[TEMP CHECK] Raising the Safety Module’s Slashing Percentage

[XenophonLabs]

Title: [TEMP CHECK] Raising the Safety Module’s Slashing Percentage

Authors: @tncintra & @maxholloway

Created at: 2023-07-27

Summary

• Aave pays between 39,000 and 56,000 AAVE (approximately $2.75M-$3.9M in USD terms) annually in cost of capital to stakers, for funds it cannot slash.
• Increasing the slashing percentage to 100% would save Aave millions of dollars per year, and we anticipate it would have neutral-to-positive effects for the module’s insurance power.
• We recommend:
  • Raise the slashing percentage on stkAAVE from 30% to 60%, and accordingly, lower emissions. After observing impact, we either raise the slashing percentage to 100%, or revert our changes.
  • Set the slashing percentage on newly proposed pools to 100%.
• We summarize our research report below, as well as a suggested rollout plan for the new changes, and request feedback from the community on our analysis.
• Our analysis and recommendations are supplementary to most of the work done by TokenLogic and Llama on the safety module. They discuss the slashing percentage in Part 2, and emissions in Part 4 of their Safety Module upgrade.

Report Link: https://xenophonlabs.com/papers/aave_slashing_percentage.pdf

Motivation

The slashing percentage determines the percentage of safety module deposits that can be slashed by governance. It currently sits at 30%. In a sponsored by the Aave Grants DAO, we provide a quantitative analysis for why a slashing percentage below 100% is strictly capitally inefficient. At a slashing percentage of 30%, Aave pays between 39,000 and 56,000 AAVE (approximately $2.75M-$3.9M in USD terms) annually in cost of capital to stakers, for funds it cannot access.

Following this forum post, we intend to propose raising the slashing percentage on all Aave safety module pools, including the existing stkAAVE pool, and the new pools proposed by Llama and TokenLogic. We expect these proposals will significantly reduce the current costs of the Safety Module, and make any future pools more capital efficient. Furthermore, they will require negligible technical uplift.

However, before moving to a more formal proposal, we would first like to request feedback from the community on our analysis. We are more than happy to discuss our assumptions, and provide further justification for why we believe this is the correct decision for the DAO.

Capital inefficiency measures the reduction in emissions we can afford by having a higher slashing percentage, while keeping the module’s insurance power constant.

Modeling

In theory, stakers demand a lower APR when the slashing percentage is lower, since they have less capital at risk. However, the insurance power of the safety module is also reduced by smaller slashing percentages. Using a simple risk-return model, we show that the capital efficiency of the module (i.e., the insurance per unit of AAVE emitted, which we denote as R) increases as we increase the module’s slashing percentage.

What makes up the APR?

A standard formulation for understanding the APR of an investment is as some risk free rate r₀, plus some risk premium r_prem.

APR = r₀ + r_prem

Let’s further break the risk premium into its slashing component (the premium demanded due to slashing risk), and a non-slashing component (which might include, for example, a liquidity premium):

r_prem = p * r_slash + r_other

where p denotes the slashing percentage. Some might say a “risk free rate” does not exist per se in DeFi; this does not affect our conclusions in any way. In fact, the actual values or r₀, r_slash, and r_other do not impact our recommendation to raise the slashing percentage.

Modeling Capital Efficiency

It follows that the cost to the protocol is R = TVL * APR = TVL * (r₀ + r_other + p * r_slash), whereas the insurance power of the protocol is TVL * p. The ratio of insurance power to cost, which we denote as capital efficiency, is expressed as:

C(p) = (p * TVL) ÷ R = p ÷ (r₀ + r_other + p * r_slash)

While we can take the derivative of C(p) and show that it strictly increases for p in [0, 1] (which we do in the report), we can observe this easily by plotting C(p):

Intuitively: when we lower p we scale the module’s insurance power by a factor of 1, but we reduce the module’s costs by a factor of less than 1. This means lowering p does reduce costs, but it reduces the insurance power even more! Notice that this analysis was entirely agnostic of the token[s] held in the pool.

Specification

So far, this exercise has been largely theoretical; we made some assumptions, which we challenged and documented in Section 2 of our report, and drew conclusions from the modeling. These economic models on their own are not enough to know the outcome with certainty. Therefore, we suggest a gradual rollout over the next few months where we raise the slashing percentage, gather data, and decide whether to continue raising it or revert back to 30%.

Proposal Overview

Acknowledging that economic models are imperfect, we recommend doubling the slashing percentage from 30% to 60% on the stkAAVE pool, and waiting some time to gather data on whether our claims hold true. If the module’s capital efficiency increases as expected, we may raise the slashing percentage to 100%. Otherwise, we may revert it back to 30% if desired.

Higher capital efficiency means that the DAO can reduce the emissions to the safety module, while keeping the insurance coverage constant, or perhaps even increasing it. For this reason, we recommended in our report to raise the slashing percentage to 60% on the stkAAVE pool, and lower emissions by 80 AAVE/day, from 550 to 470. Using our model, we show that this would retain at least as much insurance power in the module, despite spending less on rewards. This would result in annual savings of 29200 AAVE, or ~$2M USD. We later discuss how this is affected by recent snapshots from TokenLogic and Llama, where they proposed lowering emissions from 550 to 200 AAVE/day.

These are based on 150 days of historical data extracted from this Flipside dashboard and deposit rates from Aavescan. Our data extraction is discussed in the Appendix of our report.

Proposal Considerations

Llama and TokenLogic have already passed a snapshot, along with several other snapshots, to lower emissions to the stkAAVE pool to 200 AAVE/day. This impacts our ability to measure the impact of this recommendation. Measuring impact is crucial: it allows us to test whether higher slashing percentages actually lead to improvements in capital efficiency, allowing us to confidently set slashing percentages at 100%.

