[RFC]: Restart Enzyme vault & diversify ETH across staking providers

[Moss]

Hi all, @moss @elisafly @eek637 @buendiadas here, posting from Avantgarde Treasury (the Treasury Management arm of Avantgarde Finance). Looking forward to hearing your thoughts and open to receiving feedback and adapting the proposal based on it.

INTRODUCTION
The Capital Pool’s position in the M11 Credit WETH Pool has been unwound. The steps to do so were clearly outlined here by @Dopeee, and have resulted in 13,218 ETH idle in your NXMTY Enzyme vault. This chunk currently represents ≃9% of the total Capital Pool.

Despite the above, the strategy of holding assets in an Enzyme vault where they can be managed by an individual or a firm in a trustless way rather than by governance can be viewed as successful. This setup demonstrated the value of being able to react to market events in real-time rather than via token vote. From a technical perspective, the implementation has been working live in production and holding significant value for more than 6 months (Aug 2022) with no technical or contract issues.

That said, there are some lessons around portfolio construction that can be taken away from the experience. Below, we suggest a new strategy to reboot and put assets to work for the Mutual.

OBJECTIVES
The objectives of the strategy are:

• Maintain directional long exposure to ETH
• Earn rewards from ETH staking
• Hedge against risks through due diligence, diversification & continuous monitoring

INFRASTRUCTURE
We propose to keep the assets in the Enzyme vault due to the following benefits:

• The non-custodial and trustless nature of the vault setup
• The transparency and 24/7 on-chain reporting capability
• The interoperability of Enzyme with Nexus due to dev work we’ve done in the past
• Operational simplicity & ability to act quickly in a rapidly evolving environment
• The possibility for Avantgarde to make future proposals that can be voted by Nexus and implemented on chain via the Enzyme vault and its trustless delegation.
• The optionality to sell ETH into stables in support of matching currency of assets & exposure if/when the DAO decides to move in that direction (though our view is that medium term there should be an ETH-denominated capital pool and a DAI-denominated capital pool, separate from each other).

STRATEGY
We propose:

• 25% of the ETH is used to buy Rocket Pool ETH rETH
• 75% of the ETH is staked natively amongst 3-5 top-tier node operators with the aim of diversifying the staking portfolio both geographically and operationally.

Important notes:

• The implementation would take place after the Shanghai fork in order to avoid any illiquidity issues.

• Fees charged by Enzyme/Kiln are substantially lower (8%) than fees charged by liquid staking derivatives (10 - 25%).

• Deliberate exclusion of stETH: The NexusCapital Pool is already exceeding the 20% threshold and is also covering against Lido’s risk. We’ll monitor the situation and adjust recommendations if the situation changes.

• Deliberate exclusion of cbETH for the time being: Given the current enforcement action that has been initiated, we think that the centralised redemption process of cbETH could represent an extra layer of risk. We will continue to monitor the situation and adjust if things change.

The capital allocation on the total Nexus capital pool would result in something like this:

AUM RAMP UP
After the implementation of this first step, we propose a quarterly assessment of the status of the vault and the allocation of an additional 4,000 ETH per quarter so that ideally after 12 months the vault reaches an allocation that is in line with the criteria of max 20% of the capital pool per single manager/counterparty/protocol, as clearly outlined in your investment philosophy. This will help increase the overall performance of the Capital Pool.

FEES
0.5% fee (includes Enzyme protocol fee and Avantgarde management)

MONITORING & REPORTING
The monitoring can be done in real time 24/7 from the Enzyme app and we can also increase the level of customization with a dedicated whitelabel page that can be linked directly to https://nexusmutual.io/.

RISKS
Based on the list of risks laid out in the investment philosophy, and the fact that the capital allocated on the Enzyme vault is less than 10% of the capital pool, we believe this strategy qualifies for the lower risk bucket.

Risk	Score	Lower Risk	Medium Risk	Higher Risk	Comments
Illiquidity Risk	◯◯	Can be liquidated in 72 hours or less	Can be liquidated in 7 days or less	Liquidation takes more than 7 days	Locking up capital for periods of time presents risk to the balance sheet should the mutual need those funds to pay claims
Basis Risk	◯◯	ETH denominated	High to medium correlation with ETH	Little to no correlation with ETH	The balance sheet is largely ETH denominated. ETH is effectively ‘cash.’ Investing in other tokens introduces basis risk
Protocol Risk (DeFi Safety Score)	◯	DeFi Safety Score >=80%; simple design	DeFi Safety Score >=80%; Newer protocol; composability (2 layers)	DeFi Safety Score >=70%; 3 or more protocol layers	Putting funds in a vault or liquidity pool to earn a yield opens up our funds to risk of loss from smart contract hacks
Liquidation Risk	◯◯	No liquidation risk	Max 20% expected loss in liquidation	20%+ expected loss in liquidation	In the case of lending, Nexus collateral could be at risk of liquidation
Leverage	◯◯	No leverage	Max 30% net leverage	30%+ net leverage	Leverage may either be an explicit component of a particular strategy, or embedded leverage (Options, Futures, etc.)
Counterparty Risk	◯◯	10% exposure to a single counterparty or less	20% exposure to a single counterparty	20%+ exposure to a single counterparty	Other additional qualitative & quantitative measures of counterparty risk may be used to assess investments & managers
Economic risk	◯◯	Negligible possibility of loss as a result of investment	Limited loss possibility	Unlimited loss possibility	Some investments may result in losses on a short term basis, e.g. impermanent loss.

