Co-movement in Asset Market: Does Global Financial Cycle Works?
Empirical Evidence in Indonesia
Sri Andaiyani and Saadah Yuliana
Faculty of Economics, Universitas Sriwijaya, Palembang, Indonesia
Keywords: global financial cycle, risk aversion, asset, stock
Abstract: The movement of global financial risk is more volatile and procyclical during US unconventional monetary
policy. Indonesia, as one of the important EMEs in the world, also received higher capital inflows. Asset
markets of home country is more vulnerable to global risk aversion movement. T herefore, this study
attempts to analyze the impact of global financial cycle on asset markets in Indonesia using Vector
Autoregressive model (VAR). The empirical findings of this study are twofold. First, Global financial cycle
has a significant effect on stock and exchange rate markets. This result is consistent with Indonesia as an
open capital account country that remain vulnerable to the global financial cycle. Second, robustness check
reveals strong evidence that co-movement in Indonesian’s asset markets is affected by global financial cycle
as proxy the VIX index.
1 INTRODUCTION
After financial crisis 2008, the central bank of
the United States, known as the Federal Reserve or
the Fed, injected the unprecedented amount of
liquidity through large-scale asset purchases
(LSAPs). Fluctuations in global financial condition
related to unconventional monetary policy in the
United States also made investors switched to other
investment assets in the emerging market economies
(EMEs) such as bond markets and stock markets. In
EMEs, Stock price and bond price have increased
during the economic recovery in advanced
economies. The ability of financial institutions to
keep the effects of risk in global markets determine
the performance of the financial institution. If the
financial institution in a country can control the
global risk arising from the developed countries, the
financial market conditions in developing countries
will be better. The financial market risks consist of
movements in interest rates, stock price index,
commodity price index, or the exchange rate. This is
supported by a statement from Fratzscher et al
(2013) that the global externality effects of monetary
policy decisions in developed countries do affect the
developing countries.
Figure 1: Global financial cycle (VIX Index)
Source: CBOE VIX index, Datastream
Movement of global financial risk is more
volatile and procyclical (figure 1). Asset markets of
home countries are more vulnerable to sudden rise in
global risk aversion. According to Rey (2015), there
is two global factor that drives the movements in
capital flows and asset prices across EMEs. The first
is the global financial cycle that reflects both
aggregate volatility of asset markets and degree of
risk aversion of markets. The second set of global
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
90,00
652
Andaiyani, S. and Yuliana, S.
Co-movement in Asset Market: Does Global Financial Cycle Works? Empirical Evidence in Indonesia.
DOI: 10.5220/0008443606520658
In Proceedings of the 4th Sriwijaya Economics, Accounting, and Business Conference (SEABC 2018), pages 652-658
ISBN: 978-989-758-387-2
Copyright
c
2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
factors is US monetary policy that has a significant
effect on cross-border credit flows. A tightening of
US monetary policy leads to increase global risk
aversion, a fall in cross-border lending and a fall in
asset prices at the global level (Miranda-Agrippino
and Rey, 2015).
Indonesia, as one of the important EMEs in the
world, has received higher capital flows since the
global financial crisis. Capital inflows could have a
significant effect on asset prices, including property
prices (Falianty, 2016). Nevertheless, there are no
empirical studies of the spillover effect from the
global financial cycle to Indonesian’s asset markets.
Therefore, this research attempts to analyze the
effect of the global financial cycle on asset markets
in Indonesia. In this study, the Indonesian asset
markets financial markets, comprising the exchange
rate market, the stock market, and the bond market
are considered.
Figure 2: Asset Markets in Indonesia
Source: Datastream
Recent studies on global financial cycle argued
that global financial risk may have the impact on
asset markets. Miranda-Agrippino and Rey (2015),
Banerje, Devereux and Lambardo (2016) and Rey
(2015) argued that global financial factor explains
the important role of the variance of a large cross-
section of returns of risky asset price around the
world. Other studies that global risk sentiment has a
significant effect on fragile five asset markets. When
market sentiment deteriorates, equity prices fall and
local currencies depreciate, while LC government
bond yields and sovereign CDS prices increase
(Yildirim, 2016).