Furthermore, the safety module currently sits in limbo between successful snapshots and putting forward AIPs to implement them. Given that our research intersects with an existing snapshot, and to avoid confusion in the DAO, we have chosen to delay our proposal and instead request feedback on the analysis. We are working with several community members on how to best structure a proposal to increase the slashing percentage and accordingly lower token emissions. We have found there is significant interest in experimenting with a higher slashing percentage, and we request feedback from community members on when and how we may best put forward our proposal. The details are outlined below.

Other Considerations

Inelasticity

An assumption we make in our modeling is that stakers are elastic to changes in risk and yield. If stakers are fully elastic, then we expect them to react to changes in risk and yield as we have modeled. However, staking AAVE in the safety module is currently the only venue for earning yield on AAVE while also retaining voting rights. Lack of alternative investments that retain voting power likely makes stakers inelastic: an increase in risk does not lead to stakers withdrawing a portion of their stake. Therefore, it is likely that we observe an even greater improvement in capital efficiency than we estimate in our model.

AAVE Price

An important consideration is how the slashing percentage affects AAVE price in the case of a protocol shortfall. We argue that, if the module must be slashed, the option to slash more AAVE is positive for AAVE price, and therefore positive for tokenholders generally. If governance can’t cover the shortfall because the slashing percentage is too low, it reserves the right to mint and auction more AAVE to meet demands. This is discussed in the Aavenomics documentation for the Safety Module, dubbed a Recovery Issuance Event.

Even if no AAVE is minted, the inability to cover a shortfall will itself incur negative price impact on the token. We have analyzed the impacts of slippage and sentiment on token price impact following a shortfall extensively in our review of dYdX’s safety module. See our analysis here.

In short, a higher slashing percentage increases the chances that governance will (1) be able to cover the shortfall without (2) having to mint and auction more AAVE. These two effects lead us to the conclusion that a higher slashing percentage is good for tokenholders. Crucially, a higher slashing percentage does not mean that governance must slash all tokens in the safety module, it merely gives governance the option.

Integrating with Previous Proposals

Raising the slashing percentage allows us to lower emissions if our goal is to keep the insurance power of the module constant. This allows governance to reap the benefits of a more efficient safety module, without compromising security. Using the current emissions of 550 AAVE/day, our conservative estimate for how much we reduce AAVE spend is 80 AAVE/day, reducing emissions to 470 AAVE/day. However, a previous snapshot from TokenLogic and Llama discussed lowering emissions to 200 AAVE/day, and re-distributing the remaining emissions to newly proposed pools.

Given their proposal is at the snapshot phase, we are awaiting community feedback on our analysis before we put forward a snapshot. We are doing so to avoid confusion and operational overhead within the DAO. Once we have received feedback from the community, we will put forward snapshots to raise slashing percentages on current and new pools, and lower emissions from whatever level they might be at that time.

Disclaimer

This research was funded by a grant from the Aave Grants DAO.

Next Steps

1. Engage in discussion around the analysis and recommendations suggested herein.
2. Following feedback from the forums, put forward [a] snapshot[s] to raise slashing percentages and accordingly lower emissions.

Copyright

Copyright and related rights waived via CC0.

[A_J]

The whole concept of risk management revolves around risk aversion. My observations are below from the view of staker considering risk management.

SM is a sensitive component of AAVE ecosystem.

In simple terms what I understand is that raising slashing to 60% will equal the same risk as of today (at par with 30% slashing) only if the staking yield doubles in value i.e 12%

If the staking yield goes beyond 12% then staking risk is less than the current 30% slashing.

If the staking yield remains below 12%, then risk for stakers is higher than the current 30% slahsing.

In my view, stakers with a risk centeric approach will withdraw stkaave and find many better risk averse opportunities to generate yield in the ecosystem. This will result in less funds in SM which will be detrimental to the ecosystem.

Generally, I believe that while discussions on slashing % are important, more focus (and investments) should be on avoiding possible risks leading to black swan events.

I am curious to know how much AAVE supply is with governance at present?

[Figue]

Thank you for the proposal! Re-thinking the Safety Module should be a big priority for the DAO.
As one of the founders of Paladin, we have been building on top of the SM for 2+ years.

There are 4 things I want to insist on before any such proposal goes further:

• Growing the slashing of the SM will significantly reduce its TVL. This observation comes from the notable outflow of stkAave during the CRV liquidation shenanigans last year;
• Current problem of the SM is that it is exclusively made of AAVE, and when Aave needs the SM, the gov token will lose value, which will reduce the firepower of the SM in order to cover shortfalls. For this reason I believe priority should not be on increasing the slashing percentage but on diversifying the assets in the SM. @Dydymoon and @TokenLogic have pushed multiple proposals in this direction recently and they should be the current focus;
• Aave v3 separates the risk quite significantly and I am not sure a billion dollar insurance fund is really necessary, maybe the SM can be expanded into something greater at some point;
• Last but not least, the tokenomics rework will push more people into the SM to mint discounted GHO, doubling or tripling the slashing seems extremely opposite to the direction tokenomics have taken.

Looking forward to this discussion

[tncintra]

Thanks for the replies @A_J, @Figue! I will address each point below.

---

@A_J

In simple terms what I understand is that raising slashing to 60% will equal the same risk as of today (at par with 30% slashing) only if the staking yield doubles in value i.e 12%

This is a natural first approach to thinking about the APR, but it ignores the other components that make up the risk premium (yield) of the safety module. One component is a slashing risk, another is the risk free rate and another is a liquidity premium. If you assume these latter two components (the “cost-of-capital”) are non-zero, then doubling the slashing percentage would lead to a less-than-double APR as stakers manage their risk tolerance:

APR = r₀ + r_other + p * r_slash

Consider an example: suppose there exists a risk-free investment that investors can make with AAVE, at an APR of 10%. The APR for the SM would then be higher than 10%, perhaps 15%. This does not mean that the slashing risk is priced at 15%, it means that we have to pay investors at least 10% yield for them not to invest in the risk-free investment. The leftover 5% is the slashing risk. Therefore, if the slashing risk doubles, we would not see the SM APR go to 30%, we would see it go closer to 20%.

