ACALIN Julien

Individual reading commentary on The forgotten history of domestic debt

Why do sovereigns default in local currency if they have the ability to print money?

Introduction
[...] The proposition [is] that countries without a printing press are subject to self-fuelling crises in a
way that nations that still have a currency of their own are not. The point is that fears of default, by
driving up interest costs, can themselves trigger default and that because there is a crossing-theRubicon aspect to default, once a country crosses that line it will probably impose fairly severe losses
on creditors. A country with its own currency isnt in the same position: even if it is pushed into
some inaction, theres no red line that need be crossed.
Paul Krugman, The Printing Press Mystery, The conscience of a liberal, August 17, 2011

It is common belief that investing in government debt issued in local currency is close to riskfree since a government has (1) the ability to raise taxes usually paid in local currency, (2)
better access to a stable domestic capital market, as well as (3) some capacity to print
money and monetize its debt. In their paper, Kenneth Rogoff and Carmen Reinhart provide
us with historical time series data on external and domestic public debt 1 for 64 countries
ranging back to 1914. Their main findings are twofold:
First, domestic debt is and was large for the 64 countries for which they have long time
series, domestic debt averages almost two-thirds of total public debt.
Second, where there are domestic debts there is also domestic default. Examples in the last
decades include Venezuela, Russia, Ukraine, Ecuador, Argentina and Jamaica. According to
the authors, this phenomenon appears to be far too common to justify the extreme
assumption that governments always honor the nominal face value of domestic debt. Some
studies 2 even suggest that defaults on local-currency debt are more common than on
foreign-currency debt over the last twenty years (Table 1).
Regarding this result, one might ask: why do sovereign default in their local-currency debt if
they have the ability to print money? In order to answer this question well see how money
printing can reduce debt, present the limits of this mechanism and then the reasons why a
sovereign would prefer to default instead of printing money.

Domestic public debt is defined as issued under home legal jurisdiction. In most countries, over most of their history, it has been denominated in
the local currency and held mainly by residents. Yet the sample used includes some defaults on debt denominated in foreign currency and some
cases where the currency denomination is unknown.
2

See for example, A. Jeanneret, E. Paget-Blanc and Souissi S., Sovereign Defaults by Currency Denomination, 2014.

HEC Paris

ACALIN Julien

I/ Debt monetization, seigniorage and tax inflation

In order to meet their inter-temporal budget constraint, some governments borrow money by
issuing bonds. We limit here our analysis to local currency bonds issuance. The government
Treasury must then pay off government debt either with money it already holds, with taxes
collected from the public, or by financing it by issuing new bonds, in order to raise the funds
required to repay bonds that have come due. In other words, a sovereign government that
has issued debt in its own currency, which it cannot (or does not wish to) repay through
taxation, has a choice between defaulting on its bonds and printing the money required to
pay them off. This last option is called debt monetization.
To understand how printing money can reduce debt, we should consider two representations
of the standard debt sustainability equation.
Firstly, we start with the most common representation:

The primary budget balance represents the government global budget balance (income expenses), excluding interest payments on its debt. This primary budget balance includes
seigniorage (the difference between the face value of money and its costs of production). As
the cost of printing money is close to zero, we can consider that the value of seigniorage is
equal to the change in the monetary base. So if the monetary base increases, the primary
balance surplus increases, reducing the ratio of government debt to GDP in period t+1.
Secondly, lets consider the following equation:

The central bank may purchase government bonds by conducting an open market purchase,
i.e. by increasing the monetary base through the money creation process. In exceptional
times, a central bank could provide direct monetary financing of the budget deficit (a different
concept from normal seigniorage) and actively seek to generate higher inflation to
reduce the real value of the outstanding debt stock. This process of financing
government spending is called debt monetization. Central banks are usually forbidden by law
from purchasing debt directly from the government (this is the case for the ECB). Monetizing
debt is thus a two-step process where the government issues debt to finance its spending
and the Central bank purchases the debt on the secondary market, holding it until it comes
due, and leaving the system with an increased supply of money.
Money printing always reduces the price of bonds and imposes an inflation tax which shifts
resources away from households and bondholders towards government. According to Das,
Papaioannou, and Trebesch (2012)3, between 1800 and 2007 there have been 150 cases of
yearly inflation beyond 20% that allowed for de facto reduction of debt denominated in local
currency.
However, printing money present limits and can have undesirable effects on the national
economy, such as a change in asset prices, a redistribution from savers to debtors, price
inflation/currency depreciation, constrained future access to credit markets, and so forth.
3

U. Das, M. Papaioannou, and Trebesch C., Sovereign debt restructurings 19502010: literature survey, data, and stylized facts, IMF Working
Paper, WP/12/203, 2012

