Addressing the current issues regarding AuraBal

Addressing the current issues regarding AuraBal

Hi everyone, There has been some discussion in the Discord, and I thought it would be a good thing to start a thread here to better keep track of the ideas and the different points made. As most of you may already know, AuraBal, Aura’s liquid wrapper of the BAL token, has been off-peg for quite some time, meaning a slower AuraBal accumulation and stagnation of Aura’s dominance.

Solutions to regain the peg

There has been already a few ideas that emerged. Regarding the peg:

  • @Solarcurve proposed to redirect more AuraBal farmed to AuraBal stackers to increase the incentives.
  • @Tritium proposed a buyback mechanism: “If AuraBAL is more than 2.5% off Peg (Or some other number) then all BAL earned by the farms should be used to purchase auraBAL on the open market and this auraBAL should be paid out as a reward instead of BAL. If AuraBAL is not off peg, emit nakid BAL so there is some exit potential“

How about solving the underlying problem?

Currently, we’re giving a lot of $Aura to acquire Bal. At the rate of 3 Aura/Bal farmed, 25% fees, 1.5$/Aura, this means that 3.6 Aura gets us 0.25 BAL. In dollar terms, to get 1$ of BAL, we pay 4$ of Aura. This is not sustainable, and poor financial management for the DAO. Getting more TVL through bizdev, while good, won’t solve this race to the bottom for Aura.

Given that the yield boost provided by Aura is still substantial (+75% of the initial yield at the time of writing, fees included), I believe an alternative (but not excluding) solution could be to further increase the fees taken.

Having 2x the fees (so 50% fee on the $BAL yield) would give us much more breathing room, and wouldn’t affect the yield LPs get: the boost would be of 50% (instead of 75% before). So, an LP getting a yield of 17.5% on Aura, and 10% on Balancer would see it reduced to 15% on Aura. Given the current general yield environment for Defi, it won’t reduce drastically Aura’s attractiveness.

Competition

Other players allow users to farm through their liquid lockers (StakeDAO and Tetu). Right now Aura is competitive regarding their proposal, mainly because they have very low TVL, which maxes their Bal boost. However, this won’t last is they get liquidity, and they don’t distribute incentives as high as Aura, so we have maneuvering room for the moment.

And then?

We can then use the fees collected to apply the solutions proposed by Solarcurve or Tritium.

5 Likes

Given that increasing Aura’s veBal ownership would also increase the boost seen by Aura stakers, increasing the fees Aura charges and using that additional Bal to fix the peg and then buy up more veBal (and thereby increasing the boost LPs get) seems like a win for everyone.

As to these three ideas, we can do all three and see what happens no?

2 Likes

Great post, @Yakitori, and a lively discussion today in the Discord server general channel. Let’s keep it up. I’ll save the balance of my commentary for now, but I think the proposals made today were actually slightly little different.

@solarcurve proposed a bit of a different mechanism. I think he proposed to have the auraBAL yield that goes to vlAURA to be comprised by automatically purchasing auraBAL from the market in order to support the peg. Currently, the auraBAL that goes to vlAURA positions is minted, not purchased.

I believe solar and/or you also suggested increasing the 25% fee that is currently charged by Aura on boosted BAL yields for BPTs that are staked on Balancer through Aura.

I also suggested, if feasible from a technical implementation perspective, to increase the 25% fee for BPTs that are staked on Balancer through Aura only for those Aura pools that are not receiving vlAURA votes but are receiving Aura emissions. This could be done in conjunction with any of the buyback proposals.

For more information on the fee, see the Aura docs, AIP-4, and AIP-13.

2 Likes

the highest impact change would be using all BAL allocated to auraBAL (from performance fees and from passive veBAL fees) to market buy auraBAL, then give that to auraBAL stakers. or even better, have native autocompounding like LOOKS does.

This should also be comparatively the simplest change to make. StakeDAO for example already does this.

I’m sure we could get away with raising the 25% performance fee even higher but personally luke warm on the impact that would have.

