EXTERNAL BORROWING AND INFLATION IN TURKEY FROM 2003-15;
A SIMPLE LINEAR REGRESSION ANALYSIS
CONTENTS
Abstract
1.
Introduction
2. Literature
Review: Interaction between External Debt and Inflation
3.
External Borrowing and Inflation in Turkey from 2003-15
4.
Methodology and Findings; Regression Analysis for the Effects of External Debts
on Inflation in Turkey from 2003-15
4.1.
Tests for the Variables of the Analysis
4.2. Model 1: Simple Linear Regression Analysis for
EXB and CPI
4.3. Model 2: Simple Linear Regression Analysis for
EXB and PPI
5.
Conclusion
Bibliography
Appendices
ABSTRACT
An economy using external resources
can aim at several targets e.g. growth, public financing, covering a deficit in
the balance of payments. However, external/foreign debt/borrowing (EXB) may
result in some negative impacts such as a vicious cycle of increase in external
debt, a decline in economic growth, huge budget deficits and an imbalance of
payments in addition to inflation. This study examines the influence of
external debts on inflation in Turkey from 2003-
Keywords: External/foreign
debt/borrowing; growth; inflation; consumer price index, producer price index;
regression.
JEL Classifications: E31; F34; H6
1. Introduction
External/foreign debt/borrowing is one of the financial resources from
which any economy lacking the benefits of internal savings can profit. External
debt stock is defined by the Central Bank of Turkey (2015-b) as “the remainder
of current and unconditional liabilities used at any time by the residents of
an economy owed to nonresidents, and which requires payment of principal and/or
interest on a due date.” The foregoing is classified as short or long according
to its term. Short term debt includes the credit which is due up to one year
(365 days), whereas long term debt becomes due in excess of one year. As stated
by Adıyaman (2006: 22), there are further classifications concerning external
debt, one of which is related to the borrower. If the borrower is a government
the debt is classified as public, whereas the debt is private if the borrower
is other than a government. In this context, where ‘borrowing’ is generally
considered as accepting money or similarly valuable objects to be returned
after a specific time, government borrowing can also be defined as obtaining
credit by a government or a governmental institution from sources other than
its own.
The other sort of external
borrowing, private borrowing, is that executed by private institutions such as
banks, companies etc. for various purposes; the financing of their projects,
budgets, and foreign transactions for example. As the majority of such debts
are, in fact, guaranteed by national governments (the Undersecretariat of the
Treasury in Turkey’s case), they can also be considered, indirectly, as public
debts because, should they not be repaid by the private institutions, they are
ultimately nationalised and paid by the government. Chile's debt
nationalisation in 1982 is an example (Kim and Zhang, 2012: 121). Chilean total
foreign debt reached as high as 20 billion dollars in 1982, of which two thirds
was private debt incurred by leading domestic private banks. When the Latin
American economic crisis led to the cutting of new loans six top private banks
failed and the government of Chile nationalised the debts as it was assumed to
be responsible for private external borrowing.
According to Evgin (2000:
24), financial resources for a country’s capital accumulation can be obtained
from three sources: domestic individual and institutional savings by lowering
consumption; net foreign exchange flows through foreign trade and touristic
activities; and foreign savings resulting from international assistance and
borrowing. Lessards (1986: 3) states that countries need external financing or
external borrowing for the following reasons: inadequate internal savings;
industrialisation and development efforts that require financing; dependence on
external assistance due to low industrial production resulting from the
importation of intermediate goods; inadequate foreign trade; the balance of
payments and the amount of national foreign exchange; excessively large military
expenses; public sector deficit; expensive domestic financing compared to
foreign financing; the economy being open to short term capital flows; and the
necessity of rendering external debts which become due.
One of the main aims of
external borrowing resulting from any of the above reasons, is to provide
growth in an economy. States are, therefore, willing to accept external debts
in order to increase economic growth. In Evgin’s view (2000: 25), however, a
state is also like an individual who, to maximise its productivity, accepts
external borrowing to a point where its marginal social utility is equal to its
marginal social cost. Furthermore, for developing countries there is an upper
limit to the efficient use of foreign resources. This limit is called
‘absorption capacity’ and according to those who assert this opinion, external
debt should be received for only as long as it increases productivity in an
economy. An increase in investments made through external borrowing may be
subject to the ‘Law of Diminishing Returns’. That is, any increase in the
volume of production following each new investment will decrease gradually with
time and may eventually fall below the principal and interest service.
Continuation of external borrowing above this limit results in a net loss for
an economy.
Fuhmei (2009: 282-283) refers
to the conventional wisdom that low income and low savings rate countries could
grow faster with foreign capital inflow on condition that this international
borrowing is used for productive purposes. In this context, financial resources
should always be allocated to encourage accumulation of physical capital and to
stimulate private investment as this will lead to economic growth. In this
context, Fuhmei, in his paper on the relationship between public sector foreign
borrowing and economic growth, reaches the following conclusion: only under
circumstances of moderate income tax rates to guarantee the solvency of
external loans, and households having the patience to substitute consumption between
different periods, can government finance fiscal deficits by borrowing from
abroad, thereby enhancing investment and economic growth.
Prokop and Baranowska-Prokop
(2012: 321) examine the efficiency of foreign investment borrowing and its
effect on the economic growth of Poland in the 1970s. Based on their
econometric analysis, they conclude that the efficiency of foreign investment
borrowing was relatively high, which means proving/confirming that external
sources can provide economic growth. To show the positive effect of external
borrowing on economic growth, Burguet and Ferna´ndez-Ruiz (1998: 328) studied
countries Malaysia, Indonesia, and especially South Korea. As these countries
sustained high growth rates for the years from 1965-1989, their respective
annual rates of GDP per capita growth being, on average, 4, 4.4, and 7%, and
they can be given as good examples. Their economic structure changed
dramatically in this period: the share of manufacturing doubling in the first
two cases and tripling in the case of Korea. And in Indonesia, as in Malaysia,
development expenditures such as irrigation projects, village works or school
programmes were an important component of the development process.
Conversely, external
borrowing can result in some negative problems for an economy. In this context,
Akdiş (2003: 15) states that it is not possible for a government to adequately
perform its basic duties such as the provision of education, health, security,
and justice services - which absorb approximately 50% of its budget, and when
almost all taxation incomes are reserved to service interest debt. An indebted
country which has budget deficits enters a vicious cycle of repeated borrowing
in order to pay back its accrued debts. External borrowing may also result in
other negative problems such as a disequilibrium in income distribution and
taxation, shortcomings in savings and investment mechanisms and so on. External
borrowing can be a root cause of inflation as well.
Material and Method
The effects of external
borrowing on inflation in Turkey have been studied before. This paper, however,
will focus specifically on the period 2003-15. There are two main reasons for
selecting the said years, the first of which is to see if the positive
relationship between external borrowing and inflation continues to exist during
this period as Turkey gradually improves its economic structure immediately
following the most dramatic effects of the South Asian Economic Crisis that
began in 1997 and which was deeply felt in Turkey from 1999-2001. The second
reason is that inflation indices began to be calculated more systematically
from 2003 onwards.