One of the contributions of this study is to give
more information about the transmission of the
global financial cycle to Indonesian’s asset markets.
The transmission of global financial cycle provides
an overview of investor’s sentiment in asset markets.
Besides, I employ data frequency using weekly and
monthly data to see whether the global financial
cycle has a deferent impact or not on asset markets.
An important question in this study is whether this
increased co-movement of global financial risk
provides evidence of contagion. Defining contagion
as a significant increase in cross-country co-
movement of asset returns (Dungey and Gajurel,
2014).
2 LITERATURE REVIEW
Blanchard et al. (2010) explain that an increase
in global financial risk was an important channel
through which the crisis was propagated to emerging
economies. The empirical studies from Longstaff et
al. (2011) suggest that global factors explain a large
fraction of the variation in the international interest
Co-movement in Asset Market: Does Global Financial Cycle Works? Empirical Evidence in Indonesia
653
rate. The recent study by Yildirim (2016) also shows
that global financial risk factors have hit the asset
markets in the emerging fragile five countries.
Global financial risk aversion has sharply increased,
the exchange rate of local currencies depreciated,
government bond yield and country risk premium
also increased significantly, but stock prices
decreased.
Recent studies have investigated the effects of
the global financial cycle to asset markets and
macroeconomic conditions. Using two analytical
approaches – turning-point analysis and frequency-
based filters, Drehmann, Borio and Tsatsaronis
(2012) find that global financial cycles are best
captured by combinations of credit and property
prices, while equity prices do not fit the picture well.
The theory that explains the transmission of the
global financial cycle to asset market in EMEs
country is international investor risk appetite due to
market imperfections or the behavior of international
investors. Information asymmetries make investors
more uncertain about the actual economic
fundamentals of a country (Dungey and Gajurel,
2014). Falianty (2016) discuss the impact of capital
flows on the property market and the impact of
macroprudential policies represented by Loan to
Value (LTV) regulation on the property market in
Indonesia. She finds the significance of GDP to
Property Price Index (PPI). Capital flows (CF) and
LTV regulation have not significantly affected the
property price index, even for CF have the
marginally significant effect to PPI.
Yildirim (2016) provides theoretical framework
between global financial risk, capital inflows, and
EMEs asset prices. The result finds that global
financial risks depend on the strength of a country’s
macroeconomic conditions. In other words, these
impacts vary across asset classes and countries.
Some researchers focus on the risk-taking channel of
monetary policy to explain these links. In this case,
Bruno and Shin (2015) build the model by focusing
on the functioning of this channel via the banking
sector. The model suggests that movement in US
unconventional monetary policy are transmitted
internationally via shifts in global risk aversion,
which drive the asset prices in EMEs by affecting
leverage of financial intermediaries, bank lending,
and thereby, portfolio inflows into their economies.
Other studies from Lizarazo (2013) develops a
model for small open economies taking into account
risk-averse international investors with decreasing
absolute risk-aversion preferences, which is
consistent with the typical features of investors in
EM financial markets. The model provides a
possible mechanism to explain the links between
investors' characteristics (risk aversion and wealth),
capital inflows to EMs, and EM asset prices, notably
sovereign risk premiums and bond prices. Based on
the mechanism, as international investors become
more risk-averse, sovereign CDS prices move higher
while capital inflows to EMs and their bond prices
decrease (Yuldirim, 2016).
3 THEORETICAL FRAMEWORK
Since financial crisis, asset markets in the world
have become increasingly integrated with large
portfolio flows. But Global banks, namely asset
managers, have an important role in the process of
internationalization. This study follows theoretical
model from Miranda-Agrippino and Rey (2015).
They explain a theoretical framework in
international asset pricing where the risk premium
depends on the wealth distribution between
leveraged global banks and asset managers that have
more fund. It can help to interpret the data in the
best way.
Miranda-Agrippino and Rey (2015) assume that
there are two types of investors: global banks and
asset managers. Global banks are affected entities
that fund themselves in dollars because they operate
in the global capital markets. They can borrow at a
rate of US risk-free rate and a lever to buy risky
assets in dollars. Investors are risk neutral with
constraints Value-at-Risk (VaR) and will then be
imposed the rules. Risk neutrality is an extreme
assumption that might justify the fact that investors
benefit from a guarantee, either because they are a
universal bank that is part of the guarantee scheme,
or because the risk of failure is greater.