5 Likes

This is indeed a good option, but in my opinion we shouldn’t stop here.

I’m not really a fan of coining the 25% of Bal that is used to mint veBal as a “performance fee”. It’s not; Aura is not a yield optimizer or aggregator. Rather, we’re slowly exchanging Aura against veBal, at rate of 14.4 Aura per Bal acquired. Aura is not a charity aiming a relieving distressed farmers in need of a good yield.

On the long run, this slow grind is the biggest driver for Aura’s dominance over the Balancer governance. Aura’s emissions are widely too high for the amount of Bal it gets, and as a result, it’s not possible to follow a sustainable path for it, or start a flywheel. We don’t acquire enough veBal to counter the dilution, and the only strategic move is to sell the yield, which further reduces the incentives – a loop of death is forming.

Increasing the fees would at least slow down the dilution, and put all of this value to a good use. No one in the competition is incentivizing as we do, even with larger fees, so it’s not really a danger. The same goes with bribes: the dilution and Aura price is the main driver for their decrease over time. We shouldn’t release the prey for the shade and continue on building a strong foundation.

Think about it this way: by doubling the fees, we can keep on the current accumulation while having only 50% of the current TVL, and half the Aura emissions. Of course, we’ll have more than 50% of the current TVL, so overall the protocol will gain much more than it may lose. We have enough room for increasing the fee: we need to do it.

Otherwise, Aura’s price will converge toward 1/4th of its current value, which is equal to the fees it is exchanged against.

As for the autocompounder, llama.airforce does it already. It’s not a silver bullet, and it won’t solve the accumulation/dillution problem.

2 Likes

I posted this in the Discord, but will re-post here for full visibility.

Note that Convex has just put up a snapshot vote that aims to tackle this same problem. They are setting a 2% Treasury fee which is offset by lowering other fees – these Treasury fees are then used to swap CRV for cvxCRV in order to defend their peg.

Link to Convex proposal for reference:
https://snapshot.org/#/cvx.eth/proposal/0xb5106880bec1fbd0e94674ebe2f9595ec7de61e55312f609456db2f901c227c1

I would be strongly in favor of a simple mechanism similar to what they’re proposing there. Obviously the fee allocation would be different though, especially since Aura isn’t as far along in the emissions curve as Convex.

3 Likes

Market buys are a great idea.

It’s is better to auto-compound. Emitting to stakers will create a loop of sell pressure.

3 Likes

I think the solutions that @solarcurve and @Tritium proposed are a great way to start solving auraBAL depegging issue, both solutions will work hand in hand the way I see it and def think it is a good start. I also think that in whatever solution we propose moving forward, it shouldn’t be too complex, as more complexity can put us in deeper situations than we are already in.

1 Like

(post deleted by author)

Looks like Balancer is considering big changes to fee structure to ensure survival. If proposal A goes through, we’ll likely see the APR from staking AuraBal drop anyways regardless of any actions taken from this discussion.

Interesting to see Solar assessing whether veBal holders care more about “revenue from holding veBal” vs. “influence I get from directing emissions”.

1 Like

After discussing with the team, my initial calculations appear to be wrong (thanks @0xMaha, for pointing it out). Aura’s minting is done after the fees have been deducted, which means that for 1 BAL farmed, we mint 0.75·3.6 Aura.

This is great, however – it means that increasing fees decreases the inflation as well! For Aura to stay relevant, we need to have Fees < Inflation in dollar terms, with a safe margin.

At 2.07$ per Aura, here are the calculations regarding the yield increase (total yield = Bal yield * yield increase) granted by LP choosing to go through Aura:
25% Fee: +58%
33% Fee: +40%
50% Fee: +6%

Formula: (100-fee)*aura_per_bal * aura_price / bal_price - fee

This means that we have less room than I previously thought :sweat_smile:

I’ll summarize the ways that have been mentioned to ensure that AuraBal peg is kept tight:

A) Use some of the Bal yield to buyback AuraBal when it’s off peg, and distribute the yield to AuraBal stakers.

