Indigo Peg Reservoir - Revisiting the Redemption Pool

[Serotonin]

Summary

This purpose of this discussion is to examine the interest and feasibility of using CDP interest to fund an Indigo Peg Reservoir (IPR). The IPR would be an allocation of funds that would be used to buy and/or sell iAssets on the market in a fashion that helps to maintain their peg.

Suggested parameters

• 35% of CDP interest to go toward funding an IPR

• 20% market cap limit on the treasury’s balance of any given iAsset

• IPR will continue to be funded at the proposed rate until a DAO vote is approved to change or suspend it

Benefits of an IPR

• Aids in achieving and maintaining iAsset peg

• Diversifies the treasuries POL

• Funds can be used to increase the liquidity for iAssets on DEX

• A better peg incentivizes more users to open CDPs (ex: short sellers/ADA longs) and thereby increases the DAO’s revenue stream via the interest they pay

Changes from the Redemption Pool Temp Check

A previous Temp Check Discussion proposed the idea of a “Redemption Pool” which was a pool seeded with ADA that would be used to allow users to “redeem” iAssets that were below peg for their pegged value less some fee for use of the pool. Over 6 months have passed and between now and then and Indigo has launched their V2 platform along with a number of additions and revisions to the protocol parameters and fees. Due to the addition of interest for CDPs, the DAO now has a proper revenue stream that is able to support tools such as the pool that was outlined. As it now appears more feasible, I am re-initiating and proposing this idea along with a number of modifications:

First, the name “Redemption Pool” appeared to be a bit misleading as well as led to confusion as people thought that it might directly apply to redeeming CDPs. I will now refer to it as an Indigo Peg Reservoir or “IPR” as it is a reservoir of funds used to act as a buffer to maintain iAsset peg and the term “pool” might imply that it is a liquidity pool on an AMM (that is not necessarily how it has to be implemented).

Second, while identical to charging a fee, assets could be purchased or sold at a specified price where the “spread” between the price at which they are offered to be purchased and sold on the market can be seen as the fee charged.

Third, the purchase and resale of iAssets for the purpose of assisting in peg maintenance not necessarily occur within the same pool or strategy. We can separate the functions into two broader categories of iAsset purchases and iAsset sales/redistribution. In times that iAssets are under peg, ADA or other treasury assets can be used to either directly purchase iAssets off of the market or place bids on exchanges for purchase if price drops further. Once purchased, these iAssets can be

• Held as part of a strategy to diversify POL

• Used to deepen liquidity in a liquidity pool (LP) or market making (MM) strategy

• Used in lieu of/in addition to INDY emissions to reward stability pool (SP) stakers, INDY stakers, and/or liquidity providers

Estimates

In the month and a half since v2 has launched, the DAO treasury has accumulated 85k ADA (~$38k USD) in interest as well as 1M ADA (~$450k USD) in outstanding interest. Normalized to a 30 day time span, this is roughly 56k ADA ($25k USD) in interest accrued and 667k ADA (~$300k USD) in outstanding interest per month. We may anticipate that as time goes on we can expect the proportion of accrued interest to outstanding interest to increase as CDP owners have to begin paying down their interest or adding additional collateral.

At the time of writing, iUSD is ~9% below peg, which would require ~800k ADA in immediate buy side volume across all DEX on Cardano to return to peg. This implies that with the current interest entirely going towards purchases of iUSD, these purchases might take slightly over 14 months to regain peg. As mentioned, however, we might anticipate that a more significant portion of the outstanding interest is repaid and so this “repeg timeline” would be substantially reduced. Supposing 1/3 of the outstanding interest per month turns into accrued interest per month then this would imply and accrued interest of slightly under 400k ADA/mo. At the same 100% utilization rate for iUSD purchases this reduces the “repeg timeline” to 2 months.

There are a number of broad assumptions baked into these estimates and this only includes purchases needed for the largest iAsset (iUSD). These estimates do not take into consideration additional funds needed to maintain a peg once it is achieved or maintain the peg of other iAssets, however, they also do not take into consideration markets choosing to purchase iUSD in anticipation of the DAO proactively directly pursuing a repeg with iUSD purchases. Of course, we cannot request the use of 100% of interest to pursue iAsset purchases (as this leaves no funds for other initiatives) and so it is recommended to utilize a much smaller 35% of interest funds for an IPR. This utilization rate would be more sustainable as a long term rate and allow for an actual “reservoir” of ADA to accrue in times of need for future depegs.

