Interdependence between Islamic capital market and money
market: Evidence from Indonesia
Gandhi Anwar Sania1 and Imam Wahyudib2
a
,b
Inspectorate General, Ministry of Finance Republic of Indonesia
Department of Management, Faculty of Economic, University of Indonesia
Abstract
This study investigate VAR Toda-Yamamoto causality test between macroeconomic variables
and Islamic financial market. The purpose of this study is to analyze the information content
of Islamic capital market and Islamic money market return with respect to several
macroeconomic indicators. Using bivariate method, we found that Islamic capital market
index (JII) has more content information than Islamic money market index (SBIS). The
exchange rate and interest rate (SBI) significantly affected JII. Otherwise, none of the
macroeconomic variables found to significantly affect SBIS. Using multivariate method,
SBIS has more content information (SBI, industry production index, and inflation rate) than
JII (exchange rate). Contradiction in these findings indicates the presence of (i) interaction
between the macroeconomic variables, (ii) interaction between the financial market and the
macroeconomic variables, and (iii) interaction between the Islamic capital market and money
market. Further, by considering these interactions, SBIS tends to have more information
contents than JII, thus more suitable for use as a barometer of monetary policies in Indonesia.
Key words: Islamic finance, capital market, money market, monetary, causality.
1. Introduction
In financial theory, the performance of money markets and stock markets are supposed to be
negatively correlated (Friedman, 1988). An increase in the interest rate will be followed by
increase in the opportunity cost of the cash holding, and it will trigger the substitution effect
1
Correspondence address: Juanda II Building, floor 5, Inspectorate General, Ministry of Finance Republic of
Indonesia, telp. (+628159792633); email: [email protected] or [email protected].
2 Correspondence address: Department of Management Building, floor 2, Faculty of Economics, University of
Indonesia, telp. (+6221 7272425); fax. (+6221 7863556); email: [email protected].
This paper was also presented at The 15th Malaysian Finance Association Conference on June 2013.
1
on stocks and other interest bearing securities. At the same time, the increase of interest rate
in the market will increase the discount rate in the valuation model as reflected on the
nominal risk-free rate, therefore leads to an inverse relationship between interest rates and
stock prices (Maysami and Koh, 2000). The change of interest rate is positively correlated
with the growth in money supply and expansion of the real sector (Mishkin, 2004), and it
leads to an increase in the firm’s cash flows and stock price, but when the level of increase in
the cash flow is not same with the increase in discount rate as well as in the nominal risk-free
rate (DeFina, 1991), the increase in the discount rate will decrease the stock price (Maysami
and Koh, 2000). This relationship emerged as the result of the application of interest rate
regimes. On the other side, when the uncertainty of returns on equity is high and resulted in
lower stock market performance, investors immediately shift their funds to money markets
through investments in fixed income instruments. It is natural to assume that investing in
money markets is much safer than investing in stock markets, when the market is volatile.
Both money markets and stock markets are believed to play an important role in transmitting
monetary and fiscal policies in a country.
The main actors in money markets are banks, in which they can act as an issuer of
financial instruments, the buyer or the seller. Before 1992, only conventional banks existed in
Indonesian banking industry. But since 1992, Islamic banks have emerged into the
Indonesian banking system and the regime of dual monetary systems began to apply. Unlike
conventional banks, Islamic bank’s fundamental idea is to abolish the system of interest and
to apply the system of profit-loss sharing. In the profit-loss sharing system, the return is
calculated based on the real income of the clients’. In this system, the growth of circulated
money in the economy should follow the growth of the output occurring. Thus, in the area
where the dominant actor is Islamic banks, it is expected that the money market should have
independent performance over the movements in market interest rates. The dual monetary
system implies that Bank Indonesia as the regulator are responsible to manage the dual
monetary policies; the conventional monetary policies through Bank Indonesia Certificates
(SBI) and the Islamic monetary policies through Bank Indonesia Sharia Certificates (SBIS),
all at once (Ascarya, 2012).
Since 2000, JII has been established as a reference to the performance of Islamic
stocks in the Indonesian capital market and should be established based on the principles of
Islamic finance. However, in practice, both returns are indicated to be driven by the same
factors, namely SBI rate (see Figure 1). In fact, SBI rate is a form of interest rate, and as such
also usury, which is explicitly excluded from any Islamic financial system. Thus both the
2
returns of JII and SBIS should be independent from the movement of SBI’s return. Moreover,
JII and SBIS should reflect the principles of profit-loss sharing based on actual returns on the
real sector. Therefore, JII and SBIS should be more influenced by macroeconomic factors
than the SBI rate. To prove this, we will examine the relationship between macroeconomic
variables and returns of JII and SBIS. Then, to see the effectiveness of the use of either JII or
SBIS as an indicator of the success of the central bank's monetary policy, we will examine
both factor to see which of the two has more information content.
The rest of this paper will be arranged as follows. Section 2 is the institutional
background and literature review. Section 3 consists of a conceptual framework. Section 4 is
the data. Section 5 is the methodology. Section 6 is an analysis of the results. Section 7 is the
conclusion and managerial implications.
2. Institutional background and literature review
2.1 Indonesian Islamic stock market
Islamic capital market, as a part of Islamic economic system, serves to increase efficiency in
the management of resources and capital, as well as to support investment activities (Ali,
2005). The products and activities of capital markets should reflect the principles of Islam,
based on the principles of trust, and the presence of real assets or activities as an underlying
object. In addition, transactions in the capital market should be managed in a fair and
equitable distribution of benefits. Prohibition of riba, gharar, maysir, tadlis and ikraha in all
activities related to muamalah is the fundamental principle supported by some other
principles, namely the principle of risk sharing, prohibition of speculative behavior,
protection of property rights, transparency, and fairness in agreement contracts (Iqbal and
Tsubota, 2006). Some fundamental differences between conventional and sharia capital
markets are shown in Table 1.
Table 1 Differences between investment in Islamic capital markets and conventional capital
markets
No.
Islamic
Conventional
1
Limited to sectors not prohibited/ included in
negative lists of sharia investment and not on
the basis of debt (debt-bearing investment)
Based on sharia principles that encourage the
application of profit-loss sharing partnership
schemes
Prohibiting various forms of interest,
speculation, and gambling
Free to choose either the debt-bearing
investment or the profit-bearing investment
across sectors
Based on the principle of interest
2
3
3
Allowing speculation and gambling which
in turn will lead to uncontrolled market
fluctuations
No.
4
5
Islamic
Conventional
The existence of sharia guidelines governing Investment guidelines in general in capital
various aspects such as asset allocation, market legal products
investment practices, trade and income
distribution
There is a screening mechanism of companies which must follow sharia principles
Performance of Islamic stock markets in Indonesia is measured by the Jakarta Islamic
Index (JII). JII was established on the 2nd of July 2000 and consisted of 30 stocks based on
several criteria established by the Capital Market Supervisory Agency and Financial
Institution (Bapepam & LK) in accordance with the Decree of the Head of Bapepam & LK
Number: KEP-181/BL/2009 on the Issuance of Sharia Securities which then was amended by
the Decree of the Head of Bapepam & LK Number: KEP-208/BL/2012 on Criteria and
Issuance of Sharia Securities Lists. Additionally since 2003, the National Sharia Board
(DSN), MUI has issued Fatwa Number: 40/DSN-MUI/X/2003 on Capital Markets and
General Guidelines for the Implementation of Sharia Principles in the Capital Market.