Therefore, we would observe an increase in insurance power, and a marginal increase in APR demanded.

@A_J

Generally, I believe that while discussions on slashing % are important, more focus (and investments) should be on avoiding possible risks leading to black swan events.

We are in agreement that further R&D on avoiding possible risks is crucial to the long-term success of Aave. However, it is our understanding that lowering costs, and increasing Aave’s runway is also detrimental to funding these efforts. This research aims to reduce the cost of insurance by ~$3M annually with no additional technical uplift for the DAO.

@Figue

Growing the slashing of the SM will significantly reduce its TVL. This observation comes from the notable outflow of stkAave during the CRV liquidation shenanigans last year;

We expect a reduction in the TVL of the module. This is intentional: by increasing the slashing percentage we get more insurance out of each unit of AAVE staked in the SM, this effect is more significant than the expected reduction in TVL, leading to more insurance at a lower cost.

That is, we are not optimizing for TVL, but for insurance power. We believe this is the correct approach, but are interested in hearing your thoughts. We argue that non-slashable TVL is not productive for the Aave ecosystem, and is not something token holders should pay ~$3M a year for.

---

I hope this helps clarify the essence of our research, and motivates our proposal to try a higher slashing percentage. We believe this experiment poses significant upside (ultimately on the order of ~$3M a year), with little downside since the change can be easily reverted if necessary.

Happy to discuss further!

[MarcZeller]

Hello, and thank you for presenting this proposal.

Although we have a significant stake in the Safety Module and benefit from it, we support efforts to both increase risk (through a higher slashing percentage) and reduce rewards in the interest of overall sustainability.

We’ve noticed some overlap with the work TokenLogic is doing in relation to this proposal. We’re looking forward to their feedback on this TEMP CHECK before forming a final opinion.

Our general stance is supportive, and we’re eager to see this discussion progress further.

our general vision at the ACI is that Safety incentives are not sustainable and an AAVE-Only Safety module would not be the most optimal self protection system for the Aave users.

as a sidenote, we’re witnessed several mention of the Safety module as “insurance”, that term has legal weight in many juridictions and the DAO is not recognized yet as a legal entity, it’s impossible for a DAO to provide “insurance”. the term “self-protection smart contract” is more fit with our personal comfort zone.

[fig]

I’m excited for this proposal and for the community to view it seriously.

A new team, supported by Grants with a history of delivering high quality work in other DAOs.

We’ve been impressed by Xenophon’s work in dYdX and are intrigued by the potential benefits of of saving Aave ~$2.1M per year, a substantial amount.

At this rate, it feels worth exploring and rethinking the status quo.

[EzR3aL]

Hi, did you think about the implications this could have with GHO and it’s growth?
Because right now if you stake you get a discount on GHO and you earn yield. If slashing will be higher I could imagine people won’t stake and borrow discounted GHO. Because they may don’t see the risk/reward as suitable.

What if these people won’t mint GHO overall then and this stops the growth of it? Stablecoins are a money printer and we should consider this being a risk to GHO, especially it just recently launched.

Maybe you did consider this and could give an answer.

Thanks

On a side note, I think if the DAO could save 2-3m per year this would be great. These funds could be used to get yield out of other sources that are sustainable such as eth staking and distribute a percentage of this to staker to encourage them to stake.

[ProfitMaxi]

If the slash rate goes to 100% or you cancel all the incentives I would expect that the SM goes to zero. The current risk reward is ok, but if you change any of those variables then a serious change to the tokenomics is needed at the same time

[Michigan_Blockchain]

Thanks XenophoneLabs for presenting such an impressive piece of work. The decomposition of staking yield and the focus on capital efficiency, as reflected in insurance power per unit of AAVE, effectively models the efficacy of the AAVE safety module. This well-conceived and succinct model underscores the need to reduce costs, with emissions being a significant factor in this regard.

We support this proposal as a long-term strategy to enhance the protocol’s sustainability. Additionally, we believe this approach should be considered for other potential staking assets, as discussed in Llama’s Safety Module series posts.

Setting the bar at 60% seems a sensible starting point; it provides ample time for the community to monitor and assess the responses from the Safety Module depositors.

[monet-supply]

Generally supportive of raising the max slashing param to put more risk on the safety module stakers. However I want to call out an assumption that I think is not accurate:

In the event that insurance power decreases, we revert our change.

I think it is unlikely that reverting the change on max slashing would result in all or most of the people who withdraw putting their aave back in. This parameter change has some degree of risk, don’t think there’s any way around it unfortunately.

[A_J]

“We argue that, if the module must be slashed, the option to slash more AAVE is positive for AAVE price, and therefore positive for tokenholders generally.” @XenophonLabs

1. Is this argument valid only in the case of shortfall event?

2. If so, What is your view on AAVE price movement (a) as a consequence of the proposed implementation (b) in the absense of a shortfall event?

[0xbilll]

Aave Grants is hosting an AMA this Thursday to discuss @XenophonLabs proposal with their team, @TokenLogic and anyone else from the community who is interested. Please join us!

AaveGrants.lens

[tncintra]

Thanks for the feedback so far, and we are excited for the AMA on Thursday!

---

@monet-supply

I think it is unlikely that reverting the change on max slashing would result in all or most of the people who withdraw putting their aave back in. This parameter change has some degree of risk, don’t think there’s any way around it unfortunately.

We agree there is some degree of risk. However, we suspect that the user profile of someone who is elastic to the slashing percentage will just as soon come back when the slashing percentage decreases as they left when the slashing percentage increased, especially given the lack of alternative investments for AAVE.

