[Circle of Tokens] Liquidity mining: data based insights

Hello Sovryn’s,

We propose to use this thread to collect and discuss data-backed insights on liquidity mining. The aim is to create a collection of information that allows to generate actionable insights on the future direction of liquidity mining. This should help us to decide on questions that have been brought up, e.g., in
this thread.

To kick this off, let me start providing preliminary insights from graphql data. I took a look at the following questions and aim to provide an overview on the data without jumping to conclusions:

  • Who is providing liquidity in the protocol, are LPs mainly providing small amounts or do we have a few whales?
  • How did net liquidity evolve over time?
  • How much is the protocol paying for liquidity by pool?

Liquidity providers: fish or whale?

Key takeaways: liquidity providers

  • The number of users providing liquidity varies significantly by pool: BTC/SOV has by far the most users (9.8k), ETH/BTC (0.5k) is a distant second. The other 3 v1 pools (BNB/BTC, MYNT/BTC, XUSD/BTC) have roughly 0.2k users.
  • Whales are key liquidity providers for Sovryn: In all pools the mean liquidity provision is substantially larger than the median liquidity provision. For example, 50% of the LPs (4950 LPs) in the SOV/BTC pool provided less than 0.056 BTC each, while the largest LP provided 240BTC. This pattern is slightly more marked for the MYNT, BNB and ETH pools, as compared to the SOV and XUSD pools.

Overview on liquidity provider data

‘(WR)BTC/SOV’ pool

Total liquidity added (in BTC) per user:

count mean std min 25% 50% 75% max
9899 0.810 4.979 0.000 0.011 0.056 0.303 242.238

‘(WR)BTC/ETHs’ pool

Total liquidity added (in BTC) per user:

count mean std min 25% 50% 75% max
550.0000 0.8889 3.7350 0.0001 0.0232 0.0648 0.2699 54.6667

‘(WR)BTC/BNBs’ pool

Total liquidity added (in BTC) per user:

count mean std min 25% 50% 75% max
206 1.52901 5.77893 0.00020 0.02889 0.12845 0.49666 49.51670

‘(WR)BTC/MYNT’ pool

Total liquidity added (in BTC) per user:

count mean std min 25% 50% 75% max
203 0.7608 3.9793 0.0000 0.0048 0.0417 0.1661 49.9294

‘(WR)BTC/XUSD’ pool

Total liquidity added (in BTC) per user:

count mean std min 25% 50% 75% max
270 2.2394 7.2128 0.0000 0.0164 0.1140 0.9297 64.1756

Net provided liquidity

Key takeaways: net provided liquidity

  • Net total provided liquidity in BTC (all liquidity additions of all LPs minus all liquidity removals of all LPs per pool, expressed in BTC) is largest for the SOV/BTC pool (roughly 500BTC), followed by the ETH/BTC pool (roughly 400 BTC), XUSD/BTC (roughly 275 BTC), BNB/BTC (roughly 210 BTC), and MYNT/BTC (roughly 100 BTC).
  • Net total provided liquidity is falling since November 2021 for the ETH/BTC pool, the BNB/BTC pool and the XUSD/BTC pool. Net total provided liquidity in BTC for the MYNT/BTC and the SOV/BTC is roughly stable since 2022 (slight decrease for the SOV pool, slight increase for the MYNT pool).
  • Activity levels vary a lot: The BNB/BTC, ETH/BTC, XUSD/BTC and MYNT/BTC pools show several weeks (up to almost 2 months) in a row without liquidity removal/addition events.

Overview on net provided liquidity data

Cost of liquidity

Key takeaways: cost of liquidity

  • The weekly BTC rent the protocol pays per BTC of liquidity provided (BTC value of SOV reward/net provided liquidity in BTC) varies up to a factor of 4 between pools. As of May 2022 (in % of BTC):
    • (WR)BTC/ETHs pool: 0.28%
    • (WR)BTC/XUSD pool: 0.40%
    • (WR)BTC/SOV pool: 0.41%
    • (WR)BTC/BNBs pool: 0.78%
    • (WR)BTC/MYNT pool: 1.07%
  • “Short-term” trends (last 3 months):
    • The relative cost of liquidity in SOV is relatively stable for the SOV/BTC pool since the end of 2021.The relative cost of liquidity in BTC is decreasing slightly.
    • For the BNB/BTC and ETH/BTC pool the relative cost of liquidity paid in SOV increased substantially. The relative costs in BTC increased since March 2022 for the BNB pool, going hand in hand with a decrease in BNB net liquidity over that period.
  • “Long-term” trends:
    • The relative cost of liquidity in BTC went down in all pools; No uniform picture for the relative cost of liquidity paid in SOV.

