Current changes were made during the panic of the market, with not much discussion behind them.
Keep in mind our previous liquidation ratio was 66.67%. That has now become our locking ratio. Balanced started out with a 20% reward LTV and 25% maximum borrow amount LTV.
The locking ratio serves as an additional buffer for borrowers and Balanced. Borrowers are meant to keep their loans below this threshold in order to not lose access to their collaterals, and Balanced wants to prevent liquidations as they negatively affect the peg of bnUSD.
Current rebalancing mechanism does not fair well with high leverage positions.
Say there are two borrowers- Borrower A with a LTV ratio of 20%, and Borrower B with a LTV of 60%. When reverse rebalancing occurs, both borrowers have their LTV adjusted according to the size of their loan over a series of rebalancings. Both borrower positions are increased by 10%. Borrower A now has a LTV ratio of 22%, and Borrower B a ratio of 66%.
It will be very hard for Borrower B to micro manage the rebalancing meanwhile Borrower A might not even notice it, and will be happy he has bought sICX at a premium.
The previous scenario doesn’t take into account the drop in ICX price. If you combine the drop in there as well, Borrower B will be affected even more by the rebalancing, which will most likely lead into liquidation.
Decreasing the locking ratio protects both the borrowers and Balanced.
While what you are saying is technically correct, the fact that, as it stands, borrows don’t have to worry about the negative impacts that rebalancing can have simply because they can’t get locked nearly as easily is what is important.
It’s the locking portion of the rebalancing process that makes it the most dangerous. While liquidations aren’t a “good” thing, I don’t think they are having as much impact on the peg as you suggest they are. In fact, liquidations at this point would only increase the demand for bnUSD, bringing the peg down (and more inline with 1.00 usd).
I digress, though. In my opinion forcing users to adhere to the LTV locking is re-introducing the root issue unnecessarily, especially since there is ongoing discussion on how to resolve that. Until there is a concrete decision and action taken to address the rebalancing loop, the locking LTV needs to be as loose as possible.
Once users can sell ICX to pay their debt regardless of LTV, or a better rebalancing approach are in place, lowering the LTV will be no for me.
I’m still not sure what the argument for lowering the LTV would be outside of rebalancing either, especially with bnUSD demand as high as it seems to be right now.
While liquidations aren’t a “good” thing, I don’t think they are having as much impact on the peg as you suggest they are. In fact, liquidations at this point would only increase the demand for bnUSD, bringing the peg down (and more inline with 1.00 usd).
A liquidation is equal to someone selling sICX for bnUSD.
5,8m sICX being resold to borrowers is bound to have an impact on the peg.
It’s the locking portion of the rebalancing process that makes it the most dangerous.
It is supposed to be dangerous. It’s a warning for the user to not do something that is hurtful for himself and the platform.
This right here should be the main focus. Being able to sell locked collateral and paying off a part of your loan will allways give people the option to derisk their position.
Someone selling sICX for bnUSD is how people also manage their loan risk, so liquidation is really no different then “responsible borrowers” doing what they need to to avoid it. And yes, selling sICX (regardless of how it happens), will impact the peg and did, hence the constant rebalancing. The vast majority of people who were or almost liquidated (including myself) were only liquidated because they couldn’t sell their sICX after it was locked and they got pinned between the loan and the rising liquidation price.
Until there is a solution to this problem, the LTV should remain higher. A higher LTV doesn’t have any impact OTHER than it gives people the freedom to use the protocol without the risks above.
It is supposed to be dangerous. It’s a warning for the user to not do something that is hurtful for himself and the platform.
Not to come off as crass, but this is kind of absurd in my opinion. The rules of the system are defined by the smart contract. We don’t need to punish people for doing what they will anyway.
Ultimately if people get liquidated and the locking mechanism isn’t preventing them from managing their loan, then there is no excuse they can muster. But selling sICX to bnUSD isn’t “dangerous” for the platform, it’s a core feature and requirement, even if it’s done by liquidation (which protects the protocol).
Anyway, this horse is beat quite dead. I think lowering the LTV without much discussion or short notice is far more dangerous than leaving it at this point.
