Mortgage Arrears, Strategic Default and Repossessions
1 like628 views
The document analyzes mortgage arrears, strategic defaults, and property repossessions, highlighting a strong correlation between mortgage arrears and negative equity. It notes that despite various legal interventions over the past decade, the rate of repossessions remains very low, and two-thirds of mortgages in arrears have not been restructured. The analysis also indicates that the absence of repossession risks contributes to increased mortgage defaults, particularly among borrowers with negative equity.
Mortgage Arrears, Strategic Default and Repossessions
- 1. Mortgage Arrears, Strategic Default and Repossessions Notes on the mortgage arrears, negative equity, causes of strategic defaults and the rate of repossession of properties whose mortgages are in arrears Alan McSweeney http://ie.linkedin.com/in/alanmcsweeney
- 2. Mortgage Arrears, Default and Repossessions Page 2 Contents Introduction.......................................................................................................................................... 2 What is Strategic Mortgage Default? ..................................................................................................... 4 Sources of Mortgages and Their Defaults ............................................................................................... 4 Irish Mortgage Data Sources.................................................................................................................. 5 Strategic Default Analysis Material ....................................................................................................... 6 Predictors of Mortgage Arrears and Strategic Default ............................................................................ 6 House Price Data Notes....................................................................................................................... 11 Legislative and Other Interventions in the Mortgage Market ............................................................... 13 Central Bank of Ireland Restructured Mortgages and Mortgages in Arrears Not Restructured ............. 16 Repossessions ...................................................................................................................................... 21 Local Authority Mortgage Defaults ..................................................................................................... 23 Mortgage Arrears Cohort Analysis ....................................................................................................... 30 IFRS 9, Non-Performing Loans and Repossessions .............................................................................. 33 Summary............................................................................................................................................. 34 Summary Arrears in mortgages appear to be closely correlated with the amount of negative equity. In the last 10 years, there have been many legal and regulatory interventions that have affected the way in which properties whose mortgages are in arrears can be repossessed. The repossession route is still long, slow and expensive. Two thirds of mortgages in arrears have not been subject to any form of restructuring. The rate of and thus the risk of repossessions is extremely low. The correlation between the number of arrears and the number of repossessions is very low. IFRS 9 will cause banks to sell non-performing loans in bulk rather than attempting the time-consuming and expensive process of trying to engage with a core of non-engagers that have been in arrears for some time. A very high proportion of Local Authority mortgages are in arrears. Many of these arrears are more than 20 years old.
- 3. Mortgage Arrears, Default and Repossessions Page 3 Introduction These notes are a macro-level analysis of the issues of mortgage default and repossessions. In February 2018, the Central Bank of Ireland published a paper 1RT18 The Impact of Repossession Risk on Mortgage Default (https://www.centralbank.ie/docs/default-source/publications/research- technical-papers/1rt18-the-impact-of-repossession-risk-on-mortgage-default-(o%27malley).pdf). The objectives of this Central Bank of Ireland paper were: This paper evaluates the claim that reducing repossession risk for homeowners leads to an increase in mortgage default. Economic theory predicts that borrowers will be more likely to default on their mortgages if their homes cannot be repossessed. After the financial crisis, commentators frequently cited the lack of repercussions as one of the contributing reasons for high mortgage arrears in Ireland. I evaluate this claim by examining how mortgage arrears evolved during the recent "Dunne judgment" period in Ireland. The legal judgment effectively removed the ability of banks to repossess homes in the event of mortgage default. The terms of the judicial decision meant that nearly every mortgaged household in the country could no longer lawfully have their homes repossessed from mid-2011 to 2013. Crucially for the analysis in this paper, a group of mortgaged households experienced no change in their repossession risk: they were exempt from the ruling of Justice Dunne. This aspect of the ruling allows to me to conduct a quasi- experimental evaluation of the impact of removing repossession risk on mortgage default: default rates for borrowers who had their repossession risk removed are compared to similar borrowers who experienced no change in the repossession regime in the event of default. Though not the ultimate cause of the arrears crisis, I find that the removal of repossession risk led to an immediate increase in mortgage default for affected borrowers. Borrowers experiencing very low or negative levels of home equity are the most likely to default in response to the removal of repossession risk. However, this notion of purely strategic default is moderated by evidence that these "strategic defaulters" were also more likely to be in financial difficulty before the ruling. They were more likely to have missed a payment before the Dunne judgment, have lower incomes and also face higher interest rates on their mortgages. Policy implications are straightforward. Impediments to home repossession by banks reduce a borrower’s incentive to fulfil the terms of their mortgage. While a policy aiming to reduce repossession risk may benefit borrowers, it would also increase the mortgage default rate. When considering changes to repossession law, policy-makers must trade off the benefits from lower home repossessions with the moral hazard cost I have identified in this work. These notes contain some comments on this paper. They examine mortgage data from a wider range of sources that did the Central Bank to assess what are the determinants of strategic mortgage default. They also look at the wider range of legal and regulatory interventions that may have affected the issue of repossessions than were considered by the Central Bank. These notes are just a very minor contribution to the wide range of analyses on this subject
- 4. Mortgage Arrears, Default and Repossessions Page 4 What is Strategic Mortgage Default? Strategic default is a very emotive topic. It is difficult to have a rational conversation about it. Media coverage tends to vary considerably and inconsistently. Strategic default also has different meanings and results in different outcomes in different jurisdictions. So an Irish definition should be agreed initially. In my view, strategic default happens when a borrower decides to stop making repayments on a loan even though the borrower has the financial ability and resources to make the payments (won’t pay rather than can’t pay). This can be temporary or permanent. It can be in pursuit of a short-term gain or a part of a longer-term strategy to have some or all of the loan or accumulated arrears or both written off. For a borrower to stop repaying a mortgage deliberately there would have to be a set of circumstances where the probability of the borrower suffering negative consequences from their actions would be low. This can happen in circumstances where some or all of the following apply: Where the lender has no recourse on the mortgage loan other than the property Where if the lender takes charge of the collateral they amount they recover is substantially less than the loan amount Where the process for the lender to recover the arrears or enforce a penalty or recover the underlying collateral is slow, complex, expensive or difficult Where the lender feels that the value of the property will increase over time so waiting will increase the amount recovered Where the lender will suffer a disadvantage by repossessing and selling the property and having to crystallise the loss on their balance sheet Sources of Mortgages and Their Defaults There are three sources of residential mortgage lending in Ireland: 1 Regulated financial institutions such as retail banks and mortgage lenders 2 Local authorities 3 Credit unions The breakdown in outstanding mortgage balances across these sectors is:
- 5. Mortgage Arrears, Default and Repossessions Page 5 The values for regulated financial institutions include PDH (Primary Dwelling House) and BTL (Buy- To-Let). The balance amounts are in millions. These numbers exclude commercial residential property lending. Different types of strategic default may occur for each of these three lender types, depending on factors such as the assertiveness stance of the lender in relation to pursuing arrears. The Central Bank paper looked at the effect of repossessions on the PDH area. Irish Mortgage Data Sources Public mortgage arrears, house price and related data are available from multiple sources such as: Department of the Environment - http://www.housing.gov.ie/search/sub-topic/house-prices-loans- and-profile-borrowers CSO - http://www.cso.ie/px/pxeirestat/Database/eirestat/House%20Prices/House%20Prices_statbank.asp Central Bank - https://www.centralbank.ie/statistics/data-and-analysis/credit-and-banking- statistics/mortgage-arrears BPFI - https://www.bpfi.ie/publications/bpfi-mortgage-drawdowns/ The problems with this publically available data are: It is at a high level and not very detailed or granular The start and end dates of the time series are not consistent The data intervals are not consistent across the time series with some series containing monthly data and others quarterly The level of detail is inconsistent – national, regional, property type, property age So combining the data requires that it be done at the highest common denominator which tends to be national data for all property types for new and second hand properties quarterly.
