The EU’s Relaxed Approach to Bank Stress Testing

The newly released stress test of selected EU-area banks by the Committee of European Bank Supervisors (CEBS) is flawed in its methodology and the results are not a reliable indicator of EU bank sector soundness. A stress test should examine the impact on net portfolio value of extreme but plausible shocks to the key variates explaining net portfolio value.  The CEBS report states proudly and repeatedly that it uses extreme but plausible shocks, and this is true, but it ignores the key-variates criterion of a well-designed stress test.

We know that historically some of the key drivers of systemwide bank failures are falls in asset prices (particularly property prices), sovereign debt defaults, and liquidity freezes. Sovereign debt defaults and liquidity freezes are explicitly excluded in the CEBS test. Falls in asset prices are included only to the extent that they are generated by the two shocks used in the test: a fall in GDP growth rates and an increase in bond market credit spreads. This specification gives rise to a very relaxed stress scenario.

Consider for example the case of Ireland. The baseline one-year property price change for Ireland in the CEBS model is minus 13 percent. This is reasonable as a baseline case although it seems a bit optimistic. What is not reasonable is the supposedly extreme-but-plausible adverse scenaria – a one-year decline of minus 17 percent! With most independent analysts predicting continued big falls (e.g., Morgan Kelly, Kevin O’Rourke, Ronan Lyons) this should be substantially larger. The extreme but plausible scenario should have a fall of minus 50 to minus 60 percent over the next two years; minus 17 percent in the first year is not extreme enough for a stress test.

Similarly, in the cases of other vulnerable markets which have experienced big property price bubbles since the advent of the Euro, the report’s adverse scenario relies on much too mild property price declines. For example, the extreme-but-plausible adverse scenario for Spain uses a property price decline of only minus 5 percent.

The CEBS report explicitly excludes the possibility of a Greek sovereign debt default. The argument made in the report is that such an event is rendered implausible by the creation of the EU bailout funds, in particular the 110 billion joint EU-IMF fund for Greece and the 440 billion European Financial Stability Fund available to all members. Yet Ken Rogoff was quoted only one week ago (PBS Newshour 19 July 2010) stating that he believes a Greek default is not merely likely but, over the medium horizon, is virtually inevitable: direct quote: “They are going to restructure their debt, they are going to default.” Rogoff’s views are widely respected around the world and he is not inclined toward exaggeration. How can the CEBS describe as implausible a view that is so widely held by so many respected analysts? Does not this view at least qualify as extreme-but-plausible?

There are two problems with the CEBS report’s claim that the bailout funds have rendered a Greek sovereign default implausible. One, the bailout funds are supposed to provide conditional support, premised on Greece (or other troubled sovereigns) continuing with the painful austerity measures laid out in the bailout fund agreements. Does the CEBS believe that the conditionality in the bailout fund agreements is not credible, and that the bailout funds will be accessed in times of panic now matter how the defaulting government behaves? If so, this shows a refreshing lack of faith by one EU institution for the promises made by another.

The second problem with the bailout funds is that they are not extensive or long-lasting enough to permanently eliminate the risk of default. These expensive big funds will serve to slow down a Euro-area sovereign default, but not to prevent one eventually. In the interim, the funds will be another big subsidy to the banking industry, and might serve as a juicy source of profits for hedge fund managers willing to bet against the EU political establishment.

Here is my suggestion for an alternative specification for an EU-area bank stress test. It uses a four-year rather than two-year horizon. It relies on two extreme but plausible shocks: a Greek sovereign debt market default/restructuring and country-specific average inflation-adjusted property prices falling to their 1999 levels. Note that the second shock requires bigger property price falls in countries like Spain and Ireland that experienced property price bubbles post-1999 and smaller falls for countries like Germany and France which did not have these bubbles.

129 thoughts on “The EU’s Relaxed Approach to Bank Stress Testing”

  1. @Gregory

    Agreed that the residential property tests look weak. Here is a summary of the various recent projections/predictions and assuming the EU expect the bottom to be at the end of 2011

    Residential Property In Ireland Peak to Decline from
    Source Date Trough start Jan 2010
    EU Adverse Stress Test July 2010 46% 21%
    EU Benchmark Stress Test July 2010 42% 15%
    Kevin O’Rourke July 2010 70% 54%
    Moodys July 2010 45% 20%
    Goodbody July 2010 50% 28%
    Jim Power (1) July 2010 40% 12%
    S&P June 2010 41% 14%
    Moodys May 2010 46% 21%
    DKM May 2010 45% 20%
    David McWilliams April 2010 64% 47%
    Fitch Mar 2010 45% 20%
    Paddy Power (2) Jan 2010 40% 12%
    IT Pundits Jan 2010 38% 9%
    Anglo Plan (3) Nov 2009 47% 23%
    Morgan Kelly Jan 2009 80% 71%

  2. Surely the 13 and 17% falls are y-o-y falls, not peak to trough. Who is suggesting that property prices will be down 50-60% from year end 2009 levels? The banks have already taken the hit for earlier falls.

  3. @Gavin S

    The Permanent TSB Index at the end of Dec 2009 was 95.6 for national prices. At peak it was 139.5. 50% off of December 2009 would give you 47.8 which would represent a 65% (1-47.8/139.5) drop from peak. Morgan Kelly, David McWilliams and I suppose Kevin O’Rourke are projecting/predicting at that level.

  4. @ Gregory

    wouldnt a 50-60% fall in property prices take them well south of 1999 levels? Like im thinking more like 1990 levels, no? (if we have had a 50% fall already, this would imply a stess test looking at 80% eventual falls. This would imply average house prices around 65k).

  5. @Eoin

    Can’t lay my hands on pre-99 PTSB indices but the national index was 57.1 in April 1999 which would represent a 59% drop from peak of 139.5 in Jan and Feb 2007 and a 40% decline from the end Dec2009 of 95.6.

  6. If property prices have fallen 40% from peak as of December 2009 and we have a extreme-but-plausible fall of 75% from peak then the remaining fall (from the December 2009 base) is

    .60 x (1 + percentage fall) = .25

    giving a projected post-2009 decline of 58.33%.

    Morgan Kelly (UCD working paper December 2009) has a baseline forecast of 66.7% falls with larger falls plausible. He mentioned in his oral presentation an “extreme but plausible” fall of minus 80% which I have moderated to minus 75%.

  7. @ Eoin

    Please note that I said “inflation adjusted 1999 levels” and there was a moderate degree of Irish inflation during some of the earlier years of this decade. This increases the forecast average house price in current Euros.

  8. @Gregory

    As of December 2009 residential property prices as calculated by PTSB’s HPI had fallen from a peak of 139.5 in Jan/Feb 2007 to 95.6 – a 31.5% fall.

    If Morgan Kelly’s prediction/projection of say 75% from peak is examined then that represents a fall from 139.5 to 34.9 (139.5*0.25). To get from the end Dec 2009 index of 95.6 to 34.9 requires a further fall of 63.5% (95.6*0.635=60.7; 95.6-60.7=34.9).

    Morgan Kelly, David McWilliams and I suppose Kevin O’Rourke have projections that are substantially worst than the EU’s adverse scenario. Fair enough you might say, they are extremes but others such as Moodys, Fitch, DKM are also at the adverse scenario. And the government’s own central projection in the Nov 2009 Anglo restructuring plan was also worse than the adverse scenario.

    So yes, by reference to other projections/predictions the EU adverse case is not very adverse at all.

  9. @ Jagdip Singh

    Thanks for the careful and well-document analysis, it is much appreciated, especially because you agree with me!

    Your careful derivation of an adverse-scenario 63.5% fall from December 2009 matches well with my rough-work adverse scenario 58.33% fall. Your figures is better documented than mine but use essentially the same source.

  10. Sorry to repeat this question in two threads but why were only AIB and BOI included in the CEBS?

  11. @ G & J
    So lads, am I right in saying you have plucked a 75% peak to trough fall in property prices from the ether and calculated with enormous precision that we have another 60% to go or to be precise 58.3% to 63.5%. This seems to me to be as plausible as the adverse scenario from the CEBS. r put another way, all these forecasts are made up.

  12. @Rob S

    Because they are the biggest. The sample tried to capture 50% of each national banking sector expressed in terms of total assets.

  13. @ Rob

    from CEBS:

    “In each of the 27 Member States, the sample has been built by including banks, in descending order of size, so as to cover at least 50% of the respective national banking sector, as expressed in terms of total assets.”

    This also included foreign subs operating here, ie if (say) AIB and BOI had 40% of the market, and Ulster had 10%, they stopped at AIB and BOI as Ulster would also be included via the RBS figures.

  14. How many of these housing price forecasts have been plucked from the air?

    The average includes a glut of shobox apartments and at the other end of the market, expensive houses that went to ridiculous levels?

    Commercial property returns fell 1.4% in Q2 but there was a slight increase in Q1.

    A forecast over the medium term to say 10 years is more realistic than say for two years.

    During the boom, Irish land prices became the most expensive in Europe, partly because so little farm land came on the market.

    As regards housing, a small volume of sales now determines value and to factor in a steep plunge in prices in the short-term would likely realistically require a scenario like a meltdown resulting from quitting the euro.

    Sticky prices take many years to fall and that may well happen with unemployment high over a long period.

