Calibrating for target consumer behavior?

[sbenthall]

Building on #205 ...

Running some quick local tests:

• setting the aNrmInitMean to something closer to a steady state level of wealth
• under Usual expectations
• with a perfectly Lucas-pricing market

The consumers mainly buy, and rarely (but sometimes) sell. While this is an improvement over never-ever-selling, it perhaps is not the kind of consumer/retail investor behavior that will create exciting interactions with AMMPs.

Do we have some empirical target for retail investor behavior? I believe this came up here: #199

I should say this pattern holds even if I crank the starting wealth way higher. The retail investors do not rapidly 'eat down' to a steady state asset level. Rather, they are probably mainly reinvesting dividends earned on their existing investments (since we don't do this automatically).

(Moreover, as we increase starting market assets, we also correspondingly increase the volumes at which the agents are reinvesting and buying. This will increase the retail investor market impact in AMMPS.)

Lowering the rate of return of the risky asset (see #230) might address this. But it also might not.

[sbenthall]

Superseded by #243 as we seem to be committed to the positive labor income.