Proposal: Re-enable LM Rewards on YFI

TL;DR

  • Re-enable LM rewards on the YFI pool for 12 weeks.
  • YFI has attracted a liquidity pool of $36M at the time of writing.

Voting Instructions

Discussion

Yearn currently has a $36 million Liquidity Pool $18M of which is YFI tokens according to Dune analytics. Yearn LM rewards expired about 2-3 weeks ago and since then, the pool still retained 75% of its TVL while increasing the fees generated for the protocol.

Exhibit A:

YFI provides as much revenue and up to 2x revenue for the protocol compared to recently extended LM Rewards tokens (AAVE & SNX)

Exhibit B:

YFI provides as much volume and up to 2x volume for the protocol compared to recently extended LM Rewards tokens (AAVE & SNX)

Exhibit C:

Since the end of LM Rewards for YFI, the pool retained ~75% of its TVL (24M down to 18M)

Additional data:

Since late march, Yearn increased its Total Value Locked from ~1B to ~3B currently.

The fundamentals behind the governance token has been increasing in a fast phase, providing YFI with more volatility and more volume within the same time period.

I propose to re-enable the LM rewards for 12 weeks to keep liquidity on Bancor and attempt to encourage even more deposits & for current deposits to remain longer within the protocol.

Competing with others AMMs is harsh, we have the advantage of offering single sided IL protected liquidity provision, enabling similar LM Rewards as on the previous period would return an 8-12% on the deposited YFI which will prevent deposits from leaving Bancor and going into other AMMs such as Sushi that provides ~40% APY on it’s YFI/WETH pool with no IL protection.

Sources:

1 Like

We actually opened up a huge amount of space in the YFI pool when rewards were still active - it was never filled (and is still unfilled). I think if we were to reboot the rewards, it is unlikely to attract any additional YFI liquidity. It is also unlcear if the pool should be made any larger - the APY gets lower as the pool gets larger; however, the IL scales linearly. It is also more difficult to recreate LM on a pool after it has expired, as opposed to extending it or starting it from scratch.

  1. “YFI’s space not being able to be filled up”

I don’t think there is a case for rejecting this proposal based on this point, why?
a) AAVE available space: $9.7M (LM to be extended)
b) SNX available space: $13.6M (LM just extended)
c) YFI available space: $7.2M (3 weeks without LM rewards and $6M lost in TVL during the same period)

source: Bancor Network

  1. “It is also unclear if the pool should be made any larger - the APY gets lower as the pool gets larger; however, the IL scales linearly.”

The statement is correct, but the same argument could be applied to any of the tokens listed within the protocol. And based of the numbers provided on the proposal (Exhibit A & B), YFI is one of the best positioned pairs across the latest LM Rewards program extensions.

  1. “It is also more difficult to recreate LM on a pool after it has expired, as opposed to extending it or starting it from scratch.”

Where there is a will there’s a way. I’m not too familiar with the LM Rewards contracts from that perspective, can’t it be started from scratch again? either via: a) making users re-stake their positions in a new contract b) migrating contracts (huge security issues)

1 Like

Not quite - is part of the pool. To start from scratch, we need to make a new pool and migrate.

See above comment. Renewal is easy - remaking is hard.

Another important point is that YFI had a huge run - it was the first pool to have LM activated. The program on the YFI pool ran from early November, to March. AAVE and SNX are not quite as long as that. In any case, I do think that constantly renewing LM on these already deep pools is missing the point.

1 Like

Okey, understandable. The proposal could still be posted on snapshot and it would become a matter of willingness/availability rather than anything else, right?

Those stats are related to point 1) of your initial comment, I understood the renewal/remaking stuff.

This point is completely understandable. I am particularly biased on having LM Rewards on the pool as I’ve been trying to get in the pool since early days and I wasn’t able (was only able to join from the BNT side), so I have a point to make here:

YFI’s single sided deposits were filled within hours of being opened and it wasn’t until the 11th of February that more capacity was added, at which point there were only 6 weeks left of LM Rewards. So every single LM program that was done before that date, they were all just free money for the people with deposits already inside, no one within the YFI ecosystem was able to join in them apart from the last one (mid February until end of March).

