[Discussion] Uniswap Liquidity Incentive Plan

This seems reasonable to me, I would still provide liquidity. But I doubt it make much difference. Usually I will put liquidity in a pool long term cause I choose assets I trust and my gains grow as longer I stay within the pool. If one choose to leave a pool it’s either because the money is needed for something else (e.g. liquidity mining elswhere;-)) or trust in one of the assets is gone, especially in the latter case I wouldn’t stay whatever the panelty is and in the first case it has to be nearly the same as the proposed incentive elsewhere, which seems unlikely for fees accumulated in less than a month
Anyway it seems worth thinking about, if it’s technically not too hard to implement. I would appreciate to see some rewards to the uni protocol (or the remainig pool participents) even if it doesn’t prevent pool hopping.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

The simplest, most equitable criteria for liquidity incentives is that the same amount of money is paid per $1M of liquidity in each rewarded pool. There are a number of ways to measure this, whether it be via the relative amounts of slippage or impermanent loss or strictly by the amount of UNI paid out in dollar terms for liquidity. The latter term is more intuitive and will be easier to consider for a heuristic Snapshot vote, while the former is more accurate but requires more data processing and cleaning. For this vote, I suggest we use the latter and do the following:

  1. Consider the entire UNI budget spent right now (54 UNI/block, since right now we have 13.5 UNI / block / pool) and divide it by 2 (27 UNI / block)
  2. Note that USDT, USDC tended to stay near 25% of the budget spent (which meets our criteria), so their relative fraction should be unchanged (6.75 UNI / block)
  3. For the latter two pools, note that the ratio of the Dai cost to WBTC cost averaged 1.76x with a standard deviation of 0.38. We can roughly say that if we compensate WBTC 2x what the Dai pool receives, then we should equilibrate to having Dai / WBTC have the same cost, so we give WBTC 9 UNI / block and DAI 4.5 UNI block
    -----BEGIN PGP SIGNATURE-----
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=axi/
-----END PGP SIGNATURE-----
uni_frac_spent

4 Likes

Nice math, I suppose it would help to maintain the btc surge :wink:, but not sure if we need it to maintain uniswap functionallity. Although liquidity droped shortly after the incentice ended, it stabilized now and trading volume is little changed. I don’t see the need for a new incentive programm right now, but however this might change in the future.
As a customer of uniswap (lp and trading) I need enough liquidity to allow for low slippage (right now it seems enough, still more than competitors), pools of all the tokens I like to trade (uni is superior to all competitors) for a one stop get all you want experience and low gas fees (not higher on uni compared to others on mainnet). As a lp I need enough trading volume to justify the risk of impermanent loss (also superior on uni). If we are seeing higher trading volume in relation to liquidity volume at competitors or if we see a drain in pools, these would be the criteria for me to think about new incentives.

edit: also if liqudity fell so much that slipage for the most traded pairs increase for the avarege trade size would be a reason for incentive, haven’t figured out the numbers, but this is ofc something worth to look at to me.

1 Like

Favoring WBTC makes a lot of sense to me. To reiterate what this change suggests:

  • WBTC/ETH - 9 UNI / block
  • USDC/ETH - 6.75 UNI/block
  • USDT/ETH - 6.75 UNI/block
  • DAI/ETH - 4.5 UNI/block

Over a longer-time frame this breaks down to roughly:

  • WBTC/ETH - 1.67M UNI/mo
  • USDC/ETH - 1.25M UNI/mo
  • USDT/ETH - 1.25M UNI/mo
  • DAI/ ETH - 0.83M UNI/mo

I am in favor of this distribution!

1 Like

what is the reasoning to favor wbtc, just because btc has the highest market cap?

1 Like

In other words, you suggest bringing the UNI distribution APYs for participating pools closer to each other.

This would end up increasing the disparity in liquidity between the pools.

As it is a change to the status quo, I believe it needs more justification.

This criterion doesn’t make much sense to me.

There are diminishing returns on every new $1M of liquidity.

$1M worth of liquidity in a $50M pool is much more valuable than $1M worth of liquidity in a $900M pool.

However, if we find the diminishing returns criterion as the most crucial, we shouldn’t continue the liquidity mining program for the proposed pools. And start rewarding other pairs instead.

I believe there should also be a criterion in place that we use to measure overincentivization .

For example, spending all the treasury money on the WBTC-ETH pair doesn’t look right, even if it would bring larger liquidity to the platform.

Maybe we could explore the correlation of Volume and Liquidity among the pairs?

In my opinion, if we have an inverse correlation between Liquidity and Volume, it is a clear sign that the pair is heavily overincentivized.
And WBTC-ETH pair is the most overincentivized pair in the initial program.

