Friday, June 28, 2013

The Anglo tapes and alternatives to the guarantee

Below the fold is a narrative that tries to pull together the revelations in the Anglo tapes this weekend as well as the piecemeal information we already have about the run-up to, and aftermath of, the blanket guarantee introduced in September 2008.  For anyone who has shown even a modicum of interest in these developments there is nothing new that follows but maybe it will help to pull a few threads together on the alternatives that were available and the decisions that were taken.

- - - - - - - - - - - - - - - - - - - - -  -

There were alternatives to the two-year blanket guarantee of almost all the liabilities of six Irish banks that was announced on the morning of Tuesday 30th September 2008.

A little more than 12 hours after its announcement, Prof. Morgan Kelly went on RTE television and accurately foretold that the guarantee would cost the State billions and opined that "there are some non-retail banks that could have been let go and nobody would have missed them".

In a speech delivered three years later the current governor of the Central Bank, Patrick Honohan, said:

“It would have been better had Anglo and INBS been put into resolution as soon as it became clear that their capital was going to be wiped-out by unavoidable losses on developer loans. This should have been evident before September 2008, but was not, leading the Government of the day to include these two failed entities in its blanket guarantee.”

Was Prof. Kelly the only one in September 2008 who knew that Anglo Irish Bank and INBS were bust and should not be saved? While it was evident to Prof. Kelly that their capital would be wiped-out we now know that people in the banks themselves, and also in the Department of Finance, knew this could happen, even if they were not saying so publicly at the time.

The recordings of internal phone calls from Anglo Irish Bank are the most recent source of this information being slowly made public. A lot of attention has been given to the request made by Anglo for €7 billion of liquidity from the Central Bank but what is most telling from the conversation between John Bowe and Peter Fitzpatrick that took place on the 18th of September 2008 is Bowe's prediction of what will happen to the bank.

“I don’t think we’re an easy sell to anybody. So, do I think it’s going to be possible to offload it? No, I don’t. What will it end up being? It could be breaking it up and selling individual books, it could be nationalisation, you know.”

Break-up and nationalisation are outcomes for a bank that is bust. Neither Bowe nor Fitzpatrick put an estimate on the scale of the losses that would lead to these outcomes. 

There were loss estimates been discussed in meetings in the Department of Finance.  The minutes of some of these meetings have been made available through the Oireachtas Public Accounts Committee. One such meeting is listed as taking place on Thursday 25th September 2008, just five days before the blanket guarantee was announced. At the meeting the then Secretary-General of the Department of Finance, David Doyle is recorded as noting:

“that Government would need a good idea of the potential loss exposures within Anglo and INBS - on some assumptions INBS could be €2 billion after capital and Anglo could be €8.5 billion.”

We know that the public announcements from the government and their officials that, at the time, they were responding to a liquidity crisis but it is clear that, in the background, they knew a solvency crisis was also likely.  At this meeting potential losses of €10.5 billion “after capital” of €8 billion were discussed as a possibility.

The two-year blanket guarantee was the eventual outcome but many alternatives were discussed as the crisis reached its zenith. Although the guarantee was introduced it is not clear how much support it had from any source.

We know that in February 2008 a memo prepared by officials in the Department of Finance stated that:

“As a matter of public policy to protect the interests of taxpayers any requirement to provide open-ended/legally binding State guarantees which would expose the Exchequer to the risk of very significant costs are not regarded as part of the toolkit for successful crisis management and resolution.”

This opposition continued in the autumn. When writing about his discussions at the time with the Minister for Finance, Brian Lenihan, columnist David McWilliams said:

“The minister called me the next day and again on Friday, the 19th, when he rang to say they were contemplating a partial guarantee. My view was that such a move would accelerate capital flight, not avert it. From then on, we spoke on a daily basis, but he still wasn't convinced and, from what I could gather, his officials in the department were dead set against a full guarantee, although they didn't seem to be coming up with an alternative.”

At a cost of €7 million, consultants from Merrill Lynch were brought in to come up with a set of alternatives. Prior to their arrival one possibility had already been eliminated - allowing a bank failure. This position was cemented at all official levels through the summer of 2008. In his subsequent report on the crisis, Prof. Patrick Honohan noted that this position “departed from the textbook” but also that it “simplified the decision making process”. This simplification may have cost the State dear.

