I would ask for some explanation as to how this helps the customer solve a problem:
-Does adding liquidity to sushi make DPI cheaper to buy?
-Does it make it so some large unserved portion of demand can now access it (I’d argue the people who know uni and sushi is almost 100% overlap)? This point is doubly hammered home when we discuss aggregators.
If the goal is to make DPI more easy to buy, the right course of action would actually be to increase Uni rewards again, as that would make the liquidity pool larger and provide better pricing. I don’t think that’s the right way, either.
A) Once liquidity is established in a pool with a good product (i.e. DPI) it will tend to stay there. The time of DeFi summer and people chasing 100000% APY on food coins is likely over.
This is a pretty big assumption and something at this level needs to be backed up by data. An LP token is a fundamentally different product than an Index token and represents a constant mix strategy.
It prevents us from reaching full decentralization. Liquidity incentives are important tool for our community and providing liquidity is one of the most important things our community members do.
I don’t see how each statement leads to the next or how it relates to the broader point.
Sorry for picking on you guys, those were just the few examples that popped out to me and I picked them kind of randomly. (You certainly don’t need to respond to them. )We are all guilty of doing this, especially myself. It’s been a super tough pill for me to swallow as I get frustrated when people put brakes on conclusions that seem so obvious to me.
But, to level up as a DAO, it is critical we level up the way we communicate and that means making sure we are starting from a correct framing of the problem, not getting attached to a solution, backing up assumptions with data, staying concise, and staying focused in our communication.
Sushi has been brought up a lot and it has been posited that doing LM on Sushi means that we get access to the Sushi user base. To me, that is a super big logical leap and super unclear as to if its true. Every investment firm I know in DeFi is short Sushi as an AMM and long Sushi on its other initiatives.
We won’t know until we try. One thing that we do know is the Sushi Community has been on support of $DPI. They do have a userbase that will beneficial to tap into. We also need the AWG to check if their is 100% cross over for wallets holding both $UNI token and $SUSHI token. Personally I do not believe so. We are a data driven organization however data wont tell us the full picture until we try, majority of the points being made are based off assumptions, and we should be wary of falling to confirmation bias. We need to experiment. Also I do believe it is encouraging the other successful DAO’s have liquidity across different DEX’s, it’s almost like diversifying liquidity risk. Also if we give equal $INDEX rewards across multiple DEX’s my assumption would be LP’s will provide liquidity based on the value of the transaction fees, intrinsic and extrinsic value of the DEX reward tokens.
I think there are two conversations going on here, which we need to separate. One is liquidity mining and liquidity incentives. The second one is the ecosystem partnerships and increasing the visibility of DPI (storefronts analogy).
Looking strictly at liquidity mining and liquidity incentives, we don’t need to incentivise Sushiswap liquidity. Our imaginary customer benefits from the large Uni pool and low slippage. There’s no need to fragment liquidity. We should gradually scale down Uni incentives while minimising the impact on the TVL in the pool. This will increase un-incentivised TVL and make the DPI-side of the Index Coop business more sustainable.
Staying with our imaginary customer, it also makes sense to continue incentivising Loopring’s L2 liquidity. I, personally, have been trading strictly on L2 (for DPI, ETH and WBTC) for the last month or so. I think more and more people will do so. Overall volume on Loopring hit $30m on Feb 22 and DPI volume specifically is quite decent as well. The problem we are trying to solve here is high fees. It’s giving our customers an option for buying DPI with no gas fees whatsoever. It’s giving them an opportunity to take advantage of wild market swings when gas is over 1000 gwei.
