NAMA Profit Projection Due to Faulty Haircut Calculation

I had thought there was something a bit odd about how the NAMA business plan extrapolated from the 50% discount applied to the first tranche to come up with its figure for the amount it would be paying for its full portfolio. After all, the fraction of Anglo loans in the first tranche was higher than for the total portfolio and these are likely to be worse in quality than the average loans from AIB or BoI.  Surely, the correct approach would have been to calculate this figure applying the institution-by-institution discount from the first tranche to the total consideration, assuming for example that the first tranche haircut for AIB is the appropriate figure to apply to the rest of AIB’s NAMA portfolio.

So, I did the calculation (here‘s the spreadsheet) and, lo and behold, it projects that NAMA will pay €41.7 billion for the portfolio, not the €40.5 billion they published in the business plan based on extrapolating from the first tranche aggregate discount.

It turns out that Jagdip had done the same calculations before me and discussed them here.  I agree with his comments there and believe this implies some pretty serious questions about the thinking that went into the production of this business plan.

Suffice to say, from here on, I will be assuming that NAMA are going to spend €41.7 billion, not €40.5 billion, and that the baseline case in which full long-term economic value is recovered actually implies a small loss rather than a profit.

Update: As Jagdip points out below, if the base case now becomes a loss of €200 million, this can be made up by not honouring €200 million of subordinated bonds. So the outcome is neither profit or loss. Of course, the “realise ten percent less than LTEV” scenario (known to some as the worst-case scenario) would now be a loss of €2 billion because the original €800 million loss in the business plan already accounted for the sub bonds not being honoured.

53 thoughts on “NAMA Profit Projection Due to Faulty Haircut Calculation”

  1. Hey guys,

    On blackberry so hard to get good look but at first glance, I have to say it is staggering. I can’t argue with your findings. Will have to think about it but at first glance it looks bad. I hope they have a good explanation. Good website on Nama by the way. First time I have seen it. Well spotted by the two of you.

  2. @Karl,

    I hope you get the opportunity to put this at a level to NAMA which might get a response.

    Because of the reported nature of the 5% subordinated debt presumably any loss of upto c€2bn would be put back to zero because the subordinated debt would not be honoured. So if the effect of this mistake puts NAMA €0.2bn in the red, presumably €0.2bn of sub debt wouldn’t be honoured.

  3. So, what are the possible explanations for this?

    1. The people that did the business plan can’t do maths.

    2. The people that did the business plan are deliberately trying to mislead us.

    3. Karl and Jagdip are wrong.

    Probably a reflection of how cynical I have become that I would find 3. above the most surprising..

  4. I am more worried about the nature of the apparent mistake than the financial implications. As you say, sub debt will cover it. When you put it with the joke of a business plan for Anglo and Nama changing historical documents as flagged on another thread, you have to worry about the ability of these people in important roles. These are multi billion euro business plans, not case studies for the leaving.

  5. It’s possible they use a second factor – maybe lender and loan type. E.g. There’s a higher portion of investment loans in the first batch than they expect in the overall portfolio.

  6. @ Ahura

    From page 3 of the plan

    “The Board is conscious of the commentary made about the average 50% discount paid on the Tranche 1 transfers and, for the purposes of its own Business Plan, it has applied this as a basis for its financial analysis and projections.”

    So it’s pretty clearly just an extrapolation of the aggregate discount.

    I can perhaps see why NAMA might not have wanted to get into a complex “multivariate” approach incorporating varying discount for different types of loans. However, extrapolating from the bank-specific discounts is incredibly simple and likely to be more accurate than just using the aggregate from tranche 1.

  7. @ Lorcan
    It is likely 1. If you were a knave you would not provide the fugures to be exposed so easily. We can take it that 3 does not apply.

    One thing is bugging me though. If subsequent tranches have more development land and less investment property in the mix surely the proportion that is performing should fall. Should not the discount rise as well as this is presumed to be lower quality stuff.

  8. @ Tull

    I agree that an alternative way to have done these projections would have been to use first-tranche figures for haircuts by type of loan. NAMA haven’t released these figures but they presumably have them.

