ARC: Extend AAVE Liquidity Mining Rewards

Over a year, it is 1/4 of the ecosystem reserve, but it’s actually only 1/20 of the total market supply. If the goal is decentralization of the protocol, this is actually pretty slow in my opinion.

Isn’t the core value offered by AAVE realized by the process of borrowing though? I think pushing users to a higher health factor is the wrong approach to prevent things like insolvency to the protocol. The user should feel comfortable taking as much of a risk as the AAVE protocol lets them which means that real solutions need to focus on getting positions liquidated on time and seting liquidation parameters appropriately.

You point out recursive borrowing, but I’m not really sure why that would be a bad thing. I think it’s wonderful that people can take whale-like positions with small amounts of capital as long as the liquidation mechanics are solid. Overall, I think it should be AAVE’s goal to invite users to play with as much liquidity as the protocol allows. If we think things are too risky, AAVE should adjust the liquidation parameters.

I think this is a little disingenuous… This was posted early in the thread that you linked in your post. Or perhaps you’re asking for it to be posted in this thread as well. In which case I can agree. There’s no reasoning in this thread.

I mostly agree with everything else you wrote though. I think these proposals could align a bit more with community desires, especially considering that there are a few threads that requested LM rewards with pretty good support but aren’t mentioned in this proposal at all. TUSD SUSD

1 Like

I don’t believe this is really true, the Safety Module is backed by users staking their AAVE (currently $746M of AAVE has been staked, which 30% ($233M) can be used during a downfall event) and AAVE/ETH Balancer pool shares (currently $260M of AAVE and $65M of ETH, where 30% ($97.5M) can be used during a downfall event). This means the Safety Module already has $330M in funds that can be used to recollateralize the system if any significant undercollateralized loans are created on the platform (which has never happened but the SM provides additional protection). Even if this entire budget was used up, the Aave protocol can issuance new AAVE tokens to cover any holes (just like MKR in the MakerDAO protocol like during black thursday).

“In case the SM is not able to cover all of the deficit incurred, the Protocol Governance can trigger an ad-hoc Recovery Issuance event. In such a scenario, new AAVE is issued and sold in an open auction for market price prioritizing the Backstop Module.” - Safety Module Documentation

As shown above, LM programs do not reduce the Safety Module budget and the reason we have seen a decrease in the value of the AAVE token cannot really be attributed to the LM program (at least I have seen no evidence of such), but more than likely because the entire crypto market has crashed since the LM program started (BTC has gone from $60k to $30k), all cryptos are highly correlated and AAVE has dropped a similar amount.

Aave has become the prominent money market precisely because it has a liquidity mining program that incentivizes adoption. If the community wants it to keep it that way, it needs to stay competitive against other money markets like Compound that have such LM programs. Seeing success and cutting the LM program in my mind is like a patient who stops taking their anti-biotics because they feel better and thinks they don’t need them anymore. Without such a program, TVL will drop and this position will be lost against competitors like Compound and even Cream when they launch a LM program.

Incentivizing low capital efficiency of loans results in lower yields for lenders, especially for larger positions (seems people would be more likely to borrow less rather than deposit more collateral). The Aave system has proven it can handle market crashes and keep the protocol collateralized. I don’t think this is really necessary as users should be able to open positions within their risk tolerance and not be disincentived from opening capital efficient positions that the protocol can handle, then would likely switch to other platforms like Compound that don’t have such a dynamic.

Money markets are a two sided marketplace. Incentivizing borrowing increases yield for lenders (particularly stablecoins), which incentivizes more deposits to capture these yields (e.g. from users and protocols like Yearn), thereby lowering rates for borrowers, increasing borrowing activity, leading to higher lending rates, and repeat. This creates a virtuous cycle which cannot be achieved by only incentivizing deposits, which would crush yields, disincentivizing deposits, not ideal for a two-sided marketplace.

To incentivize adoption and usage, as well as get users to switch from Aave v1 and other markets like Compound over to Aave. Goals were defined in the original LM program proposal, which haven’t changed, which is likely why it isn’t in this proposal. Quantifying such metrics would indeed help but I think the goals are clear: market dominance.

