The Evolving Role of Central Banks
Chapter

24 Effectiveness and Implications of Limits on Central Bank Credit to the Government

Editor(s):
Patrick Downes, and Reza Vaez-Zadeh
Published Date:
June 1991
Share
  • ShareShare
Show Summary Details
Author(s)
ALFREDO LEONE

The main purpose of this paper is to explore the role and consequences of limiting central bank lending to the government. In the first section, the legal provisions that were in force in the mid-1980s for a sample of over 100 countries are surveyed. Also, data from 22 industrial countries and 22 developing countries are analyzed to investigate actual central banks’ behavior regarding credit operations with governments. In the second section, an analytical framework, describing the budget constraints of the central bank and the government together with a consolidated budget constraint, is presented to investigate the contribution of these limitations to preserve the independence of central banks and the value of local currencies. The third section considers several issues related to the difficulties in designing and enforcing limits on outstanding central bank lending to governments. In particular, it deals with (a) the vulnerability of these limitations, (b) the implications of foreign exchange losses, and (c) possible inconvenience that these limitations may pose for central banks in conducting open market operations. The last section, taking account of previous discussions, states some conclusions and provides some suggestions for more appropriate limitations.

Common Regulations and Decision-Making Processes

The establishment of limits on central banks’ advances to the government or central banks’ holdings of government securities, or both, is a very common practice all over the world.1 These limitations take different forms, however, ranging from strict explicit rules to vague regulations that leave wide scope for discretion.

Common Rules

Limits on cash advances usually differ from those on holdings of government securities, particularly when the securities are purchased by the central bank in the open market. Also, the cost to governments, repayment conditions, and maturity features of the government securities that central banks are allowed to hold are usually explicitly established. The more common among these forms are as follows.

Limits on Cash Advances

(1) Net outstanding central bank loans to the government (in the form of cash advances) should be zero (Austria, Switzerland).

(2) No explicit limits are established, but advances to the government by the central bank require approval by the legislature (France, Norway, Korea) or a federal agency (United States) or the Central Bank Board (Turkey).

(3) No explicit limits are established, but the circumstances under which the central bank may make advances or the purposes of such advances are specified in the central bank’s law (United Kingdom, New Zealand).

(4) No explicit limit is established (Japan).

(5) An explicit limit expressed as an absolute amount in domestic currency is established, usually by a legislative body. This amount may be revised, from time to time (Germany, Greece, Sweden).

(6) A limit is established in the form of a proportion of the central bank’s liabilities (The Gambia, Mozambique).

(7) A limit is established as a percentage of government revenues (Algeria, Argentina, Canada, Israel, Venezuela).

(8) A limit is established as a percentage of government expenditures (Thailand, Spain).

Limits on Holdings of Government Securities

(1) No explicit limit is established (Italy, Norway, Sweden).

(2) No explicit limit is established for open market operations (Canada, Iceland, Ireland, the Netherlands).

(3) No explicit limit is established, but purchases and holdings of government bonds require approval by a legislative body (Korea, Switzerland) or a federal agency (United States).

(4) An explicit limit is established expressed as a proportion of the central bank’s capital and reserves (Belgium, Turkey).

(5) An explicit limit is established by the legislative body, expressed as an absolute amount in domestic currency (Germany).

(6) An explicit limit is established expressed as a percentage of central bank liabilities (Cyprus, Malta, Nigeria).

(7) An explicit limit is established expressed as a percentage of government revenues (Austria, Ethiopia, Paraguay, Peru, Tanzania).

(8) An explicit limit is established expressed as a percentage of government expenditures (Jamaica).

We now turn to real life trying to assess how these legal limitations have worked in practice in a sample of industrial and developing countries.

Common Practices: Empirical Evidence

Among the countries considered in Table 1, there is no uniform design of limits making cross-country comparisons difficult. Some countries have established separate limits for each type of government debt. Other countries use limits on advances alone, or one single limit covering different types of government debt. In some countries, the securities purchased in the open market are excluded from the limits. There are also differences with regard to the base variable used in establishing limits. Some countries use different bases for establishing limits on different types of debt, but government revenues appear to be the most commonly used base variable for these purposes.2

Table 1.Central Bank Claims on Central Governments in Industrial and Developing Countries, 1975–87(In percent)
(1)(3)(4)(5)(6)
Claims/Base(2)Claims/Net Claims/Net Claims/Net Claims/
MoneyClaims/RevenuesExpendituresBase MoneyRevenuesExpenditures
CountryMSDMSDMSDMSDMSDMSD
Industrial Countries
Australia84.322.917.26.016.35.6n.a.n.a.n.a.n.a.n.a.n.a.
Austria5.51.01.90.61.70.53.00.81.00.30.90.2
Belgium21.21.85.00.94.20.9n.a.n.a.n.a.n.a.n.a.n.a.
Canada67.313.018.43.915.73.3n.a.n.a.n.a.n,a.n.a.n.a.
Denmark195.5194.719.917.717.614.8-92.961.3-10.25.5-10.06.1
Finland12.14.61.60.61.50.6-7.79.1-1.21.3-1.21.3
France114.94.72.50.72.40.7-1.87.6-0.31.3-0.31.3
Germany13.61.84.80.74.50.64.52.14.20.63.90.5
Greece70.120.560.120.545.713.856.321.047.818.036.111.8
Iceland63.726.020.57.620.07.527.326.78.68.18.27.9
Ireland21.86.37.41.45.51.03.714.10.55.40.34.0
Italyn.a.n.a.n.a.n.a.n.a.n.a.89.49.469.029.348.920.2
Japan246.88.415.12.713.72.038.810.312.53.311.32.7
Netherlands8.92.41.30.41.20.3-9.011.3-1.31.61.21.6
New Zealand130.228.414.44.913.24.87.694.33.19.83.08.9
Norway61.832.111.35.112.15.2-61.4123.5-7.216.4-8.818.8
Portugal105.045.069.716.354.712.995.144.262.715.949.212.4
Spain53.117.135.4313.0331.139.6347.717.031.8312.7327.939.63
Sweden97.029.718.33.817.33.3n.a.n.a.n.a.n.a.n.a.n.a.
Switzerland6.82.015.23.014.83.32.14.73.711.04.010.7
United Kingdomn.a.n.a.n.a.n.a.n.a.n.a.54.032.88.88.18.17.3
United States83.81.926.53.523.03.478.71.424.93.121.63.2
Developing Countries
Argentina476.448.266.340.855.530.974.647.863.738.953.529.8
Brazil5108.457.219.46.122.45.911.642.10.97.31.18.5
Chile6185.2120.740.221.439.519.8141.6127.129.721.328.618.8
Colombia26.221.325.118.621.414.912.521.710.719.38.715.7
Egypt114.725.1103.323.190.926.1108.322.197.320.085.623.4
Ghana135.041.5189.664.1119.424.3127.340.5178.562.3112.525.0
India93.416.493.019.183.312.791.715.691.318.381.812.1
Indonesia28.27.310.63.69.93.2-63.046.1-20.914.9-19.813.9
Israel49.116.735.814.026.59.640.414.429.812.221.88.0
Kenya88.439.334.415.127.611.285.044.732.617.726.113.5
Malaysia21.013.310.35.68.34.7-6.425.2-3.913.1-3.210.8
Mexico97.112.396.020.971.619.3n.a.n.a.n.a.n.a.n.a.n.a.
Morocco84.622.850.812.336.910.583.322.849.912.336.210.4
Nepal95.526.8117.6737.0761.6711.8752.826.662.7734.7731.9713.85
Pakistan92.47.693.67.777.19.083.18.984.511.969.611.8
Peru29.318.927.821.419.712.228.419.027.021.419.112.3
Philippines70.025.739.618.939.419.047.125.926.617.026.616.9
Tanzania113.635.279.135.653.425.1n.a.n.a.n.a.n.a.n.a.n.a.
Thailand96.627.957.910.545.26.884.931.750.314.139.110.1
Venezuela17.88.07.583.288.583.78-8.511.5-3.185.68-3.786.38
Zambia283.0138.1124.954.188.940.6280.2140.6123.755.288.041.4
Zimbabwe949.221.913.16.010.65.239.226.010.16.78.45.8
Source: International Monetary Fund. International financial Statistics.Note: (1) the ratio of central banks claims on government to the stock of reserve money; (2) the ratio of central bank claims on government to government revenues: (3) the ratio of central bank claims on government to government expenditures; (4)–(6) similar ratios with central bank net claims on government replacing gross claims in the numerator of ratios (1)–(3). M = Mean; SD = Standard Deviation; net claims are defined as central bank claims to the government less government deposits with the central bank.

1977–87.

1979–87, data on government revenues and expenditures provided by the Asian Department of the International Monetary Fund.

1975–86

1977–86

1978–85

1978–84

1975–85

1975–86

1976–87

Source: International Monetary Fund. International financial Statistics.Note: (1) the ratio of central banks claims on government to the stock of reserve money; (2) the ratio of central bank claims on government to government revenues: (3) the ratio of central bank claims on government to government expenditures; (4)–(6) similar ratios with central bank net claims on government replacing gross claims in the numerator of ratios (1)–(3). M = Mean; SD = Standard Deviation; net claims are defined as central bank claims to the government less government deposits with the central bank.

1977–87.

1979–87, data on government revenues and expenditures provided by the Asian Department of the International Monetary Fund.

