RFC-000-007: Increase burning fee to 0.2% on November 16th (start of Epoch 6)

[DeFiFrog]

Name: Increase burning fee to 0.2% on November 16th (start of Epoch 6)
Category: Protocol
Status: Draft - Rejected
Scope: Motion to increase the burning fee to 0.2% and monitor impact. If passed, allow users 7 days to exit their renBTC position at the initial 0.1% fee. Measure data after fee change and determine best next step at that point.

Overview

This RFC proposes increasing the burning fee for all assets to 0.2% after 21 days from (Oct. 26th) submission of this RFC. Same as the terms of RIP-000-001, after at minimum one week, and at maximum four weeks, a new RIP will be proposed for future adjustments.

Potential outcomes/next steps after the implementation of raising burn fees:

• If raising the burning fee to 0.2% has had no impact on total income (or somehow it has had a positive impact), then the new RIP will propose raising the burning fees to 0.3%. If raising the burning fee to 0.2% has had a negative impact on total income, then the new RIP will propose lowering the burning fees back to 0.1%. The new RIP will be the point at which everyone can voice their opinion about the impact (whether it has been positive or negative for income).
• Leave burn fees at 0.2%.

Details

For additional context, see Details within RIP-000-001. After heated discussions with the community, this proposal consolidates key arguments 1) in support, 2) against, and 3) counter-arguments to support the proposal for such an increase in burn fee.

Reasons for:

• Setting a higher burn fee now allows us to strengthen the lever of lowering burn fees in the future. Lowering burn fee to 0% or providing a rebate becomes even more attractive in the future if we raise fees now. A discount on an already cheap product is less meaningful.
• Similar to our recent mint fee increase, a change of burn fee from 0.1% to 0.2% will likely result in non-material change in burn volume
• Increasing burn fee will double the rate of burn revenue and will 1) subsequently, increase earning potential and 2) theoretically, increase TVB by increasing $Ren token value and incentivizing more registrations
• Other protocols have changed fees in the past; in addition, many protocols have much higher fees than Ren does. (Uniswap at 0.3%, Coinlist wBTC at 0.25%, Curve at 0.04%). 0.1% is significantly below industry standard for our service and is insignificant/negligible compared to gains earned by RenVM users
• Mint and burn volumes can be measured independently
• RenVM currently has more flexibility to change fees now before additional major integrations at which point, we would have to more greatly consider impact to partners
• Many hypothesized that 0.1% to 0.2% mint fee would have little to no impact on volume, and as we can see from the large 1,500 BTC mints, there is indeed proof users are still using RenVM at high volumes even with the change in fee. Yes, more data here would be useful, which the 2 weeks in discussion of this RFC should allow us.

Reasons against and counterarguments:

• Experiment further on mint fee before adjusting burn fee
  • Raising the mint fee above 0.2% places us more expensive than our competitors. Bringing burn fee closer to an appropriate level brings us closer to a system equilibrium where experimenting further with mint fee can be more meaningful
• Increasing burn fee now will impact our data collection of the mint fee change
  • Although a valid concern, it is possible to fix two mispriced products simultaneously and fixing the pricing strategy of products is more important than collecting data on a single variable. The mint data would also be more meaningful with a burn fee that is closer to equilibrium.
• TVL decline velocity may decelerate
  • This is very well possible, but the only way to truly find out if a burn fee increase impacts volume is to test it. From what we’ve observed so far with mint fee change, the burn fee likely won’t impact burn volume either
• Increasing burn fee is unfair to minters that entered into a renBTC position expecting to exit their position at a 0.1% burn fee
  • Argument is invalid and can be used at any time. Fees change, and RenVM was not and is not morally or contractually obligated to keep the fee as is. However, to mitigate risk of negative perception, we can have a 7 day delay of implementing fees to allow those to exit their renBTC position if they so wish
• Increasing burn fees will eat into arbitrageurs or renBTC users’ profits
  • Arbitrageurs and minters likely make far more than 0.1% with their renBTC position. Our goal is not to be a free service, and I’m sure many RenVM users find value in our system and would be willing to pay a fair fee for utilizing it

Conclusion:

For a future state of greater decentralization of RenVM, we must reach equilibrium of TVB >= TVL/3. At time of writing, we have $327m TVL and $46m TVB. $46m !=> $109m. RenVM is out of balance by a factor of 2.4x or operating at 240% which is above capacity

To help illustrate the logic of this RFC, consider this analogy. RenVM is a factory that has been running above capacity at 240% (out of equilibrium which is fine in this stage of decentralization.) To reach equilibrium, would a factory manager wish to give away 140% of value or would he/she first attempt to raise factory capacity to meet demand? Raising factory capacity means increasing TVB (Ren value and nodes registered).

