A stablecoin can't lose a peg, because it is backed 1:1 with a fully audited deposit of the underlying currency.
"But what about -" -- then it's not a stablecoin.
"But what about -" -- then it's not a stablecoin.
So when the crypto market crashes, some get exposed as a lie, which causes the crypto market to crash...
I suspect you're wrong, unless by collateralized you mean they store other cryptocurrencies of questionable value as their collateral. In that case sure they're probably just overflowing with funds, until they actually need to use them during a crash.
It's a very different system from UST. The backing is made up from collateral from borrowers. If the price goes too far down, their position is liquidated, with borrowers losing their collateral to maintain the peg.
So the funds are all accounted for, and there's an active system to convert the funds to stabilize the peg. By design Dai will still be $1 when there's only $100 left of crypto backing the peg.
DAI may still be vulnerable to a liquidity crisis. I.e. if there are no buyers for the collateral. This is probably not an issue if DAI is relatively small, but what if it becomes the dominant stablecoin?
It’s just that actually backing your PLNC 1:1 with actual Złoty[0] you can actually retrieve, leaves very little room for the people behind the coin to get rich.
Also, the government is a highly diversified insurer that can create more the underlying asset it is insurring.
Do they have to? Is there any reason they couldn't hedge that money?
If I wasn’t fully occupied with rsync.net I would love to run some kind of utility exchange/market/bank service …
… and I’m not involved or invested in crypto in any capacity.
The Liquity stablecoin actually works that way, they don't even have an own web app.
Thus a true backed stablecoin where the operators have proper incentives.
And that's how you get sketchy companies in Bermuda holding $100 billion in Schrodinger's assets, as well as other oddities like mining companies holding billions in assets because an ETF holding billions in assets would be illegal.
In real life there's no stablecoin. Even fully backed deposits can be traded at some other amount than 1:1, for whatever reason. Totally depends on whether you actually think the exchange can be done, and there are various reasons it might not.
That's not to mention that, as what's happening with DEI, the developers could do something like halt redemption making your holdings suddenly illiquid at the pegged price.
DAI seems safe given how over-collateralized it is. But a drop of 60+% in the collateral basket used by DAI is not inconceivable. In particular what happens if USDC's custodian is found to be untrustworthy? And the value of non-pegged cryptocurrencies tend to be strongly correlated, so a mix of crypto assets is not a diverse as one would hope.
Dai has survived a 94% drop in the value of its backing collateral in 2018, and it maintained the peg well. It would, however, be a problem if the drop was _instant_ and sustained. But it should be able to handle the speed of the drop we saw in Luna.
Your point is taken regarding trading types of trust with respect to bugs or backdoors. There is certainly a level of trust required there.
Frankly, it seems to me stablecoins are a market pipedream.
https://tether.to/en/legal/#:~:text=Tether%20reserves%20the,...
But what's the sudden obsession on HN with collateralization of blockchain assets?
Algorithmic pegs are far more interesting and democratic. That scammers have discovered they can push ponzinomics out the door as well does not change this.
The dream collateralized stablecoin, audited in some transparent way, is still vulnerable to a good old fashioned heist.
The gatekeeping of all projects that don't see the bank vault as the holy grail of community economics feels unbecoming of HN to me. This is creativity and science at work.
Again, I understand that, in addition to sincere efforts to find an algorithmic solution, we get ponzi scum peddling their things. I don't understand why people buy those, and that's an element of human nature we can hopefully overcome as this process goes on.
But you're watching collateralized coins fail, and your response is, "the only valid approach is collateralization!" ?
What approaches? Do you know of a known approach that isn't flawed in some way?
Until proven otherwise, you can't create a stable coin with "pure mathematics".
DAI is imperfect, but I think it's a sincere effort. It hasn't scammed anybody and doesn't appear capable of doing so.
And more and better efforts will follow.
The most interesting stable coin is DAI. It's not particularly profitable, and it's stable and it will be stable no matter what.
I think your point is broadly sound, though. But it's not an argument against algorithmic pegs; it's an argument against stablecoins _in principle_. And on that point, I agree.
"theoretically" here means, with cooperation from the local state, who agrees not to seize the assets. And cooperation from the local organized crime syndicate, who agrees not to seize the assets. And with a consensus-driven on-chain oracle representing the auditing process, the key custody of which is entrusted to... someone.
I'm just baffled that on HN of all places we're saying we prefer violent paleo-economic solutions to mathematical verifiability.
2. Your mathematically verified coin offers no protection against the state or crime syndicate, either (see xkcd about 50 years ago).
The local state is a boogeyman here to prop up the cyclical scams of stablecoins. There is nothing mathematical verifiable in meatspace; you only have trust. Somehow despite there being 0 cases of the USG seizing legitimate crypto assets, and several failures of algorithmic stablecoins, the "boring" option is still being fearmongered as unsafe.
The mathematical verifiability of Terra was supposedly meant that it would only lose its peg once in a million years based off of thousands of simulations. Turned out, shockingly, mathematical verifiability and economics don't mix all that well and I won't be surprised if "mathematical verifiability" of any of these stable coins is NP-hard.
But it's baffling than HN prefers paleo-economic solutions to trying to solve NP-hard problems.
Yeah, but nobody believed that.
Just like now, we all know the other shoe is going to drop on Tether. We all know it.
Are you going to feign surprise when that happens too?
Of course we don't want structures in this realm that rely on the predatory societal organization of humanity's childhood.
The whole point is to move past that.
I'm not saying blockchain tech does that on its own of course - and I think we all recognize that some of the asinine and childish habits of capitalism are leaking into crypto.
Nevertheless, to be internally consistent certainly means that the role of the state (even if it has been on relatively good behavior in this limited aspect lately) needs to be taken seriously in the context of the system.
It's far better to craft solutions that are more difficult for the state to compromise. How is this even in dispute?
(Or have I misunderstood your argument?)
USDT (tether) vs UST (terra), and now DAI vs the DEI.
DAI kept its peg from ETH $1440 -> $50 last time and is battle-tested + over-collateralized. Why don't people use that?