AdiningsihAiThe Impact of Government Debt Issuance on Short-Term Interest Rates in indonesia Gadjah Mada International Journal of Business September-December 2009.
Vol.
No.
1, pp.
301Ae316
THE IMPACT
OF GOVERNMENT DEBT ISSUANCE
ON SHORT-TERM INTEREST RATES
IN INDONESIA*
Sri Adiningsih Faculty of Economics & Business Universitas Gadjah Mada This paper analyzes whether the expansionary fiscal policy funded by issuing debt instruments in financial markets will increase short-term interest rates.
If the expansionary fiscal policy increases interest rates, which decrease private spending especially investment, crowding out occurs.
This is interesting because global economic crisis has encouraged many countries to run large budget deficits to stimulate the economy.
Indonesia has also run budget deficit during this crisis and even in years before.
The impact of such a policy can be significant because IndonesiaAos debt market is still narrow and shallow.
Therefore, its capability of absorbing the government debt instruments without influencing the private sector funding is limited.
This study tests whether the crowding out occurs in Indonesia using a time series econometric model inspired by Cebula and CuellarAos model.
The Cointegration Regression and Error Correction Model (ECM) are used in this study.
Monthly data from April 2000 to December 2008 are used for overnight real interbank call money interest rates, real net government bond issues in trading, real narrow money supply, real rate of one-month Data used in this paper are collected from Statistik Ekonomi Keuangan Indonesia (SEKI) Bank Indonesia 2000-2008 or by URL: http://w.
id/web/id/Statistik/Statistik Ekonomi dan Keuangan Indonesia/Versi HTML/ Gadjah Mada International Journal of Business.
September-December 2009.
Vol.
No.
Certificate of Bank Indonesia, growth of Gross Domestic Product, and real net international capital flows.
This empirical study shows that the crowding out problem occurred in Indonesia during the This indicates that financing budget deficit in Indonesia by issuing debt instruments in the financial markets has a negative impact on the private sector.
Keywords: crowding out.
ECM.
government debt instrument Introduction An economic crisis carries the same consequence for both developing and developed countries, which is an increase in government spending to prevail over the crisis.
The large increase in government spending in a short period of time has made fiscal deficit increase rapidly.
The size of the deficit has also created an increase in national debt, especially if the economic recovery process takes a long time and huge costs.
Indonesia, which required a long process to recover from the Asian crisis, even had to spend more than 50 percent of its GDP to foster its banking industry in the 1997/ 1998 crisis, which greatly increased its national debt.
It was worsen by the fact that as an emerging market.
Indonesia did not have broad and deep financial markets.
Its fixed-income assets were undeveloped and its sovereign debt market had just started when the government issued government bonds for bank restructuring.
Therefore, it will be interesting to learn the impacts of sovereign debt issuance on private sectors in shallow and narrow financial markets, such as those in Indonesia.
The impact of excessive use of sovereign debt policy on corporate access to external financing has long been the center of public debates.
The recent economic crisis, believed to be the most severe since the Great Depression of 1930s, has made the issue even more relevant as many countries have been forced to issue fiscal stimulus packages in their efforts to prevent their economies from larger economic There is little doubt that the use of fiscal stimulus packages by many governments may unexpectedly worsen economic performance so as to have negative implications for public finance.
This is because while the economic crisis has lowered state income especially from tax revenues, expenditure has soared due to the need for stimulating the economy.
Large state budget deficits raise concerns over crowding out private investment.
This topic has been a common concern among economists as an article in The Economist on June 13 -19 .
9: .
The article expresses such concern by noting that:
GovernmentsAo thirst for funds will eventually crowd out private investment and reduce economic growth.
AdiningsihAiThe Impact of Government Debt Issuance on Short-Term Interest Rates in indonesia It is a stance that the IMF Staff Position Note .
Large government debts have been a public issue in Indonesia over the last IndonesiaAos debt in 1999 was IDR940 trillion, increased to IDR1,229.
5 trillion in 2004 and reached IDR1,700 trillion in March 2009 (Directorate General of Debt Management 2.
In the last five years, net issuance of government securities reached more than IDR80 trillion annually, 90 percent of which was in IDR (Rupia.
and the rest was in other Regarding the thin bond market in Indonesian financial markets, worries are now coming to the fore that the increasingly large public financing funded by issuing debt instruments will adversely affect fund availability to the non-financial sector .
eal secto.
Such concerns have not been allayed by the fact that, in just a decade after the 1997 economic crisis, the highest economic growth rate of the Indonesian economy was registered at 6.
3 percent in 2007, lower than the average growth per annum before the 1997 crisis .
