Subordinated Bond Strategy?

The idea of subordinated bond holders in AIB and\or Bank of Ireland potentially taking haircuts has been discussed on various occasions since the IMF-EU deal was announced on Sunday. This post examines this issue and tries to figure out what the government’s strategy is.

The Minister for Finance’s statement on this matter was as follows:

As I said in my statement on the 30th of September last, there will be significant burden sharing by junior debt holders in Irish Nationwide and Anglo Irish Bank. These two institutions had received very substantial amounts of State assistance and it was only right that this should be done.

My Department has been working with the Office of the Attorney General to draft appropriate legislation to achieve this and this is near finalisation. Parallel to this Anglo Irish Bank has run a buyback operation which will offer these bondholders an exchange of new debt for old but at a discount of at least 80%. This process is still underway and will be concluded shortly.

Obviously this approach will also have to be considered in other situations where an institution receives substantial and significant State assistance in terms of capital provided to maintain their solvency ratios. I hope to be in a position soon to announce this legislation.

The policy conditionality document goes into more detail:

Consistent with EU State Aid rules, burden sharing will be achieved with holders of subordinated debt in relevant credit institutions over the period of the programme. This will be based on the quantum of capital and other financial assistance the State commits to support specific credit institutions and the financial viability of those institutions in the absence of such support. Resolution and restructuring legislation will address the issue of burden sharing by subordinated bondholders will be submitted to the Oireachtas by end-2010. Where it is appropriate, the process of implementing liability management exercises similar to that which is currently being undertaken in relation to holders of subordinated debt in Anglo Irish Bank will be commenced by end-Q1 2011.

Now let’s bring into the mix the Central Bank’s statement from Sunday:

The Central Bank has set a new minimum capital requirement for Allied Irish Bank, Bank of Ireland, ILP and EBS of 10.5% Core Tier 1. In addition, the Central Bank is requiring these banks to raise sufficient capital to achieve a capital ratio of at least 12% core tier 1 by 28 February, 2011 in the case of AIB, BOI and EBS and by 31 May 2011 in the case of ILP.

And after this, we get

Bank of Ireland, Allied Irish Banks, ILP and EBS will be subject, as previously announced, to a stress test in March 2011 under the Central Bank’s PCAR methodology. If, as a result of the PCAR, banks are assessed to be at risk of falling below the 10.5% core tier 1 target then further capital injections will occur.

So, the following appears to be the sequence of events:

December: Resolution legislation is introduced outlining a mechanism whereby subdebt holders lose out if the State has to provide a large amount of capital to maintain solvency.

February: AIB and BoI are recapitalised, presumably mainly or totally with state money, to 12 percent core tier one ratios.

March: PCAR is completed. This will involve writing down of asset values and this will trigger the need for more capital injections if the core tier one ratio fall below 10.5%.

So how exactly might burden sharing with subordinated bondholders happen? Frankly, it’s as clear as mud.

One interpretation is the following: AIB and Bank of Ireland are capitalised to 14% and 12.5% core equity ratios by the end of February. After this occurs, there are stress tests and only if these stress tests uncover losses that wipe out the end-February levels of capital will the subdebt holders be start to lose out.

If that’s the scenario, then it would seem pretty unlikely that subdebt holders will lose out. Stress tests rarely uncover that a bank has lost 14% of its risk weighted assets.

So perhaps all the talk of subdebt holders losing is just waffle. However, the tone of the public comments on this has been to suggest burden sharing is likely.

If that’s the case, the process might have to be something like as follows: AIB and Bank of Ireland are capitalised to 14% and 12.5% core equity ratios by the end of February. Then more money is put in after the PCAR. At this point, if the combined amount of state funding that has been put in during these two stages (the “quantum of capital”) is the difference between solvency and insolvency, then the legislation triggers a mechanism through which subdebt holders lose out.

I’m no lawyer but this latter mechanism seems legally dubious. It seems to propose that the state can put €9.8 billion in equity into AIB in February and then turn around in March after some additional losses have been diagnosed and somehow ask bondholders, who are senior to equity in the liability pecking order, to take a hit even though the bank is not insolvent at that point.

