Analysts and policymakers have recently put forth the view that Eurozone nations have, in essence, issued sovereign debt in a ‘foreign’ currency – foreign in the sense that national authorities don’t control the printing presses (see De Grauwe 2011, among many others).
The importance of this point is that nations with high and increasing debt levels but without a currency of their own are inherently vulnerable to self-fulfilling, disruptive, debt crises. As put by Krugman (2011):
“Fears of default, by driving up interest costs, can themselves trigger default … there's a crossing-the-Rubicon aspect to default, once a country crosses that line it will probably impose fairly severe losses on creditors. A country with its own currency isn't in the same position: even if it is pushed into some inflation, there's no red line that need be crossed.”
The historical record warns against overplaying the idea that an ‘own currency’ be an easy way out of sovereign default.
Outright default on public debt denominated in domestic currency is far from rare, also in countries where policymakers are in principle in control of the ‘printing press’. In a long sample ending in 2005, Reinhart and Rogoff (2011) document 68 cases of overt domestic default (often coinciding with external debt default).
Of course this evidence does not disprove the argument above. In the data, it is difficult to separate fundamental default from self-fulfilling sovereign-debt crises unrelated to fundamentals. Eliminating the latter by no means implies that default cannot occur. However, the evidence does stress the importance of identifying the conditions under which (and the concrete policy strategy by which) a central bank can be effective in stemming disruptive turmoil in the debt market – a central question, as forcefully stressed by Blanchard (2013) and the IMF (2013).
We look into this issue in a recent paper (Corsetti and Dedola 2013), building on a classical contribution by Calvo (Calvo 1988). The framework we develop is meant to capture the essence of the problem faced by a country with a small probability of fundamental default, which nevertheless becomes vulnerable to the ‘crossing-the-Rubicon’ aspect of a sovereign crisis.
Is one’s ‘own currency’ all about inflationary financing?
It is often argued that countries that issue debt in domestic currency and control the printing press can always ‘pay’ debt holders with cash, without breaching the debt contract. As paying debt with cash ultimately means that debt is inflated away, the underlying idea is that the option of inflationary financing coordinates investors on a good equilibrium and eliminates self-fulfilling sovereign-debt crises. Since the option to access the printing press and inflate domestic debt at will is precluded to national policymakers in a monetary union, its members effectively issue debt in a foreign currency.
A first insight from our analysis is that the ability to inflate debt away provides no shield against self-fulfilling crises unrelated to fundamentals. The reason is straightforward – inflation is not costless. There are trade-offs between default, taxation and inflation that bound the degree to which the central bank is willing to raise inflation in response to a debt crisis. Moreover, the moment investors anticipate inflationary financing, interest rates would rise, reducing the gains from debt monetisation – up to the point of undermining its effectiveness altogether. A similar result was foreshadowed in Calvo (1988), and is also reached by Aguiar et al. (2012).1
What enables a central bank to provide a monetary backstop to the government?
The finding above is by no means equivalent to saying that central banks are powerless as to the goal of eliminating self-fulfilling debt crises. All it says is that their ability to inflate debt ex post is not the main factor. What matters may be something different.
Modern central banks in advanced economies issue liabilities (high-powered money) not only in the form of cash but especially bank reserves, often interest-bearing, which are used both as means of payments and store of value. The ‘printing press’ argument may come into play here; independent central banks stand ready to honour their own liabilities, by redeeming them for cash at their nominal value. So, while nominal liabilities issued by central banks are exposed to the risk of inflation, they are not undermined by ‘fears of (outright) default’. The interest rate on these liabilities can therefore be lower than the interest on government debt, when the latter is subject to risk of both outright default and inflation.
Because of difference in yields, when a central bank buys government bonds by issuing liabilities – reserves – it essentially ‘swaps’ default-risky with default-free debt. The possibility of this swap explains why a monetary backstop to the government can be effective, even if, from an aggregate perspective, any purchase of government debt by the monetary authorities is at best backed by their consolidated budget with the fiscal authorities – i.e., there are no additional resources to complement tax and seignorage revenues.
Does a debt/reserve swap lead to a burst of inflation?
A swap of debt for interest-bearing reserves need not have any consequence on current inflation. This is most apparent in the (currently quite relevant) instance when nominal-interest rates are already close to their lower bound, so that reserves and cash are close substitutes.