Asset managers hold a portfolio of regional
assets which are non-tradable assets in the financial
markets. It can occur due to asymmetric information.
Miranda-Agrippino and Rey (2015) stated that any
global bank will maximize expected yields to be
obtained from a portfolio of risky assets held by the
constraints Value-at-Risk. Risk values define the
upper limit predicted the number of banks suffered
losses in the portfolio. Another research from Adrian
and Shin (2014) shows the value of risk is taken out
of proportion to the standard deviation of the bank's
portfolio risk. Global banks choose portfolio as
follows;
)
s.t V a
≤
SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference
654
states vector of excess return of all risky
assets traded in the world. Risky assets are all
tradable securities such as equities and corporate
bonds. Portfolio securities of global bank portfolios
expressed
and
is the equity of the bank.
V a
= α
By following Lagrangian optimization problems
of literature Miranda-Agrippino and Rey (2015)
sought its First Order Conditions of the obtained
solution to asset demand is
. (1)
Equation (1) shows the average portfolio
allocation of the investor.
is a Lagrange multiplier
symbol. VaR constraint has the same role as risk-
averse. Furthermore, asset managers are average
standard variance of investor. They have the same
access to the assets traded by global banks. The
owners of assets are also invested in local assets
(regional) that were not traded. Asset managers
chooses their portfolio of risky assets by
maximizing:
indicate the vector of portfolio weights of the
asset managers in tradable risky assets. While
as a
fraction of their wealth that is invested by asset
managers in their regional assets. Vector of excess
returns on investments that are not traded expressed
by
, and
I is an equity of asset manager.
Therefore, the selection of optimal portfolio in risky
securities that can be traded at risk for asset
managers, namely:
(
. (2)
4 METHODOLOGY AND
DATASET
4.1 Empirical Method
Some literatures employ a Vector Autoregressive
(VAR) model to test the effects of global financial
cycle on asset markets in EMEs. Yildirim (2016)
employs VAR model especially Structural VAR
(SVAR) to analyze the impact of global financial
shock on fragile five asset markets. On the other
hand, VAR model can capture the dynamic
interaction between monetary policy, risk aversion
and uncertainty, leverage and credit flows (Rey,
2015; Akinci, 2013).
Following previous empirical model, this study
also employs VAR model to analyze the effects of
global financial cycle on Indonesia’s asset market
using daily data. This estimation techniques
preceded by several standard measures such as
stationary test or stationary stochastic process (Ajija
et.al, 2011) and determination of the optimal lag
with lag order selection criteria. Sims (1980) states
that if there is a simultaneous relationship between
variables observed, the variables should be treated
equally, no more endogenous and exogenous
variables. The VAR is used to prove an economic
theory or to find theoretical foundations from a
shock (Bilmeier and Bonatot, 2002).
VAR Model
The specification of the VAR model in reduced
form is,
Where
is a vector with all variables;
is a
contemporaneous relation among variables; A (L) is
a finite-order matrix polynomial with the lag
operator L;
is a vector of structural disturbance;
and B is a non-zero diagonal matrix. Basically, there
are several ways to place restrictions on the VAR
model, such as long run restriction, impact, and sign
restriction. This restriction helps to identify the
models and to insert the basic theory into the model.
4.2 Dataset
To investigate the response of Indonesian’s asset
prices to global financial risk shocks, this study
focuses on daily data from January 2, 2006 to
October 30, 2016. Using daily data can capture the
spillovers of external financial shocks on Indonesian
asset markets because high frequency daily data give
more information and more precise analysis.
However, lower data frequencies, like those of
weekly and monthly, are used in the literature as
follows Yilidrim (2016).
