VoxEU Column Europe's nations and regions International Finance

The EFSF: Unfinished business

The European Financial Stability Facility was set up eighteen months ago as a response to the then Greek sovereign debt crisis. This column looks at the effect of the fund on the financial system in particular bank shareholders, the holders of bank bonds, and the holders of sovereign debts.

On 9 May 2010, Eurozone countries announced the creation of the European Financial Stability Facility (EFSF) to contain the Eurozone sovereign debt crisis. But is it working?

The EFSF was set up to make loans to Eurozone governments experiencing refinancing problems, with its own debt guaranteed by individual Eurozone countries. As originally conceived, these credit guarantees were limited to €440 billion but in July this year Eurozone politicians agreed to increase the guarantee ceiling to €780 billion, providing the EFSF with an effective lending capacity of €440 billion.

In October, Eurozone leaders decided to leverage up the remaining, uncommitted lending capacity of €250 billion to more than €1 trillion to be able to backstop the debt of a major Eurozone country such as Italy. However, leveraging the EFSF has proven difficult, as investors have shown limited interest to provide the necessary funding.

Structured support

The announcement of the EFSF in May 2010 was important as it signalled a willingness of Eurozone countries to financially support distressed sovereigns in a structured way. The earlier provision of €110 in bilateral loans to Greece in 2010 had only been on an ad hoc basis.

The EFSF’s immediate impact is on governments, as it is a vehicle to organise intergovernmental loans. Effectively, the EFSF transfers the liability to repay sovereign debts from recipient countries to guarantor countries. However, as its name suggests, the EFSF also affects the banking system. Banks are immediately concerned, if the EFSF changes the valuation of sovereign debts on their books. Indirectly, the EFSF potentially alters the creditworthiness of national governments, and hence the credibility of national financial safety nets. Banks, specifically, may see the credibility of their national financial safety nets go up or down, depending on whether their countries receive or guarantee the funding.

The international nature of the EFSF, and the combined impact on countries and banks, render the EFSF more complex than purely national bank bailout schemes. Previously, Veronesi and Zingales (2010) have analysed the costs and benefits of largescale US intervention in the financial industry in 2008 against the background of a solvent sovereign. They concluded that this intervention created a substantial net benefit arising from a reduction in the probability of bank failure. Attinasi et al (2009), and Ejsing and Lemke (2009) examine bank rescue packages offered by European governments in 2008 to find that they were accompanied by sovereign CDS spread rises, suggesting a risk transfer from banks to the sovereigns.

In a new paper (Horváth and Huizinga 2011), we investigate the effects of the EFSF announcement on EU sovereigns and on banks using an event study methodology. The study relies on data on banks’ sovereign exposures from the 2010 EU-wide bank stress tests conducted by the Committee of European Banking Supervisors.

Bailing out sovereigns and banks

We consider the repercussions of the EFSF announcement for: (1) bank shareholders; (2) the holders of bank bonds; and (3) the holders of sovereign debts. The results are as follows:

1.      Bank shareholders overall lost. We find a negative shareholder return of -1.4% on average over a 5-day period surrounding the announcement date, perhaps because the size of the prospective capacity of the EFSF was rather small. However, banks gained to the extent that they were exposed to the periphery countries (Greece, Ireland, Italy, Portugal and Spain, hereafter 'GIIPS'), reflecting a revaluation of these countries’ debts. Furthermore, banks located in Eurozone countries with strained public finances declined in value, as the financial safety nets of these countries may have been improved too little to benefit bank shareholders. Finally, Eurozone banks gained relative to non-Eurozone banks, suggesting that the EFSF improves Eurozone banks’ comparative access to the financial safety net.

2.      The holders of bank bonds gained overall, as evidenced by a drop of the average bank CDS spread by 23 basis points. The gains of bank bond holders were larger for banks heavily exposed to GIIPS sovereign debt, and for banks located in countries with strained public finances, suggesting strengthened financial safety nets in these countries. Eurozone bank bond holders gained relative to bank bond holders in other EU countries, consistent with increased bank access to the financial safety net in Eurozone countries.

3.      On average, the holders of sovereign debts benefited as well, as shown by a reduction of the average sovereign CDS spread by 54 basis points. Sovereign debts of countries with banking systems exposed to periphery country debt rose in value, while sovereign debts of countries with banking systems more exposed to core countries, Eurozone debt declined in value. Sovereign CDS spreads declined for countries with strained public finances, indicating higher values of the sovereign debts of these countries. All these results suggest that the EFSF brought about a transfer of creditworthiness from the ‘Northern’ Eurozone countries to the ‘Southern’ countries, which is a main cost for the ‘Northern’ countries.

