Solving wNXM discount while maintaining reasonable MCR%

[Muir]

Introduction

The purpose of this post is to adjust the way NXM is bought back from the market.

Problem

Currently wNXM is trading at a large discount to NXM due to the MCR% remaining under 100% for an extended period of time. The only people who are able to sell to the bonding curve are bots who sell immediately when the MCR% rises above 100%. This is extracting value from members of the Mutual who are forced to sell wNXM at a large discount if they need liquidity. While the long term solution to this is to increase the MCR%, there are short term solutions that can benefit both the Mutual and members who would like to sell.

Proposed Solution

There are 3 different wNXM scenarios that need to be accounted for:

1. wNXM is trading below book value (As it is currently)
2. wNXM is trading above book value but below bonding curve price
3. wNXM is trading above bonding curve price

Scenario 1.

In the case where wNXM is trading below book value, it makes sense for the mutual to repurchase as much wNXM as possible. This increases the book value of everyone else’s NXM because the circulating supply decreases more than the Net Asset Value of the mutual. (Everyone’s NXM will be redeemable for more ETH if the mutual were to dissolve)

Scenario 2.

While wNXM is trading above book value, but below the bonding curve price, the Mutual should only be purchasing NXM from the open market (Buying wNXM and not buying on the bonding curve). To make this happen, I suggest implementing a lower MCR% threshold of 95% for the Mutual to purchase wNXM using the capital pool. Whenever the MCR% is above 95%, the mutual will purchase wNXM off the market as long as the price is below the bonding curve price. This means that arbitrage bots will not be able to redeem NXM against the bonding curve because MCR% will stay below 100% until wNXM trades in line with NXM, and that arbitrage profit will instead go to Mutual members.

Scenario 3

In the case where wNXM is trading at or above the bonding curve price, the Mutual should not be purchasing wNXM. In this scenario, people will be buying NXM from the bonding curve and selling it to the market as wNXM to equalize the prices.

Additional Notes

As the Mutual buys wNXM from the market, it also would make sense to add these tokens as liquidity to an wNXM/ETH AMM. (Bancor/Sushi/Etc). This would deepen the wNXM liquidity and earn fees for the protocol, without actually taking on any additional downside risk. (Since the Mutual would be holding eth anyways and there is a set Eth value for each wNXM via the bonding curve)

It should be noted that any attempt to buy wNXM below book value would probably be front run by speculators, so the actual capital requirements of raising the wNXM price to book value would be low. (although the more wNXM that can be bought below book value, the better for Mutual members)

The buying of wNXM could be fully automated using on chain keepers, or manually triggered by members of the Mutual. In either case, specific parameters for purchasing could be defined in a smart contract. Some thought would need to be put into ensuring that wNXM buybacks were not front run.

Conclusion

The best long term solution is to raise the MCR% via capital inflows, but this will take a significant amount of time, and it is beneficial to Mutual members to have liquidity now. The solution I propose solves the issue of liquidity while also generating significantly increased revenue compared to the current model. I would love to hear any comments or alternative solutions to this issue.

It has been mentioned that there may be Legal/Tax limitations to this approach. Any details regarding this would also be very helpful.

[Gauthier]

Hey Muir glad to have you on board and thanks for your proposal. The Investment Committee (recently set up) has been looking into this issue and I think generally the community agrees that something should be done on this front.

A few comments :

Scenario 1.
Completely make sense to buy wNXM below book value as it is value accretive for the mutual.

Scenario 2.
I don’t agree here. Buying wNXM above book value means that we are trying to artificially maintain a price for the token, using our Capital Pool. In a long enough timeframe, it leads to the distribution of our Capital Pool to the market.

alternative : we could hold at 0 value (meaning it counts as 0 in the MCR calcs) the wNXM bought below book value. This way, MCR drops below 100% and arbitrageurs are no longer able to redeem cash inflows from the bonding curve.

Liquidity
Also agree something should be done on this front. However, providing liquidity on Bancor/Sushi would mean providing liquidity for all prices. The issue is we don’t need to provide liquidity above MCR=100% prices. Also, we don’t want to provide liquidity for prices below book value. Here is a representation of what I mean :

The solution would be to provide liquidity on Uniswap v3 for the range of prices between book value and the bonding curve price.