The data were obtained from Thomson Reuters
DataStream. The unit of measurement of the data
used is quite varied. Asset markets in this study are
divided to three markets. It includes exchange rate
Co-movement in Asset Market: Does Global Financial Cycle Works? Empirical Evidence in Indonesia
655
market, bond market and stock market. For
exchange rate market, I use Indonesian rupiah to US
dollar. Then, I use 10-year government bond yield to
capture the effect of global financial cycle to bond
market. For stock market, I employ Indonesian
Stock Exchange (IDX). Moreover, the VIX index is
used to proxy global financial cycle. All variables
are measured in logarithms except for the
government bond yields., which is expressed in
percentages.
5 EMPIRICAL RESULT
In this section, empirical result will be
discussed. The testing procedure conducted to test
the data stationarity is Augmented Dickey-Fuller test
(ADF). From stationary testing, the results obtained
indicate that all variables are stationary at first
difference except VIX index. Furthermore, the
estimated VAR model followed by determining
optimal lag length in model. Determination of the
optimal lag length is important in modeling the
VAR. If the optimal lag entered is too short, it could
not explain the dynamism of the whole model. The
test results of lag length in the VAR is 2.
Table 1: Unit Root Test
Level
First Difference
VIX
-3.95401
***
-
JAKCOMP
-2.545979
-25.29732
***
GbYield
-1.626074
-20.51981
***
EXRATE
-0.268953
-31.47800
***
Note: The test critical value for 1%, 5%, and 10%
significance level are -3.478, -2.882, and -2.578
respectively. ***, **, * denote significance at 1%,
5%, 10%, respectively.
Further testing is cointegration test by Johansen
cointegration test (Johansen Test of Cointegration)
to test whether there is a long-term relationship in
the analysis that will use the VAR model. Testing is
done by comparing the value of the Max-eigen value
statistic with critical value at α = 5%. Based on the
value of the Max-eigen value statistic on Johansen
cointegration test (See Table. 2), it can be concluded
that there is no cointegration relationship among the
variables in the long term. Furthermore, I estimate
the data with VAR model. The analysis of the VAR
model in this study can focus on relationship of asset
markets variable with VIX index.
Table 2: Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None
0.170125
28.15859
30.81507
0.1021
At most 1
0.121005
19.47533
24.25202
0.1891
At most 2
0.094104
14.92338
17.14769
0.1024
At most 3
0.003401
0.514492
3.841466
0.4732
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
By estimating VAR on daily data, this study
indicates that asset markets in Indonesia are affected
by global financial cycle. These markets are more
vulnerable with capital inflows that related to global
financial cycle. A sharp increase in global financial
risk during this period decreased stock price and
depreciated Indonesian currency. However, global
risk aversion does not have effect on government
bond yield. Consequently, Indonesia must offer high
coupon rates on government bonds to attract investor
appetite. As the result, global financial cycle shock
has a significant impact on Indonesian’s asset
markets. Global financial cycle has a positive and
significant effect on exchange rate market. It is
relevant with flexible exchange rate regime which is
applied in Indonesia. As in countries with more
flexible regime in exchange rate, the effect of global
financial cycle on exchange rate volatility can be
quite large (Ananchotikul and Zhang, 2014). The
VAR estimation results may not provide a
comprehensive analysis because the evidences that
examine a significant relationship among the
variables are very limited.
Hence, this study employs the Impulse Response
Function (IRF) to examine the impact of shock of
innovation variable to other variables. The
estimation using the assumptions that each of
innovation variables do not correlate with one
another, so that a shock effect may be direct.
SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference
656
Besides, being able to determine the effect and
duration of the shock, the IRF approach can also be
used to determine how long the shock effect will
end. Figure of impulse response will show a
response of a variable due to shock from other
variables until some period after the shock. If figure
of impulse response shows the movement that
getting closer to the point of equilibrium
(convergence) or return to the previous equilibrium,
it means that the response of variable of a shock will
disappear, so that the shock does not leave a
permanent effect on these variables. Figure 2
presents the response global financial cycle shock on
asset prices in Indonesia. It shows the response
Indonesian’s asset markets variables to a 1-standard
deviation increase in global risk aversion (i.e., a1-
standard deviation increased in the VIX). due to
more simpler and the result of estimation relatively
similar with least square regression model.