Now, let’s think about ecosystem partnerships. Who is the customer here? Well, to some degree it is us, Index Coop. How do we benefit? Sushi are building some interesting products, beyond the AMM. Like Punia said, investment firms are short their AMM and long their upcoming products. Having a close relationship with the team would allow us to explore DPI integrations in their new products, like the Bento Box, for example. They are also willing to allocate $1m of Treasury funds to DPI. Can we work with them to facilitate broader Treasury allocations to DPI across projects in the Yearn ecosystem? Furthermore, projects in the Yearn ecosystem tend to build on top of each other. Would having low DPI liquidity on Sushi hurt us? At the end, this does come back to our actual customer.
Ignoring the potential of ecosystem partnerships is missing forest for the trees. Sure, it might not be quantifiable or have immediate customer benefits, but risk/reward is asymmetric here. There’s limited risk (if we fine-tune Sushi incentives to minimise impact on the Uni pool) but potentially significant upside down the line.
To conclude, I think we should incentivise Sushi but at a level that won’t draw too much liquidity from Uni. We can run this as a one-off, 30-day or 60-day program, ideally as we work with Sushi to also increase Onsen rewards for DPI.
P.S. This is a different conversation entirely but it’s an elephant in the room. We have an enormous amount of INDEX that needs to be distributed. It’s true that there are a lot of ways to solve this problem, but we are not actively pursuing any of them as far as I know. In the meantime, the less we distribute today, the more we will have to distribute tomorrow. So we should keep this in mind as we discuss liquidity mining and Sushi.
@puniaviision these threads are extremely important. This kind of free flowing debate is exactly how we level up as a DAO. If you look at the forum for any of the best DAOs their forums are filled with these kinds of debates - sometimes the debate is quick and easy. Sometimes it evolves over 100s of posts as the community works through a problem together.
If Index Coop has passionate community members that they are willing to spend hours building out analysis and having long debates regarding a nuanced and complex aspect of our strategy - we as a community are WINNING.
These issues are complicated and extremely new, the only way to come up with really good community decisions is to have these conversations and to let these conversations flow and evolve. I am very convinced by some of @LemonadeAlpha 's points - far more convinced than at the start of this debate.
We are talking through complex and multi-layered issues. We all need to leave our pre-conceptions at the door and accept that we don’t have the answer to some of these problems.
But on your broader point of leveling up community discussion I agree. We need to improve the analysis and date and problem framing etc. And those things are happening right now! Both you and @gregdocter has done an amazing job of facilitating free-flowing community discussion while keeping people focused and on point. Any outsider reading this forum would be blown away by how the quality and engagement of these conversations in our community have evolved over the last months.
Sometimes its easy to forget that none of us has 5-years of experience working at a DAO - no-one especially myself is very good at it right now. Everyone is still learning. And we are learning together. I promise in 2-years when Index Coop has $5 or $10 bn AUM and we are running the largest crypto asset manager in the world (which Index will become), we will look back at these conversations and laugh about how little any of us really understood about DeFi.
@LemonadeAlpha and @verto0912 this is essentially where I am at as well. I have no special attachment to SushiSwap liquidity mining for $DPI. If we can do it cheaply and easily I think we should, if not it is certainly not a big deal. Doing some sort of month long limited liquidity mining in conjunction with Sushi being listed in $DPI is an easy win for both us and them - especially if both communities promote it and fund it. And if we can fund it for cheap even better!
The broader point I am emphasizing is our need for a long-term liquidity strategy for our products. None of us have the perfect strategy for our liquidity. Conversations like this are how we actively develop that strategy.
My final point: We are building a community of world-class talent. To even find this community or contribute to it you have to be EXTREMELY good. There are maybe 2000 people in the world actively working at DAOs and capable of having conversations at this level on these subjects. IMHO @overanalyser probably understand DeFi liquidity as well as anyone not named Hayden or OxMaki. People don’t just walk off the street capable of designing a fully functioning DAO Smart Treasury like @DarkForestCapital is doing.
We talk a lot about attracting and retaining top talent. We have the top talent - and we continue to bring people like @Mringz(who with @verto0912 spearheaded our Pool Together initiave during his second week at the Coop) into our community.
