Thanks for writing this temperature check Lito!

I agree 100% that the Uni v3 Staker contract is not a good idea as it does incentivize a lot of opportunistic single-tick farmers, who are the least aligned with the project and create downward pricing pressure.

At Visor, we can work together with the Uniswap team to determine the right range to provide liquidity at. Our main Hypervisor contract interacts directly with the Uni v3 core contracts and we can dynamically set +/- x% range around the current price tick and automate rebalances when the price moves +/- y% from the center of the range. Alternatively, we can apply our own Auto Regressive Strategy which is designed to mitigate IL by widening ranges during high volatility periods and narrowing them during lower volatility time periods.

Because all our LPs would be within the same range, it wouldn't be possible to game the LM mechanics through providing large single-tick ranges.

Happy to answer more questions about our LM program and would love to be part of the solution.

Glad to see you chime in Monet and I agree on all points!

Re: which pairs to incentivize, I was actually thinking of only including key stabecoin/ETH pairs similar to the first Uni liquidity mining propgram in my next proposal for the Consensus check, just for the sake of making it as uncontentious as possible. But there’s definitely an argument to be made to include tokens of 1-2 key projects on each L2 (e.g SNX on Optimism).

Regarding widening the range for LP’s I agree 100% as well. This should be an opportunity to introduce normal users to Uni V3 lp’ing - not just gift $UNI to whales and market makers. I think most V3 LP schemes work in a way that you get more rewards the closer your range is to the spot price. I will get in touch with the Gelato and Visr teams as well as other experts to see if this can be tweaked so that everyone get’s the same rewards as long as they are within a certain range (e.g +- 20%).

Thanks for writing this temperature check Lito!

I agree 100% that the Uni v3 Staker contract is not a good idea as it does incentivize a lot of opportunistic single-tick farmers, who are the least aligned with the project and create downward pricing pressure.

At Visor, we can work together with the Uniswap team to determine the right range to provide liquidity at. Our main Hypervisor contract interacts directly with the Uni v3 core contracts and we can dynamically set +/- x% range around the current price tick and automate rebalances when the price moves +/- y% from the center of the range. Alternatively, we can apply our own Auto Regressive Strategy which is designed to mitigate IL by widening ranges during high volatility periods and narrowing them during lower volatility time periods.

Because all our LPs would be within the same range, it wouldn't be possible to game the LM mechanics through providing large single-tick ranges.

Happy to answer more questions about our LM program and would love to be part of the solution.

Glad to see you chime in Monet and I agree on all points!

Re: which pairs to incentivize, I was actually thinking of only including key stabecoin/ETH pairs similar to the first Uni liquidity mining propgram in my next proposal for the Consensus check, just for the sake of making it as uncontentious as possible. But there’s definitely an argument to be made to include tokens of 1-2 key projects on each L2 (e.g SNX on Optimism).

Regarding widening the range for LP’s I agree 100% as well. This should be an opportunity to introduce normal users to Uni V3 lp’ing - not just gift $UNI to whales and market makers. I think most V3 LP schemes work in a way that you get more rewards the closer your range is to the spot price. I will get in touch with the Gelato and Visr teams as well as other experts to see if this can be tweaked so that everyone get’s the same rewards as long as they are within a certain range (e.g +- 20%).

I strongly support this proposal. Uniswap has been very careful with choosing what chains to deploy on, opting to only deploy to chains that are "ideologically aligned" with the Ethereum ecosystem. So far, this has only been Arbitrum and Optimism.

Both Arbitrum and Optimism don't have native tokens, and must therefore rely on the application layer to incentivize adoption. Uniswap is definitely the best positioned project to provide this incentivization given that it a) has the largest market cap of any DeFi project and b) has a massive treasury.

This chart from https://cryptofees.info/protocol/avalanche shows the total fee revenue on Avalanche, with the purple dot representing the start of their liquidity mining campaign. Liquidity mining is an incredibly powerful tool to bootstrap a network.

Responding to some of @Buckerino's points:

I´d say there is a better path to take than liquidity mining such as OlympusDAO has implemented

OlympusDAO has definitely pioneered a great alternative to standard liquidity mining. It would be very interesting for Uniswap to consider bonded purchases of UNI/ETH LP tokens into the Uniswap treasury.

