Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
ANALYSING MACROECONOMIC EFFECTS OF FINANCIAL SHOCK DURING GLOBAL
FINANCIAL CRISIS AND ITS CONTAGION ON EMERGING MARKETS
A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy
By
Hummaira Jabeen
DEPARTMENT MANAGEMENT SCIENCES Faculty of Commerce, Economics and Management
Sciences ISRA UNIVERSITY, HYDERABAD
December 2019
ANALYSING MACROECONOMIC EFFECTS OF FINANCIAL SHOCK DURING GLOBAL
FINANCIAL CRISIS AND ITS CONTAGION ON EMERGING MARKETS
By
Hummaira Jabeen
NAME OF SUPERVISOR AND CO- SUPERVISOR
Dr. Nadeem Qureshi (Supervisor) Assistant Professor
Dr. Hakimzadi Wagan (Co-Supervisor) Assistant Professor
PROF. DR. Hameedullah Kazi (Co-Supervisor)
iii
ACKNOWLEDGEMENTS
I would like to express my very great appreciation to assistant Professor
Dr. Muhammad Nadeem Qureshi; my research supervisor, who gave me the
golden opportunity to do this thesis and for his valuable and constructive
suggestions during the planning and development of this research work.
My deep gratitude to Professor Hakimzadi Wagan, research Co-
supervisor, for her patient guidance, enthusiastic encouragement and useful
advice and critiques on this research work and assistance in keeping my progress
on schedule. She helped me a lot in doing this Research and under her
supervision I came to know about so many new things. I am really thankful to her.
Her willingness to give her time so generously has been very much appreciated.
My special thanks to Prof. Dr. Hameedullah Kazi, my research co-
supervisor for his professional guidance and valuable support.
Finally, I wish to thank my parents for their support and encouragement
throughout my study.
iv
ABSTRACT
Financial system working at a global level is in a constant state of evolution.
This state of Globalization carry benefits for the economies but with the same
token also carries cascading defaults and failures.
The purpose of this study is to explore the transmission of U.S. monetary
policy shocks during mortgage crisis 2007 to emerging countries. In this study
classification of emerging markets by Financial Times Stock Exchange (FTSE) is
employed. FTSE classifies emerging markets into advance emerging markets
which include Brazil, Czech Republic, Hungary, Malaysia, Mexico, Poland, South
Africa, Taiwan, Thailand and Turkey and the secondary emerging markets namely
Chile, China, Colombia, Egypt, India, Indonesia, Pakistan, Peru, Philippines,
Russia and UAE.
For working on the attainment of the purpose, VAR methodologies are used.
At first level, transmission of monetary policy is focused; this has been studied at
two levels. Transmission arising from national monetary policy and the impact of
monetary policy shock arising from United States on the economy of emerging
markets in a time varying context. This objective is achieved using the TVP-VAR
model with stochastic volatility.
At second level, impact of financial shock has been studied. One of the key
outcomes of the US financial global crisis is that due to financial innovations we
are unable to capture the broader horizon of financial conditions with just few
variables. Keeping this view in front, objective of this study is to offer an empirical
v
assessment of the effects of the financial conditions of the United States upon
macro-economy of the emerging economies using standard Vector Auto-
Regression (VAR) Models. This objective is achieved by utilizing Financial
Conditions Index of Brave and Butter (2011) for the assessment of impact upon
macro-economy of the emerging markets as being classified by Financial Times
Stock Exchange (FTSE).
At third level, in this study, an index is created using wide range of macro-
economic and financial variables over a long horizon for the Pakistan using a time
varying model developed by Koop and Korobilis (2014). This method develops
and forecasts financial conditions index.
It is being found that ISLM framework is partly applicable in many cases.
Prize puzzle exists more in county specific monetary policy as compared to
international contagion. Response of output is aligning with the theory at country
level but not in international level. Hungary, Turkey, Malaysia at country level fully
aligns with the theory. Mexico and Turkey fully deviate from theory in the case of
contagion impact.
It is also found that countries do have impact of financial conditions of the
US but in many cases, Impact die off with the time. Extend vary from county to
country but impact does exist on the macro-economy of the emerging countries.
The countries that are having Free Trade Agreement with the US are having
strong and long-term response. Bilateral partners' response dies off with the
passage of time except Russia and Hungary. In the last but not the least, financial
vi
condition index of the Pakistan is able to give a true picture of the economy.
Forecasting of the index of the macroeconomic variables is close to the reality.
Key Words
Financial conditions, Bayesian analysis, financial forecasts, monetary policy
shock, emerging economies, TVP-VAR analysis, financial conditions, emerging
markets, SVAR.
JEL Classification
C53, G17, G19, C58, E02, F62, G01
vii
ABBREVIATION AND SYMBOLS
Symbol/Abbreviation Term
𝐴𝑦------------------------------------------------------------------------------------general matrix
A, Fi,𝐵𝑖 ,𝐶𝑖 , 𝑔𝑖, 𝐷𝑖-------------------------------------------------- vector/matrices of coefficients
𝐺0 ---------------------------------------------------------------------- (n*1) vector of constants
𝐺1-------------------------------------------------------------------- (n*n) matrix of coefficients
𝑝𝑡---------------------------------------------variable indicating the Monetary Policy stance
𝑣𝑡𝑝, 𝑣𝑡
𝑦, ut, εt----------------------------------------------------------------------structural shock
𝑥𝑡----- (n*1) vector of financial and economic variables for the construction of FCI
yt ---------------------------------------------------------- (k*1) vector of observed variables
𝑍𝑡-------------------------------------------vector of non-policy macroeconomic variables
Σ (Sigma)-----------------------------------------------------------------------------------addition
⊗------------------------------------------------------------------------------ Kronecker product
𝜆𝑡𝑦--------------------------------------------------------------------------regression coefficients
𝜆𝑡𝑓------------------------------------------------------------------------------------factor loadings
𝑓𝑡----------------------------------------------------------------latent factor interpreted as FCI
yt--------------------------------------------------------------vector of observed variables
ct---------------------------------------------------------------------------------------------intercept
BIS---------------------------------------------------------Bank of international settlements
viii
CPEC----------------------------------------------------China Pakistan economic corridor
DOH---------------------------------------------------Dornbusch overshooting hypothesis
EKF---------------------------------------------------------------------Extended Kalman filter
EME-------------------------------------------------------------Emerging market economics
FATVPVAR-Factor Augmented-Time varying parameters vector auto regression
FAVAR----------------------------------------Factor augmented vector auto regression
FCI-------------------------------------------------------------------Financial condition index
FOMC------------------------------------------------------Federal open market committee
FTSE--------------------------------------------------------Financial times stock exchange
GARCH----------------Generalized Autoregressive Conditional Heteroskedasticity
GDP-------------------------------------------------------------------Gross domestic product
GFC-----------------------------------------------------------------------Global financial crisis
IFS------------------------------------------------------------International financial statistics
IMF---------------------------------------------------------------International monetary fund
IPI------------------------------------------------------------------Industrial production index
IS-LM---------------------------------------------------Investment savings-liquidity money
KFS---------------------------------------------------------------------Kalman filter smoother
KSE-------------------------------------------------------------------Karachi stock exchange
MCI------------------------------------------------------------------Monetary condition index
MCMC-------------------------------------------------------------Markov chain Monte Carlo
NFCI------------------------------------------------------National financial condition index
OECD--------------------Organization for Economic Cooperation and Development
OLS----------------------------------------------------------------------Ordinary least square
OPEC-----------------------------------Organization of Petroleum Exporting Countries
ix
PCA------------------------------------------------------------Principal component analysis
PSX------------------------------------------------------------------Pakistan stock exchange
REER-----------------------------------------------------------Real effective exchange rate
S&P----------------------------------------------------------------------Standard’s and poor’s
SBP---------------------------------------------------------------------State bank of Pakistan
SDG---------------------------------------------------------Sustainable development goals
SME----------------------------------------------------------Small and medium enterprises
SRO-----------------------------------------------------------------Statutory regulatory order
SVAR------------------------------------------------------Structural vector auto regression
TVPFAVAR-Time varying Parameters-Factor augmented vector auto regression
TVP-VAR----------------------------Time varying parameters-vector auto regression
UK--------------------------------------------------------------------------------United Kingdom
US/USA------------------------------------------United States/ United states of America
VAR---------------------------------------------------------------------Vector auto regression
x
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ------------------------------------------------- iii
ABSTRACT ------------------------------------------------------------------- IV
ABBREVIATIONS & SYMBOLS ----------------------------------------- VII
TABLE OF CONTENTS --------------------------------------------------- X
LIST OF TABLES ----------------------------------------------------------- XV
LIST OF FIGURES --------------------------------------------------------- XVI
CHAPTER - I ………………………………………………………. INTRODUCTION ……………………………………………………
01
1. Shock and Contagion---------------------------------------------------- 02 1.1 Historical Context of shock and Contagion ------------------ 03 1.2 Types of Shock ------------------------------------------------------ 04
1.2.1 Monetary Policy Shock ------------------------------------ 04 1.2.1.1 Theories on Monetary Policy ------------------- 04 1.2.1.2 Transmission of Monetary Policy -------------- 07 1.2.1.3 Monetary Policy Anomalies --------------------- 08
1.2.2 Financial Shock --------------------------------------------- 09
2. Emerging Markets ------------------------------------------------------- 10
2.1 Emerging Markets Classification ------------------------------- 11 3. Gap in the literature ----------------------------------------------------- 14 4. Problem Statement ------------------------------------------------------ 15
5. Objectives ------------------------------------------------------------------ 15
6. Scope of the Study ------------------------------------------------------ 16
7. Structure of the Study--------------------------------------------------- 17
CHAPTER II…………………………………………………………. LITERATURE REVIEW…………………………………………….
18
1. Monetary Policy Shocks------------------------------------------------ 18
2. Financial shocks---------------------------------------------------------- 25
2.1 Literature on Development of FCI------------------------------- 25
2.2 Literature review on FCI Transmission ------------------------ 29
3. Gap in the Literature----------------------------------------------------- 32
CHAPTER III ……………………………………………………….. METHODOLOGY……………………………………………………
33
1. Methodology of the First Objective ---------------------------------- 33
1.1 Hypotheses of First Objective ----------------------------------- 33
xi
1.2 Scope of the Objective -------------------------------------------- 36
1.3 Methodology --------------------------------------------------------- 37
2. Methodology of the Second Objective ------------------------------ 39
2.1 Scope of the Objective -------------------------------------------- 41
2.2 Methodology --------------------------------------------------------- 42
2.2.1 VAR Model --------------------------------------------------- 42
3. Methodology of the Third Objective --------------------------------- 44
3.1 Rational of the Index ----------------------------------------------- 45
3.2 Scope of the Objective -------------------------------------------- 46
3.3 Econometric Method ----------------------------------------------- 47
3.3.1 Multivariate Models of TVP-FAVAR -------------------- 49
4. Conclusion ----------------------------------------------------------------- 50
CHAPTER IV ……………………………………………………….. MONETARY POLICY SHOCK TRANSMISSION IN EMERGING MARKETS…………………………………………….
51
1. Introduction----------------------------------------------------------------- 51
2. Identification of Monetary Policy (MP) Shock---------------------- 54
2.1 Theoretical Framework-------------------------------------------- 55
2.2 Data & Choice of Variable----------------------------------------- 56
3. Results---------------------------------------------------------------------- 59
3.1 Simultaneous Relation--------------------------------------------- 60
3.2 Stochastic Volatility------------------------------------------------- 60
3.3 Impulse Responses------------------------------------------------- 60
3.3.1 Impulse Response of Domestic Markets-------------- 61
3.3.1.1 Brazil ------------------------------------------------- 62
3.3.1.2 Colombia--------------------------------------------- 63
3.3.1.3 Czech Republic------------------------------------- 64
3.3.1.4 Hungary ---------------------------------------------- 65
3.3.1.5 Malaysia --------------------------------------------- 66
3.3.1.6 Mexico ----------------------------------------------- 67
3.3.1.7 Pakistan---------------------------------------------- 68
3.3.1.8 Peru --------------------------------------------------- 69
3.3.1.9 Phillipines ------------------------------------------- 70
3.3.1.10 Poland ---------------------------------------------- 71
3.3.1.11 Russian Federation------------------------------ 72
3.3.1.12 Turkey ---------------------------------------------- 73
xii
3.3.2 Impulse Response of International Contagion------- 74
3.3.2.1 US to Brazil------------------------------------------ 74
3.3.2.2 US to Colombia ------------------------------------ 75
3.3.2.3 US to Czech Republic---------------------------- 76
3.3.2.4 US to Hungary ------------------------------------- 77
3.3.2.5 US to Malaysia ------------------------------------- 78
3.3.2.6 US to Mexico --------------------------------------- 79
3.3.2.7 US to Pakistan-------------------------------------- 80
3.3.2.8 US to Peru ------------------------------------------ 81
3.3.2.9 US to Phillipines ----------------------------------- 82
3.3.2.10 US to Poland -------------------------------------- 83
3.3.2.11 US to Russian Federation---------------------- 84
3.3.2.12 US to Turkey -------------------------------------- 85
4. Conclusion------------------------------------------------------------------ 86
CHAPTER V………………………………………………………… UNITED STATES’ FINANCIAL CONDITIONS AND MACRO-ECONOMY OF EMERGING MARKETS………………………….
88
1. Introduction ---------------------------------------------------------------- 88
2. Identification---------------------------------------------------------------- 91
2.1 Financial Conditions and Forecasts of Macro-Economy-- 91
2.2 Selection of Variables---------------------------------------------- 91
3. Results---------------------------------------------------------------------- 99
3.1 Brazil ------------------------------------------------------------------- 99
3.2 Chile -------------------------------------------------------------------- 101
3.3 Czech Republic ----------------------------------------------------- 103
3.4 Greece ----------------------------------------------------------------- 104
3.5 Hungary --------------------------------------------------------------- 106
3.6 India -------------------------------------------------------------------- 108
3.7 Malaysia --------------------------------------------------------------- 110
3.8 Mexico ----------------------------------------------------------------- 112
3.9 Pakistan---------------------------------------------------------------- 114
3.10 Poland --------------------------------------------------------------- 116
3.11Russian Federation ----------------------------------------------- 118
3.12 South Africa -------------------------------------------------------- 120
4. Conclusion------------------------------------------------------------------ 122
CHAPTER VI ………………………………………………………. Construction of Financial Condition Index for Pakistan……
124
xiii
1. Introduction----------------------------------------------------------------- 124
2. Data-------------------------------------------------------------------------- 126
3. Construction of Financial Condition Index-------------------------- 128
3.1 Estimating financial Position using FCI------------------------ 131
3.1.1 1971-77 Era--------------------------------------------------- 131
3.1.2 1977-88 Era--------------------------------------------------- 132
3.1.3 The Era of Structural Adjustments --------------------- 133
3.1.4 2001 to onwards -------------------------------------------- 134
4. Forecasting of Macro-economic Variables------------------------- 138
4.1 Inflation----------------------------------------------------------------- 138
4.2 Exchange Rate------------------------------------------------------- 139
4.3 Monetary policy (short term interest rate) -------------------- 141
4.4 Gross Domestic Product (GDP) -------------------------------- 142
4.5 Stock Market---------------------------------------------------------- 143
4.6 Forecasting under other Variants of the Model-------------- 153
5. Conclusions---------------------------------------------------------------- 147
CHAPTER VII ………………………………………………………. DISCUSSIONS ……………………………………………………...
148 148
CHAPTER VIII ……………………………………………………… CONCLUSIONS …………………………………………………….
168 168
CHAPTER IX ………………………………………………………. RECOMMENDATIONS ……………………………………………
173
REFERENCES………………………………………………………
179
Appendix IV-A-----------------------------------------------------------------
192
Appendix IV-B----------------------------------------------------------------- 216
Appendix IV-C ---------------------------------------------------------------- 232
xiv
LIST OF TABLES Chapter Page
I-1 Country Classification of FTSE ------------------------------------------ 13
III-1 Hypotheses of First Objective -------------------------------------------- 35
III-2 Hypotheses of Second Objective --------------------------------------- 41
III-3 Hypotheses of Third Objective ------------------------------------------- 44
IV-1 Variables and Transformation-------------------------------------------- 57
IV-2 Results of Transmission of Monetary Policy ------------------------- 87
V-1 Variables and Transformation-------------------------------------------- 93
V-2 Results ------------------------------------------------------------------------- 122
VI-1 Variables and Transformation-------------------------------------------- 127
VI-2 Chronical Exchange rate in Pakistan ----------------------------------- 140
VI-3 Chronical GDP --------------------------------------------------------------- 142
VIII-1 Status of Hypotheses of First Objective ------------------------------- 169
VIII-2 Status of Hypotheses of Second Objective --------------------------- 170
VIII-3 Status of Hypotheses of Third Objective ------------------------------ 172
IX-1 Findings and Recommendations----------------------------------------
173
xv
LIST OF FIGURES Chapter Page
IV-1 Posterior means of time-varying impulse response of
Brazil---------------------------------------------------------------------
62
IV-2 Posterior means of time-varying impulse response of
Colombia----------------------------------------------------------------
63
IV-3 Posterior means of time-varying impulse response of
Czech Republic--------------------------------------------------------
64
IV-4 Posterior means of time-varying impulse response of
Hungary-----------------------------------------------------------------
65
IV-5 Posterior means of time-varying impulse response of
Malaysia-----------------------------------------------------------------
66
IV-6 Posterior means of time-varying impulse response of
Mexico-------------------------------------------------------------------
67
IV-7 Posterior means of time-varying impulse response of
Pakistan-----------------------------------------------------------------
68
IV-8 Posterior means of time-varying impulse response of
Peru---------------------------------------------------------------------
69
IV-9 Posterior means of time-varying impulse response of
Philippines -------------------------------------------------------------
70
IV-10 Posterior means of time-varying impulse response of
Poland ------------------------------------------------------------------
71
xvi
IV-11 Posterior means of time-varying impulse response of
Russian Federation -------------------------------------------------
72
IV-12 Posterior means of time-varying impulse response of
Turkey -------------------------------------------------------------------
73
IV-13 Posterior means of time-varying impulse response from
US to Brazil------------------------------------------------------------
74
IV-14 Posterior means of time-varying impulse response from
US to Colombia-------------------------------------------------------
75
IV-15 Posterior means of time-varying impulse response from
US to Czech Republic----------------------------------------------
76
IV-16 Posterior means of time-varying impulse response from
US to Hungary--------------------------------------------------------
77
IV-17 Posterior means of time-varying impulse response from
US to Malaysia -------------------------------------------------------
78
IV-18 Posterior means of time-varying impulse response from
US to Mexico ---------------------------------------------------------
79
IV-19 Posterior means of time-varying impulse response from
US to Pakistan -------------------------------------------------------
80
IV-20 Posterior means of time-varying impulse response from
US to Peru ------------------------------------------------------------
81
IV-21 Posterior means of time-varying impulse response from
US to Philippines ----------------------------------------------------
82
xvii
IV-22 Posterior means of time-varying impulse response from
US to Poland ---------------------------------------------------------
83
IV-23 Posterior means of time-varying impulse response from
US to Russian Federation-----------------------------------------
84
IV-24 Posterior means of time-varying impulse response from
US to Turkey ---------------------------------------------------------
85
V-1 Transmission to Brazil----------------------------------------------- 100
V-2 Transmission to Chile------------------------------------------------ 102
V-3 Transmission to Czech Republic---------------------------------- 103
V-4 Transmission to Greece--------------------------------------------- 105
V-5 Transmission to Hungary------------------------------------------- 107
V-6 Transmission to India------------------------------------------------ 109
V-7 Transmission to Malaysia------------------------------------------- 111
V-8 Transmission to Mexico -------------------------------------------- 113
V-9 Transmission to Pakistan ------------------------------------------ 115
V-10 Transmission to Poland -------------------------------------------- 117
V-11 Transmission to Russian Federation ---------------------------- 119
V-12 Transmission to South Africa ------------------------------------- 121
VI-1 Factors estimation using TVP-FAVAR-------------------------- 128
VI-2 Factors estimation using FAVAR--------------------------------- 129
VI-3 Factors estimation using FA-TVP-VAR-------------------------
-
130
xviii
VI-4 FCI estimation using TVP-FAVAR-------------------------------- 135
VI-5 FCI estimation using FAVAR-------------------------------------- 136
VI-6 FCI estimation using FA-TVP-VAR------------------------------- 137
VI-7 Forecasted Inflation ---------------------------------- 139
VI-8 Forecasted and Actual Exchange Rate------------------------- 141
VI-9 Forecasted and Actual Discount Rate--------------------------- 142
VI-10 Forecasted and Actual Gross Domestic Product------------- 143
VI-11 Forecasted and Actual Stock Market --------------------------- 143
VI-12 Forecasting using FA-TVP-VAR---------------------------------- 144
VI-13 Forecasting using FAVAR------------------------------------------ 145
VI-14 Forecasting using TVP-FAVAR----------------------------------- 146
1
CHAPTER I
INTRODUCTION
"Finance is, as it were, the stomach of the country, from which all the other
organs take their tone." (William Edward Gladstone, 1858)
This chapter shed light upon conceptual framework. After elaborating
conceptual framework, problem statement and structure of the study are
discussed.
The financial system is the combination of banking, non-banking institutes,
different types of financial markets and regulatory authorities. This has a crucial
part in the market-based economy. This part of the economy utilizes idle
resources for the sake of capital formation by using a wide range of different
financial tools (James, 2007). This function of finance is not the recent
phenomena; the history and economic literature are full of such examples that
have acknowledged the crucial role of finance for economic development. Walter
Bagehot (1873) had long ago accepted the crucial role of finance for the growth
of the economy specifically for England. Some other worth studying work on this
area are of Goldsmith (1969), Mckinnon (1973) and last but not the least King and
Levine (1993) who established the association between financial and economic
development in 80 countries.
A sound economy requires such a financial system that would facilitate the
smooth circulation of funds between different agents of the economy with the help
2
of different instruments. This can only be happening when the economy is
functioning properly. Financial institutes take the risk at times and work in a
dynamic environment. This state of evolution of the financial system is resulting
in deregulation, innovation, and globalization. But this positive side carries its cost
with it.
Financial system working at a global level is in a constant state of evolution
that at a time results in bubble burst, system failure, the crisis that results in the
spiral death of the institute. These failures of financial systems at an economy
level cost very high. Most recent manifestation can be seen in the form of the
recent crisis of mortgage that started as the local problem in the United States
and sooner spread to other regions of the world. This contagion nature of financial
failure costs very high for the many economies around the world. So its very
crucial for the authorities to maintain and work for the stability of the financial
system (Ghani, 2013).
1. Shock and Contagion
Likewise, to earthquakes, countries do face sudden movements. These
types of turbulence in the economy are known as shocks that may result in crises
(Zumbach, et al. 2000). Due to globalization, such a crisis does not have
implication for a single country. They do have an impact on other financial and
economic institutes and at times may create a death spiral. Open economies are
more prone to get affected by these events. When such correlations of economic
events in neighboring and cross-border economies rise exceptionally in crisis
3
times comparative to the links during normal eras, this is known as contagion.
Contagion impact among economies can be spread through diverse networks.
1.1 Historical Context of Shock and Contagion
Financial history is full of examples of the financial crisis. A quick inspection
of economic history recommends that misconduct of money and credit has led to
financial crisis and several explosions over the centuries. This is true in both
advanced and emerging markets.
Prominent events are great depression in 1930s, the savings and loan
debacle of the 1980s, the Continental Illinois Bank and Trust Company in 1984,
stock market crash in 1987, the Wall Street Crash of 1987, the dotcom bubble,
European Exchange Rate Mechanism Attack in 1992, Mexican Peso Collapse in
1994, East Asian Crisis in 1997, Long Term Capital Management (LTCM) crisis
in 1998, Turkey (2001-02), Argentina (2001), Russian collapse in 1998, Brazilian
devaluation in 1999, technological crisis in 2000, East Asian currency crisis in
1997-98, which grew in Thailand then overcame the Indonesia, Korea, and
Malaysia and last but not the least US mortgage crisis 2007 that sooner spread
in other regions of the world and are considered to be the most damaging financial
shifts since the Great Depression of 1930. Due to its intensity and monetary cost
around the world, the 2007 Global Financial Crisis (GFC) has some similarities to
the great depression of the 1930s (Ghani, 2013). One common element that is
found in all these crises is how these spread from one country to others countries
not of a similar nature but of different size and structure around the world.
4
1.2 Types of Shock
This study is an attempt to study the shock on emerging markets. There may
be many types of shock. Specifically, the focus of this study is the following types
of shock:
Monetary policy shock
Financial shock
1.2.1 Monetary Policy Shock
Monetary management is the central activity of almost all central banks.
Economic fluctuations due to monetary policy are known as monetary policy
shock. In order to understand monetary policy shock, we need to know the
science of monetary policy. For this reason, following session will discuss theories
on monetary policy, anomalies of monetary policy and transmission of monetary
policy that ultimately result in shock.
1.2.1.1 Theories on Monetary Policy
Classical economists (Prominently Adam Smith and David Ricardo) wrote
extensively on the money. The quantity theory of money is amongst oldest
surviving doctrines in economics. It describes the relationship between monetary
and real variables. In the classical model, aggregate demand is equal to
aggregate supply. Price and aggregate demand are having inverse relation and
no relationship exists between price and aggregate supply because the real
output is equal to potential output. This is also known as the neutrality of money.
It means the change in money stock does not lead to change in the level of
production, employment or income. Monetary policy thus cannot be an effective
5
means of influencing real macroeconomic variables such as output, employment
or investment (Meenai, 2012).
According to Karl Marx, money has not any use value but It serves as
universal equivalence. It is acknowledged representative of wealth in a capitalist
society. On account of the crisis, he said that money in simple circulation gives
rise to the monetary crisis. It promises to pay are not realized at a large scale, this
will lead to a chain reaction. within the context of an advanced financial system,
the monetary crisis might be caused by breaks in industrial and commercial
transactions. Origin of crises lies in contradictions of capitalist production, which
manifest themselves in persistent tendencies of overaccumulation and
underconsumption. The crisis appears as a monetary phenomenon and money
become scared or worthless. Creditworthiness collapses, financial markets
plummet (Meenai, 2012).
Keynesian Orthodox Model was presented by John Maynard Keynes who
presented it as an attempt to overthrow the conventional wisdom of those who
claimed to be the inheritors of the classical model. He criticized the classical
model on following grounds:
1. Potential output is not equal to actual output; classical economist ignores
the problem of lack of effective demand and existence of under full employment
equilibrium in a capitalist economy.
2. In long run we all are dead; classicist ignores the time of adjustment.
3. The classical model ignores the problem of monopolization.
6
4. Determination of aggregate demand is a complex process and depends
upon many other factors. Classicists have oversimplified it.
5. Real income and profit expectations are more important than interest
rate as a determinant of savings and investments.
According to Keynes money is not neutral. Increasing money supply,
lowering interest rate can influence aggregate production, employment, and
investment. Governments by proactive monetary and fiscal policy can eliminate
unemployment and enable a capitalist economy to achieve full employment
equilibrium. Capitalism is not a self-regulatory system. It needs regulations
(Meenai, 2012).
Modern Neo Classical Monetary Theory or The monetarists, led by Milton
Friedman and Alan Meltzer, emerged as the response of failure of the Keynesian
system and started reaffirming faith in the quantity theory of money. Monetarism
drew inspiration from several empirical studies during the 1970s (Meenai, 2012).
In summary, although money-related concepts existed from the diminishing
of barter system. John Maynard Keynes was the first person that worked on formal
monetary policy and stressed its role in economic stability. According to him, the
central bank by controlling money supply can have an impact on the economy.
Modern Neo Classical monetary overthrew Keynesian theory because according
to the Keynesian systems of monetary management does only work in the short
run, it is not applicable in long run (Meenai 2010). This process does not end here
but it's in continual progression; researchers and economists are working in this
7
area on a continual basis and working on understanding the nature of monetary
policy using theoretical and empirical approach.