Fully agree with you. This proposal has its merits, however I don’t like how it was pushed, hence voted against it. I feel this needs to be discussed further. I also agree with Scott that we should wait until the rebalancing mechanism is enhanced.
Here is some data I gathered.
Currently there are around 230 borrower addresses above 45% LTV
The day we implemented the 65% LTV, around 280 addresses were above 45% LTV.
Only ≈50 people have used the possibility to lower their LTV so far.
But I still believe the high LTV should be addressed ASAP.
A person can take out a 67% LTV loan right now against sICX. If the price of ICX drops ≈24%, that person will have effectively shorted ICX and damaged the functionality of Balanced.
With a 10% rebalancing threshold, the bnusd peg can be expected to drop as low as 0.85-0.87, that’s where the 24% comes from.
The bad debt of that person will also have trouble finding volunteers who are willing to retire their bnusd for it, due to the high price of bnusd.
And historically, the odds of us not going any lower aren’t in our favour.
Edit: The person would be liquidated after a 21% drop.
The only thing critical for Balanced is not having debt be bad, and that is a function of the liquidation threshold and allowable bnUSD range.
I am not feeling the example given, if ICX falls 24%, that scenario is only true if noone trades sICX at all. In reality what happened when ICX fell was people sold ICX for bnUSD raising bnUSD price over 1USD instead of having bnUSD fall below 1USD.
Additionally, in your example, that is the 100% best time to have no lock. Let us go through the example again. If ICX falls 24%, first of all for bnUSD to have a 0.85-0.87 USD peg is extremely unlikely as the peg is 0.9 so even allowing for some gap for the rebalancers, it wouldn’t go to 0.85. Either way, lets say it does. So bnUSD is 0.85 now, due to ICX both falling, and noone selling it, or even worse people mass buying it with bnUSD as it falls.
Okay so bnUSD is 0.85USD, remember this means if ICX ~ 0.5USD, that its selling for around ~0.588USD on the DEX. Thats in plain ICX terms. The borrowers (if not locked) can withdraw collateral, sell sICX/ICX and repay debt, both stabilising the peg, and profiting this strange occurance.
In reality, during a crash, the price of sICX/ICX on the DEX buying ICX for more than CEX prices or more importantly for the price in the past is extremely unlikely.
EDIT: I also want to point out this selling ICX to repay debt would also happen automatically as the rebalance point is 0.9, so as it dips a bit below that, rebalancing would sell the collateral and reduce debt, both stabilising the peg and reducing their risk. No lock just means they can do rebalancing on their own.
Just wanted to highlight this seperately, If they are shorting ICX they are gaining when it falls. I am not seeing the profit for the people being liqdated in this scenario.
Ok, I see that I misunderstood the word peg, and used it in the wrong context.
I used the buying power of sICX against bnUSD.
Would you please be so kind and redo the calculations so that the price of bnUSD is 1.10.
Keep in mind rebalancing is not automatic, it’s real people calling the function. And they only do so when it’s profitable. A single rebalancing call costs 0.13 ICX. A swap on balanced costs 0.3% and the slippage depends on the size of the swap. Currently a 1000 sICX swap to bnUSD has a 0.04% slippage
For rebalancing to be profitable, all those previous costs have to be covered x2.
I have seen those, but what is the expected implementation time on those if they pass?
Current outlook is that the devs are now focusing on writing the smart contracts in Java, which is going to take around 2 months.
Are they willing to implement those changes in python in the meanwhile, if they will have to do it in Java later anyways?
Continuous rewards has already been pushed into Java.
Once the rebalancing contract is re-written in Java, some of these smaller changes can be implemented, such as asymmetric peg and rate limiting. The LTV sorting needs to happen on Loans, which will take longer to translate into Java. I expect rebalancing to be done this month
Seeing as the DAO is rather split on which direction to take going forward, I think more opinions need to be expressed by people who have remained quiet.
My opinion is that the current rebalancing mechanism works as intended. If we don’t want to decrease locking ratio, we should re-introduce the rewards-threshold at around 20-30%. And maybe increase the borrower rewards,