In this paper, a literature
review is given first and then a simple linear regression analysis is made
using both CPI and PPI. However, before the regression analysis the
autocorrelation, causality, and heteroscedasticity of the variables were
examined to remove the spurious regression problems by checking the levels of
integration of data set. The aim in using both indices is to see the nature of
the effects of external debt on consumers and producers in Turkey from 2003-15.
At the end, some proposals are made towards lowering the negative effects of
external borrowing on inflation and other economic aggregates. While sources
for litearature review include articles, reports etc., those for regression
analyses involve data provided by the Statistical Institute of Turkey and the
Treasury of Turkey.
2. Literature Review: Interaction between External Debt and Inflation
There are two opposing views with regard to the effects of external
borrowing on prices. While one view asserts that borrowing causes an
inflationary effect, the other asserts that it results in a deflationary effect
by playing a restrictive role in an economy (Adıyaman, 2006: 37). The common characteristic of theoretical approaches
is in their agreement that there is an interaction between borrowing and
inflation. The majority of such approaches propose that both internal and
external borrowings have inflationary effects (Sugözü and Yiyit, 2010: 371).
This is confirmed by several studies one of which is by Ulusoy and Küçükkale
(1996: 23) who made an econometric analysis based on data of Turkey from
1965-94. Using the Granger Causality Test they found that foreign borrowing
increases inflation in Turkey.
In a paper by Karakaplan
(2009: 215), the following two hypotheses are tested: The first states that the
external debt is less inflationary if financial markets are well developed; the
second is that the effects of the determinants of inflation are heterogeneous
across countries in their extent and signs. For this purpose, using an
unbalanced panel data set that includes 121 countries in different groups
(Latin American, European Union, high inflation, and transition countries) for
the period 1960−2004, his analysis offers robust empirical support for
these hypotheses.
Cardoso and Fishlow (1990:
324) state that inflationary deficit finance leads inevitably to two types of vicious
circle. First, if government prices are adjusted with delays and income taxes
are collected on the basis of incomes earned one year before (Olivera-Tanzi
effect), higher inflation itself increases the budget deficit, inducing even
larger increases in money. Second, the share of the inflation tax in output is
inversely related to velocity. Since velocity increases with inflation,
increasing budget deficits will require further increases in money creation as
velocity responds to increasing inflation rates. This is a vicious circle, and
when external borrowing is made it causes an increase in the money supply
followed by inflation, which in turn further increases the need for foreign
debt.
Ulusoy and Küçükkale (1996:
23) mention that external debt, however acquired, increases the cash capital
accumulation of a debtor country in terms of foreign exchange. If this excess
cash is shifted to unproductive areas due to an insufficiency of investment
incentives and/or a high propensity to consumption, it results in increases in
domestic prices - a phenomenon of inflation. The same effect will be seen if
the debts are used in infrastructural investments because the expenditure for
such an investment will immediately stimulate consumption (accelerator), while
the contribution of the investment to production (multiplier) will be revealed
later. They state that using external resources for public financing and import
financing can also be seen as applications which accelerate inflation.
Demir and Sever (2009: 14) say that when it comes to borrowing in terms of
public financing, external borrowing by the state and the use of foreign
exchange are generally considered. The relationship between budget deficit and
borrowing becomes more evident in those economies that have weak capital
markets and lack borrowing possibilites. Since internal borrowing possibilities
are limited, public financing need is covered by external borrowing. Duran (1996: 450), finds that external debts result in an inflationary
effect when public financial deficit is met by exchanging foreign exchange
reserves with national currency used for public expenditures. As this causes
emission, the result is an increase in aggregate demand. In addition, the use
of foreign exchange generated through the external borrowing mechanism for the
purpose of public financing narrows import capability and thus has a negative
effect on the aggregate supply. As aggregate supply decreases to below
aggregate demand this leads to inflationary pressure.
As stated by Akdiş (2003: 7),
the Public Sector Borrowing Requirement (PSBR) demolishes the public financing
balance and increases inflationary pressure. The Public Sector then endeavours
to cover its deficit either by raising its net pecuniary liabilities or by
borrowing from the private sector through bond sales. In that case, there will
be a direct relationship between public sector net pecuniary liabilites and
money stock. Thus, when the public deficit increases, money stock has to be
increased as well unless the deficit is covered through bond sales. This direct
relationship between public deficit and money stock becomes the most important
factor indicating the character of the inflation phenomenon. Akdiş says,
therefore, that PSBR and its continuity supports inflationary increases. As an
extension of this relationship, Demir, Çevik, and Beşer (2005: 264), in an empirical
analysis about Turkey, find that PSBR is in positive relation with interest
rates and inflation.
Duran (1996: 436) asserts that although PSBR is one of the important
reasons for inflation it is not the sole reason. As PSBR affects inflation
through emission, it plays an increasing role while budget deficits are covered
by credit mechanisms and when there is a disequilibrium between equity and
foreign sources. Demir and Sever (2009: 24), in their paper concerning the relationship
between budget deficits and borrowing by Turkey, Azerbaijan, Kazakhstan, and
Kirgyzstan, find the following: In Azerbaijan, increasing budget deficits raise
external debt depending upon insufficiencies in national savings. In Turkey,
Kazakhstan, and Kirgyzstan, however, PSBR is, in some periods, met by other
resources (internal borrowing, tax or emission) rather than by external
borrowing.
In their study whose findings regarding sustainability of fiscal deficit
have an important bearing on macro-economic policies, Chaudhary and Anjum
(1996: 784) focus on analysing the sustainability of fiscal deficit in
Pakistan. In this context they indicate that inflation, unemployment, increasing
debt burden, and debt-servicing are linked to fiscal deficit. Thus, there is a
need to keep the fiscal deficit within a limit consistent with other
macro-economic variables like inflation, debt etc. They say that doing so may
help to stabilise the economy and resolve the related economic problems.
Evgin (2000: 11) states that
one of the influences of foreign debt increases is rising interest rates. A
state may have to increase the interest rates of its bonds to cover a budget
deficit. An upward tendency in interest rates increases the share of interest
service in budgetary expenses and this raises budget deficits. The rise in
interest rates leads to negative effects on consumption and investment
expenses. In their study on Australia, Makin and Narayan (2012: 1) examine the
impact of capital inflow on interest rates. They show that rising net capital
inflow has had a statistically significant negative impact on domestic real
interest rates in Australia, an Asia-Pacific economy that has borrowed heavily
from abroad since the mid 1980s.
Ulusoy and Küçükkale (1996:
23) state that it is also possible to meet inflation phenomena during repayment
of external debts. A country liable to repay its due debts has to increase its
export revenues. This requires one of the simplest solutions; devaluation.
While devaluation increases exports it makes imports expensive, resulting in
cost increases in foreign input-using sectors. These costs are reflected in
prices, leading to an inflationary process. On the other hand, as a result of
an increase in exports, a shrinkage in supply will occur in some sectors and
this too will cause increased pressure on prices. They indicate that the
economic crisis in 1994 (5th April) in Turkey happened in just this way. In the
said period, the inflation rate was very high (150%) while growth rate was
negative (-6%).