1.2.1.2 Transmission of Monetary Policy
“Describes how policy-induced changed in nominal money stock or short-term
nominal interest rate impact real variables such as aggregate output and
employment.” Ireland (2005)
One main task of the central bank is to maintain price stability. Most countries
adopt explicit inflation targeting. Structural structure and list of tools the central
bank uses for achieving its targets are known as monetary policy. The most
central bank uses qualitative policy. Its instruments help in direction-finding
market interest rate and handling interbank liquidity.
Transmission channels are not working disjointedly but jointly intensifying
their outcomes. Working depends upon the current structure of the economy and
financial system (Klacso, 2013). The regime of monetary policy is chosen
according to the structure and current stance of the economy. Regimes offer an
arrangement of monetary policy choices making process and also make it stress-
free to connect decisions to the public. Basic monetary policy positions are:
Nominal anchor
Monetary targeting
Exchange rate targeting
Inflation targeting
It is a general agreement amongst many economists that in short run
monetary policy may have a noteworthy impact on real economy. Indeed, many
findings have confirmed Friedman and Schwartz (1963) that movement that
8
comes in real output by monetary policy actions last for two years (Christiano et.
al 1994). But the problem is how monetary policy exercises its effect, on this
pointless agreement exists among economists. To a huge degree, empirical
examination of results of monetary policy has treated the monetary transmission
method as “Black Box” (Bernanke and Gertler, 1995).
The conservative view is that monetary policymakers use short-term rates
for influencing the cost of capital and consequently sending functions. But
estimation indicates that it has an impact on the long run. This gap led to exploring
other means (Bernanke & Gertler, 1995). From the above discussion, it can be
said that monetary policy is a complex phenomenon. It is not transmitted through
one channel but many.
12.1.3 Monetary Policy Anomalies
Monetary policy does not transmit to all the economies evenly and according
to theories presented earlier. Transmission process varies from economy to
economy and at the point, serious deviations are being observed from standards.
Few major anomalies in the transmission process are as follows:
a. Exchange Rate Overshooting:
The exchange rate overshooting or Dornbusch overshooting hypothesis was
presented by Rudiger Dornbusch in 1976. That time this was considered as the
birth of modern international macroeconomics (Rogoff, 2002). For studying
exchange rate movements under this hypothesis, the hypothetical
macroeconomic framework is built with the purpose of developing a theory that
may be employed for observing mega fluctuations in exchange rate. This model
9
is a hybrid model of Mundell-Fleming in the short run and in the long run having
features of flexible price model with endogenous expectations. According to this
hypothesis, the exchange rate will overshoot when the short run response to a
disturbance is higher than long-run response (Dornbusch, 1976).
b. Prize Puzzle:
According to many standard macroeconomic theories (e.g. IS-LM framework
or Monetarist), the contractionary monetary policy will result in declining prices.
But Sims (1992) using real data empirically proved deviation of these phenomena
and found that contractionary monetary policy is resulting in rising prices. He
named this behavior price as a prize puzzle. After him, many studies confirm this
behavior and found supporting evidence of the prize puzzle (Javid and Munir,
2010).
The preceding session has shed light on the monetary policy on different
aspects. Transmission of monetary policy specifically in turmoil time period to
other sectors of the economy is known as monetary policy transmission. Similar
nature of shocks is being studied in this study and how they have a
macroeconomic impact.
1.2.2 Financial Shock
Movements in financial condition index are named financial shock in this
study. In the second half of the study financial shocks are being studied. In this
study meaning of financial shock is movements in financial condition index (FCI).
FCI is being studied at two levels in this study. At first level transmission to
emerging markets are being studied. For the transmission purpose, FCI
developed by Brave and Butters (2011) is employed. This index is updated on
10
weekly basis on the official website of the federal reserve bank of Chicago titled
National Financial Conditions index. This is a weighted average of 105 indicators
of financial activity of broader coverage of money, debt, equity market and
traditional and shadow banking. It’s a useful indicator for monitoring financial
stability and forecasting purposes. Positive values of this index mean tighter than
average and negative values indicate looser than average conditions adjusted for
economic conditions. At second level a case of emerging market is being taken
and the index is formed for that country. After the construction of FCI, it is being
tested and it impacts upon macroeconomic variables are being studied.
2. Emerging Markets
In 1981 Antonie Van Agtmael; an economist at World Bank’s International
Corporation devised the term ‘Emerging market’ in a conference referring for
those countries, which cannot be defined existing criteria’s such as Asian Tigers
of that time namely, Thailand, South Korea, Taiwan and Hong Kong were playing
a major part in the global economy. So these sort of countries needs a different
classification. Moreover, that time he was working on ‘Third World Equity Fund’
for that he needs an attractive name for the investor’s attention. So Emerging
economies were considered those countries that were in shift from developing to
developed economies. (Serban, Borisov and Dobrea 2012).
2.1 Emerging Markets Classification
In this study emerging markets’ classification of Financial Times Stock
Exchange (FTSE) Rusell is being followed. This Global Equity Index series was
started in 1985 when FT-actuaries world index was formed. This was a joint
11
venture of financial times and institute of actuaries. In the start country
classification was not done, index only used to cover countries.
In the start division between developed and emerging markets were more
based on subjective analysis and the major focus was wealth. A transparent
system was not available. In 2003, a consultant proposed a structural framework
for the classification of markets. This was:
1. Quality of the market (of rules and regulations);
2. Materiality (country needed to be of a significant size for the inclusion);
3. Consistency and Predictability (for this purpose ‘watch list’ was formed that
would serve as a barometer for the promotion and demotion of the country);
4. Cost limitation (consideration of cost while the implementation of any change
while assessing any country);
5. Stability (phase based approach for the country introduction in the list and its
promotion);
6. Market access (liquidity for the international investors).
The outcome of the meeting was available in 2003. These criteria provided
strong support for the assessment of the quality of the market. These rules were
implemented in 2004. FTSE’s formal procedure for the assessment of the markets
is as follows:
1. Quality of market matrix that would work as a benchmark for the market
judgment;
2. Questionnaire for the regulatory bodies; whose response would help in market
assessment; a new FTSE Russell country classification advisory committee was
12
formed who would report FTSE Russell Policy Advisory Board. This committee
would conduct an objective assessment against benchmark;
3. Watchlist for the country was formed;
4. Engagement with market policy was formed;
5. Annual basis judgment system was formed;
6. Clear communication and implementation timetable was schedule for taking
necessary actions.
Since rules formed back in 2003, the transparent country classification
system is being implemented at FTSE global equity indexes. Review of the
country is being done on annual basis (September of each year). As an outcome
of these exercise countries is being classified into developed, advanced emerging
markets, secondary emerging markets and frontier markets. (FTSE Russell 2015)
Evaluation is being done on annual basis, however, in September 2016 it was
decided that FTSE Russell would not change the classification of the countries in
September 2017. Countries classification in September 2016 is given in table II-
1. (FTSE Russell 2016)
13
Table I-1: Country Classification of FTSE (September 2016)
Developed Advanced
Emerging
Countries
Secondary
Emerging
Frontier
Australia
Austria
Belgium/Luxembourg
Canada
Denmark
Finland
France
Germany
Hong Kong
Ireland
Israel
Italy
Japan
Netherlands
New Zealand
Norway
Portugal
Singapore
South Korea
Spain
Sweden
Switzerland
UK
USA
Brazil
Czech Republic
Greece
Hungary
Malaysia
Mexico
Poland
South Africa
Taiwan
Thailand
Turkey
Chile
China
Colombia
Egypt
India
Indonesia
Pakistan
Peru
Philippines
Qatar
Russia
UAE
Bahrain
Bangladesh
Botswana
Bulgaria
Côte d’Ivoire
Croatia
Cyprus
Estonia
Ghana
Jordan
Kenya
Latvia
Lithuania
Macedonia
Malta
Mauritius
Morocco
Nigeria
Oman
Palestine
Romania
Serbia
Slovakia
Slovenia
Sri Lanka
Tunisia
Vietnam
Source: FTSE Russell 2016
14
3. Gap in the Literature
In the light of reviewed literature, it can be stated that studies are available
of monetary shock but studies covering emerging markets specifically national
and international context still are lacking or are in less number. This is an area
where this study fits and will cover.
From the literature, it can be established that FCI does have implications for
the economies but it can also be seen that a great deal of studies is on advanced
economies. There is a lack of studies covering emerging economies. Secondly, a
major focus of the studies is on forecasting economic activity or growth (e.g. GDP)
or interest rates. Other strong variables like exchange rate are not employed.
Furthermore, there is a lack of studies on response analysis of emerging markets
of the financial conditions of the advanced countries. This study adds the literature
on shocks transmission using the SVAR model with bootstrap after bootstrap
method.
As an advanced version of MCI, FCI’s with the passage of time, by using
more variables for the index formation seems a wise strategy for gauging
economic and financial statements and as a policy tool and seems to have
improved the forecasting power (Hatzius, et al. 2010). With its limitations, still, in
an evolving state, it’s serving as a more realistic tool for decision makers and as
a policy tool especially in a time of crisis.
As it is evident, there is a rich literature covering FCI. But little consideration
is giving to the factor of time variation and emerging market. By filling the gap for
emerging market FCI, this study offers a contribution to the literature on emerging
15
markets specifically Pakistan.
Major GAP in the existing literature is as follows:
1. Studies on advanced countries;
2. Lack of studies on transmission of financial conditions of advanced country
and macro-economy of emerging markets;
3. Use of weak proxies;
4. Models based upon the assumption of homoscedasticity;
5. Non-availability of FCI for emerging countries;
6. FCI development using the assumption of homoscedasticity and constant
parameters.
4. Problem Statement
Globalization carries with it increased financial interdependencies among
many countries. Such linkages may result in cascading defaults and failures. This
creates the need to study the linkages among the financial markets and the impact
of these crises on the markets. Understanding the major reasons and the extent
of these crises is very much important for designing regulatory responses that
may defuse cascades before they happen.
5. Objectives
In the light of Gap in the literature and problem statement, this study aims
to work on the following objectives:
1. To gauge the time-varying effect of the monetary policy;
2. To find out the effect of U.S. financial shocks on emerging markets;
3. To develop and test FCI for Pakistan.
16
6. Scope of the Study
Testing for shock and contagion assists stakeholders and watchdogs to
understand how information special effects move to unconnected economies.
During crisis time period, such as in mortgage crisis, a cross-border contagion of
financial conditions may have strong implications for the financial stability. So it is
required to provide timely assessment of the correlation, transmission, and
contagion of policy actions and financial and economic conditions. So that
authorities may develop contingency plans for mitigating negative consequences
of these events. Transmission of volatility and contagion are hot areas of debate
and research due to their strong implications for monetary policy, financial
landscape, risk measurement, capital requirements, asset pricing and economic
assessment.
Current thesis majorly based upon saltwater economics. The term salt
water and fresh water was first employed by Robert E. Hall in 1976 to differentiate
between two major schools of thoughts of economics. Freshwater economists are
believer of free market while saltwater economists are believers of Keynesian
Economics (Gordon 2003).
Moreover, this study is on emerging countries. Advanced emerging
markets and Secondary Emerging as classified by FTSE (Financial Times Stock
Exchange) are part of this study. Quarterly data is being employed and necessary
transformation is done in the light of objective (details are available in relevant
chapter). The time span is also set according to objective and availability of data;
as data availability in case of emerging markets is a real challenge. Moreover, for
17
analysis purpose time series econometrics and Bayesian econometrics is being
employed. With this scope, shock and contagion arising from monetary policy and
financial condition index are being studied in this thesis. Following headings
discuss in length major definitions and classification mechanism in use.
7. Structure of the Study
A rise of the global crisis and sooner its spread to other regions of the world
has highlighted the need to study the shock transmission. With this background,
this study is an attempt to study the shock transmission from different
perspectives. This study is covering a wide range of shock related issue in major
three chapters such as monetary policy shock transmission both at national and
international transmission, financial shock transmission at international level
transmission and in the last an index is developed for an emerging economy and
its transmission on the economy is studied.
Chapter two and chapter three are linked with each other. Chapter two
covers the literature review. This chapter covers major studies available on
transmission and contagion. While chapter three covers methodology employed
in attainment of all three objectives. Chapter four till six discuss results of the
cases of the transmission and contagion followed by discussion, conclusion, and
recommendations chapters respectively.
This chapter has presented background knowledge of the major concepts
of the study. Moreover, it discusses problem statement and how study has been
structured.
18
CHAPTER II
LITERATURE REVIEW
In this chapter, literature will be discussed by following theoretical
background. This chapter will be presenting critical analysis of the available
literature upon monetary and financial shock.
In the recent history of finance, we may found different incidents of financial
stress. But the very little amount of studies is available directed towards
understanding the impact of such crisis on monetary policy transmission and
transmission of a recently developed tool financial condition index and their
impact upon economic activity. This study purposes to link two aspects of the
literature on the national and international impact of:
Monetary policy shocks
Financial shocks
Preceding sections shed light in the light of the literature on both types of
shocks.
1. Monetary Policy Shocks
Monetary policy transmission is an area of study from a number of
decades. This study is an attempt to bridge two major areas of the monetary policy
transmission; preceding literature covers theoretical and empirical literature on
monetary policy transmission. Both kinds of literature have developed and
employed the standard method and time-varying models for the study of the
transmission mechanism.
19
Theoretical literature has tried to develop relationships in different ways.
Sims (1980) criticizing the existing exercises of studying macroeconomy; offer
new field that produced better results. This started a new era in the macro-
economy study and using method developed in 1980, Sims (1992) studied the
monetary policy transmission in the light of existing theories. He found deviation
at points in the response of macroeconomic variables and named that behavior
as prize puzzle. Later in the same study, he also proposed a solution for the
removal of the puzzle. Bernanke and Gertler (1995) developed a model showing
the shock arising from the conditions of the balance sheet that were resulting in
output fluctuations and also found that negative shocks are having a greater
impact than positive shocks. Azariadis and Smith (1998) developed a model
where the economy was able to switch between different regime namely higher
interest rates, worsen balance sheet conditions, weaker banking lending and a
finance free zone with low financial stress. They found that the result of a
response to all the situations of shocks is non-linear in nature.
Allen and Gale (2000) conducted a study for the development of
contagious model. They displayed financial contagion as an equilibrium
phenomenon, aims to provide micro-economic foundation of financial contagion
by focusing upon liquidity preference shocks. Liquidity preference shock
imperfectly correlated across regions.
Ciccarelli and Rebucci (2003) measured contagion by using Bayesian
Time Varying Coefficient Model. Here timing of the contagion was unknown and
heteroskadicity and omitted variables were presented, they modeled cross-
market linkages changing randomly upon simulated and actual data. They applied
20
the framework upon full and limited information set and used for investigating
positive and negative contagion. They found contagious impact of Argentinian
crisis upon Chilean economy.
Rigobon and sack (2004) asserted that with changes in monetary policy
response arising in asset prices is problematic by the endogeneity of policy
decisions and also that interest rate and asset prices do react to many other
variables. For this reason, they developed an estimator based on the
heteroscedasticity that occurs in high-frequency data. They demonstrated that
response to changes in monetary policy of asset prices can be identified based
upon surge in the variance of policy shock on the day of Federal Open Market
Committee (FOMC)’s meeting and Chairman’s monetary policy testimony. They
establish that a rise in short-term interest rate stock prices decline and the yield
curve goes upward but with the passage of time it gets smaller. Results also
indicated that estimations of event-study contain biases that make estimated
effect appear on T-bill yield large and on stock prices smaller.
Macro-economy does have time variation in its impact. With this objective,
Primiceri (2004) developed a model for studying the impact of monetary policy on
growth and inflation. He proved that monetary policy does have time variation in
its transmission.
Gai (2013) developed a contagion model of financial systems by using
network theory. His model captured two important channels of contagion in
financial systems. Glasserman and Young (2013) proposed a framework that
focused upon the network defined by liabilities between financial institutions. They
analyzed the probability of contagion and the expected losses generated by
21
contagion when the joint distribution of shock is given. Elliott et.al (2014) modeled
contagions and cascades of failures among organizations.
Empirical evidence does have mixed results. Goldstein (2005) examined
the impact of the growth slowdown in China and US and its link with the global
financial conditions. He found that growth slowdown in theses countries could
result financial crisis in emerging countries.
Neri and Nobili (2010) studied the transmission of US monetary policy
shock to the Eurozone. They found that international transmission works through
exchange rate, commodity price, short term interest rate and balance sheet and
they found that contractionary monetary policy decreases the value of Euro and
commodity prices that create demand in euro area and result in expansion in euro
area.
Bagliano and Morana (2010) assessed the mechanism of great recession
spillover to advance and emerging countries using FVAR.
Nakajima (2011) conducted a study on Japanese economy for finding the
time-varying impact of the monetary policy. He estimated monetary policy over
three decades using a TVP-VAR model with stochastic volatility. He found
evidence of performance difference that clearly indicating that during three
decades Japanese economy has gone in major structural shift thus highlighting
the strong impact of monetary policy upon macro-economy. Yuksel et.al (2013)
studies the Taylor-type monetary policy rule with TVP specifications on the
Turkish economy. Time-Varying Parameters of the model was estimated using
structural Extended Kalman Filter (EKF). In the light of the consequences, they
claim that changes in the risk preferences of the firms and household need to
22
reflect in monetary policy including market interest rate as the risk attitude of the
households can do this. Moreover, they found that EKF performs better than
standard Kalman Filter for the phenomena under study.
Todorov (2012) conducted a study for finding the international linkages and
transmission of shocks between US and frontier markets. His study focused on
stock market assessment. By using Generalized Auto-regressive Conditional
Heteroscedasticity (GARCH model on daily data from 20 countries, he found the
limited exposure of frontier markets to US shock. Results from a TVP model
indicated that statistically strong impact of US returns on frontier markets.
Kazi et.al (2013) conducted a study for finding the changing transmission
of monetary policy shock in 14 (OECD) nations. By employing TVP-FAVAR
method they studied the 265 variables’ response. They found that US monetary
policy is having a strong negative impact upon growth. They also found that
transmission to growth has increased since the 1980s and size of the impact of
monetary policy shock during turmoil time was higher than normal periods and
kept on changing over time. Shock decreased the share price in many OECD
countries and asset, trade and interest rate channel were the prominent ones for
the shock propagation to the rest of the economy.
Fornari and Stracca (2013) conducted a study for finding the quantitative
impact of financial shocks on financial and real variables. Their study comprised
of a panel of 21 advanced countries from the time period of 1985 till 2011. They
found that financial shocks can be classified from other shock kinds and also that
they do have a strong effect upon macro-economy. It was also found that financial
structure and development is not a strong contributor to the shock transmission.
23
Moreover, it was found, financial shocks have an impact not only in crisis but also
in normal time.
Ghani (2013) conducted a study for finding out macroeconomic impact of
global financial crisis (GFC) upon emerging countries and policy response by
emerging market economies (EMEs). Global financial crisis impacted EMEs with
different intensity. GFC exposed strengths and weaknesses of paradigm of
development in EMEs based upon liberalized capital account and improved
macroeconomic conditions. EMEs are exposed to crisis in the presence of
financial liberalization reforms without adequate regulatory framework and
country specific characteristics also play a role.
Hab et.al, (2014) analyzed two types of contagion namely information
spillover and liquidity risk premium. These both shocks were initiated by US sub
prime segments and have impacted price determination in other markets. They
studied these shocks upon open-ended property funds. They found that in the
beginning, liquidity risk premium plays a role of contagion but with the passage of
time information spillover comes in action. As a result, they confirm that both types
of shocks are main drivers of the contagion across markets.
`Fu & Lio (2015) analyzed the influence of monetary policy on the direction
then spread of investment dynamic adjustment in the china. Marfatia (2015)
studied the influence of monetary policy on yield curve using the Cooley and
Prescott (1976)’s process of time-varying response coefficients. Results indicate
that there exists noteworthy time variation in response to bond rates during the
1989-2008-time period. Yiu et.cl (2010) conducted a study for finding the
relationship between Asian and US stock market by using the principal
24
component method. Results of Asymmetric Conditional Correlation model
indicated that US market is having a contagion impact upon Asian markets.
Barakchian (2015) studied the spillover of US monetary policy upon
Canada using global vector auto regression (GVAR). Kim (2001) concluded that
expansionary monetary policy results positive output in G6 but Bluedorn and
Bowdler (2001) in case of G7 and Scrimgeour (2010) found that positive monetary
policy results in positive short-term interest rate in four countries in America’s.
Cross and Nguyen (2016) studied global oil price shock upon china’s output using
time varying parameter vector auto regression (TVPVAR) model. They found that
impact is small and temporary in nature.
Rogers et.al (2018) assessed the relationship between monetary policy
and macro economy at the time of zero lower bound using the structural vector
auto regression model. They calculated effects of monetary policy shocks upon
expectations.
Furceri (2018) studied the impact of monetary policy shock upon income
inequality. They found that contractionary monetary policy increases income
inequality.
Arias (2019), with the help of SVAR, studied the impact of monetary policy
shocks. They found that with the increase of rate output decrease and they also
found that, during great moderation, policy shocks are contractionary in nature.
In studying the shock transmission to the stock market, Yiu et.cl (2010)
conducted a study for finding the relationship between Asian and US stock market
by using the principal component method. Results of Asymmetric Conditional
Correlation model indicated that US market is having a contagion impact upon
25
Asian markets. Todorov (2012) conducted a study for finding the international
linkages and transmission of shocks between US and frontier markets. His study
focused on stock market assessment. By using Generalized Auto-regressive
Conditional Heteroscedasticity (GARCH model on daily data from 20 countries,
he found the limited exposure of frontier markets to US shock. Results from a TVP
model indicated that statistically strong impact of US returns on frontier markets.
Ehramann and Fratzsher (2009) analyzed the transmission of US Monetary Policy
Shock to global equity market by taking data of 50 economies. They found that a
100bp increase in monetary policy results 2.7 decreases in return on average.
They also found heterogeneity of the transmission and also found that the
economies, which are open and relatively liquid markets are more prone to the
transmission. Markwat et.al (2009) proved that stock market contagion operates
as domino effect. He found that global crashes do not occur all of sudden but are
preceded by local and regional crashes.
2. Financial Shock
In this study, financial shock means changes in financial conditions index
(FCI). In case of transmission from US to emerging markets, index developed by
Brave and Butter (2012) for the US is in use.
2.1 Literature on Development of FCI
Its been long in practice to use a single variable as a policy tool. Scholars
have used different variables for serving this purpose. In this area like Friedman
and Anna (1963) used monetary aggregates for measuring monetary policy
shocks. Sims (1992), and Bernanke and Blinder (1992) employed interest rate,
26
use of quantity of non-borrowed reserves by Christiano and Eichanbaum (1992),
usage of M1 by Fung and Kasumovish (1998), and use of term spread by Oliner
& Rudebusch (1996). Use of single variable with its easiness have many
limitations and in many cases results in different types of puzzles namely
exchange rate, price or interest rate puzzle [Sims (1992); Dornbusch (1976)]. So
due to lack of agreement on true representative as policy tool raises the question
of the validity of this practice.
In this scenario, a composite measure of any sort seems to be an obvious
solution to this problem. Such an effort started with the development of monetary
condition index (MCI). This use of this measure has been in practice in many
central banks like Canada and New Zealand. But after time, questions start raising
on its validity also as both variables changes so quickly, so in this scenario, it is
hard to find out the tight or loose monetary conditions. Some also criticized the
ground that there may be some other strong variables in place of these for policy
tool (Freedman 1994).
Future attempts were made by using more variables for index formation
like Bernanke and Mihov (1998) developed an index for the United States and this
method was applied to Canada by Fung and Yuan. Hatzius (2000) used a large
number of variables for index formation. Stock and Watson (2002) using Dynamic
Factor Models developed an index for forecasting purpose. Bernanke et.al, (2005)
developed Factor-Augmented Vector Auto Regression (FAVAR) models using
two different methods for the effective study of monetary policy and concluded
that this approach results in better results. The concept of time variation is also
27
being studied in index formation. Koop and Korobilis (2014) developed multiple
indexes for the forecasting purpose.
Updated and according to macroeconomic variables, indexes were formed
later on by Guichard and Turner (2008), Goodhard and Hofmann (2001), Gauthier
et.al, (2004) Mayes and Viren (2001) and Swiston (2008) by using VAR developed
an index for the United States.
Beaton (2009) developed two growth based FCIs. One was developed
using a structural vector error correction model (VECM) and other using
macroeconomic modeling approach. They concluded that contractionary financial
conditions do have an impact on the economy. They studied the link between
financial shock and economic growth at zero lower bounds. They created
equivalency of FCI with interest rate. They concluded that tight financial conditions
do impact upon GDP growth till 40 percent.
Hatzius et.al, (2010) explored the connection amongst financial conditions
and economic activity. They studied prevailing practices based on single
indicators and FCI then proposed a new method for FCI development. The
analysis represented the strong predictive power of FCI.
Gomez et.al, (2011) constructed FCI for Colombia using 21 variables with
the help of the PCA method. They evaluated the predictive power of FCI and
found that it performs better than individual variables. They also found that it could
serve as a leading indicator for early warning indicator thus can be used as a
useful indicator for representing financial stability and macro-prudential
supervision. Matheson (2011) construed FCI for the US and Euro area using DFM
techniques and found good forecasting abilities of the FCIs.
28
Nombulelo et.al, (2012) constructed FCI for South Africa. FCI outperforms
than the benchmark in forecasting exercise.
Debuque-Gonzales and Gochoco-Bautista (2013) construed FCI for Asian
countries namely Hong Kong, China, Japan, Korea, Malaysia and Singapore
using PCA by the following the methodology of Hatzius et.al, (2010). They also
constructed regional based FCI. They found the strong predictive ability of FCI
than AR based models.
Angelopolou (2013) constructed two different FCIs for the euro area using
a wide range of the variable. One FCI with monetary policy and second without
monetary policy. Moreover, country-based FCI was also constructed. This
practice was done to see the impact of monetary policy. The indices represented
a true picture of financial conditions of the eurozone since its creation. In this
practice, the symmetric impact of monetary policy was observed.
Erdem and Tsatsaronis (2013) constructed an index for forecasting
purpose. They found that financial factors do have strong implications for GDP
but weak for the inflation.
Koop and Korobilis (2014) developed FCI using FAVAR models with TVP
and stochastic volatility for forecasting purpose. Time variations allowed a change
in weights attached with variables and DMA/DMS method allow the variable
change over time. They concluded that this way of studying macro-economy
produces better results. Areosa and Dutra (2016) constructed an FCI using the
methodology of Brave and Butters (2011) and the same was later applied for the
forecasting purpose in the case of Brazil.
29
Muraru (2015) developed an index for Romania using three different
methods namely weighted average, Principal Component Analysis, and Dynamic
Factor Model. Results indicate that regardless of the method employed index is
working as an instrument capturing a broader picture of the financial situation of
the economy. So it can be employed for forecasting purposes.
Authorities have also created an index for studying the financial conditions
of economies. Hong Kong Monetary (2010) built an index for the Hong Kong and
China for studying the episodes of stress in mentioned economies. Another index
formed by Monetary Authority of Singapore (2009) for studying the economic
conditions of Asian countries (China, Republic of China, Thailand, Taipei,
Philippines, Malaysia, Korea, Republic of Korea, Indonesia, and India).
International Monetary Fund has also constructed Asia based index for studying
the economic conditions of Asian countries.
Use of principal component analysis for the index formation has been the
practice of many. English et. al (2005) by using more than forty variables
estimated index for the US, UK, and Germany. Hatzius et.al (2010) developed an
index for the US by using 45 variables covering all major financial and economic
variables. Brave and Butters (2010)’s index comprises more than 100 variables
and captures a broad horizon of the economic and financial landscape.