    However, they have released the bank-specific haircuts and the variations in these likely reflect variations in types of loans (some banks being more into development land than others). NAMA just chose to ignore this useful information at their disposal (and already in the public domain) when putting their projections together.

  9. And if you assume that the loans transferred in the first tranche were slightly worse, so that the haircut to be applied to future loans is 5 percentage points better, then NAMA will spend €45.8 bn compared to their planned €40.5 bn.

    To be honest, a mistake like the one you’ve outlined is criminal in its sheer stupidity. It says a lot for the complexity of their calculations and certainly would give one less comfort that their ‘worst case’ scenario is anywhere near worst case.

  10. @ Tull

    52% of the first NAMA transfer were investment loans with that category only expected to account for 30% of total transfers. You would also expect the first transfer to be the “cleanest” in terms of complexity, personal guarantees, cross-shareholdings, asset backing etc. And of course property prices continue to fall.

    It seems highly likely that subsequent transfers will have greater haircuts than those accounted for in the NAMA business plan, the regulators stress test and broker models. This is not priced into the market value of the two main banks at the moment. On a seperate point, neither are “haircuts” on the NAMA bonds the banks hold which theoretically should be accounted for in the forthcoming EU bank stress tests.

  11. This is very strange becasue we have recently found out we are going to be paying less for Anglo’s loans! It now looks like Anglo will be taking a 55% haircut.

    There appears to have been a misapplication of the discount whereby the aggregate haircut was applied to the aggregate consideration, i.e. 49.6% discount applied to €81bn.

    Karl Whelan has applied the specific institution haircut to that institution’s loans, i.e. (AIB €23bn discounted by 42.3%, BOI €12bn discounted by 34.7%, EBS, €1bn discounted by 35.7%, INBS €9bn discounted by 58.2%, Anglo €36bn discounted by 55%) [I am using my own figures here.]

    Karl Whelan’s approach seems a more logical approach. This may or may not be valid depending on whether the banks applied the same criteria to smaller loans. I expect they did because they had a level of “leverage” that was acceptable to them. Only a loan by loan analysis will tell but it seems the best guess at this stage.

    I would caution however that those who were most aggressive and grew most might be those who ended up with the greatest leverage. Not everybody will have maxed out their leverage even if the bank was willing to offer the cash. Therefore, discounts may well get smaller.

    The recent news is that we are paying less and getting less cash-flow. However, schoolboy errors in the business plan are damaging to confidence. The reference to a “weighted average discount” of 50% is incorrect. I calculate the weighted average discount as 48.5%, (i.e., ((InstitutionA_discount/InstitutionA_Total_Loans)+(………))/81).

  12. @Zhou

    I agree. It’s the schoolboy errors in these reports that is the problem. Alan Dukes is in the Irish Times giving out about the coverage of the EU reaction to the business plan submitted in March saying that they have a new that is addressing their fears. He doesn’t seem to understand that the problem is that the first business plan was an embarassment with regard to the assumptions made. For a financial institution to send that to the EU does not reflect well on Anglo’s new management.
    Same goes with this NAMA business plan. I realise they are probably trying to keep things simple and we don’t know what is going on behind closed doors but messing up a basic assumption for no obvious reason does not reflect well on them.

  13. Is the same percentage of later tranches likely to be foreign (ie UK) property?

    My gut says bigger players were more likely to be in the UK market and more likley to be in tranche 1

    probably a red herring

  14. @Gavin S
    “I agree. It’s the schoolboy errors in these reports that is the problem. Alan Dukes is in the Irish Times giving out about the coverage of the EU reaction to the business plan submitted in March saying that they have a new that is addressing their fears. He doesn’t seem to understand that the problem is that the first business plan was an embarassment with regard to the assumptions made. For a financial institution to send that to the EU does not reflect well on Anglo’s new management.”

    I think it goes even further than this.

    If the Anglo business plans used assumptions that were so far outside the bounds of probabilities how are they supposed to be in any way credible?

    If you project that new anglo and old anglo will make combined profits akin to Anglo at the peak of the bubble of course this split is going to be least costly option to the state. If realistic assumptions are used then the cheapest option may be to wind the bank down.