Table I: Interaction Mechanism between External Debt and Inflation
|
Subject |
Usage
Area/Mechanism |
Impact |
Economic Fact |
Result |
|
1) External debt |
Unproductive investments |
Production insufficiency |
Demand>Supply |
Inflation |
|
2) External debt |
Infrastructural investments |
Production insufficiency (Accelerator effect) |
Demand>Supply |
Inflation |
|
3) External debt |
Public financing |
Increase in emissions (Multiplier effect) |
Demand>Supply |
Inflation |
|
4) External debt |
Public financing |
Decrease in imports; (Decreases in machinery&equipment imports lower
production capacity) |
Demand>Supply |
Inflation |
|
5) External debt |
Public financing |
Increases in interest rates cause a decrease in
investments (Crowding out effect) |
Demand>Supply |
Inflation |
|
6) External debt requirement |
Public financing |
Increases in budget deficits cause increase in
emissions |
Demand>Supply |
Inflation |
|
7) External debt repayment |
Devaluation |
Increase in export volume (Concentrating on foreign markets and neglecting the
domestic markets) |
Demand>Supply |
Inflation |
|
8) External debt repayment |
Devaluation |
Decrease in imports (Decreases in machinery&equipment imports lower
production capacity) |
Demand>Supply |
Inflation |
|
9) External debt repayment |
Devaluation |
Increase in import cost (Increases in the costs of production factors) |
Demand>Supply |
Inflation |
As can be seen from Table I,
external debt results in inflationary effects in many aspects. However, it
should be noted that the main economic fact underlying this relationship is the
insufficiency of supply to demand.
3. External Borrowing and Inflation in Turkey from 2003-15
Quarterly total external debt
stock of Turkey from 2003-15 is available in Table II.
Table II: Quarterly External Debt Stock of Turkey from 2003-15 (million
USD)
|
Quarter |
Amount |
|
Quarter |
Amount |
|
Quarter |
Amount |
|
Quarter |
Amount |
|
2003 Q1 |
130,931 |
|
2007 Q1 |
214,220 |
|
2011 Q1 |
301,994 |
|
2015 Q1 |
393,135 |
|
2003 Q2 |
135,040 |
|
2007 Q2 |
224,492 |
|
2011 Q2 |
313,683 |
|
2015 Q2 |
405,223 |
|
2003 Q3 |
138,722 |
|
2007 Q3 |
236,444 |
|
2011 Q3 |
312,123 |
|
|
|
|
2003 Q4 |
144,161 |
|
2007 Q4 |
250,012 |
|
2011 Q4 |
303,931 |
|
|
|
|
2004 Q1 |
144,800 |
|
2008 Q1 |
265,048 |
|
2012 Q1 |
316,747 |
|
|
|
|
2004 Q2 |
147,353 |
|
2008 Q2 |
287,156 |
|
2012 Q2 |
322,691 |
|
|
|
|
2004 Q3 |
153,105 |
|
2008 Q3 |
291,984 |
|
2012 Q3 |
327,496 |
|
|
|
|
2004 Q4 |
161,139 |
|
2008 Q4 |
280,957 |
|
2012 Q4 |
339,042 |
|
|
|
|
2005 Q1 |
160,322 |
|
2009 Q1 |
265,563 |
|
2013 Q1 |
352,109 |
|
|
|
|
2005 Q2 |
162,686 |
|
2009 Q2 |
268,180 |
|
2013 Q2 |
367,803 |
|
|
|
|
2005 Q3 |
166,472 |
|
2009 Q3 |
271,275 |
|
2013 Q3 |
373,499 |
|
|
|
|
2005 Q4 |
170,750 |
|
2009 Q4 |
268,963 |
|
2013 Q4 |
389,146 |
|
|
|
|
2006 Q1 |
185,545 |
|
2010 Q1 |
267,487 |
|
2014 Q1 |
388,244 |
|
|
|
|
2006 Q2 |
191,622 |
|
2010 Q2 |
265,741 |
|
2014 Q2 |
402,368 |
|
|
|
|
2006 Q3 |
197,246 |
|
2010 Q3 |
284,062 |
|
2014 Q3 |
397,781 |
|
|
|
|
2006 Q4 |
208,108 |
|
2010 Q4 |
292,057 |
|
2014 Q4 |
402,720 |
|
|
|
Source: Treasury of Turkey
(2015), Public Finance Statistics.
Retrieved on 15 October 2015 from the Treasury of Turkey Web site: http://www.treasury.gov.tr/en-US/Stat-List?mid=738&cid=12&nm=684.
As can be seen in Table II,
the external debt stock of Turkey increased gradually between 2003 and 2015.
The amount of debt eventuated as 130,931 million US Dollars in early 2003
reaching 405,223 million US Dollars in the midst of 2015.
Many ratios are used to
calculate the external indebtedness rate of a country. The commonly accepted
external indebtedness ratios can be classified into four groups as shown in
Table III.
Table III: Commonly Accepted External Debt Ratios and Turkey (%)
|
Ratios |
Commonly Accepted Ratios |
Debt Ratios in TR (2014) |
|
External Debt/GDP |
30-60 |
50.4 |
|
External Debt/Exports |
165-275 |
255.5 |
|
External Debt
Service/Exports |
18-30 |
31.2 |
|
Interest Service/Exports |
12-20 |
6.3 |
Source:
·
Treasury of Turkey (2015), Public
Finance Statistics. Retrieved on 15 October 2015 from the Treasury of
Turkey Web site: http://www.treasury.gov.tr/en-US/Stat-List?mid=738&cid=12&nm=684.
·
Central Bank of Turkey (2015-a), Balance
of Payment Statistics. Retrieved on 16 October 2015 from the Central Bank
of Turkey Web site: http://www.tcmb.gov.tr/wps/wcm/connect/TCMB+EN/TCMB+EN/Main+Menu/STATISTICS/Balance+of+Payments+and+Related+Statistics/Balance+of+Payments+Statisticss/.
·
Statistical Institute of Turkey (2015-b), National Accounts. Retrieved
on 15 October 2015 from the Statistical Institute of Turkey Web site: http://www.turkstat.gov.tr/UstMenu.do?metod=temelist.
·
Akdiş, Muhammet (2003), Türkiye’nin
Borç Gelişimi; Sorunlar-Öneriler. Active, 30, 3.
·
Calculations were made with data obtained from the Treasury, Central Bank,
and Statistical Institute of Turkey.
In accordance with the data
for 2014, the following interpretations could be made by considering the
commonly accepted external debt ratios of Turkey in Table III:
·
External Debt/GDP in Turkey is 50.4% which remains within normal limits.
Turkey, therefore, takes its place in the table of medium level indebted
countries in terms of this ratio.
·
External Debt/Exports in Turkey is 255.5% which is also between the
accepted limits. It shows that the export volume of Turkey allows it to cover a
certain amount of its external debt stock.
·
External Debt Service/Exports in Turkey is 31.2%. This ratio is above the
upper limit. Although it is not in a very risky position, Turkey has a fragile
capacity to render its principal and interest rate by its export revenues.
·
Interest Service/Export in Turkey is 6.3%. This is below even the lowest
level of the commonly accepted ratio. This proves that the external debt
interest can be paid easily through export gains.
As Karagöz (2007: 100)
mentions, the World Bank takes two main measures into account with regard to
borrowing. The first is ‘External Debt/GDP’ and the second ‘External Debt
Service/Exports’. These two measures show the repayment capacity of a country.