2.2 Literature Review on Shock Transmission
This study aims to add on the literature on the impact of financial condition
index on economies. Mortgage crisis 2007 with its significant negative impact on
leading economies has highlighted the need for the better understanding of the
30
link between financial conditions and macro-economy. For this reason, a
considerable work is available from world’s eminent researchers. Brave and
Butters (2011) constructed an FCI for the USA using a large number of variables
and prove that this index is able to forecast short term and medium term economic
activity. Gumata et.al (2012) constructed an index for South Africa and also found
that this index is having strong predictive power in the short run. Hatzius et.al
(2010) also constructed an index for the USA and found that relative predictive
power of the index is unstable and this index performs well in unusual financial
stress time period and this is able to forecast economic conditions especially
during stress time. Gonzales and Bautista (2013) constructed FCI for five Asian
Markets. They concluded that FCI predicts economy more than benchmark AR
models. FCI was helpful in forecasting economy.
Eickmeier et.al, (2011) employed the FCI developed by Hatzius et.al,
(2010) for studying the international transmission during 1971-2009 of financial
shock using the TVP-FAVAR method. They found that positive US financial
shocks do have a positive impact upon the growth of countries under study (US,
Canada, the UK, France, Germany, Italy, Spain, Japan and Australia). Moreover,
they found that transmission has increased since the 1980s, size of shock has
varied over time indicating time variation is shock transmission and changing
financial landscape in last few years in the US are major reasons for the
international transmission specifically during the crisis.
Alessandria and Mumtaz (2017) hypothesized that the links between credit
markets and real economy tighten in a crisis. Balcilara et.al (2016) used a
previously constructed index for finding its ability to forecast the South African
31
economy. They found that the response of economy is non-linear to financial
conditions. While discussing the response of individual variables they found that
among T-bills, output and inflation response of inflation are highest during the
crisis.
Beaton et.al (2009) studies the effect of financial shocks at zero lower
bound like in the current crisis on real activity. They found that impact may be
amplified at higher interest rates during the financial crisis. Opschoor et.al (2014)
studied to find the impact of financial conditions on the stock market by using
Bloomberg FCI. They found that worst financial conditions are associated with
high volatility and correlation between stock return.
Shocks do not arise from the index; shocks arising from policies are well-
studied phenomena and it’s being proven specifically in case of transmission. Kazi
et.al (2013) conducted a study for finding the changing transmission of monetary
policy shock in 14 (OECD) nations. By employing TVP-FAVAR method they
studied the 265 variables’ response. They found that US monetary policy is having
a strong negative impact upon growth. They also found that transmission to
growth has increased since the 1980s and size of the impact of monetary policy
shock during turmoil time was higher than normal periods and kept on changing
over time. Shock decreased the share price in many OECD countries and asset,
trade and interest rate channel were the prominent ones for the shock propagation
to the rest of the economy.
Todorov (2012) conducted a study for finding the international linkages and
transmission of shocks between US and frontier markets. His study focused on
stock market assessment. By using Generalized Auto-regressive Conditional
32
Heteroscedasticity (GARCH model on daily data from 20 countries, he found the
limited exposure of frontier markets to US shock. Results from a TVP model
indicated that statistically strong impact of US returns on frontier markets.
3. Gap in the Literature
In the light of reviewed literature, it can be stated that studies are available
of monetary shock but studies have majorly focused upon advanced economies
whereas crisis do not know about the boundaries. Moreover, crisis has been
studied using static method whereas impact is time varying. In keeping these
issues in front, this study has been done that covers emerging economies using
time varying methods.
Moreover, from literature it can be established that FCI do have implications
for the economies but it can also be seen that a great deal of studies are on
advanced economies. There is a lack of studies covering emerging economies.
Secondly, major focus of the studies is on forecasting economic activity or growth
(e.g. GDP) or interest rates. Other strong variables like exchange rate are not
employed. Furthermore, there is a lack of studies on response analysis of
emerging markets of the financial conditions of the advanced countries. This study
adds the literature on shocks transmission using SVAR model with bootstrap after
bootstrap method.
This chapter discusses literature upon monetary and financial shock. It
discuss in length literature upon the development and transmission mechanism
of the financial shock. After presenting critically analyzed literature, gap in the
literature has been discussed.
33
CHAPTER III
METHODOLOGY
This chapter will discuss about the research methodology used in each
objective. Study consists of three objectives and after setting objectives
hypotheses is being develop for each objective followed by methodology to test
the hypotheses. Objectives of this study are as follows:
1. To gauge the time varying effect of the monetary policy
2. To find out the effect of U.S. financial shocks on emerging markets
3. To develop and test FCI for Pakistan
1. Methodology of the First Objective
Hypotheses are developed in the light of objectives; now onwards-objective
wise methodology is shared.
1.1 Hypotheses of First Objective
This section develops hypotheses for the first objective that is ‘To gauge
the time-varying effect of the monetary policy’. This objective is further break down
into following sub-objectives:
1. To find out the time varying impact of country specific monetary policy
before, after and during crisis time period upon growth and inflation.
2. To find out the time varying impact of US monetary policy shocks in
emerging countries before, after and during crisis time period upon growth
and inflation.
34
In the light of sub-objectives four hypotheses have been developed. First
two hypotheses are addressing country specific shock and last two hypotheses
are addressing case of contagion.
1. A time-varying contractionary monetary policy has a negative impact
on growth.
Weise (1999) on UK data and Atanasova (2003) on US data found that
contractionary and expansionary monetary shocks do have an impact on the
economy. McCallum (1991) found that output responds more to a contractionary
monetary policy. Cover (1992) found that due to contractionary monetary policy
output declines whereas expansionary monetary policy does not have a
significant impact upon output. Similar results were found by Morgan (1993);
Thoma (1994); Rhee and Rich (1995); Kandil (1995); Karras (1996) and Balke
(2000). Economies behave differently in the time of crisis as compared to the
normal time period. While Nakajima (2011) confirmed the time-varying nature of
the Japanese market but Yio (2010) found no strong evidence of time-varying
nature of volatility on the US & Asian markets. So it can be inferred that different
opinion exists on the transmission of monetary policy.
2. Price puzzle exists in monetary policy transmission.
Sims (1992) found prize puzzle in monetary policy transmission while
Hanson (2004) did not found the evidence of the prize puzzle in monetary policy
transmission. So it can be said that the presence of puzzles is proven in the light
of light but the greater agreement does not exist.
35
3. The systematic US monetary policy has a positive impact on the
growth of emerging economies.
International transmission does exist but it is degree vary from county to
country e.g. Todorov (2012) and Kazi et.al (2013).
4. The expansionary monetary in the US creates prize puzzle in emerging
economies.
As stated above monetary policy transmission to the prizes do have mixed
results in the transmission.
Table III-1: Hypotheses of First Objective
S# Hypotheses Test Impulse variable
Response variable
1. A contractionary monetary policy has inverse impact on growth.
TVP-VAR Model on
Normalized variables
Monetary policy
Growth and
inflation
2. Price puzzle exists in monetary policy transmission.
TVP-VAR Model on
Normalized variables
Monetary policy
Growth and
inflation
3. The systematic US monetary policy has positive impact on growth of emerging economies.
TVP-VAR Model on
Normalized variables
Monetary policy
Growth and
inflation
4. The expansionary monetary in US creates prize puzzle in emerging economies.
TVP-VAR Model on
Normalized variables
Monetary policy
Growth and
inflation
Source: Author’s compilation
36
1.2 Scope of the Objective
Time span for the objective is 1995Q1-2012Q2. Countries from Advance
emerging countries are Brazil, Czech Republic, Hungary, Malaysia, Mexico,
Poland, Turkey and countries from Secondary emerging markets are Colombia,
Pakistan, Peru, Philippines, and Russia.
Impulse variable in the study is monetary policy. Monetary Policy
Instruments for the study are Money market rate, Central Bank Policy
Rate/Discount rate, and T-bill rate. Prominent researchers have also used these
variables. E.g. Use of Money market rate by Rosoiu & Rosoiu (2013); use of
Central Bank Policy Rate/Discount rate by Nakajima (2011) & Modenesi & Araujo
(2012); and use of T-bill rate Sims (1992) as a policy tool for the monetary policy.
Response variables are growth and inflation. Proxy for the Growth is Gross
Domestic Product (Real Index); same variable has been used by Bernanke &
Gertler (1995) and proxy for the Inflation is Consumer Price Index (Real) just like
have been employed by the Sims (1992), and Winkelried & Gutierrez (2015).
37
1.3 Methodology
VAR is a method used for forecasting purpose. VAR models are like
simultaneous equations. Here we consider many endogenous variables together
but each variable is explained by its lagged value, usually there is no exogenous
variable in this (Gujarati, 2008).
VAR models have gone through many developments and still, this process
is going on. Cogley and Sargent (2001) was the pioneer who developed VAR
model with time-varying coefficients. This model was criticized by the Stock (2001)
on the ground of the assumptions employed in the study related to the constant
variance of the VAR’s structural shock. In response to this criticism, Cogley and
Sargent (2001) modified their model using Stochastic Volatility. Stochastic
Volatility originally proposed by Black (1976) is having a significant place in TVP-
VAR models. In the line with this, Primiceri (2005) proposed the TVP-VAR model
with time-varying parameters. In the context of Bayesian inference of TVP-VAR
model with stochastic volatility, employment of Markov Chain Monte Carlo
(MCMC) method makes estimations feasible. Bayesian econometrics is based
upon the rules of probability. All the econometric analysis such as parameter
estimation, model comparison, prediction and such activities follow same
probability rules. So we may say that rules of probability are universal in nature
(Koop, 2003).
To work on the stated objectives that is to estimate Monetary Policy shock
in a time-varying nature. To work on this, TVP-VAR model proposed by Nakajima
(2011) is estimated in this study. For illustrating identification of structural shock
38
in TVP-VAR model it is convenient to present non-policy and policy variables by
a k-dimensional vector of variables (yt).
In this case stated two equations earlier can be written in following TVP-
VAR model proposed by Joushi Nakajima (2011) is as follows:
Ayt = F1yt−1 + ⋯+ Fsyt−s + ut, t=s+1,…,n, (Equation III-1)
Where;
yt= (k*1) vector of observed variables,
A, F1, …, Fs= (k*k) matrices of coefficients
ut=structural shock
it is assumed that ut~ N(0, ΣΣ) where
Σ = (
σ1 0 ⋯ 00 ⋱ ⋱ ⋮⋮ ⋱ ⋱ 00 ⋯ 0 σk
)
Simultaneous relations of the structural shock are specified by recursive
identification, assuming that A is lower-triangular,
A = (
1 0 ⋯ 0a21 ⋱ ⋱ ⋮⋮ ⋱ ⋱ 0
ak1 ⋯ ak,k−1 1
)
Equation (3) can be rewritten as the following reduced form VAR model:
yt = ct + B1yt−1 + ⋯+ Bsyt−s + A−1Σεt, εt~N(0, Ik), (Equation III-2)
Where
Bi = A−1Fi for i=1,….,s.
Stacking the elements in the rows of the Bi’s to form β (k2s ∗ 1 vector) and
defining Xt = Ik ⊗ (yt−1′ , …… , yt−s
′ ) where ⊗ denotes the Kronecker product, the
model can be written as
39
yt = Xtβ + A−1Σεt, (Equation III-3)
All the parameters of equation (2) are constant. In order to extend this
model to TVP-VAR, parameters need to be time-varying.
Consider the TVP-VAR model stochastic volatility specified by
yt = Xtβt + A−1Σtεt, t=s+1,….,n, (Equation III-4)
where the coefficients and the parameters are all time-varying. To model
the process of time-varying parameters, Primiceri (2005) is being followed, let at =
(a21, a31, a32, a41, … , ak,k−1)′ be a stacked vector of the lower-triangular elements
in At and ht = (h1t , … , hkt )′ with hjt = log σjt2 , for j=1,…,k, t=s+1,….,n. it is
assumed that the parameters if equation (3) follow a random walk process as
follows:
βt+1 = βt + uβt, at+1 = at + uat, ht+1 = ht + uht,
(
εt
uβt
uat
uht)
~
(
0,(
I 0 0 00 Σβ 0 0
0 0 Σa 00 0 0 Σh
)
)
For t=s+1,…,n, where
βs+1~N(uβo, Σβ0), as+1~N(uao, Σa0) and hs+1~N(uho, Σh0). For more details, refer
Nakajima (2011).
2. Methodology of the Second Objective
This section develops hypotheses for the first objective that is ‘To find-out
the effects of the financial conditions of the United states upon macro-economy
of the emerging economies’.
40
a. There is an exchange rate puzzle like Dornbusch’s exchange rate
overshooting in the transmission mechanism.
Link of the foreign exchange rate with other rates has been an area of
interest of scholar around the world. Such as Sanchez (2005) established a link
between the exchange rate and interest rate in the small open economy. His
results indicated that an increase in interest rate results in the contractionary
foreign exchange rate and vice versa. Sichei (2005) confirmed Dornbusch’s
Hypothesis in the case of South Africa; while Tu & Feng (2009) rejected this
hypothesis in the case of U.S. and Germany.
b. FCI reflects information of stock market in the long run.
Zeng (2010) found the strong response of monetary policy to the monetary shock.
Todorov (2012) found that lagged US stock data don’t have an impact on the frontier
market but expected to have implications for frontier markets.
c. FCI reflects information on short-term interest rate in the short run.
Montagnoli and Napolitano (2005) found that financial condition index is a good
indicator for giving information in the short run. Brave and Butters (2011) found that FCI
can be used for short and medium term forecasting.
d. FCI reflects information on long-term interest rate in the long run.
This chapter covered in-depth literature on the subject under study. After
identifying a gap in the literature, objectives were set that preceded with
hypotheses. Upcoming chapters will deal with individual objective and empirically
will test the hypotheses.
41
Table III-2: Hypotheses of Second Objective
S# Hypotheses Test Impulse variable
Response variable
1. There is exchange rate puzzle like Dornbusch’s exchange rate overshooting in transmission mechanism.
SVAR model on
normalized variables
Financial condition
index
Exchange rate
2. FCI reflects information of stock market in the long run.
SVAR model on normalized variables
Financial condition index
Stock market
3. FCI reflects information of short term interest rate in the short run.
SVAR model on normalized variables
Financial condition index
Short term interest rate
4. FCI reflects information of long term interest rate in the long run.
SVAR model on normalized variables
Financial condition index
Long term interest rate
Source: Author’s compilation
2.1 Scope of the Objective
Countries from the advanced emerging markets are Brazil, Czech
Republic, Hungary, Malaysia, Mexico, Poland, and South Africa and countries
from Secondary emerging markets are Chile, Greece, India, Pakistan, and
Russian Federation. Impulse variable is Financial condition index developed by
Brave & Butters (2011) and response variables are Economic & financial variables
namely Short term Interest rate, Long term interest rate (same variable have been
used by nakajima (2011) in his study), Stock prices (similar variable can be found
in the study of Agha et.al (2005)), Real effective exchange rate (same variable for
the same purpose have been used by Correa & Caetano (2013).
42
2.2 Methodology
To analyze the spread of shock to the economies majorly two methods are
in use namely Structural Macro Models and VAR models. Sims (1980) criticized
the structural models in his seminal work to a great deal and VAR models are
proposed as alternatives. VAR models are in great use by notable researchers for
the study on transmission mechanism (e.g. Brave and Butters (2011) and many
others).
2.2.1 VAR Model
Vector Autoregression Models (VAR) is employed for economic analysis.
In this study the vector, autoregression model proposed by Barsky and Sims
(2012) employed. VAR models have been employed to analyze shocks of
different natures such as Sims and Zha (2006) used VAR model to study the
money impact upon output; Blanchard and Perotti (2002) fiscal policy impact; and
by Gali (1999) to study the relation between technology shocks and worked hours.
VAR models multivariate and linear demonstration of a vector of observables on
its own lags and in other case other variables as constant or trend. In VAR models,
we make explicit identification assumption for isolating estimation of the behavior
under study.
𝑦𝑡 = [𝑦1,𝑡𝑦2,𝑡 …𝑦𝑛,𝑡]′ (Equation III-5)
Where:
𝑦𝑡= Vector with the value of n variables at time t
As reduced form VAR, it can be written as:
43
𝑦𝑡 = 𝐺0 + 𝐺1𝑌𝑡−1 + 𝐺1𝑌𝑡−1+…. +𝐺1𝑌𝑡−1 + 𝜀𝑡 (Equation III-6)
Where:
• 𝐺0=(n*1) vector of constants
• 𝐺1=(n*n) matrix of coefficients
• 𝜀𝑡=(n*1) vector of white noise innovation
• 𝐸[𝑒𝑡] = 0
• 𝐸[𝑒𝑡𝑒𝑡′] = Ω(𝑛𝑜𝑡 𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙)𝑖𝑓 𝑡 = 𝜏 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0
Assumption about error term:
• 𝐸[𝑒𝑡𝑒𝑡′] = Ω for t ≠ τ
In matrix notation:
• 𝑦𝑡 = 𝐺1𝑌𝑡−1 + 𝑒𝑡
𝑖𝑛 𝑡ℎ𝑖𝑠 𝑠𝑡𝑢𝑑𝑦:
𝑦𝑡 =
[ 𝑦1,𝑡
𝑦2,𝑡
𝑦3,𝑡
𝑦4,𝑡
𝑦5,𝑡]
=
[ 𝐹𝐶𝐼𝐼𝑅
𝐺𝐵𝑅𝑅𝑆 ]
(Equation III-7)
IR is the short-term interest rate, GBR is the Government Bond Rate (long-
term interest rate), R is the exchange rate and S represents stocks. VARs are
performed by ordinary least square (OLS) equation by equation. Residuals take
on recursive ordering.
44
3. Methodology of the Third Objective
This section develops hypotheses for the first objective that is ‘To develop
and test FCI for Pakistan’.
a. FCI helps in measuring financial shocks.
Hatzius (2011) found that FCI is a good representative of economic activity.
Koop and Korobilis (2014) and many others found the similar results.
Table III-3: Hypotheses of Third Objective
S # Hypotheses Test Impulse variable
Response variable
1. FCI helps in measuring financial shocks timely
TVP-FAVAR with
its restrictions
on normalized variables
Index 1. Consumer price index
2. Discount rate 3. KSE 100 index 4. Gross domestic
product 5. Real effective
exchange rate
Source: Author’s compilation
45
3.1 Rational of the Index
Financial Condition Index (FCI) may serve for many purposes. For
example, it can be used to find out early signs of bad financial conditions Gomez
et.al, (2011); Muraru (2015) or could serve as a forecaster of the economy
(Nombulelo et.al, (2012); Bautista (2013); Erdem and Tsatsaronis (2013). It is now
in practice of many financial institutes (IMF, Goldman Sachs, and Bloomberg) and
authorities (federal reserve bank of Chicago and many other banks) to develop
FCI for the market watch. Estimation of FCI ranges from simple weighted average
method to developed sophisticated methodology. Keeping in mind the existing
practice of developing the financial index, the chief empirical input in the literature
of this study is to develop an FCI for an emerging market using the most recent
approach.
Development and usage of FCI deal with the variable choice for FCI and
its link with macro-economy. These need to think about changing state. For this
reason, a method of index development by Koop and Korobilis (2014) is utilized.
Indexes are created using a wide range of macroeconomic and financial variables
over a long horizon for Pakistan. They developed a method using extensions of
Factor models and presented multiple forms of the index. The rationale of using
this method is that it is able to capture the time-varying nature of the variables so
can give a better picture of financial conditions.
46
3.2 Scope of the Objective
FCI for the Pakistan from 1969 Q1-2016Q1 using following variables:
1. Equities
2. Gold
3. Import volume
4. Export volume
5. Goods, deflator/unit value of export
6. Goods, deflator/unit value of import
7. Industrial production index
8. Deposit rate
9. Government Bond rate
10. Money market rate
11. Producer price index
12. Total reserve excluding gold & foreign reserves
13. Total consumption
14. T-bill rates
15. KMI-30
16. All Share Index
17. M1 (Currency)
18. National Saving Amount Outstanding
19. Schedule Bank Amount outstanding
20. Total fixed capital formation
47
3.3 Econometric Method
This study employs the methodology of Koop and Korobilis (2014) for the
development of FCI. In their method, the factor model is based on two connecting
equations. Equation one helps in extracting FCI from financial and economic
variables 𝑥𝑡 and second equation deals with interconnectivity of FCI and macro-
economic variables 𝑦𝑡.
TVP-FAVAR model
𝑥𝑡 = 𝜆𝑡𝑦𝑦𝑡 + 𝜆𝑡
𝑓𝑓𝑡 + 𝑢𝑡 (Equation III-8)
[𝑦𝑡
𝑓𝑡] = 𝑐𝑡 + 𝐵𝑡,1 [
𝑦𝑡−1
𝑓𝑡−1] + ⋯…………+ 𝐵𝑡,𝑝 [
𝑦𝑡−𝑝
𝑓𝑡−𝑝] + 𝜀𝑡 (Equation III-9)
Where
𝑥𝑡=An (n*1) vector of financial and economic variables for the construction of FCI
𝑦𝑡=An (s*1) vector of macroeconomic variables [in this empirical work 𝑦𝑡 =
(𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑟 𝑝𝑟𝑖𝑐𝑒 𝑖𝑛𝑑𝑒𝑥, 𝐾𝑆𝐸 100 𝑖𝑛𝑑𝑒𝑥, 𝐷𝑖𝑠𝑐𝑜𝑢𝑛𝑡 𝑅𝑎𝑡𝑒, 𝐺𝑟𝑜𝑠𝑠 𝐷𝑜𝑚𝑒𝑠𝑡𝑖𝑐 𝑃𝑟𝑜𝑑𝑢𝑐𝑡,
𝑅𝑒𝑎𝑙 𝐸𝑓𝑓𝑒𝑐𝑡𝑖𝑣𝑒 𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒 𝑟𝑎𝑡𝑒)′]
𝜆𝑡𝑦=Regression coefficients
𝜆𝑡𝑓=Factor loadings
𝑓𝑡=Latent factor interpreted as FCI
48
𝑐𝑡= Intercept
𝐵𝑡,1, … , 𝐵𝑡,𝑝=VAR coefficients
𝑢𝑡 𝑎𝑛𝑑 𝜀𝑡=Zero-mean Gaussian disturbances with time-varying
covariance𝑉𝑡 𝑎𝑛𝑑 𝑄𝑡. For more details, refer Koop and Korobilis (2014).
49
3.3.1 Multivariate Models of TVP-FAVAR
The complete model defined in equations (one and two) is the TVP-
FAVAR. some limits on the TVP-FAVAR are also considered here that as a result
give other famous multivariate models. Those are:
Factor-augmented time-varying parameter VAR (FA-TVP-VAR):
The specification is gained from the TVP-FAVAR model beneath the limit
that the loadings are constant. In this case, the first equation in the earlier model
defines a normal factor model, whereas the other equation is a TVP-VAR
augmented with the FCI.
Factor-augmented VAR (FAVAR): This model is obtained from the TVP-FAVAR
under the restriction that both𝜆𝑡and 𝛽𝑡are time-invariant.
All presented models are having heteroskedastic covariances.
Algorithm for the calculation of TVP-FAVAR is as follows:
1. Initialization of all the parameters, 𝜆0, 𝛽0, 𝑓0,𝑉0,𝑄0and gaining of the
principal components estimates of the factors, 𝑓��.
2. Estimation of the time-varying parameters𝜃𝑡 given 𝑓��,than estimation of 𝑉𝑡,
𝑄𝑡, 𝑅𝑡, 𝑊𝑡using Variance Discounting and finally estimate 𝜆𝑡and 𝛽𝑡, given
(𝑉𝑡, 𝑄𝑡, 𝑅𝑡, 𝑊𝑡), using the Kalman Filter and Smoother
3. Estimate the factors ft given 𝜃𝑡using the Kalman Filter Smoother (KFS).
50
4. Conclusion
This chapter discuss in length methodology being employed in the
attainment of each objective.
This chapter discusses hypotheses developed in the light of literature
followed by scope and methodology. All three objectives are interconnected yet
differ.
That’s why three different methods have been employed for studying the
objectives of the study.
51
CHAPTER IV
MONETARY POLICY SHOCK TRANSMISSION IN
EMERGING MARKETS
1. Introduction
Central banking has been transformed, in practice and in theory……(Wolf,
2012)
After setting objectives in last chapter, this chapter will be discussing first
objective in length. This chapter will be discussing results of the first objective.
Global financial crisis (2007) started initially as a subprime mortgage
problem in the United States (US). With a high default rate of the subprime
mortgage, economies have suffered disastrous losses in coming years [Fornari
and Stracca (2013); Olmo and Sanso-Navarro (2014)]. In its initial state, this fear
was there that this crisis would spill over to the rest of the world economies [see
e.g. Ciccarelli et.al (2013); wolf (2012)]. With time, many fears come true and
largely negative impact was seen in debt markets, real estate, and bond market
and upon other macroeconomic variables. This is a true case of contagion
indicating how a local problem turns into a global crisis [(see, e.g., Hab (2014);
Wagan and Ali (2014)]. This global response of crisis has initiated a heated debate
among researchers on the causes of the crisis. It is believing of many researchers
and market watchers while discussing this global level response of financial crisis
that monetary policy is actively responsible for the transmission [Fatima (2013)]
and possibly it’s a source of contagion in time of crisis [Kazi et.al (2013)].
52
With this background, this can be said that responsibilities of the
central/federal/state (state bank henceforth) banks are higher than past; they
need to consider the international impact in the form of contagion in their decision-
making process in order to minimize impact at the macroeconomic level. It
believes that this approach of decision-making will result in the better conduct of
Monetary Policy objectives [Blinder (2010); Borio (2009)].
Above discussion indicates that among many of the objectives of the state
bank, one is to work on achieving financial stability hence economic growth. In
order to work in this objective, state bank authorities require accurate measure on
the effect of Monetary Policy upon economy then possible sources of contagion.
Studies available on this line majorly have employed Vector Auto-Regression
(VAR) models for empirical evidence [Luporini (2008); Best (2013); Rodofo
Cermeno and Polo (2012); Aleem and Lahiani (2011); Zakir and Malik (2013);
Phiromswad (2015)]. VAR approach is an econometric method mainly used for
economic analysis. In the line with it, TVP-VAR is a new approach of this area for
studying economic issues. It was proposed by Primiceri (2005) who employed in
t for studying the systematic and non-systematic Monetary Policy of the United
state.
This study reassures the argument of the Primiceri (2005) that the state of
the economy tends to vary over time so is true for the transmission and contagion
process. By permitting parameters to change over an interval, Monetary Policy
shocks may closely be observed. This can be done using TVP-VAR models where
it is assumed that parameters for the low first-order random walk process that
allow them a long-lasting and short-lived shift in parameters and the economic
53
structure can be studied in a flexible and vigorous manner.
Motivated by the influential work of the Primiceri (2005), the objective of
this paper is to present an empirical proof on the Monetary Policy shock in a time-
varying method in emerging markets. Moreover, this study is also having a base
upon the study of Sims (1992) in that he asserts that economists and researchers
have not clear clue about the size and extent of the effects of the Monetary Policy
on aggregate activity. This view is still true in the case of many countries.
Economists have agreed on this part that monetary authorities are capable
enough to control short-term interest rate hence can have an influence on
aggregate activity. There is formal statistical and theoretical evidence on this view
(e.g. (Nakajima, 2011); and many others.) in this study this hypothesis is being
tested at two levels; firstly as contagion arising from United State from Monetary
Policy; secondly transmission at country level from Monetary Policy. Both are
being tested on emerging markets.
Evidence available in this area is majorly in the case of advanced countries
[Jannsen et.al (2015); Marfatia (2015)]. This study is an attempt to fill the gap in
the case of emerging markets. This study addresses the following questions:
1. How enormous is the influence of US Monetary Policy shock on the
aggregate economy of the emerging markets?
2. How big is the effect of the Monetary Policy shock at country level on the
aggregate economy in case of emerging markets?
For studying these questions, I employed TVP-VAR model, proposed by
Nakajima (2011). This model will create an impulse response of the economy
54
arising from the Monetary Policy Shocks. This is being studied by giving
contractionary Monetary Policy Shocks. It is being found that Monetary Policy
shocks at both levels are being transmitted to the aggregate economy both at
national and international level. Extent may vary from country to country but its
true for all the emerging countries.
The rest of the chapter is structured as follows.
Section 2 deals with identification of monetary shock.