USEFUL LINKS
Avantgarde’s bio, Nexus Mutual’s Enzyme vault , Kiln <> Enzyme deck, Kiln’s validators stats, Nexus’s investment philosophy, Nexus’s Capital Pool, Mismatch currency of assets <> exposure explained

GOVERNANCE
The proposal will be put forward through the Nexus Mutual governance starting with this Request for Comments that allows for a period of review and feedback by the community. We are open to your comments / questions and to adapting this proposal accordingly.

If the governance process approves this investment, the implementation steps as outlined in the Technical section will be put in place by the Avantgarde Treasury team.

[Moss]

This comment is reserved for changelog and recap of the feedback points raised by the community.

[GordonGekko]

Given that we are starting from where we are starting from, broadly agree with ideas and intents. Would note that the Mutual offers cover on Enzyme and likely both Lido and Rocketpool (and another ~10 LSTs to follow). Can you detail the involvement of Enzyme in direct staking options and whether this is increasing the recursive exposure of the Mutual if direct staking is pursued in the Enzyme format? What would be the tokenised on-chain representation of the direct Kiln staking? $lsETH?

Kiln Liquid Staking

If via an LST, are there differential withdrawal queue implications? Also, what are the implications of further increasing the number of LSTs in the portfolio to further spread smart contact risk and further retard the power law centralisation of Ethereum Validators under the control of dominant LSPs? Also, does use of Enzyme and direct staking preclude generating extrinsic yield on LST holdings?

[buendiadas]

Hey @GordonGekko, good questions, I’m @buendiadas from Avantgarde Finance. Here’s our attempt to answer your questions:

---

Can you detail the involvement of Enzyme in direct staking options and whether this is increasing the recursive exposure of the Mutual if direct staking is pursued in the Enzyme format

The Enzyme Vault will allocate the incoming ETH into Kiln deposit contracts and Liquidity Providers. After that it will keep track of the net asset value (NAV) of the fund, adjusting the share price in accordance. The share price will include consensus and execution layer rewards from the nodes, and also the value of any other proposed tokens.

What would be the tokenized on-chain representation of the direct Kiln staking? $lsETH?

There is no Liquid Staking Token that represents our Kiln staking position. Instead, Enzyme vault shares will play a very similar role to the one played by Liquid Staking Tokens, allowing Nexus to mint or redeem tokens at the current NAV.

If via an LST, are there differential withdrawal queue implications?

This is in our opinion one of the benefits of using direct staking instead of using Liquid Staking Tokens. The Enzyme Vault will have associated a number of staking validators that will be solely dedicated to this specific purpose. They will follow the general mechanics of the ETH2 withdrawal queue, avoiding the potential congestion of the LST queue.

Also, what are the implications of further increasing the number of LSTs in the portfolio to further spread smart contact risk and further retard the power law centralisation of Ethereum Validators under the control of dominant LSPs

• In terms of smart contract risk, allocating funds into different LSTs clearly diversifies the risk. If one of the Smart Contracts in a LST experiences a security breach, it is very unlikely to affect the Smart Contracts of the others
• When it comes to power law centralization of Ethereum Validators, direct staking allows us to strategically select the node operators we work with. We can avoid those located in regions where we already have significant exposure
• In addition to the above mentioned risks, direct staking enables us to diversify our exposure to different Ethereum clients by distributing resources among them

Also, does use of Enzyme and direct staking preclude generating extrinsic yield on LST holdings?

I am not sure if I understand the question, do you mean if we can generate yield via LPs or farming? In that case, the answer is that direct staking positions won’t generate extrinsic yield via LPs or farming, however they can potentially generate a better yield than LSTs due to reduced fees.