From a different point of view, since the first measure indicates revenue
generating capacity and the second shows the foreign exchange-gaining
possibility of an economy, they are significant for both internal and external
borrowings. While for Turkey the ‘External Debt/GDP’ lies between the commonly
accepted rates, the rate of ‘External Debt Service/Exports’ was above the said
limit as of end of 2014. The external debt service should, therefore, be
tackled with care.
Table IV: Quarterly CPI in Turkey from 2003-15 (%)
|
Quarter |
Rate |
|
Quarter |
Rate |
|
Quarter |
Rate |
|
Quarter |
Rate |
|
2003 Q1 |
96.37 |
|
2007 Q1 |
136.64 |
|
2011 Q1 |
183.74 |
|
2015 Q1 |
252.64 |
|
2003 Q2 |
99.75 |
|
2007 Q2 |
139.68 |
|
2011 Q2 |
188.40 |
|
2015 Q2 |
259.92 |
|
2003 Q3 |
100.49 |
|
2007 Q3 |
139.17 |
|
2011 Q3 |
188.69 |
|
|
|
|
2003 Q4 |
103.39 |
|
2007 Q4 |
144.63 |
|
2011 Q4 |
198.95 |
|
|
|
|
2004 Q1 |
105.51 |
|
2008 Q1 |
148.68 |
|
2012 Q1 |
203.02 |
|
|
|
|
2004 Q2 |
107.15 |
|
2008 Q2 |
154.12 |
|
2012 Q2 |
206.14 |
|
|
|
|
2004 Q3 |
108.61 |
|
2008 Q3 |
155.38 |
|
2012 Q3 |
205.76 |
|
|
|
|
2004 Q4 |
113.13 |
|
2008 Q4 |
160.44 |
|
2012 Q4 |
212.42 |
|
|
|
|
2005 Q1 |
114.60 |
|
2009 Q1 |
161.12 |
|
2013 Q1 |
217.65 |
|
|
|
|
2005 Q2 |
116.38 |
|
2009 Q2 |
162.90 |
|
2013 Q2 |
220.52 |
|
|
|
|
2005 Q3 |
117.20 |
|
2009 Q3 |
163.67 |
|
2013 Q3 |
222.85 |
|
|
|
|
2005 Q4 |
121.75 |
|
2009 Q4 |
169.60 |
|
2013 Q4 |
228.30 |
|
|
|
|
2006 Q1 |
123.86 |
|
2010 Q1 |
176.09 |
|
2014 Q1 |
235.09 |
|
|
|
|
2006 Q2 |
127.56 |
|
2010 Q2 |
177.92 |
|
2014 Q2 |
241.25 |
|
|
|
|
2006 Q3 |
129.89 |
|
2010 Q3 |
177.39 |
|
2014 Q3 |
243.44 |
|
|
|
|
2006 Q4 |
133.71 |
|
2010 Q4 |
182.20 |
|
2014 Q4 |
248.30 |
|
|
|
Source:
·
Statistical Institute of Turkey (2015-a), Inflation&Price. Retrieved
on 15 October 2015 from the Statistical Institute of Turkey Web site: http://www.turkstat.gov.tr/UstMenu.do?metod=temelist.
·
The quarterly rates were calculated by the monthly values of the
Statistical Institute of Turkey. Monthly values are available in the Appendix
I.
As shown in Table IV,
starting from 2003-15, there has, with minor exceptions, always been an upward
tendency in inflation, in terms of CPI. This holds good for PPI too, as can be
seen in Table V.
Table V: Quarterly PPI in
Turkey from 2003-15 (%)
|
Quarter |
Rate |
|
Quarter |
Rate |
|
Quarter |
Rate |
|
Quarter |
Rate |
|
2003 Q1 |
97.33 |
|
2007 Q1 |
136.39 |
|
2011 Q1 |
185.61 |
|
2015 Q1 |
239.35 |
|
2003 Q2 |
101.09 |
|
2007 Q2 |
139.11 |
|
2011 Q2 |
189.52 |
|
2015 Q2 |
247.45 |
|
2003 Q3 |
98.93 |
|
2007 Q3 |
140.55 |
|
2011 Q3 |
192.79 |
|
|
|
|
2003 Q4 |
100.80 |
|
2007 Q4 |
142.63 |
|
2011 Q4 |
200.56 |
|
|
|
|
2004 Q1 |
106.34 |
|
2008 Q1 |
147.81 |
|
2012 Q1 |
203.22 |
|
|
|
|
2004 Q2 |
110.86 |
|
2008 Q2 |
161.42 |
|
2012 Q2 |
203.51 |
|
|
|
|
2004 Q3 |
109.63 |
|
2008 Q3 |
161.88 |
|
2012 Q3 |
202.23 |
|
|
|
|
2004 Q4 |
115.48 |
|
2008 Q4 |
158.61 |
|
2012 Q4 |
206.32 |
|
|
|
|
2005 Q1 |
115.63 |
|
2009 Q1 |
156.52 |
|
2013 Q1 |
207.30 |
|
|
|
|
2005 Q2 |
119.50 |
|
2009 Q2 |
158.89 |
|
2013 Q2 |
209.67 |
|
|
|
|
2005 Q3 |
121.38 |
|
2009 Q3 |
159.50 |
|
2013 Q3 |
215.19 |
|
|
|
|
2005 Q4 |
122.25 |
|
2009 Q4 |
162.58 |
|
2013 Q4 |
219.67 |
|
|
|
|
2006 Q1 |
123.83 |
|
2010 Q1 |
167.85 |
|
2014 Q1 |
231.78 |
|
|
|
|
2006 Q2 |
130.63 |
|
2010 Q2 |
173.33 |
|
2014 Q2 |
233.41 |
|
|
|
|
2006 Q3 |
135.66 |
|
2010 Q3 |
173.42 |
|
2014 Q3 |
236.12 |
|
|
|
|
2006 Q4 |
135.41 |
|
2010 Q4 |
177.18 |
|
2014 Q4 |
237.82 |
|
|
|
Source:
·
Statistical Institute of Turkey (2015-a), Inflation&Price. Retrieved
on 15 October 2015 from the Statistical Institute of Turkey Web site: http://www.turkstat.gov.tr/UstMenu.do?metod=temelist.
·
The quarterly rates were calculated by monthly values of the Statistical
Institute of Turkey. Monthly values are available in Appendix II.
Considering the inflationary
process of Turkey from 2003-15 in terms of both CPI and PPI, it may be asserted
that there is a positive relationship between external debt and inflation.
However, this requires to be tested. For this purpose a simple linear
regression analysis has been made under the following title.
4. Methodology and Findings; Regression Analysis for the Effects of
External Debts on Inflation in Turkey from 2003-15
Here, the effect of external
borrowing on inflation rates in Turkey is measured through a simple linear
regression analysis. In this context, both CPI and PPI are used. The aim is to
confirm that external borrowing by Turkey had a positive effect on the
inflation rate between 2003 and 2015. 50 observations of EXB, CPI, and PPI were
used in the analysis. EXB data were collected from the Public Finance
statistics of the Treasury of Turkey, while CPI and PPI data were taken from
the Inflation&Price statistics of the Statistical Institute of Turkey. It
should be noted that as there were only monthly data for CPI and PPI, the
quarterly rates were calculated and used by the Author in the analysis. Monthly
values are available in Appendix I.