    In 1981, the avg s/h house price in Dublin was €42,193 and €51,450 in 1986.

    Inflation was high in the first 2 years of this period and then fell.

    In a stress test, a Greek default is reasonable but house prices falling to an avg Dublin price of €210,610 is based on a return to recession?

  15. @ Rob

    if you add the Irish subs of Danske, Rabo, HBOS and RBS to the AIB and BOI figures, then you’re probably looking at closer to 70% of all Irish banking assets being covered. Dont know how they treated the non retail subs operating here though, specifically something like DEPFA? If this was treated as being “in” the Irish figures, you’re probably looking at 80-85% of “Irish” banking assets being covered.

  16. @ROB

    As far as I know, the stress test covered a sample of 91 institutions covereing 60% of EU bank assets. So it was not exhaustive. I assume siz would rule out IN. I assume Anglo was left out because it would not be possible to conduct a stress test unless you had a starting balance sheet and a P/L projection for the next two years.

  17. @Eoin

    As Depfa is a 100% owned sub of Hypo Real Estate and the tests were carried out on a consolidated basis at the highest levels of the Group, I assume Depfa is included in Hypo’s and therefore Germany’s figures.

  18. @Michael Hennigan

    “Commercial property returns fell 1.4% in Q2 but there was a slight increase in Q1.”

    What do you mean by this and more importantly what significance does this have to property prices?

    Commercial capital prices are the subject of a few indices but the NAMA LEV Regulation seems to treat the IPD/SCS index and the JLL IPI index as superior to other sources. Let’s take JLL’s index

    http://www.joneslanglasalle.ie/ResearchLevel2/Index%20Report%20Q2%202010.pdf (free registration required) with the Irish Times report if you get frustrated with JLL’s website and registration process http://www.irishtimes.com/newspaper/commercialproperty/2010/0714/1224274649342.html

    Commercial capital values fell 2.2% in Q1 and 4.7% in Q2 – ie the rate of decline is accelerating. So what has a improvement in yields got to do with anything? haven’t we been there before in Sept 2009 with Brian Lenihan calling the bottom because of record yields?

    To be clear, commercial capital values in the State are continuing to fall and at an accelerating rate based on Q2, 2010.

  19. The banks needed to account for at least 50% of the market they served. Where foreign banks constitute enough of a domesic market, that was considered enough. As noted above, AIB, BoI and foreign banks already included in their home markets constituted enough to meet the requirement for Ireland.

    The problem with this strategy is that the very worst banks in Ireland, Anglo and INBS, would have been excluded no matter what. As they were deemed systemic, we can only suppose that they on their own could have brought down the rest of the banking system 😉

    The result of the cherry-picking of the big is to exclude the dangerous. The vast majority of the banks closed by the FDIC in the US were small-medium banks. The lesson is that they are the ones likely to be over-extended, through stupidity or looting. They also constitute a significant counterparty risk, not individually, but in aggregate. To take the example from Anglo again, a bank that accounted for 23% of banking assets is going to be responsible for more than 50% of the total banking system losses. The relationship for INBS is even worse.

    The stress test sampling renders the overall exercise worse than useless. It is like only testing vegetarians for the effects of hormone reared meat on humans…

  20. I don’t think I should be listed as an independent observation/source of predictions since when the journo from the Tribune asked me the house price question I basically told him I agreed with Morgan (which I do).

  21. BTW, I completely agree with Greg’s points, and especially the point that a proper stress test would take seriously the possibility of sovereign default at least in Greece (as well as much bigger property price falls in places like Ireland).

  22. Cmon, people, what did you expect?

    This was a propaganda exercise , pure and simple.

    I’m sure that there was a fair amount juggling of the criteria / evaluation algorithms to make sure that just the right number of banks were failed, in order to give the impression that the majority were sound.

    And I love the way accountants borrow terms like ‘stress test’ from the hard sciences and engineering to give their inventions an aura of solidity.

  23. @ Kevin O’Rourke

    Everyone seems to be quoting you it is not just me but I guess that you have not written a paper on it yet. I have limited knowledge of property markets so I was just providing a list of my “betters”.

    You can quote me that this was a substandard stress test. But there are much better authorities for house price forecasting; I just try to read what has been written and form a reasonable judgement from that.

    @ Michael Hennigan

    Thanks for the thoughtful comments. Yes the CEBS adverse scenario certainly is meant to correspond to a recession of some sort — that is implicit in the GDP growth projections, etc.

    @ Tull McAdoo and Michael Hennigan

    In my limited experience (well I did write a book Portfolio Risk Analysis Princeton University Press, see chapter 10 section 2, and I did work for four years in the City of London implementing portfolio risk analysis models) the forecasts in stress tests are usually “picked from the ether” in some sense. If the correct values are known that is called historical simulation it is a different technique. Perhaps large trading desks (we did not have that many trading desk clients who would open up to us at that level of detail) use more refined techniques but the institutional portfolio managers with whom I worked just guesstimated what was a bad-ish but possible outcome. I at least cited from experts such as Kevin ..um.. well Morgan Kelly!

  24. @ Gregory

    Ah so you admit that a forecast of 70% HPD is plucked from the air. Did you pluck it in 2006 (a la Dr K) or just follow the herd in 2009. Also what are your lotto numbers please?

  25. @ tull mcadoo

    No you misquote me — in my response to you, I was quoting your earlier comment, and saying that in fact that is how it is often done in bausiness-application stress tests. I got the forecast indirectly from roughly the same list of forecasts shown above by Jagdip Singh. I did not know all of the citations he provides, but knew some of them and that was my source. So, no, in my case it was not “plucked from the ether” it was from reading experts plus analysis of “adverse scenario” low-end forecasting. In the phrase “plucked from the ether” I was quoting you and noting that such an informal “intuitive feel” procedure is not unusual in portfolio risk stress tests.

  26. Gregory

    “Your careful derivation of an adverse-scenario 63.5% fall from December 2009 matches well with my rough-work adverse scenario 58.33% fall. Your figures is better documented than mine but use essentially the same source”

    Just to clarify, are you forcasting a 58% house price fall from Dec 2009 levels.? Is this your own forecast or is this an average of other people’s observations? This would represent a 75% fall from peak? Also when did you make this forecast at the peak?

    @ Brian
    welcome back, hope you enjoyed your holiday. Pithy as ever I see.

  27. Indeedee tull. V refreshed. Tell us, where do YOU see the houses going, when? Again Gregory (as so many others) has put out a figure with some analysis, and the hurlers on the ditch carp up. So, put up or shut up….please.

  28. I agree that 17% seems low for an adverse stress test but I can’t see the logic in saying that house prices could fall 55-65% from dec 09 in 2011 and 2012. For that to happen, would house prices not only have to fall but at the same pace as 2008 and 2009? Is that a trend we are seeing? Would have thought 20-25% would be nearer the average of the predictions. Anyway, would have thought it was the unemployment and GDP predictions that the banks should be really concerned about.

  29. Brian,
    I have no knowledge of where house prices go from here. If the current economic persists along with tight mortgage market conditions and give the level of over supply, I suspect we go lower -probably peak to trough declines of 60%, but that is only a guess. I am only plucking a number from the ether or extrapolating a trend.

    What I am curious to know from Gregory is whether his forecast is model driven or is just the consensus plus a margin. I am not clear.

  30. @ tull mcadoo

    If we use your peak-to-trough decline of 60% and my peak-to-2009 decline of 40% then this gives a 2009-to-trough decline of

    (1-.4) x (1 – Y) = .4

    giving Y = 33.3% decline for 2009-to-trough. I agree as others have pointed out that this decline could possibly take more than two years. Keep in mind we are doing a stress test so this is supposed to be a “bad” outcome not a most likely outcome.

    My main point was that the CEBS test is supposed to use pessimistic forecasts for the key variates driving net portfolio value, and it focusses on the wrong variables.

    It was a “pessimistic” variant of a consensus forecast, reflecting what is needed in this type of application. You are supposed to use a potential bad outcome.

  31. Greg,
    Great post! I agree that the treatment of sovereign default is flawed. They allow for the fear of default to be reflected in increased risk premia, but rule out the possibility of default itself. This is quite bizarre and unnecessily marred the tests in my view.

    However, it doesn’t seem to matter that much for the Irish banks given the relatively low holdings of sovereign debt (though AIB does hold over billion of Irish government debt). While conceivable, I see a 75 percent fall in house prices as well out in the tail. The question of how far out in the tail you should go for the adverse scenario does not seem independent of the pass mark on the test. I see 6 percent for tier 1 as being relatively high, given that a regulatory minimum of 4 percent — itself well clear of insolvency. With all the bad news we have had to take on the banks — and especially on Anglo — it is hard not to be cynical. But I think we did get some modestly good news on Friday.

  32. John
    ~the problem is one of contagion/covariance. Right now a greek default/restructuring would instantato mark irish bonds down. How much is unknown but a lot.
    In addition the muted reaction of the corporate markets on these tests tells me no real information was released.

  33. @Brian Lucey
    An additional problem is selection bias. In addition to my post above on the proportion of banks in an economy tested, there is the problem of what was tested. Only the sovereigns in the trading books were haircut, even if the bank had masses more of the same sovereign in their hold-to-maturity assets.

