1/ Thoughts on causality:
Suggesting policy changes may be harder than it looks. For education, the factors with the largest effect sizes may not have a clear policy implication.
We're also no longer treating factors as signals but as causes by which...
Suggesting policy changes may be harder than it looks. For education, the factors with the largest effect sizes may not have a clear policy implication.
We're also no longer treating factors as signals but as causes by which...
2/ we can obtain the effects we want (so confounding factors matter a lot more).
The goals are also murkier than for factor investing. In education, how do we measure success and tradeoffs? How long do we wait for results?
The goals are also murkier than for factor investing. In education, how do we measure success and tradeoffs? How long do we wait for results?
3/ What implicit metaphysical assumptions do we embed in the goals we choose?
Is the imposition of legislation to be considered an implementation risk, a cost, or both? How do we measure these difficult-to-determine frictions?
Is the imposition of legislation to be considered an implementation risk, a cost, or both? How do we measure these difficult-to-determine frictions?
4/ How much should we value the near certainty of increased freedom vs. the opaque cost and uncertain results of legislation?
To what extent do behavioral and agency problems play a role in successful implementation?
To what extent do behavioral and agency problems play a role in successful implementation?
5/ These seem like thorny problems that make factor investing look like a walk in the park.
Perhaps one implication is that if we think we know enough to make policy changes, we should be even more confident in factor investing (and that there should be many more quants).
Perhaps one implication is that if we think we know enough to make policy changes, we should be even more confident in factor investing (and that there should be many more quants).