Issuers/Public Companies shall meet the qualitative and quantitative criteria in order that their
stocks can be classified in the JII group.
a. Qualitative criteria, issuers/public companies do not conduct business activities in the
field of gambling and games classifies as gambling; trade banned according to sharia;
usury financial services; trading risks containing gharar and maysir; producing,
distributing, trading, and/or providing forbidden (haram) goods or services;
b. Quantitative criteria, the total interest-based debt is compared to the total assets which
are not more than 45%, or the total interest income and the other forbidden (haram)
income are compared to the total revenue and other income which is not more than 10%.
JII was established in order to serve as a benchmark for the performance of
investments in Islamic stocks. In addition, it is expected to increase investors’ confidence in
developing Islamic equity investments. Based on the data from Bapepam & LK (now the
Financial Services Authority), the performance of Islamic stocks indices (JII) indicated
significant growth. Growth occurred concerning JII monthly closing value of 662.189%, i.e.
70.459 points in January 2002 into 537.031 points in December 2011. Figure 1 shows the
value movement of the JII closing index and the SBI and SBIS over the period of 2002-2011.
In 2008 - 2009, JII closing values decreased dramatically, and at the same time there was an
increase (although not significant) in the returns of SBI and SBIS. Apparently during this
period, the subprime mortgage crisis was occurring in the U.S. This joint incident indicates
4
that the crisis in the U.S. has affected Indonesia through the transmission mechanism in
several sectors such as trade, investments, and banking.
Figure 1 Movement of the JII closing index, SBI and SBIS over 2002-2011
(in Rupiah)
JII
SBI rate
SBIS rate
600
500
400
300
200
100
0
Data sources: JII closing index from Bloomberg; SBI and SBIS from Bank Indonesia. SBI
and SBIS are presented in percentage multiplied by 10.
2.2 The theory of market integration
Based on the difficulty and the easiness of capital flows in a country, market integration
could be classified into three categories: full integrated markets, mild segmentation and full
segmentation (Husnan and Pudjiastuti, 1994; Choi and Rajan, 1997). The more restrictions
given by the government, the more it will lead to segmented markets and vice versa. Beckers,
Connor and Curds (1996) defined market integration through three approaches: based on
restrictions in investment from international investors, such as regulatory, fiscal (taxes), or
administrative impediments; based on the consistency of the stock pricing across capital
markets; and based on the correlation between the stock returns across different capital
markets.
Armanious (2007) stated that a capital market is considered to be integrated with
another capital market if they have an ongoing balance relationship. In other words, there are
co-movements among those capital markets indicating the presence of integration among
those capital markets. Thus, one capital market can be used as a measuring instrument to
5
predict the return of the other capital market(s). Heaney, Hooper and Jaugietis (2002)
considered the integration of regional capital markets as a condition in which investors can
buy and sell stocks in any market without any restrictions. Identical securities can be issued,
redeemed, and sold in all the markets in the region (Dorodnykh, 2012). This condition
implies that identical securities are issued and sold at the same price in any capital market, of
course, after adjusting the value to the one of the applicable currency.
2.3 The effects of macroeconomic factors on the financial markets
Macroeconomic represents the interaction among the labor market, the goods market, and the
market of economic assets as well as the interaction among countries in the trading activities
of one particular country and others (Dornbusch, Fischer, and Startz, 2008). In globalization,
the shock occurring within the economy of a given country can lead to contagion effects of
the economies of the other countries. When the domestic economy is integrated with the
world economy, it becomes vulnerable to any turmoil within the global economy.
In the business scope, macroeconomic factors influence the increase or the decrease in
the company's performance, both directly and indirectly. Likewise, Yusof and Abdul Majid
(2007) stated that the conditions of the capital markets as a leading indicator for the economy
of a country are intertwined with the country's macroeconomic conditions. In other words,
when a country's macroeconomic conditions change, investors calculate their impact on the
company’s performance in the future, and then make the decision to buy or sell the stocks
they own. This action of stock trading results in the changes of stock price, and ultimately
affects the indices in the stock market.
2.4 Transmission of monetary policy in Indonesia
Among the main factors influencing macroeconomic conditions is the monetary policy. To
achieve and maintain the stability of the Rupiah by controlling inflations, Bank Indonesia
may conduct monetary policy using two financial instruments that have similar functionality,
but different in features. The first is Bank Indonesia Certificates (SBI) which belongs to an
interest-based instrument. The second is Bank Indonesia Sharia Certificates (SBIS) which is
an Islamic-based instrument. The use of SBI and SBIS, as a policy rate, is essential for the
Central Bank as monetary policy makers. With these instruments, Bank Indonesia can
influence a bank’s funding and financing behaviour through the interbank money markets,
through both the conventional instrument (PUAB) and the Islamic instrument (PUAS), and
ultimately affect the cost of the banking funds in distributing credits or financing.
6
Furthermore, the expansion of credits and financing will increase the outputs and affect the
rate of inflations (Ascarya, 2012). The chronology of dual monetary policy transmission
(sharia and conventional) in Indonesia can be shown in Figure 2.
In Indonesia, Ningsih (2011) discovered that the transmission route of sharia
monetary policy is similar to one of the conventional monetary policy; where PUAS becomes
the operational target driven by SBIS in line with Hidayat (2007), where he found that SBI
affect SBIS positively. In addition, Wahyudi and Rosmanita (2011) explained that shifts in
the convergence trend of SBIS’s returns and SBI’s interest rates were triggered by an attempt
of equal treatment made Bank Indonesia over Islamic banks and conventional banks.
Typically, the transmission mechanism of monetary policy through the interest rate channel
starts from the short-term interest rates and then spreads to the medium-term interest rates
and to the long-term interest rates (Moschitz, 2004).
Figure 2 Transmission route of dual monetary policy running by Bank Indonesia
Transmission in financial sector
SBI
PUAB
Interest
rate
Deposits &
credits
SBIS
PUAS
Profit/loss
sharing
Deposits &
financing
Monetary
policies
Transmission
in real sector
- real assets price
- property price
- gold price
- others commodity price
- etc
Financial assets price
- bond yield
- sukuk yield
- stock price
- mutual funds
- etc
Consumption
Production
Investment
Aggregate
demand &
supply
Output gap
Inflation &
Rupiah’s
stabilization
The transmission mechanism of monetary policy began when Bank Indonesia change
its instruments and further affected the operational goals, the intermediate goals and the final
goals. Monetary policy transmitted through the interest rate channel can be described in three
stages. First is the transmission in the financial sector. Changes in monetary policy through
changes in monetary instruments (SBI and SBIS) influence the returns of PUAB (PUAS),
deposits and credits (financing), and ultimately affect the price of financial assets (Ascarya,
2012). Second is the transmission from the financial sector to the real sector, in which the
impact depends on consumption and investments. The effect of interest rates on consumption
7
occurs through the income effect and the substitution effect. Interest rates will also affect the
cost of capital, which in turn will affect investments and production. These effects of the
interest rate will have an impact on aggregate demands, and if it is not followed by the
aggregate supply, it will cause the output gap and finally affects inflations and the stability of
the Rupiah.