    Now, selective default can mean many things, but it takes some stretch of the imagination to think that a state is going to only default on the debt in the banks’ trading books…

    I’ll repeat, these tests are worse than useless – worse because who is going to believe the ‘real’ stress test the next time? Wolf? What wolf? Ah, you’re always saying that.

  34. Hogan
    Indeed. I discussed this on Drivetime – ignoring 905 of the book makes no sense. I called them a Charade…..what more can I do? or you?

  35. The “Pass” Results for AIB and BoI in the EU Capital Stress Tests are very misleading….bordering fanciful, almost absurd!
    Consider….If AIB and BoI are indeed so well capitalised, then it should be very easy, at this stage, for the Irish Government to remove the State Guarantee on Customer Deposits.

    I think you’ll agree that, if the Government at this point even reduced the Guarantee on Deposits to a Deposit ceiling of €100K , there’d be an instant huge rush to get Deposits out of the banks!
    So much for the Banks being well capitalised!
    Stress Test Eurocrats…. GET REAL!!!

    In the case of AIB, the EU’s Capital Stress test assumed AIB raising the €7.4bn required re-cap by this year-end. This is more smoke and mirrors!
    But what’s new?! The Government has a proven 100% preference for mis-information and untruths!
    The Eurocrats might have considered giving both Banks the following new Grade… fianna “FAIL” !!

    By the way AIB’s correct immediate min re-cap requirements are €10bn (not €7.4bn stated by the Minister and the Central Bank and Head of Regulation in March 2010, based on Loans assets information supplied by the Bank at the end of 2009 and very early in 2010 which we now know was unreliably optimistic).)
    And BoI’s correct immediate min re-cap requirements are €6.5bn (not the €3.65bn stated by the Minister and the Central Bank and Head of Regulation in March 2010, based on Loans assets information supplied by the Bank at the end of 2009 and very early in 2010 which we now know was unreliably optimistic).)

  36. Beans, lot of beans. Tins of tuna. Rice too. Get some plastic boxes to store it in and some vermiculite (to soak up any moisture that might make it go off).

    And we can keep saying it, I suppose. Eventually the EU and the ECB will have to come up with a credible solution to over-indebted nations and badly invested ones. A little bit from column A and a bit from column B, I suppose – perhaps a “pay back the principle, but you’re on interest free terms until then”.

  37. @ Hogan

    i think its a bit harsh to say they’re worse than useless. One of the very important facts to come out of this is a full breakdown of soveriegn debt holdings held by each bank. You can use that info to make your own stress test scenario. And while only 7 failed, it seems like another 17 or so were either within 1% of failing (per results) or would have failed if banking book had suffered sovereign haricutting (per Citi). As such, a lot of conclusions can be drawn from the markets from this away from the headline 84 vs 7 scoreline.

    @ Tull

    dont you usually have a band of dastardly acolytes helping you out at times like these?

  38. @Eoin
    That’s fair enough; DYODD and all that.

    The main problem I see with it is the loss of confidence down the line when a bank that passed the stress test runs into difficulties. The Fed has ensured this doesn’t happen by defecating cheap money all over the US banks. Can you see this being the case in the EU? JCT in his TUGboat down the Danube casting wads of cash to each bank?

    At the moment, I can’t either.

    Which means that some bank that has passed the stress test is going to get into bother. And Handesblatt or Le Monde will run banner headlines along the lines of “Mon Gott in Himmel: Tout Les Landesbanken Sont Kaput”.

  39. @ eoin,

    “Brian Lucey calls Stress Tests a charade”. Another one for the archive. I am with you on this. The loan loss assumptions are reasonable across Europe. Sovereign Default is an academic obsession. Who knows if or when anybody will default or restructure. What the CEBS did was reasonable. The stress tests are not perfect but a charade is a bit of a stretch.

  40. @ tull mcadoo, Michael Hennigan, John McHale

    After pondering your comments and others, I agree that my casual suggestion (in considering the Irish example of a broader problem with the test) of a 75% peak-to-trough adverse-scenario fall for Ireland was perhaps too extreme. Perhaps 70% would be better? Or what would be a reasonable figure? But this slight change does not alter my main point which is that the property price falls in Annex 3 in the stress test are much too mild, and inconsistent with the GDP falls that they assume in Table 2. They focussed on the wrong variates (or rather ignored some key variates) and this gave a wrong impression.

    @ Bond, Eoin Bond

    Good point that this report did provide lots of useful risk information, as long as thoughtful market participants read “below” the over-generous picture it gives of EU bank sector soundness. That is why bank share prices bounced up a bit.

  41. Gregory,

    Just so I understand
    *you are taking consensus forecasts which are for peak to trough falls of 40-50% and adding on a “stress margin” to get to you 70% P2T decline.
    *you correctly observe that the 17% and 5% resi declines in 2010-11 are not overly stressful
    There is no problem with this approach. The only problem is the quality of the forecasts going into your model. In some cases it is difficult to see the basis for these numbers. Some were made at the peak and some are just trend following-“technical analysis”.

    Out from that, it is still unclear what the difference between 50% and 60% decline means to bank capital other than it is not a positive. Will it tip BOI over the edge and force nationalisation. Another commentator said the trend in GDP and unemployment might be more insightful.

    It might also be worth pointinh out the Mr ELderfield’s stress tests are said to be more onerous than the CEBS in respect of implied loan losses.

    Lastly, beware of overinterpreting. Yesterday we had the stress tests, last night we had more clarity on Basle 3 and this morning we had decent results from UBS & Deutsche. In addition there has been a creeping narrowing of spreads in Euroland without much ECB buying. In a market domminated by pessimism, optimism is staging a fight back for the mo. No doubt we will be told by Cassandra and her pals that this is all rigged.

  42. looks like some market people are not buying it. From Telegraph-

    “As analysts sift through the wealth of new detail from the tests, they are baffled by the chaotic criteria.

    “We have a ludicrous worst-case scenario that Greek house prices fall by 2pc in 2011: when you first read it you think their must be a typo,” said David Owen from Jefferies Fixed Income.

    Austria’s worst-case is a 2.7pc rise in house prices, or zero for Poland, and -2pc for Italy. Mr Owen said these assumptions would be demolished by a serious recession. Yet the tests assume that all eurozone states would contract at the same rate in a downturn. In reality, Club Med states and Ireland would almost certainly fare worse since they are already coping with the triple effects of debt-deleveraging, lost competitiveness, and fiscal tightening.

  43. Hat tip to podubhlain!

    Yes Mr Owen is confirming exactly one of my main points.

    My own examination of Annex 3 (adverse scenario house price percentage changes) made me think that the Irish adverse scenario forecasts in that table are not the worst of the lot. I noticed Spain and Poland but did not know enough to critically evaluate Austria and Italy as he has.

    So I suppose that these Annex 3 property price forecasts are almost all too mild for an adverse scenario.

  44. @ P

    It is amazing that the Torygraph managed to find somebody in the City with a Eurosceptic spin. Are you a Gaelgeoir that takes the Torygraph. That is eclectic!!!

  45. @tull

    Still need a good right wing perspective.

    just noticed RTE spinning at its most ludicrous – headline on article in business section
    “German consumer mood improves in August”

  46. As enlightened as the debate is, or can be, on irisheconomy.ie there seems to be a dearth of debate on commercial property. Maybe that’s because we all have homes and a passing interest at least in residential prices. But even the economists seem to ignore commercial property when holding forth with their views.

    The Society of Chartered Surveyors and the Investment Property Databank have released their indices for Quarter 2 for the Irish Commercial property market and are available free (without registration here).

    http://www.ipd.com/OurProducts/Indices/Ireland/tabid/435/Default.aspx

    The SCS/IPD index is sufficiently well thought of to be referenced by NAMA’s LEV Regulation as a source to be used for analysis and projection.

    The index shows that capital prices have fallen by 3.5% overall in Q2, 2010 (compared with a fall of 1.8% in Q1) and the Q2 fall is close enough to the 4.7% fall published by rival index JLL’s. To be clear commercial property prices are still falling in the State and the pace of fall is picking up by reference to the Q2 indices.

  47. @ POD

    In this case, the Ministry for Information is correct. The index which measures German consumer confidence did improve in May.

  48. This comment may have been made by others but surely one very obvious stress test for the two big irish banks would run as follows: would the consequences of an immediate lifting of the guarantee be disastrous? Anybody prepared to argue in the negative? I thought not. So, on this criterion, they are as bust as they ever were.

  49. @simpleton

    Perhaps a better question to pose would be what would happen if the authorities stepped back from the market everywhere. The global banking system would grind to a halt and the world economy would slip seamlessly back into recession.

    So our banks are bust without official support..big swing. Bar HSBC every body else’s are bust as well if you apply the same standards. The developed world is over leveraged & central banks are providing a cushion of liquidity & easy money to cushion the deleveraging.

    I did not realise you were a desciple of Andrew Mellon!!!

  50. @tull
    I was not suggesting a liquidation of the banks, just a thought experiment about stress testing. Tests were devised that the banks all, mostly, passed. This thread has been about whether or not the tests were designed to be passed. I guess we will never know.
    As market participants we can imgagine tests that the banks would not pass. But with such exercises, as with the EU tests, we can begin to posit a rank ordering of degrees of pass. So stress testing does, in my view, have some utility. The absolute level of stress doesn’t matter so long as we can gauge relarive rankings – pass or fail doesn’t matter. In a sense it’s like setiing any pass mark. I’m interested in the pupil that scores 30% and the one that scores 60% and the information that conveys. Not whether they have passed or failed according to some arbitrary standard.