Taking account of this position Merrill Lynch presented six options in a report provided to the Minister at 6pm on Monday September 29th.

  1. Immediate Liquidity Provision
  2. State Protective Custody
  3. Secured Lending Scheme
  4. Good Banks/Bad Banks
  5. Consolidation of Financial Institutions
  6. Guarantee of Six Primary Regulated Banks

Merrill Lynch did not favour a full guarantee. In fact, in a meeting the previous Friday when they presented their preliminary findings Merrill Lynch warned of the dangers of the guarantee and stated that it “could be a mistake”. In their final report they said:

“The scale of such a guarantee could be over €500bn. This would almost certainly negatively impact the State's sovereign credit rating and could raise issues as to its credibility. The wider market will be aware that Ireland could not afford to cover the full amount if required.”

We know that some of the other options proposed by Merrill Lynch were given some thought. The idea of putting Anglo and INBS into state protective custody using preference shares (a form of nationalisation) was put on the table but the reported reaction of the Taoiseach Brian Cowen was that “we are not f*****g nationalising Anglo”. The consolidation of financial institutions was proposed when Anglo tried to initiate merger discussions with both AIB and Bank of Ireland. Neither was willing to entertain even beginning such discussions.

The tapes released this week show that Anglo’s management were keen to avail of immediate liquidity or secured lending. The €7 billion of liquidity that John Bowe suggested the bank needed could have been provided by the Central Bank. While this would not have stemmed the flow of deposits from Anglo it would have provided the funding for the bank to remain open.

Of course, we know that the €7 billion figure was pretty arbitrary to say the least and was part of a strategy which Bowe described in one of the tapes as:

“The strategy here is - you pull them in, you get them to write a big cheque, and they have to support that money... if they saw the enormity of it up front... they might say the cost to tax payers is too high... if it looks big enough to be important, but not too big that it spoils everything, then I think you have a chance.”

While these has been justified opprobrium to the attitude of Anglo’s management, the reality is that if the €7 billion of liquidity had been provided (or even more than that) the cost to the State of the Anglo disaster might have been significantly less than that incurred as a result of the guarantee actually introduced.  The indications are that there are lots more recorded conversations to be released but as of yet there is no evidence that Anglo sought the introduction of the guarantee.  In fact, in an interview with in November 2011, David Drumm says that Anglo did not look for any guarantee.

“We asked for a certain amount of money to be secured, we were never given an answer. The two weeks, I call it the two weeks of Lehman, because it was that between Lehman and the guarantee being issued, went by with us running around trying to get the Central Bank, the government, we put messages into the Department of Finance because the governor kept telling me ‘you know I am relying on the department of finance here and I have to keep asking them’. He wasn’t getting answers from them. So they were all gone into, I think of it like balls of little mercury, sort of scattered. They just weren’t joined up.

We came to the 30 of September, the 29 of September and we run out of money and I was down at the Central Bank and they said OK you are going to have to ask for emergency funding. Fine how do I do that? They sent me a draft of letter, I got it typed up, I signed it, we need two billion tomorrow to be able to open the doors. So signed that letter and then that was a long day, as you know. I could talk all day about that. We were down at BOI asking them to merge with us. But the end of the day was asking the Central Bank for the emergency funding, two billion. I went home, I ate my dinner and I went to bed and when I woke up in the morning I had several voicemails and messages on my phone and one of them was from Pat Neary to tell me that the government had guaranteed. Then of course I heard it on Morning Ireland. Anglo Irish did not ask them to do that, we asked for a secured loan.”

This is supported by the conversations in the tapes released this week.  Anglo was looking for the liquidity to ensure that their doors stayed open.

There were two sources from which this liquidity could have been provided to Anglo. The National Treasury Management Agency could have provided it by converting part of the National Pension Reserve Fund into cash. This was discussed but discounted as it was not part of the remit of the NTMA to support the banking industry.

The second possibility was the provision of a Secured Lending Scheme (SLS) or plain Emergency Liquidity Assistance (ELA) by the Central Bank of Ireland. It is the job of the Central Bank to support the banking industry.