Section 3 deals with the results section.
Section 4 with conclusion.
2. Identification of Monetary Policy (MP) Shock
In literature, there is no consensus on the identification problem of
exogenous Monetary Policy shock from endogenous components of the Monetary
Policy. In this study, the identification strategy proposed by Bernanke and Mihov
(1998) is employed. In light of this strategy, it is assumed that some good single
measure for Monetary Policy is available. In this scenario, the “true” structure of the
economy can be modeled as follows:
𝑍𝑡 = ∑ 𝐵𝑖𝑍𝑡−𝑖 +𝑘𝑖=0 ∑ 𝐶𝑖𝑝𝑡−𝑖 +𝑘
𝑖=0 𝐴𝑦𝑣𝑡𝑦 (Equation IV-1)
𝑝𝑡 = ∑ 𝐷𝑖𝑍𝑡−1 +𝑘𝑖=0 ∑ 𝑔𝑖𝑝𝑡−𝑖 +𝑘
𝑖=1 𝑣𝑡𝑝 (Equation IV-2)
Where
𝑍𝑡=Vector of non-policy macroeconomic variables
𝑝𝑡=Variable indicating the Monetary Policy stance.
55
𝐵𝑖 ,𝐶𝑖 ,𝑔𝑖, 𝐷𝑖=vector/matrices of coefficients
𝑣𝑡𝑝, 𝑣𝑡
𝑦=Structural shock
𝐴𝑦=General matrix (following Bernanke (1986), in eq. 1 structural shock is
multiplied by general matrix, so that shock may enter in more than one equation).
Monetary policy shock is defined as the unexpected change in the short-
term interest rate of the central banks. The inspiration to use short-term interest
rate as a proxy of Monetary Policy comes from Sims (1992) who have used for the
study of Monetary Policy transmission mechanism in the USA. Apart from him,
many prominent researchers have used short-term interest rate as a proxy of
Monetary Policy namely Nakajima (2011); Primiceri (2005) and Bernanke and
Blinder (1992) among many.
2.1 Theoretical Framework
Majorly it was in the 50s and 60s when standards emerge those emphases
on the role of Monetary Policy in the economic structure. The Keynesian and
Monetarist school of thought in length have discussed role of Monetary Policy. As
a result t, ISLM framework has emerged. In this framework, it is assumed that any
innovation in Monetary Policy does have an impact on the economy. In this study
as it is earlier identified that short-term interest rate will represent Monetary Policy
Shocks. Under monetarists and ISLM, explanation monetary contraction will
create declining output and monetary contraction is deflation. If these responses
were not created, then it would know as puzzle e.g. price puzzle.
56
2.2 Data & Choice of Variable
For studying the Monetary Policy shock variables include the gross domestic
product (GDP) and inflation. The time span for the study is 1995Q1–2012Q2. All
series are downloaded from the website of the International monetary fund (IMF),
FRED — St. Louis Fed, Bank for International Settlements (BIS) accounts and
State Bank of Pakistan. For finding stationary in series Phillips Perron (2001) test
is employed. The data consists of quarterly variables for the US and emerging
countries namely advanced emerging markets and Secondary Emerging as
classified by FTSE (Financial Times Stock Exchange). Detail on variable and
transformation is given in table IV-1.
57
Table IV-1: Variables and Transformation
Country Name
Variable Name Transformation Source
Brazil Money Market Rate Log difference IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Natural Logarithms IMF
Colombia Discount Rate Log difference IMF
Gross Domestic Product Log difference IMF
Consumer Price Index Natural Logarithms IMF
Czech Republic Money Market Rate Log difference IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Log difference IMF
Hungary Discount Rate Log difference IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Natural Logarithms IMF
Malaysia Money Market Rate Log difference IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Log difference IMF
Pakistan Money Market Rate Log difference IMF
Gross Domestic Product Natural Logarithms SBP-
Paper
(2013) &
Arby
(2008)
Consumer Price Index Log difference IMF
Mexico T-Bill Rate Log difference IMF
Gross Domestic Product Log difference IMF
Consumer Price Index Natural Logarithms IMF
Peru Discount Rate Natural Logarithms IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Log difference IMF
58
Philippines Money Market Rate Natural Logarithms IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Log difference IMF
Poland Money Market Rate Log difference IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Natural Logarithms IMF
Russian
Federation
Money Market Rate Natural Logarithms IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Log difference IMF
South Africa Central Bank-Policy
Rate
Log difference IMF
Gross Domestic Product Log difference IMF
Consumer Price Index Log difference IMF
Thailand Money Market Rate Log difference IMF
Gross Domestic Product Log difference IMF
Consumer Price Index Log difference IMF
Turkey Discount Rate Log difference IMF
Gross Domestic Product Natural Logarithms IMF
Consumer Price Index Log difference IMF
Source: Author’s compilation
As a proxy of the growth, we have used GDP (gross domestic product, Real,
Seasonally Adjusted-index in units), downloaded from IMF except in case of
Pakistan, for Pakistan data on GDP is taken from a paper by Hanif et. al (2013)
who have quartered national accounts of Pakistan. Consumer Price Index (CPI)
(index in units of the base year 2010) as a proxy of inflation taken from IMF for all
the countries.
59
3. RESULTS
This section covers the results of TVP-VAR model for the emerging
markets. Model is based on three variables and data frequency is quarterly.
Number of lags is four and it is assumed that Σβ is a diagonal matrix. The following
priors are assumed for the i-th diagonals of the covariance matrices:
(Σβ)i−2 ~ Gamma (20, 0.01)
(Σa)i−2 ~ Gamma (4, 0.01)
(Σh)i−2 ~ Gamma (4, 0.01)
For the initial state of the time-varying parameter, rather flat priors are set;
μβ0=μa0 = μh0=0, and Σβ0=Σa0 = Σh0=10*1. For computing the posterior
estimates, M=10,000 are sampled from where 1,000 are discarded. The results
in table and figure show that the MCMC algorithm produces posterior draws
efficiently (for details refer appendix IV-A).
Table and Figure report the estimation results for the selected parameters
of the TVP-VAR model for the variable set of the model for each country. The
results in table and figure show that the MCMC algorithm produces posterior
draws efficiently. Part one indicates the results of sample autocorrelations,
sample paths in part two, and posterior densities in part three.
This is done at two levels:
1. At the domestic level
2. At the international level
60
3.1 Simultaneous Relation
One of the characteristics of TVP-VAR model is a time-varying
simultaneous relation. Simultaneous relation of all the variables stays constant
over time (for details refer appendix IV-B).
This is also divided into two parts
3. At the domestic level
4. At the international level
3.2 Stochastic Volatility
The stochastic volatility of inflation and output exhibit a stable trend in all
variables. (For details refer appendix IV-C).
3.3 Impulse Responses
The impulse response is used to observe the dynamics of the model. for
the TVP-VAR model, the responses are computed at all points in time using the
estimated time-varying parameters. The impulse response is generated in the
following way:
Before the crisis at 1998 Q4-represented by the dotted line
During crisis at 2008 Q4-represented by dash line
After crisis 2012 Q2-represented by the solid line
This is done on the following:
Impact of Domestic Monetary Policy upon economic variables
Impact of US. Monetary policy upon economic variables
61
3.3.1 Impulse Response of Domestic Markets
The responses are drawn in a time- series manner by showing the size of
the impulses for before crisis, during the crisis and after the crisis over time.
Impulse response before the crisis is somewhat different but during the crisis and
after crisis response is the same. Impulse response of prices is shown in graph
title 𝜀𝑖 ↑→ 𝑝. Impulse response of the inflation is visible in graph title 𝜀𝑖 ↑→ 𝑥.
62
3.3.1.1 Brazil
Figure 1 shows the impulse responses of the time-varying responses for
the TVP-VAR model. The impulse responses of output to a positive interest rate
shock (𝜀𝑖 ↑→ 𝑥) stay negative and volatile most of the time and impulse response
of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is also negative and of varying
nature. After the crisis response goes down after initial positive response while
response before the shock initially goes negative followed by a positive response
that stays negative afterward. In summary, the response of both variables is
negative most of the time.
Fig.IV-1: Posterior means of the time-varying impulse response of Brazil
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
63
3.3.1.2 Colombia
Figure 2 is showing the impulse response of the Colombian economy. The
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) is followed
by a negative in the start that shows positive followed by the stable response of
zero boundaries. Before the crisis, response shows a heap sort of response in the
medium term that soon dies off. The response of price (𝜀𝑖 ↑→ 𝑝) is volatile in
nature and negative most of the time. Before the crisis response shows a positive
response but of volatile.
Fig.IV-2: Posterior means of the time-varying impulse response of Colombia
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
64
3.3.1.3 Czech Republic
Figure 3 is showing the impulse response of Czech-Republican economy.
The impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) shows
the highest volatility but the response is positive most of the time. The response
of price (𝜀𝑖 ↑→ 𝑝) is also positive but less volatile as compared to output. In
summary, the response of both variables is positive and volatile.
Fig.IV-3: Posterior means of the TVP impulse response of Czech Republic
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
65
3.3.1.4 Hungary
Figure 4 is showing the impulse response of the Hungarian economy. The
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
negative and volatile most of the time and impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is positive and less volatile. The response of
aftershock is of a stable nature.
Fig.IV-4: Posterior means of the time-varying impulse response of Hungary
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
66
3.3.1.5 Malaysia
Figure 5 is showing the impulse response of the Malaysian economy. The
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
negative and volatile most of the time and impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is mixed and of varying nature.
Fig.IV-5: Posterior means of the time-varying impulse response of Malaysia
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
67
3.3.1.6 Mexico
Figure 6 is showing impulse response of Mexican economy. The impulse
responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay negative and
volatile most of the time. The response of before shock is positive compared to
other shocks. Same behavior in case of price (𝜀𝑖 ↑→ 𝑝) is observed, volatile and
negative and positive response of before of the crisis.
Fig.IV-6: Posterior means of the time-varying impulse response of Mexico
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
68
3.3.1.7 Pakistan
Figure 7 is showing impulse response of Pakistani economy. The impulse
responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay positive and
volatile most of the time and impulse response of inflation to positive interest rate
shock (𝜀𝑖 ↑→ 𝑝) is also positive and of varying nature; this response is less volatile
compared to price.
Fig.IV-7: Posterior means of the time-varying impulse response of Pakistan
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
69
3.3.1.8 Peru
Figure 8 is showing the impulse response of Peru economy. The impulse
responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) is mixed and volatile
most of the time and impulse response of inflation to positive interest rate shock
(𝜀𝑖 ↑→ 𝑝) is volatile in the short run followed by a stable response; shock in during
crisis remains positive all the time while before and after crisis response goes
towards stability after initial negative response.
Fig.IV-8: Posterior means of the time-varying impulse response of Peru
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
70
3.3.1.9 Philippines
Figure 9 is showing the impulse response of the Philippines economy. The
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
negative and volatile most of the time and impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is negative in short run while positive in long run and
of stable nature. Before crisis response after initial positive response turns in a
negative response and is of a stable nature.
Fig.IV-9: Posterior means of the time-varying impulse response of
Philippines
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
71
3.3.1.10 Poland
Figure 10 is showing the impulse response of Poland economy. The
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) is mixed
and volatile most of the time and impulse response of inflation to positive interest
rate shock (𝜀𝑖 ↑→ 𝑝) is also positive and of varying nature.
Fig.IV-10: Posterior means of the time-varying impulse response of Poland
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
72
3.3.1.11 Russian Federation
Figure 11 is showing the impulse response of the Russian economy. The
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
negative and volatile most of the time and impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is also negative in short run followed by positive and
this response is of varying nature.
Fig.IV-11: Posterior means of the time-varying impulse response of Russian
Federation
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
73
3.3.1.12 Turkey
Figure 12 is showing the impulse response of Turkey economy. The
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
negative and volatile most of the time and impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is mixed and of varying nature.
Fig.IV-12: Posterior means of the time-varying impulse response of Turkey
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
74
3.3.2 Impulse Response of International Contagion
3.3.2.1 US to Brazil
Figure 13 is showing the impulse response of the Brazilian economy to US
Monetary Policy. The impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) is positive and volatile most of the time; the response of before crisis is
more volatile than other responses and impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is positive and of stable nature. Before the crisis
response is initially positive followed by a negative response after an initial
positive response that turns into a positive response that’s stable in nature. During
and after response initially negative than remain positive afterward.
Fig.IV-13: Posterior means of time-varying impulse response from the US to
Brazil
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
75
3.3.2.2 US to Colombia
Figure 14 is showing the impulse response of the Colombian economy to
US Monetary Policy. The impulse responses of output to a positive interest rate
shock (𝜀𝑖 ↑→ 𝑥) is positive and near zero; the response of before response is more
volatile compared to other responses. The impulse response of inflation to
positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is positive but of varying nature and same
like output, response of price before crisis is more volatile.
Fig.IV-14: Posterior means of time-varying impulse response US to Colombia
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
76
3.3.2.3 US to Czech Republic
Figure 15 is showing the impulse response of the Czech Republican
economy to US Monetary Policy. The impulse responses of output to a positive
interest rate shock (𝜀𝑖 ↑→ 𝑥) is positive and volatile most of the time and impulse
response of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is also positive and
of varying nature.
Fig.IV-15: Posterior means of time-varying impulse response from the US to the
Czech Republic
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
77
3.3.2.4 US to Hungary
Figure 16 is showing the impulse response of the Hungarian economy to
US Monetary Policy. The impulse responses of output to a positive interest rate
shock (𝜀𝑖 ↑→ 𝑥) stay positive and volatile most of the time and impulse response
of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is also positive and of varying
nature.
Fig.IV-16: Posterior means of time-varying impulse response from the US to
Hungary
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
78
3.3.2.5 US to Malaysia
Figure 17 is showing the impulse response of the Malaysian economy to
US Monetary Policy. The impulse responses of output to a positive interest rate
shock (𝜀𝑖 ↑→ 𝑥) stay positive and volatile most of the time and impulse response
of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is also positive but of stable
nature; the response of before crisis is somewhat volatile.
Fig.IV-17: Posterior means of time-varying impulse response from the US to
Malaysia
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
79
3.3.2.6 US to Mexico
Figure 18 is showing the impulse response of the Mexican economy to the
US Monetary Policy. The impulse responses of output to a positive interest rate
shock (𝜀𝑖 ↑→ 𝑥) stay positive and volatile most of the time and impulse response
of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is negative and of varying
nature but less as compared to output.
Fig.IV-18: Posterior means of time-varying impulse response from the US to
Mexico
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
80
3.3.2.7 US to Pakistan
Figure 19 is showing the impulse response of Pakistani economy to US
Monetary Policy. The impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) stay positive and volatile most of the time and impulse response of
inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is mixed and of varying nature.
Fig.IV-19: Posterior means of time-varying impulse response from the US
to Pakistan
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
81
3.3.2.8 US to Peru
Figure 20 is showing the impulse response of Peru economy to US Monetary
Policy. The impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
negative and volatile most of the time and impulse response of inflation to positive interest
rate shock (𝜀𝑖 ↑→ 𝑝) is also negative but of stable nature; during crisis response shows
a positive response followed by a stable negative response.
Fig.IV-20: Posterior means of time-varying impulse response from the US
to Peru
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
82
3.3.2.9 US to Philippines
Figure 21 is showing the impulse response of Philippines economy to US
Monetary Policy. The impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) is positive and volatile most of the time and impulse response of inflation
to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is positive and of varying nature; the
response of before crisis is more volatile compared to other responses.
Fig.IV-21: Posterior means of time-varying impulse response from the US to
the Philippines
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
83
3.3.2.10 US to Poland
Figure 22 is showing the impulse response of Poland economy to US
Monetary Policy. The impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) is positive and volatile most of the time; before crisis response shows
highest volatility and impulse response of inflation to positive interest rate shock
(𝜀𝑖 ↑→ 𝑝) is positive but of stable nature.
Fig.IV-22: Posterior means of time-varying impulse response from the US
to Poland
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
84
3.3.2.11 US to Russian Federation
Figure 23 is showing the impulse response of the Russian economy to the
US Monetary Policy. The impulse responses of output to a positive interest rate
shock (𝜀𝑖 ↑→ 𝑥) is positive and volatile most of the time and impulse response of
inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is negative initially that is followed
by near zero response of stable nature.
Fig.IV-23: Posterior means of time-varying impulse response from the US
to Russian Federation
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
85
3.3.2.12 US to Turkey
Figure 24 is showing the impulse response of the Turk economy to US
Monetary Policy. The impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) stay positive and volatile most of the time and impulse response of
inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is also positive but of stable
nature. The response is near zero.
Fig.IV-24: Posterior means of time-varying impulse response from the US
to Turkey
Note: Time-varying responses for Before crisis at 1998 Q4 (Dotted Line),
During Crisis at 2008 Q4 (Dashed Line), and after crisis at 2012 Q1 (Solid Line)
horizons for the TVP-VAR model where impulse variable is monetary policy (εi)
and output (x) and inflation (p) are response variables.
86
4. Conclusion
The key driver of this study is to identify the impact of Monetary Policy upon
major economic variables of emerging markets. For this purpose, the predictive
power of Monetary Policy is tested upon major macroeconomic variables namely
Growth and prices (inflation). Impulse response affirms the predictive power of
the MP of macroeconomic movements in emerging markets.
Under the monetarist/ISLM, framework interest rate surprises represent
Monetary Policy shock. This states that monetary contraction (positive Monetary
Policy shock) creates declining output and increasing prices. From the results it
can be seen that diversity of the result is visible under the ISLM framework.
Results has been summarized in table IV-2.
87
Table IV-2: Results of Transmission of Monetary Policy
National Monetary Policy International Monetary Policy
Country Output Price Puzzle
ISLM Output Price Puzzle
ISLM
Advanced Emerging Countries
Brazil Yes Yes Partly
Applicable No No
Partly Applicable
Czech Republic
No No Partly
Applicable No No
Partly Applicable
Hungary Yes No Fully
Applicable No No
Partly Applicable
Malaysia Yes No Fully
Applicable No No
Partly Applicable
Mexico Yes Yes Partly Applicable
No Yes Not Applicable
Poland No No Partly Applicable
No No Partly Applicable
Turkey Yes No Fully Applicable
No Yes Not Applicable
Secondary Emerging Countries
Colombia No Yes Partly Applicable
No No Partly Applicable
Pakistan No No Partly Applicable
No No Partly Applicable
Peru No No Partly Applicable
Yes Yes Partly Applicable
Philippines Yes Yes Partly Applicable
No No Partly Applicable
Russia Yes Yes Partly Applicable
No Yes Partly Applicable
Source: Author’s compilation
This chapter discusses monetary policy shock in length. This chapter also
present results at national monetary policy and contagion arising from US
monetary policy towards emerging economies.
88
CHAPTER V
UNITED STATES’ FINANCIAL CONDITIONS AND
MACRO-ECONOMY OF EMERGING MARKETS
1. Introduction
“when the US sneezes, the rest of the world catches a cold” Barachian
(2015).
This chapter is in continuation of last chapter. This chapter will be
discussing second objective in length.
There is mounting evidence that economies around the globe are more
integrated especially in the last few years. Financial development and
liberalization is a major reason for this integration. Due to financial integration, we
may observe co-movement among variables at international level; extent may
vary but its true for many countries especially at the macroeconomic level. This
co-movement is also true in the time of turmoil such as the collapse of housing
price resulted not the only collapse of investment domestically but also a strong
wave of crisis not in nearby countries but also far away countries were also
observed. One possible reason that seems at the front is that transmission of
shock results in contagion effects (Kazi et.al 2013).
It’s a well-established believe of many researchers and economists that
key reason behind this contagion is the active role of U.S. policies e.g. Ehramann
and Fratzscher (2009); Carmassi et.al (2009). One more strong reflection of crisis
is that in the presence of financial innovations, its difficult to capture financial
89
horizon with the help of a small number of variables as traditionally being
applied. Policymakers, researchers, regulators and other market watchers have
all acknowledged the significance of the interconnectivity of traditional and newly
developed financial markets and its link with the economy. Thus it can be said
that monitoring of financial stability requires an understanding of the evolving
paradigm of the financial world and its link with the economy (Brave and Butters
(2010). After the initial hit of the crisis in advanced countries, financial turmoil hit
to emerging economies; where stock market, exchange rate and sovereign debt
all came under pressure [(Balakrishnan et.al, 2009); (Calvo et.al, 2008) and many
others].
The question of whether financial conditions predict economic activity has
a long history in economics. The conclusion by Stock and Watson (2003) “some
asset prices predict inflation or output growth in some countries in some periods”
epitomizes the common view among econometricians that financial indicators are
too noisy and erratic to be exploited for macroeconomic forecasting. Yet
macroeconomists have got to the conclusion that financial shocks are a vital
source of business cycles [(Balcilara et.al 2016); (Alessandria and Mumtaz 2017);
(Opschoor 2014); (Jermann and Quadrini 2012)]. This suggests that financial
information should be utilized in the right conditions for the right prediction of
macroeconomic fluctuations. In order to work on this objective, most of the
empirical evidence is based upon vector auto-regressions (VAR) models (Potts
and Yerger, (2010); Owyang and Wall (2009); Boivin and Giannoni (2006) among
many). It is being observed that little attention is a pain on the question to what
extent financial conditions of advanced countries have an impact upon macro-
90
economy of the emerging economies.
The identified gap in the literature is filled by studying the transmission of
U.S. financial conditions in emerging markets. Having strong implications for the
emerging markets, this thesis is an effort to offer an empirical assessment of the
outcomes of the financial conditions of the United States upon macro-economy of
the emerging economies. To work on this objective, the paper investigates
whether brave and Butters (2011)’s FCI has an influence on the exchange rate,
interest rates, and stock markets.
Barsky and Sims (2012) use Structural Vector Auto Regression (SVAR)
method to study the response to confidence innovations of news, animal spirit and
pure noise in the New Keynesian framework. In the same way, the SVAR model
is used to examine the impact of United States’ financial conditions upon macro-
economy of emerging economies. Barsky and Sims (2012) model use confidence
innovations’ shock to other variables. Whereas, in this study, FCI is used and
response of macroeconomic variables is studies. The core drive of employing U.S.
financial condition jolts is to recognize the predictive power of financial conditions
of major economic variables of emerging markets. For this purpose, the predictive
power of financial conditions is tested upon major macroeconomic variables (short
term and long term interest rate, exchange rate and stock markets). Impulse
response affirms the predictive power of the financial conditions of
macroeconomic movements in emerging markets. This finding echoes the
findings of Brave and Butters (2011) who established that it is likely to use a
financial condition index to increase upon predictions of events of economic
activity over short and medium forecast horizons. Similar findings can be seen in
91
the study by Koop and Korobilis (2014), Hatzius et.al (2010), Debuque and
Bautista (2013) and many others.
This study offers following contribution to the literature. First, most of the
studies on the transmission of financial conditions are available in advanced
countries such as Hatzius (2010), Brave and Butter (2011) and many others. This
study covers emerging countries for studying the possible contagion impact
arising from advanced countries. Secondly, previous studies have majorly
focused upon aggregate activity for studying the macroeconomic impact, in this
study other major macroeconomic variables are being used for the assessment
of contagion impact.
This section is followed by segment 3 identification, segment 3 results and
finally by segment 4 discussion and conclusion.
2. Identification
2.1 Financial Conditions and Forecasts of Macro-Economy
One of the objectives of this thesis is to discover the response of
macroeconomic variables arising from a surprise move in united states’ financial
conditions. It is established in the literature that to work on such objective it is
required to run a VAR model so in the same way a VAR model is developed
comprising FCI and macroeconomic variables and for considering the partial
derivatives of macroeconomic conditions at different horizons with respect to
innovations in Financial conditions. This can be taken like a generalized impulse
response function in Pesaran and Shin (1998). Here orthogonalize is the only
shock means FCI not ordering of macroeconomic variables. It is orthogonalized
92
first in the model.
Furthermore, impulse response analysis is done. It traces the effects of
structural shocks on endogenous variables. By impulse response function we may
see the mechanism by which shock spread over time. For finding impulse
response we have employed Kilian (1998) bootstrap after bootstrap method. For
this method VAR is estimated using OLS and 1000 draws for impulse response
are generated for bootstrap, then the bias-corrected estimator is calculated that
later on is employed for generating 2000 new draws using bootstrap.
2.2 Selection of Variables
Transmission channels are not working separately but their mutual effect
gets amplified, moreover, this depends upon the state of the financial system and
economy (Klacso 2013). Mapping link between the economy and the financial
system has great significance and urgency, especially after the crisis. Majority of
econometric models for forecasting majorly have used interest rate. This strategy
may work in the normal time period but in crisis time period, with this single
variable, we are unable to seize all the connections between financial structure
and economy. For this reason, many authors have suggested using an index
indicating a financial condition for the study of the transmission mechanism.
The aim of the study is to study the impulse response of emerging markets
of the financial market of the United States. In order to work on it, Vector
autoregressive (VAR) systems are estimated for data from Brazil, Chile, Czech
Republic, Greece, Hungary, India, Malaysia, Pakistan, Mexico, Poland, Russian
Federation, and South Africa. Data is taken from the International Financial
Statistics (IFS) database and from other official sources. detailed descriptions
93
about the data source, time frame and transformation is given in the following
table.
Table V-1: Variable and Transformation
Country Name and
Time Span
Variable Name
indicator Transformation Source
US Chicago Fed National Financial Conditions Index, Index, Weekly, Not Seasonally Adjusted
index (weekly)
Percentage Change
FRED
Brazil 1993Q1-2016Q2
Money Market Rate
Percent per Annum
Level IMF
Interest Rates, Savings Rate
Percent per annum
Level IMF
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Chile 2004Q3-2016Q1
Discount Rate
Percent per Annum
Logarithm difference
IMF
Savings Rate
Percent per Annum
Logarithm difference
IMF
The total share price
Logarithm difference
FRED
94
for All shares
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Czech republic 2000Q2-2016Q1
Money Market Rate
Percent per Annum
Logarithm difference
IMF
Government Bonds
Percent per Annum
Logarithm difference
IMF
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Greece 1997Q3-2016Q4
T-bill rate Percent per Annum
Level IMF
Government Bonds
Percent per Annum
Logarithm difference
IMF
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Hungary 2001Q1-2016Q1
Discount Rate
Percent per Annum
Level IMF
Government Bonds
Percent per Annum
Logarithm difference
IMF
95
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
India 2005Q1-2015Q3
Discount rate
Percent per Annum
Logarithm difference
IMF
Government Bonds
Percent per Annum
Logarithm difference
IMF
The total share price for All shares
Logarithm difference
FRED
BIS effective exchange rate-Real (CPI-based), Broad Indices
Monthly averages; 2010=100
Logarithm difference
BIS
Malaysia 2000Q2-2016Q3
Money Market Rate
Percent per Annum
Logarithm difference
IMF
Government Bonds
Percent per Annum
Level IMF
KLCI Bursa Malaysia
Logarithm difference
Finance.Yahoo
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Pakistan 1997Q3-2016Q1
Money Market Rate
Percent per Annum
Logarithm difference
IMF
96
Government Bonds
Percent per Annum
Logarithm difference
IMF
KSE 100 index
Logarithm difference
SBP
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Mexico 1995Q1-2016Q2
T-Bill Rate Percent per Annum
Logarithm difference
IMF
Government Bonds
Percent per Annum
Logarithm difference
IMF
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Poland 2001Q1-2016Q2
Money Market Rate
Percent per Annum
Level IMF
Government Bonds
Percent per Annum
Level IMF
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Russia 1999Q1-2016Q2
Money Market Rate
Percent per Annum
Level IMF
97
Government Bonds
Percent per Annum
Level FRED
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
South-Africa 1990Q1-2016Q2
Central Bank-Policy Rate
Percent per Annum
Level IMF
Government Bonds
Percent per Annum
Logarithm difference
IMF
The total share price for All shares
Logarithm difference
FRED
Real Effective Exchange Rate, based on Consumer Price Index
index Logarithm difference
IMF
Source: Author’s compilation
The estimation is done on quarterly data with four lags using the rule of
thumb. For unit root analysis Ng and Perron (2001) are employed. All datasets
are standardized.