    In so doing, we can probably imagine one more irish bank slipping into state ownership later this year. And draw other conclusions about other banks.

  51. Simpleton

    Agreed, the relative rank ordering in the tests is what is important. The two Irish banks are in the bottom quartile. BOI might have enough capital, has a cleaner balance sheet and probably will be able to issue govt guaranteed term funding pretty soon. AIB was only able to pass the test because it is assumed it willbe able to raise the required capital. Even I could pass an exam on the the basis that I would answer the corrections correctly at some point in the future.

    It amazes me that nobody has commented on the fact that the Irish Permanent TSB did not even take the exam. As the anchor of the Third Force surely we should know how much capital is needed by that hobbled entity. It is almost as if we dare not ask the question for fear of the answer.

    There is no discussion of the strategic issues facing banking in this country. We have one institution that is barely fit for purpose. Two of our clearers are owned by foreign entities that are busy deleveraging. AIB is in drydock. IP-TSB/EBS/IN is a collection of bin ends with a broken business model and inadequete capital. Anglo is an out of control zombie quango. Above this rag bag collection, credit allocation decisions are being made by bureaucrats and politicians. Yet this site obsesses about who is really funding NAMA, how big the house price fall will be & whether a stress test is really a stress test.

  52. Indeed, the stress tests published by the CEBS have received criticism and in my opinion, rightly so.

    A problem with the stress tests – the main one pointed out by Wolfgang Munchau in the FT – is that there was no provision for the possibility of soverign default. I think that this is a very valid point. OK, the possibility of a Greek default may be unlikely BUT it is plausible that the event could occur. The Basel committee recommended in Jan 2009 that: “stress tests should be geared towards the events capable of generating most damage whether through size of loss or through loss of reputation. A stress testing program should also determine what scenarios could challenge the viability of the bank (reverse stress tests) and thereby uncover hidden risks and interactions among risks”.

    Furthermore, the Basel committee also made the following recommendation: “stress testing programs should cover a range of scenarios, including forward looking scenarios, and aim to take into account system-wide interactions and feedback effects”.

    It seems to me that the 2 recommendations that I have mentioned above have been put forward but what actually took place under the stress tests reveals some inconsistency. i.e. they did not use a sovereign default scenario in the tests.

    On the point about stress tests having to be ”forward looking” (as opposed to backward looking VaR), some questions have to be asked. Was historical data used (i.e. objective probabilities) – in which case the tests are not forward looking – or were the tests forward looking in which case subjective probabilities would have had to been used? Being “subjective probabilities”, this obviously asks if the stress tests are of any value.

    Another point that has been well voiced is that assumptions that variables are normally distributed (I’m not sure if this was the case with the CEBS tests) lead to unrealistic results. For example a 5 standard deviation in a market variable happens about once in every 7,000 years under a normal distribution…. but in practice, it is not uncommon to see a 5 standard deviation once or twice every ten years. A specific example is October 1987 when the S&p500 moved by a standard deviation of 22.3. Closer to home, in the UK yields on 10 year bonds moved 8.7 standard deviations on 10 April 1992.

    Here is an interesting issue that I would like to bring to the table – although I am not certain if this is applicable as I have no real expertise in the area of Finance nor do I know the exact details of the CEBS stress tests but it may be worth noting – is that because Banks wish to keep their regulatory capital as low as possible they could in fact use ”mechanistic hedging” as what happened when the Danish Financial Supervisory Authority introduced its ”traffic light” system in June 2001. What happened then was that investment banks developed products – so called traffic light options which would generate pay-offs if the companies were to fail the scenarios being tested – to help companies to keep a green light status. The companies being tested could simply purchase the products and ensure that they would attain green light status, rather than red light status, thus companies were not subject to extra monitoring as a result from the stress tests. If this did/did not happen or could have happened in the case of the CEBS stress tests I am not sure but this does point out that stress tests are far from perfect programs.

    They are just some of my thoughts on this topic and I would like to add the following 2 points for consideration:-

    Eight banks scored a tier 1 capital ratio of 6% – the minimum to pass the stress test.

    Citigroup carried out their own stress test on European banks’ entire sovereign exposure. They extended the CEBS stress haircut to banks’ valuation of all their debt holdings (not just those held in the trading book). Under these conditions 24 banks failed including AIB (5.5%) suggesting that AIB would require a capital injection of €343m.

  53. For those of us less convincd than JTO of our unique position at the top of the evolutionalry tree, there is reassuring comfort in the response of the Irish authorities to our banking crisis.

    “In a financial panic, the government must respond with both speed and overwhelming force. The root problem is uncertainty—in our case, uncertainty about whether the major banks have sufficient assets to cover their liabilities. Half measures combined with wishful thinking and a wait-and-see attitude cannot overcome this uncertainty. And the longer the response takes, the longer the uncertainty will stymie the flow of credit, sap consumer confidence, and cripple the economy—ultimately making the problem much harder to solve. Yet the principal characteristics of the government’s response to the financial crisis have been delay, lack of transparency, and an unwillingness to upset the financial sector. ”

    All that was written by an ex chief economist of the IMF about the US government. But, of course, could have been written about any one of a number of jurisdictions. The great pity about our ‘me too’ non-policy response has been the missed opportunity. There was a uniqueness about the problems that we faced: they were plain, vanilla and very easily understood. That meant that political courage could have seen them dealt with differently and speedily. Still does actually. As someone else asked this week: how much is too much (bail out money) for Anglo? Can the defenders of the Bond Brotherhood answer this question?

  54. Simpleton,

    That argument was relevant last year. Anglo should have been nationalised on the night of the guarantee and the process you outline started then.
    There are now now virtually no bond investors left in Anglo other than those issued under the guarantee. I repeat for effect, there are virtually no outside investors holding senior Anglo debt. If you want to discuss the creditors taking a hit please ring the ECB and/or its puppet in Merrion St. There is noboby left to burn at the stake. The 25bn or 30bn that Anglo will cost has gone to money heaven.

  55. @Tull
    So the state remains on the hook solely for Anglo’s deposit book. That will cost what it costs. Why incur the additional costs (admitedly more symbolic than anything else) of paying Dukes & co their salaries?

  56. @ Tull

    i think the IPTSB is less a capital issue and more a funding issue – the old business model of wholesale funding is completely broken. The issue for them is how to pivot, relatively quickly, from being wholesale funded to being much more retail/business funded. They have capital issues/requirements for sure, but they would seem to be much more transperent and their losses much more vanilla in nature than for AIB or BOI. But after the Third Force consolidation happens, then yes, a public stress testing might not be a bad idea to assure the markets.

  57. @Tull
    “It amazes me that nobody has commented on the fact that the Irish Permanent TSB did not even take the exam. As the anchor of the Third Force surely we should know how much capital is needed by that hobbled entity. It is almost as if we dare not ask the question for fear of the answer.”
    Maybe thats because those who thought about this before and after read the CEBS website which stated which banks would be included. As has been explained in some detail elsewhere and here, the 50% of market cutoff meant AIB/BOI were all that made that cut.

    @Simipleton
    “As someone else asked this week: how much is too much (bail out money) for Anglo? Can the defenders of the Bond Brotherhood answer this question?”
    Interesting you should say that. A long tweet discussion between myself , others and Dan Boyle (GP Spokesperson on Finance, President of GP, and Government Senator) last night finally winkled out of him that in his view €25b was too much. As we are close to that now at 22.something plus the possibility that as/if losses go over that the CB 11.whatever comes into the losses Dan seems to be saying that GP are coming to the end of their tether. But, personally, I will believe that when I see it.

  58. @ Simpleton

    I will forward your query to Dep Lenihan next time I meet him at the imaginary parliamentary party meeting. Amazingly, Anglo is still recruiting staff.

    @eoin,
    Wrong on capital, I am afraid. Permo is short about 1bn in capital and there is an unspecifued hole in the EBS. When you say funding problem it seem so trivial and solvable. However a loan to deposit ratio of 300% is an anachronism in the modern world. In addition they have a profitability problem-3/4 mortgages are trackers.

  59. @ simpleton

    i didnt say they dont have a capital problem. I said thats the lesser of their problems. Its also far more transperent due to the vanilla make up of their assets.

    Also, two quick clarifications: tracker mortgages represent 60% of their book and loan to deposit ratio is 245% and falling. Bringing in EBS and INBS would see this drop below 200% i believe.

    I’m not “trivialising” their problems, but there’s no need to exaggerate them either.

  60. @Tull
    ““Brian Lucey calls Stress Tests a charade”. Another one for the archive. I am with you on this. The loan loss assumptions are reasonable across Europe. Sovereign Default is an academic obsession. Who knows if or when anybody will default or restructure. What the CEBS did was reasonable. The stress tests are not perfect but a charade is a bit of a stretch.”

    Well, ok. Not to compare myself to him but Nouriel Roubini is not terribly convinced on Stress tests. And whats that he says is missing…..lets read on …

    “Several aspects of the testing process make us skeptical about the results. We’re still not convinced of the European banking sector’s resilience and therefore are concerned about the scale of the potential liabilities that stressed sovereigns need to backstop amid a low-growth environment.