Global financial cycle does not have any effect
on 10-year government bond yield. Investors are not
interested to Indonesian government bond yet. To
attract investors' appetite, Indonesia had to offer
expensive coupons for its sovereign bonds
through Finance Ministry Regulation No.
91/PMK.010/2016 in May 2016. Moreover, high
coupon rates on government bonds also influence
Indonesia's corporate bond yields as investors are
not interested in corporate bonds that carry a
significantly lower coupon rate compared to the
government bonds (corporate bonds tend to use the
government bond yield as reference).
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
1 2 3 4 5 6 7 8 9 10
Response of CBOEVIX to CBOEVIX
-5
0
5
10
15
20
1 2 3 4 5 6 7 8 9 10
Response of D(EXRATE) to CBOEVIX
-.01
.00
.01
.02
.03
.04
1 2 3 4 5 6 7 8 9 10
Response of D(GBYIELD) to CBOEVIX
-16
-12
-8
-4
0
4
1 2 3 4 5 6 7 8 9 10
Response of D(JAKCOMP) to CBOEVIX
Response to Cholesky One S.D. Innovations ± 2 S.E.
Figure 3: Impulse response function
By contrast, the global financial shock has a
positive and significant impact on LC government
bond yields in five fragile emerging economies-
Brazil, India, Indonesia, South Africa, and Turkey
(Yildirim, 2016; Ananchotikul and Zhang, 2014).
Historically, the Indonesia Government Bond 10Y
reached an all times high of 20.76 in October of
2008 and a record low of 4.99 in February of 2012.
The other empirical evidence shows that Indonesian
stock markets are affected by the global financial
cycle. This finding is consistent with Yldirim
(2016), Ananchotikul and Zhang (2014), Chudik and
Fratszcher (2011) that global financial risk caused a
decline in stock prices. Particularly, the effect of the
global factor on stock market volatility is correlated
with the financial openness of the country, as
measured by total financial liabilities as a percent of
GDP. The more exposed a country is to external
fund flows, the greater is the volatility spillover
deriving from higher global risk aversion to the
domestic equity market.
This study clearly appears that an increase in
global financial risk is acknowledged by investor
sentiment in EMEs especially in Indonesia. This
result is consistent with Indonesia as an open capital
account country that remains vulnerable to the
global financial cycle. In addition, the VIX index as
a representative of the global financial cycle tends to
boost a tightening monetary policy (Miranda-
Agrippino and Rey, 2015).
Robustness checks
In this part, to have the strong analysis of these
findings, I employ the robustness check using the
same model but different data frequencies including
weekly and monthly data. Yildirim (2016) argued
that there is an emerging consensus that data
frequency matters in examining the link between
financial variables. Therefore, I check whether the
difference of data frequency has a similar impact to
the empirical result of recent literature.
This study estimates the VAR model with
weekly and monthly data to check whether these
findings depend on the data frequency or not. The
empirical results are similar to the previous result
with daily data in this study. These results confirm
that the global financial cycle has a significant
impact on Indonesian’s asset markets. Furthermore,
the Indonesian currency has depreciated when the
global financial cycle sharply increased.
6 CONCLUSION
This empirical results in this study support some
literature about the impact of the global financial
cycle on co-movement asset markets in EMEs.
By estimating VAR on daily data, this study
indicates that asset markets in Indonesia are affected
by global financial cycle. These markets are more
vulnerable with capital inflows that related to global
Co-movement in Asset Market: Does Global Financial Cycle Works? Empirical Evidence in Indonesia
657
financial cycle. A sharp increase in global financial
risk during this period decreased stock price and
depreciated Indonesian currency. But global risk
aversion does not have an effect on government
bond yield. Consequently, Indonesia must offer high
coupon rates on government bonds to attract investor
appetite. In addition, robustness checks in this study
are consistent with empirical findings on daily data.
This conclusion is consistent with the fact that
the Indonesian financial market is still strongly
affected by foreign financial markets, so if there is a
shock in the global financial market, that will easily
cause panic among domestic investors. Bank
Indonesia, as policymakers, send clear signals to
stand ready to supply the foreign exchange and at
the same time buy the bonds that foreign investors
wish to unwind, and thus avoiding herding behavior
and contagion of escalating capital reversals.