Third is the transmission in the real sector through asset prices. Asset prices, as a
measure of economic activities, are measured in two ways: wealth effects and yield effects
(Mishkin, 2004). In addition, asset prices also contain information content that can describe
the expectations of revenue streams and the direction of inflation in the future. At least there
are two theories related to the monetary policy transmission mechanism through asset prices:
Tobin's Q theory of investment and wealth effects of consumption. In Tobin's Q theory,
monetary policy is transmitted through the influence of equity assessment. The ratio Q is
defined as the market price of the company relative to the replacement cost of capital.
Monetary expansion will increase the firm's stock price expectations, raise the ratio Q,
encourage firms to invest, and eventually have an impact on aggregate demands and supplies.
Based on the wealth effects of consumptions, monetary policy will affect consumer-spending
decisions. Based on the life cycle hypothesis from Ando and Modigliani (1963), consumption
is not constant in the long run given that wealth—measured by financial assets (such as
stocks, bonds, and deposits) and real assets (such as gold, properties, and land)—is not
constant forever. Monetary expansion will increase the price of financial assets and real
assets, causing consumers’ wealth to rise, and it will ultimately increase consumption and
aggregate demand.
In addition to transmission through interest rates and asset prices, monetary policy
transmission mechanism may take place through the exchange rate channel. Perfect capital
mobility will lead to the relationship between short-term interest rates and the exchange rates
(Mundel, 1962). This is called the interest parity relationship, in which differences in interest
rates between two countries equal to the expected changes in the exchange rates between the
two countries. If the equilibrium has not been reached, capital inflows in the country with
high interest rates will continue to happen.
3. Conceptual framework
Research on the relationship among macroeconomic variables, financial markets, and the
price of oil has been done for a long time, especially in developed countries such as the U.S.,
8
the U.K., and Japan. Preliminary research on stock market returns by Castanias (1979) and
Hardouvelis (1988) found that economic announcements to be the factor influencing stock
market returns. Furthermore, Levine and Zervos (1996), Hooker (2004), as well as Chiarella
and Gao (2004) stated that stock market returns are influenced by macroeconomic indicators
such as GDP, productivity, and interest rates.
Regarding the relationship between capital market returns and oil price movements,
Sadorsky (1999), Filis (2010) and Degiannakis, Filis and Floros (2013) stated that the
determinant factor of the capital market returns is the oil price. Park (2008) found that almost
all countries in America and Europe (except for Finland and the UK) empirically indicated
that the capital market returns responded significantly to the oil price. In almost all oilimporting countries, shock in oil prices significantly contributed negative to the capital
markets. Meanwhile, among the oil-exporting countries, only Norway that showed a
significantly positive response to the stock markets over the oil-price shock— where UK was
insignificant and Denmark was significantly negative. Furthermore, Park (2008) explained
that in the oil-importing countries, changes in the oil price contributed a greater impact on
stock market returns than the interest rates.
In the financial markets in Asia, Abdul-Rahman, Sidek and Tafri (2009) examined the
interaction of selected macroeconomic variables (i.e., money supply, exchange rates,
reserves, interest rates, and the industrial production index) and conventional capital market
returns (KLCI) in Malaysia using VAR methodology. They found that there was a cointegration relationship between conventional capital market returns and the rest of the
selected macroeconomic variables. Conventional capital market returns were very sensitive to
changes in macroeconomic variables. Furthermore, in accordance with the variance
decomposition analysis, it was shown that the conventional capital market had a strong
dynamic interaction with the reserves and the index of industrial production.
Abdul Majid and Yusof (2009) using the methodology of autoregressive distributed
lag (ARDL) examined the long-term relationship between the Islamic capital market returns
(KLSI) and selected macroeconomic variables (i.e., money supply, exchange rates, interest
rates, the industrial production index, and the Fed rate). They found that the macroeconomic
variables that significantly affected the stock price movements belonging to KLSI were
exchange rates, money supply, interest rates, and the Fed rate.
Hussin, Razak and Yahya (2012) conducted a study on the relationship between
exchange rates, oil prices, and the Islamic capital market (FBMES) in the period of January
2007 – December 2011 in Malaysia by using VAR methodology. They found that the Islamic
9
stock returns had a positive and significant correlation with oil prices, but it had a negative
and insignificant relationship with the exchange rate of Malaysian ringgit to the U.S. dollar.
Furthermore, through the test of Granger causality, it was found that oil prices could predict
short-term sharia stock returns. However, regarding the exchange rate of Malaysian ringgit to
the U.S. dollar, both for short terms and long terms, there was insufficient evidence to be able
to predict the movement of sharia stocks in Malaysia.
Based on the empirical findings from previous studies, we formulate the conceptual
framework of this research in to build the model and the analysis as follows.
Figure 3 Conceptual framework
Production
Index
Output Gap
Domestic
Inflation
Pressure
Outstanding
Money
Inflation
Foreign
Inflation
Pressure
Exchange
Rate
SBI Rate
Expected
Return
Return of SBIS
Stock Price
Oil Price
The Fed
JII
4. Data
This study uses monthly macroeconomic data over the periods of January 2002 – December
2011. Indonesia's macroeconomic data used consists of money supply (M2), Rupiah
exchange rates against USD (ER), SBI rate (rSBI), Industrial Production Index (IPI), and
inflations (INF). Whereas, the global macroeconomic data consists of the world oil price
(OP) and the Fed rate (FRR). The data on sharia financial markets used in this study are the
closing value of the Jakarta Islamic Index (JII) and the return of SBIS (rSBIS).
5. Methodology
5.1 The causality test of Toda-Yamamoto (1995)
In the Granger causality test, to predict the time interval at the co-integrated variables will
result in spurious regression, in addition, F-test becomes invalid unless the variables are co-
10
integrated (Shirazi and Abdul Manap, 2005; Akçay, 2011). Although these problems can be
solved through the method of error correction model (Engle and Granger, 1987) and vector
autoregressive error correction model (Johansen and Julius, 1990; Johansen, 1991), but these
approaches are considered not practical and sensitive to the value of the parameters in a
limited number of samples and the results are unreliable (Zapata and Rambaldi, 1997).
Toda-Yamamoto (1995) developed Granger causality test procedures to estimate the
augmented VAR with the asymptotic distribution of the statistic MWald. These procedures
can be done when there is co-integration within the models. Büyükşalvarci and Abdioğlu
(2010) said that the methods developed by Toda-Yamamoto (1995) can be used to find the
long-term causality among unintegrated variables in the same order. In addition, the causality
test procedures of Toda-Yamamoto (1995) make the Granger causality test easier, and do not
require the existence of co-integration testing or the transformation of VAR into VECM.