If Aave wants to stay competitive against other markets like Compound who offers LM rewards for such collateral, then yes. LM rewards incentivize people to switch their loan positions from other platforms to Aave and incentivizes people who don’t use money markets to start using them. As a two sided marketplace, deposits and borrows should both be incentivized.

Interesting idea, but would increase governance overhead proportionally. I think this is a bit of scope of this proposal but would be interesting to hear this fleshed out in another proposal.

I feel this ignores the all of the people who have publicly supported such LM programs like myself and others, who see the competitive advantages that such LM programs provide. This proposal wasn’t created by the Aave team but a community member, so not sure where this assumption comes from. This proposal was created to get community feedback such as yours.

I agree with many of your other points regarding adding other tokens to the program and recursive borrowing (although this doesn’t really introduce risks, just is capital not being used efficiently, not sure what the solution here would be as we need to incentivize both sides of a two sided marketplace to get that virtuous cycle that has lead to Aave becoming the #1 protocol by TVL).

I would be interested in seeing some of the metrics quantified from the data we now have from the previous LM program like the retention rate and number/percentage of people switching platforms versus using money markets for the first time. In general, a LM program enshrines Aave’s position as the largest money market which is more important during its early life (which justifies such capital usage) compared than later when it has a greater network effect through liquidity and integrations.

3 Likes

Loving the discussion here!

I’m mostly in accordance with @pakim249 and @ChainLinkGod on the topic. I think that the current Liquidity Mining program has been essential to drive Aave at the top of the rankings across liquidity protocols, and it’s too early to stop now. Especially as all the farming opportunities dried up, liquidity now concentrates, of course, in protocols that provide long term, valuable governance power redistribution (that’s why curve climbed the TVL rankings so quickly).
What i personally care about is long term sustainability of the protocol and ecosystem growth. I see liquidity mining propaedeutic to both (growth of the ecosystem reserve thanks to the increased outstanding debt and leading market metrics).
I usually avoid talking about price but since it looks like is the concern of many, will give my personal opinion - i don’t see strong correlation between the current AAVE price and the LM program. Aave reached the ATH way after the LM program had begun and we can’t ignore the fact that the market dropped 50+% and there is strong correlation between assets.

The aave reserve is on it’s way to collect 30M+ in stablecoins by the end of the year. That’s a multiyear runaway for protocol development and ecosystem growth, and even at the current market APY of aroun 2% it would generate 600+K/year out of interest only (the stablecoins collected by the aave reserve are automatically deposited in Aave).

As a community we need to aim long term, the future of finance will play out in the next 10 years and the aave protocol will be the leader.

5 Likes

@ChainLinkGod regarding the LINK emission, it is true that currently compound offers 2% LM APY on link, but it must be noted that it only has 60M worth of LINK deposited. Aave has more than 5 times that amount, if the same deposits would be in compound it would have a comparable LM APY.

2 Likes

I also support @pakim249 and @ChainLinkGod viewpoint. The liquidity program incentive is what made me deposit and use AAVE for the first time. I compared the APR of different platforms and also different strategy to accumulate more ETH or discover another great token to hold. AAVE offered a similar return then others AND had in my opinion a better long term vision and community lead projects. So I deposited here and I plan to earn AAVE and keep every single AAVE earned.

It reminds me of Paypal 20$ referral program when they first started, it seemed expensive at the time but it got them their first few million users and where they are today. So in the light of my personal experience, the LP program seems successful in creating new AAVE long term holder and onboarding new user of the platform.

2 Likes

Glad to see the message stirred some reactions and was not ignored this time.

I want to focus on clarifying an absolutely critical point that seems deeply misunderstood in this discussion, what I call the dilution of the safety module budget.

Ignore AAVE price going down, since indeed it’s not directly correlated to the liquidity mining. Even consider that the deposit to the safety module stayed the same (they lowered).

The safety module is to be used in case of critical failure, so we measure its effective coverage against the borrow side of Aave. I did the maths in February when I published my risk assessment article:

Let’s redo the maths for today’s situation, all in USD:

  • Safety Module Total Deposit = $ 1.05B
  • Budget that can be used (30%) = $350M
  • Total Borrows on V1 + V2 = 56 M + 6.3 B
  • Effective coverage offered by the safety module = 16.6%, roughly 3.5x times less that the figure observed in February.

