[RFC]: Matching Currency of Assets and Exposure

Investment
1

Introduction

This post is intended as a conversation starter and a temperature check on the relative weightings of the currencies in the capital pool and how the community sees transitioning to a longer-term state.

As a disclaimer, the post is a solo effort, so doesn’t in any way represent team views - in fact I’d encourage other full time contributors to lay out their thoughts below.

There’s been a few conversations about this, but no clear way forward has been established yet. With v2 approaching, this seems like an important topic to discuss, and given the widely differing views so far, I believe it’s worth having this conversation as early as possible.

Keen to hear whatever thoughts the community have.

Current Situation

While initially the majority of our covers were ETH denominated, this has moved to stables over time. About 75% of our cover value, or about ~$138m of exposure now is denominated in DAI (and until the recent ETH price rally, it was closer to 80-85%). Meanwhile DAI assets in the capital pool are ~$10m (under 5% of total capital pool), mostly converted in anticipation of FTX claims.

This situation could leave us exposed in large claim events that hit DAI liabilities, especially ones that lead to significant ETH price drops & force us to liquidate significant amounts of ETH at low prices to pay claims.

On the other hand, the significant size of the capital pool compared to exposure mitigates this somewhat, and a good portion of the membership base enjoy having ETH exposure as a feature of their position in NXM.

Target State

The mutual should eventually be operating at a better level of capital efficiency, with MCR driven entirely by cover amounts, and holding the amount of capital necessary to back claims.

In order to do back claims safely and avoid compounding potential large claim events when operating at maximum capital efficiency, we need the currency denominations of the assets backing the (probability-weighted) liabilities to largely match each other - this is standard risk management practice in insurance world.

Topics for discussion

Here are some topics for debate:

  • Should we be selling ETH and transitioning into stables now?
  • If not now, over what timeframe, or based on what triggers?
  • Should we be targeting a specific ETH price for conversion?
  • At what level of capitalisation do we feel uncomfortable with the current situation?

Some loose thoughts from me:

Long-term we have to be currency matched, so something has to happen eventually. I’m also sure that the community & tokenholders are largely bullish ETH, so selling in the middle of a bear market is unlikely to be met with huge support.

Therefore, a plan for increasing currency matching based on market ETH price, and possibly, levels of capitalisation, might be a good way to go.

3 Likes
2

Thank you for getting this discussion started, Rei!

This is the key part of your post imho:

In order for members to form an opinion, it would be helpful to have some figures around that:

  1. when are we at “max. capital efficiency”?

  2. where does the mutual stand now?
    The answer to #2 seems to be above 88% according to the Capital Pool page on https://nexustracker.io/capital_pool
    There is also reminder of the definition on the page:

Capital Efficiency Ratio (Active Cover Amount / Capital Pool Size)

Plus we can see as well how this figure has evolved over time

And maybe also address:
3. what impact could we expect from the tokenomics proposal?

1 Like
3

Personally a large part of the attraction of exposure to the mutual was that I retained eth exposure. Whilst I appreciate and agree with the need for asset liability matching I would find it deeply problematic if the material changes to the current asset mix that would be required aren’t tied in with allowing capital that wants to retain eth exposure to leave the mutual. Any changes to the asset mix now are more incendiary than the revised tokenomics discussion.

If validator slashing is likely to be an increasingly large business line is this not more likely to be denominated in eth?

Another potential solution would involve capping the amount of dai/stablecoin denominated coverage for smart contract risks and beyond that coverage has to be in eth - cognisant that this would reduce demand but would be more inline with the desires of the capital providers.

Realise this also would not be acceptable to some of the “real world” insurance being talked about but I would rather a smaller eth capital pool mutual than a larger Dai based one.

1 Like
4

A thought about how to improve matching in the short-term would be to enact a buyback and to convert any additional value captured by it to Dai to improve the current asset/liability situation.

1 Like
5

Thanks for putting this up Rei.

Here is my current view on the matter :

ALM mismatch is very important when the mutual is levered, which is not the case. Currently, the worst that can happen is if ETH price suddenly crashes, a gap appears and some risks go above the 20% cap. This creates additional risk for the mutual. However, at that point the mutual will stop selling cover for those risks. Since average cover duration is rather short, the additional risk will disappear over a short timeframe.

As a result, I don’t think we should sell ETH and transition to stables now.

I see two main triggers :

  1. Cover demand growth after v2. If it changes the leverage profile of the mutual, we should consider selling ETH.
  2. Price of ETH.

Minimum $10k / ETH.

More seriously : I don’t think price targets are a good way to manage this but I wouldn’t settle for anything less than previous ATH.

Unless of course cover demand or release of capital to members change the leverage profile of the mutual significantly.

So I think we’re pretty much on the same page with this.

This is something that we’ve heavily debated within the Investment Hub so would like to hear other members thoughts on this! :slight_smile:

4 Likes
6

From an insurance perspective, which I believe is the right framework to look at this through, Rei’s proposal is right. We should be matching our currencies with our exposure identically.

