If this were the goal, then why not accomplish it by removing dissolve delay as a neuron parameter? There is no need for neuron transfer if there is no dissolve delay.
I’m not claiming I agree with this goal, but I’m interested in why people want to enable a neuron marketplace instead of removing the root cause of why people feel there is a need for a neuron marketplace.
Hi @skilesare. I wanted to mention something here in slightly longer form that I also mentioned to you on Twitter.
I like your thought exercise but I feel that it is vague in one key area. You mentioned that this theoretical VC neuron has arranged a contract where a buyer has purchased the rights to control a 50M ICP neuron 2 years in the future. This is essentially a form of an option/futures contract. There are standard ways of pricing such contracts. Under simplifying assumptions, one can use the time value of money with a flat discount rate in order to calculate the present value of a future asset. The Black-Scholes formula can accommodate more complicated assumptions.
My point here is that there is that there should be a way to get an exact price that a market participant will be willing to pay for the contract you are describing. If neuron transfers could be tracked, then you could penalize the voting power of a transferred neuron. This penalty should be exactly proportional to the amount of ICP that this VC “stole” from the network by selling early. Since both the buyer and the seller are aware of this restriction, it will be reflected in the purchase price of the contract, thereby negating the potential profit by the VC.
I may be missing important security/engineering details here but I find this idea very exciting. It seems to align the voting incentives with the economic ones.
I think that this also addresses your point @wpb. There are definitely unique market dynamics that are made possible by introducing the delay. You could, in principal, retain the financial incentive for long-term investors while simultaneously making neurons more liquid by introducing an offsetting transfer penalty. This is not possible with no delay.
Yeah…I think so…provided people don’t route around that mechanism. Generally, you have to post collateral when you have some outstanding commitment like that. To extend that idea to neurons, every neuron would need to be non-splitable and not matureable until completely dissolved. In that case we could likely calculate what your payment should have been vs. what it was and deduct the difference.
But realistically, people like harvesting from their neurons…but once it is gone, it is gone.(I guess you could track the maturity vs principal and cap it if the amount left in the neuron ever gets close to a theoretical max penalty.)
If everything had PoK for regular neurons and Controller tracking for canister neurons, and we could implement this system it could be pretty interesting. It will much better when the rates are lower because it would take you many years to get to your locked principal total in rewards. At that point you could re post the collateral to keep your bonuses.
With that system you’d be able to sell/transfer at any time, but some of your locked principal would be forfeited.
I do not understand this at all. Why would you impose a 20% discount on neurons without PoK and not a 100% discount?
Transferable neurons that are not PoK have no place in our governance system.
Why would we imperil our governance system I don’t get it. Why do we want a liquid market around neurons without PoK?
If people want to just buy liquid tokens they can do that. If people want to take part in governance, well then they have to lock up their tokens. I’m sorry that should be the way it is. Otherwise don’t stake.
I am not entirely sure whether this analogy is completely accurate. To me, NNS governance is similar to shares in traditional markets.
If a long-term investor initially commits to investing for the long haul but then decides to sell their shares midway due to some reason, does that constitute a breach of agreement?
In my opinion, the IDgeek marketplace could serve as a facilitator for the ICP ecosystem to find a harmonious point through its market discovery system. IDgeek market helps neurons find a matching profile of an investor.
Regarding whether all neurons will be locked for eight years because they have the same liquidity as short-term neurons, I believe that once people realize that eight-year neurons are more valuable than shorter-term ones, there will be price discovery, and eight-year neurons will be sold at a premium price.
Hmm. I am not sure I see a need for collateral here. I basically think of the future rewards as the collateral. In my mind, there should be no outstanding commitment to the network by a neuron’s former owner once the neuron has been transferred.
From a pure trading perspective, I would think that if you stake X ICP for 8 years, and then sell ownership of your neuron at year Y (Y < 8), the total ICP accumulated after the transfer should be equivalent to the amount of ICP you would have received by staking X ICP for N years and then letting it dissolve for the remaining M years, so that N+M=Y and N is maximized. I wonder if there is a way to work backwards from there.
Pricing it this way is clearly not fair to the network because by staking in an 8 year neuron instead of in an N year neuron and never dissolving it, you have obtained additional voting power during those Y years at no additional cost. The amount of extra voting power obtained is essentially equal to:
Extra Voting Power = (8_years_dissolve_delay_bonus - N_years_dissolve_delay_bonus) * (1 + Y_years_age_bonus - N_years_age_bonus) * X * Y,
ignoring the accumulation of age bonuses. There was also additional maturity earned during those Y years. I am not sure what the exact formula for that is, but I think it can be calculated similarly. Maybe (Extra Voting Power Obtained) + (Extra Maturity Earned) is a basic starting point for a transfer penalty?
It does become more complicated once you consider compounding of the earned maturity. I think you could still have the payouts though. You would just need enough staked ICP + staked maturity remaining in your neuron to cover the cost of a transfer if you wanted to do that.
Just to circle this back to @bjoernek’s OP:
Phase 1: I support lifting the restrictions on canister controlled neurons as a starting point. I think it presents an interesting new space of possibilities to explore, and from what I am reading here it seems like this does not pose any immediate or likely future threat to the health of the network.
Phase 2 & 3: I think that some form of reduced reward/voting power mechanism for transferable/transfered neurons is interesting. I am not sure I am settled on what exactly is the correct trajectory here. I like the idea of a temporary loss of voting power in the case of a transfer as mentioned by @Seers but I am not sure how to avoid the circumvention of this through a hierarchy of canisters, etc.