    If you are willing to use any assumption without reference to reality then you will always be able to justify any decision.

  15. @christy

    The new Tranche 1 data shows that proportion of UK assets in T1 is substantially greater (38%) included than previously expected (30%).

  16. “The reference to a “weighted average discount” of 50% is incorrect. I calculate the weighted average discount as 48.5%, (i.e., ((InstitutionA_discount/InstitutionA_Total_Loans)+(………))/81).”

    …unless, of course, NAMA has applied weighting according to different criteria (a complex “multivariate” approach as referred to earlier by KW)…

  17. @zhou_enlai
    “…unless, of course, NAMA has applied weighting according to different criteria (a complex “multivariate” approach as referred to earlier by KW)…”

    Could be the case zhou, I suppose we cannot know with the level of information being released.
    At the moment what we are trying to do is akin to piecing together a 1,000 piece jigsaw with 500 pieces missing and no picture.

  18. Here is a related issue:

    If we ignore the the effect of discounting and simply add back the €2.64 billion expenses to the projected NPV in the orginal business plan we get a value of €7.44 billion as the total contribution to expenses and “profit” over the ten years.

    If, for the new plan, we add back the updated forecast of expenses (€1.6 billion) to scenario A’s NPV, we get €2.4 billion.

    The difference (€5.04 billion) is a indicator of the deterioration in the underlying outlook for Nama’s activities in just nine months.

    This is before considering the adjustment proposed by Karl and Jagdip or cosidering scenario C. If this trend were to continue then Nama is in real trouble.

  19. @Brian

    That’s an excellent good point but if I could express it in a different way. The “profit before operating expenses” has dropped from €7.44bn to €2.6bn (€1bn NPV + €1.6bn expenses?).

    So if it were not for savings to the operating expenses of €1bn (€2.64bn last Oct cf €1.6bn today) then NAMA’s base scenario “profit” would have been wiped out completely.

    Well done for drawing attention to how NAMA’s basic operation has deteriorated.

  20. BTW, we need be careful about calculations with NPV (future cash flows discounted) and numbers expressed in 2010 money.

    For your amusement would anyone like to calculate the NPV of the following cashflows using the 5.5% discount rate that is used in the new Business Plan.

    2010 +4bn
    2011 +2.42bn
    2012 +2bn
    2013 +1bn
    2014 +1bn
    2015 +1bn
    2016 -2bn
    2017 -2bn
    2018 -2bn
    2019 -2bn
    2020 -7bn

    Total net cashflow -3.58bn (ie NAMA spends more cash than it takes in)

    I wish I could offer a prize for the answer though the surprise or the satisfaction of knowing the result might be reward enough.

  21. +1.07964?

    The key here is that 0.945^10 = 0.57.

    Of course, counter to this example, NAMA is more likely to generate positive cash flows in later years

  22. @Karl,

    My calculator might be somewhat out but yes effectively the NPV would be €1bn even though the cash flow in nominal terms could be strongly negative.

    Without a split of cashflows by year, we don’t presently know what NAMA intend at all. And as any business plan reviewer will tell you, the early years should be more easily predictable than the later years and the further out in a plan horizon a figure is, the more risky it is.

    So NAMA could have a major negative cashflow and still have an NPV of €1bn. Though of course, it could have a major positive cashflow as well, but if it did I would have thought they would have let us know.

    This demonstrates that without a cashflow by year (and preferably a P&L and BS by year) we just don’t know what NAMA is planning to deliver.

  23. @Karl,

    In terms of NAMA’s cashflows by year in the new plan, we don’t know them but what do you take from the following “Debt reduction targets” on page 10

    Debt Reduction Targets
    Year % of NAMA debt paid down
    2013 25%
    2015 40%
    2017 80%
    2018 95%
    2019 100%

    Does this mean that 25% of 80bn will be paid down in by 2013 (or €20bn)? It is be a fair assumption that the more recoverable debt will be paid off first (and will have higher recovery rates). If it does *tend* to mean that, then does that not mean that cashflows are now more positive in the early years? Of course this is little more than conjecture based on an undefined term “debt reduction target” but it could mean NAMA are now forecasting stronger cashflow in the earlier years. Without a cashflow by year we don’t know. So I wouldn’t write off a strongly negative cashflow over the 10 years, which could still have a positive NPV.

