Decentralization to Minimize Custodial Risks

Development
#1

I just wonder if we can make custodian anonymous. Now those custodians have real name FIAT accounts on exchange. So if government wanna control just several of them, the whole system may fail.

My suggestion:

Several big custodians hold large amount of NBT in public or private addresses. Small NSR holders borrow small amount of NBT with their NSR pledged to Nunet. For instance, 10000 NSR holders with everyone providing liquidity of 1000NBT on exchange by Bot.

The lending contract is short-term, i.e. one week. When Nu need decrease NBT supply, big custodians simple stops lending contract, within one week, the supply will be fully controlled.

Two merits:

  1. Our VIP(big custodians) will hide safely without FIAT account.
    2)Thousands of small LPC with their FIAT accounts on exchange, immune to hack/stolen and run off. How can u imagine 10000 accounts hacked at the same time on various exchange and one government could not control so many people all over the world.
3 Likes
Another paid article about BitShares?
New mechanism to reduce NBT supply: NBT Burning
Proposal for "transfer message function" and "Public address"
#2

Who holds the parked NBT?

Can we park NSR? If you park some NSR and get some NBT, that’s lending business, and you become a small LPC. In this way, thousands LPC will emerge.

As a NSR holder, I’ll do that even without premium because pegging itself is the support of NSR price.

Of course I may get some revenue selling at 1.003$ and buy 0.997$.

#3

While I like your idea of distributing the risk of being attacked peronally or having the accounts frozen (or worse) by utilizing large amounts of custodians this leads to some kinds of problems.

It will be horrible to assess the trust level for thousands of people before granting these amounts of NBT (trust problem).
Currently more than 1 NuBot per exchange and trading pair doesn’t work. That might be changes but I still doubt that it’s (technically) a good idea to have thousands of NuBots hit the exchanges. I wonder if the NuBot is that scalable (technical problem).

Although you could try to solve the “trust problem” by taking NSR as security that will be paid back once the contract has been fulfilled (how to verify this?), I’d rather not incentivize taking NSR out of it’s place which is to mint (for securing the NBT/NSR block chain) and to decide which way to go by vote.

[edit]
I think this discussion has drifted away from the initial proposal.
For continuing a discussion about a mechanism that tries to address the “trust problem” this topic might be appropriate, because I think this process can be reverted as well.

The “small LPC topic” could be discussed here: Smaller LPCs are best for now. Submit a proposal!

These are just suggestions, but they would help to keep the discussion better structured.
Otherwise it will become hard to find these discussions once this forum is more populated than it is right now.
[/edit]

#4

Another form of collateralization is to actuallly use NSR to buy NBT – the custodian needs to put up some amount (1:1, at least in value) in order to be able to “unlock” their NBT grant. If they can’t for some reason, someone else could, otherwise the NBT grant would eventually be destroyed by the protocol after some period of time.

The NSR used to buy the NBT would be distributed to a reserve of some sort (smart contract?)

I’ve started to think about ways to address this. So far my thoughts are very rough, but I might as well share them in hopes that commentary will make them better or show me large logical holes that need to be addressed.

#5

My concern is two issues

  1. profit model. If we have no good profit business, we will run out of money sooner or later, i.e. a ponzi scheme result.

  2. trust problem. Nu is a DAC, distributed autonomous corperation. If we can’t get decentralized, we will just be worse than E-gold. Firstly people don’t trust anonymous several persons holding their FIAT as LPC. Secondly, government can easily locate “anonymous” LPC with real name account on exchange, and ruin the whole system.

This is dilemma, if you expose yourselves to public, people may trust you but government will control your fate; if you hide and get centralized, people don’t believe in you. The only way out is decentralized. Thus we must achieve large quantity such as 10,000 small LPCs.

#6

I don’t think we’re quite ready to scale to that point, right now, but I like the way that you’ve framed this.

#7

Several wealthy LPCs are primary targets of hackers and corrupted exchanges besides the government. US government have ruined E-gold and liberty reserve.

Liberty Reserve was a Costa Rica-based centralized digital currency service that billed itself as the “oldest, safest and most popular payment processor … serving millions all around a world”.[2] The site had over one million users when it was shut down by the United States government. Prosecutors argued that due to lax security, alleged criminal activity largely went undetected, which ultimately led to them seizing the service.[3]

In May 2013, Liberty Reserve was shut down by United States federal prosecutors under the Patriot Act after an investigation by authorities across 17 countries. The United States charged founder Arthur Budovsky and six others with money laundering and operating an unlicensed financial transaction company. Liberty Reserve is alleged to have been used to launder more than $6 billion in criminal proceeds during its history.[4] Liberty Reserve has been reopened by a Canadian firm Liberty Reserve Canada LLC and plans to still operate similar money transfer services but complying to laws and regulations.[1]

#8

I also think this is a great idea, because it follows the concept of decentralization and elimination of trust as far as possible!

Challenges / problems

  • we’re not ready to scale to that point (thousands of custodians)
  • technical problems (nore than one NuBot per exchange and trade pair)
  • a framework to get this done (direct exchange between NSR/NBT, depositing securities, two-direction burn model: NSR->NBT/NBT->NSR, etc.)

The reason why we need to face this callenge has been lined out very well by @Sabreiib while I was still typing these lines. The (exchange) accounts of the custodians are the place that can be attacked by governments and other entities.

Allow me to beg one more time: can we please continue this discussion in an appropriate topic?
This has really nothing to do with “transfer message function” and “Public address”…

#9

Ben, those parked NBT are held by protocol, not by any person.

Can we park NSR and get held by protocol? I mean smart contract.

