The ECB as a lender of last resort: the final delusion.

I will start this site with a text I wrote one year ago on ECB purchases of sovereign debt. This text was turned down by a major economics web site on the grounds that I am incompetent. Indeed, the crisis has triggered a new orthodoxy among the macroeconomic mainstream that (i) any idea of an inflation risk is ludicrous and (ii) it makes no sense to look at the quantity of money. My current view is that (i) we are getting the same inflation rate in a recession as we used to have during an expansion, and talks of deflation have not been validated by reality, and that (ii) the market has absorbed large injections of money because, given all the policy uncertainties, the demand for liquidity is very large. Remember,though, that liquidity is just an option to spend in the future. When people finally make up their minds and decide to allocate their liquidity to more productive uses, we may witness very sharp rises in goods and/or asset prices.

Anyway, here is the text.


Given the current difficulties facing the overindebted eurozone governments, an increasing number of experts  recommend that the ECB should simply purchase these liabilities. This sounds like a miracle cure, as markets would suddenly infer that there exists a buyer of last resort for these liabilities and would no longer have a reason to ask for a spread reflecting their perceived probability of default.

Alas, just as the German and French governments do not have a deep enough pocket to be able to commit to bail out all the troubled countries, and therefore are incapable of avoiding contagion, so  does the ECB have a limited ability to intervene.

Forget that such purchases are precluded by the Maastricht treaty (The European Union now has a long tradition of breaking its own rules).

The ECB has two kinds of assets:

-The financial claims it has acquired as a counterpart to the monetary base (i.e. the euros it is issuing).

-Its credibility, i.e. the belief that it is able and wants to achieve a low inflation rate in the Eurozone, which ensures that holding euros is a reasonable proposition.

To bail out Greece or Italy, the ECB must deplete its stock of at least one of those two assets.

Purchasing public debt while maintaining its credibility means that such purchases are not associated with a change in monetary policy.  This is what we call “sterilized intervention”. When acquiring Greek or Italian debt, the ECB does not issue Euros but instead sells some other asset (for example Gold or foreign reserves).  Unfortunately, the assets of the central banks are insufficient for any sizable bailout to take place. They are valued at some 2,200 bn EUR but a large part of it consists of loans to the banking sector that are necessary for it to get liquidity. At most roughly 1,000 bn EURs could perhaps be used, which is 3 times Greek debt but only 50% of Italian debt. This illustrates how the recent attack on Italy changes the equation completely:  It was once believed that bailing out Greece could eliminate contagion, but the countries that are next in line under the contagion scenario are too large, whether we consider national governments or the ECB; for this precise reason that scenario could not be eliminated even under a 100% Greek bailout. Note that at present the ECB only holds a minute 30 billion EUR in national debt.

An interesting question is: how much redistribution between citizens of the Eurozone is going on when the ECB purchases Greek or Italian debt? The question is relevant because the ECB is not democratically appointed. A non representative entity redistributing money between social groups is tantamount to “taxation without representation”. This was considered as a form of tyranny two centuries ago. If Greece and Italy were perfectly solvent, then selling French or German bonds to purchase Greek or Italian debt would be irrelevant. This is because the proceeds from those assets are rebated to the national governments according to a sharing rule based on population and GDP, in which there is no role for the geographical origins of the assets. However, if there is default, then it is associated with a loss for the other shareholders of the ECB, which means that implicitly the taxpayers of those countries have been forced to bailout the creditors of Greece and Italy with their own money. At the end of the day, there is no difference between ECB monetization of Greek debt and direct bailout by the taxpayer. In principle, if markets price the default risk correctly, debt purchases at market prices by the ECB or the German taxpayer can be “fair”, but in such a case they do nothing to alleviate the burden on the Greeks; instead they forcibly transfer ownership of debt from private to public creditors.

The alternative option would be to increase the monetary base and print the euros needed to finance the bailout. This would in effect deplete the other asset of the ECB, its credibility.

There seems to be a widely held opinion that since there is a recession and we are not far from a liquidity trap, people would just hold the extra euros and the money financing of the debt would not be inflationary.  That is, printing money would just afford us a free lunch, while all other exit options entail considerable pain. If so, then all the ongoing austerity packages must be really stupid, as is the current fight  between Euro politicians about who will bail out whom and by how much.

There are two typical  “free lunch” arguments.

Argument 1: The ECB’s balance sheet has expanded considerably during the crisis, and we observe no clear adverse consequence of that, and in particular no inflationary pressures.

The answer to that argument is that the ECB has increased the monetary base in response to the liquidity crisis in the banking sector, that is, there has been an exceptional  surge in the demand for central bank money, due to a failure of the banking system to continue creating liquidity through deposits and loan. Indeed, broader monetary aggregates like  M2 have not exploded during the crisis, instead they grow at a stable rate which is lower than before the crisis (see  In other words, the end user demand for liquidity has not increased; the ECB has just made up for the fall in the banking sector’s ability to generate liquidity. Thus it does not follow that an increase in the supply of central bank money for mere fiscal motives would be accommodated by the private sector.

