AIP 31 - Change in ANGLE emissions

Hey everyone!

As most of you already know, ANGLE emissions are a very important aspect in the protocol. It is a way for the protocol to grow, and to distribute the tokens to a wide user base. However, it also affects the token price as some of incentivized users dump their rewards.

After this recent tweet from Dan Robinson, we decided to revise our original distribution efficiency analysis available here (see V1 - April 2022).

The new version is available in the V2 - October 2022 page. Here, we lay down our conclusions, which will be split into two votes: AIP 31 #1, and AIP 31 #2. The goal is to discuss and move to a vote with AIP 31 #1 in priority, and then follow with AIP 31 #2.

AIP 31 #1 - Kill ANGLE rewards for Perp

The complete study can be consulted publicly here.

This is a screenshot of the analysis of ANGLE distribution on USDC, DAI, and FRAX Perpetuals, though DAI receive almost no rewards anymore.

Conclusion

Basically, our thinking is that ANGLE paid to Angle Perpetuals is not what is attracting them, as we can see from the low payout yield for USDC Perps, currently at 0.62%. This means that the position size of a USDC Perp is only getting paid 0.62% / year in ANGLE, which is relatively insignificant compared to recent EURUSD price action.

On FRAX’s side, there is an additional negative component: the hedge ratio is at a maximum but veANGLE/sdANGLE holders keep voting fro FRAX Perps, so that we have been overpaying them for a while.

We think all Perps gauges should be killed, so that Perps stop receiving ANGLE rewards.

AIP 31 #2 - Reduce global emissions

In general, we think there is a case to reduce emissions by 20 - 30%.

ANGLE emissions are a tool for growth and decentralization of the user base. In the middle of this bear market, we can see that not many new users are coming to DeFi, and people receiving those incentives are mainly returning users.

For this reason, and to keep tokens to be distributed in the future when growth comes back to DeFi, we think the protocol should reduce its emission rate by 20-30%.

Details

A setRate() function can be called in the AngleDistributor to reduce the emission rate by a specific factor. This would reduce the weekly quantity of ANGLE tokens taken from the AngleDistributor. The remainder could be used to keep distributing tokens for a longer period of time at the same rate, or increase the emissions in the future when growth and users come back to DeFi.


Other aspect of the analysis

The agEUR pools analysis tells us that agEUR-USDC UniV3 is well above average in terms in terms of TVL (agEUR) / $ANGLE and revenue, while agEUR-ETH is well below. The lower TVL could be coming from the fact that it’s much harder to be profitable by LPing on agEUR-ETH in general, especially with decreasing ETH prices.

Another important point is that any trade above 100,000 agEUR with ETH is routed at least at 90% through the agEUR/USDC - USDC/ETH path.

We think we should increase agEUR-USDC rewards and lower agEUR-ETH ones.


Looking forward to hear your thoughts and feedbacks! That would help a lot if you could spit your comments in two parts, one for each vote of the AIP. Thanks!

Thanks for the proposal @tuta!

On the rationale of killing perp gauges, I’d add the fact that if we’re not to be enough hedged, it’d still be possible to re-activate the gauges and take advantage of an announcement effect of having ANGLE distributed again to perps.

As for the reduction in incentivization rate, I’d start with a 20% reduction, just so that we can then see if it’d be worth increasing or decreasing again.

Hello @tuta ,

Very interesting, thank you!

Concerning the agEUR pools:

  1. I don’t see Angle rewards for agEUR/EUROC LP ??? Where do these 450 000 Angle come from? Isn’t it 3EUR LP instead?

  2. I’m not convinced with the agEUR/Angle metric to compare Angle rewards distribution efficiency! (What is important is the low slippage seen by end user and not the amount of agEUR in the LP, no?)

  3. The result of the change of distribution rate for Angle is difficult to evaluate as there are plenty of effects and counter effects ! I guess veAngle and sdAngle would vote for their own short term interest!

Regards,

1 Like

Hey 0x09, thank you for your comment!

1. Curve incentives

The Curve Incentives have been switched from the 3EUR to the agEUR-EUROC pool following AIP 25 (governance post, vote)

2. agEUR/$ANGLE metric

I think both metrics are important. However, the pool slippage depends on how efficient LPs are and not on how much ANGLE are being distributed on the pool. What we were trying to gauge in this analysis was if the revenue from a pool justified its ANGLE rewards. Then of course, this needs to be interpreted with the added value of each pool.

3. Difficult to evaluate a change in ANGLE distribution rate

I agree on this. It’s mainly a bet at this point but one I think might pay off in the long-term, and reduce the pressure on ANGLE price in the short-term!

I have several thoughts on these proposals.

