Morgan Kelly on the Size Distribution of Irish Mortgages

Morgan has a new working paper titled “A Note on the Size Distribution of Irish Mortgages”.  From the conclusions

Our analysis here has shown that instead of 10,000 million plus mortgages, there were probably fewer than 2,000 and these were mostly for investment rather than own home mortgages. However, looking at the 11,000 largest mortgages from the bubble peak of 2006-08, we find that the total is €9 billion. We do not know how many people held more than one mortgage, but it does not seem implausible that the total indebtedness of the 10,000 people with the largest mortgage debt is in the region of €10 billion.

Morgan uses statistical methods to estimate the number of mortgages valued at over €1 million because the Department of the Environment do not make these data available.

While the question of how many mortgages there are over €1 million is not particularly important, it might be worthwhile for the Central Bank or CSO to make available more detailed figures on the size distribution of mortgages. As there are quite a few banks in the Irish mortgage market, the usual argument about confidentiality hardly applies as an issue when releasing aggregate figures.

28 thoughts on “Morgan Kelly on the Size Distribution of Irish Mortgages”

  1. More a national problem of residential landlord mortgage arrears (occupied and unoccupied) than residential owner-occupier mortgage arrears then? Genuine question.

    Another sting in the tail (no pun intended) is that many of these landlord investor mortgages are inextricably tied up with SME and professional businesses, with the businees (or what is left of it) being the only cash generator to pay these ‘investor’ mortgages.
    One hell of an Irish stew! The maths, needless to say are, beyond me.

  2. Of more interest would be breakdown by approval year, LTV at approval, and value range.

    I’d like to see this for all property, and related loans for 2000-2010…my overwhelming suspicion is that 2006 was the year of greatest craziness, with a huge number of deals

  3. These 10,000 borrowers with mortgage debt of 1 million per capita are a key beneficiary of ongoing bank debt forbearance/forgiveness. Is that an appropriate use of scarce public resources?

  4. Much as Kramer doing Moviefone said “Why don’t you just tell me the name of the film you’d like to see?”, why doesn’t the Central Bank just tell us the proportion of investor mortgages that were interest only? This “ecological inference” is basically just an assumption-dependent guess.

  5. @Gregory Connor
    It is not debt forbearance, it is legislative forbearance. The current bankruptcy laws make it uneconomic to bankrupt people. The failure to introduce a proper personal insolvency scheme is the sop that is being provided.

    @Frank Galton & Karl Whelan
    I suspect we don’t have the information because if we did, the conclusions would be higher loan losses than the Central Bank have budgeted for. Presumably Blackrock had that information when they gave their analysis. Am I alone in wondering why there was such a fat tail given by the Central Bank to lifetime losses? Surely over time losses should come down? Instead we have a picture of some losses now, some losses later and some losses after a while. All perfectly manageable.

    Nicely fits the money that is available…

  6. @Hogan
    re: ‘It is not debt forebearnace’ ?

    I would disagree. Banks first make provisions, then they write-off/do not pursue debts.
    Thes decisions to ‘write off/ not pursue’ have always been taken by banks. But they are a little nervous now, because the problem is very large and the executives, mindful of their jobs, are very careful about playing by the rules. The problem is and was, they does not appear to any rules and probably never was.
    Not to pursue = debt forebearance=write off.

  7. @ Greg

    “These 10,000 borrowers with mortgage debt of 1 million per capita are a key beneficiary of ongoing bank debt forbearance/forgiveness.”

    Let’s not mix up two different sets of mortgages.

    MK writes “there were probably fewer than 2,000 and these were mostly for investment rather than own home mortgages.”

    The recent debate has been about owner-occupied mortgages.

    That said, foregiveness or no, many of the investment mortgages will not be paid back in full, and there’s little gained from pretending otherwise.

  8. @Joseph Ryan
    Making decisions on losses is well above the pay grade of the responsibility-emasculated middle managers that exist in the banks. Not unless they have a performance incentivised target for it, anyway. Forbearance may be contributing to losses, but if it is agreed at least for a period, then those losses are hidden under restructuring costs. Meanwhile the problem has gone nowhere.

  9. @Hogan:
    So at what level in the banks have ‘bad’ debts been written off over the years? And what criteria was used? Is there a paper trail, I wonder.

  10. Morgan uses statistical methods to estimate the number of mortgages valued at over €1 million because the Department of the Environment do not make these data available.