---

@EzR3aL

Hi, did you think about the implications this could have with GHO and it’s growth?

At its core, the discount model for GHO in the safety module acts as an incentive for securing the Aave protocol, not as a tactic to grow GHO itself, per the original announcement:

@AaveLabs

Holders of stkAAVE can mint GHO at a discounted rate, meaning they will pay a lower interest rate on the GHO that they borrow. Therefore, there is an incentive to help secure the Aave Protocol as AAVE is staked (stkAAVE) into the protocol’s Safety Module.

However, this is an important point for discussion; thanks for flagging it . We are currently discussing with other community members, and we will provide a better informed opinion soon.

---

@A_J

“We argue that, if the module must be slashed, the option to slash more AAVE is positive for AAVE price, and therefore positive for tokenholders generally.” @XenophonLabs

1. Is this argument valid only in the case of shortfall event?

2. If so, What is your view on AAVE price movement (a) as a consequence of the proposed implementation (b) in the absense of a shortfall event?

We can’t comment on the price of the AAVE token, but we can comment on the impact this proposal has on the costs to the Aave DAO. This proposal will reduce AAVE emissions/inflation, and generally improve the financial well-being of the protocol. We hope the community can decide for itself how this would affect the price of the AAVE token, outside the context of a slashing event.

[Dydymoon]

Thanks for participating in the discussion to improve the Safety module, however I disagree with this proposal.

Disclaimer: I’m a DeFi strategist at Llama & TokenLogic. I’ve led the SM upgrade research & proposal at Llama during the past months, and currently working on the ARFCs and implementation details following the successful TEMP CHECKs of part 1 to 5. This comment is my personal view.

@XenophonLabs

• Increasing the slashing percentage to 100% would save Aave millions of dollars per year, and we anticipate it would have neutral-to-positive effects for the module’s insurance power.

• We recommend:

• Raise the slashing percentage on stkAAVE from 30% to 60%, and accordingly, lower emissions. After observing impact, we either raise the slashing percentage to 100%, or revert our changes.

• Set the slashing percentage on newly proposed pools to 100%.

• We summarize our research report below, as well as a suggested rollout plan for the new changes, and request feedback from the community on our analysis.

It seems most of the feedback I shared with you last week are not taken into account in this proposal, so I’ll reexplain here why I’m against the majority of this proposal, and the different topics around it.

Anyone could say “If users are slashed more (i.e increase to 100%), the cover is higher and cheaper” and ofc in theory it’s correct … if depositors don’t leave with a increased risk + a reduced yield, which you seem to not only underestimate, but also use your only assumptions to propose upgrading one of the biggest components of the Aave protocol.

About your recommendation:

• Strongly against increasing the slashing on StkAAVE not now and not in the future either. There are several reasons for it which I’ll explain below.

• Against increasing the slashing to 100% on any assets at the moment, for the obvious reason that we want to bootstrap the new categories.

• Open to considerations to increase to 100% for non Aave assets in the future depending on how depositors behaved on volatile & stable LPs with the 45% & 60% slashing (Btw these % will still be open for discussion in the upcoming ARFC)

• About emissions, our goal was to not overspend compared to the current incentives budget and estimate the potential max TVL & APR depending on this budget value, which should decrease in the same time that the Aave DAO is accumulating strategic voting power & earning GHO.
  Against reducing to 80 AAVE/day for StkAAVE even with the same 30% slashing until GHO rewards can be added, as the TVL might already be significatively impacted by the reduction from 550 to 200 AAVE/day.

@XenophonLabs

Our analysis and recommendations are supplementary to most of the work done by TokenLogic and Llama on the safety module. They discuss the slashing percentage in Part 2, and emissions in Part 4 of their Safety Module upgrade.

The main goals of the current Llama proposal are the following:

• Improve assets diversification and total value of the SM Cover
• Implement several categories with different parameters
• Update the slashing parameters according to categories (leading to better capital efficiency)
• Reduce correlation between AAVE token price and potential shortfall event
• Reduce total cover cost & overall incentives distribution
• Improve liquidity on Aave native assets

There are additional considerations which can be found in the TEMP CHECK posts.

As a short reminder, we proposed 3 categories in the SM

• Single assets: 30% Slashing - LM, Budget partially rebalanced on other categories
• Volatile LPs: 45% Slashing - Emissions directed by votes incentives & voting power
• Stables LPs: 60% Slashing - Emissions directed by votes incentives & voting power

This proposal clearly goes against the correlation reduction between a potential slashing event and the Aave token price. Since it’s basically assuming that depositors will stay even by increasing the risk x2 or x3 and reducing the yield at the same time, it makes an unviable assumption on the potential TVL, which can also impact the total cost if the goal is to retain this amount. This proposal does not take into account the Aave liquidity & impact of a potential shortfall event either.

@XenophonLabs

It currently sits at 30%. In a sponsored by the Aave Grants DAO, we provide a quantitative analysis for why a slashing percentage below 100% is strictly capitally inefficient.

Capital efficiency is not the only metric you should be watching for your analysis, especially since as you said it yourself, it doesn’t make sense if the capital can’t be seamlessly sold on the market.

It’s impossible to predict what depositors will do, especially on new categories & assets not even implemented yet. Increasing the slashing to 45% for volatile LPs and 60% for stable LPs can be the first step before another increase, and it’s also what you need to perform an actual analysis on depositors (in)elasticity.

The 80/20 Aave-wstETH is one of the volatile LPs proposed, and should be a very interesting case study for you considering the slashing would be increased from 30% to 45%, while the yield would be lower (8 to 12% APR) for a reduced TVL.

However, considering this pool also contains Aave, in the future I’ll also support reducing the slashing, compared to other volatile pools proposed such as 80/20 wstETH/GHO for exemple.

Stable pools should begin at 60% slashing, so it’s another interesting change for you to monitor depending on how depositors act with the current estimated yield (18 to 22% up to 50M+ TVL).