It is also recommended, that in the long term the DAO treasury not hold more than 20% of the supply of iAssets and they should be sold off or redistributed as incentives before that point (though this is admittedly a good “problem” to have if it were to come to that point).

Execution

An IPR would be most easily deployable on an orderbook DEX (such as Axo) where orders can be specified at a given price and partial fulfillment is allowed. A strategy can be relatively easily created that would offer bids at or just below the current market rate and an oracle can be utilized to offer asks at or near the pegged price. If market buys are the approach preferred, a DCA strategy can also be easily deployed that would purchase iAssets at a given interval. This ensures that average purchase price of iAssets is close to the average price over a given duration and also makes it difficult for bad actors to game the purchases (for example via a sandwich attack).

Alternatively, a pool could be established on an AMM (such as SundaeSwap) that allows for a dynamic adjustment of fees. The fee structure could be setup in such a way that the effective price to purchase or sell assets from/to the pool would mimic the market rate/peg price, respectively.

Finally, if IPR funds are used to only purchase iAssets (and subsequently either held or redistributed) then the process is simplified and a simple DCA order should suffice without the need to rely on an oracle for the iAsset’s peg price.

[bitgist]

I would suggest a more conservative starting allocation of 10% of CDP interest to the IPR if this concept is developed more and pursued. A much heavier focus on incentivizing Stability Pool (SP) participation should be considered beforehand. SP incentives are more likely to attract new users to engage with the protocol by adding buy pressure of iAssets on DEXs for use in SP staking.

Focusing on incentives that bring in new participation is likely to improve the protocol’s long-term health and stability without overly relying on existing users’ fees. The incentivized participation in SPs is likely to add buy pressure on DEXs helping the peg of iAssets.

The redemption mechanism has shown promise to provide aid to the peg of iUSD and should be investigated more instead of using treasury funds/user fees to provide buy pressure on DEXs to help with peg price.

It has to be addressed that augmenting rewards with iAssets accumulated by the IPR may result in some number of recipients exchanging these rewards for other assets which will simply negate some portion of the buy pressure the IPR is attempting to provide with this mechanism.

Key points:

• Allocate 10% of CDP interest to the IPR if pursued.
• Dedicate a bigger proportion of CDP interest to Stability Pool rewards and possibly other incentives.
• Explore aggressive CDP redemption mechanisms and adjustable/dynamic RMR.
• Prioritize attracting new users over redistributing fees/rewards of existing users.

[solitude]

This should get some traction in discussion.

A couple points

Such pegging mechanism is a pretty good use of funds. However it needs to caution against giving a false sense of health in setting of other parameters such as interest rate or redemption fee.

This means it’s probably good for secondary example 0.95 peg and should be doing nothing above it.
Secondly this should require oracle synchronisation with protocol. Would suggest for such mechanism to live on the protocol instead and allow the arbitrageurs to do the weight lifting instead of via any dexes.
Such pool will probably get tapped out pretty quickly in the first couple months until positions become normally distributed similar to RMR taking its own time for interest and market volatility to work.

The DAO will also have to consider from its own position and requirements before deciding whether to give those to POL.

[Serotonin]

I think that 10% would be too conservative as it would take excessively long for the purchases to have a substantial effect on the peg.

As for focusing on SP incentives, we have historical data that suggests that this doesn’t address the depeg. SP rates were 35-40% APR and still this wasn’t enough of an incentive to put a dent into the depeg. I think this goes to show that relying on incentives doesn’t always guarantee the behavior desired. Additionally, I think the funds would go much further by directly purchasing iAssets which have a direct effect on the depeg rather than relying on secondary effects. Finally, purchasing iAssets means that the DAO retains the value of the interest accrued, it is simply exchanging it into another asset. If the ADA is distributed to the SP as an incentive then this interest is “lost” by the DAO as it is given up to the SP for nothing directly in exchange.

