At one point last month, BES subordinated bonds sank to below 70% of face value and its senior debt to 95% of face value, while yields on Portuguese government bonds rose from their June low of 3.35% to 4%.
Investors worried that for a country with net external debt at the end of 2013 equivalent to 119% of GDP – among the highest in Moody’s sovereign ratings universe – even one still with €6.4 billion of undrawn bank restructuring funds, it would not be good for these to be mostly used up just weeks after Portugal emerged from its bailout programme.
|Bank of Portugal governor |
But there’s another way to look at it: how exposed are European banks still to potential losses from large holdings of European government bonds?
After the financial crisis, regulators required banks to build up holdings of government bonds as liquidity buffers that could quickly be turned into cash if otherwise well-run banks should suddenly be cut off from short-term market funding during a systemic crisis. This was an entirely reasonable response to the way in which the drying up of previously readily available market funding spread contagion through banking systems in 2008. Unfortunately, it is built on two highly questionable assumptions: that government bonds are liquid and that they are high quality.
The extremity and suddenness of movements in both BES cash bonds and even Portuguese government bonds demonstrates how far displayed prices can move up and down on very low volumes of trade in increasingly illiquid secondary markets. And given the recent history both of bailouts of European sovereigns and private sector involvement in debt restructuring, the notion that government bonds are risk free is clearly laughable.
A recent report from Barclays shows that European banks’ holdings of government bonds have risen from €1.2 trillion on the eve of the financial crisis in June 2008 to €2 trillion today. As a percentage of their total assets, government bonds have grown from under 5% in 2008 to over 10% today for Italian banks, from around 3% to 10% for Spanish banks and from under 1% to nearly 8% for Portuguese banks.
It is a glaring weakness in the current system of bank regulation that the Basel Committee’s standardized approach allows banks to apply zero risk weights to sovereign bonds that the same banks’ own sophisticated risk models would require more realistic economic capital weightings against.
The sovereign-to-bank feedback loop has worked entirely to the advantage of sovereigns, banks and policymakers in the last two years. As international investors withdrew from periphery sovereign bonds after 2011, domestic banks stepped in as the marginal buyers, so capping rises on government yields. Growing optimism that the ECB would, if pressed, do whatever it took to protect the euro from break up including buying government bonds, drove yields down further this year, allowing banks to profit from current income and capital gains on bonds bought when they yielded over 6% that are now changing hands at 3.5%.
But the question has become pressing: how much longer can allowing these exposures to be zero risk weighted possibly be commensurate with sound risk management practice, especially given that, for many peripheral country banks, increasing exposures tend to be so heavily concentrated in home country government bonds?
Note that KBC’s first quarter results published in May disclosed that risk weighted assets had been affected by the National Bank of Belgium’s request to remove the possibility of applying a zero weight to domestic sovereign exposures of Belgium, the Czech Republic, Slovakia and Hungary. This may be a sign of big and contentious changes to come. It remains to be seen how the Basel Committee will react.
In a speech earlier this year on the leverage ratio, Stefan Ingves, chairman of the Basel Committee on Banking Supervision, acknowledged that “crises occur when enough people convince themselves that something is ‘low-risk’, only to learn afterwards that they had underestimated the risk involved”.
On that, we can all agree.