1975–86

1977–86

1978–85

1978–84

1975–85

1975–86

1976–87

Table 1 gives information on the behavior of six different ratios. The information on outstanding central bank claims and net claims on governments used in the calculation of those ratios presented covers all categories of central government debt, such as advances, treasury bills, and other government or government-guaranteed paper. The table consists of the means and standard deviations of those different ratios for the period 1975–87. Low values for means imply that on average central banks were able to enforce narrow limits in their lending to governments during the period considered. Low values for means, together with low values for standard deviations, imply that those limits were generally kept narrow at all times over the 1975–87 period. The importance of keeping narrow limits on central bank lending to governments at all times will be underscored below. High values of means combined with high standard deviations imply, on the contrary, very permissive and erratic central bank policies regarding government financing.

It is clear that the behavior of these ratios differs substantially among the countries included in the sample. Some countries (such as Austria, France, Germany, the Netherlands, and Switzerland) have kept limits narrow and stable. Others (noticeably Greece, Italy, Portugal, and most developing countries) showed relatively high and variable observed ratios. Some countries showed low and stable ratios even though the explicit legal limits did not appear to be very strict (the Netherlands). In some others, the ratios attained such high levels that it is very likely that legal allowances were transgressed. The next section attempts to analyze the implications of these different behaviors.

The available data from the countries in the sample also show a high correlation between means and standard deviations of ratios. This implies that generally more permissive access to central bank credit by governments had led to more instability in the ratios making central bank policies regarding government financing more uncertain.3Table 2 and Chart 1 illustrate this point by analyzing the behavior of the ratio of net outstanding central bank credit to the government over government revenues for the sample of countries considered in this paper.

Table 2.Central Bank Net Claims in Terms of Government Revenues

(Regression outputs for the relationship between mean and standard deviation)1

AllIndustrialDeveloping
CountriesCountriesCountries
Constant6.0003374.0935629.934198
Standard deviation of Y Est8.1464065.2866229.845206
R squared0.6514130.4961950.581942
Number of observations442222
Degrees of freedom422020
X coefficient(s)0.2672510.2402820.234973
Standard deviation of coefficient0.0301660.0541390.044532

Simple ordinary-least-squares regressions were run of the form: Standard deviation = a + b Mean.

Simple ordinary-least-squares regressions were run of the form: Standard deviation = a + b Mean.

Chart 1.Central Bank Net Claims in Terms of Government Revenues

(In percent)

An Analysis of the Effectiveness and Implications of Limits

In this section the effectiveness and implications of limits on central banks lending to governments will be analyzed using as the starting point the budget constraints specified in detail in Appendix II.4

On Budget Constraints and Limits

For the purposes of this section the three equations for budget constraints will be written in a more simplified way than in the appendix, as follows:

Government budget constraint

Central bank budget constraint

Consolidated budget constraint

Given that

net credit from central bank (in equation (1)) = net credit to the government (in equation (2)),

it follows that:

Equation (1), the government budget constraint, shows that the government current operational deficit, the interest payments on outstanding government net debt (the accumulated result of government deficits incurred in the past), and the acquisition of foreign assets and other net assets by the government determine its total financial requirements. To satisfy these requirements, the government may borrow domestically or abroad, facing the market demand functions and implied costs for its debt instruments, or may borrow from the central bank.5

As shown by equation (2) the central bank will increase its own financial requirements in increasing its lending to the government. Its financial requirements will also be affected by its operational expenditures, the interest payments on outstanding central bank net indebtedness (domestic and foreign), its lending to the economy (financial and other institutions), and the acquisition of foreign assets and other net assets. To satisfy these requirements, the central bank may borrow domestically or abroad or may print money, facing the market demands for its debt instruments and for base money, respectively.

The consolidated budget constraint, equation (3), shows that the aggregate financial needs of the government and the central bank result from their operational deficits, the interest payments on the outstanding consolidated net debt, the central bank credit to the economy, and the acquisition of foreign assets and other net assets by the government and the central bank. To satisfy these aggregate financial requirements, the consolidated public sector (government and central bank) may borrow domestically or abroad or may print money, facing the demand functions for their debt instruments and for base money, respectively.

Thus, the availability of credit from the central bank represents an alternative for governments to satisfy their financial requirements. It may represent a very convenient alternative if costs and conditions on borrowing from the central bank are more favorable than those prevailing in domestic or foreign financial markets. The availability of cheap central bank credit for the government will not reduce financial costs for the consolidated public sector but will imply lower costs for the government and higher costs for the central bank, which will face market conditions for its financial needs or may result in a higher inflation rate when government financing is not appropriately sterilized. Moreover, the availability of cheap central bank credit may encourage governments to increase expenditures, jeopardizing macroeconomic stability and deteriorating the financial position of central banks.6

From the previous analysis, it is clear that governments’ fiscal decisions may limit the independence of central banks by affecting their monetary policy decisions and their financial positions. Thus, the existence of legal limits on central bank lending to governments may represent a protective barrier that will contribute to guaranteeing a certain degree of independence for central banks. Appropriately defined limits would allow central banks to follow monetary policies independent of government fiscal decisions and protect their financial positions, particularly when access to central bank credit by governments is cheap relative to market conditions.

Traditionally, as explained before, central bank laws have established limits on the total amount of outstanding central bank credit to the government. Most commonly, these limits have taken the form of a fixed amount in domestic currency (determined periodically) or a proportion of government revenues. These limits imply that the flow of central bank credit to the government during a given period is also limited. It is limited to a fixed amount in domestic currency (equivalent to the difference between the annual limit established for the current year and the outstanding debt at the end of the previous year) as shown in the following relationships:

or

where

L̄(t) = maximum stock of outstanding government debt with the central bank allowed at the end of period t,

L*(t − 1) = outstanding stock of government debt with the central bank at the end of period t − 1,

T(s) = total government revenues during period s (s = t or t − 1); and

α = a constant.

Effectiveness of the Limits and the Independence of Central Banks

Are the kinds of limits commonly established in central bank laws effective enough to guarantee their independence from fiscal decisions and to promote macroeconomic stability? In this section, the question will be addressed from an analytical perspective leaving other aspects for the following section.

Let us first suppose that, at a given time t, all interest-bearing outstanding debt is of the same maturity period. Suppose also that the government runs a deficit at that time and that credit from the central bank is interest free. Then the government budget constraint (equation (1)) can be rewritten as follows:

where

CBCGt(t) = flow of central bank net credit to the government at time t;

[GD(t) − GD(t − 1)] = change in outstanding government net debt outside the central bank: Net borrowing at time t;

D(t) = government deficit at time t;

R = nominal interest rate on outstanding government debt outside the central bank;

AOA(t) = net acquisition of foreign assets and other assets at time t.

Rearranging and dividing all terms by the nominal gross domestic product at time t, that is, by GDP(t), the following first-order difference equation results:

where lower-case letters represent the ratio of the nominal variables with regard to nominal GDP, and

y = rate of growth of real GDP;

r = real interest rate.

Equation (5) shows that if the government deficit and the net acquisition of other assets exceed the availability of credit from the central bank and the real rate of interest exceeds the economy’s rate of growth, then the government debt-to-GDP ratio will grow without limit as shown in Chart 2. Given that the demand for government debt is finite, it will be impossible for the government to sustain this process forever. At some point in the future, it will have to take a decision of reducing the deficit, selling foreign or other assets, or increasing its demand for central bank credit. Let us next see which are the alternatives for the central bank when the government decides to follow the last mentioned, and usually easiest, route.

Chart 2.The Debt Growth Process: An Example

(In percent of gross domestic product)

Following similar procedures as in the case of the government budget constraint, the central bank budget constraint (equation 2) can be rewritten as follows:

where

cbd(s) = ratio of outstanding central bank debt-to-GDP at time s (s = t or t − 1);

cboe(t) = ratio of central bank operational expenditures to GDP at time t;

cbaoa(t) = ratio of flow of central bank acquisition of foreign and other assets to GDP at time t;

cbcg(t) = as before;

cbce(t) = ratio of central bank credit to the economy to GDP at time t;

pbm(t) = ratio of printing of base money to GDP at time t;

Equation (6) is also a first-order-difference equation. It tells us that if central bank expenditures and central bank financing of the economy and the government exceed the printing of base money and, as before, the real interest rate exceeds the economy rate of growth, the central bank interest-bearing debt will grow without limit, as shown in Chart 2. This process is not sustainable in the longer run given the finiteness of the demand for any kind of debt. At some point in the future, the central bank will be forced to reduce its expenditures, sell some of its foreign or other assets, reduce the financing to the economy (crowding-out effect) or the government, or increase the printing of money in order to reverse the process of ever-increasing debt.

The increase in the printing of base money appears to be the easiest route for the central bank to attend the financing requirements of the government and the economy without increasing its own interest-bearing debt and drifting into the explosive process depicted above. But, can the central bank increase the printing of base money without limit? To answer this question let us rearrange equation (6) a little bit, as follows:

where

[bm(t) − bm(t − 1)/(1 + p*)(1 + y)] = pbm(t);

bm(s) = the stock of base money in terms of GDP, at time s (s: t or t − 1);

p* = p(t)lp(t − 1) = the inflation rate;7

y = as before.

Next, let us define q(t) as the demand for base-money holdings in terms of GDP. Let us adopt the conventional assumption that q(t) is a negative function of the inflation rate:

where

In equilibrium the demand for base-money holdings should be equal to the observed stock of base money. Also, the assumption of a constant inflation rate implies that q(t) will also be a constant in this equilibrium. Then replacing equation (8) into equation (7) and rearranging, the following expression results:

A specific case of equation (9) is presented in Chart 3. It can be observed that there is a maximum revenue that the central bank may collect by printing base money.8 Beyond that maximum, any increase in the supply of base money will result in more inflation and less revenue.