5.2 VAR Toda-Yamamoto bivariate model
Bivariate models are used to test the causality of each financial market variables (JII and
rSBIS) with each macroeconomic variable in turn. The equation of Toda-Yamamoto bivariate
model is:
𝑑1max
𝑁
lnJIIt = 𝛽0 + ∑𝐾
𝑘=1 𝛽k lnJII𝑡−𝑘 + ∑𝑑1=1 𝛽K+d1 lnJII𝑡−𝐾−𝑑1 + ∑𝑛=0 𝛿n rSBIS𝑡−𝑛 +
∑𝑑2max
𝑑2=1 𝛿𝑁+𝑑2 rSBIS𝑡−𝑁−𝑑2 + 𝜏1 𝑋𝑖𝑡 + 𝜀𝑡
.
𝑑1max
𝑁
rSBISt = 𝜃0 + ∑𝐾
𝑘=1 𝜃k rSBIS𝑡−𝑘 + ∑𝑑1=1 𝜃K+d1 rSBIS𝑡−𝐾−𝑑1 + ∑𝑛=0 πn lnJII𝑡−𝑛 +
∑𝑑2max
𝑑2=1 π𝑁+𝑑2 lnJII𝑡−𝑁−𝑑2 + 𝜑1 𝑋𝑖𝑡 + 𝜐𝑡
.
where: k and n are the lagged period in VAR model; d1max and d2max are the integration’s
order of time series data obtained using stationarity test procedures; Xj is is the
macroeconomic variable where i = lnM2, lnER, rSBI, IPI, INF, FFR or lnOP; β0 and θ0 are
the constant; βk, δn, τ1, θk, πn and φ1 are the coefficient of variables; εt and υt are the white
noise disturbance term that assumed that E(εt) or E(υt) = 0, and E(εt υt) = 0.
5.3 VAR Toda-Yamamoto multivariate model
Multivariate models are used to test the causality of financial market variables (JII and
rSBIS) with all the macroeconomic variables simultaneously. The equation of TodaYamamoto multivariate model is:
𝑑1max
𝑁
lnJIIt = 𝛽0 + ∑𝐾
𝑘=1 𝛽k lnJII𝑡−𝑘 + ∑𝑑1=1 𝛽K+d1 lnJII𝑡−𝐾−𝑑1 + ∑𝑛=0 𝛿n rSBIS𝑡−𝑛 +
𝐼
∑𝑑2max
𝑑2=1 𝛿𝑁+𝑑2 rSBIS𝑡−𝑁−𝑑2 + ∑𝑖=1 𝜏i X 𝑖𝑡 + 𝜀𝑡
11
.
𝑑1max
𝑁
rSBISt = 𝜃0 + ∑𝐾
𝑘=1 𝜃k rSBIS𝑡−𝑘 + ∑𝑑1=1 𝜃K+d1 rSBIS𝑡−𝐾−𝑑1 + ∑𝑛=0 πn lnJII𝑡−𝑛 +
𝐼
∑𝑑2max
𝑑2=1 π𝑁+𝑑2 lnJII𝑡−𝑁−𝑑2 + ∑𝑖=1 𝜑i X 𝑖𝑡 + 𝜐𝑡
.
where: k and n are the lagged period in VAR model; d1max and d2max are the integration’s
order of time series data obtained using stationarity test procedures; Xi are the
macroeconomic variable where i = lnM2, lnER, rSBI, IPI, INF, FFR and lnOP; β0 and θ0 are
the constant; βk, δn, τ1, θk, πn and φ1 are the coefficient of variables; εt and υt are the white
noise disturbance term that assumed that E(εt) or E(υt) = 0, and E(εt υt) = 0.
5.4 Stationarity testing
Toda-Yamamoto (1995) causality test starts with testing the stationarity of data to determine
the order of integration of time series data into stationary ones. Based on these test results, the
value of dmax which is a vital component in determining time intervals on Toda-Yamamoto
(1995) methods is generated. In this study, the stationarity test was done by using two
methods: the augmented Dickey-Fuller test (ADF test) and Phillips-Perron test (PP test).
5.5 The optimum lag (k) testing using the VAR model
There are several parameters to determine the optimum lag (k), namely the Akaike
Information Criterion (AIC), Schwarz Information Criterion (SIC), and Hannan Quinn
Criterion (HQC). In this study, the optimum lag test was conducted by SIC and HQC. SIC is
more consistent in the lag selection compared to AIC. The greater the sample number in the
model is, the more conical to the SIC probability of the selection of the appropriate lag.
Conversely, AIC is not consistent with the number of the samples used (Shittu and Asemota,
2009). Hurvich and Tsai (1989) stated that HQC equals to SIC, they are consistent in the
selection of the lag in the estimation model. If the SIC and HQC show differences in the
selection of lag, this study will refer to the SIC.
5.6 Establishment of optimum lag based-VAR system
Toda-Yamamoto causality test (1995) was preceded by the establishment of the VAR model
based on the new optimum lag (p). The p-value is derived from the value of k plus dmax (p = k
+ dmax). If the stationary time series data in the first derivation (dmax = 1) and the optimum
time interval of VAR model (k) equal to 1, the time interval for the Toda-Yamamoto equals
to 2 (p = k + dmax = 1 + 1 = 2).
12
5.7 Estimation of VAR models with the methods of seemly unrelated regression (SUR)
VAR models (bivariate and multivariate) that have been formed, then are estimated using the
SUR method with an interval p = k + dmax. Rambaldi and Doran (1996) said that MWald test
to test Granger causality can improve efficiency when the SUR method is used in the
estimation (Akçay, 2011). Furthermore, Alaba, Olobusoye and Ojo (2010) stated that the use
of SUR model is more efficient than ordinary least square (OLS) and two-stage least square
(2SLS). SUR method takes into account the correlation between errors in the equation
measured at the same time, while the OLS assumes errors as something independent (Alaba
and Olobusoye, 2010).
5.8 Restrictions on the parameters with the modified Wald (MWald) test
After estimating the VAR models (bivariate and multivariate) with SUR model, the next step
is to make restrictions on the parameter by using the MWald test procedures. This MWald
test becomes the differentiation between Toda-Yamamoto (1995) causality test and the
Granger causality test.
6. Analysis Results
6.1 Analysis of stationarity test results
The results of the stationarity test of each variable can be seen in Table 2. Based on the ADF
test it was found that all the variables are not stationary at level. Whereas in the PP test, there
is only one variable that is stationary in nature at the level, which is LnER. Thus, differencing
across all variables to obtain stationary data is necessary. At the first differencing, seven
variables have been stationary, which is LnJII, rSBIS, LnER, rSBI, IPI, INF and LnOP, and
the rest, which is LnM2 and FFR, have not been stationary. After conducting the second
differencing, all variables have been stationary.
