Looking at my prior proposal with respect to your goals:
High level goals:
1) Make it more profitable to hold shares than names
Done
2) Make it profitable to participate in price discovery
No need to pay people to "participate in price discovery" with my proposal price discovery happens on its own from people getting domains that they have a real interest in using.
3) Make it profitable to bid high early
If no one else wants the domain or has expressed interest, I should be able to get it very cheap regardless of how much I personally like it, I shouldn't have to bid unnecessarily high.
the shares will be deflationary and thus someone who attempts to buy a bunch of names early will face higher and higher holding costs until they put the name up for auction again and reset the lease rate.
...Lease rate remains fixed until name is re-auctioned
Are you saying that once someone leases a domain they can continue to lease it at that rate for as long as they like?
Names remain available from the squatters who always have the names for sale.
What makes you sure that the squatters will "always have the names for sale"? I think my proposal ensures this much better and basically makes squatting a complete waste of time in the first place.
Lastly, allow someone to exit their lease early and recover the balance of their term. Otherwise squatters have no incentive to give up a name prior to the end of their lease.
If people can just "exit their lease early and recover balance" then they can bid on whatever auction they want and if they "accidentally win" they can just "exit early" with no loss. Am I missing something?
I feel like you are redefining lease if there is no commitment.
My proposal already offers a big incentive to give up the name prior to the end of lease: Giving it up quickly is the only way to profit from the sublease. If they wait till their lease is up, they get no profit.
As far as bidding rules go you want to achieve the following:
1) recognize that bidders are taking a major risk that someone started an auction at a low price and didn't have any intent of bidding higher. This particular attack causes a lot of people to lose money and costs the attacker nothing (they make money when they get outbid)
2) because there is no way to tell the difference between this "fake initial bid" and a legitimate bid the best we can do to prevent this kind of attack is to make all bids costly. If you get outbid then you only get 95% of your bid back. This will encourage individuals to bid the winning bid first, rather than risk being outbid.
I think my proposal does away with the need for this complexity