  24. @ Jagdip

    25% of 40.5bn, right?

    The timing of these is sort of odd. I’d assume essentially nothing would be paid off this year, so that would leave a schedule for debt payoff of:

    2011-2013: 8.33% per year
    2014-2015: 7.5% per year
    2016-2017: 20% per year
    2018: 15%
    2019: 5%

    Make sense of that if you can. I guess we could make assumptions about annual fees and interest inflows and outflows and back out a cashflow model — sound like your kind of thing.

  25. @Karl

    “25% of 40.5bn, right?” Not necessarily, it could be the gross €81bn – as NAMA say in quite a few places in the plan they will pursue the entire amount. So 25% recovery by 2013 could be €20bn (100% recovery) then for the next year it might be 3.75% recovery and 3.75% write-off.

    As regards the rest, I don’t think it’s possible to predict cashflows because 25% “performing” has no meaning, nor is there any default rate or indication when NAMA will start selling (I hear rumours they will be selling this year as opposed to 2014 in the draft Plan) – NAMA CEO told the Irish Times this morning that quote “Twenty-five per cent is cashflow generating – the rest has some cashflow generating” If you said there were 100 people in a room and 25 had red jerseys and the rest had blue jerseys, would you think that 75 had blue jerseys? So the implication is that all of them are generating something. Do NAMA even know what “performing” means?

    http://www.irishtimes.com/newspaper/finance/2010/0709/1224274346854.html

  26. Given the inadequacy of Nama’s published business plan, I wonder does it have a different one for internal use and, if so, is it clearer or different?

    In either case, surely it should have been published? How acceptable is it for Nama to have two separate plans but keep one hidden? Surely, there should be only one plan.

  27. If you take the hypothesis that the biggest borrowers knew most about the industry they were in, had the competencies and capability within their organisations to deliver projects and manage the finished asset and in doing so earn their targeted returns – can we assume that things may get worse as we move through the tiers of borrowers to those that were not as ‘capable’ i.e. if the people that we thought were good got it so wrong how will the rest have faired? In this case the haircut may be bigger as things progress particularly as it may be that the best ‘development product’ is in the hands of the bigger borrowers and the opportunity cost for nama of not completing these projects (as opposed to the smaller projects associated with the lower tier of borrowers) is largest?

  28. @ Karl,

    I accept you point. When I saw no detailed projections in Plan 2.0, I’d assumed the couple of scenario outputs (table 4) weren’t worth analyzing. Typically summary outputs from cashflow models are too complex to step back into. It is possible that the “Amount of debt securities to be redeemed by NAMA” is illustrative and not taken from NAMA’s cashflow model.

    For example, what does “Amount recovered from Assets” include? Given NAMA plans to extend credit to complete projects – how/where is this being repaid? How much net interest income comes from these new loans? Too many unknown unknowns. Plan 2.0 is intentionally inadequate.

  29. @ BARRY T,

    the converse is probably true, hte largest borrower in hindsight knew least about what was going on . They just levered up, with the knowledge that the good times would roll. True they probably were in the best position to hire the best help but that no longer applies. I bet you will find the best quality in the middle tiers of the scale. Thes guys were probably trying to limit their exposure to an overheating market.

    Of cours the best talent of all is not in NAMA, these are the few that spotted the ponzi scheme and sold out. There are a few of those.

  30. This is a nonsense, which I see Gene Kerrigan of the Sindo citing as gospel. Not to mention Rachel English on RTE drooling on this latest expose of NAMA by the academic elite.

    For the purposes of the very broad brush NAMA business plan it doesn’t matter one jot whether one assumed a 50% discount or a 40% discount or a 60% discount. It doesn’t matter how spuriously accurate your spreadsheet derivation of that percentage.

    The premise has always been that NAMA will pay LTEV whatever their assessment of that on a case by case basis turns out to be. The key assumption in the business plan is that the assessment of LTEV would be borne out by experience and this is what matters, not whether in the end of the transfer process the haircuts turn out to be 40%, 50%, 60% whatever.