Finally, there isn’t a reason why both couldn’t occur. For example, the IPR could purchase iUSD on the market and then redistribute that iUSD to the SP as an added incentive (or in lieu of a portion of the INDY rewards) as I outlined here

Used in lieu of/in addition to INDY emissions to reward stability pool (SP) stakers, INDY stakers, and/or liquidity providers

[Serotonin]

I agree that having an internal pool (something like the SP) would be safer than placing funds on a DEX. I was more thinking about utilizing existing infrastructure. What would be the cost associated with standing something up (development time, etc.)? Would it possibly have some of the contention issues if multiple users were to try and exchange assets at the same time?

[bitgist]

The reason I suggest starting with a conservative portion of the accrued interest is to allow for other improvements to utilize the new interest mechanism while also ensuring the long-term sustainability of the protocol. Allocating too large a portion of the treasury funds to the IPR may create an over-reliance on this mechanism without evidence that it will work as intended.

Also, the proposal to replace or augment rewards with iAssets purchased by the IPR raises several concerns:

1. The unintended consequential sell pressure which would undermine the effectiveness of the mechanism as I mentioned above.

2. Implementing such a reward distribution system may require significant effort to address issues such as determining the appropriate portion of collected interest to distribute and ensuring consistent value distribution based on either oracle or market prices of the iAssets. This complexity could divert resources from other protocol improvements.

I believe that focusing on optimizing fee and incentive structures holds more promise than relying on creating direct buy pressure to assist a re-peg. Mechanisms like redemptions and interest rates need to be dynamic and responsive to market conditions to create a more resilient protocol and ensure long-term success.

[Serotonin]

1. The unintended consequential sell pressure which would undermine the effectiveness of the mechanism as I mentioned above.

If the Indigo DAO purchases iUSD, distributes it as a reward, and then those assets are sold then the effectiveness of the purchase is slightly deteriorated, but it’s still arguably more effective at addressing the depeg than, say, using those funds to purchase INDY and then hoping that the modestly increased SP APR will be a significant enough enticement to have users buy iUSD to deposit into the pool. Either way, there isn’t anything new here. Anytime you distribute any asset as an incentive/reward, there’s the very real possibility of that asset being sold. I think there’s less of an incentive to sell iUSD that is depegged and has a reference “value”, than there is to sell a more speculative asset such as INDY.

2. Implementing such a reward distribution system may require significant effort to address issues such as determining the appropriate portion of collected interest to distribute and ensuring consistent value distribution based on either oracle or market prices of the iAssets. This complexity could divert resources from other protocol improvements.

That’s not a complicated or time consuming issue. It’s no different than any vote to distribute funds. You just say X% of the interest is going towards buying iAssets and those are then distributed as if it were the X% in ADA or, if there is a buildup of iAssets that haven’t been allocated for anything, there can be a specific proposal to distribute X amount of the iAssets.

[AdaBeacon]

My understanding is the interest rates were put in place in attempts to peg the assets.
So I think it’s suitable to use the interest charged to CDP users in order to feed back into establishing asset price before giving that money to $INDY holders.
I keep hearing a lot of people saying establishing peg is their #1 priority and overall I believe it will benefit all parties. I would advocate for this.
I’m curious how we could measure the corrective effect of the IPR vs other methods vs what it would have been if nothing was changed?

[Serotonin]

Unfortunately, you can’t really know for certain what the effect of one approach is vs the other. We can really only try to reason which is better and try it, but there’s no sure fire way to say “this is the market effect that this approach will have” until we actually go through with it and see.

On the rationale side, we can see that both the INDY market cap and iUSD market cap are currently almost equivalent.

So we can reason that every 100kADA spent purchasing INDY would have the same effect on INDY price as 100kADA spent purchasing iUSD would have on its price. Now, just for the sake of round numbers, let’s say we have enough ADA to shift the price of either asset by 10%. Now, if we shift the INDY price up 10% then how much of an effect do you think a 10% increase in rewards for the SP and LP would incentivize people to buy iUSD? I’m suggesting that it wouldn’t be anywhere near enough to also cause a 10% increase in iUSD price.
But if instead we use that ADA to shift the price of iUSD up 10% instead, then the peg is essentially restored! We don’t need to rely on some second hand incentive to try and get people to buy more iUSD, when we can create the demand for it ourselves directly.