Disclaimer:

  • LP reward data cannot (yet) be queried as they accrue by pool/user with the graph. The data here is based on weekly LP reward communications we collected on the sovryn blog. If anyone has accurate historical reward timeseries, please reach out.
  • LP fees (= forgone income of stakers, also a cost from the protocol’s point of view) has not been included in this assessment, the data can also not directly be queried by pool with graphql. Looking into this.

Overview on cost data

[Update:] Updated figures in this section on 13/05, correcting the SOV / BTC pool costs

Cost of liquidity: last 3 months

Cost of liquidity: all data

As proposed by @dseroy and @Sacro, an interesting next step would be to experiment with different reward parameters of a pool (size of rewards, vesting period of rewards) to get LP sensitivities. The data can also be looked at to decide whether it makes sense to stop rewards all together for some of the pools.

Happy to get feedback/inputs/questions/more data & big thanks to Betsy for giving access to the graph.

Stay Sovryn!

15 Likes

Great data and awesome write up @bananas_in_the_sky !!

Would be interesting to look at cost of liquidity vs. trade volume and protocol revenue.

Thanks for this great post! I’m not much of a data analyst, but I have some more general questions I hope you (Circle of Tokens) can answer.

  • How do you regard the overall state of the liquidity mining?

  • How do you view the very skewed distribution of liquidity between fish and whales? It seems to me to be quite vulnerable, but I don’t know how this distribution is within other platforms.

  • Is it/should it be a goal to incentivize more fish to provide liquidity or are more whales better?

I took an initial look at cost of liquidity vs trade volume.

Bang for buck

Key takeaways: cost per trade volume

  • Weekly cost in BTC per BTC of trade volume (BTC value of SOV reward/weekly trade volume in BTC) varies up to a factor of 36 between pools as of 2022-05-07:
    • (WR)BTC/XUSD: 0.007877 BTC
    • (WR)BTC/ETHs: 0.028068 BTC
    • (WR)BTC/SOV: 0.069628 BTC
    • (WR)BTC/BNBs: 0.115053 BTC
    • (WR)BTC/MYNT: 0.291259 BTC

**Disclaimer ** Same disclaimer as above: (i) sov reward data are not directly from graphql, I just took the promotion numbers from the ddap and accounted for the change from 15k to 30k per week for the SOV/BTC pool; (ii) fees are not (yet) part of the cost estimate.

4 Likes

Adding complementary analysis. Below is a table showing the fees generated per month by each swap pair. ConversionFee is the swap fee (0.25%) paid to LP’s and protocolFee is the swap fee (0.05%) paid to Bitocracy stakers.

So for example, in April of 2022 swapping from wrBTC > XUSD earned Bitocracy $2,725.29.

The take-away is none of these swaps generate all that much revenue whatsoever, but in particular tokens like BNBs, MYNT (and arguably ETHs) are especially immaterial. Throwing 15K SOV a week at these tokens which probably aren’t part of our long-term strategic vision is a bit non-sensical. As such I think it makes sense to drastically reduce and remove these LM rewards.

