Proposal to Enact a Variable Interest Rate on CDPs

[Serotonin]

This proposal was put together before the Indigo team announced their plans today that have been in the works for an interest mechanism. It is my take on applying interest to open CDPs, so it would be good to see the team and the DAOs response as well as to understand what the similarities and differences are between my implementation and theirs. This is very much still a skeleton draft and many details need to be nailed down. As such, this post might see a number of modifications as details are clarified.

The Proposal

This proposal is to enact a variable interest rate on open CDPs which is determined by the market price of an iAsset relative to its peg. This interest rate would be 0% when iAssets trade at or above peg and would increase as the market price extends below the peg. This interest functions as an added incentive for CDP owners to scale back or close their positions when the iAssets they’ve minted trade below peg.

This is the core of the proposal, of which many adaptations could stem. Details and nuances of a specific implementation are identified after the Background.

Background

The recent turn in sentiment within crypto and the run up of ADA from $0.23 to $0.45 has resulted in the prominent use of minting or borrowing iUSD and shorting it as a method of leveraging one’s ADA position. The result has led to a substantially reduced iUSD price on decentralized exchanges (DEX) from its peg of $1 for a prolonged period of nearly 3 months to date. A recent proposal was passed to increase the minimum collateralization ratio (MCR) of iUSD from 120% to 150% in an attempt to slow down the supply of newly minted iUSD in hopes that it would result in a reversal of the iUSD price towards its peg.

The initial result appears to have been largely successful in halting the minting of new iUSD, however, it appears that users are still using lending platforms to short iUSD and leverage their ADA positions, so the iUSD price on DEX has remained volatile in a range between $0.82 and $0.94. Such drastic deviations from peg are problematic for iAssets if they are to be treated by others and other protocols as “stablecoins” as it hampers their adoption and utility as such within the ecosystem.

The DAO has also recently approved a redemption mechanism where, once implemented, users will have the ability to redeem ADA collateral for iUSD at a $0.97 equivalent value from CDP owners with the lowest collateralization ratios (CR), so long as those CDP holder’s CR is below a preset threshold known as the redemption margin ratio (RMR). This will enable a direct arbitrage path for iUSD that is not currently available within the Cardano ecosystem, however, this is only available in the event that there exist CDP owners with CR below the RMR (currently planned to start at 200%). As such, there exists the possibility that the DAO will need to vote to increase the RMR in order to keep this arbitrage path open in order to maintain an iAsset DEX price that is close to its peg. This could leave the DAO in the same precarious position it currently finds itself: one in which they must formally submit and vote on a proposal in an to attempt to control the overall iAsset supply or demand to stabilize its price.

Taking a step back to examine the economic dynamics at play, we can identify the various flows of assets that cause or result from a depeg. When iAssets are trading on DEX at their peg, there is a balance of supply and demand for those iAssets that is consistent with the intent of the protocol to enable their use as “stablecoins”. When an iAsset trades below its peg it is because there is an oversupply or lack of demand. One example of a cause of oversupply is the case of minting an iAsset to leverage one’s ADA position. If users are bullish on ADA’s price relative to an iAsset they can short the iAsset by opening a CDP and selling that iAsset on the open market, increasing ADA exposure. In the process of selling those iAssets for less than their peg, they are introducing excess supply into the ecosystem. Conversely, high reward incentives for depositing iAssets into protocols have resulted in high demand for iAssets and led to them trading well above their peg.

In the process of maintaining or restoring a peg, “the controller” (whether it be a person, DAO, algorithm, or other entity) is effectively attempting to relocate assets from the supply or demand side to its counterpart. Any reallocation of assets results in a transfer of value from those on the side with too many assets to those on the side with too few (for an iAsset peg equilibrium). This value transfer is why any changes the controller makes are most likely to be met with contention. In the instance of the redemption mechanism the value is taken from CDP owners in the form of the ADA they want to retain and using it to burn iAssets which are in excess supply.

Interest

What should the interest rate curve look like?

There is value in having contract terms that are as simple to understand as possible to this proposal suggests having a linear relationship between the relative depeg (iAsset market price / iAsset peg price) and the interest rate. The slope of this curve is initially set to X, but will be a parameter easily modifiable by the DAO.

