The anatomy of the eurozone bank run
May 20, 2012 4:21 pm
by Gavyn Davies
A bank run is now happening within the eurozone. So far it has been relatively slow and prolonged, but it is a run nonetheless. And last week, it showed signs of accelerating sharply, in a way which demands an urgent response from policy-makers.
.
The fear of bank runs is deeply ingrained in all economists who know anything about the genesis of the Great Depression in the United States in the early 1930s. Then, the failure of the Bank of United States in December 1930 led to multiple bank runs across the country. Bank failures in the following two years wiped out personal savings and greatly exacerbated the collapse of demand in the economy.
The classic account of the crisis, by Milton Friedman and Anna Schwartz, concluded that the collapse was largely the fault of the Federal Reserve, which failed to provide enough liquidity to keep the banks functioning and thus end the panic. After the crash, the establishment of the Federal Deposit Insurance Corporation was intended to ensure that deposit holders never again had to live in fear that their savings would be in jeopardy. What are the lessons for the eurozone?
.
The risk of bank runs is an inevitable feature of any banking system which takes short term deposits from lenders, and then makes longer term loans to borrowers. This “maturity transformation” is arguably the key function of a bank, but it brings with it the risk that depositors will ask for their money in large numbers at a time when the bank does not have the liquid resources to meet these demands.
.
This can cause a rational stampede to withdraw money, since depositors who wait until late in the day can lose all of their money. The pioneering work of Diamond and Dybvig in 1983 showed that banking systems are characterised by multiple equilibria. They demonstrated that it is possible for the system to shift without much warning from a good, stable equilibrium to a situation in which it is rational for deposit holders to demand their money immediately.
.
In some ways, the withdrawal of deposits from banks in the periphery of the eurozone is simply a slow motion version of a rational bank run in the Diamond/Dybvig tradition. For example, deposits in Greek banks have fallen by about a third since the beginning of 2010. Deposits in Irish and more recently Spanish banks have also been falling.
.
Depositors have been shifting their money to “safer” banking systems, notably those in Germany. The extent of these declines and other financing difficulties for the banks is illustrated by the rise of Target 2 imbalances contained within the ECB’s balance sheet, which reflect the mechanism through which these cross-border flows have been financed:
.
.
The standard central bank response to this problem is to provide enough liquidity to solvent banks to ensure that deposit holders are always able to withdraw their money, in which case the panic is eventually supposed to subside. The ECB has done this to the full extent required; in this respect it is not possible to criticise the ECB for failing to fulfill the role of lender of last resort to the entire system. Yet the continuing drain on deposits has not been halted.
The reason for this, of course, is the fact that the eurozone is not a single nation state, even though it does have a single central bank. This has two consequences. First, the underlying fear of depositors in the periphery is not simply, or even mainly, one of bank failure. Instead, they probably fear the devaluation of their deposits relative to those in core economies if the euro should break up.
Therefore, the run is being caused by concerns about exchange rate risk, not necessarily by the fear of bank failure as such. This makes it much more complicated to deal with, since it is very difficult to offer guarantees against future exchange rate losses to today’s depositors. Germany would not want to stand behind such guarantees to Greek and Spanish citizens in the event of a euro break-up.
Second, the bank run is greatly increasing the scale of potential transfers between nation states which until recently have been disguised within the ECB balance sheet. As deposits are withdrawn from Greek banks, the ECB replaces these deposits with liquidity operations. If these are standard repo operations, such as those undertaken in the LTROs in December and February, then the ECB is directly assuming risks which the Greek private deposit holder is no longer willing to hold. If the liquidity is injected via Emergency Lending Assistance, then the Bank of Greece is theoretically assuming the risk, rather than the ECB as a whole. But in the event of a euro break-up, these losses would ultimately fall on the ECB itself.
The problem is that this potentially exposes the ECB to much bigger losses than anything which has been contemplated so far by the core economies.
.
.
Up to now, the ECB has been willing to inject liquidity to cover the financing needs of the periphery banks as the inter-bank market has dried up. If instead, they have to contemplate providing semi-permanent funds to cover large further withdrawals of bank deposits, the size and timescale of the injection becomes extraordinarily large.
As the table (from David Mackie of J.P. Morgan) shows, the outstanding bank deposits of the periphery are many times larger than the current exposures of the ECB and core governments to the periphery. It seems inconceivable that core countries like Germany will be willing to expose themselves to these risks if the deposit flight continues.
.
Mario Monti apparently took a plan to the G8 summit to offer jointly-funded guarantees on bank deposits to apply across the entire eurozone. This would certainly help, but whether it would be sufficient to eliminate fears of exchange rate losses if the euro were to disintegrate is another matter. To fix that problem, belief in the integrity of the euro as a single currency needs to be restored. The bank run could bring matters to a head.
0 comments:
Publicar un comentario