Implementation
The team will have better insights about this but development resources are a real bottleneck. In order to automate everything, it would require to do direct integrations with the Capital Pool and lots of development.
But it would still possible to do it custodially/with a multisig and it would be way easier.

Ultimately, if the lights are green on the legal/tax front, it will be up to the community to vote on this! Happy to have this discussion!

[Muir]

I really like the idea of using uniswap v3 to concentrate liquidity between book value and bonding curve value. Very little additional ETH would be required to do this if the wnxm is added at book value.

The reason I suggest purchasing wnxm above book value is because it is currently being done at a much less favorable rate via the bonding curve. You could still limit purchases above book value to an MCR% threshold to prevent ETH being drained. I think using excess eth above the MCR% to buy back from the market is a good narrative and provides some positive price pressure to keep wnxm above book value.

You could also alternatively just not allow redemptions on the curve and continuously buy wnxm up to the book value, which would increase the book value over time.

I agree that for a short term solution it may make more sense to just make any trades from a multisig, although I’m not familiar with how it would get access to the capital pool.

One additional note is that bought back wnxm needs to be effectively burned to raise the book value of NXM. I think we could still LP with it though as long as it is considered burnt.

[target]

Something to think about: how would ‘profits’ from purchasing wNXM, burning / redeeming equivalent NXM and the net gained ETH be distributed to members? It cannot go into the capital pool, would it go to a seperate contract? What would be criteria to distribute it from here, if everyone wants to sell with this dedicated contract, there would be no point in setting this up. Having a queue system also might be a bit complicated and not considered ‘fair’.

Correct me if I am wrong, but all these proposals not necessarly provide more value to members as you cannot know in what direction wNXM will go, it can keep on going down in wNXM/NXM ratio anyway.

I would not complicate things, let the market decide the value and just focus on marketing and education, e.g. wNXM can be unwrapped at a huge discount to purchase covers.

[Muir]

Purchasing wnxm below book value and burning it increases the book value of nxm for everyone. AKA everyone benefits without actually having to distribute anything. There’s really no reason to not purchase below book value, the question is really what to do with that wnxm and whether to stop at book value.

I’m not sure why Eth profits from market making couldn’t be added to the capital pool, but if that is the case it would complicate LPing in the case where all wnxm is converted to eth.

It doesn’t really matter which direction the price of wnxm goes if the LP is set up right. At book value the mutual would be 100% wnxm and purchasing any below book that is available. At bonding curve price, the mutual would be 100% eth. The price of wnxm would naturally float between those prices, and the book value would rise as capital is added to the pool and wnxm is bought below book.

[Josh]

Very interesting proposal. Thank you, Muir! While I think there are a lot of great suggestions here, my experience with governance proposals and negotiations suggests that we should really focus on passing a clear, actionable item that meets dev resources and the community’s ability to understand. Given this, Scenario 1 + Uni v3 would be a great first proposal. This scenario, which exists now, is understandable and the impacts are much clearer to NXM holder value + fund operations. Additionally, the consequences of the proposal in this scenario offer more certainty in terms of what incentives will propagate when compared with Scenario 2. Finally, getting Scenario 1 + Uni v3 passed would be a benchmark for future tweaks and a study of its effects would give more confidence with what to do in Scenario 2.

On another note, if the multi sig provides wNXM in the Uni v3 pool, is it really burned? What if big trades are made over this pool, even if within the suggested price bandwidth? Couldn’t there be a scenario where transactions fundamentally unburn it?

Happy to hear pushback or tweaks to the above! Thank you again.

[Muir]

Thanks Josh, appreciate the input. I think scenario 1 + uni v3 would be a good starting place, and maybe dealing with the bonding curve and additional buybacks could be separated into its own proposal. (I’m leaning more in favor of just disabling selling on the bonding curve instead of buying back above book, and using wnxm as the primary tool for liquidity. Book value would increase naturally overtime for all holders as more capital enters the pool, and MCR% rises)

In regards to uni v3, you are right that adding to the uni v3 pool wouldn’t actually burn the wnxm. Instead, the mutual would be essentially selling that wnxm back to the market above book value. This is even more beneficial to book value than just burning it, assuming that Eth in the LP could be considered part of the capital pool.

I think the mutual might be hesitant to “sell back” to the market. But overtime this will lead to deeper liquidity for the wnxm/eth pair, increase book value, and if it’s done right it won’t require much ETH from the capital pool.

