The Effect of Housing on Portfolio Choice

• Author: ADAM SZEIDL,RAJ CHETTY,LÁSZLÓ SÁNDOR
• Published date: 01 June 2017
• Date: 01 June 2017
• DOI: http://doi.org/10.1111/jofi.12500

THE JOURNAL OF FINANCE •VOL. LXXII, NO. 3 •JUNE 2017

The Effect of Housing on Portfolio Choice

RAJ CHETTY, L ´

ASZL ´

OS´

ANDOR, and ADAM SZEIDL∗

ABSTRACT

We show that characterizing the effects of housing on portfolios requires distinguish-

ing between the effects of home equity and mortgage debt. We isolate exogenous

variation in home equity and mortgages by using differences across housing markets

in house prices and housing supply elasticities as instruments. Increases in property

value (holding home equity constant) reduce stockholdings, while increases in home

equity wealth (holding property value constant) raise stockholdings. The stock share

of liquid wealth would rise by 1 percentage point—6% of the mean stock share—if a

household were to spend 10% less on its house, holding fixed wealth.

HOUSES ARE THE LARGEST ASSETS OWNED by most households, but the impact

of housing on financial markets remains unclear. Many models predict that

housing tends to reduce the demand for risky assets because it increases a

household’s exposure to risk and illiquidity (Grossman and Laroque (1990),

Brueckner (1997), Fratantoni (2001), Flavin and Yamashita(2002), Chetty and

Szeidl (2007)). But empirical studies have not found a systematic relationship

between housing and portfolios in practice (Fratantoni (1998), Heaton and

Lucas (2000), Yamashita (2003), Cocco (2005)).

This paper reconciles the theory with the data. We first show, using a simple

model, that it is critical to separate the effects of property value from the ef-

fects of home equity to characterize the effects of housing on portfolios. We then

show empirically that exogenous increases in mortgage debt induce substan-

tial reductions in the share of liquid wealth held in stocks, while exogenous

increases in home equity wealth raise stock ownership. Our empirical findings

differ from those of prior studies both because we separate the effects of mort-

gage debt and home equity wealth and because we account for the endogeneity

of housing choice in our empirical analysis.

∗Raj Chetty is with Stanford University and NBER, L´

aszl´

oS

´

andor is with Luxembourg

School of Finance, and Adam Szeidl is with Central European University and CEPR. We’ thank

Thomas Davidoff, Edward Glaeser, Albert Saiz, Stephen Shore, anonymous referees, and nu-

merous seminar participants for helpful comments. Gregory Bruich, Calin Demian, Andras Ko-

maromi, Jessica Laird, Keli Liu, James Mahon, Jeno Pal, Juan Carlos Suarez Serrato, and Philippe

Wingender provided outstanding research assistance. We are grateful for funding from National

Science Foundation Grants SES 0522073 and 0752835. Szeidl thanks the Alfred P. Sloan Foun-

dation, and the European Research Council under the European Union’s Seventh Framework

Program (FP7/2007-2013), ERC grant agreement number 283484, for support. The authors do not

have any potential conflicts of interest, as identified in the Journal of Finance disclosure policy.

DOI: 10.1111/jofi.12500

1171

1172 The Journal of Finance R

We structure our empirical analysis using a stylized two-period model of

portfolio choice that incorporates both the illiquidity and price risk effects of

housing. Our model is a stylized version of richer models of housing and port-

folio choice (Cocco (2005), Yao and Zhang (2005), Vestman (2012)). We use it

to characterize the distinct effects of exogenous changes in property value and

home equity,generating predictions about the causal i mpact of these variables.

We show that increases in property value (holding home equity wealth fixed)

generally reduce the stock share of liquid wealth by increasing illiquidity, in-

creasing exposure to risk, and reducing the present value of lifetime wealth.

In contrast, increases in home equity (holding property value fixed) raise the

stock share of liquid wealth with constant relative risk aversion (CRRA) prefer-

ences through a wealth effect. Since property value is the sum of mortgage debt

and home equity, increases in mortgage debt (holding home equity fixed) are

equivalent to increases in property value, and also reduce stockholdings. The

main lesson from the model is that distinguishing between property value

and home equity is critical when studying the causal effect of housing on

portfolios.

