The Irish Economy

Some media personalities and political pundits have over-stated the case for placing the blame for Ireland’s current economic mess on the government’s recent policy decisions.  The two big recent decisions of the Irish government (insured deposits and new bank capital) appear to me quite defensible.  If I may bring some controversy to the blog, here are three statements making the media rounds that I think are false:

1.  The primary cause of the Irish economic crisis was bad decisions by Irish policymakers and banks.

The statement above is false since the primary cause of the Irish economic crisis is the US-generated global credit crisis. This credit crisis in turn was caused by disastrously bad decisions by US policymakers (Congress, the quasi-state agency Fannie Mae, the SEC and Federal Reserve) and the U.S. finance industry (mortgage originators, ratings agencies, investment and commercial banks).  The too-weak oversight by the Irish central bank and financial regulator left the Irish bank sector very vulnerable to an external shock, as did the Irish government through its tax-and-spend policies, but these are both secondary not primary causes of the economic crisis.  It is a counterfactual and impossible to scientifically test, but I speculate that in the absence of the U.S. credit crisis, the Irish economy would have experience a somewhat bumpy “soft landing” from its 2002-2005 excesses. Without the US-generated crash, the 2009 situation would have been nothing like what we are in.

2.  The government decision on September 30^th to insure all bank deposits was obviously foolhardy and inconsistent with careful economic analysis.

This statement is false since it might possibly have been foolhardy but that is not obvious. As Brunnermeier notes in his recent JEP article, the 2007-8 credit crisis brought a new type of bank run, what he calls an “investment bank run.”  This is a bank run between institutions, where banks lose trust in one another and the relatively strong banks attempt to cut all credit ties to the weaker ones. As the Lehmann Brothers disaster shows, this can be a very destructive type of bank run due to the complex interlocking relationships between large banks.  Going back to the classic Diamond-Dybvig model, a bank run can be stopped by the monetary authority providing or promising monetized liquidity to all depositors.  As soon as the depositors realize that this monetized liquidity is available, the bank run ends.  There is a long-term moral hazard problem of course, but bank runs can be stopped in this way, and it usually works in practice.  So the government’s action was consistent with reasonable economic analysis of the situation.  Unfortunately Diamond and Dybvig wrote their paper before the advent of the Euro, so they do not explain what happens when the national government does not control monetary reserves.  Still, it seems a defensible policy move under the circumstances.

3. The capital injection into the two big banks is wasted and/or inadequate since they are obviously worthless on a net-value basis.  The current share prices are just the speculative value of an out-of-the-money call option.   

The second sentence of this statement might be true (Patrick Honohan made this point earlier on the blog) but the first sentence seems false.   Relatively big banks in small economies probably have considerable economic value even when their market value is near zero or effectively negative on a net basis.  The two big banks will pull through their current “negative value state” and eventually return to being profit-making institutions, with appropriate government support during the current crisis.  Banks are almost always insolvent on the basis of current liquidation value.  The fundamental nature of a bank is that it is a device for changing liquid deposits into illiquid loans.   Government capital support for the banks at this stage seems defensible.  The alternative of full nationalization of all the banks (not just the rogue bank Anglo Irish) carries too many risks for the economy.

The two main Irish bank shares fell back again today following the announcement of the details of the recapitalization — down 16 and 14 per cent respectively. There could be lots of reasons. To begin with there was the extraneous factor of the back-to-back deposits between Anglo and ILP mentioned in a previous post. ILP fell back 15 per cent as well.

Then there is the possibility that shareholders expected a more lenient deal? But how lenient could that have been? The interest rate on the preference shares is stiff enough, but not out of line with prevailing practice in other countries and anyway was well-flagged.

To all intents and purposes, however, the share prices are close to zero — down over 95 per cent on their peak.

My purpose in writing, though, is to point out that even though the preference shares are senior to equity, an injection sufficient to assure solvency going forward could nevertheless have been expected to lift ordinary share prices.

I suspect this is not a well-known effect. Permit me to present a very simple model.

Thus, suppose that there are just three periods. For convenience, assume our bank begins with zero capital.

In period 1, the government decides the amount S it will inject through purchase of preference shares.

In period 2 we discover the true state of the world, i.e. the size of the loan losses (H high in the bad state, L low in the good state). If the losses exceed the funds the government injected, then the bank is liquidated and the shareholders get nothing; If the losses are equal to or less than injection, then the bank continues in operation.

In period 3 the bank, if still in operation, earns franchise profits Z on the rest of its business. It is then wound up; the government receives its injection back if possible. Any surplus goes to the shareholders.

Clearly, if the values H and L are known and if the government injects any amount equal to or less than L, the market value of the shares at the end of period 1 is zero. (Of course, the analysis assumes rational market expectations.)

If the government injects more than that, the market value of the shares at the end of period 1 is p*max{0, (Z-L)}, where p is the probability of the good state. A longer expression gives the share value if the injection S is higher than H.

The point is that even an injection S that is only sufficient to ensure the bank’s survival in the good state will, when announced, increase the market value of the shares.

The Irish Government injection of yesterday was insufficient to do that.