Maybe Some People Shouldn t Own (3) Homes

Transcription

1 Maybe Some People Shouldn t Own (3) Homes Christopher Foote Lara Loewenstein Jaromir Nosal Paul Willen The views expressed in this paper are those of the authors and do not necessarily reflect those of the Federal Reserve Bank of Boston, Cleveland or the Federal Reserve System. June 2018

2 Motivation Great Recession: large boom-bust cycle in housing market

3 Motivation Great Recession: large boom-bust cycle in housing market Mortgage balances 12, , , , , ,000.0 Home Mortgages 0.0 Apr-98 Aug-99 Dec-00 Apr-02 Aug-03 Dec-04 Apr-06 Aug-07 Dec-08 Apr-10 Aug-11 Dec-12 Apr-14 Aug-15 Dec-16

4 Motivation Great Recession: large boom-bust cycle in housing market Mortgage balances Bust: delinquencies and foreclosures 12, ,000.0 Home Mortgages Balance 30-day Delinquency Rate , , , , Apr-98 Aug-99 Dec-00 Apr-02 Aug-03 Dec-04 Apr-06 Aug-07 Dec-08 Apr-10 Aug-11 Dec-12 Apr-14 Aug-15 Dec

5 Motivation Great Recession: large boom-bust cycle in housing market Mortgage balances Bust: delinquencies and foreclosures House prices HPI 30-day Delinquency Rate Apr-98 Aug-99 Dec-00 Apr-02 Aug-03 Dec-04 Apr-06 Aug-07 Dec-08 Apr-10 Aug-11 Dec-12 Apr-14 Aug-15 Dec

6 Motivation Great Recession: large boom-bust cycle in housing market Mortgage balances Bust: delinquencies and foreclosures House prices Homeownership does not follow same pattern 12, , , , ,000.0 Home Mortgages Balance 2,000.0 Homeownership Rate 0.0 Apr-98 Aug-99 Dec-00 Apr-02 Aug-03 Dec-04 Apr-06 Aug-07 Dec-08 Apr-10 Aug-11 Dec-12 Apr-14 Aug-15 Dec

7 Motivation Great Recession: large boom-bust cycle in housing market Mortgage balances Bust: delinquencies and foreclosures House prices Homeownership does not follow same pattern What lies behind these patterns? Prior research points to relatively wealthy, prime households In particular: households financing multiple properties: mortgage investors Bhutta (2015), Foote, Loewenstein, Willen (2017), Haughwout, Lee, Tracy, Van der Klauw (2011), Albanesi, DeGiorgi, Nosal (2017) Behavior of mortgage investors much more responsive to the cycle than owner occupiers

8 This Paper Quantify the role mortgage investors played in the housing crisis On disaggregated geographical level Analysis may point towards ways to avoid future boom/bust episodes

9 This Paper Quantify the role mortgage investors played in the housing crisis On disaggregated geographical level Analysis may point towards ways to avoid future boom/bust episodes 1 Did mortgage investor activity substantially increase in the housing boom? yes, more so in sand states

10 This Paper Quantify the role mortgage investors played in the housing crisis On disaggregated geographical level Analysis may point towards ways to avoid future boom/bust episodes 1 Did mortgage investor activity substantially increase in the housing boom? yes, more so in sand states 2 Were mortgage investors a disproportionate fraction of delinquencies and foreclosures? yes, more so in sand states

11 This Paper Quantify the role mortgage investors played in the housing crisis On disaggregated geographical level Analysis may point towards ways to avoid future boom/bust episodes 1 Did mortgage investor activity substantially increase in the housing boom? yes, more so in sand states 2 Were mortgage investors a disproportionate fraction of delinquencies and foreclosures? yes, more so in sand states 3 Is there indication that these defaults were strategic? yes, most of them keep one mortgage current

12 This Paper Quantify the role mortgage investors played in the housing crisis On disaggregated geographical level Analysis may point towards ways to avoid future boom/bust episodes 1 Did mortgage investor activity substantially increase in the housing boom? yes, more so in sand states 2 Were mortgage investors a disproportionate fraction of delinquencies and foreclosures? yes, more so in sand states 3 Is there indication that these defaults were strategic? yes, most of them keep one mortgage current 4 Was there misreporting of investor status? no

13 5 Was investor activity a major driver of house prices? it depends This Paper Quantify the role mortgage investors played in the housing crisis On disaggregated geographical level Analysis may point towards ways to avoid future boom/bust episodes 1 Did mortgage investor activity substantially increase in the housing boom? yes, more so in sand states 2 Were mortgage investors a disproportionate fraction of delinquencies and foreclosures? yes, more so in sand states 3 Is there indication that these defaults were strategic? yes, most of them keep one mortgage current 4 Was there misreporting of investor status? no

14 Data Federal Reserve Bank of New York Consumer Credit Panel/Equifax Data quarterly, 1999:Q1-2017:Q4 Use 1% sample about 2.5 million individuals each quarter Over 600 variables all aspects of financial liabilities: by type of account, $, # delinquent behavior: by severity, type of debt public record items: court judgements, collections, etc. credit score, age, ZIP code All variables are adjusted for joint accounts