The €7 billion request discussed by John Bowe and subsequent requests by David Drumm were not acceded to. The decision not to provide ELA to Anglo, according to Prof. Honohan, was “open to question” and he further says that while this would not have been ultimately decisive it could have helped buy some time.

“while use of ELA would only have been a temporary solution, it might have bought some breathing space while other possibilities were being explored to address the unprecedented situation that many - not only in Ireland - were facing.”

The provision of this “breathing space” would have cost money but it might not have cost €35 billion that the rescue of Anglo and INBS through the two-year blanket guarantee ultimately cost.

There were alternatives to this guarantee but they were all ruled out for one reason or another. All official levels were opposed to an outright bank failure. Nationalisation was ruled out. Both AIB and Bank of Ireland were opposed to any mergers. Neither the NTMA nor, most crucially, the Central Bank were willing to provide the necessary liquidity to cover the loss of deposits that Anglo was experiencing.

As the alternatives were ruled out the introduction of a guarantee became ever more likely. But even then there were alternatives. The option chosen was a two-year blanket guarantee of almost all the liabilities of the covered institutions. An alternative would have been a partial guarantee of some liabilities. As the problem being address was a deposit flight it was unnecessary to include bonds and other term investments as these were “locked in” – they could not fly out the doors. As Prof. Honohan noted “their inclusion complicated eventual loss allocation and resolution options.”  The inclusion of some subordinated debt has close to nothing to support it.

Merrill Lynch also favoured the provision of emergency liquidity to Anglo. Their report concludes that “the extension of a discrete liquidity advance is important to stabilise Anglo (and possibly INBS) and avoid immediate contagion risk.” If this was made available the breathing space provided may have allowed the time for a European response to the fast developing crisis to have evolved.

In fact, at the EU Summit on the 12th of October 2008 it was decided that some form of guarantee was an appropriate response to the banking crisis. The summit statement said:

“To this aim, Governments would make available for an interim period and on appropriate commercial terms, directly or indirectly, a Government guarantee, insurance, or other similar arrangements for new medium term (up to 5 years) bank senior debt issuance. Depending on domestic market conditions in each country, actions could be targeted at some specific and relevant types of debt issuance.”

This was very different to the guarantee introduced in Ireland a fortnight earlier. This proposal was for a guarantee of new bank debt rather than existing debt in the case of the Irish guarantee. This is unlikely to have been extensive enough to stem the tide of deposits from Irish banks, but allied with ECB and Central Bank liquidity it would at least have kept resolution options open.  The panic that expanded as September 2008 progressed was partly as a result of the failure of the Central Bank to act as a “lender of last resort”.

As we now know the blanket guarantee was itself only temporarily effective. It provided an immediate liquidity solution to a long-term solvency problem. The Central Bank was eventually required to provide massive amounts of Emergency Liquidity Assistance to Anglo and Irish Nationwide, while the outflow of deposits that resumed close to the end of the guarantee meant that the reliance of the covered Irish banks on central bank liquidity peaked at over €150 billion – nearly 100% of GDP.

The central bank funding was provided but as a result of the sovereign guarantee almost all of the losses made by the banks in excess of their shareholder equity was covered by the State rather than the banks’ creditors.  Close to €25 billion of shareholder equity was eliminated and around €13 billion of losses were subsequently imposed on subordinated bondholders.

So an equally important question to why the guarantee was introduced is why the Central Bank of Ireland did not provide the liquidity assistance that Anglo sought.  Maybe we could have saved a billion or ten if Shane Ross’s advice had been followed and Sean Fitzpatrick was made governor of the Central Bank.

Or maybe we wouldn’t.  The problem with the guarantee was that it hugely limited the ability of bust banks to be put into resolution.  The retail banks (AIB, BOI, EBS & PTSB) were always going to be rescued and the cost of doing so would probably be similar in any scenario with or without a retrospective blanket guarantee. 