My point being here: those LM programs were done for marketing (which is lame but understandable) and not to truly attract new capital/users.

2 Likes

Yeah - nothing prohibiting it from being voted on.

I totally see your point of view. At the same time, this sounds personally motivated. It is important that these proposals be argued as a benefit to the protocol, rather than whom is entitled to LM rewards.

It was to attract new capital and users. At the time, it was assumed that deeper pools = more trades. An underlying assumption to the approach of Bancor with v2.1 is that DEX aggregators like 1inch would become popular - they haven’t. Swappers seem content to get worse trade value on other DEXes, revealing an imperfect understanding of how profits are derived from TVL. This means that, unfortunately, deeper pools are not necessarily the best strategy anymore, and we have to pivot our LM rewards towards a more productive use case. It seems that increasing the token inventory (i.e. have more pools, rather than deeper pools) is an experiment worthy of pursuit.

I am working on a new LM program with @MichalHerzyk which should help to generalize the LM program, and properly incentivize LP behavior in a manner that is conducive to protocol profitablilty.

It isn’t personally motivated, it is to explain from my anecdotical POV how the process was at the time (0 available space for 4 months + keep pushing LM programs), I can actually go back to discussions I’ve had with other folks within the YFI community and provide a more accurate timeline if necessary.

Why LM Rewards program were kept being pushed if there was no more space in the single sided liquidity provision?

Maybe new users through aggregation protocols, but the LM program for YFI could have been way more successful from my POV if the protocol would have co-invested on the pool earlier (as I commented in the previous answer).

This is factually incorrect, problems originate elsewhere, but I’m not going to go into product stuff because its not related to this topic (and I don’t want to be tagged as someone that is just FUDing / speaking non-sense about Bancor)

100% agree with this from the standpoint that going all-in into a strategy doesn’t make much sense**, from that perspective there is a ton of stuff to be improved on how inflation within the protocol is coordinated/managed.

** It would make sense if governance really goes ahead and decides to make the “liquidity black hole” theory a reality, which IMO would work perfectly on how we are currently positioned in the market

19th of November
Already 0 space within the YFI pair, and your comment there was already not supporting the idea of increasing the caps

30th of November
First Liquidity Mining Program for YFI proposed (from BNT team)

2nd of February
Proposal to increase co-investment cap

11th of February
Proposal to increase co-investment cap passes enabling a ton of space but just under 5 weeks LM Rewards

How is it possible that we went from “YFI community is starving from space in the pool” to “their community doesn’t want LM rewards, don’t worry about that pool” in just under 3 weeks?
image
(February 27)
a) There was still time for a YFI LM Rewards extension (community was requesting it)
b) Pool capacity wasn’t getting filled (there was a clear issue going unnoticed)

1 Like

At this point I think that the proposal has sufficient data and additional information around YFI, the risks & motivation behind it, that it would be cool if we make it progress to LEVEL 2.

Do let me know if you need anything else from my side.

1 Like

Yeah man - I think it is good for level 2. I’ll move it up end of the week, to give the community some time to comment.

During that period we were discussing with the YFI team that they were going to deposit $100,000,000 in YFI tokens, and they weren’t interested in LM rewards. We created the space for them, but they never deposited any tokens.

Notice that these comments reflect the idea that TVL was the #1 priority. We now know this was a misapprehension, or at least that the data suggests growing the deep pools with LM rewards achieves very little.

I would still like to move our discussion to how the LM program in this case is being used to specifically benefit the protocol. We don’t have to - I will push this to Snapshot for you either way.

I am of the opinion that we should let all the other communities that benefit our platform have a chance at a liquidity mining campaign at least once during the 72 week trial period. This is how we grow our user base and convert people over since most are not aware of our value proposition (IL protection and single sided deposit) that differentiate us from other DEXes. I am typically not hugely in favor of renewals unless there is a strong argument (e.g. The Aave renewal is useful since we are applying for having BNT onboarded as collateral Proposal to extend LM on AAVE for an additional 12 weeks - #10 by glenn). Can a similar value proposition be made from the YFI community to the Bancor community that would make for a stronger argument?

