There has been quite a lot of change in the DeFi markets over the past few months and especially for Nexus with massive cover growth, a large increase in the capital pool and now activity calming down a bit.
As you may have noticed the MCR% has been slowly decreasing for the past several weeks. While I don’t see any fundamental issues here, more of a short term vs long term situation, it got me thinking if we need to change anything with the incentive mechanics. And in particular can we use the situation to improve incentives more broadly and encourage the right behaviour.
Nexus will work best, and will scale best, when it has lots of members who are focussed on the long term. Therefore I believe the time is right to consider introducing a mechanism to gain greater alignment over longer time frames which also strongly encourages re-capitalisation of the pool when required. Specifically I am proposing the following:
If the MCR% is below 130% members may lock their NXM for a 1 year period and receive bonus NXM tokens (newly minted), with the bonus rate based on how far below 130% the MCR% actually is.
Members will be able to stake for Risk Assessment as well as lock for bonuses at the same time, so this won’t compete with staking capital.
Bonuses to be under the same 1 year lock as the original tokens.
Locked tokens can’t relock to get another bonus.
Potential Downsides:
Additional NXM inflation.
Worst case is that existing long term holders access NXM bonuses but material capital doesn’t flow into the pool.
While this is a relatively small change, it will delay other work.
Key Benefits:
Provides a strong additional incentive to recapitalise the pool at lower MCR%, especially after any large claims.
If large claims were to occur, there is the potential for a race condition to start and reduce the pool to 100% while the pending claim is being assessed. This would provide a significant off-set to that.
Materially reduce the probability of ending up “withdrawal locked” at 100% MCR
Provides a potential path to greater decentralisation by having a long term aligned subgroup that could perform actions currently performed by the Advisory Board.
Attracts more wNXM back into the mutual to access bonuses.
The proposal brings together a variety of aspects I’ve been thinking about for a while but it does entail a material change to economics so please chime in with thoughts.
This is a loyalty mechanism and I think it’s critical to get the short-termers separated from the long-termers. I’d even have tranches, like a CD or hedge fund, for 12, 18, 24 months. Allows us to have a predictable yield on NXM.
Also, presumably, those NXM could still be used for staking and governance despite the lock up.
While I understand the desire to incentivize long term participation, I don’t think this is an effective approach.
I doubt the people willing to accept a 1 year lock of their tokens are the same people that are going to be concerned with (and selling during) short term price declines.
I don’t think it would attract anyone who isn’t already planning to hold long term to do so, its largely preaching to the choir.
I also believe these long term timelocks cause future problems as people identify large holders leaving timelock and anticipate a sell off.
@Hugh I think this is a great idea but we need something more flexible than 1yr to attract outside capital which is the short term priority right now in my view.
Perhaps you could apply a different multiplier on the bonus for the time-lock? For example, give people the option to come in now for a month at 1.0x (130%-cMCR), 1.1x for 3 months etc.
Either way I think allowing a bonus with a more flexible time-lock will attract new capital and the benefits of this far outweigh the inflation in the short term.
My rationale is that it rewards long-term capital at the expense of short-term capital. At the end of the day, short-term speculators don’t really do much for Nexus – they pile into the bonding curve (or wNXM) in up markets, and then the price comes right back down when they take profits or the market cools. In doing so, they burn newcomers and price out parties that we want to properly incentivize to contribute to the capital pool longer term. Long term holders are more excited to invest in something that went from $3 to $18 than something that went from $3 to $70 to $18. I have always been in favor of optimizing for slower but real growth – e.g. it’s better to look like Berkshire Hathaway stock (consistent strong price performance) rather than Tesla (volatile retail mania) – after all, the brand Nexus is building is for safety and reliability.
