Friday, February 1, 2013

Deficit-, debt- and expenditure-impacting banking measures

A recent post over on Notesonthefront has attracted a lot of attention on “the cost of the banking crisis”.  The figures from Eurostat show that Ireland contributed 42% of the EU total and have been widely quoted including this prominent piece in The Irish Examiner

42% of Europe’s banking crisis paid by Ireland

Ireland has paid 42% of the total cost of the European banking crisis, at a cost of close to €9,000 per person, according to Eurostat.

This is not the correct interpretation of the Eurostat figures.  The Eurostat figures used to support the claim give one impact of the banking crisis on public finances, that is:

1. The impact on the flow of annual government deficits

It is not the case that this reflects the full cost of the banking crisis.  As will be discussed below not all of the measures introduced in response to the banking crisis are deficit impacting.  An additional impact that could be considered is:

2. The impact on the stock of the gross government debt

There is no reason to expect #1 to be the same as #2.  There may be transactions with the banking sector that are counted as deficit increasing but if they are funded from existing resources they will not add to the stock of debt. 

It will also be the case that there may be transactions which are not counted as deficit increasing but if funded with borrowed money they will add to the stock of debt.

Eurostat have produced figures on #1 as they can measure the revenue and expenditure flows on an annual basis.  However, they will not produce statistics on #2 for the simple reason that money is fungible.  Governments borrow money because their overall expenditure exceeds their revenue.  It is difficult to attribute changes in debt to a single expenditure item because reductions in any expenditure would reduce the need to borrow.  That doesn’t mean it isn’t attempted though!

The Eurostat figures show that between 2008 and 2011, measures related to the banking crisis contributed €41 billion to Ireland’s general government deficits.  By the end of 2011, the government had contributed around €63 billion to the banking sector and, of this we can guess that around €47 billion was with “borrowed” money (essentially it is non-NPRF money, but the assumption is that this expenditure increased borrowings).

The above paragraph shows a third, a more complete, measure of the impact of the banking crisis on the public finances:

3. Total expenditure incurred by general government as a result of the banking crisis.

Again Eurostat are not going to produce statistics on this as much of the expenditure will be by public investment funds (such as the NPRF) or by special purpose vehicles (such as NAMA).  A lot of these are “off-balance sheet” transactions.

So for Ireland, at the end of 2011, the figures were:

  1. €41 billion
  2. c. €47 billion
  3. €63 billion

The first figure has received lots of attention but it is actually the smallest.  The second figure is a bit of a guess and although the largest the final figure could yet be under-stated. 

Number 3 will be clouded by the use of special purpose vehicles which initially keep the expenditure outside the government sector.  NAMA has spent €32 billion acquiring loans with a nominal value of €74 billion from our delinquent banks.  NAMA is not going to lose €32 billion but a shortfall of, say, €5 billion is possible on its operations which will have to be made good with expenditure by the government.

In Ireland in 2011 #1 makes up about two-thirds of #3.  In other countries the gap between #1 and #3 is likely to be even greater.  In part, this is down to the type of bailout adopted in Ireland, rather than the total cost.

For example, we know that the UK has contributed £66 billion to just two banks: Lloyds and RBS.  That is around €80 billion which would count in #3 (full expenditure) but the figure for the UK in the Eurostat deficit data is just €11 billion.  Where did the other €69 billion go?

To answer this question we must explore what Eurostat measured when they provided figures for the deficit-impacting measures introduced in response to the banking crisis.  This all comes back to a Eurostat decision published in July 2009.

The key is whether a measures is considered as a “financial transaction” or a “capital transfer”.  Financial transactions are not deficit impacting; capital transfers are deficit impacting.  The impact of both in the debt is not objective, while their impact on expenditure is unambiguous.  So what is a “financial transaction”?  Per the Eurostat decision:

The valuation of financial transactions: In principle the ESA 95 provides for financial transactions (which do not impact on the government deficit) to be recorded "at the transaction values, that is, the values in national currency at which the financial assets and/or liabilities involved are created, liquidated, exchanged or assumed between institutional units, or between them and the rest of the world, on the basis of commercial considerations only" (paragraph 5.134).

However it is acknowledged in paragraph 5.136 that "in cases where the counterpart transaction of a financial transaction is, for example, a transfer and therefore the financial transaction is undertaken other than for purely commercial considerations, the transaction value is identified with the current market value of the financial assets and/or liabilities involved".

It does, of course, leave something of a grey area but we can see that if a financial transaction is done at the “current market value” it does not impact on the government deficit, whereas if it happens above the “current market value” the transaction value is identified and the amount above that is considered a capital “transfer”. 

The Eurostat decision goes through different forms of banking support and shows whether they impact on the deficit or not.

