The Determinants of Indonesia’s Business Cycle 215

UDK: 336.7:005.44(594)
DOI: 10.2478/jcbtp-2020-0029

Journal of Central Banking Theory and Practice, 2020, special issue, pp. 215-235

Berry A. Harahap *, Pakasa Bary **, * Department of Economic

and Monetary Policy at Bank
Anggita Cinditya M. Kusuma *** Indonesia, Jakarta,
Indonesia

The Determinants of Indonesia’s E-mail:

berry@bi.go.id
Business Cycle1
** Department of Economic
and Monetary Policy at Bank
Indonesia, Jakarta,
Abstract: This study investigates the determinants of Indonesian’s Indonesia
business cycle using the global vector autoregressive (GVAR) ap-
E-mail:
proach, by including spillover responses within 33 countries with
pakasa@bi.go.id
2000 bootstrap replications. The results show that Indonesia’s busi-
ness cycle is influenced by both domestic and external factors. In
*** Department of Economic
addition to exogenous shocks from output, the dominant domestic
and Monetary Policy at Bank
factors are monetary policy and price competitiveness. The domi-
Indonesia, Jakarta,
nant external factors are global economic activity and liquidity con-
Indonesia
ditions, particularly those originating from the Chinese economy.
Spillovers from a number of economies appear to shape Indonesia’s E-mail:
economic fluctuations. The paper discusses such relevant spillovers. anggita_cmk@bi.go.id

Keywords: Business cycle; Trade relations; Global vector autogres-

sive approach.

JEL Classification: E32; F44; E30; C22

The authors thank two anonymous referees in journal reviewing

1

process, Reza Anglingkusumo, Maryam Akbari Nasiri, Bernard

Njindan Iyke, Paresh K. Narayan, and participants at BMEB confer-
ence in Bali, Indonesia in August 2019. The views expressed in this
paper are those of the authors and do not necessarily represents the
views of Bank Indonesia.
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
216 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

1. Introduction
This paper assesses the determinants of Indonesia’s business cycle. Business cycle
can be very helpful for policy makers as well as the private sector, as it provides a
tool for estimating short-term economic behavior, for evaluating the outcomes of
certain policy decisions in different markets, or for assessing the implementation of
policies in accordance with the cycle phase (Jimenez, 2001). Moreover, the awareness
of current state of the economy is undeniably important, for instance Tarsidin, Id-
ham, and Rakhman (2018) suggested that by knowing the state of the economy and
projections of upcoming conditions, the policy maker can determine the appropriate
policy response to adopt. Hence, a comprehensive understanding of business cycle
determinants enables policy makers to design effective policy programs. Apart from
this, it enables private sector to develop effective strategies for their businesses.

Although the analysis of business cycle is commonly conducted in number of stud-

ies (see for example, Burn and Mitchell, 1946; Walsh, 1999; Comin and Gertler 2006;
Male, 2011; Drehmann et al, 2012; Altuğ and Bildirici, 2012 and Han 2019), study
about the determinants of business cycle remain limited and elusive. Jimenez (2001)
argues that the external sector plays a key role in shaping small and open business
cycle dynamics, where the origin of business cycles in developing countries are un-
derstood as impulses originating from developed countries that condition the evolu-
tion of business cycles, especially the turning point. In other words, business cycles
in developing countries are arguably not purely endogenous processes. The interna-
tional origin of economic fluctuations in small open economies operate along two
channels: (i) through the variation in relationships with major trading and financial
partners (Schmitt-Grohe, 1998); and (ii) through random events, such as the oil crisis
(Jimenez, 2001). For business cycle analysis, the first variation is important because
persistence occurs from time to time2.

Shocks originating from major trading and financial partners are usually distrib-
uted to small open economies through two transmission mechanisms, trade and fi-
nancial channels. The trade channel is mainly related to changes in the small open
country’s exports (demand factors), while the financial channel relates to variations
in domestic interest rates (cost factors) because of changes in world interest rates
(Schmitt-Grohe, 1998). These transmission mechanisms are consistent with the in-
tuition that a small open country’s business and financial cycle determinants are
closely related to its economic conditions, as well as to global economic conditions
(Jimenez, 2001). Garratt, Lee, and Shields (2013) find support for this intuition by
showing that economic performances across countries are linked through inter-

Impulses coming from trading/financial partners are typically propagated in the small country
2

through commercial and financial channel and it is more regularly occur rather than random
events such as oil crisis (Jimenez, 2001).
 The Determinants of Indonesia’s Business Cycle 217

national trade and capital markets such that business cycle fluctuations in a given
country is usually transmitted to others. Dufrénot and Keddad (2014) in their study
with a Markov-switching approach also provide evidence that the signals contained
in some regional and global leading business cycles can impact business cycle in
some ASEAN countries, furthermore their finding suggests that ASEAN economies
are characterized by a strong dependence on external demand. In the Indonesian
context, Silalahi, Wibowo, and Nurliana (2012) show that international shocks affect
the bank lending.

In view of the preceding, our aim is to identify determinants of the business cycle
in Indonesia, a small open economy. Prior studies (Alamsyah, Adamanti, Yumanita,
and Astuti, 2004; Wimanda and Djuranovik, 2014a; 2014b; Andaiyani and Falianty,
2017) consider aspects on the country’s business or financial cycles. Some other stud-
ies, for instance Kim, Kose, and Plummer (2003), examined the similarities and dif-
ferences of business cycle characteristics of the Asian countries and compares the
cyclical regularities in this region with those of G7 countries. Narayan (2011) study
the role of permanent and transitory shocks in determining Indonesia’s business cy-
cle through using a simple real business cycle model and found that at business cycle
horizons permanent shocks explain the bulk of variations in income, consumption
and investment for Indonesia. Moreover, Moneta and Rüffer (2006) examined the
extent and nature of synchronisation of business cycle in ten East Asian countries.
They found that cross-country spill-over effects explain only a small part of the co-
movement in the region and economic linkages with Europe and North America
may have contributed to the observed synchronisation. However, none of those stud-
ies closely examines the determinants of the business cycle in certain economy. Dutu
(2016) comes close, by showing that a decelerating trend of Indonesia’s economic
growth is reinforced by slower world growth.3 In spite of this, a number of questions
including which countries play the key role in Indonesia’s slow economic growth,
what is the contribution of domestic factors relative to external factors in the slow
growth, among others, remain unanswered. Our study attempts to provide answers
to these questions.