On this point, to address @nexudm’s concerns, I think there’s quite a bit of capital on the sidelines waiting for the market to stabilize, and there’s also significant growing institutional interest in DeFi blue chips like Nexus. This discount is a good way to make some of those institutional buyers move. Think about it like this: everyone has a different model for how low they think the market could go. If the consensus opinion was another 25% down from here, and all of a sudden there is a limited time opportunity to get a 25-50% discount on NXM, it just became optimal to pull the trigger on that investment now rather than wait and play the market.
So my sense is that this does properly incentivize new capital, but my one suggestion would be in the vein of what @gyoung10 suggests. If the lockup period and multiplier are more dynamic, that addresses a broader group of potential investors. Someone might be happy to do a 2 year lock but needs to ensure a higher discount to do it. If you look at veCRV it’s fairly surprising how long on average people are comfortable locking for, and since longer dated locks are strictly beneficial for Nexus it would probably be good to accommodate a range of preferences.
Yes, still available for Risk Assessment and Governance (but not claims assessment due to a technical restriction).
Different tranches can work. The bonus for eg 6 months would need to be roughly 50% of the bonus for 12 months, otherwise it massively compounds inflation spend and nobody would lock for longer periods. I struggle to see any real benefit below 6 months, as we want to use this subgroup for long term incentive alignment in the future, and 6 months is still a pretty short time period in an insurance context.
But I take the feedback that more flexibility on lock periods is a highly desired feature!
I think this is a great idea. I think incentives work anyways in keeping capital in protocols but incentives combined with locking is even more powerful because it takes the benefits of incentives and moves it into the hands of long-term investors and away from short term farm-and-dumpers. Not only are the long term investors most aligned with the success and vision of the protocol but they are also far less likely to immediately dump that yield for a quick buck. I suspect anyone willing to lock up for 1+ years are also likely to take that yield and put it back into the mutual as pool staking to compound it.
I will echo what has already been said that multiple lock up lengths would be really attractive. I think there are different categories of long term holders, and some may be willing to lock up much longer than others. That just strengthens the capital base of the mutual for that much longer into the future. One way is to create different lock up lengths, and each has their own multiple. The longer the lockup the higher the multiplier.
Another idea is to make it dynamic so the multiplier slowly increases over time and hits a ceiling after 2,3 or even 5 years, but the user can unlock at any time they want. For example, it could start at 1.1x and increase by 0.0003x every day until it caps out at around 1.5x in 4 years (you could even increase the multiplier by a tiny fraction every block). I think the advantage with this is that it takes away the fear of an irrevocable lockup, but also makes the capital very sticky, especially as lockup time increases. The higher the multiplier goes the more costly it will become for the user to pull the capital and lose the multiplier they had been building up. For this to work the smart contract would need to store a timestamp on each deposit, and when the user finally pulls and collects their yield it is calculated on every deposit “tranche” that user made and its respective multiplier. You could simplify it a bit by requiring the user to withdraw all funds if they want to withdraw (i.e. all or nothing), which triggers the yield payout. Otherwise to allow a user to withdraw fractional amounts you would need some FIFO or FILO system to remove amounts from the tranches, starting at one end (oldest or newest), and allowing anything from small amounts up to the entire thing to be removed. But that seems quite a bit more complex to manage.
An added thought that could make an already attractive lock up incentive system even more attractive would be to give those who lock up some sort of NFT or ERC20 lockup token tied to the lockup length, and these lockup tokens are given some utility in the future. So not only are NXM lockers receiving a NXM yield but they are also receiving some sort of utility token which will have its own value adds and special privileges. Maybe it’s discounts when purchasing cover, maybe it’s giving pool stakers a little boost in yield, and/or maybe it’s giving early access to new features. It could literally be anything and a lot of things. Sky’s the limit. There would be plenty of time to brainstorm those value adds, write the code to use them, and unlock their value.
I want to add that I am not supportive of anything less than a year. A 6 month lockup is only 3 months longer than the pool staking lockup, so it adds to inflation while hardly adding any benefit.
