The European Sovereign Debt Crisis

Key Takeaways

• The market priced six credits as one. Through every month of 2005, 2006 and the first half of 2007, the ten-year yields of Germany, Ireland, Spain, Portugal, Italy and Greece sat inside a band of 18 to 35 basis points. In June 2007 the whole range was 4.56 to 4.80 percent.
• The two countries that fell hardest owed least. Irish general government debt was 23.9 percent of GDP in 2007 and Spain's 35.7 percent, both under Germany's 63.7 percent. Ireland ran a surplus in 2007 and a deficit of 32.1 percent of GDP in 2010, almost all of it the cost of its banks.
• The single monetary policy tightened into the stress. The ECB raised its main refinancing rate to 1.25 percent from 13 April 2011 and 1.50 percent from 13 July, then reversed both by 14 December. Italy's ten-year peaked at 7.06 percent in November 2011.
• Liquidity came before any promise about solvency. The three-year operation settled on 22 December 2011 lent 489.2 billion euros to 523 counterparties and the second on 1 March 2012 lent 529.5 billion to 800, a net injection the ECB put at roughly 500 billion.
• The most effective instrument was never used. Outright Monetary Transactions, described on 6 September 2012 with no ex ante quantitative limit, has never bought a bond.
• Sovereign yields healed years before bank credit did. The gap between what a Spanish and a German company paid to borrow peaked at 1.34 percentage points in November 2013, and Italy's at 1.54 points in January 2013.
• Cyprus set the precedent that a bank's uninsured creditors pay. In March 2013 its two largest banks were restructured, 1.4 billion euros of subordinated debt was written down and Bank of Cyprus was recapitalised by converting deposits into shares. Capital controls ran from April 2013 to April 2015.
• The recovery clocks ran for different lengths. German real GDP regained its pre-crisis level in the first quarter of 2011, the euro area not until the second quarter of 2015. Spanish unemployment peaked at 26.4 percent while Germany's fell to 4.2 percent.

What Was the European Sovereign Debt Crisis?

The euro had existed for a decade without anyone answering a basic question about it: what happens when a member state cannot pay. The treaties barred members from assuming each other's debts and barred the central bank from financing governments. There was no fund, no procedure and no legal template for a rescue. None of that mattered while every euro area government borrowed at nearly the same price, and it began to matter in late 2009, when an incoming Greek government restated the country's public finances. Eurostat's current figure for the Greek general government deficit in 2009 is 15.4 percent of GDP against a euro area average of 6.3 percent.

What followed was not contagion in the mechanical sense but the repricing of a category. Investors had held these bonds assuming the currency of issue and the credit of the issuer were separable problems, and that membership had solved the second. Greece showed it had not. Five countries ended up in programmes and no two arrived the same way. Ireland guaranteed its banks' liabilities in September 2008 and spent two years discovering what it had guaranteed. Portugal lost market access in spring 2011 after a decade of weak productivity growth. Spain never took a macroeconomic programme, borrowing 41.3 billion euros solely to recapitalise banks. Cyprus had a banking sector of 550 percent of GDP heavily exposed to Greece. Italy, the largest stressed sovereign, took no programme and still spent two years at the centre of the crisis, because its bond market was too large for anything then in existence to refinance.

Chronology of the acute phase

Dated events with the monthly average ten-year government bond yield of the country most directly involved, and the German ten-year for comparison. Yields are OECD monthly series retrieved from FRED.

Read the last two columns together rather than row by row. The German ten-year falls almost monotonically across the episode, from 4.09 to 0.72 percent, while the country in focus swings violently. Capital did not leave the euro area. It moved inside it, from the periphery to the core, and Germany was paid to be the destination.

Why Does This Page Cover More Than Greece?

Greece is where the crisis started, where the losses were largest, and the country almost every retelling collapses the whole episode into. That collapse loses most of what an investor should take from the period, because the Greek story is the least representative in the set. Consider two facts it cannot hold. Greek general government debt was 104.6 percent of GDP in 2007 and Italian debt 103.5 percent, almost identical, yet the market went on to charge them wildly different prices. And Ireland at 23.9 percent and Spain at 35.7 percent had the two lowest ratios in the group, and both ended up borrowing from the rescue funds.