- 6. Mortgage Arrears, Default and Repossessions Page 6 Detailed mortgage data is available privately to financial regulators and to individual financial institutions that can perform detailed analyses. This data is not available publically. Strategic Default Analysis Material There is much excellent material available on the subject of strategic default, both in Ireland and elsewhere, of which the following is a very small subset: Irish Mortgage Default Optionality, Economics, Finance and Accounting Department Working Paper Series n243-13.pdf, Department of Economics, Finance and Accounting, National University of Ireland – Maynooth - http://repec.maynoothuniversity.ie/mayecw-files/N243-13.pdf Moral and Social Constraints to Strategic Default on Mortgages - http://www.financialtrustindex.org/images/Guiso_Sapienza_Zingales_StrategicDefault.pdf Recourse and Residential Mortgage. Default: Theory and Evidence from. U.S. States. WP 09-10R - https://www.richmondfed.org/~/media/richmondfedorg/publications/research/working_papers/200 9/pdf/wp09-10r.pdf Mortgage Modification and Strategic Behavior: Evidence from a Legal Settlement with Countrywide - https://chicagounbound.uchicago.edu/cgi/viewcontent.cgi?referer=https://www.google.ie/&httpsre dir=1&article=1003&context=housing_law_and_policy Can't Pay or Won't Pay? Unemployment, Negative Equity, and Strategic Default Can't Pay or Won't Pay? Unemployment, Negative Equity, and Strategic Default - https://www.bostonfed.org/publications/research-department-working-paper/2015/cant-pay-or- wont-pay-unemployment-negative-equity-and-strategic-default.aspx Mortgage Modifications after the Great Recession New Evidence and Implications for Policy - https://www.jpmorganchase.com/corporate/institute/document/institute-mortgage-debt- reduction.pdf The Effect of Debt on Default and Consumption: Evidence from Housing Policy in the Great Recession - https://scholar.harvard.edu/files/noel/files/ganong_noel_housing_2017-12-16.pdf Resolving a Non-Performing Loan crisis: The ongoing case of the Irish mortgage market - https://www.centralbank.ie/docs/default-source/publications/research-technical-papers/10rt17--- resolving-a-non-performing-loan-crisis-the-ongoing-case-of-the-irish-mortgage-market.pdf Some defaults are deeper than others Understanding long-term mortgage arrears - https://www.centralbank.ie/docs/default-source/publications/research-technical-papers/research- technical-paper-05rt15.pdf Predictors of Mortgage Arrears and Strategic Default It is possible to do some very simple analysis with the limited set of publically available information to determine the likely predictors of strategic default.
- 7. Mortgage Arrears, Default and Repossessions Page 7 This table below combines data from: Department of the Environment Quarterly Average Second Hand Property Price by Area http://www.housing.gov.ie/sites/default/files/publications/files/form_41d-price-sh-property- area-by-qtr_1.csv National New House Prices by agency - by quarter http://www.housing.gov.ie/sites/default/files/publications/files/form_42f-agency-new-house-price- by-qtr_1.csv CSO HPM06 Residential Property Price Index by Month, Type of Residential Property and Statistic http://www.cso.ie/px/pxeirestat/Statire/SelectVarVal/Define.asp?maintable=HPM06&P Language=0 Central Bank Residential Mortgage Arrears and Repossession Statistics: Data https://www.centralbank.ie/docs/default-source/statistics/data-and-analysis/credit-and-banking- statistics/mortgage-arrears/mortgage-arrears-data/moa-data.xlsx This data applies to residential data for Primary Dwelling House (PDH) mortgages rather than Buy To Let mortgages. The last interval for which all data is available is 2016Q1. I converted monthly data to quarters by a simple average rather than a more complex interpolation. The Estimated Positive/ Negative Equity is derived by the difference between the CSO House Price Index for the quarter relative to 2016Q1 multiplied by the average property price for that quarter taken from the original property price, So a property bought in 2005Q1 for €298,986 has a notional value of 0.8236 times its original price in 2016Q1 (CSO index at 2016Q1 82.97 / divided by CSO index at 2005Q1 100.73) or €246,245. This represents an estimated negative equity of €52,741. This is a rather artificial value. The calculations of this are: Interval Average New and Second Hand Home Price Nationally CSO House Price Index CSO House Price Index Relative to 2016Q1 Estimated Positive/ Negative Equity Central Bank PDH Arrears Number 2005Q1 €298,986 100.73 0.8236 -€50,206 2005Q2 €323,921 102.87 0.8065 -€58,766 2005Q3 €320,316 106.87 0.7764 -€67,504 2005Q4 €342,193 111.57 0.7437 -€81,860 2006Q1 €340,765 113.60 0.7303 -€86,719 2006Q2 €368,758 118.10 0.7025 -€102,234 2006Q3 €378,175 125.20 0.6627 -€117,909
- 8. Mortgage Arrears, Default and Repossessions Page 8 Interval Average New and Second Hand Home Price Nationally CSO House Price Index CSO House Price Index Relative to 2016Q1 Estimated Positive/ Negative Equity Central Bank PDH Arrears Number 2006Q4 €366,516 127.93 0.6485 -€121,028 2007Q1 €371,333 130.17 0.6374 -€127,067 2007Q2 €379,008 130.63 0.6351 -€131,166 2007Q3 €366,391 130.47 0.6359 -€126,263 2007Q4 €359,288 129.73 0.6395 -€122,789 2008Q1 €352,293 127.00 0.6533 -€116,219 2008Q2 €350,409 123.53 0.6716 -€110,061 2008Q3 €330,820 120.23 0.6900 -€98,788 2008Q4 €317,416 114.07 0.7274 -€82,536 2009Q1 €291,166 106.07 0.7822 -€60,145 2009Q2 €280,331 98.77 0.8400 -€42,252 2009Q3 €251,627 94.67 0.8764 -€30,153 63,619 2009Q4 €242,044 92.47 0.8973 -€24,205 69,647 2010Q1 €244,447 89.30 0.9291 -€16,801 76,100 2010Q2 €272,153 86.27 0.9617 -€9,691 82,377 2010Q3 €273,145 83.73 0.9908 -€2,340 86,362 2010Q4 €266,387 80.07 1.0362 €9,094 89,234 2011Q1 €269,965 76.40 1.0860 €22,197 94,518 2011Q2 €268,989 72.47 1.1449 €36,760 102,397 2011Q3 €249,696 68.13 1.2177 €52,239 110,597 2011Q4 €243,668 64.33 1.2896 €68,338 118,464 2012Q1 €243,558 61.47 1.3498 €81,130 122,941 2012Q2 €248,271 60.03 1.3820 €91,528 128,197 2012Q3 €247,499 60.73 1.3661 €86,596 141,389 2012Q4 €240,357 61.23 1.3549 €81,839 139,224 2013Q1 €242,233 59.47 1.3952 €92,954 142,118 2013Q2 €254,868 59.50 1.3944 €95,535 142,892 2013Q3 €253,732 62.87 1.3197 €78,250 141,269 2013Q4 €253,627 64.67 1.2830 €69,116 136,558 2014Q1 €250,369 65.70 1.2628 €64,024 132,217 2014Q2 €266,782 69.37 1.1961 €50,281 126,005 2014Q3 €261,930 74.97 1.1067 €27,308 117,889 2014Q4 €256,397 77.30 1.0733 €18,875 110,366 2015Q1 €260,564 77.20 1.0747 €19,679 104,693 2015Q2 €268,012 78.93 1.0511 €13,848 98,155 2015Q3 €270,065 81.47 1.0184 €5,087 92,361 2015Q4 €268,682 82.67 1.0036 €1,020 88,292 2016Q1 €270,848 82.97 1.0000 €0 85,989 2016Q2 83.90 82,091 2016Q3 87.53 79,562 2016Q4 89.73 77,493 2017Q1 90.77 76,422 2017Q2 92.77 73,706 2017Q3 97.83 72,489 2017Q4 100.37 70,488 2018Q1 101.97 This is fairly crude. National property prices for all property types and ages are used. This is also a one-dimensional view of estimated negative equity.