Chart 3.Central Bank Revenue from Printing Money: An Example

In summary, both governments and central banks face structural constraints that limit their financing possibilities. Going beyond those structural constraints will produce explosive processes of debt growth or serious inflationary problems. These constraints are best represented by the following consolidated budget constraint (a transformation of equation (3)):9

where

cgd(s) = gd(s) + cbd(s), s: t or t − 1;

cdef(t) = d(t) + cboe(t);

caoa(t) = aoa(t) + cbaoa(t);

cbce(t) = as before;

pbm(t) = as before.

Thus, limiting the scope for central bank lending to governments to very narrow limits would force governments to pursue prudent fiscal policies in order to avoid explosive episodes of government debt growth. In this way, the limits would help central banks to avoid inflationary processes or the explosive growth of their own debt.

The existence, however, of narrow limits on a government’s access to central bank credit at a given time t is only a necessary condition to avoid the kind of explosive and inflationary processes mentioned before. Unfortunately, it is not a sufficient condition. For those processes to be avoided, the scope for central bank lending to the government should be kept within very narrow limits at all times. This requires institutional arrangements that effectively discipline the fiscal authority. As shown before, bond-financed expansive fiscal policies may not be sustainable in the longer run. Under these circumstances, central banks will be unable to control either the growth rate of the monetary base or inflation forever if it is known that sooner or later governments will be financed by printing money when placement of additional government or central bank debt becomes unfeasible. This was already clearly understood at the time of the currency reform taken in Germany, against a background of historical experience with two hyperinflations.10 This is also the message of a theoretical paper by Sargent and Wallace (1981).

Some Cross-Country Experience on Deficits, Debt Growth, and Government Borrowing from Central Banks

Let us now illustrate the analytical framework considered above with the experience in some industrial and developing countries regarding government deficits and central bank behavior. For these purposes, Table 3 presents information on the average values for government deficits, government borrowing from central banks, and printing of money, all in terms of GDP, and the economy growth and inflation rates corresponding to the 1975–87 period in industrial and developing countries, respectively.

Table 3.Government Deficits and Central Bank Behavior in Industrial and Developing Countries, 1975–87
Government NetGrowth in
Borrowing fromPrinting ofGross DomesticInflation
Deficit1Central BanksMoneyProductRate
(In percent of GDP)(In percent)
Industrial Countries
Australia2.50.530.63.09.8
Austria4.50.020.62.04.8
Belgium9.00.110.42.26.4
Canada4.30.370.43.47.7
Denmark2.32-0.500.52.28.4
Finland1.50.060.72.99.3
France2.1-0.140.42.29.1
Germany2.00.020.51.93.5
Greece6.82.443.42.718.3
Iceland2.90.722.73.942.1
Ireland12.3-0.141.33.212.3
Italy11.91.872.22.713.7
Japan2.630.220.64.44.6
Netherlands4.60.150.51.64.7
New Zealand6.3-0.140.11.513.7
Norway1.3-1.460.63.98.8
Portugal10.64.963.62.820.5
Spain4.321.392.12.014.1
Sweden4.80.890.61.88.9
Switzerland0.40.050.61.13.2
United Kingdom4.1-0.350.42.010.8
United States3.70.330.42.76.7
Developing Countries
Argentina45.98.410.70.0271.0
Brazil53.70.42.73.4116.1
Chile60.12.40.63.126.5
Colombia1.70.32.34.023.0
Egypt13.07.67.0n.a.14.9
Ghana5.15.43.90.0558.2
India6.31.81.85.07.0
Indonesia1.3-0.91.36.112.2
Israel14.38.021.12.8125.5
Kenya4.41.41.15.112.3
Malaysia10.6-0.041.25.84.0
Mexico7.14.74.73.553.0
Morocco10.01.51.74.09.0
Nepal4.371.31.53.58.9
Pakistan7.61.92.26.18.4
Peru4.41.44.82.374.4
Philippines2.20.41.13.213.9
Tanzania7.43.32.32.024.1
Thailand3.91.10.96.76.5
Venezuela-0.780.41.62.112.5
Zambia13.95.82.10.0521.9
Zimbabwe98.40.30.92.412.7
Source: IMF, International Financial Statistics.

Deficit = revenues + grants received - (expenditure + lending - repayments).

The figure is for 1975–86.

The figure is for 1979–87; data on government revenues and expenditures provided by the Asian Department of the International Monetary Fund.

The figures are for 1977–86.

The figures are for 1978–85.

The figures are for 1978–84.

The figure is for 1975–85.

The figure is for 1975–86.

The figures are for 1976–87.

Source: IMF, International Financial Statistics.

Deficit = revenues + grants received - (expenditure + lending - repayments).

The figure is for 1975–86.

The figure is for 1979–87; data on government revenues and expenditures provided by the Asian Department of the International Monetary Fund.

The figures are for 1977–86.

The figures are for 1978–85.

The figures are for 1978–84.

The figure is for 1975–85.

The figure is for 1975–86.

The figures are for 1976–87.

Chart 4 illustrates the observed relationship between government deficits and borrowing from central banks in the sample of countries selected.11 It can be observed that in countries with relatively low government deficits, say of less than 5 percent of GDP, most central banks were able to keep relatively low lending-to-GDP ratios. On the contrary, the proportion of countries enforcing low lending to governments when deficits exceed 5 percent of GDP is much lower.

Chart 4.Government Deficits and Credit from Central Banks

(In percent)

Chart 5 shows the relationship between central bank lending to governments and printing of base money.12 It seems that a close positive association exists between these two variables. In fact, in many countries in the sample, the printing of base money has been, on average, equal to government borrowing from central banks over the 1975–87 period. One possible interpretation of this result is that, in the long run, other sources of base-money expansion were relatively insignificant.

Chart 5.Central Bank Credit to Government and Printing of Money

(In percent)

Chart 6 offers a sort of consolidated picture. Again, government deficits representing less than 5 percent of GDP seem to have no effect on the printing of money. This means that in these cases governments have, so far, been able to place debt instruments outside the central bank without facing the structural constraints underscored in the previous subsection or that the rate of growth in those economies has exceeded, so far, the real interest rate on government debt instruments. For deficits exceeding 5 percent of GDP, it seems likely that the monetary authorities will end up printing money. In addition, Chart 7 shows that, among the countries in the sample selected, printing money in excess of 2 percent of GDP a year may become troublesome in terms of inflation rates.

Chart 6.Government Deficits and Printing of Money

(In percent)

Chart 7.Inflation and Printing of Money

(In percent)

There are some countries in the sample, however, that have relatively high government deficits and relatively low printing of money,13 that is, relatively high government borrowing outside the central bank (Belgium, Ireland, Italy, Malaysia). Taking account of the different growth rates presented in Table 3 and assuming a real interest rate of about 3 percent, it seems possible that some of these countries may be facing in the medium-term debt-growth processes of the kind depicted above, requiring significant fiscal adjustments.14

Complexities in Designing and Enforcing Limits for Central Banks’ Lending to Governments

Vulnerability of Legal Limits

The establishment of legal limits on central bank claims on governments does not necessarily guarantee that central banks will be safe from pressure from governments looking for borrowing alternatives. Even in cases where the established limits are entirely respected, central banks may end up providing the required financing to governments through indirect mechanisms that “legally” circumvent the legal constraints.

If the legal limits are not clearly defined or they do not include all possible forms of government debt with the central bank, it will be relatively easy for governments to issue debt instruments not affected by the limitations and sell them to central banks particularly when financial needs exceed other available borrowing alternatives. Also, central banks may provide financing to banks (private and public) or to business or individuals for them to buy government debt. These operations will be reflected in an increase in central bank credit to the economy (or in the acquisition of other net assets) in the central bank accounts and would not be affected by the usual legal limitations on central bank credit to the government.

In some countries, governments have forcefully placed long-term debt instruments in the portfolio of different institutional investors (particularly, in social security systems or provident funds) at low interest rates. In this way, governments were able to avoid both limitations on financing from central banks and, also, market-related interest rates when borrowing. The main implication of this procedure is the decapitalization of those institutional investors. The limit to this source of deficit financing is given by the size of the capital of these institutions, and sooner or later the government will have to recapitalize them. From this perspective, and particularly if this debt is not marketable and not remunerated at market rates, the placement of government debt under these conditions is another way of postponing fiscal discipline.

The experience of some countries shows that imagination to circumvent legal limits on central bank lending to governments has proved to be almost unlimited. Following the analysis of the previous section, we should emphasize, however,. that circumvention of legal limits only transfers to central banks the task of looking for borrowing alternatives. At an aggregate level, the public sector will have to limit its financial requirements to the structural constraints imposed by the demands for money and government debt or the economy will drift into explosive debt-growth and inflationary processes. Moreover, it is frequently the case that central banks’ profit-and-loss accounts deteriorate as a consequence of this intermediate activity tailored to finance governments.

Implications of Foreign Exchange Losses

In many countries, central banks have also been intermediaries for foreign financial resources. In many cases, central banks have on-lent foreign financial resources to the government and other domestic borrowers in domestic currency and at low interest rates. These on-lending conditions were generally insufficient to cover the costs of borrowing abroad and the credit and exchange rate devaluation risks leading to foreign exchange losses for the central bank.