So this page deliberately spends its length elsewhere. The Greek chapters that dominate every popular account, the 2009 restatement of fiscal data, the conditionality attached to the two Greek programmes, the March 2012 bond exchange with private creditors and the 2015 referendum, are summarised here only where the rest of the currency area turned on them. What gets the room instead is what a Greece-only account leaves out: how property bubbles in Ireland and Spain became sovereign problems without either government overspending; why the ECB was raising rates while three members sat inside rescue programmes; the Cypriot depositor bail-in and the capital controls that followed inside a shared currency; why a company in Madrid still paid more than one in Munich for a bank loan in 2013 under the same policy rate; and the rescue funds, collective action clauses and banking union built while the crisis ran.

One distinction is worth making explicitly. Greece was small enough that its default was arithmetically survivable for the rest of the bloc. What made it systemic was its precedent, not its size. Once a euro area government had restructured, every holder of Italian and Spanish paper had to price the possibility that they might too. The transmission was informational rather than financial, which is why the response that eventually worked was also informational: a statement about what the central bank would do, not a transfer of money.

Why Did Six Governments Borrow at the Same Price Before 2008?

Before the crisis, the euro area sovereign bond market behaved as though the credit question had been abolished.

Range of monthly average ten-year government bond yields across Germany, Ireland, Spain, Portugal, Italy and Greece. The final column is the entire spread from the cheapest borrower to the dearest, in basis points.

Twenty-four basis points is the sort of gap that separates two issues from the same issuer with slightly different maturities. It cannot carry information about the relative solvency of six sovereign states, one with debt of 104.6 percent of GDP and another 23.9 percent. The market was pricing an assumption that monetary union had removed both currency risk and, by extension, credit risk. The first half was correct as a matter of contract. The second was an inference nobody had written down.

The consequence was not merely cheap government borrowing. Private borrowers in the periphery borrowed at core interest rates while their economies grew and inflated faster, so real interest rates were lowest where credit growth was already fastest. Cypriot private indebtedness reached 310 percent of GDP in 2011, up from 213 percent in 2007. Convergence in the price of money produced divergence in the economies using it.

The pattern that repeats. An arrangement removes an obvious risk, the market infers a related but different risk has also gone, and the inference is priced for years without being tested. It is the mistake that made a currency peg feel like a guarantee in the Asian financial crisis. Ask it of any implausibly tight spread: what is the guarantee, who wrote it, and what does it promise?

How Did Ireland and Spain Reach a Rescue Owing Less Than Germany?

Because their governments did not create the debt. Their banks did, and the governments took it on.

General government consolidated gross debt and budget balance, both as a percentage of GDP, from Eurostat's government deficit and debt dataset. A positive balance is a surplus.

Spain ran a budget surplus of 1.9 percent of GDP in 2007, larger than Germany's, and Ireland 0.3 percent, also larger. Both sat comfortably inside every fiscal rule the euro area had. Three years later Ireland recorded a deficit of 32.1 percent of GDP in a single year, which is not a number any tax or spending decision produces. It is the accounting entry for putting capital into failed banks, and the ESM records over 60 billion euros injected into the Irish banking system.

The Irish sequence is the exact inverse of the story the crisis is usually told as. A property boom financed by bank borrowing collapsed; the banks became insolvent; in September 2008 the government guaranteed their liabilities before anyone knew the size of the hole; and in November 2010 the sovereign itself needed rescuing. Spain's version is slower and less total: its savings banks had concentrated lending in construction and property development, and by 2012 recapitalising them cost more than the sovereign could raise while its own yields were rising. Portugal entered with a higher ratio and a different problem, and Italy with the highest ratio of all and never lost market access. Four countries, four diseases, one label.

What Is the Sovereign Bank Doom Loop, and Which Way Did It Run?

The doom loop is the mutual dependence between a government and the banks headquartered in its territory. It has two legs, and the euro area ran both at once in different countries, which is why the crisis resisted a single diagnosis.