    Of course there would be some merit to the argument made here if NAMA blindly paid over 50% for the rest of the loans, but that was never going to be the case, it was always going to be a case by case assessment.

    Karl’s statement that the 1.2bn “error” has no contra is patently wrong. If indeed 1.2bn “extra” is paid it is because the assessment of LTEV is 1.2bn larger than that very broad aggregate assumption and the underlying premise of the plan is that this will wash its face.

  31. @Brian Woods II

    If NAMA produced a P&L, Balance Sheet and Cashflow by year in its Business Plan then it might be possible to test the effect of paying €41.7bn compared with €40.5bn. For my part, I am willing to concede that it is possible that if NAMA’s cost of buying the loans is €41.7bn instead of €40.5bn then it *might* recoup the increased costs through increased revenue and the net effect is nil.

    On the other hand, if the loan default rate in the new Business Plan is higher than the draft Business Plan – you might remember that they used an unsubstantiated 20% last October, 2009, then NAMA may have to rely on the sale of the assets backing the loans to a greater extent than previously planned. That being the case there is increased risk to *any* increased cost of acquisition. More defaults = more risk from managing and selling assets.

    What Karl and I independently drew attention to, was the fact that NAMA says it has based its new Business Plan on the evidence of tranche 1. There is no indication that there is any forward planning for any differences between future tranches and tranche 1 – gobsmackingly incompetent in my view, but whatever – so I therefore expect NAMA to calculate its plan accurately by reference to its own internal methodology. If calculating the amounts payable by reference to the haircuts *by financial institution* in tranche 1 shows €41.7bn, then that’s what NAMA should be showing, particularly so because NAMA should be trying to be prudent in its cost planning.

    There is an additional point from my end and that is that the Financial Regulator has told the banks to use the haircuts in tranche 1 to plan capital requirements. That means he has told the banks to plan to receive €41.7bn from NAMA (ie apply the tranche 1 haircuts to their planned NAMA loans). If NAMA’s €40.5bn is accurate then the Financial Regulator should be saying “sorry lads, but to be on the prudent side you need to assume you will receive €1.2bn LESS from NAMA and you need to explain where you going to patch that capital hole”.!

    NAMA made a mistake by reference to its internal methodology – pure and simple. Whether it has an effect on the NPV is debatable (in the absence of a P&L, Balance Sheet and Cash flow by year, I’d say the debate is near futile). Drawing attention to a €1.2bn mistake is at the very least public spirited. For me, the NAMA management honeymoon is over, the buck for the current Business Plan needs stop somewhere and Brendan McDonagh , via the NAMA Board, is the obvious candidate. And figures need to stop having the “back of an envelope” feel to them. As another “nerd” (Brian Flanagan) on here pointed out, if it were not for the €1bn reduction in operating costs from last October 2009 (and NAMA can’t blame the banks for inaccuracies in those last October) then the present central scenario €1bn would be largely wiped out. So I hope irisheconomy.ie continues to debate, examine, challenge and on occasion expose NAMA’s operations.

  32. @ Jagdip

    Congrats on being quoted not once but twice by Gene Kerrigan.

    My point is that Karl is totally overcooking the importance of predicting the actual haircut.

    A typical business plan will project,say, that it will buy 100,000 widgets for 10M and sell them at a margin of 5% making 0.5M profit. A nerd in the finance department points out that if we had based our projections on an analysis of sales per postal district we would be buying 8,523 more widgets easily wiping out the 0.5M profit. The nerd is of course more accurate in his prediction for purchases but has clearly drawn the wrong conclusions for the bottom line. This seems to be Karl’s argument though I see that you don’t quite take that line.

    Of course it is very easy for a Professor to hoodwink the likes of Kerrigan by producing a spreadsheet which appears to make fools of the NAMA bean counters. It is unfair and Karl knows it.

  33. @ BW2

    Let me see if I understand this: It is unfair of me to produce “spuriously accurate” (great phrase) calculations and I know it? Sorry Brian but you’ve completely lost me.

