULANAN: Perkembangan Moneter, Perbankan dan Sistem Pembayaran, Triwulan II - 2007

BULLETIN OF MONETARY ECONOMICS AND BANKING

Directorate of Economic Research and Monetary PolicyBank Indonesia

PatronPatronPatronPatronPatronBoard of Governor Bank Indonesia

Editorial BoardEditorial BoardEditorial BoardEditorial BoardEditorial BoardProf. Dr. Anwar Nasution

Prof. Dr. Miranda S. GoeltomProf. Dr. Insukindro

Prof. Dr. Iwan Jaya AzisProf. Iftekhar HasanDr. M. Syamsuddin

Dr. Perry WarjiyoDr. Halim Alamsyah

Dr. Iskandar SimorangkirDr. Solikin M. JuhroDr. Haris Munandar

Dr. Andi M. Alfian Parewangi

Editorial ChairmanEditorial ChairmanEditorial ChairmanEditorial ChairmanEditorial ChairmanDr. Perry Warjiyo

Dr. Iskandar Simorangkir

Executive DirectorExecutive DirectorExecutive DirectorExecutive DirectorExecutive DirectorDr. Andi M. Alfian Parewangi

SecretariatSecretariatSecretariatSecretariatSecretariatToto Zurianto, MBA

MS. Artiningsih, MBA

The Bulletin of Monetary Economics and Banking (BEMP) is a quarterly accreditedjournal published by Directorate of Economic Research and Monetary Policy-BankIndonesia. The views expressed in this publication are those of the author(s) anddo not necessarily reflect those of Bank Indonesia.

We invite academician and practitioners to write on this journal. Please submityour paper and send it via mail to: [email protected]. See the writing guideon the back of this book.

This journal is published quarterly; January √ April √ August √ October. The digitalversions including all back issues are available online; please visit our stable link:√http://www.bi.go.id/web/id/Publikasi/Jurnal+Ekonomi/. If you are interested to sub-scribe for printed version, please contact our distribution department: Publicationand Administration Section √ Directorate of Economy and Monetary Statistics,Bank Indonesia, Building Sjafruddin Prawiranegara, 2nd Floor - Jl. M. H. ThamrinNo.2 Central Jakarta, Indonesia, Ph. +62-21-3818202, Fax. +62-21-3802283,Email: [email protected].

BULLETIN OF MONETARY ECONOMICSAND BANKING

Volume 14, Nomor 2, October 2011

QUARTERLY ANALYSIS: The Progress of Monetary, Banking and Payment System,

Quarter III - 2011

Author Team of Quarterly Report, Bank Indonesia

Does Financial Development Absorb or Amplify the Shock?

Meily Ika Permata, Ibrahim, Hidayah Dhini Ari

Bank Capital Inflows, Institutional Development and Risk: Evidence from Publicly -

Traded Banks in Asia

Wahyoe Soedarmono

Competitive Conditions in Banking Industry: An empirical Analysis of the Consolidation,

Competition and Concentration in the Indonesia Banking Industry between 2001

and 2009

Tri Mulyaningsih, Anne Daly

Relative Effectiveness of Indonesian Policy Choices During Financial Crisis

Tumpak Silalahi, Tevy Chawwa

107

141

103

127

177

103QUARTERLY ANALYSIS: The Progress of Monetary, Banking and Payment System, Quarter III, 2011

QUARTERLY ANALYSIS:The Progress of Monetary, Banking and Payment System

Quarter III - 2011

Author Team of Quarterly Report, Bank Indonesia

The Board of Governor Meeting of Bank Indonesia on October 11, 2011 decided to

lower the BI rate by 25 bps to the level of 6.5%. Bank Indonesia will also maintain the

stabilization of Rupiah particularly from the impact of global financial market shock. The decision

is in line with the inflation expectation of below 5% on current and next year. Furthermore,

these policies are meant to anticipate and to mitigate the negative impact of the global economic

and financial slowdown on Indonesian economy. Looking ahead, the Board of Governor will

continue to evaluate the global economic and financial performance and use the interest rate

as well as the mix of monetary and the other micro prudential policies to mitigate the possible

slowing down of Indonesian economic performance, especially on achieving the inflation target

of 5% + 1% in 2011 and 4.5% + 1% in 2012.

The Board of Governor put a high concern on the high risk and uncertainties in global

financial market and the trend of global economic slowdown due to the debt and fiscal problem

in Europe and US. Main attention is addressed to the short term impact through the financial

channel in the form of the stock market weakening, the increase of debt risk indicator, and

the capital reversal pressure by global investor from emerging countries, including Indonesia.

Meanwhile, the global economic performances also weaken as indicated on the slowing

production activities and retail sale, accompanied by the decrease of consumer»s confident in

developed countries and also numbers of correction on international commodity prices. On

the other side, the inflation pressure has decreased, although the inflation of emerging countries

remains high hence shifts the monetary policy response to neutral or accommodative.

Looking ahead, on general the Board of Governor foresee a slower output growth in

developed countries, the lower international trade volume, and lower global commodity prices.

Meanwhile, the large excess liquidity and risk perception of investor will still drive high capital

inflow into emerging countries, including Indonesia, both in foreign direct investment (FDI) or

in portfolio investment.

104 Bulletin of Monetary Economics and Banking, October 2011

The Board of Governor considers the fundamental of national economic and banking in

Indonesia are still strong among higher concern about the world economic prospect. The

output growth on 2011Q4 is predicted to be higher, especially driven by consumption and

investment that the output will reach 6.6% in 2011. So far, the impact of global economic

shock is limited to the financial sector, while the real sector is relatively not affected. However,

the slowing global economic is predicted to influence the domestic economic performance in

2012, both on financial market and international trade.

The domestic economic growth in 2012 is predicted to be in the range of 6.2% - 6.7%.

The source of this growth is driven by strong consumption and increasing investment.

OnSectorallevel, all sectors are predicted to have positive growth. The main growth contributors

are Manufacture, Trade Hotel and Restaurant, and Transport and Communication.

The Balance of Payment (BOP) in 2011Q4 is predicted to gain surplus following the high

pressure due to the capital outflow in previous quarter. In aggregate for 2011, BOP is predicted

to have a large surplus. The main sources of this surplus are the increasing surplus of capital

and financial transaction, both in the form of portfolio and direct investment. In line with this

the reserve on the end of September 2011 was USD114.5 milliard or equivalent to 6.5 months

of import and government debt service payment. This reserve is more than enough to support

the stabilization of Rupiah.

The exchange rate of Rupiah in 2011Q3 experience a pressure, especially on September

2011 where the Rupiah depreciate by 2.42% (ptp) to the level of Rp8,790 per USD with

increasing volatilities. However, this depreciation is in line with other currencies on regional

peer»s member. Among several factors, the pressure on Rupiah is caused by the increase of

global risk due to the increasing concern on the global economic prospect. In addition, the

increase of foreign exchange demand to pay import also contributed to the pressure of Rupiah.

Looking ahead, Bank Indonesia will continue to keep the stability of Rupiah»s exchange rate to

support the macroeconomic stability.

The inflation pressure continues to decrease. CPI inflation on 2011Q3 was 1.89% (qtq)

or 4.61% (yoy), lower than the previous year. The lower inflation pressure came from the

volatile food and administered price, along with the supply improvement, the lower international

food prices and the minimum government intervention concerning strategic commodities.

Meanwhile, the core inflation pressure excluding the hike of gold price is relatively well managed,

because of the appreciation of Rupiah on previous period and the sufficient supply. With

these facts, in 2011 the inflation will be lower than 5%.

105QUARTERLY ANALYSIS: The Progress of Monetary, Banking and Payment System, Quarter III, 2011

In 2012, the inflation will be under control and is expected to be lower than 5%, along

with the correction on international commodity prices and the weakening of world economy.

The stability of banking system will be also maintained with better intermediary function,

despite of the global financial shock. Banking industry stability are also well maintained as

reflected on the Capital Adequacy Ratio (CAR) which is far above the requirement 8% and

Non-Performing Loan which is lower than 5% gross. In addition, the distribution of credit

continues to finance the real economic activities where its growth reached 23.8% (yoy) until

the end of September 2011. Bank Indonesia maintains the banking system stability and

encourages its intermediary function while still keeping the prudent banking practice by directing

the credit allocation for productive sectors in order to optimize the economic growth among

the uncertain of global economic.

The reliability and efficiency of payment system also contributes on macroeconomic

performance in Indonesia. The support of the payment system is indicated on the availability

of the Bank Indonesia Real Time Gross Settlement (BIRTGS), Bank Indonesia-Scrip less Security

Settlement System (BI-SSSS), and National Clearing System that reach 100%. Besides, the

reliability of card payment system and electronic money, arranged and processedoutside Bank

Indonesia are also well maintained. Bank Indonesia can also supply the fiat money even though

there wassignificant increase of money demand during the Holy Month of Ramadhan and

«Idul Fithri.


This page is intentionally left blank

107Does Financial Development Absorb or Amplify the Shock?

DOES FINANCIAL DEVELOPMENTABSORB OR AMPLIFY THE SHOCK?

Meily Ika PermataIbrahim

Hidayah Dhini Ari 1

This paper analyzes the role of financial development on economic output in Indonesia. Using

vector autoregressive method, the results confirm the positive impact of financial development on output

growth. The interaction between the financial development and the shock either in financial or real sector

shows that the financial development has a positive role to dampen the negative impact of the shock on

the output growth, while strengthen the positive one. Another variable on the model, which significantly

affect the output growth are excess money, term of trade, and the price. Compare to these variables, the

marginal effect of financial development on output is smaller.

Keywords : Financial development, shock, output volatility, VAR.

JEL Classification : E44, O16

1 Meily Ika Pertama ([email protected]), Ibrahim ([email protected]) and Hidayah Dhini Ari ([email protected]) are researchers in EconomicResearch Bureau, Directorate of Economic and Monetary Policy Research. The author thanks to Dr. Noer Azzam, Dr. Perry Warjiyo, Dr.Iskandar Simorangkir and other colleagues in DKM for the excellent input and suggestion.

Abstract


1. INTRODUCTION

Macroeconomics stability is a necessary condition for sustainable economic development.

Therefore, efforts to understand the sources of macroeconomic instability condition is one of

the challanges of the important attention in economics.

Role of the financial sector is associated with economic growth that has long been a

subject research, especially after Schumpeter (1912)2. Since then, continues studies have been

conducted to examine the interaction between financial development and output growth.

Based on review of various conducted studies , the relationship between financial development

and output growth tend to be inconclusive.

Some views believe that the financial development will drive the growth (initiated by

Schumpeter, 1911 and Gurley and Shaw, 1995) because financial sector can overcome the

problem of financial constraint, contribute to the allocation of resources more efficiently, channel

funding and other activities related to risk sharing and financial innovation, be the medium for

the monetary policy transmission (Cecchetti and Krause, 2001 and Krause and Rioja, 2006), and

can overcome the problem of imperfect capital market) (Bernanke and Gertler (1989), Greenwald

and Stiglitz (2003) and Kiyotaki and Moore (1997). Financial development is also believed to

have positive role in reducing the volatility of macroeconomic variables,(Dynan, Elmendorf and

Sichel (2006), Denizer, Iyigun, and Owen (2002), Harvey and Lundblad (2006), Aghion et. al

(2005), Aghion et. al (2009), Cecchetti, Lagunes and Krause (2006), and Mendicino (2007)).

However, Bacchetta and Caminal (2000), Easterly et al. (2002) and Kuneida (2008), show

that the financial constraint can be a factor that restrain or even exacerbate the shock impact

that occured in the economy. Meanwhile, Lopez and Spiegel (2002), denizer et al. (2002), Silva

(2002), and Tharavanji (2007) actually found negative relationship between financial development

and output volatility. Other researchers such as Tiryaki (2003), Beck (2006) & Guryal et al.

(2007) did not find significant relationship between financial development and growth volatility.

The output volatility shown on the figure, suggest a need to conduct empirical testing

about the role of financial sector in macroeconomic stability, especially GDP growth in Indonesia.

As one of the developing countries with a growing financial sector, the study about the role of

financial sector in the context of macroeconomic stability is expected to assist the formulation

of a more appropriate monetary policy. When compared to some developed countries, the

output volatility in emerging countries is generally higher, except the period of global crisis.

Among emerging countries, Indonesia»s output volatility post-crisis in 1997 looks relatively smaller.

2 Schumpeter, J.A., 1991. The Theory of Economic Development. Cambridge, Mass: Harvard University Press


Figure 1. Output Volatilityin Developing and Emerging Countries

Explicitly, this paper identifies the relationship between the role of financial sector and

output volatility in Indonesia. To measure the financial development we use the ratio of credit

to private sector towards GDP. This measure commonly used in previous studies. The conclusive

relationship between financial developmentsto macroeconomic stability is expected to be a

valuable input in formulating better monetary policy.

The second part of this paper reviews the theory and literature, the third part review the

methodology and empirical models are used, while the results and analysis are described in the

fourth. Conclusions, implication and policy recommendation will close the presentation.

II. THEORY

The view that financial development will drive the growth (started by Schumpeter, 1911

and Gurley and Shaw, 1955) is based on the view that the financial sector can overcome

financial constraint. In addition, the existence of financial sector contributes to the allocation of

resources, both financial and non-financialmore efficiently. Financial deepening and development

of can reduce volatility in economy, through its ability to channel financing and other activities

related to risk sharing and financial innovation.

Financial sector contributes in stabilizing macroeconomic condition in general through

its ability to be the medium for the monetary policy transmission (Cecchetti and Krause, 2001

and Krause and Rioja, 2006). Bernanke and Gertler (1989), Greenwald and Stiglitz (1993) and

0

1

2

3

4

5

6

7

8

1994

Q4

1995

Q3

1996

Q2

1997

Q1

1997

Q4

1998

Q3

1999

Q2

2000

Q1

2000

Q4

2001

Q3

2002

Q2

2003

Q1

2003

Q4

2004

Q3

2005

Q2

2006

Q1

2006

Q4

2007

Q3

2008

Q2

2009

Q1

2009

Q4

China Indonesia BrazilThailand EU USA


Kiyotaki and Moore (1997) carry out a study and confirm the view above, by showing that

financial sector contribute to overcome the problem of imperfect capital market, and there by

reduce output volatility in economy.

Dynan, Elmendorf and Sichel (2006) and Denizer, Iyigun, and Owen (2002) and Beakaert,

Harvey and Lundblad (2006) present the results of study showing that financial development

have a positive role in reducing the volatility of macroeconomic variables. Aghion et al. (2005)

and Aghion et al. (2009) confirm that credit constraint contribute to enlarge the shock impact

in economy through the choice of type investment made by businesses. Cecchetti, Lagunes

and Krause (2006) and Mendicino (2007) prove that the consumption credit have a positive

role in overcoming liquidity constraint at the household level so that it can reduce the volatility

of economic growth.

However, Bacchetta and Caminal (2000) show that financial constraint can be a factor

that restrain or even exacerbate the shock impact that occured in the economy, depending on

the type of shock that occured in the economy. In line with Bacchetta and Caminal, Easterly et

al. (2002) and Kuneida (2008) indicate the characteristic of relationship between financial

development and economic growth tends to be nonlinear. In this case, financial development

will reduce macroeconomic volatility to a certain point, beyond that point, more credit to private

sector will increase the volatility.

Meanwhile, Lopez and Spiegel (2002), Denizer et al. (2002), Silva (2002), and Tharavanji

(2007) actually found a negative relationship between financial development and the volatility

of economic growth. More interestingly, Tiryaki (2003), Beck (2006) and Guryay et al. (2007)

did not find significant relationship between financial development and volatility growth.

Based on several conclusions from various research above, the relationship between

financial development and output volatility is still ambiguous or inconclusive. That characteristic

of relationship depends on the source of the shock on economy, whether it comes from the

real or monetary sector, and how an economy responds to the shocks .

One simple model about financial development was built by Bacchetta and Caminal

(2000), and then developed and modified by Beck (2006). The model assumethe economy

consists of consumers and entrepreneurs. Every entrepreneur has the same access to production

technology, represented by f (k), with f (0) = 0, f ‘(k) > 0, and f “(k) < 0. Although each

entrepreneur has same access to production technology, entrepreneur basically has a different

level of wealth b. As much as β ratio of the entrepreneur is entrepreneur with high wealth. And

as much as (1 - β) ratio of the entrepreneurs are those with low wealth. High group is assumed

to be able to meet the financing needs of its investment and has excess funds in the bank and


Meanwhile, Low has limited financing so that had to borrow funds with interest rate rL.

The existence of assymetric information and moral hazard potentially cause the Low group to

bear an agency cost ϕ. With this condition, Low profit maximization is achieved at:

(1)

(2)

(3)

get the interest rate rD. In addition, High is assumed to have no financing constraint in financial

market, and thereby High profit maximization is achieved at :

Thus, the relative marginal productivity between High and Low is :

The greater the difference between rL and rD and ϕ, the higher ratio kH / kL and the

greater the difference of marginal productivity between Low and High. The higher ϕ, the greater

effect of real location funds between High and Low. This real location funds will influence

aggregate productivity and ultimately economy»s output as a whole.

The financial intermediaries, in this case is bank, is assumed to work in a perfect competition

condition, without cost, and only holds the asset in the form of credit. However, the bank

deposit is subject to reserve requirement of the monetary authority, that is τ. Thus, the credit

can be allocated by bank to Low (1 − τ ) is multiplied by the deposits of High. Thus, the increase

of τ will reduce funds available for loan forthe Low, and vice versa..... Total aggregate of loans

from bank is:

(4)

where bH is internal funds of High and kH is the desired level of investment of group High.

While kL and bL, respectively, is the level of investment desired by Low and their internal funds.

Low internal funds areassumed to be very small so that it cannot meet the level of their desired

investment.

Furthermore, assuming there is no possibility of default, then the ratio of rL and rD is only

influenced by the size of reserve requirement, τ , hence:


With the existence of assymetric information that generates agency cost, then Low

naturally will always face sub-optimal investment condition. This means the realized level of

investment will always below the desired level. In this case, agency cost stated as:

(5)

(6)

Where ω is a function of several technological parameters, which are exogenous. This

means that agency cost is a negative linear function of ratio between Low»s internal funds

owned and their desired level of investment. The greater internal funds owned, the smaller the

agency cost charged by bank to Low. Furthermore, in this case, is:

Equation (6) shows that Low will be exposed to credit-constrained condition, so that the

level of investment that might be achieved will lower than the desired one. Low»s level of

investment will decrease with the increaseof rL, ω, and the leverage ratio .

Based on the equations above, then the marginal productivity of High and Low can be

elaborated in following equation:

(7)

Thus, the market clearing condition in financial market is:

(8)

Based on equation (7) and (8), we can seethat the relative investment kL / kH will increase

along with the increase of ratio bL / kL and the relative ratio of internal funds bL / bH and

declining agency cost ω, and reserve requirement τ.


xt = A

0 + A

1 x

t-1 + A

2 x

t-2 + .... + A

p x

t-p + e

t

Based on the equations above, we can obtain some interpretation below:

1) The impact of shocks that occur in the real sector will be largerin the presence of assymetric

information condition on capital market. The impact was even greater in line with the

increase in agency cost ω. This is shown in equation (7), where the investments of the High

will be greater than the Low, so that causing the Low marginal productivity is higher than in

High. The difference of marginal productivity will be greater with higher agency cost; a

condition found in the capital market conditions that characterized assymetric information.

With in creasing magnitude of the difference of marginal productivity, then the impact of

shocks to the economy»s output will be greater.

2) The impact of shocks coming from monetary policy affects the supply of loan able funds will

have smaller impact on the conditions where there is assymetric information on the capital

market. Relaxing the monetary policy in the form of reduction of level reserve requirement,

would increase the amount of loan able loan that will eventually degrade. Never the less,

the reduction in interest rates will increase the leverage and eventually agency cost for the

Low. This in turn will reduce some of the positive impact of reserve requirement reduction.

Mean while, the decline in agency cost as a result of the growing of financial development

will further streng then the positive impact of reduction in reserve requirements to the

output.

These should be the the base of our hypothes is to be tested on this paper.

III. METHODOLOGY

This study will use an econometric approach to test the testable hypothesis by utilizing

the method of vector autoregression (VAR). We use this VAR method to test the case of Indonesia.

VAR model is a linear function analysis of past data movement of a set of variables

(endogenous variable) in the same period (t=1,...,T). A VAR model of order (p) can be represented

in the following equation:

Which et is a vector from error term that meet the standard condition; E (e

t) = 0, and E (e

t e’

t)

= Ω.

VAR model is often used in macroeconomic analysis. However, although the VAR

approach has the advantage in modeling dynamic behavior of economic variable and


forecasting, many criticism directed against VAR approach which tends to be a-theory, where

the lack of restriction in the VAR model lag structure is associated with the lack of a structure

underlying the relationship among variables in the VAR system.This would create difficulties

in interpreting the results.

The theoretical model described previously has provided a hypothesisto be tested. But

the test to perform is not solely based on a reduced form derived explicitly from the model

above, but rather we use empirical ad hoc model in order to accommodate some other control

variables that are not captured explicitly in theoretical model.

Some of the variables to be included in the model are real GDP growth and price

movements, in order to illustrate the volatility of macroeconomic variables. Also included are

variables in real and monetary sectors. For the real sector, we can use terms of trade variable for

instance, while for monetary sector we may use excess money. Financial development ( FD ) is

measured with the credit ratio to GDP. Meanwhile, to capture a market imperfection/assymetric

information, we can usea variable that can explain the existence of risks such as interest spread

between lending and SBI rate.

The data used in this study are quarterly data, covering first quarter period of 1997 to

second quarter of 2010. The variables involved are, (i) Real GDP, representing the output,

(ii) Consumer Price Index, (iii) Term of trade, representing the shock from real sector, (iv)

Excess money, represents shock from monetary sector, (v) Ratio of credit to GDP, measuring

the financial development, (vi) Interest rate spread between lending rate and the SBI rate,

measuring the magnitude of risk due to the existence of market imperfection/assymetric

information.

IV. RESULTS AND ANALYSIS

In general, the macroeconomic indicator in Indonesia at the beginning of the observation

period had a relatively high volatility movement. During the period ofcrisis in 1998, the economic

growth contracted by more than 13.1% and inflation jumped to 69.8% followed by a variety

of other indicators such as the ratio of credit to GDP, the excess money, and the risk indicators

represented by the spread of lending rates to SBI (Figure 5). These volatilities were getting

smaller along with the improvement of monetary management.

By looking at the more detail movement on standard deviation of the macroeconomic

and the financial indicators, we can obtain a clearer picture. In general, the volatility of

GDP with other variables has same direction. The correlation of growth volatility with the


average of credit to GDP ratio reached 81% and the correlation between growth and

inflation volatility by 78% (Figure 6). Volatilityin the figure is calculated by using the three

years moving standard deviation for each indicator, then we calculate the pair correlations.

Based on the initial hypothesis on the co-movement among these variables, we proceed to

use the VAR method.

Prior the VAR estimates we run the stationary tests on each variable by using the

Augmented Dickey Fuller (ADF) unit root tests. The test results show that GDP and IHK variables

are non-stationary in levels (Table 1). Based on these unit root tests, the variables selected for

inclusion in the VAR model is dLnPDB, dLnIHK, dLnTOT, dLnFD, risk growth and excess money

growth.

Figure 2.The Main Indicators Development

Figure 3.The Financial Development (dln FD)

Figure 4.Excess Money

Figure 5.Spread Credit Interest Rate-SBI

-20

0

20

40

60

80

PDB

Inflasi

1997Mar

1998Mar

1999Mar

2000Mar

2001Mar

2002Mar

2003Mar

2004Mar

2005Mar

2006Mar

2007Mar

2008Mar

2009Mar

2010Mar

1997Mar

1998Mar

1999Mar

2000Mar

2001Mar

2002Mar

2003Mar

2004Mar

2005Mar

2006Mar

2007Mar

2008Mar

2009Mar

2010Mar

0.4

0.2

0.0

-0.2

-0.4

-0.6

-0.8

-0.10

-0.12

1997Mar

1998Mei

1999 2000 2001Jul

2003Sep

2004Mar

2005Mei

2006Jul

2007Sep

2008Nov

2010Jan

40

30

20

10

0

-10

-20

-30Nov Jan

1997Mar

1998Mar

1999Mar

2000Mar

2001Mar

2002Mar

2003Mar

2004Mar

2005Mar

2006Mar

2007Mar

2008Mar

2009Mar

2010Mar

20

10

0

-10

-20

-30

-40

-50


Optimum lag order for VAR procedure showed mixed results. Based on the Schwartz

Information Criteria, the optimal lag is 1, while the Akaike Information Criteria and Hannan

Quinn Information Criteria generate optimal lag is 6. However, the lag order 6 is not selected,

since the number of observation is only 54. This study also did not follow the Schwart optimum

lag with lag order 1, but sets the use of lag order 2 to better capture the variables dynamics.

Based on lag structure test, it is known that the estimated VAR with lag order 2 is stable with all

roots smaller than1 and is in the unit circle (Figure 7).

Figure 6.The Connection Development of Main Indicator

0.00

0.50

1.00

1.50

2.00

2.50

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

1990Mar

1990May

1992Jul

1993Sep

1994Nov Mar May Jul Sep NovJan

1996 1997 1998 1999 2000 2001Mar May Jul Sep NovJan

2003 2004 2005 2006 2007 2008Jan

2010

Average of Credit/GDP (3 years)

Rolling StDev of GDP Growth(3 years -yoy) (RHS)

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00Rolling StDev of Inflation(3 years - yoy)

Rolling StDev of GDP Growth(3 years -yoy) (RHS)

1990 1991 1992 1994 1995 1997 1999 2000 2001 2002 2004 2005 2006 2007 2009 20101996Apr Jul Oct Jan AprJanApr Jul OctJanApr Jul OctJanJun SepMar

Table 1.The Test Result of Unit Root Test

Variable Level (P-value)

GDP 1.0000dLnPDB 0.0967GDP Growth 0.0943CPI 0.8050dLnIHK 0.0006Inflation 0.3043TOT 0.0258dLnTOT 0.0020TOT Growth 0.0057FD 0.0163dLnFD 0.0024FD Growth 0.2049Risk 0.0000Growth Risk 0.0000Excess Money Growth 0.0002


The results of Granger Causality/block exogeneity test showed that the simultaneous

movement of inflation, TOT, excess money, financial development and the risk are the

explanatory variable of the GDP movement (Table 2). However, individually the risk can not

explain the GDP movement, but we keep it on the model since the risk is the control variables

that describe the magnitude of risk due to the existence of market imperfection/assymetric

information.