However, as Granger and
Newbold (1974: 111-112) pointed out, since regression analysis
with time series data may lead to spurious regression problems if the data are
non-stationary, the levels of integration of the data set should be checked
before starting the analysis. In this context, the autocorrelation, causality,
and heteroscadasticity of the variables were examined. For these, linear unit
root tests by Dickey-Fuller
(Augmented Dickey-Fuller: ADF) and Phillips-Perron (PP) plus Granger Causality
Tests were applied. Also heteroscadasticity of the variables was tested. The aim of such tests was to figure out whether
regression results were unbiased and efficient. The tests were performed
through the EViews 8 while the regression analyses were performed through the
MS Excel.
4.1. Tests for the Variables of the Analysis
According to the
Dickey-Fuller (ADF) Unit Root Test (1981) the presence/absence of unit root is
very significant in figuring out whether a time series is stationary. The
series is appropriate for the analysis if it has a unit root and can be removed
by the differencing method. Here the
‘‘τ (tau)’’ statistic of the Monte Carlo Study by Dickey-Fuller (1979) is
used. If the absolute value of ‘‘τ (tau)’’ exceeds the absolute critical
values by Dickey-Fuller or MacKinnon Dickey-Fuller, the assumption of
stationarity of time series cannot be rejected. If “Ho: p=1” is rejected then
the time series is stationary.
The Dickey-Fuller Test assumes that error terms are statistically
independent and have constant variance. Therefore, one should be sure that
there is no correlation between error terms and they have constant variance
(Altunöz, 2013: 187). Phillips-Perron (1988), broadened this assumption of
Dickey-Fuller. They ignored the independence and homogenity assumptions of
Dickey-Fuller and supposed weakly dependent and possibly heterogenously
distributed data. Thus, it is clear that Phillips-Perron did not take into
consideration the restrictions on the assumptions of error terms when
developing Dickey-Fuller t-statistics.
Table VI: Linear Unit Root Test Results
|
Variable |
ADF Statistics (Level) |
PP Statistics (Level) |
|
EXB |
-2.922 |
-2.922 |
|
CPI |
-2.926 |
-2.922 |
|
PPI |
-2.928 |
-2.928 |
|
Variables on
first differences (constant and inconstant) |
||
|
EXB |
-3.506 |
-3.506 |
|
CPI |
-3.518 |
-3.506 |
|
PPI |
-3.510 |
-3.506 |
·
MacKinnon 5% critical value: Level (constant): -2.9;
First difference (constant and inconstant): -3.5.
·
Lags were determined in accordance with the
Schwarz Information Criterion.
·
Tests were performed through EViews 8.
According to the
results in the Table IV, ADF-t statistical values for EXB, CPI, and PPI exceed
Mackinnon’s (1991) critical value of 5% significance level. Therefore; EXB,
CPI, and PPI variables are stationary according to first differences. All
variables are stationary although they are at different significance levels. In
other words, the variables used in this analysis do not contain unit roots and
there is no contrariness for the predictions.
However, for the
autocorrelation problem lag numbers are used.
Table VII: Lags for Autocorrelation Problem
|
Lags |
LM-Statistics |
Probability |
|
1 |
5.960573 |
0.7439 |
|
2 |
3.529407 |
0.9396 |
|
3 |
6.454472 |
0.6937 |
|
4 |
5.482571 |
0.7904 |
|
5 |
8.745679 |
0.4611 |
|
6 |
3.569288 |
0.9374 |
As available in
the Table VII, autocorrelation problem is solved when a lag length of 2 is
used.
The polynomial can
be assessed as an indicator of the stationarity of the model as well.
Figure I: Inverse Roots of AR Characteristic
Polynomial
As can be seen in
Figure I, the position of inverse roots of the AR Characteristic Polynomial of
the Model also shows that there is no problem in terms of the stationarity of
the Model. As none of the inverse roots are outside the unit encirclement, the
established VAR system is stable and there are no different variances. Thus,
the Model is stable in this context.
At this stage, the variables should be settled from
outer to inner in the VAR Analysis prediction. For this, the Granger Causality
Test that can be done through VAR Analysis with positive significance test
results following the determination of appropriate lag numbers is applied.
Table VIII: Granger Causality Test Results
|
Model |
Statistic (χ2) |
Lag |
Probability |
Causality |
|
PPI→CPI |
0.608191 |
2 |
0.7378 |
Unavailable |
|
CPI→PPI |
8.678073 |
2 |
0.0130 * |
Available |
|
CPI→EXB |
1.638277 |
2 |
0.4408 |
Unavailable |
|
PPI→EXB |
2.368730 |
2 |
0.3059 |
Unavailable |
|
EXB→CPI |
0.031739 |
2 |
0.9843 |
Unavailable |
|
EXB→PPI |
2.989692 |
2 |
0.2243 |
Unavailable |
·
Lags were determined in accordance with the
Akaike Criterion.
·
* shows 5% significance level.
As can be seen in the Table VIII, there is a causality between CPI→PPI
at 5% significance level. However, there is no relation between other variables
(PPI→CPI, CPI→EXB, PPI→EXB, EXB→CPI, EXB→PPI).
The results of the
White Heteroscedasticity Test applied to determine whether the variance of
error terms is constant for whole sample, are shown below.
Table
IX: White Heteroscedasticity Test Results
|
Chi-sq |
df |
Probability |
|
59.85760 |
72 |
0.8458 |
It is seen in the Table IX that the variance of time
error term is constant for all observations. That is, there is no variance
problem (p=0.8458>%5). In this case, “Ho” is accepted and an inconstant
variance problem is not available (Null Hypothesis: no heteroscedasticity).
4.2. Model 1: Simple Linear
Regression Analysis for EXB and CPI
This model includes 50 observations for the period 2003 to 2015. The
regression analysis summary outputs are available in the Appendix III.
Variables of the model are as follows:
Dependent
variable (Y) : CPI
Independent
variable (X) : EXB
As the
relationship between the variables of the model is positive, a linear
regression analysis is applied.
Y = b0 b1X + Ɛ
Inflation
(CPI) = b0 + b1 (EXB) + Ɛ
Y = 21.75424209 + 0.000546646
X + Ɛ
Standard
Error : (4.083334534) (1.46847E-05)
tstatistics : (5.327567924) (37.22559373)
R2 =
0.966521239
Adjusted R2 =
0.965823765
Assessments of the results are as follows:
·
b0:
21.75; Even if there is no EXB, there will be a CPI of 21.75.
·
b1:
0.00055; 1 unit EXB causes a 0.00055 unit increase in CPI.
Now, we try to figure out if
the model is significant. For this purpose F-test shall be applied. Here are
the hypotheses:
·
H0: b=0 (It is not significant that the model best fits the population from
which the data were sampled; that is Model is not
significant).
·
H1: b≠0 (It is significant that the model best fits the population from
which the data were sampled; that is Model is
significant).
If F value > critical value
of F distribution, then the null hypothesis is rejected. As F value=1385.745
> critical value of F distribution=1.61, we reject the null hypothesis. That
is the Model is significant which means that EXB increases CPI.
It is time to check whether
the coefficients are statistically significant. For this purpose, T-test shall
be applied and, in this context, the values of tstatistics and ttable
shall be compared. Here are the hypotheses:
·
H0: b=0 (a unit change in X does not make any change in Y; that is, there is no
correlation between these two variables).