Table 2 Results of stationarity test
Research
variable
LnJII
rSBIS
LnM2
LnER
rSBI
Level
-1.22
-1.18
1.43
-2.57
-2.75*
ADF test
1 different 2nd different
-8.44***
-10.00***
-12.91***
-8.42***
-2.46
-7.83***
-9.08***
-11.01***
-4.90***
-13.12***
st
13
Level
-1.21
-2.48
2.78
-2.93**
-2.49
PP test
1 different
-8.58***
-15.10***
-11.47***
-8.99***
-4.93***
st
2nd different
-22.70***
-39.05***
-47.43***
-20.28***
-15.34***
Research
variable
IPI
INF
FFR
LnOP
Level
0.61
-2.65*
-2.79*
-2.03
ADF test
1 different 2nd different
-4.62***
-8.74***
-8.92***
-10.38***
-2.80*
-14.41***
-8.62***
-12.11***
st
Level
-2.64*
-2.47
-0.97
-2.17
PP test
1 different
-23.56***
-8.95***
-14.49***
-8.60***
st
2nd different
-185.45***
-27.81***
-68.60***
-36.50***
Note: *** indicate significant at 1%, ** significant at 5%, and * significant at 10%.
6.2 Analysis in determining optimum lag
In determining the optimum lag on the VAR system, we employ SIC and HQC criteria (see
Table 3). Using the lag optimum test results (k) on the bivariate and multivariate VAR and
the maximum order of integration (dmax), Toda-Yamamoto lag is calculated by summing the
two components (see Table 4).
Table 3 Summary of lag optimum VAR using SIC and HQC criteria
Optimum lag (k)
SIC
HQC
2
0
1
1
0
4
0
2
1
1
1
1
4
1
3
0
2
2
0
4
0
3
2
2
2
2
4
2
1
1
2
2
Bivariate: LnJII and other variables
LnJII and LnM2
LnJII and LnER
LnJII and rSBI
LnJII and IPI
LnJII and INF
LnJII and FFR
LnJII and LnOP
Bivariate: rSBIS and other variables
rSBIS and LnM2
rSBIS and LnER
rSBIS and rSBI
rSBIS and IPI
rSBIS and INF
rSBIS and FFR
rSBIS and LnOP
Multivariate
LnJII and all variables
rSBIS and all variables
14
Table 4 Calculation of Toda-Yamamoto lag
Optimum lag (k)
Maximum order of Toda-Yamamoto lag
integration (dmax)
(p=k+dmax)
Bivariate: LnJII and other variables
LnJII and LnM2
LnJII and LnER
LnJII and rSBI
LnJII and IPI
LnJII and INF
LnJII and FFR
LnJII and LnOP
2
0
1
1
0
4
0
2
1
1
1
1
2
1
4
1
2
2
1
6
1
3
0
2
2
0
4
0
2
1
1
1
1
2
1
5
1
3
3
1
6
1
1
1
2
2
3
3
Bivariate: rSBIS and other variables
rSBIS and LnM2
rSBIS and LnER
rSBIS and rSBI
rSBIS and IPI
rSBIS and INF
rSBIS and FFR
rSBIS and LnOP
Multivariate
LnJII and all variables
rSBIS and all variables
6.3 Content Analysis using bivariate VAR model: LnJII and rSBIS
After the degree of integration (dmax) and the optimum lag (k) of each relationship pattern
between rSBIS and LnJII with macroeconomic variables are obtained, both in bivariate ways
and in multivariate ways, then the next step is to develop a new VAR model based on the lag
method of Toda-Yamamoto (p = dmax + k). After the new VAR model is formed, estimation
employing the method of seemingly-unrelated regression (SUR) to examine the causality
relationship of each macroeconomic variable with LnJII and rSBIS. The null hypothesis is
that either rSBIS or LnJII does not result in one-way Granger causality to any
macroeconomic variables. The criteria to reject the null hypothesis employ the Wald statistic
test. Rejection of the null hypothesis suggests that there is a causality relationship between
LnJII and rSBIS with the macroeconomic variables. The results of Toda-Yamamoto (1995)’s
bivariate causality test with the interval (p) are shown in Table 5.
15
Table 5 Result summary of Toda-Yamamoto’s bivariate causality test
Dependent variable
LnM2
LnER
rSBI
IPI
INF
FFR
LnOP
LnJII
Wald test Probability
0.91
0.34
8.51
0.0035***
5.49
0.019**
0.13
0.71
0.31
0.57
2.24
0.13
2.47
0.12
rSBIS
Wald test Probability
0.54
0.46
0.21
0.65
0.07
0.79
1.05
0.30
0.41
0.52
0.66
0.42
1.22
0.27
Note: *** indicate significant at 1%, ** significant at 5%, and * significant at 10%.
Based on Table 5 it can be concluded that with the significance level of 5%, the
sharia capital market returns incorporated in the JII have more information content related to
macroeconomic variables than the returns of SBIS. Based on the results of Toda-Yamamoto’s
bivariate causality test, LnJII leads to one-way Granger causality over two macroeconomic
variables, namely changes in the exchange rate of Rupiah against the U.S. dollar (lnER) and
SBI interest rates (rSBI).
The test results in Table 5 can be explained using the mechanism of monetary policy
through the exchange rates as shown in Figure 2 and 3. Mundel (1962) described how
monetary policy affects exchange rates and therefore affects output and inflation. Mundel
(1962) showed that perfect capital mobility will lead to the relationship between short-term
interest rates and exchange rates. The differences in interest rates between two countries
equal to the expected changes in the exchange rates between the two countries. Capital inflow
in countries with high interest rates will continue to occur if the equilibrium has not been
reached.
In other words, if Bank Indonesia carries out monetary operations by making changes
in interest rates in domestic money markets, there will be differences in interest rates at home
and abroad. It will have an impact on the magnitude of the flow of funds from and to abroad,
one of which can be represented from the movement of stock markets (in this study, JII).
Such movement of capital flows is closely related to the supply and demand of money which
in turn will influence exchange rates and inflation targets. Then, Bank Indonesia will respond
to the rate of inflation through the monetary instrument namely SBI to obtain the level of
inflation target targeted.
16
On the other hand, regarding the test of SBIS’ returns of the same significance level of
5%, as shown in Table 5, rSBIS does not causes one-way Granger causality toward the
macroeconomic variables tested. From the results of Toda-Yamamoto’s bivariate causality
test, it can be concluded that the JII contains more information related to macroeconomic
variables than the SBIS. This suggests that the JII is more able to describe the movement of
macroeconomic variables if compared to SBIS. Therefore, based on the test results of the
bivariate model, the JII is worth considering as a policy indicator better than rSBIS.
6.4 Analysis of multivariate VAR models using the new time intervals (p)
The use of only two variables in the previous testing system of Toda-Yamamoto causality
methods of bivariate models remains possible, allowing misspecification in the model. Thus,
it is necessary to estimate the multivariate model (Kassim and Abdul-Manap, 2008). The
multivariate model is a causality test of Toda-Yamamoto method conducted between LnJII
and rSBIS with the overall macroeconomic variables. The causality test results of TodaYamamoto method of multivariate models are presented in Table 6 below.
Table 6 Result summary of Toda-Yamamoto’s multivariate causality test between
LnJII/rSBIS toward the macroeconomic variables
LnJII
rSBIS
Dependent
variable
Wald test
Probability
Wald test
Probability
LnM2
2.60
0.11
1.20
0.27
LnER
6.65
0.01***
0.19
0.66
rSBI
2.83
0.09*
0.15
0.70
IPI
0.00
0.99
4.62
0.03**
INF
0.90
0.34
0.14
0.71
FFR
0.01
0.93
0.76
0.38
LnOP
0.39
0.53
0.01
0.90
Note: *** indicate significant at 1%, ** significant at 5%, and * significant at 10%.