Figure 7.AR Roots Graph

Table 2.VAR Granger Causality/Block Exogeneity Wald Tests

ExcludedExcludedExcludedExcludedExcluded Chi-sqChi-sqChi-sqChi-sqChi-sq dfdfdfdfdf ProbabilityProbabilityProbabilityProbabilityProbability

DLNIHK 17.48 2 0.000DLNTOT 11.80 2 0.003EXCMON 8.52 2 0.014DLNFD 9.53 2 0.009GRISK 1.09 2 0.580AllAllAllAllAll 56.7356.7356.7356.7356.73 1010101010 0.0000.0000.0000.0000.000

Inverse Roots of AR Characteristic Polynomial

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

4.1. Impact to GDP Growth

The results ofimpulse response shows that financial development have a positive impact

in improving GDP growth and significant in quarter 2 (Figure 8). Meanwhile, an increase of

inflation will reduce GDP growth significantly during quarter 2 to 5. The increase of TOT will


2 4 6 8 10 12 14 16 18 20 22 24

Accumulated Response of DLNPDB to NonfactorizedOne Unit EXCMON Innovation

-.024

-.020

-.016

-.012

-.008

-.004

.000

.004

2 4 6 8 10 12 14 16 18 20 22 24

Response of DLNPDB to NonfactorizedOne Unit EXCMON Innovation

-.004

-.003

-.002

-.001

.000

.001

.002

2 4 6 8 10 12 14 16 18 20 22 24

Accumulated Response of DLNPDB to NonfactorizedOne Unit DLNTOT Innovation

-1,2

-0,8

-0,4

0,0

0,4

2 4 6 8 10 12 14 16 18 20 22 24-.5

-.4

-.3

-.2

-.1

.0

.1

.2

.3

Response of DLNPDB to NonfactorizedOne Unit DLNTOT Innovation

2 4 6 8 10 12 14 16 18 20 22 24

Accumulated Response of DLNPDB to NonfactorizedOne Unit DLNIHK Innovation

-3

-2

-1

0

1

Response of DLNPDB to NonfactorizedOne Unit DLNIHK Innovation

-.8

-.6

-.4

-.2

.0

.2

.4

.6

2 4 6 8 10 12 14 16 18 20 22 24

also significantly reduce the GDP during these cond to third quarter. In addition, the increase

of excess money will also reduce GDP significantly in quarter 2. However, the impact of

increased risk due to market imperfection/assymetric information to GDP tends to be not

significant.

Figure 8. The Shock Impact of Inflation, TOT development, Excess Money Growth,Financial Development and Risk Changes against GDP Growth


2 4 6 8 10 12 14 16 18 20 22 24-.00002

-.00001

.00000

.00001

.00002

.00003

Accumulated Response of DLNPDB to NonfactorizedOne Unit GRISK Innovation

2 4 6 8 10 12 14 16 18 20 22 24-.000008

-.000006

-.000004

-.000002

.000000

.000002

.000004

.000006

Response of DLNPDB to NonfactorizedOne Unit GRISK Innovation

2 4 6 8 10 12 14 16 18 20 22 24

Accumulated Response of DLNPDB to NonfactorizedOne Unit DLNFD Innovation

-.4

-.2

.0

.2

.4

.6

.8

2 4 6 8 10 12 14 16 18 20 22 24-.08

-.04

.00

.04

.08

.12

.16

Response of DLNPDB to NonfactorizedOne Unit DLNFD Innovation

Figure 8. The Shock Impact of Inflation, TOT development, Excess Money Growth,Financial Development and Risk Changes against GDP Growth (continued)

In aggregate, the rise of inflation, TOT and the excess money growth will reduce GDP

growth, while financial development will increase it. Cumulatively, an increase of financial

development of 1% will give additional increase in GDP of 2.4% in 2.5 years (Figure 9). In

contrast, a decrease of financial development of 1% will cause accumulation of GDP decline of

2.4% in 2.5 years.

Furthermore, using the variance decomposition analysis shows that excess money

growth, TOT development and price movement are variables that contribute most in

explaining the GDP movements in long-term, respectively by 27%, 11% and 9% (Table 3).

While financial development contributes only about 2.6% as well the risk factor only

contribute less than 1%.


4.2. Impact to Inflation

Financial development and the increase of risk due to market imperfection/assymetric

information do not influence inflation significantly. A significant factor in influencing inflation

is the increase in TOT, which will raise inflation for 5 quarters, from quarter 2 to quarter 6. In

addition, the increase in excess money will also reduce GDP significantly for 11quarters, from

quarter 2 to quarter 12. The results of impulse response showing that increase in TOT and

excess money growth accumulatively will increase inflation.

Figure 9. The Impact ofFinancial Development Against GDP Growth

Cumulative Increase in GDP %0.30

0.20

0.10

0.00

-0.10

-0.20

-0.30

Shock Increased 1% Financial Development

Shock Decline 1% Financial Development

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

1 0,026 100,000 0,00 0,000 0,000 0,000 0,000 2 0,033 81,750 5,535 11,579 0,623 0,438 0,075 3 0,040 70,384 8,575 13,504 6,495 0,374 0,669 4 0,044 62,215 9,940 12,844 13,955 0,423 0,622 5 0,046 56,466 10,170 11,625 20,609 0,574 0,555 6 0,048 52,796 9,954 10,869 24,969 0,876 0,536 7 0,049 50,784 9,659 10,685 27,004 1,308 0,560 8 0,049 49,859 9,440 10,822 27,481 1,800 0,598 9 0,050 49,475 9,314 10,998 27,326 2,262 0,624 10 0,050 49,281 9,249 11,078 27,129 2,629 0,633

Period S.E. DLNPDB DLNIHK DLNTOT EXCMON DLNFD GRISK

Table 3.Variance Decomposition of DLNPDB


2 4 6 8 10 12 14 16 18 20 22 24

Response of DLNIHK to NonfactorizedOne Unit EXCMON Innovation

-.008

-.004

.000

.004

.008

.012

2 4 6 8 10 12 14 16 18 20 22 24-.01

.00

.01

.02

.03

.04

.05

.06

Accumulated Response of DLNIHK to NonfactorizedOne Unit EXCMON Innovation

Cumulated Response of DLNIHK to NonfactorizedOne Unit DLNTOT Innovation

2 4 6 8 10 12 14 16 18 20 22 24-2

-1

0

1

2

3

2 4 6 8 10 12 14 16 18 20 22 24

Response of DLNIHK to NonfactorizedOne Unit DLNTOT Innovation

-0.8

-0.4

0.0

0.4

0.8

1.2

2 4 6 8 10 12 14 16 18 20 22 24-8

-4

0

4

8

12

Accumulated Response of DLNIHK to NonfactorizedOne Unit DLNPDB Innovation

-2

-1

0

1

2

3

2 4 6 8 10 12 14 16 18 20 22 24

Response of DLNIHK to NonfactorizedOne Unit DLNPDB Innovation

Figure 10.The Impact of Changes in GDP Growth, TOT, Excess Money Growth,

Financial Development and Changes in Risk on Inflation


2 4 6 8 10 12 14 16 18 20 22 24-.3

-.2

-.1

.0

.1

.2

Response of DLNIHK to NonfactorizedOne Unit DLNFD Innovation

2 4 6 8 10 12 14 16 18 20 22 24-.00008

-.00006

-.00004

-.00002

.00000

.00002

.00004

.00006

Accumulated Response of DLNIHK to NonfactorizedOne Unit GRISK Innovation

2 4 6 8 10 12 14 16 18 20 22 24-.000012

-.000008

-.000004

.000000

.000004

.000008

.000012

.000016

Response of DLNIHK to NonfactorizedOne Unit GRISK Innovation

2 4 6 8 10 12 14 16 18 20 22 24-1.6

-1.2

-0.8

-0.4

0.0

0.4

0.8

1.2

Accumulated Response of DLNIHK to NonfactorizedOne Unit DLNFD Innovation

1 0,021 22,875 77,125 0,000 0,000 0,000 0,000 2 0,032 42,434 38,577 17,416 1,221 0,058 0,293 3 0,040 39,753 33,418 18,054 8,582 0,047 0,146 4 0,045 29,168 32,137 22,651 15,769 0,164 0,111 5 0,048 23,780 31,123 21,890 22,814 0,171 0,202 6 0,049 21,424 28,777 19,244 30,189 0,189 0,177 7 0,051 20,988 26,287 18,068 34,240 0,197 0,221 8 0,052 20,345 25,022 18,978 35,073 0,197 0,385 9 0,053 19,593 25,173 20,268 34,112 0,209 0,646 10 0,053 19,030 25,862 20,954 33,133 0,216 0,804

Period S.E. DLNPDB DLNIHK DLNTOT EXCMON DLNFD GRISK

Table 4.Variance Decomposition of DLNIHK

Figure 10.The Impact of Changes in GDP Growth, TOT, Excess Money Growth,Financial Development and Changes in Risk on Inflation (continued)


Based on the variance decomposition analysis for inflation, the increase of excess money,

TOT and GDP movement are variables that contribute most in explaining the movement of

inflation in the long run respectively by 33%, 21%and 19% (Table 4). While financial

development and risk factor only contribute less than 1%.

4.3 The Impact of Interaction between Financial Development with the Shockin Real Sector (TOT) and Monetary (Excess Money)

The result of impulse response that incorporates interaction between the shock in the

real sector and monetary with financial development shows that financial development has a

positive role in dampening the negative impact of the shock on GDP growth (Fig.11 and 12). In

contrast, financial development will help increasing (amplify) the positive impact of the shock

on economic growth.

Cumulatively, increase of TOT by 1% would reduce GDP by 0.4% within 4 years, but if at

the same time there was also an increase of financial development by 1%, the cumulative

impact of the decline in GDP in the period of 4 years will tend to be smaller, only amounting to

0.17%. In contrast, if there is a TOT decreased of 1%, then cumulatively will provide an additiona

lincrease in GDP in the period of 4 years of 0.4%. Interestingly, if the TOT decline by 1% is

accompanied by increased financial development by 1%, the cumulative impact of increase in

GDP during the period of 4 years will be higher, amounting to 0.64%.

Figure 11. Impacts of Shock Interaction betweenTOT Development and Financial Development on GDP Growth

Cumulative Increase in GDP %

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Increased 1% TOT shock is accompanied by increase of 1% Financial DevelopmentIncreased 1% TOT shockThe decrease of 1% TOT shock is accompanied by increase of 1% Financial DevelopmentShock decline 1% TOT


Cumulatively, the growth of 1% excess money would reduce GDP by 1% within 2 years,

but if at the same time there was an increase financial development of 1%, the cumulative

impact of the decline in GDPin the period of 2 years will be smaller, only amounting to 0.75%.

In contrast, if there is a decrease of 1% excess money, then cumulatively would raise GDP in

the period of 2 year sat 1%. However, if the decrease of 1% excess money is accompanied by

increased financial development by 1% then the cumulative impact of increase in GDP during

the period of two years will be higher, amounting to 1.25%.

Figure 12. Impact of Shock Interaction betweenExcess Money and Financial Development on GDP Growth

Cumulative Increase in GDP %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20-1.50

-1.00

-0.50

0.00

0.50

1.00

1.50

Shock Increased 1% Excess Money which is accompanied by an increase of 1% Financial DEvelopment

Shock Increased 1% Excess Money

Shock Decline 1% Excess Money which is accompanied by an increase of 1% Financial DEvelopment

Shock Decline 1% Excess Money

V. CONCLUSION

Financial development and economic growth hasa positive relationship, where the

increasing financial development will have a positive impact on economic growth. However,

the impact of risk increase due to market imperfection/assymetric information is not significant

to GDP.

The interaction between the shock in real and monetary sector with financial development

shows that financial development has a positive role dampening the negative shock impact on

GDP growth, while the positive shock impact on economic growth will be amplified.


Some factors that influence the movement of long-term growth is the increase of excess

money, TOT and the price movement. Meanwhile, although the financial development has a

positive role in the development of economic growth but its contribution is smaller compared

to these factors.

Related to the price movement, financial development and increased risk due to market

imperfection/assymetric information are not significant in influencing the inflation. This is in

line with the finding that financial development and risk factors do not contribute greatly in

explaining inflation dynamics in the long run.


REFERENCES

Bacchetta, Philippe & Ramon Caminal (2000), ≈Do capital market imperfections exacerbate

output fluctuations?∆. European Economic Review, No. 44, pp. 449-468.

Beck, Thorsten, Mattias Lundberg & Giovanni Majnoni (2006), ≈Financial Intermediary

Development and Growth Volatility : Do Intermediaries Dampen or Magnify Shocks?∆. Journal

of International Money and Finance, Volume 25, Issue 7, pp. 1146-1167.

Enders, Walter (2004), Applied Econometric Time Series, Wiley Series in Probability and Statistics.

John Wiley & Sons, Inc.

Greene, William H. (2008), Econometric Analysis. Prentice Hall.

Guryay, Erdal, okan Veli Safakli & Behiye Tuzel. (2007). ≈Financial Development and Economic

Growth: Evidence from Nothern Cyprus∆. International Research Journal of Finance and

Economics, Issue 8.

Kuneida, Takuma (2008). ≈Financial Development and Volatility of Growth Rates : New Evidence∆.

MPRA Paper No. 11341.

Schumpeter, J.A. (1911),The Theory of Economic Development, Cambridge, Mass. Harvard

University Press.

Stock, J.H. and M.W. Watson (2001), ≈Vector Autoregression∆.Journal of Economic Perspectives,

15, 4.

127Bank Capital Inflows, Institutional Development and Risk: Evidence from Publicly - Traded Banks in Asia

BANK CAPITAL INFLOWS,INSTITUTIONAL DEVELOPMENT AND RISK:

EVIDENCE FROM PUBLICLY - TRADED BANKS IN ASIA

Wahyoe Soedarmono1

This paper examines the relationship between bank capital inflows and financial stability. Using a

sample of publicly-traded commercial banks in Asia over the 2002-2008 period, our empirical results

show that higher banking inflows measured by the share of foreign liabilities in banking reduces systematic

risk, but increases bank-specific risk and total risk. A deeper investigation further suggests that an increase

in total risk and bank-specific risk is driven by strong institutional development. Specifically, higher foreign

liabilities in banking exacerbate bank-specific risk and total risk in countries with greater economic freedom.

Hence, the reinforcement of prudential regulations is necessary to overcome bank-specific risk and total

risk, particularly when the countries move to a more liberal economic environment.

1 The author is a PhD candidate in Money, Banking and Finance at the University of Limoges, France. The views expressed in this paperis the author»s and do not reflect those of the author»s affiliation. The author can be contacted through the following email:[email protected] or [email protected]

Abstract

JEL Classification : : : : : G21, G28, G38

Keywords : Banking Globalization, Economic Freedom, Capital Market Measures of Risk


I. INTRODUCTION

Despite a large number of studies examining the impact of foreign participation on bank

risk, the current wave of financial globalization and crises again highlight the increasing needs

for better understanding how foreign participation affects financial stability through channels

other than banking ownership or penetration per se. For instance, foreign participation in

banking can be through the presence of foreign managers, foreign debts, or even foreign

customers that demand services from domestic banking institutions.

This paper is the first to analyze the link between foreign participation and bank stability

in the Asian context through financial globalization channel by which foreign counterparts can

play their role in influencing banks» behavior. In this context, we focus on the role of banks»

foreign liabilities, where higher foreign liabilities in banking mean that foreign counterparts

have higher incentives to monitor banks» behavior. Higher foreign liabilities can also be associated

with higher technological innovation and better risk management in banking, so that banks

are able to access financing from international market beforehand. However, higher foreign

liabilities may induce banks to become prone to exchange rate depreciation as observed in the

1997 Asian crisis.

As the wave of financial globalization in Asian banks emerges over the last decades,

foreign participation in banking can also come from foreign debts. Sahminan (2007) works on

the Indonesian banking industry and shows that banks with higher ratio of foreign currency

assets to foreign liabilities are less exposed to exchange rate depreciation which in turn reduces

banks» insolvency risk. Our paper is close to Sahminan (2007) who examines the impact of

exchange rate depreciation on bank stability, but we focus on the role of foreign liabilities in

banking regardless of the exchange rate depreciation aspects. More specifically, our contribution

is threefold.

First, we work on a cross-country setting and focus on the post-1997 Asian crisis period,

while Sahminan (2007) focuses on the Indonesian banking industry in the pre-1997 Asian crisis

period. Second, it is admitted that banks» activities are now increasingly linked to financial

market and thus, contemporary banking crises are due to risks related to financial market

activities. In this regard, the present paper considers various risk measures based financial market

data instead of focusing on banks» balance-sheet indicators. Third, this paper augments the

analysis by examining the role of institutional development. Institutional development has indeed

become an important dimension in attracting foreign participation through the reinforcement

of shareholders» protection and the level of international playing field2. In the Asian context,

institutional development also played a critical role during the 1997 Asian crisis, where Furman


and Stiglitz (1998) point out that liberalized countries with weak institutional quality are the

ones hardest hit by the 1997 Asian crisis.

In order to investigate the issues raised in this paper, we work on a sample of publicly-

traded banks in seven Asian countries during the 2002-2008 period3. We choose Asian countries

whose data on banking globalization is available, and that are likely to acquire a particular

attention because of their capacity in attracting foreign capital during the last decade, as well

as their openness in permitting foreign participation in banking. These include India, Indonesia,

Hong Kong, Japan, South Korea, Thailand, and the Philippines. The rest of this paper is organized

as follows. Section 2 presents the theory,section 3 describes hypotheses and econometric

methodology. Section 4 discusses empirical findings and presents some robustness checks.

Section 5 concludes and highlights policy implications.

II. THEORY

Regarding the link between foreign participation and bank stability, the first strand of

literature focuses on the direct link with foreign entry or foreign ownership. From a broad

sample of emerging economies, Demirgüc-Kunt and Detregiache (1998) find that the presence

of foreign banks reduces the likelihood of banking crisis. In a similar vein, Detregiache and

Gupta (2004) document that the presence of foreign banks has a stabilizing effect before and

during financial crisis. Levy-Yeyati and Micco (2007) further suggest a positive relationship

between foreign bank entry and bank stability in Latin America. In the emergent economies of

Central and Eastern Europe, Dinger (2009) finds that the presence of foreign banks reduces the

risk of aggregate liquidity shortages. Foreign banks are also shown to have better position than

domestic banks without foreign participation, since foreign banks can provide modern banking

activities with better technological innovation and risk management, as well as access to

international financial markets (Berger et al., 2001; Bonin et al., 2005). On the contrary, some

have argued that domestic banks are better than foreign banks because domestic banks are

unlikely to suffer from a home-bias risk that exacerbates agency problems due to cultural

differences and regulatory environments (Berger et al., 2001; Lensink and Naaborg, 2007).

2 Institutional development also affects the link between financial development and economic growth, so that financial sector doesnot always boost economic growth. Specifically, threshold effects in the finance-growth nexus are likely to exist in developingcountries, which in most cases, lack of institutional development (Augier and Soedarmono, 2011).

3 The reason why we focus on the Asian banking industry is that Asian banking has experienced substantial changes in terms offoreign participation, particularly after the 1997 Asian crisis period. This includes a rapid growth of foreign ownership in bankingbecause of bank mergers and acquisitions (M&As), as well as an increase in foreign direct investment entering Asian banks. SeeSoedarmono et al. (2011a) for further discussion.


The second strand of literature is associated with the role of bank competition. Jeon et al.

(2011) highlight from a sample of Asian and Latin American banks during 1997-2008that

higher foreign participation in terms of bank ownership enhances competition in the banking

market through spillover effects from foreign partners to domestic counterparts.However,

empirical results on the implications of foreign participation through the link between bank

competition and financial stability are mixed. Working on a broad set of commercial banks

across developing countries over the 1999-2005 period, Ariss (2010) finds that the higher bank

market power, the lower risk and the higher profit efficiency of banks, although higher market

power deteriorates cost efficiency. On the other hand, Soedarmono et al (2011a) focus on the

Asian banking industry and find that banks in less competitive market tend to have higher

insolvency risk due to moral hazard effects.

This paper employs a set of dependent variables to assess financial stability. Specifically,

we use three types of risk measure coming from financial market data. These include total risk

(TRISK), systematic risk (BETA) and idiosyncratic risk or bank-specific risk (SRISK). To calculate

these risk measures, we initially construct the standard market model as follows4.

where ri,j,t

is bank i stock return at country j and at day t, while r mj,t

is the daily market

returns computed on the basis of domestic market index at country j. In the meantime, ri,j,t

and

r mj,t

are calculated as follows

and

where ri,t

and rm (j),t

are daily banks» stock price and total market index, respectively. In

constructing Equation (1), we impose a criterion to deal with low bank trading volume.

Specifically, we eliminate banks whose trading days are less than 70% of the total number of

trading days. For each year during the 2002-2008 period, moreover, Equation (1) is estimated

by applying the Panel Least Squares method. Hence, we obtain Equation (1) for each bank i

year by year.

TRISK is the annual standard deviation of daily bank stock returns during the 2002-2008

period, where the measure of daily bank stock returns is expressed by ri,j,t

. BETA is the annual

(1)

(2)

4 The model specification follows Bautista et al (2009).


systematic risk or the beta, β, coefficient in Equation (1). Systematic risk is a risk linked to

financial market activities and thus, is often referred to as market risk or non-diversifiable risk.

SRISK is the annual bank-specific risk or idiosyncratic risk which is diversifiable through several

risk managements at the bank level. SRISK is represented by the annual residual terms of the

standard market model, ε, as presented in Equation (1).

Variable of interest in this study is the level of globalization in banking activities and the

degree of economic freedom. We use country-specific data from the International Investment

Position section of the International Financial Statistics as a proxy of banking globalization.

Specifically, we use the share of international liabilities in banking sector over total international

liabilities (BLIAB). Greater BLIAB is associated with higher foreign participation in banking activities

and thus, representing greater financial globalization in banking. In parallel, the degree of

economic freedom is assessed by the Economic Freedom index (FREEDOM) coming from Heritage

Foundation. FREEDOM is a composite index of 10 indicators ranking policies in the areas of

trade, government finances, government interventions, monetary policy, capital flows and foreign

investment, banking and finance, wages and prices, property rights, regulation and black market

activity. The index scores from 0 to 1 with higher scores indicating policies being more conducive

to competition and economic openness.

Several control variables are also included in this study. First, we include the ratio of total

loans to total assets (LOAN) to account for bank opacity. Greater LOAN is likely to exacerbate

asymmetric information between banks and borrowers, while LOAN is the major source of

bank risk. Furthermore, we incorporate the ratio of total deposits to total assets (DEPO), as

bank deposits are also the major source of risk, particularly when market discipline is weak. We

also incorporate the ratio of loan loss reserves to total gross loans (LLR) as a proxy of credit risk,

where higher LLR is expected to increase total risk, systematic risk or bank-specific risk (Agusman

et al, 2008). Following Agusmanet. al. (2008) as well, we incorporate the ratio of equity to

total assets (EQTA) and the ratio of liquid assets to total assets (LIQUIDITY) to account for

leverage risk andliquidity risk, respectively. Higher EQTA and LIQUIDITY are expected to enhance

financial stability.

III. METHODOLOGY

There are two hypotheses tested in this paper. First, we aim to test if there is a relationship

between globalization in banking and financial stability as measured by total risk, systematic

risk and idiosyncratic risk. Second, we examine whether or not institutional development affects

the impact of banking globalization on financial stability. Let i, j, and t be bank index, country


index, and time index, respectively, then both steps are respectively presented in the following

equations.

RISK consists of TRISK, BETA and SRISK, while Control represents a set of control variables.

Initially, we calculate risk indicators coming from financial market data. We then run the

regressions of banking globalization on these indicators. Moreover, we also investigate the

impact of the interaction term between banking globalization and institutional development

on financial stability. To estimate such relationships, we use the Fixed Effect regressions with

heteroscedasticity-consistent standard errors.

In the present paper, we use bank-specific and country-specific data. For bank-specific

data, we retrieve banks» financial indicators from BankScope Fitch IBCA over the 2002-2008

period. Our initial sample consists of 189 publicly-traded commercial banks in seven Asian

countries. These include India, Indonesia, Hong Kong, Japan, Philippines, South Korea and

Thailand. In the meantime, we retrieve banking globalization data as country-specific indicator

from International Financial Statistics provided by the IMF. Other country-specific indicator such

Table 1.Descriptive statistics

TRISK 0.024158 0.022083 0.135767 0.00037 0.01215 1181BETA 0.832851 0.876176 1.967998 -0.28694 0.403701 1181SRISK 0.020088 0.01783 0.134859 0.000369 0.011475 1181BLIAB 0.191862 0.182324 0.396951 0.021616 0.111458 1323LOAN 0.58088 0.600438 0.886473 0.032652 0.1294 1219DEPO 0.867947 0.889310 0.970079 0.073737 0.096575 1225LLR 0.034881 0.02017 0.80149 5.00E-05 0.05337 1135EQTA 0.068095 0.05796 0.57868 0.00009 5.089816 1226LIQUIDITY 0.042605 0.001 0.57 0.0001 0.101373 916FREEDOM 0.629607 0.637 0.90 0.512 0.094772 1820

Variables Mean Median Maximum Minimum Std.Dev. Obs.

Source : Author»s calculationNote : TRISK is the indicator of total risk measured by the annual standard deviation of daily bank stock returns. BETA is the annualbeta coefficient from the standard market model, in which BETA represents market risk or systematic risk. SRISK is bank-specific riskmeasured by the annual residual term of the standard market model. BLIAB is the aggregate share of foreign liabilities in banking overtotal foreign liabilities. LOAN is the ratio of total loans to total assets. DEPO is the ratio of total deposits to total assets. LLR is the ratioof loan loss reserves to total gross loan. EQTA is the ratio of total equity to total assets. LIQUIDITY is the ratio of liquid assets to totalassets. FREEDOM is the Economic Freedom index retrieved from Heritage Foundation.

(3)

(4)


as institutional development comes from Heritage Foundation, where we assess institutional

development by using the Economic Freedom index. In order to assess financial stability based

on financial market data, we retrieve daily bank stock prices and daily total market index during

2002-2008 from Thomson Datastream International.

Before we run regressions, we impose restrictions to our dataset to deal with outliers and

missing values. We exclude all values that are less than zero for LLR, EQTA and LIQUIDITY. In

Table 1, we present the descriptive statistics of all clean variables used in this study.

IV. RESULT AND ANALYSIS

4.1. Empirical results

Table 2 shows the relationship between foreign liabilities in banking (BLIAB) and total risk

(TRISK). It can be seen that higher foreign liabilities in banking tend to exacerbate total risk.

Table 2.The relationship between banking globalization, economic freedom and total

BLIAB 0.0129*** 0.0131*** 0.0135*** 0.0101*** 0.0081** 0.0143*** -0.3622***(2.083) (2.152) (2.215) (2.606) (2.142) (3.499) (-4.369)

LOAN -0.0275*** -0.0258*** -0.0195 -0.0194 -0.0212 -0.0321***(-5.048) (-4.643) (-1.443) (-1.451) (-1.242) (-4.421)

DEPO -0.0125* 0.0061 -0.0148 -0.0093 -0.0126(-1.655) (0.6927) (-1.431) (-0.7805) (-1.179)

LLR 0.0559*** 0.0562*** 0.0541*** 0.0505***(3.901) (4.504) (4.316) (5.147)

EQTA -0.0627*** -0.0667*** -0.0698***(-2.909) (-2.738) (-3.735)

LIQUIDITY 0.0161 0.0172**(1.565) (2.464)

FREEDOM -0.0354(-1.079)

BLIAB*FREEDOM 0.5964***(4.944)

Obsevation 1181 1118 1118 1043 1043 859 859Adj R-square 0.42 0.42 0.42 0.44 0.45 0.44 0.51

ExplanatoryVar.