There is also a reluctance to abandon the bonding curve price, but I don’t think it’s relevant anymore except for the minting of new nxm. If v2 is as successful as people hope, the book value, and therefore the price that nxm can be redeemed at via wnxm, will increase continuously with this new strategy.

[target]

@Muir

Purchasing wnxm below book value and burning it increases the book value of nxm for everyone. AKA everyone benefits without actually having to distribute anything. There’s really no reason to not purchase below book value, the question is really what to do with that wnxm and whether to stop at book value.

Thank you for clarifying, however is that not exactly what these arbitrage bots are doing too? They are burning the NXM reducing supply.

Still does not help the members to sell their NXM so not sure what we would achieve here. It does not solve the Problem / desired outcomes put forward.

I am also a bit fearful we might mess up here (e.g. introduce a potential scenario we have not thought about). Surely if people value NXM, they would buy wNXM coming to the same outcomes. And that could come with education and more exposure.

[Muir]

No, the arbitrage bots are buying wnxm below book and redeeming it above book. The nxm is burnt but the capital pool also reduces in size by a greater amount than book value. This means the arbitrage bots are keeping the book value constrained to the MCR limit.

If the mutual purchases below book, EVERYONE in the mutual benefits. If we purchase below book value and burn, the decrease in outstanding tokens is larger than the decrease in capital pool size. Each NXM then has a larger claim to the remaining underlying eth. If we decide to LP with the bought wnxm instead of burning it, all of it will eventually be converted to eth at higher than book value, which increases the book value for all holders.

Buying below book value will create a psychological floor for the price of wnxm, and it will likely trade between book value and bonding curve value based on the expected rate of increase of the capital pool. People who want to exit can always exit at book value or above (like $100 right now), but I think the desire to do so will be much lower if people know they have will have liquidity when they want it. It is pointless to try to allow people to redeem at the bonding curve price, there isn’t enough money to let people do so. BUT, This strategy will eventually lead to the convergence of the NXM bonding curve price and the wnxm price.

I agree, there is always potential to mess up something like this. The main concern I see is running out of allocated ETH to buy wnxm below book price. Technically enough ETH is available in the capital pool to buy at or below book, but the MCR needs to stay reasonable. While this case would end up benefitting all holders in terms of book value, it would be damaging to the mutuals ability to provide cover.

I appreciate the questions and concerns, please let me know if any of this didn’t make sense.

[Josh]

Thanks Muir! Interesting to hear your updated thinking on Scenario 2.

I want to focus on your second paragraph. I’m happy to see you agree with me in that we are technically not burning wNXM. And, while yes, we could just go this route and increase liquidity like you say, I think we lose some of that potential behind your idea. Burning it, or essentially removing it from the outstanding supply, is central to what you have proposed. Therefore, I would love to work with you and the community on looking into DeFi 2.0 solutions to understand how we can use this would-be burned wNXM to increase liquidity without increasing supply. We could:

• Use a Rari Fuse pool
• Convert the wNXM into Toke (via a treasury deal with the Tokemak team). We then petition to have a Tokemak reactor where we can direct liquidity to the ETH-wNXM pool on a 10:1 basis without having to sell the wNXM (this would be kept in the Tokemak treasury)
• Assess opportunities pertaining to OHM bonds
• Discuss potential collaboration with Alchemix on some type of suitable product

Using the primitives of DeFi 2.0, I really think that we can obtain this goal that the proposal so boldly set forth without having to be exposed to ways in which we are not even doing what we propose.

Let me know your thoughts.

[Gauthier]

Hey Josh, thanks for your input!

My view about what to do with wNXM after a buyback is that we have options :

1) burn it
2) hold it for future use
3) LP with it
4) send it to Community Fund to fund grants
…

@Josh

Burning it, or essentially removing it from the outstanding supply, is central to what you have proposed

I don’t think it is correct. We are in crypto and not in traditional markets. I’d argue that the psychological effect of guaranteed liquidity through the buyback has much more impact than the “accounting” perspective (due to the inefficient market we operate in). There is no guarantee that burning wNXM is the key to improve wNXM value or set a floor. As reference, the MKR buyback-and-burn program.

As a result, I believe that the buyback should be an ongoing thing instead of a one-time event : “every time wNXM trades below book value, Nexus can buy back”. Starting this way would at the very least allow us to assess if we can set a proper floor, without destroying value in the process.