- 9. Mortgage Arrears, Default and Repossessions Page 9 Also, the estimate of negative equity is effectively a proxy for (1 - CSO House Price Index). This in turn is a proxy for overall economic circumstances. So the correlation could be viewed as being between number of arrears and the state of the economy. Falling property prices are a consequence of a generally falling economy. The apparent link between mortgage default and negative equity may therefore not be strictly causal but indicative of a shared third common factor of general economic conditions. The following chart shows the number of PDH arrears on the left vertical axis mapped against the estimated negative equity on the right vertical axis. Negative equity is estimated with respect to 2016Q1. A positive value for negative equity means the property has an estimated negative equity. A negative value indicates an estimated positive equity. There certainly appears to be a relationship between the number of arrears and the estimated negative equity, Simple measures of correlation such as the correlation coefficient or R2 (which is just the square of the correlation coefficient) yield values such as 0.956 and 0.914. However, it is more than possible that this correlation is an example of spurious regression in two time series where there is no causal relationship between the two series but each are related to a common third series such as the overall economy. There are complex methods for determining if a statistical relationship is due to spurious regression. These are outside the scope of these notes. The following shows Ireland’s GDP and GDP and the unemployment rate from 2017Q1.
- 10. Mortgage Arrears, Default and Repossessions Page 10 These measures all have their issues. GDP and GNP arte distorted by the effect of multi-nationals. The rate of unemployment is affected by the various activation and education programmes and the number being allocated to disability payments. The unemployment rate, being a proportion of the employable population, will also be affected by emigration that increased during the financial crisis that decreased the size of the employable population. It is possible to attempt to correct for these factors. The adjustment would be complicated by factors such as those who emigrated would be less likely to be mortgage holders and therefore less likely to be in mortgage arrears. This is outside the scope of these notes. The following chart tracks the unemployment rate shown in red on the right hand axis with the number of arrears shown in blue on the left axis. At best, it could be said that any relationship indicates that the increase in the number of arrears tracked the increase in the unemployment rate at a lag of 2 or more years.
- 11. Mortgage Arrears, Default and Repossessions Page 11 House Price Data Notes The Department of the Environment published quarterly property price data for both new and second- hand properties at the links listed above. New property prices are available separately from 1975. Second-hand property prices are available separately from 1978. The CSO retrospectively published new and second-hand monthly property price data from 2010 in the time series HPM02 Residential Dwelling Property Transactions by County, Dwelling Status, Stamp Duty Event, Type of Buyer, Type of Sale, Month and Statistic http://www.cso.ie/px/pxeirestat/Statire/SelectVarVal/Define.asp?maintable=HPM02&PLanguage=0. These two sets of property price data do not agree. The following table shows the two data sets with the CSO monthly price average averaged over the quarter. Quarter Department of the Environment New Department of the Environment Second Hand CSO New CSO Second Hand 2010Q1 €226,245 €247,534 €244,210 €190,959 2010Q2 €226,833 €279,839 €212,800 €190,392 2010Q3 €230,868 €280,315 €219,423 €194,841 2010Q4 €229,531 €272,638 €210,309 €161,716 2011Q1 €241,749 €274,750 €202,256 €167,053 2011Q2 €232,174 €275,233 €189,240 €160,220 2011Q3 €226,215 €253,678 €188,864 €168,737 2011Q4 €225,067 €246,823 €169,002 €144,800 2012Q1 €215,587 €248,302 €185,498 €141,736 2012Q2 €227,376 €251,815 €179,850 €141,129 2012Q3 €221,123 €251,972 €184,619 €167,139 2012Q4 €216,810 €244,350 €166,451 €152,757 2013Q1 €225,340 €245,098 €151,251 €151,637 2013Q2 €224,432 €260,030 €172,018 €155,931 2013Q3 €232,083 €257,404 €176,995 €172,730 2013Q4 €231,011 €257,462 €165,410 €182,428 2014Q1 €234,098 €253,128 €172,858 €159,719 2014Q2 €241,912 €271,000 €172,639 €178,620 2014Q3 €247,398 €264,394 €199,967 €205,165 2014Q4 €258,989 €255,958 €210,472 €188,278 2015Q1 €267,517 €259,385 €195,888 €189,604 2015Q2 €275,235 €266,787 €206,389 €183,984 2015Q3 €285,015 €267,530 €234,526 €202,021 2015Q4 €298,551 €263,616 €253,016 €193,636 2016Q1 €309,703 €264,258 €267,172 €197,551 2016Q2 €273,188 €200,899 2016Q3 €282,714 €216,816 2016Q4 €298,384 €212,193 2017Q1 €299,364 €214,622 2017Q2 €309,452 €216,723 2017Q3 €315,823 €234,758 2017Q4 €325,086 €232,367 2018Q1 €343,541 €228,941 The following chart shows this information visually.