Indonesia is a small open economy. Hence, its business cycle is arguably influenced
by both domestic and global factors. We take this into account by employing the
GVAR approach and sample covering 33 countries that account for about 90% world

Studies, such as Simorangkir (2012), and Anwar and Ali (2018), developed early warning sys-
3

tems to detect downturns. For a survey of these studies, see Padhan and Prabheesh (2019).
Others such as Juhro and Iyke (2019a, b) developed financial condition indexes to track the
direction of the financial system and the economy as a whole or demonstrated that consumer
confidence is an important driver of consumption expenditure, which determines business cy-
cle fluctuations in Indonesia.
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
218 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

GDP and over a period of 1979Q2 to 2016Q44. We show the contribution of these
determinants by generating forecast error variance decompositions (FEVDs) via
2000 bootstrap replications. The estimates suggest that the Indonesian business cy-
cle is indeed influenced by domestic and global factors. Apart from the exogenous
shock from output (productivity shock), the dominant domestic factors are mon-
etary policy and price competitiveness. With regards to the global factors, we find
global economic activity and liquidity conditions, particularly those origining from
China, to be influential. This is consistent with Sznajderska (2019) who states that
since China plays an important role in traditional global trade and the global supply
chain, a slow down in its economy may have indirect effects on the rest of the world,
especially through neighboring countries. We also find shocks originating from a
number of relatively remote economies play a sizeable role in Indonesia’s business
cycle fluctuations. From a more general perspective, this suggest that spillover effects
and indirect relations via a third country are important to measure, as they provide
a clearer picture of the fluctuations in an economy. We provide some counterfactual
analysis to ‘normalize’ the effect of China’s rebalancing.

Our paper relates to the broader literature discussing transmission of external

shocks such as Stock and Watson (2002); Kose, Otrok, and Whiteman (2003, 2008);
Crucini, Kose, and Otrok (2011); Mumtaz, Simonelli, and Surico (2011); Vasishtha
and Maier (2013); Jo (2014); Solomos, Papageorgiou, and Koumparoulis (2016); Kose,
Otrok, and Prasad (2012); Inoue, Kaya, and Ohshige (2015); and Cashin, Mohaddes,
Raissi (2017a). Crucini, Kose, and Otrok (2011), using the dynamic factor model, find
that oil prices, productivity, and terms of trade are the drivers of business cycles.
Vasishtha and Maier (2013) find, using the factor-augmented vector autoregressive
(FAVAR) model with more than 260 series for 20 OECD countries, that Canada is
mainly exposed to shocks against foreign activity and commodity prices. Solomos,
Papageorgiou and Koumparoulis (2016) find, using the generalized linear model
(GLM), that the total value of traded shares, private sector debt, and net inflows
of FDI are significant determinants in the fluctuations in business cycles in 12 Eu-
ropean countries. Inoue, Kaya, and Ohshige (2015) find, using GVAR model with
33 countries, that China’s slowdown mostly affects commodity exporters such as
Indonesia, and countries that are heavily dependent on exports, that include Japan,
Malaysia, Singapore and Thailand. Consistent with Inoue, Kaya, and Ohshige (2015),
Cashin, Mohaddes, and Raissi (2017a) find, using the GVAR model, that negative
output shocks from China have the greatest influence on diversified commodity ex-
porters and ASEAN-5 countries (Indonesia, Malaysia, Singapore, Thailand, without
the Philippines).

GVAR approach contains 33 countries, covering about 90% of world output as stated in the
4

GVAR Handbook (Mauro & Pesaran, 2013). List of the countries is displayed in Table 1.
 The Determinants of Indonesia’s Business Cycle 219

Our paper is most comparable to Boschi and Girardi (2011) and Boschi, Marzo, and
Salotti (2015), since they explore the determinants of business cycles using GVAR as
well. Our paper differs from theirs in at least three ways. First, it focuses on Indone-
sia, whereas they focus on Latin America and the Euro Area. By focusing on Indone-
sia, we add an important developing country to the existing sample in the literature.
Second, this paper is not classifying countries from a certain region into one group,
particularly to analyze the origin of the shocks. This is due to one objective of this
paper is to determine countries which have a major role to economic fluctuation in
Indonesia. Third, our paper extends the sample period in Boschi, Marzo, and Salotti
(2015) by 10 years. Hence, our estimates, are perhaps more precised.

The paper proceeds as follows. Section 2 discusses our data and methodology. We
discuss our main findings in Section 3. Section 4 sets forth our conclusions and pol-
icy recommendations.

2. Methodology and Data

We analyze the determinants of Indonesia’s business cycle using GVAR model de-
veloped by Pesaran, Schuermann, and Weiner (2004) and later by Dees, Di Mauro
Pesaran, and Smith (2007). GVAR model combines time series, cross-sectional, and
factor analysis techniques to explain macroeconomic and financial phenomena.
Technically, the model is an aggregation of country-specific vector autoregressions
(VARs), whereby domestic variables are related to foreign variables that are specific
to each country (Pesaran, Schuermann, and Weiner, 2004; Dees, Di Mauro Pesaran,
and Smith, 2007). The foreign variables are connected to the domestic variables of a
given country through trade, financial, or other patterns that are considered appro-
priate for that country (Dees, Di Mauro Pesaran, and Smith, 2007).

