I stumbled upon this Ethereum address. Etherscan calls it the ENS-BurnAddress. Currently, the account holds more than $9,000,000 US. My guess is that this is used to impose a cost on each bid so people don't spam the system.
What exactly is the ENS-BurnAddress? Does it make sense to simply throw away $9,000,000 (couldn't it be given to charity)? As unlikely as it is, what happens if someone coincidentally finds the private key?
This article explains why: https://motherboard.vice.com/en_us/article/new5k8/why-did-this-company-burn-dollar26m-worth-of-ether
But the ENS auctions come with an interesting twist. If you bid on a domain name and lose, all of your ether will be returned to your wallet minus 0.5 percent of your bid value. This 0.5 percent will be transferred to the ENS burn wallet, which is inaccessible to everyone, including the ENS developers. All the ether sent to this wallet is removed from the ethereum network forever.
This bidding style is known as a Vickery auction, which was designed so that bidders would bid truthfully. Basically, it prevents bidders from jacking up the price since they don't know how others are bidding on the item up for auction. This means they must bid based on the value that they actually attribute to item.
The Burn Address: 0x000000000000000000000000000000000000dEaD
@MSwezey links an interesting article, but I think this is the most important highlight:
After the 72-hours are up, everyone who bid on that domain must reveal the value of their bid within 48 hours or else they will lose the entire value of their bid.
(emphasis mine)
I haven't run the numbers yet, but I'd bet on this being the higher contributor to the dead ether than the 0.5% burn.
If you hang out in the gitter channel, (or more importantly, were there during the first name rush), you would see that losing the whole bid was surprisingly common. In roughly descending order, people:
All of these things caused people to lose the entirety of their bids. Especially painful is that people bid very high, thinking it was safe: that the "worst case" was that they could get their ether back in a year. They learned a tough lesson that the worst case was losing the full bid. (losing more than the winning bidder paid, even!)
I'm highly in favor of dramatically increasing the reveal period in the next iteration, to at least a week or two. I don't think I've convinced the maintainers of that though, most importantly: Nick Johnson*.
* Nick does amazing work. When he disagrees with me, I tend to wonder if I'm wrong. Although I think I'm right this time :P