    In a Strategy Flash available exclusively to clients, we address a crucial deficiency of the stress tests: the absence of a sovereign default scenario. Optimistic commentators point to the rebound in U.S. markets after the Supervisory Capital Assessment Program (SCAP), which faced significant criticism, as Europe’s equivalent is now. But while the U.S. SCAP tests modeled the key concern of the market—future property risk—and forced banks to recapitalize, the European tests did neither. The former would have meant stress testing sovereign debt in a default scenario. The only valid argument for not doing so is that the ECB could monetize the debt, with the €440 billion European Financial Stabilization Facility (EFSF) serving as a fiscal fallback plan for sovereigns whose solvency concerns limit their bank backstopping capacity. But in the event of several sovereigns drawing on this fund—which we see as not entirely unlikely—the guarantees would be worthless.

    We also take issue with the Tier 1 ratio of 6% the CEBS used, which we see as a poor benchmark that can hide hybrid instruments that have proven to lack loss absorption capacity on a going concern basis. What’s more, while only seven of the 91 lenders failed to pass this bar, there were at least 10 other marginal fails, with Tier 1 ratios of 6.3% or less. Had the threshold been 7%, 24 banks would have failed the test, notes an RGE Critical Issue parsing the results.

    The capital shortfall figure likewise warrants closer inspection: For Spain, it includes €14.4 billion of public funds committed and pre-approved before the stress tests. Effectively, the tests rely on ongoing government support (almost €200 billion in capital injections, amounting to 1.2% of the aggregate Tier 1 ratio) and sovereigns’ guaranteed backup from the EFSF and ultimately the ECB. But utilizing the EFSF, which was originally intended for liquidity support, for bank recapitalizations could expose the sovereign to refinancing risks and rising spreads. “In situations of insolvency,” Dr. Roubini asserts in a June 28 RGE Analysis, “official support not only fails to prevent the eventual default, but also exacerbates the trouble, causing more damage to the country and even to its creditors.”

  61. @tull
    “It amazes me that nobody has commented on the fact that the Irish Permanent TSB did not even take the exam.”
    It is not just PTSB. There is one in every country. Particularly in every small country with a large foreign bank presence. Look at the parameters of the test again. They have selectively ignored the banks most likely to cause problems if they get in difficulty – small-medium, local market, hybrid (like bancassures).

  62. Again as it is germaine to the stress test, commercial property was assumed to fall in Ireland in the benchmark case by 14% and in the adverse case by 17% in 2010. The latest index for commercial property reported on at length in today’s Irish Times tells us we’re down by about 6% so far this year but the pace of falls in capital values is picking up. And remember that commercial property is, as reported in the article 58% off peak and we are back to 1999 price levels. If it can happen with commercial, why not residential?

    http://www.irishtimes.com/newspaper/commercialproperty/2010/0728/1224275602729.html

    One reason I’d suggest is that commercial transactions tend to be reported, particularly big ones. So people know the “going rates”. With the absence of a House Price Database declines might take way longer.

  63. @tull
    “That argument was relevant last year. Anglo should have been nationalised on the night of the guarantee and the process you outline started then.
    There are now now virtually no bond investors left in Anglo other than those issued under the guarantee. I repeat for effect, there are virtually no outside investors holding senior Anglo debt. If you want to discuss the creditors taking a hit please ring the ECB and/or its puppet in Merrion St. There is noboby left to burn at the stake. The 25bn or 30bn that Anglo will cost has gone to money heaven.”

    Unless I am wrong won’t there be over €7bn between subordinated and senior bondholders left in Anglo when the guarantee expires?

  64. @DE
    Dan Boyle is suggesting on the twitter that (goes to check)…banks may have to restructure debt, but not senior debt. That seems to me to suggest sub debt. But maybe Dan is wrong.
    @Jagdip
    yes. Japanification looms.

  65. @BL

    Hardly breaking news as they have already restructured most Sub Debt through buybacks. You mad not like the haircuts given but that’s another argument. Most of the deep sub debt Holders and shareholders in all the banks have already suffered losses (about 50% of the cost of the bailout if I remember correctly).

  66. @DE

    There are 2bn subbies certainly. But you ask Anglo or the DOF how much of the 4-5bn senior debt is held by outside investors. If you get an answer perhaps you could tell us. If they tell you there is 4-5bn senior debt outstanding, repeat your question.

  67. @ Hogan

    And your point is? Just because the Irish stress test was incredible does not mean that the CEBS stress tests are not credible. In form and substance the latter is problem closer to the US stress tests of 2009. Tell me were they a charade?

  68. People are placing too much importance on these tests. They were never going to solve the uncertainty out there about European Banks and Soverign Debt.

    If they had done a test showing what would happen if Greece, Ireland, Portugal and Spain all default and came to conclusion that most European Bank’s would need capital in that event, we would be sitting here going ‘WOW, what a surprise’.

    They may not have been the most testing of examinations as has been pointed out but they weren’t a charade. Analysts now have more information than they did a week ago. That alone makes the exercise worthwhile.

  69. @tull
    I’ll tell you after they stop pumping money into the US banks. As I already said, credibility is the problem. The US has pumped money into the US banks to ensure that the 19 they stress tested do not fail and appear to be succeeding. The BoE have had to move a step further allowing individual loans to be used as collateral without the need to package them.

    There is no mechanism in the EU for the ECB to support the banks beyond their liquidity measures which were already tapped out (those who had eligible assets had already posted them).

    Without measures to increase the cheap money the banks are getting, there will be more bank failures. The stress test is a cosmetic exercise – as previous Irish stress tests showed. Nothing wrong with this provided there is a plan B which recognises stress early and effectively deals with it. The Fed and the BoE have done this in their markets. I don’t see a mechanism for the ECB to do it in the eurozone. This is where a cosmetic exercise becomes a charade.

  70. @ Hogan

    i think its funny how lots of people have said “but if the T1 threshold was 7%, 24 would have failed” or “per Citi’s analysis, if there ha been haircuts into the trading book, 24 would have failed” etc etc, but we wouldn’t be able to make these statements without all the information garnered, analysed and released from the stress tests! As such, while the pass/fail results can be dismissed or somewhat discounted/ignored, the stress tests themselves have proven to be at least reasonably useful to many people in the market already (for “muted reaction” read double digit % increases in moast bank stocks and big gains in peripheral Eurozone govt bonds), and are far from being “a charade”.

  71. @Eoin
    Come on, you know that the danger that a non-negligible risk poses is much greater than the probability of it happening. If a bank that has been stress-tested as fine gets into trouble, then all banks ranked ‘below’ it and some slight ‘above’ it are going to be flushed. Without a mechanism to keep a bank that has been ranked as ‘pass’ always a ‘pass’ this is providing a short-term bounce at the expense of wider trouble in the future.

    The reason it’s a charade is not because of anything to do with the haircuts or the projections, but because it doesn’t address two fundamental problems:
    1. Selection bias in picking only 60% of the banks concentrating on the big ones. The small ones (like Anglo and INBS) are more likely to come to a sticky end.
    2. The lack of a system wide mechanism for resolving or supporting banks beyond the measures already in progress (which seem to be pretty much at their limit).

    If the financial crisis had as simple a solution as banks raising capital on their own, many more would already be doing it. They are not. In the current market, an unforced (by regulators) request for capital is something of a death-knell.

    Most sell-side analysts are doing what they always do when presented with a set of numbers – looking in the direction the numbers point them; like the duracell bunny, they are indefatiguable, noisy, fluffy and irritating.

  72. @Hoganmahew

    I don’t get your point about selection bias. How small did you want them to go? There are thousands of small to medium sized banks in places like Germany, Italy and Spain that people haven’t even heard of. Of course the market only cares about the big ones. I would rather lose 50 crappy Spanish saving banks or even INBS than one of the banks on this list.

    By all accounts the regulator’s test here was more robust and INBS was included in that. Don’t see why they had to include it in a CEBS report. It’s chicken feed in a European context.

  73. @Gavin S
    Who lends money to the small banks? Who are the bondholders? How small is too small? Anglo was 23% of the Irish market. INBS about 10%. Would you allow small banks to go bust and firesale their assets? What would this do to the rest of the banks?

    All or nothing. That’s pretty much the way the banking system works. That’s why the FDIC holds on to assets if it can’t get a reasonably price for them. Not because they think they are going to turn a profit, but because selling distressed assets has a knock-on effect.

    Tell me, had you heard of DEPFA Bank before 2007? Or HRE? What about WestLB? Or SachsenLB?

    As I say again, it is about what backs up the stress test. There is nothing. It is as much a charade as the Irish guarantee of the banks. Cheapest bailout in the world. It must have worked because the interbank markets are still expensive for the banks… oh wait, something wrong there…

    PS I’ve no doubt about Mr. Elderfield’s test on current projections, as I’ve already said. It contains the appropriate caveats…

  74. @Hog

    Yeah, I have heard of all of them because they are big market players and I deal with them.

    Have you ever heard of Banca Popolare di Spoleto, Bawag bank, Bank Girotel, Caja General de Ahorros de Canarias?

