[Withdrawn] T3 Trusted Custodianship @Nagalim

[Cybnate]

What is that?

The T1-T3 custodian can be held accountable end-to-end for balancing liquidity and will be able to proof that to the Shareholders. They can use existing tools as the NuBots gateways model to speed up the delivery process to almost immediate.

The T3 custodian will make deals on an ad hoc bases with all kinds of P1/P2 providers. It will be hard to make these deals transparent and effective as in the liquidity ends up in the right place at the right time. Practically ad hoc deals needs to be made and some else need to bring it to the exchange. Many layers, many risk regarding transparency and speed of delivery.

[masterOfDisaster]

The T3 custodian will make deals on an ad hoc bases with all kinds of P1/P2 providers. It will be hard to make these deals transparent and effective as in the liquidity ends up in the right place at the right time.

A solution would be to use T3 custodians to fund gateways.
That wouldn’t (directly) eliminate the risks associated with using gateways (although that can be dealt with), but would increase the speed of transactions by far (compared to requiring FLOT).

If the T3 custodians act based on verifiable (and recorded) thresholds it’s possible to check that they do their job right.

In case they are compensated based on the volume they transfer (with collateral in place a compensation for that is required as well) it needs to be assured that they do so as often as necessary (incentive for that: compensation), but not as often as possible (they are held liable for their actions and will be punished if the transfer more often than required).

This would be something in between the (withdrawn) propsal by @Nagalim and the current situation.

I think it’s overall more complicated because Nu might want the gateway operators to provide a collateral if the gateways aren’t meant to be used for emergencies only, but on regular basis instead.
The collateral would primarily be an insurance against gateway operator misbehaviour.
If the collateral would be extended to compensate exchange defaults, the compensation, which operators did request, would be at least an order of magnitude bigger.

Plus you need gateways for all supported exchanges.

The gateway operators will create costs (compensation) for Nu while T3 custodians don’t necessarily - the compensation for transactions might be paid from the spread. And the share of the spread that goes to Nu might be enough to compensate the fixed T3 custodian costs (for the collateral).

In the end there will a lot of complexity, but I lack to see the benefits.
Maybe I’m overlooking something obvious.

[mhps]

Can a P3 custodian deal with unlimited number of individual LPs? Even if so, it wouldn’t be cost effective hence very time consuming.

That is where T3 custodians are expected to shine. They are for-profit agents living off comissions. They have a natural need to reach out to many LPs to get more business. Every T3 custodian is one person, being able to make rapid decision and act on it. FLOT is a globally distributed group whose best reaction time record is 6 hours.

I envisage that there will be ~10,000 traders served by many 100s T1-T2 LPs in several 10s pools running on ~10 exchanges. The LPs are served by several T3 custodians. One FLOT, in somewhat predictable and routine – therefore transparent – operations, deals with T3 custodians who FLOT can afford to check.

[masterOfDisaster]

That hits the nail on the head.
How to create a scalable solution if not that way?
It’s simple, efficient, scalable and provides proper incentives with little need to monitor, track and record (on Nu side).
FLOT does what it’s best suited for and so do T3 custodians.

[Cybnate]

I think it’s overall more complicated because Nu might want the gateway operators to provide a collateral if the gateways aren’t meant to be used for emergencies only, but on regular basis instead.The collateral would primarily be an insurance against gateway operator misbehaviour.If the collateral would be extended to compensate exchange defaults, the compensation, which operators did request, would be at least an order of magnitude bigger.

Collateral is certainly required and keeps people honest. I bet it won’t be more expensive than a T3 custodian trying to find T1 and T2 LPs to make a deal. With the current lack of real competition the risk on ‘shady’ deals is high. There is a hidden cost to that which might take a while to become clear due to lack of transparency.
To have transparency some rules/complexity is required for a T1-T3 custodian. However that would be a lot easier to agree with a handful T1-T3 custodians than T3s having to negotiate with several T1/T2 providers. It is just moving the problem to a T3 custodian and then hope it will all magically work out. It won’t and if it does it will most likely be at a high cost / commissions and high spreads on exchanges making NuBits less attractive to trade with.

I envisage that there will be ~10,000 traders served by many 100s T1-T2 LPs in several 10s pools running on ~10 exchanges. The LPs are served by several T3 custodians. One FLOT, in somewhat predictable and routine – therefore transparent – operations, deals with T3 custodians who FLOT can afford to check.