Central banks in many countries have also incurred foreign exchange losses by extending exchange rate guarantees for payments related to imports and foreign borrowing by domestic borrowers (both public and private). Under this scheme, central banks grant (generally for a specified premium) a guarantee of foreign exchange at a certain price on a given future date. These guarantees have no immediate significant effect on either the profit-and-loss account or the balance sheet of central banks. In many countries, however, where the exchange rate devaluation exceeded the premium, exchange rate guarantees have eventually resulted in very large losses for central banks.

Foreign exchange losses usually become an important item among the assets of central banks. This is so because, in fact, the accumulated losses represent an additional form of government debt with central banks. Recalling the discussion earlier, it is clear, however, that at a consolidated level foreign losses arise from the misallocation of foreign financial resources (which made foreign currency liabilities greater than foreign currency assets15) or the unrealistic pricing of exchange guarantees (which made expenditures in foreign currency higher than income in foreign currency). Sooner or later, through increased taxes or reduced government expenditures or experiencing explosive inflationary processes, society will have to bear these losses.

The most significant impact of foreign exchange losses occurs when foreign debts or exchange guarantees become due. At that time, the government will have to buy foreign exchange at the current exchange rate to honor foreign debts or exchange guarantees. If, because of misallocation or mispricing, the government is not able to generate enough resources from the investment of its foreign liabilities or from premiums on exchange guarantees, a fiscal surplus in other government operations will be needed to compensate foreign exchange losses. Otherwise, the government will increase its borrowing from domestic or foreign lenders, or from the central bank, contributing to the kind of debt-growth or inflationary processes depicted above.

When foreign exchange losses of significant magnitude arise, governments usually appeal to central banks for financing. The usual case is that central banks end up honoring government liabilities in foreign currency while governments do not provide the necessary counterpart in domestic or foreign currency. In these cases, existing legal limits on central bank lending to the governments may be transgressed. Presumably, this explains why in many countries government debt to central banks arising from these losses were usually explicitly or implicitly disregarded for the purposes of the legal limits. In addition, if this form of government debt becomes nonperforming or its pricing is unrealistic relative to the costs that the central bank faces for borrowing from domestic or foreign markets to compensate the impact of government losses, then the central bank will incur losses of its own.16

The main conclusion of this section is that foreign exchange losses represent a fiscal problem: sooner or later they will require a fiscal solution. Otherwise, either directly (by increasing government debt) or indirectly through the central bank (by increasing central bank debt or the printing of money) foreign exchange losses will contribute to unsustainable debt-growth or inflationary processes. From this perspective, government debt with central banks arising from these losses should be included in the legal limitations to central banks lending to governments. Permanent regulations of this kind will encourage governments to avoid misallocation of resources and mispricing of government services.

The main message is, thus, that foreign exchange losses should be avoided. Many countries, however, are surely incurring foreign losses today as a consequence of past decisions. At the time, when these losses are realized they will probably exceed any (existing or potential) reasonable limitation on government borrowing from central banks. It is obvious that this legacy cannot be absorbed, at least completely, by current limitations. It will be important in this case to set up a strategy to absorb them over time without generating losses to central banks.

These inherited losses should be considered by the fiscal authorities when elaborating the government budget. They should be able to determine, given structural constraints,17 how these losses will be borne by current and future generations through increasing taxes (or reduced current government expenditures) or increasing borrowing. At the same time, any required financing from the central bank should be remunerated at market rates and honored timely.18

Implications for Open Market Operations

Limits on central bank lending to the government may also become binding when the central bank implements open market operations with treasury bills or other government paper for monetary policy reasons or it intervenes in the secondary market to support the prices of government debt instruments.19

We have already seen that central banks’ open market operations and other interventions in government securities markets may imply an indirect way of government borrowing from the central bank resulting from permanent deficient fiscal conditions. In this section, it is assumed that central bank intervention is required “temporarily” to offset excessive fluctuation of security prices reflecting either an excess or a shortage of liquidity, which are not a consequence of the fiscal position of the government.

Let us suppose that an exogenous shock creates excess liquidity in the economy. Then, short-term interest rates will tend to fall. The fall in interest rates may encourage consumption, creating short-term pressures in the goods markets and in the current account of the balance of payments. It may also promote capital outflows of short-term funds. Moreover, the fixed interest rate on long-term government securities will become more attractive, encouraging the demand for these papers and increasing excessively their price. In this case, the central bank will have to intervene by selling part of its stocks of bills and securities. This case, of course, does not pose any problem for the central bank with regard to the legal limits on its holdings of government paper.

Problems may, however, arise under conditions of liquidity shortages. Then short-term interest rates will make the interest rate on long-term paper increasingly unattractive. This will lead to a fall in the demand for long-term paper. This may also discourage consumption and encourage inflows of short-term capital, hence, creating off-setting forces that will gradually push the short-term interest rates to lower levels. However, this adjustment path may take a long time, and the price of long-term government paper meanwhile may decrease excessively. To avoid this effect, the central bank may be forced to intervene by acquiring bills and securities from the market.20 In this case, however, the required action of the central bank may be limited by the prevailing legal limits on its holdings of government paper.

Thus, the main issue to be considered with regard to the implications of legal limits on central bank intervention in the open market is how these limits may be modified or interpreted to allow more flexible monetary management while, at the same time, assuring long-term monetary control.

A technique that helps to minimize excessive but transitory fluctuations in interest rates on government paper when various lending limits and other liquidity requirements are in effect is the use of “repurchase agreements” supported by appropriate accounting conventions. A repurchase agreement is the purchase of a security from another party, who agrees to buy it back at a specified future date and price. In practice, the repurchase agreement is equivalent to a loan from the buyer to the seller of the securities, with government paper serving as collateral. For example, if a liquidity shortage emerges, the central bank may select competitive offers to buy government paper, with the agreement to sell the securities back to the original sellers at a specified future date.

This mechanism will allow central banks to avoid excessive interest rate volatility and, particularly, excessive fluctuation in the price of long-term government paper. Moreover, under generally accepted accounting conventions, government securities acquired by the central bank with a repurchase agreement would be excluded for the purposes of the legal limits on the central bank holdings of government paper. Under these conditions, central bank sales of government securities would not reduce the central bank’s holdings of government paper and, likewise, central bank purchases of government paper would not increase its holdings of government paper. This is so because the ownership of a paper sold under a repurchase agreement remains with the seller, even though it may be temporarily deposited with the buyer.

While this mechanism of repurchase agreements is useful to deal with temporary shortages and surpluses of liquidity, it is not prudent to use them (possibly exceeding legal limitations on central bank holdings of government paper) for permanent shortages or surpluses of liquidity. Liquidity problems of a permanent nature should be clearly identified by central bankers and trigger correcting mechanisms of a permanent nature, such as the modification of reserve requirements, sales of long-term central bank paper, or changes in the rediscount policy.

Recommendations

The problem then is how to design appropriate limitations that will contribute to macroeconomic stability and, at the same time, will allow central banks to perform monetary policy independent of fiscal actions while protecting their financial positions. Previous discussions suggest that it is unlikely to attain these objectives by limiting only government borrowing from central banks without having some kind of limitation on the growth of net government debt. Previous discussions also suggest that limitations on central bank lending to governments should be carefully designed to protect central banks from excessive and cheap financing of governments but, at the same time, without limiting the independence of central banks to perform monetary policy appropriately (which may require operations with government securities).

Thus, in looking for a solution to this problem, it seems appropriate to consider two different kinds of limitations: (a) some form of limitation on total government debt; and (b) appropriately defined limits on central banks lending to governments.

Design of Limitations

The design of the suggested limitations requires careful consideration. In the light of previous discussion, it seems important to address the following questions: (a) How should limits be defined and which variables should be involved in the definition of these limitations? (b) How comprehensive should the limitations be? and (c) How could legal limits be complemented to protect the financial position of central banks?

Definition of Limits

Suggestions for the definition of limits on total government debt and on government borrowing from the central banks is discussed next. The implications of these suggestions for fiscal discipline and central bank independence are also considered.

Limits on Total Government Debt

A prudent and elemental principle in lending establishes that the ratio of payments of loans installments to income should not exceed a certain given percentage. In addition, the amount of payments is a proportion (which depends on interest rate levels and debt maturity) of the total outstanding debt. Appendix III shows that application of these elemental principles and relationships to government borrowing results in a particular form of limitation: total outstanding government debt should not exceed a proportion of government revenues.21

Thus, simple prudential principles make it recommendable to link total outstanding government debt (government debt outside the central bank plus central bank holdings of government debt) to government revenues.22 From this perspective, it does not seem to be safe to link total outstanding government debt to other variables, such as government expenditures or central bank liabilities.

Given the diversity of possible economic environments, it seems difficult to establish a precise proportion of general validity linking total government outstanding debt to government revenues. It does seem possible, however, to state some general principles based on the previous analysis.

In countries facing economic growth rates lower than real interest rates, governments should not borrow. In these cases, the equilibrium value for the stock of government debt is zero. In fact, they should make a fiscal effort to avoid deficits. Even in countries where real interest rates are lower than the economy’s growth rate, increasing borrowing requirements by governments may end up by increasing market interest rate levels more than the growth rate and, thus, increasing the likelihood of drifting into explosive processes. Thus, a limit on outstanding government debt should also be established in these cases.

Limits on Government Borrowing from Central Bank

Even if a limit on total government debt is set along the lines suggested in previous discussions, it is also important to set limits on government borrowing from the central bank. This is so because at any time the monetary authority could be forced to purchase all the fiscal authority’s outstanding bonds. This would create a dangerous expansion of the monetary base or of interest-bearing central bank debt even within the margins allowed by the previously suggested limit on total outstanding government debt.