5 percen.
IndonesiaAos debts also continue to rise, although its share to GDP has shown a downward trend from 89 percent in 2000 to 33 percent in 2008.
Outstanding government debts for the period of 2004-2009 are shown in Table 1.
In the last five years, the Indonesian government has become increasingly dependent on the financial markets in the form of treasury bonds and bills to finance its deficit.
Prior to the period, the Indonesian government used external debts such as the World Bank.
Asia Development Bank, and bilateral loans from Japan.
France.
Table 1.
Outstanding Government Debts, 2004-2009 Total central
government debt
n billion US$) Total central
government debt
quivalent in billion IDR) March
2009***
1,229.
50 1,313.
1,302.
16 1,389.
1,636.
74 1,700.
Securities in IDR Exchange rate (IDR/US$.
11,575
Source: Directorate General of Debt Management.
Ministry of Finance Indonesia .
Note: * Preliminary.
** Very Preliminary.
*** Very very preliminary, as of March 2009 Gadjah Mada International Journal of Business.
September-December 2009.
Vol.
No.
Table 2.
Size and Composition of Emerging East Asian Local Currency Markets in 2008 (% of GDP) People Republic of China Hong Kong.
China Indonesia Republic of Korea Malaysia Philippines Singapore Thailand Vietnam Total Emerging East Asia Total Government Corporate Source: Asia Bond Monitor.
First Quarter .
ADB Germany, and U.
The change in deficit financing strategy from using external debts to issuing government bonds, from the political vantage point, is very popular among Indonesians as they mostly oppose external debts.
They suspect that loans are never free from political interests of creditor countries, and some hold the view that conditionality which often accompanies the disbursement of external loans does not benefit Indonesia.
However, the recent budget deficit of around one percent to two percent of GDP is financed by issuing debt instruments in the financial markets.
To that end, it is imperative to have a good understanding of the impacts that financing deficit by issuing debt instruments in the financial markets have on the Indonesian economy.
This is particularly true given the fact that bond market in Indonesia is still thin.
Table 2 shows that the size of IndonesiaAos bond market is the smallest in emerging East Asia.
The development of bond market in Indonesia started a decade ago when the government issued bonds tailored to recapitalize banks, in the aftermath of the banking crisis.
Literature Review Crowding out is defined by Investopedia as .
An increase in interest rates due to rising government borrowing in the financial markets.
When expansionary fiscal policy increases interest rates, which decrease private spending especially investment, crowding out occurs.
In a full employment economy, crowding out can occur easily, and in this case it is a full AdiningsihAiThe Impact of Government Debt Issuance on Short-Term Interest Rates in indonesia crowding out.
In the classical case, an increase in government deficit leads to full-fledged crowding out.
However, in an economy below the full employment level, crowding out occurs not in the form of a full crowding out.
The rise in interest rates occurs even in an economy characterized by unemployed This is because whenever a rise in the aggregate demand leads to an increase in income, the increase in income induces a rise in savings.
However, the rise in savings can not finance a larger budget deficit without completely displacing private borrowing.
In this way, interest rates will rise as the budget deficit increases.
As a result, crowding out occurs even in an unemployment economy.
In an unemployment economy, there is a possibility that even if the budget deficit increases, crowding out may not occur.
In this case, there is a possibility that the deficit induces an increase in output, which precludes a rise in interest This happens if the monetary authority accommodates fiscal expansion by raising the money supply, which in turn prevents a rise in interest Such a policy is referred to as monetizing budget deficit (Dornbusch and Fisher 1990:149-.
So far, the discussion on crowding out has focused on the demand side.
However, from the classical perspective, supply side is also used as an explanation for the crowding-out effect.
In this case, fiscal expansion is posted to increase demand, which leads to firms experiencing excess demand for goods, sparking an increase in prices rather than The firms increase prices until excess demand is eliminated and reaches the full employment level of At such a level of output, real balances decrease while interest rates rise induces a reduction in private spending to make rooms for an increase in government spending (Dornbusch and Fisher 1990:219-.
Crowding out creates problems when the government spends more money and finances it by borrowing money in the financial markets using debt securities.
These induce a rise in market interest rates, which leads to private sector difficulties in raising external financing.
This occurs because government debts are considered risk-free assets coupled with the fact that the government pays the market interest rates.
Nonetheless, at some point, as government borrowing increases, the private sector and individuals no longer obtain sufficient funds in the market.
In that event, competition pushes market interest rates higher.
Temporary government deficit due to the issuance of debt instruments in the market, which will induce an increase in interest rates in the event of an increase in government consumption, substitutes private consumption as long as the substitution is less than one for one (Barro 1.