This is precisely what I meant by “dilution of the safety module budget”. Even if the budget stays the same, it got diluted by the massive increase in borrowing. This is the main reason why I’m worried about borrowing side incentives as it’s essentially driving a reduction of efficiency of the insurance budget — especially if nothing is done to stimulate deposits to the SM simultaneously.

To clarify the rest of my points, allow me to restate that I do not oppose liquidity mining per se, and I recognized the results of the first wave. What I oppose is the untargeted dimension (both sides, all major tokens) & the excessive budget, amongst other concerns.

10 Likes

Thanks for all the great input! Aligned with @ChainLinkGod, @pakim249, and @Emilio.

@ChainLinkGod Fair points on the allocation towards the LINK market. 4 stkAAVE is relatively low, but this can also be seen as a starting point before scaling rewards up further. LINK would be the first asset added to the program. Once added, we can monitor borrow/lend activity and further refine. On new assets in general, the risk DAO may be best positioned to add/remove assets in a scalable process.

Given the liquidity mining program will end this week, we are supportive of extending the current program by 4 weeks to give the community more time to deliberate. There is a lot to unpack on revising the distribution. By extending the program for a short period of time, we avoid a sudden pause to the rewards while giving the community enough time to revise the distribution (i.e add/remove assets, adjust the allocation across assets).

5 Likes

In favor of extending to allow the community to have a more articulated discussion.

1 Like

Very interesting community discussion so far. I have to say that liquidity mining has been quite interesting experiment. Would also favour on continuing the liquidity miming program for 4 weeks to give time for the community to reach wider consensus.

8 Likes

In my opinion, there are some valid and constructive points in the discussion and probably the proposal should be refined a bit.
It could be a good idea to start a series of Snapshot votes proposing different models, considering as mentioned the option of liquidity mining on others assets like TUSD and SUSD, to really understand how is the community sentiment about it. Taking into account the flexibility of Snapshot, sounds the most natural next step.
In what rewards the temporary extension, I’m also in favour, as:

  1. The current liquidity mining is quite neutral and can be sustained for 4 weeks by the Aave DAO treasury without problem.
  2. It is quite important for the community to take a really well reasoned decision on the topic. Those 4 weeks will give room for it.
6 Likes

Agreed with the extension because we do not have the time to produce a well crafted counter proposal before the end of the vote.

However, let the community realize that this extension of one month is essentially a one month enforcement of the revised proposal (which extrapolate the initial 3 months proposal and its budget over the course of a year).

It’s not a satisfying decision, and it’s taken with a light consensus, aka a failure of governance. This proposal should have been submitted much earlier (especially considering how rushed it is) to allow ample time for comments and discussions.

I’m willing to help lead the effort for the revised proposal, starting right after the end of EthCC. Other people who commented on this post reached out to help too already, feel free to do it too if you’re interested in getting involved.

7 Likes

i don’t agree that is a failure of governance, the community needs to take responsibility when discussing a topic, @Anjan-ParaFi did in this case but others could have done earlier, everybody knew the 3 months period was ending. Agree though that this topic should be discussed more thoroughly. I’m fully in support of avoiding a collapse of the fundamental market metrics (TVL and outstanding debt) at this moment of market turmoil. The proposal needs to be put on chain though, so if others community members don’t agree they can always vote against

11 Likes

This is Ratan, President of Blockchain at Berkeley.

Very insightful to see the discussion in the community here - we agree with @Anjan-ParaFi in that we don’t see any significant drawbacks in extending the program by a month so the community can come to a consensus. Overall, we’re supportive of extending the LM program in a way that incentivizes more liquidity and usage to come into the ecosystem, as we’re still at a very early point in Aave’s potential growth. As @ChainLinkGod has mentioned, we need to enable Aave to remain competitive to LPs and we see LM as the most effective way to do so.

2 Likes

I believe it’d be great to all agree on defining the safety ratio to maintain. Is it 16%? Is it 35%?
This will help us make accurate decisions for the next steps.