The seemless interaction of foreign with domestic variables within the GVAR model
makes it popular in many applications including macro stress testing (Al-Haschimi,
Dées, Mauro, and Jancokova, 2014), analysing the growing importance of China in
the world economy (Cesa-Bianchi, Pesaran, Rebucci, and Xu, 2012; Cashin, Mohad-
des, and Raissi, 2017a), analysing the global macroeconomic transmission of weather
shocks (Cashin, Mohaddes, and Raissi 2017b), assessing the impact of commodity
price shocks (Mohaddes and Pesaran, 2016, 2017; Cashin, Mohaddes, Raissi, and
Raissi, 2014; Mohaddes and Raissi, 2019), assessing the impact of US monetary poli-
cy shocks (Harahap, Bary, Panjaitan, and Satyanugroho, 2019), analyzing other real
and financial sector shocks (Chudik and Fratzscher, 2011; Eickmeier and Ng, 2015),
and forecasting macroeconomic indicators (Favero, 2013).

The GVAR model is developed as follows. For each country, the conventional VAR
model is expanded to capture foreign variables. These foreign variables are con-
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
220 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

structed as weighted averages of foreign variables that correspond to the domestic

variable, typically using trade weights5. Suppose there are N + 1 countries in the
global economy with index i = 0,1,2, ... N, where country 0 is used as numeraire (or
reference) country. Then, individual VARX* (pi , qi) for each country (where VARX*
denotes a VAR augmented with foreign variables) is as following:

1)

where xit is a vector of domestic variables (i.e. GDP, CPI, interest rates and real ex-
change rate) of size ki × 1, is the vector of foreign variables with size li × 1, and εit is
a serially uncorrelated and cross-sectionally weakly dependent process. The foreign
variables are computed as weighted averages of the corresponding domestic vari-
ables of all countries, with the weights also being country-specific; that is

2)

where ωij are the weights such that ωii = 0 and .

The value of ωij for country i is built based on the flow portion of country j to the total
flow received by country i. This represents the trade relationship between country i
and j. Country-specific foreign variables are considered as weakly exogenous, and
“long-run forcing” in the model (Dees, di Mauro, Pesaran, and Smith, 2007). This
means the coefficient of the error correction term is set to zero in the foreign variable
equation. Thus, the dynamics of the foreign variables are not affected by long-equi-
librium paths, unlike those of domestic variables. For each country, the parameters
are estimated using reduced-rank regressions and ordinary least squares (OLS).

The estimation of the GVAR model is completed using the country-level parameter
estimates to indicate the global parameters. Although the estimation is carried out
separately for each country, the GVAR model is solved simultaneously, because of
sampe-period dependence between domestic variables xit and foreign variables .
The estimates of the GVAR model can be used to obtain impulse responses.

If , then equation (1) can be expressed as:

3)

where Ai = (Iki – ˄i0 ), Bis = (Φis ˄is). ai0 and ai1 are parameters.

From equation (2), zit = Wi xt where Wi is quality matrix with size (ki + li) × k defined
from specific qualities of countries, ωij . Therefore, equation (3) can be expressed as:

For more comprehensive explanation about GVAR model, please refer to Mauro and Pesaran
5

(2013).
 The Determinants of Indonesia’s Business Cycle 221

4)

and individual state models are grouped together to become a global model xt, which
is:

5)

Multiplying (5) by , we obtain the following equation:

6)

Specifications (5) can be solved recursively to obtain future values ​​and to obtain im-
pulse responses. Equation (6) can also be solved recursively and the variance gener-
ated and decomposed. This is the so-called FEVD, which we use to assess the contri-
bution of each business cycle determinant to overall fluctuations. Boschi and Girardi
(2011) use FEVD to analyze the determinants of business cycles.

Data used in the GVAR model are de-trended, while the noise components are nor-
mally captured as residuals. Hence, technically, the estimated relationships among
the variables are actually the relationships between their cyclical components. There-
fore, determinants of business cycle are always examined as determinants of aggre-
gate macroeconomic variables. For example, Kose, Otrok, and Whiteman (2003) use
three alternative variables, which are gross domestic product (GDP), investment and
consumption. Holland and Scott (1998) use alternative variables, which are GDP,
investment, consumption, real wages, and total working hours. It is known that one
indicator, GDP, includes both investment and consumption. Therefore, we follow
Claessens, Kose and Terrones (2012) and Boschi and Girardi (2011) to use GDP as
our measure of economic activity, and hence discussing determinants of GDP as the
determinants of business cycle.6

The GVAR data set covers economic and quarterly financial variables for 33 econo-
mies during the period 1979Q2-2016Q4 (Table 1). These 33 countries cover more
than 90% of world GDP (Mauro and Pesaran, 2013). The dataset is obtained from

6
As it includes consumption and investment, GDP represents other alternative variables as well.
In addition, Claessens, Kose and Terrones (2012) argues that output or GDP is the best indicator
available to measure economic activity. Moreover, Luthfiana and Nasrudin (2018) also stated
that GDP is considered as the best measure of economic performance.
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
222 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

Mohaddes and Raissi (2018). Structural break unit root test was done using the
Narayan and Popp (2010, 2013) test and results are available upon request.
Table 1. Variables in GVAR
Country-specific variables
Variable name Period
Real GDP 1979Q2 – 2016Q4
Inflation 1979Q2 – 2016Q4
Real equity prices 1979Q2 – 2016Q4
Real exchange rate 1979Q2 – 2016Q4
Nominal short-term interest rate 1979Q2 – 2016Q4
Nominal long-term interest rate 1979Q2 – 2016Q4
Global variables
Variable name Period
Oil price 1979Q2 – 2016Q4
Raw material price 1979Q2 – 2016Q4
Metal price 1979Q2 – 2016Q4

3. Results
The preliminary analysis entails certain features on the GVAR model. Table A2 and
A3 reports tests of the model’s soundness. Overall, the statistical tests show that the
estimated GVAR model is stable, and hence can for analyzing the business cycle de-
terminants. For instance, the weak exogeneity test shows that the external variables
are weakly exogenous in most cases. Contemporaneous effects on domestic variables
are generally in line with the presence of shocks on the same external variables. In
addition, the persistence profile of system-wide shocks in all Indonesian cointegra-
tion equations falls to zero exponentially.