Many economists have made extensive research on whether increasing government budget deficit influences the private sector (Carlson and Spencer .
Plosser .
Hoelscher .
Barth et al.
Evans .
Tanzi .
Barro Gadjah Mada International Journal of Business.
September-December 2009.
Vol.
No.
Barth et al.
Hoelscher .
Barro .
Zahid .
Cukierman and Meltzer .
Ostrosky .
Cebula .
Arteta and Hale .
Cebula and Cuellar .
Trebesch .
Some economists find that there are some impacts while some others find no influence of the deficit on the private sector.
Research on the impact of state budget deficit on the rate of interest has also received extensive attention.
In a study by Cebula and Cuellar .
using quarterly data for the period of 1973.
4 in the U.
, it is revealed that the federal budget deficit, expressed as a percent of GDP, has a positive and statistically significant impact on the ex-ante real interest rate yielded on MoodyAos Baa-rated corporate bonds.
PlosserAos .
study using U.
also indicates that the capital markets are not indifferent with respect to the level of government expenditures as higher interest rates are associated with increases in government purchases.
Carlson and SpencerAos .
study shows that crowding out occurs in real rather than nominal terms.
TanziAos .
study in the U.
shows that fiscal deficit and .
the level of public debt positively influence interest rates.
Lower fiscal deficit levels are generally associated with somewhat lower interest rates or at least being equal.
Hoelscher .
shows that larger deficit increases the slope of the yield curve in the U.
However, no relationship between short-term rates and deficits is found, while long-term rates increase with higher deficits.
Therefore, deficits push private borrowers into short-term markets for funds.
Consequently, longterm capital spending projects are reduced and the rate of economic growth Zahid .
shows that between 1971 and 1980 in the U.
government budget deficit was defined appropriately to reflect the government excessive demand for funds from the non-government public, and the counter-cyclical variations in the deficit figures were adjusted, a significantly positive impact of deficits on real interest rates is established.
CebulaAos .
empirical results (U.
show that federal budget deficits have a positive and significant impact on the ex-ante real interest rate yielded on ten-year Treasury notes.
MoodyAos a-rated long-term bonds, and MoodyAos Baa-rated long-term corporate bonds.
To the extent that private capital formation is sensitive to such interest rates, the above findings imply the possibility of at least some degree of Aucrowding outAy.
Nonetheless, the results of the subject have differed among researchers.
Some studies find no impact of government deficit on private sectors spending.
Barro .
in his research using U.
data for the period of 1983-1987 shows that Ricardian equivalence theorem, which shows the substitution of a budget deficit for current taxes .
r any other rearrangement of the timing taxe.
, has no impact on the aggregate demand for goods.
In other words, budget deficits and taxation have equivalent effects on the economy.
AdiningsihAiThe Impact of Government Debt Issuance on Short-Term Interest Rates in indonesia Therefore, there is no effect on investment, and no burden of public debt.
the other hand.
Hoelscher .
shows that there is no measurable correlation between government borrowing and short-term rates.
There is no evidence of significant relationship between federal borrowing and short-term interest rates for the post-WWII period in the U.
The principal determinants of short-term rates are: expected inflation, monetary factors, and economic activities in the economy.
Private expenditure is sensitive only to shortterm rates, meaning that federal borrowing does not have financial crowding out effects.
Evans .
, using three-period data during Civil War.
World War I.
World War II, and postwar periods in the U.
, shows that evidence does not substantiate the paradigm that large deficits produce high interest rates.
Evidence shows strong supports for a negative association between the two variables rather than a positive one.
This paper should not be considered to be supporting deficit spending, as Barro reveals that the government ought to run whatever deficit or surplus necessary to flatten the expected future profiles of its marginal tax rates.
Barro .
shows that, using British data from early 1700s through World War I, because of the exogeneity of the deficits, interest rates showed no special movement at that Ostrosky .
shows that, unlike the finding of Cebula .
, the federal deficit did not have a significant impact on the nominal interest rate in the U.
In light of this foregoing, it is evident that the impact of government deficit on interest rates to this day is far from clear.
Some studies indicate that increasing government debts in the financial markets induces a rise in interest rates, while others show no impact.
Therefore, the impact of government debt on interest rates remains a debatable point.
There is still a need for better knowledge of the impact of government debt acquired through the issuance of debt instruments in the financial markets on interest rates and the economy.
Therefore, to achieve high economic growth in the future, it is important for Indonesia to have a better understanding of the influence that the issuance of government debt instruments has on the economy.
Methodology Model The model used this study is inspired by previous studies on the subject, especially the one developed by Cebula and Cuellar .
