VoxEU Column EU policies

Why an ESM programme could be a kiss of death: Recovery values and subordination

Spain, needing a bailout for its banks, was granted a vague promise by EZ leaders for up to €100 billion. The details remain obscure, yet they matter enormously. This column argues that the so-called subordination effect of fresh official lending could put Spain on the slippery road to ruin. It argues that if sovereign bonds must be bought, this should be done in the secondary market which, would be pari passu with private investors and thus avoid the subordination trap.

‘Recovery value’ is a crude means used by some investors when evaluating sovereign default risk. 

  • The ‘recovery value’ calculation compares the size of the foreign debt (excluding equity, which is loss absorbing) of the country to its total foreign assets.

The difference between these two figures gives the amount the country would not be able to repay if it had to liquidate all its assets to pay of its foreign creditors (but not foreign equity holders).

  • The ratio gives one the ‘recovery ratio’, i.e. how many cents on the euro foreign creditors could expect in this kind of situation.

Rough calculations of these ratios are listed in column (3) of table 1. These numbers provide thus the ratio of (gross) foreign debt to total (gross) foreign assets where (gross) foreign debt is defined as total (gross) foreign liabilities minus inward FDI and minus portfolio equity in the country.

Table 1. Basic data for recovery ratios

Source: Own calculations based on Eurostat, IMF, ECB & European Commission. Total foreign assets, total foreign debt and non loss absorbing debt in billion Euro.

Table 2 shows the recovery ratios in as a share of two things:

  • The overall recovery rate in which no distinction is made between private and official creditors, and;
  • The recovery value for private investors, which must be adjusted for that part of the debt which is senior and will thus be repaid before private investors.

The overall recovery values vary widely among the GIIPS, with a low of close to 60% for Greece. But for Ireland the hypothetical recovery value would even be above 100% because the net debt of the country is more than covered by loss-absorbing equity. Even on a liquidation basis Ireland looks thus like a safe bet.

Italy is not quite in the same situation as Ireland, but given its modest net debt its hypothetical recovery rate would be very close to 100%. Portugal has a slightly higher overall recovery rate than Spain because a higher proportion of its external liabilities are in the form of loss absorbing equity.

This variability in recovery rates points to the importance of FDI and other ‘non debt creating’ capital inflows which can mitigate the vulnerability of a capital importing country.

A critical factor in the recovery value calculation is the amount of official lending since most of this is is ‘senior’ to private debt; in plain English loans from the IMF, the ECB and probably even the EFSF (certainly the ESM) get paid off first. This lowers the recovery value for private investors. (A comparison of the prices of local law bonds to foreign law bonds provides some indication of the relevance of the phenomenon.1

The second column of Table 2 shows the estimated private recovery rate given the known amounts of official funding. For Greece the private recovery rate falls to 35% if one takes into account only EFSF and the bilateral loan facility. If one takes into account the funding Greek banks have received from the ECB as well the recovery rate the private sector can expect actually goes to essentially zero.

Table 2. Overall vs private recovery ratios in the GIIPS

Source: Own calculations based on Eurostat data for the international investment positions.

As the amounts of official lending are increasing continuously it may make more sense to look at what would (will) happen if the official funding of the GIIPS continues.

The series of figures below shows the evolution of the hypothetical recovery value for the GIIPS2 countries as a function of the amount of official financing they receive (all sources combined). It is apparent that the relationship is not linear;

  • Initially official, senior financing has only a small impact on recovery values as the total amount is small relative to the total debt of the country;
  • Once the total amount is large, even small additional amounts can have a large impact on recovery values.

This implies that the argument that the €100 billion provided to Spain will not have a strong subordination effect is misplaced.

Slippery slope and vicious cycle effects of subordination

Even if this sum has only a small impact it will still increase the need for further support, which in turn will lead to larger subordination effects. There is thus a danger that the combined impact of the EFSF programme for bank recapitalization, higher ECB lending to Spanish banks and the SMP leads to a spiral of ever increasing risk premia and ever larger injections of official financing – which then leads to higher risk premia.

This seems to be a danger for Spain, also but to a smaller degree for Portugal, but not for Italy whose recovery value remains rather high even for large values of official financing.

Secondary market purchases by the EFSF (or even by the ESM) should not have this negative effect since they should be pari passu with private investors. Secondary market purchases should thus be preferred to a full EFSF/ESM programme or the SMP (after the ECB asserted absolute seniority in Greece) or even regular ECB lending to banks for which the ECB is also likely to assert seniority.

Figure 1. Greece: Recovery rate for private sector (%) as a function of the rescue package (% GDP)


Figure 2. Portugal: Recovery rate for private sector (%) as a function of the rescue package (% GDP)


Figure 3. Spain: Recovery rate for private sector (%) as a function of the rescue package (% GDP)


Figure 4. Italy: Recovery of private sector (%) as a function of the rescue package (% GDP)


Ghezzi, Piero (2012), "Official lending: Dispelling the lower recovery value myth", VoxEU.org, 21 June.

1This note concentrates on recovery values, not present values of long dates official credits at favorable rates. The argument made recently by Piero Ghezzi (Ghezzi 2012) that official lending could enhance recovery values if it is at below market rates does thus not apply when insolvency actually occurs because at that point the official credits become due immediately.

2As an interesting aside one should note that in the case of Ireland the subordination effect does not work because the recovery rate would anyway be above 100%.

525 Reads