As long as the big CEXs do not adopt L2s, nothing is truly going to change

This may be a bit of a chicken-and-egg problem, CEXs don't support L2s since they don't have users, and users aren't on L2s because there's insufficient CEXs. A liquidity mining campaign is at least a worthwhile experiment to stir up some noise around Ethereum L2s.

One final thought I'll add: AMMs like Uniswap truly succeed with long-tail assets. It may be worth waiting for Arbitrum and Optimism to fully remove the whitelists from their token bridges before incentivizing adoption of L2s.

All 3 instances of the Uniswap protocol sit in areas that are expensive/hard to reach for the masses.

We know that L1 has no chance of becoming affordable and will likely only get more expensive as time goes on. Which should make the Arbitrum/Optimism instances our primary focus.

Just because this move benefits other teams as well doesn't mean we shouldn't go ahead with it. Uniswap has already benefited from both of the rollup projects and will continue to benefit from their success.

It is justified to move to the consensus check stage. This will allow a more layed out plan on mining pairs, and how Hop Protocol see's a program unfolding.

Reasons why this should move to a consensus check stage:

1. The number of token holder's voting with smaller amounts of UNI voted by large for this to pass to next stage.

2. The voting period was very short at ~2 days.

3. UCLA blockchain is signaling support (after voting had closed) with 5 million.

4. @BOR4 who has <4 very large delegated UNI. Voted down the proposal right before vote closed, and did not engage in proposal dialogue or give reasoning on behalf of delgated votes.

Instead of repeating a temperature check, and to allow more information to be unveiled on how such a program would be released a consensus check makes more sense.

I also encourage the proposer to check out sybil.org and inform large delegate's via twitter about a consensus check and to at the very least to get involved in the discussion (on behalf of those who have entrusted UNI votes to their engagement).

Well, this represents a conundrum: on one side, nobody can really stop anybody from making a second temperature check.

However, the question stands whether this would set a good precedent. It would be preferrable to avoid making it look like, "We will vote as many times as we want to get the result we want." type of thing. Hypothetically, another set of delegates could come after the second temperature check saying they missed that one and would have voted against it. I think having one binding temperature check on an issue should be enough to avoid this situation. Also, it should be said I am talking from a biased position since Im not in favor of the proposal.

So, of course, I will leave it up to your judgement to decide, but I do believe this is the basic principle for voting which should not be violated to preserve order.

I looked at the data from the original liquidity mining, which has discouraged other mining proposals. I believe there is a misconception that the original liquidity mining had no benefit.

Supporting documentation:

In the above chart, we can see that volume had tapered down after the large volumes of 2020 Defi Summer.

The liquidity mining program was launched on September 18th 2020 and ended November 17th 2020 (indicated by the liquidity mining arrow, and the end date by the new settling point)

The key improvement lines are:

Purple Line: eth-wbtc Orange Line: eth-dai

September 16th (2 days before liquidity mining):

eth-wbtc: 17,793,554.0260 eth-dai:31,953,224.9506

November 18th (new settling point):

eth-wbtc: 206,514,310.6676 eth-dai: 176,986,579.1124

Post liquidity mining (new long time floor):

eth-wbtc: 206,514,310.6676/17,793,554.0260= 11.6x increase in liquidity TVL

eth-dai: 176,986,579.1124/31,953,224.9506= 5.54x increase in liquidity TVL

Sources:

https://uniswap.org/blog/uni/

https://dune.xyz/queries/13500

I strongly support this proposal. Uniswap has been very careful with choosing what chains to deploy on, opting to only deploy to chains that are "ideologically aligned" with the Ethereum ecosystem. So far, this has only been Arbitrum and Optimism.

Both Arbitrum and Optimism don't have native tokens, and must therefore rely on the application layer to incentivize adoption. Uniswap is definitely the best positioned project to provide this incentivization given that it a) has the largest market cap of any DeFi project and b) has a massive treasury.

This chart from https://cryptofees.info/protocol/avalanche shows the total fee revenue on Avalanche, with the purple dot representing the start of their liquidity mining campaign. Liquidity mining is an incredibly powerful tool to bootstrap a network.

Responding to some of @Buckerino's points:

I´d say there is a better path to take than liquidity mining such as OlympusDAO has implemented

OlympusDAO has definitely pioneered a great alternative to standard liquidity mining. It would be very interesting for Uniswap to consider bonded purchases of UNI/ETH LP tokens into the Uniswap treasury.