Cebula and Cuellar provide recent empirical evidence on the impact of federal budget deficit on ex-ante real interest rate yielded on MoodyAos Baa-rated corporate bonds.
Cebula and Cuellar Aos model uses ex-ante real short-term interest rate.
M1 money supply, net international capital inflows, and unemployment rate as variables.
Developed from the model, the basic model to be employed in this study is as follows:
Gadjah Mada International Journal of Business.
September-December 2009.
Vol.
No.
yieldt= 0 1rgbt 2m1t 3sbit 4gpdbt 5ncit At .
puabt= 0 1dagangt 2m1t 3SB1t 4gpdbt 5ncit t .
A puabt= A yieldt= real average interest rate yield
(%),
A rgbt = real net government bond issues .
illions of Rupia.
A m1t = real M1 money supply .
illions of Rupia.
A sbit = real rate of one-month Certificate of Bank Indonesia/ Sertifikat Bank Indonesia
(SBI) (%)
A gpdbt = real growth of Gross Domestic Product (GDP) (%) A ncit = real net international capital flows .
otal of financial and capital account.
illions of Rupia.
A At = the error term real interbank call money interest rate (%) A dagangt = real net government bond issues in trading .
illions of Rupia.
A m1t = real M1 money supply .
illions of Rupia.
A SB1t = real rate of one-month Certificate of Bank Indonesia/ Sertifikat Bank Indonesia
(SBI) (%)
A gpdbt = real growth of Gross Domestic Product (GDP) (%) real net international capiA ncit = tal flows .
illions of Rupia.
the error term A t = This study uses overnight interbank call money interest rate as the proxy for real market interest rate .
This is deemed necessary because of insufficient data on yielded rate of corporate bonds traded in Indonesia.
The rates reflect short-term market rates, hence they can be used to proxy market rates that should be paid by the private sector when borrowing funds in the market.
As the proxy for government external financing in the market or real net government bond issues .
, we use the outstanding government bonds portfolio in trading purpose.
Therefore, the basic model (Equation .
is estimated as follows:
This study uses monthly data .
4 Ae 2008.
collected from the Economic and Financial Statistics (SEKI) of Bank Indonesia 2000-2008 or on website: http://w.
web/id/Statistik/Statistik Ekonomi dan Keuangan Indonesia/Versi HTML/ All data are monthly data, except the data of Net Capital Inflow (NCI) and GDP, which are quarterly.
Consequently.
NCI and GDP data should be interpolated to be monthly data before being However.
NCI data are available only from 2000, so the interpolation can only be conducted from the fourth month of the year 2000.
The Data AdiningsihAiThe Impact of Government Debt Issuance on Short-Term Interest Rates in indonesia interpolation method used is as follows (Insukindro 1.
PDBJan2000 = 1/3 [PDBfirstquarter of 2000 (-1/.
x (PDBfirstquarter of 2000 - PDBfourth quarter of 1999 )] by dividing data with consumer price index in 2002 as the base year and multiplying them with 100, whereas real interbank rate and SBI rate are calculated by subtracting them with Analysis Results PDBFeb2000 = 1/3 [PDBfirst quarter of 2000 .
x (PDBfirst quarter of 2000 - PDBfourth quarter of 1999 )] PDBMar2000 = 1/3 [PDBfirst quarter of 2000 .
x (PDBfirstquarter of 2000 - PDBfourth quarter of 1999 )] PDBDec2000 = 1/3 [PDBfourth quarter 0f 2000 .
x (PDBfirst quarter of 2000 - PDBthird quarter of 1999 )] Data of all variables are in IDR (Rupia.
, except of NCI which are in USD.
Thus.
NCI data on mid-rate USDRupiah must be transformed into Rupiah.
Values of all variables are real Real GDP.
NCI, government bond position, and M1 are calculated The model used in this research is Engle-GrangerAos Error Correction Model (ECM).
The first step of conducting Engle-GrangerAos ECM is to carry out the unit root test, which aims at establishing whether data are stationary or otherwise.
If a time series is stationary, its mean, variance, and autocovariance .
t various lag.
remain the same no matter at what point we measure them (Gujarati 2003:.
One of the unit root tests often used is the Dickey-Fuller (DF) and Augmented Dickey-Fuller (ADF).
Unit root test of level data produces a unit root for all variables used in the study (Table .
This implies that there is a need for conducting the first-difference unit root Table 3.
Unit Root Test: Level Variable Lags ADF Inference Dagang Gpdb Nci Puab Sbi0 unit root unit root unit root unit root unit root unit root Gadjah Mada International Journal of Business.
