This discussion post outlines a plan to increase the capacity of the Kintsugi ecosystem, which is currently bottlenecked by the limited supply of collateral, preventing growth and blocking the exit of vaults from the system. The proposed solution is to use tokens from existing yield-generating products, such as staking derivatives, as collateral for the bridge. The short-term solutions could be to increase the ceiling for LKSM and adding vKSM and qToken as collateral. Another proposed solution for the mid-term would be adding LP tokens as collateral and more generally, any yield generating, structured product long-term.
Kintsugi capacity is currently close to zero and blocks the growth of the ecosystem and does not allow for vaults to leave the system. While the high demand is favorable for the ecosystem, the limited supply of collateral poses an obstacle for further growth. There may be different reasons for the limited supply of collateral.
However, one of the main reasons for this being the bottleneck to the growth are the opportunity costs of vault operators. By depositing collateral such as KSM or USDT, they have to forgo the yield they would get when staking their KSM on the relay chain, depositing it into a lending protocol or providing liquidity into an AMM. Therefore, the bridge would have to compete with these other yield generating for the users capital.
Rather than trying to compete with the existing yield generating products, these yield generating tokens can be re-used as collateral for the bridge. While this comes with additional risks for the protocol as well as for the vault operator, it offers a sustainable solution to attract enough capital to scale the bridge efficiently.
Requirements/Dependencies: None, deeper liquidity for LKSM (optionally)
It could be proposed to increase the ceiling for LKSM to a higher amount to attract more capital because vault operators can re-stake their LKSM in vaults, making the operation of a vault much more capital efficient and attractive.
The current ceiling is at 55,000 LKSM (~$190,000) due to a lack of liquidity. However, this assumed that the full amount would be liquidated at once and that this would require to swap the total amount from LKSM into KSM into a stable coin or other token. These assumptions are extremely conservative and could be relaxed to allow for more capacity while simultaneously increase the risk of an under-collateralization of kBTC during extreme market events. How the assumptions could be relaxed:
It is unclear though, how much of the newly created LKSM capacity would be filled up with new capital as it may happen that existing vaults would replace their existing KSM vaults with LKSM vaults.
Requirements/Dependencies: DIA oracle market price feed, coingecko price feed — time: 2-3 weeks , deeper liquidity for LKSM (optionally) — time: 4-5 weeks
Apart from Acalas LKSM, Bifrost offers the same solution with their vKSM staking derivative. Under the current assumption that resulted in the LKSM ceiling of 55,000, the ceiling for vKSM would be around 2,500 vKSM (~$65,000) due to it’s lower liquidity. This is, of course, based on the same assumptions as stated above for LKSM, which could be relaxed to allow for a higher ceiling for vKSM.
Interlay team is closely working with Acala and Bifrost to improve the liquidity of both tokens to facilitate more efficient liquidations.
Requirements/Dependencies: Adding support for qTokens as vault collateral on the UI — time 2-3 weeks
Another restaking solution would be to add the supply tokens of Kintsugi’s/Interlay’s lending market as collateral. Users that supply tokens to the lending market receive qToken, which can be used as collateral in the lending market. E.g. if a user deposits USDT into the lending market, he receives qUSDT, which is interest bearing and can freely be transferred by the user unless he decides to lock the tokens as collateral to borrow from the lending market.
Similar to that, these qTokens could be registered as vault collateral. This could solve two issues at once:
While the above solution would, in theory, also work with lend tokens of other lending protocols, allowing only qTokens from Kintsugi as collateral would (a) encourage vault operators to deposit their tokens into Kintsugis lending market rather than a competitor, (b) give the DAO more control over the parameters of the underlying (interest rate, target utilization, etc.).and (c) avoid adding additional protocol risk.
Requirements/Dependencies: Adding support for LP tokens on the parachain and vault client — time: unclear
Similar to the other solutions, the bridge could also accept LP tokens as collateral. Although this would require further research and modelling of the thresholds, the value of an LP token is well defined as a function of the price and amount of the tokens in the pool. LP tokens also possess two properties that make them an interesting collateral asset:
However, the downside of the second property is, that liquidations could potentially dry up the liquidity of a pool, so that the ceiling for LP tokens need to be set accordingly and regularly monitored.
Apart from the existing tokens, this ‘restaking’ approach could be used with any other kind of yield generating tokens in the future. This could be any kind of structured products such as index funds, liquid crowd loan tokens etc., given there is sufficient liquidity to liquidate those tokens or a method to redeem them for the (liquid) underlyings.
This section is supposed to give an overview of the additional risks which the vaults and the protocol would bear compared to directly locking the ‘base’ token as collateral.
Generally speaking, the bridge would be exposed to the risks as the vault operators, although some of the risks can be mitigated to a great extent. While protocol risk is usually relatively unlikely to materialize, the effect on the price of the token that relies on the protocol in question is very severe in most cases.
The liquidity and and depeg risk can, to a certain extent, be reduced by adjusting the collateral thresholds upwards to account for the additional risk.
Kintsugi is a blockchain ecosystem that needs more collateral to grow. The limited supply of collateral is preventing the growth of the ecosystem. The current collateral options have high opportunity costs for vault operators.
To solve this problem, yield-generating tokens from other products can be used as collateral for the bridge.
In the short term, the ceiling for LKSM can be increased, vKSM and qToken can be added as collateral.
In the mid-term, LP tokens can be added as collateral.
In the long term, any yield-generating tokens can be used as collateral.
There are additional risks for the vaults and the protocol, such as protocol risk, depeg risk, and liquidity risk.
1/ LKSM vs vKSM
"Under the current assumption that resulted in the LKSM ceiling of 55,000, the ceiling for vKSM would be around 2,500 vKSM (~$65,000) due to it’s lower liquidity."
Ok so this is basically false as vKSM is the largest LSD on Kusama: vKSM as 2x more liquidity than LKSM.
https://stats.bifrost.app
vs
https://stats.karura.network/liquidstaking
This being said, yes vKSM has to be pushed as a collateral for Kintsugi vaults.
2/ Offer the possibility to deposit collateral for the non-tech guy
The main issue for vaults, as non super tech guy, is that you need to know how to install and run a vault.
The bottleneck is mainly due to this, because it's strictly restricted to tech guys currently.
-> we all have KSM or LKSM or vKSM.
I've got plenty of vKSM but i can't use them in Kintsugi vaults because i can't run a vault, i tried but it's beyond my tech knowledge.
A solution would be to have the pool system like DOT/KSM nomination pools or Khala nomination pools.
A vault that could be run for the community, where anyone could deposit in the vault.
So even if you add 10 or 20 new tokens as collateral, the guys running vaults will always be the same in the end, your bottleneck persists.
-> To my opinion, this is the main bottleneck issue, it's restricted to a base of users which is too small to run vaults.
Edited