15 Data Federal Reserve Bank of New York Consumer Credit Panel/Equifax Data quarterly, 1999:Q1-2017:Q4 Use 1% sample about 2.5 million individuals each quarter Over 600 variables all aspects of financial liabilities: by type of account, $, # delinquent behavior: by severity, type of debt public record items: court judgements, collections, etc. credit score, age, ZIP code All variables are adjusted for joint accounts Mortgage Investor: individual with more than 1 first lien on real estate property

16 Data Federal Reserve Bank of New York Consumer Credit Panel/Equifax Data quarterly, 1999:Q1-2017:Q4 Use 1% sample about 2.5 million individuals each quarter Over 600 variables all aspects of financial liabilities: by type of account, $, # delinquent behavior: by severity, type of debt public record items: court judgements, collections, etc. credit score, age, ZIP code All variables are adjusted for joint accounts Mortgage Investor: individual with more than 1 first lien on real estate property Plus HMDA

17 Source: NY Fed Consumer Credit Panel/Equifax Mortgage Investors In Boom Investors share increases from 20% to 30%+ in boom Share q1 2005q1 2010q1 2015q1 Share of value of mortgages held by people with 2+ mortgages Share of value of mortgages held by people with 3+ mortgages Share of value of mortgages held by people with 4+ mortgages

18 Source: NY Fed Consumer Credit Panel/Equifax Mortgage Investors In Boom Investors share increases from 20% to 30%+ in boom Share q1 2005q1 2010q1 2015q1 Share of Mortages Held by People with 2+ Mortgages Share of People With Mortgages who have 2+ Mortgages

19 Mortgage Investors In Boom Investors share increases from 20% to 30%+ in boom Average number of mortgages goes up as well Number Avg. Number of Mortgages Held by Investors with 2+ Mortgages Source: NY Fed Consumer Credit Panel/Equifax

20 Mortgage Investors In Boom Investors share increases from 20% to 30%+ in boom Average number of mortgages goes up as well Sand vs non-sand states exhibit large differences Sand States Non-Sand States Share Share q3 2004q1 2008q3 2013q1 2017q3 1999q3 2004q1 2008q3 2013q1 2017q3 Share of value of mortgages held by people with 2+ mortgages Share of value of mortgages held by people with 3+ mortgages Share of value of mortgages held by people with 4+ mortgages Share of value of mortgages held by people with 2+ mortgages Share of value of mortgages held by people with 3+ mortgages Share of value of mortgages held by people with 4+ mortgages Source: NY Fed Consumer Credit Panel/Equifax Source: NY Fed Consumer Credit Panel/Equifax

21 Mortgage Investors In Bust Delinquency rate among investors double that of owner occupiers Share Delinquency Rate Among Owner Occupiers Share Delinquency rate among Investors with 2+ Mortgages Source: NY Fed Consumer Credit Panel/Equifax Source: NY Fed Consumer Credit Panel/Equifax

22 Mortgage Investors In Bust Delinquency rate among investors double that of owner occupiers Large increase especially in sand states Probability Transitions into default Non-Sand States Probability Transitions into default Sand States 2000q1 2005q1 2010q1 2015q1 People, < 2 Investors, q1 2005q1 2010q1 2015q1 People, < 2 Investors, 2+ Source: NY Fed Consumer Credit Panel/Equifax Source: NY Fed Consumer Credit Panel/Equifax

23 Mortgage Investors In Bust Delinquency rate among investors double that of owner occupiers Large increase especially in sand states Sand States Non-Sand States Share Share q3 2004q1 2008q3 2013q1 2017q3 1999q3 2004q1 2008q3 2013q1 2017q3 Share of value of 90+ day delinquent mortgages held by people with 2+ mortgages Share of value of 90+ day delinquent mortgages held by people with 3+ mortgages Share of value of 90+ day delinquent mortgages held by people with 4+ mortgages Share of value of 90+ day delinquent mortgages held by people with 2+ mortgages Share of value of 90+ day delinquent mortgages held by people with 3+ mortgages Share of value of 90+ day delinquent mortgages held by people with 4+ mortgages Source: NY Fed Consumer Credit Panel/Equifax Source: NY Fed Consumer Credit Panel/Equifax

24 Mortgage Investors: Strategic Default? Most delinquent investors keep at least 1 mortgage current Share of Delinquent Investors with 2+ Mortgages With at least 1 Current Mortgage Share Source: NY Fed Consumer Credit Panel/Equifax

25 Mortgage Investors: Strategic Default? Most delinquent investors keep at least 1 mortgage current Most of those keep exactly one current Share of Delinquent Investors with 2+ Mortgages With exactly 1 Current Mortgage Share Source: NY Fed Consumer Credit Panel/Equifax

26 Mortgage Investors: Misrepresentation? Investment purposes mortgages are associated with more stringent lending requirements: higher interest rates, downpayments Incentive to misreport: is it in the data?