We need many small LPCs, one hundred LPCs is OK, one thousand is excellent.

My imagination:

The granted custodians generate large quantity NBT and park them with zero interest rate. Many small NSR holders sign smart contract with protocol, they park their small quantity NSR to protocol for one week, but NSR is still in their hands and can vote/mint, but can’t transfer to other addresses. Then small shareholders unpack and get 50%(this ratio can be voted) value parked NBT by system voting. With borrowed NBT, small shareholders become small LPCs. Of course there is incentive being a LPC.

One week later, if any side won’t renew the smart contract, small LPCs must return the borrowed NBT to redeem pledged NSR, if they can’t, protocol/smart-contract will keep their NSR. If large NSR holders think they should decrease supply, just vote to cancel those short term contract, within one week, supply will decrease to ideal level.

Many many small LPCs is decentralized style, Nu system can prevent hackers/government and the public believe in those small LPCs because of pledged NSR.

#10

masterOfDisaster， we can talk trust problem in other thread.

#11

I’ll split off the comments to their own topic.

Leakage detection
#12

Thank you @Sabreiib for bringing out this idea and thanks @masterOfDisaster and @Ben.

How about this idea -

  1. NBT grant for decentralized LPCs are generated but locked.
  2. A share owner can unlock some NBT grant via a special BUY transaction attached with a bot ID. After selling, the share owner gets some LPC-NBT (1:1 to NBT) which can not be circulated in NBT network.
  3. The bot with the ID sees in the block chain that someone has bought NBT for it, it unlocks the number of NBT and put it in the wall fund pool.
  4. The spread earned by the bot is distributed, in a form of LPC dividends, to those who contributed NBT proportionally, after the exchange and the bot-operating custodian have collected fees.
  5. The share owner can sell its LPC-NBT via a similar SELL transation. The number of NBT will be taken off the wall fund pool. The LPC-NBT will be destroyed.

This way we only need one bot for one exchange. Shareowner/LPCs don’t need to be borthered with daily bot operations. If the governemnt gets hold of one LPC and ask her to handout all shares, all can be lost is the LPC’s fund. There is no way an LPC can run away with the fund. LPC-NBT are like shares of a bot and could in theory be traded.
The complicated thing might be in determining the price for BUY and SELL. Another problem is that most share holders only have, say, hundres to thousands of dollars’ worth of NuShares, while the bots needs millions of dollars of liquidity. Shareholds may not provide enough NBT the market needs with their 1:1 collaterals.

To complement sending NBT to the bot, the bot can accept USD transfer from members on the same exchange who can also submit a NBT address for collecting LPC dividends. Actually the bot could even accept NBT transfer to the peg fund from other exchange users, with a promise of rewarding dividends. The difference between NBT from exchange members and the NBT received from the LPC is that the latter is from the grant – the risk is on different parties. Third party bot could exist because NuNet is providing a risk umbrella.

Leakage detection
New mechanism to reduce NBT supply: NBT Burning
#13

Thank you @mhps for joining this important discussion and for sharing your great ideas!
I very much like to see this develop into a direction that helps mitigating risks that are inherent with using the current custodian model.
For those who just want a brief overview, there’s a tl:dr section :wink:

I still see some “centralization” by having one NuBot that controls the tx-IDs and by that the cash-flow.
This would put the NuBot in the focus of attacks (more than it currently is; attacking custodians would be more simple).
Even if the NuBot could be considered an attackable entity it would provide some benefits to go this way:

  • the custodian (human being) that currently needs to be considered a point of attack would no longer be needed that urgently (or at all!)
  • the custodians would be assisted by the new NuBot and there would be lesser need to trust people with a huge grant
  • by avoiding fiat transactions and utilizing crypto coin/asset transactions the NuBot could automatically verify the transaction to unlock the NBT
  • especially payment with PPC could be nice to create a stock of PPC for dividend payments (while other crypto thingies should be avoided because of the volatility…)

might not be that big problem because the NuBot is already capable of determining rates to maintain the non-NBT/USD trading pairs.

There could be an option for not placing the bought NBT in a wall but having them sent directly to an address of the buyer. Because the buyer then needs to sell the NBT him/herself, an appropriate charge would be a appropriate (sounds a little bit like custodian job, right?).

I try to sum it up in my words. Maybe I didn’t understand it right and write nonsense here :wink:

[tl;dr]

Using the NuBot for automating the distribution of NBT (directly by the NuBot!) would ease maintaining the (sell) walls and by “buying NBT directly from the NuBot” it would be like going to an exchange office to trade whatever (PPC preferred!) for NBT.

[/tl;dr]

The shareholders wouldn’t be needed to distribute the NBT (by custodian grant). They would decide (vote!) on the amount of NBT to be made available for the NuBot (similar to now where shareholders vote for custodian grants). So there would be times where this “exchange office” would have no NBT to trade…

A requirement for that would be a NuBot function that can’t be attacked/abused easily. I’m not comfortable with that single point of failure. I like automation but I’m aware that things can go horribly wrong if the AI of the NuBot is confronted with situation for what it’s not prepared…
A NuBotNet (multiple NuBots that use multisig tx to secure the funds even if single NuBots are compromised?) would make me feel much more comfortable at least in terms of DDOS prevention and successful single NuBot attacks.
The hardening of the NuBot and ensuring that the NuBot can’t be “robbed” are technical issues which I can’t properly assess. But I fear that this might be a problem that can’t be mitigated easy if ever.

This sounds technically very complicated and I bet there are a lot of flaws. But at least it sounds like a way to cancel out some trust and centralization aspects.

New mechanism to reduce NBT supply: NBT Burning