---

Would note that the Mutual offers cover on Enzyme and likely both Lido and Rocketpool (and another ~10 LSTs to follow)

• The Nexus pool is covering Enzyme (it has done for a while) but we believe it is within the acceptable risk constraints for the DAO. In any case the cover buyer is inactive now since it was Celsius (redeemed almost everything from Enzyme) and in any case the cover rolls of in May anyhow.
• Currently the nexus pool (as far as we can tell from the Dune board) is not actively covering Rocketpool hence the proposed focus on rETH. It isn’t covering Lido either but given the Nexus collateral pool is already quite long stETH (c.20%) we deliberately left it out of our portfolio construction to stay within provided risk guidelines. The idea here is that we’ll continue to monitor the exposure periodically and adjust allocations if active cover suddenly spikes on a protocol that we’re using.

I hope I answered your questions! Let us know if you have any other questions.

[Rei]

Hey @Moss and Avantgarde team.

Thank you for the proposal & suggestions put forward, indeed important to use that chunk of ETH wisely.

Few requests for additional detail on some aspects:

Why is Enzyme a good route to making these investments? What value are we getting out of the 0.5% fee?

It made complete sense for Maple from my perspective - it was a bespoke smart contract integration that we didn’t have to develop and Enzyme managed the MPL reward selling/re-investing process on our behalf.

What’s the reason we wouldn’t just buy rETH using our CoW Swap integration and hold it directly? Same with Kiln.

What exactly do you mean by ‘Avantgarde management’ that’s included in the 0.5% fee? Could you elaborate on the management services provided, e.g. are you proposing an active monitoring process that would have an early warning system for withdrawing/changing assets?

Additional detail on the suggested investments

Would you mind analysing & summarising the two proposed underlying investments in a bit more detail for the community?

Specifics on why Rocketpool and/or Kiln instead of other alternatives (excluding Lido for the reasons you already outlined):

• Liquidity & options for withdrawal. How quickly can we get our investment out? What proportion can we withdraw at what value in what timeframes?
• Governance. What’s their set-up, who controls the funds, who are the key people, how do they make decisions, what are the critical points of failure.
• Validator set & associated risks
• Reward estimates & fees. How rewards are expected to change over time with some range estimates. Who earns the fees & how are they used?
• Detailed Risk Assessment. Comments on smart contract risk, governance risks, rETH de-peg possibilities & scenarios where we would lose funds or have limited liquidity. Also would be valuable to note what percentage we would make up of their total size.
• Oracles: Understand that we have the Capital Pool <> Enzyme vault link established, assume you just value the ETH in Kiln directly as ETH. What oracle do you use for rETH?

[GordonGekko]

It may be useful to break down what will become a recurring discussion into several threads:

1.) Integration(s) of Treasury with Oracles/NAV via Custodians/Asset Managers - Enzyme and ?

2.) Given broad consensus on (liquid?) staking ETH, what format should be used?

3.) Which LSTs should be whitelisted?

================================================

1.) It would be useful if someone can describe the existing integration with Enzyme, outlining the problem it was intended to solve, and the ramifications of this integration in terms of our options for staking ETH in various formats. Does it mean that ALL Treasury Assets must be held in Enzyme? Should the Mutual continue to offer coverage on Enzyme as VAR in Enzyme grows? What are other custodial/asset manager options for the Treasury where additional integrations would allow spreading risk beyond just Enzyme?
2.) Given that our Treasury can deal in full 32 ETH blocks, staking options for ETH would include:

• Direct staking of ETH via an institutional staking operator (like Kiln/$lsETH or RockX/$uniETH), but maybe also more decentralised direct staking option like EtherFi
• A diversified collection of individual whitelisted LST holdings held directly in the Treasury wallet
• Passive LST ‘indexes’ holding an (expanding) mix of LSTs, but harvesting only intrinsic yield (unlikely to offset additional contract risk and .25% fee)
• Active LST ‘baskets’ that also auto-harvest optimised extrinsic yield, AND automate peg arbitrage across a large number (10+) of underlying whitelisted LSTs → much higher yield to offset contract risk and fees (forthcoming)

3. Currently, there are about ~20 (and counting) credible LSTs soon to be available. Excluding CEXes and stragglers, here is a potential (white?)list:

Lido - $stETH
RocketPool - $rETH
Stader Labs - $ETHx
Stakewise v3 - $osETH
Frax - $sfrxETH
Liquid Collective - $lsETH
pStake - $stkETH
Ankr - $ankrETH
Swell - $swETH
Diva Labs - $divETH
Hord - $hETH
Stake.Link - $sdlETH
BiFrost $vETH
Manifold - $mevETH
SharedStake - $vETH
Tranchess - $qETH
Redacted - $btETH

Confronted with such an extensive list, the natural reaction is to say, ‘well, we will just go with the top one or two’. This is a very bad take. The Mutual’s entire future is bound up with the battle to maintain and increase the decentralisation of POS Ethereum. Bitcoin maxis claim that power law concentration is inevitable. Our Treasury investment policies should set an example for all other DAOs regarding Ethereum ‘ESG’ responsible investing. We should be held to that standard.