It can also be interesting to revisit the slashing parameters to potentially increase it on stable pools a few months after the category is launched.

@XenophonLabs

Following this forum post, we intend to propose raising the slashing percentage on all Aave safety module pools, including the existing stkAAVE pool, and the new pools proposed by Llama and TokenLogic. We expect these proposals will significantly reduce the current costs of the Safety Module, and make any future pools more capital efficient. Furthermore, they will require negligible technical uplift.

While increasing the slashing on all assets could reduce the costs, it also brings an important risk of significantly reducing the SM TVL, which would obviously not be capital efficient.

I’ve explained above why I don’t think increasing the slashing to non-Aave pools too much at the moment is a good idea (but could be in the future). I’ll now explain why I believe that the slashing on AAVE single asset shouldn’t be increased at all:

1. Decrease the correlation between token price & shortfall event risk:

As explained above, one of the goals of our proposal is to increase the cover diversity with liquid & resilient assets. Considering this task will take time, I believe it’s currently not possible to fully remove the Aave token from the SM. However, we can avoid increasing the slashing to begin with. Then, imo the % changes should be to decrease it, which will also reduce the risk for holders.

2. Reduce risks and improve attractivity associated to staking Aave:

Since the first iteration of the Aave tokenomics, the only way to receive rewards and governance power is to protect the protocol by taking the slashing risk, along with the risk that if an important shortfall event actually happens, the Aave price would most likely go to 0.

Associated with an improved tokenomics (discounted interest rate on GHO borrows for stakers), and GHO revenues distribution incoming, less risks could lead to more attractivity on the Aave token, so a potential price increase which would improve the SM incentives budget value, as well as other expenses denominated in AAVE & reduce the amount spent from the ecosystem reserve.

3. Avoid the “Governors are also depositors” Conflict of interest:

Up to this point, the Safety module slashing has never been used. It could have been used following the CRV excessive debt issue, but the DAO decided to use stables from the treasury as it was a relatively small portion of the holdings, and I believe it was ok since not highly impacting the treasury.

If the amount to repay was $30M+, the rational decision should be to use the SM to at least partially cover the issue since it was created and costing money to the DAO for this reason. However, Aave stakers are also the ones voting to whether the SM should be used or not:

Considering that the Aave treasury is growing (~ 30M$ of non AAVE assets) and should grow even more with GHO earnings, voters could take selfish decisions to not use the SM to avoid loosing funds, highly impacting the treasury and protocol health.

Reducing the slashing on Aave single assets & LPs over time would help to reduce both the impact of this conflict of interest, and the correlation to the Aave price in case of shortfall, so Xenophon proposal requesting to increase the slashing on StkAAVE goes fully against this.

@XenophonLabs

However, before moving to a more formal proposal, we would first like to request feedback from the community on our analysis. We are more than happy to discuss our assumptions, and provide further justification for why we believe this is the correct decision for the DAO.

Interesting considering that while you guys agreed with most of my feedback on our last call, most if not all are not included here.

@XenophonLabs

Specification

So far, this exercise has been largely theoretical; we made some assumptions, which we challenged and documented in Section 2 of our report, and drew conclusions from the modeling. These economic models on their own are not enough to know the outcome with certainty.

This proposal is entirely theoretical as mostly based on your assumptions that depositors will take more risks for less rewards. It does not take into account many variables useful to actually modelize the changes and draft some estimations, such as the new yield mechanism for LPs in our proposal, an estimation of the TVL drop, no estimation on the aave token price impact in case of shortfall so explaining that you get more by taxing more is not really helpful tbh.

@XenophonLabs

Proposal Overview

Acknowledging that economic models are imperfect, we recommend doubling the slashing percentage from 30% to 60% on the stkAAVE pool, and waiting some time to gather data on whether our claims hold true. If the module’s capital efficiency increases as expected, we may raise the slashing percentage to 100%. Otherwise, we may revert it back to 30% if desired.

Outside the fact that raising StkAAVE slashing should be a no go, do you realize that the SM is mostly composed of Aave tokens atm, and you want to risk a significant drop of the TVL simply based on your unviable assumption that stakers won’t leave with more risks and less rewards ?

Why not just gather data on the case studies outlined above and present a proposal in a few months with an actual analysis of changes in TVL ?
Definitely open to work on it together if you’re interested as these analysis will be needed.

@XenophonLabs

Higher capital efficiency means that the DAO can reduce the emissions to the safety module, while keeping the insurance coverage constant, or perhaps even increasing it.

One of the goals of our proposal is also to reduce the AAVE emissions to the SM over time, however it’s explained by the fact that Aave DAO would accumulate voting power used to vote on smBPT gauges, progressively recycling more AAVE with vote incentives received.

@XenophonLabs

For this reason, we recommended in our report to raise the slashing percentage to 60% on the stkAAVE pool, and lower emissions by 80 AAVE/day, from 550 to 470. Using our model, we show that this would retain at least as much insurance power in the module, despite spending less on rewards

Your model is assuming that Aave stakers will stay in the SM risking 60% of their holding to receive ~ 0,9% APR (for the same TVL assuming no one would leave). How can this make sense ?

@XenophonLabs

We later discuss how this is affected by recent snapshots from TokenLogic and Llama, where they proposed lowering emissions from 550 to 200 AAVE/day.

There was no comment from Xenophon Labs on any of the TEMP CHECKs.

Looking forward to seeing you on the ARFC so we can avoid double proposals.

@XenophonLabs

Proposal Considerations

Llama and TokenLogic have already passed a snapshot, along with several other snapshots, to lower emissions to the stkAAVE pool to 200 AAVE/day. This impacts our ability to measure the impact of this recommendation.