Central-bank purchases of government paper, of course, may affect future inflation and inflation expectations, as agents may anticipate discretionary attempts by the central bank to raise seignorage revenues and inflate away nominal debt in the future. Indeed, the interest rate on reserves will generally be sensitive to these expectations: the larger the anticipated monetary expansion in the future, the higher the market-determined nominal-interest rate on reserves at the time the central bank issues them.
But monetary authorities need not raise future inflation anywhere near the possibly elevated levels required to ‘debase’ their own nominal liabilities. Inflation will have to rise only insofar as government payments to the central bank will fall short of the interest bill due on reserves (net of the present discounted value of seignorage, at the desired level of inflation). Other things equal, once the economy is shielded from self-fulfilling debt crises, the government is in a much better position to service its debt on the balance sheets of the central bank.
Monetary interventions in the debt market are not a mechanical back-door entry for inflation. But for this reason, the design of interventions is not unconstrained.
Credibility of a monetary backstop: The role of cooperation between fiscal and monetary authorities
To what extent, and how, can a central bank use the power and prerogative of swapping risky debt with less risky liabilities to stabilise the sovereign-debt market?
The conventional view on central bank interventions is typically shaped by the received wisdom on the lender of last resort: monetary authorities should stand ready to respond to unjustified market pressure on sovereign borrowing costs, with unlimited debt purchases at a preannounced rate. De facto, the central bank should be willing to step in and fully replace private lending to the government. As is well understood, when the strategy works, no actual purchase of debt is required ex post: the threat of an intervention is enough to coordinate expectations on the economically sound equilibrium.
However, just announcing the intentions to intervene is no guarantee of success. For the strategy to work, the prospective contingent interventions must be credible, i.e., both feasible and desirable from the point of view of the monetary authorities. Doubts about the willingness of the central bank to carry out its threat would obviously undermine its effectiveness, and prompt the market to test the policymakers’ resolve. Incredible announcements can easily be counterproductive, as they may fuel the very self-fulfilling expectations that they are meant to prevent.
Because of the likely size of interventions, the conditions for the conventional strategy just described to be credible are quite strict. In practice, the probability of fundamental fiscal stress causing default, however small, may not be nil. Even if the central bank’s interventions could in principle eliminate self-fulfilling debt crises, it is still possible for the monetary authorities to suffer losses on their debt holding. Large interventions may not be credible if prospective losses foreshadow the need for the printing press to service interest-bearing reserves.
As we show in our paper, a way to ensure that a monetary backstop is effective is cooperation between the fiscal and monetary authorities: the two authorities should agree on common objectives and act under a consolidated budget constraint. A concrete implication of cooperation is that the treasury must be willing to cover (or avoid) possible losses by the central bank resulting from its interventions in the sovereign-debt market. In order not to result in higher, undesired inflation, the fiscal authority should be ready to make these losses good.
Arguably, fiscal and monetary cooperation and/or unconditional fiscal support to the central bank are key conditions most critics of the euro as an ‘incomplete’ monetary union (e.g. lacking political union) have in mind when praising the advantage of retaining a national currency. Cooperation between the central bank and the fiscal authorities, if only limited to financial stress situations, is clearly much more difficult in a currency union among essentially independent nations.
A monetary backstop can be effective also in an ‘incomplete’ monetary union
Once we acknowledge the essence of the mechanism by which interventions work under fiscal and monetary cooperation, the next logical passage is to recognise that impediments to full budget consolidation are, after all, not specific to a monetary union. In countries with their own currencies, for instance, it is often the case that institutional arrangements or political constraints rule out or limit fiscal transfers to monetary authorities. Independent central banks are typically wary to ask for fiscal support. They are expected to generate revenues for the treasury, or at the least to break even: they are clearly held responsible for losses they incur as a consequence of their policies.
For many a reason, the scale of interventions in the sovereign-debt market is necessarily bounded: banning contingent transfers from the treasury, a central bank’s losses can be no larger than (the present discounted value of) seignorage revenues, and the gains from inflating debt, at the rate of inflation it is willing to purse according to its own plans.
The reality of fiscal and monetary interactions suggests that the key question is actually similar for both monetary unions and countries with their own currency: can a central bank provide an effective backstop when the scale of interventions that can be credibly announced covers only part of the financing needs of the government?
