Drastically increase BALN-sICX rewards

First this is a proposal that only makes sense with multi pool routing, so please think of it with that frame of reference in mind.

On PancakeSwap, a successfull inflationary platform token based DEX, their reward for CAKE-BNB is by far the largest, by a gigantic factor. If you examine their pools in general, many are CAKE-X or BNB-X, with the stablecoin-token pairs also taking a moderate chunk.

They can do this because routing kind of merges more of the TVL into a single blob (while generating more fees).

As a side benefit as well, more platform token stays locked in pools.

But why should the pool with least IL get the most rewards?

Well that is looking at it from the wrong perspective, all DEXs pay for liquidity with rewards, the pool with the lowest APR is the best (for the platform) because its ‘buying’ the most TLV with the least amount of rewards. As the pool with the lowest rewards, its currently the best ‘bang for buck’ for Balanced. Increasing the rewards does not ‘waste’ the reward, it turns into larger TLV as yield seeking is an inherently very efficient process.

With BALN-sICX as the central big liquidity pool ala CAKE-BNB, other pairs will just need to pair off one of them.
This is even more apparent for the sICX-bnUSD pool which needs less liquidity to get a good price for DEX users because of pegging. Not saying it needs to be tiny it should be sizable, but we should look to working examples like Pancake and Quickswap to figure out a nice ratio.

CAKE-BNB is 40x, BNB-BUSD is 8x, BNB-USDT is 4x. Adding the stable pairs together their platform-native pair is 3.3 times their native-stable pair rewards.
You might be interested to learn CAKE-USDT and CAKE-BUSD only add up to 4x, the same analogy for BALN would be 10 times the rewards for BALN-sICX to BALN-bnUSD + BALN-stable combined.
This matches up with the BALN-bnUSD rewards being highest, ie. the most expensive liquidity for Balanced to pay for.

Quickswaps farm page is a ugly mess and the search sucks, so I gave up sorry.

Pros that I can see:

  • Cheaper TLV for platform ~ more TLV ~ better slippage stats
  • More BALN locked in LPs
  • BALN-sICX is a more ‘long’ stance which I believe is good to encourage versus X-bnUSD which is a more bearish stance.
  • less IL for people that blindly throw their money at pools that clearly do not align with their investment philosophy because it has the largest number. I am divided on this, it may not be a pro.

Cons:

  • Rebalancing will happen a bit more often, but with a lower amount, as a smaller sICX-bnUSD pool will take less to move both ways. (if this is a significant problem sICX-bnUSD can be special and kept higher than industry average)
  • If there isn’t enough BALN to fufil demand for the higher rewards in the LP, BALN price will increase. (Is this even a Con?! Fake Con)

Addendum: For those unfamiliar with multipliers, you can just this of them as points and the total rewards allocated to all farms is divided to all farms in the ratio of their multipliers to each other.

3 Likes

This seems to be a good proposal. I think the biggest advantage would be that more of the native tokens are locked in the pool. I would have to study it a little more and do a little more research.

1 Like

You make a good case here. I do agree with some of the points you’ve made.

However comparing Balanced to other dexs isn’t comparing apples to apples.

The sICX/bnUSD rewards are vital to the use case and utility of bnUSD. Atleast for now.

The other platform you mentioned don’t have the ability to mint their own stable coin.

I think it’s vital to support our own stable coin as much as possible.

Also I do strongly agree with supporting sICX/BALN pair. They are the life blood of the Balnced platform.

Over time as new pairs are added to Balnced, we need to prioritise our main pairs. (Same as pancakeswap has done).

The sICX/ICX is a great example of low rewards, which will be great for get something similar with new pair coming iUSD/bnUSD.

For reference the CAKE/BNB pair is currently around 50% APY. And the sICX/BALN is also around the 50% APY.

The pools we currently have running. They need to remain our legacy pools. Which we need to support.

However if anything we need to take a small amount of rewards away from them. To accommodate the new pair with rewards.

I do not believe bnUSD needs supporting to that level.
It already has a use case, you can buy sICX and BALN with it. I am not proposing removing rewards for them, just a reduction.
Addtionally as stables go, their ultility is ‘is pegged to the USD’, as long as that remains true, it will have a use.

I don’t have hard numbers for this but anecdotally just from looking at it everyday, bnUSD spends more time over 1 USD than below 1 USD. This indicates to me that there is likely enough demand for bnUSD.