@XenophonLabs

Furthermore, the safety module currently sits in limbo between successful snapshots and putting forward AIPs to implement them. Given that our research intersects with an existing snapshot, and to avoid confusion in the DAO, we have chosen to delay our proposal and instead request feedback on the analysis. We are working with several community members on how to best structure a proposal to increase the slashing percentage and accordingly lower token emissions.

For context, you reached out mid-may with the goal of requesting a grant on this topic.

We had a first call but iirc, the research had not really started yet on your side since you were waiting to know if your grant would be approved.

I explained the proposal we worked on, and invited you to comment on the TEMP CHECK about the slashing parameters concerns, as the proposals have been on the forum since mid-April. We decided to wait because of not many comments on such an important topic, but you didn’t share your thoughts so we pushed the parts (1-2-3) of the snapshot for voting early June (part 4 & 5 came later on).

I only heard back from your early July with the repost linked above. EthCC week was a few days later, but I took the time to read your report and have another call to share my feedback on your research & the state of our proposal, and literally no news since this call.

I thought we agreed that you would comment on the upcoming ARFC, so quite surprised to see you push another proposal contradictory instead of working together.

@XenophonLabs

Other Considerations

Inelasticity

An assumption we make in our modeling is that stakers are elastic to changes in risk and yield. If stakers are fully elastic, then we expect them to react to changes in risk and yield as we have modeled. However, staking AAVE in the safety module is currently the only venue for earning yield on AAVE while also retaining voting rights. Lack of alternative investments that retain voting power likely makes stakers inelastic: an increase in risk does not lead to stakers withdrawing a portion of their stake. Therefore, it is likely that we observe an even greater improvement in capital efficiency than we estimate in our model.

• First of all, this should be explained a lot earlier in the proposal since it’s what you base your entire estimations on.

• Then, as explained above I think its unreasonnable to believe that stakers will stay twice more risk and more than 5x less yield.

• But most importantly, this basically brings a risk of having less participation in the governance (which is already more than limited with around 20 voters per proposal including delegates ofc) by motivating depositors to just leave the SM to use their AAVE elsewhere (as collateral, in liquidity pools etc) bringing them more yield and less risks, but highly impacting the SM TVL if the diversification is unsuccessful.

• This also increases the conflict of interest outlined above between governors & depositors.

• And finally, even if the slashing was raised on AAVE to 100% leading to more capital available (so called capital efficiency in your proposal), there is not enough existing liquidity to sell it all without literally killing the token price and highly impacting the protocol and DAO treasury health.

This assumption is not only very light but also quite dangerous.

@XenophonLabs

We argue that, if the module must be slashed, the option to slash more AAVE is positive for AAVE price, and therefore positive for tokenholders generally.

Wait what ? You know slashing a portion of funds deposited in the SM is actually selling the token right ?

@XenophonLabs

If governance can’t cover the shortfall because the slashing percentage is too low, it reserves the right to mint and auction more AAVE to meet demands. This is discussed in the Aavenomics documentation for the Safety Module, dubbed a Recovery Issuance Event.

Are you aware that minting more AAVE literally dilutes all token holders, and increases even more the liquidity limits risks outlined above right ? It’s actually the reason why the governance could not cover a shortfall atm, not because of a low slashing %.

Yes a recovery issuance event is possible, but it’s like the last line of defense of the DAO. If this happens, it would mean the shortfall happening is so important that the full SM cover + potentially the available part of treasury + potentially the available part of the ecosystem reserve would not be enough to cover the loss, in which case the Aave token price will crash close to 0 and minting more token could be worthless.

Luckily, Aave DAO has a growing treasury and several revenue streams, so I guess the Aave governance will always consider using the treasury before a new mint of AAVE tokens. However, the treasury should also always be considered after the SM cover.

Anyway, I believe this proposal needs to be reworked and I’m looking forward to seeing Xenophon Labs comment on the ARFC and work together on improving the Safety Module.

[tncintra]

Hi @Dydymoon, thank you for your feedback. We’re glad to be able to have this discussion on the merits of our ideas transparently in the forums. I believe there are some misunderstandings regarding our analysis, and I would like to elaborate on them below.

Although we disagree with a lot of the post, let’s start with some important points where we agree with Dydy.

1. Agreed that stkAAVE should not be slashed before other staked pools.
2. Agreed that diversifying the safety module is a good thing.
3. We think a lot of the other work on the safety module has value, and that it’s a totally fine and healthy thing for multiple researchers to have input on the various aspects of the module.

---

The Effect of Staker Elasticity

Dydy presented several accusations that our analysis assumes that stakers won’t leave despite us raising risks and lowering rewards:

@Dydymoon

This proposal is entirely theoretical as mostly based on your assumptions that depositors will take more risks for less rewards.

@Dydymoon

Outside the fact that raising StkAAVE slashing should be a no go, do you realize that the SM is mostly composed of Aave tokens atm, and you want to risk a significant drop of the TVL simply based on your unviable assumption that stakers won’t leave with more risks and less rewards ?

This is a misunderstanding of our approach. We are claiming that stakers will take more risks and will demand more rewards. Under our analysis, stakers will withdraw some stake, with the APR reaching equilibrium a couple percentage points above the current 6%. At this new (and higher APR), the TVL will be lowered. However, despite a lower TVL, we will still be able to extract more insurance power, since the slashing percentage is higher. This is what allows us to lower emissions while keeping insurance power approximately constant. Perhaps this is better visualized in the below bar graph*:

We have elaborated on this analysis in our responses to A_J and Figue, please refer to that post if interested.

Our proposal assumes stakers are perfectly elastic, and concludes that we would still achieve a more capitally efficient safety module. If stakers are not perfectly elastic, then the improvement to capital efficiency would be amplified. Dydy’s accusation that our research relies on inelasticity is incorrect.