Leg one runs from banks to the sovereign. Banks fail, the government recapitalises or guarantees them because the alternative is a payments system failure, and the cost lands on the public balance sheet. Ireland is the pure case, Spain and Cyprus variants of it. The government's credit deteriorates in proportion to the size of its banking system relative to its economy, which is why a small country with an oversized banking sector is exposed in a way a large country with the same banks is not.

Leg two runs from the sovereign to banks. Domestic banks hold large quantities of their own government's bonds. When those bonds fall, the banks lose capital, their funding costs rise, their lending contracts, and their capacity to absorb further government issuance shrinks, which raises the sovereign's borrowing cost again. Italy and Spain ran this leg hard through 2011 and 2012.

Two features made both legs worse. A euro area government issues debt in a currency it does not control, so no monetary authority has an unconditional obligation to make its payments. That is the condition of an emerging market borrowing in dollars, and it is why Japan can carry a far higher debt ratio without a market crisis, as the Japanese asset bubble case study describes. And supervision was national in 2010, so the supervisor assessing a country's banks reported to the government that would have to rescue them.

The contrast with the 2008 financial crisis is exact. There, banking losses landed on a sovereign whose central bank could and did buy its debt without limit, so the question was how much loss the state would absorb, never whether it could. In the euro area that question had no answer for two years.

How Far Apart Did Euro Area Borrowing Costs Actually Get?

Peak monthly average ten-year government bond yield during 2008 to 2015, the German ten-year in the same month, and the gap.

The peaks did not happen together, and the sequence says something the aggregate does not. Ireland peaked first, in July 2011, eight months after entering a programme. Being rescued did not lower its borrowing cost: Irish yields went from 6.42 percent in October 2010 to 8.22 percent in November, the month the package was agreed, and kept climbing for another eight months. A programme replaces market funding; it does not persuade the market that the country will be able to return.

Italy and Spain peaked at much lower absolute levels, and those levels mattered far more. Seven percent became a watched threshold because Greece, Ireland and Portugal had each requested assistance not long after crossing it. When the Italian ten-year reached 7.06 percent in November 2011 the arithmetic was simple: Italian debt was 119.1 percent of GDP that year, and no rescue fund then in existence could refinance it. Spain's July 2012 peak came closest of all, with two countries in the same currency, under the same central bank, at the same policy rate, priced 555 basis points apart. Part of that gap was not a credit judgment but a judgment about whether both would still be using the same currency at maturity.

Why Did the ECB Raise Interest Rates in 2011?

This decision looks most obviously wrong in hindsight and is the most instructive about how a single monetary policy behaves over an unequal currency area. With Greece, Ireland and Portugal inside rescue programmes, the ECB raised its main refinancing rate to 1.25 percent from 13 April 2011 and 1.50 percent from 13 July.

The stated reason was inflation, and the numbers were real. The ECB's own annual report for 2011 records euro area consumer price inflation averaging 2.7 percent for the year, up from 1.6 percent in 2010, and peaking at 3.0 percent between September and November against an aim of below but close to 2 percent. Reading an aggregate that mixed a booming German economy with three contracting programme countries, a central bank looking at 3.0 percent had an argument for tightening. The reversal took five months.

ECB official interest rates and effective dates, from the ECB's published rate history.

The lesson is not that the increases caused the Italian crisis, which had several causes running at once. It is structural. A single policy rate is set for an average, and no member experiences the average. In 2011 the same 1.50 percent applied to a German economy that had already regained its pre-crisis output level and to a Portuguese economy that would not regain its own for another seven years.

What Did the Rescue Funds Lend, and to Whom?

The euro area started the crisis with no rescue vehicle and built three in three years. The Greek Loan Facility of 2010 pooled bilateral loans from other euro area governments; the European Financial Stability Facility was created the same year as a temporary special purpose vehicle; and the European Stability Mechanism, a permanent treaty-based institution, took over from 2012. The European Commission lent alongside them through the European Financial Stabilisation Mechanism, and the IMF joined every programme except the Spanish one.

Amounts lent by each creditor, from the European Stability Mechanism's own programme pages. Figures in billions of euros.