month/year _fromToken.symbol _toToken.symbol Sum of _conversionFee $ Sum of _protocolFee $ Sum of Volume
01/2022 BITP WRBTC $18.14 $- $18,137.48
BNBs WRBTC $1,105.46 $221.09 $442,185.99
DOC WRBTC $114.07 $- $96,387.29
ETHs WRBTC $1,660.63 $332.13 $664,251.84
FISH WRBTC $136.07 $27.21 $54,426.17
MOC WRBTC $10.11 $2.02 $4,043.62
MYNT WRBTC $951.42 $190.28 $380,567.59
RIF WRBTC $788.86 $157.77 $315,544.75
rUSDT WRBTC $1,374.86 $- $1,374,856.73
SOV WRBTC $4,535.09 $907.02 $1,814,035.78
WRBTC XUSD $9,137.39 $1,827.48 $3,654,956.22
SOV $2,642.74 $528.55 $1,057,094.96
ETHs $2,159.73 $431.95 $863,891.53
rUSDT $1,397.84 $- $1,397,836.32
BNBs $853.28 $170.66 $341,311.27
RIF $843.97 $168.79 $337,586.19
MYNT $583.31 $116.66 $233,324.72
DOC $107.01 $- $95,145.61
BITP $14.22 $- $14,222.87
MOC $7.90 $1.58 $3,161.19
FISH $7.39 $1.48 $2,957.67
XUSD WRBTC $9,370.29 $1,874.06 $3,748,115.10
01/2022 Total $37,819.77 $6,958.73 $16,914,040.88
02/2022 BITP WRBTC $167.41 $- $167,408.91
BNBs WRBTC $3,519.97 $703.99 $1,407,988.81
DOC WRBTC $273.75 $- $243,571.98
ETHs WRBTC $7,618.74 $1,523.75 $3,047,495.19
FISH WRBTC $257.49 $51.50 $102,996.08
MOC WRBTC $131.29 $26.26 $52,516.69
MYNT WRBTC $1,418.76 $283.75 $567,504.19
RIF WRBTC $1,315.43 $263.09 $526,171.95
rUSDT WRBTC $4,362.19 $- $4,362,186.87
SOV WRBTC $9,566.91 $1,913.38 $3,826,765.08
WRBTC XUSD $32,130.45 $6,426.09 $12,852,178.59
ETHs $7,271.95 $1,454.39 $2,908,778.32
SOV $7,111.83 $1,422.37 $2,844,733.47
rUSDT $4,370.23 $- $4,370,231.32
BNBs $3,101.06 $620.21 $1,240,425.14
RIF $1,390.28 $278.06 $556,112.39
MYNT $841.55 $168.31 $336,620.24
DOC $274.76 $- $246,649.00
BITP $168.79 $- $168,787.24
MOC $123.84 $24.77 $49,536.77
FISH $35.40 $7.08 $14,158.34
XUSD WRBTC $32,922.58 $6,584.52 $13,169,031.25
02/2022 Total $118,374.66 $21,751.51 $53,061,847.81
03/2022 BITP WRBTC $83.52 $- $83,521.45
BNBs WRBTC $3,879.47 $775.89 $1,551,786.96
DOC WRBTC $269.29 $- $261,028.76
ETHs WRBTC $4,235.82 $847.16 $1,694,329.90
FISH WRBTC $365.22 $73.04 $146,088.53
MOC WRBTC $165.75 $33.15 $66,301.72
MYNT WRBTC $1,347.07 $269.41 $538,827.45
RIF WRBTC $812.98 $162.60 $325,190.67
rUSDT WRBTC $4,472.86 $- $4,472,859.28
SOV WRBTC $9,238.23 $1,847.65 $3,695,290.89
WRBTC XUSD $28,334.62 $5,666.92 $11,333,848.43
SOV $7,420.64 $1,484.13 $2,968,254.81
ETHs $4,870.49 $974.10 $1,948,197.46
rUSDT $4,456.27 $- $4,456,274.88
BNBs $3,982.01 $796.40 $1,592,804.43
MYNT $881.43 $176.29 $352,573.22
RIF $675.38 $135.08 $270,153.41
DOC $263.39 $- $258,506.24
MOC $184.92 $36.98 $73,967.62
BITP $81.79 $- $81,788.86
FISH $14.21 $2.84 $5,683.82
XUSD WRBTC $28,847.97 $5,769.59 $11,539,189.80
03/2022 Total $104,883.34 $19,051.24 $47,716,468.57
04/2022 BITP WRBTC $84.17 $- $84,168.04
BNBs WRBTC $975.79 $195.16 $390,315.62
DOC WRBTC $203.97 $- $202,026.64
ETHs WRBTC $4,950.29 $990.06 $1,980,116.30
FISH WRBTC $144.69 $28.94 $57,874.36
MOC WRBTC $88.50 $17.70 $35,399.20
MYNT WRBTC $749.79 $149.96 $299,915.44
RIF WRBTC $441.33 $88.27 $176,530.85
rUSDT WRBTC $2,649.28 $- $2,649,278.72
SOV WRBTC $5,484.85 $1,096.97 $2,193,940.19
WRBTC XUSD $13,626.47 $2,725.29 $5,450,586.70
SOV $6,714.49 $1,342.90 $2,685,795.01
ETHs $5,231.97 $1,046.39 $2,092,787.00
rUSDT $2,666.61 $- $2,666,605.10
BNBs $1,379.12 $275.82 $551,648.58
MYNT $1,026.00 $205.20 $410,399.22
RIF $391.99 $78.40 $156,797.09
DOC $212.02 $- $201,065.43
MOC $99.77 $19.95 $39,908.89
BITP $80.03 $- $80,031.26
FISH $45.71 $9.14 $18,282.33
XUSD WRBTC $12,656.63 $2,531.33 $5,062,653.25
04/2022 Total $59,903.45 $10,801.48 $27,486,125.21
Grand Total $320,981.22 $58,562.95 $145,178,482.48
4 Likes