Another indication that the GVAR model is sound can be verified through the mag-
nitude of residual correlation (Dees, di Mauro, Pesaran, and Smith, 2007). The resid-
ual correlation in each vector error correction model augmented with foreign vari-
ables (VECMX), reported in Table A3, looks quite small (i.e. they are in the range of 0
to 0.3). This shows that the estimated GVAR model is quite sound in accommodating
factors that influence the endogenous variables. In addition, the low residual cor-
relation indicates that the model is quite effective in explaining reciprocal relations
between countries (Sun, Heinz, and Ho, 2013).

The estimation results are then used primarily for variance decomposition analysis.
Following Boschi and Girardi (2011), we interprete the variance decomposition anal-
 The Determinants of Indonesia’s Business Cycle 223

ysis on GDP as the business cycle determinants. Table 2 reports the variance decom-
position of Indonesia’s output. The results indicate that the Indonesian business cycle
is influenced by domestic and global factors. In addition to output, the dominant
domestic factors are short-term interest rates and Real Exchange Rate (RER)—con-
tributes around 15% to the business cycle fluctuations. An alternative interpretation
is that the monetary policy stance and international competitiveness are important
determinants of Indonesia’s business cycle.

Given that the variables used in GVAR are demand side variables, then when refer-
ring to business cycle theory, the result implies the business cycle is explained more
by preference shocks (Holland and Scott, 1998). Productivity shocks are represented
by the output variable, GDP, itself. Hence, in the short term, the business cycle is
more influenced by productivity shocks, while, in the medium term, it is more influ-
enced by shift in preferences or preference shocks. Our findings are consistent with
Holland and Scott (1998), who find that correlations of productivity with macroeco-
nomic aggregate variables tend to decrease over time, while correlations of prefer-
ence shift with macroeconomic aggregate variables tend to increase over time.

Table 2. Variance Decomposition of Indonesia’s GDP

Foreign
Domestic
(rest of the world)
GDP Inflation RER Interest Rate GDP Interest Rate
1 0.6119 0.0223 0.0958 0.0456 0.1330 0.1437
4 0.2676 0.0153 0.1441 0.1573 0.1294 0.1514
8 0.1485 0.0094 0.1432 0.1680 0.1282 0.1622
12 0.1055 0.0075 0.1328 0.1608 0.1258 0.1739

The table reports the decomposed variances of the GDP equation. It shows the contribution
of each variable to GDP. The domestic variables and the foreign variables are GDP, inflation,
RER, and interest rate. The sample is made up of 33 countries over the period of 1979Q1 to
2016Q4.

The dominant global factors affecting the Indonesian business cycle are the global
short-term interest rates and the output of rest of the world, each contributing around
16% and 13% to the business cycle fluctuations. The implication is that Indonesia’s
business cycle is influenced by global liquidity, monetary policies, and economic ac-
tivity. The estimates further suggest that macroeconomic spillovers, especially pro-
ductivity shocks, from China are the most important for Indonesia’s business cycle.
This is consistent with data. The slowing down phase of Indonesia’s economic growth
since the global financial crisis is strongly related to China’s economic growth that
experienced rebalancing. Figure 1 shows that, in general, China’s economic growth
precedes Indonesia’s.
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
224 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

Figure 1. GDP Growth in Indonesia, China, United States (% yoy)

The figure shows GDP growth in Indonesia vis-à-vis China and the United States’ growth. GDP
growth is defined as the annual percentage change of real GDP in each country for the period
2004 - 2018.

This study focuses on discussing FEVDs as it attempts to explain determinants of

Indonesia’s business cycle. However, figure 2 shows several impulse responses for re-
porting and illustrative purposes, as it is usually reported on literatures using GVAR.
The figure shows that GDP shock from China and Japan have positive impact on
Indonesia’s GDP. However, they have different magnitude and persistency.

Figure 2. Impulse Responses – China and Japan’s GDP to Indonesia’s GDP

The figures show impulse response of China’s GDP (left) and Japan’s GDP (right) on Indonesia’s
GDP. The range is derived from bootstrap replications. The solid line represents median, while
dotted lines represent lower and upper bounds.
 The Determinants of Indonesia’s Business Cycle 225

Figure 3 shows spillovers from other countries to Indonesia. The spillovers from the
US continue to gain importance in the Indonesian business cycle, especially in the
medium term. In addition, spillovers from ASEAN countries, especially those origi-
nating from Malaysia, have a large influence on the Indonesian business cycle fluc-
tuations. The important role played by spillovers from other developing countries to
the business cycle fluctuations of a developing country is consistent with Boschi and
Girardi (2011), who find this to hold true for Latin America. We find that spillovers
from developing countries contribute greatly to Indonesia’s business cycle fluctua-
tions, when compared with spillovers from developed countries (Table 3).

Figure 3. Variance Decomposition – Other Countries’ Contribution (All Variables) to

Indonesia’s GDP

The figure shows variance decomposition of Indonesia’s GDP overtime (quarters) after the
shock. The factors shown are the ones which come from several countries’ macroeconomic
variables that include GDP, inflation, interest rate, and real exchange rate.