Apologises for multipost, but I feel I need to stress this, I believe increasing rewards for BALN-sICX will not increase the rewards APR, but instead increase TLV to match the new rewards. Additionally this increase in TLV will be larger than the TLV loss by reducing it in another pool, this is the base argument I have.

The pool with the lowest APR will give us the most TLV per reward given while there is still demand. And I believe there is still much demand for BALN-sICX that we have not tapped into.

I think we will have to agree to disagree on this one arch.

1 Like

Total daily rewards for LP: 37199.02 BALN
Approximate TVL: 57,060,000 sICX

sICX/ICX

  • APY: 12.4%
  • Total supply:
    36,703,230.84 ICX
  • Total daily rewards:
    7,439.8 BALN
    20%
  • Approximate TVL:
    35,500,000 sICX
    62%

sICX/bnUSD

  • APY: 72.2%
  • Total supply:
    5,342,109.47 sICX
    7,682,905.11 bnUSD
  • Total daily rewards:
    13,019.66 BALN
    35%
  • Approximate TVL:
    10,700,000 sICX
    19%

BALN/bnUSD

  • APY: 110.64%
  • Total supply:
    2,144,819.79 BALN
    5,017,238.69 bnUSD
  • Total daily rewards:
    13,019.66 BALN
    35%
  • Approximate TVL:
    6,980,000 sICX
    12%

BALN/sICX

  • APY: 56.49%
  • Total supply:
    1,204,420.97 BALN
    1,942,814.09 sICX
  • Total daily rewards:
    3,719.9 BALN
    10%
  • Approximate TVL:
    3,880,000 sICX
    7%

I posted data on how the rewards structure currently sits (above). I feel this may be handy to facilitate discussion.

Reminder, the sICX/BALN pair was the first new addition to the DEX, and it’s low reward allocation was probably an effort to minimize the disruption to the original LP pair’s rewards. It could also be argued the potential for less impermanent loss risk may have had a hand in the decision.

Also: I feel new DEX pairs are not important to this conversation, or maybe they are, but more so when they happen. Currently, how will new pairs be incentives? Is OMM incentivizing the OMM pairs that will be available soon?

To sum up, the BALN/sICX pair has

  • The lowest TVL with 7% of the LP TVL
  • The lowest daily rewards with 10% of LP rewards
  • The second lowest APY at about 56%

Here I’ll go over one way of making this change, and how the differing values of the changes can effect the other pools.

In these examples I keep the sICX/ICX pair’s rewards as is. This leaves the 2 bnUSD pairs, with a combined 70% of the daily rewards, as the options to adjust in boosting the BALN/sICX pair’s rewards. I changed both pairs evenly, but that doesn’t need to be the case. These are not the only options, but it does give a general idea of the relationship between rewards, TLV, and APY.

Option A:
Decrease the 2 bnUSD pairs down to 30% of LP rewards from 35% (each), and increase BALN/sICX pair’s rewards up to 20% of LP rewards from 10%.
This is how that change would effect APY:

BALN/sICX 56% to 112%, +56%
BALN/bnUSD 110% to 95%, -15%
sICX/bnUSD 72% to 62%, -10%

Option B:
Decrease the 2 bnUSD pairs down to 32.5% of LP rewards from 35% (each), and increase BALN/sICX pair’s rewards up to 15% of LP rewards from 10%.
This is how that change would effect APY:

BALN/sICX 56% to 84%, +28%
BALN/bnUSD 110% to 102.5%, -7.5%
sICX/bnUSD 72% to 67%, -5%

Option C:
Decrease the 2 bnUSD pairs down to 34% of LP rewards from 35% (each), and increase BALN/sICX pair’s rewards up to 12% of LP rewards from 10%.
This is how that change would effect APY:

BALN/sICX 56% to 67.2%, +11.2%
BALN/bnUSD 110% to 107%, -3%
sICX/bnUSD 72% to 70%, -2%

I don’t know how I feel about this proposal. My concerns are around how this change will effect the liquidity for bnUSD pools and how that will effect price movement and rebalancing.

Just wanted to add some of the numbers I’ve been looking up, they are a lot rougher than yours, so anyone wanting to rely on them, please check their validity with the respective platform.