3 Likes

I personally don’t think it’s necessarily healthy to put much weight on the safety module ratio, largely because the safety module is for the most part comprised of AAVE tokens. AAVE correlates highly with its popular collateral assets, and correlates weakly with its popular debt assets. This is very bad in the case of a shortfall event which would dilute the safety module to unpredictable levels. This correlation also spills over into the relationship between AAVE and the safety ratio, which means that this ratio can fluctuate as meaninglessly as bitcoin.

And so I argue that the safety module TVL should not be viewed as a measure of security of the protocol. Rather, it should be viewed as a measure of confidence in the protocol, and a penalty mechanism for irresponsible governance. Because TVL simply fails as a metric for the ability of AAVE to cover a shortfall event.

I think a good start to protecting AAVE from shortfall events was the collection of a portion of reserve interest for the treasury, but this isn’t formally incorporated into the risk framework of AAVE. Another option is to simply outsource the risk to another protocol so that the focus of governance can be narrowed and AAVE can have the cost of their risk quantified by a third party. Of course, the safest and most capital efficient option is to simply set parameters to never suffer a serious shortfall event though. Which is all to say, I don’t think maintaining the safety module ratio should get in the way of a liquidity mining proposal.

What I wanted to stress out is that ratio lowered by a factor of 3 to 4x from almost 60% pre-LM to roughly 15% nowadays.

I’m not an actuary, but I’m already uncomfortable at 15%: it means that if we suffer a critical issue affecting more than 15% of the total borrow, the Safety Module budget will be insufficient to compensate for the loss — and we’re not even accounting for the massive and to be expected drop in price in AAVE token if that happened.

We need indeed more diversification of assets in the safety module but that’s another discussion.

Regarding the liquidity mining proposal, I’m finalizing a new proposal including several adjustments to make sure the LM plan is sustainable and synergistic with Aave’s best long term interest. I hope it will provide a better base for the discussion, and incorporates most of the most recurring feedback shared here. I’ll share as soon as possible, probably early next week.

@TokenBrice i don’t think we should stress about the ratio safety capital/outstanding debt, if the protocol grows organically (even without liquidity mining) that ratio is gonna shrink no matter what. Outstanding debt is always an excellent metric for the protocol, in terms of income generated and usage, therefore we should always aim to maximize it

1 Like

Your point is correct, indeed the coverage ratio of the safety module will shrink as outstanding debt grows if no additional deposits are made.

Yet I would contest the second part:

Outstanding debt is always an excellent metric for the protocol

Yet, using outstanding debt as a metric lead us for instance to regard single-asset-based recursive borrow-deposit loops as positive for the protocol, are they really?

So far, Aave is the only major money market that has not suffered a critical failure resulting in loss of funds for users. I’d like it to stay as such, and even better — unlike other money markets — be safe enough that we could reimburse affected users if a loss-of-funds event was to occur.

I believe it’s critical to keep growing Aave’s credibility and reputation within the space. What we have is both the safest and most innovative major money market, we need to tread carefully to keep it as such.

With this in mind, even if it’s not one of the main parameters in the decision, I believe the effective coverage offered by the safety module is still a relevant factor to look at while taking decisions related to the growth of Aave.

I’ve gathered the feedback to establish what a counter-proposal incorporating them would look like.


This proposal expands on the first one published by Anjan to incorporate recurring community feedbacks shared in the discussion regarding the first proposal and now its extension.

Hinder recursive behavior

They currently account for almost half of the platform TVL while providing little utility to the protocol. I personally would like us to stop incentivizing borrowing, but it’s not a view shared by the broader community.

Instead, I’d suggest we move forward with Alex’s suggestion: factor the LM rewards on the basis of the net deposit per asset (= total deposited on asset X - total borrowed on asset X). This mechanism will ensure that users looping on USDC for instance are not extracting a neat APY despite providing no value to the platform outside of the fees collected.

Review the supported asset list & weighting

There are several mismatches in the assets suggested and their balance. First, I don’t think the LINK budget is appropriate considering they have been amongst the top supporters and users of Aave since launch.

→ Raise LINK budget. Adjust distribution 95% supply / 5% borrow.

Then, I’d also like us to question and discuss and eventually vote on the inclusion of GUSD in the liquidity mining program. I don’t think this discussion occurred during the exchanges for the first proposal.