    When a man breaks out log log charts and integrals leading to incomplete gamma functions, you know he means business. Statistics hardly come into it.

    Kelly has clarified his 10,000 “cup of coffee” borrowers in this paper, using hard analysis of the available data. It is clear that overall, these 10,000 mortgages do account for almost €10 biillion euros, though only ~2,000 are over €1 million. The question is, who took out these 2,000 loans, and are they paying them back?

    From the paper:

    It is notable that, although the total number and value of mortgages fell between 2006 and 2008 , the share of mortgages going to large loans rose which is shown by the falling parameter alpha.

    I wonder how all those personal and interbank loans factored into all of these? Maybe there were a lot of people re-mortgaging properties as the whole thing was going belly up?

  11. To clarify, when I said 10,000 borrowers with 1 million per capita mortgage debt, I meant 10 billion lent to the top (in terms of mortgage size) 10,000 mortgage borrowers. That is different from the 2,000 borrowers with above-1 million mortgages. They have a substantially higher per-capita mortgage size.

    I agree with Karl Whelan’s comment above that these are not the intended beneficiaries of proposed mortgage debt forgiveness plans. However I would argue that they are a major beneficiary of the actual ongoing mortgage forbearance/forgiveness policies. I conjecture (without firm evidence to back it up) that perhaps there is more credit tied up in rolling over their delinquent loans than in those of owner-occupiers? Maybe someone can crunch out that ratio from central bank data?

  12. @OMF

    ‘Kelly has clarified his 10,000 “cup of coffee” borrowers in this paper, using hard analysis of the available data. It is clear that overall, these 10,000 mortgages do account for almost €10 billion euros, though only ~2,000 are over €1 million. The question is, who took out these 2,000 loans, and are they paying them back?’

    No wrong question – the real question is why is God’s holy name with net rental yields in Dublin in the more affable areas close to 1% in the 1st quarter of 2006 were these loans given in the first place. Banks lending decisions dictate this market not consumers.

    Had the banks being doing their job they would have simply said ‘NO. The metrics don’t work now please piss off’.

    But the banks, as we know didn’t say this. Thanks to MKs ongoing and very insightful analysis (constrained as it is because of the fears in the CBI re data releaes), screwed this market up beyond repair for ALL market participants.

    The fact that the banks have a long term interest in ensuring the mortgage/property market remains relatively stable and sound is in all our economic interests, including obviously theirs.

    With the facts emerging as they are indicates my long held belief that all the lending banks in the RoI market since c2000 have in effect been negligent in the manner in which they have allowed the property market to develop. With every passing day this is becoming more and more evident.

    Given the fact (and they are now facts – statistical assumptions aside) does it really make any logical sense that banks who lent money on phoney models and flooded the market with significant excess supply and guaranteed almost zero chance of price recovery (at least not before 2040 per the CBI) in the event of a downturn – that they have now the legal right to pursue anyone for the full 100% of the value of the mortgage? The banks very own actions they have ensured this simply cannot happen by the normal disposal method.

    As has been said here many times before banks price property, not consumers or estate agents or developers but banks. Those in control of the leverage in a leverage dependant market control the underlying asset prices.

    Given the ongoing disaster in pricing and hardship for hard pressed cinsumers, surely it now makes sense, as I’ve been suggesting here and elsewhere for the best part of three years to reprice these transactions based on longer term normalised yield metrics as a basis as an effective stab at debt forgiveness. Forgiveness should not based on someone’s ability to pay, I repeat it should not, but it shoould be based on a mortgage value based on a property priced house at the origination of the transaction. I stress properly valued property.

    It is simply not true to suggest, as many sadly do, that a property is only worth what someone was willing to pay for it. This is a nonsense statement. Yet time and time again it is trotted out as fact. It’s nothing of the sort. Property transactions in 98% of cases only happen because a lending bank allows it so. Given this, a property is actually only worth what a bank is willing to lend against it – nothing at all to do with the action of the consumer or what a seller believes it worth. It completely depends on the lending bank.

    Given the way in which the market has evolved surely it would now seem reasonable to ask the obvious questions as to why in a market in which c4 times the average annual demand (2004-2007) was actually being supplied was experiencing rising prices? Those better informed should be able to throw some light on this.

    The only way in which this makes any sense, in my view, is where the banks were breaking all the rules in relation to compliance and ensuring the long run market integrity as a whole was put at significant risk. We now know this to be the case with recent price falls in Allsops averaging c70% from the top.

