Real Estate Investors and the Housing Boom and Bust

Real Estate Investors and the Housing Boom and Bust Ryan Chahrour Jaromir Nosal Rosen Valchev Boston College June 2017 1 / 17

Motivation Important role of mortgage investors in the housing boom and bust (households with multiple first lien mortgages) homeownership rate peaks in 2004, but mortgage balances keep rising Homewnership Rate and Mortgage Balances 12,000.0 70.0 69.0 10,000.0 68.0 8,000.0 67.0 66.0 6,000.0 65.0 64.0 4,000.0 63.0 62.0 Home Mortgage Balances 2,000.0 Homeownership Rate 0.0 2001-01 2001-07 2002-01 2002-07 2003-01 2003-07 2004-01 2004-07 2005-01 2005-07 2006-01 2006-07 2007-01 2007-07 2008-01 2008-07 2009-01 2009-07 2010-01 2010-07 2011-01 2011-07 2012-01 2012-07 2013-01 2013-07 2014-01 2014-07 2015-01 2015-07 2016-01 2016-07 61.0 60.0 59.0 2 / 17

Motivation Important role of mortgage investors in the housing boom and bust (households with multiple first lien mortgages) homeownership rate peaks in 2004, but mortgage balances keep rising fraction of mortgage balances of investors up from 20s to 35% delinquencies follow from 5 up to 25-30% of 90+ days late balances Albanesi, DeGiorgi, Nosal (2017) effects stronger in bubble states Effects of the rise (and fall) of mortgage investors largely unexplored are major factor in the narrative of the crisis limited literature empirical so far Bhutta (2015), Foote, Loewenstein, Willen (2016), Haughwout, Lee, Tracy Van der Klauw (2011), Albanesi, DeGiorgi, Nosal (2017) 2 / 17

Motivation Important role of mortgage investors in the housing boom and bust (households with multiple first lien mortgages) homeownership rate peaks in 2004, but mortgage balances keep rising fraction of mortgage balances of investors up from 20s to 35% delinquencies follow from 5 up to 25-30% of 90+ days late balances Albanesi, DeGiorgi, Nosal (2017) effects stronger in bubble states Effects of the rise (and fall) of mortgage investors largely unexplored are major factor in the narrative of the crisis limited literature empirical so far Bhutta (2015), Foote, Loewenstein, Willen (2016), Haughwout, Lee, Tracy Van der Klauw (2011), Albanesi, DeGiorgi, Nosal (2017) Can mortgage investor behavior explain the dynamics of the housing debt, delinquency and prices leading to and in the Great Recession? 2 / 17

Motivation Why are mortgage investors different than owner-occupants? incentives to buy/sell dependent more on expected returns and financial conditions not living in house simplifies sell/default decisions: purely financial easy out: more eager to go in are investors more sensitive to financing conditions in their purchase decisions? creating instability in housing market due to higher willingness to go in and out? We analyze these questions in an equilibrium framework which captures main tradeoffs of owner-occupants versus investors revisit sensitivity of house price boom and bust, evolution of balances and default extends prior work which finds need for expectations for boom Kaplan, Mitman, Violante (2016) 3 / 17

Model We work with a transparent 3 period model Preferences are modeled as in KMV(2016): Households derive utility from consumption and housing services s i0: U i = u(c i0, s i0) + βu(c i1, s i1) + β 2 v(c i2) Utility of owned house is bigger than rented house of same size (H) { ωh if own s = H if renting, ω > 1 For tractability we assume separable log utility u(c, s) = ln(c) + ln(s) v(c) = ln(c) Credit constraints: no uncollateralized borrowing 4 / 17

Model Period 0: Households born with heterogeneous wealth levels W i choose: Consumption c i0, savings b i0 (borr. constraint b i0 0) Whether to buy a primary residence or rent - To buy a house need to put a down payment of 1 θ, so mrtg balance m i0 θp 0 - Cannot buy epsilon house, but otherwise size unrestricted: H i0 1 - Rental units are of fixed size 1 and rent costs d 0 Whether to buy any investment property I i0 So the household faces the budget constraint: W i0 + I i0 (d 0 φ) = c i0 + b i0 + H i0 (p 0 m i0 ) + d 0 1 Hi0=0 5 / 17

Model Period 1: Households come into the period with savings b i0r Primary residence H i0 with associated mortgage balance m i0r m i0 Investment properties I i0 with associated mortgage balance m I i0r m,i i0 In addition they receive stochastic income y i1 Choose Consumption c i1 and savings b i1 In terms of primary residence, choose whether to - Refinance or service the mortgage on their primary residence - Sell and buy another primary residence of different size or rent - Default on the primary residence and rent In terms of investment properties, choose whether to - Refinance or service that mortgage - Sell the investment property - Default on the investment property 6 / 17