Passionate debates like this driven by world-class community members and contributors is evidence of what a strong DAO Set Protocol has built and grown.
This is great! Let’s dig into the statements here.
Problem: High gas fees on Ethereum are restricting certain subsection of our users from accessing DPI.
Solution:* L2s provide a way to buy and sell tokens for low gas costs. Building out a liquid pools for DPI on L2s will help us acquire more users. Currently, Loopring is the most active L2 so that is where we are focusing our efforts.
We will know this works if: We see a net inflow of users holding and trading DPI on Loopring wallets. (I don’t know if this is possible, but this is directionally the type of KPIs we would want to set based on the problem. Not just net liquidity or net volume.)
Considerations: Loopring does not have mass appeal. Do users who come to our website, try to buy the product, churn because gas fees are too expensive even know about Loopring? Should we be educating them? Should we add a flow to the website to complete the loop?
This is just a sample, typically you do a brainstorming activity among all of these points and simplify the complete initiative in a Product Requirements Doc. Happy to share how we are approaching it.
*Note: Typically we start with the problem and brainstorm a bunch of different solutions before picking one, this is too make sure that we don’t have tunnel vision and are missing other low cost/high effective ways to deliver the solution to our customers. Will skip this for now.
Let’s dig in here. We need to go deeper on what the actual problem is and we can’t consider ourselves as the customer. We’re not paying ourselves.
Sample Product Statement:
Problem: Through user interviews, we have gathered that extrinsic productivity is a key considerations whales make when choosing to buy DPI. Getting integrated with DeFi protocols like Compound, Maker, and Aave has been incredibly difficult and taken us months.
But, for the sake of discussion, let’s assume the problem is itself we aren’t integrated into the Yearn ecosystem. The logical jumps that I see are:
DPI is already on Sushi and being incentivized, how does incentivizing it with $INDEX change the relationship?
Is incentivizing a token pair the best way to build a relationship with another project (maybe we can pay an ambassador or be more active on their forums)?
Similarly, what is the asymmetric upside to this relationship and what is the best way to access it?
Do we know that building a relationship is a key point to accessing future products like Bento Box?
Are we precluded right now from being integrated from future Sushi/Yearn products? (We are already integrated in CREAM which is part of the Yearn ecosystem.)
Let’s get clarity on these points. It would be great if those that believe in an integration with Sushi to step back, consider the problem, and consider multiple avenues to that potential solution.
@LemonadeAlpha - I really like the logic reading over your posts. The conversation and engagement on this thread is great.
I agree Liquidity Mining is not necessarily the best bang for buck right now in terms of growing AUM but it is assisting in DPI retention. CEX listings will move the dial more. I really like the front of the funnel targeting, going after the big wins. However, we can not look past what is happening within the ecosystem. Small holders are being priced out of the market by gas fees on layer 1. From the survey we did a few months back now, we know DPI has a lot of small holders. This is an entire segment of the market.
When DPI is a >$1B AUM product, it will be spread through out the ecosystem. When the fund is this >1B being across multiple DEX will be a plus. None of our competitors have 2 decent liquidity pools. We know as time goes on, the direction of each DEX is diverging. An example: different layer 2 solution incorporation could mean different timelines to achieving L2 to L2 transactions. I agree with @BigSky7 point on relationship building is very true. To @Mringz point we don’t have data to know how sticky our client base is with out incentives. A natural progression would be re-allocate incentives and see what capital follows it. I with to echo @verto0912 concluding comments and extend on it to go beyond 60 days. I want to see a full program 60-90 days setup at the start and left to run its course.
The data around the UniSwap DPI-ETH LP staking contract is not great. We are seeing the number of users depositing funds into the staking contract fall and we are experiencing net outflow of funds. Combined with the INDEX price movements of Jan/Feb this meant APY was still relatively high, great work on this @overanalyser. Reducing the INDEX token allocation is likely to further reduce the number of users interacting with that contract. With gas prices as they are, any person with a small balance is unlikely to deposit fresh capital with a short time horizon. This could act to keep capital in the staking contract, so good retention but is unlikely to bring new users or have any positive affect on AUM. Charts can be found in IIP-014 chat on discord.
