| In this dissertation I explore theoretically and empirically the extent to which frictions associated with housing investment can reduce the gap between theoretical predictions and empirical findings about portfolio choices. It complements the existing literature on the impact of housing on stock investments by considering choices of housing, the effect of a costly mortgage, the mobility rate as well as housing adjustment costs.; In chapter one, I calibrate a model in which the household endogenously transits from renting to mortgage-financed homeownership. The household can later adjust its housing status either voluntarily or because of a forced move, by paying transaction cost. The study shows that homeownership crowds out stock market participation. Young and middle-aged households, regardless of whether they are currently homeowners, have much less stock holdings than predicted by traditional models without housing. Therefore, introducing frictions associated with housing into standard models can partially resolve the portfolio choice puzzle. Homeowners' portfolio shares are also very sensitive to the liquidity of housing assets and the exogenous probability to move. In response to a high transaction cost, a homeowner may save more in the riskless bond to pay for the fixed housing adjustment cost. However, a higher transaction cost also lowers the probability of housing adjustments, reducing the need for higher precautionary savings. When the probability of exogenous moving is low, a homeowner's risky asset holdings exhibit a U-shape relative to transaction cost, reflecting these two offsetting effects.; Chapter two presents empirical evidence on the cross-sectional variation of portfolio holdings from the Survey of Consumer Finances. I look at the portfolio choice of homeowners and renters as a function of wealth, households' characteristics, risk tolerance and exposure to real estate risks. I find that renters are more risk averse in asset allocations when they expect to purchase a house in the near future; homeowners hold proportionally less stock when a larger proportion of their wealth is concentrated in the form of housing. These results are consistent with the idea that for many households, financial portfolios are biased towards safer assets because of the decision to purchase a home. |