[BraveNewDeFi]

@GordonGekko

It would be useful if someone can describe the existing integration with Enzyme, outlining the problem it was intended to solve, and the ramifications of this integration in terms of our options for staking ETH in various formats.

The Enzyme vault was originally created to manage the investment in the M11 Credit wETH pool on Maple Finance. See the summary in the docs. The Enzyme team provided engineering work to create the integration with Maple Finance, so the mutual could deposit and manage the wETH in the lending pool.

Chainlink Proof of Reserves was used as the oracle, so the price of the assets deposited in Maple Finance could be called and tracked by the capital pool, which is critical since the NXM token price is based on the value of the assets held in the capital pool.

@GordonGekko

Does it mean that ALL Treasury Assets must be held in Enzyme? Should the Mutual continue to offer coverage on Enzyme as VAR in Enzyme grows?

No, it does not. The wETH in the Enzyme vault could be withdrawn back to the capital pool if members voted to take that action. You can see that the stETH is held direclty in the capital pool. Investment assets can be held directly in the capital pool, but each new investment requires engineering resources.

@GordonGekko

Confronted with such an extensive list, the natural reaction is to say, ‘well, we will just go with the top one or two’. This is a very bad take.

The constraint for any potential LST and/or directing staking allocations in the capital pool comes down to:

• Available liquidity. Can the mutual access the majority of the invested capital in the event these assets need to be liquidated to meet claim obligations? If so, is there enough liquidity to swap from a LST back to ETH without significant slippage?
• Reliable oracle. This goes back to my comment above. Because the NXM token price is based on the value of the underlying assets held in the capital pool, a relaible oracle needs to be available that isn’t prone to manipulation.

The other consideration: how much are the fees for each of the LST or direct staking options? What do members receive in exchange for the fee charged by each protocol, staking provider, or potential treasury management service?

@GordonGekko

Our Treasury investment policies should set an example for all other DAOs regarding Ethereum ‘ESG’ responsible investing. We should be held to that standard.

I agree with this, and diversifying exposure to staking providers also spread the exposure to risk broadly. The technical constraints are the main blocker for smaller staking providers. It’s a balance between working to decentralize the network, while ensuring there is a reasonable, timely path to liquidity from any of these proposed investments should claims obligations exceed the amount of liquid ETH held in the capital pool.

[elisafly]

Thanks for your feedback and insights. We’ve combined some of the responses for efficiency

Why is Enzyme a good route to making these investments? What value are we getting out of the 0.5% fee?

There are two points here;

• The bespoke smart contract that we built for Maple was to enable moving nimbly between positions (on Enzyme) whilst still having a Proof of Reserve Oracle reported back to the Capital Pool in order to correctly value NXM.
• The key value proposition is that the vault is actively managed and monitored. We would build a portfolio that is diversified both geographically and operationally (in particular thinking about the concentrated risks of hosting providers). Note that the vault will hold both LSTs and natively-staked ETH. This requires substantial diligence, research and ongoing monitoring. Furthermore, Nexus’s exposure is not static and changes constantly so it is necessary for the positions to be actively managed in order not to duplicate risks or take them above threshold. e.g. if tomorrow Nexus sells 20% cover on Rocketpool, and the DAO is already long rETH, Nexus would need to react in order to avoid overshooting the defined risk thresholds. This is an ongoing process that requires efficient active management.

Would you mind analyzing & summarizing the two proposed underlying investments in a bit more detail for the community?

Every piece of research we do on the validators looks at all the aspects you are asking about. We’d be happy to share an example of one of our investment memos privately and/or jump on a call for you to meet our team and get comfortable with their abilities.

Note that the number of possible options is large and we have not assessed them all yet but we are working on building out a research and monitoring library, which will not be made public.

Confronted with such an extensive list, the natural reaction is to say, ‘well, we will just go with the top one or two’. This is a very bad take. The Mutual’s entire future is bound up with the battle to maintain and increase the decentralisation of POS Ethereum. Bitcoin maxis claim that power law concentration is inevitable. Our Treasury investment policies should set an example for all other DAOs regarding Ethereum ‘ESG’ responsible investing. We should be held to that standard.

@GordonGekko, we couldn’t be more aligned on this. However, as was pointed out, oracle reliability and liquidity are must-haves that can’t be overlooked.
We are going out of our way to ensure that we can provide maximum decentralization to Nexus without compromising on quality. This week alone for example, we’ve spoken to several lawyers to understand the various risks facing hosting providers. We’ve also looked into bare metal based validators. We are looking into the integration of Stakewise V3 in order to add any validators permissionlessly whilst still being able to utilize the Proof of Reserve oracle. This would let NXM be priced accurately without requiring any extra engineering resource from the mutual.

