Cyclical Retirement and Housing Wealth Effect: evidence and theory

In the literature of macroeconomics with heterogeneous agents, there is one broad topic about the interaction between idiosyncratic risk and aggregate fluctuations. I was inspired by this to study the response of households to exogenous aggregate shocks, in particular, housing price risk in this paper and how this shock affects the retirement decision over both life-cycle and business cycles. Understanding this would help the government to choose the right program to insure the old generation during econo mic downturns.

The empirical part of this project uses the U.S. Monthly Current Population Survey 1978-2009 data to analyze the cyclical behavior of retirement. The main finding is that the retirement probability for the old is positively correlated with both the cyclical parts of housing price and home ownerships. The behavior of retirement is pro-cyclical after controlling for asset prices, labor market condition and many other factors.

The theoretical part of this project is going to account for both qualitatively and quantitatively the magnitude of housing wealth effect in people's retirement decision and household portfolio choice. I am going to set up an overlapped generation, incomplete market and general equilibrium model with heterogeneous hous eholds. The households face idiosyncratic labor productivity shocks and aggregate housing shocks each period and make decisions about consumption, savings, housing holdings and retirement. The model is calibrated to match the observed characteristics of t he economy, such as the housing wealth to households net worth ratio, home ownerships rate, the average retirement age for home owner and home renter etc. I am going to use the Finite Elements Method to solve the Stochastic Dynamic Programming problem.

T he project would be conducted during my visiting period to the Department of Economics, University of Minnesota