Is the interest rate compounding?

No, while variable, unpaid interest does not compound. It would be a simple interest rate (updated daily) that is charged on the principle balance only.

What is the interest charged in?

Interest charged would be applied to the CDP owner and modify their CR.

You can either charge interest in the iAsset (increase their debt) or charge interest in ADA (reduce their collateral). This proposal suggests charging interest in ADA for multiple reasons.

• Charging interest in ADA allows for flexibility in how that value is used and what parties it is effectively dispersed to. The interest can go to a DAO fund which could be used for any number of purposes, including stabilizing the peg of other iAssets (even if they were not the ones that originally generated the revenue). If interest is charged in the iAsset, its value is inconsistently transferred according to the repayment or liquidation rules, depending on which occurs. In the case of repayment (assuming the excess iUSD debt is burned) value goes broadly to iUSD holders in the form of the buyback that occurred which raised the iUSD market rate. In the case of liquidation, value goes to Indy stakers and the SP in the form of a higher liquidation threshold (liquidation occurs at a higher ADA price).
• CDPs are already holding the collateral so the DAO is guaranteed to receive the interest even if the CDP is liquidated (per 1), therefore,
• It is possible to take the interest generated in ADA as it accumulates rather than having to wait until CDP repayment or liquidation.
• Typically, ADA is a CDP holders “safe” asset (asset they are attempting to maximize) so it is more of an incentive to repay their debt and/or close their positions.
• CDP holders that are longing ADA effectively “share” some of these rewards when ADA does well and it is less harmful to their position is ADA fares worse.
• A Mismatch Between Ecosystem Debt and Minted iAssets that could possibly lead to market manipulation type tactics to inflate iAsset prices is not guaranteed to occur (depending on the specifics of how the DAO uses the funds).

Who gets this interest?

As indicated by 1) above, interest would go to a DAO controlled wallet. This fund could then be used by the DAO for any variety of uses including stabilizing iAsset pegs (see Redemption Pool), furthering development, or creating additional incentives for activities such as liquidity providing (LP). Optionally, a portion of the interest could also be given to INDY stakers.

What asset is the interest applied to?

The purpose of the interest is to make it more expensive to maintain the excess debt (iAssets) in the ecosystem. As such, interest should be applied to the amount of iAssets minted in the CDP rather than the amount of collateral. If interest were to be applied to the collateral, this would promote less healthy collateralization ratios.

How often is the interest applied and claimed to the DAO wallet?

Interest would be continuously accrued most likely daily, via snapshots using an oracle which validates some median market rate for the iAsset. Technical details of the smart contract design remain to be fleshed out, but automated withdrawals of interest appear an impossibility in a trustless manner. The best alternative might be to have the CDP locked to all modifications unless the CDP owner agrees to and approves settlement and a transfer of the accrued interest as part of the CDP modification they are attempting to make.

(Discuss implications of such a complex smart contract)

Criticisms if applying interest

One might argue that having the DAO charge interest on iAssets that are under peg rewards the DAO for failing to do its job of developing robust stablecoins and that it incentivizes the DAO to be complacent with such depegs. While this is a fair criticism and it is healthy for individuals to be wary of the rationale of actors in the space, it’s also important to understand that the economics at hand and the overall benefit to the protocol are the underlying rationales for implementing interest.

Example Scenarios
The parameters used here are for the sake of simplicity of understanding and are NOT the specific parameters that are proposed.

Interest Rate Slope: 1% APR / % iAsset value below peg.
CDP iAsset: iUSD
CDP Owner Collateral: 200,000 ADA
CDP Owner Debt: 100,000 iUSD
ADA price: $1 (throughout entire example)

Day	iUSD market rate
1	$0.85
2	$0.9
3	$0.95
4	$1
5	$0.95
6	$0.95

Over the first 5 days, this CDP owner would accrue 95.891 iUSD worth of interest
Day 1: 15% APR * 100,000 iUSD / 365 = 41.096 iUSD
Day 2: 10% APR * 100,000 iUSD / 365 = 27.397 iUSD
Day 3: 5% APR * 100,000 iUSD / 365 = 13.699 iUSD
Day 4: 0% APR * 100,000 iUSD / 365 = 0 iUSD
Day 5: 5% APR * 100,000 iUSD / 365 = 13.699 iUSD