First, with respect to the HA gauges, I don’t think you can come to hard conclusions about user behavior from these data. The reason is that I have been responsible for 50%-60% of the USDC HA open interest for most of this year. I have used the perps to hedge my large exposure to agEUR, as well as to bet against the Euro, so the rewards were not as important to me. However, I have recently turned neutral/slightly bullish on the Euro, so I’ve closed my perp positions, leaving the protocol quite underhedged. Speaking for myself, I am unlikely to open new perps without ANGLE rewards until I turn bearish on the Euro again. If the protocol is hedged during a Euro downtrend, but underhedged on an uptrend, the protocol loses quite a bit of money.

I’m not saying we shouldn’t kill the HA gauges, but just keep in mind the risk of not incentivizing HA’s if the Euro goes up in value, since that is when we actually need the HA’s, and the data is skewed by one large (and pretty talented) trader :wink:

On the subject of shifting more rewards to USDC/agEUR vs ETH/agEUR, what is the mechanism for doing this? The rewards are decided by veANGLE votes, so unless we change the functionality of veANGLE I’m not sure how we accomplish this. I’d like more clarification here.

And finally, on the subject of reducing emissions, I think this is probably a good idea. However, we still need to improve the value proposition of locking ANGLE vs dumping it. Otherwise I don’t think this will have a very large effect. I’d like to see the change in revenue sharing to veANGLE holders implemented along side this reduction in emissions for maximum effect (AIP-17). AIP-17 received broad support, in particular using Tuta’s proposed realignment of fee sharing, and I think it is time it is implemented, since there is significant crossover between that proposal and these proposals, and the goal is the same (reduce ANGLE sell pressure and improve capital efficiency of emissions).

2 Likes

Thanks for your reply and thoughts here @GTRminator! And congrats on your HA profit!

Some thoughts on my side:

  • You’re right about the risk of decreased HA rewards if Euro increases again. I personally believe that killing HA gauges is something which should not be permanent and we could activate them back again if the trend goes in a wrong direction.
    On top of that, I think we still have a lot we can do (with oracle work) to reduce HA fees so that even people who want to take small positions for short periods of time can do it.
  • There’s no hard coded mechanism to shift ETH/agEUR to USDC/agEUR. Just an insight from the analysis which will on my side lead me to allocate more voting power to USDC/agEUR with respect to ETH/agEUR.
    When we have a Chainlink feed on Euler, I think I’ll support the fact of killing the ETH/agEUR gauge which is inefficient
  • About the redistribution of the fees, I’m sorry if we haven’t made much progress on this over the last few months, and I take my responsibility for the status quo. Thing is that this proposal is harder to implement technically than the other proposals because the protocol was not coded to allow for that, and I prefer working on building products that can 10x the adoption of Angle rather than working on improvements that can will more short term effects.
    Maybe an alternative technical option could be to increase the interest going to veANGLE holders so that it matches in some way the fees the protocol makes. Not the same setup, but can have the same effect.
    But, putting that aside, I still wonder to what extent increasing amounts distributed to veANGLE holders will reduce the ANGLE sell pressure. APY for veANGLE holders is at the moment 24%, and increasing fees distributed would increase the APR, but to what extent will we have people more people buying (or not dumping) and locking?
    And believe me I’m a veANGLE+sdANGLE holder, so it’s in my short term interest to increase fees distributed, but what I have in mind now is the value of the ANGLE token in 4 years rather than in the coming months (even though both matter)
1 Like

Could you expand a little bit on the difficulty in shifting fee revenue to veANGLE holders, if it’s possible to do so without being too technical?

If it is indeed too large of a protocol change to do right now, I would be open to redistributing interest instead if it can have a reasonably similar impact to veANGLE income. I’m certainly interested in focusing on 10x improvements as well, so I don’t want to spend an inordinate amount of dev resources on incremental improvements either.

I will push back a little bit about increasing veANGLE revenue being a short term benefit though. I think it has powerful long term positive impacts. In a bull market, rising prices combined with high apr’s create a powerful flywheel effect of higher veANGLE income → less dumping → higher ANGLE price → more agEUR usage for farming → higher veANGLE income, and repeat.
And speaking for myself, I would be much more inclined to lock more ANGLE vs selling it if the veANGLE apr was higher, particularly in a bear market when prices are unlikely to rise.

Would it be possible to publish an analysis of the effects of shifting interest to veANGLE vs shifting interest + shifting protocol fees, so we can see the difference and weigh the benefits of both approaches?

Thanks for your quick reply!

For the explanation, every time interest accrues to the protocol, there is in the code a parameter responsible for doing the split. For fees we don’t have that, and on top of that fees accrue to many different contracts which makes it hard to have a centralized accounting place.

