Increase the collateral ratio

Situation: We’ve had several important changes in the protocol in the last few weeks, mostly based on the votes and the new Platypus pool. However, the effect on H2O peg and MELT price has been quite tiny if any, at least up till now. There are several reasons for it. First, the current market conditions are not ideal but still MELT underperforms. Second, the Platypus pool is in pair with USDC which evidently led to only depositing H2O from the Curve pool, but also transferring USDC from the Curve pool so that in effect nothing really happened. And third, even though there is the H2O savings pool, the MELT rewards attached to it are too small to attract necessary H2O drain from the Curve pool (plus most of the deposits had been pre-minted H2O before the MELT fee increased, not transfers from Curve; no foul play, though).

So we are back to the issue at hand - the H2O supply is too high. And it is evident now that it is majorly due to over-looping. Mind, looping in Super Vaults II is one of the Defrost selling points and it certainly has its solid place within the protocol. However, the initial conditions are dragging us down now and we are stuck due to these conditions. What conditions? Majorly the super-low collateral ratio which is now set at 105% for the stablecoin LPs. Most, or rather practically all H2O has been minted with either no or practically no MELT fees so it’s over-abundance had no structural effect of MELT.

Now why is this historical burned a problem? If we were now at the peg of $1, it would be great for loopers because they would mint H2O, change for another stablecoin, stake, mint more H2O, etc. That’s how we arrived at 0.93. However, back in the day, there were no MELT fees or almost no MELT fees so the only limiting factor was the collateral ratio set at 105%. This led to misbalanced Curve pool full of H2O that people don’t want to put more money to because they’re afraid they would only get H2O, or they just let their H2O sit there as the final step of looping. Without the MELT fees, it had been just a harvesting token, price slowly falling down, thus also rewards in USD terms. It’s a vicious circle. And we will not get out of it unless we reduce the amount of H2O in circulation. How do we do this?

We have tried with the previous steps, and I believe they are good and correct ones in the Defrost complexity. However, they will not work if the starting point is H2O at 0.93. We must be more aggressive trying to get it back to peg or very close to it.

Let’s increase the collateral ratio to 110%. This means that most of the borrowers of H2O will need to repay part of their liability. How do they do this? If they have it in the savings pool, they will withdraw part of it and repay. APR goes up a bit. If they have it looped, they will need to de-loop partially. Curve ratio becomes healthier. If they sold H2O for something else and invested elsewhere, they will need to buy H2O to repay. Curve ratio becomes healthier. If they forget about it, they will be likely liquidated which gives chance to liquidators, buying H2O, driver the price up. This last thing is obviously a bit problematic but truth be told, if you invest in crypto and specifically if you put your money as collateral, you should check it frequently and re-balance. We are practically staring into the abyss, so we shouldn’t really look back whether someone checks the page or not. If we inform and repeat on Telegram, Discord, Twitter and have a banner on the page, it’s on us/them. I would go in two steps - up to 107.5% two weeks since the announcement, and up to 110% in another 2 weeks. Actually, even the one step “shock” to 110% with 2 weeks announcement should be fine.

Getting back to the peg will motivate loopers to loop again, but now we have the MELT fees that should make sure that this activity actually benefits the platform. It is this last thing into the puzzle that’s missing. And we should go for it.

With respect to potential decollateralization, I copy what I posted on Telegram:

Ok, so I did some basic calculations. Let’s go. Currently, we have 17.144m H2O minted, 18.33m USD in collateral. The ratio is thus around 1.07. The difference between minted and collateral is 1.18m USD. We have this UST bad loan of 803k USD as a loss. Let’s say we move to 1.10 collateral instead of the current 1.05. And let’s assume people will keep their average 0.02 cushion so that the average collateral would move to 1.12. To get precisely to the breaking point where the overcollateralization equals the bad loan loss, we would need to go down to 6.7m H2O, i.e. people would need to repay/unloop more than 10m of H2O, i.e. around 60% of all H2O currently minted. Do you think this is likely to happen due to increasing the ratio from 1.05 to 1.10? I hardly think so. If you have some alternative calculation, I’m surely ready for discussion.

Nobody seems to want to answer the question about what happens when we boot these vaults and instantly become undercollateralized due to TVL dropping since we have that outstanding UST debt though. In a perfect world, sure. But we have 900k in debt that needs to be taken into consideration before we make rash decisions that could absolutely rekk us.

Most of that collateral is in just a couple of vaults. The entire vault is liqudiated, or nothing. Which means we will be undercollateralized. Undercollateralization likely means platform death. The owner of these vaults (yes, single owner) hasnt moved their funds to date, why would we assume they would statt actively managing them and increase their collateral accordingly if we were to make this change? The risk is too high to rely on a single actor behaving the way we want them to.