Table 3. Variance Decomposition – Contribution of Other Countries’ GDP to Indonesia’s GDP

Developed Developing
Quarter China US Japan ASEAN
Countries Countries
1 0.023 0.002 0.007 0.035 0.051 0.088
4 0.043 0.002 0.005 0.025 0.036 0.098
8 0.046 0.003 0.004 0.022 0.033 0.099
12 0.040 0.003 0.003 0.021 0.034 0.095

The table shows comparison of variance decomposition of Indonesia’s GDP which originates
from other countries.

Table 4 ranks countries by their contribution to Indonesia’s business cycle fluctua-

tions. The China is the most important contributor to Indonesia’s output fluctua-
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
226 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

tions. A number of Asian and non-Asian countries are also significant contributors
to Indonesia’s business cycle. Contributions from countries such as Malaysia, Brazil
and Mexico are unexpected. Such a new finding can be explained by the superiority
of the analytical methods that pay attention to the impact of indirect spillovers. The
GVAR considers all the spillover effects—directly and indirectly—from global trade
relations, which, so far, is difficult to interpret using descriptive data, which only
shows direct relations. A closer look at the trade matrix suggests that Mexico–China
and Brazil–China trades are very high, and for Brazil it is even higher than Brazil’s
trade with the US and Canada (Table A4). This fact, combined with Indonesia–China
trade relations, which is also dominant, can cause changes in Mexico and Brazil’s
output to have a significant impact on Indonesia’s output.

Table 4. Ranking of Influences on Indonesian Output

Rank Country Contribution on fourth quarter after shock
1 CHINA 0.0432
2 MALAYSIA 0.0138
3 BRAZIL 0.0066
4 MEXICO 0.0056
5 KOREA 0.0055
6 JAPAN 0.0049
7 THAILAND 0.0048

The table shows countries that contributes the most on variance decomposition of Indonesia’s
GDP.

Table 5 shows how other countries respond to spillovers from China. As can be seen,
two ASEAN countries, namely Singapore and Thailand, are even more sensitive to
spillovers from China, than Indonesia.

Table 5. Ranking of Sensitivity on China’s GDP

Ranking Country Response after four quarters
1 Singapore 0.0074
2 Thailand 0.0056
3 Indonesia 0.0056
4 Malaysia 0.0056
5 Peru 0.0039
6 Japan 0.0035
7 Saudi Arabia 0.0032
8 Turkey 0.0027
9 India 0.0021

The table shows comparison of impulse responses due to 1 sd shock to China’s GDP.
 The Determinants of Indonesia’s Business Cycle 227

Figure 4 provides further illustration of China’s significance in the Indonesian busi-

ness cycle. This figure is a counterfactual analysis of China’s output. The assumed
counterfactual scenario is that China’s output growth remains above 7% after 2015
(i.e. approximately 7.3%, to be more specific). Given this scenario, the path of Indo-
nesia’s business cycle becomes expansive, with an average growth of 5.6% through-
out 2016-2018, and reaching almost 6% in the second quarter of 2018. This shows that
the counterfactual scenario may cause different results on the path and acceleration
speeds in the Indonesian business cycle.

Figure 4. Indonesia GDP Growth - Counterfactual and Actual

The figure shows actual and counterfactual GDP growth of Indonesia. The counterfactual
GDP growth is computed using GVAR estimates by assuming that China’s GDP growth has not
decelerated.

4. Concluding Remarks
This study analyzed the determinants of the Indonesian business cycle within a
GVAR model. Using a sample of 33 countries over a period of 1979Q2 to 2016Q4 and
2000 bootstrap replications, it generated FEVDs for GDP. It finds that the Indone-
sian business cycle is influenced by domestic and global factors. With regards to the
domestic factors, monetary policy and price competitiveness are most influential in
Indonesia’s business cycle fluctuations. For the global factors, global liquidity, mon-
etary policy, and economic activity contributed largely to the business cycle.

Among the external sources of fluctuations, those from China are dominant. A
shocks originating from number of countries, Asian and non-Asian, are also im-
portant in Indonesia’s business cycle. The main countries are Malaysia, Brazil and
Mexico, whose impacts are rather surprising. This may be attributed to the fact that
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
228 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

the GVAR model considers direct and indirect global spillovers from trade relations,
which are not evident in descriptive data.

These findings are meaningful in the real business cycle theory. Since the model
captures demand-side variables, Indonesia’s main business cycle determinants in the
short term are productivity shocks, whereas, in the medium term (starting at 1 year),
the fluctuations are driven by demand-side or preference shocks. Holland and Scott
(1998) associate the demand side with a preference shift between work and leisure.
This indicates that policies that affect the demand side will have a lagging impact,
while policies that directly affect productivity, such as the application of new tech-
nologies, will have relatively instant effects.

The influence of China on Indonesia’s business cycle is buttressed using a counter-

factual analysis. The study finds that if China’s did not experienced rebalancing in its
economy after 2015, Indonesia may have a higher output growth followed by a phase
of expansion of the business cycle at a higher level.

Because China’s rebalancing is permanent with the changing phase of its economic
development, Indonesia needs to look for new sources of external growth.

Monitoring the conditions in other developing countries will be more important

than previously believed. For instance, Latin American countries, which are consid-
ered the farthest from Indonesia, should be monitored since they appear relevant to
Indonesia’s economic fluctuations. Furthermore, given that external factors contrib-
ute greatly to the Indonesian business cycle, if external conditions do not support the
business cycle expansion phase, the business cycle can be pushed up through accom-
modative monetary policy, or by price competitiveness. The demand-side policy will
take time to be felt within the economy. However, the impact can be immediate via
the application of new technologies.
 The Determinants of Indonesia’s Business Cycle 229