The experience in the sample of countries considered in this paper shows that, in general, government borrowing from central banks has translated into monetary accommodation, that is, the expansion of the monetary base. It also shows that the margins for monetary accommodation without troublesome inflationary consequences are indeed very narrow. At the same time, any limitation on governments borrowing should leave central banks enough flexibility to perform open market operations as required by the main objective of monetary policy, that is, the safeguarding of the internal and external value of domestic currency.

These considerations lead us to the following set of general principles for limiting government borrowing from central banks: (a) the government should undertake its financing mainly in the open market; (b) only a very limited amount of direct lending to the government by the central bank should be permitted. This limit will represent the maximum degree of monetary accommodation allowed to finance the government. Given that a prudent limit to monetary base expansion is represented by the economy’s real rate of growth, it seems advisable to limit total direct (nominal) government net borrowing from central banks to the economy’s real rate of growth;23 and (c) the central bank should be allowed to buy and sell government securities in the open market as required by its monetary policy objectives. It is recommended that total outstanding net government debt in the central bank (cash advances plus holdings of government paper by the central bank) should not, however, exceed a proportion of total government revenues. Any additional requirement for purposes of monetary policy could be handled through transitory repurchase agreements in conjunction with appropriate accounting conventions.

Implications of Suggested Limitations

It is clear that, according to the budget constraint (equation (1)), the first limitation (the limit on total outstanding government net debt) implies a limit on total government financial needs (the sum of (1) the government operational budget deficit; (2) interest payments on outstanding net government debt; and (3) the acquisition of foreign and other net assets by the government). Thus, this limitation is intended to contribute to macroeconomic stability. It does not, however, provide any guidance to regulate government borrowing from central banks. This role is played by the other two limits.24

Limiting the growth of net government debt will contribute to making limits on government borrowing from central bank credible and sustainable in the long run. The suggested limits on central bank lending to governments will protect central banks’ independence without curtailing their ability to perform monetary policy appropriately. The conjunction of these limitations will contribute to fiscal discipline at all times.

Comprehensiveness of Limitations

The limits previously suggested should be clearly specified so as to minimize the likelihood of cheating, constrain discretionary management, and permit appropriate scrutiny. This requires the specification of rules defining the forms of government debt affected by the limitations and the methodology guiding the calculations.

The limit on total outstanding government debt should comprehend all forms of outstanding government and government-guaranteed debt (foreign or domestic) net of government financial assets (foreign or domestic). Limits on total net government outstanding debt with the central bank25 should also comprehend all forms of government debt with the central bank, such as advances, treasury bills, government and government-guaranteed paper, and other forms of credit originating, for example from foreign exchange losses.

For the purposes of these limits, government revenues could include both tax and nontax revenues. Also, government revenues should be calculated as the average value of total tax and nontax collections during some previous period (for example, the three previous fiscal years). For the purposes of the limit on government direct borrowing from the central bank (in nominal terms), the rate of the economy growth should be calculated as the average annual growth rate of real GDP during previous years (for example, during the last three years).

Complementary Regulations

The previous limitations will contribute to long-run macro-economic stability and central bank independence. Even within these limits, the financial position of central banks may still be threatened if lending to governments is made under inappropriate conditions. To minimize this possibility, some complementary regulations may be useful.

Within the margins allowed by the limits previously suggested, central banks should preferably acquire government securities in the open market. Any other form of government borrowing from the central bank should be exceptional and bear market-related interest rates and conditions.

Conclusions

At least in the sample of countries selected in this paper, the experience with legal constraints to limit central banks’ lending to governments has been diverse. In some countries where laws apparently leave more scope for discretionary management, actual central bank lending to governments has remained within narrow and stable limits. On the contrary, in some other countries where legal limitations look more strict, the available data seem to indicate that those limits have been transgressed.

The determination of maintaining fiscal discipline at all times, that is, to keep public sector financial needs within the margins allowed by the kind of structural constraints underscored earlier, seems to have played a key role in facilitating a predictable and prudent central bank behavior regarding the financing of governments. From the experience of the countries selected in this paper, it is difficult, however, to determine if actual legal limitations on central banks lending to governments have contributed to encouraging fiscal discipline. In some cases, these limitations seem to have contributed to limiting government financial needs. In countries with chronic fiscal deficits, legal limits have become unsustainable in the long run. In some others, where fiscal discipline has been a permanent feature of actual behavior, legal limitations appear to be unnecessary.

It has also been underscored that, in general, central bank financing of governments has been a major source of monetary base expansion. In addition, the experiences of the countries in the sample shows that whenever the expansion of the monetary base exceeded some structural constraints, it resulted in relatively high inflation.

In summary, the establishment of legal limitations on government borrowing from central banks looks like a necessary condition to avoid excessive expansion of the monetary base or central banks interest-bearing debt and, thus, to avoid troublesome debt-growth or inflationary processes in the long run. The experience of some countries, particularly those with chronic fiscal deficits, seems to indicate, however, that the establishment of this kind of limitation has not been sufficient to guarantee fiscal discipline. Persistence of fiscal disorder usually makes these limits unsustainable in the long run.

The central bank laws of individual countries were also used in this paper:
Statute
CountryStatuteDate
Anguilla1East Caribbean Central Bank Agreement1983
AlgeriaStatutes of the Central Bank of Algeria1962
Antigua and Barbuda1East Caribbean Central Bank Agreement1983
ArgentinaCharter of the Central Bank, Argentine Republic1977
AustraliaReserve Bank Act 19591982
AustriaNational Bank Law, 19551984
BahamasThe Central Bank of Bahamas Act, 19741974
BahrainDecree law No. 23 of 19731981
BelgiumThe Organic Law of the Bank of Cape Verde, 19761976
BelizeCentral Bank of Belize Act 19821982
Benin2Charter of the Central Bank of West African States1981
Burkina Faso2Charter of the Central Bank of West African States1981
BurundiStatutes of the Bank of the Republic of Burundi1976
Cameroon3Articles of the Bank of Central African States1984
CanadaBank of Canada Act1980
Cape VerdeOrganic Law of the Bank of Cape Verde, 19761976
Central African Republic3Articles of the Bank of Central African States1984
Chad3Articles of the Bank of Central African States1984
Congo3Articles of the Bank of Central African States1984
Côte d’Ivoire2Charter of the Central Bank of West African States1981
CyprusCentral Bank of Cyprus Laws of 1963 and 19791979
DenmarkThe National Bank of Denmark Act, 19361966
Dominica1East Caribbean Central Bank1983
EcuadorLaw on the Monetary System1981
EgyptCharter of the Central Bank 1957 with Decree No. 488 of 19761976
El SalvadorCentral Reserve Bank of El Salvador1982
Equatorial Guinea3Decree Law No. 1/1980 of February 91980
EthiopiaMonetary and Banking Proclamation No. 99/19761976
FinlandRegulations for the Bank of Finland1966
FranceLaw No. 73-7 of January 3, 1973 with Decree Law 73-102 of January 30, 19731973
The GambiaThe Central Bank of The Gambia Act, 1971 with The Central Bank of the Gambia (Amendment) Act, 19781978
GermanyThe Law Concerning the Deutsche Bundesbank, 19571978
GhanaBank of Ghana Act, 19631971
Grenada1East Caribbean Central Bank Agreement1983
GreeceThe Statutes of the Bank of Greece1966
GuineaDecree No. 126/PRG1960
GuyanaBank of Guyana Act1982
IcelandThe Central Bank of Iceland Act1979
IraqLaw No. 64 of 19761976
IrelandCentral Bank Act, 19711971
IsraelBank of Israel law1982
ItalyThe Statute of the Bank of Italy, 19361978
JamaicaBank of Jamaica (Schedule)1977
JapanThe By-Laws of the Bank of Japan1960
JordanThe Central Bank of Jordan law, 19711979
KoreaThe Bank of Korea Act1977
KuwaitLaw No. 32 of 19681980
Lao People’s DemocraticStatutes, National Bank of Laos (New Law 1990)1955
Republic
LesothoCentral Bank of Lesotho Act 19821982
LibyaBanking Law No. 4 of 19631971
MadagascarOrdinance No. 73-0251973
MalawiReserve Bank of Malawi Act1982
MalaysiaCentral Bank of Malaysia Ordinance, 19581982
Mali2Charter of the Central Bank of Mali1973
Charter of the Central Bank of West African States1981
MaltaCentral Bank of Malta Act, 19671981
MauritaniaStatutes of the Central Bank of Mauritania1975
MauritiusBank of Mauritius Act1981
Montserrat1East Caribbean Central Bank Agreement1983
MoroccoStatutes of the Bank of Morocco1962
MozambiqueCharter of the Central Bank1975
Myanmar (Burma)The Union Bank of Myanmar Act, 1952 (New Law 1990)1952
NetherlandsBank Act 19481980
Netherlands AntillesThe Central Bank Statute 19851985
New ZealandThe Reserve Bank of New Zealand Act, 1964 (New Law 1989)1974
Niger2Charter of the Central Bank of West African States1981
NigeriaCentral Bank of Nigeria Act, 19581962
Central Bank of Nigeria By-Laws, 19591962
NorwayAct relating to Norges Bank and the Monetary System1985
OmanLaws of the Sultanate of Oman, 19741975
Papua New GuineaCentral Bank Act, 19731977
ParaguayDecree Law No. 181952
PeruOrganic Law of the Central Reserve Bank of Peru1983
PhilippinesThe Central Bank Act1979
PortugalBanco de Portugal Organic Law (No. 644/75)1976
RwandaDecree-law No. 06/81 of February 16, 19811981
Sao Tome and PrincipeOrganic law National Bank of Sao Tome and Principe1976
Senegal2Charter of the Central Bank of West African States1981
Sierra LeoneBank of Sierra Leone (Amendment) Act, 19781978
SomaliaSomalia National Bank Law, 19681968
SpainBank of Spain Law, 18/19621975
Sri LankaMonetary Law Act1979
St. Kittsand Nevis1East Caribbean Central Bank Agreement1983
St. Lucia1East Caribbean Central Bank Agreement1983
St. Vincent1East Caribbean Central Bank Agreement1983
SudanBank of Sudan Act, 19591962
SurinameBank Ordinance Act, 19591962
SwedenSveriges Riksbank Act1983
SwitzerlandNational Bank Law, 19531982
TanzaniaThe Bank of Tanzania Act, 19651971
ThailandBank of Thailand Act, B.E. 2485 (1942)1962
Royal Decree Reg. Bank of Thailand, B.E. 24851979
Togo3Charter of the Central Bank of West African States1981
Trinidad and TobagoCentral Bank Act, 19641978
TunisiaCentral Bank of Tunisia Law No. 58-90 of 19581975
TurkeyLaw No. 12111970
UgandaThe Bank of Uganda Act, 19661971
United Arab EmiratesCentral Bank Law1980
United KingdomThe Bank of England Act, 18191819
The Bank of England Act, 19461946
United StatesFederal Reserve Act1983
VanuatuThe Central Bank of Vanuatu Act1982
VenezuelaLaw of the Central Bank of Venezuela1984
Viet NamOrdinance No. 48, 19541954
Western SamoaCentral Bank of Samoa Act 19841984
Yemen Arab RepublicCentral Bank of Yemen Law, 19711971
Yemen People’s DemocraticBanking System Law (No. 36 of 1972)1973
Republic
ZaïreBank of Zaire Statutes1976
ZambiaBank of Zambia Ordinance, 19651971
ZimbabweReserve Bank of Zimbabwe Act1984
New Laws 1989 and 1990
ChileOrganic Law Establishing the Central Bank of Chile1989
Lao People’sThe Law Conc. with Creation of The Bank of1990
Democratic RepublicLao People’s Democratic Republic
Myanmar (Burma)The Central Bank of Myanmar Law1990
NamibiaCentral Bank of Namibia Act1990
New ZealandReserve Bank of New Zealand Act1989

