Europe Investment Report 2011 — The Central Bank Became the Interbank Market

By 2011 the euro-area interbank market had substantially stopped functioning across borders, and the reason was location rather than credit.

A bank in a stressed member state found that:

• Other banks would not lend to it unsecured, regardless of its own capital position.
• Its wholesale funding was unavailable or prohibitively expensive.
• Its deposits were leaving, moving to banks in member states perceived as safer — the redenomination flight the Global Investment Outlook 2011 describes.
• And its access to markets was priced off its sovereign, because of the guarantee and the sovereign bonds on its balance sheet.

Meanwhile banks in core member states had surplus deposits and nowhere to put them, since they would not lend across the border.

This is a market failure with a specific shape: funds existed and demand existed, and no transaction occurred because the intermediating market had stopped.

The central bank stepped into the middle, and the mechanism is worth being precise about because it was widely misdescribed:

• Banks needing funding borrowed from the central banking system, against eligible collateral.
• Banks with surplus funds deposited them with the central banking system.
• The central bank was therefore intermediating what banks had previously done directly.

The cross-border balances that accumulated within the central banking system were the accounting record of this substitution. They rose because the private market had stopped, and they would have risen in exactly this way regardless of anyone's intentions.

The second theme is that collateral rules became the operative policy instrument. When funding is available only from the central bank, the list of assets it will accept determines what banks can fund — and therefore what they will hold. This is a powerful allocative lever that operates without any interest rate decision.

Fragmentation by location

The breakdown deserves setting out because it reversed the single market's central premise.

In an integrated market, a bank's funding cost reflects its own credit quality. Two equally sound banks in different member states pay the same rate, which is what monetary union is supposed to deliver.

By 2011 this had inverted. A bank's funding cost was determined principally by where it was.

The channels connecting a bank to its sovereign:

Sovereign bond holdings. Banks held large quantities of their own government's debt, encouraged by the zero risk weight described in the Europe Investment Report 2009. A falling sovereign damages bank capital directly.

The guarantee. A bank's deposits are guaranteed by its home state, so the guarantee's value depends on that state's fiscal capacity — the national safety net problem from the Europe Investment Report 2008.

Collateral. Banks funded themselves against sovereign bonds. When those bonds fell in value or lost eligibility, the funding capacity fell with them.

And the economy. A stressed sovereign's economy is contracting, so the bank's domestic loan book deteriorates.

All four operate simultaneously and in the same direction.

A bank in a stressed member state was four separate exposures to its sovereign wearing one balance sheet. There was no way to be a sound bank in a distressed country, which is what made the fragmentation impossible to escape through good management.

The consequence for the real economy was that identical businesses faced materially different borrowing costs depending on which member state they were in — which is a broken monetary union, since the whole purpose is a single monetary condition. The Europe Investment Report 2014 covers this at its most acute.

What the cross-border balances actually were

The balances that accumulated between national central banks became the most misinterpreted statistic of the crisis, and the mechanism is genuinely simple.

How the system works normally:

• A payment from a bank in one member state to a bank in another moves through the central banking system.
• The receiving country's central bank records a claim; the sending country's records a liability.
• In normal conditions these net out, because payments flow both ways and because banks recycle funds through the interbank market.

What changed:

• Payments flowed persistently in one direction — deposits and funding moving from stressed to core member states.
• The interbank market did not recycle them back, because core banks would not lend to stressed ones.
• So the balances accumulated rather than netting.

What they represent:

• They are the mechanical consequence of one-way payment flows without private recycling.
• They are not a transfer, a loan decision, or a subsidy — no one chose their size.
• They are a measure of the private market's failure, not a policy.

The balances were a thermometer being blamed for the fever. They rose because banks stopped lending to each other across borders, and they would have fallen the moment that resumed — which is eventually what happened.

The legitimate concern underneath the misinterpretation is real, and worth stating fairly: the balances represent exposure if a member state were to leave the currency union. In that event, the claims would be against a national central bank in a different currency. This is the redenomination exposure from the Global Investment Outlook 2011, appearing in the central banking system's own accounts.

So both readings have something to them. The balances were not a policy choice, and they did represent an exposure that only materialises in a scenario the policy existed to prevent.

The practical value is as an indicator: rising balances mean the private cross-border market is not functioning, and they are published monthly and free by the ECB and by national central banks. They remain one of the better available measures of euro-area financial integration.

Collateral rules as policy

When central bank funding is the only funding, the rules governing it become the binding constraint on bank behaviour — and this is an underappreciated policy lever.

How central bank lending works: banks borrow against collateral, from a defined list of eligible assets, each subject to a valuation haircut.

Three parameters, each a policy instrument:

Eligibility. Which assets qualify at all. An asset that is not eligible cannot be used to raise funding, so it is far less useful to a bank that may need liquidity.

Haircuts. How much less than face value the collateral is credited. A larger haircut means less funding per unit of collateral, which reduces the asset's funding value without excluding it.

And rating thresholds. Minimum credit ratings for eligibility, which link the central bank's collateral policy to the judgments of rating agencies — an uncomfortable dependency that became acute as programme countries were downgraded.