- 12. Mortgage Arrears, Default and Repossessions Page 12 The CSO prices are significantly lower than those published by the Department of the Environment. It is outside the scope of these notes to reconcile these differences. The Central Bank of Ireland also changed historical data in their arrears time series without any notification or explanation. The following table lists differences between the data published in September 2017 and December 2017. Mar-17 Jun-17 Sep-17 Number Balance € (000) Arrears € (000) Number Balance € (000) Arrears € (000) Number Balance € (000) Arrears € (000) Total outstanding classified as restructured - at end of quarter -95 -18,571 -675 -26 -7,570 -448 -19 -6,704 -546 Interest Only Interest Only - up to one year Interest Only - over one year -4 -760 -9 -3 -431 -14 -3 -424 -14 Reduced Payment (less than interest only) -51 -9,994 -930 -18 -3,889 -289 -7 -1,654 -262 Reduced Payment (greater than interest only) Term Extension Arrears Capitalisation -118 -16,495 202 -117 -18,012 -388 -133 -21,266 -588 Payment Moratorium Deferred Interest Scheme Split Mortgage -52 5 -58 8 -58 9 Permanent Interest Rate Reduction -81 -13,991 -239 -70 -12,065 -125 -73 -12,584 -151 Temporary Interest Rate Reduction Trade Down Mortgages Other 159 22,721 296 182 26,885 360 197 29,282 460 of which are not in arrears 1 -3,136 16 -189 31 1,387
- 13. Mortgage Arrears, Default and Repossessions Page 13 Legislative and Other Interventions in the Mortgage Market The intent of the Central Bank of Ireland paper was to determine the impact, if any the Dunne Judgment had on repossessions. This was singled-out because of its apparent effect of freezing repossessions because of the uncertainty it introduced However, there have been many legislative, regulatory and other administrative interventions into the mortgage market during the interval of the financial crash. Any analysis of repossessions should look at the range of these interventions in their entirety to attempt to determine their impact on mortgage default behaviour. These legislative interventions were: Land and Conveyancing Law Reform Act 2009 (http://www.irishstatutebook.ie/eli/2009/act/27/enacted/en/html) enacted on 1 Dec 2009 – S.I. No. 356/2009 - Land and Conveyancing Law Reform Act 2009 (Commencement) Order 2009 (http://www.irishstatutebook.ie/eli/2009/si/356/made/en/print) November 2010 Cooney Report - Irish Expert Group on Mortgage Arrears and Personal Debt. The Cooney Report introduced the Central Bank MARP process. January 2011 Central Bank of Ireland Mortgage Arrears Resolution Process (MARP) https://www.centralbank.ie/docs/default-source/Regulation/consumer-protection/other-codes-of- conduct/24-gns-4-2-7-2013-ccma.pdf September 2011 Inter-Departmental Mortgage Arrears Working Group http://www.finance.gov.ie/wp-content/uploads/2017/08/170828-Keane-Report-30-September- 2011.pdf July 2013 Central Bank of Ireland Code of Conduct on Mortgage Arrears (CCMA) https://www.centralbank.ie/docs/default-source/Regulation/consumer-protection/other-codes-of- conduct/24-gns-4-2-7-2013-ccma.pdf Start Mortgages & Ors v Gunn & Ors – the Dunne Judgment http://www.courts.ie/judgments.nsf/6681dee4565ecf2c80256e7e0052005b/89f3e895ac665956802578da 002fcd0e?OpenDocument&Highlight=0,Gunn – 25 July 2011 Land and Conveyancing Law Reform Act 2013 (http://www.irishstatutebook.ie/eli/2013/act/30/enacted/en/html) enacted on 31 Jul 2013 - S.I. No. 289/2013 - Land and Conveyancing Law Reform Act 2013 (Commencement) Order 2013 (http://www.irishstatutebook.ie/eli/2013/si/289/made/en/print) December 2013 Report of the Expert Group on Repossessions - http://www.justice.ie/en/JELR/ExpGroupReportFinal.pdf/Files/ExpGroupReportFinal.pdf There was also the Personal Insolvency Act, 2012 http://www.irishstatutebook.ie/eli/2012/act/44/enacted/en/html which had an effect on how arrears were handled. Elements of this Act came into effect on different times:
- 14. Mortgage Arrears, Default and Repossessions Page 14 1 Mar 2013 - S.I. No. 63/2013 - Personal Insolvency Act 2012 (Commencement) (No. 2) Order 2013 http://www.irishstatutebook.ie/eli/2013/si/63/made/en/print 18 Jan 2013 - S.I. No. 14/2013 - Personal Insolvency Act 2012 (Part 6) (Commencement) Order 2013 http://www.irishstatutebook.ie/eli/2013/si/14/made/en/print 31 July 2013 S.I. No. 285/2013 - Personal Insolvency Act 2012 (Commencement) (No. 3) Order 2013 http://www.irishstatutebook.ie/eli/2013/si/285/made/en/print Dec 2013 S.I. No. 462/2013 - Personal Insolvency Act 2012 (Part 4) (Commencement) Order 2013 http://www.irishstatutebook.ie/eli/2013/si/462/made/en/print Any analysis of the effects of external legal interventions on arrears and on repossessions should look at the relatively active legal and regulatory landscape over the 8 year interval being analysed. It would be difficult to determine the impact of a single intervention in the context of these series of interventions that introduced doubt and uncertainty into the repossession process. The following overlays the times of some of these key legislative interventions that may have affected mortgage arrears on the original mortgage arrears and negative equity data. What we are trying to analyse and estimate here is whether borrowers in arrears change their behaviour in response to external positive or negative incentives. So, is negative equity a predictor of arrears and thus an indicator of strategic default or do the two share a common third cause? Can the mortgage holders in market be segmented into a number of groups, such as:
- 15. Mortgage Arrears, Default and Repossessions Page 15 Those who will resume paying their mortgage when they are able to do so Those who resume paying their mortgage when the negative equity is such they see a benefit in paying to retain the asset Those who do not resume paying their mortgage because they see they do not see any consequences to not paying Intuitively, the size of the negative equity could lead to a decision on behalf of the borrower to stop repaying the loan because the underlying asset has not value and a view that they are only throwing good money after bad. When the borrower stops making loan repayments, the property cannot be sold. This would be accompanied by a view from the lender that repossession of a poor quality distressed asset would lead to them having to recognise their losses after selling the recovered property at a significant loss and a view from the borrower that this is also the case. This would then be combined with a complex, lengthy, cost and uncertain repossession process that acts as a barrier to repossessions. Conversely, when the negative equity falls, the borrower is more likely to make repayments because the load to value ratio is reduced. The borrower may be more inclined to accept a small loss when selling the property in order to be able to more to a more desirable property. The borrower may also perceive that the lender is more likely to seek repossession as the loss they would have to accept is reduced. This intuitive insight may however be simplistic and is certainly speculative. As I said above, the estimated negative equity amount is simply a proxy for (1 - CSO House Price Index). By way of comparison, this chart shows the Department of Environment national new and second hand property price mapped against the CSO national House Price Index, shown on the right vertical axis. The divergence in the average price of new and second-hand property is an interesting, if minor, secondary observation. The following shows the two dimensional view of the progress of average estimated national negative equity for domestic property bought in any quarter at all subsequent quarters. It really demonstrates the massive trough of negative equity that occurred at the depth of the financial crisis.
- 16. Mortgage Arrears, Default and Repossessions Page 16 Central Bank of Ireland Restructured Mortgages and Mortgages in Arrears Not Restructured The PDH arrears statistics https://www.centralbank.ie/statistics/data-and-analysis/credit-and-banking- statistics/mortgage-arrears contains the following rows: A Total mortgage arrears cases outstanding - at end of quarter which are:
- 17. Mortgage Arrears, Default and Repossessions Page 17 B Total outstanding classified as restructured - at end of quarter C of which are not in arrears So the number of mortgages that are in arrears and have been restructured is B – C. The number of mortgages that are in arrears and have not been restructured is A – B + C. The following table shows the calculations for the last five quarters: Classification Dec-16 Mar-17 Jun-17 Sep-17 Dec-17 Total Residential Mortgage Loan Accounts Outstanding 736,894 734,106 732,439 731,119 729,722 Total Mortgage Arrears Cases Outstanding 77,493 76,422 73,706 72,489 70,488 Total Outstanding Classified As Restructured 120,944 120,641 120,047 119,051 118,477 Restructured Not In Arrears 94,441 94,335 94,656 94,012 92,999 In Arrears Restructured 26,503 26,306 25,391 25,039 25,478 In Arrears Not Restructured 50,990 50,116 48,315 47,450 45,010 Total Mortgage Arrears Cases Outstanding % 10.52% 10.41% 10.06% 9.91% 9.66% In Arrears Not Restructured % 65.80% 65.58% 65.55% 65.46% 63.85% From this, the following can be seen: A large number of mortgages have been restructured without being classified as being in arrears: A much smaller number of mortgages that are in arrears have been restructured A larger number of mortgages in arrears have not been restructured - nearly twice the number that are in arrears and have been restructured This chart profiles the numbers of mortgages to which various types of restructuring have been applied over time:
- 18. Mortgage Arrears, Default and Repossessions Page 18 It also shows the number restructured mortgages. The number of restructured mortgages reached a peak in September 2015 and has remained roughly at this level since then. The key trends in the type of restructuring arrangement are isolated in the following chart: These key trends are: Split Mortgage increased to the current position where they account for nearly one quarter of restructuring arrangements Arrears Capitalisation now account for one third of arrangements Reduced Payment (Greater Than Interest Only) have fallen to less than 5% of arrangements Reduced Payment (Less Than Interest Only) have fallen to less than 0,5% of arrangements Interest Only - Up To One Year have fallen to less than 1,5% of arrangements This chart profiles the balances of mortgages to which various types of restructuring have been applied over time:
- 19. Mortgage Arrears, Default and Repossessions Page 19 It also shows the balance of the restructured mortgages. The balance of restructured mortgages reached a peak in September 2015 and has been dropping since then. The following chart contains a profile of the numbers of mortgages in arrears that have been restructured and those that have not. Roughly two thirds of mortgages in arrears have not been and have never been restructured. The following chart contains a profile of the balance of mortgages in arrears that have been restructured and those that have not. The vertical axis amounts are in thousands.