Argument 2: During a crisis many assets are under-priced and the bail-out pays for itself when the securities that are purchased are eventually sold back at a profit. Thus, the US Fed and US government have made money in the bailouts of Bears Stern and General Motors. Similarly, the ECB could just purchase public debt and sell it back at a profit if and when one exits the crisis. After this operation is completed, the ECB’s balance sheet would be back to normal.

This argument is an ex-post facto rationalization. Surely it was not obvious ex-ante that the price would go up and that the bail out would be profitable. And there are plenty of examples of loss-making government investments. If governments were better at arbitrage than private markets, we would need more speculation by the governments and we could eventually get rid of taxes and fund public expenditures with the gains from such speculation. Why did not governments buy the Internet and housing bubbles and then sell those assets at the peak of those bubbles? Not only would it have been profitable but it would have stabilized asset prices. The answer is that no one knew for sure whether these were bubbles and where the asset prices were headed.

A more specific version of this argument is that during a financial crisis, some assets are clearly underpriced because of their liquidity premium – people do not want to buy them because they fear being unable to sell them back when needed. By using its deep pocket, the government could possibly inject liquidity into the market while making a capital gain (in some sense the government is acting as an insider trader by making a capital gain on the news about its own decision to inject liquidity, but that is for the common good).  The bail-out would then pay for itself, the pieces of papers issued in exchange for the bonds would be withdrawn in due time, much as if this speculative operation had been financed by leverage.

The question is whether the argument applies to Euro-zone sovereign debt and I doubt it. What markets are seeing is insolvency due to poor growth and demographic perspectives, on the one hand, and political stalemate in attempts to reduce public expenditure, on the other hand. For example, France has not been able to run a budget surplus since the early 1970s. The austerity cure of the Barre government between 1978 and 1981 led to a quick reduction in deficits despite the country’s dismal economic performance, but the government lost the subsequent election, not least because of the bitter pill administered by Barre.

Yet another point is that attacks on sovereign debt are self-fulfilling: as expectations of default occur, interest costs go up, which in turn accelerates the growth in the stock of debt, thus leading to inevitable default. If only people believed that the debt would be repaid no matter what, for example by ECB monetization, then spreads would never go up in the first place, and the crisis would be avoided. For this argument to be valid, it must be that holders of public debt would make no losses under the attack scenario, and therefore have no incentive to participate in the attack.  Given the sheer amount of money that would have to be created, I doubt it entirely. Rather, the Eurozone would join a club of emerging, unstable countries (such as many Latin American ones) where money printing is a substantial source of fiscal revenues and, as people will eventually want to get rid of their excess money balances to purchase goods and services, the price level will, in the long run, rise in line with the rate of money growth. This means that the Euro would depreciate in the short run as people would try to exchange their Euros against sounder currencies. Initial holders of public debt denominated in Euros would lose money, not because of default, but because of such depreciation. Hence the commitment to monetization would fail to kill the attack scenario.

Currently the monetary base is around 1400 billion Euros, i.e some 14% of Euro area GDP. To bail out say half of Italian debt plus the entire Greek debt (which does nothing for Portugal, Ireland and Spain,), this amount would have to double. It is hard to believe that private agents would not get rid of this liquidity as fast as they can.

But the worst consequence is that we would be in a new policy regime where each Euro area country will have an interest in running large deficits, in order to increase its share of the seignoriage generated by debt monetization, at the expense of taxpayers in other countries. A while ago Sargent and Wallace argued, in their “unpleasant monetarist arithmetics” paper, that a tight monetary policy without budgetary discipline would eventually be defeated, as the central bank would be eventually compelled to accommodate budget deficits with monetary policy. This is exactly the scenario that may happen right now in the Eurozone, only that it will be made worse by the fiscal race to the bottom of national governments trying to beat each other to the ECB booty.

How much inflation can we expect? Suppose public debt purchases by the central bank lead us to a policy regime where debt financing for many governments in the area has become too costly, in particular because the bailouts have left them with no incentives to set their accounts straight (and it would be foolish to do so under such a policy regime). Then, in addition to the initial bailouts, the central bank will have to finance the budget deficits of the Euro area countries. Suppose, say, that  these deficits are equal to 3% of the euro area GDP. With a monetary base of 14%, this means that to purchase the flow of new debt issued by the governments, the ECB would have to increase the monetary base by 3%/14% = 21 %. Even allowing for real growth (and nobody expects much of it in the coming decade) at a generous 3% rate, this means an 18% inflation rate in the long run[1].  Even ignoring the chances that this could degenerate into hyperinflation, the Euro will be dead before that point is reached.

During the accession phase to the Euro, economists were seriously debating its costs and benefits. Some of them pointed out that member countries diverged notably in their ability to raise taxes, and for that reason some countries like Greece and Italy needed inflation to finance their public expenditures, while at the same time others like Germany insisted that the ECB mandate should be to insure a stable price level. This conflict of interest this is proving to be a fatal flaw in the design of the EMU.

[1] This may be less because of the secular growth in the demand for liquidity, as evidenced by the fact that broad aggregates grow faster than nominal GDP. However we do not know that this trend will continue, as there is no reason why it should.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s