[geofft]

This feels like an abstracted form of arbitrage: the markets have figured out that there's a quantifiable risk difference between men and women, but they have not distinguished cis and trans folks in their model, and not figured out whether this is a hormones-at-birth difference, or a current-hormones difference, or a socialization difference, or what. They have also certainly not figured out how to distinguish actual trans folks from arbitrageurs (this is something that I think the law ought not to try too hard to figure out, on limited-government grounds, but the markets would love to take as much data as they can and feed it into a model). Once they figure it out, they'll price this person as a person, and he won't be able to swing it as dramatically.

Gender is a pretty crude (though effective) axis to make this sort of risk distinction on. An insurance company that can accurately figure out which men are lower-risk to insure than the average man will probably have a pretty good time in the market. (And get a head start on any policy changes about not discriminating on gender.)

[allenz]

> Once they figure it out, they'll price this person as a person, and he won't be able to swing it as dramatically.

Ronald Coase, of Coase theorem fame, believed that markets are rarely efficient: "Coase argued that real-world transaction costs are rarely low enough to allow for efficient bargaining and hence the theorem is almost always inapplicable to economic reality."[1] In this case, the inefficiency persists because the costs of implementing a more detailed model would far exceed the benefits.

[1] https://en.wikipedia.org/wiki/Coase_theorem