As a sidenote, it appears that there is a misunderstanding regarding the reduction in emissions we are proposing, which we should clarify:

@Dydymoon

Against reducing to 80 AAVE/day for StkAAVE

@Dydymoon

Your model is assuming that Aave stakers will stay in the SM risking 60% of their holding to receive ~ 0,9% APR (for the same TVL assuming no one would leave). How can this make sense ?

We are not suggesting we would lower emissions to 80 AAVE/day, we are suggesting to lower it by 80 AAVE/day assuming that we are at the current 550 AAVE/day. If the emissions at the time of our proposal are at the new 200 AAVE/day (assuming the ARFC passes), then we will re-evaluate how much we would change emissions.

Notice that the data in the graph is conservative: we use USDC deposit rates from Jan/Jun as our risk-free-rate (about 2% on avg) which is significantly below other potential risk-free-rates, such as treasuries (5%) or ETH liquid staking (4%). Assuming a higher risk-free-rate, the yield demanded for slashing is proportionally lower, which would lead to a lower increase in APR!

---

Diversification

We are all in full agreement that having more diverse assets in the safety module is good for the DAO, and we support the idea of transitioning away from stkAAVE. Furthermore, we agree that we should slash the non-AAVE funds first to avoid damaging AAVE price. In fact, if we really cared about AAVE price, we would set the slashing percentage on the non-AAVE pools to 100% to prevent AAVE from needing to be slashed in the first place.

@Dydymoon

Against increasing the slashing to 100% on any assets at the moment, for the obvious reason that we want to bootstrap the new categories.

Let’s be specific, we want to bootstrap the insurance power of new strategies — not the TVL — which is exactly what this proposal will do. We could, of course, bootstrap liquidity in the new modules by setting their slashing to 0%, but this would be a waste of Aave’s treasury, since we get no insurance out of it. As stated above, we are fully in favor of diversifying the safety module away from AAVE, and each of those modules would be more effective if they had a higher slashing percentage.

---

AAVE Price

@Dydymoon

This proposal clearly goes against the correlation reduction between a potential slashing event and the Aave token price.

We have elaborated clearly on why increasing the slashing percentage increases the probability that Aave will be able to cover a shortfall and prevent the minting of new AAVE token under the Recovery Issuance module.

Assuming the new pools are created, this is still the case, since governance can choose which pools to slash first (e.g., stkAAVE can be more “senior” to staked ETH, and therefore stkAAVE gets slashed as a last resort). And, as we have stated, setting their slashing percentages to 100% will further prevent having to auction AAVE.

@Dydymoon

Are you aware that minting more AAVE literally dilutes all token holders, and increases even more the liquidity limits risks outlined above right ? It’s actually the reason why the governance could not cover a shortfall atm, not because of a low slashing %.

Yes a recovery issuance event is possible, but it’s like the last line of defense of the DAO. If this happens, it would mean the shortfall happening is so important that the full SM cover + potentially the available part of treasury + potentially the available part of the ecosystem reserve would not be enough to cover the loss, in which case the Aave token price will crash close to 0 and minting more token could be worthless.

Let’s be clear about two points:

1. A higher slashing percentage does not require governance to auction all of the AAVE up to that max slashing, we can always slash less than the limit. However, the optionality of slashing more is strictly positive for the DAO in trying to cover a shortfall, since the DAO would only exercise that option if it needed to.
2. Minting more AAVE is bad for the protocol. By increasing the slashing percentage, we increase the DAO’s ability to avoid having to do this.

Raising the slashing percentage does not mean the DAO will have to flood the market with more AAVE, it means that the DAO will have the option to auction more AAVE if necessary to prevent having to mint more AAVE or not being able to cover the shortfall, either of which might be worse.

---

Final Note

@Dydymoon

There was no comment from Xenophon Labs on any of the TEMP CHECKs.

Looking forward to seeing you on the ARFC so we can avoid double proposals.

Our grant to study the safety module was awarded after Llama’s TEMP CHECK. We will certainly be participating in this upcoming ARFC.

It’s understandable that @Dydymoon is not totally on board with this yet. We’re DeFi researchers too, and we’re reasoning about a complex system here. We believe this discussion is a healthy way for the DAO to reach an informed decision on its expenditures. We look forward to ironing these concepts out in our AMA tomorrow!

[A_J]

@stani what are your thoughts on this temp check? This is as important as it gets

[Dydymoon]

Thanks for the feedback !

@tncintra

Although we disagree with a lot of the post, let’s start with some important points where we agree with Dydy.

Agreed that stkAAVE should not be slashed before other staked pools.

I didn’t said one category should be slashed before another one, idk if it’s a good idea but can be to be explored, I think you suggested it later on this comment.

Also observing that you didn’t answer some points outlined above.

@tncintra

Dydy presented several accusations that our analysis assumes that stakers won’t leave despite us raising risks and lowering rewards:

This is from your report: “Based on 150 days of data on the Aave safety module, we estimate that the slashing percentage of 30% is costing the Aave DAO between 2 and 4 million dollars a year. To address this, we propose raising the slashing percentage on the stkAAVE pool and cutting emissions to the stkAAVE module, keeping its insurance power approximately constant.” This is more risk for less rewards.

As for the TVL assumptions (also from your report): “In reality, there are a number of idiosyncratic reasons why stakers might not express (or have any) risk aversion. For example, stakers might not have immediate access to their tokens or they might be “Aave-maxis” and wish to backstop the protocol in case there is a shortfall. As mentioned, a large driver for inelasticity in the stkAAVE module is likely to be governance. Staking into the AAVE-only safety module is currently the only way to earn yield on AAVE while retaining governance power (as far as the authors are aware). A great piece of evidence for stakers being inelastic to risk and returns is that the stkAAVE tvl has been roughly constant over the last 150 days, despite significant fluctuations in AAVE price and a turbulent crypto market with an uncertain risk profile”

@tncintra

This is a misunderstanding of our approach. We are claiming that stakers will take more risks and will demand more rewards. Under our analysis, stakers will withdraw some stake, with the APR reaching equilibrium a couple percentage points above the current 6%. At this new (and higher APR), the TVL will be lowered. However, despite a lower TVL, we will still be able to extract more insurance power, since the slashing percentage is higher. This is what allows us to lower emissions while keeping insurance power approximately constant. Perhaps this is better visualized in the below bar graph*:

It doesn’t change the fact that it’s based on assumptions of the TVL before and after (btw wondering why this graph wasn’t in your report/proposal before).