    Given this background it was hugely disappointing to watch last nights debate on the VB programme where these overwhelming background market influences were barely even discussed.

    The nature of the current debt forgiveness debate has centred on who deserves it more and the lack of revised bankruptcy laws. My view is that these issues are actually avoiding the real issue that being the case that the evidence now suggests there has been a ‘pricing fraud’ committed against the average consumer (particularly new build).

    Against that background why should any consumer have to go through a potentially demonising bankruptcy procedure (new or old) when the average consumer was at all times a price taker in a market where the leverage providers failed to do their job – and now the consumer is asked to shoulder 100% of the loan or potentially have to go to court because he/she can’t get out of their situation because of a saturated market all of which is completely out of his/her control?

    It is simply a daft situtaion whereby the Regulatiors, Govt and banks alike fail, it seems to get this central issue and that is the banks, who are the architects of the demise of their own bread and butter market should in any common law envirnoment hold any right to impose any legal charge against suffering consumers.

    Is it not reasonable to assume that in the case of any individual debt action the law should seek to find evidence that because of one partys wider market activities i.e. the bank in question, in saturating the market with excess supply, that that action has had the effect of imposing hardship on all particpants thereby ensuring one side could simply not entangle themselves from their obligations even if they wanted to?

    My view is that Seamus Coffey’s statement last night that an across the board debt forgiveness program will not happen may prove to be wide of the mark.

    This is a market which has been mis managed by the banks for a decade and that’s a larger structural problem than I believe Seamus is currently forecasting.

  13. I find this offering from Morgan Kelly truly pathetic. Brendan Burgess who has been heavily involved in examining the problem of mortgage distress immediately lambasted MK’s “10,000 over €1M” claim. He exposed his source for what it was, an anecdotal piece in the Irish Times. BB showed that the number of >€1M borrowers was indeed a fraction of 10,000. He also pointed out that the real culprits in the interest only BTL market were BoS and their problems have nothing to do with the Irish taxpayer.

    At least MK has admitted in this latest piece that he hasn’t a clue what default behaviour will be in this category. A far cry from the certain Armageddon of two weeks ago. Of course nothing like a few Wienbulls and Paretos to make sure the media will not run with this retraction.

  14. @Yield

    You make a good argument as to the deciding factor in the price of a house.

    I thought that the Seamus Coffey approach of confining a structured program to those in arrears made some sense. While your thinking has a a good deal of logic to it, it is hard to see how it could be contemplated at this stage.
    The Gurdgiev analysis of the fact the stress test only go to 2013 and that international research points to a lot of defaults in 2015/2016/2017 was very worrying.

    As for Brendan Burgess, he may as well have been a spokeman for the banks in my view. The report of his his group was sheer nonsense. It amounted to squeesing the lemon until the pips squeek for about five year and when there in no more juice, thrown the lemon peels (aka families) onto the road. Just to avoid moral hazard, you understand.

    The difficulty is that when the banks are back in private hands in 2016/2017, any residual cost will be dumped back on the taxpayer. Again.

  15. So a professional owes 750K interest only on a BTL now worth 500K. Very painful but unless indeed they can’t afford a cup of coffee this mortgage will be repaid. This is almost a good news story for society, these professionals are going to have to work harder and longer to rebuild their pensions.

    MK was right not to trust the earlier internal stress tests of the banks and he deservedly gets credit for his “ghost estates” comment amongst others. But the PCAR test in March was transparent and independent. One presumes Blackrock considered carefully the 10,000 highest value personal mortgages and with much more data at their disposal than MK – no need for them to do Pareto distribution fits.

  16. @BWII

    All banks who operated in the RoI over the past decade are implicated including BoS. To suggest, as Brendan Burgess did last night that these other non Guaranteed banks are of no concern to us is plain wrong. The excess supply sitting on the market effects all participants if that excess supply was fianaced through KBC, BoS or Ulster Bank it makes no difference it has still added to the supply of stock which has imparied the recovery of the market.

    The same Brendan Burgess sat on the Mortgage Review Group and were adamant that ‘moral hazard’ would be too overwhelming a factor to consider debt forgiveness and the fact that it hadn’t worked in any other part of the world. I’d venture to say that the state of the lending practices here and the reasons why we experienced rising pricing at a time where we were tripping over new houses was probably something the Review Group were also unlikely to have experienced anywhere else in the world – so drawing conclusions where none should correctly be drawn suggests to meet that Mr Burgess opinions should be taken with a fair degree of salt.

