Quantitative easing in euro zone requires shared risk
After much hesitation and amid intense controversy, the European Central Bank is stumbling towards a program of sovereign bond purchases. This monetization of government securities does not come without tail risks, since euro-area member states can default on their public debt, as the Greek experience has shown. In order to immunize the shareholders of the ECB, namely the national central banks, against potential losses from such an operation, a simple accounting technique has been proposed: purchases would not be undertaken jointly, but rather would be made on the accounts of individual national central banks of the euro zone. The Bundesbank would buy Bunds, the Banca d’Italia BTPs, the Bank of Greece would purchase Greek government bonds – and each institution would keep these assets on its own balance sheet, and, crucially, at its own risk. Proponents of this solution suggest that it would allow all shareholders of the ECB to enjoy the benefits of quantitative easing (QE) without having to assume the inherent risks.
Unfortunately, this belief in risk-free joy is fundamentally flawed and offers false comfort. Leaving government bond purchases on the balance sheets of national central banks would not prevent cross-country loss-sharing in the event of another sovereign default in the euro area. On the contrary, it could even increase the probability that such an accident would happen again.
Normal monetary policy operations in the euro area are mainly carried out by national central banks, but all income generated and any credit losses incurred from these operations are distributed amongst ECB shareholders according to the capital key. Thus, the 2008 default of Bankhaus Lehman Brothers could have created losses for all national central banks in the euro area, although this bank had carried out its refinancing transactions exclusively with the German Bundesbank. A similar problem might arise if QE were to operate as a “normal” monetary policy operation. If a member state defaults on securities held by any member of the eurozone, all participants would share losses on a pro rata basis.
In the absence of formal risk-sharing of sovereign bond purchases, a member state’s default on its government bonds would at first sight impact only its own national central bank and likely bankrupt it. However, this would have little practical implications, since central banks can easily operate with negative equity. A number of central banks, including the Czech National Bank, have for years carried out their tasks successfully while having a negative capital position. The only consequence of the national central bank’s negative capital position might be a temporary stop of dividend payments to the government, but even this can be easily avoided as a government can always recapitalize its own central bank with a perpetual zero coupon bond.
Much more relevant is the fact that the liquidity created by the national central bank through its government bond purchases would remain in the system after the default. Euros will have been printed to pay for the purchase of government bonds and will continue to affect the price level in the eurozone as a whole. At present, this is the desired effect of quantitative easing. However, when economic conditions normalize again, the national central bank whose government defaulted will find it difficult to do its share in mopping up excess liquidity by selling securities, since its holdings of defaulted government bonds will have lost their value. The rest of the euro zone would then have to undertake larger sterilization operations in order to keep money supply under control. This would imply lower seigniorage income for all other member states as well. Losses emanating from government bond purchases would thus end up being shared anyway.
Demands to limit cross-border risk-sharing by not carrying out the ECB’s sovereign bond purchases as a joint monetary policy program are thus misguided. They fall victim to the analytical error of focusing exclusively on the central bank’s asset holdings, while price dynamics as well as decisions on costly sterilization policies are driven by the monetary liabilities of the central bank. Furthermore, turning national central banks into large-scale holders of their own countries’ government bonds could actually increase the likelihood of another payment default in the euro area. Under this setting, the national government as main beneficiary of any debt reduction would be in a position to appoint the leadership of the institution that would decide on this debt forgiveness, i.e. the board of the national central bank. Over the past years, the vicious circle between commercial banks and sovereigns has done much to destabilize the euro area. It would be extremely unfortunate if another diabolical loop were created between national central banks and their own governments. The spirit of the European Treaties is pointing exactly into the opposite direction, namely towards a separation of monetary policy from the problems that national fiscal policy might create.
It has also been suggested that a government might more easily be tempted to default on its debt under explicit risk-sharing because the resulting losses would be borne by a foreign entity. This argument is misguided as well. If the sovereign bond purchase program entails full risk-sharing, then the cost of an individual member state’s decision to default on government bonds held across the euro zone would very likely be a forced exit from the euro area. By contrast, a member state would probably be able to default on government bonds held exclusively by its own national central bank while keeping the euro as a currency. Quantitative easing without risk-sharing would thus increase the danger of strategic sovereign defaults.