top of page

ECB's quantitative easing

Friday, October 2, 2015


Question 1:

Do you agree that the design of the ECB's QE programme reduces its effectiveness? 



Question 2:

Do you agree that the structure of the ECB's QE programme makes the Eurozone more fragile and increases the risk of one country leaving the euro?













Will the risk-sharing arrangements within the European Central Bank’s quantitative easing (QE) programme reduce its effectiveness? According to the latest monthly survey of the Centre for Macroeconomics (CFM) reported in this column, our panel of experts are exactly evenly divided. The written responses suggest that this divergence reflects differences in views about the channels through which QE operates.


In January 2015, the European Central Bank (ECB) announced a substantial increase in its asset purchase programmes.[1] Under its existing programmes, the ECB had been buying asset-backed securities and covered bonds of around €10bn in total per month. Following the January 2015 announcement, the total assets purchases would increase to €60bn per month between March 2015 and September 2016. The total assets to be bought in this window are €1.1 trillion, 9% of the stock of central and other government debt of Eurozone nations, widely referred to as quantitative easing (QE).[2]

The additional €50bn of asset purchases per month is through the Public Sector Purchase Program (PSPP). This consists of €6bn of debt securities of EU supranational institutions and €44bn of debt securities of sovereign, national agencies and national utilities. These national securities will be bought in proportion to the Eurozone's national central banks' shareholdings of the ECB (in effect, in proportion to the size of national economies).

Risk-sharing in the ECB’s QE

In this CFM Survey, our interest is in the risk-sharing arrangements within the QE programme. The ECB indicated that the credit risk of the €6bn debt of the supranational EU institutions and €4bn of the national debt securities would be shared across the Eurosystem according to shareholdings.[3] The credit risk of the remaining €40bn of national securities would remain with the national central bank of the issuer. This is in contrast to the ECB's earlier Securities Market Programme (SMP) in 2010-12, which involved the acquisition of €220bn public and private debt securities from Greece, Ireland, Italy, Portugal and Spain to be held to maturity. Profits and losses are to be shared across national central banks according to the ECB’s shareholdings rather than the borne by the national central bank of the issuing government.

Effectiveness of QE

According to press reports, the decision to allocate the major fraction of national securities back to national central banks reflects a compromise decision.[4] A number of arguments have been put forward in favour of this approach:

  • First, this is a direct means of coupon payments being kept within national borders.

  • Second, as long as fiscal policy is nation-specific, any credit risk should also stay within borders. If credit risk is perceived as shared across the Eurozone, national governments may be less inclined to implement reforms.

  • Third, as QE increases in size, the ECB may risk its solvency, whereas the national central banks would (potentially) continue to have the fiscal backing of their governments.

  • And fourth, more generally, risk-sharing may have no effect on the efficacy of QE since it has no bearing on the total amount of liquidity (money base) created or where this money will flow.

There are also arguments in favour of greater risk-sharing:

  • First, a rationale for the European Banking Union is to break the link between governments and national banking sectors (the so-called 'doom loop'). Requiring national central banks to hold greater amounts of their sovereigns’ debt may re-establish the link. If the government were to default, this may make the national central bank insolvent and depositors less certain about repayment, possibly leading to capital flight.

  • Second, QE with limited risk-sharing where a government’s solvency is in doubt might increase the cost of market funding relative to a QE programme with more risk-sharing. Giavazzi and Tabellini (2015) point out that if it is recognised that the government cannot default on its bonds to its central bank, then the central bank becomes a senior creditor and private investors become junior creditors. This would increase the cost of market funding in high-risk countries.

Q1: Do you agree that the design of the ECB's QE programme reduces its effectiveness?

Thirty-five of our panel of experts replied to the question. Leaving aside the five who expressed no opinion, or who neither agreed nor disagreed, our respondents were exactly split between those who agreed and those who disagreed with the proposition. Taking account of self-declared expertise, 54% agreed with the proposition.

Those who agreed with the proposition tended to emphasise interest rates as the channel through which QE affects activity. Panicos Demetriades (Leicester) noted that several countries are likely to face higher borrowing costs and David Bell (Stirling) described it as ‘entirely fanciful that the market will ignore the allocation of risk.’