As long as the big CEXs do not adopt L2s, nothing is truly going to change

This may be a bit of a chicken-and-egg problem, CEXs don't support L2s since they don't have users, and users aren't on L2s because there's insufficient CEXs. A liquidity mining campaign is at least a worthwhile experiment to stir up some noise around Ethereum L2s.

One final thought I'll add: AMMs like Uniswap truly succeed with long-tail assets. It may be worth waiting for Arbitrum and Optimism to fully remove the whitelists from their token bridges before incentivizing adoption of L2s.

All 3 instances of the Uniswap protocol sit in areas that are expensive/hard to reach for the masses.

We know that L1 has no chance of becoming affordable and will likely only get more expensive as time goes on. Which should make the Arbitrum/Optimism instances our primary focus.

Just because this move benefits other teams as well doesn't mean we shouldn't go ahead with it. Uniswap has already benefited from both of the rollup projects and will continue to benefit from their success.

It is justified to move to the consensus check stage. This will allow a more layed out plan on mining pairs, and how Hop Protocol see's a program unfolding.

Reasons why this should move to a consensus check stage:

1. The number of token holder's voting with smaller amounts of UNI voted by large for this to pass to next stage.

2. The voting period was very short at ~2 days.

3. UCLA blockchain is signaling support (after voting had closed) with 5 million.

4. @BOR4 who has <4 very large delegated UNI. Voted down the proposal right before vote closed, and did not engage in proposal dialogue or give reasoning on behalf of delgated votes.

Instead of repeating a temperature check, and to allow more information to be unveiled on how such a program would be released a consensus check makes more sense.

I also encourage the proposer to check out sybil.org and inform large delegate's via twitter about a consensus check and to at the very least to get involved in the discussion (on behalf of those who have entrusted UNI votes to their engagement).

Well, this represents a conundrum: on one side, nobody can really stop anybody from making a second temperature check.

However, the question stands whether this would set a good precedent. It would be preferrable to avoid making it look like, "We will vote as many times as we want to get the result we want." type of thing. Hypothetically, another set of delegates could come after the second temperature check saying they missed that one and would have voted against it. I think having one binding temperature check on an issue should be enough to avoid this situation. Also, it should be said I am talking from a biased position since Im not in favor of the proposal.

So, of course, I will leave it up to your judgement to decide, but I do believe this is the basic principle for voting which should not be violated to preserve order.

I looked at the data from the original liquidity mining, which has discouraged other mining proposals. I believe there is a misconception that the original liquidity mining had no benefit.

Supporting documentation:

In the above chart, we can see that volume had tapered down after the large volumes of 2020 Defi Summer.

The liquidity mining program was launched on September 18th 2020 and ended November 17th 2020 (indicated by the liquidity mining arrow, and the end date by the new settling point)

The key improvement lines are:

Purple Line: eth-wbtc Orange Line: eth-dai

September 16th (2 days before liquidity mining):

eth-wbtc: 17,793,554.0260 eth-dai:31,953,224.9506

November 18th (new settling point):

eth-wbtc: 206,514,310.6676 eth-dai: 176,986,579.1124

Post liquidity mining (new long time floor):

eth-wbtc: 206,514,310.6676/17,793,554.0260= 11.6x increase in liquidity TVL

eth-dai: 176,986,579.1124/31,953,224.9506= 5.54x increase in liquidity TVL

Sources:

https://uniswap.org/blog/uni/

https://dune.xyz/queries/13500

I dont see the liquidity mining that was provided during that time as equivelant to this one. The first mining program was more of a reaction to sushiswap's mining program. Yes, many sold UNI. I do think we had more engagement at the time in the forums; as UNI had just been launched.

I see this program as being proactive, and a way to educate and incentivize new users to the L2 ecosystem (while also reimburse cost of trying L2 ecosystem) VS the old reactive liquidity mining that incentivized established users who moved around to different farming to just sell.

I am in support of this temperature check, and it will be good to see how the liqudity program is actually structured (I would be on board with a bonding progam as well if that seems to make more sense). It is better to be in a position of setting the tone on onboarding new users, rather than to be on the heels of another protocol.

I am in full support of this proposal. I see this as a win-win for both L2 adoption, and Uniswap being seen a portal to this ecosystem. I am keen to see how the liquidity mining would be structured.

