Euro, Au Revoir? PIIGS’ Term Structure of Default Probabilities

Should the “Greek case” infect the other high (public and foreign) debt – low competitiveness countries, the PIIGS (Greece, Ireland, Portugal, Spain, and Italy), the stability of the euro and the cohesion of the Euro zone would be in jeopardy. Are we on the verge of a new domino effect, a replay of the European crisis of 1992, but this time with a flight from the sovereign debts, rather the currencies? Or, absent painful reforms, are the PIIGS in for a slow agony of the fixed exchange rate and fiscal tightening? And which will be, after Greece, the next target?1

The Worst Case Scenario

A roll-over crisis, namely the refusal of underwriters to renew the maturing debt, would  precipitate Greece into default, forcing a restructuring of the debt, and sending the signal that even euro countries may fail. In the worst case scenario, the flight from debt will spread to all PIIGS, and that would not be a pleasant ride: the  PIIGS’s short-term foreign debt (not total!) represents 23% of the Euro-zone’ GDP (16 countries!). A European bail out of such a scale would be almost impossible: there is simply no “federal” European government to squeeze French and German tax payers and finance the PIIGS’ debts. The ECB may come to the rescue by monetizing the debts, but this would be a blatant violation of its statutory duties: would Germany and France be willing to be part of such a club? In this scenario, the debt crisis would lead to a banking and corporate crisis throughout Europe. All in a situation where the economy of the old continent is already prostrated and its public finances in disarray. Euro, au revoir.

Default probabilities How likely is this scenario? Figure 1 shows the recent evolution of the cumulative probabilities of default that are implicit in CDS spreads, one to five years maturity, on sovereign default 2.

Figure 1



Click graph to enlarge

Source: author’s calculations on data from Data Stream

From the graph we see that (a) these probabilities are highly correlated across countries, especially during periods of frenzy (for example, February 2009), (b) since November ‘09, markets have started to discriminate between PIIGS, with Irish credibility improving (the convincing fiscal adjustment plan: cumulative probability of default on Feb 18, 2010, p = 12.5%  ), Greece in free fall (for the bogus budget data,  p = 24.7%), Portugal getting worse, from January 2010 (for the fragile government, p = 12.8%),  Italy  and Spain slowly worsening ( with default probabilities at 10.1%  and 10.5%, respectively).

The timing of the crisis Few observers have noted that markets not only differently price default risks to GIPSI’s debts, but also different risk profiles over time. In Greece and Portugal, the risks 3 are now skewed towards the short term (one year), while for Italy and Spain towards the medium term. Figures 2 and 3 show that in the first two countries the likelihood that a crisis within a year (currently at 7.8% and 3.5% respectively) has increased substantially. However, when we look at more distant periods (corresponding to maturity of 2 or more years), the risks fall off quickly. The time profile for the likelihood of Spain and Italy is very different (see Figures 4 and 5). For these countries it is relatively low within the year (around 2%), but is perceived on the rise.

Figure 2




Source: author’s calculations on data from Data Stream

How can we explain these differences? After all, the short term economic outlooks of GIPSI are not too dissimilar. I will advance two explanations, which are not mutually exclusive: 1) markets consider Spain and Italy as the “second line” of attack: if Greece and Portugal fall, they will come next 2) Unlike Greece and Portugal, Spain and Italy are seen as “too big too fail”. This reduces demand for insurance, the premia and default probabilities that they embody.

(the Italian version of this article has appeared on   and )

Figure 3



Source: author’s calculations on data from Data Stream

Figure 4




Source: author’s calculations on data from Data Stream

Figure 5


Source: author’s calculations on data from Data Stream


1 I thank Giulio Trigilia, student of doctorate in economics at Collegio Carlo Alberto in Turin for having processed the data presented here and for interesting discussions on their interpretation.

2 The methodology used here is described in U. Cherubini, “Structured products and credit risk”, (mimeo), University of Bologna, (2006).

3 More precisely, the hazard rates, i.e. the probability that a default would occur say in the second year given that it has not occurred before.