You’re right on the increased ANGLE price, I agree that it’s important to have a good price for building network effects.

As for the effects of shifting interest to veANGLE vs. shifting interest + protocol fees, you can take a look at this chart we have in the analytics where we breakdown protocol revenues by fees and interest. I don’t think that in the chart Core interests includes as well revenue distributed to veANGLE or not.

A lot of interesting thoughts, thanks @GTRminator !

Part 1: killing Perps gauges

I agree that this very light data can’t make us reach any obvious conclusion. They point towards the conclusion of ANGLE rewards being insignificant for Perps, but there is also a part of guesswork.

I think the relevant question to ask then is: “Would you open new perps while being bullish on Euro only for ANGLE rewards?”

Talking for myself, the answer is a clear no, and I believe it’s the case for most people that would be looking to speculate directionally on EURUSD. If the Perps holders end up being only delta neutral farmers (short EURUSD on Angle, long EURUSD elsewhere) to dump ANGLE, it means that the model simply doesn’t work.

My opinion is that there is a possibility of seeing some Hedged amount even with a bullish EURUSD trend (hedging for the sake of it for example), and that it isn’t a few bps of ANGLE rewards that will make the difference.

Incentivizing agEUR-USDC rather than agEUR-ETH

Nothing to add to @sogipec 's comment:

Part 2: reducing ANGLE emissions

I still think AIP 17 makes sense, as I believe that 1) delta neutral Perps would work much better with some source of yield other than ANGLE tokens, and 2) the current distribution of revenue is not ideal.

However, I have two other aspects in mind:

  1. It requires much deeper work than this, and I believe AIP 31 (Part 1 & 2) is already a first step towards implementing a change in how Angle revenue is being redistributed between the different actors.
  2. I don’t think our goal should be to increase revenue distribution to veANGLE holders yet. They are already earning 50% of all interest generated from strategies. There is no real way to increase their revenue here without impacting SLP, which would be better done through the complete rework of AIP 17.

In general, I see AIP 31 as cutting the cost which should come first, and AIP 17 as reorganizing the use of revenue, which should come 2nd. I agree we should start to evaluate how difficult / long this change would be to get a better idea of the feasibility and the potential timeline.

Didn’t we conclude that SLP’s are getting paid too much though? As a first step, couldn’t we redirect a portion SLP interest to veANGLE?

Also (not directly related), I think it would be a good idea to begin thinking about accumulating CRV and/or CVX with some of the surplus while prices are low. This would allow us to reduce the amount of ANGLE paid as bribes as well.

All in all, I generally agree with the direction here. I just wanted to be sure that AIP-17 hasn’t been forgotten about, and that we work towards a more optimized fee and interest split.

2 Likes

SLPs are indeed getting paid a lot, but they are also one of the major sources of revenue of the protocol.

About acquiring CRV and CVX, the protocol already owns some CRV and CVX, I am however unable to say whether prices are high, low or if current valuation is fair. With the AMOs, the protocol is going to start accumulating more and more CRV and CVX (at a low rate though but still).

1 Like

Both votes are now live:

I just want to add that with the new univ3 reward system it has been difficult to put funds in agEUR-ETH or agEUR-USDC and get a boost in StakeDao because they only support the gelato pool and not the custom ranges. I love the custom ranges but if I can’t get a boost on it it’s not worth it. If that were fixed it might increase TVL for the same amount of emissions

1 Like

Hey @PouncerK7s, are you the owner of pouncer.eth?

If yes, would you mind sharing the reason why you voted against reducing ANGLE emissions (Part #2) ?

Thanks for pointing it out @Tokenwood. I guess having Stake DAO integrating Gamma strategies would help. Have you tried asking them ?

I asked Chago a while ago and he mentioned it was technically difficult to implement and not sure it was worth it.

Hi,

I’m currently not an Angle Protocol SLP, I just hold veANGLE, so I’m comfortable with my position, which I’ll defend next.

We cannot forget that the Angle Protocol goes further than a CDP or stablecoin minting protocol. The Angle protocol is also an investment powerhouse, having said that, I do not agree that the remuneration of the SLP is exaggerated, since they are a source of liquidity for the protocol, namely for its investment strategies (greater liquidity = greater volume in the strategies investment = higher volume of interest income = higher equity growth and higher distributions to veANGLE holders).

2 Likes

Hi @sogipec,

Not sure about the weak correlation between veANGLE revenue and ANGLE dumping that you mentioned.

Please give me your feed back on my forum discussion “Protocol Interest Distribution to veANGLE holders (classical veTokenomics)”.

Thanks

Just answered your governance post