    Yes, I would allow small banks go bust. Whatever about Anglo, Irish Nationwide was never of systemic importance except for the headline risk that it’s failure would generate.

    Ask Washington Mutual bondholders if that is what the FDIC do when a bank fails.

    As for the guarantee and funding costs. Want do guess at what the difference is between getting funding at expensive levels and getting it at all, especially in late 2008 and early 2009. Yes, it’s the guarantee. The Guarantee is not the cost of the banking bailout. The recapitalisation is the cost of the banking bailout.

  75. @ Hogan

    How quaint of you to think that the ECB is at the limit of its measures. In 2008, it was unthinkable that the ECB would provide the level of liquidity that it has provided threough REPO. It was even more unthinkable that the ECB would buy sovereign bonds, never mind bank paper in the secondary market. You would never have thought that they would provide 60bn toIrish banks against ABS NAMA bonds. But they have. You must have immense insider knowledge to rule out further action.

    The criticism of the stress test on the two Irish is carping and smacks of the not being able to see the wood for the trees. Where is the discussion of the the macro implications of membership of the Euro? Where is the discussion of the competitve landscape of banking in Ireland? Where is the discussion of the desirability of the state being involved in credit allocation, property management etc?

  76. @Gavin S
    Only heard of two of them I’m afraid. But then I am an amateur 😀

    What would the effect be of 10 bn of INBS C&D and commercial property loans heading out in a firesale? (On the other banks, on the deposit guarantee, on ‘confidence’).

    You might want to ask the Washington Mutual bondholders whether they consider themselves assets of the bank or liabilities…

    Don’t get me wrong, I’m a “let them bust”er myself. But there is a difference between you and me saying that x is systemic but y is not, and the lender of last resort saying it. Effectively the ECB are saying the other 40% of the banking market doesn’t matter, while they have no mechanism in place to either protect the 60% that does matter or resolve the 40% that doesn’t.

  77. @tull
    “How quaint of you to think that the ECB is at the limit of its measures.”
    How amusing that you have it backwards again. It is the banks that are at their limits of eligible collateral for the current measures. The changes at the Fed and BoE have largely been to relax collateral regulations (at first to ECB levels and then beyond (the SLS for example)).

    I agree that the ECB has more bullets. The problem is that they are locked safely away in Berlin. While the gun is on holiday in Club Med. Perhaps there is a grand plan, but it appears that the ECB is locked between those who think there shouldn’t be action and those who want it – note, I don’t know that any believes that further action will not be required.

  78. What did the FDIC do with Washington mutual and their assets? They did a firesale to jpmorgan and burnt the bondholders. I don’t really understand your point about the ecb. What do you want from them?

  79. @Gavin s
    They sold them to JPM for a song. Which apparently has a chorus. Doesn’t the settlement last March mean that the bondholders are going to be made whole?

    The point is that the assets that WaMu bank held were not placed on the open market. There was an arranged sale without a market clearing price.

    I want the ECB to create a template for a bank resolution regime that each central bank/financial regulator in each eurozone state would have to implement. And the ECB to provide central (not NCB) funds to support resolution processes.

  80. @Hogan

    You will be a while waiting for a central resolution regime. There is no single template for regulation of institutions. The Germans are refusing to sign off oon the Basle 3 proposals. Our own govt. could legislate for a resolution mechanism but it appears unwilling. Perhaps you should stand for election.

  81. Sorry hogan, as a Washington mutual bondholders, we are certainly not being made whole! I think we would all to see a bank resolution scheme in place but I don’t see that being the responsibility of the ECB. A resolution scheme is a complex legislation so what works in France might not work in Ireland because of the ways our laws ate written. As Tull says, there just doesn’t seem to be the political will. Also, don’t see why you would want to use ECB funds to support the process. I would prefer to see a fund created and funded by banks that would be used.

  82. Peter Mathews has his views on stress tests here
    http://bit.ly/aNVDT6
    “I think you’ll agree that, if the Government at this point even reduced the Guarantee on Deposits to a Deposit ceiling of €100K , there’d be an instant huge rush to get Deposits out of the banks!
    So much for the Banks being well capitalised!
    Stress Test Eurocrats…. GET REAL!!!

    In the case of AIB, the EU’s Capital Stress test assumed AIB raising the €7.4bn required re-cap by this year-end. This is more smoke and mirrors!
    But what’s new?! The Government has a proven 100% preference for mis-information and untruths!
    The Eurocrats might have considered giving both Banks the following new Grade… fianna “FAIL” !!

    By the way AIB’s correct immediate min re-cap requirements are €10bn (not €7.4bn stated by the Minister and the Central Bank and Head of Regulation in March 2010, based on Loans assets information supplied by the Bank at the end of 2009 and very early in 2010 which we now know was unreliably optimistic).)
    And BoI’s correct immediate min re-cap requirements are €6.5bn (not the €3.65bn stated by the Minister and the Central Bank and Head of Regulation in March 2010, based on Loans assets information supplied by the Bank at the end of 2009 and very early in 2010 which we now know was unreliably optimistic).)”

  83. @Gavin s
    Apologies. It is the WaMu Inc (holding company) bondholders who look like they will be paid in full:
    http://finance.yahoo.com/news/Washington-Mutual-Senior-prnews-3010810595.html?x=0&.v=1

    There may not be the political will for a resolution regime and laws may differ, which is why an external agency and a central template for a process is required. Without measures in place, we will see more ad-hoccery. The ECB, like the Fed in the US or the BoE in Britain, are the only ones with the financial muscle to fund, or at least underwrite, what would effectively be chapter 11 bankruptcy for banks.

    I don’t like that the ECB should perform this role, but I don’t see anything else capable of doing it. I would expect that it would have to be funded by a bank asset or financial transactions tax.

  84. @ Brian
    Is there a link to a spreasheet shiwing his assumptions?

    When will Mr Matthews call for the replacement of Dr honohan and Mr Elderfield? It is one thing ot be foolish a la Hurley and Neary but if Mr M can stand up his numbers then the present governor of the CB, the CEO and the FR have to go.

  85. @tull
    No would seem to be the answer to that one (been trying to find them myself). I have presumed it relates to the fact that the NAMA haircuts are wider than was anticipated during the stress test and so, to leave the banks with adequate capital to function, more new capital will be required.

  86. @ Hogan

    has to be bigger than that. BOI is transferring 12bn to NAMA. A 3bn additional requirement would imply a 25 percentage point increment to the haircut, putting it into the Nationwide bracket. It would be a 10 point increment approx to AIB’s haircut.

    My 2nd point still stands, if Matthews can stand up his numbers. Honohan and Elderfield would be guilty of a major effort to mislead investors, even worse than the previous incumbants. Given their lofty valuations, perhaps their stock is a short.

  87. First I have ever heard of these numbers. Interested to know where he is getting them from. As Tull said, end of a lot of peoples careers if that’s the case.

  88. @tull
    Well, the stress test was completed in March and the tranche 1 loan haircuts were released in May. I’m not sure what the expectations were for the PCAR, as others have commented, they are a little opaque. I did find this:
    http://www.financialregulator.ie/press-area/press-releases/Documents/30%20March%202010%20Prudential%20Capital%20Assessment%20Review.pdf
    So I believe it is possible that the reduced scope of NAMA (so more worthless stuff left on the bank balance sheets?), the increased haircuts (beyond NAMA’s initial expectations as well?) and the knock on to the retained portion of development loans means that the PCAR on which Mr. Elderfield and Mr. Honohan made their speeches to Parliament has been superceded by events.

  89. Still doesn’t account for the extra €3 billion that he says boi need. That means boi need double what the regulator and every analyst report I have seen say that they need. Doesn’t add up but he might be able to prove me wrong. Nothing surprises me anymore. Unfortunately!

  90. @Gavin s
    “Nothing surprises me anymore. Unfortunately!”
    😀 Tell me about it. I see conspiracies of opaqueness everywhere. Spoof and bluster, spin and twaddle. Lies, all lies surround me. Paraniod? No. Parlonoid…

    There is that weird pending 5bn derivative loss in BoI’s accounts that I’ve never seen an adequate explanations for… http://online.hemscottir.com/ir/bir/ar_2010/ar.jsp?page=170&zoom=1.0&layout=single
    (see p.295 – it seems to be 6 bn now?). I doubt it, though. I think Mr. Mathews is from a pre-derivative banking age, hence his belief that a bank can just be wound down quickly.

  91. Hi Hoganmahew

    That’s not a pending loss. It is just a liquidity analysis and so that table is just looking at outgoing cash flow’s according to contractual maturity date. From what I can see from their accounts, they have roughly a €1 billion “potential loss” on a €390 billion derivative book as of the 31st Dec 2009.

  92. @ Brian Lucey

    re Banker Mathews

    (a) he suggests paying 47bn for the NAMA loans. Aren’t we paying c.42-43bn now?
    (b) i couldn’t find a spreadsheet with his AIB/BOI figures – honestly may have missed it, but i couldn’t find it.
    (c) he makes a lot of ‘definitive’ statements which are actually only his analysis/opinion, ie “My previous letter (November 7) clearly shows how the NAMA experiment will result in losses on loans recoveries of not less than €11.65bn”. (even though this ‘loss’ should actually be negated by the now smaller consideration being paid to the banks…)
    (d) he uses an awful lot of capital letters, for no particular reason, in his blog, this always makes me uneasy.