This RFC proposal was deliberately kept high-level because there’s a lot of research (and possibly subsequent development work) to be done to get you (and us) happy with the diversification and decentralisation aspects.

The key point is that currently the wETH is idle and there isn’t a clear plan to move forward. With funds already deployed into an Enzyme vault that utilizes the PoR oracle, there is a readily available structure to do it nimbly and our team is offering to support that. We’re proposing to take the lead on researching a path whose aim is 100% aligned with Nexus: diversification, active risk monitoring and contribution to the decentralization of the Ethereum network.

[BraveNewDeFi]

Thank you for your detailed answers, @elisafly

Some follow ups:

The key value proposition is that the vault is actively managed and monitored.

Using the Maple investment as an example:

Pros

• Enzyme provided the development resources to create a bespoke integration complete with Chainlink’s Proof of Reserve oracle, so the capital pool could track the price of the Maple deposit. This allowed the Engineering team to have limited involvement, so their resources could be focused elsewhere.
• The Avantgarde team harvested the MPL rewards and swapped them to wETH on a regular basis to capture that value in wETH terms. Your team also deposited the initial wETH in the Maple pool.

Cons

• Between the initial deposit date in August 2022 and the aftermath of FTX (November 2022), the Nexus Mutual community was unaware that the concentration of lenders in the pool had gone from ~5-6 funds to largely just 2 (Orthogonal and Auros) with limited liquidity allocated to the Amber Group and Flow Traders. In the case on actively monitoring this sort of development, members were not aware and once it become well known that the mutual was exposed to these two firms, it was too late to withdraw without sizeable losses to the initial investment.

The above was a unique situation, but defining what active monitoring and management would mean in terms of ETH staking would be helpful. Since natively staked ETH can be fully withdrawn but it’s subject to a queue, which current is ~16 days (Source: rated.network). I imagine this queue for full withdrawals is longer if there are more requests. LSTs can be swapped on the market to access liquidity in lieu of withdrawing, but that can come with it’s own pros/cons.

I think that Enzyme could provide a good solution, but before moving to a formal proposal, some of the specifics on what active montioring and management mean specific to this investment type, how Avantgarde would respond in a situation where risk dramatically increases for a staking provider in a short period of time or if the mutual needs to access liquidity to meet claim obligations would give members a better idea of the value that Avantgarde Treasury can provide.

Smart Contract Risk

Furthermore, Nexus’s exposure is not static and changes constantly so it is necessary for the positions to be actively managed in order not to duplicate risks or take them above threshold. e.g. if tomorrow Nexus sells 20% cover on Rocketpool, and the DAO is already long rETH, Nexus would need to react in order to avoid overshooting the defined risk thresholds. This is an ongoing process that requires efficient active management.

Currently, Nexus Mutual does not provide Protocol Cover or ETH Staking Cover for Rocket Pool. If the mutual had exposure to rETH, those potential lines of business would be more heavily scrutinized, and it’s likely that the available coverage for those types of risk would be limited.

While Rocket Pool isn’t listed, currently Enzyme V4 is listed. That potential for concentration risk would force a decision, imo, that if Enzyme was used to manage a large portion on the capital pool, Enzyme V4 would need to be delisted from Nexus Mutual to avoid overlapping risk exposure.

On that note: I wouldn’t be supportive of allocating 20% of the capital pool to the Nexus Mutual Enzyme vault, as members would have diversified exposure to risk across multiple staking providers or LST protorocols, but it would be a sizeable concentration of risk in the Enzyme smart contracts if an exploit on Enzyme were ever to occur (knock on wood that never happens). At most, I would be comfortable with the amount of ETH currently held in the Enzyme vault, but I would not like to see an increase.

At this point, I’m confident rETH could be held directly in the capital pool and there is now an rETH/ETH oracle available through Chainlink, so the reliable oracle problem has been resolved for Rocket Pool ETH. If members were inclined to see an investment allocation to rETH, available liquidity would be the only real constraint and, if members approved a reasonable allocation through governance, rETH could be purchased through the SwapOperator in the capital pool, with ETH traded for rETH via CowSwap.

If the Enzyme vault were used, it would be largely beneficial to pursue staking options that would not be available due to a lack of a reliable oracle source (this is where Chainlink Proof of Reserve would be beneficial) and which require more complex transactions. Minority staking solutions, where the mutual can advance diversity, while reducing overall fees for the mutual would both help promote geographic, execution client/consensus client, and staking provider diversity and offset the 0.5% management fee for the services provided by Avantgarde.