Based on this result, we investigate empirically the effects of property value

and home equity wealth on portfolios. As emphasized by Cocco (2005)and

Vestman (2012), both portfolios and housing are endogenous choices that are

affected by unobserved factors such as future labor income and preferences.

Thus, one cannot identify the causal effect of housing on portfolios using cross-

sectional variation across households. We address this endogeneity problem

using three research designs.

We begin with a research design that instruments for property values and

home equity using current and year of purchase home prices in the individual’s

state, calculated using repeat-sales indices. The current house price index is

naturally a strong predictor of property values. However, current house prices

also create variation in household wealth: increases in house prices increase

home equity wealth. To isolate the causal effect of owning a more expensive

house while holding wealth fixed, we exploit a second instrument, namely, the

average house price at the time of purchase. Individuals who purchase houses

when prices are high tend to have less home equity and a larger mortgage.

We control for aggregate shocks and cross-sectional differences across housing

markets by including state and year fixed effects, and thereby exploit only

differential within-state variation for identification.

We implement this cross-sectional instrumental variables (IV) strategy us-

ing microdata on housing and portfolios for 80,392 households from the Survey

of Income and Program Participation (SIPP) panels spanning 1990 to 2008.

We use two-stage least squares (2SLS) specifications to estimate the effect of

property value and home equity on the share of liquid wealth that a household

holds in stocks. We find that housing has a large effect on the share of stockhold-

ings. A $10,000 increase in property value (holding fixed home equity wealth)

causes the stock share of liquid wealth to fall by 0.6 percentage points ($275)

or by 3.9% of mean stockholdings in the sample. This estimate is stable and

statistically significant with p<0.05 across a broad range of specifications. In

The Effect of Housing on Portfolio Choice 1173

contrast, a $10,000 increase in home equity (holding fixed total property value)

increases the stock share of liquid wealth by 4.3% through a wealth effect.1

Our first research design suffers from two potential concerns that could lead

to biased estimates. The first is omitted variable bias: state-level house price

fluctuations may be correlated with other factors such as local labor market

conditions that directly impact portfolio choice. The second is selection bias due

to the endogenous timing of housing purchases: people who buy houses when

prices in their state are relatively high, for example, may have different risk

preferences from those who buy when prices are lower, potentially generating a

spurious correlation between stock shares and house price indices. We address

these two concerns using two refinements of the research design, each of which

exploits a subset of the variation used for identification in the first design.2

Our second research design, which constitutes the central identification

strategy of the paper, addresses omitted variable bias by isolating variation

in house prices driven by supply constraints. Here, we instrument for property

values and home equity using the current and year of purchase national av-

erage house price interacted with state-level housing supply elasticity, as mea-

sured by Saiz (2010), based on land availability and regulations. Intuitively,

fluctuations in the national housing market generate larger price fluctuations

in states with inelastic housing supply,leading to differential variation in house

prices across states over time. This strategy yields estimates that are similar

to those under the first design. We estimate that a $10,000 increase in prop-

erty value causes a 4% reduction in the stock share of liquid wealth, while a

$10,000 increase in home equity (holding fixed total property value) raises the

stock share by 4.8%. The elasticity of the stock share of liquid wealth with

respect to outstanding mortgage debt is −0.2, while the elasticity with respect

to home equity wealth is 0.3. These portfolio changes are driven by both the ex-

tensive and the intensive margins: changes in mortgage debt and home equity

wealth induce changes in both the probability of owning stock and the amount

of stock held conditional on stock ownership.

Our third research design addresses selection effects by using panel data to

study how portfolios for a given household change around the purchase of a

house. We test whether individuals who buy a larger house reduce their stock

share of liquid wealth more than those who buy smaller houses. We again

instrument for the change in property value using the state-level house price

index at the time of home purchase. This panel strategy complements the cross-

sectional approaches in two ways. First, it provides evidence that households

actively change the composition of their financial portfolios depending on the

amount they invest in a house. Second, it further mitigates concerns about the

1To facilitate comparison between samples with different rates of stock market participation

and hence different mean stock shares of liquid wealth, we report results in both percentages and

percentage points throughout the paper.

2Although the refined designs rely on weaker identification assumptions, we start with the

first design because it exploits all of the variation in average house prices rather than a narrow

subset of the variation, which demonstrates that the effects of housing on portfolio choice that we

document hold quite broadly.