Proxy for the financial condition is FCI from the USA developed by Brave
and Butters (2012), it comprises a weighted average of 105 indicators of financial
activity those wide-ranging exposure of the financial system of Money markets,
Debt and Equity markets, Traditional and “Shadow” banking system and useful in
monitoring financial stability and forecasting.
98
For each country, proxy for short interest rate R is money market rate for
Brazil, Czech Republic, Malaysia, Pakistan, Poland Russian Federation and
discount rate for Chile, Hungary, India, T-bill rate for Greece, Mexico and central
bank policy rate for South Africa, proxy for long-term interest rate is saving rate
for the Brazil and a government bond rate (GBR) for the rest of the countries, a
real effective exchange rate (index) REER as proxy of exchange rate, and a stock
prices S proxy of stock market. for Pakistan and Greece government bond was
missing at some points. It was handled using interpolation. For this econometric
method is employed named cubic spline. High-frequency variables are converted
to low-frequency variables using period-end values.
99
3. Results
Here, it is tried to lengthen the time span of the data to assess how the
U.S. financial conditions are transmitted to emerging markets. For this reason, in
this study impact of FCI to emerging economies upon following channels is
studied:
Interest rate channel (short term and long term)
Exchange rate channel
Stock market channel
In the current study, it is being claimed that financial conditions have
predictive power of macroeconomic variables. In this session, results of financial
shocks on economies are presented. The shaded zones signify one standard
error bias-corrected bootstrap confidence bands of Kilian (1998).
3.1 Brazil
The impulse response functions in case of Brazil are presented in figure 1.
Time span is 1993q1-2016q2. This allows seeing the financial condition index’s
(FCI) impact mechanism clarifying by demonstrating the response of the system
to a shock in the measure of financial condition index. An innovation to FCI has
implications for all the variables in short term. A one standard deviation innovation
in FCI in short-term interest rate is followed by strong impact volatile and rapidly
building but a temporary response. In terms of long-term interest rate response is
also volatile but strong, highly positive in the start and rapidly building like short-
term interest rate but temporary in nature. Exchange rate overshoot in the start,
the initial negative response is being followed by a positive response that ends
100
shortly. The response of the stock market is also low and volatile and temporary
in nature. The response of all the variable is temporary in nature.
Fig V-1: Transmission to Brazil
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
101
3.2 Chile
The time span for the study is 2004 Q3 - 2016 Q1. The reason behind the
shorter span is the unavailability of data. The impulse response functions in Chile
is presented in figure 2. This allows seeing the financial condition index’s impact
mechanism clarifying by illustrating the response of the system to a shock in the
measure of financial condition index. An innovation to FCI has implications for all
the variables in long run. A one standard deviation innovation in FCI in short-term
interest rate is followed by strong and volatile and rapidly building, the response
is of a permanent nature and positive in long run. The response of long-term
interest rate is negative in the start being followed by positive stable and
permanent response. The response of exchange rate is like Dornbusch
overshoot; it overshoots in the start but stabilize in the long run. Stock markets
respond volatility in the start but the positive and stable response in long run. The
response of all the variables is of a permanent nature. The response of all the
variable is temporary in nature.
102
Fig V-2: Transmission to Chile
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
103
3.3 Czech Republic
The impulse response functions in case of Czech Republic is presented in
figure 3. The time span for the study is 2000 q2-2016q1. This allows seeing the
financial condition index transmission mechanism clarifying by illustrating the
response of the system to a shock in the measure of financial condition index. An
innovation to FCI has implications for the variables in short term. The response of
interest rate is volatile and negative in the short run but positive and stable in long
run. Long-term interest rate responds positively in short-term but this response is
of a temporary nature. Exchange rate behaves negatively in short-term but
stabilizes in the long run while the stock market behaves positively in the short
run and stabilize afterward. The response of variables is of a temporary nature.
The response of all the variable is temporary in nature.
Fig V-3: Transmission to the Czech Republic
104
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands. Chile
3.4 Greece
The impulse response functions in the case of Greece is presented in figure
4. The time span is 1993 q3-2016q4. This allows seeing the financial condition
index transmission mechanism clarifying by illustrating the response of the system
to a shock in the measure of financial condition index. A one standard deviation
innovation to FCI has strong implications for the short-term interest rate in short-
term, after the initial hike, it is being followed by a positive and stable response in
the long run. The response of long-term interest rate, exchange rate, and the stock
market is of a temporary nature, volatile in the start but stable in the long run. The
response of exchange rate is overshooting; Response of all the variable is
temporary in nature.
105
Fig V-4: Transmission to Greece
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
106
3.5 Hungary
The impulse response functions in the case of Hungary is presented in
figure 5. The time span is 2001q1-2016q1. This allows seeing the financial
condition index transmission mechanism clarifying by illustrating the response of
the system to a shock in the measure of financial condition index. A one standard
deviation innovation to FCI has implications for the variables in the long term. The
response of short-term interest rate is strongly positive in short-term followed by
a negative response that stabilizes in the long run, the response is of a permanent
nature. The response of long-term interest rate and the stock market is volatile
throughout the time span; a positive high response being followed by negative
making a V-shaped response in short-term that keeps a volatile movement in the
long run. Exchange rate responds negatively initially followed by volatile but long-
run relation. The response of all the variable is temporary in nature.
107
Fig V-5: Transmission to Hungary
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
108
3.6 India
The impulse response functions in the case of India is presented in figure
6. time span 2005 q1-2015q3. This allows seeing the financial condition index
transmission mechanism clarifying by illustrating the response of the system to a
shock in the measure of financial condition index. An innovation to FCI has
implications for the variables in the long run. The response of short-term interest
rate is negative initially being followed by a positive response that’s permanent
nature. Response if the long-term interest rate is volatile in the short term that
tends to stabilize in the long run. This is true in the case of the exchange rate and
a stock market that is volatile initially but strong and positive in the long run. The
response of all the variable is temporary in nature.
109
Fig V-6: Transmission to India
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
110
3.7 Malaysia
The impulse response functions in the case of Malaysia is presented in
figure 7. Time span 2000q2-2016q3. This allows seeing the financial condition
index transmission mechanism clarifying by illustrating the response of the system
to a shock in the measure of financial condition index. An innovation to FCI has
implications for the variables in short term. One standard deviation of innovation
in FCI is being followed by a negative response in short-term but positive in the
long run that faded away with time, response of long-term interest rate is volatile
in short-term that also faded away in the long run, exchange rate is highly volatile
in the short run that tends to fade away in long run and response of stock market
a positive initial response is being followed by stable response that tends to fade
away. The response of all the variable is temporary in nature.
111
Fig V-7: Transmission to Malaysia
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
112
3.8 Mexico
The impulse response functions in the case of Mexico is presented in figure
8. 1995 q1-2016q2. This allows seeing the financial condition index transmission
mechanism clarifying by illustrating the response of the system to a shock in the
measure of financial condition index. An innovation to FCI has implications for the
variables in short term. The response of short-term interest rate is negative
followed by a positive, the long-term interest rate is volatile, the exchange rate is
positive than negative and the stock market is negative followed by positive. The
response is of short-term nature of all the variables. The response of all the
variable is temporary in nature.
113
Fig V-8: Transmission to Mexico
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
114
3.9 Pakistan
The impulse response functions in the case of Pakistan is presented in
figure 9 with the time span of 1997q3-2016q1. This allows seeing the financial
condition index transmission mechanism clarifying by illustrating the response of
the system to a shock in the measure of financial condition index. An innovation
to FCI has implications for the variables in short term. The one standard deviation
shock arising in FCI is followed by the volatile response of short-term interest rate
that faded away with time. The response of long-term interest rate dies after initial
positive and negative wave. Exchange rate dies also the first negative than
positive response. The response of stock market is strong in the start but die off
with the time. The response of all the variable is temporary in nature.
115
Fig V-9: Transmission to Pakistan
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
116
3.10 Poland
The impulse response functions in case of Poland is presented in figure
10. 2001q1-2016q2. This allows seeing the financial condition index transmission
mechanism clarifying by illustrating the response of the system to a shock in the
measure of financial condition index. An innovation to FCI has implications for the
variables in long term. The small impact effects are followed by quickly building
response short term and long term interest rate and it’s of permanent nature. The
response of exchange rate volatility and the stock market is volatile that tend to
stabilize with time. The response of all the variable is temporary in nature.
117
Fig V-10: Transmission to Poland
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
118
3.11 Russian Federation
The impulse response functions in case of Russian Federation is
presented in figure 11. Russia 1999q1-2016q2. This allows seeing the financial
condition index transmission mechanism clarifying by illustrating the response of
the system to a shock in the measure of financial condition index. An innovation
to FCI has implications for the variables in long run. The response of interest rate
is strong, after an initial positive response it tends to stabilize and it’s of permanent
nature. The response of the exchange rate and the stock market is volatile but it's
of a temporary nature. The response of all the variable is temporary in nature.
119
Fig V-11: Transmission to Russian Federation
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
120
3.12 South Africa
The impulse response functions in the case of South Africa is presented in
figure 12. Time span 1990q1-2016q2. This allows seeing the financial condition
index transmission mechanism clarifying by illustrating the response of the system
to a shock in the measure of financial condition index. An innovation to FCI has
implications for the variables is mixed. One standard deviation shock of FCI is
being followed by the strong response of short-term interest rate that is permanent
in nature. The response of long-term interest rate, exchange rate, and stock
market dies off after the initial response. Exchange rate overshoots. The response
of all the variable is temporary in nature.
121
Fig V-12: Transmission to South Africa
Note: These are Impulse Response Functions from a five-variable VAR
with FCI, short term interest rate (IR), long term interest rate (GBR), exchange
rate (ER) and stock market (S). FCI is ordered first. The shaded areas are one-
standard-error confidence bands.
122
4. Conclusion
The main purpose of using U.S. financial condition shocks is to identify the
predictive power of major economic variables of emerging markets. For this
purpose, the predictive power of financial conditions is tested upon major
macroeconomic variables namely short term and long term interest rate,
exchange rate and stock markets. Impulse response affirms the predictive power
of the financial conditions of macroeconomic movements of the US in emerging
markets. Results have given in the following table.
Table V-2: Results
Short Term
Interest Rate
Long Term
Exchange Rate
Stock Die Off With Time
Country
Advanced Emerging Countries
Brazil Temporary and Weak
Temporary and Weak
Temporary and Weak
Temporary and Weak
Yes
Czech Republic
Weak but Permanent
Weak but Permanent
Weak but Permanent
Weak but Permanent
Yes
Hungary Strong and Permanent
Moderate and Permanent
Moderate and Permanent
Moderate and Permanent
No
Malaysia Moderate and Permanent
Moderate and Permanent
Moderate and Permanent
Moderate and Permanent
Yes
Mexico Weak but Permanent
Weak but Permanent
Weak but Permanent
Weak but Permanent
Yes
Poland Strong and Permanent
Strong and Permanent
Moderate and Permanent
Moderate and Permanent
No
South Africa
Yes Except In Case Of Short Term
123
Interest Rate
Secondary Emerging Countries
Chile Strong and Permanent
Strong and Permanent
Strong and Permanent
Strong and Permanent
No
Greece Strong and Permanent
Strong and Permanent
Moderate and Permanent
Moderate and Permanent
Yes
India Moderate and Permanent
Moderate and Permanent
Moderate and Permanent
Moderate and Permanent
No
Pakistan Temporary and Weak
Temporary and Weak
Temporary and Weak
Temporary and Weak
Yes
Russia Moderate and Permanent
Moderate and Permanent
Weak but Permanent
Weak but Permanent
No
Source: Author’s compilation
This chapter discusses transmission of the US financial shock to the
emerging countries. This chapter also discuss results of the second objective of
the study.
124
CHAPTER VI
CONSTRUCTION OF FINANCIAL CONDITION
INDEX FOR PAKISTAN
1. Introduction
This chapter will be discussing last objective. An index is created for the
special case on the emerging countries named Pakistan. The construction and
prediction of the index is also presented.
One of the learning lessons of recent mortgage crisis commonly known as
the global financial crisis is that broad financial conditions due to innovations in
financial landscape are difficult to capture by using a small number of variables
that cover only a few traditional financial markets. In the light of different episodes
of crises policymakers, regulators and financial market stake-holders have
affirmed the link between traditional and newly developed markets and the link
between financial and nonfinancial market. Closer watch on financial stability is
essential for understanding such links. Moreover, many of the econometric
system, that can be used to forecast or simulate a shock’s impact, do have the
very weak financial background. In many situations, they are based on single
variables; interest rate in many cases. This practice creates problems especially
in a time of crisis time periods. For this reason, index on financial conditions are
the best for this purpose (Brave and Butters 2010), (Hatzius, et al. 2010); Koop
and Korobilis (2014).
125
Financial Condition Index (FCI) may serve for many purposes. For
example, it can be used to find out early signs of bad financial conditions Gomez
et.al, (2011); Muraru (2015) or could serve as a forecaster of the economy
(Nombulelo et.al, (2012); Bautista (2013); Erdem and Tsatsaronis (2013). It is now
in practice of many financial institutes (IMF, Goldman Sachs, and Bloomberg) and
authorities (federal reserve bank of Chicago and many other banks) to develop
FCI for the market watch. Estimation of FCI ranges from simple weighted average
method to developed sophisticated methodology. Keeping in mind the existing
practice of developing the financial index, the chief empirical input in the literature
of this study is to develop an FCI for an emerging market using the most recent
approach.
Development and usage of FCI deal with the variable choice for FCI and
its link with macro-economy. This need to think about changing state. For this
reason, a method of index development by Koop and Korobilis (2014) is utilized.
Indexes are created using a wide range of macroeconomic and financial variables
over a long horizon for Pakistan. They developed a method using extensions of
Factor models and presented multiple forms of the index. The rationale of using
this method is that it is able to capture the time-varying nature of the variables so
can give a better picture of financial conditions.
Results show that constructed FCIs do have predictive power for
macroeconomic variables. This index correctly forecasts major macroeconomic
variables and indicates that they both moves in the same direction and correctly
forecast.
126
FCI constructed here may serve as a decision-making tool in place of
monetary policy stances those are employed as a policy tool. This FCI can also
be utilized for identifying the historical development of any phenomena and the
current state of the system and may utilize for forecasting other macroeconomic
sectors.
Following segment deals with the econometric method of FCI
development, followed by data and model sets, estimation of FCI and forecasting
of macro-economy and in the last but not the least conclusion and discussion.
2. Data
In this study, for the index formation twenty variables covering major
financial and economic variables. All the variables are stationary. Data is being
stationary using Phillips Perron. Table 5.1 provides detail on the stationary, data
source and other detail related to data. The time span for the study is 1969 Q1 till
2016 Q1. All models use four lag. Forecasting is done of macroeconomic
variables namely Consumer price index, gross domestic product, real effective
exchange rate, discount rate and KSE 100 index. Some variables do not start
from 1969 and some have missing values; this is being dealt with Kalman filter.
127
Table VI-1: Variables and Transformation
S# Name Transformation Source
1 Equities Logarithm difference IMF
2 Gold Logarithm difference IMF
3 Import Volume Logarithm difference IMF
4 Export Volume Logarithm difference IMF
5 Goods, Deflator/unit value of import
Logarithm difference IMF
6 Goods, Deflator/unit value of export
Logarithm difference IMF
7 Industrial Production Index Logarithm IMF
8 Discount Rate First difference IMF
9 Bond Rate First difference IMF
10 Money market Rate Level IMF
11 Producer Price Index Logarithm difference IMF
12 Total Reserves Logarithm IMF
13 Currency Logarithm difference IMF
14 CPI Logarithm difference IMF
15 T-bill rate First difference IMF
16 KSE 100 index Logarithm difference Finance.Yahoo
17 GDP Natural Logarithms SBP-Paper (2013) & Arby (2008)
18 Real Effective Exchange Rate Logarithm difference IMF
Source: Author’s compilation
To summarize, the models which yield FCI are the TVP-FAVARs, FA-TVP-
VARs, FAVARs.
Hyperparameters and initial conditions in the paper are being set as set by
Koop and Korobilis (2014). Following them, TVP-FAVAR models and its restricted
versions are being obtained by setting following forgetting/decay factors:
TVP-FAVAR: 1-0.96; 2-0.96; 3-0.99; 4-0.99
Heteroscedastic FAVAR: 1-0.96; 2-0.96; 3-1; 4-1
FA-TVP-VAR:1-0.96; 2-0.96; 3-1; 4-0.99
128
3. Construction of Financial Condition Index
FCI is being constructed using all three methods of FAVAR. Factors are
being estimated and in the same order, FCI is being estimated for the time span
of 1969 Q1 till 2016 Q1.
Figure 1 till 3 are representing factors estimated using all the variables. As
it is depicted that estimates are quite similar in all models namely TVP-FAVAR,
and-TVP-VAR.at some points minor difference does exist but these differences
are not very strong.
Fig VI-1: Factors estimation using TVP-FAVAR
129
Fig VI-2: Factors estimation using FAVAR
130
Fig VI-3: Factors estimations using FA-TVP-VAR
131
3.1 Estimating financial Position using FCI
In order to find out the validity of constructed FCI, its movement is being
checked with historical events and if it is succeeded to illustrate the true shape of
financial history, that index is considered to be strong. To work on it, Figure 4 to
figure 6 shows the construction of the FCI. The estimates from TVP-FAVAR,
FAVAR, FA-TVP-VAR are quite similar. This indicates that all the indices are
showing similar conditions of financial and economic history. Next job is to closely
monitor history, for this, we need to have a look at history and see whether it
covers major events of the history.
3.1.1 1971-77 Era (Post Dhaka Fall period)
Sample time frame starts from the end of a decade of development and
start of bad luck years. In its background, the second five-year plan (1960-65) was
a huge accomplishment due to political stability. It was generally believed that
South Korea acquired many ideas from the second five-year plan and
implemented these to achieve a high degree of success. During the 1960s,
Pakistan achieved food autarky. New variations in wheat and rice were
introduced. The Gross National Product was 8.3%, second highest in Asia after
Japan. The payback periods were the shortest and in many cases only seven
months. Pakistan achieved record growth and as a result of examining the record,
this decade could be said best performing in the history of Pakistan. Pakistan was
considered to be a model capitalist economy in the 1960s. But the 1965 war with
India proved an economic setback for Pakistan. Socio-economic tensions and
social upheavals and industrial strikes coupled with several other factors besieged
the country.
132
The decade of development was followed by Dhaka fall due to that
industrial base shrinks. In December 1973 inflation reaches its highest 37.8
percent. Bhutto’s economic program was considered to be a failure by his critics.
In this time rupee was devalued by 120 percent till may 1972. In 1973,
Organization of the Petroleum Exporting Countries(OPEC) price increases played
havoc with Pakistan’s import bill and balance of payments deteriorated.
The period after 1973 saw a serious worldwide recession affecting
Pakistan’s exports.
Recurrent domestic cotton crop failures and floods in 1973, 1974 (along
with pest attacks) and 1976 affected Pakistan’s main exports. The 1970s were a
turbulent period in Pakistan’s economic history. Due to socio-economic
compulsions, industries, financial institutes, agro-based sectors, social
infrastructure projects, shipping services and many other productive sector
industries were nationalized without appropriate cushioned logistics. Payments to
wage earners went up without productivity increase.
The outcome of nationalization measures ended incentives for private
investment. Hundreds of public enterprises were established with political motives
resulting in corruption and inefficiency (Saeed, 2013); (Zaidi, 2010).
3.1.2 1977-88 era (The second military regime)
Martial law was imposed in 1977 and the rulers never addressed the
economic agenda as a priority. General Zia’s time period was more liberal in
economic terms than predecessors. Remittances from the middle east and aid
from abroad helped launch Pakistan’s second economic revolution. 1984
government launched money whitening scheme. 1988 first structural adjustment
133
agreement with IMF. The results were that towards the latter part of the 1970s
and a major part of 1980s, Pakistan operated on the famous principle of ‘business
as usual’.
3.1.3 The era of structural Adjustments (1988-2001)
During the 1990s, the country lost on several fronts, especially in economic
growth. The economic performance was dismal. Despite public announcements
of self-reliance, the government’s actions continued to undermine their intentions.
Borrowings increase but export declined. Loan defaulters were not prepared to
return borrowed money. There were about 5,000 sick industrial units. Fixed
income earners received no incentives to protect their wages. The business
community was exasperated with the multiplicity of taxes at different levels.
Indeed, the decade of the 1990s excluding may 28, 1998 when Pakistan became
a nuclear power, didn’t emerge as the economically sound period for the country.
after nuclear test many nations imposed economic sanctions on Pakistan. More
precisely 90’s era is known as return of democracy and era of structural
adjustment. In this time period, Economic liberalization and stabilization were
adopted by the authorities and privatization was encouraged, among these for the
encouragement of the exports tariff rates were lessen. Moreover, during this time
macroeconomic crisis happened, high taxation was in action, trade reforms
resulted in deindustrialization and rupee devalued on a continual basis, inflation
rates were high and privatization was done without proper policy (Saeed 2013).
in such a volatile time period, volatility can also be seen in the index. Its been
changing all the time.1991-PM begun the economic liberalization program. 1991-
KSE 100 indexes were launched.
134
3.1.4 2001 to Onwards
2008-government enters IMF stabilization program to ward off the balance
of payment crisis. The stock market hit 15760 points on April 20 but in just four
months KSE plunges 55pc, wiping 136.9 billion off market value. Market touched
the lowest point. In the end of 90’s and early years of 2000’s acceleration in
economic growth is seen, along with this industrial production increased, export
earnings raised, upsurge in investment, and foreign exchange reserved increased
(Zaidi 2005) This can be seen from the graph that during this time index goes up
but again is in volatile condition.in the mid-2000’s economic growth picked up
more, in this time period sound macroeconomic fundamentals were achieved.
(Zaidi 2005). From the graph, we may see fluctuation but towards an upward
trend. In last of 2000’s impact of global crisis started to show its impact upon
Pakistan’s economy and downward trend was observed in major variables. This
is observable from the graph. 2013-China Pakistan Economic Corridor(CPEC)
formalized, IMF approves $6.7 billion loan package to help Pakistan revive the
ailing economy. A recovery phase can be seen after the impact of the crisis that
sustained till 2016, fluctuating and with the upward movement of the index.
Overall, significant periods of crisis in financial history are well captured as are
periods of relative calm.
The volatility of all the models is the same for all figures. This is indicative
of the graph that financial conditions are volatile and changing over time. This is
true for Pakistan. Throughout the history due to political instability, financial
conditions are not sound over time in Pakistan. More precisely, year wise following
135
are significant events in the time span for which index is created. This timeline is
truly indicating up and down in the history and at the same time fluctuations can
be seen in the graph of the FCI.
Fig VI-4: FCI estimation using TVP-FAVAR
136
Fig VI-5: FCI estimation using FAVAR
137
Fig VI-6: FCI estimation using FA-TVP-VAR
138
4. Forecasting of Macroeconomic Variables
Preceding session test the index by its ability to forecasts the major
macro-economic variables.
4.1 Inflation
Pakistan has traditionally been a low inflation country. Consumer price
index annual changes on average 2 to 3 percent during the 1950s and 1960s.
during the Bhutto Government in the 1970s, the CPI rose on average by 20
percent every year. The 1980s saw a major deceleration of inflationary pressure
and price increases were restricted to 7% per annum. On the other hand, policy
liberalization period up to the end 1990s observed acceleration. Inflation was 10%
on average during mild liberalization and about 12% per annum during the era of
intensive liberalization. The rate of inflation was halved during 1998-2003. The
Musharraf period (1999-2008) can be divided into inflationary episodes. During
the ending period in 2004, inflation was low but it doubled during 2004-08,
reaching double figures in 2008 (Meenai, 2010).
Figure VI-7 is showing the results of forecasted inflation. Movements in
inflation are almost similar to forecasted and actual. Ups and downs are almost
the same. So, it can be said that index is a true representative of the economy.
139
Fig VI-7: Forecasted Inflation
4.2 Exchange Rate
The exchange rate is a policy instrument which can be used to affect
almost all constituents of the balance of payments. A depreciating currency may
or may not stimulate exports, discourage imports, workers’ remittances, jack up
interest rates stimulating external capital inflows and raising the burden of
government debt (Meenai, 2010). The Pakistani rupee has depreciated from Rs.
3.3 to the dollar in August 1947 to Rs. 116 to the dollar in March 2018. (source:
Official Website of Statistical Bureau of Pakistan).
140
Table VI-2: Chronical Exchange Rate in Pakistan
Year Exchange Rate Regime PKR per
US $
1955 Fixed Exchange Rate (1947-1982) 4.76
1972 11
1973 9.9
1982 10.1
1982 10.55
1983 Managed Float (1982-1998) 12.7
1984 13.4
1985 15
1986 16
1987 17
1988 17.59
1989 19.2
1990 21.4
1991 22.4
1992 22.4
1993 25.9
1994 30.1
1995 30.85
1996 33.56
1997 39.99
1998 43.195
1999 Two-Tier Exchange Rate System (Multiple Exchange Rate from July 1998 till may 1999)
50.05
2000
Dirty Float: State Bank Pakistan (SBP) defending the Exchange Rate within a Narrow Band from 19 may 1999 till July 2000; from Managed Float to Floating Exchange Rate Regime since 20 July 2000 (Free Float Regime since 2000)
51.77
2001 58.4
2002 61.4
2003 58.4995
2004 57.57
2005 59.657
2006 59.85
2007 61
2008 71.46
Source: Author’s compilation
141
Movements in the exchange rate are almost similar to forecasted and
actual. It can be said that Index has given a realistic view of the exchange rate.
Fig VI-8: Forecasted Real Effective Exchange Rate
4.3 Monetary policy (short-term interest rate)
In mid of the 1990s, SBP shifted monetary policy management method.
Before the 1990s it used to change monetary policy on an ad hoc basis but after
1990s it shifted towards market-oriented monetary management. Despite sincere
efforts from monetary authorities’ monetary policy have failed to meet its objective
in Pakistan (Meenai, 2012). As it is evident from the graph that monetary policy is
kept throughout the time.
142
Fig VI-9: Forecasted Discount Rate
4.4 Gross Domestic Product (GDP)
Historically, the growth rate of Pakistan has been good. On average, it has
been near to five percent annually during the last six decades. In its regional level,
its been at two percent from the 1960s till 1980s; however, since 1993 its low then
the regional average.
Table VI-3: Chronical GDP
1950s 1960s 1970s 1980s 1990s 2000-06 1950-2006
GDP 3.5 6.8 4.8 6.5 4.6 5.4 5.2
Source: Official Website of Statistical Beaurue of Pakistan
It is evident from graph that FCI is able to forecast GDP.
143
Fig VI-10: Forecasted Gross Domestic Product
4.5 Stock Market
The stock market of Pakistan has been volatile in most of the time as
depicted from the following graph.
Fig VI-11: Forecasted Stock Market
144
4.4 Forecasting under other Variants of the Model
Fig VI-12: Forecasting using FA-TVP-VAR
145
Fig VI-13: Forecasting using FAVA
146
Fig VI-14: Forecasting using TVP-FAVAR
Forecasting graph is similar under all the models and constructed FCI is
able to closely predict the macroeconomic situation of Pakistan. Thus, can be
used for the prediction of Pakistan’s economy.
147
5. Conclusions
After mortgage crisis or more commonly known as global crisis 2008, it’s
been core job of policymakers and other stakeholders to have an eye on the
financial conditions of the economy for better study of the economic situation of
any country. A recently developed tool serving this purpose is financial condition
index. Literature has no clear rule for the variable selection and method for the
index formation but its importance has been acknowledged and backed by
empirical support.
By knowing such an important job of the Financial Condition Index, this
study is an attempt to develop an index for Pakistan as Pakistan don’t have such
an instrument for market and economy watch. Using the financial and economic
variables those are considered representative of a developing state develops the
index.
This chapter has presented in length methodology and analysis of the
index. After presenting index, prediction of the major macroeconomic variables is
also presented.