Based on Table 6, from the results of Toda-Yamamoto’s multivariate causality test, it
can be concluded that with the significance level of 5%, LnJII triggers one-way Granger
causality over macroeconomic variables, namely changes in the exchange rates of Rupiah
against the U.S. dollar (lnER). With the significance rate of 10%, LnJII results in one-way
Granger causality over the interest rates of the SBI (rSBI). This multivariate test result is
consistent for the macroeconomic variables and supports the previous results of TodaYamamoto (1995)’s bivariate causality test (Table 5). In Toda-Yamamoto’s bivariate test,
17
LnJII leads Granger causality over LnER and rSBI at the same significance level of 5%.
Different from the results of Toda Yamamoto’ bivariate causality test on rSBIS over
macroeconomic variables (Table 5), based on the the results of Toda-Yamamoto’s
multivariate causality test in Table 6 it can be assumed that with the significance level of 5%
rSBIS triggers one-way Granger causality over macroeconomic variable specifically the
Industrial Production Index (IPI). IPI is a proxy for economic growth reflecting the quantity
of national industry products by assessing the physical quantity of the output. In other words,
IPI represents the changes in the economy of Indonesia and becomes one of the indicators for
the economic growth of Indonesia. The presence of the relationship of one-way Granger
causality between the rSBIS and the IPI indicates that Islamic finance, especially sharia
money markets along with their systems and instruments, have something to do with the
development of the real sector that has a real impact on the economy.
The rSBIS leads to Granger causality over only one macroeconomic variable, namely
the Industrial Production Index (IPI). It can be explained that the transmission of sharia
monetary policy through the financing way in Indonesia Islamic banking has the ultimate
goal of economic growth and the stability of the money value (Ascarya, 2012)). Simply put,
this can be written down as follows:
IPI = f (IFIN, IDEP, PUAS, SBIS)
Where: IPI is the industrial production index, IFIN is the sharia banking financing; IDEP is
the DPK of Islamic banks; PUAS is the interest rate in a certain day in the Islamic interbank
money market; SBIS is the returns of Bank Indonesia Sharia Certificate.
From the function of IPI, and Figure 2, it can be seen that when the monetary
authority, in this case BI, establishes policies concerning the interest rates of SBI, it will
affect the returns of SBIS and further affect the returns of PUAS. This PUAS will affect the
level of the sharing of savings, deposits and financing (PLS). When the sharing of
savings/deposits set by the bank increases, then by assuming the ceteris paribus of the
number of the DPK of Islamic banking (IDEP) owned by the bank will also increase.
Furthermore, the bank will use this excess of liquidity by distributing it in the form of
financing or credit (IFIN) to move the real sector of Indonesia (IPI).
From the results of Toda Yamamoto’s multivariate causality test, it can be concluded
that if LnJII and rSBIS are tested simultaneously against macroeconomic variables, with the
significance level of 5%, LnJII has more influences on LnER while rSBIS has more
influences on IPI.
18
6.5 Analysis of VAR model formation between macroeconomic variables and LnJII/rSBIS
To see whether the causality relationship that occurs between LnJII/rSBIS and
macroeconomic variables is one-way or two-way, we conducted the Toda-Yamamoto (1995)
causality test with the null hypothesis that macroeconomic variables do not lead to one-way
Granger causality over LnJII or rSBIS. Results summary of Toda-Yamamoto (1995)’s
bivariate causality test between macroeconomic variables with LnJII and rSBIS at the
intervals (p) that have been obtained from Table 3 and 4, are shown in Table 7.
Table 7 Results summary of Toda-Yamamoto’s bivariate causality test between
macroeconomic variables and LnJII/rSBIS
LnJII
rSBIS
Independent
variable
Wald test
Probability
Wald test
Probability
LnM2
0.70
0.40
0.06
0.81
LnER
2.83
0.09*
0.58
0.45
rSBI
2.06
0.15
0.12
0.73
IPI
0.03
0.87
0.00
0.96
INF
2.13
0.14
0.80
0.37
FFR
0.07
0.80
0.00
0.96
LnOP
1.15
0.28
0.40
0.53
Note: *** indicate significant at 1%, ** significant at 5%, and * significant at 10%.
From Table 7, it can be seen that none of the macroeconomic variables that causes the
one-way Granger causality over both LnJII and rSBIS with the significance level of 5%.
However, if the significance level increases to 10%, based on the result of Toda-Yamamoto
(1995)’s bivariate causality test, it can be seen that LnER causes one-way Granger causality
over LnJII. This result supports the research conducted by Abdul Majid and Yusof (2009)
that one of the variables significantly affecting the stock movement of Kuala Lumpur Syariah
Index (KLSI) is the exchange rate (REER), in addition to Board Money Supply (M3),
Treasury Bill Rates (TBR), and Federal Fund Rate (FFR).
Table 8 Results summary of Toda-Yamamoto’s multivariate causality test between
macroeconomic variables and LnJII/rSBIS
Independent
variable
LnM2
LnER
rSBI
IPI
LnJII
Wald test
2.07
0.88
3.10
0.29
rSBIS
Probability
0.15
0.35
0.08*
0.59
19
Wald test
0.44
0.71
5.33
0.05
Probability
0.51
0.40
0.02**
0.82
LnJII
rSBIS
Independent
variable
Wald test
Probability
Wald test
Probability
INF
3.25
0.07*
4.09
0.04**
FFR
1.61
0.20
0.19
0.66
LnOP
5.60
0.02
0.00
0.97
Note: *** indicate significant at 1%, ** significant at 5%, and * significant at 10%.
Table 8 shows the results of Toda-Yamamoto (1995)’s multivariate causality test,
with the significance level of 5%, indicate that rSBI and INF cause one-way Granger
causality over rSBIS. On the contrary to findings in table 7, none of the macroeconomic
variables that leads to one-way Granger causality relationship over LnJII. However, if the
significance level is increased to 10% and the variables of INF and rSBI simultaneously
cause one-way Granger causality over LnJII and rSBIS. Table 9 presents the summary of the
method of Toda-Yamamoto (1995) causality test between LnJII and rSBIS with
macroeconomic variables at the significance level of 5%.
Table 9 Results summary of two-way causality test of Toda-Yamamoto’s bivariate
Macroeconomic variable
LnM2
LnER
rSBI
IPI
INF
FFR
LnOP
LnJII
>--<


>--<
>--<
>--<
>--<
rSBIS
>--<
>--<
>--<
>--<
>--<
>--<
>--<
Table 9 summarizes the results of two-way causality test of Toda-Yamamoto
(1995)’s bivariate methods and shows that there is no macroeconomic variable having a
causality relationship with rSBIS while LnJII has a one-way relationship with LnER and
rSBI.