·
H1: b≠0 (a unit change in X makes a significant change in Y; that is,
there is a correlation between these two variables).
If the value of a is 0.05, then ttable
is 1.68. In this case if tstatistics > ttable, it means
that the coefficients are statistically significant. As tstatistics
for b1 =37.23 > ttable for b1=1.68, coefficient b1 is statistically significant
which means that there is a positive correlation between EXB and CPI.
Another measure to interpret
the model is the coefficients of determination:
R2 = 0.966
Adjusted R2 = 0.966
Values of coefficients of
determination show the strength of the relationship between the dependent and
independent variables. Both R2 and Adjusted R2 are high (97%)
which means that the model is reliable. In a word, these coefficients show that
the 97% increase in the CPI is explained by EXB in Turkey for the period
2003-15.
4.3. Model 2: Simple Linear
Regression Analysis for EXB and PPI
This model also includes 50 observations for the period 2003 to 2015. The
regression analysis summary outputs are available in Appendix IV. Variables of
the model are as follows:
Dependent
variable (Y) : PPI
Independent
variable (X) : EXB
As the
relationship between the variables of the model is positive, a linear
regression analysis is applied here too.
Y = b0 + b1X + Ɛ
Inflation
(PPI) = b0 + b1 (EXB) + Ɛ
Y = 30.01116 + 0.000508
X + Ɛ
Standard
Error : (3.559362) (1.28E-05)
tstatistics : (8.431611) (39.66731)
R2 =
0.970398
Adjusted R2 =
0.969781
Assessments of the results are as follows:
·
b0:
30.01; Even if there is no EXB there will be a PPI of 30.01.
·
b1:
0.00051; 1 unit EXB causes a 0.00051 unit increase in PPI.
For showing the significance
of the Model we apply F-test. As F value=1,573.496 > critical value of F
distribution=1.61, the null hypothesis is rejected. That is the Model is
significant which means that EXB increases PPI.
As another component of model assessment for checking whether the
coefficients are statistically significant, T-test shall be applied.
As tstatistics for
b1 =39.67 > ttable for b1=1.68, coefficient b1 is statistically
significant. That is, EXB increases PPI.
The interpretation of determination coefficients is given below:
R2 =
0.970
Adjusted R2 = 0.970
In this model, both R2 and Adjusted R2 are high (97%)
which means that the model is reliable. That is, the 97% increase in PPI is
explained by EXB in Turkey in the said period.
5. Conclusion
Different views on borrowing can
be found. As quoted by Tuna (2014), while David Ricardo defined public
borrowing as “an awful scourge invented at any time to torment the people”,
nearly 100 years later, Lorenz von Stein, a Finance Officer in Germany,
opposing this idea said that “a debtless country either does fewer things for
its future or demands many things from the moment”. Considering these
approaches it is clear that on the one hand, while inflation as a result of
external borrowing becomes a means to torment the people, on the other hand it
is also the result of investment, public financing, growth etc.
In this paper, both effects
have been examined for Turkey. Firstly, the simple linear regression analyses
confirm that the use of external borrowing (EXB) has resulted in increased
inflation rates. These analyses show that both the consumer price index (CPI)
and the producer price index (PPI) have been affected by the EXB in Turkey from
2003-15. That is, EXB has increased both CPI and PPI through various
mechanisms. Secondly, it is obvious that external borrowing has had some
positive results on some economic aggregates such as investments, the public
budget and growth. However, these have also had indirect effects on inflation
due to some negative aspects of Turkey’s economy. One of which may, be the lack
of well organised financial markets in addition to other shortfalls. The main
factors for foreign debt being a cause of inflation in the economy of Turkey
may be a misuse of these sources for unproductive investments, huge infrastructural
investments, public deficits, and an imbalance of payments.
Ulusoy and Küçükkale (1996:
24), while considering that external debts used for infrastructural investments
cause inflation, say that if external debts were used to finance income-generating
investments (especially for gaining foreign exchange), the debts could be
repaid and factor endowments increased in favour of capital. Thus, it would be
possible to provide growth without accelerating the inflationary process.
Duran (1996: 442) says that
direct income-generating public investment expenditures could be financed by
external borrowing as they provide direct revenue for the servicing of
principal and interest. However, maturity of the debt should be equal to the
terms of return on investment. The most significant component of such financing
is the difference between the real interest rate and the return on investment
ratio. He adds that it is advantageous to finance investment by external
borrowing provided the real interest rate is either negative or less than the
return on investment ratio. In any case, taxation would be preferable to long
term, unmeasurable and indirect income-generating public investments.
Evgin (2000: 13) emphasises
that while there have been brilliant successes in decreasing inflation in
several countries, increases in external debt may frustrate these results and
that strong budget discipline together with monetary stability measures are the
sole solution for success in this respect. An expansionary monetary and budgeting
policy may result in economic crises as happened in Germany in the 1920s and in
the USA in the 1930s. She asserts that continuous economic growth is possible
only through implementation of a stable monetary policy and strong budget
discipline. Only these policies can decrease interest rates by dashing
inflationary expectations and lowering public debt burden to a bearable level.
Karagöz (2007: 109) asserts
that covering a deficit in the balance of payments rather than dealing with the
shortfall in internal savings has been the main reason for the need for
external borrowing in Turkey. In his study he states that by providing balance
of payments, new financial resources should be generated and current debts be
repaid before further debts are incurred. For this purpose, while export
revenues are increased, import expenditures should be decreased. Moreover,
tourism revenues need to be increased and direct foreign investments fostered.
Furthermore, internal borrowing with its positive effect on internal savings
and with less foreign exchange risk, may be preferable to external borrowing.
Sugözü and Yiyit (2010: 371), on the other hand, in agreement with Classical
Economists say that borrowing should be the last financial choice made and only
then under obligatory circumstances such as the need for financing huge amounts
of investment due, for example, to natural disasters, and war.
* Assist. Prof.
Dr. Mehmet Behzat Ekinci
Economics, FEAS,
Mardin Artuklu University.
** “External Borrowing and Inflation in Turkey Between
2003 and 2015: A Simple Linear Regression Analysis”, International Journal of
Economics and Financial Issues, 2016, 6(1), 45-54, https://www.econjournals.com/index.php/ijefi/article/viewFile/903/pdf.
Bibliography
Adıyaman,
Ahmet Turan (2006), Dış Borçlarımız ve
Ekonomik Etkileri. Sayıştay Dergisi, 62, 21-45.
Akdiş,
Muhammet (2003), Türkiye’nin Borç
Gelişimi; Sorunlar-Öneriler. Active, 30, 1-16.
Altunöz, Utku (2013), Türkiye’de Enflasyon,
Büyüme ve Finansal Derinleşme İlişkisinin Ampirik Analizi, Kahramanmaraş Sütçü
İmam Üniversitesi İİİBF Dergisi, 3:2, 175-195.
Burguet,
Roberto; Ferna´ndez-Ruiz, Jorge (1998), Growth
through taxes or borrowing? A model of development traps with public capital.
European Journal of Political Economy, 14, 327–344.
