Have you and the team thought about what happens when someone does decide to dump? Does this result in them claiming the apy earned by other peoples ICP in the 8yr?
Ok, letâs play that scenario. ICP âpumpsâ, the first nICP user who âdumpsâ it all gets a lot. Or if it has nICP/USDT, the pool benefits. It brings the nICP price down, and the arb bots benefit. At no point the nICP holders get anything from the price action if they do nothing.
With cICP, they do. The cICP/USDT pool brings cICP/ICP rate down, and the buy-back stream starts compounding at a higher rate, unless someone just buys the cICP. Then the cICP is liquid, and someone can just sell cICP.
During a very serious rate increase that wipes out nICPs liquidity multiple times, nICP holders get nothing as well, and if they become illiquid, they canât benefit. If they want to unlock and sell, they wait 6 months, but by then the rate they wanted to sell at - couldâve gone down. With cICP, all holdersâ benefit.
The only way nICP is more profitable is if the ICP rate goes very high for more than 6 months, so they can sell and then crashes down and never recovers. But - how is the IC benefiting from all of that really? Itâs not. The point of locking tokens and getting APY is so you donât need to mess with the market, and you are here for the long-term success of the IC. The small up and down movements donât matter that much to you.
Its kind of funny that adam just market dumped a massive amount of nicp but the market instantly bought the discount.
With nICP you will always be able to cash out your original stake plus the apy share of the neuron. You do not have to rely on the dex to allow you to leave your position.
With Cicp you must rely on the market to let you get icp back with no gurantee you will exit with your original amount much less any apy for the others using your vp.
Two very different strategies that ultimately end up in the same place. Like you said the people who wish to use their liquid tokens for defi give up the rights to their vp. So the only way to retain your vp while having the option to exit with cicp is to do nothing with it and hope there is enough liquidity for you to exit with apy gains. If you choose to do some defi fun with cicp it puts you in the exact same place that nicp starts at. No voting rights and giving the votes to whomever you participate with.
Two different strategies - yes. But they donât end up in the same place.
nICP is for the short term strategies. cICP is for the long term strategies.
Cicp has a goal to keep the vp with the staker.
You remove the right to the icp in the neuron in order to give them the voting rights. The user must keep the cicp in the wallet and use it for nothing else to retain this vp. Once they choose to do anything with it they then give up the right to the vp and lose the apy associated with it if the market dumps.
One relies on people just holding the âliquid tokenâ as if it were staked in order to earn apy. the other allows you to earn apy while playing with defi opportunities in the ecosystem. But in the end when they are used for defi they both lose the voting rights associated with the token.
Alex, I am starting to understand your problem with it better.
I guess you arenât worried about converting cICP into something liquid like ICP, but more about being able to trade cICP between entities, while it can vote.
But really, you already have that market.
Letâs examine the current setup:
- NNS dapp locked 8-year neurons can be sold using the Internet Identity marketplace.
- Canister neurons can be sold locked with a controller transfer.
- Only the ones directly connected to a private key canât be sold, because both previous and the new owner will have the key.
- WTN is pretty much pure voting power without stake
- WTN neurons are locked and canât be transferred, but the Internet Identity holding them can, probably sold inside the II marketplace as well.
- WTN neurons were sold before SNS launch and can be sold in the future, just need to pass the II to someone
- There are WTN canister neurons that can change controllers to be sold.
What I think shouldnât be done is separating the voting power component from the stake and selling that separately. All other chains donât seem to have a problem with the transfer of a token that is the stake and the vote. You can think of ETH as such a token, if one gets enough of it, they can then stake it and have half of the network validators, which is the governance system there.
Thanks Anvil. Yes, I think it all comes down to VP transferability. The NNS is designed to make transferability hard. VP cannot be traded in a fungible manner (having to sell your entire II or canister, makes it hard). The IC is unlike any other chain (itâs governance heavy). If the NNS is compromised, quite literally everything is vulnerable.
Assume a highly resourced actor/organisation wants to destroy the IC, and imagine we live in a future where code contributions from external contributors is commonplace. This person/organisation could mint lots of cICP and acquire a large amount of influence on the network. They could simultaneously contribute changes to the codebase that improve the IC, all the while preparing the stage for a malicious action in the future (which is hard for anyone to spot). Assume theyâve managed to introduce some sort of backdoor, kill switch or other vulnerability that only they know how to exploit, and only they know is there (because of the phased approach they have taken to introduce it). By having a large portion of VP on the network they are far more likely to be considered trusted contributors (why would they try to destroy a network on which they have such a large stake?).
Now imagine they begin selling their cICP in large batches, striking a balance between maximising their sale price while exiting their position a soon as possible (say, within a month or two). Once theyâre out, they can trigger their malicious code to action whatever their attack is intended to do.
why would they try to destroy a network on which they have such a large stake?