  93. @Gavin s
    Thanks for that. Explains a lot. Did you notice that the cash collateral they put up is also in their “deposits from banks” section of their accounts. Is this standard accounting for this, do you know?

  94. @Jagdip Singh

    Agreed that the residential property tests look weak. Here is a summary of the various recent projections/predictions and assuming the EU expect the bottom to be at the end of 2011

    Residential Property In Ireland Peak to Decline from
    Source Date Trough start Jan 2010
    EU Adverse Stress Test July 2010 46% 21%
    EU Benchmark Stress Test July 2010 42% 15%
    Kevin O’Rourke July 2010 70% 54%
    Moodys July 2010 45% 20%
    Goodbody July 2010 50% 28%
    Jim Power (1) July 2010 40% 12%
    S&P June 2010 41% 14%
    Moodys May 2010 46% 21%
    DKM May 2010 45% 20%
    David McWilliams April 2010 64% 47%
    Fitch Mar 2010 45% 20%
    Paddy Power (2) Jan 2010 40% 12%
    IT Pundits Jan 2010 38% 9%
    Anglo Plan (3) Nov 2009 47% 23%
    Morgan Kelly Jan 2009 80% 71%

    ESRI/TSB figures for Q2 just out. They show the smallest quarterly decline since the onset of recession, just 1.7pc in the quarter (slightly over 0.5pc a month), bringing the total decline since the 2007 peak to 35.3pc. Not good news for the more pessimistic forecasts in this list.

  95. Germaine to the residential property prices at the top of this thread, the latest PTSB index indicates that residential property dropped by 1.7% in Q2 which brings the six month drop to 6.5% compared with the benchmark scenario in the bank stress test of 13% for 2010 (benchmark) and 17% (adverse).

    http://www.esri.ie/irish_economy/permanent_tsbesri_house_p/

    Given the last Irish Banking Federation figures for mortgages in Q1 showed that 4224 mortgages were provided and although the latest report from Q2 isn’t available recent press reports suggest lending criteria have been tightened up. PTSB has what? 20% of the mortgage market? On Q1 data they would have had 840 mortgages. How many for Q2? And these are spread across Dublin and Outside Dublin and by different size of home. At what point does their analysis become statistically suspicious?

  96. @ Eoin, GavinS,
    I await the call from Mr. Matthews for the resignation of Honohan & Elderfield for running a fraudulent stress test.

    @BL
    CDS on the 2 Irish banks are in this week.AIB -100bps BOI-80bps. Other EZ banks in as well but not as much. Bond markets seem to be fooled by this charade as do equity markets- SX7P up by 7% since Friday close. Probably only bear closing. Will you be callingfor the resignation of Dr H?

  97. @JTO

    With the Eurostat confirmation that our population grew at 6,000 in 2009 (the lowest 12-month growth since 1991) and with an excess supply of housing at 35-350k and new homes being built at 6-15k/year I would suggest that on a supply:demand basis houses have some way to fall yet.

    http://europa.eu/rapid/pressReleasesAction.do?reference=STAT/10/110&format=HTML&aged=0&language=EN&guiLanguage=en

    The PTSB index shows that for Q1/2 the decline is 1/2 the full year benchmark scenario. It is likely to be based on a very small sample given the further decline in mortgage lending and the fact that PTSB has what? 20% of the market?

    IMO, NAMA itself as the biggest putative property owner in the State with a mandate to sell at the LEV and holding itself out to be the rescuer of developers is probably the biggest obstacle to property reaching its bottom.

  98. @ At Tull

    He seems to be making his assumptions based on his superior judgement. He talks about it here at the joint commitee but doesn’t go into much detail.

    http://debates.oireachtas.ie/DDebate.aspx?F=FIJ20100505.xml&Page=1&Ex=H2#H2

    I honestly don’t know what to make of him to be honest. He talks about himself in the third person a fair bit which always worries me! His reply here to NAMAwinelake’s good blog shows he doesn’t really understand NAMA either.

    http://bankermathews.com/2010/07/22/response-to-namawinelake-blog/#more-272

    Having said that, can we please put a muzzle on people like Frank Fahy. The guy is clueless.

  99. @ Jagdip

    few things:

    1. most of those looking for the apocalyptic house price falls are also factoring in massive negative demographics (think Joan Burton and her “500k will leave” forecast). You admit that demographics in 2009 were still in fact positive.
    2. similar accepted demographics assume that more houses will in fact be required in the coming years due to an eventual change in our social make up from baby boomer to old ager.
    2. house prices are falling at a slower pace at the half way stage of 2010. Given basic ceterus paribus trend dynamics, wouldn’t the full year figure likely (and i use the term admittedly very softly) be less than the benchmark figure?
    3. PTSB is still one of the biggest (THE biggest?) mortgage lenders, so to say its “a very small sample” is a touch miseading.

  100. @Eoin
    I dunno. The PTSB is almost closed for business.

    From their 2009 accounts –
    Irish residential lending:
    2008: 4.2bn
    2009: 0.9bn
    81% decline.
    UK residential lending:
    Closed in 2008
    Commercial lending:
    Closed in 2008
    Consumer finance:
    2008: 1.1bn
    2009: 0.3bn
    72% decline

    EBS and AIB have been crowing about increasing market share. Unfortunately, I can’t find any statistics about this. The IBF seem to have pulled this section from their reports and pulled all the old stuff too (unless it is now behind the paywall?).

    While I can believe that PTSB continue to hold 20% of residential mortgages, I don’t think the idea that they are issuing 20% of new mortgages is tenable.

  101. @Eoin

    Is the expression “looking for apocalyptic house price falls” a tad provocative? Given that 450,000 net migrated (immigration – emigration) during the Celtic Tiger years (1994 – 2008), could we have a reciprocal reversal during the next decade? I dunno, won’t it come down to how well our economy is performing relative to our competitors and there are upsides and downsides. However in terms of the next few years, wouldn’t the betting be that we have negative net migration? For what it’s worth the ESRI seem to think so. So I don’t think you need to be looking for apocalyptic house price falls to conclude that on balance there will be large scale negative net migration.

    Even with 6,000 net new souls in 2009 (though I’d be willing to put money on there being a drop in the 12-month period to April 2010 when the CSO publish their estimates in Sept), new (in the sense of presently unbuilt or vacant) houses will be required. The question is how the overhang plus new build compares with new demand. I’m suggesting that for the next 3-7 years at least that the dynamic should tend to depress house prices.

    The Q2 fall according to PTSB was 1.7%, Q1 was 4.8% and Q4, 2009 was 8.3% so you could say the trend is towards a stabilisation. I’m saying that there are factors at play now which might not be in a few months and those factors are tending to prop prices up above their natural level. NAMA would be at the fore. What developer is selling property at any price now when NAMA may assist the developer in riding out the downturn and selling at a better price when the LEV returns? What developer with partly completed estates is selling at whatever he can get when NAMA have €5bn to help him complete his project? What bank is pushing their lenders to sell at any price and consequently recognise a loss in their books and put in jeopardy the value of NAMA-bound loans? There are other factors at play. As we’re talking about scrying the future, all we can do is declare the factors we judge to be in play. And I think that further bigger falls are in prospect.

    Permanent TSB had 20% I think of the market. Does anyone know what they have now? If you examine the Q1 IBF mortgage stats, you’ll see that lending had fallen off the cliff and recent media reports imply further tightening in credit and lending criteria. If PTSB had 20% of the market then they would have had 800-odd mortgages to examine in Q1. What’s the number for Q2? Given that we don’t seem to know the total number of transactions (PTSB mortgage plus others mortgages plus non-mortgage transactions) how would you assess margin of error? There is plainly a point at which the number of transactions examined will expose your results to significant error. There were 4,224 new mortgages on property in Q1,2010 compared with 32,000 in Q4,2005. Are there sufficient PTSB mortgages now to enable a defensible survey?

  102. Gurgdiev is going on about today’s house price figures in much the same way as Jagdip. Its amazing how the doom industry in Ireland always asks us to ignore any statistics that aren’t to its liking. This occurs over and over again. So, we get in the past month alone:

    gdp up 2.7pc in Q1 – ignore it, it was all exports
    car sales up almost 100pc in June – ignore it, it was the scrappage scheme
    manufacturing output up 9.5pc in May – ignore it, it was all chemicals.
    merchandise exports up 8pc in May – ignore it, it was all chemicals again
    Irish banks pass EU stress tests – ignore it, they fiddled the tests
    house prices fall slows down in Q2 – ignore it, the sample was too small

    Re the slowdown in the house prices fall to 1.7pc in the latest quarter, it may well be case that we shouldn’t make too much of it. Several quarters would be needed before we could say definitely that prices were levelling out. However, it fits in with recent rental data. Both the CSO and Irish Property Watch report rents levelling out or even increasing very slightly in the first half of 2010. Just as a matter of interest, if house prices fell by 1.7pc every quarter from now on, it would be almost 10 years before the fall from peak reached 60pc and almost 20 years before the fall from peak reached 80pc, Morgan Kelly’s prediction. So, even if the Great One is proved correct in his forecast of an 80pc fall from peak, I may well not be around to congratulate him.