Overall, it’s an interesting proposal and I’m looking forward to allocating more idle (w)ETH to staking solutions, so more of the capital pool can be made productive. As with anything, the devil is in the details and it’s important for members to understand certain specifics, so they can provide feedback before this moves to a formal Nexus Mutual Protocol Improvement Proposal (NMPIP).

Engineering Support

It would also be beneficial to understand if/how much the Enzyme/Avantgarde Treasury team would need support from the mutual’s Engineering team to move this forward, should members approve this proposal. If you can comment on the potential technical requirements and if the mutual’s Engineering team would need to be involved (and, if so, at what level), that would be an important factor for members to consider, since there’s strong support to prioritize the Tokenomics Revamp initiative.

[elisafly]

Thanks for all the comments @BraveNewDeFi! To respond to some of your points;

Between the initial deposit date in August 2022 and the aftermath of FTX (November 2022), the Nexus Mutual community was unaware that the concentration of lenders in the pool had gone from ~5-6 funds to largely just 2 (Orthogonal and Auros) with limited liquidity allocated to the Amber Group and Flow Traders. In the case on actively monitoring this sort of development, members were not aware and once it become well known that the mutual was exposed to these two firms, it was too late to withdraw without sizeable losses to the initial investment.

The Maple strategy was not one that we proposed and our mandate was to write the code to bridge the two protocols and then to execute on the MPL yield strategy that was initiated by the community. We were not tasked with actively monitoring or managing the portfolio on a discretionary basis but in hindsight, it would have been a valuable service given how the pool concentration changed. Actively monitoring does add to our cost base substantially (good monitoring tools are not cheap and require substantial attention and maintenance). However, we believe that we can still do this competitively because there are synergies in other parts of our business that can help bring costs down here.

What does active monitoring and management look like?

This is a great question:

From a management perspective we’d be looking for the following:

• Past track record at staking
• Regional exposure of hosting provider
• Any changes in the Nexus cover profile that require us to adjust our own exposure (eg. selling cover on rETH might mean reducing exposure hypothetical exposure to rETH).
• Diversifying between hosting services vs bare metal providers (to the extent possible)
• Ensuring Geographic diversity
• Reacting to any cause for concern (eg. slashing occurs which is unexplainable and/or likely to be repeatable)
• (LST’s only) reacting to any concerns with respect to protocol security.
• Understanding, monitoring and negotiating fees charged by validators (where applicable)

From a monitoring perspective;

• Define which risk elements need to be monitored and build an internal dashboard for this
• Build an alert system to notify us automatically
  • Eg. Actively monitoring regulatory events, security alerts, governance proposals, withdrawal queues, LST liquidity, hosting service outage, slashing and reacting as necessary.

“While Rocket Pool isn’t listed, currently Enzyme V4 is listed. That potential for concentration risk would force a decision, imo, that if Enzyme was used to manage a large portion on the capital pool, Enzyme V4 would need to be delisted from Nexus Mutual to avoid overlapping risk exposure.

I wouldn’t be supportive of allocating 20% of the capital pool to the Nexus Mutual Enzyme vault, as members would have diversified exposure to risk across multiple staking providers or LST protorocols, but it would be a sizeable concentration of risk in the Enzyme smart contracts if an exploit on Enzyme were ever to occur (knock on wood that never happens). At most, I would be comfortable with the amount of ETH currently held in the Enzyme vault, but I would not like to see an increase.”

I understand your concern here but for context, the Enzyme cover in question was purchased by Celsius before they filed for bankruptcy. It was for a vault they managed which subsequently they had to close due to the bankruptcy procedures. We can either delist enzyme insurance from Nexus or limit it to a much smaller number (eg. $2-5m). Also I believe the Celsius exposure rolls off in May and will not be renewed. In any case, there shouldn’t be a reason not to increase AUM up to 20% of the Capital Pool over time as long as we are aware that we have to limit exposure on the protocol insurance side…

If the Enzyme vault were used, it would be largely beneficial to pursue staking options that would not be available due to a lack of a reliable oracle source (this is where Chainlink Proof of Reserve would be beneficial) and which require more complex transactions. Minority staking solutions, where the mutual can advance diversity, while reducing overall fees for the mutual would both help promote geographic, execution client/consensus client, and staking provider diversity and offset the 0.5% management fee for the services provided by Avantgarde.

This makes a lot of sense. We can certainly build a portfolio based on a mandate of direct staking options that do not have oracles and let the DAO interact with rETH directly.

It would also be beneficial to understand if/how much the Enzyme/Avantgarde Treasury team would need support from the mutual’s Engineering team to move this forward, should members approve this proposal. If you can comment on the potential technical requirements and if the mutual’s Engineering team would need to be involved (and, if so, at what level), that would be an important factor for members to consider, since there’s strong support to prioritize the Tokenomics Revamp initiative.