The core result of our analysis is that a monetary backstop may nonetheless work. Here is the main argument. As is well understood, the decision to default entails economic costs in terms of financial and output disruptions. These costs can be expected to be in large part unrelated to the size of haircut on debt holders, reflecting the far-reaching consequences of the government decision to breach debt contracts. A large fixed component in the economic costs of default may indeed explain why we do not observe debt repudiation that often, but only at a relatively high level of debt. With a low stock of debt, the benefits from imposing haircuts on bond holders (e.g. in terms of lower taxation and higher supply of public services) are small relative to the disruptive consequences of a default on economic activity. Only when debt grows large default becomes an attractive option.
As we have seen above, by buying high-interest government debt against the issuance of low-interest monetary liabilities, the central bank can reduce the effective interest bill of the treasury, hence the level of primary surplus required to guarantee solvency and the associated economic distortions. With fixed costs of repudiation, there will be some level of sovereign-debt purchases at which the debt costs will be low enough, that the government will prefer to honour its liabilities in full, rather than defaulting. If a central bank can credibly stand ready to intervene up to that level, it will be able, so to speak, to walk the markets back across the Rubicon.
The insight underlying this strategy is straightforward. Instead of threatening to fully replace investors and de facto become the sole lender to the government, the central bank takes advantage of the fact that continued market access with participation of private investors imposes discipline on the government.
As the required scale of intervention is not determined by the whole financing needs of the government, central bank losses on holdings of sovereign-debt contingent on fundamental default may actually be contained. They may well be fully covered by (current and future) seignorage revenue at the desired level of inflation.
Implications for a monetary union
Our main conclusions resonate with the widespread view that a central bank has indeed the power to backstop the fiscal authority, although for different reasons than many observers invoke:
- What ultimately matters is the ability of the central bank to swap default-risky debt with reserves, reducing the cost of public indebtedness;
- By purchasing government paper on a sufficient scale, central banks can alter the trade-offs between default, taxation and inflation, up to making the decision to default suboptimal vis-à-vis a reduced debt burden.
This power is quite strong when the monetary and fiscal authorities operate under the mutual understanding that possible losses by the central bank on interventions in the sovereign-debt market are jointly managed with the treasury. Absent this understanding, the size of monetary interventions is effectively constrained by (the present discounted value of) seignorage revenues and the gains from inflating debt, at the desired rate of inflation of the central bank.
The analysis here bears key lessons for the current debate on sovereign default in a monetary union:
- Countries in a monetary union could indeed be more vulnerable to debt crises, to the extent that the common central bank cannot count on the joint support of national fiscal authorities;
- The common central bank, however, still has the power to engineer successful interventions.
In doing so, it will have to weigh the benefits and costs of providing a backstop to countries exposed to debt crises, possibly drawing on seignorage accruing to all countries in the union.
Editor’s note: The views expressed in this column are those of the authors and do not necessarily reflect those of the European Central Bank.
Aguiar M, M Amador, E Farhi and G Gopinath (2012), “Crisis and Commitment: Inflation Credibility and the Vulnerability to Sovereign Debt Crises”, mimeo.
Blanchard, O (2013), “Rethinking Macroeconomic Policy”, VoxEU.org, 9 April.
Calvo G (1988), “Servicing the Public Debt: The Role of Expectations”, American Economic Review, 78(4), 647-661.
Corsetti G and L Dedola (2013), “The mystery of the printing press: self-fulfilling Debt Crises and Monetary Sovereignty”, CEPR DP.
De Grauwe, Paul (2011), “The European Central Bank: Lender of Last Resort in the Government Bond Markets?”, CESifo Working Paper No. 3569, September.
IMF (2013), “Rethinking Macro Policy II: First Steps and Early Lessons”, conference, 16-17 April.
Krugman, Paul (2011), "The Printing Press Mystery", blog, 17 August.
Reinhart, Carmen M and Rogoff, Kenneth S (2011), “The Forgotten History of Domestic Debt”, The Economic Journal 121(552), 319-350.
1 As a matter of fact, Calvo (1988) had already worked out an example in which inflation rates and thus seigniorage and monetisation may be driven by self-fulfilling beliefs under special assumptions. We further show that the problem is much more general.