can we get a snapshot for option 3 then? they’re all pretty similar
A poll has been added to the initial post in this thread. Additionally, everyone should be aware of the following thread:
which provides more details on these options.
3 Likes
Also, reposting this video with my new thoughts on the issue:
1 Like
Just my opinion on the voting of the different options put forward. If we’re going off the majority for a clear winner I believe 2 votes is actually skewing the results somewhat. Assuming people are picking one of the 3 models as their particular favourite and then opting for the option with fee generating features and vortex increase as their 2nd choice is actually making that option look the like the most voted for when I believe a 1 vote poll would have the yield throttler as the clear winner.
Just to summarise, if we’re using this poll to determine our next path forward it would be better with a 1 vote person for a more accurate poll.
Hope that makes sense
Second that. Can we get a new one?
This. Yield throttling implemented NOW. Can figure out the rest later. I also voted for yield throttling and increased revenue.
1 Like
Hey, this is a temp check vote that is nonbinding since the Bancor DAO will ultimately have to vote via snapshot for an option. A temp check vote using ranked choice voting might be a good idea to get a sense of what $vBNT holders are leaning towards before a final decision is made (maybe for this upcoming Sunday @foxsteven FYI).
What I have gathered so far is that the “DAI Remittance” and “Protection Fund” models aren’t the favorite choices for voters (based on this poll). Most appear to like the yield throttling model and also increasing fees on the vortex while focusing on yield generating features. FYI, there is nothing that states that we can’t work on both simultaneously (a recovery model and yield generating features). I can redo the poll and make it single choice after 24 hours (tomorrow essentially).
3 Likes
I don’t have an issue with the yield throttle that people are mentioning in this video. My main issue is and continues to be the lack of fee generating potential here.
The main fee generating feature that many seem to be bullish on is protocol-level arbs. However, I don’t think there’s much value there simply due to the price impact on the Bancor leg of the trade. I understand that our “competitive advantage” is saving on the fees, but the bigger issue is the price impact incurred when doing this at scale.
Let’s take for instance a 100,000 USDC trade for ETH on Bancor.
This simple trade incurs 7.37% price impact. So let’s say we avoid the 1% fee, that gets the price impact down to maybe 6.37%. So the price differential between Bancor and another AMM has to be greater than at least 6.37% in order for there to be profitability. And the total value from this arb is really only $100,000 * (difference in AMM prices) * (1-other AMM’s fee %) * (1 - price impact of other AMM) * (1 - price impact on Bancor’s pool). Thus the price impact on the Bancor leg of the trade severely limits the profit potential. And this assumes that we can get the delta even that far wide. Bear in mind there will be normal traders who see the lower price on Bancor and will place trades there via DEX aggregators and prevent the delta from getting too large.
It gets even worse when we do larger trades at even larger scale. If you try to trade $1,000,000 USDC, you get 44.16% price impact. That makes large scale arbs virtually impossible.
Probably arbs would work best in the 10,000 USD per trade range so as to not suffer tremendous price impact. But then you run into gas issues eating into your profit margins.
I think what people are failing to realize is that liquidity is directly correlated to how much arb income is possible here. Lower liquidity results in higher price impact which increases the cost per arb.
The profit per arb is also capped due to the competitiveness of the trade. If we set the fee tier at 1%, we only have a 1% cushion before we get squeezed by the competition. And we have to have the delta get larger and hope that no one places trades to close out the delta before a profitable trade is possible.
3 Likes
I’m not sure how a mechanism of protocol arb would work?
Wouldn’t dynamic conversion fee which responds gradually to the curve of a demand surge work just as effectively to share any arb opportunity and be a more simple mechanism, as it only requires internal protocol activity monitoring & calc moving average and use a fee formula for higher than average demand (relative to an agreed MA time frame)?
BTW The idea was initially explored in Bancor chat in 2019.
Results of the poll which was just closed:
\
I have added a new one that only allows for single-choice selection. It will close on Sunday 7/24.
2 Likes
Theoretically charging more fees per increase in demand would lead to more profits. However, you’re discounting the fact that AMMs like Uniswap have static fees at 0.05% for their most popular ETH-USDC pool which has TONS of liquidity and concentrated liquidity which takes all volumes away when using a DEX aggregator.
The only volume that you’ll get is the arb volume that brings the Bancor AMM price to parity with other AMMs. That’s not going to be much.
Well it looks like Yield Throttling is going to be the winner. How long will this take to implement? Zeno’s video just said it was quickest but what is an actual time frame, and once it’s implemented, how fast should we expect to see results and the deficit start to fill in?
1 Like
Do we have any statistics on what percentage of Bancor volume has been?
Take a look at the DEX volume landscape in this dashboard.