The way this conversation is going, I think the fastest way we can achieve the goals being discussed is to integrate Stakewise V3 which will give us a permissionless way to point to different validators. This will likely take 8-10 weeks (integration + audit) but opens up all the direct staking possibilities. This also gives us enough time to build the necessary research library for the starting portfolio as well as the monitoring infrastructure.

To answer your question more directly, we do not foresee the need for the involvement of the Mutual’s Engineering Team, which can stay 100% focused on the current priorities at hand.

One additional question I have is whether Nexus DAO would like us to take on a more holistic role and monitor the entire risk of the DAO regularly. This is something not currently proposed in the scope of our work. I’m thinking it might be useful to have periodic reports (eg. monthly) which highlight how the exposure has changed based on new premiums sold, current cover exposure, positions rolling over. Together with that, we could provide recommendations necessary for the DAO to remain within its risk thresholds. For this extra assignment it might make more sense to explore a fixed monthly fee (vs. AUM based).

[BraveNewDeFi]

@elisafly

The Maple strategy was not one that we proposed and our mandate was to write the code to bridge the two protocols and then to execute on the MPL yield strategy that was initiated by the community. We were not tasked with actively monitoring or managing the portfolio on a discretionary basis but in hindsight, it would have been a valuable service given how the pool concentration changed.

Thank you for the clarification! It’s important context, and I know a few members shared that they thought Avantgarde was actively monitoring the Maple position for the mutual, so I really appreciate the clarity here

@elisafly

I understand your concern here but for context, the Enzyme cover in question was purchased by Celsius before they filed for bankruptcy. It was for a vault they managed which subsequently they had to close due to the bankruptcy procedures. We can either delist enzyme insurance from Nexus or limit it to a much smaller number (eg. $2-5m). Also I believe the Celsius exposure rolls off in May and will not be renewed. In any case, there shouldn’t be a reason not to increase AUM up to 20% of the Capital Pool over time as long as we are aware that we have to limit exposure on the protocol insurance side…

My comment wasn’t about the Celsius cover buy per se, but just best practices wrt to risk management. Since the mutual would be exposed to Enzyme v4 smart contract risk, providing cover for that risk would potentially amplify losses should an exploit occur. Having 20% of the mutual’s capital in an external smart contract and then allocating to other strategies with smart contract risk creates additional exposure to risk. The more smart contract layers, the more risk the mutual is exposed to.

@elisafly

To answer your question more directly, we do not foresee the need for the involvement of the Mutual’s Engineering Team, which can stay 100% focused on the current priorities at hand.

If there is a proposal to increase the ETH allocation to the Nexus Mutual Enzyme vault, then there would be some need for their involvement since the capital pool cannot send out funds without smart contract work being performed. If the allocation remains as-is, then I don’t believe the Engineering team would need to be involved, but I’ll leave that to @Hugh or someone else from the Foundation team to confirm.

@elisafly

One additional question I have is whether Nexus DAO would like us to take on a more holistic role and monitor the entire risk of the DAO regularly. This is something not currently proposed in the scope of our work. I’m thinking it might be useful to have periodic reports (eg. monthly) which highlight how the exposure has changed based on new premiums sold, current cover exposure, positions rolling over. Together with that, we could provide recommendations necessary for the DAO to remain within its risk thresholds. For this extra assignment it might make more sense to explore a fixed monthly fee (vs. AUM based).

Currently, the DAO R&D and Marketing teams are working on Dune Dashboards that track cover exposure in detail, as well as analytics about staking across the various staking pools. Previously, members voted to end an ongoing partnership with dedaub due to cost and its impact on the DAO treasury holdings. You can read through the Reviewing treasury expenditures | Dedaub smart contract monitoring service for more info. My educated guess would say the potential for Avantgarde to do this would be highly dependent on the fixed cost and how Avantgarde would focus on risk (e.g., changes in risk at the smart contract layer or changes in overall exposure to risk based on cover sales?). The protocol does have risk exposure limits and the Advisory Board is able to change certain parameters to better control exposure to global risk levels should risk change quickly for certain products. You can read more about capacity controls in the docs.

Overall, I just want to say that I appreciate your responses and the clarify you’ve provided, @elisafly

[Moss]

Thank you for your comments @BraveNewDeFi, we also appreciate your efforts to provide thorough and constructive feedback.

Comments on the ongoing thread

Regarding the combined smart contract risks of the proposed strategy, we think it’s important to put things into further context.

