Presentation on Ireland’s Bank Debt

Today, I’m giving a presentation titled “Ireland’s Bank Debt and What Can be Done About It” at the IIEA’s conference Exiting the Crisis. The slides from my talk are available here as a PowerPoint slideshow and here in PDF.

The presentation is based on this earlier blog post, though I think that post failed to emphasise enough that I prefer a restructuring of the promissory notes within the current ELA arrangement to replacing ELA with a loan from the ESM.

The slides were prepared prior to today’s EU banking announcement but are, if anything, more rather than less relevant as the summit did actually make some progress towards EU risk-sharing in relation to bank recapitalisation costs.

Despite this morning’s euphoric claims from Enda Kenny and Eamon Gilmore that this is a “seismic change” and a “deal changer”, I would recommend serious caution. Ireland’s bank debt is a complex sui generis issue and its resolution will not fit easily into the tools the EU is now developing to deal with banks elsewhere. There is lots of negotiating still to be done and premature celebrations can be counter-productive.

What Is Ireland’s Bank Debt and What Can Be Done About It?

In both the run up to the referendum and its aftermath, there have been widespread references to something called “Ireland’s bank debt” with widespread agreement among all right-thinking folk that “something should be done about it” particularly in light of ongoing discussions about the use of EU funds to recapitalise Spanish banks.

One problem with this discussion, as two recent pieces by John McHale and Seamus Coffey discuss, is that there is a lot of fuzziness in discussions of the bank debt issue with a lack of clarity about what people mean by bank debt and what they expect to be done about. Here I want to clarify the size of the Irish state’s outlays and commitments to the banking sector and consider some of the steps that could be taken to reduce the burden associated with these outlays.

Total Outlays and Commitments

The figure of €64 billion has been widely circulated as the cost of bailing out Ireland’s banks. I’m happy to be corrected but I believe the most recent estimate of the total outlays in terms of cash and promissory note commitments comes from this written answer from Michael Noonan on 18th April.  The answer contains a handy table that I reproduce below and it shows the total figure at the slightly-lower €62.8 billion.  Of this, €34.7 billion was supplied to the institutions that comprise the IBRC, while the other €28.1 billion was spent recapitalising and acquiring full ownership of AIB/EBS and Irish Life and Permanent and acquiring preference shares and a minority ordinary shareholding in Bank of Ireland.

IBRC was recapitalised with €30.7 billion in promissory notes and €4 billion in exchequer resources. €20.7 billion from the National Pension Reserve Fund was used to acquire preference and ordinary shares in AIB and Bank of Ireland. Additional exchequer resources were used to spend €4.7 billion on AIB and €2.7 billion on ILP.

So the total outlay and commitments to the Irish bank bailout can be broken into three components:

  • €30.7 billion in promissory notes of which two installment payments have been made (or one made and one kinda-sorta-maybe made depending on your point of view) with many more to come.
  • €20.7 billion from the NPRF has been invested in acquiring ownership stakes in AIB and Bank of Ireland.
  • €11.4 billion of additional exchequer resources have been spent on IBRC, AIB and ILP.

It may be stating the obvious but these figures show that the €63 billion figure does not simply refer to a unique “bank debt” waiting to be defaulted on or renegotiated. The €20.7 billion from the NPRF could have been used to finance deficits instead of borrowing money from the sovereign bond markets and the troika. Similarly the €11.4 billion of cash balances used was largely obtained via borrowing.  However, there are no particular sovereign bond issues or slices of the borrowings from the troika that are specifically earmarked “bank debt.”   The part of the debt that is clearly and unequivocally bank debt, however, is the IBRC.

Let’s look separately at these two areas, the IBRC debt and the rest of the outlays.

AIB, Bank of Ireland and Irish Life and Permanent

The government invested a combined €25.4 billion in acquiring the shareholdings in AIB and Bank of Ireland currently held by the NPRF.  The NPRF’s latest set of accounts values these holdings at €9.36 billion. Those accounts don’t break down the valuations between the two banks but these accounts from 2011:Q3 value the holdings at €9.6 billion, of which €7.1 billion is allocated to the ownership of AIB. The book value of equity in AIB at the end of 2011 was €14.6 billion.  Given weak operating profits and ongoing loan loss writedowns, the NPRF valuation looks highly optimistic.

The government also own Irish Life and Permanent, which had equity with a book value of €3.5 billion at the end of 2011. Given the bank’s serious ongoing problems, I’d be surprised if it had a market value much above zero.

So the Irish government has invested €28.1 billion in bank shares that are likely worth less than €9 billion now.  This calculation omits €2.0 billion received from Bank of Ireland in preference share dividends, the repurchase of warrants and the sale of ordinary shares to Wilbur Ross’s consortium. Still, you’re looking at a loss in the region of €17 billion, which is over 10 percent of GDP.