It is great to think big, but we should act small first. Rule one of a start-up. Scaling up along these lines is always possible. I’m ok with FLOT’s as is, the problem is the middlemen associated costs and high risk of lack of transparency in T1-T3.

However, reality check learns that there are no competing proposals and as said before getting something in place is better than nothing. Therefore I will add any T3 proposal to my feed as something needs to be done other than continuing discussing this. I hope some of us are able to monitor transparency in T1-T3 with T3 custodian because I don’t see a way to do that properly in a scalable and cost-effective way. Good luck to any T3 custodian to succeed in such environment. When T1-T2 is happy the Shareholders won’t and vice-versa.

[Nagalim]

What’s a ‘shady’ deal? If I sell nbt for $1.001 and give all the money to Nu, why do shareholders care whether we did the deal in the shade or in a nearby well lit park?

[masterOfDisaster]

When T1-T2 is happy the Shareholders won’t and vice-versa.

Ideally both will be equally (un)happy

Allow me to try summing up why I like the (collateralized) T3 custodian proposal

Demand for Nu to have T3 custodians
The T3 custodian scheme creates a transition layer on T3 connecting T4 with T1-2.
That’s what Nu desperately needs.
Using FLOT to balance liquidity on demand is neither feasible nor sustainable.
At the moment there’s nothing else.

Incentive for T3 custodians to do their job right
If T3 custodians are paid by commission for making deals, they have an incentive to make them.
Deals are reported on blockchain and if liquidity data is recorded, it can be checked that conditions to make deals were met.
If T3 custodians don’t behave according to the terms, they will face consequences.

Costs and risks
Can be shifted after the fancy of Nu and the T3 custodial offers:

• T3 custodians can be funded by Nu. Nu takes the risk, but saves money (no compensation for collateral required).
• T3 custodians can be funded by Nu, but need to provide a collateral (full range or part of the funds). Nu has lesser risk, but needs to pay for the collateral.

[Nagalim]

costs

Please tell me what costs there are? I’ve shown that I’m competitive with NulagoonTube for customers and I earn Nu money. Where’s the cost other than 0.2%? Do you really thing you can get 5 signers to agree on a price for a lower overhead than just me? I have no real overhead, unlike NuLagoonTube; my setup is basically just a btc and nbt wallet. This won’t be an exchange like NuLagoonTube, this is me making deals for Nu. And what’s more, it gives us more of that much needed legitimate T3 liquidity that seems so hard to define outside of NuLagoon.

The thing is, all this is is formally defining the position JR was in just a couple months ago. This is a response to him stepping out. This isn’t new, it’s just formalizing a tried and true method we’very been using for over a year now.

So about this ‘shady deals’ thing. Are we worried about money laundering? Do we need to make a whitelist? A whitelist is actually the solution to the T3 custodian using their own funds. It was rejected on moral grounds of creating a whitelist. Why have we never been worried about dirty bitcoin before, but all of a sudden we are now?

Are there any downsides in trying

As far as I can tell, this proposal was rejected on the ephemeral grounds of ‘we don’t need it’. I personally thought this motion was a no-brainer and absolutely uncontroversial, but I was clearly wrong. The new proposal should draw much more controversy because it is uncollateralized. But I wouldn’t be surprised if it passes where this one fails because shareholders seem to like taking the bad deal recently.

I feel like I’m in a continuous version of the trick at mecone, but I’m jumping up and down pointing at the pile with meat and y’all choose the bone pile anyway.
https://en.m.wikipedia.org/wiki/Trick_at_Mecone

[masterOfDisaster]

As far as I can tell, this proposal was rejected on the ephemeral grounds of ‘we don’t need it’. I personally thought this motion was a no-brainer and absolutely uncontroversial, but I was clearly wrong. The new proposal should draw much more controversy because it is uncollateralized. But I wouldn’t be surprised if it passes where this one fails because shareholders seem to like taking the bad deal recently.

