Spain & Banking Union Report 2012 — Whatever It Takes

Greece invited a fiscal reading of the European crisis: a state had spent beyond its means and the market eventually noticed. It was a comfortable diagnosis because it located the fault in national behaviour and implied a remedy in national discipline.

Spain made that reading untenable. Before 2008 Spain ran a budget surplus. Its public debt ratio was materially below Germany's. On every fiscal metric the framework used to police the currency union, Spain was a model member. By 2012 it was at the centre of the crisis and requesting European assistance for its banks.

The explanation is that Spain's problem was never fiscal in origin. It was a private credit boom, concentrated in property, intermediated substantially through savings banks whose governance made them poorly suited to resist it. When the boom reversed, the losses landed on institutions the state was obliged to support, and the support was what created the sovereign problem.

This is the doom loop described in the Greece & the Periphery Report 2010, running in the opposite direction. In Greece, a sovereign problem impaired the banks. In Spain, a banking problem impaired the sovereign. The circuit is the same and it can be entered from either end — which is the finding that matters, because it means fiscal discipline alone cannot prevent it.

Two things followed, and 2012 is remembered for the first while the second did more.

A commitment repriced the market. A statement that the central bank would do what was necessary within its mandate — paired with a mechanism that made it operational — removed the possibility of a funding failure. Spreads compressed without the facility being used. That is only possible where the problem is an equilibrium rather than a fundamental, and it is the cleanest demonstration in the archive that the two are different things.

Banking union followed. Supervision moved to the European level and a resolution framework was constructed to allow a bank to fail without the national sovereign absorbing it. This addressed the mechanism rather than the symptom — and it was left incomplete in the one respect that mattered most.

Why Spain is the diagnostic case

The value of Spain is that it holds the fiscal variable constant and lets you observe the rest.

Spain's pre-crisis position, on the metrics the union actually monitored:

• Budget in surplus in the years immediately preceding the crisis.
• Public debt below the union's reference threshold, and below Germany's.
• Compliant with the fiscal framework in a way several core members were not.

What was accumulating instead was private:

• A very large construction and property sector as a share of output and employment.
• Rapid household and corporate credit growth, funded increasingly from abroad rather than from domestic deposits.
• A substantial external deficit, which is the accounting counterpart of that foreign funding.

The framework did not monitor any of this, and that is the structural lesson. The union's surveillance was built around fiscal deficits and public debt because those were understood as the risks a currency union creates. The risk that actually materialised was a private-sector capital-flow imbalance, invisible to a rulebook watching budgets.

Spain complied with every rule that existed and accumulated the exposure that mattered. A monitoring regime is a statement about which risks you expect, and it is silent about the others by construction.

The external deficit deserves particular emphasis, because within a currency union it looks harmless. There is no exchange rate to defend and no currency crisis to fear, so a persistent deficit reads as capital flowing to where returns are highest. What it actually represents is a cross-border funding dependency — and when that funding reversed, the effect was the same as a sudden stop in an emerging market, minus the devaluation that normally resolves one. The Global Investment Outlook 2013 describes the same sudden-stop mechanism in emerging markets during the taper episode.

The savings banks and the governance problem

The institutions at the centre of the Spanish episode were not principally the large international banks. They were the cajas — regional savings institutions — and their structure is the mechanism.

What made them distinctive:

• No conventional shareholders. Governance sat with boards drawn substantially from regional political and social bodies rather than from owners of capital.
• Strong regional concentration, both geographically and in the sectors their regions depended on.
• Limited ability to raise external equity, because there was no ordinary equity to issue.

Each of those becomes a specific vulnerability in a property boom:

• The discipline that restrains lending in a boom is the owner's fear of loss. Where the board's incentives run toward regional development and local employment, the case for continued lending is politically compelling and the counter-argument has no obvious advocate.
• Concentration removed the diversification that might have absorbed a regional downturn — the same common-factor failure the US Housing & Mortgage Report 2008 describes, at institutional rather than pool level.
• Inability to raise equity meant that when losses arrived, recapitalisation had to come from the state. A shareholder-owned bank can dilute its owners. A caja could only be rescued.

That last point is the direct link to the sovereign. The institutional form determined that losses would land on the public balance sheet, not because of any policy choice made during the crisis, but because of a structure decided long before it.

The consolidation response compounded it initially. Merging weak institutions into larger ones produced entities that were harder to resolve and more systemically significant, without necessarily improving asset quality. Combining several impaired balance sheets does not produce a sound one; it produces a larger impaired one whose failure is less tolerable.

How a property bust becomes a sovereign crisis

The transmission from private losses to sovereign risk is worth stating explicitly, because in a currency union it is unusually direct.

1. Property prices fall, and development loans against land and unfinished projects become the worst exposures — land being the most cyclical collateral and the least liquid.
2. Bank capital is impaired, and for institutions unable to issue equity the only recapitalisation source is the state.
3. The state's contingent liability becomes actual, adding to public debt — the deterioration is caused by the rescue, not by prior spending.
4. Sovereign spreads widen, and domestic banks holding sovereign debt take a further markdown, per the Greece & the Periphery Report 2010 circuit.
5. Credit contracts, deepening the recession, raising unemployment, reducing revenue and widening the deficit further.
6. The debt ratio rises through the denominator as well as the numerator.