    In the absence of hard data MK is forced to use the methods he does.

  17. @BW11

    One presumes Blackrock considered carefully the 10,000 highest value personal mortgages and with much more data at their disposal than MK – no need for them to do Pareto distribution fits.

    If Blackrock or the CBI did this kind of analysis, why did they not produce that data. In fact is this not the kind of data that any ‘Expert Group’ should demand to see before it came up with proposals to deal with the issue.

  18. @ Yields or Bust

    MK lobbed one of his periodic hand grenades a few weeks ago – that here was another black hole for the Irish taxpayer. Now, in a fog of gobbledygook math, he completely retracts.

    The fact that some of the main lenders in this category are not covered institutions is entirely relevant, I can’t see how you can argue differently.

  19. @BWII

    In all asset markets supply matters. How that supply comes to the market is irrelevant. The fact is it’s there and is currently having a seriously negative detrimental effect on pricing which affects all participants.

    The debt forgiveness debate, as I argue above, has boiled down to who deserves it and why not me? To me this is completely missing the point – the broader point is the banks have contributed directly to this problem by their actions and they deserve to pay the penalty, the law currently protects their legal right to exercise their charge regardless of their actions in the broader market – this is clearly wrong and is ultimately illogical.

    Why a market controlled by the leverage suppliers, regardless of where the came from, can not see what effects their actions were having for the long term stability and integrity of the overall market, is still the great mystery to be solved by the politicans and Regulators alike.

    Its a matter of complete indifference what the names above the credit suppliers doors were, the facts are that their models were all flawed and rental yields were telling all players to stop supplying credit many years before the crash came. It’s a bankers job to understand when the market is telling you to stop, the consumers visit such property shops a couple of times in their adult lifetime, they are novices, banks lending into this market are expected, by the law and Regulators alike to be expert and competent. We now know this to be a complete fairytale yet it seems the consumer is the party who desreves to be demonsied through the bankruptcy procedures. I’m at a loss to understand why.

    The excessive supply is directly contributing to the mortgage stress because in a normal environment disposals would potentially allow relief. The current situation is not allowing for such an out clause and therefore the pain goes on – all players have contributed to this situation. Quite honestly it’s very hard to analyse it any other way.

  20. @ YoB

    I think you a tad off topic. MK argued a few weeks ago that there was another hidden hole for the Irish taxpayer in 10,000 professionals with mortgages of over €1M who cannot afford the price of a cup of coffee. He has revised that down to 2,000 and still doesn’t mention that the taxpayer is not on the hook for e.g. BoS loans.

    On your ideé fixe, yes one can certainly blame the banks for their part in the property bubble. But it is oversimplistic to say that banks actually set the price level, otherwise I presume they would not have allowed the bottom to fall out of the market.

  21. @BWII

    No not off topic at all – I read the MK article and without the data I agree he may well be off the mark but that doesn’t change the overall agrument -banks contributed to their own demise, mortgages such as these be they 10,000 or 2,000 of them made no sense on a yield basis regardless and its the banks job to get these calls right.

    The banks always have the last say in a leverage driven market and the numbers were screaming NO but they still went ahead and released the credit.

    The additional supply is affecting all market partipants and particularly those through no fault of their own have been denied a normal out because of the banks overall market actions. These are all the pertinent facts which are staring us in the face and no amount of faster bankruptcy procedures will change them. The banks hold the keys to this market in every respect and they screwed up royally.

    The fact that banks made a complete arse of the price setting over the past decade is precisely why the bottom has fallen out of the market. Perhaps you haven’t heard but they’re broke – directly as a result of their mispricing the risks in the market. The quaint and simplistic view is actually to believe that the estate agents and consumers value properties sadly that’s just being a tad naive.

    House on sale priced by local estate agent at €200k: Two pillar banks Bank A is only willing to lend on an average long term yield of 7% as is B – proxy house generating €1,000 rent per month on a net yield (11/12 of Gross) house is valued at €157k – In a very constrained market does the estate agents price come to the banks or the other way around? I’ll think you’ll find the price comes to where the money is at. The banks dictate the amount of cash available for the deal – they dictate the price – it makes no difference what the estate agent believes what the price is. Yield is yield.