In response to @Buckerino:

I still dont see a reason why the UNI holders should subsidize it for everyone else. What do the token holders get in return?

As we have seen governance has not had much participation. Having a liquidity mining program is a great way to also get new users engaged with Uniswap through another avenue of distributing out the token into new paricipants. Yes, some may sell, but other's will see the value in uniswap and the onboared ecosystem to become more involved. The sushiswap, and olympusdao forums have active dialogues in governance forums, while also having distribution model of tokens into the market.

Discounted OlympusDAO bonds provide a better alternative where the UNI holders providing the liquidity incentive do not get stiffed. I do not see any reason why this would not be a better alternative.

I also think the OlympusDAO model of liquidity distribution through bonds is a good idea. It also makes sense with uniswap TWAP's to allow more security of oracle pricing through stable liquidity. I am unsure if this the avenue for this particular proposal, or a seperate one.

I still dont see a reason why the UNI holders should subsidize it for everyone else. What do the token holders get in return?

The performance of the protocol has not been in alingment for quite some time with the performance of the token. It has been proven that the liquidity will not stay once the incentives run out.

Discounted OlympusDAO bonds provide a better alternative where the UNI holders providing the liquidity incentive do not get stiffed. I do not see any reason why this would not be a better alternative.

I think there is a general misconception of what the UNI treasury is capable of doing. Whilst I do agree that not every subsidy has to have a ROI attached to it (UGP, DeFi Fund), this one certainly would need to have it. We are talking about 10Ms worth of USD in subsidies for the success of unaffiliated companies with effectively nothing in return for the people who are supposed to foot the bill.

Effectively, the Uniswap treasury is only worth as much as the liquidity on the market allows it to be. Currently, any argument saying the treasury is worth more than 180-200M USD is absolutely absurd and imaginary to say the least. Proposing to cut double digit value out of it for vague benefits is outright provocative/taunting at best.

I dont think its good use of the funds. The 11B treasury you are pointing to is largely a myth. A paper put out by Monet.supply and UCC.Hasu show just how much "11B is worth". Its not even 1B...more like 150-200M USD which would get distressed even more if a liquidity campaign would ensue.

Liquidity mining is a great way to bootstrap liquidity onto a freshly launched project, but it proves to be quite parasitic in the long run. I´d say there is a better path to take than liquidity mining such as OlympusDAO has implemented. The problem with liquidity mining is that it leaves a negative footprint once it ends. Liquidity is loyal only to the person who pays. Why should UNI token holders bear this burden? What do they get in return besides the fact that they will get utterly dilluted?

The real problem for L2s is like you outlined in your point - fiat-on ramps. As long as the big CEXs do not adopt L2s, nothing is truly going to change. Liquidity mining will just burn through the UNI holders´ value.

Also, I think this is a duplicate thread. This was discussed in the past already...but I think it was for a migration from V2 to V3.

Doing a liquidity mining campaign instead of OlympusDAO liquidity model represents inefficiencies which are going to be exploited. Inefficiencies worth millions of dollars. Why waste resources? It does not make sense to throw money out of the window if better options are on the table.

Regarding your point on governance not having much participation, UNI actually had liquidity mining campaign a year ago. It did not revive the governance part of the token, did it? All it did was provide temporary boost whilst sucking out value from the token holders and then left.

This would be the exact same result. UNI holders would lose value and get little in return.

I dont see the liquidity mining that was provided during that time as equivelant to this one. The first mining program was more of a reaction to sushiswap's mining program. Yes, many sold UNI. I do think we had more engagement at the time in the forums; as UNI had just been launched.

I see this program as being proactive, and a way to educate and incentivize new users to the L2 ecosystem (while also reimburse cost of trying L2 ecosystem) VS the old reactive liquidity mining that incentivized established users who moved around to different farming to just sell.

I am in support of this temperature check, and it will be good to see how the liqudity program is actually structured (I would be on board with a bonding progam as well if that seems to make more sense). It is better to be in a position of setting the tone on onboarding new users, rather than to be on the heels of another protocol.

I am in full support of this proposal. I see this as a win-win for both L2 adoption, and Uniswap being seen a portal to this ecosystem. I am keen to see how the liquidity mining would be structured.

In response to @Buckerino:

I still dont see a reason why the UNI holders should subsidize it for everyone else. What do the token holders get in return?