The asymmetry that makes this a currency-union problem specifically: a country with its own currency facing this sequence would see its exchange rate fall, cushioning the contraction through exports, and its central bank would backstop the sovereign's funding. Spain had neither. The adjustment fell entirely on domestic wages and employment, and the funding depended on a market that had begun to doubt it.

Requesting European assistance for the banks was itself destabilising, which is the trap in its purest form. Assistance channelled through the state increased sovereign debt, which worsened the spread, which impaired the banks holding sovereign bonds. A rescue routed through the impaired guarantor transmits the problem it is meant to solve — and this is precisely why banking union's central idea was to route it elsewhere.

The commitment that repriced without being spent

The decisive moment of 2012 was a statement, followed by a mechanism that made it credible. The economics of why that worked are the important part.

The problem, restated from the Greece & the Periphery Report 2010. Sovereign borrowing costs are self-fulfilling: a market that doubts solvency demands a premium that creates insolvency. Two equilibria exist over identical fundamentals — a good one where borrowing is cheap and the debt path stabilises, and a bad one where it does not.

What a credible backstop does. It does not improve the fundamentals. It removes the bad equilibrium from the set of possible outcomes. If investors know a buyer will step in at a ceiling, the panic scenario cannot occur, so pricing it stops being rational.

Why the commitment did not need to be exercised. In a multiple-equilibria problem the announcement is the intervention. Once the bad outcome is off the table, the market moves to the good one on its own, and no purchases are required. Spreads compressed substantially without the facility being drawn.

A backstop that is never used is not a backstop that was unnecessary. It is a backstop that worked, because its entire function was to make the scenario it insured against impossible to reach.

Two conditions made it credible, and both were essential:

• Unlimited capacity in principle. A facility with a stated maximum invites the market to test it. The commitment's power came from having no ceiling a speculator could exhaust.
• Conditionality. Support was tied to a programme, which addressed the objection that a backstop removes the incentive to adjust. That constraint is what made the commitment politically survivable, and therefore believable.

The general lesson recurs throughout this archive. The Global Investment Outlook 2008 observes that a funding run stops the moment a credible lender exists, because withdrawal stops being rational. This is the same mechanism at sovereign scale — and it explains why the European crisis appeared to end abruptly rather than to heal gradually.

What banking union actually severed, and what it left

Banking union was the structural response, and assessing it requires separating what it fixed from what it did not.

What it addressed:

• Supervision moved above the national level, removing the conflict in which a national supervisor oversees banks whose failure would fall on its own sovereign — a supervisor with a reason not to look too hard.
• A resolution framework was built to impose losses on a failed bank's creditors rather than on taxpayers, so a bank failure need not become a fiscal event.
• Common standards reduced regulatory divergence that had let comparable risks be treated differently across members.

What it left unresolved, and this is where the loop survives:

• Sovereign exposures kept their zero risk weight. The single mechanism that made domestic sovereign debt the cheapest asset a bank could hold — and therefore produced the concentration — was not changed. It remains the largest unfinished item.
• Common deposit insurance was never completed. Deposit guarantees stayed national, so a depositor's protection still depends on the strength of their own sovereign. That is the doom loop preserved in the one place it is most visible to the public, and it is why deposit flight across borders remains possible.
• Legacy assets stayed on national balance sheets, so the framework governed future failures more than existing damage.

The honest assessment is that the loop was attenuated rather than broken. A bank failure is far less likely to become a sovereign crisis than it was in 2012. But a bank still holds its own sovereign's debt without capital against it, and its depositors are still insured by that sovereign. The mechanism has been made harder to enter, not removed — which is worth knowing whenever European bank exposure is assessed.

What an allocator could act on

Do not read fiscal compliance as safety. Spain satisfied every rule the union monitored and accumulated the exposure that nearly destroyed it. A surveillance regime tells you which risks were anticipated; the unmonitored ones are where the position builds.

Watch the external balance inside a currency union. A persistent current-account deficit within a monetary union looks benign because no currency can break. It is a cross-border funding dependency, and its reversal behaves like a sudden stop without the devaluation that normally ends one.

Ask how an institution recapitalises before it needs to. A bank that cannot issue equity has only the state as a capital source, which pre-determines that its losses become public. Institutional form, decided long in advance, dictates who bears the loss.

Distinguish a fundamentals problem from an equilibrium problem, because the trades differ entirely. An equilibrium problem can resolve abruptly and cheaply on a credible commitment. Positioning for slow fundamental repair will miss the move, and it will miss it violently.

Treat an unexercised backstop as evidence it worked. Facilities that go undrawn are routinely cited as unnecessary. Where the mechanism is expectations, the absence of use is the measure of success.

Check whether a structural reform addressed the mechanism or the symptom. Banking union moved supervision and resolution upward, which is mechanism. It left the zero risk weight and national deposit insurance, which is where the mechanism still lives. Reforms should be assessed against the specific circuit they were built to break.

What 2012 established

• The doom loop can be entered from the banking end, so fiscal discipline is not sufficient protection — Spain's public finances were exemplary and irrelevant.
• Private capital-flow imbalances inside a currency union are the unmonitored risk, and an external deficit is a funding dependency rather than a benign flow.
• Institutional governance determines who absorbs losses, and institutions that cannot raise equity convert private losses into public debt by construction.
• A credible commitment can resolve a self-fulfilling funding crisis without expenditure, which distinguishes an equilibrium problem from a fundamentals problem definitively.
• Banking union attenuated the loop without severing it, leaving zero-risk-weighted sovereign exposure and national deposit insurance intact.