Aiming to increase TVB by increasing Ren value and nodes registered by reaching a more appropriate fee structure is overall value accretive for RenVM. Aiming to decrease TVL by decreasing burn volume or offering burn rebates is overall value dilutive for RenVM. If attempting to raise TVB is unfruitful, we can then, in the future, reduce our newfound, higher burn fee to better incentivize burning.

As a side point, potentially better for a separate RFC, polishing of and increasing marketing for wBTC.cafe to increase burn volume through wBTC might be an interesting lever to consider as we aim for equilibrium.

P.S. special thanks to @mitche50 for helping consolidate arguments for and against.

Implementation

1. Measure sentiment of this RFC on Nov 2nd; if positive, submit RIP on Nov 9th
2. If there is support for the RIP, make announcements to raise awareness of the increasing fees, so users may exit their renBTC position at 0.1%, if they wish
3. Increase burn fee to 0.2% on Nov 16th

Nothing is needed to implement this change. RenVM governance (currently controlled by the Ren team as per https://github.com/renproject/ren/wiki/Phases 5) can set the burning fee with a call to the renBTC (and other) smart contracts.

Sentiment polling as requested by @frilledshark (not official vote as this is a RFC not a RIP)

• Increase burn fee in the near future.
• Decrease burn fee in the near future.
• No fee adjustment in the near future.
• Focus on mint fee in the near future.

Edit: added a poll as requested by @frilledshark. First time making a poll in discourse, and he let me know I made it without showing votes (my bad). Can’t add show votes after polls been made unfortunately.

[loong]

This RFC is a more in-depth version of RFC-000-006 Raising Burn Rates 0.2%. It is closer - if not already - to being ready for RIP, so I recommend most discussion happen here. Leaving as an RFC for now, because no precedent has been set for burn fees yet.

[akinsawyerr]

I am in support. I like the idea of higher burn fees to incentivzes increased renBTC liquidity and support fees more in line with the industry.

[MaxRoszko]

I am strongly against this proposal:

1. The mint and burn fee already adds up to 0.3%, and could increase even further pending follow-ups on the increasing minting fee RIP. Such a high total fee already disincentivizes strongly against high-volume in and out trading, which would only get worse if the burn fee also increases. The preferred usage of RenVM is quick in and out mints and burns, to keep the TVL low but volume high.

2. Increasing the burning fee goes back on an implied promise to people minting renBTC previously, as they expected the burn fee to remain at 0.1%, meaning if they want to burn back BTC when the burn fee increases, they are hit with more fees than they expected which could have a negative reaction community- and marketing-wise.

3. If we every are to progress to the more decentralized phases, we need to get the TVL vs TVB in check, and increasing the burn fee goes directly against this endeavor, as it disincentivizes burning. Protocols like MakerDAO care about decentralization when evaluating adding new assets, and likely so for other DeFi projects. Additionality if we want to get out of the way for regulators, we also want to progress towards full decentralization as soon as possible. Increased burn fees goes against this.

The long play here is to get RenVM adopted and economically secure, not short term rent seeking, and for that lower fees, particularly for the burn step, works in our favor.

[Renchad]

I am against this also for now we need more integrations our only integration is a dying yield farming platform

[DeFiFrog]

Thanks for the response, Max. Here are some of my responses to your concerns.

@MaxRoszko

The mint and burn fee already adds up to 0.3%, and could increase even further pending follow-ups on the increasing minting fee RIP. Such a high total fee already disincentivizes strongly against high-volume in and out trading, which would only get worse if the burn fee also increases.