@BraveNewDeFi

Since the mutual would be exposed to Enzyme v4 smart contract risk, providing cover for that risk would potentially amplify losses should an exploit occur. Having 20% of the mutual’s capital in an external smart contract and then allocating to other strategies with smart contract risk creates additional exposure to risk. The more smart contract layers, the more risk the mutual is exposed to.

The code base of Kiln (and Stakewise V3 already, which is already available to us) is extremely simple and therefore their additional SC risks are marginal in the context of this discussion. In other words, while we certainly cannot argue that there would be no additional SC risks, we can confidently defend the idea that those risks would be minor and not comparable to looping assets through multiple complex smart contracts in the way that it is assumed here.

@BraveNewDeFi

If there is a proposal to increase the ETH allocation to the Nexus Mutual Enzyme vault, then there would be some need for their involvement since the capital pool cannot send out funds without smart contract work being performed

With regards to the need for extra engineering work:

• For depositing new funds into the Enzyme vault: we’re not aware of such obstacles but we’re happy to look into the code and help wherever needed to make it happen and minimize the workload on your side.

• For allocating funds already on the Enzyme vault: we can certainly say that there is no effort needed to execute the strategy with the funds that are already in the vault.

In conclusion, we are positive that a gradual and progressive AUM ramp-up would be beneficial for both parties, as it would allow the mutual to increase the utilisation rate and flexibility of the capital pool over time without substantially increasing the overall smart contract risk exposure. However, if there is a broader consensus to keep the size as it is for the moment, that is also fine with us.

Further considerations

We understand that the additional services that we may offer are currently outside of the scope of this proposal so we will evaluate with the community at a later stage.

We understand that having a more detailed investment memo about Kiln (and in the future about Stakewise V3) is key for the whole community to get comfortable with our risk assessment process as well as serving as future reference to avoid misunderstandings and confusion. In order to move forward this proposal to “vote readiness”, we’ll produce a more formal investment dossier that can be viewed and referenced to by the whole community.

[Moss]

Hi @BraveNewDeFi @GordonGekko @Rei @Hugh and all,

as promised in our latest post, please find our investment memo, which includes our criteria, objectives and key results. You can find the full PDF report here and the most salient part of the report here below.

Please let us know if you have any comments or if you want to see anything more/else from this report. Happy to amend it if required.

Also, we’re keen to hear more feedback on the AUM ramp up before proceeding to the next phase of the governance process.

[Rei]

Thanks @Moss

We spoke about this privately, but wanted to re-emphasise that the most likely way in which Enzyme can add value here is with managing the direct staking integrations at a combined fee lower than the main LSDs on the market.

With regards to the ramp-up of AUM, imo the most productive immediate conversation would be about using the current assets already in the Enzyme pool. I personally believe ~9% of the pool would be a reasonable amount to deploy in such a strategy and committing to a ramp-up of some sort would only complicate the proposal at this stage. This doesn’t preclude us from using the Enzyme vault in the future for the right opportunities, but I would be against committing to a schedule.

Look forward to a governance-ready proposal & thanks again for all the work you guys have put into this!

[GordonGekko]

I think the entire LST investment strategy of the Mutual needs a master framework, likely dividing upcoming staked ETH investments into tranches emphaisizing different aspects of yield, risk, liquidity and composability. No single tranche can emphasise all 4 of these objectives with a single holding. But a mix of staked ETH tranches could achieve a blended result. Given the sunk cost of the Enzyme integration, this vault may be useful for holding emerging LSTs that do not have oracles and cannot report directly to NAV/book value. But the Mutual is large enough to spread contract risk and gain liquidity by investing directly in a diversified spread of proliferating LSTs, as long as the LST has an oracle reporting to NAV. Those LSTs that don’t (yet) can go into the Enzyme vault in the meantime to encourage decentralisation. So let’s decide how much is 100% of the staked ETH portfolio, how much is fee savings oriented via direct staking (but no liquidity and composability), and how much liquidity and composability oriented via various tranches of LSTs, likely including Lido (capped), Rocket Pool, Swell (has oracle already) etc. Advisors need to propose how many emerging LSTs should go into the Enzyme vault in the meantime. I am not a believer in passive LST indexes, because they are not offering any LSTs that do not have oracles, and they are passively managed, so do not pursue peg arbitrage or other extrinsic yield strategies that would justify their fees. The burden is on the investment advisors to propose a combination of direct staking deals AND LSTs the Mutual cannot (yet) hold directly, and the process by which this would be managed, as more successful LSTs will eventually migrate from Enzyme to direct holdings by the Mutual.

[Moss]

@BraveNewDeFi @Rei @GordonGekko thanks for your feedback. For the future records, the discussion of this thread has now been moved to [RFC2]: Restart Enzyme vault & diversify ETH across staking providers.