A member of the East Caribbean Central Bank.

A member of the Central Bank of West African States.

A member of the Bank of Central African States.

A member of the East Caribbean Central Bank.

A member of the Central Bank of West African States.

A member of the Bank of Central African States.

Appendix I Limitations on Lending from the Central Bank to the Government
CountryType of LoanDecision Process and Regulations
A. Limit: As Described
AustriaLoans to the governmentThe Federal Republic, the Länder or the municipalities in no way, either directly or indirectly may draw on the National Bank’s funds for their own purposes, without providing the counterpart in gold or foreign exchange.
BelgiumGovernment securitiesThe National Bank may discount, buy and transfer short- and medium-term securities issued or guaranteed by the Belgian State or the Luxembourg State or issued by institutions whose liabilities are guaranteed by the Belgian State or the Luxembourg State. It may also buy and sell national long-term public securities quoted on the Stock Exchange. Holdings of national public securities may not exceed an amount equal to its capital, reserves and amortization accounts.
CanadaSecurities issued or guaranteed by the governmentFor the purposes of its open market operations there is no established limit.
FranceAdvances and loans to the stateAgreement drawn up by the Ministry of Economy and Finance and the Governor of the Bank of France, authorized by the Board of Directors and approved by the legislature.
IcelandAdvancesThe Central Bank may advance short-term loans to the treasury. Such loans shall be paid up within three months after the end of each fiscal year through borrowing or other acquisition of funds outside of the Central Bank.
Government securitiesTreasury bills, bonds, and other securities, which are issued by the treasury and bought by the Central Bank in the securities market or from financial institutions in order to promote balance in the money market, shall not count as loans to the treasury.
IrelandGovernment securitiesThe Central Bank may buy, hold, or sell securities of or guaranteed by the state which have been offered from public subscription or tender before being bought by the Bank, and are officially quoted on stock exchanges.
ItalyGovernment securitiesThe Bank of Italy may invest funds in securities issued or guaranteed by the state.
JapanAdvances to the governmentThe Bank of Japan may make advances to the government without collateral and may subscribe to or take up government loan issues.
KoreaAdvances and holdings of government bondsAuthorization by the National Assembly.
NetherlandsGovernment securitiesThe central bank may buy and sell Dutch Treasury paper and debt instruments quoted on the Amsterdam Stock Exchange and issued or guaranteed as to interest and principal by the government.
New ZealandLoans to the governmentThe Reserve Bank shall make loans to the government and on such conditions as the minister decides from time to time, in order to ensure the continuing full employment of labor and other resources of any kind. The Bank may also buy and sell securities issued or guaranteed by any government or issued by any local authority or public body, and such other classes of securities as may be approved from time to time by the minister.
NorwayShort-term credits to the governmentThe Norges Bank may grant seasonal and other short-term credits direct to the government within specific limits stipulated by the Storting. In special cases, it may also grant long-term credit direct to the government within specific limits stipulated by the Storting. The Bank may also purchase and sell treasury bills and government and government-guaranteed bonds. Purchases of these instruments are not covered by the previous limitations.
PortugalGovernment securitiesIn accordance with the guidelines for monetary, financial, and exchange rate policy established by the appropriate authorities, the Bank of Portugal may buy and sell securities issued by the Portuguese State.
SwitzerlandAdvances and government securitiesThe National Bank may make payments on behalf of the Confederation but only up to the limit of the credit balance of the Confederation with the Bank. Moreover, to prevent excessive recourse to the money and capital market, the Federal Council may require authorization of public issues of domestic treasury bills and debt instruments of any kind. The National Bank shall establish the overall amount of issues to be authorized in a given period.
TurkeyAdvances to state economic enterprises and administrationsThese advances should be made against bills or treasury-guaranteed bills of a maximum of nine months’ maturity. The maximum limits of the bills that may be accepted in this manner and applicable discount and interest rate shall be decided by the Board of Directors.
Purchases of state bondsIn order to regulate money supply and the liquidity requirements of the economy the Bank may purchase and sell state bonds, state internal loan bonds, bonds of the public agencies and institutions, and the sound bonds quoted on the stock exchange. These open market operations shall not exceed five times the total of the bank’s capital and reserve funds.
United KingdomLoans to the treasuryThe Bank of England may lend any sums that the treasury has the power to borrow under the National Loans Act, 1968, for providing the sums required to meet any excess of payments out of the National Loans Fund (the account of the treasury at the Bank of England), over receipts into the National Loans Fund, and for providing any necessary working balance in the National Loans Fund.
United StatesHoldings of government securitiesEvery Federal Reserve bank may buy and sell bonds, notes, or other obligations that are direct obligations of the United States or that are fully guaranteed by the United States as to the principal and interest, without regard to maturities, but only in the open market. Besides, Federal Reserve banks may buy and sell, at home or abroad, bonds and notes of the United States having maturities from date of purchase that do not exceed six months, and bills, notes revenue bonds and warrants with a maturity from date of purchase of not exceeding six months, issue in anticipation of the collection of taxes or in anticipation of the receipt of assured revenues by any state, county, district, political subdivision, or municipality in the continental United States. They may also buy and sell in the open market, under the direction and regulations of the Federal Open Market Committee, any obligation that is a direct obligation of, or fully guaranteed as to principal and interest by, any agency of the United States.
Treasury borrowingMoreover, the Secretary of the Treasury may borrow from any Federal Reserve bank, subject to the approval and rules and regulations of the Federal Open Market Committee, any obligation of, or fully guaranteed as to principal and interest by, any agency of the United States, and to sell any such obligation in the open market lor the purpose of meeting the short-term cash needs of the treasury. Not later than six months after the date of sale of such an obligation, the Secretary of the Treasury shall repurchase such obligation and return such obligation to the Federal Reserve bank from which such obligation was borrowed. The aggregate of the face amount of obligations borrowed under their authority shall be included during the period of such borrowing as part of the public debt subject to the limitation imposed by Section 21 of the Second Liberty Bond Act (a fixed amount in U.S. dollars).
B. Limit: Absolute Amount in Domestic Currency
GermanyCash advancesThe maximum limit on the cash advances is set annually by the legislative body and includes such treasury bills as the Deutsche Bundesbank has purchased for its own account or to the purchase of which it has committed itself.
Holdings of treasury bills and treasury bondsThe Bundesbank may, for the purpose of regulating the money market, purchase and sell on the open market at market prices treasury bills and treasury bonds issued by the Federation (mobilization paper). Mobilization paper shall not be counted toward the credit ceiling mentioned before, but there exists also a maximum limit (set by the legislative body) for the holdings of mobilization paper by the Deutsche Bundesbank.
GreeceAdvancesThe Bank of Greece may make temporary advances in drachmas to the government for expenditure in the annual state budget, provided that the whole of the advances outstanding at any one time shall not exceed a given amount in domestic currency. All advances shall be repaid not later than at the end of the quarter following the close of the fiscal year in respect of which such advances were made.
NetherlandsAdvancesFor the temporary strengthening of the treasury’s position the central bank may make advances to the government against sufficient security of treasury notes provided that the issue or pledge of such notes has been permitted by law.
SwedenCreditsThe Riksbank may open credits in current account for a period not exceeding 12 months against the pledge of the Swedish State up to a fixed amount in kronor.
Government securitiesThe Riksbank may buy and sell Swedish Government bonds and other Swedish Government securities.
Country NameType of LoanPercentBase for