That’s not to say that we can’t have more ETH than dictated by the above, but anything above that should be classified as an investment and we all need to be aware that it brings with it extra risk.

One way to frame this might be to allow us some additional % leeway to invest.

So, if we say we can be 20% overweight one currency.

If 80% of exposure is in DAI, 20% is in ETH, the maximum exposure we could have would be 60% DAI, 40% ETH.

This gets us a good way to where we know we need to be, while giving some leeway for staying long a currency if we desire. As we become more capital efficient (read: less overcapitalized), that 20% would need to come way down. But right now, with our current capital efficiency we can afford to have some excess ETH. Having said that, I still think we should start moving in the right direction, but we don’t have to go all the way to a 1:1 match immediately.

3 Likes
7

I agree with your take but everyone who put into the capital pool added eth and expected eth to be their underlying exposure. Now you can argue that people put eth into the mutual and the mutual must do what it thinks is best to create value and that’s a reasonable perspective but changing the capital mix would only be uncontroversial if people were able to leave the mutual freely and that is not the case. If/when tokenomics changes pass and mutual trades closer to book then this would be the correct time to revisit this.

Spitballing thoughts involve creating separate eth and stablecoin capital pools similar to how Maple has/had Eth or USDC pools so that people can get the exposure they want but cognisant this is not a short-term solution.

3 Likes
8

On the split capital pools, you’d lose a ton of capital efficiency so it’s not a realistic solution IMO.

Hear what you’re saying on the first point though. Ultimately, our risk exposure is not what it was a year or two ago. The vast majority of covers right now are denominated in DAI, and so it will eventually be necessary to switch the holding majority DAI. We don’t need to do that now, but know that if you stay in the mutual and the ratio of DAI / ETH covers stays the same, while we max out capital efficiency, our capital pool must match our risk exposure, meaning you will be having mainly DAI exposure.

In the short term, we’re currently not in a position where we must sell ETH. We’re over capitalized, and so we can theoretically just continue holding primarily ETH, there’s downsides and upsides here but that’s up to members to decide. But just know that in the long-term, this is not sustainable and it must change. Having said that, in reality, it’s highly likely that tokenomics will change before we need to do any significant currency rebalancing.

3 Likes
9

Hope Eth based slashing coverage grows very quickly…

4 Likes
10

Same! I think it’s a reasonably big business line.

However, I would caution that I believe other business lines that we hope to enable over the next 6-12 months via syndicates are likely to be bigger, and would be denominated in stablecoins. I do not believe that our exposure will swing back towards ETH in the long-term unless something external happens to demand for usage of stablecoins. Just want to give my take as someone working on the revenue side, that I see this as only a one-way street, which is that our ETH exposure will very likely only drop over time as we write more cover.

1 Like
11

@BraveNewDeFi can we transition this into a snapshot vote about whether or not we should be selling ETH to DAI, currently? Not value as to how much, but to get a sense check on what it is that community members want to do, today.

I believe this vote would be very closely aligned to the broad topic of the OP, so, find it reasonable that we construct the vote that way.

1 Like
12

Yes, we can. I try to avoid putting signalling votes up on Thursday or Fridays, so it would be good to draft the text for the vote and post to Snapshot on Monday.

For signalling votes, best practices is at least three (3) days for voting.

I’d recommend using several options in the signalling vote. It might make sense to give a few options: sell ETH for DAI at current prices, sell ETH for DAI when ETH price is higher, do not support selling ETH for DAI. Might help inform the discussion, but ultimately, up to @Rei or whomever puts up the Snapshot proposal.

1 Like
13

If we are taking about long term solutions to transition into in the future, wouldn’t it be simpler and better to have 2 separate capital pools, one in DAI, one in ETH, allowing people to deposit and redeem capital in one or the other pool independently, maybe offer different insurance products for each currency based on capacity (Uniswap: cover available in DAI $10m, cover available in ETH 5k), having different investments for each pool, different yields, and cross-funding pools only in case of need?

1 Like
14

Yeah, I would like to see part of the temperature check / snapshot asking whether the mutual should be pursuing non-eth lines of business. I get that this is where there is potentially a lot of growth but I struggle to believe that this is inline with the interests of the vast majority of current members. The mutual could potentially have a very interesting proposition for people looking for a return on stablecoins but forcing conversion of eth to dai to enable this business to be written is something that also needs to be discussed.

2 Likes
15

We already are. 80+% of covers are denominated in DAI, that ship sailed a year ago. Not sure it’s really feasible to turn around now, given that would mean destroying a huge % of our revenue.

To the general point, please see “On the split capital pools, you’d lose a ton of capital efficiency so it’s not a realistic solution IMO.”