I think anyone who doesn’t want to lock up for at least 1 year already has their benefit of being able to exit earlier than lockers. They are essentially giving up yield opportunity for that privilege.
I am generally on board with this mechanism to re-incentivize capital into the mutual.
After reviewing the community’s thoughts on this, I’d like to propose the following:
Linear reward bonus (Agree with Hugh) with 36% APY at 100% MCR and 12 month tenor
Optionality for various tranches for LT locking bonus (3/6/12/24 month options) with commensurate yields. For eg.
3 month tenor: Bonus % multiplier = 1.06x
6 month tenor: 1.1x
12 month tenor: 1.2x
24 month tenor: 1.3x
Locked NXM should also be allowed to stake for Risk Assessment (not clear from Hugh’s initial post if this can be done).
Rationale
NXM has never faced a scenario where MCR has neared 100% since inception. A mechanism to provide a liquid buffer for the capital pool is vital to protect the mutual in times of distress. However, the tradeoffs of higher inflation must be carefully considered.
Why shorter timeframes?
Capital efficiency should always be prioritized and we should only be paying for protection of the capital pool when it is needed. For example: If NXM faces a temporary shortage of capital for 1 month but strongly recovers after that, we only really need the capital for that first month. Paying a high fixed yield for 12 months for really only 1 month worth of capital protection is highly inefficient.
Furthermore, I echo the sentiment that NXM holders (even long term ones) have different time horizons. Offering tenors from 3 - 24 months would thus incentivize a broader stakeholder base.
I strongly recommend this should be put to a vote.
I would lock up for more than a year if rewards incentivized me to do so. CRV/VECRV is simple and encourages long locks. I would reinvest/compound rewards/payouts but it would make sense to pay them out at regular intervals
If there is a rush to to lock up for one year when MCR is below 130%, it will dilute people who will not be able to lock up if the MCR goes quickly back up. Could it lead to the same problem we witnessed with the KEEP rewards?
Would you be required to lock your entire holdings or could you lock up a %?
I personally believe this would remove some of the concerns around 6 months vs 12 months. You could lock up 50% to earn bonuses and leverage risk on the remaining 50% for short-term fluctuations. Even if this isn’t the case, I’m strongly supportive.
I like the idea. Is there a particular reason for the linear model though?
What we are really trying to achieve I assume is avoiding a lockup when MCR% <100. So in this scenario we should incentivise behavior above 100% MCR%. In other words having an optimum at 100% or greater.
Borrowing from biology for a sec, bacterial growth rate looks like this:
The idea is really simply to incentivise capital to enter before lockup can occur, but still have incentives in case it does drop yet in decreasing order. Why?
An outside speculator would probably not enter at MCR% < 100% knowing they would have to wait for it to rise again before getting out. So a quick buck might turn into a half hodl.
An investor looking at that will prob rightly assume that we are not able to grow organically or horizontally in this case.
I also like the idea of time tranches and agree that it should not be less than 6 months. Particularly because in that shorter time frame it is imperative that we find a solution to the staking lock problem of 90 days. Once we have that ‘tokenized stake’ plus bonus incentives, capital should flow in more organically.
On a final note I think that providing a lock incentive might be good even for latecomers who might want to jump in long term even when MCR% > 130%.
Yes, locked NXM will still be able to participate in Risk Assessment and Governance (but not Claims Assessment due to a technical restriction)
@charlie you’ll be able to choose how many tokens you lock, it won’t be all or nothing.
On multiple time frames:
If we allow long term locking, > 24 mths, the mutual is likely overpaying for the benefit of long term capital via extra inflation. The marginal benefit from a 2 year lock vs a 5 year lock is likely quite small for the mutual. So I don’t think we should offer over 2 years (likely at any point), as much as I like long locks for other purposes.
Relatedly, it’s quite hard to get the bonus levels correct on the first attempt, and whatever can be gamed will be gamed. If we end up setting bonus levels that are too high and we have a 18/24 mth option we likely end up with a significant amount of one-off inflation that will be very challenging to repay. I’d like to offer 18/24mths but the consequence of getting it wrong seems relatively high.