In addition to the PancakeSwap numbers in my initial post, here are some more:

Bancor - in APR as BNT
Bancor lists the APR for BNT seperate from the trading fees, so this is what I will use
ETH/BNT - 55% - TLV - 495 million - 1 year of rewards - 272m
USDC/BNT - 41% - TLV - 46 million - 1 year of rewards - 18.8m
USDT/BNT - 52% - TLV - 37 million - 1 year of rewards - 19.2m
DAI/BNT - 47% - TLV - 36 million - 1 year of rewards - 16.9m
Bancor does not have ETH/stable pairs, in fact those 3 are the only stablecoin pairs it has.

This gives a rough reward rate ratio between ETH/BNT and BNT/stables as 272:55 ~ 4.9
BALN/sICX to BALN/bnUSD ratio = 0.2856

SushiSwap - Reward in Sushi/day
SUSHI/WETH - 14604 - TLV - 515m - APY - 14.18%
DAI/WETH - 6748 - TLV - 245m - APY - 13.76%
USDC/WETH - 9127 - TLV - 460m - APY - 9.93%
WETH/USDT - 7302 - TLV - 268m - APY - 13.61%
It does not have a SUSHI/stable farm I could find.
Total ETH/stable rewards - 23177

Ratio of reward SUSHI/ETH : ETH/stable = 0.63
Ratio of reward BALN/sICX : sICX/bnUSD = 0.2857

ApeSwap

BANANA-BNB - 40x
BANANA-BUSD - 15x
BNB-BUSD - 4x
Please don’t stone me for this entry, it is 11th on this list, Top Automated Market Maker (AMM) Coins by Market Capitalization - CoinGecko, just behind QuickSwap

QuickSwap - in QUICK/day
The search doesn’t work for me, so I will just throw out some pairs I see
ETH-USDC - 56 - TLV - 131m - APR - 24%
ETH-USDT - 35
DAI-ETH - 23
ETH-MATIC - 50 - TLV - 98m - APR - 31%
ETH-QUICK - 22
MATIC-USDC - 12
MATIC-QUICK - 35
USDC-QUICK - 14 - TLV 8.4m - APR - 135.8%
USDC-USDT - 12
DAI-USDC - 5

Yes thats right, they are incentivising stable/stable more than Quick/stable.
I didn’t transcibe all the values but I hope this also demonstrates a rough idea.

Extra Notes/Opinions
Uniswap currently doesn’t offer liquidity incentives.

Smaller sized ‘satellite’ pools around a central large ‘main’ pool appears to be an extemely common design. Technically, I believe any pair can serve as that ‘bulk’ of liquidity, but selecting the one with easiest to access liquidity and largest extra beneficial side effects (like locking platofrm tokens) seems the best idea.

Closing
I hope this gives a rough indication of all the top DEXes following the ‘give larger total reward amount to pairs that will end up with lower reward rates, and lower total reward amount to pairs that will end up with higher reward rates’. The rates, or APR of rewards, directly represents the market opinion on the desireability of providing liquidity for that pair.

Ie. Let’s attract the most liquidity with the least amount of rewards.

ETH pair/ SICX will we need to incentive as well? let say we introduce it by next month align with ICE.

Another data set that might be important to note is supply/allocation of BALN.

  • Current total supply: 11,020,382 Baln
  • Current staked supply: 4,290,600 Baln 39%
  • Current Baln/bnUSD Pool: 2,137,195 Baln 19%
  • Current Baln/ICX Pool: 1,208,633 Baln 11%

@arch thanks for taking the time to write this proposal, but it’s going to be hard to come to common ground because I disagree with your core thesis here as I mentioned in this tweet. Having said that, we spoke an length on Discord, you made some good points, but ultimately I am sticking with the logic outlined in my tweet and below.

LP rewards, in aggregate, are paying for liquidity, that I will certainly agree with. But from there, we need to think about who needs to get paid the most. Personally, I am comfortable holding both BALN and sICX, I don’t have a strong opinion as to which one I hold more of at any given time, and I’d be open to LPing the BALN/sICX with hardly any incentive at all, just fees.

The main thing to think about, as I mentioned in the tweet, is correlation between two assets. Stablecoin-stablecoin pools are another example of pools that don’t necessarily need a strong incentive, just enough to compete with yields elsewhere. I think of LP rewards very programmatically:

High correlation = low rewards
Low correlation = high rewards

Then you also need to consider opportunity cost, which is where the rewards for the ICX-only pool.

1 Like

I 100% fully agree with this, as does the market, the reward rate for the high correlation pair are lower than the reward rate for the low correlation pair.

