[Gnso-newgtld-wg] Process Flow for Vickrey Auction Model
kurt at kjpritz.com
Mon Dec 9 23:46:08 UTC 2019
Cool flowchart; really well thought through.
I have these thoughts:
1) Doing some reading several weeks ago about Vickrey (second-price) auctions several weeks ago when this topic was discussed, I quickly realized that selecting from among various auction models is a complex task. There are papers illustrating where Vickrey auctions (as opposed to sealed-bid, top-priced auctions) were misapplied and led to results contrary to the policy goals. These missteps occurred in simple spectrum auctions where there was one asset type up for bid as opposed to our complex environment where there are multiple assets available, all valued differently and where the existence of some have an effect of the value of the others. The literature that I read, albeit casually, stated that these were complex issues to be thought through by qualified economists.
I think it would be better to state our policy goals, mention that we have discussed sealed-bid auctions as a possible way to attain those policy goals (as compared to graduating bids) and also mention that we have discussed Vickrey auctions as a possible alternative to the more usually employed top-priced auctions.
I understand the flow chart describes timing and does not differentiate between top-price and Vickrey sealed bid auctions but I added this comment as the Vickrey model was a center of the the discussion.
2) With regards to the timing of the sealed bid, I think having the submission of the sealed bid so far in advance of the asset acquisition would work against many of the policy goals. As I understand it, the TLD bidder could be awarded many months or more than a year after the bids are submitted, There will be a limit on the number of delegations monthly, objections, IRPs and similarity check disputes that delay the process.
There will be great changes in information during that time and those changes will affect the value of the asset for which bids are received. The value of a potential TLD will change based on the number of TLDs for which there are applications, the number of TLDs addressing the same market segment, the types of companies bidding on the TLDs, how the market place evolved over that year or more, how the Guidebook and rules have been changed by ICANN, and what kind of objections or IRPs there are.
The bids should be close in time to the award so that the bidders can fairly assess the value of the asset. A bid made months or years in advance will always be “wrong” because the change in information will change the value of the asset. At least, this would result in a party over paying or underpaying. More likely, bidders will drop out if they have realized they have grossly overpaid - this after years of planning and investment. It will certainly be daunting to smaller firms or new entrants.
If a sealed bid auction is preferred (Vickrey or not), it should be done at a time when all parties can adequately value the asset for which they are bidding.
3) Regarding our policy goals,
A) I am not sure sealed bid auctions prevent over-paying - as opposed to an incremental auction where one party bids $1 over the closest rival. It might be the case, but it might not be. In any case, I think value is in the mind of the beholder and we should not be trying to prevent an unintended outcome by an uninformed applicant, and we should not force our valuation methods on others. By many definitions, one over or under pays nearly every time there is some sort of auction.
B) I also do not know exactly how to prevent parties from making private agreements. Parties in contention can talk and then some can “drop out.” It is difficult to write and enforce a rule set to prevent this.
C) It seems that a policy goal should be to put the TLD in the hands of those where it would be the most effectively utilized, That is the economic theory behind auctions
I hope this is helpful.
> On Dec 6, 2019, at 1:05 PM, Rubens Kuhl <rubensk at nic.br> wrote:
>> Em 6 de dez de 2019, à(s) 17:24:000, lists at christopherwilkinson.eu <mailto:lists at christopherwilkinson.eu> escreveu:
>> Good evening:
>> At first sight, this flowchart confirms my opinion that there should be NO auctions. The proposed procedure is so complex that, at best, it will inevitably be associated with considerable delays and costs. Not to speak of the lack of transparency. Who would be responsible for the publicly accountable recording of the whole procedure and its outcome?
>> If there is no other resolution, strings in such contention should simply be withdrawn and held over for a future round.
> Such a mechanism could be called King Solomon's method, since it's equivalent to what he did to adjudicate a dispute between two mothers. We have experience using that method in previously registered .br domains, and it only made almost 10 thousand names to never be available. What we did to finally release those names into the market was exactly to move to an auction model.
> Whatever mechanism we decide on, even choosing an applicant by chance like a lotto game or doing a beauty pageant, needs to be something that is able to reach a decisive result.
> Gnso-newgtld-wg mailing list
> Gnso-newgtld-wg at icann.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Gnso-newgtld-wg