Source: Author»s calculation.Note: Dependent variable is TRISK which represents the indicator of total risk and measured by the annual standard deviation of dailybank stock returns. BLIAB is the aggregate share of foreign liabilities in banking over total foreign liabilities. LOAN is the ratio of totalloans to total assets. DEPO is the ratio of total deposits to total assets. LLR is the ratio of loan loss reserves to total gross loan. EQTAis the ratio of total equity to total assets. LIQUIDITY is the ratio of liquid assets to total assets. FREEDOM is the Economic Freedomindex retrieved from Heritage Foundation. Estimations are carried out using the Panel Fixed Effect regressions by considering White»sheteroscedasticity-consistent standard errors. Constants are included but not reported. ***,**,* indicate significant at the 1%, 5%and 1% levels, respectively.

Dependent Var. : TRISK

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7


This finding is robust to various control variable modifications as presented in Model 1 to

Model 6. Moreover, we examine how the interaction term between institutional development

and foreign liabilities in banking affects total risk, as shown in Model 7. It shows that the

positive impact of banks» foreign liabilities on total risk is dependent on institutional development.

More precisely, only in countries with higher economic freedom, the positive relationship between

banking globalization and total risk holds. This further suggests that in countries with lower

economic freedom, higher foreign liabilities in banking reduce total risk.

Meanwhile, Table 3 presents the impact of banks» foreign liabilities (BLIAB) on systematic

risk (BETA). By estimating different model specifications as presented in Model 1 to Model 6, it

shows that banks» foreign liabilities have a stabilizing effect in terms of reducing systematic

risk. But, the link between banks» foreign liabilities and systematic risk is no longer significant

when we take into account the role of institutional development as shown in Model 7. Economic

freedom does not seem to influence the relationship between bank globalization and systematic

risk.

Table 3.The relationship between banking globalization, economic freedom and systematic risk

BLIAB -1.984*** -2.032*** -2.041*** -2.129*** -2.163*** -2.161*** -3.105(-11.689) (-11.924) (-11.972) (-12.319) (-12.493) (-11.786) (-1.335)

LOAN -0.6573*** -0.6977*** -0.5690*** -0.5684*** -0.4779** -0.4672**(-4.303) (-4.485) (-3.161) (-3.164) (-2.381) (-2.297)

DEPO 0.2874 0.1649 -0.1927 -0.2190 -0.2168(1.361) (0.6918) (-0.6684) (-0.7301) (-0.7216)

LLR -0.2086 -0.2051 -0.1372 -0.0799(-0.7807) (-0.7693) (-0.515) (-0.2903)

EQTA -1.073** -1.109** -1.153**(-2.197) (-2.134) (-2.198)

LIQUIDITY -0.1841 -0.1932(-0.9408) (-0.9849)

FREEDOM -0.6407(-0.6964)

BLIAB*FREEDOM 1.245(0.3679)


ExplanatoryVar.

Source: Author»s calculation.Note: Dependent variable is BETA which represents the indicator of systematic risk and measured by the annual beta coefficient ofthe standard market model. BLIAB is the aggregate share of foreign liabilities in banking over total foreign liabilities. LOAN is the ratioof total loans to total assets. DEPO is the ratio of total deposits to total assets. LLR is the ratio of loan loss reserves to total gross loan.EQTA is the ratio of total equity to total assets. LIQUIDITY is the ratio of liquid assets to total assets. FREEDOM is the EconomicFreedom index retrieved from Heritage Foundation. Estimations are carried out using the Panel Fixed Effect regressions by consideringWhite»s heteroscedasticity-consistent standard errors. Constants are included but not reported. ***,**,* indicate significant at the1%, 5% and 1% levels, respectively.

Dependent Var. : BETA



In terms of the link between banks» foreign liabilities (BLIAB) and bank-specific risk (SRISK),

Table 4 presents our empirical findings for a number of model specifications. Higher foreign

liabilities in banking increase idiosyncratic risk, but again this relationship is dependent on

economic freedom. Only in countries with greater economic freedom, higher foreign liabilities

in banking worsen bank-specific risk. This may be due to the fact that under higher economic

freedom, foreign liabilities held by banks may be excessive, as banks are free to raise funding

from international market. Hence, the countries with greater economic freedom can be more

exposed to exchange rate depreciation that in turn exacerbates bank-specific risk.

Table 4. The relationship between banking globalization,economic freedom and idiosyncratic risk (bank-specific risk)

BLIAB 0.0221*** 0.0212*** 0.0216*** 0.0174*** 0.0157*** 0.0201*** -0.1618**(3.969) (3.945) (4.019) (3.204) (2.896) (3.228) (-2.098)

LOAN -0.0362*** -0.0344*** -0.0295*** -0.0294*** -0.0331*** -0.0386***(-7.515) (-7.027) (-5.197) (-5.227) (-4.876) (-5.719)

DEPO -0.0125 0.0054 -0.0128 -0.0082 -0.0099(-1.885) (0.7135) (-1.416) (-0.810) (-0.9901)

LLR 0.0546*** 0.0548*** 0.0524*** 0.0503***(6.489) (6.555) (5.817) (5.519)

EQTA -0.0545*** -0.0589*** -0.0602***(-3.558) (-3.347) (-3.462)

LIQUIDITY 0.0131** 0.0136**(1.969) (2.099)

FREEDOM -0.0141(-0.4608)

BLIAB*FREEDOM 0.2894**(2.579)


ExplanatoryVar.

Source: Author»s calculation.Note: Dependent variable is SRISK which represents the indicator of idiosyncratic risk and measured by the annual residual term of thestandard market model. BLIAB is the aggregate share of foreign liabilities in banking over total foreign liabilities. LOAN is the ratio oftotal loans to total assets. DEPO is the ratio of total deposits to total assets. LLR is the ratio of loan loss reserves to total gross loan.EQTA is the ratio of total equity to total assets. LIQUIDITY is the ratio of liquid assets to total assets. FREEDOM is the EconomicFreedom index retrieved from Heritage Foundation. Estimations are carried out using the Panel Fixed Effect regressions by consideringWhite»s heteroscedasticity-consistent standard errors. Constants are included but not reported. ***,**,* indicate significant at the1%, 5% and 1% levels, respectively.

Dependent Var. : SRISK


This finding is somehow consistent to Sahminan (2007), where there is a positive

relationship between exchange rate depreciation and insolvency risk for banks with higher

share of foreign liabilities. To the extent that only idiosyncratic risk plays a significant role in


capturing bank instability in Asia (Agusman et al, 2008), our findings are also consistent with

Agusman et al (2008). Specifically, systematic risk in the present paper is not an important

source of instability due to higher banks» foreign liabilities, regardless of the level of countries»

institutional development. Overall, the empirical findings suggest that investors in countries

with higher economic freedom are more concerned with total risk and bank-specific risk in this

regard.

With regards to control variables, higher banks» loan portfolio (LOAN) is associated with

lower total risk, lower systematic risk and lower bank-specific risk. This suggests that bank

lending activities are not source of instability. This result contradicts with the nature of bank

loan portfolio which is opaque. Presumably, Asian banks suffer from a managerial self-interest

problem in which managers in banks with higher asymmetric information might be driving

banks to become safer by holding less risky loan portfolios (Bris and Cantale, 2004; Soedarmono

et al, 2011b). Bank deposits (DEPO) also does not seem to be the source of instability. In line

with Agusman et al (2008), higher loan loss reserves ratio (LLR) is associated with higher total

risk and bank-specific risk. Meanwhile, the link between bank capitalization (EQTA) and RISK

fulfills our expected sign. Higher bank capital ratio reduces total risk, systematic risk and bank-

specific risk.

4.2. Robustness checks

In order to further ensure for robustness linked to variable omission issues, we also perform

some sensitivity analyses5. First, it is shown that Model 6 and 7 from Table 2, 3 and 4 suffer

from observation loss when we incorporate LIQUDITY as control variable. To ensure that the

empirical results obtained are not due to observation bias, we exclude LIQUIDITY and re-estimate

Model 6 and 7 in all three cases. However, this consideration does not change our empirical

results discussed in Section 4.1. Second, due to the fact that our bank sample comes from

countries with different level of macroeconomic environment, we also consider the influence

of economic development and inflation rate to account for this dimension. In other words, we

include the real gross domestic product (GDP) and inflation rate (INF) as control variable. On the

whole, our empirical results are not altered with regards to the link between banking

globalization, economic freedom and financial stability as measured by total risk, systematic

risk and idiosyncratic risk.

5 The results are not presented in the paper, but are available upon request.


V. CONCLUSION

In the aftermath of the 1997 Asian crisis, financial globalization in Asian banking in the

form of FDI and foreign ownership, as well as a rapid growth of foreign capital inflows entering

Asian countries due to the 2008 credit crisis and the 2010 European debt crisis, urgently require

better understanding whether or not greater foreign participations enhance financial stability.

Nowadays, such an increasing trend of foreign participations also show that Asian countries

are already in a better position in terms of institutional development, so that they have been

successful in attracting foreign participations. However, there is no attempt to investigate the

impact of foreign participations and institutional development on financial stability in the Asian

context.

This paper attempts to fulfill such a gap by assessing the impact of foreign participation

on financial stability through channel other than foreign participation commonly used in the

previous literature, such as bank foreign ownership, foreign bank entry, or bank competition.

In the present paper, we consider the indicator of foreign participation measured by the aggregate

share of foreign liabilities in banking over total foreign liabilities. Using a sample of publicly-

traded commercial banks in seven Asian countries during 2002-2008, the empirical results

from the Fixed Effect regressions show that higher foreign liabilities in banking reduces systematic

risk, but exacerbates bank-specific risk and total risk. However, we further shows that such

findings only hold for countries with greater economic freedom, suggesting the needs for

enhancing prudential regulations when banking in a more liberal environment are more

globalized in terms of their abilities to raise financing from international financial market. To

this end, bank-specific risk and total risk can be reduced, bank failure can be prevented, and

the systemic risk of bank failure can be avoided.


REFERENCES

Agusman, A., Monroe, G.S., Gasbarro, D., Zumwalt, J.K., 2008. Accounting and capital market

measures of risk: Evidence from Asian banks during 1998-2003. Journal of Banking and

Finance 32, 480-488

Ariss, R.T., 2010. On the implications of market power in banking: Evidence from developing

coutnries. Journal of Banking and Finance 34 (4), 765-775

Augier, L., Soedarmono, W., 2011. Intermédiation financière, croissance et effet de seuil. Revue

Economique (forthcoming)

Bautista, C., Rous, P., Tarazi, A., 2009. The determinants of bank stock return»s co-movements

in East Asia. Economics Bulletin 29(3), 1596-1601

Berger, A.N., DeYoung, R., Genay, H., Udell, G.F., 2001. Globalization of financial institutions:

Evidence from cross-border banking performance. Federal Reserve Bank of Chicago, Working

paper 1999-25

Bonin, J.P., Hasan,I.,Wachtel, P., 2005. Bank performance, efficiency and ownership in transition

countries.Journal of Banking and Finance29, 31-53

Bris, A., Cantale, S., 2004. Bank capital requirement and managerial self-interest. Quarterly

Review of Economics and Finance 44, 77-101.

Demirgüc-Kunt, A., Detregiache, E., 1998.The determinants of banking crises in developing

and developed countries.IMF Staff Papers 45 (1).

Detregiache, E., Gupta, P., 2004. Foreign banks in emerging market crises: Evidence from

Malaysia, IMF Working Paper 04/129

Dinger, V., 2009. Do foreign-owned banks affect banking system liquidity risk? Journal of

Comparative Economics 37, 647-657

Jeon, B.N., Olivero, M.P., Wu, J., 2011. Do foreign banks increase competition? Evidence from

emerging Asian and Latin American banking markets. Journal of Banking and Finance 35,

856-875

Lensink, R.,Naaborg, L., 2007. Does foreign ownership foster bank performance.Applied Financial

Economics17, 881-885

Levy-Yeyati, E., Micco, A., 2007. Concentration and foreign penetration in Latin American

banking sectors: Impact on competition and risk. Journal of Banking and Finance 31, 1633-

1647


Sahminan, S., 2007.Effects of exchange rate depreciation on commercial bank failures in

Indonesia.Journal of Financial Stability 3, 175-193

Soedarmono, W., Machrouh, F., Tarazi, A., 2011a. Bank market power, economic growth and

financial stability: Evidence from Asian banks. Journal of Asian Economicsdoi: 10.1016/

j.asieco.2011.08.003.

Soedarmono, W., Rous, P., Tarazi, A., 2011b. Bank capital and self-interest managers: Evidence

from Indonesia. LAPE Working Paper, Université de Limoges



141Competitive Conditions in Banking Industry: An empirical Analysis of the Consolidation, Competition andConcentration in the Indonesia Banking Industry between 2001 and 2009

COMPETITIVE CONDITIONS IN BANKING INDUSTRY:AN EMPIRICAL ANALYSIS OF THE CONSOLIDATION,

COMPETITION AND CONCENTRATION INTHE INDONESIA BANKING INDUSTRY BETWEEN 2001 AND 2009

Tri Mulyaningsih 1

Anne Daly 2

Few large banks dominate the Indonesia banking industri. Furthermore, in the past ten years, there

were a series of mergers and acquisitions in the banking market. The facts cause implications on competition.

In this paper, we examine these issues exploiting an unconsolidated annual financial report of all commercial

banks between 2001 and 2009. The Panzar-Rose method is employed to examine the banks behavior in

competition. Estimates indicate that banks in all three subsamples, large; medium-sized and small are

working in a monopolistically competitive market. The analysis of market concentration supports the

conventional view that concentration impairs competition. The study shows that the most competitive

market was the medium-sized banks because it was least concentrated. In contrast, the large market was

more concentrated thus it was less competitive. The consolidation policies driven by the Central Bank

reduced market concentration because mergers and acquisitions were mostly conducted by the medium-

sized and small banks. Further the improvement of market share distribution and the increasing capacity

of the merging banks enhanced competition in the Indonesia banking industry.

Keywords: Banking, market competition, market structure

JEL Classification: D43, G21

1 Doctoral Student, Faculty of Business and Government, University of Canberra Corresponding author: [email protected] Professor in Economics, Faculty of Business and Government, University of Canberra Corresponding author: [email protected]

Abstract


I. INTRODUCTION

The Indonesia banking industry experienced a structural change in 1990 where the number

of banks increased significantly. The change in market structure was promoted by the

deregulation policy in the banking sector in 1980s. Through the second deregulation in October

1988 the government removed barriers to entry to the banking sector. The minimum capital

requirement was lowered and it was easier for banks to obtain licenses to undertake foreign

exchange operations (McLeod, 1999). In addition, if the license was obtained, all branches

were automatically eligible to provide foreign exchange services.

After a series of deregulation policies during 1980s, the authorities decided to reduce

the speed of the banking expansion. During 1995 and 1996, the Central Bank introduced

some policies related to the prudential aspect of banks. McLeod (1999, p. 281-285) explained

that during those periods the Central Bank re-regulated the industry by re-imposing controls of

Table 1.The Microeconomic Policies in Indonesia Banking 1983 - 1997

Y e a r Policy Details

Removing the control on the interest rates of time deposits of State Owned Banksand the lending of all banks.

1. Opening the banking industry to the new private banks and the new joint venturebanks by lowering minimum capital requirement.

2. Removing business restrictions such as easing requirement of the existing banksto expand their branch easily; limit on interbank borrowing and permitting banksto introduce savings deposits product of their own design.

1. Opening the market by permitting foreign investors to purchase shares indomestic banks listed on the stock exchange

2. Partial privatizing by permitting State Owned Banks to list on capital market

Re-RegulationRe-RegulationRe-RegulationRe-RegulationRe-Regulation

1. Re-imposing the controls of bank lending;2. Extending the control to bank involvement with commercial paper issues;3. Extending the supervisory to non-bank finance company; increased the required

reserve ratio;4. Tightening of licensing of new bank branches;5. Imposing the fines on banks which expand more rapidly than permitted;6. Increasing the required reserve ratio and tightening prudential regulation

DeregulationDeregulationDeregulationDeregulationDeregulation

June 1983

October 1988

February 1992

1995-1997

Source: McLeod (1999, p. 293-295) and Chua, BH (2003)


bank lending (in 1995); extending the control to bank involvement with commercial paper

issues (in August 1995); extending supervision to non-bank finance companies (December

1995); increasing the required reserve ratio (February 1996); tightening of licensing of new

bank branches (June 1996); increasing the required reserve ratio and tightening prudential

regulation (in April 1997). Further, the 1997 economic crisis had heightened the importance of

re-structuring the industry as well as increasing prudential concern.

The consolidation was started on December 1997. The 1997 economic crisis generated

distress in the banking industry. In order to improve the performance of the state owned banks,

the Central Bank decided to merge four State Owned Banks. The authorities also closed the

operation of twenty three banks in 1997. The liquidation policies contributed to the reduction

on the number of banks. The consolidation process was continued by the introduction of the

Indonesia Banking Architecture (API) on January 2004. The policy essentially encourages banks

to reach their economies of scale as well as fosters the creation of a healthy banking system.

Table 2.The Microeconomic Policies in Indonesia Banking 1997 - 2010

Y e a r Policy Details

1. Banks liquidation (23 banks).

2. Banks recapitalization

3. Merger of 4 State Owned Banks became Bank Mandiri

Privatization of the bailout banks under the Indonesia Banking Restructuring Agency

(IBRA)

The introduction of the Indonesia Banking Architecture (API)

A series of mergers and consolidation conducted in banking to comply with Single

Presence Policy and Minimum Capital Requirement.

ConsolidationConsolidationConsolidationConsolidationConsolidation

1997

2003

2004

2004 - 2010

Source: Chua, BH (2003) and Central Bank of Indonesia (2010)

The banking architecture provides direction of the Indonesia banking system development

in the next ten years. The policy aims to create a strong, healthy and efficient banking structure.

In order to achieve its objectives, the Central Bank developed six pillars which are creating a

healthy domestic banking industry; inventing an effective system for regulating the banking

sector based on international standards; increasing the monitoring function of central banks

based on international standards; creating a strong banking industry with highly competitive


banks and fostering good corporate governance; realizing the proper infrastructure to support

the creation of a healthy banking system; and lastly increasing consumers» empowerment and

protection (Bank of Indonesia, 2010).

We suspect that at least two policies under the Indonesian Banking Architecture will

directly affect the market structure and competition in the banking industry. First is the minimum

capital requirement as regulated under the Central Bank Regulation number 10/15/PBI/2005.

The second policy is the single presence policy as explained in the Central Bank Regulation

number 8/16/PBI/2006.

Under Indonesian Banking Architecture, banks must increase their capital which is aligned

with their business scale (Bank of Indonesia, 2010). Bigger capital will enable banks to maintain

their business and risk develop the infrastructure such as information technology and increase

the scale of support of the expansion of credit capacity. Capital consists of in-paid capital and

disclosed reserves. According to the regulation, all existing banks, including the banks established

by regional governments, should attain minimum capital of Rp100 billion by 31 of December

2010 otherwise the Central Bank will impose several restrictions to the banks. In order to

increase capital, banks are allowed to receive additional capital injections from existing owners,

Figure 1.The Vision of Indonesian Banking Architecture Policy in 2014

Capital(Trillion)

Micro Bank Bank with Limited Business

Bank with Certain Focus

Corporate Retail Others

NationalBank

50

0.1

10

International

Local

Source: The Central Bank of Indonesia, (2010), Indonesian Banking Architecture, www.bi.go.id


merge with other banks, be acquired by bigger banks, or sell their shares on the capital market

(Bank of Indonesia, 2010). In the next ten or fifteen years, the Central Bank plans to reduce the

number of banks to 60, which consisting of 2-3 international banks, 3-5 national banks, and

30-50 specialized banks (Bank of Indonesia, 2010).

The single presence policy was introduced to re-arrange the structure of bank ownership

(the regulation of Bank of Indonesia, Number 8/16/PBI/2006). Single ownership refers to the

condition where one party is just eligible to be a main shareholder in one bank. The regulation

applies to shareholders who owned more than 25 per cent and those with ownership less than

25 percent but can control the banks. According to the rule, there should be an adjustment of

ownership structure by transferring the ownership (some or all) to one bank. Therefore they

only become a main shareholder in one bank. In addition, banks under the same ownership are

encouraged to merge. At last, they can set up a Bank Holding Company.

The study aims to assess the impact of consolidation policies in the banking industry on

market condition. The consolidation policy was started from the 1997 economic crisis and was

further enhanced by the introduction of the Indonesia Banking Architecture on January 2004.

Referring to Table 3, a large number of mergers and acquisitions took place in the Indonesia

banking industry between 1997 and 2010. One merger occurred in large bank market conducted

by the Bank Niaga and Bank Lippo in order to comply with the single presence policy. In addition,

there were 7 mergers in medium-sized banks and 7 mergers conducted by small banks.

A series of mergers and acquisitions reduces the number of banks. This raises some

important issues such as, whether a smaller number of banks reduced or increased the degree

of concentration in the banking industry; whether a smaller number of banks created a more

concentrated market or improved the market share distribution. Further we are also interested

in understanding the competition in the banking industry during the implementation of

consolidation policy, and whether consolidation generates a significant change in the

competition. Lastly, this paper would like to examine the link between the degree of market

concentration and the nature of competition during the consolidation process in the Indonesia

banking industry.


Table 3.The list of mergers and acquisition between 2000 and 2010

Bank Category Merging Bank

PT Bank Niaga

PT Bank Lihalo

Bank Dai-Ichi Kanggo

Bank IBJ Indonesia

Bank Bali

Bank Artha Media

Bank Universal

Bank Prima Express

Bank Patriot

PT Bank Sumitomo Mitsuo

Indonesia

Sakura Swadarma Bank

UFJ Indonesia Bank

Tokai Lippo Bank

UFJ Indonesia

PT Bank of Tokyo

Mitsubishi

Bank Hagakita

Bank Haga

Bank Rabobank Duta

Bank Buana

Bank UOB Indonesia

Bank Pikko

Bank CIC

Bank Danpac

Bank Artha Graha

Bank Inter-pacific Tbk.

Commonwealth Indonesia

Artha Niaga Kencana

Bank Multicor

Bank Windu Kentjana

Bank Harmoni

International

Bank Index Selindo

Bank Haga

Bank Hagakita

Bank OCBC

Bank NISP

Large Banks

Medium-Sized

Banks

Small Banks

Source: Banks» Annual Financial Report Published by the Central Bank of Indonesia

No Year Nama of New Bank

1

1

2

3

4

5

6

7

1

2

3

4

5

6

7

2008

2000

2001

2001

2001

2006

2008

2010

2001

2001

2004

2005

2007

2007

2008

2008

2009

PT Bank CIMB Niaga Tbk

PT Bank Mizuho Indonesia

PT Bank Permata Tbk

PT Bank Sumitomo Mitsuo Indonesia

UFJ Indonesia Bank

PT Bank of Tokyo Mitsubishi UFJ Ltd.

PT Bank Rabobank International

Indonesia Bank

PT Bank UOB Buana Tbk

PT Bank Mutiara Tbk

PT Bank Artha Graha International

Tbk

PT Bank Commonwealth

PT Bank Windu Kentjana

International Tbk.

PT Bank Index Selindo

Rabobank Duta Bank

PT Bank OCBC-NISP Tbk


II. THEORY

Many studies were interested to assess the impact of consolidation on concentration and

competition. Further, some journal articles also discuss the relationship between the degree of

concentration and competition. There are two prominent approaches dealing with this issue,

which is structural and non-structural (Bikker and Haaf, 2002).

According to the first approach, there is a direct relationship between market structure,

firms behavior and industry performance. This approach is based on traditional structure-conduct-

performance hypotheses (SCP). The second approach is known as the non-structural approach.

It has a different view on the relationship between those three elements. As explained by

Shaffer (1994a), competitive outcomes such as efficient prices can be established in un-

concentrated as well as highly concentrated markets therefore the link between market structure

and performance is not linear. As the relationship among market structure and performance is

not linear, they suggested focusing on the competitive conduct of banks instead of relying on

the market structure (Bikker and Haaf, 2002).

2.1. Structural Approach

The first foundation of traditional structure-conduct-performance approach was built by

Manson (1939). He concluded that fewer firms in the market will generally lead to less competitive

conduct, in terms of higher prices and reduced output levels, simply likes a monopolist model.

Furthermore, a concentrated market will produce less competitive performance, where the

price ratio to cost will be higher at the expense of lower consumer welfare. The small number

of firms might also facilitate firms to collude with their competitors. The collusion aims to boost

price, so it will be much higher than marginal cost (Yeyati and Micco, 2003b).

On the other hand, the competitive market will produce an efficient outcome as price

equals marginal cost. Thus, an increase in firm numbers will lead to more competitive conduct

by lowering price and reducing firms» profitability. In addition, the structure-conduct performance

(SCP) approach believed that the competitive market, which is produced by a low concentration

in the market, will deliver higher consumer welfare (Shaffer, 1994a). Efficient firms are able to

produce a higher output with lower prices.

Bain (1951) tested the structure-performance hypotheses empirically on American

manufacturing industry during period 1936 to 1940. His research confirmed the structure-

performance hypotheses. By using a Z score test, he compared the profit rates of a group of

firms that worked in a high concentrated market and those that worked in less concentrated


market. He found that the profits rates of firms in industries of high seller concentration should

on average be larger than those firms in industries of lower concentration.

Calem and Carlino (1991) found a correlation between market concentration and

performance in banking. Bain employed profit to assess market performance, Calem and Carlino

tried to use price to reflect market performance. Their conclusion supported traditional structure-

performance hypotheses that the concentrated market might contribute to collusive behavior

(Calem and Carlino, 1991). Further, the research claimed that the concentrated market is inefficient

because prices are higher than marginal cost. It is also not equitable because concentrated

market produces higher profits at the expense of consumers (Berger and Hannan, 1989).

The structural approach has become the foundation of antitrust policy in many countries.

The US Department of Justice has long adhered to this view by maintaining an explicit policy of

challenging mergers between rival firms that result in concentration levels above certain threshold

(Shaffer, 1994b). The Indonesia Legislation number 5, 1999 employs the structural approach to

determine the unlawful conduct of firms based on its impact to competition. The Law applies

to all industries. Further, according to the Government Regulation number 28/ 1999 on article

8 (2), we can find that the merged banks could not have the asset (after merger) of more than

20 percent of total assets in the banking industry. Thus, 20 percent is perceived as the threshold

where the merger would bring a negative effect into competition.

2.2. Non Structural Approach

The first argument against the traditional structure performance hypotheses came from

Demzets (1973) and Pelzman (1977). According to them, the source of concentration is efficiency

instead of market power. Their finding is labeled as Efficiency-Structure hypotheses. They

explained that the difference in firm-specific efficiencies within markets can create unequal

market shares and high level of concentration. The difference in efficiency might be derived

from superior management and production technology (Neuberger, 1997).