However the rate is determined by the market, not the nomimal total rewarded amount. We reward quadruple to the low correlation pair, yet the reward rate is merely double. In fact the reward reward rate has little to do with the total reward, and everything to do with how the market values supplying liquidity to that pair (risk/goals/correlation/access to other liquidity markets/etc.)

A reminder that doubling the total nominal BALN/sICX reward pool, taking it all from BALN/bnUSd represents only a fall of 25% of total reward pool, and should be a net gain of 2m TVL, represent more locked BALN, more TVL on DEFI marketing lists.
Additionally after a shift like that the relationship outlined above still holds true, the low correlation will STILL be higher rwards. Noone has yet outlined why this current exact ratio is the goldilocks ratio, and no improvements may be had to increase TVL.

First, I see no evidence that this change would increase TVL. Where are you getting a 2m TVL increase from? Also, why would more TVL in sICX/BALN pool be beneficial to Balanced?

Second, by prioritizing the sICX/BALN pool, we would be diminishing the attractiveness of our core service/product, minting bnUSD.

Having huge TVL in the sICX/BALN pool is horrible for balanced. We should want that sICX locked as collateral.

The bnUSD pools encourage locking collateral and minting bnUSD.

The sICX/BALN pool encourages people to withdraw (or not deposit) collateral.

At risk of repeating myself, the ‘reward rate’ of a pair, the APR, is independant of the ‘reward allocation’ of the pair.
The ‘price of liquidity’ is directly represented by the reward rate, the APR of a pool. The market decides the price, by supplying liquidity. We decide, ‘how much liquidity should we buy’ via the total allocation.

At the time of writing, the market has indicated, directly TOLD us, that the price of liquidity in BALN/sICX is half that of BALN/bnUSD, and also a bit cheaper than the price of liquidity in sICX/bnUSD.
Not only is it displayed on our own platform, as the data I gathered before has shown, there can be HUGE vast differences in the price of liquidty, as shown on the stats from ALL the DEXs in the top 11 DEFI platforms.

Stable/stables have the least reward rates, and lo and behold, if you calculated the liquidity(TVL) per reward you would realise, they are the cheapest, followed by major pairs, followed by speculative token pairs.

The price of liquidity is NOT set by the purchaser(Balanced). It is set by liquidity suppliers, the AMOUNT of liquidity purchased, that is decided by the purchaser(Balanced) and is represented by the reward allocation.

The liquidity market is not a supermarket

Did that make sense? Don’t get me wrong, I love the sICX/BALN pool. Increasing the rewards in that pool would make my position that much sweeter.

Right now I have about:
30% collateral
52.5% sICX/BALN
25% BnUSD/BALN (about 7.5% is loan)

If the changes here happen, I would probably rebalance to:
12% collateral
90% sICX/BALN (about 2% from loan)

You are one participant in the market, and you have your goals/investment thesis/bullishness or bearishness on BALN/sICX/and indeed bnUSD.

The total supplied liquidty per reward, is a reflection of the agregate willingness of the entire market to supply liquidity at that price. There WILL be variations, but the market decides the price. This price can of course change, with a change in the sentiment of the market. But long term trends in the pools in both Balanced, and other platforms are clear, lower IL pools get more liquidty per reward.

This isn’t as obvious in a closed ecosystem like ICX, but on chains where the fight for liquidity is fierce, liquidity shoppers are very discerning, and vaults allow them to have almost no exposure to the underlying reward token (for pairs not including that token)

EDIT: We as individual investors tend to view an increase in total reward as a increase in reward rate, as we are fully invested. But the reality is there exist vast amounts of liquidity not yet present, and yes part of my initial argument also contained the facet where if we believe the market prices liquidity in BALN/sICX at a cheaper price, if we increase the rewards on that pool, for lack of BALN (if there is a lack) it will cause the price of BALN to increase instead. And then find an equilibirum that roughly matches todays ratios. Or it can slowly absorb a larger share of inflation, reducing the price impact of inflation. Whatever the equalising mechanisms are, the markets price on liquidity of that pair, is both clearly known, and has been stable for a long time

I am not fluent in finance, my apologies. Is this statement sort of the point:

Increasing rewards to sICX/BALN will be the most efficient way to gain higher TVL (more TVL for less rewards paid).

My point is, I don’t think doubling the reward will bring substantial new capital to Balanced or ICON. I believe this change will just move the current capital around.

Further, I believe this change prioritizes a secondary product at the expense of our primary product.