Check out Bancor’s portion of trade volume for its top 10 tokens. It’s only getting 0.2% volume via DEX aggregators. That’s the volume that only gets allocated to Bancor when it’s price deviates so far from other AMMs that finally, it’ll get some volume to bring it to parity with other DEXes.
Oh and that is outdated. It’s probably less now given all the liquidity withdrawals since then.
BNT yield throttling is the obvious choice from the three options. I’ve got some questions and observations:
- Capital efficiency: How would be the off curve funds used? Can they be used? Are they just a idle reserve?
- LPs no longer compound their earnings if their earnings are stored off curve, this is negative IMO. I like the idea that I don’t have to do anythning to grow my investments, with cancelling the coumpounding my fees won’t earn fees on fees. Aren’t the withdrawals more expensive because the funds come from two sources (+ the ILP), pool and the off curve vault
-
Can the ILP reserve (ILPR) and the LP revenue split be changed on a per pool basis? (20% goes to off curve ILPR; 80% goes to LPs - can be changed for each pool?)
Explanation:- It is certain that there will be ILPR:IL deviations between each pool. Some pools could have 10:1 (means surplus) ratio and some 2:3 (namely the large pools) and the system as whole would be still in a deficit: 48 of pools in surplus, 3 pools in deficit → whole system in deficit → network fee remains high
- In pools with large surplus the (ILPR fee):(LP fee) split should be in favour of the LPs [10:1 ILPR ratio on a small pool could be quickly depleted by huge deposits and rapid price movements]
- OG LPs pay the most, they build up reserves, which the future depositors enjoy
- Would be the LP compensation at withdrawal from a pool in surplus be 100% and in pools with deficit <100%?
- How dificult would it be to implement in the future?
-
Could the network fee be used to fill up the individual ILPRs? The collected network fee would be swapped into a TKN and deposited into a ILPR off the curve to decrease the deficit.
- This would mitigate the per pool ILPR divergence; help to fill up the ILPR of the pools that are in more deficit quicker and reduce network fee afterwards.
- Somewhat alleviates the LPs burden of filling up the ILPR
- It would decrease the BNT price and possibly increase IL; bad if the IL is caused by BNT price fall, good if caused by BNT price increase [this lead me to this idea]
- How dificult would it be to implement in the future?
Hi everyone, I think the best way to solve this problem is to make the system function as intended.
Let’s assume that LINK holders are 45% underwater. This 45% should be compensated with BNT token if they would like to withdraw them.
These BNT tokens need to be locked and staked with some kind of vesting schedule to avoid the selling pressure. The owners of these locked BNT holders should gain some extra yield from the fees. That will help ease the pain.
I don’t know why the community came up with all these extra complicated plans for recovery. I agree that the benefits of those plans can be observed over the long term, but I don’t think the project can survive that long. We need something simple and can be implemented easily and quickly. All the other ideas to make the system more robust and create more yield can be implemented later on.
5 Likes
Hey guys, a bit of research and some potential solutions. Since you are all very familiar with Bancor you can skip to the solutions part.
I think Bancor can distribute the tail risks in the protocol to LPs up-front. The important aspect would be the long-term survival of the network and remove expectations for and the possibility of a death spiral. It would also add important value alignment between BNT holders and depositors. Perhaps a discussion could offer depositors a governance token for staking so they have a say in the network.
The other creates a more dynamic insurance pricing model for TKN LPs. The idea being as the network gets more risky, the cost of protection should rise.
Really looking forward to any discussion, thanks!
8 Likes
I agree with you, if bancor can make itself more sustainable then this will lead towards repairing the deficits.
Fee generation is one major topic that the contributors seem focused on, but I appreciate your take on granting the BNT token additional utilities which will help to strengthen the BNT token, stimulate organic growth, encourage staking and hodling etc. Sharing additional trading fees w/ bnt lps when it becomes optimal for the specific pool, pausing ILP when bnt minting is in excess of x%, perhaps as I mentioned in the other thread for adding additional features for bnt, perhaps bnt can also be used to purchase ILP.
I hope that others will be able to add onto your thoughts here and the focus towards repairing bancors issues will involve adding features that make the bnt token more desirable.
1 Like
Thank you for posting this - very interesting read
1 Like
Yeah I think most of the solution I am seeing are “how do we raise enough fees to pay the IL insurance” but that is not using the real magic of the Bancor design.
If you can drive BNT outperformance, your protocol deficit will fall. Meaning it’s really a confidence game at the moment. Now you could do some short-term thing about let’s juice the numbers for BNT so it outperforms. but I am not advocating for that. I think you put in a real fix that cuts the long tail from the protocol and that will help confidence tremendously.
Certainly could be paired with yield-throttling or something in the short-term to get outperformance and then returned to a more LP-friendly base case longer term.
1 Like