- 20. Mortgage Arrears, Default and Repossessions Page 20 Again, roughly two thirds of mortgages balances in arrears have not been restructured. This shows a residual high-level of non-engaged or not-engaged mortgages in arrears. The following chart shows the proportions of arrears that have been restructured and not restructured over the interval:
- 21. Mortgage Arrears, Default and Repossessions Page 21 This illustrates the substantial number of mortgages in arrears have yet to be restructured. This points to a core of non-engaging mortgage holders who are in arrears. Repossessions The following chart shows the percentage of mortgages in arrears and the number of repossessions per quarter as a percentage of mortgages in arrears. This chart shows the number of repossessions – those as a result of a legal process and those voluntarily surrendered – on the right vertical axis and the number of mortgages in arrears. It also overlays the dates of legal interventions listed above to provide a context for any changes.
- 22. Mortgage Arrears, Default and Repossessions Page 22 Intuitively, the legal interventions do not appear to have impacted repossessions. This may not be the complete picture. It may be that the various legal interventions caused repossession legal cases in progress to be abandoned. It may have been that financial institutions were preparing to initiate greater volumes of repossessions and that the Dunne judgement and other interventions caused them to stop this activity. However this speculation cannot be verified. The low rate of repossessions, particularly those on foot of legal action are so very low. The following chart overlays the previous chart with an estimated linear extrapolation of the Legal Repossession values from the intervals Sep 09 to Sep 11 for the intervals before the Dunne judgement occurred to the later intervals Dec 11 to Jun 13.
- 23. Mortgage Arrears, Default and Repossessions Page 23 The rate of repossessions is very low. The risk of involuntary repossession is very low. The total number of PDH repossessions over the measured interval of over eight years is 8,506. Of these 2,860 repossessions occurred on foot of an order. The balance of 5,646 were voluntarily surrendered or abandoned. The correlation between the number of arrears and the number of repossessions is very low. Local Authority Mortgage Defaults While Local Authority mortgage lending is not regulated by the Central Bank (or any other regulator), it provides an interesting potential contrast to causes of arrears. The subject of Local Authority arrears was not covered in the Central Bank paper. Local Authority home loan statistics are available from http://www.housing.gov.ie/housing/statistics/house-prices-loans-and-profile-borrowers/local-authority- loan-activity. The profile of PDH loan arrears in the latest home loan statistics for December 2017 produced by the Central Bank at https://www.centralbank.ie/statistics/data-and-analysis/credit-and-banking- statistics/mortgage-arrears is: Number Balance Number % Of Total Value % Of Total Total Home Loans 729,722 €98,521,574,000 In Arrears Over 90 Days 48,433 €9,694,109,000 6.64% 9.84% In Arrears Over 720 Days 28,946 €6,417,284,000 3.97% 6.51% The similar arrears profile for the latest Local Authority home loans for Q4 2017 is: Number Value Number % Of Total Value % Of Total
- 24. Mortgage Arrears, Default and Repossessions Page 24 Total Home Loans 15,893 €913,370,742 In Arrears Over 90 Days 3,807 €400,867,568 23.95% 43.89% In Arrears Over 720 Days 1,534 €76,234,873 9.65% 8.35% The arrears data is in different formats over different intervals. The following table aggregates the available data into a single view from Q1 2010 to Q3 2017. The blank cells are because of discontinuities in the Local Authority home loan data. I have filled-in some of the blanks. The arrears data is in different formats over different intervals. The following table aggregates the available data into a single view from Q1 2010 to Q3 2017. The blank cells are because of discontinuities in the Local Authority home loan data. I have filled-in some of the blanks.
- 25. Mortgage Arrears, Default and Repossessions Page 25 Year and Quarter Total Loan Book Loans In Arrears Between 0 - 90 Days Loans In Arrears Between 90 - 180 Days Loans In Arrears Over 180 Days Loans In Arrears Over 180-360 Days Loans In Arrears Over 360-720 Days Loans In Arrears Over 720 Days Number Value € Average Outstanding Loan Amount Number Value € Number Value € Number Value € Number Value € Number Value € 2010Q3 24,405 €1,345,610,918 €55,137 1,735 €54,358,247 4,325 €114,895,907 2010Q4 23,909 €1,316,137,267 €55,048 1,577 €56,666,806 4,488 €120,978,701 2011Q1 23,560 €1,304,652,037 €55,376 1,506 €49,503,439 4,501 €126,300,626 2011Q2 23,112 €1,290,622,536 €55,842 1,656 €60,956,340 4,541 €132,395,697 2011Q3 22,616 €1,277,119,336 €56,470 1,543 €62,702,965 4,656 €141,561,228 2011Q4 22,008 €1,263,112,738 €57,393 1,629 €66,727,283 4,715 €153,608,643 2012Q1 22,390 €1,261,818,255 €56,356 1,556 €64,930,329 4,759 €161,657,541 2012Q2 22,071 €1,246,408,605 €56,473 1,479 €60,413,426 4,801 €152,201,999 2012Q3 21,061 €1,236,974,251 €58,733 1,463 €59,437,595 4,870 €174,420,725 2012Q4 20,802 €1,226,067,605 €58,940 1,391 €58,781,758 4,919 €180,382,194 2013Q1 20,409 €1,212,398,496 €59,405 1,340 €55,658,799 4,975 €191,380,443 2013Q2 20,339 €1,203,272,046 €59,161 1,307 €54,285,309 4,968 €193,245,375 2013Q3 20,015 €1,189,764,959 €59,444 1,289 €51,994,280 4,921 €194,289,971 2013Q4 19,788 €1,177,152,139 €59,488 1,243 €52,194,614 4,892 €197,172,461 2014Q1 19,065 €1,161,007,843 €60,897 1,241 €50,236,833 4,774 €195,889,598 2014Q2 19,083 €1,141,160,222 €59,800 1,174 €47,437,069 4,749 €192,745,007 2014Q3 18,912 €1,126,834,181 €59,583 1,113 €45,952,435 4,714 €197,365,093 2014Q4 18,679 €1,108,323,965 €59,335 1,037 €42,862,988 4,631 €198,245,421 2015Q1 18,559 €1,098,248,646 €59,176 1,038 €57,726,874 4,494 €300,720,759 1,128 €74,497,753 1,184 €83,536,914 2,182 €142,685,592 2015Q2 18,364 €1,087,983,057 €59,245 995 €55,037,600 4,380 €299,543,884 1,078 €68,882,579 1,124 €82,222,507 2,178 €148,438,798 2015Q3 18,079 €1,070,892,303 €59,234 951 €52,688,282 4,261 €286,575,011 1,040 €63,030,660 1,110 €80,232,295 2,111 €143,312,057 2015Q4 17,845 €1,061,931,494 €59,509 892 €47,129,682 4,091 €281,477,264 1,003 €60,211,997 1,039 €73,656,948 2,049 €147,608,318 Q4 2015 17,992 €1,047,135,254 €58,200 3,476 €199,593,982 863 €48,014,763 3,776 €237,796,796 943 €56,887,485 995 €67,455,871 1,838 €113,453,440 Q1 2016 17,669 €1,028,012,545 €58,182 3,617 €204,791,379 935 €49,472,276 3,896 €231,120,410 942 €55,718,426 1,024 €62,387,189 1,930 €113,014,795 Q2 2016 17,411 €1,014,204,922 €58,251 3,323 €185,421,158 845 €43,378,856 3,741 €221,162,762 897 €51,853,886 978 €59,375,103 1,866 €109,933,773 Q3 2016 17,182 €993,443,770 €57,819 3,187 €176,770,253 820 €41,746,370 3,668 €208,526,260 877 €49,197,556 953 €56,961,879 1,838 €102,366,825 Q4 2016 16,884 €970,555,810 €57,484 3,336 €193,307,698 875 €44,645,924 3,497 €193,118,924 840 €47,320,923 924 €53,160,748 1,733 €92,637,254 Q1 2017 16,591 €953,660,780 €57,481 3,451 €202,843,304 818 €42,515,138 3,355 €183,303,815 801 €44,642,757 892 €51,092,998 1,662 €87,568,059 Q2 2017 16,352 €940,913,690 €57,541 3,398 €199,637,542 763 €37,560,638 3,266 €177,840,951 782 €43,777,973 866 €49,347,330 1,618 €84,715,649 Q3 2017 16,107 €924,553,638 €57,401 3,375 €199,007,989 761 €37,769,868 3,192 €170,945,610 768 €43,151,107 852 €47,542,445 1,572 €80,252,058