References
1. Alamsyah, H., Adamanti, J., Yumanita, D., & Astuti, R.I. (2004). Siklus
Keuangan Indonesia. Working Paper Bank Indonesia. WP/8/2014.
2. Al-Haschimi, A., Dées, S., Mauro, F., & Jancokova, M. (2014). Linking Distress
of Financial Institutions to Macrofinancial Shocks. ECB Working Paper Series,
1749 (December 2014).
3. Altuğ., S. and Bildirici., M. (2012) Business Cycles in Developed and Emerging
Economies: Evidence from a Univariate Markov Switching Approach. Emerging
Markets Finance and Trade, 48, 4-38.
4. Anwar, S., & Ali, A., (2018) ANNs-based early warning system for Indonesian
Islamic banks, Bulletin of Monetary Economics and Banking, 20, 325-342.
5. Andaiyani, S., & Falianty, T. A. (2017). Asean Credit Growth And Asset Price
Response To Global Financial Cycle, Bulletin of Monetary Economics and
Banking, 20(2), 203-228.
6. Boschi, M., & Girardi, A. (2011). The contribution of domestic, regional and
international factors to Latin America’s business cycle. Economic Modelling,
28(3), 1235-1246.
7. Boschi, M., Marzo, M., & Salotti, S. (2015). Domestic versus international
determinants of European business cycles: a GVAR approach. Empirical
Economics, 49(2), 403-421.
8. Burns, A.F., & Mitchell, W.C. (1946). Measuring Business Cycles, National
Bureau of Economic Research, New York.
9. Cashin, P., Mohaddes, K., & Raissi, M. (2014), The Differential Effects of Oil
Demand and Supply Shocks on the Global Economy. Energy Economics, 44,
113–134.
10. Cashin, P., Mohaddes, K. and Raissi, M. (2017a), China’s slowdown and global
financial market volatility: is world growth losing out? Emerging Markets
Review, 31, 164-175.
11. Cashin, P., Mohaddes, K., Raissi, M. (2017b), Fair Weather or Foul? The
Macroeconomic Effects of El Nino. Journal of International Economics, 106, 37-
54.
12. Cesa-Bianchi, A., Pesaran, M.H., Rebucci, A., & Xu, T. (2012), China’s
emergence in the world economy and business cycles in Latin America.
Economía, 12(2), 1-75.
13. Chudik, A. & Fratzscher, M. (2011). Identifying the global transmission of the
2007-09 Financial crisis in a GVAR model. European Economic Review, 55,
325-339.Claessens, S., Kose, M.A., & Terrones, M. (2012). How Do Business and
Financial Cycles Interact? Journal of International Economics, 87(1), 178-190.
14. Comin, D. & Gertler, M. (2006) Medium-term business cycles. American
Economic Review, 96(3), 523-51.
15. Crucini, M.J., Kose, M.A., & Otrok, C. (2011). What are the Driving Forces of
International Business Cycles? Review of Economic Dynamics, 14, 156-175.
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
230 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

16. Dees S., di Mauro, F., Pesaran, M.H., & Smith L.V. (2007), Exploring the
international linkages of the euro area: a global VAR analysis, Journal of
Applied Econometrics, 22(1), 1–38.
17. Drehmann, M., Borio, C. & Tsatsaronis, K. (2012), Characterising The Financial
Cycle: Don’t Lose Sight of The Medium Term!, BIS Working Paper, 380.
18. Dufrénot, G., & Keddad, B. (2014). Business cycles synchronization in East
Asia: A Markov-switching approach. Economic Modelling, 42, 186-197.
19. Dutu, R. (2016). Why has economic growth slowed down in Indonesia? An
investigation into the Indonesian business cycle using an estimated DSGE
model. Journal of Asian Economics, 45: 46–55.
20. Eickmeier, S. & Ng, T. (2015) How Do Credit Supply Shocks Propagate
Internationally? A GVAR Approach. European Economic Review, 74, 128-145.
21. Favero, C. (2013) Modelling and forecasting government bond spreads in the
Euro area: a GVAR model, Journal of Econometrics, 177, 343–356.
22. Garratt, A., Lee, K., & Shields, K. (2013). Global recessions and output
interdependencies in a GVAR model of actual and expected output in the G7.
Oxford University Press.
23. Han., Q. (2019) International Real Business Cycles of the Chinese Economy:
Asymmetric Preference, Incomplete Financial Markets, and Terms of Trade
Shocks, Emerging Markets Finance and Trade, 55, 1926-1953.
24. Harahap, B., Bary, P., Panjaitan, L. & Satyanugroho, R. (2019) Spillovers
of United States and People’s Republic of China Shocks on Small Open
Economies: The Case of Indonesia. In: Yoshino, N., Chantapacdepong, P., &
Helble, M. (Eds) Macroeconomic Shocks and Unconventional Monetary Policy:
Impact on Emerging Markets. Oxford University Press.
25. Holland, A. & Scott, A. (1998). The Determinants of UK Business Cycles. The
Economic Journal, 108(449), 1067-1092.
26. Inoue, T., Kaya, D., & Ohshige, H. (2015) The impact of China’s slowdown on
the Asia Pacific region: an application of the GVAR model, Policy Research
Working Paper Series 7442, The World Bank.
27. Jo., G. (2014) Transmission of U.S. Financial Shocks to Emerging Market
Economies: Evidence from Claims by U.S. Banks. Emerging Markets Finance
and Trade, 50, 237-253.
28. Jimenez, J.F.P. (2001). Business Cycles in Small Open Economies: The Case of
Costa Rica. Institute of Social Studies, Working Paper 330, February 2001.
29. Juhro, S.M., and Iyke, B.N., (2019a) Monetary policy and financial conditions
in Indonesia, Bulletin of Monetary Economics and Banking, 21, 283-302.
30. Juhro, S. M., & Iyke, B. N. (2019b). Consumer confidence and consumption
expenditure in Indonesia. Economic Modelling. https://doi.org/10.1016/j.
econmod.2019.11.001.
31. Kim, S. H., Kose, M. A., & Plummer, M. G. (2003). Dynamics of business cycles
in Asia: differences and similarities. Review of Development Economics, 7(3),
462-477.
 The Determinants of Indonesia’s Business Cycle 231