If the CDP owner then wanted to modify their CDP on the 6th day, as part of the adjustment contract they would have to sign to send 95.891 ADA (as the settlement rate is 1 ADA / $1 and the peg is $1 / iUSD) from their CDP collateral. Note that even though iUSD goes for $0.95 on the market on day 6, the settlement price for interest is dependent on the peg price of the iAsset, not what the oracle determines the market rate is.
Before the CDP owner decides to settle the interest on day 6, their CR is 199% as the 95.891 ADA worth of interest is deducted from their collateral in the calculation for their CR (199,904 ADA / 100,000 iUSD).

[Cooltown-SM]

I would only ever consider this if it was exclusively for new CDPs or newly added funds to a CDP after instated. Everyone on Indigo who opened a CDP, read the 2% and agreed to that. If you’re going to increase it, then either grandfather the previous rules or make it apply to new CDPs moving forward.

I’m not getting this recent attack of the socialists within the Indigo community. Every day it’s like “change the MCR to 200%!” or “Increase the fees!”. Fees are good, but they should be relatively small compared to the service offered. This proposal is just a tax increase in disguise. If you want to create a redemption pool fund and that was your main priority, just take 1% of the liquidation fees or a small percentage of the total 2% fee when you close the CDP, and fund a redemption pool. Why is it always increase the tax rate after everyone signed up? The truth seems to be you just want to raise the fees, probably because you bought a tonnnnnnn of INDY, right?

This constant proposing of changing the rules to suit someone’s emotional opinion is getting pretty gross. This is no different than Satoshi coming online tomorrow and saying “Just kidding! There are 50 billion more BTC supply!” The best part of bitcoin is that the supply is supposedly locked. The best part when I opened a CDP is that I owed 2% interest when I close it. Ok, that is what I agreed to, so I opened it. If you change that rule after the fact, it seems like you have a legal issue on your hands.

Far more important that any single argument you have made here, is that code and blockchain are considered truth, because they aren’t expected to be changed at a whim or someone’s emotional opinion. I’m not saying we don’t have slow methodical and calculated progress, but you guys lately are just flooding the DAO proposals with statements that would make the world’s central bankers proud.

We all locked in terms on a fixed rate loan of 2%. The fact that you want to propose adding an additional % which is wildly variable is beyond hilarious to me.

I’m literally 1-2 more of these kind of proposals away from making a proposal to lock all MCRs and interest mechanisms until INDY V2 launches. Watch that one pass with 90% approval.

[James]

Changing MCR every now and then doesn’t instill much confidence and such a tedious process to boot. Although, we only changed iUSD MCR twice if i remember correctly? It didnt even do much with regards to the peg. It just makes Indigo not capital efficient.

For the fee, it did say on the docs and in the WP that the 2% fee is modifiable by the DAO. Instead of a variable fee every now and then, probably an additional interest every epoch (probably about 0.03% per epoch?) could be added after a year has passed (which equates to an additional 2.19% APR for the succeeding years?) instead of an indefinite 2%? It might encourage people to close their CDP after a year which in turn generate revenue for the Protocol. People could leave high MCR CDP open for several years (if Indigo is still alive) and the protocol wouldn’t be able to generate revenue from that CDP (if the user lost their key and their CDP is very healthy, that CDP could be there forever). 2%-2.2% APR is very competitive currently.

[juanCardano]

Currently, you are liquidating at 150% and plan to increase the liquidation of ADA to 200%. This is excessively high and costly, a real madness. In addition, a 2% fee with annual compound interest is another nonsense. Were you aware that there are other protocols with an MRC of 120%, even in Cardano? And not only that, more competition is coming. There are also more DEX and CEX protocols, even outside the Cardano ecosystem. Before, your rates were reasonable, but now I feel inclined to quickly abandon this protocol.