Since fees were mentioned, I was curious to see how Aave, SNX, and YFI ranked against each other given that each liquidity mining program runs for ~12 weeks and we mint 10K BNT per week (I think all these pool get LM rewards at the medium level). That’s roughly 120K total BNT for a campaign assuming that everyone gets rewards at 1X but since most people stick around for the full 12 weeks then we actually mint a max of 2x120K BNT = 240K BNT and that’s ~$1.68M (or $840K at 1x) assuming BNT price at $7.00.

Full data is at the bottom, I started from the week of 12/28/2020 until now and got the following:

Token Total Fees Total Volume
AAVE $252,653.17 $135,637,961.80
SNX $288,283.13 $153,246,681.17
YFI $337,713.45 $169,551,424.21

Going by the above, YFI did have the highest total fees in this period but the cost to us was a lot higher (we minted 120K-240K BNT in rewards). By this measurement alone, I wouldn’t have personally renewed any of these pools and I only voted in the renewal for one of them (not because of the fees argument but mostly due to a strategic decision).

With that said, fees are not everything and user acquisition is my opinion is more important. If we have to chose between extending rewards on existing pools vs enabling LM rewards for a new pool that would potentially let us bring hundreds to thousands of new users to our platform then we should always vote for the latter since ultimately that’s what we need for growth. Even though we might not be able to make up the cost of rewards in fees during the LM period for a pool, in the long run those new users might stick around and make Bancor their forever home (e.g. you mentioned that:

“Yearn LM rewards expired about 2-3 weeks ago and since then, the pool still retained 75% of its TVL while increasing the fees generated for the protocol.”

maybe the value proposition of Bancor is strong and they really do love what we provide?
).