Two things are visible there that the headline narrative hides. Greece received roughly 288 billion euros across three programmes, more than the other four countries combined. And Spain took a single loan for a single purpose, drawn down between December 2012 and February 2013 and completed the same year, so grouping it with Greece in one sentence misdescribes both. Ireland left in December 2013 and Portugal in May 2014, both having returned to the bond market first. Portugal's average loan maturity was extended to 21 years from 14 in April 2013, which is the quiet mechanism by which a rescue becomes affordable. Not forgiveness, but time.

What Did the Two Three-Year LTROs Change?

By late 2011 the problem had stopped being purely about governments. Banks in the stressed countries were losing access to wholesale funding, which meant a solvent bank could fail for reasons unrelated to its loan book. The ECB's response was to lend against collateral for an unprecedented term. The first three-year operation settled on 22 December 2011 and lent 489.2 billion euros to 523 counterparties, including 45.7 billion shifted out of an earlier twelve-month operation. The second settled on 1 March 2012 and lent 529.5 billion to 800 counterparties, a combined net injection the ECB put at around 500 billion.

What this did and did not do is the interesting part. It removed the funding failure mode almost immediately: no euro area bank was going to fail for want of three-year money after March 2012, and Italian yields fell from 6.81 percent in December 2011 to 5.05 percent in March. But it addressed liquidity, not solvency. By June 2012 the Spanish ten-year was back at 6.59 percent and the Italian at 5.90 percent, both higher than in March. Cheap term funding does not make a bank's borrowers repay or a government's debt sustainable. It buys time.

It also deepened the second leg of the doom loop. A bank able to borrow at 1 percent for three years and buy its own government's short-dated paper at a much higher yield has an obvious trade, and it strengthened exactly the link between national banks and national sovereigns that made the system fragile.

Did One Sentence in July 2012 End the Crisis?

On 26 July 2012, speaking at a conference in London, the President of the ECB said that within its mandate the ECB was "ready to do whatever it takes to preserve the euro", and added that it would be enough. The usual telling of that remark is misleading in two directions at once.

It understates it, because the sentence was not a rhetorical flourish. It was followed on 2 August by an announcement of intent and on 6 September 2012 by a published framework. Under Outright Monetary Transactions the Eurosystem could buy a member state's bonds in the secondary market, focused on maturities of one to three years, with no ex ante quantitative limit on size, and accepting the same treatment as private creditors rather than the seniority it had effectively claimed earlier. It required the country to be in an EFSF or ESM programme under strict conditionality, which is what makes it a monetary policy instrument rather than a fiscal transfer.

It also overstates it, because much was left undone. Spanish and Italian yields did fall sharply, the Spanish ten-year from a monthly average of 6.80 percent in July 2012 to 5.34 percent in December. But euro area real GDP did not trough until the first quarter of 2013, two quarters after the framework was published, Spanish unemployment not until February 2013, and Cyprus had its entire crisis after the speech.

The most effective instrument was never used. Outright Monetary Transactions has never purchased a bond. Its whole effect came from being credible and available. A commitment works that way only when the institution can plainly deliver, has the legal authority, and faces no constraint the market can count. Remove any of those and the same words are just words.

Why Was Cyprus Resolved by Taking Money From Depositors?

Cyprus in March 2013 changed the rules for everyone else and is almost entirely absent from popular accounts of the crisis. The setup was extreme even by the standards of this period. The Cypriot domestic banking sector, including cooperative credit institutions, amounted to 550 percent of GDP, and its two largest banks, Bank of Cyprus and Cyprus Popular Bank, known as Laiki, had expanded heavily into Greece. The state was too small relative to its banks to rescue them the way Ireland had tried.

The first attempt was extraordinary. On 16 March 2013 the agreement announced a one-off stability levy applying to resident and non-resident depositors, both insured and uninsured. Levying insured deposits would have broken the promise on which deposit insurance rests across the entire European Union, and the Cypriot House of Representatives did not adopt the proposal. On 18 March the Central Bank of Cyprus declared a bank holiday, extended until 28 March, during which cash withdrawals at ATMs generally continued under normal individual limits, with Laiki imposing a cap of 260 euros a day.

