A Life-Cycle Model with Individual Volatility Dynamics

A large literature has studied how the presence of uninsurable labor-income risk affects the patterns of savings and portfolio allocation over the life cycle. For example, workers in risky companies, occupations, or industries may have a larger incentive to accumulate wealth to insure against adverse events, such as unemployment, and to prepare for retirement. Moreover, they are likely to hold different investment portfolios, e.g., how much they invest in risky assets and how much of their investment is directed toward liquid versus illiquid accounts. In models with heterogeneous agents, income risk is usually represented by a probability distribution over income draws with a constant variance. Nonetheless, there is increasing evidence that labor-income risk is itself idiosyncratic. For example, Meghir and Pistaferri (2004) use income data from the Panel Study of Income Dynamics to show that there is strong support in favor of income dynamics with a time-varying volatility. Guvenen, Karahan, Ozkan, and Song (2015) show that an income process where variance switches stochastically between low and high regimes can match several higher-order of income moments including the high kurtosis of earnings in the U.S. data. Chang, Hong, Karabarbounis, Wang, and Zhang (2020) use administrative data from Statistics Norway to calibrate a life-cycle model with stochastic volatility in earnings and explore its implications for portfolio choice.