==

There was a question in the survey about restructuring the existing pools based on APYs, and here’s no clear preference there:

restructuring existing pools

==

We could state that we want to reward the Bitcoin community with the same APY we reward the other top-4 pools. Or that we want to bring more whale trading on WBTC on Uniswap.
But that would be political statements. I can’t detect a valid quantitative argument yet, though.

If we think about the slippage, the slippage below <0.01% is not even displayed in the UI. And 0.01% slippage already looks quite negligible given the fact that users pay 0.3% fee on top of it.

6 Likes

Thanks for sharing here @Mr_Po - this is the first I had seen the survey and it definitely shines some light on where forum participants minds are at.

I especially found this line to be interesting.

Still, seeing as the other three pools are all stablecoins, you could almost view the rewards as BTC & stablecoins, rather than WBTC, DAI, USDC and USDT.

I personally don’t feel too strongly about how the rewards are distributed among these pools, so long as we can agree to only include (at most) these four pools for the next chapter as we consider other pools and incentive structures.

As a small LP, I don’t really want to move my liquidity around all the time(gas fees)

Is there a technical reason that regular LP pool fees(from the 0.3 fee on all transactions) aren’t resumed on my liquidity when the UNI rewards end?

I’d love to just leave my liquidity on Uniswap permanently, collect UNI when they’re being given out, but then automatically collect regular fees when they’re not. As it stands right now I would have to move it manually, which on my scale makes it essentially cost prohibitive.

I’m fairly new to this but am I missing something?

1 Like

I have to oppose, here some reasons why they are not overincentivized.
a. BTC and ETH are the motherships of crypto econcomy. They offer the highest security for everyone, most widespread network. Most trusted foundation.
b. BTC and ETH are traded everywhere, much more than those nonsense tokens. The demand to trade is simply higher.
c. For the HIGH risk of impermanent loss, we are underincentivized, even on WBTC Pools.
BTC will move a lot due to halving. We are loosing money. And if you interact with uniswap you pay a lot of fees.
d. Next door are higher ROI decentralized exchanges. Users follow profit. Market force.

We can also make a further distinction:
BTC, centralized stablecoins, decentralized stablecoin.
There might be a case to be made to include only one centralized stablecoin in the program.
The justification for this change would be strategic and ecosystem-related.
It seems beneficial for DeFi ecosystem on Ethereum if DAI and WBTC grow adoption.

That said, when it comes to paying for ecosystem development, I think Sushiswap is already doing it just fine. And adding UNI rewards on top of Sushi rewards would just turn that into a pointless zero-sum game.
It wouldn’t add the liquidity to DeFi or retaining liquidity to Uniswap, it would just move the liquidity that doesn’t stay from one incentive program to the other. I mean the liquidity with a very low retention rate, the farming money.

I’m definitely not suggesting adding rewards to pools other than the 4 at hand in this topic. However, I think it is quite useful to look at other pools to have reference points. I’ll come back at it later with some numbers.

2 Likes

Your points a. and b. are totally correct. Point c is something every liquidity provider has to consider and right now there seems enough liquidity in eth/wbtc pool, so the fees seem to be enough incentive for them. Regarding higher ROI elsewhere: I would differentiate between true ROI due to the genuine protocol and incentives which are not sustainable in long term. If the ROI is temorarily higher elsewhere (like sushi in it’s first two weeks) ofc much mony moves for it. But I am sure it will come back to the place where the income is more sustainable, and this, right now is uniswap even without incentive.

These are strategic / political arguments.
When I talked about overincentivization, I meant it as a quantitative argument in terms of volume/liquidity ratios.

The IL on WBTC-ETH pair is lower than on USD-ETH pairs, because ETH+BTC moves in USD terms are correlated.
The lower IL on WBTC-ETH pair also explains why it had bigger liquidity while being a less profitable pair in terms of UNI distribution APYs. Meaning: LPs really don’t need any noticeable APY from volume fees on this pair as long as they get UNI distribution that gives them 10% APY.

If we take a look at it in isolation, the current 10% APY on WBTC-ETH pair could be viewed as LPs bet that most of the times ETH won’t rise above 2.5x+ or fall below -60% compared to BTC.
It is more complicated than that, but the general concept is there.

The statement that you make - that even the most overincentivized (comparatively) pool is underincentivized, implies that the economic design of Uniswap is not sustainable. I don’t think this statement has a solid foundation.

While BTC is certainly important, Uniswap’s core value proposition is that it is a trustless exchange where everyone can create a pair. And ETH is the main reference coin in the network.
Being able to trade ‘the nonsense tokens’, being able to list your token without the need of exchange’s permission is what Uniswap brings to the table in the first place.