27 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors

28 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors No difference in the aggregate Investors Owner-Occupiers Mil.# Mil.# HMDA CCP Source: NY Fed Consumer Credit Panel/Equifax; HMDA

29 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors Large differences on disaggregated level Investors AZ Owner-Occupiers AZ Thous.# HMDA CCP Thous.# HMDA CCP Source: NY Fed Consumer Credit Panel/Equifax; HMDA

30 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors Large differences on disaggregated level Thous.# Investors CA HMDA CCP Thous.# Owner-Occupiers CA HMDA CCP Source: NY Fed Consumer Credit Panel/Equifax; HMDA

31 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors Large differences on disaggregated level Investors FL Owner-Occupiers FL Thous.# HMDA CCP Thous.# HMDA CCP Source: NY Fed Consumer Credit Panel/Equifax; HMDA

32 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors Large differences on disaggregated level Investors NV Owner-Occupiers NV Thous.# HMDA CCP Thous.# HMDA CCP Source: NY Fed Consumer Credit Panel/Equifax; HMDA

33 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors Large differences on disaggregated level Investors IL Owner-Occupiers IL Thous.# HMDA CCP Thous.# HMDA CCP Source: NY Fed Consumer Credit Panel/Equifax; HMDA

34 Mortgage Investors: Misrepresentation? Compare number of mortgages in CCP versus HMDA adjusted for joint and second liens owner-occupiers versus investors Large differences on disaggregated level location of the borrower (CCP) versus location of the property (HMDA) if a lot of out-of-town investors, HMDA>CCP if a lot of investors investing out-of-town CCP>HMDA

35 Mortgage Investors and House Price Dynamics Estimate a Markov-switching process for house price growth at CBSA level Investigate the covariates of estimated breaks across CBSAs in the spirit of Ferreira and Gyourko (2017)

36 Mortgage Investors and House Price Dynamics Estimate a Markov-switching process for house price growth at CBSA level Investigate the covariates of estimated breaks across CBSAs in the spirit of Ferreira and Gyourko (2017) Average outcomes around boom/bust breaks by quarter

37 Source: NY Fed Consumer Credit Panel/Equifax Mortgage Investors and House Price Dynamics Estimate a Markov-switching process for house price growth at CBSA level Investigate the covariates of estimated breaks across CBSAs in the spirit of Ferreira and Gyourko (2017) Average outcomes around boom/bust breaks by quarter HPA House Price Growth Predictive Margins of realqtrcat with 90% CIs Start of Boom = Quarter 1

38 Booms Hazard of becoming a 2+ investors high prior to boom Non-Sand States Sand States Probability Hazard of Becoming Investor with 2+ Mgts Start of Boom = Quarter 1 Probability Hazard of Becoming Investor with 2+ Mgts Start of Boom = Quarter 1 Source: NY Fed Consumer Credit Panel/Equifax

39 Booms Growth of investor share of mortgages for 2+ investors prior to boom Share of Mgts Held by Investors with 2+ Mgts Share Start of Boom = Quarter 1 Source: NY Fed Consumer Credit Panel/Equifax

40 Booms Increase in the number of mortgages held by investors Share Average Number of Mgts Held by Investors with 2+ Mgts Start of Boom = Quarter 1 Source: NY Fed Consumer Credit Panel/Equifax

41 Booms: HMDA Share of investor mortgages growth pre-post-boom Non-Sand States Sand States Share of Investor Mortgages in HMDA, # Share of Investor Mortgages in HMDA, # Share Start of Boom = Year 1 Share Start of Boom = Year 1 Source: HMDA

42 Busts Share of investor mortgages growth pre-bust Non-Sand States Sand States Share of Mgts Held by Investors with 2+ Mgts Share of Mgts Held by Investors with 2+ Mgts Share Start of Bust = Quarter 1 Share Start of Bust = Quarter 1 Source: NY Fed Consumer Credit Panel/Equifax

43 Busts Hazard of becoming an investor drops after busts Non-Sand States Sand States Probability Hazard of Becoming Investor with 2+ Mgts Probability Hazard of Becoming Investor with 2+ Mgts Start of Bust = Quarter Start of Bust = Quarter 1 Source: NY Fed Consumer Credit Panel/Equifax

44 Busts Share of delinquent balances held by 3+ investors Non-Sand States Sand States Share of Delinquent $ Held by Investors with 3+ Mgts Share of Delinquent $ Held by Investors with 3+ Mgts Share Start of Bust = Quarter 1 Share Start of Bust = Quarter 1 Source: NY Fed Consumer Credit Panel/Equifax

45 Busts: HMDA Share of investor mortgages growth pre-bust in Sand states Non-Sand States Sand States Share of Investor Mortgages in HMDA, $ Share of Investor Mortgages in HMDA, $ Share Start of Bust = Year 1 Share Start of Bust = Year 1 Source: NY Fed Consumer Credit Panel/Equifax

46 Conclusions Outsize role of mortgage investors in the housing boom and bust Different incentives for default delinquency behavior consistent with strategic default Geographical pattern of residence versus investment purchases some states are magnets for out-of-state investors some generate a lot of outflows of real estate investment dollars importance of investigating loan-level data ongoing work on LPS dataset: study originations and delinquencies Patterns relative to booms and busts in house prices