Mortgage Decisions of Households: Consequences for Consumption and Savings

This PhD thesis addresses three financial problems regarding the mortgage choice of the household and its effect on household consumption and savings.
The first chapter, How do Interest-only Mortgages Affect Consumption and Saving over the Life Cycle?, examines how the financial flexibility offered by interest-only (IO) mortgages affects households’ consumption and saving decisions over the life cycle. This paper is resubmitted to the academic journal Management Science. Using a unique data set with detailed information on Danish households and their mortgages, we show that young and old households are more likely to use IO mortgages compared to middle-aged households. Young households use IO mortgages because they expect higher future income, old households because IO mortgages allow them to circumvent an otherwise binding liquidity constraint. Through different channels, IO mortgages thus facilitate consumption smoothing for young and old households. Our detailed data also allow us to examine how households with IO mortgages differ from households with repayment mortgages in terms of leverage, debt and asset composition, and pension contributions.
The second chapter, The end is near: Consumption and saving decisions at the end of interest-only periods, studies the consumption behavior of IO borrowers around the end of IO periods, where amortization starts or a refinance takes place. Using Danish register-based household data on IO borrowers containing detailed information on their mortgages, we find a positive average consumption response when the borrower refinances to a new IO mortgage, whereas it is negative in response to starting amortization. For households with expiring IO mortgages, we show a significant variation in the consumption response across age and level of consumption during the IO period, indicating that consumption smoothing over the mortgage life depends on these borrower characteristics. Young borrowers use the extra liquidity in the IO period to repay bank debt, whereas others mostly tend to use it on consumption. At expiration, we find that either borrowing constraints force IO borrowers to start amortization rather than rollover to a new IO mortgage, or IO borrowers start amortization voluntarily to minimize the cost of debt. Our findings have implications for regulation of IO mortgages.
The third chapter, Double Jeopardy: Households’ consumption responses to shocks in stock and mortgage markets, investigates how household consumption is affected by shocks in the stock and the mortgage markets. Households adjust consumption downwards following negative shocks to their stock holdings. Households also lower consumption following exogenous increases in mortgage debt payments. But what is the impact of simultaneous adverse shocks in both markets, such as in the 2008 financial crisis? Using detailed Danish household data we find that the reduction in consumption doubles if households are highly exposed to both the stock and the mortgage market. We also find that the negative effects persist over time. It has a severe effect on consumption as households with a high-risk profile in the asset market also tend to have high exposure in the debt market. We discuss underlying reasons behind our results and their implications for macroprudential policies.