The second attempt became the template. Resolution legislation was adopted on 22 March 2013 and Laiki entered resolution on 25 March: its performing assets, insured deposits and central bank liquidity assistance moved to Bank of Cyprus, while its uninsured deposits and remaining assets stayed in the legacy entity. Around 1.4 billion euros of subordinated debt was written down, and Bank of Cyprus was recapitalised by converting deposits into equity without public money. The banking sector shrank to 350 percent of GDP, and capital controls imposed when the banks reopened were not fully lifted until April 2015.

Three things follow that matter well beyond Cyprus. The distinction between insured and uninsured deposits stopped being theoretical: European Union rules protect deposits up to 100,000 euros, and in Cyprus that line separated full protection from conversion into bank shares. Capital controls turned out to be possible inside a monetary union, so for two years a euro in a Cypriot bank was not fully interchangeable with a euro in a German one. And shareholders and uninsured creditors absorbed losses before any public money was committed, the principle later written into the European Union's bank resolution framework. The comparable American protection is FDIC deposit insurance, and in the 2023 Silicon Valley Bank episode the authorities went the other way, invoking a systemic risk exception so uninsured depositors were made whole. Same fact pattern, opposite decision.

Why Did Credit Stay Expensive in Spain and Italy After Yields Fell?

This is the most under-told part of the episode, and where the sovereign bond charts most badly mislead. A central bank sets one policy rate, but what reaches a business is the rate its bank charges. Between 2011 and 2014 those two things came apart, and the divergence outlasted the bond market panic by more than a year.

Composite cost of borrowing for non-financial corporations, new business, monthly, from the ECB's MFI interest rate statistics. The final column names whichever of Spain, Italy, Portugal and Ireland sat furthest above Germany in that month, in percentage points. Ireland is in that comparison but has no column of its own, to keep the table readable.

Start with the first row, because it is startling. In June 2007 a Spanish company borrowed more cheaply than a German one, 5.10 percent against 5.53 percent. Now look at when the gaps peaked. Spain's widest gap over Germany was 1.34 percentage points, in November 2013. Italy's was 1.54 points, in January 2013. Ireland's was 1.57 points, in November 2014. Portugal's was 3.51 points, in August 2012. Only Portugal peaked while the sovereign bond panic was still running.

That is the transmission mechanism of monetary policy breaking, and it is why the ECB kept easing long after the headlines stopped. The same 0.50 percent policy rate applied in Spain and Germany in November 2013 and produced borrowing costs 1.34 percentage points apart. A business does not experience the policy rate. It experiences the price its own bank quotes, which reflected that bank's funding cost, its sovereign bond holdings, its loan book and the credit risk of borrowers in a depressed economy. All four were damaged in exactly the countries where cheap credit was most needed.

Where to look when a crisis appears to be over. Sovereign spreads are the fastest-moving and most-quoted measure of stress, which makes them the first to recover and the worst single indicator of whether the problem is fixed. Bank lending rates, non-performing loans and credit volumes recover on a much slower clock, and they are what a real economy feels. See the credit cycle and refinancing.

What Did the Euro Area Build After 2012?

The answer was institutional, and most of it was built while the crisis ran.

A permanent rescue fund. The European Stability Mechanism replaced the temporary EFSF as a treaty-based institution with paid-in and callable capital. In the current consolidated treaty its authorised capital stock is 714,674.7 million euros, and the treaty's recital sets an initial maximum lending volume of 500,000 million euros.

A restructuring procedure written into the bonds. Article 12(3) of the ESM Treaty requires collective action clauses in all new euro area government securities with maturity above one year, from 1 January 2013. Such a clause lets a qualified majority of bondholders bind the minority to a change in terms, which matters because there is no bankruptcy procedure for a sovereign. It is a consequential admission: the euro area wrote the machinery for an orderly default into its own government bonds.

Supervision moved above the national level. On 4 November 2014 the ECB assumed responsibility for supervising euro area banks under the Single Supervisory Mechanism, directly supervising 120 significant banking groups representing 82 percent of euro area banking assets at that date and setting standards for roughly 3,500 smaller banks. That attacks the first leg of the doom loop, because the supervisor no longer answers to the government that would pay for a failure.