    Your point, as far as I understand it, seems to be that there will be an additional €1.2 billion in LTEV to be collected to accompany this €1.2 billion in additional payment.

    I don’t understand this at all. My point, which perhaps you just wish to misunderstand, is that the evidence from the first tranche points to NAMA paying €1.2 billion more than they are currently projecting.

    We have not been allowed any insights into how NAMA are currently valuing the loans but we do know that the valuations are also based on information from the first tranche — there is no reason to believe that the faulty payment projection has any counterpart in these valuations.

    To sum up, based on the first tranche information, I expect NAMA to pay €1.2 billion more than the business plan projects. I have no reason to expect a corresponding larger figure for the LTEV of these loans than reported in the business plan.

    Now from your perspective, safe in your firm belief that NAMA will “wash its face” (ugh) you might not care about whether any figures in the business plan are accurate. All the same, your criticism of those of us who would like the current estimates to be as accurate as possible seems a little misplaced.

  34. @KW that is so wrong!! JS Help! you know what I am saying.

    The premise of the business plan is that NAMA will pay LTEV on a case by case basis and that LTEV will be realised. Okay there is lots to debate there but that’s not the point.

    It really doesn’t matter whether the business plan projects it will pay 40.5bn, 41.7bn, 50bn, 30bn. Doesn’t matter, do you get that? The assumption is that what will be paid will be the correct LTEV, but that is a different debating point.

    Gene Kerrigan’s article had a cartoon with the manager of the Bad Bank saying Bad Banks can be expected to get their sums wrong by 1.2bn. That is oh so misrepresentatitive of what your spreadsheet implies.
    You are simply showing that a more plausible extrapolation would give an LTEV 1.2bn higher but it is a complete non sequitur that ergo the business plan should be projecting a bottom line 1.2bn less.

  35. @Karl

    Or put another way. When this process is over I would be most surprised if NAMA pays over exactly 40.2Bn. I’d expect at least 1bn projection error either way. But no more will I say that NAMA will be 1bn better off if it turns out to be 39.2bn than the other way round.

  36. @ BW2

    I agree that NAMA could have written their business plan by working back from LTEVs to estimate what they were going to pay — which is the implicit business plan you have in mind. However, other considerations go into the haircut. For this reason, NAMA have decided to base their business plan on extrapolating from the first-tranche haircut. Feel free to disagree with that premise if you wish but that’s a different set of criticisms of the business plan.

    As it is, I feel we have little choice other than to evaluate the business plan based on its extrapolation from the first tranche to arrive at figures for (a) How much will be recovered and (b) How much will pay out. The methodology being used is such that poor extrapolations for (b) have no effect on (a).

    As you’d say yourself, do you get that?

  37. Bw2
    Sure what’s a few billion anyhow, is that your argument?
    Isn’t Nama togit and so it can and will pay what’s needed to keep bondholder skin in the game and stop or delay the evils of state ownership.

  38. @Karl It seems to me from reading the scenario matrix on page 25 of the latest plan that everything is calculated by reference to LTEV. Debt issued is equal to LTEV and the asset scenarios are (A) LTEV recovered (B) LTEV plus 10% (C) LTEV less 10%. It scarcley matters what the case by case LTEV actually turns out to be as the value of debt issued and assets recovered will move in tandem in these scenarios. You are assuming that NAMA’s projection of asset cashflows is independent of the debt they issue in LTEVs, it is clear from this scenario matrix that they are wholly inter-related.

  39. @Brian Woods II

    Sorry, I don’t think I will be much use to your argument. My view is that the €1.2bn extra cost *may* put the €1bn “central scenario” NPV at risk, though I concede that without better financial statement information in the business plan it is not at all definitive.

    Upon reflection though, I am satisfied that the suggestion that it wipe out NAMA’s NPV is given prominence. NAMA are getting too easy a ride with the new Business Plan. Putting aside the obvious absence of the usual information you’d expect from a business plan, the banks are being blamed 100% by NAMA and the DoF for providing information last year which turned out to be inaccurate. Neither NAMA nor PwC who stress-tested loans have any responsibility it seems.

