VoxEU Column EU policies Europe's nations and regions

The ECB’s bond purchase programmes and the limits of national risk-sharing

QE in the Eurozone is unusual in that the risks of sovereign debt defaults are shared between the ECB and the national central banks. This column argues that if such risk sharing were applied to the Outright Monetary Transactions programme, it could potentially create insolvency problems for countries with large public debts, especially in a low-growth scenario.

The ECB’s long-awaited programme to buy Eurozone sovereign debt began on 9 March. This plan, ‘quantitative easing’ (QE), is supposed to boost the ECB’s fight against deflation and stimulate European economies by triggering, among other things, a decline in the euro exchange rate. To appease German opposition to QE, the ECB announced that the risks of sovereign debt defaults associated with QE would be shared between the ECB and the national central banks, as these central banks will hold 92% of the sovereign bonds bought through the programme on their balance sheets.

In this respect, QE differs importantly from the ECB’s Outright Monetary Transactions instrument, as announced in September 2012. Outright Monetary Transactions enables the ECB to purchase the debts of distressed Eurozone sovereign states as a stabilising mechanism. In the case of Outright Monetary Transactions, the ECB has stated its intention to hold purchased bonds on its own balance sheet, effectively sharing the associated risks among all Eurozone member states.

By accepting the principle of national risk-sharing in the case of QE, the ECB has created an important precedent that might be applicable to all future ECB bond purchase programmes, including Outright Monetary Transactions. At the very least, the advantages and disadvantages of national risk-sharing are likely to be weighed whenever the ECB initiates a sovereign debt purchase programme in the future.

How far can the ECB go in purchasing Eurozone sovereign debts if risks are substantially shared with the national central banks (see Benink and Huizinga 2015)? In practice, sovereign debt purchases with national risk-sharing are limited by the loss absorption capacities of the national central banks. We provide calculations of these loss absorption capacities using national central bank balance sheet information and prior estimates of the valuation of future seigniorage for the Eurozone as a whole by Buiter and Rahbari (2012).

All national central banks are found to be able to bear any losses stemming from sovereign debt purchases under the current round of QE. In a favourable high-growth scenario, all national central banks can equally bear any losses from an Outright Monetary Transactions-type intervention resulting in the acquisition of 25% of sovereign debts outstanding by the national central banks. The limits of the loss absorption capacities of some national central banks, however, are reached in a less favourable macroeconomic scenario, or if the national central banks have to acquire higher percentages of their sovereign debts. Outright Monetary Transactions generally requires unlimited sovereign bond purchases and is incompatible with national risk-sharing with the national central banks.

The loss absorption capacities of national central banks

The balance sheets of national central banks contain several accounting entries that can be loss-absorbing. These are:

  • Capital and reserves;
  • Provisions;
  • Revaluation accounts;
  • Banknotes in circulation; and
  • Retained profits for the current year.

For the 17 national central banks belonging to the Eurozone in 2013 (excluding Latvia and Lithuania, which joined in 2014 and in 2015) we represent the sum of these loss-absorbing items in Column 1 of Table 1.

Table 1. Loss absorption capacities of national central banks, in billions of euros

Source: Calculations based on data from NCB balance sheets in 2013 and Buiter and Rahbari (2012).

The national central banks, in addition, can rely on income from future seigniorage stemming from the incremental issuance of euro banknotes. These future seigniorage revenues are distributed by the ECB to the national central banks according to their capital shares in the ECB. Buiter and Rahbari (2012) provide a benchmark estimate of the present value of future seigniorage in a non-inflationary environment for the entire Eurozone of €2,065 billion, assuming a rate of economic growth of 1% per year, and a discount rate of 4%. Using this estimate and applying the national capital shares in the ECB, we can calculate the present value of expected future seigniorage for each national central bank as provided in column 2.1 Expected future seigniorage on average represents 61% of total national central bank loss absorption capacity in Column 3.

Applying a somewhat lower discount rate of 3.5%, Buiter and Rahbari (2012) find a higher present value of future non-inflationary seigniorage for the Eurozone of €3,717 billion, yielding correspondingly higher estimates of national central bank loss absorption capacities as seen in Column 4. A higher economic growth rate of 1.5% similarly yields higher overall expected future seigniorage of €3,817bn, with also higher national central bank loss absorption capacities given in Column 5. Conversely, a lower economic growth rate of 0.5% reduces projected Eurozone future seigniorage to €1,273bn, with also lower national central bank loss absorption capacities as provided in Column 6.

QE and potential national central bank losses

In the current round of QE, the ECB is scheduled to purchase about €800 billion of euro country sovereign debt in proportion to the national capital shares in the ECB, with 92% of these purchases transferred to the balance sheets of the respective national central banks. The resulting exposures to national public debts of the national central banks are presented in Column 1 of Table 2. The exposure of the Banca d’Italia, for instance, is projected to be €131.4 billion, given that its capital share in the ECB is 17.86%. The Banca d’Italia’s calculated loss absorption capacity ranges from a low of €472.8 billion in the low-growth scenario in Column 6 of Table 1 to €927.1 billion in the high-growth scenario in Column 5.The Banca d’Italia thus could sustain a full write-down of the public debts acquired under QE even in the low-growth scenario. All other national central banks similarly can be seen to be able to bear losses equal to the public debts acquired under QE in any of the four scenarios.