The books were opened first. Before taking over, the ECB assessed the 130 largest euro area banks as at 31 December 2013 and published the results on 26 October 2014: a capital shortfall of 25 billion euros at 25 banks, asset values requiring adjustment of 48 billion, and non-performing exposures rising by 136 billion to a total of 879 billion. Its adverse scenario would have depleted 263 billion euros of capital, taking the median common equity tier 1 ratio from 12.4 to 8.3 percent. That total deserves a pause: more than a year after the crisis is conventionally said to have ended, the system still carried 879 billion euros of non-performing exposures.

Why Is Redenomination Risk Not the Same as Default Risk?

Default risk is the possibility that a borrower fails to pay what it owes. Redenomination risk is the possibility that it pays exactly what it owes, in a different currency, worth less. These are separate risks, and the euro area is the only place in modern finance where a large investment-grade bond market carried both at once.

The mechanics matter. A Spanish government bond issued under Spanish law is denominated in euro, but if Spain left the monetary union its own legislature could redenominate obligations governed by its own law. A bond issued under English law could not, which is why the governing law clause stopped being a technicality in 2011 and became a priced characteristic. A holder was no longer only asking whether Spain would pay, but what they would be paid in. That is why an instrument aimed at redenomination worked when funding operations had not.

The lesson transfers to any cross-border position. The Asian financial crisis is the mirror image: borrowers owed dollars they could not print, and devaluation made local-currency earners insolvent. The euro area inverted it, because the borrower owed a currency it could not print and the danger was that it would start printing a different one. Both are covered from the portfolio side in international investing.

How Far Apart Were the Recovery Clocks Inside One Currency?

Asking when the crisis ended produces a different answer for every measure and country, so the question needs a stated definition. The one used here is when quarterly real GDP first regained its own pre-crisis peak.

Chain-linked volume real GDP, seasonally and calendar adjusted, quarterly, from Eurostat data retrieved through FRED. Peak is the highest quarter before 2010. Percentages are computed from the levels in those series.

The German row reframes everything. Germany's 2008 to 2009 contraction of 7.0 percent was larger than the euro area's 5.7 percent, because Germany is an export economy and world trade collapsed. Yet Germany was back to its previous peak by the first quarter of 2011, four and a quarter years before the euro area aggregate managed it. The 2008 recession hit Germany harder than the average; the sovereign debt crisis did not touch it. The euro area also had a second recession most American accounts omit: aggregate real GDP peaked again in the first quarter of 2011 and fell about 1.8 percent to a trough in the first quarter of 2013.

Harmonised unemployment rate, monthly, OECD series retrieved from FRED. The low is the lowest month from 2007 onward.

Germany's entry has to be written differently, because German unemployment has no crisis peak. Its highest reading of the period is the first month of it, and it fell more or less continuously to 4.2 percent by the end of 2015, while Spain's more than tripled. One currency, one central bank, one policy rate, and labour market outcomes moving in opposite directions for most of a decade. That is the euro area's unresolved problem stated as data, and no monetary instrument addresses it.

Which Signals Were Readable Before 2010, and Which Only Afterwards?

Signals classified by whether an investor could have acted on them using information available before the Greek restatement of late 2009.

The structural read was available and the timing read was not. An analyst in 2007 could have said that pricing six sovereigns identically was wrong and that the euro area had no mechanism for a member state in trouble. Both were correct, and neither told you Ireland would need help in November 2010 and Italy none at all. Being early here meant being early by years, while the carry on peripheral bonds stayed positive, a gap between correct and profitable discussed in cognitive biases in trading.

Common Myths About the European Sovereign Debt Crisis

"It was caused by governments spending too much." This describes Greece reasonably and Portugal partially. It does not describe Ireland, with debt of 23.9 percent of GDP and a surplus in 2007, or Spain, at 35.7 percent with a surplus of 1.9 percent. Both were below Germany, and in both, causation ran from private borrowing to bank losses to public debt.

"The euro crisis is the Greek crisis." Greece received more than the other four programme countries combined, so the emphasis is understandable. But Spain's programme was a 41.3 billion euro bank recapitalisation loan with no macroeconomic conditionality, Cyprus was resolved by converting deposits into equity, and Italy, arguably the most dangerous case for the currency, never took a programme at all.