Manson implicitly assumes that the concentrated market facilitates firms to boost price.

Therefore market, where the price is higher than the marginal cost, perceives as less efficient.

Shaffer (1994b) argued that lower price is not a good indication to measure the market efficiency.

He defended his idea with the explanation that any extra efficiency among leading firms in a

competitive market would tend to show up as lower, not higher, prices.

SCP assumes that there is a single-way relationship between structure, conduct and

performance. Hence, market structure affects the firms» behavior further the behavior influences


the market performance. The causal relationship becomes unclear because firms» decision to

enter the market might be affected by the expectations of the degree of competition in the

market stage (Vesalla, 1995, p. 18). The non-structural approach argued that both the market

structure and firms» conduct are endogenous, since there is a feedback effects from conduct

back to market structure (Vesalla, 1995, p.18). In addition, market performance may also

influence the firms» decision to enter the market.

The non-structural approach claimed that the firms» profit is a poor measure of market

power. Vesalla (1995) argued that market power and profits are not necessarily positively

correlated (p. 19). The monopolists tend to be less efficient and it will affect the firms» profitability.

Hence inefficient behavior will reduce the profit rather than increase the firms» profit.

After that some researchers tried to use the Lerner Index to measure the market

performance. Shaffer evaluated the usage of the Learner Index as a measurement of market

performance. Shaffer (1994b) proved that the Learner index is a poor proxy of social welfare

because the linkage between the index and total welfare is not monotonic. The decreasing

number of firms reduces consumers surplus, but it increases producer surplus. Thus, mergers

may increase total surplus if the cost efficiency derived from mergers is higher than the reduction

of consumer surplus. Further, Shaffer (1994b) suggested measuring the cost structure as an

impact of consolidation policy, whether the cost structure becomes more efficient or not.

The non-structural approach argues that the link between concentration and competition

is not linear because it depends on various factors. The contestable market developed by Baumol

et al (1982) was referred by some researchers to explain competitive pricing in a concentrated

market. The main characteristic of contestable markets is free entry and exit. In such markets

an entering firm can attract customers by charging a low price and can recover any cost of

entry while abandoning the market if older firms retaliate by under-pricing in turn. As entry and

exit are free, the market might produce competitive pricing because less efficient firms can be

excluded from the market and replaced by the more efficient entrant.

Next, the non-structural approach focused on the information derived from competitive

behavior of firms. They believed that in learning about the market, we need to concentrate on

firms conduct instead of market concentration. In assessing competition, Panzar and Rose

(1987) developed a method based on the firm cost structure3. The method tries to distinguish

the competition level in markets by finding the relation of firms» revenue to factor price changes.

3 Panzer, J.C., Rose, J.N., (1987) developed a measurement to assess the competition in the market by observing the banksbehaviour. The H-statistic reflects the elasticity of changes on input price to banks revenue.


They derived a generally applicable testable implication of monopoly profit maximizing behavior.

Their model is focusing on the sum of factor elasticity of the reduced form of the revenue

equation. The factor elasticity capture the effect of a proportional shift in average, total, and

marginal cost curve even when cost data themselves are unavailable.

They concluded that in monopoly markets, the value of elasticity is negative because an

increase in input prices will increase marginal costs, reduce equilibrium output and subsequently

reduce revenues; hence elasticity will be zero or negative (Bikker and Haaf, 2002). In perfect

competition, the value of elasticity is unity, because the average cost will shift proportionally

with the changes in factor price.

Many studies employed the Panzar-Rose (PR) method to measure the impact of

consolidation policies on competition in the banking industry. In the early 1980s, the deregulation

of deposit interest rates in some developed countries had heightened competition and led to a

wave of mergers and acquisitions. Shaffer (1982), Molyneux et al (1994), Bikker and Haaf

(2002) and Bandt and Davis (2000) examined the impact of the banking consolidations in the

United States and European countries on competition.

The consolidation in the US banking market was mostly driven by markets as «merger

was perceived as one way to improve their diversification, efficiency or possibly market power»

(Shaffer, 1994, p. 3). By using samples from banks in some states in the United States between

1979 and 1980, Shaffer (1982) proved that the banking market was competitive. The study

found that in most of the market, the significance test of the elasticity rejected the existence of

market power.

In addition to that, the European banks also faced structural changes after the

implementation of the European Single Market, included the financial services market. Under

the policy, there was limitless access to enter markets of the other member countries (Molyneux

et al, 1994). Molyneux et al (1994) tried to assess the impact of consolidation on competition

in the early stage of implementation of the Single Market policy. They focused on assessing the

banks behavior between the period 1986 and 1989. They had quite similar findings to Shaffer

(1982). The competitive environment in the banking industry was well sustained. Monopolistic

competition was established in countries such as Germany, United Kingdom, France and Spain.

However, the banking market in Italy was performed under monopoly or conjectural variation

short run oligopoly.

More recent data was employed by Bandt and Davis (2000). Quite similar findings appeared

with Molynuex et al (1994). Bandt and Davis (2000) aimed to measure the level of competition

across different groups of banks, which are large, medium and small during the period of 1992


to 1996. According to them, the large banks in German and France were working under

monopolistic competition, while the smaller banks were acting as a monopoly. Regarding Italy,

both small and large categories were working under monopoly market.

Extending the coverage of countries, Bikker and Haaf (2002) tried to investigate the

competitive behavior in 23 developed countries all over the world during 1989 to 1998. The

banking market in 23 countries was indicated to be working under monopolistic competition.

The findings supported De Band and Davis (2000) where competition among larger banks was

stronger than other groups.

Quite a fewer studies have been conducted in emerging countries. Gelos and Roldos

(2002), Claessens and Laeven (2004) and Yeyati and Micco (2007) assessed the competitive

behavior of banks in some emerging countries. It is important to conduct the study for emerging

markets as we might find quite different conditions compared to mature markets. The Latin

American countries also experienced consolidation with a substantial number of mergers and

acquisitions. The consolidation was firstly initiated by government through the Central Bank.

Further, it was also driven by market forces. Gelos and Roldos (2002) explained that the three

largest private banks in Brazil initiated the consolidation to sustain their competitiveness. In

addition, the five largest private banks in Argentina gained substantial market share from the

combination of organic growth and acquisitions.

Consolidation did increase concentration in emerging market of Latin American countries.

However, the concentrated market did not lead to a less competitive situation. This conclusion

was generated from three different studies of Gelos and Roldon (2002), Claessens and Laeven

(2004) and Yeyati and Micco (2007). The Contestable market principle might be helpful to

explain this situation. The removal of the restriction of entry to the market effectively preserved

the competitive environment (Classens and Laeven 2004).

On the other side, the consolidation which was mostly driven by government, as happened

in most Asian countries, showed different outcomes. Instead of increasing the concentration,

the wave of mergers and acquisitions in Asian countries reduced concentration (Gelos and

Roldos, 2002). In relation to competition, the less concentrated market in the Asian banking

market was perceived as competitive.

This illustration of findings under the non-structural approach clearly supports the notion

that the relationship among market concentration, firms conduct and market performance

may not linear. The consolidation itself, which was taken from broad varieties of policies, might

create concentrated or less concentrated markets. Further, competition in concentrated market

might be sustained by lessening the entry barrier into the market (Classens and Laeven 2004).


(1)

(2)

2.3. Panzar and Rose Approach

As illustrated in the above section, we could not rely on market structure information to

determine the competition level in the banking market. It might be misleading to depend on

the information about market structure as some studies proved a non-linear relation between

market structure and competition. Moreover, Bikker and Haaf (2002) proved that concentration

indices such as the Concentration Ratio (CR) and the Herfindahl-Hirschman Index (HHI) appear

to be inversely correlated with the number of firm. The countries with smaller number of banks

tend to have a higher concentration level.

In order to measure competition in the market, this study will employ a method used

by Panzar and Rosse (1987). The method is based on properties of reduced form revenue

equations at the firm level, and the data of revenues and factor prices (Panzar and Rose,

1987). Therefore the method will directly assess the competitive behavior of banks to define

the market structure.

The Panzar-Rose method will calculate the sum of elasticity of the reduced form revenues

with respect to factor prices. This sum of elasticity is given the symbol H (Vesalla, 1995). The

value of elasticity will provide information about banks conduct, and furthermore it will

determine the structure of the market. The assumption underlying this method is that market

power of banks is measured by the extent to which changes in factor prices (unit costs) are

reflected in revenue earned (Vesalla, 1995). If the industry is competitive the elasticity will be

high, otherwise the elasticity will be low or even negative in the case of monopoly and

collusive oligopoly.

The properties of H allow us to distinguish empirically between common imperfect

competition theories of price formation as characterizations of the competitive behavior of

Indonesia»s banks, whether monopoly or perfect collusion, monopolistic competition or perfect

competition (Vesalla, 1995).

The Panzar-Rose empirical model assumes that banks have a log-linear marginal cost and

revenue function (Bikker and Haaf, 2002).


Where OUT is output, n is the number of banks, FIP denotes factor input prices, and EXirevenue

and EXicost

are other variables affect banks» revenue and cost functions, respectively. The empirical

application of Panzar and Rose approach assumes a log-linear marginal cost function, where

dropping subscripts referring to bank»i (Bikker and Haaf, 2002).

Further, the Rose Panzar model assumes a profit maximizing for individual banks. The

profit maximizing banks will produce at the level where marginal cost is equal to the marginal

revenue, yielding the equilibrium value for output:

In the empirical analysis, the following operationalization of the reduced-form revenue

equation is used (Bikker and Haaf, 2002, p.2196):

Where, TIR is the ratio of interest income to the total balance sheet; AFR is the price of funding

(funding rate); PPE is the price of personnel expenses (wage rate); PCE is price of capital

expenditure (capital price); OI is ratio of other income to the total balance sheet; and BSF is

bank-specific exogenous factors, such as the risk component, differences in the deposit mix

and size (banks» real assets) (Yeyati and Micco, 2007, p.1637).

Under this condition, H shows the sum of the elasticity of the reduced-form revenue

function with respect to factor prices.

The level of competition determines the value of H, whether monopoly or perfect collusion,

monopolistic competition or perfect competition. Below is a formula to calculate H, where H is

the sum of elasticity, which consists of the elasticity of revenue with regard to funding cost

changes ( β ), elasticity of revenue with regard to changes on human resource expenditure ( γ )

and elasticity of revenue with regard to capital price changes ( δ ).

(3)

(4)

(5)


Panzar and Rose proved that under monopoly, an increase in input prices will increase

marginal costs, reduce equilibrium output and subsequently reduce revenues; hence H will be

zero or negative (Bikker and Haaf, 2002). In other words, markets where monopoly power

exists will yield a negative relation between those two variables as gross revenue responds in

the opposite direction to a change in a unit cost (Vesala, 1995). The same outcome is also

found in monopolistic competition without the threat of entry, i.e. with the fixed number of

banks. The market consists of some banks; however there are barrier to enter the market,

therefore the number of bank is unchanged. Vesalla (1995) proved that in such market H is

zero or negative, where similar with Panzar-Rose findings on monopoly market.

In analyzing monopolistic competition, the Panzar-Rose approach is based on the

comparative statics properties of the monopolistic competition Chamberlinian equilibrium model

(Bikker and Haaf, 2002). In the limit case of the monopolistic competition model, where banks»

products are regarded as perfect substitutes of one another; the Chamberlinian model produces

the perfectly competitive solution, as demand elasticity approaches infinity (Bikker & Haff: 2002:

pp. 2195). In a perfectly competitive market, increasing input prices will increase average cost

Figure 2.H in perfect competition

P1AC

P1

P0

y* y

AC1 (W1)

AC0 (W0)

(6)H = β + γ + δ


P MC

AC

B

MR

Pc

P

y y c y D

proportionally. Exit of some banks increases the demand faced by each of the remaining banks,

leading to increases in prices and revenues equivalent to the rise in costs (Bikker and Haaf,

2002). Finally, the value of H in perfectly competitive markets is equal to one.

In the case where monopolistic competition model recognizes product differentiation

the value of H is positive but less than one. As shown in figure 3, monopolistic competition is

an intermediate case between monopoly and perfect competition. Monopolistic competition

produces «excess capacity» where yi* is produced at average cost (point A) higher than minimum

average cost (point B). In addition, price pi* exceeds marginal cost (MC), compared to competitive

solution pc (Vesalla, 1995).

Figure 3.Monopolistic Free Entry (Chamberlinian) equilibrium

Banks in Chamberlinian equilibrium are not free from market power but they cannot

receive supernormal profits, since banks price at average cost (Vesalla, 1995). Therefore, the

value of H in this condition is positive but less than one. It means that the changes to input

prices positively affect revenue but are less than one.

This model is more consistent with the observation that banks tend to differentiate

themselves by various product quality and advertising, although the core services provided by

them are fairly homogenous (Vesalla, 1995, p. 50). The product differentiation is important in


order to create less elastic demand (Tirole, 1988). As product is less differentiated, less market

power is exercised and the value of H will be higher. High switching cost is also the source of

market power. The account holders of the banks bear high switching cost to move from one

bank to others. If customers plan to switch the supplier then they must create a new account

number. It implies a lot of paperwork. In addition, the customer must inform about the changes

to their business relation (Canoy et.al, 2001).

At last, in an oligopoly market, the value of H may also be positive, where there are

strategic interactions among fixed number of banks (Bikker and Haaf, 2002). However, in the

case of perfect collusion in an oligopoly, the Rose and Panzar method produces a negative

value for H, similar to the monopoly model.

Table 4.Summary of the Discriminatory Power of H

Competitive environment

Monopoly equilibrium: each bank operates independently as under monopoly profit

maximization conditions (H is decreasing function of the perceived demand elasticity) or

perfect cartel.

Monopolistic competition with free entry equilibrium (H is an increasing function of the

perceived demand elasticity).

Perfect competition. Free entry equilibrium with full efficient capacity utilization.

Value of H

H < 0

0 < H < 1

H = 1

Source: Bikker and Haaf (2002, p. 2195)

There are five assumptions need to apply the Panzar-Rose method. First, banks are treated

as single product firms that act as financial intermediaries. Therefore banks produce interest

revenues by employing intermediate funds, labor and capital as inputs (De Bandt and Davis,

2000). Second, we have to assume that higher input prices are not associated with higher

quality services that generate higher revenues. Gelos and Roldos (2001) explained that if the

correlation exists, there might be bias in interpreting H. The third assumption is that the market

is in equilibrium in the long run. The fourth and fifth assumptions are considering banks as

profit maximization institutions and they have normally shaped revenue and cost functions

(Gelos and Roldos, 2002, p. 13-14).


III. METHODOLOGY

3.1. Empirical Model

In assessing the competition in the Indonesia banking market between 2001 and 2009,

we will run a regression model to estimate the following reduced form revenue equation. In

following equation, the four explanatory variables outside the bracket represent bank specific

factors.

Table five provides detail information about the variable specification. It explains the

definition and proxy used to measure each variable.

The study will employ a pooled regression, more specifically the Fixed Effect Model. There

are some reasons underlying the application of the pooled regression model. First, since the

application of a cross-section regression implicitly assumes that all banks have access to the

same factor markets therefore they only differ in terms of the scale of operation (De Bandt and

Davis, 2000, p. 1050) the cross-sectional regression is unable to capture the time-series dimension

inside the model. Vesalla (1995) concluded that it is difficult to infer whether the changes on

competition over time are statistically significant if we rely on the estimation of H using the

(7)

Table 5.Variable Spesification

Ratio of annual interest income to total assets

Ratio of annual interest expenses to total deposit

Ratio of annual wage and salary expenses to total assets

Ratio of other expenses to fixed assets

Ratio of others income (operational and non-operational incomes) to total assets

Equity divided by total assets

Loans divided by total assets

Ratio of interbank deposits to total deposit

Ratio of demand deposits from customers to total deposit and short term

funding

TIR (revenue)

AFR (funding rate)

PPE (wage rate)

PCE (capital rate)

OI (other income)

EQ (capital risk)

LO (loan risk)

BDEP (deposit mix)

DDC (deposit mix)

VariableVariableVariableVariableVariable Variable SpesificationVariable SpesificationVariable SpesificationVariable SpesificationVariable Spesification


cross-section regression for each year. Some studies such as Vesalla (1995) and Yeyati and

Micco (2003) suggested consolidating the individual cross sections into a single pool data set.

Polled estimation will produce more reliable estimates of H as it examines the behavior of

banks over time.

The second consideration is that the study will also explore the relation between

concentration and competition across time. Therefore we emphasize the dynamic dimension,

which cannot be incorporated in the cross-section method (Yeyati and Micco, 2003a).

The third benefit of the implementation of pooled regression method relies on the ability

to capture the non-time-varying determinants of banks» revenues. The application of the Fixed

Effect Model allows the inclusion of bank fixed effects that can be used to control the

heterogeneity between banks that are not captured in the model. The Fixed Effect Model will

treat the heterogeneity of the non-time-varying determinants of revenues by entering the cross-

section dummies (for each bank). Therefore the model will introduce the different intercepts

capturing the banks-specific variables that are not explicitly addressed in the regression

specification (De Bandt and Davis, 2000).

In order to capture the impact of the consolidation policy on the banks» behavior, this

study tested the changes on the input prices coefficients. Referring to Gelos and Roldos (2002)

the test can be implemented by dividing the observation period into two sub-periods and

interacting the input price variables (ln(AFR), ln(PPE) and ln(PCE)) with a dummy variable that

takes the value of one in the second sub-period» (p. 15). The interaction variables will show

whether the consolidation policies significantly changed the banks» behavior; if the interaction

variables generate significant values, they indicate a structural break in the statistical relationship

between revenues and input prices (Gelos and Roldos, 2002, p. 15). In addition, the value of

interaction variables will determine the direction of changes on competition. If they are positive,

we can conclude that the consolidations increase competition or otherwise. Further if the H-

statistics is positive between 0 and 1 and the cumulative value of the interaction variables is

positive, it implies a stronger competition (Vesala, 1995, p. 56)4.

The first period is from 2001 to 2003 and the second period starts from 2004. This study

chooses 2004 as a year marking of structural break because starting from 2004; the Central

Bank of Indonesia formally enforced the consolidation process by introducing the Indonesia

Banking Architecture policy.

4 It refers to Bikker and Haaf (2002, p. 2203). They explained that the result of Vesala (1995) implies that the interpretation of H-statistics between 0 and 1 is a continuous measure of the level of competition. Further Bikker and Haaf (2002) described thehigher value of H can be used as an indication of the stronger level of competition (p. 2203).


One of the assumptions under the Panzar and Rose method is that the market is in

equilibrium (Claessens and Laeven, 2003). We need to check that the assumption is fulfilled by

running the test on market equilibrium. Referring to some studies, the market equilibrium test

must be able to validate that the Panzar-Rose statistics can deliver reliable results (De Bandt and

Davis, 2000).

In order to have an equilibrium test, some studies suggest testing whether the input

prices are related to the industry return. Here, we will modify the reduced form revenue equation

by replacing the dependent variable with the ratio of net income to total assets as an endogenous

variable (De Bandt and Davis, 2000).

The equilibrium is defined as a condition where the value of equilibrium E-statistics is

zero. The E-statistic is defined as the summation of β, γ, dan δ. Further, we can employ a F-test

to verify the statistical significance of the test whether the E = 0 (Claessens and Laeven, 2003).

3.2. Data

Data were obtained from the Central Bank of Indonesia. It consists of the unconsolidated

annual balance sheets and income statements of the commercial banks. The sample contains

128 banks for each year approximately; however the exact number in the sample for each year

varies. The yearly variation on the number of banks is caused by mergers, acquisitions, bank

liquidations and banks entry during the observation period. In the case of mergers and

acquisitions, the database only keeps the data of the new institution which is usually the bigger

bank.

Table 6 provides detail information about the number of bank for each category. According

to the database, the Central Bank divides commercial banks into six categories. Currently

Indonesia has 4 State Owned Banks, 68 local private banks which are comprised of banks

which engaged with foreign exchange services and those which are not. Even though the

Central Bank has grouped the 68 banks into local banks, actually some of them were bought

by overseas investors during the privatization policy in the late 1990s. There are 20 joint ventures

banks, where the owners are foreign and local investors. There are 11 foreign banks which are

usually branch offices of foreign banks. Finally, we have 25 regional banks that usually operate

(8)


in their province. The shareholders of the regional banks are the government of the province

and the municipalities.

Compared to other developing countries in South East Asia, the number of banks in

Indonesia is large. However, the market is concentrated as a few large banks dominate it. The

banks» balance sheet shows that on average 3 from 113 banks control more than forty percent

of the industry between 2001 and 2009. Descriptive statistics information about the distribution

of»the assets and capital support the notion of the highly concentrated market. Table 7 illustrates

the distribution of the assets and equity from all banks in 2009. In terms of assets, a quarter of

market is dominated by single bank which is Bank Mandiri with assets 370 trillion rupiah and

Table 6.Sample Size

4

39

29

20

11

25

128

State Owned Banks

Private Banks - Foreign Exchange

Private Banks - Not engage in foreign exchange

Joint Venture Banks

Foreign Banks

Regional Banks

Total

Type of Banks (based on the Central Bank Categorization)Type of Banks (based on the Central Bank Categorization)Type of Banks (based on the Central Bank Categorization)Type of Banks (based on the Central Bank Categorization)Type of Banks (based on the Central Bank Categorization) Number of BanksNumber of BanksNumber of BanksNumber of BanksNumber of Banks

Source: The Indonesian Banking Statistics, 2009, the Central Bank of Indonesia

Table 7.The Descriptive Statistics of Asset and Equity All Banks in 2009 (in Million Rupiah)

StatisticsStatisticsStatisticsStatisticsStatistics AssetAssetAssetAssetAsset EquityEquityEquityEquityEquity

Mean

Median

Q4 (100% data)

Q3 (75% data)

Q2 (50% data)

Q1 (25% data)

Skewness

Kurtosis

Jarque-Bera

Probability of JB

22,158,195

3,978,396

370,310,994

13,356,445

3,923,234

1,523,057

4,44

23,26

2,242,80

0,00

2.267.144

547.938

35.108.769

1.370.931

539.862

164.954

4,16

21,00

1.803,03

0,00


equity more than 35 trillion rupiah. The asset of Bank Mandiri is 15 times larger than the

average and the equity is 17 times larger than the average.

Since the market is positively skewed, the median is a preferred measure of central

tendency. Table 7 shows that all statistics to assess the normality distribution of data are violated.

The values of skewness are positive both for assets and equity, implying the positively skewed

distribution. The JB values are positive and significantly different from one, indicating that the

data rejects the normality assumption. Referring to the data, half of the bank (56 banks) had

equity less than 550 billion rupiah. In addition, more than 66 percent (85 banks) had equity less

than 1,371 billion rupiah. On the other hand, a very small portion of banks - less than 1 percent

or only 7 banks - has equity more than 10 billion.

The above figures will be used to calculate break point to divide the sample based on the

size. The sample division into large, medium and small banks is important to capture the real

behavior of banks in competing with other banks in their categories. Large banks are assumed

to work at the national or even international level, while small banks focus on local or specific

industries and often do not have customers outside their defined market.

5 The categorization is based on the data of Indonesia banking in 2009.

Table 8.Three Categories of Banks based on the Value of Equity

CategoryCategoryCategoryCategoryCategory Equity (IDR)Equity (IDR)Equity (IDR)Equity (IDR)Equity (IDR) Sample Size Sample Size Sample Size Sample Size Sample Size 44444

> 10 trillion

1 - 10 trillion

< 1 trillion

7

36

70

113113113113113

Large Banks

Medium-sized Banks

Small Banks

TotalTotalTotalTotalTotal

The value of equity in 2009 will be employed as a basis to break the sample into large,

medium, and small banks. Large banks are those with equity more than 10 trillion rupiah. The

medium-sized banks are working with the capital from 1 to 10 trillion rupiah. Finally, the small

banks are those with equity less than 1 trillion rupiah. In terms of the number of banks, the

small bank category dominates the market; however the market is controlled by the larger

group.



4.1. Competition in Banking Industry

The finding is robust using the two different measurements of revenue with respect to

the level of competition in the Indonesia banking industry. Some studies employed the ratio of

interest income to total asset to measure the banks» revenue. Bikker and Haaf (2002) explained

that this proxy is more consistent with the assumption underlying the Panzar and Rose method,

where banks are assumed to be a single product firms that acts as financial intermediaries.

Therefore the main business of banks is financial intermediation.

On the other hand, De Bandt and Davis (2000) argued that as a response of tighter

competition, banks increase their interest to non-interest income activities √ including asset

management income, mutual funds and insurance (p. 1047). They proposed to employ both

measurements on the ratio of interest income to asset and the ratio of total revenue to asset as

a validation method to test if the model is robust.

Overall both specifications produce similar outcomes. However the interest income

specification provides some larger coefficients. We choose to employ the first proxy because

the Indonesian banking industry still relies heavily on interest-based-income. The data of the

annual financial report of all banks between 2001 and 2009 shows 88 percent of banks» income

was originated from interest income. Therefore the first proxy will capture the real portrait of

competition in the Indonesian banking industry. In addition, the study includes other income

which means income originated from non-interest activities as one of the explanatory variables

in order to capture the growing trend of non-interest based income in banks revenue.

One critical assumption under the Panzar-Rose method is whether the observations are

in long-run equilibrium. As described in the previous part, we have conducted an equilibrium

test by running a regression of the revenue reduced-form by replacing the dependent variable

by return on equity. We employ the method introduced by Claessens and Laeven (2003) in

deriving the Return on Asset (ROA). As ROA can take on a small (negative) value, they suggested

computing the dependent variable ROA» by using the formula of Ln (1+ROA) (p. 11). The

estimation showed that the joint coefficient of β, γ and δ is not significantly different from

zero (please refer to the appendix for the complete estimation result). Therefore, the Panzar-

Rose method is suitable to assess competition in the Indonesia banking market between 2001

and 2009.

The Panzar-Rose model has been applied to around 128 banks between 2001 and 2009.

The redundant test was employed to examine whether the introduction of an intercept for

different cross section units in order to account the heterogeneity across banks are significant.


The restricted F-test and x2 test of fixed effect are significant. They confirm that fixed effect

model is more reliable capturing the information about what contributes to the difference

across banks. Thus the model can capture the information of what contributes to the differences

across banks. Please refer to appendix C for detail statistical results on redundant fixed effects

tests.