The key issue is the failure to put Anglo and INBS into resolution and to distribute the cost of covering more of its losses to the banks’ creditors.  The failure of Anglo and INBS was always going to cost the State money. At a minimum insured depositors would have to be made good and it is probable that all retail depositors would have been repaid in full.  But there were creditors in the bank in September 2008 who could have shouldered some/more of the burden.

There is no doubt that the guarantee was an impediment to this happening but there is also little to indicate that if the “breathing space” that ELA could have provided was available it would have been used effectively.   Almost five months after the guarantee was introduced PwC presented a report to the Minister for Finance on the status of Anglo (at a cost of €5 million) and concluded that:

“Under the PwC highest stress scenario, Anglo’s core equity and tier 1 ratios are projected to exceed regulatory minima (Tier 1 – 4%) at 30 September 2010 after taking account of operating profits and stressed impairments.”

This conclusion is incredible given that the previous September there were meetings in the Department of Finance discussing potential losses of €8.5 billion in Anglo “after capital” and phone calls within Anglo itself talking about its possible break-up or nationalisation. 

Patrick Honohan is right that Anglo and INBS should have been put into resolution but what is still not clear is how significantly this was prevented by the guarantee.  Was the conclusion of PwC taken as fact or was it used as cover for a strategy that had become locked-in because of the guarantee?  If the strategy followed was indeed as a result of the guarantee how much did it cost? €5 billion? €10 billion? More?

One things has remained constant in the four-and-a-half years since the guarantee was introduced – there are more questions than answers.  Any inquiry into the decision to provide the guarantee must also include a careful analysis of why each of the alternatives available was rejected.  We may be waiting a good while yet though.


  1. It seems pretty clear from the unwillingness of the CBI to provide ELA to Anglo that they thought Anglo was insolvent.

    It has been clear for some time that the NTMA thought the same.

    The dogs on thought so too...

    So the state guaranteed one bank that it knew to be insolvent, probably two (in the form on INBS) and possibly more (as rumours swirled around the bowl about the inability of the other banks to fund themselves and the scale of their likely losses).

    This is reckless politicians at the top of the blame pyramid, no?

  2. This smacks of Bi Partisan Payola - we need look at the people who are protected in this country and their interrelationships with the banks. Its a very tangled web.

  3. Just a few brief points based on Chs 9 and 10 of Antoin and my recent book ( Fall of Celtic etc...FCT).

    1. You refer to the DoF draft of Feb 08. However, as noted in FCT(p. 180) the April draft contained a significant amendment re special circumstance leading to aid being "unavoidable". This is important because it shows that six months before D day DoF were beginning to think of the unthinkable...

    2. I dont think they could easily ( or even at all?) put the banks into resolution at or soon after Sept 08. The legislative proposals had been abandoned - for good or bad reasons - some months earlier (FCTp.181). Indeed it was only in early 2011, after troika prodding, that they passed the law.

    3. Based on a detailed reading of all the ML pieces, ( Sep 28 memo plus earlier Power points) , my conclusion was that their views were unclear, ambiguous and inconsistent (FCT pp209-211) on the overall guarantee idea and aspects of it.

    4. I dont think either the present or even the previous incumbents of the CBI Governorships would agree that "the job of the CBI was to support the banking industry" ( although the Drummer might have thought so !). Mind you, pressures from politicians pushed in this direction, unfortunately (FCT p.87)

    5. I think you might have had in mind that the CBI should have been a "lender of last resort". Fair enough, but remember , if the banks were ajudged somehow to have been insolvent, the CBI was and is explicitly prohibited by their and ECB rules from extending ELA. Hence the PN to get around this. ( and the row in Cyprus as to whether someone lied earlier about the Cypriot banks being solvent in order to squeeze money out of the ECB ) Pat N, for that reason alone, had to assert/believe the banks were solvent . See FCT pp 201-202..


    1. Hi Donal,

      These observations are most welcome.

      In the post I clearly did not give the issue the full-length treatment it deserves and was undoubtedly selective in the extracts I quoted. But the only point was to be selective! And maybe I wasn't clear enough about the alternatives on the night of the 29th of September 2008 and the alternatives in the months and weeks leading up to the end of September 2008. It was the elimination (or non-progression) of alternatives in the run-up that did much to make the guarantee decision on the night itself "unavoidable" to a certain extent.