    Having worked in a lending bank for many years believe me when I say the banks have got their sums wrong for a decade and are soley to blame for this mess.

  22. @Brian Woods 11

    re

    I find this offering from Morgan Kelly truly pathetic.

    Kelly has now supported his claim that the top 11,000 mortgages amount to approx €9 billion.
    Do you have anything other than bluster and bulls**t to refute this claim.

  23. @ Joseph Ryan

    Kelly claimed 10,000 professionals owed more than €1M each. This was immediately refuted by people who have actually studied the figures.

    He now says that 8,000 of these actually owe somewhat less than €1M. To most lawyers, accountants, doctors etc. in prime career, a 750K loan, say, against a 500K property is a major setback for their pension plans but entirely manageable. This is not another taxpayer black hole.

    One notes that MK has not been able to use the Pareto distribution to prove that most of these professionals cannot now afford a cup of coffee, which was the other key pin in his most recent grenade.

    Have you anything other to add rather than puerile ** comments?

  24. Yes.

    If as you contend, but which is assuredly incorrect that

    “most lawyers, accountants, doctors etc. in prime career, a 750K loan, say, against a 500K property is a major setback for their pension plans but entirely manageable. ”
    then these people are part of overpaid elite that are suckling at State’s teat and are very undertaxed. That is if you can these 500K properties anymore.

    Many lawyers and accountants are out of work or struggling. The only reason doctors are super-remunerated is that the State is so supine as to tolerate it. This is also particularly true of ‘tribunal lawyer’ brigade who have milked the State to near death over the past 20 years.

    And I put a lot more credence in Kelly’s figures of 11000/€9 billion that I would put on comments from Brendan Brugess. His statement in today’s independent is reminiscent of McCreevy’s version of economics.

    Negative equity is a messy problem for couples who split up who can’t afford to live separately but who can’t afford to take the loss on the sale. But for most people, while negative equity makes them feel poorer, it has no other impact on their cash flow or lifestyle.

    If negative equity has no impact on cash flow or lifestlye then economics has surely been turned on its head.

  25. @brian woods II: easy there fella – you are beginning to read like one of burgess fawning contributors on askaboutmoney.com – maybe you might like to propose banning someone eh?

  26. @ brianwoodsII why not have a go at reconciling the BTL stats: march 2011 PCAR BTL was €24.5bn for four banks – what was BOSI’s figure before it handed in its lending license? It takes more than a few phone calls to people who’ll talk to you to work out how much is owed, to whom and by whom.

  27. @ Joseph Ryan

    As I understand it, these property deals offered by the Private Banking arms of the banks were to what you and I would regard as “fat cats”. One can empathise with your argument that these cats would not have been nearly so fat, say, in the UK. But that is off the topic as to their ability to pay for their ill advised investments. Regard it as some form of poetic justice that while Ireland uniquley made these cats fat it has also uniquely whipped the cream away from them.

    Undoubtedly there are struggling young lawyers who cannot afford the price of a cup of coffee, but be assured the private bankers would have regarded these as beneath their radar screens.

  28. I think that it is time that someone made some statistical comments on Professor Kelly’s note on the size distribution of Irish Mortgages. I would ask the following questions

    1. Has the Weibull distribution ever been used as a distribution for mortgages? Certainly this is not suggested in Johnson et al (1994). The grouped data here are not sufficient to justify the use of a Weibull distribution.

    2. It is suggested that this distribution is used to measure a long right tail. Then it is suggested that the distribution does not have sufficient weight in the right tail and the distribution is used for the data excluding the right tail. The value of k found implies a tail that decays rapidly

    3. The Weibull parameters are estimated using a regression procedure. Using a graph the fit of the extreme right hand side point is rejected by Professor Kelly as an outlier. Yet the fit is only 1.3 standard deviations from the observed value. This is hardly justification for regarding this point as an outlier

    4. The Wiebull distribution used has a lower bound of zero. The Wiebull distribution can have a third parameter which in this case in this case represents a lower bound for a mortgage. The assumption here is that the minimum mortgage is zero. I don’t think that this is a realistic assumption. The effect of using the zero lower bound is to have some very low value mortgages in the thoeretical model. Thus the process used will underestimate the average mortgage in the lower range and thus overestimate the average in the range covered by the Pareto distribution.

    Basically there is not sufficient data to get at the truth about the upper tail of the distribution of these mortgages. The only way forward is for the Department to release the actual data.

Comments are closed.