As we have seen governance has not had much participation. Having a liquidity mining program is a great way to also get new users engaged with Uniswap through another avenue of distributing out the token into new paricipants. Yes, some may sell, but other's will see the value in uniswap and the onboared ecosystem to become more involved. The sushiswap, and olympusdao forums have active dialogues in governance forums, while also having distribution model of tokens into the market.

Discounted OlympusDAO bonds provide a better alternative where the UNI holders providing the liquidity incentive do not get stiffed. I do not see any reason why this would not be a better alternative.

I also think the OlympusDAO model of liquidity distribution through bonds is a good idea. It also makes sense with uniswap TWAP's to allow more security of oracle pricing through stable liquidity. I am unsure if this the avenue for this particular proposal, or a seperate one.

I still dont see a reason why the UNI holders should subsidize it for everyone else. What do the token holders get in return?

The performance of the protocol has not been in alingment for quite some time with the performance of the token. It has been proven that the liquidity will not stay once the incentives run out.

Discounted OlympusDAO bonds provide a better alternative where the UNI holders providing the liquidity incentive do not get stiffed. I do not see any reason why this would not be a better alternative.

I think there is a general misconception of what the UNI treasury is capable of doing. Whilst I do agree that not every subsidy has to have a ROI attached to it (UGP, DeFi Fund), this one certainly would need to have it. We are talking about 10Ms worth of USD in subsidies for the success of unaffiliated companies with effectively nothing in return for the people who are supposed to foot the bill.

Effectively, the Uniswap treasury is only worth as much as the liquidity on the market allows it to be. Currently, any argument saying the treasury is worth more than 180-200M USD is absolutely absurd and imaginary to say the least. Proposing to cut double digit value out of it for vague benefits is outright provocative/taunting at best.

I dont think its good use of the funds. The 11B treasury you are pointing to is largely a myth. A paper put out by Monet.supply and UCC.Hasu show just how much "11B is worth". Its not even 1B...more like 150-200M USD which would get distressed even more if a liquidity campaign would ensue.

Liquidity mining is a great way to bootstrap liquidity onto a freshly launched project, but it proves to be quite parasitic in the long run. I´d say there is a better path to take than liquidity mining such as OlympusDAO has implemented. The problem with liquidity mining is that it leaves a negative footprint once it ends. Liquidity is loyal only to the person who pays. Why should UNI token holders bear this burden? What do they get in return besides the fact that they will get utterly dilluted?

The real problem for L2s is like you outlined in your point - fiat-on ramps. As long as the big CEXs do not adopt L2s, nothing is truly going to change. Liquidity mining will just burn through the UNI holders´ value.

Also, I think this is a duplicate thread. This was discussed in the past already...but I think it was for a migration from V2 to V3.

Doing a liquidity mining campaign instead of OlympusDAO liquidity model represents inefficiencies which are going to be exploited. Inefficiencies worth millions of dollars. Why waste resources? It does not make sense to throw money out of the window if better options are on the table.

Regarding your point on governance not having much participation, UNI actually had liquidity mining campaign a year ago. It did not revive the governance part of the token, did it? All it did was provide temporary boost whilst sucking out value from the token holders and then left.

This would be the exact same result. UNI holders would lose value and get little in return.

After discussing these ideas with Dan Robinson and a few other people, I now realize that implementing my earlier proposal may require too big of a change to the Staker contract.

Note sure what's the status of that proposal is, but I agree that one way forward could be to mandate that liquidity is deployed to a minimum "floating" range -- ie have the position's (upperTick-lowerTick) be larger or equal than a fixed value be centered at any price.

Choosing that range could be tricky, but it could be based on the 1-std expected move. For instance, a 1-std expected move for a 100% annualize volatility after 21 days is ±24%. So limiting the LP range to ±24% would only require a rebalancing every 21 days, on average.

By comparison, a 1-tick position in the 0.3% pool would need to be rebalanced every 20 minutes (clearly impossible to do unless you run bots), but successfully running this strategy would results in 34x more rewards than ±24% LP positions.

BTW, I suggested 21 days, but the number of days could be changed to whatever according to

• Expected move = (annual volatility) * sqrt(days/365)

TL;DR: UNI liquidity mining rewards should be based on the amount of time the price is within the LP token's range.