148
CHAPTER VII
DISCUSSIONS
This chapter will be presenting discussion upon all the objectives. All the
objectives will be discussed in length.
1. Discussions of First Objective
The key driver of this objective is to identify the impact of Monetary Policy
upon major economic variables of emerging markets. For this purpose, the
predictive power of Monetary Policy is tested upon major macroeconomic
variables namely Growth and prices (inflation). A transmission of monetary policy
is being analyzed at country level and for studying contagion impact of US
monetary policy upon emerging markets has been studied. This has been
conducted at two groups of countries namely advanced emerging countries and
secondary emerging countries.
1.1 Advanced emerging countries
Brazilian government introduced Plano Real in 1994. This plan
transformed economy completely and brought sustainability of economic growth.
Economy was praised after this plan due to price stability, fiscal responsibility,
rapid growth, was result of macro-economic reforms started in 1990’s in the form
of real plan, banking reforms, privatization, greater openness of the economy
towards foreign direct investment. As a result of all these efforts, inflation came
down. As a result of real plan, increasing interest rates stabilized currency. This
149
attracted foreign investment. After success of this initiative, formal inflation
targeting policy was introduced. By looking at its, growth rate during 1995-2003,
it was weaker by international standards but during 2000’s, remarkable growth
can be seen in Brazil. A slump in growth during 2009 due to global crisis but a
jump in 2010 is also observable. After first decade of 2000’s Brazil experienced
decline in economic situation (Amann & Baer, 2012).
In the study of monetary policy transmission of Brazil, direct response is
observable in both variables. The impulse responses of output to a positive
interest rate shock (𝜀𝑖 ↑→ 𝑥) is not aligned with theory as it stays negative and
volatile most of the time but impulse response of inflation to positive interest rate
shock (𝜀𝑖 ↑→ 𝑝) is align with theory as it is negative most of the time. Luporini
(2012) found similar results while analyzing the monetary policy transmission in
Brazilian economy using VAR model. He found that tight monetary policy results
in downward trend of GDP growth rate and inflation. Inflation and exchange rate
does have an impact but after an interval. In the study of United State’s monetary
policy transmission in Brazil, the impulse responses of output to a positive interest
rate shock (𝜀𝑖 ↑→ 𝑥) is positive and volatile and impulse response of inflation to
positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is positive and of stable nature.
1997’s currency crisis shattered Czech Republic’s economy severely. As
a mitigating plan, two austerity packages and revitalization programs were
introduced from the government. As a result of these efforts, growth can be seen
during first decade of 2000. Growth rate of the economy was on decreasing trend
from 2009 till 2013 and from 2014 to onwards; it again started towards increasing
150
trend. In 2015, its GDP was ranked as highest in European Union. Despite of
being part of European union, Global financial crisis doesn’t have an impact upon
this economy mainly due to two reason; first strong banking sector as a result of
1990’s crisis and presence of national currency. (Havlat, Havrlant, Kuenzel, &
Monks, 2018); and (Abdel-Salam, 2017).
In the study of monetary policy transmission of Czech Republic, direct
response is observable in both variables. The impulse responses of output to a
positive interest rate shock (𝜀𝑖 ↑→ 𝑥) shows the highest volatility but the response
is positive most of the time that is not align to the standard. The response of price
(𝜀𝑖 ↑→ 𝑝) is also positive but less volatile as compared to output. Response of
both variables deviates from standard in both cases. Both response deviates from
the standard. Koerner (2015) found opposite results while analyzing the
transmission of monetary policy in Czech Republic using recursive VAR, SVAR
and structural vector error correction model (SVECM). He found in all the models
that tightening monetary policy results in decline of output and inflation. In the
study of United State’s monetary policy transmission to the Czech Republic, the
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) is positive
and volatile most of the time and impulse response of inflation to positive interest
rate shock (𝜀𝑖 ↑→ 𝑝) is also positive.
As a result of transition from 1990 till 2004, Hungary shifted from planned
economy to market economy. In the form of reform, government announced
Bokros package for saving country from the financial collapse and aims to
stabilize economy. Many of the resulting measures worked as shock therapy and
151
put the country on the road of sustainable growth and in succession country joined
European Union in 2004 (Csizmadia, 2008).
In the study of monetary policy transmission, direct response is observable
in both variables. The impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) are aligned with theory as it stays negative but impulse response of
inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) deviates from standard that is
positive. Balazs (2005) found partly similar results while analyzing the monetary
policy transmission in Hungarian economy using SVAR method. He found that
contractionary monetary policy results in appreciation of exchange rate, lowering
output and inflation. In the study of United State’s monetary policy transmission,
the impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
positive and impulse response of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝)
is also positive.
Malaysia is an economic sound country and outperforming in the region
since 1970. Since 1970, economy has gone to a number of transformations. As a
result of Asian financial crisis in 1997-98, country strengthen its macro-prudential
policies and many other policies aims toward stability of the financial and
economic sector. These reforms also helped in recovering from global financial
crisis of 2008 and country managed to have a sound economic growth (Koen,
Asada, Nixon, Rahuman, & Arif, 2017).
The impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥)
stay negative and impulse response of inflation to positive interest rate shock (𝜀𝑖 ↑
→ 𝑝) is mixed. In the study of United State’s monetary policy transmission, the
impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay
152
positive and volatile most of the time and impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is also positive.
Mexican economy improved macroeconomic policies as a result of 1994
crisis that’s why country was not much affected by South American Crisis 2002.
However, country was severely hit by the mortgage crisis of 2008. Mexico is an
export oriented economy (OECD , 2017).
The impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥)
stay negative and volatile most of the time. The response of before shock is
positive compared to other shocks. Same behavior in case of price (𝜀𝑖 ↑→ 𝑝) is
observed, volatile and negative and positive response of before shock. In the
study of United State’s monetary policy transmission, the impulse responses of
output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥) stay positive and impulse
response of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is negative.
Poland, eighth largest economy of the European Union (EU), is pursuing
policy of economic liberalization since 1990. Poland was the only country in
European union to avoid recession of 2007. Till 2017, Polish economy was the
only country of the EU that was able to maintain economic growth since 26 years
(OECD, 2018).
The impulse responses of output to a positive interest rate shock (εi ↑→ x)
is mixed and volatile most of the time and impulse response of inflation to positive
interest rate shock (εi ↑→ p) is also positive and of varying nature. In the study of
United State’s monetary policy transmission, the impulse responses of output to
a positive interest rate shock (εi ↑→ x) is positive and impulse response of inflation
to positive interest rate shock (εi ↑→ p) is positive.
153
Turkey is at center of Eurasia (political and economic area where Europe,
Former Soviet Union and Middle East intersect). Geography of the Turkey creates
unique business opportunities for the country. Since 1994, country has
experienced three major economic downturn, among them crisis in 2001 was the
severest (Arguden, 2007). Following crisis of 2001, country experienced
economic growth as a result of structural change policies. From 2007 to onwards
government growth slowed down as government spending was the mainstay of
the economy (Acemoglu & Ucer, 2015).
The impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥)
stay negative and volatile most of the time and impulse response of inflation to
positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is mixed and of varying nature. Durran et.al
(2012) estimated the monetary policy transmission in turkey using GMM model.
They found that increase in monetary policy results in decline of stock price and
increase in government bond yield. In the study of United State’s monetary policy
transmission, the impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) stay positive and impulse response of inflation to positive interest rate
shock (𝜀𝑖 ↑→ 𝑝) is also positive.
1.2 Secondary Emerging Markets
Between 1989-1992 Colombia went on an unprecedented period of
change in economic policy and reforms. As a result of these initiatives, country
enjoyed fairly good economic growth in first half of 1990s, but country faced its
first economic recession in late 1990s in the midst of Asian and Russian crises.
This recession and real estate bubble resulted in major banking crisis. In early
years of 2000, economy started to recover (Steiner & Vallejo, 2010).
154
The impulse response of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥)
is negative in the start but positive afterwards. The response of price (𝜀𝑖 ↑→ 𝑝) is
volatile in nature and negative most of the time. In the study of United State’s
monetary policy transmission, the impulse response of output to a positive interest
rate shock (𝜀𝑖 ↑→ 𝑥) is positive and the impulse response of inflation to positive
interest rate shock (𝜀𝑖 ↑→ 𝑝) is positive.
During the 1990s, the country lost on several fronts, especially in economic
growth. The economic performance was dismal. Indeed, the decade of the 1990s
excluding may 28, 1998 when Pakistan became a nuclear power, didn’t emerge
as the economically sound period for the country. After nuclear test many nations
imposed economic sanctions on Pakistan. More precisely 90’s era is known as
return of democracy and era of structural adjustment. In this time period, the
authorities adopted Economic liberalization and stabilization and privatization was
encouraged, among these for the encouragement of the exports tariff rates were
lessen. Moreover, during this time macroeconomic crisis happened (Saeed,
2013). In the end of 90’s and early years of 2000’s acceleration in economic
growth is seen, along with this industrial production increased, export earnings
raised, upsurge in investment, and foreign exchange reserved increased (Zaidi,
2005).
The impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥)
stay positive and volatile most of the time and impulse response of inflation to
positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is also positive and of varying nature. Agha
et.al (2005) found opposite results while studying the monetary policy
transmission in Pakistan using VAR method and found that tightening monetary
155
policy results in reduction of price and output. In the study of United States’
monetary policy transmission, the impulse responses of output to a positive
interest rate shock (𝜀𝑖 ↑→ 𝑥) stay positive and volatile most of the time and
impulse response of inflation to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is mixed and
of varying nature.
Inflation in Peru in the form of Fuji shock resulted in macro-economic
stability, prudent fiscal spending, High international reserve accumulation,
External debt reduction, Achievement of investment grade status and Fiscal
surpluses (Martinelli & Vega, 2018). Being such a sound country Peru is having
direct impact of the United State’s monetary policy upon its output. This behavior
may be attribute to the trade agreement that happens between United States of
America and Peru on April 2006 (Federal Register, 2009). Under this agreement,
obstacles to trade were eliminated, and measures were taken to fostering private
investment in and between United States and Peru.
The impulse responses of output to a positive interest rate shock (εi ↑→ x)
is mixed and volatile most of the time and impulse response of inflation to positive
interest rate shock (εi ↑→ p) is volatile in the short run followed by a stable
response. In the study of United State’s monetary policy transmission, the impulse
responses of output to a positive interest rate shock (εi ↑→ x) stay negative and
volatile most of the time and impulse response of inflation to positive interest rate
shock (εi ↑→ p) is also negative but of stable nature.
The Philippines once was a model of development and second to japan in
East Asian countries. It hit badly with global financial crisis of 2008. But economy
started to recover in 2010s and it started to continue after wards (Hays, 2008).
156
The impulse responses of output to a positive interest rate shock (𝜀𝑖 ↑→ 𝑥)
stay negative and volatile most of the time and impulse response of inflation to
positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is negative in short run while positive in long
run and of stable nature. In the study of United State’s monetary policy
transmission, the impulse responses of output to a positive interest rate shock
(𝜀𝑖 ↑→ 𝑥) is positive and volatile most of the time and impulse response of inflation
to positive interest rate shock (𝜀𝑖 ↑→ 𝑝) is positive and of varying nature.
Starting years of transition from Soviet Central Planned Economy were not
easy for Russia. It was a time of economic chaos. Years 1999-2008 can be seen
with impressive Russian economic growth. This process ends with the hit of global
financial crisis and resulted recession in the country (Cooper, 2009).
The impulse responses of output to a positive interest rate shock (εi ↑→ x)
stay negative and volatile most of the time and impulse response of inflation to
positive interest rate shock (εi ↑→ p) is also negative in short run followed by
positive in the long run. Ono (2013) found the monetary transmission in Russian
economy using VAR method that monetary shocks have positive impact upon
growth. In the study of United State’s monetary policy transmission, the impulse
responses of output to a positive interest rate shock (εi ↑→ x) are positive and
volatile most of the time and impulse response of inflation to positive interest rate
shock (εi ↑→ p) is negative initially but weak positive response afterwards.
In studying the monetary policy transmission of US to the rest of the world,
researchers are having dynamic results but mostly endorsing this fact that
contagion exist despite the usage of method and data. Extent of contagion may
vary from sector to sector, economy to economy but in the light of evidences it
157
can be said that contagion does exist specially of the last Global financial crises.
Neri and Nobili (2010) found using the VAR method the evidence of the
transmission between Eurozone and US. Bagliano and Morana (2010) found the
evidence of spillover of US financial shocks to the advanced and emerging
countries using FVAR method. Todorov (2012) found the evidence of spillover
effects in financial markets of frontier markets from the US using the TVP-VAR
model. Kazi et.al (2013) found the evidence of US monetary policy shock to the
OECD nations in the light of TVP-FAVAR model. Fornari and Stracca (2013) in
studying the advance economies found the propagation of shocks not in turmoil
time period but in normal time. Ghani (2013) concluded that emerging economies
were exposed to shock from advance countries due to the weakness in regulatory
framework. Hab et.al (2014) found the evidence on information spillover and
liquidity spillover from the US to the open-ended property funds of other
economies. Yiu et.cl (2010) conducted a study for finding the relationship between
Asian and US stock market by using the principal component method. Results
indicated that US market is having a contagion impact upon Asian markets.
Barakchian (2015) studied the spillover of US monetary policy upon Canada using
global vector auto regression (GVAR). Kim (2001) concluded that expansionary
monetary policy results positive output in G6 but Bluedorn and Bowdler (2001) in
case of G7 and Scrimgeour (2010) found that positive monetary policy results in
positive short-term interest rate in four countries in America’s. Cross and Nguyen
(2016) studied global oil price shock upon china’s output using time varying
parameter vector auto regression (TVPVAR) model. They found that impact is
small and temporary in nature.
158
In studying the shock transmission to the stock market, Yiu et.cl (2010)
conducted a study for finding the relationship between Asian and US stock market
by using the principal component method. Results of Asymmetric Conditional
Correlation model indicated that US market is having a contagion impact upon
Asian markets. Ehramann and Fratzsher (2009) analyzed the transmission of US
Monetary Policy Shock to global equity market by taking data of 50 economies.
They also found heterogeneity of the transmission and also found that the
economies, which are open and relatively liquid markets are more prone to the
transmission. Markwat et.al (2009) proved that stock market contagion operates
as domino effect. He found that global crashes do not occur all of sudden but are
preceded by local and regional crashes.
Impulse responses of the emerging markets affirm the predictive power of
the Monetary Policy of macroeconomic movements in emerging markets. This
finding echoes the findings of Nakajima (2011), Primiceri (2005) who established
that Monetary Policy is being transmitted upon macro-economy.
Under the monetarist/ISLM, framework interest rate surprises represent
Monetary Policy shock. This states that monetary contraction (positive Monetary
Policy shock) creates declining output and increasing prices.
Looking at the individual variable response at the national level, with
positive Monetary Policy shock output declines in Brazil, Hungary, Malaysia,
Mexico, Philippines, Russian Federation and Turkey. Economies those deviate
from the theory are Colombia, Czech Republic, Pakistan, Peru, and Poland.
At the international level, positive Monetary Policy shock creates declining
output only in Peru, for the rest of the countries namely Brazil, Colombia, Czech-
159
republic, Hungary, Malaysia, Mexico, Pakistan, Philippines, Poland, Russian-
federation and turkey standard deviate. Inflation in Peru in the form of Fuji shock
resulted in macro-economic stability, prudent fiscal spending, High international
reserve accumulation, External debt reduction, Achievement of investment grade
status and Fiscal surpluses (Martinelli & Vega, 2018). Being such a sound country
Peru is having direct impact of the United State’s monetary policy upon its output.
This behavior may be attribute to the trade agreement that happens between
United States of America and Peru on April 2006 (Federal Register, 2009). Under
this agreement, obstacles to trade were eliminated, and measures were taken to
fostering private investment in and between United States and Peru. Direct
response of the output as a result of monetary policy shock may be due to this
agreement.
The response of Price to positive Monetary Policy shock is positive in Czech
Republic, Hungary, Pakistan, Peru, Philippines, Poland, Russian Federation and
Turkey. Standard deviates in Brazil, Colombia, and Mexico, and somehow in the
short run in Philippines and Russian Federation. In international level, prices
increase in Brazil, Colombia, Hungary, Malaysia, Pakistan, Philippines, and
Poland. The response is mixed in Czech-republic, Russian Federation, and
turkey. Prize puzzle exists in Mexico, Peru and somewhat in Russian Federation.
Sims (1992) conducted a study on the US by using short-term interest rate and
CPI and Industrial Production Index (IPI) and provides evidence of prize puzzle.
This study also affirms the results of agha et.al (2005).
160
From a theoretical perspective, this study provides evidence for prize
puzzle in emerging markets to some extent. From a practical perspective, this
study is beneficial for the researchers, economists and market practitioners.
This study is being limited by the availability of data. This study was started
with the objective to work in emerging markets. Emerging markets do lag due to
lack of specialized and well-maintained data. This study may be extended in the
future for broader results and application with the availability of data. We were
limited by data time span. Future work can be more reliable once these limitations
are overcome.
This study is an attempt to study the international transmission of US
Monetary Policy and standard Monetary Policy on macro-economic of emerging
markets. It is being found that Monetary Policy does have an impact upon macro-
economy of emerging markets; the extent may vary but its true for all the
countries. Past studies majorly covering advance countries; this study is an
attempt to extend the studies on emerging markets.
2. Discussions of Second Objective
2.1 Advanced Emerging Countries
Brazil, Czech Republic, Hungary, Malaysia, Poland and South Africa are
having bilateral relations with the US (Executive Office of the President, ). For that
reason, financial conditions of the US are having on impact upon the Brazil in
short term yet volatile. A one standard deviation innovation in FCI creates strong
but temporary response of short-term interest rate, long-term interest rate, and
exchange rate and of the stock market. Response of the macro-economy of
161
Czech Republic is of short term in nature and more stable as compared to Brazil.
Long run impact of US financial conditions is observable in the macro economy
of Hungary. Malaysia and the US are having bilateral relations. Impact of US
financial conditions upon macro-economy of the Malaysia is of temporary and
short-term nature. An innovation to FCI has implications for the Polish economy
in the long term. Response of South African economy upon the financial
conditions of the US is mixed. Short-term interest rate respond in the long run
while other macro economic variables behave in the short run. Mexico and the
US are having foreign relations, for that reason, an innovation to FCI has
implications for the macro economy of the Mexico in short term.
2.2 Secondary Emerging Countries
Chile a stable and prosperous nation, in 2006 it was having highest
nominal GDP per capita in Latin America (World Economic Forum, 2009). Chile
is having Free Trade Agreements and on strong bilateral relations with the US
(Executive Office of the President, ).
Chile is strategic alley of the US. That’s why; an innovation to FCI has
implications for the macro-economy of the Chile in long run.
Greece has been center of Eurozone debt crisis. It has the highest level of
public debt in Eurozone and high budget deficit. It was the first country that came
under intense market pressure and turn to international monetary fund (IMF) and
other states for financial assistance. By adopting Euro as currency, borrowing cost
decreased dramatically. Interest rate on government bond dropped till 18% from
1993-98. As a result of European Union membership, capital inflow increase and
162
convergence criteria was adopted. But with the passage of time investors lost its
confidence and it was on its peak in 2009. Global financial crisis and related
downturn of economies worsen the situation (Nelson, Belkin, & Mix, 2011). US
and Greece are on bilateral relations. Response of the macro economy of the
Greece upon the financial conditions the US is temporary in nature except stock
market.
Relation of the US and India are international relations. An innovation to
FCI has implications for the macro economy of the India in the long run. Pakistan
and Russian federation are having bilateral relations with the US. Innovations to
the financial conditions of the US are having implications for the macro economy
of the Pakistan in short term while in case of Russia response is in the long run.
Financial conditions does have implications for the economies and can be
used a representative of the economy. Major studies are available at country level.
Eickmeier et.al, (2011) employed the FCI developed by Hatzius et.al, (2010) for
studying the international transmission during 1971-2009 of financial shock using
the TVP-FAVAR method. They found that positive US financial shocks do have a
positive impact upon the growth of countries under study (US, Canada, the UK,
France, Germany, Italy, Spain, Japan and Australia).
Alessandria and Mumtaz (2017) hypothesized that the links between credit
markets and real economy tighten in a crisis; financial indicators might be
particularly useful in forecasting the macroeconomic outcomes associated with
episodes of financial distress. To capture the state of financial markets they
employed the Financial Condition Index (FCI) constructed and maintained by the
Chicago Fed. Balcilara et.al (2016) used a previously constructed index for finding
163
its ability to forecast the South African economy. They found that the response of
economy is non-linear to financial conditions. Opschoor et.al (2014) studied to
find the impact of financial conditions on the stock market by using Bloomberg
FCI. They found that worst financial conditions are associated with high volatility
and correlation between stock return.
The main purpose of using U.S. financial condition shocks is to identify the
predictive power of major economic variables of emerging markets. For this
purpose, the predictive power of financial conditions is tested upon major
macroeconomic variables namely short term and long-term interest rate,
exchange rate and stock markets. Impulse response affirms the predictive power
of the financial conditions of macroeconomic movements in emerging markets.
This finding echoes the findings of Brave and Butters (2011) who established that
FCIs do have strong predictive power in short-term and long-term horizons.
Similar findings can be seen in the study of Koop and Korobilis (2014), Hatzius
et.al (2010), Debuque and Bautista (2013) and many others.
Looking at the individual variable response, we can see that response of
short-term interest rate is initially negative and volatile in the start but positive and
stable in the long run in case of Brazil, Chile, Czech Republic, Greece, India,
Malaysia, Mexico, Pakistan, Poland, and South Africa. The exception of this trend
is Hungary and the Russian Federation where the initial positive response is being
followed by a negative response that with time shift into a positive and stable
response.
The response of long-term interest rate is volatile that tends to stabilize with
time and this is of permanent nature in case of Chine, Czech Republic, Greece,
164
Hungary, India, Poland, and Russian Federation. Volatile yet of the temporary
nature of response can be seen in the case of Brazil, Malaysia, Pakistan, and
South Africa.
The response of stock market is volatile in short-term in case of Brazil,
Czech Republic, Greece, Malaysia, Mexico, Pakistan, Russian Federation, and
South Africa while volatile and of permanent nature in Chile, Hungary, India, and
Poland.
The response of the exchange rate is like overshooting in many countries.
These finding echoing the findings of overshooting hypothesis proposed by
Dornbusch (1976) who established that immediate response of exchange rate
towards a disturbance is higher than its long-run response. Such response is
being observed in the case of Brazil, Chile, Greece, Hungary, India, Malaysia,
Mexico, Pakistan, Poland, Russian Federation, and South Africa. Exception from
this response is the Czech Republic who does not overshoot whose reasons are
unknown now.
From a theoretical perspective, this study confirms the hypothesis being
tested by notable researchers’ that state that financial condition index does have
predictive power of macro-economy and specifically confirms Dornbusch’s
exchange rate overshooting hypothesis in most of the economy.
From a practical perspective, this study is beneficial for the researchers,
economists and market practitioners.
This study is being limited by the availability of data. This study has started
the objective to work in emerging markets. Emerging markets due to lack of
specialized and well-maintained data. This study may be extended in the future
165
for broader results and application with the availability of data. We were limited by
data time span. It was after 2000 that data on most countries was available.
Furthermore, data were available on different frequencies that also limit us with
the application. Future work can be more reliable once these limitations are
overcome.
This study is an attempt to study the international transmission of US
financial conditions on macro-economic of emerging markets. It is being found
that financial conditions do have an impact upon macro-economy of emerging
markets; the extent may vary but its true for all the countries. Past studies majorly
covering advance countries; this study is an attempt to extend the studies on
emerging markets.
3. Discussion of Third Objective
After mortgage crisis or more commonly known as global crisis 2008, it’s
been core job of policymakers and other stakeholders to have an eye on the
financial conditions of the economy for better study of the economic situation of
any country. A recently developed tool serving this purpose is financial condition
index. Literature has no clear rule for the variable selection and method for the
index formation but its importance has been acknowledged and backed by
empirical support.
By knowing such an important job of the Financial Condition Index, this
study is an attempt to develop an index for Pakistan as Pakistan does not have
such an instrument for market and economy watch. Using the financial and
economic variables those are considered representative of a developing state
develops the index. A developing state does have an issue with the data. This
166
data issue limits us with the variable choice. The developed index is able to serve
as a barometer of the economy as it’s so near to historical event of Pakistan. So
it may say that it may serve better than monetary policy based estimation results
and can give a more realistic picture of the economy both at short term and long-
term basis. It is constructed over a long horizon that it could serve a better
representative of the economy.
Gaglianone and Areosa (2016) constructed a FCI for the Brazilian
economy using the method of Brave and Butters (2011) and Aramonte et.al
(2013). They use the developed index for studying the economic conditions and
employed for forecasting purpose. Gomez et.al (2011) construed a FCI for the
Colombia using PCA method and used it for forecasting purpose of the macro-
economy. Goodhard and Hofmann (2001), Mayes and Viren (2001), Gauthier
et.al, (2004). Guichard and Turner (2008) and Swiston (2008), Brave and Butters
(2011), Hatzius et.al (2010) constructed an FCI for the USA, Gumata et.al (2012)
constructed an index for South Africa, Gonzales and Bautista (2013) constructed
FCI for five Asian Markets and concluded that these index can be used for
forecasting purpose. Authorities have also created an index for studying the
financial conditions of economies. Hong Kong Monetary (2010) built an index for
the Hong Kong and China for studying the episodes of stress in mentioned
economies. Another index formed by Monetary Authority of Singapore (2009) for
studying the economic conditions of Asian countries (China, Republic of China,
Thailand, Taipei, Philippines, Malaysia, Korea, Republic of Korea, Indonesia, and
India). International Monetary Fund has also constructed Asia based index for
studying the economic conditions of Asian countries. This chapter has presented
167
discussion upon each objective. All three objectives have been discussed in
length.
168
CHAPTER VIII
CONCLUSIONS
This chapter will be presenting conclusion of the study. Conclusion
gained from each objective will be discussed.
1. Objective-wise Findings
Succeeding the evaluation of literature (Chapter 3), three key research
objectives for the thesis were identified.
1. To gauge the time varying effect of the monetary policy
2. To find out the effect of U.S. financial shocks on emerging markets
3. To develop and test FCI for Pakistan
We now raise back to these objectives, summarizing the procedures used
to approach them and the key points of conclusion that can be drawn from the
research.
1.1 Findings of First Objective
Mortgage financial crisis that sooner converted into global crisis affirms the
evidences of the transmission and contagion. For this reason, its being an attempt
to study the transmission and contagion arising from monetary policy. At first level,
transmission of monetary policy towards macro-economy is being studied and at
second level, contagion of monetary policy shock arising from United States on
the economy of emerging markets is being studied in a time varying context. This
objective is achieved using the TVP-VAR model with stochastic volatility proposed
by Jouchi Nakajima (2011). Results are indicating significant impact of shock
169
upon growth and price of the emerging markets and evidences for prize
puzzle and deviation from standards are also found in transmission mechanism
in some countries. This objective was further broken down into four hypotheses.
Status of the hypotheses is given in table 6.1.
Table: VIII-1: Status of Hypotheses of First Objective
S # Hypotheses Status
1. A contractionary monetary policy has inverse impact on growth.
Accepted in case of Hungary, Malaysia, Mexico, Philippines, Russian Federation, and Turkey.
2. Price puzzle exists in monetary policy transmission.
Accepted in case of Czech Republic, Hungary, Malaysia, Pakistan, Poland, Russian Federation, and Turkey.
3. The systematic US monetary policy has positive impact on growth of emerging economies.
Accepted only in case to US to Peru
4. The expansionary monetary in US creates prize puzzle in emerging economies.
Accepted in case of Brazil, Colombia, Czech Republic, Hungary, Philippines, Poland.