- 26. Mortgage Arrears, Default and Repossessions Page 26 In summary, the current amount of Local Authority home lending is €913.4 million. This compares to the €146.3 million of Credit Unions home lending or over 6 times the value. However the total value of Local Authority home lending is only 0.92% of the PDH home loan lending by retail banks. So it is still relatively quite small. The proportion of the number of Local Authority mortgages In Arrears Over 90 Days is 3.6 times that of retail banks. The proportion of the value of mortgages In Arrears Over 90 Days is nearly 4.5 times that of retail banks. The average outstanding loan amount for Local Authorities - €57,470 - when compared to retail banks - €135,012. The In Arrears Over 720 Days numbers for Local Authorities is over 2.4 times that of the retail banks. It is possible that the different profile of borrowers availing of Local Authority loans will give rise to a higher default rate because of a possible higher risk profile. But without more information, this is just an unvalidated statement. The available information is at a very gross level: not broken down by age of loan, original loan amount, type of property and Local Authority. Local Authorities have been effectively running-down their mortgage book. The rate of new home lending is currently very low. So the relatively high-level of arrears should be assessed in the context of a very old portfolio of loans and a very low rate of home lending since the late 1980s. The current 7,147 home loans in arrears would mean that nearly every loan issued since 1991 is in arrears. 7,685 new loans were drawn-down since 2016. This is clearly unlikely. So the profile of loans in arrears must include some loans that are in arrears must be very old – 25 years and more. When you include Local Authority loans in any arrears, the details are: Year and Quarter Number of Loans Number Of All Loans In Arrears Number Of All Loans In Arrears As % Of The Total Loan Book Q4 2015 17,992 8,115 45.10% Q1 2016 17,669 8,448 47.81% Q2 2016 17,411 7,909 45.43% Q3 2016 17,182 7,675 44.67% Q4 2016 16,884 7,708 45.65% Q1 2017 16,591 7,624 45.95% Q2 2017 16,352 7,427 45.42% Q3 2017 16,107 7,328 45.50% Q4 2017 15,893 7,147 44.97% The age profile of loans in arrears combined with the relatively small value of these loans would lead to the reasonable inference that economic circumstances are not a major contributor to arrears. This indicates a possible additional driver of loan arrears such as poor Local Authority lending and collection practices combined with a belief that if a Local Authority property is repossessed the occupier will have to be rehomed by the Local Authority.
- 27. Mortgage Arrears, Default and Repossessions Page 27 The following table aggregates Local Authority lending home lending by year and combines it with overall lending and average national home price across all property types. This gross level of granularity provides little detailed insight other than general trends. The figures are for loans approved and drawn- down. Year Local Authority Number of Home Loans Approved Local Authority Total Value of Home Loans Approved €M Local Authority Number of Home Loans Paid Local Authority Total Value of Home Loans Paid €M Average Local Authority Loan Approved Average National Home Price Local Authority Loan Approval % Of Average National Home Price National Total Loans Approved National Total Value Approved €M Local Authority % Of Total Number Of All Home Loans Local Authority % Of Total Value Of All Home Loans 1976 5,113 26.5 6,732 32.1 €5,190 €15,564 33.35% 25,240 242.6 20.26% 10.94% 1977 5,884 39.1 5,021 23.9 €6,646 €18,754 35.44% 24,540 277.2 23.98% 14.11% 1978 8,370 72.0 5,697 41.1 €8,601 €24,082 35.72% 26,777 358.3 31.26% 20.09% 1979 8,950 91.5 6,943 63.4 €10,229 €29,387 34.81% 30,051 29.78% 19.30% 1980 10,381 132.2 7,998 88.8 €12,733 €34,967 36.41% 28,728 522.4 36.14% 25.30% 1981 8,971 135.4 7,826 121.3 €15,088 €40,167 37.56% 29,485 632.5 30.43% 21.40% 1982 7,595 123.2 8,126 125.8 €16,217 €44,060 36.81% 26,824 646.4 28.31% 19.05% 1983 5,333 87.4 6,150 98.8 €16,381 €44,448 36.85% 30,257 803.9 17.63% 10.87% 1984 4,959 87.1 5,365 88.4 €17,565 €45,419 38.67% 28,852 787.4 17.19% 11.06% 1985 5,046 95.4 5,135 91.3 €18,898 €46,542 40.60% 31,203 879.7 16.17% 10.84% 1986 6,646 146.7 5,444 108.1 €22,067 €48,256 45.73% 30,091 876.6 22.09% 16.73% 1987 7,186 128.0 8,299 186.0 €17,811 €48,151 36.99% 31,874 986.3 22.55% 12.98% 1988 2,062 45.1 4,444 78.1 €21,860 €52,450 41.68% 42,543 1430.0 4.85% 3.15% 1989 1,489 33.5 2,343 39.4 €22,512 €58,178 38.70% 45,090 1777.8 3.30% 1.89% 1990 1,085 26.5 1,369 26.0 €24,459 €65,541 37.32% 34,812 1491.9 3.12% 1.78% 1991 1,254 31.2 1,278 27.4 €24,909 €66,914 37.23% 37,058 1611.7 3.38% 1.94% 1992 1,167 29.2 1,269 27.4 €25,025 €69,264 36.13% 44,433 2013.6 2.63% 1.45% 1993 759 18.8 871 18.7 €24,759 €69,883 35.43% 45,390 2161.0 1.67% 0.87% 1994 489 12.2 634 13.2 €24,927 €72,732 34.27% 50,204 2445.0 0.97% 0.50% 1995 371 10.3 403 9.5 €27,763 €77,994 35.60% 49,288 2666.2 0.75% 0.39% 1996 313 9.9 376 9.5 €31,629 €87,202 36.27% 61,006 3677.0 0.51% 0.27% 1997 225 7.2 259 7.2 €32,000 €102,222 31.30% 64,652 4424.1 0.35% 0.16% 1998 173 6.2 211 6.1 €35,838 €125,302 28.60% 68,925 5654.9 0.25% 0.11% 1999 112 5.0 141 5.0 €44,643 €148,521 30.06% 78,572 7692.7 0.14% 0.06% 2000 156 9.5 113 4.7 €60,897 €169,191 35.99% 80,856 9003.7 0.19% 0.11% 2001 192 14.9 155 10.7 €77,604 €182,863 42.44% 69,062 8732.6 0.28% 0.17% 2002 218 20.0 224 17.6 €91,743 €198,087 46.31% 93,136 14359.3 0.23% 0.14% 2003 162 13.9 215 16.4 €85,802 €224,567 38.21% 97,888 17446.1 0.17% 0.08% 2004 171 16.0 215 16.2 €93,567 €249,191 37.55% 104,305 21019.2 0.16% 0.08% 2005 147 16.3 193 14.3 €110,884 €276,221 40.14% 120,037 27753.3 0.12% 0.06% 2006 166 20.3 242 20.8 €122,289 €305,637 40.01% 114,593 31382.2 0.14% 0.06% 2007 57 6.8 92 7.8 €119,298 €322,634 36.98% 88,747 24064.1 0.06% 0.03% 2008 50 5.5 75 7.3 €110,000 €305,269 36.03% 55,879 15140.3 0.09% 0.04% 2009 65 8.3 75 7.0 €127,692 €242,033 52.76% 27,924 6431.1 0.23% 0.13% 2010 92 12.1 69 7.4 €131,522 €228,268 57.62% 20,021 4,153.9 0.46% 0.29% 2011 110 12.3 106 11.0 €111,818 €230,303 48.55% 12,834 2427.7 0.86% 0.51% 2012 174 17.3 149 14.7 €99,425 €220,415 45.11% 17,769 3225.0 0.98% 0.54% 2013 212 20.0 143 11.4 €94,340 €228,216 41.34% 19,258 3709.3 1.10% 0.54% 2014 222 22.7 171 13.1 €102,252 €246,378 41.50% 31,897 6187.2 0.70% 0.37% 2015 298 32.4 247 22.8 €108,725 €281,432 38.63% 32,236 6326.0 0.92% 0.51% 2016 330 37.7 257 26.6 €114,242 €313,483 36.44% 35,037 7284.8 0.94% 0.52%