2 Likes

Unicorn is all about governance, and governance is all about community involvement, and community involvement is about community inclusiveness, and community inclusiveness is all about community incentives, and community incentives borders on UNI circulation. A greater percentage of people have come to be actively involved in Unicorn governance through the airdrop and liquidity incentives, and this is what governance is all about…reward for loyalty. With more participation and loyalty from end-users, Uniswap will attain a high level of adoption that will in turn result in sustainable growth, UNI valuation will organically grow as a follow-on effect. Buy and burn CANNOT bring sustainable growth to UNI valuation and favour Unicorn.

2 Likes

Personally, I’m in favor of @tarun’s suggestion to shift a bit of incentives from the DAI to the WBTC pool.

Over the course of this last distribution, DAI liquidity increased a lot less than liquidity in the other pools - this leads me to believe that DAI/ETH liquidity is less responsive to incentives, and Uniswap gains a bit less from incentivizing DAI/ETH equally with the other pools.

WBTC is the only candidate for incentivized BTC liquidity, and attributing 1/3 of the program rewards to it with 2/3 to USD stablecoins (1/4 USDC, 1/4 USDT, 1/6 DAI) seems reasonable from a business standpoint. Over time, as other BTC implementations get more established, Uniswap may be able to diversify BTC incentives as well.

As for USD stablecoins, each of the current options offers something different and I support maintaining all 3 as part of the incentive scheme. USDC is trusted and well regulated, USDT is the most widely used stablecoin, and DAI offers decentralization.

If the community decides that pools still need additional incentives, I like the idea that was mentioned not needing 3 redundant pools for stable coins. Someone mentioned mUSD being unappealing as a forced exposure to USDT. What if USDT Uni staking wasn’t included and shifted to the other pools. A dollar is a dollar at the end of the day and the only difference between them is risk and compatibility. The USDT pool is already pretty large, why not promote others that have more universal trust instead?

You seem to like long posts and i have to say you go off target here and there.
Of course uniswap should list every token users desire. Even complete nonsense token pairs.
I never said we should stop it. Even if scam tokens can harm the image!! Just dont give any incentive on top.

  • Give incentives for the most used/secure/widespread crypto tokens
  • Calculate impermanent loss into the equation or many people will withdraw and hold liquidity
  • Most of your correlation aspects are true. yet they dont give reason to not incentivize WBTC.
    Everything is interdependend. Whole uniswap system is build on arbitrage trading. So?
1 Like

ETH/BTC is actually one of the most correlated pair of assets so it has one of the lowest risk of impermanent loss.

1 Like

It seems that our approaches’ main difference is that @monet-supply and @tarun seem to focus more on the absolute numbers in liquidity growth, and I’m more focused on the retention rates of liquidity and volume.

Here’s some numbers to consider:



The way I see it, both the liquidity programs Uniswap LPs enjoyed are quite successful overall.

The main achievement of the UNI liquidity mining program (LMP) is that it brought DAI and WBTC on par with the other two of the top-4 pairs on the exchange.

However, it appears that there’s not much room to grow from here, especially with 50% of the incentives.

If we consider liquidity before and after UNI LMP (see “4 difference with 2” column), it signals that rewarding USDC and USDT was less beneficial in comparison to rewarding WBTC and DAI.

The volume’s retention rate is quite impressive post LMP, but it looks like the volume is close to its cap at this point in time. So additional subsidies are unlikely to be efficient for the most part (see: “3 difference with 2”).

As for liquidity, as you can see in the “Difference with 3” column, WBTC and USDC have lower liquidity retention rates than the other two pairs so far. It could signal that a bigger percentage of money in these pairs stayed there temporarily to enjoy the farming incentives.

==

I think it makes sense to bring the discussion more to the quantitative realm, so we can figure out how good this decision is from the business perspective. It seems that there are some models to explore to guesstimate what we’re trying to achieve with the program - and what efficiency we expect.

Of course, even if all the metrics say that the continuation of LMP is inefficient as a business proposal, it doesn’t mean that the community shouldn’t accept it due to other reasons.

Some of the possible reasons to continue distribution are:

  1. enough self-rewarding interest in the community, i.e. there are enough UNI holders who want to participate in LMP and rip the benefits of it.
    If we view LMP from the self-interest perspective, the current proposal seems like a good middle-ground between the part of the community that wants to mine and the part that doesn’t.
  2. the distribution argument I’ve stated above
  3. other strategic considerations, some of which are more quantifiable, some are less.
5 Likes


Calculate the jumps. Be sure its not a stable relation. (150% swing potential)

I totally agree to your conclusion that right now incentive isn’t necessary to catch volume. Anyway, if there is some I will be happy to take it, but imho the UNI can better be hold till there is a good rational to spend it.