Hi all–

The goal of the liquidity mining program should be to increase active liquidity. What I mean by active liquidity? Let’s look at a few pools on mainnet:

First, ETH-USDC-0.3%: info[dot]uniswap[dot]org/#/pools/0x8ad599c3a0ff1de082011efddc58f1908eb6e6d8 (I can't embed links or media it seems, otherwise I would have posted a screenshot).

We’ve had a rather large downward price move in recent days and the “liquidity peak” is away from the current price. However, there are still enough funds locked at the current tick (>2M USDC in active liquidity) to limit slippage and be resilient against large buys-sells.

On the other hand, let’s look at the ETH-MKR-0.3% pool: info[dot]uniswap[dot]org/#/pools/0xe8c6c9227491c0a8156a0106a0204d881bb7e531

There is only 1.87 ETH lock at the current tick, that’s only $7,500 out of the $355m in TVL (!!). As a result, a relatively small transaction can drastically shift the price of MKR. Clearly, most LPs do not care about rebalancing, or perhaps they’re OK holding MKR, but the amount of active liquidity is way too small compared to the TVL.

The problem with the current LP staking contract is that it rewards users according to the effective liquidity, which is basically proportional to the amount of fees collected. Whales and market makers can easily shift their LP position in response to large price moves because gas costs will not matter as much once you cross a certain level (>$10,000?).

Smaller players, on the other hand, will rapidly get wiped out because burn+rebalance+re-deploy may incur $100’s in gas fees.

How do we solve this? How can smaller players that do not have time to actively manage their LP position still be rewarded for contributing liquidity in a manner that is proportional to their actual stake?

I am proposing to reward all LPs who have any amount of liquidity at the current tick.

For example, if the current price is 4,280, then locking 1 ETH in the ETH-USDC pool between 1,000-32,000 will receive as much UNI reward as locking 1 ETH between 4275-4300. But LPs that deployed 1 ETH to either 1000-4000 or 4500-4535 or any range that do not contain the current price will receive nothing.

Why do it like that?

1. It’s relatively easy to monitor that time spent inside any given range by calling snapshotCumulativesInside.
2. Whales and market makers that actively rebalance their position already benefit by collecting a disproportionate amount of the collected trading fees. They have their cake.
3. A more conservative position will receive as much reward as a more aggressive one, but the conservative option does not suffer IL because it won’t need to be rebalanced as frequently. Easier to manage for most users.
4. By rewarding according to the time spent “in-the-money”, aggressive LPs would actually receive a lower relative fraction of the rewards because the price will leave a narrowly defined range (ie. 1-tick) after small price movements whereas a wider one will remain “in-the-money”.
5. Using a fixed % range may still leave pools under-liquid in case of large price moves. What’s the best percentage? 20% is way too wide for DAI-USDC, but may be too narrow for ETH-WBTC. What’s the best range? Even a 1-standard deviation expected price move will still be “wrong” 32% of the time.
6. The fixed-range option still rewards power-users and market makers and will leave smaller LPs with a much lower reward.

But clearly, making everyone deploying to the Full-Range (which is worse than Uniswap v2 in terms of capital efficiency) would not contribute sufficient liquidity to decrease slippage, etc.

What’s a good limit for the maximum range?

I suggest that any position whose range is chosen so that the capital efficiency is greater than 2x that of Uniswap v2 should be rewarded. By the way, the calculation for capital efficiency improvement over Uniswap v2 is:

(Capital efficiency) = sqrt(r)/(sqrt(r)-1) = 1/(1-1/sqrt(r))

So a Full-range liquidity would receive zero rewards, and any liquidity whose range factor r = sqrt(upperTick/lowerTick) is smaller than r < 4 will receive liquidity mining rewards.

The r < 4 criteria could be changed (a 4x price move is indeed unlikely for most assets), but r=4 LPs would still be performing 2x better than Uniswap v2.

After discussing these ideas with Dan Robinson and a few other people, I now realize that implementing my earlier proposal may require too big of a change to the Staker contract.

Note sure what's the status of that proposal is, but I agree that one way forward could be to mandate that liquidity is deployed to a minimum "floating" range -- ie have the position's (upperTick-lowerTick) be larger or equal than a fixed value be centered at any price.

Choosing that range could be tricky, but it could be based on the 1-std expected move. For instance, a 1-std expected move for a 100% annualize volatility after 21 days is ±24%. So limiting the LP range to ±24% would only require a rebalancing every 21 days, on average.