Table 10 Results summary of two-way causality test of Toda-Yamamoto’s multivariate
Macroeconomic variable
LnM2
LnER
rSBI
IPI
INF
FFR
LnOP
LnJII
>--<

>--<
>--<
>--<
>--<
>--<
20
rSBIS
>--<
>--<



>--<
>--<
Table 10 summarizes the results of two-way causality test of Toda-Yamamoto (1995)’s
multivariate methods and shows that LnM2 has no causality relationship with LnJII and
rSBIS; LnER has a one-way relationship with LnJII and no causality relationship with rSBIS;
rSBI has no causality relationship with LnJII but it has a one-way relationship with rSBIS;
IPI has no causality relationship with LnJII but it has a one-way relationship with rSBIS; INF
has no causality relationship with LnJII but it has a one-way relationship with rSBIS, and
both FFR and LnOP have no causality relationship both with LnJII and rSBIS.
7. Conclusions and Policy Implications
Through the examination of the Toda-Yamamoto (1995)’s VAR causality relationship
between macroeconomic variables and Islamic financial markets, this study is aimed to
examine and compare the information content related to the macroeconomic variables
contained in the sharia capital market (JII) and sharia money market (SBIS). Between the two
Islamic financial variables, one that contains more information is the better candidate for
policy indicators.
Several conclusions can be drawn from these research findings. First, that the twoway bivariate test results between the returns of JII and the ones of SBIS with one of
macroeconomic variables in turn indicate that the JII has two types of information content
related to macroeconomic variables, namely changes in the exchange rate of Rupiah against
dollar (LnER) and changes in interest rates of SBI (rSBI). In the other hand, the test results of
the SBIS variable show that SBIS does not have any information content related to
macroeconomic variables. Second, that the two-way multivariate test results between JII and
SBIS with the whole macroeconomic variables simultaneously indicate that the JII only has
one type of information content related to macroeconomic variables, namely changes in the
exchange rate of Rupiah against dollar (LnER). While SBIS has three types of information
content related to macroeconomic variables, namely changes in the interest rates of SBI
(rSBI), the Industrial Production Index (IPI), and inflation (INF). Third, through the two-way
bivariate test, JII is able to describe the movement of macroeconomic variables better as it
has two types of information content related to macroeconomic variables, namely LnER and
rSBI, while SBIS fails to describe the movement of macroeconomic variables at all. Fourth,
through the two-way multivariate test, SBIS has more information content than JII does. In
other words, Sharia Money Market can describe the movement of the macroeconomics more.
21
From the perspectives of policy makers, these different findings provide some insights
and a variety of policy implications. For example, it was found that LnJII led to one-way
Granger causality over two macroeconomic variables, namely changes in the exchange rate
of Rupiah against the U.S. dollar and the interest rates of SBI. Mundel (1962) shows that
perfect capital mobility will lead to the relationship between short-term interest rates and the
exchange rates. Differences in interest rates between two countries equal to the expected
changes in the exchange rates of the two countries. Capital inflow in countries with high
interest rates will continue to occur if the equilibrium has not been reached. In other words, if
Bank Indonesia carries out monetary operations by making changes in interest rates in
domestic money markets, there will be differences in interest rates at home and abroad. This
will have an impact on the magnitude of the flow of funds from and to abroad, one of which
can be represented from the movement of stock markets (in this study, JII). Such movement
of capital flows is closely related to the supply and demand of money which in turn will
influence exchange rates and inflation targets. Then, Bank Indonesia will respond to the rate
of inflation through the monetary instrument namely SBI to obtain the level of inflation target
targeted.
The second implication is that a relationship of one-way Granger causality between
the returns of SBIS and the ones of IPI was found. IPI is a proxy for the economic growth
reflecting the quantity number of national industrial products by assessing the physical
quantity of the output. In other words, IPI represents the changes in the economy of Indonesia
and becomes one of the indicators for the economic growth of Indonesia. This indicates that
Islamic finance, along with their systems and instruments, has something to do with the
development of the real sector that has a real impact on the economy. That when the
monetary authority, in this case Bank Indonesia, establishes policies concerning the interest
rates of SBI, it will affect the returns of SBIS and further affect the returns of PUAS. This
PUAS will affect the level of the sharing of savings, deposits and financing. When the
sharing of savings/deposits set by the bank increases, then by assuming the ceteris paribus of
the number of the DPK of Islamic banking (IDEP) owned by the bank will also increase.
Furthermore, the bank will use this excess of liquidity by distributing it in the form of
financing or credit to move the real sector of Indonesia.
The third implication is that the contradiction in these findings indicates the presence
of (i) interaction between the macroeconomic variables, (ii) interaction between the financial
market and the macroeconomic variables, and (iii) interaction between the Islamic capital
market and money market. Further, by considering these interactions, SBIS tends to have
22
more information contents than JII and more suitable for use as a barometer of monetary
policies in Indonesia.
8. Suggestions for Further Research
This study has several limitations. For further research, we provide some input as suggestions
that can be adapted. First, about the methodology, which is the Augmented VAR model
developed by Toda and Yamamoto (1995). Although it can project the causality relationship
among the research variables, it is unable to identify the lagging or how long it takes for a
variable in influencing the other variables. Thus, the other types of VAR models can be used
to examine the lagging of each research variable in influencing the other variables. Second, in
relevance to macroeconomic variables, this study uses the variable M2 indicating the number
of broad money supply, the Industrial Production Index (IPI) as a proxy for economic growth,
and OP as the representation of the world oil price in the WTI market (the U.S.). For further
research, similar and identical variables may be selected such as Brent market (UK) for world
oil prices, M1 for narrow money supply, and other variables that can reflect economic growth
considering that the variable IPI does not accommodate the output value of UKM.
References
Abdul Majid, M. S., and R. M. Yusof (2009). Long-run relationship between Islamic stock
returns and macroeconomic variables: An application of the autoregressive distributed
lag model. Humanomics, Vol. 25, No. 2, pp. 127-141.
Abdul Rahman, A., N. Z. M. Sidek and F. H. Tafri (2009). Macroeconomic determinants Of
Malaysian stock market. African Journal of Business Management, Vol.3, No. 3, pp. 95106.
Akçay, S. (2011). Causality relationship between total R&D investment and economic
growth: Evidence from United States. The Journal of Faculty of Economics and
Administrative Sciences, Vol. 16, No.1, pp. 79-92.
Alaba, O. O., E. O. Olubusoye, and S. O. Ojo (2010). Efficiency of seemingly unrelated
regression estimator over the ordinary least square. European Journal of Scientific
Research, Vol. 39, No. 1, pp.153-160.
Ali, S. S. (2005). Islamic capital market products: Developments and challenges. Occasional
Paper No. 9 (IRTI), pp. 1-92.
23
Ando, A., and F. Modigliani (1963). The “life-cycle hypothesis of saving: Aggregate
implications and tests. American Economic Review, Vol. 53, No. 1, pp. 55-84.
Armanious, A. N. R. (2007). Globalization effect on stock exchange integration. SSRN
Working Paper (http://dx.doi.org/10.2139/ssrn.1929045).
Ascarya (2012). The transmission channel and the effectivity of the dual monetary policies in
Indonesia. Monetary Economic and Banking Bulletin, Vol. 14, No. 3, pp. 283-315.
