There's a two way arb in ETFs which keeps the price tracking accurate: - if the ETF trades higher than bitcoin on crypto exchanges: the market makers can simply buy bitcoin on exchange, sell the ETF on traditional exchanges: do this enough and the price difference narrows. - if the ETF trades lower than bitcoin on crypto exchanges: I can just buy the ETF, ask to redeem to bitcoin and pocket the difference. Sooner or later this price discrepancy goes away.
Can only redeem to cash.
https://www.etfcentral.com/news/two-important-things-need-kn...
How are you protected against losing your investment if they can not deliver?
> Price tracking
Does Bitcoin support a high enough frequency of transactions for long enough to make this possible for years into the future? Does it not become too hard to mine at some point?
The custody provider (e.g. Coinbase is the custody provider for most of these ETFs) should be able to simply show wallet addresses for all the Bitcoin they hold.
> simply show wallet addresses for all the Bitcoin they hold
DTC's database is not where the Bitcoin is. It's still on Bitcoin's decentralized blockchain, just like always. That's the ultimate authority no matter how many centralized forms you stack on top of it.
At least with this structure, there are separate, distinct entities for custody, market making, transfer agent, etc. to reduce the incentives to self-deal and print fake paper securities and commodities.
So a huge improvement on the current state of affairs, where FTX or Binance handles all of these functions, with no segregation.
If we ignore that:
Dumbest version: You give everyone a list of all the liabilities. They make sure they are in there themselves. Anyone left out isn't entitled to the assets in an insolvency, so there is an incentive to check.
This isn't very efficient, giving everyone the whole list. But make a hash tree with each node carrying a hash and the sum of liabilities below it. Show each user their log2() sized proof. Much more efficient.
This leaks information about the distribution of ownership. Replace the values with pedersen commitment (and range proofs on their sums), and only open the top and bottom ones.
There is a fine writeup on the subject here: https://web.archive.org/web/20170928054354/https://iwilcox.m...
(Archive.org link because Zak died a couple years ago stomach cancer at age 40 due to an undetected chronic H. pylori infection, and his website didn't out survive him for long)