The impact of the EFSF announcement on the values of bank sovereign debt portfolios, bank shares, and bank bonds has been significant. Using information on bank CDS spread changes, we calculate that the event increased the valuation of the sovereign debt portfolios of 32 major EU banks by around €10 billion. In addition, the announcement increased the combined value of these banks’ shares and bonds by around €25 billion. These numbers suggest that the EFSF had a direct effect on the value of the bank shares and bonds of the concerned banks – through higher valuations of the sovereign debt portfolios – of €10 billion, and an indirect effect – through improved access to the financial safety net – of €15 billion.

An inefficient tool

The distribution of the gains for banks was tilted towards bank bond holders rather than bank shareholders. This is appropriate, as bank shareholders should not be rewarded for their banks’ risky exposures to sovereign debts.

Banks with risky Eurozone sovereign exposures, however, benefited regardless of whether they were distressed or located in the Eurozone. Thus, the EFSF makes inefficient use of limited public resources to the extent that it aims to promote financial stability inside the Eurozone. In recognition of this, the instrument set of the EFSF was expanded in July 2011 to include loans to Eurozone governments with the purpose of recapitalising distressed banks. In principle, such a loan can be provided to a country that is defaulting on its general debt, or to a country that has negotiated a haircut with its creditors such as Greece. Hence, the proceeds of such a loan can be targeted to bank recapitalisation rather than to general debt repayment. Acharya et al (2011) calculate that the cost of recapitalising EU banks now would be between €200 billion and €500 billion based on sovereign debt discount estimates. Thus, the current resources of the EFSF should be sufficient to honour requests from Eurozone governments to recapitalise their banks.

Limited in size

The size of the EFSF – as of July 2011 – was insufficient to bail out both Eurozone banks and Eurozone sovereigns that are ‘too big to fail’. Hence, the decision of Eurozone leaders in October 2011 to leverage up the EFSF’s remaining, uncommitted lending capacity of €250 billion to arrive at an enlarged lending capacity of at least €1 trillion. In recent weeks, the EFSF has worked to achieve this degree of leverage, but apparently unsuccessfully. In practice, leveraging up the available resources may only attain an effective lending capacity of between €500 billion and €750 billion. This implies that the EFSF would be too small compared to the size of Italy’s sovereign debt of €1.9 trillion. To bring about a larger EFSF, Eurozone countries may need to increase their credit guarantees to the EFSF, even if this comes at a cost of deteriorating credit ratings for the guarantor countries. Larger credit guarantees to the EFSF are warranted, if they are accompanied by the acceptance of far-going EU oversight on budget and reform policies in recipient countries.

In the absence of an expanded EFSF, only the ECB has the capacity to backstop the Eurozone sovereign debt market. Significant debt purchases by the ECB would imply a monetisation of the debt, leading to higher inflation and an erosion of savings everywhere in the Eurozone.

This makes the EFSF route to backstopping the Eurozone sovereign debt preferable to the ECB route. However, the EFSF route appears to require additional action by politicians, while the ECB does not. Eurozone politicians should hasten their efforts to expand the EFSF, even if this implies larger credit guarantees to the EFSF.

Editor’s Note: The author also discusses his research on the EFSF in a recent Vox Talk with Viv Davies of CEPR here.


Acharya, Viral, Dirk Schoenmaker, and Sascha Steffen (2011), “How much capital do European banks need? Some estimates”, VoxEU.org, 22 November.

Attinasi, Maria-Grazia, Cristina Checherita, and Christiane Nickel (2009), “What explains the surge in Eurozone sovereign spreads during the financial crisis of 2007-09?”, Working Paper 1131, European Central Bank.

Ejsing, Jacob and Wolfgang Lemke (2009), “The janus-headed salvation: Sovereign and bank credit risk premia during 2008-09”, Working Paper 1127, European Central Bank.

Horváth, Bálint and Harry Huizinga (2011), “Does the European Financial Stability Facility bail out sovereigns or banks? An event study”, CEPR Working Paper 8661.

Veronesi, Pietro and Luigi Zingales (2010), “Paulson's gift”, Journal of Financial Economics, 97:339-368.

420 Reads