Given the demographics of our client, skewed to smaller holders, and gas costs the way they are, it seems logical to me to reallocate funding from the UniSwap LP staking contract to the Loopring. This supports smaller holders and enables us to target that segment of the market. If we stop UniSwap LP incentives and increase/maintain Loopring incentives then we will learn more about our holders, like who is in it for yield. I also don’t see our competitors on Layer 2, it is good being the first mover and leader.
@LemonadeAlpha The Uni LP has not increased in liquidity with decreasing incentives. The number of DPI/ETH units are WAY down from November. There is about the same $ amount in liquidity as there was in November, even after a 3-4x increase in the market. So liquidity has actually cratered from what the rest of the market has done in that time, following reduction in index incentives.
And I will admit my bias, I am an ETH/DPI LP, but it we stop incentives, I’ll have to pull liquidity because INDEX rewards are not covering IL + gas fees, and I presume others will follow. Also, I would push back at the statement that INDEX is just for paying other people to dump. I still have all my index.
If we do stop incentives on the UNI LP, I think it would be prudent to give a heads up so smaller LPs don’t get kicked in the teeth.
This chart shows the quantity of DPI-ETH LP tokens in existence and the quantity deposited in the staking contract yielding INDEX tokens.
After genesis liquidity mining incentives stopped - The quantity of DPI-ETH LP tokens in existence appears to be very correlated to the number of LP tokens deposited into the staking contract. Perhaps the liquidity on UniSwap is more correlated to liquidity mining incentives than we thought.
This chart shows the daily annualised yield (APY) on the staking contract. It takes into consideration the value the DPI-ETH LP token and the INDEX token price. The data is daily from early December.
First of all, I would like to congratulate all of us for this very valuable discussion.
I have certainly not understood everything and do not want to comment on everything.
But I would like to make the following comment: I will certainly keep my DPI/ETH uni-pool even after the index incentive, because 1) I take the 0.3% fees rewards and 2) because I am deeply convinced of both products (ETH, DPI).
It would also be interesting to know how long DPI holders hold them on average. Out of conviction or out of compulsion (gas).
I can also imagine that many larger investors also like to chase the incentives. First uniswap pool index rewards. Then loopring. Now sushi. Tomorrow… so I have my doubts as to how far incentives will bring new buyers into the DPI. I have always kept my uniswap pool, certainly for convenience, for cost reasons (gas), and connectedness (uniswap, AirDrop).
Big data is certainly always right and important to use, if you can get that data and interpret it right.
User stories, like mine, are very subjective and individual, they help us as personalities, which we are ourselves, to understand others better. And then to draw the right conclusions from them. I have no doubt that this happens here at the Coop.
Regarding shepherding our resources, it obviously makes sense to think about what we spend our time, energy and financial reserves on. In some ways, I could argue that INDEX tokens are our most abundant resource, and so justify less scrutiny as to expenditure (with Dev time being the most scares resource we have and so needing the strongest case for each programme).
If we look at the problems we have been trying to solve:
Distribute 9% of INDEX in a neutral way - - Initial liquidity mining
Maintain $30 M of liquidity for large purchases -Dec, Jan and Feb liquidity mining.
Bootstrap liquidity on L2 - - Loopring Initial liquidity mining.
For the main Uniswap pool, the pool has averaged 14% fees over the last 30 days (croco.finance) and the INDEX rewards are generating 15% (see above).
So the market is settling with about 30% APY for the ETH:DPI pool. If we stop liquidity mining, we can expect rational (!) LP’s to move out of the pool until yield from fees approaches 30%. i.e. a 50% reduction in the number of LP tokens which will give us ~$25 M USD in the pool for large trades.