32. Kose, M.A., Otrok, C., & Prasad, E.S. (2012). Global Business Cycles:
Convergence or Decoupling? International Economic Review 53(2), 511-538.
33. Kose, A., Otrok, C. & Whiteman, C.H. (2003). International Business Cycles:
World, Region, and Country-Specific Factors. American Economic Review
93(4), 1216 – 1239.
34. Kose, M.A., Otrok, C., & Whiteman, C.H. (2008). Understanding the evolution
of world business cycles. Journal of International Economics 75(1), 110-130.
35. Luthfiana, P.S., and Nasrudin. N. (2018). Disaggregation and Forecasting of
the Monthly Indonesian Gross Domestic Product (GDP). Bulletin of Monetary
Economics and Banking, Volume 20, Number 4, April 2018.
36. Male., R. (2011) Developing Country Business Cycles: Characterizing the
Cycle. Emerging Markets Finance and Trade, 47: 20-39.
37. Mauro, F., & Pesaran, M.H. (2013). The GVAR Handbook: Structure and
Applications of a Macro Model of the Global Economy for Policy Analysis.
Oxford University Press.
38. Mohaddes, K. & Raissi, M. (2019). The U.S. Oil Supply Revolution and the
Global Economy. Empirical Economics, 57(5), 1515-1546.
39. Mohaddes, K. & Pesaran, M. H. (2016). Country-Specific Oil Supply Shocks
and the Global Economy: A Counterfactual Analysis. Energy Economics, 59,
382–399.
40. Mohaddes, K. & Pesaran, M.H. (2017). Oil Prices and the Global Economy? Is it
Different This Time Around?. Energy Economics, 65, 315–325
41. Mohaddes, K. & Raissi, M. (2018). Compilation, Revision and Updating of the
Global VAR (GVAR) Database, 1979Q2-2016Q4. University of Cambridge:
Faculty of Economics (mimeo).
42. Moneta, F., & Rüffer, R. (2009). Business cycle synchronisation in East Asia.
Journal of Asian Economics, 20(1), 1-12.
43. Mumtaz, H., Simonelli, S., & Surico, P. (2011). International comovements,
business cycle and inflation: A historical perspective. Review of Economic
Dynamics, 14, 176-198.
44. Narayan, P. K. (2011). An Analysis of Indonesia’s Business Cycle. The Indian
Economic Journal, 59(1), 162-173.
45. Narayan, P. K., & Popp, S. (2010). A new unit root test with two structural
breaks in level and slope at unknown time. Journal of Applied Statistics, 37(9),
1425-1438.
46. Narayan, P. K., & Popp, S. (2013). Size and power properties of structural break
unit root tests. Applied Economics, 45(6), 721-728.
47. Padhan, R., and Prabheesh, K.P., (2019) Effectiveness of early warning models:
A critical review and new agenda for future direction, Bulletin of Monetary
Economics and Banking, 22, 457-484.
48. Pesaran, M.H., Schuermann, T., Weiner, S. (2004) Modeling Regional
Interdependencies Using a Global Error-Correcting Macroeconometric Model,
Journal of Business & Economic Statistics, 22, 129–162.
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
232 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

49. Schmitt-Grohe, Stephanie (1998) The international transmission of economic

fluctuations: effects of U.S business cycles on the Canadian economy. Journal of
International Economics, 44: 257-287.
50. Silalahi, T., Wibowo, W. A., & Nurliana, L. (2012). Impact of global financial
shock to international bank lending in Indonesia. Bulletin of Monetary
Economics and Banking, 15(2), 77-114.
51. Simorangkir, I. (2012). Early warning indicators study of bank runs in
Indonesia: Markov-switching approach. Bulletin of Monetary Economics and
Banking, 15(1), 3-40.
52. Solomos, D., Papageorgiou, T., & Koumparoulis, D. (2016). Financial Sector and
Business Cycles Determinants in the EMU Context: An Empirical Approach
(1996-2011). European Research Studies Journal, 16(2), 34-58.
53. Stock, J.H., & Watson, M.W. (2002). Has the Business Cycle Changed and Why?
NBER Macroeconomics Annual, 2002, 159-230.
54. Sun, Y., Heinz, F. F., & Ho, G. (2013). Cross-Country Linkages in Europe: A
Global VAR Analysis. IMF Working Papers,13/194.
55. Sznajderska, A. (2019). The role of China in the world economy: evidence from
a global VAR model. Applied Economics, 51(15), 1574-1587.
56. Tarsidin, Idham, & Rakhman, R. N. (2018) Nowcasting Household
Consumption and Investment in Indonesia. Bulletin of Monetary Economics
and Banking, Volume 20, Number 3, January 2018.
57. Vasishtha, G., and Maier, P. (2013). The Impact of the Global Business Cycle on
Small Open Economies: A FAVAR Approach for Canada. The North American
Journal of Economics and Finance, 24, 191-207.
58. Walsh, C. E. (1999). Changes in the Business Cycle. FRBSF Economic Letter,
1999(16).
59. Wimanda, R. E., & Djuranovik, L. (2014a). Siklus Bisnis and Siklus Finansial
Indonesia: Revisited. Catatan Riset Grup Riset Ekonomi Bank Indonesia No.
16/9/DKEM/GRE/CR.
60. Wimanda, R. E., & Djuranovik, L. (2014b). Business Cycles and Financial Cycles
in Indonesia: Analysis by Components. Catatan Riset Grup Riset Ekonomi Bank
Indonesia No. 16/8/DKEM/GRE.
 The Determinants of Indonesia’s Business Cycle 233