"It ended in July 2012." Euro area real GDP kept falling until the first quarter of 2013. Spanish unemployment peaked in February 2013, the Cypriot resolution came in March 2013, the Spanish corporate borrowing gap peaked in November 2013, and euro area banks still carried 879 billion euros of non-performing exposures at the end of that year. What ended in mid-2012 was the repricing of sovereign bonds, one market among several.

"Depositors are always protected in developed markets." Cyprus in 2013 is the counterexample, inside the European Union, in a country using the euro. Uninsured depositors at Bank of Cyprus had deposits converted into shares, and the initial proposal would have levied insured deposits too, rejected by the Cypriot parliament rather than ruled out by any rule.

"A currency union removes country risk." It removes exchange rate risk between members, which is neither the same thing nor implies it. Removing the exchange rate concentrates the adjustment elsewhere: into unemployment, into wages, and into the price a domestic bank charges a domestic borrower. Spanish unemployment reached 26.4 percent while Germany's fell to 4.2 percent, under one policy rate throughout.

What a Reader Can Actually Carry Forward

What generalises

• A spread compressed to nothing is a claim, and you can name it. Twenty-four basis points across six sovereigns in June 2007 asserted that currency union had removed credit risk. Nobody wrote that down and no institution promised it. When a spread narrows below what the underlying difference explains, ask what promise is being priced and who made it.
• Ask who pays if the banks fail, and how big they are relative to that payer. Ireland's banking losses were survivable in absolute terms and not relative to a small economy that had guaranteed them. Cyprus's banking sector was 550 percent of GDP. That ratio is public and needs no forecast. Risk management covers sizing exposure against the entity that would absorb it.
• Borrowing in a currency you do not control is a distinct risk class. It links euro area members to emerging market dollar borrowers, and explains why a debt ratio comfortable for one sovereign is dangerous for another. The question is not how much a government owes but whether anyone is obliged to make its payments.
• Sovereign spreads recover first and mean least. Government yields normalised in 2012 and 2013 while corporate borrowing costs, unemployment and non-performing loans were still deteriorating. To know whether a crisis is over, look at the price of credit to real borrowers.

What does not generalise

• The unused commitment. It worked without buying anything because the ECB could plainly deliver, had defined the legal framework, and faced no countable constraint. Most verbal interventions have none of those properties.
• The depositor bail-in. Cyprus happened in a country whose banks were far too large for its state and where an unusual share of deposits was uninsured and non-resident. The 2023 American response to a comparable bank went the other way.

The question worth asking now

Not whether the euro will survive, which nobody can answer and which has been forecast wrongly for fifteen years. A narrower one: for every holding you own, what entity would absorb a loss, and is it large enough relative to that loss to be credible? For a bank deposit above the insured limit that means the bank's own balance sheet. For a government bond it means whether anyone is obliged to fund the issuer in its currency of obligation. Both are answerable today, with published data.

References

Nothing on this page is quoted from a secondary account of the crisis. Every yield, spread, unemployment rate and output level below was recomputed by Swoopr Investment from the underlying statistical series, and every institutional figure was read out of the issuing body's own document. All sources were retrieved on 26 August 2026.