- 28. Mortgage Arrears, Default and Repossessions Page 28 The number of loans paid is smaller than the number of loans paid. Also loans paid will be at a lag relative to loans approved. The following charts use the loans approved numbers and so over represent the Local Authority loan numbers. Local Authority home lending, from a peak in 1981 where it accounted for over 36% of homes and over and 25% of the loan home amount, has fallen to where it represents less than 1%. However, the recent affordable homes initiative, being delivered through Local Authorities, may cause this to increase. In this context, the apparently unregulated lending practices of Local Authorities may come under examination. Another hidden factor that would be worth considering in terms of Local Authority home lending is not just lending practices but some form of index of lending efficiency – number of loans issued and operated for numbers of personnel relative to other lenders. The average Local Authority loan approved is considerably than the average mortgage and the average non-Local Authority property purchase price. The latter may be due to Local Authorities selling their housing stock at sub-economic prices. The following chart shows the numbers of home loans approved nationally and by Local Authorities, It shows that Local Authorities effectively stopped large scale lending after 1987
- 29. Mortgage Arrears, Default and Repossessions Page 29 The following chart shows the proportion that Local Authority loans represented as a percentage of the number and value of home lending. Again, it shows the rapid fall after 1987. The following shows a profile of RFSP PDH and Local Authority mortgages in arrears for over 90 days with the RFSP numbers shown on the left vertical axis and the Local Authority numbers shown on the right vertical axis.
- 30. Mortgage Arrears, Default and Repossessions Page 30 While the proportion of loans in arrears is much greater for Local Authorities, the arrears percentage profiles are very similar. The R2 value of these two series is 0.9096 indicating a high statistical correlation. The same comments as before apply to the possibility of spurious regression Local Authority mortgage lending and collections practices are potentially important as Local Authorities may be the vehicle for increased mortgage lending for social and affordable housing in the future. The high rates of arrears and their long duration may be due to poor collections practices by Local Authorities. These numbers also do not take into account the cost of home loan issuing and loan management by Local Authorities. Local Authority mortgage lending is currently not regulated by the Central Bank. Mortgage Arrears Cohort Analysis These are just some notes on a possible approach to analysing movements between arrears cohorts. Central Bank mortgage arrears data is measured at 90 day intervals. Arrears are classified into 90 day cohorts: Not In Arrears In Arrears Up To 90 Days In Arrears 91 To 180 Days In Arrears 181 To 360 Days In Arrears 361 To 720 Days In Arrears Over 720 Days Repossessed
- 31. Mortgage Arrears, Default and Repossessions Page 31 So, for example, a new mortgage issued during Interval 1 could move to the either of the cohorts Not In Arrears or In Arrears Up To 90 Days at the end of Interval 2. Similarly, the number of mortgages in the cohort In Arrears 720+ Days could move to any of the cohorts in the next measurement interval: 1. Repossessed – the mortgaged property is repossessed and so no longer mortgaged 2. 720 + Days – their arrears continue unpaid and get older 3. In Arrears 361 - 720 Days – the older arrears are paid but some of the newer arrears remain 4. In Arrears 181 To 360 Days – the older arrears are paid but some of the newer arrears remain – 5. 91 - 180 Days – the older arrears are paid but some of the newer arrears remain 6. 1 – 90 Days – the older arrears are paid but some of the newer arrears remain 7. Not In Arrears – all arrears are paid off Schematically, the possible movement between cohorts is:
- 32. Mortgage Arrears, Default and Repossessions Page 32 Modelling this movement at this level of detail would be quite complex and would add little insight as the probabilities with which elements of one cohort move to another cohort change over time. The underlying data was volatile over the interval being measured. You could model this using simple matrix algebra. The numbers of mortgages that fall into the various categories would be represented as: Mortgage Paid Off New Mortgage Not In Arrears In Arrears Up To 90 Days In Arrears 91 To 180 Days In Arrears 181 To 360 Days In Arrears 361 To 720 Days In Arrears Over 720 Days Repossessed A1 B1 C1 D1 E1 F1 G1 H1 I1 The probabilities of moving between cohorts at the end of an interval could be represented as a second matrix. This shows the values for the cohort In Arrears Up To 90 Days. Mortgage Paid Off New Mortgage Not In Arrears In Arrears Up To 90 Days In Arrears 91 To 180 Days In Arrears 181 To 360 Days In Arrears 361 To 720 Days In Arrears Over 720 Days Repossessed BD CD DD ED FD GD HD Where: BD = proportion of New Mortgages that go into In Arrears Up To 90 Days CD = proportion of existing mortgages Not In Arrears that go into In Arrears Up To 90 Days DD = proportion of already In Arrears Up To 90 Days remain In Arrears Up To 90 Days ED = proportion of In Arrears 91 To 180 Days move to In Arrears Up To 90 Days FD = proportion of In In Arrears 181 To 360 Days move to In Arrears Up To 90 Days GD = proportion of In Arrears 361 To 720 Days move to In Arrears Up To 90 Days HD = proportion In Arrears Over 720 Days move to In Arrears Up To 90 Days Multiplying the two matrices would give you: Mortgage Paid Off New Mortgage Not In Arrears In Arrears Up To 90 Days In Arrears 91 To 180 Days In Arrears 181 To 360 Days In Arrears 361 To 720 Days In Arrears Over 720 Days Repossessed A2 B2 C2 D2 E2 F2 G2 H2 I2