  1. Recapitalisation operations
  2. Lending
  3. Guarantees
  4. Purchase of assets and defeasance
  5. Exchange of assets
  6. Classification of certain new bodies
  7. Recording of certain transactions carried out by public corporations

Eurostat did not need to provide a separate decision for the general government gross debt measure it produces.  The debt is just the sum of all of the liabilities of the general government sector.  It does not matter what the money is used for.  All that matters is whether a liability exists or not.

For example, when the Promissory Notes were created in 2010 they were classed as a “loan to government” from Anglo/INBS and would immediately be added to the general government gross debt.  There was some issue of whether they would count in the 2010 deficit but the counter transaction to the loan was recorded as a “notional capital transfer” as the government promised to repay a €31 billion loan without first receiving the money from the bank as is the case with a typical loan.

Here is the full set of recapitalisation payments made to the banks since 2009, classified as “financial transactions” or “capital transfers”.

Bank Recapitalisation Payments

The payments under financial transactions were not deficit-increasing, whereas those recorded as capital transfers were deficit increasing.  It remains to be seen what value can be realised through the sale of the assets acquired through the financial transactions. 

There have been some sales already.  In 2011, €1.1 billion was realised from the sale of a 35% ordinary shareholding in BOI, while in January 2012 the €1 billion contingent capital note in BOI was sold at close to par.  There was also a transaction in 2010 that saw €1.7 billion of the preference shares in BOI converted into ordinary shares.

It can be seen that 75% of the capital transfers total arises from the Promissory Notes issued in 2010.  No other country has used such a scheme to prop up a bust bank and the loan loss figures mean that any mechanism devised would have been recorded as a capital transfer.

The transactions are split between those undertaken by the NPRF (directed by the Minister for Finance) and those undertaken by the Exchequer (directly by the Minister for Finance).

All of the financial transactions involving preference and ordinary shares in AIB and BOI were done through the NPRF.  It should be noted that the €3.5 billion of preference shares in both AIB and BOI bought in 2009 by NPRF was funded with €4 billion from the NPRF and a “front-contribution” of €3 billion from the Exchequer to the NPRF.  All the contingent capital notes transactions as well as the ordinary shares in PTSB and Irish Life were funded, and now held, by the Exchequer.

Most of the capital transfers were provided by the Exchequer but the €6 billion capital transfer provided to AIB in July 2011 was split with €2.3 billion coming from the Exchequer and €3.7 billion coming from the NPRF.  This €3.7 billion from the NPRF was a deficit-increasing expenditure (though didn’t impact on the debt as it came from pre-existing funds).

So has Ireland carried 42% of the total EU cost of the banking crisis?  Impossible to say.  We do know that Ireland has incurred 42% of the deficit-impacting measures introduced in response to the crisis.  But that is not the same thing as the total cost. 

In Ireland’s case, the Promissory Notes have pushed up the capital transfers to an extraordinarily high figure relative to other EU countries.  In fact the explanatory notes to the data say (on page 12) that:

The only case where government liabilities increased much more than government assets is Ireland. This can be explained by the fact that most interventions have been immediately recorded as deficit-increasing government expenditure and not as financial transactions.

Most EU countries have generally recapitalised their banks using financial transactions (purchase of shares and other instruments).   Ireland didn’t have any money to buy anything in Anglo and, as stated above, Anglo was nursing such loan losses that all efforts to keep it solvent were going to be recorded as capital transfers anyway.

Throughout the EU it remains to be seen what the assets acquired through these the financial transaction approach to recapitalising their banks will actually be worth.  If losses relative to the purchase price are crystallised on the sale of these assets then the difference will be recorded as a capital transfer and Ireland may not be such an outlier.

How much of the £66 billion pounds provided to Lloyds and RBS will be returned to the UK Exchequer? Will they get back much of the £14 billion (€17 billion) capital contributions that Lloyds through Bank of Scotland(Ireland) and RBS through Ulster Bank have made to their loss-making Irish subsidiaries?  In total, the UK has made a cash outlay of around €145 billion to rescue its banks.  See question on “current level of support” in this set of FAQs.

For the moment though, Ireland is extreme when it comes to the deficit-increasing impact of the banking crisis.  It makes a good headline and the extent and cost of the disaster in Ireland will always be high relative to other EU countries.  But it should not be thought that other countries have escaped lightly from the banking crisis.  They have simply gone about it in a different manner and haven’t used Promissory Notes.  The true cost of this period will only emerge over the next decade or longer when their investments and special purpose vehicles are unwound.


  1. Great work Séamus. Your point is valid. The Eurostat data is what it says. If anything Ireland's 'total public' cost may be relatively under-stated going by the nature of capital transactions in other EU countries. Also, IMF data are enlightening here

    if it were not for Cyprus the Irish bailout would be the largest by far in Europe in recent times.

    Tom Healy


    Interest rebuttal of Unite's 42% argument in the comment section of the attached article. Unite's figure is rubbished.