I think the biggest issue is that you’re not optimizing for TVL as mentioned later, when the goal is to increase the amount of cover so it can actually have an impact.

If I understand this graph correctly, it assumes that the StkAAVE TVL will have a 50% drop, but since you’re slashing twice more the insurance amount remains constant.

However, this highly impacts the SM TVL, creates more risks on the Aave token and doesn’t fix the inefficiency of the SM since the liquidity is not enough to cover this kind of selling pressure so still not the best solution imo.

@tncintra

As a sidenote, it appears that there is a misunderstanding regarding the reduction in emissions we are proposing, which we should clarify:

For the APR graph, I indeed misunderstood the amount of incentives you proposed to allocate to StkAAVE (80 instead of 470 AAVE/day) probably because of your mention that this proposal is supplementary to the Llama one, when it’s actually contradictory.

By allocating 470 AAVE/day to StkAAVE only, it’s not only possible to add another single asset to diversify this category (ETH/LST) or to incentivize the volatile & stable pools without increasing the max amount of incentives allocated, which we avoided in our proposal.

@tncintra

urthermore, we agree that we should slash the non-AAVE funds first to avoid damaging AAVE price. In fact, if we really cared about AAVE price, we would set the slashing percentage on the non-AAVE pools to 100% to prevent AAVE from needing to be slashed in the first place.

As mentioned above, I guess you agree with yourself here, considering I didn’t propose to slash one category before another one.

For setting the slashing to non Aave assets to 100%, I already answered in the previous comment,I disagree now as it can have a significant impact if done directly but could be explored in the future, based on an analysis of the changes related to the new categories.

@tncintra

Let’s be specific, we want to bootstrap the insurance power of new strategies — not the TVL — which is exactly what this proposal will do.

It’s exactly the issue here. The current SM TVL could theoretically cover around 1,37% of the Aave protocol TVL (excluding the liquidity issue making this impossible).

Our proposal focuses on increasing the potential total cover amount compared to Aave TVL, improving the liquidity resilience in the SM & improving the capital efficiency with the higher slashing parameters for volatile & stable LPs categories.

@tncintra

We could, of course, bootstrap liquidity in the new modules by setting their slashing to 0%, but this would be a waste of Aave’s treasury, since we get no insurance out of it.

That would not make sense, since the goal is to use these non Aave assets to increase the capital efficiency, and the liquidity resilience in the SM in case of a shortfall event.

@tncintra

As stated above, we are fully in favor of diversifying the safety module away from AAVE, and each of those modules would be more effective if they had a higher slashing percentage.

Yes and a higher slashing than single assets is exactly what we proposed: 45% on volatile & 60% on stable LPs compared to single assets at 30% to begin with.

Even the AAVE/wstETH 80/20 is included for now considering that’s the only LP atm, which will also be interesting for you to observe in the coming months.

@tncintra

Assuming the new pools are created, this is still the case, since governance can choose which pools to slash first (e.g., stkAAVE can be more “senior” to staked ETH, and therefore stkAAVE gets slashed as a last resort). And, as we have stated, setting their slashing percentages to 100% will further prevent having to auction AAVE.

You still self agree and repeat the same arguments here.Decide if a category should be slashed before another one is another topic that wasn’t in the initial proposals iirc.

That’s an interesting idea tbh but even assuming stkAave would be slashed last, this would really depend on the total SM TVL and the impact of the potential shortfall event, which is hard to predict.

Still don’t think implementing the new categories with 100% slashing from the start is a good idea, as it will most likely refrain depositors.

@tncintra

Raising the slashing percentage does not mean the DAO will have to flood the market with more AAVE, it means that the DAO will have the option to auction more AAVE if necessary to prevent having to mint more AAVE or not being able to cover the shortfall, either of which might be worse.

I understood what you proposed the first time, repeating it doesn’t make it a better idea.

• Minting more AAVE is very dangerous as it dilutes all holders, it’s bad for the protocol, and should be used as the last line of defense, but in reality arriving at this point would probably lead to a worthless mint as the AAVE token should be already very close to 0.

• Increasing the slashing on Aave assets would not fix anything with the current market conditions as slashing more AAVE does not mean you can sell more AAVE, but it also increases the risks on the token, the correlation between token price & shortfall event, and the conflict of interest between governors & depositors outlined in my previous comment.

@tncintra

We will certainly be participating in this upcoming ARFC.

Looking forward to it !

[tncintra]

Thank you everyone for joining the AMA, and thanks @0xbilll for organizing. Here is a recording for those interested: https://twitter.com/AaveGrants/status/1687083027422179330?.

[tncintra]

We will publish a snapshot vote to raise the slashing percentage on the existing stkAAVE pool and the newly proposed pools. We decided to wait to submit our snapshot until the existing proposals by @Llamaxyz and @TokenLogic finish going through the AIP process. This allows us to gather some data on the new state of the world, leading to a better estimate of the impact of raising the slashing percentage and a better estimate on how much we can lower emissions by. We also hope this minimizes governance overhead regarding the safety module.

We are happy to answer any questions until then. Thank you for the detailed discussion so far.

[ApuMallku]

Aho
We think that a revamp of the staking system is needed. It is not effective to use the native token for the safety module. We will appreciate if we can start a new topic around how to improve the safety module as a whole, instead of raising the slashing.

We will vote for NAY.