By comparison, a 1-tick position in the 0.3% pool would need to be rebalanced every 20 minutes (clearly impossible to do unless you run bots), but successfully running this strategy would results in 34x more rewards than ±24% LP positions.

BTW, I suggested 21 days, but the number of days could be changed to whatever according to

• Expected move = (annual volatility) * sqrt(days/365)

TL;DR: UNI liquidity mining rewards should be based on the amount of time the price is within the LP token's range.

Hi all–

The goal of the liquidity mining program should be to increase active liquidity. What I mean by active liquidity? Let’s look at a few pools on mainnet:

First, ETH-USDC-0.3%: info[dot]uniswap[dot]org/#/pools/0x8ad599c3a0ff1de082011efddc58f1908eb6e6d8 (I can't embed links or media it seems, otherwise I would have posted a screenshot).

We’ve had a rather large downward price move in recent days and the “liquidity peak” is away from the current price. However, there are still enough funds locked at the current tick (>2M USDC in active liquidity) to limit slippage and be resilient against large buys-sells.

On the other hand, let’s look at the ETH-MKR-0.3% pool: info[dot]uniswap[dot]org/#/pools/0xe8c6c9227491c0a8156a0106a0204d881bb7e531

There is only 1.87 ETH lock at the current tick, that’s only $7,500 out of the $355m in TVL (!!). As a result, a relatively small transaction can drastically shift the price of MKR. Clearly, most LPs do not care about rebalancing, or perhaps they’re OK holding MKR, but the amount of active liquidity is way too small compared to the TVL.

The problem with the current LP staking contract is that it rewards users according to the effective liquidity, which is basically proportional to the amount of fees collected. Whales and market makers can easily shift their LP position in response to large price moves because gas costs will not matter as much once you cross a certain level (>$10,000?).

Smaller players, on the other hand, will rapidly get wiped out because burn+rebalance+re-deploy may incur $100’s in gas fees.

How do we solve this? How can smaller players that do not have time to actively manage their LP position still be rewarded for contributing liquidity in a manner that is proportional to their actual stake?

I am proposing to reward all LPs who have any amount of liquidity at the current tick.

For example, if the current price is 4,280, then locking 1 ETH in the ETH-USDC pool between 1,000-32,000 will receive as much UNI reward as locking 1 ETH between 4275-4300. But LPs that deployed 1 ETH to either 1000-4000 or 4500-4535 or any range that do not contain the current price will receive nothing.

Why do it like that?

1. It’s relatively easy to monitor that time spent inside any given range by calling snapshotCumulativesInside.
2. Whales and market makers that actively rebalance their position already benefit by collecting a disproportionate amount of the collected trading fees. They have their cake.
3. A more conservative position will receive as much reward as a more aggressive one, but the conservative option does not suffer IL because it won’t need to be rebalanced as frequently. Easier to manage for most users.
4. By rewarding according to the time spent “in-the-money”, aggressive LPs would actually receive a lower relative fraction of the rewards because the price will leave a narrowly defined range (ie. 1-tick) after small price movements whereas a wider one will remain “in-the-money”.
5. Using a fixed % range may still leave pools under-liquid in case of large price moves. What’s the best percentage? 20% is way too wide for DAI-USDC, but may be too narrow for ETH-WBTC. What’s the best range? Even a 1-standard deviation expected price move will still be “wrong” 32% of the time.
6. The fixed-range option still rewards power-users and market makers and will leave smaller LPs with a much lower reward.

But clearly, making everyone deploying to the Full-Range (which is worse than Uniswap v2 in terms of capital efficiency) would not contribute sufficient liquidity to decrease slippage, etc.

What’s a good limit for the maximum range?

I suggest that any position whose range is chosen so that the capital efficiency is greater than 2x that of Uniswap v2 should be rewarded. By the way, the calculation for capital efficiency improvement over Uniswap v2 is:

(Capital efficiency) = sqrt(r)/(sqrt(r)-1) = 1/(1-1/sqrt(r))

So a Full-range liquidity would receive zero rewards, and any liquidity whose range factor r = sqrt(upperTick/lowerTick) is smaller than r < 4 will receive liquidity mining rewards.

The r < 4 criteria could be changed (a 4x price move is indeed unlikely for most assets), but r=4 LPs would still be performing 2x better than Uniswap v2.