ECB liquidity is not a free lunch
March 4, 2012 12:20 pm
by Gavyn Davies
The initials LTRO, barely ever discussed prior to last December, now form the most revered acronym in the financial markets. Before the first of the ECB’s two Longer Term Refinancing Operations in December, global equity markets lived in fear of widespread bankruptcies in the eurozone financial sector. Since LTRO I was completed on December 21, equities have not only become far less volatile, but have also risen by 11 per cent.
With LTRO II completed last week, over €1tn of liquidity has been injected into the eurozone’s financial system. Private banks were permitted to bid for any amount of liquidity they wanted, the collateral required was defined in the most liberal possible way, and the loans will not fall due for three years. Any bank that might need funds before 2015 should have participated to the hilt, thus eliminating bankruptcy risk for a long time to come.
What can there possibly be not to like about this? A few things. Some observers point to the danger of a zombie banking system, kept alive artificially as a wing of the central bank. And, in a much-publicised “private” letter to Mario Draghi in February, Jens Weidmann, Bundesbank president, expressed concerns that the latest two LTROs will expose the ECB to potential losses which will undermine its capital base.
Of course, this can only happen in dire circumstances, under which banks borrowing from the ECB or the national central banks go bust, and the collateral held by central banks proves insufficient to cover the associated debts. Even in the case of Greece, this has not yet happened, because the ECB has been fully protected from losses on Greek government bonds up to now. But it is fairly easy to imagine that it could happen, especially if there were a generalised collapse in the euro, involving sovereign defaults in several of the indebted economies at the same time.
It is now widely recognised that a central bank cannot become insolvent in the same way that a private company can. Even if it incurs losses on its assets which more than completely eliminate its equity, it can never find itself in a position where it is unable to settle its debts, at least in its own domestic currency. Most of the liabilities of a central bank come in one of two forms: banknotes, and commercial banks’ deposits at the central bank. It is impossible for the private sector to force the central bank to exchange these assets for any other asset (like gold, for instance), and in any event the central bank can create more of each of them at will. Hence it can never become illiquid.
Admittedly, a central bank can encounter a negative equity (ie insolvent) position by taking write-downs on its assets, and it cannot eliminate this simply by creating central bank money. This is because printing money increases its liabilities as well as its assets, thus leaving net capital unchanged.
But over time it would expect the interest earned on its assets (eg government bonds) to be much higher than that paid on its liabilities (eg banknotes), so an increase in the size of the balance sheet would normally be expected to generate extra profits for the central bank. It could take some time, but these profits, which are called seigniorage, will in the long run compensate for any losses made by the ECB on its LTROs.
So why worry? One reason is that the ECB’s future seigniorage revenue should be seen as an asset of the governments of the member states. Therefore if it is used up in LTRO losses, the long-term income of the member states from the ECB will be reduced, and in effect their net government debt will be increased.
Seigniorage revenue is an asset of the governments, like future tax revenues. A decision to put the future seigniorage revenue at risk is in fact exactly equivalent to putting at risk future tax revenues by making inter-government loans which might incur subsequent defaults. Jens Weidmann is correct to point this out, though Mrs Merkel seems to be much more favourably disposed to these heavily disguised, quasi fiscal transfers than he is.
The second reason to be concerned is the potential risk to inflation. There is a long run safety limit to seigniorage, which is determined by the amount of central bank money which the private sector will be willing to hold if inflation remains at its 2 per cent target. Recent estimates by Willem Buiter at Citigroup and Huw Pill at Goldman Sachs both suggest that the net present value of this asset is over €2tn, many times bigger than the official capital reserves of the ECB of only €80bn.
These calculations have led many economists to conclude that the ECB could lose more than E2tn on its LTRO and other lending operations before it risks creating inflation through excessive creation of central bank money. Since this would imply a much greater loss than anything that appears likely at present, the LTRO operations have been widely deemed to be “non inflationary”.
This conclusion is not axiomatically true. If the central bank brings forward its future seigniorage revenue by creating a lot of money now, it seems clear that it could cause inflation to rise in the short term while still remaining well within its long-term E2tn limit for central bank money creation. This is analogous to a situation in which the government creates inflation by running a large budget deficit in the short term, while remaining solvent in the long term.
I am not saying that the ECB has already done this. It has not. But I am saying that it must be careful not to do this in the future. The LTROs were the right thing to do in the difficult circumstances of the time, but they are not a free lunch.
0 comments:
Publicar un comentario