Given the imminent release of the exchange issuance contracts to cover large trades, we have effectively managed to bridge the gap to the point where trading fees sustain the liquidity. i.e. we can stop uniswap liquidity mining and goto 100% unincentivised DPI.
Great debate guys, love all the data and seeing your thought processes. I’m not going to give an opinion on overall LM because you all are way deeper in the data than me but i do have an alternate solution to propose.
So far you all are basically proposing the same thing - selling assets (INDEX) for liquidity - just on different platforms (uni vs sushi) with different numbers.
What if we sell debt for liquidity?
Instead of paying $1M / month in straight INDEX, what if we take out an uncollateralized stablecoin loan on Iron Bank and use that to add liquidity as the Coop? This is similar to Aave Credit Delegation except we dont need a counterparty, its purely smart contracts. Lets assume a pretty bad interest rate of 30% APY (current is ~3%), we can take out a $40M stablecoin loan, add that to DPI/ETH pool, and still be paying the same $1M / month as today with almost the exact same amount of liquidity ($40M vs $48M now).
I call this Liquidity Minting instead of Liquidity Mining because it’s coming from “nothing” instead of extracting resources.
We can even kill many birds with one stone.
Put this liquidity in Sushi to form that partnership
Farm SUSHI with DPI/ETH to payback interest over time
Iron Bank might require us to purchase insurance instead of putting up collateral which is something thats been discussed already
This plan might be more suited to bootstrapping new products like CGI than established products like DPI. Curious to hear your guys thoughts on it. This somewhat overlaps with Smart Treasury discussion (cc @DarkForestCapital) so might want to move this idea to a separate post.
Less expensive than current programs on a spent per of liquidity basis (depending on variable interest rates)
Less debating, governance overhead, and comms about LM programs
Hey all. What a great discussion above, I can’t keep up with all the details, but I did have a nice call with a few owls last week or so, where we discussed listing DPI-ETH orderbook pair, in addition to the DPI-ETH AMM pool. That is up!
Orderbooks have big benefits vs AMMs (contrary to popular belief these days :).
Drawing professional traders/ algorithmic market makers
Allow limit orders from takers
Just allowing a general different trading behaviour
Orderbooks are actually Loopring’s roots. And on Layer 2, orderbook trading is beautiful, emulating CEX experience.
However, they are more difficult to bootstrap. It requires an active MM to run strategies on the books via API, not just passive LPs depositing. (Can be done manually of course, just not as competitively).
We have our orderbook liquidity mining program, which incentivizes users to place tight resting limit orders on the books. Relayer takes snapshots of the books every x minutes, and users with orders within the predefined spread (say, 1.5%) accrue rewards. From your (Index Coop) POV, it’s the same: allocate INDEX rewards to be paid to these MMs. Read the details here (this is from a long time ago, but rules still stand).
I propose for the coop the allocate rewards to a DPI-ETH orderbook liquidity campaign. This will bring in more ‘sophisticated’ or at least traditional style traders to MM DPI, and also give takers a different experience as well. As @overanalyser mentioned, soon AMM and orderbook liquidity will be unified from a taker’s POV, which will be awesome. I would propose doing both AMM and orderbook liquidity incentives for the coming round (or the round after, since it may be too late for coming round in a few days for you to reach consensus). This will increase overall liquidity, and invite yet another type of market participant, arbers, to reconcile the prices on books vs pools.
One thing to add: Loopring Exchange is integrated into MM software like Hummingbot (which just hooks into our orderbooks, not pools). This means that a DPI-ETH LM campaign could draw in their community of hundreds (or thousands?) of programmatic market makers. https://docs.hummingbot.io/exchange-connectors/loopring/
I’m not sure retention via liquidity mining is the kind we want to target. We’ve rallied around this idea of unincentivized AUM because its the more sustainable path though it may be (short-term) harder.