    Whose responsibility is it to plan operating costs? NAMA you would think. They got those wrong to the same relative degree as “performing loans” (40% to 25% is a 40% relative drop – €2.6bn to €1.6bn is also a 40% relative drop). Of course what wonderful people for reducing expenses, except does anyone believe the reduction is down to NAMA’s prowess or negotiating skills (if it was then let’s move Brendan McDonagh to the HSE or Dept of Education) – it’s more likely to be down to not being able to plan properly.

    And for me the term “central scenario” is new. You’d normally say “base plan”, ie that which you consider most likely. Why would NAMA depart from the usual naming convention (usual to me and also it would seem to NAMA if you read their guidance notes for developers). This has the whiff of spin and abnegation of responsibility to it. So for me I have happy that what is plainly a mistake (nerdy, spurious, bean-counterish, whatever) be given prominence.

    Given that NAMA are no longer answering journalists questions about the Buisiness Plan (See below), I also hope the Committee for Finance and the Public Service (which sits until the end of July) summons Brendan McDonagh to account for his Business Plan. According to the Indie ALL tranches might be transferred by the end of Sept so now is the time to examine, influence and change behaviour which might have ramifications for decades.

    “These details were included in the draft business plan so why have they been omitted now? We’ve already learned of a sharp drop-off in performing loans – largely because the banks have continually lied throughout the whole process and are still doing so – but the extent to which this is due to short-term income coming to an end is unclear as Nama isn’t saying, despite requests from the Sunday Tribune for the information.” http://www.tribune.ie/business/news/article/2010/jul/11/neil-callanan-nama-vs-the-rest-of-europe/

  40. @JS I will agree with all that NAMA projection skills look sloppy. But my basic point is that Gene Kerrigan is totally wrong in saying that because NAMA is more likely to issue debt for 41.4 than 40.2 that there is yet another hole of 1.2 in its plans. He based this assertion on his interpretation of Karl’s spreadsheet.

    @BL Obviously I do not communicate very well. I am not at all saying that a few bn here or there is negligible. The fact is that NAMA now projects to pay c. 10Bn LESS than it planned last November. Clearly this does not mean they will make an extra profit of 10Bn. Why then does the comparison with the 40.2 made in March imply a loss of 1.2Bn?

  41. I am going to be a bit lazy here and ask a question rather than read the NAMA business plan.

    It is estimated NAMA will pay €1.2bn more than the business plan suggests if the per institution discounts remain constant. It is suggested that this is a more appropriate estimate for inclusion in the business plan. It is suggested that the difference of €1.2b wipes out the profits.

    My question is whether the cash-flow and income has been under-estimated by extrapolating the aggregate figure for the first tranche in the same way that the consideration has been under-estimated? If it is the case that the price paid and profit were both reduced by extrapolating aggregate figures, then setting the addtional €1.2bn against the profits may not be valid.

  42. @ Zhou

    The quick answer to your question is that there is no evidence of any relationship between the underestimation of how much is to be paid and the quite separate valuation process. I tried to explain this to BW2 above. He has decided not to agree with me. I have decided to agree to disagree.

  43. Umberto Eco has said that being able to search through documents using technology means people don’t bother reading the whole piece and therefore don’t get the full picture. The below extracts are based on a word search in Adobe Acrobat – I have not read the full document.

    However, I think they show that the assumptions used to estimating cash-flow and value may be consistent even if a oper institution basis might have been better. It is not valid to subsequently apply a different assumption to one variable (price paid) and not to the other (income/profit).

    The big worry is that the business plan does refer to 50% as a ‘weighted’ average discount. This suggests there is more to this that meets the eye. Unfortunately, the small print is not available.

    “Based on the Tranche 1 weighted average discount of 50%, it is assumed that NAMA will issue €40.5bn as consideration in securities to the participating institutions for the €81 bn of eligible assets acquired from them – €38.5bn (95%) in senior debt and €2bn (5%) in subordinated debt. The overall NAMA discount on the €81bn portfolio will depend on the results of a loan-by-loan valuation.”

    “25% of eligible assets in the portfolio are assumed to be cash-producing during NAMA’s life (this assumption is based on analysis of the first tranche of acquired loans).”