This question doesnât make sense anymore in the context of something like cICP (if cICP holders are to be afforded NNS VP).
Iâm still very unclear on why you wouldnât just make proportional representation relate to NTN stakers, rather than cICP holders (after all NTN is the governance token for the protocol - so they have the ultimate power at the end of the day anyway). At the moment NTN is fairly centralised, but I would happily start buying up NTN knowing that each token I hold contributes to my influence on the NNS (in order to further decentralise it - given that the NNS is also currently highly centralised). With proportional representation, I know that each token I hold makes a difference (unlike something like WTN).
1 Like
The other chains also have governance, just not on-chain.
(â Liquid Staking Tokens (LST) on other chains)
Their governance tokens unlocking in 1 day, changing owners, and voting again arenât causing them problems. What NNS offers is rewards for voting if you lock in liquidity, which is still valid with cICP. You can think of ICP as an almost stablecoin you can stake to contribute, where inflation gets adjusted to keep it hang around in place, while cICP is getting an undilutable share like what you get in ETH and other chains after staking. Maybe best if a few changes are made so ICP becomes a full-blown XDR pegged stablecoin - kinda of wild, just brainstorming here.
Interesting scenario, but it feels somehow unrealistic, everyone will spot them leaving cICP - if they try to do it in 2-3 months with an 8y locked neuron yield buyback stream and double check everything theyâve done. Now that youâve pointed out the problem, we can look for that. After all, itâs open-source. People shouldnât be able to put backdoors at all.
NTN doesnât need to hold the ultimate power at all. Itâs governing the voting system, just to make sure that whatever changes happen, it can upgrade and keep it working. Sure, if NTN is voting, that will make us rich, but we want to make something fair instead of selling pure voting power. I would consider NTN voting only if we merge cICP and NTN into one, but this will be pretty hard, and we would have somehow summoned lots of liquid staking tokens that arenât there - so no way. Like I said, we canât split voting power and stake. Splitting of these things and node providing has gotten IC into trouble.
If we get all proposals around upgrading the voting system - critical, then 33% get to reject changes, while cICP holders vote however they want inside the NNS. If we give cICP holders a neuron that votes on the voting system and can reject upgrades - the whole thing gets more decentralized, and the DAO is definitely not âownerâ of cICP or holds ultimate power. So, in a way, not NTN voting with cICP stake, but cICP holders voting in Neutrinite DAO.
But thatâs exactly what youâre proposing.
If youâre really wedded to the idea of cICP holders having the VP using an unstaked token (one thatâs liquid and tradeable), then I canât get behind this idea (though I very much want to).
What if the VP remains with cICP holders, but only if they stake their cICP? Staking cICP could mean having it allocated to a liquidity pool thatâs governed by the NTN DAO (thereâd need to be another token in the mix though to form the trading pair). I think this would be complicated. Staking could also just mean locking it to get your VP back, and potentially higher yield than theyâd get for the same VP by staking directly on the NNS (if the node provider rewards aspect pans out).
Iâve been thinking about the node provider aspect quite a bit, and Iâm planning to post a separate thread shortly about NP clustering, and the need for some sort of strongly modelled concept on the NNS for node beneficiaries (not just for cICP, but as a generally useful concept).
Not really. Liquid staking - you are liquid while staking. Whatâs locked is the liquidity, which comes in slowly, not the token. You are economically incentivized to vote in the best long-term interest of the IC, or you arenât going to get any value out of your cICP.
We already have the worst solutions out - liquid democracy, where you transfer voting power from people with skin in the game, to people without in large amounts. @wpb held ~9% of total vp on code proposals from Synapse and CodeGov and nothing bad happened (so far). WTN is buying and selling pure voting power, which is way worse. Unless you come up with a realistic, dangerous scenario where cICP opens up new attacks, you wonât change my mind. What we need to do is make sure people who have ICP and stake it, directly vote themselves (possibly using AI agents in the long run that follow reviewersâ feeds) so the value of their VP stays with them, and anyone who wants to get it has to buy ICP. One of the benefits of AI agents voting is that the prompts can be public, malicious intent is visible, and the system is trustless. The biggest danger so far isnât in someone installing backdoors, but in people with no skin in the game working hard to dilute the ones with skin in the game. You canât be seriously thinking that stripping cICP owners of voting power, because they may harm the NNS by gaining trust, installing undetectable backdoors, is solving their problems. Iâve thought about locking cICP, but it will make things cumbersome, possibly disable SNSes voting with their cICP, and require a special app instead of working with all wallets. So, let me know if you have a scenario in which cICP holdersâ voting is going to harm the NNS, and then they donât lose value from that.
Iâve already given you a scenario, and if you donât think itâs realistic then I would argue that youâre not being security-focused enough.