@DeFiFrog

Other protocols have changed fees in the past; in addition, many protocols have much higher fees than Ren does. (Uniswap at 0.3%, Coinlist wBTC at 0.25%, Curve at 0.04%).

It appears the market is fine with this level of friction at other protocols. Curve is the only one with lower fees, but their mechanism is a less value add of a step (swapping between like-assets on the same chain), and their swaps occur more frequently. For every 1 mint and 1 burn of renBTC at RenVM, I estimate that there will be up to 10 Curve swaps of that renBTC.

---

@MaxRoszko

Increasing the burning fee goes back on an implied promise to people minting renBTC previously, as they expected the burn fee to remain at 0.1%, meaning if they want to burn back BTC when the burn fee increases, they are hit with more fees than they expected which could have a negative reaction community- and marketing-wise.

@DeFiFrog

Argument is invalid and can be used at any time. Fees change, and RenVM was not and is not morally or contractually obligated to keep the fee as is. However, to mitigate risk of negative perception, we can have a 7 day delay of implementing fees to allow those to exit their renBTC position if they so wish

If this were to be an issue as you pose it, this would also accelerate our push to equilibrium of TVB/TVL.

---

@MaxRoszko

If we every are to progress to the more decentralized phases, we need to get the TVL vs TVB in check, and increasing the burn fee goes directly against this endeavor, as it disincentivizes burning. Protocols like MakerDAO care about decentralization when evaluating adding new assets, and likely so for other DeFi projects.

@DeFiFrog

This [disincentivizing burning] is very well possible, but the only way to truly find out if a burn fee increase impacts volume is to test it. From what we’ve observed so far with mint fee change, the burn fee likely won’t impact burn volume either

@DeFiFrog

Setting a higher burn fee now allows us to strengthen the lever of lowering burn fees in the future. Lowering burn fee to 0% or providing a rebate becomes even more attractive in the future if we raise fees now. A discount on an already cheap product is less meaningful.

I believe my proposal and arguments for this burn fee change do not directly conflict with the interest of partner organizations such as MakerDAO. The best way to determine this, IMO, is to consider their opinions directly rather than to speculate how we believe they would view the proposal.

[Ruzo11]

One “against” Point I’d like to add which hasn’t made it to your list, is that if raising the burn fee suddenly, people could feel like their btc is held captive. If given a long enough headsup that burn fee would be increased, I wouldn’t mind setting the burn fee to 0.2%.

Edit: my bad, I overread the part where DeFiFrog mentioned the 7 day notice period. Imho that period is too short and should be 2 weeks.

Edit 2: I’m really in favor of dynamic fees so people get used to having fees being dynamic. Non dynamic (burn) fees (unless being negative, which poses other issues), even when low (or 0) don’t incentivise burning, and might lead to a so called cobra effect. People would just continue to sit in renbtc as they would have to later mint again, and pay a fee again. (Or, when non dynamic, i,e switching every weeks between 0.1% and 0.2%, could nudge people to burn every week.)

Btc being safest on the Btc chain is no valid argument for Defi degens who chase crazy high yield, and thus no nudge for burning.

[mitche50]

As mentioned in the RFC I had helped discuss this with @DeFiFrog, so it should be fairly obvious that I’m in support. This is a detailed explanation of thought and I believe the arguments are strong.

Now is the time to set the norm for what fees will be moving forward, and give us some padding that will allow us to drop fees and still collect a meaningful amount on burning.

[Maggie]

I am against this proposal. I support @MaxRoszko comments but also believe it is too soon to start changing burning fees when we have not yet capture enough data from the minting fee increase.

I also believe we cannot pretend fixing the TVL/TVB issue through fees only. This would be equivalent to a government trying to fix all issues of a country through tax. We need more integrations and use cases for RenVM or ren assets capable to increase volume through RenVM. Once people can see more partners using RenVM and a constant volume flowing, this will be give more confident to doubtful investors.

Finally, increasing the burning fee in this precise moment when a hacker just used RenVM to exit makes it look like we want to profit from this type of scenarios which I believe is bad PR.

[Alexander]

Before I address the arguments, I want to create some baseline for my reasoning.