Appendix

Table A1. List of countries

Area Countries
Americas Argentina, Brazil, Canada, Chile, Mexico, Peru, USA
Euro (includes Austria, Belgium, Finland, France, Germany, Italy,
Europe
Netherlands, Spain), Norway, Sweden, Turkey, United Kingdom.
Australia, China, Indonesia, India, Japan, Korea, Malaysia,
Asia-Pacific
New Zealand, Philippines, Singapore, Thailand
Africa and Middle East South Africa, Saudi Arabia

Table A2. Weak Exogeneity Test

Equity Real exchange Short term Long term
Country Fcritical Output Inflation
price rate interest rate interest rate
Argentina 3.1 1.0 1.0 0.3 3.3 1.4
Australia 2.7 1.9 2.4 0.9 0.6 2.6
Brazil 3.1 0.4 1.4 0.4 0.3 3.7
Canada 2.7 5.2 4.0 0.1 1.8 0.2
China 3.1 0.2 0.8 0.0 1.6 1.3
Chile 3.1 0.9 0.1 1.0 0.7 1.5
Euro 3.9 0.6 3.1 1.1 0.4 0.1
India 3.1 3.0 0.7 1.5 2.3 1.0
Indonesia 2.7 1.6 0.5 0.6 1.2 0.2
Japan 3.1 2.1 0.7 0.1 0.4 0.5
Korea 2.7 0.2 1.1 1.8 1.0 1.5
Malaysia 3.1 2.7 3.7 2.4 4.0 0.3
Mexico 3.1 0.4 4.2 0.3 1.1 1.4
Norway 2.7 3.1 1.9 1.1 0.2 0.9
New Zealand 2.7 4.3 0.6 0.7 0.8 0.1
Peru 3.1 0.4 1.0 0.7 3.0 0.2
Philippines 2.7 0.4 0.7 0.5 2.1 3.2
South Africa 3.1 0.2 0.3 0.1 2.1 0.4
Saudi Arabia 3.9 0.2 0.0 2.1 0.0 1.5
Singapore 3.9 1.4 0.3 6.1 2.6 2.3
Sweden 3.1 0.7 0.7 0.0 0.2 0.6
Switzerland 2.7 1.9 0.7 1.7 0.6 0.2
Thailand 3.1 0.5 0.9 0.3 0.8 0.7
Turkey 3.9 1.7 0.8 0.0 0.2 0.0
United Kingdom 3.1 2.8 0.5 0.1 0.7 1.2
United States 3.1 0.7 3.3 0.3

The table reports F-statistics that indicate weak exogeneity of foreign variables.
 Journal of Central Banking Theory and Practice · Special Issue Proceedings from 13th BMEB International Conference in Bali:
234 Maintaining Stability, Strengthening Momentum of Growth Amidst High Uncertainties

Table A3. VECMX Residual Correlation

Equity Real Short term Long term
Output Inflation
price exchange rate interest rate interest rate
argentina 0.01 0.05 -0.02 0.03 0.01
Australia 0.03 0.02 0.04 0.19 0.02 0.01
Brazil 0.04 -0.04 0.13 -0.03
Canada 0.00 0.04 0.04 0.14 0.09 -0.02
China -0.09 -0.02 0.03 0.02
Chile 0.01 0.01 0.04 0.16 -0.02
Euro -0.01 0.05 -0.11 0.29 0.07 -0.09
India -0.02 0.01 -0.03 0.14 0.04 -0.01
Indonesia -0.02 0.02 0.08 0.03
Japan -0.02 0.02 -0.11 0.13 0.00 -0.06
Korea 0.02 0.05 -0.05 0.14 0.06 -0.06
Malaysia -0.01 0.02 0.01 0.19 0.04
Mexico 0.03 0.01 0.03 0.02
Norway -0.01 0.03 0.05 0.28 0.01 0.02
New Zealand 0.05 0.03 0.00 0.22 0.04 0.02
Peru 0.02 -0.04 0.05 0.03
Philippines 0.01 0.00 0.01 0.15 0.02
South Africa 0.05 0.03 0.06 0.19 0.03 0.00
Saudi Arabia 0.00 0.04 0.06
Singapore -0.02 0.03 0.00 0.24 0.01
Sweden 0.02 0.06 0.00 0.23 0.00 0.02
Switzerland 0.02 0.05 -0.02 0.26 -0.01 0.01
Thailand 0.01 0.00 0.01 0.20 0.04
Turkey 0.01 0.00 0.12 0.03
United Kingdom -0.01 0.00 0.00 0.19 0.04 0.00
United States -0.04 0.07 -0.01 0.04 -0.01

The table reports average cross-sectional correlation of VECMX residuals.

 The Determinants of Indonesia’s Business Cycle 235

Table A4. Trade Weight Matrix

United United
Country Brazil China Chile Euro Indonesia Japan Mexico Peru
Kingdom States
Brazil 0.00 0.03 0.08 0.02 0.01 0.01 0.01 0.05 0.01 0.02
China 0.23 0.00 0.28 0.16 0.18 0.27 0.10 0.26 0.10 0.19
Chile 0.03 0.01 0.00 0.01 0.00 0.01 0.00 0.04 0.00 0.01
Euro 0.20 0.16 0.14 0.00 0.08 0.10 0.07 0.13 0.48 0.15
Indonesia 0.01 0.02 0.00 0.01 0.00 0.03 0.00 0.00 0.00 0.01
Japan 0.03 0.11 0.07 0.04 0.13 0.00 0.03 0.04 0.02 0.06
Mexico 0.03 0.02 0.03 0.02 0.00 0.01 0.00 0.04 0.00 0.17
Peru 0.01 0.01 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00
United Kingdom 0.02 0.03 0.01 0.17 0.01 0.02 0.01 0.01 0.00 0.04
United States 0.18 0.21 0.18 0.18 0.09 0.19 0.68 0.22 0.14 0.00

The table shows a portion of trade weight matrix used in GVAR estimation. The number
represents trade portion of country in each column with partner country in rows.