If this were the goal, then why not accomplish it by removing dissolve delay as a neuron parameter? There is no need for neuron transfer if there is no dissolve delay.
Because you are taking out from governance your stake and by doing so increase inflation pressure. In the case of sale - locked ICPs are staying locked.
I didn’t mention this earlier, but I also thought about optionally taking a transfer fee and burning it.
Maybe thinking neurons as houses could help.
Selling houses has no place in our nation.
Why would we imperil our nation’s security? I don’t get it. Why do we want a liquid market for houses?
I would add that these are magic houses (or Airbnbs 
), minting 7% of their value in dollars under your mattress. Some businesses will be put out because they can’t compete with that return, unless, of course, you can compose them.
Sharing abstract IDs, even without personal information, can still raise privacy concerns. People might feel uncomfortable knowing their follow lists could be publicly accessible, leading to potential privacy violations or unwanted attention. It’s crucial to prioritize user privacy and consider the implications of making such information public.
Having thought about this a bit more I think we should simply allow canisters to control neurons and make no other changes.
SNS neurons can already have their controllers changed entirely so they can all be exchanged even if controlled by an II principal.
If canister controlled neurons get a 20% reduction in voting power, all that means is that it will cost 25% more to buy the same amount of voting power on a neuron marketplace, so it doesn’t prevent malicious actors buying voting power using flash loans.
Also, more than 1/4 of all ICP is locked for 8 years. That is a huge amount of voting power that won’t be able to be transferred. So even if neuron marketplaces do pop up, their total voting power will be relatively tiny (for at least the next few years).
I really think we shouldn’t add more complexity to neurons. My brother has some neurons, he signed in the the NNS app and then called me because he wanted to know about maturity (staked and available), age bonus, disbursing, spawning, etc. Neurons are already more than complicated enough. Adding PoK is adding significantly more complexity.
Some people have said that if you can trade your neuron then you will be incentivised to make short term rather than long term decisions. But even if you consider that to be true, this is such a small group of people that it doesn’t validate adding more complexity to prevent it happening.
The current solution proposed by Dfinity is canister controlled neurons receive 20% less voting power and it would still allow for a liquid market around neurons without PoK? I think the Idea behind the 20% less voting power is to deter a nefarious actor buying neurons for a governance attack (I don’t know if it would be effective - I don’t think anyone knows).
I was not suggesting anything new and I was trying to think of a simple implmentation without adding the new complexity of PoK (currently PoK does not exist on neurons). Are you saying you don’t want canister controlled neurons at all then?
For the record I don’t really want to see a neuron marketplace appear. I’m not sure what the long term effects of something like that would be. But I agree with others in this thread I would like to see canister controlled neurons but I think PoK is just too complex.
Then we should form a joint stock company. Bluntly, the NNS is absolutely NOT like governance shares in traditional markets. We could do that, but then it is almost certainly a security and not a utility token.
Joint stock companies have shares that trade freely and an incentive system for electing board members who steer the company. Board and shareholders have incentives to increase shareholder value exclusively. It is a completely different model. I don’t think you can build a world computer under the governance of this kind because it is too easy to make money in too many other ways.
I appreciate the creativity, but there is no similarity between a neuron and a house. I mean…almost none. Maybe you could say that owning a house in place gives the right to vote in that jurisdiction, but we even have laws to force you to pick one jurisdiction.
I’m probably the biggest proponent of this. I think this is the biggest threat to the network. Lots of small decisions that end up degrading the network long-term. I’d argue that while it is a small group of people now, it could become a very large group very quickly. We don’t have to look around the world very much to see how quickly sentiment, groups, organizations, and public opinion can change very quickly. After all, we are all rooting for 100x more people to be on our network as soon as humanly possible right? We don’t have much control over who those folks are, what their intentions are, whether or not they are VCs trying to subvert the network because the funded the competition or what.
We don’t have many ‘small matters’ where a few votes here and there change the direction of the network, but in a healthy, vibrant, and engaged network at the level we are reaching for we should be having those ALL THE TIME.
All of that being said, we likely have some time and this could be studied more so I’m not opposed to moving forward with canister neurons while we do that…its been too long that we’ve suppressed many of the unique funding opportunities that they provide.
Super interesting about SNS…have we seen any neuron marketplaces or moving of SNS neurons to this point?
I don’t want to throw too many gates up, but I’d really like this to start with whitelisting of canisters that can control neurons. This doesn’t seem like a huge change and it would help us closely monitor the types of applications that are being built with this power. If the registration is also tied to the controllers of the canister it could help limit the transfer ability.
Maybe this would just drive people to the tecdsa route, but if it was cheap, easy and straight forward it would likely lead to some cultural best practices:
I don’t like putting more on the NNS…perhaps this should be a separate SNS that votes on these things? Although there isn’t much economic security here so that might be gameable.
If we get away down the line and there just isn’t much controversy opening it up permissionless could eventually make sense.
Just thinking out loud here.
You’re right; there’s limited similarity between neurons and houses. While both involve connections and ownership, they operate in vastly different contexts.
You would be surprised by the amount of insight I can gain by imagining X as Y, 
. I used it a lot during my PhD to prove theorems that remained elusive for others. It requires a non-trivial amount of imagination. Anyway, I think we already reached a conclusion, so I will try not to prolong this thread any further.
I’m not against this kind of thinking…it can be the best kind! But I didn’t see it…if you have some insights from I’d be happy to hear them even if only in a dm.
Idea #42: making slashing work
I think this is a pretty nice solution: flexible, simple and more secure.
cc @bjoernek