Source: Author’s compilation
1.2 Findings of Second Objective
One of the key outcomes of the recent global crisis is that due to financial
innovations we are unable to capture the broader horizon of financial conditions
with just few variables. Policymakers, regulators, market participants and
researchers have affirmed this conjunction and have emphasis to work on this
part for enhancing our level of understanding on this part. Keeping this view in
170
front, another objective of this study is to offer an empirical assessment of the
effects of the financial conditions of the United States upon macro-economy of
the emerging economies using standard Vector Auto-Regression (VAR) Models.
This objective is achieved by utilizing financial conditions index of Brave and
Butter (2011) for the assessment of impact upon macro-economy of the emerging
markets as being classified by Financial Times Stock Exchange (FTSE). It is being
found that macro-economic variables do respond on the financial conditions of
the united stated however magnitude varies from country to country. This
response is of short-term nature rather of long term. Status of the hypotheses is
given in table 6.2.
Table: VIII-2: Status of Hypotheses of Second Objective
S# Hypotheses Status
1. Dornbusch’s exchange rate overshooting hypothesis exists in emerging markets in transmission mechanism
Accepted in case of Brazil, Chile, Hungary, India, Mexico, Pakistan, Poland, Russian Federation, South Africa.
2. FCI reflects information of stock market in the long run.
Accepted in case of Chile, Czech republic, Greece, Hungary, India, Malaysia, Pakistan, Poland, Russian Federation, and South Africa.
3 FCI reflects information of short-term interest rate in the short run.
Accepted only in case of Brazil
4 FCI reflects information of long-term interest rate in the long run.
Accepted in case of Chile, Czech republic, Greece, Hungary, India, Malaysia, Poland, and Russian Federation.
Source: Author’s compilation
171
1.3 Findings of Third Objective
One of the learning lesson of recent mortgage crisis is that broad financial
conditions due to innovations in financial landscape is difficult to capture by using
small number of variables that cover only few traditional financial markets. In the
light of recent crisis policy makers, regulators and financial market stakeholders
has affirmed the link between traditional and newly developed markets and link
between financial and non-financial market. Closer watch on financial stability is
essential for understanding such links. For this reason, index on financial
conditions are the best for this purpose. With this background, in this study, an
index is created using wide range of macro-economic and financial variables over
a long horizon for the Pakistan using a time varying model developed by Koop
and Korobilis (2014). This method develops and forecasts financial conditions
index.
Overall, significant periods of economic growth and crisis in financial
history are well captured by the index. By looking at different statistics of the
dynamic forecast, it can be derived that fit is good and graph of the forecasts
closely follows financial conditions, indicating that this index is having strong
predictive power of the major macro-economic variables. This study proposes
new dimension for the policy studies and the constructed index in this study may
help regulators, policy makers and scholars for finding the true state of the
financial conditions. Index is able to closely mark major episodes in the financial
history of the Pakistani economy, particularly those characterized by large
172
external financial and economic shocks. Status of the hypotheses is given in table
6.3.
Table: VIII-3: Status of Hypotheses of Third Objective
S# Hypothesis Status
1. FCI helps in measuring financial shocks. Accepted
Source: Author’s compilation
This chapter discuss result objective wise and present comprehensively
conclusion on each objective.
173
CHAPTER IX
RECOMMENDATIONS
This is the last chapter of the study. This will be discussing
recommendations, implications and limitations of the study.
1. Policy Recommendation
This study discusses in length case of transmission and contagion arising
from US. This study specifically follows salt-water school of economics and
majorly employ ISLM framework for studying the impact of monetary policy.
Moreover, for studying the impact of financial conditions, an index (representative
of US economy) and second index (created for the Pakistan) both have been
employed for studying macroeconomic response. As a result of this exercise and
findings of the study, following are the recommendations of this study.
Table IX-1: Findings and Recommendations
Findings Recommendations ISLM framework is partly applicable in many cases.
Monetary policy objectives need to revise. Revision of monetary objectives in accordance of time is need of time.
prize puzzle exist more in county specific monetary policy as compared to international contagion
Contagion arising from international link may be studied under the framework of ISLM.
Response of output is aligning with the theory at country level but not in international level.
Transmission arising from the monetary policy shed light on the output response in the light of ISLM framework. So, specifically in this area, ISLM can be employed.
Hungary, Turkey, Malaysia at country level fully align with the theory
These countries may be studied and policies in these countries may be designed using ISLM Framework
174
Mexico and Turkey fully deviate from theory in the case of contagion impact
These both countries should not be studied under ISLM framework in studying contagion impact. The reason of this complete deviation is known. Use of another salt water theory may define the behavior of these countries
Countries do have impact of financial conditions of the US but in many cases, Impact die off with the time.
Financial conditions of the US should be taken into consideration while designing policies specifically short-term policies.
Financial conditions of the US do have implications for the emerging countries. Extend vary from country to country but impact does exist on the macro-economy of the emerging countries.
In globalization, consideration of impact of other economies will help deals better with the disaster situations. This could be achieved by the Inclusion of financialization in decision-making.
The countries who are having Free Trade Agreement with the US are having strong and long term response
Strong international relation with any country is having its own merits and demerits. Inclusion of any clause or any such component that would act as buffer in case of disaster could help in this matter.
Bilateral partners' response dies off with the passage of time except Russia and Hungary (These both countries shifted from planned economy towards capital economy and both were badly hit from the crisis.
Initiatives are needed to eliminate the dominancy of foreign capital economies for lessen the impact of crisis. It is being found that the countries that were stick to their currency faced low extend of the crisis.
Interest rates (both short and long term) are more responsive towards financial conditions of the US
It is proven that macro-economy respond, but Interest rate responded more. So while designing interest rates, consideration of financialisation and globalization may result more realistic rates.
Financial condition index of the Pakistan is able to give a true picture of the economy.
Decision making process of the authorities should take input from Multi-factors. Decision based on single or few variables results in biased results.
forecasting of the index of the macroeconomic variables is close to the reality.
use of index may give better insight in the study of macroeconomic fluctuations.
Source: Author’s compilation
175
2. Implications
Implications of this study may be divided into two groups as this study addresses
transmission upon macro-economy from two areas namely financial condition index and
monetary policy.
2.1 Implications of Financial Condition index
During last few decades, we may see enormous growth in financial
landscape. Due to surge in financial engineering and innovation, Investors are
having more option to avail for efficient allocation of funds than ever before. But
this blessing has also its cost with it. It’s a challenge for the market and economy
stakeholder as how to deal with innovations that operate at global level.
As the results of recent crisis that is seen at global level, we may say that
more closely watched instrument is required for having mitigating plans.
Policy makers for designing more robust and realistic market plans and
mitigating strategies can use Index. Moreover, recent trend led by Basel III in
terms of regulation and supervision is to work in more systematic and macro-
prudential framework. So it can be say that a tool like FCI can serve this purpose
well and can be considered a step towards this also.
FCI is able to mimic the behavior of the economy in a more robust way. So
it can be employed as a tool to study the economy in a more realistic way.
2.2 Implications of Monetary Policy
Effectiveness of monetary policy has been under question from decades
especially under crisis. Along with it, its possible contagion is also under question.
There is a great deal of available literature covering both aspects. Thus
highlighting its importance. Timely information on monetary policy transmission is
176
useful for the policy makers and regulators for developing relevant plans.
By these objectives in hand, key contribution of this thesis is to study the
adverse impact of shocks namely monetary and financial on economy and
possible contagion arising from the origin of crisis towards emerging economies.
Results indicate that impact of shock do exist both at national level and
international level in the form of contagion. It’s also found that Financial condition
index may serve as early warning indicator in both situations.
This study shed light on monetary policy transmission at national level and
also at international level in a time varying nature meaning transmission is not
static in nature. Monetary policy transmitted in a time varying way so monetary
objectives should be designed accordingly keeping this fact in front.
Moreover, it is found that transmission process is not aligned with the
theory. Certain deviation exists from theory in transmission mechanism.
Theoretical background of this study is ISLM framework. It was found that it is
partly applicable to the most of the economies. So monetary authorities should
consider this finding while designing their policies.
Moreover, this study does have implications and address following
Sustainable Development Goal (SDGs). Sustainable development goals are set
by United Nations for the better future. This project will address Goal number eight
that state “Promote sustained, inclusive and sustainable economic growth, full
and productive employment and decent work for all” in general and specifically its
8.10 part that state “Strengthen the capacity of domestic financial institutions to
encourage and expand access to banking, insurance and financial services for
all” (United Nation General Assembly, 2014).
177
3. Limitations and Future Work
This study is a comprehensive in nature in studying the shock in emerging
markets and it has tried to cover many aspects. But no study is perfect. It is true
for this study. This study is being limited by the availability of data. Limit the
number of countries, data time span, shorter time span; data was available on
different frequencies that also limit us with the application and interpolation.
Future direction of this study can be taken by using following means:
• Study on monetary policy rules and its implications for the economy;
• FCI with more important variable e.g. CPEC, Small and Medium
Enterprises (SMEs) sector;
• High frequency FCI;
• Regional and international comparison of FCI;
• Study on different crisis and comparison.
So its plan to work on these areas in future for better study of subject under
study subject to the removal of limitations. Moreover, a good index is one that
keep on updating with time as variables impact keep on changing. It is true for the
FCI. The FCI cannot always precisely identify the direction of the evolutionof
economic activity, as GDP growth can also be affected by exogenous factors (for
example productivity shocks). Moreover, the good indicator properties of the index
are conditional on the stability of the association between real and financial
variables and of the allocated weights – therefore, given their changing nature,
the index needs to be periodically reassessed so as to capture the most relevant
combination of indicators, both in terms of their selection and of the assigned
178
weights, as the relationships between the variables change. In this regard,
estimation using Dynamic Model Averaging techniques can be quite helpful.
With this chapter, this study comes to an end formally. This chapter
discusses policy recommendations, implications and limitations of the study.
179
REFERENCES
ABDEL-SALAM, N. (2017), “Diverging Migration Tendencies of the Youth in Central Europe Case Studies of Hungary and the Czech Republic”, Master Thesis, Charles University in Prague, Faculty of Arts, Prague.
ACEMOGLU, D., AND UCER, M. (2015), “The Ups and Downs of Turkish Growth,
2002-2015: Political Dynamics, the European Union and the Institutional Slide”.
AGHA, A. I., AHMED, N., MUBARIK, Y. A., & SHAH, H. (2005), “Transmission
Mechanism Of Monetary Policy In Pakistan”, SBP-Research Bulletin, 1(1), pp. 1-23.
ALEEM, A., AND LAHIANI, A. (2011), “Monetary policy rules for a developing
country: Evidence from Pakistan”, Journal of Asian Economies, vol. no. 22, pp. 483-494.
ALESSANDRIA, P., AND MUMTAZ, H. (2017), “Financial conditions and density
forecasts for US output and inflation”, (Elsevier, Ed.) Review of Economic Dynamics, vol. no. 24, pp. 66-78.
ALLEN, F., AND GALE, D. (2000, February), “Financial Contagion”, The Journal
Of Political Economy, pp. 1-33. AMANN, E., AND BAER, W. (2012), “Brazil as an Emerging Economy: A New
Economic Miracle?”, Brazilian Journal of Political Economy, vol. no. 32(3(128)), pp. 412-423.
ANG, JAMES B. (2007), “A Survey of Recent Developments in the Literature of
Finance and Growth”, Discussion Paper, Monash University, Department of Economics.
ANGELOPOULOU, E., BALFOUSSIA, H., AND GIBSON, H. (2013), “Building a
Financial conditions index for the euro area and Selected euro area countries What does it tell us about the crisis?”, Working Paper, European Central Bank, macro Prudential research network, Frankfurt am Main.
AREOSA, W. P., AND DUTRA, W. (2016, May), “Financial Condition Indicators
for Brazil”, Working Paper, Central Bank of Brazil. ARIAS, J. E., CALDARA, D., RUBIO-RAMÍREZ , J.F., (2019), “The systematic
component of monetary policy in SVARs: An agnostic identification procedure”, Journal of Monetary Economics, vol. no. 101, pp. 1-13.
180
ARGUDEN, Y. (2007), “An Overview of the Turkish Economy: Outlook and Current Perspectives”, Policy Brief, Middle East Institute.
BACCHETTA, P., AND BALLABRIGA, F. (2000), “The Impact of Monetary Policy
and Banks' Balance Sheets: Some International Evidence”, Applied Financial Economics (10), pp. 15-26.
BAGEHOT, WALLET (1873), “Lombard Street: A Description of the Money
Market”, Great Britain: Henry S. King & Co. BAGLIANO, F., AND MORANA, C. (2010), “The Great Recession: US Dynamics
And Spillovers To The World Economy”, Working Paper, Center For Research On Pension And Welfare Policies.
BALAKRISHNAN, R., DANNINGER, S., ELEKDAG, S., AND TYTELL, I. (2009,
June), “The Transmission of Financial Stress from Advanced to Emerging Economies”, IMF Working Paper, 3.
BALCILARA, M., THOMPSON, K., GUPTAB, R., AND EYDEN, R. V. (2016),
“Testing the asymmetric effects of financial conditions in South Africa: A nonlinear vector autoregression approach”, Journal of International Financial Markets, Institutions & Money.
BARAKCHIAN, S. MAHDI (2015), "Transmission of US monetary policy into the
Canadian economy: A structural cointegration analysis." Economic Modelling 46: pp. 11-26.
BARSKY, R., AND SIMS, E. (2011), “News shocks and business cycles”, Journal of Monetary Economics, pp. 273-289.
BARSKY, ROBERT B; SIMS, ERIC R. (2012), “Information, Animal Spirits, and
the Meaning of Innovations in Consumer Confidence”, American Economic Review, 102(4), pp. 1343-1377.
BEATON, K., LALONDE, R., AND LUU, C. (2009), “A Financial Conditions Index
for the United States”, Discussion Paper, Bank of Canada, International Economic Analysis Department, Ontario.
BELONGIA, M. T., AND IRELAND, P. N. (2006), “The Own-Price of Money and
the Channels of Monetary Transmission”, Journal of Money, Credit, and Banking, 38(2), pp. 429-445.
BERNANKE, B. S. (1983), “Irreversibility, uncertainty, and cyclical adjustment”,
The Quarterly Journal of Economics, vol. no. 98(1), pp. 85-106.
181
BERNANKE, B. S., AND BLINDER, A. S. (1992, September), “The Federal Funds Rate and the Channels of Monetary Transmission”, American Economic Review (82), pp. 901-921.
BERNANKE, B., AND GERTLER, M. (1995), “Inside the Black Box: The Credit
Channel of Monetary Policy Transmission”, Journal of Economic Perspectives , vol. no. 9(4), pp. 27-48.
BERNANKE., B., AND MIHOV, I. (1998), “Measuring Monetary Policy’, Quarterly
Journal of Economics, vol. no. 113(3), pp. 869-902. BLACK, F. (1976), “Studies of Stock Market Volatility Changes”, Proceedings of
the American Statistical Association, pp. 177-181. BLANCHARD, O. J., AND PEROTTI, R. (2002), “An Empirical Characterization of
the Dynamic Effects of Changes in Government Spending and Taxes on Output”, Quarterly Journal of Economics, vol. no. 117, pp. 1329-1368.
BLANCHARD, O. J., AND QUAH, D. (1989), “The Dynamic Effects of Aggregate
Demand and Supply Disturbances”, American Economic Review, vol. no. 79, pp. 655-673.
BLANCHARD, OLIVIER J., HAMID FARUQEE, AND MITALI DAS. (2010), "The
Initial Impact of the Crisis on Emerging Market Countries." Brookings Papers on Economic Activity, pp. 263-323.
BEST, G. (2013), “Fear Of Floating Or Monetary Policy As Usual? A Structural
Analysis Of Mexico’s Monetary Policy”, North American Journal Of Economics And Finance , vol. no. 24, pp. 45-62.
BLINDER, A. S. (2010), “How Central Should The Central Bank Be?”, Journal Of Economic Literature, vol. no. 48(1), pp. 123-133.
BLOOM, N., BOND, S., AND REENEN, J. V. (2007), “Uncertainty and Investment Dynamics”, Review of Economic Studies, vol. no. 74(2), pp. 391-415.
BLUEDORN, J., AND BOWDLER, C. (2011), “The Open Economy
Consequences Of U.S. Monetary Policy”, Journal Of International Monetary Finance, vol. no. 30, pp. 309-336.
BOIVIN, J., AND GIANNONI, M. P. (2006, August), “Has Monetary Policy become more Effective?”, The Review of Economics and Statistics, vol. no. 88(3), pp. 445–462.
BORIO, C. (2008). Policy Analysis, pp. 251. BOUIS, R., RAWDANOWICZ, L., RENNE, J.-P., WATANABE, S., AND
CHRISTENSEN, A. K. (2013), “The effectiveness of monetary policy since the onset of the financial crisis”, OECD Economics Department Working Paper No. 1081.
182
BRAVE, S., AND BUTTERS, R. A. (2010), “Gathering Insights on the Forest from the Trees: A New Metric for Financial Conditions:, Chicago: Federal Reserve Bank of Chicago.
BUCH, C. M., BUCHHOLZ, M., AND TONZER, L. (2014), “Uncertainty and
international banking international Banking: Micro foundations and Macroeconomic Implications”, International Monetary Fund (IMF) and the Dutch National Bank (DNB).
CALVO, G. A., IZQUIERDO, A., AND MEJA, L.-F. (2008), “Systemic Sudden
Stops: The Relevance Of Balance-Sheet Effects And Financial Integration”, Massachusetts: National Bureau of Economic Research.
CARMASSI, JACOPO, DANIEL GROS, AND STEFANO MICOSSI (2009), "The
Global Financial Crisis: Causes and Cures", Journal of Common Market Studies 47: pp. 977–996.
CHRISTIANO, L., AND EICHENBAUM, M. (1992), “Identification and the Liquidity
effect of Monetary Policy Shocks”, Political Economy, Growth, and Business Cycles, pp. 335-370.
CICCARELLI, M., MADDALONI, A., & PEYDRO, J.-L. (2013), “Heterogeneous transmission mechanism monetary Policy and financial fragility in the euro area”, Frankfurt am Main: European Central Bank.
CLAESSENS, S., DORNBUSCH, R., AND PARK, Y. (2001), “Contagion: why
crises spread and how this can be stopped, in International Financial Contagion. Boston: S. Clasessens and K”, Forbes, Kluwer Academic Publishers.
COGLEY, T., AND SARGENT, T. J. (2001), “Evolving Post-World War II U.S.
Inflation Dynamics”, NBER Macroeconomics Annual, vol. no. 16, pp. 331–373.
COOPER, W. H. (2009), “Russia’s Economic Performance and Policies and Their
Implications for the United States”, Congressional Research Service. CROCKETT, A. (1997), “Why is Financial Stability a Goal of Public Policy?”, FRB
Economic Review, Q IV, pp. 5-22. CROSS, J., AND NGUYEN, B. H. (2016), “The Relationship Between Global Oil
Price Shocks And China’s Output: A Time-Varying Analysis”, Journal of Energy Economics.
CSIZMADIA, L. (2008), “The Transition Economy of Hungary between 1990 and 2004”, MSc. in Finance and International Business, University of Aarhus, Aarhus School of Business.
183
DEDOLA, L., AND LIPPI, F. (2005), “The monetary transmission mechanism:
Evidence from the industries of five OECD countries”, European Economic Review, vol. no. 49(6), pp. 1543–1569.
DIXIT, A., AND PINDYCK, R. (1994), “Investment under Uncertainty”, Princeton
University Press. DORNBUSCH, R. (1976, December), “Expectations and Exchange Rate
Dynamics”, Journal of Political Economy, vol. no. 84, pp. 1161-1176.
DURAN, M., ÖZCANB, G., OZLU, P., AND ÜNALMIS, D. (2012), “Measuring the impact of monetary policy on asset prices in Turkey”, Economic Letters, vol. no. 114, pp. 29–31.
EHRMANN, M., AND FRATZSCHER, M. (2009), “Global Financial Transmission of Monetary Policy Shocks”, Oxford Bulletin of Economics And Statistics, vol. no. 71(6), pp. 739-759.
EICHENGREEN, B., AND LINDERT, P. H. (1992), “The International Debt Crisis in Historical Perspective”, Cambridge: MA: MIT Press.
EICKMEIER, S., LEMKE, W., AND MARCELLINO, M. (2011), “The changing
international transmission of financial shocks: evidence from a classical time-varying FAVAR”, Discussion Paper, Deutsche Bundesbank.
ELLIOTT, M., GOLUB, B., AND JACKSON, M. O. (2014), “Financial Networks
And Contagion”, American Economic Review Journal, vol. no. 104(10), pp. 3115-3153.
ERDEM, M., AND TSATSARONIS, K. (2013, March), “Financial conditions and economic activity: a statistical approach”, Quarterly Review, pp. 37-51.
“Executive Office of the President, Countries and Regions”, Retrieved 2018, from
Office of the United States Trade Representative: https://ustr.gov/countries-regions.
“Executive Office of the President. Free Trade Agreements”, Retrieved 11 14, 2018, from United States Trade Representative: https://ustr.gov/trade-agreements/free-trade-agreements.
FATIMA, K. (2013), “Globalization, inflation and monetary policy”, Thesis,
University of Glasgow, Adam Smith Business School, Glasgow. FEDERAL REGISTER (2009, January 22), The President. Vol. no.74, pp. 13.
184
FORNARI, FABIO, AND LIVIO STRACCA (2013), “What does a Financial Shock do? First International Evidence”, Working Paper, Macroprudential Research Network, European Central Bank.
FREEDMAN, C. (1994), “The use of indicators and the monetary conditions index
in Canada”, (T. B. Cottarelli, Ed.) Frameworks for monetary stability. FRIEDMAN, M., AND ANNA, J. S. (1963), “A Monetary History of United States”,
pp. 1867–1960. FTSE RUSSELL (2016), “FTSE Annual Country Classification Review”. FTSE RUSSELL (2015), “FTSE country classification process”, Research Report. FU, Q., AND LIU, X. (2015), “Monetary policy and dynamic adjustment of
corporate investment: A policy transmission channel perspective”, China Journal of Accounting Research, vol. no. 8, pp. 91-109.
FURCERI, D., PRAKASH LOUNGANI, P., ZDZIENICKA, A. (2018), “The effects
of monetary policy shocks on inequality”, Journal of International Money and Finance, vol. no. 85, pp. 168-186.
FUNG, B. S., AND KOSUMOVICH, M. (1998), “Monetary Shocks in G-6 Countries
is There a Puzzle?”, Journal of Monetary Economics, vol. no. 42(3), pp. 575-592.
GAI, P. (2013), “The Robust Yet Fragile Nature Of Financial Systems. In P. Gai,
Systemic Risk”, United Kingdom: Oxford University Press, vol. no. 1, pp. 9-28.
GAGLIANONE, W. P., AND AREOSA, W. D. (2016), “Financial Conditions Indicators for Brazil”, Central bank of Brazil.
GALI, J. (1999), “Technology, Employment, and the Business Cycle: Do Technology Shocks Explain Aggregate Fluctuations?”, American Economic Review, vol. no.89, pp. 249-271.
GAUTHIER, C., CHRISTOPHER, G., AND LIU, Y. (2004), “Financial Conditions Indexes for Canada”, Bank of Canada Working Paper, pp. 2004-22.
GHANI, SHAZIA (2013), “Global Financial Crisis of 2007 Analysis of Origin and
Assessment of Contagion to Emerging Economies Lessons and Challenges for Financial Regulation”, PhD Thesis, Grenoble: University of Grenoble.
GLASSERMAN, P., AND YOUNG, H. P. (2013), “How Likely Is Contagion In
Financial Networks?”, Discussion Paper Series, University Of Oxford, Department Of Economics.
185
GOLDSMITH, R. W. (1969), “Financial structure and development”, New Haven, CT: Yale University Press.
GOLDSTEIN, M. (2005), “What Might The Next Emerging-Market Financial Crisis
Look Like?”, Working Paper, Institute For International Economics, Washington, DC.
GOMEZ, E., MURCIA, A., AND ZAMUDIO, N. (2011), “Financial Conditions index:
early and leading indicator for Colombia”, Essays on Economic Policy, vol. no. 29(66), pp. 177-220.
GONZALES, M. D., AND BAUTISTA, M. S. (2013), “Financial Conditions Indexes
for Asian Economies”, Asian Development Bank. ADB Economics. GOODFRIEND, M. (1991), “Interest Rates and the Conduct of Monetary Policy”,
Carnegie- Rochester Conference Series(Spring), pp. 7-30. GOODHART, C., AND HOFMANN, B. (2001, march 2-3), “Asset Prices, Financial
Condition of Transmission of Monetary Policy”, Conference on Assets Prices, Exchange Rates, and Monetary Policy. Stanford University.
GORDON, ROBERT J. (2003), “Productivity Growth, Inflation, and
Unemployment”, Cambridge University Press, pp. 226–227. GUICHARD, S., AND TURNER, D. (2008), “Quantifying the Effect of Financial
Conditions on US Activity”, Economics Department. OECD Economics Department.
GUJARATI, D. N., AND PORTER, D. C. (2009), “Basic Econometrics (Fifth ed.)”,
New York: McGraw-Hill Companies. GUMATA, N., KLEIN, N., AND NDOU, E. (2012), “A Financial Conditions Index
for South Africa”, Working Paper, International Monetary Fund, African Department.
HANIF, M. N., IQBAL, J., AND MALIK, M. J. (2013), “Quaternization of National
Income Accounts of Pakistan”, SBP Research Bulletin, vol. no. 9(1). HANSON, M. S. (2004), “The “Price Puzzle” Reconsidered”, Journal of Monetary
Economics, pp. 1-38. HAB, L. H., KOZIOL, C., & SCHWEIZER, D. (2014), “What Drives Contagion in
Financial Markets? Liquidity Effects versus Information Spill‐Over”, European Financial Management, vol. no. 20(3), pp. 548-573.
186
HATZIUS, J., HOOPER, P., MISHKIN, F., SCHOENHOLTZ, K., AND WATSON, M. (2010), “Financial Conditions Indexes: A Fresh Look After the Financial Crisis”, NBER Working Paper Series.
HAVLAT, M., HAVRLANT, D., KUENZEL, R., AND MONKS, A. (2018), “Economic
Convergence in the Czech Republic and Slovakia”. Economic Brief, European Commission, Directorate-General for Economic and Financial Affairs.
HAYS, J. (2008), “Economic History of the Philippines”, Retrieved from
http://factsanddetails.com/southeast-asia/Philippines/sub5_6g/entry-3916.html.
JANNSEN, N., POTJAGAILO, G., AND WOLTERS, M. H. (2015), “Monetary
Policy during Financial Crises: Is the Transmission Mechanism Impaired?”, Working Paper, Kiel Institute for the World Economy, Kiel.
JAVID, M., AND MUNIR, K. (2010), “The Price Puzzle and Monetary Policy
Transmission Mechanism in Pakistan: Structural Vector Autoregressive Approach”, The Pakistan Development Review, pp. 449–460.
JERMANN, U., AND QUADRINI, V. (2012), “Macroeconomic effects of financial
shocks”, The American Economic Review, vol. no.102(1), pp. 238–271. JIMENEZ, G., ONGENA, S., PEYDRO, J.-L., AND SAURINA, J. (2014),
“Hazardous times for monetary policy: What do twenty-three million bank loans say about the effects of monetary policy on credit risk-taking?”, J. Econometrica vol.no. 82(2), pp. 463–505.
KAZI, I. A., WAGAN, H., AND AKBAR, F. (2013). “The changing international
transmission of U.S. monetary policy shocks: Is there evidence of contagion effect on OECD countries”, Economic Modelling, vol. no. 30(6), pp. 90–116.
KILIAN, L. (1998), “Small Sample Confidence Intervals for Impulse Response Functions”, The Review of Economics and Statistics, pp. 218-230.
KIM, S. (2001), “International Transmission Of U.S. Monetary Policy Shocks:
Evidence From VAR'S”, Journal Of Monetary Economics, vol. no.48, pp. 339–372.
KING, R. G., AND R. LEVINE. (1993), "Finance and growth: Schumpeter Might Be Right", Quarterly Journal of Economics, vol. no. 108: pp. 717-737.