HEC Paris

ACALIN Julien

II/ Limits of printing money

a) Printing money can be ineffective
As we saw, higher inflation can be used to raise seigniorage and reduce the real value of the
government debt stock. Nevertheless, in the long run, assuming the real growth rate does
not exceed the real interest rate, the ratio of government debt/GDP will rise if the government
is running a primary budget deficit, irrespectively of the inflation rate. In other words, printing
money provides a temporary solution and does not invalidate the basic arithmetic of the
inter-temporal budget constraint. A recent Fitch special reports4 shows using debt dynamics
models that sovereigns would need to generate a large inflation surprise to have a material
impact on fiscal solvency.
Moreover, a governments ability to use surprise inflation to reduce the real value of its debt
depends on characteristics of debt structure, economy and financial system. This refers to
the concept of monetary sovereignty. In fact, the inflation tax can be ineffective if it is
anticipated and debt is relatively short term or indexed to inflation, because investors would
be able to adjust by demanding higher interest rates to compensate for their losses. This is
particularly true in the case of short-term debt: inflation is unlikely to do serious damage to a
portfolio in the course of a few months. The potential costs of inflation are then especially
problematic since the government has in this case to inflate much more aggressively to
achieve a significant real reduction in debt service payments.
b) Inflation has undesirable effects
When it decides to monetize its debt, a government should be aware of the potential costs of
doing so. If inflation increases, people will not want to hold bonds because their value is
falling. Therefore, the government will find it difficult to issue debt in its own currency at long
maturity and will have to pay higher interest rates to attract investors (loss of trust).
Then, if a country prints money and creates inflation, then there will be a decline in the value
of its currency (real depreciation). This affects not only the international value of the
governments debt promises, since debt burden would actually increase as a result of
translation effect on the foreign currency debt, but also that of all other contracts
denominated in local currency. In particular, depreciation can be very costly for a corporate
sector with a currency mismatch due to foreign currency liabilities (if the private sector
borrows from abroad in foreign currency). From the internal point of view, depending on the
speed of price-resetting (pass-through) it can affect competitiveness throughout the
economy.
More generally speaking, inflation results in a fall in value of savings, destabilizes an
economy by creating uncertainty and confusion (distortion in relative prices). Periods of
high inflation discourage firms from investing and can lead to lower economic growth,
unemployment, and in extreme cases, poor living conditions.
These effects have a high social and political cost underlining that inflating away sovereign
debt is not an easy option. Ironically, trust in the local currency gives a sovereign both the
potential to inflate away its debt and a lot to risk if it were to do so. As a result, printing
money could create more problems than it solves.

Parker E. and Riley D., Why Sovereigns can Default on Local-Currency, Fitch special report, 2013

HEC Paris

ACALIN Julien

III/ The default option

Why would a government refuse to pay its domestic public debt in full when it can simply
inflate the problem away? Defaulting on local currency debt occurs regularly because debt
monetization can be costly (high inflation and political turmoil) and may, therefore, not always
be optimal choice for a government. Sometimes, the government may view repudiation as a
lesser evil.
In a recent paper, Jeanneret, Paget-Blanc and Souissi (2014) 5 present a new database
indicating the existence of 58 sovereign default events over the period 1996-2012: 31 consist
of defaults on local currency bonds and 27 defaults on foreign currency bonds. For 15
events, they observe simultaneous defaults on both types of debt. They found that a low
level of domestic bank credit and a high inflation are characteristics that make governments
more prone to default on local currency bonds.
First, banks provide a particularly important channel for magnifying and transmitting the
impacts of sovereign risk outcomes because they hold a huge amount of sovereign debt for
liquidity purposes, including the government bonds that banks hold as required reserves
against deposits. Banks are fragile: a sovereign default would lower bank solvency through
losses on their sovereign debt holdings when haircuts are imposed on creditors. This can
have significant impacts on bank profits and survival, and by extension significant impacts on
financial intermediation activity and thus investment, output and consumption. That being
said, countries with a low level of domestic bank credit to firms are more prone to
sovereign default on local debt. In the same way, one can also argue that low level of
financial development (M2/GDP) make the default option less costly for sovereigns. As an
example, when Russia, Ukraine, Ecuador, Argentina and Jamaica defaulted on their localcurrency debt, their M2/GDP ratio was under 45% (source: Fitch). On the opposite, a
government will be loath to default on its debt if sizeable amounts of defaulted debt are held
by the local banks. Testament to this is the track record of Lebanon, whose banks have
tended to lend a higher percentage of GDP to the government than in the case of any other
country
The authors also find that the level of inflation raises the default probability on (both foreign
and) local currency debt. This finding suggests that high inflation is associated with a limited
possibility to further increase inflation through debt monetization. Defaulting thus appears to
be an optimal decision an optimal decision in periods of severe inflation.
Concerning inflation, default also has the potential to be less damaging than monetization:
while under default government action is limited insofar as the government cannot default on
more debt than it has outstanding, under a money-printing policy it is possible for the
government to print even more cash than the amount of the original debt. This is particularly
the case when there is no stable relationship between the monetary base and inflation.
Therefore, it is unclear what increase in money supply would be required to deliver a
desirable level of inflation and the government could crowd out households completely by
printing an important amount of cash.
Thus, it appears monetization has the potential to be more damaging than default in some
situations (e.g., result may be ineffective, large amounts of monetization and domestic bank
credit, low reserve ratios, economic agents exposed to a currency mismatch, high inflation,
etc.).

A. Jeanneret, E. Paget-Blanc and Souissi S., Sovereign Defaults by Currency Denomination, 2014.

HEC Paris

ACALIN Julien

Conclusion
Printing money can reduce a sovereign local-currency debt by increasing seigniorage
revenues and reducing the real value of debt through tax inflation. Yet this option is
temporary, can be ineffective and has an important economical, social and political cost. To
default on its local-currency debt, even if costly in terms of reputation, can appear as a better
solution, or a lesser evil, for a government, especially if domestic bank credit is low and/or
inflation is already high.
However, some characteristics, not developed in this paper, such as an independent
currency and a floating exchange rate, referring to the monetary sovereignty, make
sovereigns more resilient against local-currency default. Monetary sovereignty also means
that the risk of a self-fulfilling liquidity crisis is lower than for sovereigns without a central
bank 'lender of last resort'.
This can explain the differences we observe between local-currency sovereign ratings: the
risk of default on local-currency bonds exists but is not the same among countries.

Table 1: History of Sovereign Defaults by Currency Denomination

Source: A. Jeanneret, E. Paget-Blanc and Souissi S., Sovereign Defaults by Currency Denomination, 2014.

HEC Paris