September-December 2009.
Vol.
No.
Table 4.
Unit Root Test: First Difference Variable Lags ADF Inference Dagang Gpdb Nci Puab Sbi Table 5.
Cointegration Test Dependent Variable: puab Independent Variables Coefficient t statistic Constant Dagang Sbi Gpdb Nci CRDW = 0.
DF = -5.
ADF = -3.
Results in Table 4 show that the first-difference unit root test produces significant estimates, leading to the inference that the data are stationary.
To ensure that the model produces valid and unbiased estimates, the next step involves a co-integration test.
Granger Represent Theorem underlines that if all variables used in the estimation are co-integrated, the correct dynamic model to employ is an ECM (Engle and Granger 1.
Cointegration test is done using Dickey Fuller (DF) test and Augmented Dickey Fuller (ADF) test.
The estimation of residuals shown in Table 5, generated by the equations using DF and ADF tests at 1 Ae 10 percent significance levels, produces stationary outcomes which clear the way for using the Engle-GrangerAos ECM (Engle and Granger 1.
table for 1, 5, and 10 percent are 5.
58, and 4.
26, respectively, whereas ADF table for 1, 5, and 10 percent are 98, 4.
36, and 4.
06, respectively.
AdiningsihAiThe Impact of Government Debt Issuance on Short-Term Interest Rates in indonesia The following dynamic regression equation uses the co-integrated regression in the long run and ECM regression in the short run.
The ECM model used in this study is as follows:
= 1D.
3D(SBI.
6et2t-1 The results of the estimation using cointegration regression and ECM are presented in the Table 6.
Table 6 shows that the ECM equation can be analyzed further because an absolute value of ET2 (-.
is smaller than 1 and significant, which implies that the equation is valid and therefore can generate robust estimates (EngleGranger 1.
In the short run, real SBI rate and real net government bond issues have positive and significant influences on real interbank rate.
Other variables have no significant influence on the rate.
In the long run, they have the same results .
ositive impact.
The ECM results show that an increase in real SBI by 1 percent induces an increase in real interbank rate 9 percent, and an increase in real government bond by IDR1 billion induces an increase of 0.
002 percent in real interbank rate.
The ECM estimation, afterwards, is tested through Jarque-BeraAos normality test.
Breusch-GodfreyAos Correlation LM autocorrelation test.
RamseyAos Reset linearity test.
White No Cross Term homoskedasticity test, and multicollinearity test with Koutsoyiannis approach.
The tests show that the ECM estimation is free from specification error, auto-correlation, hetereoskedasticity, and multicollinierity.
However, the residuals of ECM are not normally distributed as the data are limited, but it is believed that the series will be distributed normally if the number of observation increases by the Central Limit Theo- Table 6.
The Results of Estimation ECM Variable D(DAGANG) D(M.
D(SBI) D(GPDB) D(NCI) ET2(-.
Long-run t statistic Variable Coefficient t statistic 001812 *** 1.
DAGANG
SBI
GPDB
NCI
Note: significance: *=1 percent, **=5 percent, ***=10 percent Gadjah Mada International Journal of Business.
September-December 2009.
Vol.
No.
rem (Gujarati 2003:.
Therefore, this study assumes that the residual is normaly distributed.
Conclusion This empirical study shows that the crowding out effect was evident in Indonesia between 2004 and 2008.
indicates that financing state budget deficit in Indonesia by issuing debt instruments in financial markets has a negative impact on the private sector.
Increasing state budget deficit by issuing more debt instruments in the market will increase real short-term interest rates.
The results show that there is a limit on the capacity of the financial markets to absorb debt instruments.
The government of Indonesia should manage its debt issuance better by considering the capacity of the market to absorb the debt instruments without creating a negative influence on the private sector.
As discussed earlier, the Indonesian bond market is still thin, which implies that the issuance of government debts should be constrained by the absorbing capacity of the financial markets if Indonesia does not want to hamper economic growth when issuing government debts in the To determine whether there is crowding out in the commercial banksAo credit markets disbursement, the author also checks the effect of government debt issuance on real rates of investment credit in IndonesiaAos commercial banks.
The results indicate that despite its positive sign, the issuance has no significant effect on the credit rates.
This is presumably because the rates of commercial bank investment credit are not only influenced by market interest movement but also by various components.
Some components that allegedly affect the formation of investment credit rates are risk premiums, administrative costs, long maturity, and bank profits.
In this way, the development of market interest rates is not automatically reflected in the development of commercial banksAo investment credit rates.
is estimated that the impact of crowding out on the credit rates will be noticeable when using the data of shorter period and small risk money market interest rates.
References