Let’s imagine that I request a CDP and see what they would charge me. First, if I am liquidated (today), I will lose 100% of my ADA plus a 50% penalty. Fifty percent! It’s madness. When closing the CDP, the smart contract requests a 2%, which I think will soon increase to 3% (I can understand this last part), and now the loss of the PEG of -10% (this is relative and depends on each person when asking for more debt), now they also want to charge a 2% annual fee for my open CDPs with compound interest! This is another big madness. What attracted me to Indigo was that they didn’t charge me interest on interest. My idea was to keep my CDP for at least 10 years or more. It was an alternative to banks, but with all the modifications they are making, I prefer to ask for a loan from my lifelong bank.

Your protocol is becoming expensive and every day with new proposals I feel more inclined to leave. When Indigo’s competition, LENFI, which does not change rates or liquidations, arrives and more stable coins arrive, this protocol will be the last one I will use. I no longer trust it. Now it’s very expensive.

If tomorrow they lose the peg again, are they going to raise the MRC to 250% because they couldn’t solve it or the market beat them? They are not serious. I already presented a proposal to them and this that happened I had already warned them. They should check if they have vulnerabilities. Again, be careful with the proposals they present. Think that you are the ones who have the CDPs and you don’t have how to pay them. If one asks for a CDP it is to leverage and do something with that money, like buying a house, a car, starting a business, buying more ADA, etc. Those who ask for CDPs are your customers. They are already starting to lose confidence.

My proposal was healthy, but they intend to apply a 2%, 5%, 50% to everything. You, my friends, are charging a lot. Now, with compound interest, the same thing will happen to you as to MIM, another stable coin that nobody uses anymore because they lost confidence.

If this project survives all this, I congratulate you, but I will no longer participate or add more liquidity here.

[juanCardano]

I present my proposal to you again, but with a slight adjustment. I suggest that you modify the fee that I initially proposed to you, setting it at 0.5%. However, if the PEG in the main DEXs with the most liquidity, like Minswap, is below or above 5%, this fee could increase up to 3%. Each increase would be staggered by 0.5% for every 5% that deviates from the PEG in the DEXs. For example:

When requesting iUSD, the fees in ADA would be as follows:

If iUSD is worth 0.97 USD, the fee when requesting a CDP would be 0.5% (base fee). If iUSD is worth 0.94 USD, the fee would increase to 1%. If iUSD is worth 0.89 USD, the fee would rise to 1.5%. When burning iUSD, the fees would be as follows:

If iUSD is worth 1.051 USD (a problem you will also face), the fee when burning would be 1%.

If iUSD is worth 1.35, the fee would increase up to 3%, which would be the fee limit. Remember that the accumulated fees of ADA and iUSD are allocated to the stability pool. To have more stable fees, they could be adjusted in each Cardano epoch. Because if the value is below 0.8 for just one minute, such fees should not be charged. Again, the idea is to maintain the base fee at 0.5%

My propuse:

Solution and Incentives to Accumulate synthetics (CDP) and stabilize the price on DEXs

IMPORTANT!!! If you think that my proposal would help and solve the current problem, please vote and like. I propose a fixed fee of 0.15% on the amount that is requested as a loan or is returned. For example, if I request a loan in iUSD (or any other synthetic), I pay a 0.15% in ADA. On the other hand, if I return iUSD, I pay the 0.15% fee in iUSD. This fee would only affect new positions created when requesting a loan and the amount of synthetics to be requested or returned. The fees would…

[Serotonin]

I will refer you to the popup panel that came up when you chose to respond to this post.

Welcome to Indigo Protocol — thanks for contributing!

• Be kind to your fellow community members.
• Does your reply improve the conversation?
• Constructive criticism is welcome, but criticize ideas, not people.

As mentioned, this proposal is incomplete and still in the ideation phase, it isn’t even going to a full temperature check yet.

I’m literally 1-2 more of these kind of proposals away from making a proposal to lock all MCRs and interest mechanisms until INDY V2 launches.

Perhaps I forgot to make a note about it, but would be impossible to enact this proposal before v2 launches, so then I suppose that it meets your criteria.

[Serotonin]

Changing MCR every now and then doesn’t instill much confidence and such a tedious process to boot.

This proposal does not include any adjustment to the MCR.