Table 9 provides complete information about the estimation results of the reduced-form

revenue equation. The H-statistics for all banks is .69. The F test shows that the value is

significantly different from zero thus it rejects the monopoly hypotheses. It is also significantly

different from one, thus it also rejects the perfect competition hypotheses. We can conclude

Table 9.Empirical Results using Fixed Effect Model Interest Income as Dependent Variable,

All Banks between 2001 and 2009

Funding rate

Wage rate

Capital price

Risk of loans

Risk of capital

Deposit Mix √

Demand Deposit

Deposit Mix √

Interbank Deposit

Other Income

Number of observation

R2

H (competition) MarketH (competition) MarketH (competition) MarketH (competition) MarketH (competition) Market

structurestructurestructurestructurestructure

All Banks Large Banks Medium Banks Small Banks

a) the value of F test shows that H is not significantly different from 0 and 1 (level confidence 99%)The value in the bracket is the t-statistic;*** Significant at 99% confidence level; ** significant at 95% confidence level; * significant at 90% confidence level

0.46***

(34.74)

0.21***

(11.52)

0.02

(1.50)

0.45***

(14.04)

0.002

(0.19)

0.03***

(2.06)

-0.001

(-0.60)

-0.008

(-0.83)

892

0.89

0.690.690.690.690.69aaaaa

MonopolisticMonopolisticMonopolisticMonopolisticMonopolistic

CompetitionCompetitionCompetitionCompetitionCompetition

0.49***

(18.56)

0.08***

(2.53)

0.01

(0.40)

0.17

(1.20)

0.18***

(2.68)

0.025

(0.81)

-0.021***

(-2.17)

0.04

(1.54)

63

0.94

0.590.590.590.590.59aaaaa



0.51***

(28.52)

0.26***

(10.10)

0.03***

(2.12)

0.52***

(11.51)

0.10***

(4.02)

0.007

(0.33)

-0.003

(-0.97)

-0.022

(-1.64)

309

0.94

0.800.800.800.800.80aaaaa



0.46***

(24.50)

0.19***

(7.50)

0.018

(0.88)

0.457***

(9.41)

-0.028

(-1.83)

0.02

(1.29)

0.001

(0.40)

-0.010

(-0.77)

520

0.87

0.660.660.660.660.66aaaaa




that the banking sector is working under monopolistic competition. The changes in factor price

will increase the revenue less than proportionally because the products are differentiated and a

switching cost is high. Further, competition has risen over time. This conclusion was based on

a comparison of our findings with the study conducted by Classens and Laeven (2003). According

to them, from 1994 to 2001, the H-statistic for Indonesia banking was .62.

The across groups analysis provides broader understanding of the market. The H-statistics

vary from .59 to .80 suggesting that the banking market across all groups was working under

monopolistic competition. The large banks groups was the least competitive market, with the

H-statistics .59. The medium banking group was the most competitive market. The H-statistic

for the medium bank group is quite high (.80), and it is comparable with the banking market in

developed countries and other emerging market such as Latin American bank (Yeyati and

Micco, 2007). It is quite surprising as in other countries including Europe; the banking market

of larger banks is more competitive than that of the smaller banks which usually service the

local market (Bikker and Haaf, 2002).

As expected, deposits or funds are the most important input. The coefficient for the

funding rate is the main contributor of the elasticity. This figure is consistent across different

groups. Human resources are the second main input in the banking industry. The sign of the

coefficients are also positive indicating that the increase in the wage rate is transmitted into the

revenue. The least important input is capital. The coefficients on the capital price are small, and

in some estimates, they are not significant.

Risk of loans is an important explanation of the pattern of revenue. Positive coefficients

for risk of loans imply that banks with a higher proportion of loans on their balance sheets

generate higher interest revenue per rupiah of asset. A similar conclusion is also applied to the

risk of capital variable. The coefficients are positive for large and medium banks. Thus in the

larger banks, it can be concluded that the higher portion of equity per asset earns higher

revenue. The deposit mix variables are not really significant in explaining the movement of

revenue, except the proportion of interbank deposits to total deposits in the large banks market.

The sign of the coefficient is negative which implies that the bigger the portion of interbank

deposit to total deposits, the less revenue that might be earned.

4.2. Market Structure and Competition

This section will describe the relationship between concentration and competition. The

study will employ two-frequently applied types of such indices as a proxy of market concentration.

The first index is called k-banks concentration ratio (CRk) which takes the market shares of the


k biggest banks in the market, and ignoring the remaining banks. The second index is Herfindahl-

Hirschman index (HHI), which takes market shares as weights, and stresses the importance of

larger banks by assigning them a greater weight than smaller banks (Bikker and Haaf, 2002)6.

These concentration measurements can be considered reliable for assessing market

concentration. As all proxies produces similar outcomes. The data shows that the banking

market in Indonesia has become less concentrated. This finding is surprising. Unlike the developed

countries where consolidation increased the market concentration7, policy changes reduce the

concentration in the Indonesia banking market.

Our findings are aligned with a study conducted by Gelos and Roldos (2002). They argued

that consolidation did not increase market concentration in Asian countries because the structural

changes were mostly driven by government. While in the developed countries and more mature

markets of developing countries in Latin America, mergers and acquisitions were driven by the

market.

In those markets, tight competition forced large banks to merge to enhance their

competitiveness. On the contrary, mergers and acquisitions in Indonesia banking were mostly

6 The following are formulas to calculate the concentration ratio (CR)and Herfindahl-Hirschmanindexes, where s is market share ofbanks.

7 Please refer to De Bandt and Davis (2000) and Bikker and Haaf (2002).

Figure 4.The Banking Concentration by Using Four Different Measurements

All Banks, 2001 - 2009

0.7

0.6

0.5

0.4

0.3

0.2

0.1

02001 2002 2003 2004 2005 2006 2007 2008 2009

CR3CR4CR5HHI

CRk = Σk sii = 1 HHI = Σm s

i2

i = 1


conducted by medium-sized and smaller banks in order to comply with the Single Presence

Policy and Minimum Capital Requirement. These consolidations increased the scale of economies

of the merging banks. However it did not increase industry concentration because the mergers

were dominated by small and medium-sized banks. It implies that consolidation in the Indonesian

banking market does not create more concentrated markets but it improves the market share

distribution.

In the case of large banks, the Single Presence Policy has effectively reduced concentration

by the consolidation of PT Bank Niaga Tbk. and Bank Lippo in 2008. The market share of the

merged bank, PT. CIMB Niaga, increased about 3 percent. This merger significantly reduced the

market concentration because the new merge bank was not in the top three large banks.

Therefore as the share of merge bank increased, the market share distribution was changed

where the share of biggest three banks decreased from 74 percent in 2001 to 66 percent in

2009.

The mergers and acquisitions in the medium-sized and small banks market aimed to

improve the banks» performance after the 1997 economic crisis or to comply with the Single

Presence policy or Minimum Capital Requirement policy. There were seven mergers in the

medium-size banks category between 2001 and 2009. The number of banks in the small

group also reduced. The consolidation in small banks market was dominated by the

liquidations of banks with poor performance and consolidation to comply with the Minimum

Capital Requirement policy. Please refer to table 3 for a detail list of merger between 2001

and 2009.

Table 10.The Banking Concentration across Groups 2001 - 2009

All Banks Large Banks Medium-sized Banks Small Banks

20012001200120012001 0.47 0.10 0.74 0.24 0.30 0.06 0.35 0.05

20022002200220022002 0.46 0.10 0.74 0.23 0.30 0.06 0.30 0.04

20032003200320032003 0.45 0.09 0.73 0.22 0.27 0.05 0.30 0.04

20042004200420042004 0.42 0.08 0.71 0.21 0.26 0.05 0.29 0.05

20052005200520052005 0.38 0.07 0.67 0.20 0.25 0.05 0.33 0.05

20062006200620062006 0.36 0.06 0.65 0.17 0.23 0.05 0.29 0.05

20072007200720072007 0.37 0.07 0.64 0.18 0.23 0.04 0.30 0.05

20082008200820082008 0.37 0.06 0.64 0.18 0.23 0.04 0.32 0.06

20092009200920092009 0.39 0.07 0.66 0.18 0.24 0.05 0.28 0.05

Average 0.0.0.0.0.4141414141 0.0.0.0.0.0808080808 0.0.0.0.0.6969696969 0.0.0.0.0.2020202020 0.0.0.0.0.2626262626 0.0.0.0.0.0505050505 0.0.0.0.0.3131313131 0.0.0.0.0.0505050505

CR3CR3CR3CR3CR3 HHIHHIHHIHHIHHI CR3CR3CR3CR3CR3 HHIHHIHHIHHIHHI CR3CR3CR3CR3CR3 HHIHHIHHIHHIHHI CR3CR3CR3CR3CR3 HHIHHIHHIHHIHHIY e a r


Compared to large banks, the medium-sized and small banks have a much lower degree

of concentration. They have a better market share distribution, as the biggest three banks in

the group only controlled less than 35 percent of the market share. Analogous with the large

banks, the concentration level in medium-sized banks and small banks has also reduced. But if

we compare the value of the concentration changes, the changes of the concentration level in

medium-sized banks and small banks were relatively smaller, especially if we refer to Herfindahl-

Hirschman index. The information of the market concentration and competition in the medium-

sized banks and small banks helps in understanding this phenomenon. The medium-sized and

small banks markets were highly competitive and less concentrated, therefore the mergers and

acquisitions just slightly reduced the market concentration.

There is a consistency between market concentration and competition in the three different

groups. Larger banks have the lowest degree of competition because the market is more

concentrated (CR3average

=0.69; HHI average

=0.20). On the other hand, the market of medium-sized

banks is the most competitive because it is less concentrated (CR3average

=0.26; HHI average

=0.05).

Small banks also work in a quite competitive environment because each bank has a small share

of the total market (CR3average

=0.31; HHI average

=0.05).

Table 11. Empirical Results using Fixed Effect Method with Interaction Variables(All Banks between 2001 and 2009)

Rincian

H (Competition Level)

∆H (After 2004 Consolidation Policy)

Significance test on ∆H , F test¡

Test of change in H

Market Structure

N (Number of Observation)

0.68

0.027

3.5074240.0615*

Can not reject the increasing of H-statistics

Monopolistic Competition

892

Note: *** significant at 99% confidence level; ** significant at 95% confidence level; * Significant at 90% confidence level

The outcome of fixed effect model using the interaction variables for all banks implies

that the second wave of consolidation through the introduction of Indonesia Banking Architecture

increased the value of H-statistics. It increases about 0.027 from 0.68 in the first period to 0.70

in the second period. The significance test by using F test shows that the increase on the level

of competition is significant at 90 percent of confidence level. The result confirms that the

consolidation policies enhanced the competition in the Indonesia banking industry. Consolidation


policies were effective to encourage medium-sized and small banks to merge in order to comply

with the Minimum Capital Requirement or Single Presence Policies. It significantly increased

the scale of economies of the merging banks and equipped them with bigger capacity to

compete with other banks. Further the mergers of smaller banks improved the distribution of

the market share. The improvement of the market share distribution reduced the market

concentration and enhanced the competition.

V. CONCLUSION

There are some interesting findings of this study. First, the data implies that the structure

of Indonesia banking market is vulnerable. Comparing to other East Asian countries, the number

of bank in Indonesia is larger. However, the market is concentrated in a few banks. Large banks

control a substantial share of the market. On the other hand, more than a half of banks are

small with the equity less than 1 trillion rupiah. In addition, the market concentration in the

large banks group is much higher than in smaller banks.

Secondly, banking market has become less concentrated during the implementation of

consolidation policy. This finding showed that the consolidation policy in Indonesia has similar

pattern with those implemented in other Asian countries. On the contrary, consolidation policy

in Latin American countries and other developed countries increased the market concentration.

As described by Gelos and Roldos (2002), the consolidation policy that was driven by market

will increase the concentration. On the other hand, the consolidation policy in Indonesia was

driven by the authority through the banks liquidation, mergers of the State Owned Bank and

the implementation of the Indonesia Banking Architecture. They have effectively encouraged

medium-sized and small banks to consolidate. This improved the distribution of market share

and reduced the market concentration.

Thirdly, during the implementation of consolidation policy the banking industry was

working under monopolistic competition. The H-statistic was .69. If we compare to the value

of H-statistic within 1994-2001 as calculated by Claessen and Laven (2003), the competition

has risen over time. The competition analysis by sub-groups shows that large banks were working

in the least competitive market while the medium-sized banks are working under the most

competitive market. The nature of competition in Indonesia banking was different with developed

countries where large banks were more competitive (Bikker and Haaf, 2002).

Next finding is discussing the impact of consolidation policy on competition. The estimation

shows that banking industry was more competitive in the second period of the implementation


of the consolidation policy. The increasing of the economies scale of the merging banks and

the improvement of market share distribution enhanced the competition.

Finally the study demonstrates that a concentrated market contributes to a less competitive

environment. It is possibly the reason why the large banks in Indonesia are working in a less

competitive market rather than the smaller banks. Large banks possess monopoly power to

enable them to behave as monopolist or oligopolies.


REFERENCES

Bain, Joe E. (1951). Relation of Profit Rate to Industri Competition: American Manufacturing,

1936 √ 1940. The Quarterly Journal of Economics 65.hal. 293-324

Bain, Joe E. (1956). Barriers to New Competition. Cambridge. Mass: Harvard University Press.

Berger, Allen N., and Timothy H. Hannan. (1989). The Price-Concentration Relationship in

Banking. Review of Economics and Statistics 71.Hal.291-99.

Bikker, Jacob A., and Katharina Haaf. (2002). Competition, concentration, and their relationship:

An empirical analysis of the banking industri. Journal of Banking and Finance 26.hal. 2191-

2214.

Boumol, William, John C. Panzar, and Robert D. Willig.(1982). Contestable Markets and the

Theory of Industri Structure. San Diego: Harcourt Brace Jovanovich.

Buletin Bisnis. (2007). Daftar Akuisis dan Merger Bank di Indonesia. Retrieved from website:

http://buletinbisnis.wordpress.com/2007/12/11/daftar-akuisisi-dan-merger-bank-di-

indonesia/.

Calem, Paul, and Gerald Carlino. (1991). The Concentration/ Conduct Relationship in Bank

Deposit Market. Review of Economics and Statistics 73.hal. 268-76.

Canoy, Marcel, Machiel van Dijk, and Jan Lemmen. (2001).

Competition and Stability in Banking.CPB Netherland Bureau for Economic Analysis.Netherland.

Bank Sentral Indonesia.(2010). Arsitektur Perbankan Indonesia.dikutip dari website:

www.bi.go.id.

Aturan Bank Sentral IndonesiaRepublik IndonesiaNomor 10/15/PBI/2008 TentangKebutuhan

Modal Minimum

Aturan Bank Sentral Indonesia Republik Indonesia Nomor 8/16/PBI/2006 TentangKebijakan

Single Presence.

Central Bank of Indonesia.(2009). Indonesia Banking Statistics 2009.

Chan, Donald, Christopher Schumacher, and David Tripe.(2007). Bank Competition in New

Zealand and Australia.Finsia_MCFS.The Melbourne Center for Financial Studies. Retrieved

from website: http://www.melbournecentre.com.au/

Chua, H.B. (2003). FDI in financial sector: the experience of ASEAN countries over the last

decade, in CGFS (2004), the Central Bank paper submitted by Working Group members.

Retrieved from website: www.bis.org/publ/cgfs22mas.pdf


Claessens, Stijn, Asli Demirhuc-Kunt, and Harry Huizinga.(1998). How Does Foreign Entry

Affect the Domestic Banking Market.Policy Research Working Paper.The World Bank.

Claessens, Stijn, and Luc, Laeven. (2004). What Drives Bank Competition? Some International

Evidence.World Bank Policy Research Paper.No.3113, August.

De Bandt, O., and Davis, E.P., (2000). Competition, contestability and market structure in

European banking sectors on the eve of EMU. Journal of Banking and Finance 24, 1045√

1066.

Demsetz, Harold. 1973. Industry Structure, Market Rivalry, and Public Policy. Journal of Law

and Economics 16 (April).hal. 1-10.

Gelos, Gaston., R and Jorge Roldos. (2002). Consolidation and Market Structure in Emerging

Market Banking Systems.IMF Working Paper WP/02/186.

Legislation number 5/ 1999 Regarding the Indonesian Competition Regulation

Legislation of Republic of Indonesia Number 7/1998 Regarding Banking Sector.

McLeod, Ross. (1999). Control and competition: Banking deregulation and re-regulation in

Indonesia. Journal of the Asia Pacific Economy, No 4(2), p. 258-297.

Molyneux, Phillip, John Thornton, and D. Michael Llyod-Williams.(1991). Competition and

Market Contestability in Japanese Commercial Banking.Mimeo. Washington, D,C.

International Monetary Fund.

Molyneux, Phillip, John Thornton, and D. Michael Llyod-Williams.(1994). Competition condition

in European Banking.Journal of Banking and Finance 18.hal.445-459.

Nathan, Ali, and Edwin H. Neave. (1989). Competition and Contestability in Canada»s Financial

System: Empirical Results. Canadian Journal of Economics 22 (June).Hal.576-94.

Nueberger, Doris. (1997). Structure, Conduct and Performance in Banking Markets.University

Rostock, Working Paper No. 12.

Panzer, John C., and James N. Ross. (1987). Testing for ≈Monopoly Equilibrium∆. The Journal

of Industrial Economics.hal.443-456.

Pelzman, Samuel. (1977). The Gains and Losses from Industrial Concentration. Journal of Law

and Economics 20.hal. 229-63.

Shaffer, Sherrill. (1982). A Nonstructural Test for Competition in Financial Markets. Proceedings

of a Conference on Bank Structure and Competition (Federal Reserve Banks of Chicago)

on May. hal.225-43.

Shaffer, Sherrill. (1994a). Bank Competition in Concentrated Market.Business Review.March/

April.

Shaffer, Sherrill. (1994b). Structure, Conduct, Performance, and Welfare.Review of Industrial

Organization 9.hal. 435-450.


Tirole, J. (1987). The Theory of Industrial Organization.MIT Press, Cambridge, MA.

Vesala. (1995). Testing Competition in Banking: Behavioral Evidence from Finland. Bank of

Finland Studies Working Paper No. E1.

Yeyati, Eduardo, Levy and Micco, Alejandro. (2003a). Concentration and Foreign Penetration

in Latin American Banking Sectors: Impact on Competition and Risk. Inter-American

Development Bank.Retrieved from website: http://www.oecd.org/dataoecd.

Yeyati, Eduardo Levy and Micco Alejandro.(2003b). Banking Competition in Latin America.

Latin-American Competition Forum. Retrived from website: http://www.oecd.org/dataoecd.

Yeyati, Eduardo, Levy and Micco, Alejandro. (2007). Concentration and Foreign Penetration in

Latin American Banking Sectors: Impact on Competition and Risk. Journal of Banking and

Finance 31.hal. 1633-1647.


Appendix A.Alternative Specification, Empirical Results Polled Regression Model of Fixed Effect Method,

Total Revenue as Dependent Variable (All Banks between 2001 and 2009)

Funding rate

Wage rate

Capital price

Risk of loans

Risk of capital

Deposit Mix -

Demand Deposit

Deposit Mix -

Interbank Deposit

Other Income

Number of Observation

R2

H (Competition)H (Competition)H (Competition)H (Competition)H (Competition)


a) the value of F test shows that H is not significantly different from 0 and 1 (level confidence 99%) The value in the bracket is the t-statistic.*** significant at 99% confidence level;** significant at 95% confidence level;* significant at 90% confidence level

0.43***

(35.05)

0.17***

(10.30)

0.067***

(5.32)

0.39***

(13.34)

0.037***

(3.10)

0.03***

(2.61)

-0.005***

(-2.52)

0.12***

(13.19)

892

0.87

0.660.660.660.660.66aaaaa

0.43***

(19.02)

0.065***

(2.13)

0.01

(0.58)

0.08

(0.66)

0.16***

(2.69)

0.04

(1.22)

-0.022***

(-2.66)

0.16***

(7.85)

63

0.94

0.510.510.510.510.51aaaaa

0.40***

(23.65)

0.21***

(8.91)

0.04***

(2.60)

0.41***

(10.50)

0.09***

(4.00)

0.001

(0.08)

-0.008***

(-2.55)

0.15***

(11.90)

309

0.93

0.650.650.650.650.65aaaaa

0.46***

(29.85)

0.14***

(6.93)

0.05***

(3.01)

0.41***

(10.21)

0.02

(1.52)

0.02

(1.63)

-0.003

(-1.26)

0.08***

(7.48)

520

0.86

0.660.660.660.660.66aaaaa


Appendix B.The Equilibrium Test √ Return on Asset

as dependent variable

All Banks

H = 0 cannot be rejected (level confidence 99.9%)The value in the bracket is the t-statistic.

-0.067

(-1.55)

0.16

(2.63)

-0.074

(-1.52)

-0.70

(-6.16)

0.004

(0.095)

0.093

(2.06)

-0.028

(-3.26)

0.19

(5.61)

862

0.75

0.0240.0240.0240.0240.024aaaaa

0.0720.0720.0720.0720.072

0.7890.7890.7890.7890.789

Funding rate

Wage rate

Capital price

Risk of loans

Risk of capital

Deposit Mix √ Demand Deposit

Deposit Mix √ Interbank Deposit

Other Income


R2

H (competition)H (competition)H (competition)H (competition)H (competition)

Equilibrium testEquilibrium testEquilibrium testEquilibrium testEquilibrium test

F testF testF testF testF test

ρρρρρ (1,742) (1,742) (1,742) (1,742) (1,742)


Appendix D.Empirical Results of Polled Regression Model of Fixed Effect with Time Dummies

All Banks between 2001 and 2009


a) the value of F test shows that H is not significantly different from 0 and 1 (level confidence 99%)The value in the bracket is the t-statistic.*** significant at 99% confidence level;** significant at 95% confidence level;* significant at 90% confidence level

0.43***

(24.73)

0.23***

(12.62)

0.028***

(2.1)

0.42***

(12.70)

0.016

(1.25)

0.0009

(0.069)

0.00045

(0.178)

-0.0004

(0.177)

892

0.90

0.690.690.690.690.69aaaaa

0.53***

(9.62)

0.033

(0.61)

-0.0025

(-0.063)

0.0008

(0.0061)

0.21***

(2.30)

0.03

(0.77)

-0.028***

(-2.75)

0.064

(1.66)

63

0.95

0.550.550.550.550.55aaaaa

0.41***

(17.87)

0.28***

(9.96)

0.06***

(3.34)

0.40***

(9.51)

0.19***

(6.73)

-0.046

(-1.94)

0.0005

(0.124)

-0.042***

(-2.82)

309

0.94

0.760.760.760.760.76aaaaa

0.48***

(17.61)

0.18***

(7.18)

0.015***

(0.76)

0.46***

(9.15)

-0.024

(-1.56)

0.011

(0.62)

0.002

(0.70)

-0.015

(0.70)

520

0.87

0.680.680.680.680.68aaaaa

Appendix C.Redundant Fixed Effects Tests

F-test

df (degree of freedom)

x2 statistics

df (degree of freedom)

Number of Observation


H = 0 is (level confidence 99.9%)

13.73

(112;771)

978.28

(112)

892

9.42

(6;48)

49.04

(6)

63

20.27

(35;265)

NA

(NA)

309

9.42

(69;442)

470.23

(69)

520

Funding rate

Wage rate

Capital price

Risk of loans

Risk of capital

Deposit Mix √

Demand Deposit

Deposit Mix √

Interbank Deposit

Other Income


R2

H (competition)H (competition)H (competition)H (competition)H (competition)



177Relative Effectiveness of Indonesian Policy Choices During Financial Crisis

RELATIVE EFFECTIVENESS OF INDONESIAN POLICY CHOICESDURING FINANCIAL CRISIS1

Tumpak Silalahi2 and Tevy Chawwa3

The objective of this paper is to review the impact of crisis and policy measures taken during the

crisis, to evaluate the effectiveness of those measures and to analyze the exit strategy in Indonesia. The

econometric model was used to evaluate the impact of monetary and fiscal policy to economic output

using quarterly data from 1990 - 2010. The result shows that monetary and fiscal policies have significant

impact to economic output. In the short run the changes in real GDP is significantly affected by changes

in real monetary supply in the previous three quarter and real fiscal expenditures. The lesson learned from

this research among other are that cooperation and coordination among the policy makers and the timely

responses are very important in tackling the crisis; an effective conventional monetary policy in normal

times may become less effective in a crisis thus unconventional monetary policy indeed necessary as

timely policy response and the improvement for more timely disbursement of government expenditure is

important to increase the effectiveness of this policy to stimulate economic output. Moreover, several

Indonesian exit strategy and policies to face future challenges are very important to reach the ultimate

objective of sustainable economic growth while maintaining macroeconomic stability.

1 This is a collaborative research project with several central banks coordinated by SEACEN Research and Training Centre.2 Associate Senior Economist at Directorate of Economic Research and Monetary Policy - Bank Indonesia, email : [email protected] Junior Economist at Directorate of Economic Research and Monetary Policy - Bank Indonesia, email : [email protected] . The views

expressed in this paper are those of the authors and do not reflect the views of Bank Indonesia

Abstract

JEL Classification: E52, E62, E63

Keywords: monetary policy, fiscal policy, financial sector policy, global financial crisis.


I. INTRODUCTION

Prior to the global crisis, particularly during 2007, optimistic expectation about Indonesian

economy was a general mindset of the economic forecasters. This expectation was supported

by various macroeconomic indicators which showed a remarkable achievement of the Indonesian

economy in 2007 after the Asian crisis in 1997. Indonesian GDP growth trend was increasing

continuously since 2005 and for the first time after crisis, the GDP growth reaches more than

6% in 2007. Basically, the growth in 2007 was driven by robust domestic consumption and

external demand which led to a surplus in the current account. In addition, positive sentiment

accompanied by attractive yield on rupiah portfolio investment helped encouraged capital

inflows. Shifting funds into emerging market assets contributed to a positive appreciation of

currency in the region. The surplus in current account and the rise of portfolio capital inflows in

2007 had increased Indonesian foreign currency reserves to 13% of GDP or sufficient to cover

import of goods and services for an average of seven months.

Indonesia financial market and institutions were also in a stronger condition based on

financial indicators. Lessons learned from the 1997 crisis resulted in a fairly strict implementation

of prudential regulations in the corporate and banking sectors as a result of which led Indonesian

banking industry became much sounder with a more robust foundation to absorb various

shocks in the economy. Demand pressure in 2007 was relatively high indicated by positive

output gap, although it was still far below output gap in 1996. Moreover, the government had

tried to reduce its dependency on foreign debts, both short-term and long-term. All of these

improvements resulted in Indonesia been assessed as a low risk country and it achieved the

highest ICRG (International Country Risk Guides) scores since 1997.

In 2007, fiscal policy was targeted at maintaining price stability for energy and staple

needs, while also delivering an economic stimulus. Escalating world oil prices in combination

with below-target lifting of domestic oil led to a considerable pressure in the government

budget deficit. In the monetary policy, BI»s stance could be divided into 2 periods, the period of

decline in the BI Rate (January-July 2007) and the period of flat movement in the policy rate

(August-November 2007). Bank Indonesia also applied a flexible exchange rate policy, allowing

the rupiah to move in line with economic fundamentals. To manage the volatility in the rupiah,

Bank Indonesia conducted foreign exchange market interventions on a limited scale. Although

many improvements such as better monetary and fiscal coordination were introduced to

strengthen the effectiveness of policy choices in Indonesia, the disturbances from external

shocks such as the rise in commodity/oil price and internal shocks such as crop failures or

seasonal events were important factors that influenced the Indonesian macroeconomic

conditions.