Adding more data. Here is a chart of historical APY yields for each Yield Farming pool, but adjusted for Impermanent Loss. If anyone wants to see the weekly IL for each pool, I’m happy to post that data as well. Not a huge surprise but large amounts of IL for SOV and Mynt pools, but those also have the highest APY from LM rewards so they mostly off-set.

3 Likes

I’m a LP in AMM pools, I’ve kept a significant amount over the past year. I allocate about 70% to SOV/btc, and about 30% to eth, mynt and bnb. The impermanent loss on the SOV/BTC has been brutal and the stability of the other pools have allowed me to remain (to some degree) invested in the SOV AMM, Ive watched it dwindle as the price has gone down. All of my generated SOV is staked/reinvested, I try to make up some of the loss and add to my stake… about 50/50. but either way I keep it in the protocol. My point/question is that in my experience it’s tough to just be in the SOV/BTC amm w the terrible impermanent losses (i hope it turns around which makes this exercise easy) and the secondary benefit of the other pools is stability and reinvestment of those SOV produced. I think I’d have to pull out some liquidity without the other pools and look elsewhere to offset like many did/do as the price drops. Curious what others think of the value of stability and LP retention vs the cost of SOV to the protocol, i think it is more complex than simply reducing outflow. In my case it would be losing reinvestment of that outflow and overall investment/liquidity in the system. I’m ok with whatever way it goes but worth a discussion possibly.

4 Likes

I had started writing before dseroy posted this great chart but it looks like slight gain in the BNB and ETH, big loss in SOV, mynt which has been my experience. good stuff, interesting.

I know this question is collateral, but on the assumption that the Adoption Fund will see its volume reduced by six million sov (because of the new round of investment), could this have any consequences on the analyses you are doing?

Bitocracy can decide how to spend the Adoption Fund. Then I wonder: Is there anything to prevent the Bitocracy from doing several rounds of Investment in the future that would further reduce the Adoption Fund?

@bananas_in_the_sky @Sacro @dseroy

I’m not sure I understand the question. No I don’t see it impacting what we’re doing much. If anything it locks up a ton of tokens which was something a lot of people wanted to see changed with the tokenomics.

Sorry, I disagree.

I think that if there are new rounds of investment against the Adoption Fund in the future, it could have consequences for the protocol if the Adoption Fund is reduced again. And maybe it could be measured.

I think there is nothing to prevent it, if it does not prevent it now. Or can we think that this is the last investment round after the programmatic sales?

I don’t know how else to ask the question.

The thing that prevents it now and into the future is the Bitocracy. This question is probably better suited in the GC strategic investment thread.

While I sympathize with your point, I think it’s misaligned with the reality of the situation. We don’t really have the luxury to forgo a capital raise because of our ‘rough’ outline of the Tokenomics created over a year ago. If we’re on the precipice of a sustained bear market we need to make sure we are capitalized. The treasury is healthy but it’s also not flooded with cash either. We need to ensure long-term survival. Those tokenomics categories were created very early on in the project and were basically just budget estimates. We now have a more clear picture of our financial situation, the market, upcoming costs, etc. It’s clear cut to me that we should take tokens allocated from the adoption for this raise.