I agree that getting to L2 makes sense in terms of capturing smaller-sized on-chain demand. I worry that we are overlooking a few things:
Loopring costs money and mental overhead to onboard onto
Loopring is a small ecosystem rn
You can already trade $10k for under 1%
When DPI is a >$1B AUM product, it will be spread through out the ecosystem. When the fund is this >1B being across multiple DEX will be a plus.
Agree here, unclear to me why they need to be incentivized, though. These pools have sprung up on their own merit. Sushi incentivized a pool because DPI/ETH is one of the most consistently liquid and traded pools on Uniswap. They have the pressing need for this liquidity on Sushi; we as the Coop don’t.
To @Mringz point we don’t have data to know how sticky our client base is without incentives.
If our client base isn’t sticky without incentives, then our business model is broken. And, I think there is some data. We have seen that as our lm rewards have slowed down, retention has increased. We have seen our base of AUM and holders outside of the liquidity pool ramp up in recent months.
@uncle We have been targetting against a dollar amount entry so I’m not sure that DPI unit liquidity is the best marker. I could have the wrong framework.
Also, those LM incentives were 4.5% of the total INDEX supply, per month. It couldn’t last. I am advocating for a tapering down of incentives in Uni, not to remove them
To me, this whole discussion is trying to solve a liquidity problem for the customer that doesn’t exist. If customers want to buy DPI on l2, they can. If they want to buy DPI on l1, they can.
When I look at these:
Kickstarting AUM - Essential for launch
Distributing INDEX tokens in a decentralised way - Essential for becoming a decentralised organisation.
Marketing our product
Defending AUM against competitors
Reducing slippage on large trdes
Building links with AMM communities
Expanding the visible storefront for our products (@BigSky7 ).
Only 2 are about the customer: slippage and storefront. Slippage is not going to be improved by the proposed Sushi rewards. The storefront argument isn’t strong enough to justify rewards atm. if we had some kind of data suggesting there is not significant overlap between sushi and uni users I think it would be more compelling.
To me a worthy goal seems to be to increase liquidity on Loopring such that a user can make a $20k trade with x% slippage. I would still posit that the above questions about making a sustained bet on Loopring at this point in the race are still worthwhile to answer.
All, During todays call it was suggested that we should split out the different liquidity mining programmes (UNiswap for maintaining large liquidity, Sushiswap to build bridges and Loopring to ensure we have a foot hold on L2).
I don’t think that is possible, the available liquidity will react to the incentives that are in place. We need one single strategy that covers all (with some flexibility over the longer term).
That’s why I made the proposal for an overall reduction in INDEX, but spreading it over 4 pools (UNI, Sushi, Loopring AMM and Loopring orderbook) and committing to 60 days with a planned reduction in $ value.
However, it’s was clear that the proposal didn’t attract widespread support and some fundamental disagreements.
As a result, I think the best approach would be to end all incentives and let the market find it’s natural balance. Assuming that current trading volumes continue, an APY of 30% should result in over $30M of liquidity.
If there is a sudden move to L2 (optimism etc, then I would expect the L2 exchanges to be courting liquidity - more UNI rewards?)
With the rollout (and publicizing) of exchange issuance, we should be able to serve all our users…
I would agree with @Matthew_Graham that we should have a strategy for LM’ing new products. But this is really something the methodologies should be outlining as part of the product proposal.
Due to the way that loopring incentives work, there is a cut off on being included in the next round.
So we would need to act today if we want to be included. At the moment I’m seeing 63% in favour of continuing the current rewards (1,000 INDEX over 14 days = 72 per day) and 50% preferring not to give any incentives to the order book market makers.
As such, I’ll be suggesting that we allocate funds (without an IIP / further vote on this basis).
For Uniswap and Sushiswap, we control the staking contracts and the current programme has 4 or 5 days to run so we can let the rote run another day or two before doing an IIP for a single combination.