Concerning DeFi 2.0 solutions :

• I’ve investigated OlympusDAO bonds but it only works for protocols renting their liquidity (i.e. with token incentives). Bonds convert rented liquidity to protocol owned liquidity, in exchange for a discount on the token. As Nexus doesn’t incentivize liquidity, we have no use for bonds.
• Tokemak seems interesting, worth investigating!
• I don’t really see how we could benefit from Fuse and Alchemix? I’m not that familiar with these protocols. Interested to know more

[Muir]

I think I agree with Gauthier here regarding burning, and would also add that using the bought wnxm for anything in DeFi 2.0 wouldn’t count as burning it either.

I do like the idea of partnering with either tokemak or Olympus Dao, but we do have a substantial differentiator right now: we know roughly where the price should trade. I’m not familiar enough to comment on bonds or reactors, but uni v3 would provide significantly more capital efficiency than a constant product AMM like sushi or uni v2 if we set the bounds at book value and bonding curve price.

I still think the bonding curve is relevant for new minting of wnxm once book value and bonding curve price converge. When that happens, the Uni v3 position could be constrained to extremely narrow bounds, giving deep liquidity near bonding curve price.

I also agree that buy backs need to be ongoing to effectively set a floor. Every buyback below book value would also move book value closer to the bonding curve price, as long as the wnxm is burned or sold above book.

[target]

@Muir

No, the arbitrage bots are buying wnxm below book and redeeming it above book. The nxm is burnt but the capital pool also reduces in size by a greater amount than book value. This means the arbitrage bots are keeping the book value constrained to the MCR limit.

Thank you for taking the time to respond.

Kindly help me to understand: How does the capital pool reduce in size greater than book value? I do not think this is correct, and even that is the case any member would have sold anyway, it just keeps the MCR% at 100% keeping the exact same ETH in the capital pool.

Reading also comments below, I still do not see how members benefit while we introduce additional complexities and potentially risks. All references to burning, that is exactly what is now happening anyway.

In addition, I would question the ‘ethics’ of basically front-running the free market.

We already have agreed on the Community Funding.

Can we just not keep things simple? Everything has worked just fine, we are getting covers and people are getting premiums on their stakes. Let the market decide what happens and just wait for Nexus Mutual 2.0!

Something for the community call this week?

[BraveNewDeFi]

Absolutely. We can allocate some time to discuss potential solutions to the wNXM discount issue and talk about the various approaches described in this forum discussion. I hope you’ll all join the Mutant Meetup on Tuesday at 10am EST in the Nexus Mutual Discord!

[Muir]

I’m definitely interested in discussing this further during the community call.

Kindly help me to understand: How does the capital pool reduce in size greater than book value? I do not think this is correct, and even that is the case any member would have sold anyway, it just keeps the MCR% at 100% keeping the exact same ETH in the capital pool.

The current bonding curve price is .0383 eth, while the book value of NXM is .0235 eth ( 162,411 ETH / 6,901,940 NXM). This means that selling to the bonding curve gives someone more than their “fair” share of the Mutual’s holdings. While it is a mutual “member(s)”, that are benefitting from this using their bots, it is not beneficial from the mutual as a whole, and actually detrimental to all other mutual members who are never able to sell to the bonding curve. If this was prevented, the book value would begin to increase again for all members as the capital pool grew, and everyone would benefit equally.

The proposal I have suggested wouldn’t front run anyone, and the main point is to shift the focus from this useless bonding curve price to the book value, which can be increased. Every wnxm bought below book and sold above book or burned increases the book value of everyone else’s NXM, and therefore increases the value that their NXM can be redeemed at. The mutual doesn’t have any advantage in the free market, the market is just not recognizing the book value of wnxm because of NXM’s illiquidity.

Earning premiums for staking is great, but it is useless if NXM can only be redeemed for 1/3 of its “value”.

[glenn]

Glad to see this discussion happening (I lurk around this forum since I am a wNXM holder). I am mainly active in the Bancor community and just wanted to chime in with regards to any buybacks that might happen on the open market.

We have recently had a few projects that have put their treasuries (UMA, Harvest, and BarnBridge come to mind) in their bnt-token pool since the Bancor protocol co-invest together with LPs which means that you only have to add liquidity single-sided (more capital efficient since you free up half your stack). This allows these projects to provide liquidity with their treasury and get fees on assets that would otherwise sit idle with very little risk. Some of these projects are also using their pools on Bancor for any buybacks (e.g. Harvest) on the open market as essentially any fees generated by this activity goes back to them (they are LPs in their pools).