symbol time Conversion Fees in USD Volume in USD
AAVE 12/28/2020 0:00 $221.68 $110,616.75
AAVE 1/4/2021 0:00 $1,733.30 $1,007,939.68
AAVE 1/11/2021 0:00 $3,479.43 $3,475,948.72
AAVE 1/18/2021 0:00 $3,825.41 $3,821,587.80
AAVE 1/25/2021 0:00 $5,371.40 $5,366,029.64
AAVE 2/1/2021 0:00 $11,688.16 $8,915,280.30
AAVE 2/8/2021 0:00 $11,921.27 $5,948,715.87
AAVE 2/15/2021 0:00 $27,615.70 $13,780,236.65
AAVE 2/22/2021 0:00 $29,636.08 $14,788,402.26
AAVE 3/1/2021 0:00 $28,790.37 $14,366,396.51
AAVE 3/8/2021 0:00 $22,587.47 $11,271,146.19
AAVE 3/15/2021 0:00 $12,046.95 $6,011,426.77
AAVE 3/22/2021 0:00 $11,900.03 $5,938,113.06
AAVE 3/29/2021 0:00 $16,515.26 $8,241,113.23
AAVE 4/5/2021 0:00 $17,592.65 $8,778,733.91
AAVE 4/12/2021 0:00 $37,864.10 $18,894,184.30
AAVE 4/19/2021 0:00 $9,863.91 $4,922,090.14
$252,653.17 $135,637,961.80
symbol time Conversion Fees in USD Volume in USD
SNX 12/28/2020 0:00 $2,479.38 $1,237,208.62
SNX 1/4/2021 0:00 $6,732.98 $3,459,783.15
SNX 1/11/2021 0:00 $4,308.09 $4,303,779.96
SNX 1/18/2021 0:00 $4,144.25 $4,140,101.55
SNX 1/25/2021 0:00 $4,787.77 $4,782,978.57
SNX 2/1/2021 0:00 $10,122.72 $7,724,558.57
SNX 2/8/2021 0:00 $13,164.69 $6,569,177.90
SNX 2/15/2021 0:00 $24,621.67 $12,286,211.74
SNX 2/22/2021 0:00 $38,680.99 $19,301,814.77
SNX 3/1/2021 0:00 $39,445.36 $19,683,234.03
SNX 3/8/2021 0:00 $17,602.21 $8,783,502.66
SNX 3/15/2021 0:00 $14,885.25 $7,427,740.15
SNX 3/22/2021 0:00 $20,477.86 $10,218,451.19
SNX 3/29/2021 0:00 $26,570.33 $13,258,595.51
SNX 4/5/2021 0:00 $21,867.34 $10,911,802.80
SNX 4/12/2021 0:00 $29,610.42 $14,775,601.75
SNX 4/19/2021 0:00 $8,781.84 $4,382,138.24
$288,283.13 $153,246,681.17
symbol time Conversion Fees in USD Volume in USD
YFI 12/28/2020 0:00 $13,382.23 $3,810,112.21
YFI 1/4/2021 0:00 $29,762.12 $8,612,856.64
YFI 1/11/2021 0:00 $7,166.51 $7,159,344.11
YFI 1/18/2021 0:00 $4,956.63 $4,951,676.31
YFI 1/25/2021 0:00 $3,444.45 $3,441,005.74
YFI 2/1/2021 0:00 $4,450.58 $4,446,125.24
YFI 2/8/2021 0:00 $22,350.51 $11,282,294.23
YFI 2/15/2021 0:00 $30,992.24 $15,465,125.34
YFI 2/22/2021 0:00 $21,501.71 $10,729,354.85
YFI 3/1/2021 0:00 $32,229.26 $16,082,398.62
YFI 3/8/2021 0:00 $26,332.40 $13,139,867.97
YFI 3/15/2021 0:00 $17,578.87 $8,771,858.35
YFI 3/22/2021 0:00 $13,304.35 $6,638,873.00
YFI 3/29/2021 0:00 $16,924.98 $8,445,563.89
YFI 4/5/2021 0:00 $31,632.11 $15,784,422.60
YFI 4/12/2021 0:00 $43,074.73 $21,494,290.49
YFI 4/19/2021 0:00 $18,629.77 $9,296,254.60
$337,713.45 $169,551,424.21

My dune query was a fork of the Bancor Revenue and Earnings v2.1
query from the official Bancor dashboard but adapted to show weekly fees as opposed to daily.

2 Likes
  1. Does the action of extending a program allows to modify the max amount of BNT to allocate on each period? I think that there should be a more dynamic adjustment of allocations and the fees of the pool within each LM program. Why? Because at the end of the day, if we are not going to be the ones having the deepest markets (nor concentrated liquidity market), we should go into a “okey, we will capture all the arbitrage but at a true cost for arbers” mode.
    Particularly to this proposal, what I’m suggesting is to target a higher amount of YFI to be onboarded (capping the pool = 8M USD more worth of YFI) and potentially increasing the fees of the pair (maybe 0.05 or 0.1%)

LM rewards for the YFI side targets a 6% APY @ 24M USD worth of YFI

  • this translates to $1.44M worth of BNT per year (210k BNT at current rates)
  • this translates to $120k worth of BNT per month (17.5k BNT at current rates)

Increasing the fees of the pair (taking historical data provided by you)

  • Increasing +0.05% fees implies a total of $422,141.25 in fees which annualized would be $1.26M
  • Increasing +0.1% fees implies a total of $506,570.175 in fees which annualized would be $1.52M

This is just an example but it should be more thoroughly thought.

  1. Are these accounting for the ETH or STABLE side fee?

if not, then the amount of fees provided during that period is 1.5-2x for those pairs (depending on where it got routed from). This question is mainly related to my previous topic about the target % rate + increase in pair fees, and strengthens that side of the proposal (I’m supportive of making a more extensive proposal than just to re-enable LM rewards).

  1. It would be great to have a dashboard to also track IL generated by the protocol (and have it online) to provide less biased proposals/decisions into the ecosystem, and focus on what’s truly helping it grow.
  1. If that’s the case, then let’s move forward with a proposal to subsidize gas fees for swaps and LP additions because that’s the maximum pain point that a user has to go through whenever they have to interact with the protocol.

