In the experiments on testnet, they chose that the reserve contract has to hold between four and eight times the value of the minted Djeds in ADA. So, not being able to repay the Djed holders can only happen if the price of ADA suddenly drops by more than 75%.
One of the downsides is that a lot of ADA will be locked in that contract.
They just stopped their blockchain completely. I don’t think they will recover.
Over a longer period, the Djed construction can adapt. People can buy the reserve coin (Shen) to fill the reserve.
The Djed contract mints two kinds of coins, Djed – the stable coin pegged to USD – and Shen – the reserve coin. Both are bought by transferring ADA into the reserve. You don’t borrow anything there.
You buy Djed with the current ADA/USD rate (given to the contract by some oracle). And you can also sell them again for that rate. So, the Djed contract/reserve has liabilities to the Djed holders: Number of Djed in circulation times current ADA/USD rate.
The rest of the reserve – reserve minus liabilities – is the equity – the current value of all the Shen tokens in circulation. And this has to be three to seven times the liabilities. So, we need a lot of Shen holders. They can buy Shen for the current price – equity divided by Shen tokens in circulation.
If ADA goes up, liabilities go down and Shen go up. If ADA goes down, Shen also go down up to the point, where you cannot sell them to the contract anymore, because the Djed – stablecoin – holders are privileged in this concept.
So amount of Djed depends on Shen circulation. Djed borrowed has the fixed price from the moment you converted ada/usd, when Shen price will depend on ada price as reserve.
I see it is programmed stable coin on demand and supply in a market. As I understood Djed is what people get to convert at ada/usd rate to usd. Shen is for equity as reserver or as liquidity provider and DJed is for stable coin function in crypto market.
I need 100 Djed I buy it for ADA/USD rate 1usd per ada with ADA coins.
It locks 3-7 times more Shen coins 3x100 - 7x100
If ADA/USD drops to 80 cents, DJed still maintain 1usd and only Shen down to 80cents, which means need more Shen to maintain liabilities?
Perhaps, the blogpost is more accessible than the original paper:
In the beginning the reserve is empty. You cannot buy Djed, then, because there would not be enough ADA in the reserve to be in the ×4 – ×8 range.
So, first, someone has to buy Shen, first, to fill up the reserve. We don’t have a price for Shen now, so the protocol has a minimal price as parameter. Let’s say it’s 0.2 ADA/Shen.
User A buys 1500 Shen for 300 ADA. Now, we have 300 ADA in the reserve.
Now, if the current exchange rate is 1 ADA/USD, user B can buy 50 Djed for 50 ADA. Then, the reserve has 350 ADA, 50 ADA are liabilities for the 50 Djed of user B and 300 ADA are the equity for the 1500 Shen of user A.
If ADA falls and we now have 2 ADA/USD, liabilities would be 100 ADA for the 50 Djed of user B and equity would be 250 ADA for the 1500 Shen of user A. So, the Shen would have fallen to 0.16 ADA/Shen, but, since 350 ADA reserve divided by 100 ADA liabilities is only 3.5, user A could not sell their Shen, anyway.
If ADA rises and we have 0.5 ADA/USD, liabilities would only be 25 ADA for the 50 Djed of user B and equity would be 325 ADA for the 1500 Shen of user A. So, the Shen would have risen to 0.216 ADA/Shen and user A could sell a lot of it, since 350 ADA reserve divided by 25 ADA liabilities is 14, so there is more than enough reserve.
DJED could get wrecked similar to what happened to Luna/Terra death spiral if enough users sell both DJED and SHEN simultaneously which would as a consequence tank ADA similar to how BTC reserves sold in the recent fiasco have tanked the entire market. Seeing as there are no users currently this is highly speculative. However any algorithmic stable coin is always susceptible to an abuse of the algorithm itself.
Why would that tank ADA? It would, at first, only unlock a lot of ADA to those users selling both coins.
If someone would try to stabilise the Djed, that would rather require to buy Shen and to do that would require to buy ADA on other markets. As you said, highly speculative, but sounds like it could provide a factor against tanking.
I mean theoretically if algorithmic stable coins worked, which there has been very little proof of, then yes this is how it should play out. In theory selling one should induce buying of the other. As we just witnessed a capitulation can not only send both into a death spiral but also tank the backing asset. In this case the backing asset of DJED is ADA which gives a minor whale manipulation an option to put Cardano back to nearly 1 cent without much resistance.
Repetition without reason is ridiculous. Thank me in October?
The ADA/USD exchange rate goes into the Djed system through an oracle. It is not determined by the trades with the contract. The contract just says that ADA/DJED is (as long as the reserve has enough ADA) always this ADA/USD coming from outside.
Nothing in the Djed system can directly influence the ADA/USD exchange rate.
I’m no expert but you guys aren’t convincing me after the terra fiasco. The concept of algorithmic stable coins seems very precarious. I don’t like the idea of any stable coin that isn’t directly backed by something like USD or gold at 1:1.
For me too, but i am happy djed was not the first algo stable coin, other wise might cardano we would see dead today. I hope on tesnet they will simulate these billions shorts and will look how it will worm
Last crash proofed than best stable coins are usdc and busd - practice is practice and always truth and facts coming from practice. Even most people loved usdt is not as stable as it looked like.
you cannot simulate the same that killed Terra, because Djed works differently. It does not allow to buy the stablecoin or to sell the reserve coin if the value of the reserve and equity is out of given limits.
in simplified terms Djed, based on the coefficients used in each particular implementation, guarantees stability of the stablecoin upto certain price fluctuation of the base coin. No more, no less.
I’m not here to convince anyone. I’m not affiliated with COTI or IOG and just trying to understand, how it is supposed to work. And it is supposed to work totally differently from Terra/Luna, so lumping all together under “the concept of algorithmic stable coin” seems too simple to me.
Has other problems. How do we know, how can we verify that the backing is really there? With Tether it is highly questionable. Are they, should they be allowed to invest the backing somewhere or does it have to lie around unused?
Honestly, don’t know. Simulations on testnet are also quite limited, because nobody has any real risks.