Curve.fi is the largest integration for RenVM at the moment. Curve.fi works by having a tight trading curve, that assumes assets are tightly bound around the 1:1 spot price.
Since the trading curve of Curve.fi is a bit more complex than the average one, we just need a general understanding of it.
In simple terms, Curve.fi weights certain prices and provides liquidity according to the weights at those prices. The higher the A, the more liquidity is pushed towards the 1:1 spot price.

That means a deviations from the 1:1 spot price increase the slippage paid when the assets move away from the 1:1 spot price, essentially reducing the efficiency of the market maker.

Looking at the current balances (and historical balances) of the pool, it is obvious that there is a greater renBTC to wBTC pressure than in the reverse direction.
Yield farming can easily explain the pressure to acquire wBTC.
And the price point, 1 wBTC to 1.0025-1.0035 renBTC, which is exactly 0,1% RenVM burn fee, + 0,2% (Coinlist fee) + 0,05% (Arbitrager profit).
Even after the Harvest finance hack with some of the largest burns, this is still generally true.

What does a change in burn fee mean for this?
The price point would be pushed by 0,1% more, so the spot price would be 1 wBTC to 1.0035-1.0045 renBTC.
But because of the Curve.fi market making curve, this would increase the slippage for more transactions.
This increases the cost of minting wBTC, which should generally decrease the usage of Curve.fi by more than increasing the mint fee by the same amount.

With this in mind, we can address why the reason for are not valid.

@DeFiFrog

• Setting a higher burn fee now allows us to strengthen the lever of lowering burn fees in the future. Lowering burn fee to 0% or providing a rebate becomes even more attractive in the future if we raise fees now. A discount on an already cheap product is less meaningful.

This depends on the integrations of RenVM. And those are uncertain at the moment and would be wrong to use as the base of an assumption.
The current target use case by far for renAsset is Curve, and apart from big users withdrawing, the majority of the burn volume is arbitrage.
This is obvious when you look at exchange rate between wBTC and renBTC that is around 1 wBTC to 1.0025-1.0035 renBTC.

So the argument doesn’t hold well.

@DeFiFrog

• Similar to our recent mint fee increase, a change of burn fee from 0.1% to 0.2% will likely result in non-material change in burn volume

Reverse argumentation. This is what you want to argue, so you can’t use it to show that it is true.

@DeFiFrog

• Increasing burn fee will double the rate of burn revenue and will 1) subsequently, increase earning potential and 2) theoretically, increase TVB by increasing $Ren token value and incentivizing more registrations

If we assume that any transaction generally has a mint and burn, the total fee is about 0,1% + 0,2% = 0,3%. Increasing the burn fee to 0,2% would not double the fees, but only increase it by 1/3. Only looking at one side of the equation creates the bias.

And if use the assumption I created initially, we would decrease the minting volume, which is probably hurting the adoption of renAssets long term.

With this in mind, it is probably wiser to increase the mint fee.

@DeFiFrog

• Other protocols have changed fees in the past; in addition, many protocols have much higher fees than Ren does. (Uniswap at 0.3%, Coinlist wBTC at 0.25%, Curve at 0.04%). 0.1% is significantly below industry standard for our service and is insignificant/negligible compared to gains earned by RenVM users

RenVM is infrastructure, not a project that needs to get enough funds to pay investors.
All of the mentioned projects except from Coinlist have a real cost associated with them. Uniswap have impermanent loss, Curve has depeg risk. Especially with DAI pools, where you can pay around 0,2% entrance fee. That fee has to be paid by traders.

Coinlist is difference as it is a centralised project. And they generally charge around 0,3% to 0,2%, which Coinlist is right in the middle of.

@DeFiFrog

• Mint and burn volumes can be measured independently

No. Argumentation from the first part.

@DeFiFrog

• RenVM currently has more flexibility to change fees now before additional major integrations at which point, we would have to more greatly consider impact to partners

While I agree on adjusting the fees now, I don’t think it should be done without any real argumentation. That is why I support adjusting the minting fee, as that has less consequences for our main integration where burn fee has a larger consequence.

@DeFiFrog

• Many hypothesized that 0.1% to 0.2% mint fee would have little to no impact on volume, and as we can see from the large 1,500 BTC mints, there is indeed proof users are still using RenVM at high volumes even with the change in fee. Yes, more data here would be useful, which the 2 weeks in discussion of this RFC should allow us.