2 Likes
16

I find that an incredibly arrogant attitude, the ship’s course can always be changed if the owners of the capital pool/mutual want to. Destroying revenue is inconsequential to retaining exposure to eth rather than a stablecoin and am confident this view is shared by many. I did not set out to monopolise this discussion so will stop commenting and hope others do so, but have pulled all my nxm staking except against eth slashing in response and I encourage other stakers to do so who feel similarly.

2 Likes
17

I agree…everyone here is sufficiently bullish ETH (and native ETH LSD yield) to feel comfortable with the mismatch given that (too much) business has been written in DAI. It is responsible to point out the mismatch, but also responsible for the members to say we stake ETH because we want ETH (& LSD yields) and we will collectively wear the mismatch risk for a while longer because we think that post-Shanghai there will be not only significant ETH value appreciation, but also a shift in the demand for coverage in favour of ETH-denominated risk (e.g. staking ETH).

2 Likes
18

Hi there, Mona here from Avantgarde Finance. I agree that we’re all bullish ETH etc but from a risk management perspective it isn’t right to have this mismatch.

Wondering if it might make sense to have an ETH denominated capital pool and a DAI denominated capital pool and only sell insurance matched to the capacity in each pool. The ETH denominated pool should aim to earn yield on ETH and DAI pool can earn yield on DAI. It means that capital pool suppliers can maintain exposure to their currency of choice and there is no risk that the pool can’t pay out. Thoughts?

1 Like
19

While this seems like a simple solution, there are a few complicating factors that make a two capital pool solution nearly impossible, imo.

NXM Token. The NXM token is backed by assets held in the capital pool, whether it’s 100% ETH, 50% ETH and 50% DAI, or any other combination of LSTs, ETH, DAI, or any other token. Isolating two pools and splitting between ETH (i.e., native ETH, LSTs, or any other yield-bearing ETH token) and DAI (i.e., any stablecoin or yield-bearing stablecoin token) would require two tokens if members want to choose their desired asset exposure. Because the NXM token plays a fundamental role within the protocol, this would have far reaching implications and would likely require significant development work.

Cover Demand. Right now, the vast majority of the mutual’s cover is DAI-denominated, and that tends to be the case in bear market conditiions. During bull markets, the mutual sees the majority of cover denominated in ETH. If we shifted to a two capital pool model, the mutual would then need to bootstrap the DAI capital pool. Since most of the current demand is for DAI-denominated cover, this would severely restrict the mutual’s core business, as many cover buyers want to buy protection in a stable asset. This would likely impact the flow of capital into the mutual through cover sales at a time when new staking pools are launching and members are building distribution networks on top of the protocol.

Capital Efficiency. Splitting the capital pool into two distinct pools also impacts the overall capital efficiency the mutual can acheive in the long-term. The goal of pooling all the mutual’s capital into one contract is to enable the greatest capital efficiency possible as cover sales scale over time. This leads into the next point.

New Tokenomics. As @Rei and the R&D team work to finalize the racheting AMM model for the new tokenomics, members will begin discussing the amount of liquidity that will be supplied to the virtual AMM in the planned new tokenomics. The sentiment in the Request for Comment - Tokenomics Design Detail thread is to allow any redemption of NXM for ETH and no other asset. Any other changes to NXM’s tokenomics would likely impact the timeline for the new tokenomics plan, and I’d wager that members would like to have the new tokenomics plan implemented before any further decisions are made about capital pool assets, whether it be further diversification to make a larger portion productive or to better match assets and liabilities.

I’m not a fan of a two capital pool solution, as the mutual was established with a capital pool that held both ETH and DAI. At this point, I’d be in favor of talking about diversification tranches when the price of ETH hits certain benchmarks, so members can capture the upside and hedge against potential price downside. If done correctly, a diversification strategy that happens over time with predefined goals can allow members to maintain the potential upside of ETH price movements, while better matching DAI-denominated cover exposure.

When ETH was more than double the current price, members also discussed a potential diversification into DAI, which was roundly dismissed since the vast majority of members were long-term bullish ETH. While I’m sufficiently bullish, it seems prudent to find a reasonable path to diversify over time and prepare for a future where crypto-native coverage isn’t the only product members are underwriting and distributing.

Now that V2 is live, there’s a whole market of potential risk (i.e., real world risks) that members can underwrite where the vast majority of cover will be denominated in DAI and we could see a significant increase in annualized fees from these new lines of business.

2 Likes
20

I appreciate the critiques of a two capital pool solution. They are definitely valid.

But having a single pool where underlying asset exposure is determined by cover demand means exposure of capital providers is decided by cover buyers. Which seems very hard to reconcile for capital providers. As a capital provider you’d have to be completely price-neutral on ETH and a price-taker in either direction for it to make sense.

Whichever route we ultimately take (two pool seems unlikely given the difficulties regardless of merits); I strongly feel it needs to be agreed that any significant rebalancing of the underlying asset mix should take place only after there’s ample opportunity for ETH-long capital providers to exit.

If it’s urgent to rebalance and match exposure, then that should translate to added urgency to implement the new tokenomics revamp.

4 Likes