- 33. Mortgage Arrears, Default and Repossessions Page 33 In this simple example: D2 = B1 x BD + C1 x CD + D1 x DD + E1 x ED + F1 x FD + G1 x GD + H1 x HD. You could start with some assumptions to make the model easier such as the population in one cohort only moves right to the next lowest cohort, moves to the Not In Arrears cohort by paying off all arrears or by moving to the Repossessed cohort. Then: D2 = B1 x BD + C1 x CD + E1 x ED + H1 x HD. The Central Bank mortgage arrears data only contains the more details set of arrears cohorts from Sep 2012 onwards. The proportions of numbers in one cohort in any interval relative to the number in the previous interval are: Dec- 12 Mar- 13 Jun- 13 Sep- 13 Dec- 13 Mar- 14 Jun- 14 Sep- 14 Dec- 14 Mar- 15 Jun- 15 Sep- 15 Dec- 15 Mar- 16 Jun- 16 Sep- 16 Dec- 16 Mar- 17 Jun- 17 Sep- 17 Dec- 17 Not In Arrears 0.979 0.989 0.993 0.999 1.002 1.004 1.010 1.009 1.010 1.006 1.006 1.001 1.001 0.999 1.002 1.000 1.001 0.997 1.002 1.0001.001 In Arrears Up To 90 Days 0.937 0.993 0.967 0.945 0.943 0.976 0.912 0.924 0.962 0.957 0.919 0.959 0.987 0.997 0.933 0.947 1.001 1.004 0.941 0.9931.012 In Arrears 91 To 180 Days 0.935 0.983 0.967 0.947 0.916 0.891 0.915 0.865 0.840 0.910 0.907 0.907 0.938 0.978 0.968 1.031 0.966 0.996 0.965 0.9930.981 In Arrears 181 To 360 Days 0.958 0.990 0.995 0.962 0.935 0.912 0.892 0.877 0.847 0.851 0.889 0.885 0.889 0.935 0.955 0.974 0.960 0.971 0.981 0.9761.008 In Arrears 361 To 720 Days 1.029 1.036 1.004 0.990 0.958 0.941 0.948 0.914 0.883 0.908 0.871 0.860 0.893 0.911 0.925 0.919 0.944 0.946 0.970 0.9581.064 In Arrears Over 720 Days 1.124 1.119 1.113 1.103 1.055 1.051 1.050 1.011 1.008 1.004 1.003 0.981 0.974 0.985 0.977 0.988 0.968 0.985 0.976 0.9830.915 Not In Arrears 0.979 0.989 0.993 0.999 1.002 1.004 1.010 1.009 1.010 1.006 1.006 1.001 1.001 0.999 1.002 1.000 1.001 0.997 1.002 1.0001.001 The proportions are not consistent over time. These proportions tend to indicate an inconsistency in the Central Bank mortgage arrears data. There are some discrepancies with the Central Bank mortgage data. For example, from Jun-12 to Sep-12 the numbers of mortgages increased from 765,267 to 794,275 and then fell back to 778,375 in Dec-12. The corresponding new mortgage numbers from the BPFI show just 3,983 new mortgages from Jun-12 to Sep-12 and 6,043 from Sep-12 to Dec-12. IFRS 9, Non-Performing Loans and Repossessions IFRS 9 (‘Financial Instruments’) is a new accounting standard which will succeed IAS 39, and is to be implemented for annual periods beginning on or before 1 January 2018. A key part of IAS 39 and IFRS 9 relates to impairment accounting. IAS 39 used an ‘incurred loss model’ (impairment recognised when a credit loss event occurs), while IFRS 9 introduces an ‘expected loss model’ (expected credit losses (“ECL”) recognised even if an actual loss event has yet to occur). In addition, under IFRS 9 entities
- 34. Mortgage Arrears, Default and Repossessions Page 34 must take into account future expectations/ forecasted conditions combined with historic / existing conditions in its assessment of credit impairment. It would be worthwhile tracking the impact of IFRS 9 on mortgages in arrears. Retail mortgages – PDH and BTL - account for around 57% of non-performing loans (NPLs). Impairment in IFRS 9 is based on an expected loss (EL) model. It affects the way credit losses are recognised in a bank’s profit and loss statement. Impairments are currently based on incurred losses. The main impact on banks is the need to recognise EL for all financial products. Banks will have to update their estimated losses to reflect changes in the credit quality of their assets. Essentially non-performing loans must have an impairment allowance, measured as the present value of all credit losses projected for the instrument over their full lifetime, reported on the bank’s profit and loss statement. The assumption is that all loans in arrears for 90 days or more must be classified as non-performing but this can be modified for defined cases. IFRS 9 does not state that restructured loans in arrears do not have to be so classified. In response to IFRS 9, retail banks will have to reduce their bad debts. This includes mortgages in arrears. It is easier for these banks to divest themselves of those mortgages by selling them in bulk quickly than to devote resources to attempting to engage with a core of non-engagers that have been in arrears for some time. The repossession route is still long, slow and expensive. Summary Arrears in financial service provider mortgages appear to be closely correlated with the amount of negative equity. Being in arrears is a pre-requisite to repossession on foot of a legal process. Voluntary possession dos not strictly require any arrears in order to occur. In the last 10 years, there have been many legal and regulatory interventions that have affected the way in which properties whose mortgages are in arrears can be repossessed. The repossession route is still long, slow and expensive. This has been combined with a possible reluctance by retail banks whose bail-outs were financed by tax payers engaging in widespread repossessions. Two thirds of mortgages in arrears have not been subject to any form of restructuring. The rate of repossessions is extremely low. The risk of repossession is very low. The correlation between the number of arrears and the number of repossessions is very low. The Dunne Judgment may have had an effect in stopping latent repossession actions becoming actual. But this is difficult to establish. IFRS 9 will cause banks to sell non-performing loans in bulk rather than attempting the time-consuming and expensive process of trying to engage with a core of non-engagers that have been in arrears for some time. The loans will probably be sold at a considerable discount to organisations that will pursue repossession or some other financial arrangement. This will represent a loss to the bank as well as to taxpayers that have funded those banks.
- 35. Mortgage Arrears, Default and Repossessions Page 35 A very high proportion of Local Authority mortgages are in arrears. Many of these arrears are more than 20 years old. The drivers for Local Authority mortgages arrears appear to be very different that those that apply for financial service provider mortgages.
- 36. Mortgage Arrears, Default and Repossessions Page 36 For more information, please contact: http://ie.linkedin.com/in/alanmcsweeney