On the other end of the spectrum 3 months doesn’t seem meaningful given the core objectives of the mutual and 6 feels borderline to me but still worthwhile offering (at some reduced rate). If nothing else, it widens the market of potential lockers and gives us another data point to work with.
Bringing this together, my suggestion is to start with a 6 - 12 month range and see how it goes for a bit. There is always the option to expand the ranges later but if we over-inflate we can’t go back, seems better to be a bit cautious and see how it goes first. The risk of over inflation doesn’t really exist on the shorter end of the spectrum so it seems reasonable to extend downwards first.
Implementation
We have to build this and get it audited before release. It seems like there is enough interest to fast track the build so we will continue on that path. We can continue discussing details here with a view to opening a governance proposal.
@NjB Pretty sure the bonus needs to keep increasing below 100%. There is the possibility of a large claim dropping the MCR% below 100%, if the bonus decreases it creates perverse incentives to not enter below 100% and wait for other to first.
We’re trying to:
Avoid the MCR% going below 100% in the first place, and
I personally don’t think 6 months is a long enough lock to justify an incentive. After all it’s only 3 months longer than the 90 day lockup pool stakers already have. If it is decided to do 6 months the multiple should be quite a bit lower. Keep in mind that a 6 month locker will be able to relock the incentive for a second 6 month stretch, effectively giving 6 month lockers a compounding incentive advantage that 12 month lockers won’t have, so the 6 month multiplier should be low enough that even compounding into a second 6 month lockup comes in lower than a 12 month lockup, and lower yet on top of it so a 12 month lockup has a material advantage over two 6 month compounded lockups.
Given the complexity of the current system, any implementation will open up more attack surface.
Of course on a personal level I would like to lock and get a nice 30% reward but at this point of ~100% MCR I am also not the one to sell my NXM and get out anyway. By the time this gets implemented the only beneficiaries will be the large holders that are holding it for the long run anyway. ~50% of the current supply is held by 20 addresses.
I do not think we’ll get to stay low for long.
Total Staked - 3.8 M Tokens (leveraged)
Average Leverage - 6.3 (3.8 M / 600K)
850K pending unstakes (leveraged) / 6.3 = 135 K Tokens pending to unstake.
At the current price of 0.0477 and an MCR% of 108%, with 162.5K E in the pool, given we have about 12.5K ETH worth of tokens that can be sold until we get to MCR 100%, this means we are at about 261K NXM that could potentially be bought back by the bonding curve.
We are extra 100K NXM, ~$2M if everyone who unstaked would want to exit the next days with no capital coming in.
Let the ones on the sidelines look at MCR 100% and ask themselves if the bottom is in already or maybe someone else with more conviction in the project will buy before them.
There is enough capacity at the moment and the 160K ETH are not going anywhere. Why not increase the rewards for stakers and focus on reaching more customers, integrating more projects and so on?
TLDR
Wait and see what happens, if and when we get to 100% during the next month
If implemented, offer smaller rewards for locking longer term (MCR 100% at 8-10% APY maybe?)
Implement this together with the staking funnel and nudge all users towards long term
@Hugh Thanks makes sense in regards to large claim coming in. But doesn’t this model also create the same incentive you mentioned, ie let’s wait until MCR% drops even more before getting in since bonus will be higher?
Maybe a more exponential model between 130 to 100% with diminishing returns after that? The point I’m trying to stress here is that we should encourage as much activity above 100, creating some sort of incentivized price floor.
Agree with @krugman25 on compounding issue and probably 1 yr for such bonus is justified, since what @adrian mentioned, there is no running for the exit from people at this point, unlocked stake incoming will not be material so incentives given at the ‘expense’ of current holders should bring in more long term holders and not speculators on short time frames.