Cardoso,
Eliana; Fishlow, Albert (1990), External Debt, Budget Deficits, and Inflation,
in: Jeffrey D. Sachs (Ed.) Country Debt and Economic Performance, Volume 2: The
Country Studies - Argentina, Bolivia, Brazil, Mexico, University of Chicago
Press, 318-334 (http://www.nber.org/chapters/c8948).
Central Bank
of Turkey (2015-a), Balance of Payment
Statistics. Retrieved on 16 October 2015 from the Central Bank of Turkey
Web site: http://www.tcmb.gov.tr/wps/wcm/connect/TCMB+EN/TCMB+EN/Main+Menu/STATISTICS/Balance+of+Payments+and+Related+Statistics/Balance+of+Payments+Statisticss/.
Central Bank
of Turkey (2015-b), Outstanding Loans Received from Abroad by Private Sector; Definitions
and Explanations. Retrieved on 16 October 2015 from the Central Bank of Turkey Web site: http://www.tcmb.gov.tr/wps/wcm/connect/bc4ba7ca-2241-46ae-8c6c-589c9aa1faaa/definition.pdf?MOD=AJPERES.
Chaudhary,
Mohammad Aslam; Anjum, Shahid Waseem (1996), Macroeconomic Policies and Management of Debt, Deficit, and Inflation
in Pakistan. The Pakistan Development Review, 35:4, Part II, 773-786.
Demir, Murat; Çevik, Savaş; Beşer, M. Kemal (2005), Kamu Kesimi Finansman Açıklarının Ekonomik
Etkileri: Türkiye Üzerine Bir İnceleme. Marmara Üniversitesi, İİBF Dergisi,
20;1, 247-267.
Demir, Murat; Sever, Erşan (2009), Kamu Açıkları ve Borçlanma İlişkisi: Türkiye,
Azerbaycan, Kazakistan ve Kırgızistan Üzerine Bir Uygulama, Sosyo Ekonomi,
Ocak-Haziran, 8-26.
Dickey,
David A.; Fuller, Wayne A. (1979), Distribution
of the Estimators for Autoregressive Time Series with a Unit Root, Journal
of the American Statistical Association, 74:366, 427-431.
Dickey,
David A.; Fuller, Wayne A. (1981), Likelihood
Ratio Statistics for Autoregressive Time Series with a Unit Root,
Econometrica, 49:4, 1057-1072.
Duran, Mahmut (1996), Kamu Finansman Açıklarının Optimal
Finansmanı, in: Kamu Kesimi Finansman Açıkları: X. Maliye Sempozyumu, Mayıs
1994, Antalya, İstanbul Üniversitesi Basımevi, 435-453.
Evgin, Tülay
(2000). Dünden Bugüne Dış Borçlarımız.
Hazine Müsteşarlığı, Ekonomik Araştırmalar Genel Müdürlüğü, Ankara, Haziran.
Granger,
C.W.J.; Newbold, P. (1974), Spurious
Regressions in Econometrics, Journal of Econometrics, 2, 111-120.
Karakaplan,
Mustafa Uğur (2009), The Conditional
Effects of External Debt on Inflation. Selçuk Üniversitesi İİBF Sosyal ve
Ekonomik Araştırmalar Dergisi, 11:17, 203-217.
Karagöz,
Kadir (2007), Türkiye’de Dış Borçlanmanın
Nedenleri; Ekonometrik Bir Değerlendirme. Sayıştay Dergisi, 66-67, 99-110.
Kim, Yun
Jung; Zhang, Jing (2012), Decentralized
borrowing and centralized default. Journal of International Economics, 88,
121-133.
Lessard R.
D. (1986), International Financing for
Developing Countries. World Bank Staff Working Paper, No:793, Series on
International Capital and Economic Development, Washington DC.
Mackinnon,
J. G. (1991). Criticial Values for Cointegration Tests, in: R.F. Engle and
C.W.J. Granger (Ed.), Long-Run Economic Relationship: Readings in
Cointegration, New York: Oxford University Press.
Makin, A.J.;
Narayan, P.K. (2012), Has international borrowing
or lending driven Australia's net capital inflow?. International Review of
Economics and Finance, 1-10, (http://dx.doi.org/10.1016/j.iref.2012.09.010).
Phillips,
Peter C.B.; Perron, Pierre. (1988), Testing
for a Unit Root in Time Series Regression, Biometrika, 75:2, 335-346.
Prokop,
Jacek; Baranowska-Prokop, Ewa (2012), The efficiency of foreign borrowing: the
case of Poland, in: International Conference on Applied Economics (ICOAE), Procedia
Economics and Finance, 1, 321-329, [http://dx.doi.org/10.1016/S2212-5671(12)00037-8].
Sugözü,
İbrahim Halil; Yiyit, Mehmet (2010), Borçlanmanın
Enflasyona Etkisi Üzerine Teorik Yaklaşımların Temel Özellikleri. Maliye
Dergisi, 158, 365-373.
Statistical
Institute of Turkey (2015-a), Inflation&Price.
Retrieved on 15 October 2015 from the Statistical
Institute of Turkey Web site: http://www.turkstat.gov.tr/UstMenu.do?metod=temelist.
Statistical
Institute of Turkey (2015-b), National
Accounts. Retrieved on 15 October
2015 from the Statistical Institute of Turkey Web site: http://www.turkstat.gov.tr/UstMenu.do?metod=temelist.
Treasury of
Turkey (2015), Public Finance Statistics.
Retrieved on 15 October 2015 from the Treasury of Turkey Web site: http://www.treasury.gov.tr/en-US/Stat-List?mid=738&cid=12&nm=684.
Tuna, Yusuf (2014), Türkiye’de Kamu Kesimi Açıkları. Retrieved on 25 July 2014 from the
Web site http://www.ekodialog.com/Makaleler/turkiyede_kamu_kesimi_aciklari.html.
Ulusoy,
Ahmet; Küçükkale, Yakup (1996), Türkiye’de
Borçların İktisadî Büyüme ve Enflasyon Üzerine Etkisi. Granger Nedensellik
Testi, Ekonomik Yaklaşım, 7:21, 15-25.
Wang, Fuhmei
(2009), The effects of foreign borrowing
policies on economic growth: success or failure?. Journal of Economic
Policy Reform, 12:4, 273-284, (http://dx.doi.org/10.1080/17487870903314567).
