The random walk hypothesis of consumption and time aggregation

The random walk hypothesis of consumption is tested after accounting for time aggregation bias. Lags on income and lags on a measure of wealth do not enter the regression significantly. Also, additional lags on consumption are not significant. No ARCH effects are present in the consumption residuals, and normality of the consumption disturbances cannot be rejected with a Jarque-Bera test. The life cycle-permanent income model under rational expectations is therefore not rejected by the data if one accounts for the time aggregation bias.