KLACSO, JAN. (2013), "Mathematical Analysis of the Transmission Mechanism of Monetary Policy of the National Bank of Slovakia." Ph. D. Thesis, Department of Applied Mathematics and Statistics, Comenius University, Slovakia.
187
KOEN, V., ASADA, H., NIXON, S., RAHUMAN, M. R., AND ARIF, A. Z. (2017), “Malaysia's Economic Success Story and Challenges”, Working Paper, Organization for Economic Co-operation and Development.
KOERNER, J. (2015), “Monetary Transmission in the Czech Republic after the
Transformation”, Eastern European Business and Economics Journal, vol. no. 1(3), pp.19-47.
KOOP, G. (2003), “Bayesian Econometrics”.Chichester: John Wiley & Sons Ltd. KOOP, G., AND KOROBILIS, D. (2014), “A new index of financial conditions”,
European Economic Review, vol. no. 71(C), pp. 101-116. KSEL, E. Y., METIN-OZCAN, K., AND HATIPOGLU, O. (2013), “A survey on time-
varying parameter Taylor rule: A model modified with interest rate pass-through”, J. Economic Systems, vol. no. 37, pp. 122-134.
LAURENT, R. D. (1988), “An Interest Rate Based Indicator of Monetary Policy”,
J. Economic Perspective, pp. 3-14. LUPORINI, V. (2008), “The Monetary Transmission Mechanism in Brazil:
Evidence from a VAR Analysis”, J. EST. ECON., SAO PAULO, vol. no. 38(1), pp. 7-30.
MARFATIA, H. A. (2015), “Monetary policy’s time-varying impact on the US bond
markets: Role of financial stress and risks”, North American Journal of Economics and Finance, vol. no. 34, pp.103-123.
MARKWAT, T., KOLE, E., AND DIJK, D. V. (2009), “Contagion As Domino Effect
In Global Stock Markets”, ERIM Report Series Research in Management, Rotterdam.
MARTINELLI, C., AND VEGA, M. (2018),”Monetary and Fiscal History of Peru 1960-2010: Radical Policy Experiments, Inflation and Stabilization”, George Mason University, Interdisciplinary Center for Economic Science.
MATHESON, T. (2011), “Financial Conditions Indexes for the United States and
Euro Area”, International Monetary Fund. MAYES, D. G., AND VIREN, M. (2001), “Financial condition Index”, Bank of
Finland, Discussion Paper, 17/01. MCKINNON, R. I. (1973), “Money and Capital in Economic Development”,
Washington, DC: Brookings Institution. MEENAI, S. A. (2010), “Money and Banking in Pakistan’, (J. A. Ansari, Ed.)
Karachi: Oxford University Press.
188
MORGAN, D. P. (1993), “Asymmetric effects of monetary policy”, vol. no. 78, pp.
21-33. MURARU, A. (2015), “Building a Financial Conditions Index for Romania”,
Occasional Paper, National Bank of Romania. NAKAJIMA, J. (2011, November), “Time-Varying Parameter VAR Model with
Stochastic Volatility: An Overview of Methodology and Empirical Applications”, Monetary and Economic Studies Journal, pp.107-142.
NELSON, R. M., BELKIN, P., & MIX, D. E. (2011), “Greece’s Debt Crisis:
Overview, Policy Responses, and Implications”, Congressional Research Service.
NERI, S., AND NOBILI, A. (2010), “The Transmission Of US Monetary Policy To
The Euro Area”, International Finance Journal, vol. no. 13(1), pp. 55–78. OECD. (2017), “OECD Economic Surveys Mexico”. OECD. (2018), “OECD Economic Surveys Poland”. OLINER, S. D., AND RUDEBUSCH, G. D. (1996), “Is There Broad Credit Channel
of Monetary Policy?”. OLMO, J., AND SANSO-NAVARRO, M. (2015, August), “Changes in the
transmission of monetary policy during crisis episodes: Evidence from the euro area and the U.S. Economic Modelling”, vol. no.48, pp. 155-166.
ONO, S. (2013), “The effects of foreign exchange and monetary policies in Russia”, Economic Systems Journal , vol. no. 37, pp. 522-541. OPSCHOOR, A., DIJK, D. V., AND WEL, M. V. (2014, December), “Predicting
Volatility and Correlations with Financial Conditions Indexes”, Journal of Empirical Finance, vol. no. 29, pp. 435-447.
OWYANG, M. T., AND WALL, H. J. (2009), “Regional VARs and the Channels of
Monetary Policy”, Applied Economics Letters, vol. no. 16(12), pp. 1191–1194.
PEERSMAN, G. (2004), “The transmission of monetary policy in the euro area:
are the effects different across countries?”, Oxford Bulletin of Economics and Statistics, vol. no. 66(3), pp. 285–308.
PESARAN, M. H., AND SHIN, Y. (1998), “Generalized Impulse Response
Analysis in Linear Multivariate Models”, Economics Letters, vol. no. 58(1), pp. 17–29.
189
PHIROMSWAD, P. (2015), “Measuring Monetary Policy with Empirically Grounded Restrictions: An Application to Thailand”, Journal of Asian Economies, vol. no. 38, pp. 104-113.
POTTS, T., AND YERGER, D. (2010), “Variations across Canadian regions in the
sensitivity to U.S. monetary policy”, Atlantic Economic Journal, vol. no. 38, pp. 443–454.
PRIMICERI, G. E. (2005), “Time varying structural vector auto regressions and
monetary policy”, The Review of Economic Studies, vol. no. 72(3), pp. 821–852.
RAJAN, R. G. (2005), “Has financial development made the world riskier?”,
Proceedings-Economic Policy Symposium, pp. 313–369. Kansas: Jackson Hole, Federal Reserve Bank of Kansas City.
REINHART, C. M., AND ROGOFF, K. S. (2008), “Is the 2007 US sub-prime
financial crisis so different? an international historical comparison”, American Economic Review, vol. no. 98(2), pp. 339-344.
REINHART, C., AND ROGOFF, K. (2010), “This time is different: Eight Centuries
of Financial Folly”, Princeton University Press. RIGOBON, R., AND SACK, B. (2004), “The impact of monetary policy on asset
prices”, Journal of Monetary Economics, vol. no. 51, pp. 1553–1575. RODOLFO CERMENO, F. A., AND POLO, J. O. (2012), “Monetary Policy Rules
in A Small Open Economy: An Application to Mexico”, Journal of Applied Economics, vol. no. 15(2), pp. 259-286.
JOHN H. ROGERS, J. H., SCOTTI, C., WRIGHT, J. H., (2018), “Unconventional
Monetary Policy and International Risk Premia”, Journal of Money, Credit and Banking, vol. no. 50(8), pp. 1827-1850.
ROGOFF. (2002), “Dornbusch’s Overshooting Model after Twenty-Five Years”,
Retrieved from IMF.Org: http://www.imf.org/external/np/speeches/2001/112901.html.
SAEED, K. A. (2013), “The Economy of Pakistan. Karachi”, Oxford University
Press. SERBAN, E., BORISOV, D., AND DOBREA, R. (2012), “Emerging Markets:
Concept Evolution and Business Environment Particularities”, Management Research and Practice Journal, vol. no. 4(2), pp. 5-19.
SICHEI, G. A. (2005), “An Econometric Model of the Rand-US Dollar Nominal
Exchange Rate”, Retrieved from Web.up.ac.za.:
190
http://www.up.ac.za/dspace/bitstream/2263/4509/1/Sichei_Econometric%282005%29.pdf.
SIMS, C. A. (1980), “Macroeconomics and Reality”, J. Econometrica, vol. no. 48,
pp. 1-48. SIMS, C. A. (1992, June), “Interpreting the Macroeconomic Time Series Facts:
The Effects of Monetary Policy”, European Economic Review (36), pp. 975-1000.
SIMS, C. A., AND ZHA, T. (2006), “Does Monetary Policy Generate
Recessions?”, Journal Macroeconomic Dynamics (second), pp. 231-272. STEINER, R., AND VALLEJO, H. (2010), "The Economy. In Colombia: A Country
Study”, (Rex A. Hudson ed.). Library of Congress Federal Research Division.
STOCK, J. H. (2001), “Discussion of Sargent and Cogley’s ‘Evolving Post-World
War II U.S. Inflation Dynamics”, NBER Macroeconomics Annual, vol. no. 16, pp. 379-387.
STOCK, J. H., AND WATSON, M. W. (2002, April), “Macroeconomic Forecasting
using Diffusion Indexes”, Journal of Business and Economic Statistics, vol. no. 20(2), pp. 147-162.
STOCK, J., AND WATSON, M. (2003), “Forecasting output and inflation: the role
of asset prices”, Journal of Economic Literature, vol. no. 41(3), pp. 788–829.
SWISTON, A. (2008), “A US Financial Conditions Index: Putting Credit Where
Credit is Due”, IMF Working Paper. TESOGLU, H. S. (2004), “Monetary transmission-federal funds rate and prime
rate”, Journal of Post Keynesian Economics, vol. no. 26(2), pp. 357-362. TODOROV, GALIN KOSTADINOV. (2012), “A Study of Stock Market Linkages
between the US and Frontier Markets”, PhD Thesis, Florida International University.
TU, W., AND FENG, J. (2009, February), “An Overview Study on Dornbusch
Overshooting Hypothesis”, International Journal of Economics and Finance, vol. no. 1(1), pp. 110-116.
UNITED NATION GENERAL ASSEMBLY (2014), “Open Working Group of the
General Assembly on Sustainable Development Goals”, United Nation.
191
VALENCIA, F. (2013), “Aggregate uncertainty and the supply of credit”, IMF Working Paper 13/241.
VAVRA, J. (2014), “Inflation dynamics and time-varying volatility: New evidence
and an SS interpretation”, Quarterly Journal of Economics, vol. no. 129(1), pp. 215-258.
W., D., AND HATZIUS, J. (2000), “The Goldman Sachs Financial Conditions
Index: The Right Tool for a New Monetary Policy Regime”, Global Economics Paper, pp. 44.
WAGAN, H., AND ALI, Z. (2014), “Are Emerging Markets Exposed to Contagion
from the United States: Evidence from Stock and Sovereign Bond Markets”, In Mohammed El Hedi Arouri, S. Boubaker, and D. K. Nguyen (Eds.), Emerging Markets and the Global Economy pp. 509-536, Academic Press.
WOLF, M. (2012, May 1), “The Financial Times”. WORLD ECONOMIC FORUM (2009), “The Global Competitiveness Report”. Y., A., GAMBACORTA, L., AND MARQUES-IBANEZ, D. (2014, March), “Does
monetary policy affect bank risk?”, International Journal of Central Banking, pp. 95-135.
YIU, MATTHEW S., WAI-YIP ALEX HO, AND DANIEL F. CHOIC. (2010),
"Dynamic correlation analysis of financial contagion in Asian markets in global financial turmoil." Applied Financial Economics Journal, vol. no. 20, pp. 345-354.
VONNÁK, B. (2005), “Estimating the effect of Hungarian monetary policy within a
structural VAR framework”, Working Paper. YUKSEL, E., METIN-OZCAN, K., AND HATIPOGLU, O. (2013), “A survey on
time-varying parameter Taylor rule: A model modified with interest rate pass-through”, Economic Systems Journal, vol. no. 37, pp. 122-134.
ZAIDI, S. A. (2005), “Issues in Pakistan's Economy”, Karachi: Oxford University
Press. ZAKIR, N., AND MALIK, W. S. (2013), “Are the effects of monetary policy on
output asymmetric in Pakistan?”, Economic Modelling Journal, vol. no. 32, pp. 1-9.
ZUMBACH, M. S., DACOROGNA, M. M., OLSEN, J. L., AND OLSEN, R. B.
(2000), “Measuring Shock In Financial Market”, International Journal Of Theoretical And Applied Finance, vol. no. 3(3), pp. 347-355.
192
APPENDICES Appendix IV-A
Part A: Domestic Transmission
Brazil
Table 1: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1714 0.1046 0.0413 0.4459 0.000 231.53 sb2 0.1533 0.0756 0.0586 0.3469 0.418 143.75 sa1 0.0055 0.0017 0.0034 0.0097 0.073 22.60 sa2 0.0055 0.0016 0.0034 0.0097 0.689 11.96 sh1 0.0054 0.0015 0.0034 0.0093 0.253 20.85 sh2 0.0057 0.0018 0.0034 0.0101 0.568 21.44
Source: Author’s Compilation
Figure (1) – Results of MCMC draws in Brazil- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
193
Colombia
Table 2: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1111 0.0448 0.0454 0.2160 0.045 85.24 sb2 0.2512 0.1266 0.0755 0.5481 0.000 150.22 sa1 0.0056 0.0018 0.0034 0.0100 0.761 32.64 sa2 0.0056 0.0019 0.0034 0.0097 0.349 26.08 sh1 0.0055 0.0014 0.0034 0.0090 0.252 15.09 sh2 0.0056 0.0017 0.0034 0.0100 0.664 23.45 Source: Author’s Compilation
Figure (2) – Results of MCMC draws in Colombia- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
194
Czech Republic
Table 3: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1172 0.0649 0.0358 0.2784 0.000 147.86 sb2 0.1885 0.1430 0.0432 0.5695 0.000 237.45 sa1 0.0055 0.0016 0.0034 0.0094 0.266 19.63 sa2 0.0056 0.0017 0.0034 0.0098 0.202 31.31 sh1 0.0056 0.0016 0.0034 0.0094 0.085 19.19 sh2 0.0055 0.0016 0.0033 0.0096 0.040 19.83
Source: Author’s Compilation
Figure (3) – Results of MCMC draws in Czech Republic- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
195
Hungary
Table 4: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1192 0.0706 0.0289 0.2973 0.013 193.87 sb2 0.0941 0.0483 0.0280 0.2114 0.328 127.48 sa1 0.0056 0.0017 0.0034 0.0096 0.375 21.75 sa2 0.0056 0.0017 0.0034 0.0098 0.098 24.33 sh1 0.0059 0.0069 0.0033 0.0103 0.591 29.28 sh2 0.0055 0.0016 0.0034 0.0093 0.108 26.68
Source: Author’s Compilation
Figure (4) – Results of MCMC draws in Hungary- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
196
Malaysia
Table 5: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1669 0.0693 0.0661 0.3389 0.002 127.71 sb2 0.2291 0.1091 0.0776 0.5150 0.462 165.32 sa1 0.0054 0.0015 0.0033 0.0092 0.457 17.26 sa2 0.0055 0.0017 0.0033 0.0097 0.961 26.46 sh1 0.0055 0.0015 0.0034 0.0094 0.568 23.44 sh2 0.0056 0.0017 0.0034 0.0097 0.657 27.03
Source: Author’s Compilation
Figure (5) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
197
Mexico
Table 6: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.2464 0.1589 0.0640 0.6871 0.000 249.45 sb2 0.1590 0.0952 0.0341 0.3984 0.240 216.46 sa1 0.0065 0.0120 0.0033 0.0107 0.108 34.96 sa2 0.0056 0.0016 0.0034 0.0095 0.002 21.91 sh1 0.0069 0.0131 0.0034 0.0116 0.438 42.52 sh2 0.0063 0.0077 0.0034 0.0114 0.136 42.67
Source: Author’s Compilation
Figure (6) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
198
Pakistan
Table 7: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1965 0.1187 0.0385 0.4882 0.000 178.67 sb2 0.2131 0.1355 0.0553 0.5508 0.459 162.04 sa1 0.0055 0.0016 0.0034 0.0096 0.605 18.24 sa2 0.0054 0.0017 0.0034 0.0094 0.923 19.04 sh1 0.0056 0.0016 0.0034 0.0096 0.033 20.27 sh2 0.0057 0.0017 0.0035 0.0100 0.090 25.59 Source: Author’s Compilation
Figure (7) –
Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
199
Peru
Table 8: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1500 0.0732 0.0416 0.3300 0.004 158.02 sb2 0.1498 0.0869 0.0445 0.3793 0.893 203.45 sa1 0.0055 0.0016 0.0033 0.0097 0.281 28.27 sa2 0.0055 0.0016 0.0034 0.0096 0.061 15.98 sh1 0.0055 0.0015 0.0034 0.0094 0.932 19.23 sh2 0.0055 0.0015 0.0034 0.0092 0.000 16.95
-------------------------------------------------------------- Source: Author’s Compilation
Figure (8) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
200
Philippines
Table 9: Estimation results for the parameters of the TVP-VAR model
---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1223 0.0640 0.0339 0.2715 0.001 179.52 sb2 0.2207 0.1247 0.0462 0.4946 0.991 247.99 sa1 0.0056 0.0024 0.0033 0.0101 0.838 26.73 sa2 0.0056 0.0018 0.0034 0.0102 0.685 30.40 sh1 0.0062 0.0104 0.0034 0.0101 0.198 35.53 sh2 0.0057 0.0023 0.0034 0.0104 0.773 29.45
Source: Author’s Compilation
Figure (9) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
201
Poland
Table 10: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1094 0.0578 0.0366 0.2564 0.757 153.39 sb2 0.0867 0.0477 0.0264 0.2094 0.504 122.01 sa1 0.0056 0.0018 0.0034 0.0100 0.275 29.55 sa2 0.0055 0.0015 0.0034 0.0092 0.897 18.60 sh1 0.0065 0.0131 0.0034 0.0107 0.225 48.23 sh2 0.0055 0.0015 0.0034 0.0093 0.899 15.76
Source: Author’s Compilation
Figure (10) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
202
Russian Federation
Table 11: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1661 0.0956 0.0336 0.3693 0.001 244.97 sb2 0.3464 0.1996 0.0478 0.7781 0.000 277.79 sa1 0.0055 0.0016 0.0033 0.0096 0.529 15.99 sa2 0.0056 0.0016 0.0034 0.0095 0.213 27.08 sh1 0.0055 0.0015 0.0034 0.0093 0.774 14.18 sh2 0.0056 0.0016 0.0034 0.0097 0.127 15.03
Source: Author’s Compilation
Figure (11) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
203
Turkey
Table 12: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1505 0.0800 0.0470 0.3610 0.022 178.68 sb2 0.1625 0.0848 0.0484 0.3627 0.071 159.26 sa1 0.0055 0.0016 0.0034 0.0095 0.683 20.08 sa2 0.0056 0.0017 0.0034 0.0101 0.496 27.24 sh1 0.0056 0.0016 0.0034 0.0095 0.830 17.42 sh2 0.0056 0.0017 0.0034 0.0098 0.013 25.40 Source: Author’s Compilation
Figure (12) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
204
Part B: International Transmission
US to Brazil:
Table 13: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.3040 0.1778 0.0693 0.7019 0.001 260.84 sb2 0.1976 0.1075 0.0510 0.4448 0.338 251.05 sa1 0.0056 0.0016 0.0033 0.0098 0.238 23.07 sa2 0.0055 0.0016 0.0034 0.0093 0.867 15.88 sh1 0.0058 0.0019 0.0035 0.0106 0.927 19.28 sh2 0.0056 0.0016 0.0034 0.0098 0.005 25.13
Source: Author’s Compilation
Figure (13) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
205
US to Colombia:
Table 14: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1488 0.0753 0.0392 0.3217 0.000 229.39 sb2 0.3991 0.3212 0.0854 1.1887 0.000 320.98 sa1 0.0056 0.0016 0.0034 0.0095 0.989 21.37 sa2 0.0055 0.0016 0.0034 0.0097 0.255 30.02 sh1 0.0055 0.0015 0.0034 0.0095 0.001 17.76 sh2 0.0057 0.0017 0.0034 0.0100 0.836 36.81
Source: Author’s Compilation
Figure (14) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
206
US to Czech Republic:
Table 15: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1991 0.1149 0.0536 0.4729 0.769 277.37 sb2 0.3955 0.2196 0.0720 0.8442 0.548 269.89 sa1 0.0056 0.0017 0.0034 0.0097 0.063 18.53 sa2 0.0055 0.0015 0.0034 0.0092 0.946 15.13 sh1 0.0056 0.0017 0.0034 0.0096 0.001 22.51 sh2 0.0054 0.0015 0.0034 0.0089 0.415 22.92
Source: Author’s Compilation
Figure (15) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
207
US to Hungary:
Table 16: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1650 0.0954 0.0427 0.3834 0.132 231.87 sb2 0.2215 0.1237 0.0378 0.4898 0.371 280.18 sa1 0.0056 0.0016 0.0034 0.0098 0.008 22.28 sa2 0.0055 0.0017 0.0034 0.0098 0.887 19.45 sh1 0.0056 0.0017 0.0034 0.0099 0.461 17.71 sh2 0.0055 0.0015 0.0034 0.0093 0.804 17.49
Source: Author’s Compilation
Figure (16) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
208
US to Malaysia:
Table 17: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1598 0.0727 0.0547 0.3285 0.515 193.18 sb2 0.3986 0.1412 0.1437 0.7134 0.000 178.74 sa1 0.0055 0.0016 0.0034 0.0096 0.102 16.07 sa2 0.0056 0.0017 0.0034 0.0099 0.300 22.83 sh1 0.0055 0.0016 0.0033 0.0093 0.996 15.61 sh2 0.0057 0.0017 0.0034 0.0100 0.695 22.90
Source: Author’s Compilation
Figure (17) –
Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
209
US to Mexico:
Table 18: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1195 0.0592 0.0370 0.2630 0.004 202.16 sb2 0.1434 0.0778 0.0456 0.3278 0.000 200.54 sa1 0.0056 0.0016 0.0034 0.0094 0.360 15.77 sa2 0.0057 0.0018 0.0034 0.0103 0.356 30.17 sh1 0.0056 0.0016 0.0034 0.0099 0.407 21.76 sh2 0.0056 0.0017 0.0033 0.0100 0.007 18.76
Source: Author’s Compilation
Figure (18) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
210
US to Pakistan:
Table 19: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.3759 0.2515 0.0904 1.0263 0.000 285.26 sb2 0.5145 0.2591 0.1965 1.2112 0.137 240.36 sa1 0.0056 0.0016 0.0033 0.0093 0.620 19.78 sa2 0.0055 0.0015 0.0034 0.0093 0.014 24.11 sh1 0.0056 0.0019 0.0033 0.0100 0.407 28.13 sh2 0.0054 0.0015 0.0033 0.0089 0.707 14.49
Source: Author’s Compilation
Figure (19) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
211
US to Peru:
Table 20: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1724 0.1257 0.0333 0.4905 0.000 305.25 sb2 0.2830 0.1756 0.0561 0.6776 0.000 284.37 sa1 0.0066 0.0164 0.0034 0.0100 0.279 44.95 sa2 0.0055 0.0016 0.0033 0.0095 0.934 24.77 sh1 0.0058 0.0051 0.0034 0.0095 0.433 32.00 sh2 0.0058 0.0027 0.0034 0.0101 0.576 19.93 Source: Author’s Compilation
Figure (20) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
212
US to Philippines:
Table 21: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1482 0.0789 0.0443 0.3380 0.898 203.37 sb2 0.2607 0.1469 0.0635 0.6052 0.000 263.04 sa1 0.0055 0.0015 0.0034 0.0092 0.869 13.34 sa2 0.0056 0.0016 0.0034 0.0097 0.846 21.91 sh1 0.0057 0.0017 0.0034 0.0102 0.022 9.03 sh2 0.0056 0.0017 0.0034 0.0098 0.098 18.86
Source: Author’s Compilation
Figure (21) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
213
US to Poland:
Table 22: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.1094 0.0578 0.0366 0.2564 0.757 153.39 sb2 0.0867 0.0477 0.0264 0.2094 0.504 122.01 sa1 0.0056 0.0018 0.0034 0.0100 0.275 29.55 sa2 0.0055 0.0015 0.0034 0.0092 0.897 18.60 sh1 0.0065 0.0131 0.0034 0.0107 0.225 48.23 sh2 0.0055 0.0015 0.0034 0.0093 0.899 15.76
Source: Author’s Compilation
Figure (22) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
214
US to Russian Federation:
Table 23: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.2592 0.1552 0.0395 0.5933 0.000 285.33 sb2 0.5321 0.2465 0.0951 1.0534 0.000 271.79 sa1 0.0053 0.0015 0.0033 0.0089 0.525 15.74 sa2 0.0056 0.0018 0.0035 0.0100 0.002 28.46 sh1 0.0055 0.0016 0.0034 0.0094 0.373 25.89 sh2 0.0056 0.0018 0.0034 0.0101 0.982 23.44 -------------------- Source: Author’s Compilation
Figure (23) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
215
US to Turkey:
Table 24: Estimation results for the parameters of the TVP-VAR model ---------------------------------------------------------------------- Parameter Mean Stdev 95%U 95%L Geweke Inef. ---------------------------------------------------------------------- sb1 0.4546 0.2708 0.0964 1.0415 0.000 306.92 sb2 0.1616 0.0926 0.0359 0.4092 0.747 210.17 sa1 0.0055 0.0017 0.0034 0.0097 0.071 22.76 sa2 0.0055 0.0016 0.0034 0.0097 0.693 12.01 sh1 0.0055 0.0015 0.0034 0.0092 0.208 21.40 sh2 0.0056 0.0017 0.0034 0.0100 0.580 19.23 Source: Author’s Compilation
Figure (24) – Results of MCMC draws- Sample autocorrelations (top), sample paths (middle), and posterior densities (bottom)
216
Appendix IV-B
a) Simultaneous Relation in domestic countries:
Figure (1) – Simultaneous Relation of Brazil
217
Figure (2) Simultaneous Relation of Colombia
Figure (3) Simultaneous Relation of Czech Republic
218
Figure (4) Simultaneous Relation of Hungary
Figure (5) Simultaneous Relation of Malaysia
219
Figure (6) Simultaneous Relation of Mexico
Figure(7) – Simultaneous Relation of Pakistan
220
Figure (8) Simultaneous Relation of Peru
Figure (9) Simultaneous Relation of Philippines
221
Figure (10) Simultaneous Relation of Poland
222
Figure (11) Simultaneous Relation of Russian Federation
Figure (12) Simultaneous Relation of Turkey
223
b) Simultaneous Relation in International Linkage:
Figure (13) – Simultaneous Relation from US to Brazil
Figure (14) Simultaneous Relation from US to Colombia
224
Figure (15) – Simultaneous Relation from US to Czech Republic
Figure (16) Simultaneous Relation from US to Hungar
225
Figure (17) Simultaneous Relation from US to Malay
Figure (18) Simultaneous Relation from US to Mexico
226
Figure (19) Simultaneous Relation from US to Pakistan
Figure (20) Simultaneous Relation from US to Peru
227
Figure (21) Simultaneous Relation from US to Philippines
Figure (22) Simultaneous Relation from US to Poland
228
Figure (23) Simultaneous Relation from US to Russian Federation
Figure (24) Simultaneous Relation from US to Turkey
229
Appendix IV-C
Stochastic Volatility
Figure 25- Stochastic Volatility of Brazil
Figure 26- Stochastic Volatility of Colombia
230
Figure 27- Stochastic Volatility of Czech Republic
Figure 28- Stochastic Volatility of Hungary
Figure 29- Stochastic Volatility of Malaysia
231
Figure 30- Stochastic Volatility of Mexico
Figure 31- Stochastic Volatility of Pakistan
Figure 32- Stochastic Volatility of Peru
232
Figure 33- Stochastic Volatility of Philippines
Figure 34- Stochastic Volatility of Poland
233
Figure 35- Stochastic Volatility of Russian Federation
Figure 36- Stochastic Volatility of Turkey
Figure 37- Stochastic Volatility of US-Brazil
Figure 38- Stochastic Volatility of US-Colombia
234
Figure 39- Stochastic Volatility of US-Czech Republic
Figure 40- Stochastic Volatility of US-Hungary
235
Figure 41- Stochastic Volatility of US-Malaysia
Figure 42- Stochastic Volatility of US-Mexico
236
Figure 43- Stochastic Volatility of US-Pakistan
Figure 44- Stochastic Volatility of US-Peru
237
Figure 45- Stochastic Volatility of US-Philippines
Figure 46- Stochastic Volatility of US-Poland
Figure 47- Stochastic Volatility of US-Russian Federation
238
Figure 48- Stochastic Volatility of US-Turkey