Calculation
CostRepaymentDays to Maturity
C. Limit: Percentage of Central Bank Liabilities
CyprusHoldings of government and government-guaranteed securities20Total sight liabilities
Equatorial GuineaHoldings of government securities20Total liabilities
(max, annual increase)(5)(Monthly average previous 12 months)
Gambia, TheAdvances and holdings of government securities50Average demand liabilities previous fiscal year
MalawiHoldings of government securities25Total demand liabilitiesMore than 2 years, less than 25 years.
MaltaHoldings of treasury bills and holdings of publicly issued securities of or guaranteed by the government, including any security held as collateral20Total central bank demand liabilitiesFor treasury bills, 93 days and not more than 20 years for other government securities.
MozambiqueLine of credit10Average monthly sight liabilities of first 9 months of previous fiscal yearFreeEnd of current fiscal year
NigeriaHoldings of government securities75Total demand liabilitiesMore than 2 years, less than 25 years.
D. Limit: Percentage of Government Revenue
AlgeriaOverdrafts5Previous fiscal yearService chargeMaximum 240 days each fiscal year
ArgentinaAdvances30Previous fiscal yearDiscount rate or higherTwelve months after fiscal year
AustriaHoldings of short-term treasury certificates and other bonds of the Federal Republic5Current fiscal yearDiscount rate
BahamasAdvances10Least of previous and current fiscal year
BahrainAdvances10Previous fiscal yearAs agreed by the authoritiesThree months after fiscal year
Holdings of government securities and other loans25Previous fiscal year
BelizeAdvances15Current fiscal yearThree months after fiscal year
BotswanaAdvances and holdings of government securities20Average three preceding fiscal yearsAs agreed by the authoritiesSix months after fiscal year
+ 10(In exceptional circumstances)
BurmaAdvances15Current fiscal yearMaximum six months
BurundiAdvances50Previous fiscal yearMinimum 3 percent
CanadaAdvances to the government33Current fiscal yearThree months after fiscal year
Advances to any provincial government25Current fiscal yearThree months after fiscal year
Cape VerdeOverdraft holdings of loans against treasury notes15Previous fiscal yearFreeEnd fiscal year
Central African StatesOverdraft20Previous fiscal yearMaximum 12 months
CyprusAdvances25Current fiscal yearSix months afterMaximum 12
Holding of treasury bills30Current fiscal yearfiscal yearmonths
East Caribbean Currency BoardTemporary advances5Average three preceding fiscal years
Holdings of treasury bills10Average three preceding fiscal yearsMaximum 91 days
EcuadorAdvances10Average three preceding fiscal yearsEnd fiscal yearMaximum one year
EgyptLoans10Average three preceding fiscal yearsMaximum 12 months
El SalvadorAdvances and loans30Average five preceding fiscal yearsMaximum 12 months
EthiopiaAdvances25Previous fiscal yearMaximum 3 percentSix months after fiscal year
Holdings of treasury bills and government bonds50Previous fiscal yearMaximum 10 years
GhanaAdvances10Current fiscal yearAs agreed by the authoritiesThree months after fiscal year
15(In exceptional circumstances)
GreeceOverall amount of advances, guarantees and discounted treasury bills10Current fiscal year
GuineaOverdraft10Previous fiscal yearService chargeMaximum 240 days each fiscal year
GuyanaAdvances15Average 3 preceding fiscal yearsMinimum 3 percentMaximum 350 days each fiscal year
IraqAdvances15Current fiscal yearThree months after fiscal year
IsraelAdvances20Current fiscal yearEnd fiscal year
JamaicaAdvances30Current fiscal yearThree months after fiscal year
JordanAdvances25Current fiscal yearFree
KuwaitAdvances10Previous fiscal yearEnd of next fiscal year
Lao People’s Democratic RepublicAdvances25Previous fiscal yearAs agreed by the authoritiesAs agreed by the authorities
50(In exceptional circumstances)
LesothoAdvances5Current fiscal year
LibyaAdvances10Current fiscal yearEnd of fiscal year
MadagascarAdvances15Previous fiscal year6 months after fiscal
(20)(In exceptional circumstances)year
MalawiAdvances10Current fiscal yearDetermination by central bankFour months after fiscal year
MalaysiaAdvances12.5Current fiscal yearThree months after fiscal year
MaliOverdrafts10Previous fiscal yearDiscount rateMaximum 240 days each fiscal year
15(In exceptional circumstances)
MaltaAdvances15Current fiscal yearAs agreedEnd fiscal year. If after that date such advances remain unpaid, the power of the central bank to grant further advances shall not be exercisable until the outstanding advances have been repaid.
MauritaniaOverdrafts and loans15Previous fiscal year
20(In exceptional
circumstances)
MauritiusAdvances25Current fiscal yearFour months after fiscal year
MoroccoAdvances10Maximum 240 days each fiscal year
MozambiqueOverdrafts and loans8.3Current fiscal yearMaximum discount rateThree months
Netherlands AntillesAdvances10Previous fiscal yearFree
NigeriaAdvances25Current fiscal yearDetermined by the Central BankEnd fiscal year
OmanAdvances10Current fiscal yearNinety days
Papua New GuineaAdvances and holdings of government securities20As agreedAs agreed
25(In exceptional circumstances)Maximum six months
ParaguayHoldings of treasury bills20Average three preceding fiscal yearsMaximum one year
PeruHoldings of treasury bills8.3Current fiscal yearMaximum 90 days
PhilippinesAdvances20Average three preceding yearsThree months after fiscal year
PortugalOverdrafts10Previous fiscal yearFreeEnd of fiscal year
RwandaOverdrafts5Average three3.0 percent
-5preceding fiscal4.5 percent
+ 1years6.0 percent
Sao Tome andOverdrafts5Previous fiscal yearFreeEnd of fiscal year
PrincipeShort-term loans8.3Current fiscal yearMaximum discount rateMaximum three months
Special purpose loans20(In exceptional circumstances)
Sierra LeoneAdvances20Current fiscal yearEnd of fiscal year
SomaliaAdvances and holdings of government securities35Average three preceding yearsMinimum 2.5 percent
Sri LankaAdvances10Current fiscal yearMaximum six months
SudanAdvances15Current fiscal yearDetermined by the Central BankSix months after fiscal year
SurinameAdvances10Current fiscal yearFree up to 1 million guilders
TanzaniaAdvances20Current fiscal yearMinimum 3 percentMaximum 300 days each fiscal year
Holdings of government securities25Maximum 12 months
Trinidad and TobagoAdvances15Current fiscal yearEnd fiscal year
TunisiaOverdrafts5Previous fiscal yearService chargeMaximum 240 days each fiscal year
TurkeyAdvances15Current fiscal yearAs agreed between the Ministry of Finance and the central bank
UgandaAdvances18Current fiscal yearAs agreed by the authoritiesEnd fiscal year
United Arab EmiratesAdvances10Previous fiscal yearEnd of next fiscal year
VanuatuAdvances and holdings of government securities15Average three preceding fiscal yearsMaximum six months
VenezuelaAdvances and holdings of10Average fiveEnd of fiscal year
government securities20preceding fiscal years (In exceptional circumstances)
Viet NamAdvances25Previous fiscal year
West African Monetary UnionAdvances20Previous fiscal year
Western SamoaAdvances25Current fiscal yearDetermined by theSix months after
35(In exceptional circumstances)central bankfiscal year
Yemen Arab RepublicAdvances20Current fiscal yearService charge minimum 3 percentAs agreed
Holdings of government securities15Maximum 12 months
Yemen People’s Democratic RepublicAdvances15Current fiscal yearEnd of fiscal year
Holdings of government securities25Current fiscal year
ZaïreAdvances15Average three preceding fiscal yearsMinimum 3 percentMaximum 300 days each fiscal yearHoldings of treasury notes
20Average three preceding fiscal years
ZambiaAdvances and holdings of government securities50Current fiscal yearMinimum 3 percentMaximum 300 days each fiscal yearHoldings of treasury notes
ZimbabweTwelve months after fiscal year
E. Limit: Percentage of Government Expenditure
JamaicaHoldings of government securities40Current fiscal year
ThailandAdvances25Current fiscal yearThree months after fiscal year
SpainAdvances to the treasury12Current fiscal yearFree
Appendix II Analytical Framework

We will start with the following general expression for the budget constraint faced by a typical economic unit:

where

Ex = expenditures;

ΔA = change in assets;

In = income; and

ΔL = change in liabilities.

We can use expression (1) to specify the government budget constraint as follows:

where

GOE = Total government current operational expenditures (including investment);

rG = a(1 × m) vector of nominal interest rates on government net liabilities with the central bank;

CBNCG = a(m × 1) vector of outstanding government net liabilities with the central bank;

rD = a(1 × m) vector of nominal interest rates on government domestic debt instruments outside the central bank;

GDD = a(m × 1) vector of outstanding government domestic debt instruments outside the central bank;

GFAis = stock of government foreign asset type i, valued in local currency, outstanding, at time s (i = 1,…, m; s = t, t − 1);

GFLjs = stock of government foreign liability type j, valued in local currency, outstanding at time s (j = 1,…, m;s = t, t − 1);

GOAis = stock of other government asset type i outstanding at time s (i = 1,…, m; s = t, t − 1);

GOLjs = stock of other government liability type j outstanding at time s (j = 1,…, m; s = t, t − 1);

T = tax revenues;

r* = α(1 × m) vector of (local currency equivalent of) nominal international interest rates on government net foreign assets;

GNFA = α(m × 1) vector of outstanding government net foreign assets valued in local currency;

r** = α(1 × m) vector of nominal interest rates on other net government assets;

GONA = α(m × 1) vector of outstanding other net government assets;

CBCGis = stock of central bank credit (type i outstanding at time s) to the government (i = 1,…, m; s = t, t − 1);

GDCBpjs = stock of government deposits (type j outstanding at time s) with the central bank (j = 1,…, m; s = t, t − 1);

GDDis = stock of government domestic debt instrument type i outstanding at time s (i = 1,…, m;s = t, t − 1) outside the central bank.