Additionally, the Bancor DAO is very supportive of projects that have integrated any buyback mechanism for their pools on Bancor and have voted favorably in incentivizing these pools with BNT. The DAO is also very open to increasing the co-investment limits on any of these pools to grow the depth/liquidity as required. Note that an added benefit of LPing on Bancor is that any deposits after 100 days are protected 100% from any IL experienced.

Happy to answer any questions from community members and thanks for this great thread.

[dfmerin]

Glad to see this discussion happening. Buying back wNXM below book seems like an obvious way to create value for our community akin to how Berkshire routinely buys back its shares when they trade below book value.

As Muir highlights, in the current system the bots are hostile to the community rather than helpful. It seems hard to argue against the wNXM buyback below book in an informed and good faith way. Aside from growing the book value for our members, this program would also help improve liquidity for members, increasing the NPS among members and helping accelerate growth of the mutual.

The Uni v3 pool between book value and bonding curve value is also a good idea, but perhaps these two ideas should be separated for the purposes of simplifying the governance proposal, given that the buyback program, which should be continuous IMO, would likely be easier to implement / align on.

[Josh]

Thank you for your input and for shedding light on some new opinions! I see this will be brought up on the Community Call, which hopefully means some serious research into this will commence. Therefore, I’ll refrain from adding too much more until there is certainty on what actions are being considered.

The one thing I do want to point out about this Uni v3 mechanism is that it leaves the Mutual exposed to potential attacks / market circumstances where a vicious cycle could occur that forces the Mutual to buyback and LP wNXM, while draining ETH. In a way, it is reminiscent of what happened with TRIBE / FEI in 2021 H1 when the peg was broken and the PCR had to be substantially drained to maintain it. Also, we would be exposed to oracle-based attacks or MEV, but these seem to be less likely and our knowledge of how they work provides some defense.

If we do decide to pick the Uni v3 path, an adequate study of risks and robust discussion would be necessary. If all is good after that, then this could be one of the possible great steps to getting the wNXM discount sorted.

[target]

I love these discussions and can understand not able to sell NXM is painful at the moment.

Just on another note: is Nexus Mutual not distancing themselves from wNXM as wNXM is basically another project?

I really just can’t help myself, this proposal to solve the wNXM ‘discount’ should not be a worry of any member, it is a risk we all signed up for. More covers, better education, marketing, etc will all solve this anyway. We now even have a contract to buy covers with wNXM!

Another simple suggestion, which does not change anything about the dynamics and current MCR model. The Mutual does not need to be involved at all with wNXM:

How about we allow members to sell their NXM based on the investment earnings, so we divert the investment earnings into a seperate redeem pool (currently stETH alone gives 4 ETH a day). Once that is established, it’s a matter of finding a fair system members can sell. 4 ETH a day can accumulate very quickly, a system could be setup to sell 10% max (every 90 days) of the current NXM holdings of a wallet and for max 1 ETH. A contract could lock that in (including transfer of NXM), members will accept that selling will occur no matter the MCR% and it will be executed in FIFO (from the contract). This implies:

1. The Mutual to not be involved with anything related to wNXM, no legal or tax risks / implications

2. Likely have members purchasing discounted wNXM to register having their 10% NXM sold (every 90 days), reducing supply.

3. Be a lot less complicated in my opinion, with the exact same outcome and

4. Not having to touch the current model and minimal modification to contracts (investment earnings was always an addition and bonus, the earnings just need to find a way to be diverted from the capital pool to a seperate pool, nothing else needs to be touched).

5. Members to actually be able to sell NXM! All other buyback options did not allow that, not providing a real solution to the problem that members cannot sell

6. wNXM price / discount becomes irrelevant to the whole problem.

7. All current NXM holders (even including the community fund) will have been able to sell within a few months (only about a 1000 members have more than 1 eth worth).

Edit: sorry about all the edits, wanted to clarify my points.

[dfmerin]

Any proposal like this would need to include a permanent ban on bots (e.g., burning the tokens of any member caught running one) since that is the core of the current problem and not per se solved by your proposal, but more importantly it’s not as value accretive to the mutual itself as the mutual does not get to take advantage of the current discount to book value.