B) Increase the fees and use the surplus to either buyback AuraBal, or lock more veBal (can be both depending on the peg).

C) Use a % of the incentives paid through Hidden Hand – it would likely be only a complement though – around 150k/year at 2% of the rewards.

I’m not familiar enough with the contracts to decide which is the easiest solution to implement. A technical answer from the core devs would be welcome. However, I think some action would be beneficial for Aura as a whole, to further improve the cybernetics of the protocol.

2 Likes

Thanks for sharing this. Do you know how much an increase in AuraBal minting would come from bumping the LP fee to 33%? If we had an extra 8% to work with, and we spent that minting AuraBal, how much more AuraBal would we be minting? The more veBal we capture, the higher the more boosting power we can give to the LPs. there may be some equilibrium point we can reach sooner for them if we’re locking more?

Hi @bfloor, thank you for your patience.

From the Dune dashboard (which gets the % wrong btw, Bal incentives aren’t at 145k anymore), the protocol gets around 36500 Bal per week (or 255k$/w @7$/Bal).

Citing the docs:

There is a 25% total fee on all BAL revenue generated by Balancer LPs on Aura.

  • 20.5% goes to auraBAL stakers. This is paid out as BAL.
  • 4% goes to AURA lockers. This is paid out as auraBAL.
  • 0.5% goes to the harvest caller. This is paid out as BAL.

This means that weekly:

  • 29 930 BAL (=210k$) are sent to the AuraBal stackers.
  • Increasing fees by 100 BPS (1%) would yield 1460 BAL (10k$)

AuraBal buybacks

At 16.8$/B-80BAL-20WETH (source: Coingecko), we can use 12 485 B-80BAL-20WETH to buyback AuraBal. 4 weeks of buybacks (~50k of veBal) would increase the price by 0.56%. 6 weeks → 0.82%

Overall, as @solarcurve said, the easiest choice is probably to use the Bal destined to AuraBal stackers to buyback the excess AuraBal in the pool. This has two effects:

  • We get the peg back, so willing stakers will start to lock veBal, at Aura’s great advantage.
  • AuraBal stakers see their assets go up in price, and more importantly, are ensured that exit liquidity will be here when they’ll want to exit. From a game theoretic perspective, this should create a strong incentive to enter the pool.
  • StakeDAO’s clercks stop nagging about AuraBal depeg /s

Can I haz more feez?

As for the fee increase, from what I understand, it requires redeploying a new staking contract, and do a liquidity migration…again. While I believe it’s a needed measure in the long term, we had to go through this costly process not so long ago. It might be reasonable to wait for the effects of the other proposal, and a larger update, to go forward with it.

On the long run, Aura has a good pricing power, and a fee increase would be beneficial for vlAura holders. It would increase our voting power overall and the $ veBal per $vlAura ratio, that all the incentivizooors are watching like hawks.

Generally agree that using the BAL heading to auraBAL stakers is the best path.

As for the fee increase, from what I understand, it requires redeploying a new staking contract, and do a liquidity migration…again.

So changing the fees wouldn’t need a redeployment (the fee caps got increased with the recent LP migration). The thing that is locked in (and would require a redeployment to fix) is the part where you buy back the auraBAL before distributing it to the auraBAL stakers - this BAL allocation will always go to the base auraBAL staking contract. Luckily there is a way round that though by creating a wrapper contract and having auraBAL stakers move their capital to this new wrapper contract that converts all the BAL it earns into auraBAL.

Currently:

  • auraBAL gets 20.5% of BAL emission + AURA + bbaUSD

Solution:

  • Create new wrapper reward contract for auraBAL that has auraBAL and AURA as reward tokens
  • This wrapper reward contract will use all BAL and bbaUSD earned from the base auraBAL reward pool to market buy auraBAL
  • Could potentially add gamification/stickyness to give a boost the longer you stay in (like chicken bonds-esque or some sort of veModel) that get a boost the longer they are locked in
  • Change fees to split to 18.5% auraBAL and 2% platform fee (platform fee capped at 2%). Send this extra 2% directly to the wrapper as auraBAL to make the base yield much higher. This means that there is a strong incentive for auraBAL stakers to migrate to the new contract as the rewards are higher

Code should be simple depending on the addition/complexity of the gamification part. Only need to convert the BAL/bbaUSD into auraBAL (could use cowswap if the routing worked, or at worst do 2 on chain txs wrapping into 8020bpt and swapping using a protected fn).