The global financial crisis in 2008 had reversed the previous optimistic mindset of economic

forecasters. The pressures in the global liquidity had caused a massive short term portfolio

capital outflow followed by a decline of Indonesia»s financial market performances. In the real

sectors, reflecting the input of global slowdown, exports declined and it had an indirect impact

on household and private sector»s income, leading to a decline in Indonesia»s consumption and

investment. As a result, Indonesia»s GDP growth»in 2009 declined to 4.5% (yoy).

Some policy measures were implemented in monetary, fiscal and financial sectors to deal

with the global financial crisis. Bank Indonesia had implemented an accommodative monetary

policy in order to keep a moderate growth achieve at least by maintaining financial markets

liquid which was facilitated by relatively low inflation. The policy rate was brought down in

December 2008 with the intention to decrease banks» lending rates. Some unconventional

monetary policy measures such as narrowing the interest rate corridor for standing deposit and

lending facility had also been taken to address liquidity issues. On the fiscal side, the government

provides policy response to keep domestic demand by several fiscal stimulus and trade policies.

There were also coordination between Ministry of Finance, Central bank and other institutions

in order to maintain financial and macroeconomic stability.

With this background, this paper is aimed at reviewing the policy measures taken during

the crisis, evaluating their effectiveness and analyzing the exit strategy to reach the ultimate

objective of sustainable economic growth while maintaining macroeconomic stability in

Indonesia. In turn this is expected to make a contribution to a comprehensive evaluation of the

effectiveness of the policy measures in SEACEN Economies. To handle the broad issues in the

paper, two methodologies are adopted: firstly, by descriptive analysis using simple statistics and

graphics; and secondly, by econometric model to analyze the relative effectiveness of policy

choices.Second session of this paper discusses the theory and the empirics of the crisis, session

three discuss methodology while session four present result and analysis. Session five will give

conclusion and close the presentation.

II. THEORY

As a small open economy, Indonesia could not be immune from external shock impact.

The integration in financial sector has left many countries particularly for open economy to

contagion risk. An empirical study by Santoso et al (2009) shows that Indonesia has a contagion

relationship with several countries in Asia, such as Japan, Taiwan, Korea, Hong Kong and India.

The domestic financial market moves closely with the movements in global financial markets.

The research also showed that there is no direct contagion between Indonesia stock exchange


and the Dow Jones index and NASDAQ index. Thus, if Indonesia is affected by the global crisis,

it is not a direct effect from the US market but rather indirect effects from capital markets in

Asia that share a direct relationship with the US capital market. Furthermore, the research

indicated that Indonesia is a shock absorber rather than shock transmitter, particularly with

regard to developed countries (Japan, Australia, Germany, United Kingdom and the US).

In the real sector, Indonesian export is also affected by external condition. Research by

Kurniati»s et al (2008) study show that Indonesia»s exports are most sensitive to the economic

growth of Singapore (1.19), followed by the US (0.84), Japan (0.81) and China (0.3).

Related to the current global financial crisis, Kurniati and Permata (2009) found that a

shock in global risk aversion have immediate negative impact on capital inflows to Indonesia,

particularly from portfolio investments that leading to rupiah depreciation. The impacts through

financial channel on financial variables are temporary and relatively faster to recover (self-market

correction). The second round effects of global crisis occur through trade channel. Negative

shock of US GDP growth leads to contemporaneous decline in Indonesia»s exports which

subsequently result in decreasing domestic real GDP growth, capital outflow and rupiah

depreciation. The impacts on exports seem to persist and need policy responses from the

authorities.

The effectiveness of monetary policy depends on the environment of domestic economy

and the disturbance from external shock. Study done by Arifin (1998) which analyze effectiveness

of interest rate policy for rupiah stabilization during crisis 1997/1998 in Indonesia conclude that

interest rate policy is effective for rupiah stabilization only if there are no disturbance from

other non-economic factor, such as negative rumors, mass mobilization and riots. Thus, interest

rate policy becomes less effective for reducing inflation level because inflation is also affected

by supply factor pressures such as production and distribution.

The current research done by Simorangkir and Adamanti (2010) examines the impacts of

fiscal stimulus and interest rate cut on Indonesian economy during global financial crisis using

Financial Computable General Equilibrium (FCGE) approach. The simulation results showed

that the combination of fiscal expansion and monetary expansion boosts economic growth of

Indonesia effectively. Relative to the effectiveness of fiscal expansion without monetary policy

expansion or monetary expansion without fiscal expansion, the combination of those two policies

is more effective. Another result of this paper showed that looking into the components of

GDP, the combination of fiscal and monetary expansion has a large multiplier effect, boosting

aggregate demand through increasing consumption, investment, government expenditure,

exports and imports. Meanwhile, from production side, the combination of fiscal and monetary


expansion has positive effects on increasing production of all economic sectors. This effect

comes from fiscal incentive (lower tax, lower import duties, etc) in increasing investment.

Moreover, the increase in aggregate demand also encourages enterprises to increase their

production. This paper also found that institutionally fiscal stimulus and monetary easing has

increased income and purchasing power of the poor and rich households in rural and urban

area. This increase in turn results in higher all household consumption.

Several other research related to the effectiveness of fiscal policy during crisis noted that

the important thing to concern when government gives fiscal stimulus are: (i) the coordination

between central and local government and (ii) the ability of local and central government to

disbursing the money quickly. In order to reach this, the complicated procurement processes

and administration practices will need to be simplified and made more transparent.

Survey done by The SMERU Research Institute in 2009 investigating the roles of Social

Safety Net programs and Fiscal Stimulus Program (especially infrastructure and labor intensive

program) in mitigating the impact of global financial crisis (GFC) found that:

- There are several problem related with the implementation of social safety net programs

that disturbing its effectiveness: (i) most local government officials lack of specific information

on the impact of GFC to the community, (ii) no information system on the socioeconomic

condition of community available and provided in hierarchical, periodic & systematic way,

(iii) lack of responsiveness on prices changes or other signs of crisis.

- Related with fiscal stimulus programs (FSP), several problems happened are: (i) Central

government does not formally socialize FSP to the local governments, (ii) There is no linkage

between the level of impact of GFC and the fund allocation to the regions, (iii) There is no

linkage between the impacted sector and the funded sector projects, (iv) There is important

role of Central MP to inform local governments on the availability of projects to be funded,

(v) Fund allocated to the regions also depend on the active role of local governments efforts

to lobby Central Government.

These problems also pointed in Yudo et al (2009) analysis about social safety net program,

known as JPS (Jaring Pengaman Sosial) that Indonesian government introduced to response to

the 1997/1998 crisis. An evaluation of this anti-poverty program which encompassed food

security, employment and income maintenance, and preservation of access to education and

health showed that in many cases the target groups had largely been missed and that the

effectiveness of the efforts varied across programs and regions. This suggested that some districts

were better than others in implementing common national program, further highlighting the

need for large improvements in program implementation, in particular in targeting the


beneficiaries of a particular program and raising coverage within the target groups. Further

research reaffirmed the earlier conclusion on the heterogeneity of performance both across

programs and regions. This could be due to programmed design, budget allocation across

programs and regions, and regional capacity to implement the program. One main lesson to be

drawn from these assessments is that targeting requires detailed administrative guidance as

well as community involvement if it is to be both effective and socially and politically acceptable.

Furthermore, static administrative targeting was unable to capture the newly poor or shocked

households.

As a part of emerging market economies, Indonesia was impacted by the global financial

crisis which resulted in a sudden stop in capital inflows into emerging market countries and a

decline in global economic growth. The impact on macroeconomic indicators can be grouped

into the first and second round effects as follows:

2.1. First Round Effects of Global Financial Crisis

2.1.1 Impact on BOP and Exchange Rate Movements

During quarter III 2008, global economic developments placed pressure on the Indonesia»s

balance of payments. The pessimistic outlook for the global economy in 2008 signaled by

international institutions reinforced pessimism among market actors. Investors saw gloomier

prospects and higher risk in fund placements in emerging market, including Indonesia. The

high risk perceptions of Indonesian market could be seen in indicators such as Credit Default

Swap (CDS) and Government Bond Yield that increased significantly (Figure 1 and 2). Then, in

order to avert risk, the investors moved their funds to the safe haven of US Treasuries.

Figure 1.CDS (Credit Default Swap)

Figure 2.Government Bond Yield

0

200

400

600

800

1000

1200

1400

2008 2009 2010

Jan Feb Mar Apr MayJun Jul Aug Sep Okt Nov Des Jan Feb Mar Apr MayJun Jul Aug Sep Okt Nov Des Jan Feb Mar Apr MayJun

Source : Bloomberg

5

10

15

20

25

1 YR

5 YR

10 YR

2008 2009 2010

Jan Feb Mar Apr MayJun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr MayJun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr MayJun

Source : Bloomberg


A negative sentiment caused by the turbulence in global financial markets prompted a

wave of capital outflows as seen in Figure 3. At that time, foreign investors cut back their

holding of Indonesian government securities by Rp 4.4 trillion or approximately US$ 387 million.

The foreign investors» behavior to terminate their portfolio investment then followed by domestic

investors drawing their assets and these actions brought the portfolio investment in quarter

IV-2008 to a recorded net outflow. Further, the domestic agents moved their accounts from

domestic banks to overseas banks and some of them had failed to get new foreign financing

as indicated by the other investment component that recorded a deficit. The increased deficit

of other investment also explained by higher drawing on corporate lines of credit spurred by

heavy corporate foreign exchange demand to pay for imports in 2008. In contrast, direct

investment still recorded as net surplus due to acquisition of activities of domestic banks by

foreign investors.

Global financial crisis also weakened the performance of Indonesian current account in

the quarter II until quarter IV 2008 (Figure 4). This escalation in the current account deficit

resulted primarily from falling exports as global economy contracted and the falling of export

commodity prices. Current transfers buoyed by worker remittances were also decreasing although

remain positive. In 2008, incoming transfers from worker remittances generated a surplus of

US$ 5.2 billion then decreased to US$ 4.8 billion in 2009.

The downturn in the BOP in turn triggered a strong exchange rate depreciation

accompanied by high volatility. Demand pressure for foreign currency which was derived from

foreign portfolio capital outflow and the drop in foreign currency supply because of the collapse

Figure 3.Financial /Capital Account

Figure 4.Current Account

bn US$

-6.0

-4.0

-2.0

0.0

2.0

4.0

6.0

Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun2008 2009 2010

Portfolio Investment Direct Investment

Other Investment Capital Account Balance (Quarterly)

Source : Bank Indonesia

bn US$

2009 :10.75

2008 :0.13

-2.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0CA Balance (Quarterly)

CA Balance (Annual)

Mar Jun Sep2008

Dec Mar Jun Sep2009

Dec Mar Jun2010



of export led to heavy depreciating pressure on the exchange rate. Downward pressure was

also sustained by regional currencies that weakened from the spillover effects of external

turbulence. Additionally, the side effect of Bank Indonesia»s policy in lowering reserve requirement

ratio may also gave effects to the excess rupiah liquidity in the market and led to the depreciate

rupiah values decreased. These developments placed to put pressure on the rupiah, fell to its

lowest level of Rp 12,400 per US$ in November 2008 (Figure 5).

Bank Indonesia responded by issuing a series of policies to ease the pressure and prevent

excessive volatility in the rupiah such as exchange rate stabilization, policy governing the purchase

of foreign currency through banks, policy governing the transaction of foreign currency against

rupiah and prohibition of structured product transaction. These policies will be discussed in

detail in the next section.

2.1.2 Impact on Stock Market

The investor»s behavior to withdraw their funds from emerging countries during global

crisis led to fall in the stock market index of emerging market including Indonesia. The falling of

mining and agricultural commodity prices on the world market also affected the stock market

index adversely. The Indonesian Composite Index (IDX) continued to decrease sharply and closed

at 1.355 at the end of the period 2008, a drop of 50.64% compared to the quarter II 2008

(Figure 6). With this worst performance, Indonesian Stock Exchange was placed at level 5 in

Asia and the Pacific, after Vietnam, Shanghai, Shenzhen and Mumbai.

Figure 5.Exchange Rate Movements

November :Policy governing thepurchase of foreign

currency through bank

8000

8500

9000

9500

10000

10500

11000

11500

12000

12500

13000October :

- Several exchange ratestabilization policy : RR

Ratio, Daily BalancePosition, FX Swap tenor,

Foreign currency provision-Policy governing

lowering rupiah RR ratio

December :Policy governing the

transaction of foreigncurrency against

rupiah & prohibition ofstructured product

transaction

2008


Jan Feb MarApr May Jun Jul AugSep Oct Nov Dec

2009Jan Feb MarApr May Jun Jul AugSep Oct Nov Dec Jan Feb MarApr May Jun

2010


The disappointing performance of the IDX has led to a fall in both the volume and value

of transactions in the stock market especially in the fourth quarter of 2008 (Figure 7). Number

of firms who already held principle permits to issue shares also decided to postpone their share

issuances.

To avoid further fall on the stock market performances, government decided to suspended

IDX trade on 9th and 10th October 2008, issued regulations regarding buyback, banned short

selling and limited margin trade. These policies were intended to provide time for investors to

think rationally amidst the financial market turbulence caused by the crisis.

2.1.3 Impact on Market Liquidity

The global financial pressures also led to a liquidity shortage in the domestic money

market, which was reflected in a slower pace of growth in narrow money (M1) and broad

money (M2) (Figure 8). In quarter IV-2008 average M1 and M2 growth slower at 1.5% (yoy)

and 14.9% (yoy), decline from 19.9% (yoy) and 17.2% (yoy) in the preceding quarter.

Strong perception of tight bank liquidity and spillover from global conditions was also

reflected in the rising liquidity premium, which widened progressively for longer tenors. With

the tightened conditions on the money market, some banks who customarily supply liquidity,

reviewed their credit lines and credit limits to individual counterparties. This resulted in more

uneven distribution of liquidity on the market tending towards greater segmentation, due to

loss of confidence in transaction. The movement in the overnight interbank rate remained

above the policy rate (BI Rate) during July-October 2008, alongside drastically reduced transaction

Figure 6.Jakarta Stock Index

Figure 7.Trading Volume & Value

1000

1500

2000

2500

3000

3500

Source : CEIC

2008 2009 2010Jan FebMarAprMayJun Jul AugSepOct NovDecJan FebMarAprMayJun Jul AugSepOct NovDecJan FebMarAprMayJun

0

50,000

100,000

150,000

200,000

250,000

300,000Trading Volume (mn)

Trading Value (IDR bn)

2008

Jan FebMarApr MayJun Jul AugSepOct NovDec

2009

Jan FebMarApr MayJun Jul AugSepOct NovDec

2010

Jan FebMarApr MayJun

Source : CEIC


volume and a growing spread between the highest and the lowest overnight rate (Figure 9).

Following this, banks with long positions on the money market chose to shift their liquidity to

Bank Indonesia short term investment.

2.1.4 Impact on Financial Institutions

The impact of financial crisis on Indonesian financial institutions was not as severe as in

other countries because Indonesian banks and national financial institutions» exposure to

subprime mortgages was minimal. One of the contributory factors in this regard was the

characteristics of Indonesian banks and financial institutions, which still leaned towards

conventional instruments of investment. Another factor was the quality of surveillance in the

banking sector and non-bank financial institutions as well as the capital market which has been

improved. The lessons from Asian crisis 1997 had caused Bank Indonesia to strengthen Indonesian

banking structure under Indonesian Banking Architecture as a part of financial landscape. Beside,

the banks were well disciplined in following prudent regulations which led to limit their exposure

to the larger problems associated with derivative products.

However, the tight liquidity in the money market had made it difficult for banks to manage

their fund. On October 2008, three state banks proposed a liquidity support from the government

in amount of Rp 15 trillion in total (approximately US$ 1.36 billion). Medium and smaller banks

also had more severe problems as depositors moved their fund to bigger banks because they

worried of possible bank liquidations as experienced in 1997. Under the circumstances, banks

competed to attract deposits by offering high deposit interest rates. In turn, lending rates were

Figure 8.M1 and M2 Growth

Figure 9.Daily Interbank Market Volume

%

0.0

5.0

10.0

15.0

20.0

25.0

2008 2009 2010Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun

Source : IFS-IMF

M2 Growth (yoy)

M1 Growth (yoy)

Billion of Rp

0

5000

10000

15000

20000

25000

30000

35000

40000

Jan FebMarAprMayJun Jul Ags SepOct NovDecJan FebMarAprMayJun Jul AugSepOct NovDecJan FebMarAprMayJun

2008 2009 2010



also increased. These conditions led to a decrease of the banks» performance as indicated by

Capital Adequacy Ratio and Non Performing Loans (Figure 10).

2.2 Second Round Effects of Global Financial Crisis

2.2.1 Impact on Exports

The weakening global demand and collapsing world commodity prices had deteriorated

Indonesia»s export earnings significantly, especially in quarter I 2009 (Figure 11). The growth of

exports dropped drastically to -18.73% yoy from 13.64% yoy in the same quarter 2008 (Figure

..

Figure 10.Commercial Bank Performances

%%

2.5

2.7

2.9

3.1

3.3

3.5

3.7

3.9

4.1

Source : CEIC

14

15

16

17

18

19

20

21CAR (RHS)

NPL Ratio

2008 2009 2010Mar Jun Sep Des Mar Jun Sep Des Mar Jun

Figure 11.Export and Import

Figure 12.Export Growth (yoy)

bn US$

0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0ImportExport

2008 2009 2010Mar Jun Sep Dec Mar Jun Sep Dec Mar Jun

Source : IFS - IMF

-25

-20

-15

-10

-5

0

5

10

15

20

25

Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q22008 2009 2010

13.64 12.36 10.63

1.99

-18.73-15.52

-7.79

3.67

19.99

14.60

Source : IFS - IMF


12). Indonesia»s exports were concentrated in a select group of countries and relatively lacked

of diversification of export commodities and there was a higher degree of the vulnerability of

Indonesia»s export to external shocks.

Even though there was a shift in the main destination of Indonesia»s exports to China

since 5 years ago, but the Japan and the US markets still become the most destination of

Indonesian export. Prior to crisis in 2007, based on destination country, the proportion of

Indonesian export to Japan was 21.71% and to USA was 10.67% from total exports (Figure

13). The next main destinations were Singapore (9.65%), China (8.89%) and South Korea

(6.97%). The implication of this concentration is that any slowing of economic growth in the

major export destinations as experienced in the global crisis will have an adverse impact on

Indonesian exports.

Unlike other developing countries in Asia in which exports are dominated by electronic

appliances and office machinery, the structure of Indonesia»s export is dominated by oil and gas

as well as low technology industrial product. Accordingly, the primary sector especially oil,

natural gas, mining and agricultural commodities accounts for a substantial portion in Indonesia»s

export. This dependency on primary sectors made Indonesian export more vulnerable to external

shocks especially fluctuations in international commodity prices.

The growth in export volumes in all main export»s sectors decreased (Figure 14). Annual

growth of mining sector»s export decreased sharply from 47.83% in 2006 to 7.98% in 2007

and become -0.43% in 2008. The growth of agriculture and manufacturing sectors continued

to remain negative in 2009, but export in mining sector turned positive and recorded a growth

of 13.79%.

Figure 13.Indonesian Export Destinations in 2007

Figure 14.Growth of Export Volumes by Sectors

Japan, 21.71%

China,8,89%South Korea,

6,97%

Malaysia,4,68%

India,4,54%

Australia,3,12%

Thailand,2,81%

Netherlands,2,53%

Taiwan,2.39%

Germany,2,13%

Others, 19,92%

USA,10,67%

Singapore,9,65%

Agriculture, Hunting And Fishing

Mining And Quarrying

Manufacturing

%

-40

-30

-20

-10

0

10

20

30

40

50

60

2001 2002 2003 2004 2005 2006 2007 2008 2009



2.2.2 Impact on the Industrial Sector, Unemployment and Poverty

Tradable sectors were affected worst by the fall in exports during the crisis. During

quarter I 2008 until quarter III 2009, the growth of manufacturing sector declined from above

4 % to 1.5% (yoy) on average, in line with the deterioration in manufacturing products export.

Subsectors that impacted deeply in 2008 were other non-metallic products, chemicals products

and products of wood, which their export volumedeclined by 20 √ 30% from previous year. In

2009, volume of exports in machinery and motor vehicles decreased by 44% and 36% compared

to export in 2008 (Figure 15 and 16).

Figure 15. Volume of Export Decrease inManufacturing Subsector 2008

Figure 16. Volume of Export Decrease inManufacturing Subsector 2009

Other Non-Metallic Mineral Products

Chemicals and Chemical Products

Wood And of Products Of Wood And

Office, Accounting And Computing

Coke, Refined Petroleum Products

Furniture

Textiles

Basic Metals

Rubber And Plastics Products

%-35 -30 -25 -20 -15 -10 -5 0

-31%-22%

-21%

-14%

-13%

-9%

-8%

-8%

-6%


Machinery And Equipment N.E.C.Motor Vehicles, TrailersChemicals And Chemical ProductsOther Transport EquipmentWearing Apparel; Dressing And DyeingFabricated Metal ProductsBasic MetalsElectrical Machinery And ApparatusMedical, Precision And OpticalFurnitureTanning And Dressing Of LeatherPaper And Paper ProductsOther Non-Metallic Mineral ProductsTobacco Products

%-50 -40 -30 -20 -10 0

-44.25%-36.20%

-25.90%-25.07%

-21.66%-19.85%-19.40%

-11.50%-9.54%-9.03%

-6.30%-3.99%-2.77%-1.83%


The slowdown of tradable sector»s performance in turn led to a reduction in employment.

Pressure from the global crisis compelled several companies to make changes in their operations

and upgrade business efficiency. Consequently some factories were closed or began laying off

workers, driving down purchasing power even further in 2008. According to Ministry of

Table 1.Impact of Export Deterioration on Labor Absorption

Scenario

Source: Indonesia Economic Outlook, January 2008

Labor Absorption (%)

All Sector Industrial Sector

Export total decreased by 1%

Agricultural export decreased by 1 %

Mining export decreased by 1 %

Manufacture export decreased by 1 %

-0.166

-0.009

-0.005

-0.091

-0.416

-0.001

-0.002

-0.400


Manpower and Transmigration, total number of workers that had been temporarily laid off

was of 10,306 until December 2008. Bank Indonesia»s analysis using Indonesia Input Output

Table shows that for each 1% decline in Indonesian exports resulted in 0.42% reduction in

industrial employment (Table 1).

The fallout in the export dependent sectors from the global financial crisis seems to have

put considerable pressure on prosperity levels. The impact of reductions in working hoursand

dismissals in some industries caused many households to lose their income. Additionally, farmer»s

income in the estate sector began to suffer in October 2008, following the collapse in commodity

prices. The high inflationary pressure in 2008 has also produced a reduction in real wage levels

for workers. Fortunately, the relatively buoyant economic growth until QIII-2008 helped bring

improvement in various indicators of welfare such as poverty and unemployment. Government

program designed to combat unemployment, such as the National Community Empowerment

Program (PNPM) for block grants, disbursement of Grassroots Business Credit (KUR), the

Unemployment Reduction Movement and distribution of Direct Cash Transfers, also seems to

have some positive influence in the improvement of welfare indicators.

2.2.3 Impact on GDP Growth

The decline in exports, deterioration in production and lower income simultaneously

reflected in the decline of Indonesia»s economic growth from 6.3% in 2007 to 6.0 in 2008 and

4.55% in 2009. Compared to other countries in the world, Indonesian GDP performance during

crisis were relatively remarkable. The GDP growth in 2009 was the third highest in the world

Agriculture 6.44 4.81 3.25 5.12 4.83 5.91 2.95 3.29 4.61 4.13 3.00 3.08

Mining and Quarrying (1.62) -0.37 2.32 2.43 0.68 2.61 3.37 6.20 5.22 4.37 3.08 3.77

Manufacturing 4.28 4.23 4.31 1.85 3.66 1.50 1.53 1.28 4.16 2.11 3.70 4.29

Electricity, Gas and Water Supply 12.34 11.77 10.41 9.34 10.92 11.25 15.29 14.47 13.99 13.78 8.18 4.76

Construction 8.20 8.31 7.76 5.88 7.51 6.25 6.09 7.73 8.03 7.05 7.05 7.18

Trade, Hotels and Restaurant 6.75 7.68 7.59 5.47 6.87 0.63 -0.02 -0.23 4.17 1.14 9.36 9.63

Transportation and Communication 18.12 16.57 15.64 16.12 16.57 16.78 17.03 16.45 12.22 15.53 11.92 12.91

Financial, Rental and Business Services 8.34 8.66 8.60 7.42 8.24 6.26 5.33 4.90 3.77 5.05 5.34 6.10

Services 5.52 6.51 6.95 5.93 6.23 6.70 7.19 6.04 5.69 6.40 4.62 5.25

GDP 6.21 6.30 6.25 5.27 6.01 4.53 4.08 4.16 5.43 4.55 5.69 6.17

20102008 2009

Q1 Total Q1 Q2Q2 Q3 Q3 Q4 Total Q1 Q2Q4

Table 2.GDP Growth by Sector (% yoy)

Source: Bank Indonesia


after China and India. This achievement was supported by the performance of some sectors

not related to external developments such as electricity, gas and water utilities, construction,

transport and communications, and the services sector. Growth in the electricity, gas and water

utilities sector reached 13.78% and growth in transport and communications sector reached

15.53%, respectively (Table 2). On the demand side, economic expansion in 2009 was driven

by strong domestic demand, especially consumption of both households and government that

grew by 6.21% (Table 3).

2.3 Comparison between 1997/1998 Crisis and 2007/2008 Crisis

During the last ten years, Indonesia and most other Asian countries had experienced two

financial crises. The first, Asian crisis occurred in 1997 and the second crisis known as global

financial crisis occurred ten years later in 2008. In magnitude and breadth, there are similarities

between those crises. Therefore, computation of the duration, amplitude, slope and cumulative

losses of each of the two crises will show the big picture of those events.4

Compared to 1997/1998 crisis, the impact of the current crisis on the real sector was

relatively smaller. In 1997/1998 crisis, the GDP growth continued to decline for 8 quarters and

the amplitude of the decrease was -28.53%. In fact, the GDP growth was negative in 5 quarters.

Under the GDP growth declined by 2008/2009 crisis for 3 quarters and the amplitude was only

2.16%.