Rearranging expression (2) we get:

A similar expression can be derived for the central bank budget constraint as follows:

where

ΔRM = RMt - RMt − 1;

RMs = stock of reserve money outstanding a time s (s = t, t − 1);

CBBis = stock of type i central bank bond (or bill) outstanding at time s (i: 1,…, m; s = t, t − 1);

CBFAis = stock of type i central bank foreign assets, valued in local currency, outstanding at time s (i = 1,…, m;s = t, t − 1);

CBFLjs = stock of type j central bank foreign liability, valued in local currency, outstanding at time s (j = 1,…, m;s = t, t − 1);

CBOE = current operational central bank expenditures;

rB = α(1 × m) vector of nominal interest rates on central bank bonds or bills;

CBB = α(m × 1) vector of outstanding central bank bonds or bills;

r′ = α(1 × m) vector of (local currency equivalent of) international nominal interest rates on outstanding central bank net foreign assets;

r″ = α(1 × n) vector of nominal interest rates on other outstanding central bank net assets;

CBONA = α(m × 1) vector of other outstanding central bank net assets;

re = α(1 × m) vector of nominal interest rates on outstanding liabilities of the economy with the cen-tral bank;

CBCE = α(m × 1) vector of outstanding debt instruments of the economy (banks and private sector) with the central bank;

CBCEis = stock of the liability type “i” of the economy with the central bank outstanding at time s (i = 1,…, m;s = t, t − 1);

CBOAjs = stock of other central bank assets (type i) outstanding at time s (i: 1,…, m; s = t, t − 1);

CBOLjs = stock of other central bank liabilities (type j) outstanding at time s (j = 1,…, m;s = t, t − 1); and rG, CBNCG as before.

Using (3) and (4) we can obtain an expression for the consolidated budget constraint of the government and the central bank as follows:

Appendix III Suggested Limitations

Limit on total outstanding government debt

A prudent and elemental principle in lending establishes that the ratio of payments of loans installments to income should not exceed a certain given constant. For government debt, this relationship can be written as follows:

where

Pt = payments at time t;

Tt = government revenues at time t;

k1, = a constant.

A relationship linking the amount of payments to the total amount of government debt also exists. Assuming, for simplicity, that only one outstanding loan to the government exists, this relationship can be written as follows:

where

D(t) = amount of the loan or outstanding debt;

i = interest rate per payment period;

n = number of payment periods.

Thus, relationship (1) can be rewritten as follows:

In equality (3) limits total outstanding debt to a proportion of total government revenues. This proportion is decreasing in the interest rate on government debt and increasing in the maturity of loans to the government. To some extent, a limitation of the form in equation (3) derived from simple prudential principles resembles some of the limitations we found in actual central bank laws governing the lending to governments.

The relationship expressed in equation (3) implies a limit on government-debt growth of the form:

Limits on Governments Borrowing from the Central Bank

Limit on Total Outstanding Net Government Debt with the Central Bank

This limit is similar to the one suggested for total outstanding government debt and can be expressed as follows:

where

Dcb(t) = outstanding total government net debt with the central bank.

The relationship expressed in equation (5) implies a limit on the growth of government debt with the central bank of the form:

Limit on Government Direct Borrowing from the Central Bank

The purpose of this limiting direct borrowing from the government is to indicate the maximum allowed amount of direct central bank lending to the government. This amount is also constrained by the limits given by the relationship shown in equation (6). The proposed limit takes the form:

where

CBDCG = outstanding direct central bank net credit to the government at time t (in nominal terms)

y(t) = rate of growth of real GDP at time t.

Bibliography

    AlesinaAlberto andGuidoTabellini“Rules and Discretion with Noncoordinated Monetary and Fiscal Policies,”Economic Inquiry (U.S.) (October1987) pp. 61930.

    • Search Google Scholar
    • Export Citation

    BaileyMartinJ.“The Welfare Cost of Inflationary Finance,”Journal of Political Economy (April1956) pp. 93110.

    CaganPhillip“The Monetary Dynamics of Hyperinflation,” in Studies in the Quanity Theory of Moneyed. byM.Friedman (Chicago: University of Chicago Press1956).

    • Search Google Scholar
    • Export Citation

    CukiermanAlex“Central Bank Behavior, Credibility, Accommodation, and Stabilization” (MIT Pressforthcoming).

    Deutsche Bundesbank“Bundesbank Cash Advances to the Federal Government and the Länder Governments,”Monthly Report (May1986) pp. 2226.

    • Search Google Scholar
    • Export Citation

    FischerStanleyThe Economics of the Government Budget (Bombay: Reserve Bank of IndiaFebruary1990).

    International Monetary Fund International Financial Statistics.

    ParkinMichael“Domestic Monetary Institutions and Deficits,” in Deficitsed. byJ.M.BuchananC.K.Rowley andR.D.Tollison (New York; Oxford: Basil Blackwell1987) pp. 31037.

    • Search Google Scholar
    • Export Citation

    SargentThomasJ. andNeilWallace“Some Unpleasant Monetarist Arithmetic,”Quarterly ReviewFederal Reserve Bank of Minneapolis (Fall1981) pp. 117.

    • Search Google Scholar
    • Export Citation

    WallaceNeil“Integrating Micro and Macroeconomics: An Application to Credit Controls,”Quarterly ReviewFederal Reserve Bank of Minneapolis (Fall1980) pp. 1629.

    • Search Google Scholar
    • Export Citation
*The author is Senior Economist in the Central Banking Department of the International Monetary Fund. The author is grateful to Alex A. Cukierman, Patrick Downes, Douglas A. Scott, and V. Sundararajan for their helpful advice and comments and to Caroline Cox and Amelita Concepcion for their secretarial assistance.
1Appendix I summarizes the regulations and decision-making processes related to limitations on central bank credit to governments in different countries. The summary was prepared with the research assistance of Ms. Anne Johannessen of the Central Banking Department of the International Monetary Fund.
2In addition, the average government revenues or expenditures during a previous period (the past three or five years) is usually the base variable in practice. This procedure differs from the one used in calculating the ratios in Table 1 where the current levels of revenue, expenditure, or base money were used instead.
3This finding is consistent with the analysis of Cukierman (forthcoming), Chapter 18.
5It is convenient to emphasize that current borrowing will contribute to future financial requirements through interest payments and debt servicing of outstanding debt.
6These issues relate to the problem of coordination of fiscal and monetary policies. See, for example, Alesina and Tabellini (1987).
7An equilibrium with a constant inflation rate is considered for the purposes of this analysis.
8This curve has been derived elsewhere in the literature, see for example, Bailey (1956), Cagan (1956), and Wallace (1980).
9Fischer (1990) uses a form of this equation to analyze the dynamics of the (consolidated) government debt.
11The distance between each point in the graph and the diagonal represents government net borrowing outside central banks (in terms of GDP).
12Printing of base money may also originate in other central bank operations (foreign exchange purchases, credit to the economy, etc.) and in central bank losses.
13Relative, for example, to the sample means.
14For many countries, it is difficult to obtain real interest rates values reflecting realistic market conditions, mainly because of the implementation of price and interest rate controls. For the period considered here, real average yield to maturity for government bonds was 3.9 percent in Germany, 2.7 percent in Japan, and 3.2 in the United States. Average treasury bill real rates were 2.1 percent in Germany and 1.5 percent in the United States (not available for Japan). The assumption of a 3 percent average real rate of interest for the 1975–87 period seems then reasonable.
15And income from these foreign assets lower than expenditures related to foreign liabilities.
16This is valid for any other government or nongovernment debt with the central bank.
17Including those constraints depicted above.
18In many countries, a proportion of central bank profits are distributed to the government. Then, the need to remunerate central bank loans to the government at market-related rates does not seem necessary at a first look: if, because of mispricing, central bank income from government debt is lower than otherwise, so will be central bank profits and the proportion to be distributed to the government. However, legal provisions in the case that central banks incur losses are not symmetric: in these cases central banks’ capital will be affected. It is in these cases when it becomes important to find the causes of central bank losses and determine the responsibility of fiscal authorities. The appropriate remuneration of government debt with the central bank will make this process more transparent.
19Such intervention may be needed in the case of long-term papers that carry larger risk of price variability, particularly in the early stages of market development.
20It should be recognized that by intervening in this way the central bank will reduce the incentives for the offsetting forces (reduced consumption and inflows of short-term capital) to take place.
21It is also shown in Appendix III that this form of limitation implies a limit on the growth of government debt.
22This kind of link resembles some of the limitations that are found in actual central bank laws governing the lending to governments.
23Notice that this is a limit on the growth of government debt with the central bank.
24All three limitations taken together impose some constraints to government borrowing outside the central bank.
25Net of government deposits with the central bank.

    Other Resources Citing This Publication