Instead of a variable fee every now and then, probably an additional interest every epoch (probably about 0.03% per epoch?) could be added after a year has passed (which equates to an additional 2.19% APR for the succeeding years?) instead of an indefinite 2%?

It sounds like you are suggesting a flat APR interest rate in lieu of a variable interest rate while also getting rid of the 2% withdrawal fee? I would argue that interest rates need to be able to respond to market demand, so a variable rate would be better. As for the 2% withdrawal fee, this proposal does not include any adjustment to it, however the DAO could easily vote to reduce or remove that fee altogether if this proposal were to pass.

[Serotonin]

Currently, you are liquidating at 150% and plan to increase the liquidation of ADA to 200%.

The Background section includes some recent events and provides context for the rationale of the proposal. This proposal doesn’t change any liquidation threshold parameters

now they also want to charge a 2% annual fee for my open CDPs with compound interest

I never specified a rate, but this would be variable and only apply if the iAsset were depegged to the downside.

[juanCardano]

that rate would be with compound interest, or only when applying for the CDP. I am concerned about that.
Did you read my proposal?

[Serotonin]

No, to both. Interest is not compounded as there is no “unpaid interest” to compound upon. It would be simple interest but with a variable rate instead of a constant APY. Interest is charged over time, not when the CDP is opened, similar to how lending protocols charge interest (Liqwid, Lenfi, etc.).

[juanCardano]

Can you give me an example, let’s say if ADA is worth 1 USD, if I have 100,000 ADA as collateral and I ask for 50,000 iUSD (200%), what would that interest be in 1 and 10 years. As I told you, my intention is this, I already have my CDP created, how it would affect me over the years.

Give me an example of your rate, assuming again that Cardano is always stable 1 to 1 with the dollar. (for a practical example

[Serotonin]

The slope of this curve is initially set to X, but will be a parameter easily modifiable by the DAO.

I haven’t put a hard number in yet as I’m trying to do some on-chain analysis to base the decision off of first. But, just for the sake of understanding, lets say that X = 1% APR/% depeg. This would imply that if the iUSD depeg stayed at $0.85 for a full year then a 15% interest would be charged to the CDP on the iAssets minted. This 50,000 iUSD CDP would be charged 7,500 iUSD = 7,500ADA in your scenario so the users CDP, if untouched the entire year, would then be at a 185% CR with 92,500 ADA (100k-interest) backing the 50k iUSD since the interest is “deducted” from the CDP collateral.

[none11]

Interest should not impact part of minted iAssets held in SP!
High interest will also drain SP deposits, so protocol security under question, not only circulating iAseet amount will be burned.

Discounts on interest should be implemented for INDY stakers!
Added utility or benefit =higher demand= higher price = higher SP and LP rewards = more liquidity to protocol= more stable iAssets

Strategic Adjustment of iUSD MCR

This is case of FREE leverage! Until we implement time dependent CDP fees, depeg will have wide range and continuity. At this moment, levereging @indigo costs 2% (burning fee), but if CDP is not burnet, it can stay open for ever and we have free leverage tool. (That is not a norm of industry!) -You may take a look, how much it costs to borrow in other protocols. I would suggest to implement something like following… if CDP is burned in first 40 epochs from minting, collects 2% fee for stIND…

That was my rough wiev on matter

[Serotonin]

Interesting points. I don’t think there is a problem with charging interest on iAssets in the SP. It’s somewhat equivalent to saying that the interest rate from the SP is lower and I imagine would greatly complicate the smart contracts to create some special implementation. In addition to the complexity already proposed of having to find a good way to determine and charge interest (oracles are needed for this), you’d also have to determine which CDP’s have assets that are deposited in the SP. I think this is actually impossible, because all iAssets are fungible so you can’t tell if the iAssets came from a CDP owner or someone who bought them off of the market. I suppose there’s some limited implementation that could be done where you wouldn’t be charged if the same wallet that owned the CDP also had a claim for tokens in the SP. This, however, would all be affected/broken if Indigo decided to make CDPs transferable/sellable…so long story short, I think it’s unnecessary complexity.

