Are Europe's banks safe?

"Banco Esprito Santo - bankruptcy rumors in Portugal"; "New banking scandal in Bulgaria". These two headlines from the past few weeks have reopened the question of whether Europe's banking system is safe. Wherever a bank gets into trouble in Europe, fears of a renewed financial crisis reawaken.

After depositors lost confidence in Bulgaria's Corporate Commercial Bank (Corpbank) in June on the heels of rumors of financial irregularities, the government first announced it would close the bank. Then it reversed course, and decided to save Corpbank. It went to the financial markets to borrow the money needed to fix the bank's balance sheet - about 1.4 billion euros.

"You cannot rule out more such cases - on the contrary," says Andreas Dombret, a member of the executive board of Germany's national central bank, the Bundesbank, one of the European Central Bank's eighteen member institutions.

If a country like Bulgaria, the poorest in the European Union, takes on new debts of this magnitude, it affects the country's entire economy. In the wake of the Corpbank bailout, Bulgaria had to revise its deficit target sharply upward, from 1.8 per cent prior to the bailout to 2.7 percent of the country's gross domestic product (GDP).

Tense atmosphere

In the current financial climate, with interest rates at historically low levels in the US and EU, the banking sector is particularly crisis-prone, according to Dombret: "This combination of low interest rates, low inflation, low volatility and high liquidity means that the search for yield is on again in the financial markets."

In such a heated environment, criminal offenses like the control fraud suspected in the Corpbank example, or mere management mistakes, can be enough to topple a bank into insolvency.

In an environment of thin margins, even a rumor that a bank may be illiquid can topple a bank by setting off a 'bank run'. In a bank run, many savers simultaneously try to withdraw their deposits from a bank, out of fear that the bank may be insolvent and about to close its doors.

When a bank is asked to pay out far more on a given day than it takes in, and cannot immediately raise the funds to do so - for example, by borrowing funds from other banks on interbank money markets - then the fear that the bank will close its doors can become a self-fulfilling prophecy.

Moreover, banks have interlinked balance sheets, and that means trouble at one bank can lead to trouble in others. According to economist Thomas Hartmann-Wendels at the University of Cologne, banks are always at risk of contagion. When one bank starts to wobble, confidence in other banks is often shaken as well. Often banks have made loans to each other, and get into difficulties through that channel. These are systemic effects that we generally don't see in the non-financial economy.

Hopes are pinned on the coming eurozone banking union

In an effort to calm the current turmoil in its banking sector, Bulgaria has decided to join the new European banking union - the instrument European finance ministers hope will enable them to deal more effectively with troubled banks.

A core aim of the banking union is to limit the exposure of taxpayers when banks go under. One component of the new process is the Single Resolution Mechanism (SRM), a rescue fund into which financial institutions will be required to pay contributions totaling 55 billion euros ($76 billion). The SRM would have the authority to borrow additional funds beyond that amount in an emergency.

The fund will be available to recapitalize banks if their losses exceed what their shareholders and creditors can absorb - especially banks whose failure would pose a systemic risk.

However, critics have panned the SRM as inadequate. Some of the biggest euro area banks have assets exceeding 1 trillion euros, and if only a few percent of those assets - such as loans they've made, or bonds they've bought - turn out to be bad, the resources of the SRM could quickly be overwhelmed.

This would be more likely if there were systemic reasons why assets went sour, such as a sudden rise in interest rates that leave large numbers of borrowers unable to refinance outstanding loans, or a risk of sovereign default by a country whose bonds are held by many banks.

It could prove difficult for the SRM to borrow sufficient money from private investors in a crisis. Rescuing one of the big European banks - or a number of smaller banks at the same time - could require a lot more money than the SRM can readily provide.

That would throw the problem back to taxpayers - or leave governments to deal with the consequences of letting a major bank declare insolvency and close its doors. Since the failure of Lehman Brothers, that has been a scary prospect.

Before the banking union is launched, authorities will carry out a balance-sheet analysis and "stress tests" on 128 of Europe's biggest banks. The Bundesbank's Andreas Dombret says this process will "ensure that old bad debts and vulnerabilities in banks' balance sheets are resolved before the European banking union begins its work in November".

Stress tests are scenario analyses in which examiners test the resilience, liquidity and solvency of a bank's balance sheet - for example, an examination of what might happen to a bank's solvency if there were a sudden downturn in housing prices in its domestic market, coupled to a surge in interest rates.

Are stress tests reliable indicators of bank's health and resilience?

However, Dombret is careful to say that even if a bank passes a "stress test", it could still get in trouble. There is always the question of what the examiners are looking for, he says, and makes an analogy with medical diagnosis: "If you suspect someone has a viral disease and you're trying to identify the virus, then you cannot test for every virus, because there are far too many different kinds. So you focus only on the most likely."

Economist Hartmann-Wendels also cautions against "over-estimating" stress tests. In individual cases, the results are not very meaningful: "There are a lot of accountants going around taking a magnifying glass to banks' balance sheets and trying to leave no stone unturned."

In the end, however, Hartmann-Wendels says the flood of data is too big to handle: "Then time constraints force you to fall back on a set of standard assumptions, and the situation of an individual bank can only be roughly assessed."

Who is really "too big to fail"?

Must banks necessarily be saved? Andreas Dombret points out that bank managers are part of our market economy, and "among the basic principles of a market economy is a principle that when companies and banks have no compelling business model, they drop out of the market."

Therefore, the Bundesbank's executive board is "very much of the opinion that banks, like other companies, must also be able to fail."

But before banks can be allowed to fail, it's necessary to define clearly which banks, under what circumstances, are "too big to fail" - in particular, which banks are too big and important to the system as a whole to be allowed to fail, in view of contagion risks resulting from mutual obligations with other financial institutions.

Dombret is still waiting for a "really convincing solution to this question". In any event, he says, it cannot be permitted that we remain "dependent on a few big, internationally networked banks that have the power to blackmail society, simply because they're so big".

And what will happen under the new European banking union if a bank that continues to be considered "too big to fail" does in fact become insolvent?

Dombret points to the hierarchy of liabilities set out in the banking union agreement. In case of an insolvency, losses are borne "first by the owners of the bank, then its creditors, then by the investment funds fed by the bank, and only at the end, if it remains necessary, by the taxpayer".

But it's not clear how quickly the end of this liability chain would be reached in a given case. Economist Hartmann-Wendels, at any rate, is skeptical: "I think the State will never be able to wholly escape losses and liabilities if a massive crisis occurs that affects several banks."