  103. @ jagdip,

    One of the few timeseries on house price trends indicates stabilisation. However, this does not suit the thesis. What do we conclude? There has to be a flaw with the data. Clearly it must be dismissed.
    At the peak of the boom, the followers of Dr Pangloss could see that the data was wrong. Now after a 40% plus fall, the Cassandras are playing the same game.

  104. @JTO

    “So, even if the Great One is proved correct in his forecast of an 80pc fall from peak, I may well not be around to congratulate him”

    Come on, where is the optimism!

  105. @JTO
    The decline in the rate of retail sales is increasing – it’s MoM data, so I discount it. Like I discount pretty much all other MoM data.

    Volume of mortgages issued is as low as it has been this century. If this was GDP figures, you’d be calling it a catastrophe. Ah, but now you are only concerned with price data?

    I’m still waiting for those GDP figures you promised. Many were surprised by the level of price deflation last year and by the fact that it continues this year.

  106. RTE reports that NIRSA (which is attached to NUIM) has issued a report calling for an inquiry into local planning and suggesting that there are 103,000 unoccupied housing units. Can’t find a link to it, but it might add some grist to t’mill.

    And wrt to house prices, perhaps the store of land and property-based collateral being built up by NAMA is a factor in the apparent stabilisation – and which may be at an artifically high level.

  107. @Paul,

    The link to the NIRSA report is below – they’re the people of the 350,000- -vacant- homes fame though they set out that figure in some detail in the report and also consider vacant homes using another basis (overall they and other sources seem to agree the “overhang” is 100,000+ though gross vacancies may well be over 300,000)

    By the way they conclude that housing prices will remain depressed for some time. The media have already claimed that NIRSA are predicting declines of upto 60% from peak – in fact what they say is that other economists have estimated declines in the 40-60% range – isn’t this how Kevin O’Rourke got labelled with the 70% decline?

    http://www.nuim.ie/nirsa/research/documents/WP59-A-Haunted-Landscape.pdf

  108. Qtrly Dublin Outside Dublin National

    Q3 2006 4.3% 2.6% 2.7%
    Q4 2006 1.9% 0.1% 0.9%
    Q1 2007 0.9% -0.2% -0.4%
    Q2 2007 -2.7% -3.2% -2.5%
    Q3 2007 -2.9% 0.3% -0.4%
    Q4 2007 -1.5% -3.2% -3.8%
    Q1 2008 -2.6% -2.9% -2.0%
    Q2 2008 -3.9% -2.4% -3.2%
    Q3 2008 1.6% -2.6% -1.5%
    Q4 2008 -4.2% -2.6% -2.4%
    Q1 2009 -9.0% -2.8% -4.3%
    Q2 2009 -3.5% -3.4% -3.9%
    Q3 2009 -5.7% -4.2% -3.9%
    Q4 2009 -7.5% -6.2% -7.7%
    Q1 2010 -10.3% -3.5% -4.8%
    Q2 2010 -3.5% -0.8% -1.7%

    Looking at the quarterly house price series there would have been plenty of times where it would appear we are “stabilizing” only for the pace of falls to pick up again.

    No matter what trend analysis might show by all most any fundamental measure property is still significantly overvalued.

    I think prices still have atleast another 30% to 40% to fall. Until the rental yield returns to normal prices will continue to fall.

    @JtO
    Irishpropertywatch is showing rents falling again

  109. I have downloaded the NIRSA report and I have been reading it this afternoon. I will put a separate blog entry up on it for comments but want to finish reading it first. So read the NIRSA report if interested and we will have a comment blog page on irisheconomy.ie very soon.

    Gregory C.

  110. I gather that my Comment of 26th July caused some people some concerns.

    Perhaps readers might consider the following…

    AIB originally and defensively admitted in mid 2009 that it held €24bn in NAMA bound bad loans. Apply a modest average write-down of 40%. This amounts to €9.6bn. Experience and prudence would indicate that AIB should round the write-down to €10bn and seek to build up its capital resereves by that amount. Remember, AIB will need a good cushion to buffer increasing loan losses in the sub €5m property loans and also in its Household Mortgage book and also in its general lending book.

    I have always been a proponent of conservative common sense banking based on common sense definitions and terms. I’d caution against getting hung up on formula based rules clouded by jargon definitions such as tier 1. tier 2 etc. Straightforward old fashioned capital levels and definition (e.g. equity at 10% of total assets) are more meaningful. Together with the assurance of the application of fundamental Banking principles such as prudential fractional reserving under competent trustworthy Board Directors and Managers gives depositors the confidence and trust to place their deposits with the bank. In this context it’s relevant to remind readers that AIB has consistently fallen far short in providing dependable financial information for at least the last 3 – 4 years to date.

    In the case of Bank of Ireland, once again events and patchy and unreliable information over the last 3 years lead me to be quite circumspect in assesing its re-capitalisation requirments. BoI originally and defensively admitted, in mid 2009, that it held €16bn (not the later revised downwards €12bn) in NAMA bound bad loans. Once again, apply a modest average write-down of 40%. Experience and prudence would indicate that BoI should round the write-down to €6.5bn and seek to build up its capital resereves by that amount. Remember, BoI will also need a good cushion to buffer increasing loan losses in the sub €5m property loans and also in its Household Mortgage book and also in its general lending book.

    I have always been a proponent of conservative common sense banking based on common sense definitions and terms. I’d caution against getting hung up on formula based rules clouded by jargon definitions such as tier 1. tier 2 etc. Straightforward old fashioned capital levels and definition (e.g. equity at 10% of total assets) are more meaningful. Together with the assurance of the application of fundamental Banking principles such as prudential fractional reserving under competent trustworthy Board Directors and Managers gives depositors the confidence and trust to place their deposits with the bank. In this context it’s relevant to remind readers that BoI has consistently fallen far short in providing dependable financial information for at least the last 3 – 4 years to date.

    I hope that I’ve addressed readers’ concerns and that my explanations, which are tempered and seasoned by my professional experience and judgement are clear. I’d like to thank readers for their views and comments.

    Every good wishes

  111. @ Peter Mathews

    thanks for the reply and analysis.

    However, i dont see any mention of their pre-existing loss provisioning which can be netted off against the NAMA haircuts, or any mention of the fact that they’ll have a good deal less risk weighted assets on their balance sheet post-NAMA, and so will require less capital to be held as a result. Would these not materially alter your figures for required new capital?

  112. @ Peter

    A lot of what you say is true. The disclosure to the public by the banks generally. As such it is difficult to know with certainty what exposures are on their balance sheets at any give time. However the downside of this is that one cannot be definitive about loan losses & capital requirements. Yet you are quite definitive.

    The one person who should be in position of the requisite information is the new FR, yet you seem to dismiss his assessment of the capital requirements. Does he not have more information than you do.

    Lastly, it is clear from your analysis that you seek the banks to hold more capital than any other regulator. I can think of no system in the developed world that would demand a through the cycle Equity/Assets ratio of 10x. The nearest ratio to that is the new basle 3 proposal which is tier 1/adjusted assets of 30x.

    Your demand for more capital is a fair one but I do not think it is clear from your advocacy that a key reason for the difference between you an the FR is that you would require banks to hold more capital than he does. Instead it looks like you are accusing the authorities of cooking the books to arrive at a lower capital requirement.

  113. @Peter,

    Thanks for the clarification but I have to echo what Eoin Bond and Tull have said especially on the area of pre existing provisions and capital.

    Also your idea of a leverage figure of 10x is economic suicide. If we take your argument and simplify capital as simple equity, BOI had €5.3 billion of equity Tier 1 at Dec 31 2009. I know you don’t like Tier 1 definitions etc but I assume you don’t have a problem of using the Risk Weighted Assets figure rather than plain asset figures. Under your proposals, BOI would be allowed a balance sheet size of €53 billion RWA. They had €98 billion RWA in Dec 2009 so you would require BOI to deleverage their balance sheet by 45%. Good luck getting credit into the economy with that requirement.

    We all want banks to have enough capital and be properly capitalised but there is a balance to be struck. Otherwise we would just tell to have 20% Tier 1 and never worry about them again. I don’t think just throwing numbers out there at Dail joint committees without proper analysis to back you up is the way to go. Despite whatever professional experience and judgement you might have. There is enough political grandstanding as is.

    By the way, well done on dealing with Frank Fahy the way you did!

  114. I should add that if you take BOI’s total capital position of €13.2 billion at Dec 31 2009 and their RWA of €98 billion, then they are well inside your 10x leverage limit so shouldn’t need capital.

  115. Peter,

    I would echo what the other said. Moreover if you were to apply a leverage ratio of 10x assets to equity, the cost of credit would have to go up. BOI median ROA has been around 1% plus or minus for the last 20 years and will likely be less than that in the future due to the absence of higher margin lending. Leveraged 10x gives a ROE of max 10% i.e below the cost of equity. The sector return in the rest of Europe is going to be about 13-15% in a normal environment. Therefore, Irish banks and the banking sector here will not attract capital. The net results will be some or all of the following
    *the state will not be able to recoup its equity or exit the stakes it holds or
    *ROAs will have to go up by 50% to be achieved by higher cost of credit
    *there will be less credit available anyway due to the need to deleverage

    I think 10x unweighted A/E is too low just as 30x adjusted A/E is too high.

    That said, congrats on putting Fahey in his box.

Comments are closed.