• European Central Bank: Official Interest Rates: every policy rate and effective date quoted.
• European Central Bank: ECB Decides on Measures to Address Severe Tensions in Financial Markets, 10 May 2010: the Securities Markets Programme and the sterilisation of its purchases.
• European Central Bank: Verbatim of the Remarks Made by Mario Draghi, 26 July 2012: the wording and date of the London remarks.
• European Central Bank: Technical Features of Outright Monetary Transactions, 6 September 2012: the conditionality, maturity focus, absence of a quantitative limit and pari passu treatment.
• European Central Bank: Annual Report 2011: 2011 inflation of 2.7 percent against 1.6 percent in 2010, peaking at 3.0 percent, and the operation settled 22 December 2011 with 523 counterparties and 489.2 billion euros.
• European Central Bank: Annual Report 2012: the second operation, settled 1 March 2012 with 800 counterparties and 529.5 billion euros, and the combined injection of around 500 billion.
• European Central Bank: ECB's In-Depth Review Shows Banks Need to Take Further Action, 26 October 2014: the assessment of 130 banks as at 31 December 2013 and every figure quoted from it.
• ECB Banking Supervision: ECB Assumes Responsibility for Euro Area Banking Supervision, 4 November 2014: the transfer date and the supervisory scope figures.
• ECB Data Portal: Cost of Borrowing for Corporations, Spain: the Spanish series and, in the same MIR dataset, its German, Italian, Irish and Portuguese equivalents.
• Eurostat: Government Deficit and Debt (gov_10dd): the dataset behind every debt and budget balance figure.
• European Stability Mechanism: Ireland: the September 2008 guarantee, the lender amounts, the over 60 billion euros injected into Irish banks and the December 2013 exit.
• European Stability Mechanism: Portugal: the lender amounts, the April 2013 extension of average loan maturity to 21 years from 14, and the May 2014 exit.
• European Stability Mechanism: Spain: the 41.3 billion euros disbursed between December 2012 and February 2013 for bank recapitalisation only.
• European Stability Mechanism: Cyprus: the lender amounts, and capital controls imposed in April 2013 and fully lifted in April 2015.
• European Stability Mechanism: Greece: the amounts lent under the Greek Loan Facility, the EFSF, the IMF and the ESM.
• European Stability Mechanism: Treaty Establishing the European Stability Mechanism, Consolidated Version: Article 12(3) on collective action clauses, the authorised capital stock and the initial maximum lending volume.
• European Stability Mechanism: The Revised ESM Treaty Envisages the Introduction of Single-Limb Collective Action Clauses. What Are They?: how such clauses work.
• European Commission: The Economic Adjustment Programme for Cyprus, Occasional Papers 149, May 2013: every Cypriot figure and date quoted, including the banking sector at 550 percent of GDP reduced to 350 percent, private indebtedness of 310 percent in 2011 against 213 percent in 2007, the 16 March levy proposal, the bank holiday, Laiki entering resolution on 25 March, the 1.4 billion euros written down and the 260 euro ATM limit.
• European Commission: Deposit Guarantee Schemes: the European Union deposit protection level of 100,000 euros.
• Federal Reserve Bank of St. Louis, FRED: Monthly Ten-Year Government Bond Yield Series: IRLTLT01GRM156N for Greece, with IRLTLT01DEM156N, IRLTLT01IEM156N, IRLTLT01ESM156N, IRLTLT01ITM156N and IRLTLT01PTM156N. Every yield and spread here was computed from these six.
• Federal Reserve Bank of St. Louis, FRED: Monthly Harmonised Unemployment Rate Series: LRHUTTTTESM156S for Spain, with the Greek, Irish, Portuguese, German and euro area equivalents.
• Federal Reserve Bank of St. Louis, FRED: Quarterly Real Gross Domestic Product Series: CLVMNACSCAB1GQEA19 for the euro area, with the German, Spanish, Portuguese and Greek equivalents.

Figures deliberately not stated. This page gives no TARGET2 balance for any national central bank, because no primary series could be verified while writing. It gives no haircut percentage or principal amount for the March 2012 Greek bond exchange, no figure for the Greek deficit revision of late 2009 beyond Eurostat's current outturn, no current account figure, and no percentage decline for any European equity index.

Method note: yield spreads are differences between monthly averages, so they understate intra-month extremes and are not comparable with intraday quotations. Real GDP peak to trough figures are computed from chain-linked volume levels. Debt and deficit figures are the current Eurostat outturns, which differ in places from what policymakers had at the time.

Rules that can change. Three items here are current rules rather than settled history: the European Union deposit protection level of 100,000 euros, which sits in a directive amended before and where a new crisis management and deposit insurance framework has recently entered into force; the collective action clause requirement in Article 12(3) of the ESM Treaty, which the revised treaty extends to single-limb clauses; and the ESM's authorised capital and lending volume, which changed with its membership and can be altered again. Verify each against the linked source.

This is educational content about a historical episode. It is not investment advice, it is not a forecast, and nothing here should be read as a claim about the condition of any government, bank or currency today.

Frequently Asked Questions