As indicated by the cumulative losses, the impact of the recent financial crisis on domestic

credit was smaller than that of the 1997/1998 crisis. The figure shows the cumulative losses

Table 3.GDP Growth by Demand Side

20102008 2009

Q1 Total Q1 Q2Q2 Q3 Q3 Q4 Total Q1 Q2Q4

Consumption 5.47 5.49 6.34 6.42 5.94 7.28 6.27 5.44 5.91 6.21 2.52 3.12

Gross Fixed Capital Formation 13.88 12.16 12.30 9.40 11.86 3.46 2.37 3.24 4.18 3.32 12.45 10.13

Export 13.64 12.36 10.63 1.99 9.53 -18.73 -15.52 -7.79 3.67 -9.70 19.99 14.60

Import 17.99 16.11 11.10 -3.73 10.00 -24.42 -21.04 -14.67 1.62 -14.97 22.60 17.74

GDP 6.21 6.30 6.25 5.27 6.01 4.53 4.08 4.16 5.43 4.55 5.69 6.17

Source: Bank Indonesia

4 The more explanation about similarities and differences of 1997 crisis and 2008 crisis could be seen in Appendix


Table 4. Duration, Period, Amplitude, Slopeand Cumulative Losses for Recent Crisis And 1997/98 Crisis

Source : CEIC, estimated

1997/1998 Crisis 2008/2009 Crisis

Economic Growth Recession (% yoy)Duration 8 quarters 3 quarters

Period Q1 1997 √ Q4 1998 Q4 2008 √ Q2 2009

Amplitude 28,53% 2,2%

Slope 4% 0,72%

Cummulative Losses 115,16% 4,86%

Domestic Credit Growth (% yoy)Duration 4 quarters 4 quarters

Period Q3 1998 √ Q2 1999 Q4 2008 √ Q3 2009

Amplitude 150,4% 25,2%

Slope 37,6% 6,3%

Cummulative Losses 370,8% 56,8%Stock Prices (Index)

Duration 5 quarters 4 quartersPeriod Q3 1997 √ Q3 1998 Q1 2008 √ Q4 2008

Amplitude 448 1390

Slope 90 348

Cummulative Losses 1411 2999

%

Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul NovMar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar Jul Nov Mar

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

-80.0

-60.0

-40.0

-20.0

0.0

20.0

40.0

60.0

80.0

100.0Jun-98, 90.5%

Jun-99, -59.9%

Sep-08, 34.8% Sep-099.6%

GDP Growth Indonesia (yoy)%

-20

-15

-10

-5

0

5

10

15

Dec-96, 10.28%

Dec-98, -18.26%

Dec-99, 5.36%

Jun-09, 4.08%

Jun-10, 6.17%Sep-08, 6.25%


1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 17b.Peak and Trough of Credit Growth

Figure 17a.Peak and Trough of GDP Growth

Source : CEIC, estimated Source : CEIC, estimated


consecutively by 370.8% and 56.8%. Even the duration of the credit decline was almost equal

at 4 quarters, the amplitude of 2008/2009 crisis was lower (25.2%) than that of 1997/1998

crisis (150.4%).

In the financial market, although it seems that the impact of recent crisis was bigger

because the stock market index decreased by 1390 points, in relative terms (percentage change)

it was lower (50,6%) than that of 1997 crisis (61,9%). The duration was also shorter.

III. METHODOLOGY

In examining the relative effectiveness of both monetary and fiscal policies on economic

growth, we use the Engle and Granger two step estimating procedure which allows an explicit

testing of co-integration and specification of the Error Correction Model (ECM).

Empirical model in this paper is aimed at testing the relationship of economic growth

with monetary policy, fiscal policy and other control variables.

Figure 17c.Peak and Trough of Stock Market Index



1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

500

1000

1500

2000

2500

3000

3500

Jun-97, 725

Sep-98, 276

Dec-07, 2,746

Dec-08, 1,355

(1)

where Y is a measure of economic activity, MP, a measure of monetary policy, FP measure of

fiscal policy and Z is other control variables that may affect economic activity. Finally, the general

form of the Error Correction Model (ECM) specification in this paper is:


(2)

Where:

Y = dependent variables (economic output)

X = independent variables, consists of monetary variables, fiscal variables, and other

control variables

ECM = residuals from long run relationship between variable

n = number of explanatory variable in the model

p = number of lags used to represent the short run dynamics in the model

There are several variables which could be used as proxies for of economic activity,

fiscal policy, monetary policy and control variables as outlined in the table below. We use

quarterly data from 1990 Q1 to 2010 Q2. To capture the effect of financial crisis period to the

effectiveness of policy choices, we add interaction variables of dummy recession and policy

variables. Some of variables were adjusted for seasonality using Census X12 method.

Table 5.List of Variables

Growth

Fiscal

Monetary

Inflationary

Effect

External Sector

Dummy

Recession

IFS, staff estimated

IFS

BI

BI

BI

1.1.4.1.1.1

BI

IFS

IFS

BI

IFS, staff estimated

BI

IFS

IFS

Estimated

Estimated

Government Expenditure √

Interest Payment

Prior to Q3 2005, we use SBI

1 month as proxy for policy

rate

Q1 1997 √ Q4 1998

Q4 2008 √ Q2 2009

Indicator Variable Sources Notes

GDP Real

GDP Nominal

Fiscal Balance

Government Revenue

Government Expenditure

Primary Expenditure

Primary Balance

M1

M2

Policy Rate

GDP Deflator

CPI

Exchange Rate

Current Account Balance

1997/1998 Recession

2008/2009 Recession

Notes: all variable are in logarithm, except Fiscal balance, primary balance, current account balance (because they contains negative values) and dummyrecession (binary 1/0)



4.1. Empirical Result

To characterize the time series property of the variables, we utilize the Augmented Dickey-

Fuller (ADF) and the Phillips Perron (PP) methods. This PP approach is more appropriate than

ADF since the data shows a structural break as effect of 1997/1998 crisis. Both of the ADF and

PP test indicate that most of the series are non-stationary when the variables are defined in

levels, except Fiscal Balance, Primary Balance, Policy Rate and Current Account Balance. But

first-differencing the series removes the non-stationary components in all cases and the null

hypothesis of non stationary is clearly rejected at the 5% significance level suggesting that all

variables are integrated of I(1). Thus, the next step of testing for possible cointegration relationship

will be done only with the I(1) variables.

t-stat p-values t-stat p-values t-stat p-values t-stat p-values

Real GDP RGDP -2.631 0.268 -2.139 0.032 I(1) -2.327 0.415 -8.320 0.000 I(1)

Nominal GDP NGDP -2.038 0.572 -5.929 0.000 I(1) -2.117 0.529 -5.491 0.000 I(1)

Real GDP_Adjusted RGDP_SA -2.214 0.475 -5.216 0.000 I(1) -1.921 0.634 -5.159 0.000 I(1)

Nominal GDP_Adjusted NGDP_SA -1.913 0.639 -1.957 0.049 I(1) -1.869 0.661 -2.939 0.004 I(1)

Fiscal Balance FB -4.231 0.006 -6.526 0.000 I(0) -9.025 0.000 -29.488 0.000 I(0)

Government Revenue GR 5.943 1.000 -17.544 0.000 I(1) 2.972 0.999 -42.299 0.000 I(1)

Government Expenditure GE 5.327 1.000 -20.666 0.000 I(1) 2.100 0.991 -36.833 0.000 I(1)

Primary Expenditure PRIM_GE 5.137 1.000 -20.657 0.000 I(1) 1.976 0.988 -42.757 0.000 I(1)

Primary Balance PB -5.436 0.000 -6.289 0.000 I(0) -7.132 0.000 -26.702 0.000 I(0)

Fiscal Balance_Adjusted FB_SA -3.074 0.003 -12.525 0.000 I(0) -8.552 0.000 -20.894 0.000 I(0)

Government Revenue_Adjusted GR_SA 4.040 1.000 -15.234 0.000 I(1) 3.211 1.000 -15.292 0.000 I(1)

Government Expenditure_Adjusted GE_SA 3.562 1.000 -10.192 0.000 I(1) 4.154 1.000 -18.761 0.000 I(1)

Primary Expenditure_Adjusted PRIM_GE_SA 4.400 1.000 -11.337 0.000 I(1) 3.046 0.999 -18.449 0.000 I(1)

Primary Balance_Adjusted PB_SA -1.843 0.063 -12.502 0.000 I(0) -7.297 0.000 -23.681 0.000 I(0)

M1 M1 -2.348 0.404 -1.880 0.058 I(1) -2.369 0.393 -7.610 0.000 I(1)

M2 M2 -0.933 0.947 -8.250 0.000 I(1) -0.933 0.947 -8.248 0.000 I(1)

M1_Adjusted M1_SA -1.496 0.823 -3.371 0.001 I(1) -1.641 0.768 -5.250 0.000 I(1)

M2_Adjusted M2_SA -0.864 0.955 -6.885 0.000 I(1) -0.941 0.946 -3.738 0.000 I(1)

Policy Rate PR -3.265 0.020 -7.633 0.000 I(0) -3.000 0.039 -7.646 0.000 I(0)

GDP Deflator GDPDEFL -2.125 0.524 -4.443 0.000 I(1) -1.871 0.661 -4.293 0.000 I(1)

GDP Deflator_Adjusted GDPDEFL_SA -2.271 0.444 -3.798 0.000 I(1) -1.849 0.672 -3.798 0.000 I(1)

CPI CPI -2.402 0.376 -3.049 0.003 I(1) -2.208 0.478 -4.278 0.000 I(1)

Exchange Rate ER -1.845 0.673 -5.637 0.000 I(1) -1.221 0.899 -6.479 0.000 I(1)

Current Account Balance CAB -3.021 0.003 -11.764 0.000 I(0) -3.021 0.003 -16.024 0.000 I(0)

GDP US USGDP -0.215 0.992 -2.199 0.028 I(1) -0.249 0.991 -3.144 0.002 I(1)

GDP Japan JPGDP 1.014 0.917 -3.321 0.001 I(1) 1.351 0.955 -16.482 0.000 I(1)

Variables Abbreviation

ADF Test Result Phillips Perron Test Result

Level 1st difference Level of

integration

Level 1st difference Level ofintegration

Table 6.Unit Root Test for Variables


Following the Engle and Granger two step-method, in the next step we estimate the

long run equilibrium relationship among variables by OLS and test for stationary of the residuals,

using critical values for the Engle - Granger Cointegration Test provided in Enders (2004). After

estimating several alternatives model based on the variables, we found the best long run

cointegration equations as follow:5

5 RM1_SA defined as real seasonally adjusted M1 which equal to nominal M1/GDP Deflator. RGE_SA defined as real seasonallyadjusted government expenditure which equal to nominal government expenditure/GDP Deflator. value in ( ) shows standard error.*** significant at a = 1%, ** significant at a = 5%, * significant at a = 10%. We realize that there might be endogenityrelationshipbetween RM1_SA and GDP_SA, but to be inline with the agreed methodology we assume the one-way relationship between themand use ECM. For robustness, we also use VECM and found the long run relationship between those variables.

(4)

The critical value of residual unit root test from this equation is -6.61, and given the

critical value of Engle - Granger Cointegration Test for 2 variables which is -4.123 for significance

at 1%, then variables real GDP, real M1 and real government expenditures are said to be

cointegrated.

We next switch to a short run model with an error correction mechanism in the form:

RGDP_SA = 5.13 + 0.88 RM1_SA + 0.20 RGE_SA

(0.05)*** (0.03) *** (0.25) ***

R2 = 0.94

(3)

Where X consists of real M1, real government expenditure and other control variables, meanwhile

ECM is residuals from equation (3). To address the impact of crisis, we also use dummy recession

variables CR97 and CR08 and interaction variables between recession in 2008 and policy

measures. With general to specific approach to several combinations of cointegrated variables

and lags, we found the best models as follows:


The empirical results show that Real GDP cointegrated with real M1 and real fiscal

expenditure. On the basis of this information, an error correction model was developed which

was shown to be well-specified relative to its own information set and capable of parsimoniously

representing the data set.

From the error correction models, we could conclude that in the short run the changes in

real GDP is significantly affected by changes in real fiscal expenditures and real monetary changes.

The previous real GDP changes are not significant to affect real GDP changes. The crisis 1997

decrease the real GDP significantly, meanwhile the crisis 2008/2009 effect is not significant.

The model shows that in the period of 2008 crisis, the impact of M1 changes and fiscal

expenditure to output relatively similar with the impact in the normal period. Thus, both of

variables are still significant in affecting output although in crisis period.

The result also shows a well-defined error correction term, and indicates a feedback of

7% of the previous quarter»s disequilibrium from the long run money supply, and fiscal

expenditure to economic activity. To evaluate the goodness of the model, we did some in

sample forecasting and compared the result with the actual data. The result was quite good as

shown in Figure 20. The root mean square error (RMSE) of the forecast was only 0.01.

Table 7.Error Correction Models

Real M1(-3)

Real Gov_Exp

ECM(-1)

Exchange Rate (-1)

Inflation (-1)

Crisis 1997

Crisis 2008

Real M1(-3)*Crisis 2008

Real Gov_Exp * Crisis 2008

Constant

R2

DW-Stat

SIC

0.06

0.03

-0.07

-0.05

-0.29

-0.01

0.01

0.14

0.08

0.02

0.66

1.98

-5.40

1.83

3.80

-2.04

-3.00

-5.47

-2.04

0.75

0.51

0.42

11.36

6

8

10

Variable Dependent : Real GDP

*** significant at α = 1%, ** significant at α = 5%, * significant at α = 10%

Coefficient t-stat

*

***

**

***

***

**

3

4

5

***

7

9

11


Figure 18.Evaluation of Model

The relative effectiveness of policy actions are determined by the size of the contribution

of policy instruments implemented to limit the severity of the downturn in order to achieve a

more positive outcome. While some economists believed that monetary policy should be the

first line of defense during the turbulence, others argued that fiscal policy has a more important

role particularly when conventional monetary policy measures were not sufficient in addressing

losses in output due to vulnerable in a weakening economy. Although it is not specific in the

crisis period, but the lag in the effect of monetary changes relative to fiscal policy as shown in

table 6 indicates that the impact of fiscal policy on GDP is relatively faster than monetary policy.

This result is in line with the Elmendorf and Furman (2008) which consider that a key

potential advantage of fiscal stimulus relative to monetary stimulus was that it could boost

economic activity more quickly, and true fiscal stimulus implemented promptly can provide a

larger near-term impetus to economic activity than monetary policy.

4.2. Analysis

In the empirical model above, we use M1 for indicator of monetary policy and real

government expenditure for indicator of fiscal policy. Meanwhile, after implementing Inflation

Targeting Framework (ITF) in July 2005 BI has shifted the focus of monetary operation toward

short term interest rate, replacing the former operating target of monetary base.

At operational level in the ITF, the monetary policy stance is reflected in the setting of the

policy rate (BI Rate) with the expectation of influencing money market rates and in turn the

12.2

12.4

12.6

12.8

13.0

13.2

13.4

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

LRGDP_SAF LRGDP_SA


deposit rates and lending rates in the banking system. At early stage of implementation of BI

Rate, BI rate was intended to only be reflected in the discount rate of Bank Indonesia one-

month paper (known as 1-month SBI). Since late January 2008, some gradual steps have been

taken to focus more on managing short term market interest rates around the BI Rate level.

Effective from June 9, 2008, BI has officially changed the operating target from 1-month

SBI rate to overnight (O/N) interbank money market rate (PUAB O/N). Changes in these rates

will influence output and inflation.Ω

Beside those rates, there are also several monetary instruments used in controlling the

agregate demand of the economy. Regarding to that, in this part, the analysis of the monetary

policies and fiscal policies that have been done will be broaden, not only M1 and government

expenditure. Because exchange rate also significant in affecting output, there will be analysis

about foreign exchange policy.

4.2.1 Monetary Policy

Bank of Indonesia took a series of policy measures in response the global financial instability.

Generally, the monetary policy measures are decided by taking into considerations, economic

circumstances and macroeconomic characteristic.As shown in the figure below, during the

period from January 2008 to November 2008, Bank Indonesia increased the policy rate from

8% to 9.50% in order to constrain the pressure of hiking CPI. Although the pressure from

financial stability point of view was higher as indicated by the high overnight interest rate,

banking interest rates and yield rate of government bonds were increasing due to shortage of

global liquidity at the time, BI decided to increase the policy rate in view of concern on inflation.

In line with slow down economic growth, inflation was also declined during 2008 √ 2009.

Accordingly, the policy rate has been adjusted downward since December 2008.

Further, as a response to the global financial crisis Bank Indonesia took some measures to

address liquidity issues in the financial sector. Following the declared bankruptcy of Lehman

Brothers in September 2008, the interbank market rose sharply from 6.98% to 9.37% Q-1 to

Q-III 2008. Beside that the volume of interbank rupiah transaction declined by 41% during the

period due to falling of market confidence among the banking institutions. The main objective

of the unconventional policy as shown in the table 4 was to bring the interbank interest rate to

the policy rate and feasibility its convergence with the policy rate. In order to reduce excessive

volatility in the interbank money market the interest rate corridor was narrowed on 4 September

2008 and on 16 September 2008.


3-A

pr-0

8

O/N

inte

rban

k ra

te re

plac

ed th

e 1

mon

thSB

I rat

e as

the

oper

atio

nal t

arge

t- F

asbi

Cor

ridor

: BI

Rat

e - 3

00 b

psRe

po C

orrid

or :

BI R

ate

+ 30

0 bp

s

16-S

ep-0

8

Fasb

i Cor

ridor

: BI

Rat

e - 1

00 b

psRe

po C

orrid

or :

BI R

ate

+ 10

0 bp

s

4-Se

p-08

Fasb

i Cor

ridor

: BI

Rat

e - 2

00 b

psRe

po C

orrid

or :

BI R

ate

+ 30

0 bp

s

23-S

ep-0

8

Expa

ndin

g tim

e pe

riod

ofFT

O fr

om 1

4 da

ys to

3 m

onth

4-D

es-0

8

Fasb

i Cor

ridor

: BI

Rat

e - 5

0 bp

sRe

po C

orrid

or :

BI R

ate

+ 50

bps

9-D

es-0

8

- Ope

n w

indo

w re

po o

f 2-1

4da

y te

nure

Jan-

08

Feb-

08M

ar-0

8A

pr-0

8M

ay-0

8Ju

n-08

Jul-0

8A

ug-0

8Se

p-08

Oct

-08

Nov

-08

Dec

-08

Dec

-08

Dec

-08

BI R

ate

9.25

%

Oct

-08

- Nov

-08

BI R

ate

9.50

%

Sep-

08BI

Rat

e9.

25%

Aug

-08

BI R

ate

9%

Jul-0

8BI

Rat

e8.

75%

Jun-

08BI

Rat

e8.

5%

May

-08

BI R

ate

8.25

%

Jan-

08 -

Apr

-08

BI R

ate

8%

14-O

ct-0

8

Low

erin

g Re

serv

e Re

quire

men

tRu

piah

Dep

osit

to 7

.5%

Fore

ign

Dep

osit

to 1

%

23-O

ct-0

8

Low

erin

g Re

serv

e Re

quire

men

tRu

piah

Dep

osit

to 5

%

Apr

-09

Ope

n w

indo

w re

po o

f 1 m

onth

tenu

re

Sep-

09

Ope

n w

indo

w re

po o

f 3 m

onth

tenu

re

Oct

-09

Impl

emen

tatio

n of

2.5

% s

econ

dary

rese

rve

requ

irem

ent

Jan-

09D

ec-0

8

Feb-

09M

ar-0

9A

pr-0

9M

ay-0

9Ju

n-09

Jul-0

9A

ug-0

9Se

p-09

Oct

-09

Nov

-09

Dec

-09

Jan-

09Fe

b-09

Mar

-09

Apr

-09

May

-09

Jun-

09Ju

l-09

Aug

-09

- Dec

-09

BI R

ate

8.75

%BI

Rat

e8.

25%

BI R

ate

7.75

%BI

Rat

e7.

50%

BI R

ate

7.25

%BI

Rat

e7.

00%

BI R

ate

6.75

%BI

Rat

e6.

50%

Pict

ure

1.

Mo

net

ary

Polic

y in

200

8 an

d 2

009


Like other emerging central banks, Bank Indonesia also responded to the liquidity

problem in the financial sector by reducing reserve requirement, followed by providing liquidity

facility in order to flow the fund into the financial market. Furthermore, there were coordination

between the central bank and the government taken to address confidence issues and also to

respond to the asset price burst. Several policies also had been taken to address confidence

issues and asset price burst. The description of policies taken by Bank Indonesia during the

global crisis could be seen in the appendix.

a. Evaluation of Interest Rate Policy

There were significant differences between the movements of the policy rate in 1997/

1998 crisis and the recent crisis. To contain the inflation pressure which was very high in the

1997/1998 crisis (almost 83% in Q-IV 1998), the central bank increase the SBI rate significantly

to 68% and then gradually brought down to 11.03% in Q-I 2000. Meanwhile, prior to the

recent crisis, the inflation pressure was not so high, and hence, the policy rate was raised only

up to 9.25%. At that moment, BI Policy to increase the policy rate in quarter II, III and IV 2008

1997/1998 Crisis 2008/2009 Crisis

Policy Rate (%)Duration 6 quarters more than 6 quarters

Period Q4 1998 √ Q1 2000 Q1 2009 √ Q2 2010 (still not increase yet)

Amplitude 57.73% 2.75%

Slope 10% 0.46%

Lending Rate (%)Duration 9 quarters more than 6 quarters

Period Q4 1998 √ Q4 2000 Q1 2009 √ Q2 2010


Slope 2.0% 0.3%

Inflation (% yoy)Duration 6 quarters 5 quarters

Period Q4 1998 - Q1 2000 Q4 2008 - Q4 2009


Slope 13.9% 1.9%

Table 8.Duration, Period, Amplitude and Slope of Interest Rate and Inflation



was contrary to other central banks in the region and across the world as then policy was to

lower their interest rates to address liquidity issues and reduce economic activity. BI commenced

gradually reducing its policy rate in QI 2009 and now it l stays in 6.5%.

During both of crisis periods, the magnitude of the decline in average lending rate in

the banking sector was much smaller than the decline of BI Rate which could be seen in the

wider spread between lending rate and policy rate. From micro perspective of banks, some

factors contributing to the lending rate movements included the cost of fund and risk premium

which tended to rise during the crisis, and profit margin. A previous study has showed that

the decrease in the aggregate banking cost of fund throughout 2009 tended to be slower

than the decline in BI Rate. Furthermore risk premium in the economy was still perceived to

be high and there were indications that the banking industry preferred to maintain their

profit margin. All of these contributed to reduce the power of policy rate pass through to

lending rates.

Figure 19a.Peak and Trough of Interest Rate

Figure 19b.Peak and Trough of Inflation


1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

0

10

20

30

40

50

60

70

80Policy Rate

Working Capital Lending RateSep-98, 68.76

Mar-00, 11.03Dec-08, 9.25 Jun-10, 6.5

Sep-98, 35.7

Dec-00, 17.65Dec-08, 15.22 Jun-10, 13.26


1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

%

-10.0

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0Sep-98, 82.4%

Mar-00, -1.2%Sep-0812.1% Dec-09

2.8%

Source : CEIC, estimated Source : CEIC, estimated

Non - Crisis Crisis

Q2 1996 - Q1 1997

6.12 %6.12 %6.12 %6.12 %6.12 %

Q1 2006 - Q4 2006

4.29 %4.29 %4.29 %4.29 %4.29 %

Q2 1999 - Q1 2000

8.23 %8.23 %8.23 %8.23 %8.23 %

Q1 2009 - Q4 2009

7.41 %7.41 %7.41 %7.41 %7.41 %

Table 9.Average Spread between Lending Rate and Policy Rate During Expansion Period


b. Evaluation of Lowering Reserve Requirement

Reserve Requirement Policy was affected to provide more rupiah liquidity to the banking

system to achieve liquidity constrain and reduce volatility in the interbank market. Its impact

could be seen in the interest rate of interbank market which declined after the policy

announcement. The volatility of interbank interest rate also decreased. As measured by the

standard deviation, during the period of 1 month before the implementation of the policy, the

volatility of interbank interest rate was 0.22% and it was decreased to 0.08% during the 1

month period after the implementation of the policy.

The volume of interbank transactions also increased although the impact was not as

immediate as that on the interest rate. One of the reasons for lower increase in the interbank

market transaction could be expansion in the fine tuning operations of the Central bank at that

time, which enhanced the access to liquidity from the Central bank and hence reduce the need

for borrowing from the interbank market. This was also supported by the amendment of liquidity

facility for commercial banks from central bank which provide wider access for banks to receive

funding for a longer time horizon than the usual inter-day facility. Another reason was to

reduce the loan disbursement by commercial bank which was in line with the lower demand of

credit at that time due to global crisis and hence the lower need to borrow funds.

Figure 20.Interbank Market Interest Rate

Figure 21.Volume of Interbank Market Transaction

Lowering RRPolicy, 23 Oct '08

8.80

9.00

9.20

9.40

9.60

9.80

10.00

10.20

2 0 0 8

1Aug

11Aug

21Aug

31Aug

10Sep

20Sep

30Sep

10Oct

20Oct

30Oct

9Nov

19Nov

29Nov

9Dec

19Dec

29Dec


Billion of Rp

Lowering RRPolicy, 23 Oct '08

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

2008

1Aug

11Aug

21Aug

31Aug

10Sep

20Sep

30Sep

10Oct

20Oct

30Oct

9Nov

19Nov

29Nov

9Dec

19Dec

29Dec

However, there is a side-effect of this policy. The increasing supply of rupiah has led to

the rise of money in circulation in the market (M0), and this could cause a depreciation of

rupiah.


Figure 22.Interest Rate Corridor (%)

Table 10.Spread of Policy Rate and Interbank Market Rate

Period Spread

1 week before the 1st adjustment

1 week after the 1st adjustment

1 week after the 2nd adjustment

1 week before the 3rd adjustment

1 week after the 3rd adjustment

9.08 bps

11.72 bps

27.42 bps

35.57 bps

17.59 bps

Notes

Higher

Higher

Ω

Lower

Note:1st : 4 September 2008; 2nd : 16 September 2008; 3rd : 4 December 2008Source: Bank Indonesia, estimated from daily data

Adjustment of Corridor ofOvernight Interest Rate

Adjustment ofCorridor of O/N

Interest Rate

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

2 0 0 8

1Aug

11Aug

21Aug

31Aug

10Sep

20Sep

30Sep

10Oct

20Oct

30Oct

9Nov

19Nov

29Nov

9Dec

19Dec

29Dec

Interbank O/N RateBI RateOperational Corridor


c. Evaluation of Narrowing the Interest rate corridor for Standing DepositFacility/ Lending Facility (Repo)

This policy was aimed to reduce excessive volatility in the interbank money market and

increase credibility of Bank Indonesia. The lower spread between interbank interest rate and

policy rate indicate as higher degree of credibility of central bank.

From the figure above, we could see that after the first and second adjustment of the

corridor, the spread between interbank market rate and policy rate was relatively higher and it

contradicted with the purpose of the policy. One of the reasons would be the increasing pressure

of liquidity because of higher risk perception in September following the bankruptcy of Lehman

Brother. Meanwhile, the third adjustment was successful to decrease spread between interbank


market and policy rate (Table 10). For comparison, best practices of this spread in several ITF

countries are about 20bps.

4.2.2 Fiscal policies

Like other countries, Indonesia launched fiscal policy measures as a countercyclical measure

to fight the slowdown direct effect on aggregate demand. In order to reduce the impact of

current crisis, the Indonesian government took ten steps for the purpose of economic stabilization

and securing the state budget. Additionally, the government offered a fiscal stimulus package

amounting to 73.3 trillion Rupiah or 7.56 billion US$ (2.6% of GDP) with following aims,

a. Maintaining household purchasing power to keep consumption growth above 4%.

b. Improving real sector resilience and competitiveness to prevent more worker layoffs.

c. The government issued some traded policies.

d. Creating job opportunities for unemployed/laid-off workers by launching labor √intensive

infrastructure projects

e. Social protection and poverty alleviation has decided to state in the budget for 2009.