I think if Exchequer unilaterally took from the adoption fund without informing community or putting to a vote it would be an issue. But putting it up for a vote and providing new information, like the 2022 budget projections, which clearly show our financial situation is acceptable. Things change and we must adapt.

2 Likes

Thank you David,

I think it goes without saying that I love Sovryn. And that my way of expressing it is by asking questions like the one I have asked.

As for exposing it in the other thread, it makes no sense, because this kind of questions should be answered by Sovryn people as you are doing in a timely manner. They are not questions I should be asking General_Catalyst.

I have confidence in the bitocratic system. I have suggested improvements in the past, which I know are on the agenda. In the meantime, I always ask for transparency. That is why I consider your response sufficient, and I appreciate it.

Stay Sovryn

3 Likes

Just a small clarification, I dont know how things are coded, however if I just go with whats written in the forum, the Exchequer cannot “unilaterally” take from the adoption fund without involving bitocracy.

For the development fund, team multisig has complete control so that could be utilized without involving the bitocracy.

I don’t think this is true. SIP34 here approved transfer of Sovryn Improvement Proposals | S O V R Y N of 2.2M and 9.9M from Development and Adoption funds respectively into the Exchequer wallet. It’s my understanding that up to 6.5M of those SOV will be used for the GC round (assuming it passes).

Here’s the onchain; RSK explorer

It was not clear to me which bucket of SOV will be used for this sale, if its the SOV already given to the Exchequer through SIP34 then you are right, the bitocracy is not needed in that case.

Does this take into account SOV subsides? It seems to me we should be targeting something more akin to 10% after IL - especially with SOV price being where it is.

Keinor, on the basis of the chart posted by @dseroy above, the impermanent loss has been more than completely offset by SOV rewards. Your post seems to disagree with this?

Sorry to intrude. I invested 0,25865404 rBTC. Impermanent Loss equals 0,06215404 rBTC. Rewards equal 0,02025881 rBTC.
Rewards make up 32,54% of Impermanent Loss. 67,46% of stated Impermanent Loss is still there. 24,03% of my initial investment is currently Impermanent Loss.
Taking the rewards into the calculation there is still 16,20% of my initial investment in Impermanent Loss. Don´t know how anybody could have offset Impermanent Loss with rewards at this very moment.

1 Like

Investigating. Calculations could be off let me circle back.

Update @AllRoadsLeadToSovryn which pool was this for and over what time-frame? The issue may be that my analysis is performed weekly and considers the net IL impact according to the $SOV price that historical point in time every week.

Specifically, I broke down the analysis into weekly chunks. So, I would say over a week time-frame how much IL was incurred in X pool, extrapolate it to a year; then how much rewards were accrued as earned in $SOV, extrapolate it to a year, then net out the difference to get the net yield. The key assumption here is the LP would be able to sell their rewards to off-set the IL instantly. This obviously isn’t how it operates in reality because of the 10 month vesting, but I chose to do it this way because if we’re trying to adjust rewards to compensate for IL then we have no way of projecting out into the future 10 months what $SOV price will be. We have to look at it each week and say what is the expected IL this week and how much rewards in the present day price can we compensate to try and off-set.

Here is a link to the spreadsheet which calculates the IL and yield: Liquidity Mining Analysis - IL - Google Sheets

The spreadsheet takes a weekly look at the pools since April 2021. I sourced the average balance of each token in the pool for the respective week, calculated the average price of each token during that time-frame, then re-created the weekly IL incurred vs. the APY earned in rewards (specifically at the $SOV price during that week)

Here is the formula to calculate IL; where baseQTY is amount of rBTC in the pool, tokenQty is the amount of other token in the pool, futurePriceRatio is token price ratio over the week time-frame.

Constant = baseQty * tokenQty;
hodlStrategy = ((tokenQty * futurePriceRatio) + baseQty);
lpStrategy = ((Math.sqrt(productConstant / futurePriceRatio)) * futurePriceRatio) + (Math.sqrt(productConstant * futurePriceRatio));
impermanentLoss = (hodlStrategy - lpStrategy) / hodlStrategy * 100;