Why these are powerful:

• They determine what banks hold. An asset that can be pledged for funding is worth more to a bank than an identical asset that cannot.
• They can be adjusted quickly, without a rate decision.
• And they are quasi-fiscal. Accepting weaker collateral at a modest haircut transfers risk to the central bank, which is a fiscal act conducted through a monetary instrument.

The bind this created in 2011: the collateral banks held most of was their own sovereign's debt, and that debt was being downgraded. Applying the rules mechanically would have removed the collateral of exactly the banks most dependent on central bank funding — withdrawing liquidity from the system at the worst moment.

The responses — waiving rating thresholds for programme countries, broadening eligible collateral, and accepting national credit claims — kept the funding flowing and were necessary. They also moved the central bank further into credit allocation, which is the tension the Europe Investment Report 2014 examines.

Cheap funding, domestic bonds

The long-term refinancing operations of this period had an effect that was predictable, arguably intended, and structurally problematic.

The design: very large amounts of central bank funding, at a low fixed rate, for an unusually long term — three years rather than the usual weeks or months.

What banks did with it:

• Refinanced maturing wholesale funding, which was the primary purpose and which worked.
• And a substantial portion was used to buy domestic sovereign bonds.

Why buying sovereign bonds was rational for the banks:

• The carry was large. Borrow at a low fixed rate for three years, buy a domestic sovereign bond yielding far more. The spread was substantial and the funding was locked.
• The bonds were eligible collateral, so they could be pledged back for further funding.
• The risk weight was zero, so no capital was required.
• And there was official encouragement, explicit or otherwise, to support domestic sovereign markets.

The immediate effect was positive: sovereign yields fell, the bank funding crisis eased, and the acute phase passed.

The structural effect was to deepen the doom loop:

Banks used central bank funding to increase their holdings of the sovereign debt that was the source of their fragility. The operation relieved the funding crisis and increased the concentration that had caused it.

This was understood at the time and accepted as a trade — an acute funding crisis is more urgent than a structural concentration. The concentration remained, and the Europe Investment Report 2016 covers it as an unresolved feature of the system.

The broader point about zero risk weights deserves restating: a capital framework that treats domestic sovereign debt as riskless makes this trade free in capital terms. The concentration is not a failure of bank management; it is the response the rules encourage.

Two ways to hit a capital ratio

A supervisory decision in this period produced a result opposite to its intent, and the mechanism is worth understanding because the same instrument is still used.

The requirement: major banks were told to reach a specified core capital ratio by a set date.

The intent was to strengthen the system and restore confidence, which is reasonable.

A capital ratio is capital divided by risk-weighted assets, so there are two ways to raise it:

Raise the numerator — issue equity, retain earnings, convert instruments. This strengthens the bank and leaves its lending capacity intact. It is what the requirement intended.

Shrink the denominator — sell assets, let loans run off without replacing them, reduce exposures with high risk weights. This also raises the ratio and reduces lending.

Why banks chose the second:

• Issuing equity at prices well below book value is enormously dilutive to existing shareholders, and management is accountable to them.
• The equity was frequently not available at any price for the weakest institutions.
• Accepting state capital carried stigma and conditions, per the mandatory-versus-voluntary logic in the UK Investment Report 2008 — and the scheme was not universal, so accepting it identified the bank as weak.
• And deleveraging was fast, controllable and entirely within management's discretion.

The aggregate result:

A requirement designed to strengthen banks was met substantially by shrinking them. The ratios improved, the system's lending capacity fell, and the credit contraction fed the recession that was generating the losses the capital was meant to absorb.

The deleveraging was also selective in a damaging way. Banks shed the assets that were easiest to sell and carried the highest risk weights — which meant foreign exposures, trade finance and lending to smaller borrowers, rather than the domestic sovereign debt with its zero risk weight. The capital rules determined what got cut, and they pointed away from the concentration that was the actual problem.

The design lesson, which informed later exercises including the 2014 review in the Europe Investment Report 2014, is that a ratio target must be paired with a capital-raising mechanism and ideally an absolute capital floor — otherwise the cheapest route to compliance is the one that damages the economy.

What an allocator could act on

Assess a bank in a stressed sovereign as four exposures to that sovereign. Bond holdings, guarantee value, collateral capacity and loan book quality all move together, so good management cannot escape the location.

Read cross-border central bank balances as an integration indicator. They measure whether the private cross-border market is functioning, they are free and monthly, and they are not a policy or a transfer.

Track collateral eligibility and haircut rules as policy. They determine what banks can fund and therefore what they hold, and they change without any rate decision.

Watch what banks do with term funding, not just that they receive it. Cheap locked funding plus a zero risk weight on domestic sovereign debt makes the carry trade free in capital terms, which is a predictable response rather than a surprise.

Note where a monetary union has stopped being one. Identical businesses facing different borrowing costs by member state is the observable failure, and lending rate dispersion is published by the ECB.

Distinguish a crisis relieved from a concentration removed. The 2011 operations did the first and worsened the second, and both were understood at the time.

What 2011 established

• Interbank fragmentation made funding cost a function of location, not balance sheet.
• The central bank substituted for the interbank market, and cross-border balances were the accounting record of that.
• Collateral rules became the operative instrument, determining fundability and therefore bank holdings.
• Term funding was recycled into domestic sovereign debt, relieving the funding crisis and deepening the concentration.
• Bank and sovereign crises in a currency union are one crisis, connected through four simultaneous channels.