As a thought experiment, imagine that the IC didnât have the concept of neurons. Imagine instead that unstaked ICP provides the holder with VP, and casting a vote causes fresh ICP to be minted to you (voting rewards).
- Do you see any problems with this approach?
- What purpose do you think neurons serve?
- Do you consider the concept of neurons to be unnecessary complexity?
- Do you think your perspective on these matters is aligned with DFINITY?
Note that WTN is staked to get VP (the longer you stake, the more VP). WTN holds financial value relative the the VP that it wields. Youâre wrong to suggest that WaterNeuron gives VP to individuals with no skin in the game.
2 Likes
It gives it to a specific set of individuals, who may or may not have skin in the game (spoiler: they donât)
I want to add that Iâm not a fan of the fact that WTN does not provide stakers with proportional representation. Iâd really like to see this implemented in a competitor, such as NTN (cICP)
Ok, letâs try it out. Someone stakes using cICP - 6,750,000 ICP, locking in 34 mil$ at these prices to get 3% VP. They gain trust because of their cICP stake and massive IC contributions (in the future when external contributors are a common thing). Once writing code for the NNS, which nobody except Dfinity does now, they manage to put a backdoor in the code and nobody notices. Then they somehow sell their cICP in the open market for 2-3 months and the price of cICP doesnât go crashing for them to loose a lot of value (While they currently can just do that with a neuron and sell the neuron privately). Nobody makes the connection between cICP massive unloading and their code contributions over months. Once they got out, they activate the backdoor and destroy the IC.
We are worried about that? But we arenât worried about people hitting follow on a neuron with 1 ICP and gifting it 9% vp ? Or WTN buying voting power by providing tech.
Iâm am worried about all of these things, which is why Iâm so active in NNS governance. Itâs also why I have been so active in WTN governance. Itâs also why Iâm being so active in this discussion, and why Iâm pushing for greater competition and due diligence among known neurons.
Itâs a slippery slope to point at one bad thing as justification for another bad thing.
Would you humour me and have a go at tackling these questions?
2 Likes
The controlling group can still transfer and sell their locked neurons, selling the whole control in a private deal in 5 min. WTN for everyone other than the controlling group just has speculative value, not a real long-term one. Even the % of the maturity isnât reliable. What you are advocating against here is already happening in WTN, except there you donât have the requirement for stake and itâs worse. The model you are proposing (the WTN model) is flawed. Applied to NTN is flawed again. Locking WTN provides none of the guarantees you think it does, since you can sell locked neurons.
The attack scenario you described additionally looses plausibility, since itâs faster for the attacker to lock for 6mo and use that as trust building stake and unlock after 6mo without worrying about getting somehow out through DEX liquidity. Even easier with 3mo neurons. If someone gains trust to contribute code that will take years, they can as well wait the 6mo dissolve after installing undetectable backdoor. No need for cICP at all.
What you do in the NNS is lock liquidity and vote. What you do with cICP is the same - lock liquidity and vote. Nothing has really changed. You can already sell locked neurons now.
If you think itâs a problem that you can exchange the cICP token, Iâve already pointed out that all other chains work like that already. Lock and unlock ETH in a day and take it over if you have the billions of $ required and people are willing to sell.
Unless a scenario in which someone in cICP with significant voting power votes and gets away with bad voting without loosing value comes around, all staking principles are kept.
What do you mean? Without staking WTN there is no VP.
You can only sell the neuron controller (which typically holds a bunch of other assets), which makes it harder to sell because itâs not fungible.
6mo stakers are awarded less VP than 8yr stakers. Why do you think that is?
Everything has changed. Youâre suggesting making 8yr-staked VP fungible.
Please stop pointing at other chains. IC governance is designed with well-defined constraints and youâre proposing to undermine them, with a false sense of confidence that thereâs nothing to be concerned about. Forget about potential attack vectors, and just consider incentive alignment for a moment.
2 Likes
Doesnât this apply to cicp through the devs selling their ntn to a new person?
They then can force changes and claim all the vp from the cicp stakers as they govern the protocol.
Really love the idea of matching staking to RWA and infrastructure. Just two questions, how you determine who is the service provider? Because that would be a governance issue in the future for sure. I can suggest when topology allows that the DAO itself can own it through data centers(would recommend very strict cost-benefit analysis on this). Which brings us to the NTN, is it going to be decentralized? Why not create another DAO with a sale on this. Love the idea, itâs something we can invest and contribute to in our mandate.
Canât people follow other cicp users for voting?
Essentially that would reduce capital cost to do a governance attack.
You have a lot of the same vulnerabilities but have increased the risk to icp by using a 8yr neuron for maximum vp.
Cicp can also be dumped by selling the ii, similar to wtn. Only difference is you can ONLY exit wtn by selling your entire account while you can enter and exit cicp on the open market and through id geek.
This has effectively made a new way to buy and sell vp. But with bigger risks to the network.