This is mainly related to the mint fee and can be summarised in my initial argumentation. The burn fee would be more expensive to the users of RenVM compared to a mint fee.

With this in mind, I hope better arguments for increasing the burn fee is raised, since the main one I can see is why not?, which I simply don’t agree with.

[Big]

I think it is reasonable to increase fees. Introductory rates were for users whom risked a lot by using the platform early. 0.4% fee for a mint to burn is not unreasonable for the service provided. That said, i was and am still a bigger fan of continuous fee over any tx fee changes but since this is our route, i am in favor of this rfc.
I will mention with the Harvest hack i feel the single most pressing issue on Ren right now is to push phase zero, ready or not.

[DeFiFrog]

@Alexander

Reverse argumentation. This is what you want to argue, so you can’t use it to show that it is true.

This is why I said likely. This cannot be proven until tested.

@Alexander

Increasing the burn fee to 0,2% would not double the fees, but only increase it by 1/3. Only looking at one side of the equation creates the bias.

I did not say it would double fees, I said it would:

@DeFiFrog

double the rate of burn revenue

(Rate = 0.1 to 0.2)

@Alexander

And if use the assumption I created initially, we would decrease the minting volume, which is probably hurting the adoption of renAssets long term.

This also cannot be proven until tested.

@Alexander

That is why I support adjusting the minting fee, as that has less consequences for our main integration where burn fee has a larger consequence.

This point appears to be counter to your original argument. Aren’t you arguing that any increase to fees will increase the deviation between wBTC and renBTC in curve pools?

@Alexander

The burn fee would be more expensive to the users of RenVM compared to a mint fee.

Can you add more detail to explaining why you hold this view?

@Alexander

With this in mind, I hope better arguments for increasing the burn fee is raised, since the main one I can see is why not ?, which I simply don’t agree with.

I appreciate the additional points you raised, but I don’t appreciate this quote. You make good arguments against, but to discredit the arguments for the proposal by reducing them down to “why not” is a gross misrepresentation of others in the forum. I hope others that read the correspondences in this thread understand why this statement is not conducive to fruitful and productive discourse.

[MaxRoszko]

@DeFiFrog

@DeFiFrog

Other protocols have changed fees in the past; in addition, many protocols have much higher fees than Ren does. (Uniswap at 0.3%, Coinlist wBTC at 0.25%, Curve at 0.04%).

It appears the market is fine with this level of friction at other protocols. Curve is the only one with lower fees, but their mechanism is a less value add of a step (swapping between like-assets on the same chain), and their swaps occur more frequently. For every 1 mint and 1 burn of renBTC at RenVM, I estimate that there will be up to 10 Curve swaps of that renBTC.

RenVM’s fee is additional to what users are already paying on for instance Uniswap. They might be fine with 0.3%, but 0.5% is really pushing it for any kind of high-volume trading. We can’t make a simple comparison of our fees with Uniswap because noone is minting/burning renBTC just because, they are doing it in conjunction with other Ethereum operations.

[DeFiFrog]

@MaxRoszko

RenVM’s fee is additional to what users are already paying on for instance Uniswap. They might be fine with 0.3%, but 0.5% is really pushing it for any kind of high-volume trading. We can’t make a simple comparison of our fees with Uniswap because noone is minting/burning renBTC just because, they are doing it in conjunction with other Ethereum operations.

The argument you make against my proposal of raising burn fee can also hold valid for a future proposal for an increase in mint fee. How do you view this difference, and assuming mint volume remains constant, would you then instead be in favor of an increase in mint fee?

[Alexander]

@DeFiFrog

This point appears to be counter to your original argument. Aren’t you arguing that any increase to fees will increase the deviation between wBTC and renBTC in curve pools?

I am arguing that if we want more fees, we should increase the fees on the minting side, since it is more favorable.
It keeps us in the very liquid area on the Curve.fi curve, which an increased burn fee won’t.

Essentially, my argument is increase the mint fee by 0,1% is a lesser evil compared to increase the burn fee by 0,1%.

@DeFiFrog

Can you add more detail to explaining why you hold this view?

