Once we understand consumption and saving, the next question is how the saving gets allocated across assets, and what the resulting demand implies about prices. This part runs the argument in both directions.
It begins with the portfolio problem taken in isolation. Under CARA utility the solution is a dollar amount held in the risky asset, independent of wealth; under CRRA utility the Merton-Samuelson result instead makes the risky share independent of wealth. The logic then reverses. Lucas takes consumption as exogenous and solves for the price that makes households willing to hold the market. With many assets, the same reasoning gives the consumption CAPM; applied to data, it gives the equity premium puzzle. No plausible degree of risk aversion explains the historical excess return on equity: matching it takes a coefficient of relative risk aversion of 15 to 20, while the riskless rate over the same period wants one of at most 2/3.
The part closes by asking what prices are consistent with the absence of arbitrage even when they diverge from fundamentals.