This raises the question of what could happen with these investments if ESM was to be given the power to recapitalise banks (a big if).   Personally, I would be shocked if such a decision did not come with guidelines for ESM to recapitalise banks in a manner consistent with not losing money.   For example, insolvent banks may have to be restored to a “zero capital” point by creditor write-downs or national government funding before ESM gets involved in recapitalizing banks.

Even if ESM was allowed get involved in directly recapitalising banks in this manner, that does not mean it will be allowed to simply acquire shares in banks that are already recapitalised. And even if it was, it is wildly optimistic to believe that the EU would sanction acquiring these shares for the €28.1 billion we spent on them rather than what they are currently worth.

All told, I think we should welcome the prospect of ESM acquiring AIB and ILP as well as our shares in Bank of Ireland.  We could use the funds obtained (perhaps €9 billion if my guess above is correct) to reduce debt and perhaps implement some well-judged capital programmes. We should also welcome ESM playing a role in any future recapitalisation of these institutions as this would eliminate this source the uncertainty from considerations of the Irish public finances. But Irish citizens will almost certainly still be left with the large loss on their “investment” in these institutions.


Despite putting €35 billion in to recapitalise the IBRC, the institution only has equity capital of €3 billion.  The €31 billion in promissory notes thus account for almost all of the loss the taxpayer is incurring due to the IBRC. As is now pretty well known (see details here) the payments on the promissory notes will be used to repay ELA loans from the Central Bank of Ireland and that these loan repayments will retire the money issued when the ELA loans were created.

Many people would like the Central Bank to simply write off these loans but the Eurosystem’s rules do not allow this. An alternative is to delay the repayment of the ELA. However, Eurosystem rules allow the ECB to block any renegotiation of the terms of the ELA loans and, despite months of negotiations, it seems pretty clear that the ECB are not in any way minded to allow a rescheduling of IBRC’s ELA repayments.  If lobbying from Ireland had been likely to produce a rescheduling, then it would have occurred in late March of this year.  Instead, IBRC were required to meet their scheduled ELA repayment in full and on time.

Perhaps the most realistic option for replacing the promissory notes is to use Ireland’s potential access to the European Stabilisation Mechanism (ESM) to refinance the IBRC.  For example, ESM could issue thirty-year notes to financial markets and loan the funds raised to the Irish government over the same term.  In turn, the Irish government could give these funds to IBRC as a replacement for the promissory note.  IBRC could then pay off all of its ELA debts and be quickly wound down.

I read Seamus’s contribution as arguing against an ESM deal and John’s contribution as neutral. I’m happy to say that, in the absence of any deal with the ECB on restructuring the promissory notes, I am willing to endorse an ESM deal as an obtainable second best.

What can be said in favour or against such an approach? Against this approach is the fact that the interest rate on such a loan, likely to be around 3 percent, is higher than the effective current cost of ELA, which is the ECB main refinancing rate (1 percent at the time of writing).  Another complication is that the size of such a loan from ESM would be large: €28.1 billion would be required to pay off the remaining ELA and the IBRC’s one-year loan from Bank of Ireland.  The existing European debt issuance agencies such as EFSF and EFSM have not yet placed a thirty-year bond of anywhere close to this size and this may make it difficult to achieve the desired low interest rate.

Those caveats noted, the benefits to this arrangement seem to far outweigh the costs.  An ESM loan locks in a low interest rate for a long period and may provide lower cost financing than the ELA arrangement once the ECB’s main refinancing rate moves upwards again.  More importantly, a thirty-year ESM loan would require only an annual interest payment for 29 years followed by a payoff of the principal in year 30.

This deferral of principal payments would significantly reduce Ireland’s cash financing requirements over the next decade.  It would also provide a substantial reduction in the net present value of the burden of this debt when calculated using realistic discount rates reflecting Ireland’s likely cost of funding.

For example, in this paper, I calculated that the IBRC’s ELA would be paid off by 2022 if Ireland sticks to its current schedule of paying off €3.1 billion a year in promissory notes. Using a 7 percent discount rate, this stream of planned promissory note payments plus the IBRC’s loan repayment to Bank of Ireland in 2013 have a combined net present value almost twice as large as the stream of payments associated with paying 3 percent interest on a €28.1 billion loan for 29 years and then repaying the principal in year 30. (Calculations in this spreadsheet. Since the promissory note example incorporatse €3.1 billion for repaying the IBRC’s one-year loan from Bank of Ireland next year while the ESM example has that money being borrowed from ESM).

Given the seriousness of the current problems with the public finances, one can hardly say that such a restructuring would guarantee sustainability but it would represent a significant improvement over current arrangements.

As a final thought, I’d note that each of the realistic options for lowering the burden of the bank debt, whether they involve purchases of equity stakes in live banks or refinancing the debts of dead banks, involved access to the ESM.  Those who advocated voting No to facilitate a “deal on bank debt” were advocating cutting off access to ESM.  I’m still in the dark as to how those who adopted this position believed Ireland was going to renegotiate a bank debt deal after voting No.