Others were more concerned with signalling. Sir Christopher Pissarides noted that ‘national central banks are to take on the risk but cannot monetise the debt in the event the government cannot pay and cannot affect interest rates. It's an odd situation that might deter private lenders.’ Ethan Ilzetzki (LSE) asked the question ‘if the EU is not willing to bet on its own survival, who should?’ Angus Armstrong (NIESR) went further, suggesting that limited risk-sharing is likely to increase the risk of capital flight, if there is concern about government solvency.

Those who disagreed with the proposition either emphasised the monetary base channel by which QE affects activity, or thought the consequences simply too small to matter. Wouter den Haan (LSE) noted that QE involves adding liquidity to the economy and how risks are shared has no effect on this operation. Similarly, Jagjit Chadha (Kent) said that monetary conditions are determined by the liabilities of the central bank, hence the ‘risk-sharing has little impact on the effectiveness of QE.’ Patrick Minford (Cardiff Business School) described the very notion that credit risk is not shared by the ECB as ‘far-fetched’.

Fragility of the Eurozone

The limited risk-sharing on the €40bn of additional national securities is in contrast to another major monetary union and the ECB's previous actions. In the United States, assets bought under QE are held in the System Open Market Account (SOMA) at the New York Federal Reserve and any losses are shared across the Federal Reserve System. Moreover, while ECB President Mario Draghi has made clear that ‘in Outright Monetary Transactions (OMT) full risk-sharing is fundamental for the effectiveness of that monetary policy measure’, Benink and Huizinga (2015) suggest that subsequent risk-sharing arrangements within QE may embolden those would prefer risks to be contained within national boundaries (the vote for OMT was not unanimous).[5][6]

Q2: Do you agree that the structure of the ECB's QE programme makes the Eurozone more fragile and increases the risk of one country leaving the euro?

Thirty-six of our panel responded to this question. Leaving aside the eight who had either no opinion or neither agreed nor disagreed, the votes were again evenly split between those who agreed and those who disagreed with the proposition.

Those who agreed with the proposition include Ricardo Reis (LSE and Columbia) who noted that if we consider the possibility of a country leaving the Eurozone, reneging on TARGET II and/or ELA balances, then this arrangement encourages a ‘run on their local banks and force an exit.’ Others are concerned about the signal of political commitment. Martin Ellison (Oxford) suggested that QE risk-sharing arrangements do not ‘bode well for the type of risk-sharing reforms that will ultimately be needed to keep the Eurozone together’; and Nicholas Oulton (LSE) noted that the failure of Eurozone countries to agree that ‘we are all in this together’ surely makes the exit of one or more countries more likely.

Those who disagreed with the proposition, tended to emphasise that whatever fragilities exist in the Eurozone, they are not the result of the QE programme. Ray Barrell (Brunel) supported the risk-sharing arrangements as they minimise moral hazard and so prevent higher debt by countries trying to ‘game the system’ by issuing even more debt. Costas Milas (Liverpool) expressed the view that the limited risk-sharing reflects a lack of structural reform, and that it is this lack of reform that undermines growth and creates the fragility. John Driffill (Birkbeck) acknowledged that the QE programme could have been better designed, but said that it does not increase the fragility of the Eurozone. 



Benink, H and H Huizinga, (2015), ‘The ECB’s bond purchase programmes and the limits of national risk-sharing’, 16 May 2015 (

De Grauwe, P and Yuemei Ji (2015), ‘Quantitative easing in the Eurozone: It's possible without fiscal transfers’, 15 January 2015 (

Giavazzi, F and G Tabellini (2015), ‘Effective Eurozone QE: Size matters more than risk-sharing’, 17 January 2015 (



[1] See the ECB’s press release: 'ECB announces expanded asset purchase programme', 22 January 2015. on 22nd January 2015

[2] The total nominal value of outstanding central government debt was €6,798bn and other general government debt was €608bn. See ECB

[3] The credit risk on the supranational bonds is marginal. In the case of the European Investment Bank, the 28 EU members are shareholders and have joint liability. Over 90% of the capital is callable on the member states.

[4] For example, see

[5] Quoted from the ECB’s ‘Introductory statement to the press conference (with Q&A)’, 22 January 2015.

[6] The lack of unanimity in the vote for QE is reported in the ECB’s 'Introductory statement to the press conference (with Q&A)’, 6 September 2012.

bottom of page