→ Vote on GUSD inclusion & adjust the proposal accordingly. If GUSD is removed, we could simply redirect its budget to LINK, fixing two issues in a single vote!

)

Finally, if we include LINK in the mining, it’s only natural to consider also adding AAVE who is also historically one of the top collaterals used on the platform.

Account for shorter-term LM campaigns

The LM budget is voted on a yearly basis. While this is amazing to adjust the parameters and finetune the design thinking, it makes it hard to adjust it dynamically. To achieve flexibility, I suggest we reserve a share of the total LM budget to incentivize new assets or markets, 50% for instance.

Considering that such decisions are timely, I’d see this share of the budget not controlled by the community or the DAO. Instead, it should be governed by an executive committee. Their mission would be to best allocate this budget to help kickstart new markets.

Budget distribution overview

The yearly program would be funded with a total of 600K to 803K AAVE (depending on the results of the snapshot votes).

50% of which are set aside to be allocated by the Liquidity Mining Executive Committee to help kickstart new markets and assets (such as Aave’s implementation on rollups).

The remaining 50% to be distributed as follows (assuming the adoption of all proposed adjustments). It rebalances slightly in favor of ETH/BTC/LINK/AAVE to account for the new calculation method accounting for net deposits (or borrows) per assets.

Asset Anjan's Proposed Allocation Revised Allocation Split between lenders and borrowers
DAI 15.53% 19.90% 50/50
USDC 62.23% 31.00% 50/50
USDT 17.09% 18.25% 50/50
GUSD 0.42% 0.00% 50/50
WBTC 1.93% 12.00% 95/5
ETH 2.62% 14.50% 95/5
LINK 0.18% 2.10% 95/5
AAVE 0.00% 2.25% 95/5

Total

100.00%

100.00%

Soft-incentivizing farmed StkAave holding

Since our LM program distributes tokens as StkAAVE, I think it would be wasteful to not be intentful about designing a plan that optimizes for their retention as StkAave. Instead of penalizing people who withdraw early, I’d like us to go the other way around by rewarding users who farmed StkAave, claimed them, and held them as such for several months.

We could for instance add a long-term twist to the farming by saving a share of the budget to be distributed as an airdrop to users who did not withdraw the StkAAVE they received from the liquidity mining program.

Envisioned Snapshot votes

We have yet to reach consensus on several parameters, so here are the snapshot votes I’d like the community to take part in:

  1. Liquidity Mining Budget: What should be the plan yearly budget? A - 803k stkAAVE (~27% of the ecosystem reserve) B- ~600K StkAAVE (~20% ER)
  2. Should we switch the LM reward calculation to the basis of net deposit per asset? A - Yes. B - Keep the current mechanism
  3. GUSD was included in the first liquidity mining and then intended to be renewed, but is that relevant: should we keep expanding StkAave to include GUSD in the liquidity mining program? A - Yes. B - No.
  4. Should we also include the AAVE token into the liquidity mining program? A - Yes. B - No.
5 Likes

Recursive behavior
Is the recursive behavior actually bad for the protocol? I don’t do it myself, as I prefer to use the protocol in a different manner, but is it actually hurting anything? I don’t claim to have as deep an understanding of this as you, so please take this with a grain of salt, but if I deposit x funds, and borrow y funds, and remove y from the protocol to use elsewhere, it seems ostensibly worse than if I deposit x, borrow y, and deposit y, as the funds never actually leave AAVE.

Incentivizing borrowing
Isn’t borrowing good for AAVE? I’d think that many people come to the protocol specifically for the incentivized borrowing.

Executive committee controlling 50% of LM rewards
Respectfully, I think a significant portion of the community (myself included) would be very opposed to the idea of an executive committee controlling 50% of the LM budget. It feels like centralization.

Incentivized stkAAVE holding
I really like this idea. Would it be possible to work that into the individual rewards rates? For instance, let’s say my WBTC LM reward rate is 1.25%, could it be changed to 1% + a .25% bonus done as an airdrop after x months/1 year? For those who removed their stkAAVE earlier, they aren’t so much getting a penalty as wasting potential, and the uncollected rewards from these people could be returned to the safety module.

5 Likes