The current liquidity balancing relies way to much on the gateways I operate, which have so many drawbacks that I don’t want to repeat them all.
Gladly you made a nice list:

I am blown away that you don't see the risks associated with gateways. I will enumerate them, putting the way in which this proposal alleviates that risk in {}. I will try to put most important first, but I do intend to be thorough. Exchange Default. If the exchange is hacked, we lose all funds in the gateway. This is very reminiscent of ktm and jmiller's operations that cost shareholders so much money. The only other Nu service that has this risk to shareholders is the share buyback pro…

Just imagine I would decide to be dishonest and just run with the funds sent to the gateways.
There was a time when effectively both were funded. One was funded first and before I could withdraw all, the other side was in need for funds and the other gateway was funded.
T3 custodians with collateral are so much better.

With my gateways there’s no collateral and no cost. I charge no fee, but might get tired of operating them, if they continue to get used regularly. But the full risk is on Nu’s shoulders.

(Collateralized) T3 custodians would reduce the risk and increase the speed of balancing.
I really wonder why there’s so much resistance.

How do you imagine getting the liquidity balancing under control if there’s more than just NuBits?
We need to put that on more shoulders!

[mhps]

What’s a ‘shady’ deal? If I sell nbt for $1.001 and give all the money to Nu, why do shareholders care whether we did the deal in the shade or in a nearby well lit park?

Shareholder care if T3 cusodians somehow make market spreads of nubits, or the cost to Nu if Nu has to subsidize T3, greater than needed.
Pool fee (in NBT per month) and FLOT fee (in NBT per month) can be used to form a baseline to compare efficiency of other tiers although the sustainability of FLOT has just started to show as cointoolkit matures so its cost will take still more time to show.

[Cybnate]

The thing is, all this is is formally defining the position JR was in just a couple months ago. This is a response to him stepping out. This isn’t new, it’s just formalizing a tried and true method we’very been using for over a year now.

It is not the method, it is how deals are made and what is paid for that. JL dealt with T3-T6 himself and who knows also on T1/T2. That is very effective but something we don’t want in the hands of only 1 person. Instead we are assigning a custodian to every tier. As advocated it would be better to have custodians overseeing T1 to T3 or even T4 as it would be easy for Shareholders to held them to account re performance. That will be increasingly difficult with the model we are heading for.

So about this ‘shady deals’ thing. Are we worried about money laundering? Do we need to make a whitelist?

No, nothing about money laundering. As T3 custodian you will have to negotiate with T1/T2 providers. You will have to deliver and in a very limited market it will be difficult to do so. So you might be forced to e.g. sell NBT at very low prices (if FLOT allows) or allow T1/T2 to use high spreads (if Shareholders allow). What I hope that all these deals, sell and buy prices, amounts and verifying that T1/T2 custodian delivers in time within agreed spreads will be very challenging for anyone. So my expectations this can’t be done properly and deals will be made without the required transparency from FLOT to T1. Ensuring the T1-T3 (or even T4) process is in control of the custodian(s) will ensure the custodian can be held accountable and performance measured way more easy by the Shareholders.

I will stop debating this for now as I’m feeling I’m repeating myself, let’s get on with it. There are no competing proposals and as said before doing nothing risks to stall the whole model.

I’m ok with a T3 custodian which has some community trust. Nagalim’s proposal fits in that, still hoping for others but that won’t stop me for supporting any reasonable T3 proposal.

I hope that by trialling this it will become clear that middleman will optimise their own commission by making deals with high spreads or force FLOT to sell too low which is not in the best interest of NuShareholders. I’m considering setting up my own gateways with NuBot and picking up nice profits as T1/T2 provider. But don’t complain about the spreads. Let’s learn from this together.

[Nagalim]

FLOT to sell too low

I don’t understand this concern. FLOT does not sell or buy from me, they just refill my pool. Effectively, the price I use is the same price FLOT is using, minus my commission.

Either I will make the fee (what you call the spread) too high and no one will trade with me or I’ll make it too low and not make much money for myself (it’s a balancing act that I will improve on with time), but there is no situation where Nu loses money on this if I don’t break contract. That’s because there is no situation where I sell for less than $1 or buy for more than $1.