Beckers, S., G. Connor, and R. Curds (1996). National versus global influences on equity
returns. Financial Analysts Journal, Vol. 52, No. 2, pp. 31-39.
Büyükşalvarci, A., and H. Abdioğlu (2010). The causal relationship between stock prices and
macroeconomic variables: A case study for Turkey. International Journal of Economic
Perspectives, Vol. 4, No. 4, pp. 601-610.
Castanias, R. P. (1979). Macro Information and the variability of stock market prices. Journal
of Finance, Vol. 34, No. 2, pp. 439–450.
Chiarella, C., and S. Gao (2004). The value of the S&P 500 — a macro view of the of the
stock market adjustment process. Global Finance Journal, Vol. 15, No. 2, pp. 171–196.
Choi, J. J., and M. Rajan (1997). A joint test of market segmentation and exchange risk factor
in international capital market. Journal of International Business Studies, Vol. 28, No. 1,
pp. 29-49.
DeFina, R. H. (1991). Does inflation depress the stock market? Business Review, Issue
November 1991, pp. 3-12.
Degiannakis, S., G. Filis and C. Floros (2013). Oil and stock returns: Evidence from
European industrial sector indices in a time-varying environment. Journal of
International Financial Markets, Institutions and Money (in press, accepted manuscript).
Dorodnykh, E. (2012). What is the degree of convergence among developed equity markets?
International Journal of Financial Research, Vol. 3, No. 2, pp. 1-16.
Engle, R. F., and C. W. J. Granger (1987). Co-integration and error correction:
Representation, estimation, and testing. Econometrica, Vol. 55, No. 2, pp. 251-276.
Filis, G. (2010). Macro economy, stock market and oil prices: Do meaningful relationships
exist among their cyclical fluctuations? Energy Economics, Vol. 32, No. 4, pp. 877-886.
Friedman, M. (1988). Money and the stock market. Journal of Political Economy, Vol. 96,
No. 2, pp. 221-245.
Hardouvelis, A.G. (1988). Stock prices: nominal vs. real shocks. Financial Markets and
Portfolio Management, Vol. 2, No. 3, pp. 10–18.
24
Hidayat, T. (2007). The effect of inflation on the performance of Islamic banking’s financing,
the transaction volume of interbank Islamic money market (PUAS) and the position of
outstanding Bank Indonesia wadiah sertificate (SWBI). University of Indonesia:
undergraduate thesis.
Hooker, M. A. (2004). Macroeconomic factors and emerging market equity returns: a
Bayesian model selection approach. Emerging Markets Review, Vol. 5, No. 4, pp. 379–
387.
Heaney, R., V. Hooper, and M. Jaugietis (2002). Regional integration of stock markets in
Latin America. Journal of Economic Integration, Vol. 17, No. 4, pp. 745 – 760.
Hurvich, C. M., and C. L. Tsai (1989). Regression and time series model selection in small
samples. Biometrika, Vol. 76, No. 2, pp. 297-307.
Husnan, S., and E. Pudjiastuti (1994). International diversification: Observation on AsiaPacific markets. Kelola, Vol. 3, No. 5, pp. 107-118.
Hussin, M. Y. M., F. Muhammad, M. F. Abu Hussin, and A. Abdul Razak (2012). The
Relationship between oil price, exchange rate and Islamic stock market in Malaysia.
Research Journal of Finance and Accounting, Vol. 3. No. 5, pp 2222-2847.
Iqbal, Z., and H. Tsubota (2006). Emerging Islamic capital markets: a quickening pace and
new potential. In The Euromoney Yearbooks (2006): International debt capital markets
handbook. Washington, DC: The World Bank.
Johansen, S. (1991). Estimation and hypothesis testing of cointegration vectors in Gaussian
vector. Econometrica, Vol. 59, No. 6, pp. 1551-1580.
Johansen, A., and K. Juselius (1990). Maximum likelihood estimation and inference on
cointegration — with applications to the demand for money. Oxford Bulletin of
Economics and Statistics, Vol. 52, No. 2, pp. 169–210.
Kassim, S. H., and T. A. Abdul Manap (2008). The information content of the Islamic
interbank money market rate in Malaysia. International Journal of Islamic and Middle
Eastern Finance and Management, Vol. 1, No. 4, pp. 304-312.
Levine, R., and Zervos, S. (1996). Stock Market Development and Long-Run Growth. The
World Bank Economic Review, Vol. 10, No. 2, pp. 323–339.
Maysami, R. C., and T. S. Koh (2000). A vector error correction model of the Singpore stock
market. International Review of Economics and Finance, Vol. 9 No. 1, pp. 79-96.
Mishkin, F. S. (2004). The economics of money, banking, and financial markets, the 9th
edition. USA: Prentice Hall.
25
Moschitz, J. (2004). The determinant of the overnight interest rate in the Euro area. European
Central Bank Working Paper Series No. 393/September 2004.
Mundell, R. A. (1962). The Appropriate use of monetary and fiscal policy for internal and
external stability. Staff Papers - International Monetary Fund, Vol. 9, No. 1, pp. 70-79.
Ningsih, S. (2011). Information contents regarding macroeconomic variables of conventional
money market rate and Islamic money market rate with Toda-Yamamoto causality test
method for the periode January 2003- December 2010. University of Indonesia:
undergraduate thesis.
Park, J. W., and R. A. Ratti (2008). Oil price shocks and stock market in the U.S. and 13
European countries. Energy Economics, Vol. 30, No. 5, pp. 2587-2608.
Rambaldi, A., and H. Doran (1996). Testing for Granger non-causality in cointegrated
systems made easy. University of New England: Working Papers in Econometrics and
Applied Statistics.
Sadorsky, P. (1999). Oil price shocks and stock market activity. Energy Economics, Vol. 21,
No. 5, pp. 449–469.
Shirazi, N. S., and T. A. A. Manap (2005). Export‐led growth hypothesis: Further
econometric evidence from South Asia. The Developing Economies, Vol. 43, No. 472488.
Shittu, O. I., and M. J. Asemota. (2009). Comparison of criteria for estimating the order of
autoregressive
process: A monte carlo approach. European Journal of Scientific
Research, Vol. 30, No. 3, pp. 409-416.
Toda, H. Y., and T. Yamamoto (1995). Statistical inference in vector autoregressions with
possibly integrated processes. Journal of Econometrics, Vol. 66, No. 1–2, pp. 225–250.
Wahyudi, I., and F. Rosmanita. Islamic money market: formal instrument that not optimal. In
Yusuf, W. (2011), Indonesian Islamic Economic Outlook (ISEO) 2011. Jakarta: Lembaga
Penerbit FEUI.
Yusof, M. R., and M. S. Abdul Majid (2007). Stock market volatility in Malaysia: Islamic
versus conventional stock market. Journal of KAU: Islamic Economics, Vol. 20, No. 2,
pp. 17-35.
Zapata, H. O., and A. N. Rambaldi (1997). Monte carlo evidence on cointegration and
causation. Oxford Bulletin of Economics and Statistics, Vol. 59, No. 2, pp. 285–298.
26