But we have seen with Balancer’s gas subsidies that this doesn’t help with anything in the long-term, and it just creates a fake sense of security for the platform’s tokenholders

For the long term, the best solution is to create a more gas efficient AMM, not a cheaper AMM (I don’t believe that the segmented liquidity that will go towards Arbitrum whenever Bancor launches there will be significant enough unless you guys support it with a ton of incentives)

Yes, that’s the whole point of incentivizing LM rewards on a pool, to incentivize other protocol’s Liquidity Providers to get to know the IL mechanism on Bancor. That’s why I commented that the timeline for the previous YFI’s LM rewards was really odd and just for marketing purposes, because there was 0 substance on incentivizing people without allowing to onboard more deposits:

sources:

3 Likes

I’m going to check in with the YFI community. I am interested to see if anyone is actually interested in staking more YFI. If there is some interest, I will flip my position to support.

I will reach out now and invite the YFI crew to comment.

3 Likes

I could be wrong but it doesn’t look like it. At least the latest extension to the SNX pool suggest that the function only takes the pool token and a new end time.

Yes, I think so as well but I am not sure what the backlog is at the moment on the developer side. I think making the contracts more gas efficient, porting to arbitrum, shadow tokens, and origin pools are the current focus. FYI, this is just speculation on my end so take this with a grain of salt (I have no insider information).

OK, I see what you are saying. Essentially what you are suggesting is for us to open up enough space to create a pool that’s $48M deep which would be roughly 3.5M in BNT co-investment (assuming BNT price of $7). This would open up $24M for the YFI side and if LM rewards are passed, then we should target 6% APY on the YFI side. What this means is that rather than emitting 120K-240K BNT in 12 weeks we would spread this out to meet the target (potentially 17.5K per month) 6% APY on the YFI side. We would also slightly raise the fees on the pool to potentially cover the rewards that are getting minted.

Essentially, the model that you are pitching is to grant pools LM rewards based on the fees that they are bringing to the protocol. From our end, we would emit less or more rewards based on the performance of the pool. Potentially, for YFI the proposal can be along the lines of granting another 12 week extension and emit 52.5K BNT (17.5K * 3 months) in this period. If at the end of the 12 week period we find that the fees are still there then we would evaluate and figure out how much BNT should be emitted for the next 12 weeks.

It’s hard to say, I would have to dissect the SQL query. My gut feeling is that this is the fees on the pool only so YFI-BNT swaps and vice versa. I think most people are probably swapping from stable coins or ETH in general and since that’s a multihop swap then my guess is that the fees from the first hop accumulate in those pools (ETH or stable pools) and the second hop in the respective destination pool (there are two conversion fees in this scenario that are payed).

I don’t think gas subsidies would get us much (I agree with your balancer assessment and making a more efficient AMM) and is probably better for us to put the effort into making the contracts more gas efficient (I think that’s where we will see the biggest gains and my understanding is that this is also the team’s focus). There is some good discussions going on in Trader incentives discussion that might be of interest to you as well as suggestions from the community on how we can grow the user base. As to the liquidity that might go to Arbitrum, I think it is too early to tell what will happen there. I think ultimately, whichever L2 solution gets the most tractions and projects there first might become the defacto winner and the Ethereum community will follow. It might be that Arbitrum (if they manage to launch before optimism) might have that first movers advantage and if we make the right bet and have Bancor running there before others then we would have the upper hand against other DEXes. Again, I think it is too early to tell.

I can’t really comment around the timeline of the previous LM campaign on the YFI pool. I was sitting on the sidelines back then and have only taken a more active role in the community recently (since I do love this community/project and wish for it to succeed.) Anyway, I am glad that you have taken the time to put up a good proposal for re-enabling LM rewards on the YFI pool since some of the others ones that I have seen lack in substance. It would be good for the YFI community to show their support and comment here if possible to show the Bancor community that they care about their pool.

1 Like