Fix stochastic returns simulation and improve wealth inequality analysis (#748)

- Fix simulation code to properly handle stochastic returns R_t = R(Z_t, ζ_t)
- Add ζ shock draws in simulate_household function
- Correct NamedTuple unpacking throughout simulation code
- Increase grid_max to 100 for better coverage of wealth distribution
- Calibrate return volatility (a_r=0.16) to match empirical inequality
- Increase simulation to 200,000 households for better tail estimates
- Replace Lorenz curve with histogram using log y-axis for clearer visualization
- Add exercise exploring relationship between return volatility and inequality
- Include solution showing nonlinear Gini-volatility relationship
- Fix code cell language consistency (ipython3) for Jupyter Book compatibility

Results: Gini coefficient of 0.787 (close to empirical ~0.8)

🤖 Generated with Claude Code

Co-authored-by: Claude <noreply@anthropic.com>

Author: jstac