Model Period 2: No utility from housing, only consumption Everyone gets income y 2 Houses are worth p 2 exogenous terminal value c i2 = y 2 +b 1 R+max{0, (p 2 m i1 R m i1)h i1 }+max{0, (p 2 m I i1r m,i i1 )I i1} Moving, selling and operating costs Moving cost ψ when changing primary residence Selling cost ψ I when selling investment property (ψ I < ψ) Operating cost φ when renting out investment property 7 / 17

Model Financial Intermediaries: We assume that lenders enforce the collateral constraint m it θp t, θ [0, 1] Set individual mortgage rate R m it to break-even in expectation Rental Sector: An outside rental sector provides rental units at exogenous rate ψ Investor households can also rent out their second houses, so equilibrium rent d ψ Market Clearing: Fixed supply of housing H H = H it di + I it di 8 / 17

Household optimization For this presentation we focus on analyzing time 1 optimal choices Period 2 optimization is straightforward and results in: Period 1: c i2 = y 2 +b 1 R+max{0, (p 2 m i1 R m i1)h i1 }+max{0, (p 2 m I i1r m,i i1 )I i1} If you default on one property, cannot take out another mortgage Default on primary residence and sell or default on inv property: c 1 + b 1 + d 1 + χ = y 1 + b 0 R + max{(p 1 M I 0 )I 0 χ I, 0} Default on primary residence and service inv. property mortgage: c 1 + b 1 + d 1 + χ = y 1 + b 0 R δ + rm 0,I R0,I m M0 I I 0 + I 0 (d 1 ψ) 9 / 17

Household optimization If not defaulting: Refinance vs Service vs Sell (both primary and investment) Cash flows of different options: Servicing: Primary Residence: (δ + r0 m )m 0H 0 Investment Property: (δ + r0,i)m m I 0I 0 + I 0(d 1 φ) Refinance: Primary Residence: (m 1 m 0R0 m )H 0 Investment Property: (m I 0R m,i 0 m I 1)I 0 + I 0(d 1 φ) Adjust size of either property Primary Residence: (p 1 m 0R0 m )H 0 (p 1 m 1)H 1 χ Investment Property: (p 1 m I 0R0 m )I 0 (p 1 m I 1)I 1 χ I + I 1(d 1 ψ) Sell primary residence and move to renting: d1 χ + (p1 m 0 R m 0 )H 0 10 / 17

Household optimization some intuition Why hold investment properties? They are a good financial investment (for sufficiently wealthy HHs) Housing offers high overall return Richer households have higher risk appetite than poor households They find excess returns on housing (risky asset) attractive Also, mortgages offer 1. downside protection 2. cheap access to leverage Why not simply buy a bigger primary residence? It is costly to tap into primary house equity (moving costs) Buy and sell investment properties without incurring moving cost Can make cheaper adjustments to overall portfolio investment property is better savings vehicle 11 / 17

Default Decision 0.5 0.45 0.4 0.35 Default Rate (%) Rich Primary Res. Rich Inv. Primary Resid. Investment Prop 0.3 0.25 0.2 0.15 0.1 0.05 0 0 1 2 3 4 5 6 12 / 17

Household optimization some intuition Thus, housing is an attractive investment asset In equilibrium, a positive measure of housing units held as investment property This is an important, additional component of aggregate housing demand Increases overall demand for housing and will help in generating high price-to-rent ratios Also makes housing demand and price more sensitive to shocks Changing LTV constraint leads to bigger changes in the equilibrium house price Marginal households are no longer primarily renters, but middle-to-high income HH looking to buy investment properties A given change in LTV leads to bigger change in overall housing demand Housing price more sensitive to aggregate income shocks HH are quicker to buy and sell investment properties, than primary houses 13 / 17

p1 Housing Price and LTV constraint 1.2 1.1 No Investors Yes Investors 1 0.9 0.8 0.7 0.6 0.5 0.6 0.62 0.64 0.66 0.68 0.7 0.72 0.74 0.76 0.78 0.8 thet 14 / 17

p1 Housing Price and Aggregate Income 1.1 1.05 No Investors Yes Investors 1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 0.9 0.92 0.94 0.96 0.98 1 1.02 1.04 1.06 1.08 1.1 y1 15 / 17

p1 Housing Price and Expectation of Future Price 1.3 1.2 No Investors Yes Investors 1.1 1 0.9 0.8 0.7 2.6 2.8 3 3.2 3.4 3.6 3.8 p2 16 / 17

Conclusion Housing investors (multiple property holders) are an important driver of the rise and fall in mortgage balances during the boom-bust cycle However, their effects on prices and the broader housing market has not been explored in a theoretical model We setup a transparent model where households can hold more than one house and find that It is easier to generate boom and bust episodes Housing prices are more sensitive to aggregate shocks than in the standard models Work in progress What are the distributional and welfare effects? Is there room for policy? Why or why not? 17 / 17