Table 2. Maximum losses of national central banks under current QE and potential Outright Monetary Transactions scenarios, in billions of euros

Source: Columns 2-4 reflect data on general government debt in 2013 from Eurostat.

How far can national risk-sharing be applied to Outright Monetary Transactions?

Central bank purchases of sovereign debts under Outright Monetary Transactions are potentially much larger than in the current round of QE, with correspondingly larger exposures to national public debts for the national central banks if national risk-sharing is applied. At most, we can assume that all of a Eurozone country’s general public debt is acquired through Outright Monetary Transactions, with 92% put on the national central bank balance sheet. In this event, the hypothetical exposure of the Banca d’Italia to Italian sovereign debt would be €1,903.2, or 92% of total outstanding debt of almost €2.1 trillion, as seen in Column 2 of Table 2. The Banca d’Italia, not surprisingly, cannot sustain a loss of this magnitude in any of the loss absorption capacity scenarios in Table 1. Estonia’s and Slovenia’s national central banks, however, could bear a loss equal to 92% of general public debt in the benchmark loss absorption capacity scenario, while only Estonia’s central bank can do this in the low-growth scenario.

Alternatively, we investigate the cases where Outright Monetary Transactions results in national central bank exposures to national public debts equivalent to 50% and 25% of the amounts outstanding (see Columns 3 and 4 of Table 2). These exposures represent the losses in case of a complete write-down if 50% or 25% of outstanding debts are acquired, or alternatively potential losses if somewhat higher percentages of debts are acquired and there is some recovery of the debts.

A loss of 50% of national public debt can be sustained by the national central banks of six countries (Estonia, Finland, Luxemburg, Malta, Slovakia, and Slovenia) in the benchmark scenario, and by only the national central banks of three countries (Estonia, Luxemburg, and Slovenia) in the low-growth scenario.

The lower loss of 25% of public debts, however, can be sustained by 16 out of 17 national central banks (only excluding Ireland’s) in the benchmark scenario, and by 13 national central banks in the low-growth scenario. The Banque de France, in particular, can sustain a loss of 25% of national public debt in either scenario, but the Banca d’Italia cannot do this in the low-growth scenario, reflecting the relatively high amount of Italian national government debt.


National risk-sharing with the national central banks in the current round of QE implies potential losses to the national central banks that do not call into question their solvency if expected future seigniorage revenues are taken into account. Applying the precedent of national risk-sharing to Outright Monetary Transactions, however, potentially creates national central bank insolvency for countries with large public debts, especially in the low-growth scenario. Hence why Outright Monetary Transactions, which may require unlimited purchases of the public debts of distressed Eurozone countries, is generally incompatible with national risk-sharing.

In his press conference announcing QE in January, President Mario Draghi of the ECB emphatically stated that national risk-sharing would not apply to Outright Monetary Transactions, because “in [Outright Monetary Transactions] full risk-sharing is fundamental for the effectiveness of that monetary-policy measure”. However, by conceding that national risk-sharing would be part of QE, the ECB has provided ammunition to critics of Outright Monetary Transactions who may reason that national risk-sharing should apply to all ECB purchases of public debts, including those that are part of Outright Monetary Transactions. This is unfortunate, as doubts about the effective future use of the Outright Monetary Transactions instrument make a recurrence of the sovereign debt crisis in the Eurozone more likely. The present application of the principle of national risk-sharing in the case of QE undesirably weakens the Outright Monetary Transactions instrument. This suggests that the introduction of national risk-sharing was too high a price to pay to reduce opposition to QE, particularly in Germany.

Authors' note: An earlier, less technical version of this article was published as “QE Undermines the ECB’s Crisis-Fighting Ability” on the opinion page of the Wall Street Journal on March 13-15, 2015 (see wsj.com). The authors would like to thank Philip Pilgerstorfer for excellent research assistance.


Benink, H, and H Huizinga (2015), “QE Undermines the ECB’s Crisis-Fighting Ability”, The Wall Street Journal, 13-15 March.

Buiter, W, and E Rahbari (2012), “Looking into the Deep Pockets of the ECB”, Global Economics View, Citigroup Global Markets Inc., 27 February.

Corsetti, G, L Feld, P Lane, L Reichlin, H Rey, D Vayanos and B Weder di Mauro (2015), A New Start for the Eurozone: Dealing with Debt – Monitoring the Eurozone 1, CEPR Press, March.


1 Corsetti et al. (2015) consider how the non-inflationary seigniorage income stream over a certain horizon could be used by a European stability fund to perform a debt buyback operation of highly indebted countries in the Eurozone.


2,520 Reads