As far as the effect it might have on SP reliability, I don’t share this concern. When prices are under peg, then that means that there are too many iAssets out in circulation, meaning that there are likely also too many in the SP. For example, more than half of all iUSD minted is in the SP currently. While I agree that a healthy buffer is needed in the SP, I think that >50% of the minted iAssets is a bit excessive.

Discounts on interest for INDY stakers is also an interesting idea, but once again I don’t think it’s necessary. INDY stakers could already receive a portion of the fees/interest so again, this is already equivalent to receiving a discount, but with less complexity.

[juanCardano]

jmmm, I read your example and understood it better. However, as a customer, I find the fee too high. It could be 2% each time it deviates 5% from the PEG, and this fee could be calculated in each epoch. Now, the question that arises is: where will the interests that are charged now or at that time go? It’s almost exactly my same proposal, but you created a compound interest debt instead of a simple fee. In my proposal, it is a simple fee. It’s not far from something I like, but it can create a problem before providing a solution. Because later, the problem will be buying iUSD when the market is bearish, all because you increased the debt of those who requested a CDP, in addition to the current liquidation rate.

[Serotonin]

As mentioned in the example, the 1% APR / 1% iAsset value below peg was not the rate proposed, it was for an easy to digest example scenario. However, I would argue that this rate curve is far too low, not too high. 15% APR to borrow when the iAsset is depegged to quite an extreme extent ($0.85 iUSD) is quite generous, especially when you compare it to margin interest rates in traditional finance. Note that most of the time, the iAsset price should be very near the peg, so the interest rate would be much smaller in that case.

Now, the question that arises is: where will the interests that are charged now or at that time go?

That would be up to the DAO, but I’m also working on another proposal where it would go to a redemption pool to help stabilize the peg. In this way, funds generated from those who are leveraging and causing the depeg (via this interest) are going towards something that helps restore the peg.

It’s almost exactly my same proposal, but you created a compound interest debt instead of a simple fee.

Once again, it is NOT compound interest. See Investopedia for the differences between simple and compound interest.

[iruncursebidoof]

I will leave for liqwid if this is proposed is passed as it is the singular advantage this inflexible buggy protocol has is it doesn’t charge interest but a fixed fee for withdrawal.

At liqwid I could get significant return for my invested capital, as well as get access to many different assets. I can short ada instead of just longing. I can short min or long it.

[Cardaflow]

Current rules of the game are simple and clear… and makes access to iAssets pretty easy

Have you modelled CDP owner rewards and losses and how they may behave in different market scenarii? It looks again a bit over engineered…

I invest when I understand rules of the game and risks, and I would feel that consequences, rewards from having a CDP would be too much complex to me under such proposal to keep such DeFi product in my portfolio… personally, if it passes, I will close mine…

[iruncursebidoof]

They dont seem to realize the whole protocol is based on cdp holders. That’s why Indy is hitting the skids right now.

[Serotonin]

I will leave for liqwid if this is proposed is passed as it is the singular advantage this inflexible buggy protocol has is it doesn’t charge interest but a fixed fee for withdrawal.

My whole argument is that this “inflexible buggy protocol” (as you put it) is that way because all it has is a fixed fee for withdrawal that doesn’t provide any incentive whatsoever to behave in a way that would maintain iAsset peg and doesn’t generate any revenue to help with such incentives from “forever open loans.” This is one way that something like the Redemption Pool that I also proposed could be funded. Maintaining peg is important for a number of reasons that I’ve already highlighted here, but least of all it makes the shorting that many are doing more efficient everywhere. So do you think that people are willing to short iUSD 10-15% below peg in a single trade, but would all abandon Indigo and go to Liqwid if they were charged 10-15% per year?
Liqwids fees to borrow iUSD were so low because there was so much iUSD minted and on the market. It’s not so much anymore since the MCR adjustment to 150% has made it more capital efficient to borrow on Liqwid if people want to short (which they are doing and now the borrow rate currently stands at 15% APR).
Finally, are you aware that development on an interest fee has already been approved by the most recent vote for v2? See 2. Interest Mechanism: A Soft Peg Approach. This is simply my temperature discussion of a specific implementation that I think could work.