Open questions:

  • Is the additional 2% increase in fees enough to incentivise users to migrate? If not, it’s possible to increase the allocation going to the AURA lockers and re-allocate that
  • Do we want this gamification/ve layer to provide stickiness?
3 Likes

Buying auraBAL with BAL + AURA + bbaUSD makes sense and this along with an additional 2% in yield should be a no brainer for auraBAL stakers. The yield can also be automatically compounded for all stakers which is kind of neat.

One thing I do wonder about is adding any sort of gamified lockup. Obviously it makes sense to try and defend the peg because in theory it will reduce sell pressure. But it could also be argued that:

  1. Stakers might not think the 2% additional yield is worth factoring in a lock.
  2. Because all rewards will be auraBAL its more likely Stakers will leave it auto compounding for the extra yield rather than market selling as they might do with AURA, BAL…etc. So perhaps that is already enough stickiness.
3 Likes

Made an account to add my thoughts.

Agree with solarcurve here.

I would like to add some additional thoughts to this discussion… very much alpha phase thoughts, but I think we can flesh them out.

  • Bonding vebal for discounted vlaura

It’s currently difficult for anyone w/ size to enter the aura ecosystem. Let’s take olympusdao as a case study. They’ve passed a gov proposal to acquire $1m worth of aura. Aura isn’t that liquid. Dao proposals get front ran. Byproduct of this, aura number goes up (temporarily), farmers dump it back down, current vlaura lockers get diluted. No one really wins.

What if olympus was able to acquire vlaura by bonding vebal to aura? Frontrunning wouldn’t be an issue for DAOs. Aura’s innate value increases as a byproduct of this event. Aura begins to capture more and more vebal market share. We could even configure the discount rate for Aura’s purchase of vebal to be significantly less than we currently pay w/ our farms… maybe even 0 discount seeing as there will be no slippage on the transaction. This is a long term win for the protocol.

3 Likes

Can you explain the bonding idea in more detail and maybe offer an example? I don’t follow what you’re proposing. Welcome to the discussion, sir. Don’t Aura lockers already earn AuraBal, so locking Aura creates veBal capture already? You’re proposing something different?

1 Like

This is awesome. I don’t have a strong view on which version of this we do (2% to the wrapper or send more value to Aura lockers) but I feel pretty confident we should NOT mess with gamification stuff. I just think it’s very easy to make things overly complex, sort of cheap and confusing-feeling and whatever we put in place will have people wanting to change it all the time. Better to just keep things simple but make changes to be buying AuraBal instead of just giving Aura lockers Bal.

I think one of the reasons Aura has had a good first six months is because the UI is simple, and well organized and it doesn’t feel like it’s got “traps” or “gotchas” in it. Users feel safe like they understand everything and what it does. Adding game elements risks nipping away at that “safe” feeling.

2 Likes

I fully support this initiative, that is in line with the problem that I was mentioning earlier (we pay too much to acquire veBal). Given that DAOs are much likelier to keep Aura locked for a long time, this is a much better way to acquire veBal than the current farming emissions, which are dumped at some point. Thanks for this idea!


Regarding gamification, while it sounds exciting, I agree with bfloor that keeping the UX simple to acquire more stakers outweighs the small gains of gamification regarding churn. Furthermore, keeping contracts simple means less dev time, quicker release and less attack surface.

2 Likes

We should decrease emission rate to farmers and add a bonding mechanism for people to acquire aura for a fraction of what we are paying now… they can’t get better / “safer” yield elsewhere so… it’s not like they’d bail

2 Likes