Moreover, the intervention is a way to bring about
the objective of monetary stability to be consistent
with maintaining financial system stability. By
stabilizing the foreign exchange market and
government bond market, the intervention helps in
stabilizing the overall financial markets.
Further research may extend this analysis by
giving more information about stance domestic
monetary policy that relates to global financial risk
aversion. To get more specific description associated
with the problem, the analysis of the data can also be
directed to the semi-quantitative method (a blend of
quantitative and qualitative methods), so that the
statistical facts can be synchronized with the
behavioral aspects.
REFERENCES
Ajija et al. 2011. “Cara Cerdas Menguasai Eviews”.
Salemba Empat
Bilmeier and Bonatot. 2002. Exchange Rate Pass-Through
and Monetary Policy in Croatia. IMF Working Paper
WP/02/109.
Ananchotikul, N., Zhang, L., 2014. Portfolio flows, global
risk aversion, and asset prices in emerging markets.
IMF Working Paper WP/14/156 (August).
Akinci, Ö., 2013. Global financial conditions, country
spreads, and macroeconomic fluctuations in emerging
countries. J. Int. Econ. 91 (2), 358–371.
Banerjee, R, Michael B. Devereux dan Giovanni
Lombardo. 2016. Self-oriented monetary policy,
global financial markets and excess volatility of
international capital flows. Journal of International
Money and Finance 68 (2016) 275–297
Blanchard, O.J., Das, M., Faruqee, H., 2010. The initial
impact of the crisis on emerging market countries.
Brookings Papers on Economic Activity 41 (1), 263–
323.
Bowman, D., Londono, J.M., Sapriza, H., 2015. US
unconventional monetary policy and transmission to
emerging market economies. J. Int. Money Finance.
55, 27–59.
Dungey, Mardi, & Gajurel. 2015. Contagion and banking
crisis – International evidence for 2007–2009. Journal
of Banking & Finance. Vol. 60 (C) Pages 271-283.
Chudik, A., Fratszcher,M, 2011. Identifying the global
transmission of the 2007–09 financial crisis in a
GVAR model. Eur. Econ. Rev. 55, 325–339.
Drehmann, M, C Borio and K Tsatsaronis.2012.
Characterizing the financial cycle: don’t lose sight of
the medium term, BIS Working Papers No 380.
Ebeke, C., Kyobe, A., 2015. Global financial spillovers to
emerging market sovereign bond markets. IMF
Working Paper WP/15/141.
Falianty, Telisa. 2016. Capital Flows, Macro Prudential
Policy, and Property Sector, IJABER, Vol. 14, No. 10
(2016): 6935-6958.
Fratzscher, M., Lo Duca, M., Straub, R., 2013. On the
international spillovers of US quantitativeeasing. ECB
Working Paper No. 1557.
Longstaff, F.A., Pan, J., Pedersen, L.H., Singleton, K.J.,
2011. How sovereign is sovereign credit risk?
American Economic Journal: Macroeconomics 3 (2),
75–103.
Lizarazo, S.V., 2013. Default risk and risk averse
international investors. J. Int. Econ. 89 (2), 317–330.
http://dx.doi.org/10.1016/j.jinteco.2012.08.006.
Miranda-Agrippino, Silvia dan Rey H. 2015. World Asset
Markets and The Global Financial Cycle. NBER
working paper series.
Narayan, P.K., Ahmed, H.A., Narayan, S., 2015. Do
momentum-based trading strategies work in the
commodity futures markets? J. Futur. Mark. 35, 868–
891.
Rey, Hélène. 2015. Dilemma not Trilemma: The global
Financial Cycle and Monetary Policy Independence.
NBER Working Paper No. 21162
Setiawan, Budi. 2018. LASSO Technique Application in
Stock Market Modelling: An Empirical Evidence in
Indonesia. https://doi.org/10.29259/sijdeb.v2i1.%25p
Yildirim, Zekeriya.2016. Global financial conditions and
asset markets: Evidence from fragile emerging
economies. Economic Modelling 57 (2016) 208–22
SEABC 2018 - 4th Sriwijaya Economics, Accounting, and Business Conference
658