      I think if anyone was to become Brian Lenihan/Brian Cowen at 9pm on Monday 29th of September 2008 the outcome would be as it actually turned out but I do think there were a lot of different things that could have been done before that.

      1. Yes, the DoF did put in the "unavoidable" phrase in April '08 but there was no change to the "not regarded as part of the toolkit" phrase. I think the whole point there was to ensure that one kept ahead of any crisis to ensure that such an eventuality did not become unavoidable. Failure all round I think on that one.

      2. If they were "beginning to think of the unthinkable" why did they shelve bank resolution proposals? This was an alternative but lack of action eliminated it.

      3. I don't think ML were inconsistent. They were not unequivocal but the evidence suggests they did not support the guarantee. Saying that it "could be a mistake" seems fairly strong. They did say that it would be the "most impactful from the market's point of view" but that is pretty self-evident. Of course, the markets would be delighted when possible losses were eliminated via a sovereign guarantee. The decision had to take all stakeholders into account not just "the markets" and the "could be a mistake" note from the Sep 25 meeting is fairly strong. The conclusion to the ML memo does not even mention the guarantee. The section on the guarantee calls it an "alternative". ML don't put it up in lights but my reading of their memo is that they favoured the provision of emergency liquidity to Anglo (and possibly INBS) as their conclusion states.

      4. This point was made as more of a contrast with the NTMA rather than a description of the CBoI and does feed into the next point

      5. If the CBoI ajudged the banks to be insolvent that surely is reason enough alone not to give them a guarantee. A guarantee solves a liquidity problem. I think this is a key question. On what basis was the decision made not to forward the liquidity that Anglo was requesting? Who made the decision? What information was it based on? The provision of CB liquidity to Anglo was an alternative. Somewhere it was decided not to proceed with it. The David Drumm interview from November 2011 shows that he thought this decision was made on Merrion Street rather than Dame Street. That seems unusual and it would be useful to know what actually happened to these requests.

      The Promissory Notes only came into being in March 2010 so I don't think they have much relevance to the discussion of September 2008. As we know the ultimate public acceptance that Anglo was insolvent did not come for a long time after September 2008 (the PwC report from Feb 2009 is one testament to this). But if there was private acceptance that Anglo was bust (D Doyle notes on 25th September and possibly the reluctance of CBoI to provide ELA to Anglo on solvency grounds) then the elimination of alternatives to the guarantee is even more questionable.

      This debate will run and run and will continue to do even if the best and most revealing enquiry is undertaken. There really was no good options to dealing with Ireland's banking collapse but it is possible to see alternatives that may have led to a less worse outcome - all with the benefit of hindsight of course and there are no certainties that any alternatives would have been better.


    2. Thanks for getting back to me. Just two observations in that connection.

      1. The failure to proceed with bank resolution legislation in early 08 was a trumpeting of "be prepared" by "fears" - mainly fears of a leak that could upset market jitters, I believe (see FCT)- some laziness might also have been a factor. The decision was, on balance, a mistake, especially with the benefit of hindsight, and that FCT says. However, the "fears" such as they were cant be dismissed out of hand. After all, no one thought that there was a solvency problem looming so why take risks? The amendment to the DoF draft did not, in my judgement, reflect any serious move in the direction of rethinking that issue. The entire draft was somewhat more of an academic exercise (this was my impression from talking to the authors but this was not said explicitly anywhere)

      2. I continue to be puzzled as to why provision of ELA in Sept '08 would have made any difference ultimately. After all, it started to be provided in '09 as the deposits began to run out anyway. Doing it earlier would simply have brought the final bill more visibly forward (see FCT again for a discussion). The fundamental point is that unless we were willing to burn senior bondholders at some stage, the bill was going to end up the same, albeit with variations of timing and instruments. We were not willing to burn them in '08, nor in '13, nor at any time in between (including in '10, when the US/Geithner vetoed it and we had no negotiating power since the budget was bust independently of the banking costs and we had to play ball to get the funding to keep going)

      To be continued, doubtlessly...