Appendices
Appendix I: Monthly CPI in Turkey from 2003-15 (%)
|
Months |
2003 |
2004 |
2005 |
2006 |
2007 |
2008 |
2009 |
2010 |
2011 |
2012 |
2013 |
2014 |
2015 |
|
January |
94.77 |
104.81 |
114.49 |
123.57 |
135.84 |
146.94 |
160.90 |
174.07 |
182.60 |
201.98 |
216.74 |
233.54 |
250.45 |
|
February |
96.23 |
105.35 |
114.51 |
123.84 |
136.42 |
148.84 |
160.35 |
176.59 |
183.93 |
203.12 |
217.39 |
234.54 |
252.24 |
|
March |
98.12 |
106.36 |
114.81 |
124.18 |
137.67 |
150.27 |
162.12 |
177.62 |
184.70 |
203.96 |
218.83 |
237.18 |
255.23 |
|
April |
99.09 |
106.89 |
115.63 |
125.84 |
139.33 |
152.79 |
162.15 |
178.68 |
186.30 |
207.05 |
219.75 |
240.37 |
259.39 |
|
May |
100.04 |
107.35 |
116.69 |
128.20 |
140.03 |
155.07 |
163.19 |
178.04 |
190.81 |
206.61 |
220.07 |
241.32 |
260.85 |
|
June |
100.12 |
107.21 |
116.81 |
128.63 |
139.69 |
154.51 |
163.37 |
177.04 |
188.08 |
204.76 |
221.75 |
242.07 |
259.51 |
|
July |
99.93 |
107.72 |
116.14 |
129.72 |
138.67 |
155.40 |
163.78 |
176.19 |
187.31 |
204.29 |
222.44 |
243.17 |
|
|
August |
100.09 |
108.54 |
117.13 |
129.15 |
138.70 |
155.02 |
163.29 |
176.90 |
188.67 |
205.43 |
222.21 |
243.40 |
|
|
September |
101.44 |
109.57 |
118.33 |
130.81 |
140.13 |
155.72 |
163.93 |
179.07 |
190.09 |
207.55 |
223.91 |
243.74 |
|
|
October |
102.38 |
112.03 |
120.45 |
132.47 |
142.67 |
159.77 |
167.88 |
182.35 |
196.31 |
211.62 |
227.94 |
248.37 |
|
|
November |
103.68 |
113.50 |
122.14 |
134.18 |
145.45 |
161.10 |
170.01 |
182.40 |
199.70 |
212.42 |
227.96 |
248.82 |
|
|
December |
104.12 |
113.86 |
122.65 |
134.49 |
145.77 |
160.44 |
170.91 |
181.85 |
200.85 |
213.23 |
229.01 |
247.72 |
|
Source: Statistical Institute
of Turkey (2015-a). Inflation&Price,
Retrieved on 15 October 2015 from the Statistical
Institute of Turkey Web site: http://www.turkstat.gov.tr/UstMenu.do?metod=temelist.
Appendix II: Monthly PPI in Turkey from 2003-15 (%)
|
Months |
2003 |
2004 |
2005 |
2006 |
2007 |
2008 |
2009 |
2010 |
2011 |
2012 |
2013 |
2014 |
2015 |
|
January |
94.32 |
104.46 |
114.83 |
123.51 |
135.09 |
143.80 |
155.16 |
164.94 |
182.75 |
203.10 |
206.91 |
229.10 |
236.61 |
|
February |
97.28 |
106.17 |
114.81 |
123.83 |
136.37 |
147.48 |
156.97 |
167.68 |
185.90 |
202.91 |
206.65 |
232.27 |
239.46 |
|
March |
100.40 |
108.40 |
117.25 |
124.14 |
137.70 |
152.16 |
157.43 |
170.94 |
188.17 |
203.64 |
208.33 |
233.98 |
241.97 |
|
April |
102.17 |
111.27 |
119.62 |
126.54 |
138.80 |
159.00 |
158.45 |
174.96 |
189.32 |
203.81 |
207.27 |
234.18 |
245.42 |
|
May |
101.53 |
111.24 |
119.23 |
130.05 |
139.34 |
162.37 |
158.37 |
172.95 |
189.61 |
204.89 |
209.34 |
232.96 |
248.15 |
|
June |
99.58 |
110.06 |
119.64 |
135.28 |
139.19 |
162.90 |
159.86 |
172.08 |
189.62 |
201.83 |
212.39 |
233.09 |
248.78 |
|
July |
99.04 |
108.39 |
119.33 |
136.45 |
139.28 |
164.93 |
158.74 |
171.81 |
189.57 |
201.20 |
214.50 |
234.79 |
|
|
August |
98.85 |
109.25 |
121.40 |
135.43 |
140.47 |
161.07 |
159.40 |
173.79 |
192.91 |
201.71 |
214.59 |
235.78 |
|
|
September |
98.90 |
111.26 |
123.40 |
135.11 |
141.90 |
159.63 |
160.38 |
174.67 |
195.89 |
203.79 |
216.48 |
237.79 |
|
|
October |
99.46 |
114.85 |
124.22 |
135.73 |
141.71 |
160.54 |
160.84 |
176.78 |
199.03 |
204.15 |
217.97 |
239.97 |
|
|
November |
101.15 |
115.72 |
121.40 |
135.33 |
142.98 |
160.49 |
162.92 |
176.23 |
200.32 |
207.54 |
219.31 |
237.65 |
|
|
December |
101.78 |
115.87 |
121.14 |
135.16 |
143.19 |
154.80 |
163.98 |
178.54 |
202.33 |
207.29 |
221.74 |
235.84 |
|
Source: Statistical Institute
of Turkey (2015-a). Inflation&Price,
Retrieved on 15 October 2015 from the Statistical
Institute of Turkey Web site: http://www.turkstat.gov.tr/UstMenu.do?metod=temelist.
Appendix III: Regression Analysis Summary Outputs for External Borrowing
and CPI in Turkey from 2003-15
|
SUMMARY OUTPUT |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Regression
Statistics |
|
|
|
|
|
|
|
Multiple R |
0.983118121 |
|
|
|
|
|
|
R Square |
0.966521239 |
|
|
|
|
|
|
Adjusted R Square |
0.965823765 |
|
|
|
|
|
|
Standard Error |
8.803423529 |
|
|
|
|
|
|
Observations |
50 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ANOVA |
|
|
|
|
|
|
|
|
df |
SS |
MS |
F |
Significance
F |
|
|
Regression |
1 |
107395.5926 |
107395.5926 |
1385.744828 |
4.57738E-37 |
|
|
Residual |
48 |
3720.01276 |
77.50026583 |
|
|
|
|
Total |
49 |
111115.6053 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Coefficients |
Standard Error |
t Stat |
P-value |
Lower 95% |
Upper 95% |
|
Intercept |
21.75424209 |
4.083334534 |
5.327567924 |
2.61965E-06 |
13.54414775 |
29.96433643 |
|
X Variable (Ext. Debt) |
0.000546646 |
1.46847E-05 |
37.22559373 |
4.57738E-37 |
0.000517121 |
0.000576172 |
Appendix IV: Regression Analysis Summary Outputs for External Borrowing and
PPI in Turkey from 2003-15
|
SUMMARY
OUTPUT |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Regression Statistics |
|
|
|
|
|
|
|
Multiple R |
0.985088 |
|
|
|
|
|
|
R Square |
0.970398 |
|
|
|
|
|
|
Adjusted R Square |
0.969781 |
|
|
|
|
|
|
Standard Error |
7.67377 |
|
|
|
|
|
|
Observations |
50 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ANOVA |
|
|
|
|
|
|
|
|
df |
SS |
MS |
F |
Significance F |
|
|
Regression |
1 |
92658.05 |
92658.05 |
1573.496 |
2.38E-38 |
|
|
Residual |
48 |
2826.564 |
58.88675 |
|
|
|
|
Total |
49 |
95484.61 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Coefficients |
Standard Error |
t Stat |
P-value |
Lower 95% |
Upper 95% |
|
Intercept |
30.01116 |
3.559362 |
8.431611 |
4.94E-11 |
22.85458 |
37.16773 |
|
X Variable (Ext. Debt) |
0.000508 |
1.28E-05 |
39.66731 |
2.38E-38 |
0.000482 |
0.000533 |