Following the discussions, I started with some copied statements:
In theoretical macro and finance models, it is important to have time-varying relative risk aversion (TVRRA) to generate observed moments in the macro/financial data.
A popular way to generate TVRRA is to impose habit formation preferences. Macro and finance models with habit formation preferences have successfully generated observed moments in the data.
The literature has tested habit formation preferences with the PSID data.
Brunnermeier and Nagel (2008) rejected the portfolio choice prediction due to habit formations preferences with the PSID data from 1984 to 2003.
Liu et al. (2016) also rejected the strong form portfolio choice prediction due to habit formation preferences; even though found some evidence of the weak-form prediction (associated with heterogeneous households).
Can we find a model with homogenous households such that its theoretical prediction about ρ will be in line with the evidence in the PSID data?
How robust of relationships between financial wealth and risky shares and between labor income and risky shares?
We introduce labor income to and replace the habit formation preferences with GHH preferences in the portfolio choice model in Brunnermeier and Nagel (2008). Our analytical solution shows that, with GHH preferences, households will decrease their risky shares when their financial wealth increases.
In the PSID data from 1984 to 2015, we have found robust evidence that households' risky shares are negatively correlated with (financial) wealth.
Nevertheless, our simple portfolio choice model fails to generate the robust empirical relationship between risky shares and labor income.
M. Brunnermeier, and S. Nagel, (2008). Do Wealth Fluctuations Generate Time-varying Risk Aversion? Micro-evidence from Individuals' Asset Allocation. American Economic Review, 98 (3), 713–736.