[zoro]

Nagalim, i was thinking that your T3 proposal should not compete directly with nulagoon tube!
Why?
Because you are different entities. Nulagoon provides liquidity to anyone, custodian or not.
You are going to provide liquidity only to T1-T2 custodians. You will ask how you are going to know if you trade with
a custodian or not. Good question, i don’t have an answer!
Even if a “trusted” T1-T2 custodian trade with you, there is no quarantee that funds will end up to exchanges!
That is the reason why MoD’s gateways are so popular. And i guess these gateways are the best candidates to bring
funds to exchanges.
Perhaps a T1-T3 custodian should think of setting gateways and make a contract with NU (FLOT) or use them in ALPs.
Then, the so called emergency gateways will be the norm

[Nagalim]

I don’t intend on asking people if they are custodians or not, that brings it back to a whitelist. I will certainly charge a premium fee to people I haven’t dealt with before, and I will surely lower the fee when a peg is in danger and the custodian is trusted (Zoro or NuLagoon). However, I do not have qualms about for example a business that needs 1,000 nbt a week to get it through me instead of on the open market. I know they aren’t a custodian, but their money would end up in the network one way or another and I would actually be cutting out a few middlemen at Tier 1.

[Cybnate]

I don’t understand this concern. FLOT does not sell or buy from me, they just refill my pool. Effectively, the price I use is the same price FLOT is using, minus my commission.

That is the first I hear FLOT is using a fixed market price identical for every T3 custodian (or others?) interested.

Like to refer to NMei’s passed proposal which is a great example of how a T1-T3 can operate and the Shareholders know that liquidity and balancing does occur without all the other deals with non-custodians as you mention in your last post. That were the ones I called ‘shady’ earlier. They are off exchange and not necessarily and verifiable deals favouring NuShareholders or their interests (e.g. visible volume on exchanges).

[Nagalim]

They’re verifiable because they’re on the blockchain, meaning anyone can look at the txns involving the trusted custodian addresses and know what the price was for each txn. It’s true you don’t know where the money goes, but we don’t want to limit our customers to only using nbt as a hedging instrument. Someone came in here a few months ago saying they wanted several tens of thousands of nbt per month for their business and was having trouble finding this on-exchange. That person shouldn’t need to try to siphon through T1 for such regular activities. They should be able to buy from T3 or even FLOT directly, depending on how much of a contract they want. I think that really shows the power of this type of proposal: we can have T3 custodians creating their own contracts without going through the blockchain. As long as every NBT that gets put into the wild creates $1 worth of T4 liquidity, bring on the deals of all shades of grey.

Also, remember that this is a trust-based proposal. I will of course vet the customers according to my own criteria. I am just very wary about telling shareholders I will only deal with custodians because there is no formal definition of that in an ALP. Are nmei’s funds off-exchange really T3, or are they a privately owned wallet? There is no ‘verifiable’ difference.

[Cybnate]

Are nmei’s funds off-exchange really T3, or are they a privately owned wallet?

It doesn’t really matter as we have a contract measuring T1 performance and balancing. That is what counts imo. With your proposal there is no guarantee it benefits T1/T2.

I will of course vet the customers according to my own criteria.

And mileage will vary for sure. Wish you luck with that.

[jooize]

Could someone summarize what a T3 custodian would do? Preferably in a numbered list. I’d like to grasp the effort required, and it might help others in their consideration of applying as well.

What calculations have to be made, and which tools need to be installed …? I assume there’s some kind of bookkeeping. You trade at exchanges, but also directly with people?

Would it make sense for a T3 custodian to setup a gateway for themselves?

[Nagalim]

Setting up:

1. Get an nbt client
2. Get a btc client
3. Post a proposal using an address in your nbt client. Get it passed.
4. Deal with any collateral
5. Give FLOT your btc address.

Doing a trade:

1. Find someone that wants to buy a reasonable amount right now for > $1 or sell for <$1.

2. Tell them the price, markup, and minimum and maximum volume. Tell them to deposit the money in the correct address within the next half hour or whatever.

3. Wait for a couple confirmations, then send them money. Record the price, markup, time you chose the price and on what exchange.

4. If you have too much of one side, wait 24 hours then send it to the proper FLOT address.

5. If you need more of one side, present your txn records on friday so FLOT can put in a txn saturday. Tell them how much you need.

6. Do your best to report how much T3 liquidity you’re providing manually using the liquidityinfo rpc command.

7. Remember how much you’ve earned from your markups (remember to subtract 0.1% for Nu). When it becomes a reasonable amount ask Nu to pay it, either through FLOT or an individual grant.

[Nagalim]

By the way, I highly suggest shareholders demand a fixed markup % (0.3% total, 0.1 for Nu and 0.2 for the custodian). It alleviates a lot of the trust by making the custodian’s compensation known from just the txn volume, which is verifiable on the blockchain.