redemption reseach

redemption.md

Recent redemption stats in Liquity over past 31 days (in October):

• Average redemption fee is 0.7632% (52% higher than minimum floor rate of 0.5%)
• Total 696 redemptions (current active total CDP is around 1100 as shown in https://liquity.defiexplore.com/)
• Total redemption value is around $46.38Million with maximum single redeemed size is around $1.68Million
• ICR redeemed ranges from [110.32%, 228.74%] with average at 188.38%
• Current TCR is around 290% and recommended "safe" ICR is above 240% as shown in https://dune.com/queries/3140919/5238495

Liquity team recap on an earlier depeg event in mid August

https://twitter.com/LiquityProtocol/status/1694058367394664486

Key takeaways:

• Min and max redeemed Trove CR% were 117% and 147%
• Redemptions does not cause immediate notional loss for the user. Although such redemption may result in a loss or gain, depending on ETH price movement afterwards and buy-back time.
• In the scenario of further collateral price drop, redemption protect user to avoid liquidation (premium loss) and cash-in collateral at a higher price
• Redemption fee is designed in a way that it takes a bit longer (half-life decay) for arbitraguer to arb-back depegged price thus restrict unnecessary redemptions for borrowers' sake
• Redemption fee also provide an additional revenue for Liquity LQTY stakers
• Convenient operations on CDP (no interest & fixed upfront opening fee) might "unexpectedly" help the easy mint/dump of debt token (LUSD) in favor of other high-yield choices (like sDAI & sFRAX compared LUSD) which could trigger downside depeg

My two cents learned from above Liquity events: -Redemption did its work given adequate liquidity to protect downside depeg as designed but it also pushes the minimum "safe" ICR all the way up to more than double the minimum ICR requirement of 110%.

• Redemption may look like a double-edged sword. In eBTC case, currently NO obvious external factor would act like the vampire(a new victim example is GHO from AAVE) to attract eBTC minter to dump eBTC for other assets on a large scale. It is important to have a stable "population of eBTC holders happy to keep them as such, i.e., not swapping eBTC to something (wBTC) else"
• However, if that does happen, extra tools like a Peg Stability Module (allow eBTC 1:1 swap for wBTC, etc, but need separate incentive design) might help to serve both: bringing peg back and keeping necessary-only redemption without pushing ICR too high
• There are two governance-adjustable parameters in eBTC smart contract (i.e., beta and minuteDecayFactor) for "emergency" redemption fine tune: The higher the beta and the smaller minuteDecayFactor, result in less redemption fee with shorter half-life, thus more aggressive redemption. Opposite settings then bring much gentle redemption. Note these two parameters in Liquity are hardcoded and not adjustable

100 successive redemptions simulation with various beta(each plot shows 4 different settings: 1/2/5/10) and half-life decay (slower decaying to faster decaying from top plot to bottom plot: 24h/12h/6h/1h) under the assumption that both [the time interval between two redemptions] and [the ratio of a single redemption size against total debt supply] would follow some "normal distribution".

Redemption fee rate increase with each executed redemption and decrease over time due to the decaying factor. It could be seen that with higher beta and smaller half-life decay, the redemption fee increases gently and dies down quickly to minimum base rate, i.e., this is quite friendly to redemption arbitraguers