An m-of-n multi-signature wallet or account perhaps?
@Jonza One of the challenges to any reserve account is centralized trust. What if the n holders refuse to burn the NBT, and run off with the money, instead? What if they claim it was stolen? We want to minimize the need for trust in a decentralized system, as much as possible.
I assume there is a mechanism in place right now to prevent Kiara Tamm from doing the same thing with money raised from NBT sales is there not? How was that dealt with?
Perhaps the n holders could be diversified enough such that collusion is nearly impossible? Perhaps the n holders could be Nu shareholders or trusted representatives of them that can be voted out at any time? I agree the idea is to not have to rely on trust, but I imagine the solutions to deal with this are already out there, though I’m not an expert here.
There is not such a mechanism. NuBits is not perfect (yet).
Given the problem already exists for custodians, and is relied upon at the moment for buy-side liquidity also, people here or the devs should probably think of a way to deal with this type of important issue. Then maybe that solution makes the idea of a reserve worth considering.
My major concern is less about the human element (such as a custodian), and more about what the assets can be stored in that won’t lose value over the time that they needs to stay in storage.
It’s not possible to store it in NBT, because that’s what it is intended to back-stop (should it be needed). BTC or another crypto are too volatile to make it worth while. USD or another fiat is possible, but storage becomes a major headache unless the network was able to come to an agreement with a semi-autonomous bank in a region of the World that was mostly stable.
I’m open to listening to all sorts of ideas, so if I’m missing an obvious storage mechanism, I’m all ears.
That’s part of what I meant by trust. There’s no easy solution, unfortunately.
Shouldn’t the USD simply sit on the exchanges? As described at the outset, exchange default was the primary risk incurred by the custodians. This is clearly a point of failure, but it seems such a situation must exist if NBT is to be backed by a real asset.
I think the reserve should be distributed to many reserve-custodians and there should be a setup such that if a few of thereserve-custodians runs off with the money, then 1) there impact is limited and 2) there is a cost to the custodian. The exchanges that host network bots can be reserve-custodians.
Open Transaction has designed a distributed system that exchange risks is distributed such that they could claim users are collectively risk free in that regard. We might be able to adopt similar concept to keep the reserve. I have been wondering what would come out if OT and Nu work together – the creation, exchange, and bookkeeping of money are all public and distributed.
My proposal would be to use relatively stable pegged currency for the fund
This will require a pegged crypto coin to e.g. Euro or Yuan. For now central banks peg USD, Euro and Yuan to some extent so any assets in such crypto coin would be relatively stable. Not a great option, but I think the best given the decentralised and multisignature requirements.
Edit: Or you could park it with BitUSD. Technically possible, but not very desirable and too risky.
Ben, I can more clearly see the storage issue now. Wherever fiat USD is stored, it might be subject to nationalistic government intervention or even confiscation. Stored in crypto, even a basket thereof, means there is the risk of a large fall in crypto prices wiping the reserve and taking away the ability to reliably peg NBT, so shareholders would have to commit to topping up such a reserve when its value becomes too depleted.
Having agreed with this vulnerability however, I think the current system suffers the same risk. The custodian providing dual side liquidity also needs to store their funds either in exchange USD (counterparts risk) or crypto (volatility risk). So there is a similar risk of capital loss for the custodian (hence shareholders), and potentially not being able to provide the buy-side liquidity in NBT when required to support the peg. So once again, it is still necessary to think about how to manage this risk in the present approach - its not just a reserve fund issue.
This got me thinking however, that perhaps a compromise or least-change approach would be for custodians to provide dual-side liquidity by doing the following:
- sell created NBT according to their mandate, and hold funds gained in reserve
- market-make though standing in the buy queue to buy back NBT, then adding them to the sell queue again
- when required or voted upon, to buy NBT and send them to a burn address (instead of adding to the sell queue)
Giving the custodian the mandate to burn bought NBT provides the necessary symmetry around the peg.
In this case, shareholders would receive their reward in a different form, rather than just sales of new NBT. For example, a small dividend of X% per month could be paid out of the value of the reserve over time. In this way shareholder reward is linked to the support they provide to the market rather than just selling of the product. Shareholders would vote to burn NBT when required because maintaining the peg keeps them in business to earn further reward down the track.
In a way, this is not much different to what Kiara Tamm is currently doing, except giving custodians an ability to burn as required or voted upon. Regardless of the approach, the issues raised above - i.e. - how to ensure that custodians are honest with the funds, and also to minimise volatility in the funds held - still need though around better protections.
I just wanted to assure everyone that I haven’t forgotten about this excellent proposal originated by indigoman. I have a few details to work out in my draft motion for a currency burning protocol change, after which I will present it for comment and editing.
It will leave the parking subsystem unchanged, includes other modest changes to the protocol that regard quantity of NuShares, and will not permit currency to currency exchange or exchanging NuShares for currency.
Looking forward for your post Jordan. I’ll save my comments on a lot of the ideas that we’re discussed around until I read your proposal to keep the discussion focused. I’m glad I managed to contribute to the development of Nu
We’re eagerly await the details on the draft motion. Adding a NBT burn mechanism to the protocol may drastically improve the long-term outlook of NuBits stability. Thank you for bringing this forward.
This place is awesome. You guys are great.
To be frank, I doubt about the decentralization of Nu in future.
In 2009, satoshi himself control the whole BTC net(centralized) and that was only temporary and in 2014 BTC is successful without reliable sotoshi(decentralized).
Nowadays, several big NSR holders/custodians control Nu system(centralized), but shall we trust those Nu custodians in 2019? Decentralized? This is a fatal question, IMHO.
The burning mechanism may encourage evil activity in future. An evil scenario:
In 2019, an anonymous custodian is granted for 2 million NBT and play LPC role for one month, suddenly he runs off with 1million USD and 1million NBT, then he pretends to be another person and spread the rumor that Nu can NOT support pegging any more, NSR price drops, buying wall becomes thinner. He attacks the buying wall with 0.5million NBT and threaten NSR holders that he will break buying wall and ask for burning his stolen remaining 0.5mllion NBT to get NSR.
In fact, every custodian has the incentive to run off and blackmail the whole Nu system. Yes, they have lots of NSR, but they can sell NSR first and commit crime.
You are right; unfortunately this is a scenario which is entirely possible.
On the one hand custodians need to be as anonymous as possible (at least to the “outside” of Nu) to prevent them from being attacked, blackmailed, their accounts frozen, their funds being confiscated.
On the other hand they need to be well-known (“inside” of Nu) to assess their trustability and to be able to hold them accountable (if possible…).
This is obviously a conflict of objective.
And it leaves a base for attack scenarios (what if one who knows the custodian decides to blackmail him/her?).
I can only encourage people to help thinking about, refine, develop what is discussed here: Decentralization to Minimize Custodial Risks
I am also worried by the fact that some custodians could convince the shareholders to give them large amount of nubits…but if we make sure that those custodians are also shareholders, with a granted amount of nubits in proportion to the owned shares, we should be able to prevent most of the potential frauds or manipulations unless the custodian wants to commit economic suicide; in this sense, it is similar to the conditioning that we can have with peercoin: it would be committing suicide to perform a 51% attack on peercoin’s network since it would imply holding large amount of peercoins to get a large stake necessary for the attack.