  44. @Zhou

    As Karl points out, it is my read that any difference between what actually transpires as the NAMA discount and the NAMA current projections of that would flow through on both sides when NAMA do their final business plans at the end of this exercise.

    What we do know from the adjustments already made by NAMA to their bottom line projections is that this is necessarily a horrendously unscientific exercise. As a betting man, if I was asked to make a spread on the ultimate result, I would say somewhere between 15Bn negative to break-even.

  45. @ Zhou

    I wouldn’t worry too much about the presence of the word “weighted”. After all, if you go through and weight the individual bank haircuts by the share of the bank in tranche 1’s transfers, then you will get the aggregate haircout of 50%. So that is probably all they mean.

    My point is that if you weight the individual bank haircuts by the share of the bank in the total amount being transferred, then you will not get 50%.

  46. @KW

    “My point is that if you weight the individual bank haircuts by the share of the bank in the total amount being transferred, then you will not get 50%.”

    I agree. I did the calculation too and did not get 50%.

    However, I wonder if we did the same calculation for cashflow would it work out at higher than 25%. Perhaps the poor Anglo assets are depressing the average cash-flow down as well as the average discount.

    The loans will be bought one by one. Any assumptions/estimates are for illustration purposes only. As long as they are applied consistently then they are valid, i.e. if cashflow is calculated using the T1 gross average then using the T1 gross average for price paid is appropriate.

    Apart from the paucity of information in the business plan, the real story is the deterioration in cash-flow. This is big news because it negatively affects value apart from LTEV. It also reduces NAMA’s capacity to wait for values to recover if there isn’t sufficient cashflow to cover the cost of funds.

    We don’t know how this reduction in cash-flow breaks down across institutions and asset classes. We also don’t know how much of it is due to deterioration since the valuation date though we know that has happened.

    We also don’t know how accurate estimates of cash-flow in the valuation are proving to be and whether they need revision. Worst of all, we don’t know what “cash-flow generating” or “income generating” mean. No comment has been made about revising the valuation date either.

    The €1.2bn is a side issue. Setting it off against NAMA profit is open to attack. Focussing on the €1.2bn is to allow oneself to be led down a blind-alley imho. I woudn’t be surprised if those who drafted up the business plan anticipated these criticisms of the €40.5 estimate and left it in to draw attention away from other more fundamental lacunae in the information.

  47. @Karl,

    From yesterday’s Sunday Independent, though reading it again the “process” might refer to the third tranche as opposed to the transfer of tranches – in my mind the previous estimate was that t3 would be transferred in July so I took the statement “will be handed over to NAMA” in August to mean that t3 would be completed in August. That may be the case but upon reading the article again, it is ambiguous so apologies for the unequivocal statement.

    http://www.independent.ie/national-news/doctors-lawyers-loans-go-to-nama-2254201.html

    “New figures obtained by this newspaper reveal that in the wake of the first tranche of loans totalling €15bn back in April, which was made up of the 10 builders who owed the most money, the second tranche of loans worth €13bn, from other developers, will go into Nama by the end of this month.

    Then in August, the third tranche of loans — totalling about €11bn worth of loans — including those relating to syndicates of professionals will be handed over to Nama. It is not yet clear when that process will be completed but it is estimated it could be by the end of September.”

  48. @ Jagdip

    The figures in the Indo piece refer to original book value (€15.3 billion in first tranche bought for €7.7bn)

    15+13+11=39, so this is just under half that 81 being transferred by September.

    I’m sure the banks wish it was faster — they could probably do with a few more NAMA bonds by September.

  49. The projected results for NAMA can at best be said to be somewhat educated guesses.

    I do not believe NAMA will make a profit or even be close to breaking even.

    That being said, the price paid for the assets will affect how much recapitalisation is needed and that should impact how much equity stake the Irish state will get in the banks. However, it does seem that the Irish state is determined not to get value for money by getting significantly less equity stake than what seems justified by market capitalisation.

    Overpaying is only part of the transfer of wealth, the other part is not getting equity. NAMA will make losses, equity holders in banks can make profits. End game has always been about equity in the banks.

Comments are closed.