That is part of my main argument, it moves us into the less liquid part of the Curve.fi curve, which means more slippage which means more losses to arbitragers. Those losses could have attracted users or we could have taken them as fees.

@DeFiFrog

You make good arguments against, but to discredit the arguments for the proposal by reducing them down to “why not” is a gross misrepresentation of others in the forum. I hope others that read the correspondences in this thread understand why this statement is not conducive to fruitful and productive discourse.

I completly agree with you. The style doesn’t fit the discussion. I wanted to write a conclusion a compact way. I will try to rephrase it:

The argument for increasing the burn fee (versus the mint fee) can really be boiled down to data gathering or non existing.
Why? Since every argument relies on the same argument for and against, which is:

@DeFiFrog

This also cannot be proven until tested.

Where we can prove that our main integration’s users have a cheaper experience with a higher minting fee versus a higher burning fee.

[fuzzylogic]

I agree on raising fees to address TVL imbalance issues. Our approach should be to increase investors ROI (= more Ren buying pressure) not trying to disincentivize users by offering “race to the bottom” prices to leave renBTC.

DeFi users are looking for yield optimization, and generally seem quite unconcerned about exit/entry fees. Many here voiced concerns how an increase in mint fees would impact RenVM volume. It didn’t, as the past 7 days have proven.

Furthermore, as the Harvest exploit last night showed in an impressive manner: We’re still the cheapest player around in the DeFi space.

[fuzzylogic]

Also, quick reminder.
We’re already displaying the the website that fees are subject to change:

[Alexander]

@fuzzylogic

agree on raising fees to address TVL imbalance issues. Our approach should be to increase investors ROI (= more Ren buying pressure) not trying to disincentivize users by offering “race to the bottom” prices to leave renBTC.

You would simply lock more value inside, as people would be less likely to burn renBTC to release BTC from the RenVM custodian.
Yes, an increase in return for Darknode would result in an increase in Ren price, but this trend is anything but a direct correlation. When the yearly ROI of RenVM reached 25%, the Ren price increase by 10% and fell by an equal amount later.

@fuzzylogic

DeFi users are looking for yield optimization, and generally seem quite unconcerned about exit/entry fees. Many here voiced concerns how an increase in mint fees would impact RenVM volume. It didn’t, as the past 7 days have proven.

When why specifically the burn fee?

@fuzzylogic

Furthermore, as the Harvest exploit last night showed in an impressive manner: We’re still the cheapest player around in the DeFi space.

When RenVM carry the lowest volume, it should also carry the lowest fee.

@fuzzylogic

Also, quick reminder.
We’re already displaying the the website that fees are subject to change:

Yes, but a change should be up for debate. We should not change something just because it is the first solution we think of. We should try to find the best solution to address the problem and use that as a solution.

[MaxRoszko]

@DeFiFrog

@MaxRoszko

RenVM’s fee is additional to what users are already paying on for instance Uniswap. They might be fine with 0.3%, but 0.5% is really pushing it for any kind of high-volume trading. We can’t make a simple comparison of our fees with Uniswap because noone is minting/burning renBTC just because, they are doing it in conjunction with other Ethereum operations.

The argument you make against my proposal of raising burn fee can also hold valid for a future proposal for an increase in mint fee. How do you view this difference, and assuming mint volume remains constant, would you then instead be in favor of an increase in mint fee?

In general I think we should have low fees to incentivize growth at this stage, and later have application-specific fees, where high fees would be appropriate for yield-farming cases. Increasing the minting fee now though does help chill the ever expanding TVL a little which is good, because we don’t have to collateral to back it up if we want to go decentralized. More in favor of increasing the minting fee than the burning fee.

[davoice321]

I’m against increasing the burn fee at this time. The are many good reasons outlined above why we should be conservative from a fee adjustment perspective as we gather additional data.

I’m also a fan of:

-Algorithmic, use case-based fees (higher for low velocity use cases, and lower for high velocity use cases)

-Programmatic votes by darknodes to adjust fees based on market data and utilization across various use cases, rather than less rigorous comment or sentiment counting via this forum.

I’m for waiting at least 60-90 days before considering another mint / burn fee adjustment, depending on if additional use cases come online.