In the period of crisis, the realization of fiscal balance at the end of 2008 was a significant

improvement with a marginal deficit, while in 2009, fiscal stimulus and the lower government

revenue due to economic slowdown leads to a higher fiscal deficit of 1,6% of GDP (Figure 23).

Based on its component, the largestgovern,entexpenditure in 2008 and 2009 were transfer to

region and subsidies (Figure 24).

Figure 23.Government Finance Operation

Figure 24.Total Expenditure Component

% of GDP% of GDP

-2.00

-1.50

-1.00

-0.50

0.00

0.50

1.00

1.50

2.00

2.50

2004 2005 2006 2007 2008 2009 2010*0.0

5.0

10.0

15.0

20.0

25.0

-1.0%

-0.5%

-0.9%

-1.3%

-0.1%

-1.6% -1.6%

1.7% 1.8% 1.5% 0.8% 1.7% 0.1% 0.3%

Fiscal Balance Primary BalanceExpenditure (RHS) Revenue (RHS)

Poly. (Fiscal Balance)

Source : Ministry of Finance

% of GDP

0.0

5.0

10.0

15.0

20.0

25.0

2004 2005 2006 2007 2008 2009

Transfer to Region Other Routines Social AssistanceCapital Expenditure Subsidy Interest PaymentMaterial Personnel Total Expenditure

5.7% 5.4% 6.8% 6.4% 5.9% 5.5%

2.7% 1.2%

1.6%1.6% 1.5% 1.3%

4.0% 4.4%3.2%

3.8% 5.6%2.9%

2.7% 2.4%2.4% 2.0%

1.8%

1.7%

2.3% 2.0%2.2%

2.3% 2.3%

2.3%

18.6% 18.4%20.0% 19.2% 19.9%

17.1%

Source : Ministry of Finance


The success of the government to hold fiscal sustainability during the financial crisis in

2008 was attributed to (1) over the past ten years, fiscal policy actions have reduced the high

public debt ratio. (ii) The government has taken significant measures to reduce domestic fuel

subsidies enabling it to increase spending at both the central and local government levels. (iii)

The government has increased the total spending on education.

However, in the Indonesian case, the problem in the fiscal stimulus package was that

regional government has a limited ability to complete their budgets on time. Therefore, the

stimulus package hadn»t used optimally as can be interpreted from the data for October and

December 2009 which show that only 52.1% of fiscal stimulus plan was realized. Lack of

socialization, frugal spending and slow regulation implementation led to this low absorption of

the fiscal stimulus. Thus, to increase the effectiveness of fiscal policy, it is needed to configure

an effective and understandable standard operating procedure for fiscal policy implementation,

both in central and regional areas.

To analyze the effectiveness of fiscal policy to boost economic activity after the crisis is

not an easy task, thus we try to adopt the evaluation of fiscal stimulus (FS) principles based on

Elmendorf and Furman (2008) as follows:

Table 11.Fiscal Stimulus Plan and Realization

Source: Ministry of Finance and other sources

Notes : 1) Realization until October 2009, more updated data is not available2) Realization until December 2009

1 43.0 20.5 47.7%1)

Reductions in Income Tax Rates 32 18 56.3%1)

Lower Corporate Tax Rate 18.5 12.8 69.2%1)

Lower Personal Income Tax Rate 13.5 5.2 38.5%1)

Income tax-free band raised to IDR 15.8 million 11 2.5 22.7%1)

2 13.3 3.7 27.8%1)

VAT on oil/gas exploration, cooking oil 3.5 2.5 71.4%1)

Import duties on raw materials and capital goods 2.5 0.3 12.0%1)

Payroll tax 6.5 0.1 1.5%1)

Geothermal tax 0.8 0.8 100.0%1)

3 17 14.0 82.2%Reduced price for automotive diesel 2.8 2.8 100.0%1)

Discounted electricity billing rates for industrial users 1.4 1 71.4%1)

Additional infrastructure expenditures+subsidies+government equity injection 12.2 10.18 83.4%2)

Upscaling of Community Block Grants (PNPM) 0.6 n/a73.3 38.2 52.1%

Plan Realization

(IDR Trillion) (IDR Trillion)Tax Savings

Tax/Import Duty Subsidies for Business/Targeted Households

Pro-business/Jobs subsidies + budget expenditures

TOTAL

DescriptionNo% of

Realization


7 SILPA: Excess of government budget utilization in the previous year

Table 12.Evaluation of the Effective Fiscal Stimulus Principles

Principles Explanation

Timely

Targeted

Temporary

Although government immediately gave instruction for FS,there were problems that delayed the implementation ordisbursement of the fund.The government expenditure was mostly disbursed in Q-4(Figure 26). It would have been better if the disbursementin each quarter were quite balanced. Thus in terms of thetimely principles, the Indonesia»s fiscal stimulus was lesseffective.

The biggest proportion of the FS was tax reductions. Thiscould stimulate economic output from investment andindirectly would increase employment and wages. Then,there will be increases in consumption and economic output.Moreover, the big spending in infrastructure was goodbecause it would boost a sustainable growth in longer terminstead of only short term.Thus from the targeted principles, the Indonesia»s fiscalstimulus was effective.

The source of fund for fiscal stimulus came from the excessof budget utilization (SILPA)7 in 2008 and debt.Fund from excess of budget utilization won»t affect the nextgovernment budget but the usage of debt, in the long termcould impact the budget deficit.Additionally, the budget deficit plan in 2010 still relativelyhigh (1.6% GDP) (Table 15). Thus from the temporaryprinciples, the Indonesia»s fiscal stimulus was quite effective

Explanation Measure for Indonesia»s Fiscal Stimulus

FS should not be enactedprematurely, delayed too long, orconsist of tax cuts or spendingincreases that would take too longto be implemented or to boostoutput

Tax cuts and spending increasesshould be directed so that theyprovide the greatest benefit topeople who are affected mostadversely by an economic slowdown

The FS should not increase thebudget deficit in the long run

Note: The effectiveness of means is based on authors» opinion.

Figure 25.Government Expenditure by Quarter

Figure 26.Tax Revenue versus GDP Changes

0

100,000

200,000

300,000

400,000

90 92 94 96 98 00 02 04 06 08 10

Q1Q2Q3Q4

Source : CEIC

Trillion of Rp

-

200

400

600

800

1,000

1,200

1,400

2003 2004 2005 2005 2007 2008 2009 2010*

Change of Tax Revenue (yoy)

Change of GDP (yoy)


Figure 27.Impact of Foreign Exchange Policy

Table 14.Volatility of Foreign Exchange after Policy

1 weekbefore

1 weekafter

15 October's policies 94.58 69.61 Lower

13 November's policies 349.90 244.21 Lower

16 December's policies 338.22 73.40 Lower

Std Deviation of Rp/US$FX Policies Result

- Abolishing the limit of dailybalance position of short term

foreign loan- Foreign Exchange Provision for

Domestic Corporation- The extension of FX Swap tenor

Regulation governingthe purchase ofForeign Exchangeby Banks

Prohibition ofstructured

producttransaction

8000

8500

9000

9500

10000

10500

11000

11500

12000

12500

13000

2 0 0 8

1Aug

11Aug

21Aug

31Aug

10Sep

20Sep

30Sep

10Oct

20Oct

30Oct

9Nov

19Nov

29Nov

9Dec

19Dec

29Dec

Source : CEIC, Estimated

Table 13.Impact of Fiscal Policy on Government Budget

PeriodFiscal Balanceas a% of GDP

1990-1996 Average

1997-1998 Average

1999

2001-2008 Average

2009

2010 Estimates

0.26%

-1.45%

-2.84%

-1.16%

-1.56%

-1.59%

Primary Balanceas a % of GDP

2.05%

1.13%

1.05%

1.92%

0.12%

0.28%

Source: Ministry of Finance

4.2.3 Foreign Exchange Policy

To address and avoid more Rupiah depreciation, Bank Indonesia also introduced following

foreign exchange policy measures such as extending FX swap tenor, issuing regulation about

the purchase of foreign exchange by banks, etc (details could be seen in Appendix).The foreign

exchange policy measures were effective to reduce the volatility of exchange rate during the

crisis. Although in term of level, rupiah currency continue to depreciate after several exchange

rate policy measures in October and November 2008 because of the massive capital outflow,

the volatility of the exchange rate became relatively lower after the introduction of policy

measures.


4.2.4 Condition of Indonesia Economic after the Crisis

There have been improvements in Indonesian economy during 2009 and 2010 which

boosted optimism over the sustainability of the ongoing economic recovery. The positive

economic performances include an improvement in the risk indicators, performance in stock

market, balance of payment, the strengthening of rupiah, and the fairly high economic growth.

Economic growth in quarter II 2010 has increased to 6.17%, which was 2.07% points

higher than economic growth in the same quarter on the previous year. From the demand side,

the highest contribution to economic growth came from domestic demand especially from

investment (gross fixed capital formation) 2.35% and consumption 2.06%. From sector

perspectives, the main contributors were transportation and communication (1.13%); trade,

hotel and restaurants (1.61%) and manufacturing (1.12%).

The indications that the global recovery is proceeding sooner than previously expected

have boosted optimism over Indonesian economic outlook. Such optimism is also supported by

domestic economic resilience that endured the effects of the global crisis. The increased optimism

over Indonesian economic outlook is confirmed by the upgrading of Indonesia»s rating by

international rating agencies in early 2010. These positive conditions support the empirical

result of this research that both monetary policies and fiscal policies give significant impact to

the economic output.

However, there are several challenges for Indonesian economic development. From the

external side, the challenge is primarily related to the impacts of likely strategies developed

countries to unwind the measures adopted by them in response to the global crisis, which

included monetary easing and fiscal expansion, the polarizing trend of global trade and the

Table 15.Sector Contribution to GDP Growth 2010

Tabel 16. Demand Side Contribution toGDP Growth 2010

Q1 Q2

Agriculture 0.42 0.43Mining and Quarrying 0.25 0.31

Manufacturing 0.97 1.12Electricity, Gas and Water Supply 0.06 0.04

Construction 0.45 0.45Trade, Hotels and Restaurant 1.55 1.61Transportation and Communication 1.02 1.13Financial, Rental and Business Services 0.52 0.58

Services 0.44 0.50GDP 5.69 6.17

S e c t o r 2 0 1 0Q1 Q2

2 0 1 0 Demand

Consumption 1.65 2.06

Gross Fixed Capital Formation 2.82 2.35

Net Export 1.18 0.42

Export 7.85 6.02

Import 6.67 5.60

Statistic Discrepancy 0.04 1.34

GDP 5.69 6.17

Sumber: Bank Indonesia

Sumber: Bank Indonesia


large imbalances in global economic performance. From the domestic side, challenges are

related to several issues that could disrupt the effectiveness of monetary policy, such as excess

bank liquidity, the dominance of short-term inflows in the structure of capital inflows, the

potential asset price bubble, a shallow financial market and numerous structural problems in

the real sector.

V. CONCLUSION

Indonesia has been affected by sudden stop of capital inflow into emerging market

countries and declining on global economies growth following global financial crisis. Their first

and second round effects on macroeconomic indicators were identified. The first lesson from

the recent crisis is that Indonesia as an emerging country clearly demonstrated the effectiveness

of timely monetary, fiscal and financial sector policies which helped Indonesia to recover from

the economic crisis. Indonesia and most Asian countries had experienced the two financial

crises during the last ten years. The first Asian crisis episode occurred in 1997, led to introduce

significant reforms both policy reforms and institutional reforms in the financial sector, However,

in the second crisis ten years later known as global financial crisis occurred in 2008, the reforms

introduced can be categorized as soft reform compared to those in the first Asian crisis.

The second lesson is that the closer cooperation and coordination among the policy

makers is very important in identifying and dealing with challenges posed by a global crisis. As

the Central Bank authority, Bank Indonesia had implemented an accommodative monetary

policy in order support a moderate growth with a relatively low inflation. The policy rate

commenced sliding on December 2008 with the intention to decrease bank»s lending rates.

Some unconventional monetary policy had also been taken to address liquidity issues. On the

fiscal side, the government responded to keep domestic demand by several fiscal stimulus and

trade policies. There were also coordination between Ministry of Finance, Central bank and

other institutions in order to maintain financial market and macroeconomic stability.

The policy measures taken during the crisis had been formulated in a timely manner with

the ultimate objective of sustainable economic growth while maintaining macroeconomic stability

in Indonesia. Based on error correction models, we concluded that in the short run the changes

in real GDP is significantly affected by changes in real money supply in the previous three

quarter and real fiscal expenditures. This indicated that the impact of fiscal policy to GDP is

relatively faster than monetary policy.

As conclusion of this research, policy implications that should be concerned in the future

are as follows:


1. The cooperation and coordination among the policy makers and the timely responses are

very important in tackling the crisis. Thus, in addressing the crisis, monetary policy could not

stand alone but requires coordination with fiscal policy and other sectoral policies.

2. An effective conventional monetary policy in normal times may become less effective in a

crisis because of the high degree of uncertainty particularly with pressure from external

circumstances. Thus, unconventional monetary policy indeed is necessary as timely policy

response.

3. Regarding to fiscal policy, more timely disbursement of government expenditure is important

to increase the effectiveness of this policy to stimulate economic output.


Arifin, Syamsul, (1998), ≈Effectiveness of Interest Rate Policy for Rupiah Stability during Crisis,

Bank Indonesia Buletin Ekonomi Moneter dan Perbankan, December 1998. (Available on

Bahasa).

Bank Indonesia, Indonesian Economic Outlook 2009-2014, ≈Global Financial Crisis and the

Impact on Indonesian Economy∆

Enders, Walter, (2004), ≈Applied Econometric Time Series∆, John Wiley∆& Sons, Inc, USA

Elmendorf, Douglas and Furman Jason, (2008),∆≈If, When, How: A Primer on Fiscal Stimulus∆,

Bookings Institution, The Hamilton Project Strategy Paper

Kurniati Y, et all, (2008), ≈Sensitivity of Mainstay Export Commodities to Slowing Trading Partner

Growth and Changing International Prices∆, Bank Indonesia Research Notes: October.

(Available on Bahasa)

Kurniati Y and Meily Ika Permata, (2009), ≈Transmission of External Shock to Indonesian

Economy∆.Bank Indonesia Working Paper.(Available on Bahasa).

Santoso et all, (2009), ≈Impact of Contagion Risk on the Indonesian Capital Market∆. Bank

Indonesia Financial Stability Report No 12: March.

Simorangkir, Iskandar and JustinaAdamanti, (2010), ≈The Role of Fiscal Stimulus and Monetary

Easing in Indonesian Economy during Global Financial Crisis: Financial Computable General

Equilibrium Approach∆, Presented at Call for Papers - EcoMod2010, Istanbul, July 7-10,

2010.

Yehoue, Etienne B et all, (2009), ≈ Unconventional Central Bank Measures for Emerging

Economies∆. IMF Working Paper No. 09/226

Yudo, Teguhet all, (2009),

≈The Impact Of The Global Financial Crisis On Indonesia»s Economy∆. Centre for Strategic and

International Studies (CSIS)

Bank Indonesia (2008) Economic Report On Indonesia. Bank Indonesia Annual Publication

Bank Indonesia (2009) Economic Report On Indonesia. Bank Indonesia Annual Publication

REFERENCES


APPENDIX

Note : * calculated by HP Filter Method using annual Real GDP data in Billion RpSource : IFS, CEIC, Bank Indonesia and staff estimates

69.22

4.64

19.30

70.50

2,745.83

6.3%

BB -

5.60%

10,852

Indicator

Bank Loan to Deposit Ratio (LDR)

Non Performing Loan %

Bank Capital Adequacy Ratio

ICRG

Stock Market Index (last position)

GDP Growth (yoy)

S & P Rating

Inflation

Output Gap*

1996 1997

109.26

11.82

70.00

637.43

7.8%

5.12%

149,293

Table A. 1 Key Indicators : Measuring Vulnerability of Economy to External Shocks √ 1997/98 and 2008/09 Recession

432.04

118.01

54.74

85.26

7.10

7.70

10.49

147.51

141.18

Ω

27.49

113.69

2.8

3.3

-5.5

-

27%

13%

20%

-

-

2.4%

34.1%

32.7%

Ω

6.4%

26.3%

0.7%

0.8%

-1.3%

Indicator

GDP (Current)

Exports of Goods and Services

Foreign Currency Reserves

Imports of Goods and Services

Average Monthly Imports

Months of Imports Covered

Balance on Current Account

Total Government Debt

Foreign Debt

Composition of External Liability

Short term

Long term

Debt Service Payment (foreign)

Primary Balance

Fiscal Deficit

1996 1997

US $ Billions % of GDP US $ Billions % of GDP

227.37

50.19

17.82

44.24

3.69

4.83

-7.80

Ω

110.17

Ω

22.03

88.14

2.82

4.55

1.73

Ω

22%

8%

19%

-

-

-3%

Ω

48%

Ω

9.7%

38.8%

1.24%

2.00%

0.76%


Table A. 2 Similarity and Difference between 1997/1998Crisis and 2008 Global Crisis in Indonesia

Similiarity

Both crisis were the consequence of theglobal economy, because of the economicand financial interdependence amongcountries;The impact of the crisis led to falling valueof the rupiah against foreign currencies;The impact of the crisis will affect theeconomic sectors which resulted in lossesfor the community.

Difference

1998 crisis was multidimensional witheconomic crises, political, social, ideologi-cal, defense and security , meanwhileglobal crisis is tend to caused by financialand economic crisis;1998 crisis started from currency crisis inthe Bath-Thailand while the global crisisstarted from the breakdown of Sub-PrimeMortgage in the United States;1998 economic crisis led to on societyanarchism action while the global crisis didnot;1998 crisis led to the demanding forchange of leadership, while global crisis didnot;The focus of monetary policy in 1998 crisiswas tightening, meanwhile in global crisiswas loosening.


Table A. 3Monetary Policies Taken During Financial Crisis 2008 - 2009

Policies

Policy rate increased gradually to 9.25% onSeptember 2008

Policy Rate (BI Rate) increased to 9.5%(October andNovember 2008, decrease to 9.25% (December2008) and then decreased gradually to 6.75% inJuly 2009Lowering Reserve Requirement for Rupiah currencyfrom 9.1% to 7.5% consist of 5% primary reserve(cash reserve) and 2.5% secondary reserve (23October 2008)Lowering Reserve Requirement for foreign currencyfrom 3% to 1%. (23 October 2008)

Objectives

To contain prevent inflationary pressure such assecond round effect of the fuel price hike and foodprices on other goods.To sustain business momentum amid the globaleconomic slowdown while safeguarding macro-economic stability

To provide more rupiah liquidity in to the bankingsystem

To increase USD liquidity availability to be used bybanks in their transactions with customers.

Conventional Monetary Policy

Policies Objectives

Unconventional Monetary Policy

Narrowing the interest rate corridor for Standing DepositFacility (Fasbi) to BI Rate - 200 bps (from BI Rate - 300bps and maintain Lending Facility (Repo) at BI Rate +300 bps (4 September 2008)Narrowing the interest rate corridor for Standing DepositFacility (Fasbi) to BI Rate - 100 bps and Lending Facility(Repo) to BI Rate + 100 bps (16 September 2008)

Expanding time period of Fine Tuning Operation from14 days to 3 month (23 September 2008)Amendment of Regulation regarding the Liquidity Facilityfor Commercial Banks (18 November 2008).

Open Standing Facility(repo) of 2-14 day tenure (9December 2008)Regulation regarding a Liquidity Facility for Rural Banks(BPR) (10 December 2008)

Narrowing the interest rate corridor for Standing DepositFacility (Fasbi) to BI Rate - 50 bps and Lending Facility(Repo) to BI Rate + 50 bps (4 December 2008)Open 1 month window repo (FTE) (17 April 2009)

To reduce excessive volatility in the interbank moneymarket

To reduce excessive pressure in the interbank moneymarket and maintain sufficient liquidity in bankingindustry sufficiently.

To provide a wider flexibility for liquidity managementin interbank money marketTo smooth the operation of the payment systemsupported by high-value, liquid collateral

To provide wider access to banks by offering fundingwith a longer time horizon than the inter-day fundingfacility

To allow banks suffering from insufficient liquidity toremain solvent and avoid systemic impacts

To facilitate the longer-term bank liquidity requirement

To provide an equal opportunity for rural banks tomake use of this funding facility if a short-termliquidity shortfall is experiencedTo resolve the issue of segmentation in the interbankmoney market

To facil itate the longer-term bank liquidityrequirement.


Table A. 4Policies Taken to Address Confidence Issues and Asset Price Burst

Joint Policies

Executing Government Bond buyback andpreparing a state-owned enterprise equitybuyback program.

Allowing alternative security evaluation techniquesuch as discounted cash flows beside marked tomarket value (Joint press release- Bank Indonesia,Bapepam,-LK and Accounting Association).Allowing commercial bank to switch bondportfolio from trading and available for salecategories to held to maturity category.

Maintaining a sufficient level of foreign reserves.

Putting some restrictions on short-selling on thecapital market. Limiting purchases of foreigncurrency without underlying transaction toUS$100.000 to curb speculation.

Banning trading on banks» structured product/derivative products that provide chances for bankcustomers to purchase foreign currencies includingdual currency deposits that are callable forward.

Objective

To reduce the high risk perception in Indonesianfinancial portfolio which can distort the monetarypolicy transmission mechanism, The Minister ofFinance has bought back IDR41 billion (US$3.89million) worth of Government Bond using thegovernment fund in the central bank account.

To provide market confidence for the governmentbond particularly when no market prices areavailable.To minimize the impact of Indonesian Financialturbulence by providing an opportunity to thecommercial bank to arrange portfolio categories.

To support the rupiah and focused more onpreventing too volatile movement of the rupiah.

To reduce high risk perception

To reduce speculation and exchange rate volatilityexpectation.


Table A. 5Foreign Exchange Policy

Policies

Abolishing the limit of daily balance position ofshort term foreign loan (13 October 2008)

Foreign Exchange Provision for DomesticCorporation through Banks (15 October 2008)

The extension of FX Swap tenor from a maximumof 7 days to a maximum of 1 month (15 October2008)

Regulation governing the purchase of ForeignExchange by Banks (13 November 2008).

Amendment to Bank Indonesia Regulation onConcerning Derivative Transaction (prohibition ofstructured product transaction) (16 December2008)

Coordination with Other Central Banks, such as :- Signing of a Bilateral Currency Swap

Arrangement (BCSA) between Bank Indonesiaand People»s Bank of China (23rd March 2009)

- Signing of the agreement on an increase in themaximum amount of the Bilateral SwapArrangements between Japan and Indonesiaunder the Chiang Mai Initiative (6 April 2009)

Objectives

To decrease pressures in USD purchase due totransfer of rupiah account to foreign currencyaccount by foreign customers.

To enhance assurance in fulfilling foreign currencydemand by domestic companies

To fulfill the temporary demand for USD currencyand in order to provide sufficient adjustment timefor banks/market players before actually adjustingtheir portfolio composition

To support the balance of supply and demandcondition of foreign exchange in the domesticmarketTo moderate excessive pressure on rupiah exchangerateTo mitigate foreign currency purchase for speculativepurposesTo support banks» prudential actions through KnowYour Customer Principle (KYC).

To minimize speculative foreign currency transaction

To improve trade and direct investment betweenboth countriesTo assist in providing short-term liquidity for financialmarket stabilization and help Indonesia address tightinternational liquidity.



219

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

1,000,000 to IDR 3,000,000.

3. The paper should be sent in 2 formats (1) Microsoft Word (*.doc) and (b) portable digital file

(*.PDF). The 2 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

Directorate of Economic Research and Monetary Policy-Bank Indonesia

Building Syafruddin Prawiranegara, 20th Floor

JI. M. H. Thamrin No.2 Central Jakarta, Indonesia

Ph. +62-21-3818202, Fax. +62-21- 3800394

4.4.4.4.4. To avoid missing fonts or other compatibility issue, any special characters or mathematicalTo avoid missing fonts or other compatibility issue, any special characters or mathematicalTo avoid missing fonts or other compatibility issue, any special characters or mathematicalTo avoid missing fonts or other compatibility issue, any special characters or mathematicalTo avoid missing fonts or other compatibility issue, any special characters or mathematical

expression (equations, symbol, matrix, etc.) must be written using Microsoft Equation.expression (equations, symbol, matrix, etc.) must be written using Microsoft Equation.expression (equations, symbol, matrix, etc.) must be written using Microsoft Equation.expression (equations, symbol, matrix, etc.) must be written using Microsoft Equation.expression (equations, symbol, 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 reference must obey the following rule:


a. Book:

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

Jersey.

b. Article in journal:

Rangazas, PeterRangazas, PeterRangazas, PeterRangazas, PeterRangazas, Peter. ≈Schooling and Economic Growth: A King-Rebelo Experiment with

Human Capital∆, Journal of Monetary Economics, October 2000, 46(2), page. 397-416.

c. Article in book edited by other people:

Frankel, Jeffrey AFrankel, Jeffrey AFrankel, Jeffrey AFrankel, Jeffrey AFrankel, Jeffrey A. and Rose, Andrew KRose, Andrew KRose, Andrew KRose, Andrew KRose, Andrew K. ≈Empirical Research on Nominal Exchange Rates∆,

in Gene Grossman and Kenneth Rogoff, eds.,∆Handbook of International Economics.

Amsterdam: North-Holland, 1995, page. 397-416.

d. Working papers:

Kremer, MichaelKremer, MichaelKremer, MichaelKremer, MichaelKremer, Michael and Chen, DanielChen, DanielChen, DanielChen, DanielChen, Daniel. ≈Income Distribution Dynamics with Endogenous

Fertility∆. National Bureau of Economic Research (Cambridge, MA) Working Paper

No.7530, 2000.

e. Mimeo or unpublished work:

Knowles, JohnKnowles, JohnKnowles, JohnKnowles, JohnKnowles, John. ≈Can Parental Decision Explain U.S. Income Inequality?∆, Mimeo,

University of Pennsylvania, 1999.

f. Article from web or other electronic form:

Summers, RobertSummers, RobertSummers, RobertSummers, RobertSummers, Robert and Heston, Alan WHeston, Alan WHeston, Alan WHeston, Alan WHeston, Alan W. ≈Penn World Table, Version 5.6∆

http://pwt.econ.unpenn.edu/, 1997.

g. Article in newspaper, magazine or equal periodicals:

Begley, SharonBegley, SharonBegley, SharonBegley, SharonBegley, Sharon. ≈Killed by Kindness∆, Newsweek, April 12, 1993, page. 50-56.

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

phone number.

Documents

BULLETIN OF MONETARY ECONOMICS AND BANKING · The Bulletin of Monetary Economics and Banking ... Tumpak Silalahi, ... The author thanks to Dr. Noer Azzam,