    3. Hi Donal,

      1. I think this sort of issue is one of the keys questions to be answered. There is no doubt that there were alternatives to the blanket guarantee but for one reason or another they were eliminated from the decision process. There were decisions and/or inactions in the months preceding the end of September 2008 that limited the options available by the time the crisis reached its zenith. The Honohan Report was good on this when he concluded that:

      "Still, given the perceived lack of a solvency problem at Anglo (or the other banks) on balance a guarantee seems to have been the best approach, not least because no other clear and effective medium-term solution appeared available."

      I find the use of the word "appeared" interesting. He could have said no other solution "was" available. Of course, he also said "a" guarantee rather than "the" guarantee.

      2. This is correct. Again the Honohan Report goes through this in detail. I took the extract where he said that all the provision of ELA would have bought "breathing space". In my original piece you will note that I doubt that this space would have been used effectively. The point was not so much that the ELA would have been better (there were no good solutions) but that the use of ELA was an alternative and that understanding the decision-making that led to its elimination is important. My reading of the ML report is that there were (slightly) more in favour of the provision of liquidity to Anglo than the other alternatives they presented. They were far from uneqivocal on this though.

      As to whether the cost could have been lower. There was €11 billion of senior unsecured bonds and €2 billion of dated subordinated debt in Anglo on the night of the guarantee. I don't think any of the sub-debt matured during the guarantee and was subsequently subject to a haircut as part of a debt buyback. Of the €11 billion of senior bonds, €7 billion matured during the guarantee. It is hard to know what the attitude would have been to these senior bonds in the absence of a guarantee. The evidence shows that the then government showed no willingness to even contemplate burning senior bonds right through to November 2010. There are two ways to view this. One, is that it was essentially the period of the guarantee so talking about burning senior bonds was of little relevance until that expired and, two, it was the period that Ireland remained in capital markets and as long as that access was maintained senior bonds were "fundamental" to Ireland. The second was the expressed view of the Minister for Finance.

      However, when after the guarantee expired and Ireland no longer had access to capital markets the senior bonds in Anglo and INBS were "raised" as part of the negotiations for the EU/IMF programme. Brian Lenihan expressed the view in this doorstep interview.

      I have no idea how hard this issue was pushed and you are obviously correct that we had no negotiating power by that time.
      I don't think the final bill would have been much different if an alternative to the guarantee was chosen. Possibly between €5 billion and €10 billion but only if one believes a very different path would have been chosen. Anyone who suggests that deposits (Anglo's main source of funding) should have been haircut is engaging in revisionism. I don't recall any calls for depositor haircuts.

      I think we probably agree on more than we disagree on. If you look from the perspective of the 29th of September 2008 then it is hard to see how there were alternatives to the decision taken. However, to me at any rate, it appears that there were alternatives that could have been available on that night but they were eliminated for one reason or another. I think knowing how and why these decisions were taken is important.


    4. Hi Seamus,

      I think this interchange has been very helpful and we have indeed "converged" . I agree pretty much with your summary , especially the second last para . It is right to have the reasons for some things more out in the public , even if guys like me have heard these reasons/excuses already quite a bit via the inquiries.

      This is useful since too often people who perhaps should know better make statements to the effect that the guarantee was an enormous error, the cause of all our problems etc. Aside from being simplistic and wrong, this diverts attention and responsibility from the main underlying causes. It also can mean that addressing some of those causes (systemic failure- see FCT last chapter) is left aside.

      A very small point- I thought that taking all the banks about 20% of sub debt did fall due in the first two years of the guarantee so we maybe lost out a bit more there. Cowan stated this in his Georgetown speech and I was told independently that it was correct, but I never tried to check it directly. The 20% may well have included 0% for Anglo, as you indicate.


    5. Hi Donal,

      On the sub debt. Yes, some of it did mature during the guarantee. There was around €12 billion of sub debt across the six institutions in Sep '08 but only around €1.3 billion of that matured and was paid out during the period of the guarantee - and all of that was in BOI. See this PQ on the issue.

      BOI did undertake a haircut with sub debt holders so maybe there were losses from the decision to include dated sub debt in the guarantee. These liabilities shouldn't have been covered but I don't think the decision to do so, although a mistake, was an overly costly one.