There seems to be a cake-and-eating-it problem here. On the one hand, the government wants to reassure everyone owed money by the Irish banks that their money is safe. Thus, it wants to keep the banks adequately capitalised and avoid ever having to declare a bank temporarily insolvent. On the other hand, since there are strong suspicions that the banks really are insolvent, it wants a mechanism to allow it to apply ex post haircuts to subordinated bondholders after a stress test shows that the banks have been insolvent.

Or maybe I’ve got it all wrong. I’m happy to hear from others who perhaps have a better idea than me as to what’s going on.

EU-IMF Bailout Borrowing Rates

The NTMA have released a note explaining the interest rate associated with the bailout package. The average interest rate is 5.82% and the average maturity of the borrowings is 7.5 years.

The 5.82% is presented as being comprised of an average rate of 5.7% from the IMF and EFSM and 6.05% from the EFSF.

The statement leaves a few questions unanswered about the IMF and EU rates.

Regarding the IMF rate, the IMF’s statement on Sunday night said

At the current SDR interest rate, the average lending interest rate at the peak level of access under the arrangement (2,320 percent of quota) would be 3.12 percent during the first three years, and just under 4 percent after three years.

How do we get from a weighted average of 3.12% and 4% to the government’s figure of 5.7%? The answer appears to be related to the fact that the IMF lends in SDRs at a floating rate. From the NTMA statement:

The SDR comprises a basket of four currencies, Euro, Sterling, the US Dollar and Japanese Yen. The IMF’s SDR lending rate is based on the three month floating interest rates for the currencies in the basket. In the presentation of the financial support programme the interest cost on the IMF’s floating rate SDR lending is expressed as the equivalent rate when the funds are fully swapped into fixed rate Euro of 7.5 years duration.

Since both IMF and NTMA statements must be true, this suggests that the cost of swapping a floating rate SDR loan into a fixed rate Euro loan is somewhere between 170 and 258 basis points. That seems very high to me.

Regards the EFSM rate, the NTMA tell us that it “will be at a rate similar to the IMF funds, i.e. 5.7 per cent per annum.” But presumably the EFSM is lending in euros and so there was no need to undertake a very expensive swap exercise, so this deal factors in a profit margin for the EFSF that does not apply to the IMF loan.

Finally, the mystery that is the EFSF rate is revealed. 6.05%. Less than the 6.7% that was doing the rounds last week but more than the 5.7% that I had guessed a few weeks ago and still a pretty hefty rate.

I have to confess to having no idea how this 6.05% was arrived at. The EFSF FAQ states

fixed-rate loans are based upon the rates corresponding to swap rates for the relevant maturities. In addition there is a charge of 300 basis points for maturities up to three years and an extra 100 basis points per year for loans longer than three years. A one time service fee of 50 basis points is charged to cover operational costs.

So let’s plug in the numbers consistent with a seven year maturity. The seven year swap rate is 2.85%. Add 400 basis points for the profit margin and you’re at 6.85%. And this ignores the 50 basis point service fee which, if annualised over the term, would bring the rate up to 6.92%. Finally, this also ignores the following aspect of EFSF lending. From the framework agreement:

The Service Fee and the net present value of the anticipated Margin, together with such other amounts as EFSF decides to retain as an additional cash buffer, will be deducted from the cash amount remitted to Borrower in respect of each Loan (such that on the disbursement date (the “Disbursement Date”) the Borrower receives the net amount (the “Net Disbursement Amount”)) but shall not reduce the principal amount of such Loan that the Borrower is liable to repay and on which interest accrues under the relevant Loan.

This seems to mean that we are paying the service fee and margin up front. In addition, we are going to pay interest on a cash buffer, which is money kept by EFSF that we’ll never see.

So, two thoughts here. First, it seems likely that the government negotiated in the final days to get the “profit margin” aspect of the EFSF money down. Second, I wouldn’t be surprised if the true effective cost of this loan is understated by the 6.05% figure, for instance because it doesn’t include the up-front service fee or the role played by the cash buffer.

A full detailed explanation of the EFSF rate from either the Irish government or the EFSF would be very welcome.

EU-IMF Programme Details

The Department of Finance has released the documents setting out the “policy conditions for the provision of financial support to Ireland by EU member states and the International Monetary Fund.” (Statement by the Minister here.) The documents are available here. Both the review process (including “actions for the twelfth review”) and the detailed economic and financial policies appear to be very detailed.