[Passed] NBT entry gateway on Poloniex provided by @masterOfDisaster

[masterOfDisaster]

##Intro
Dear community,

over the last weeks Nu was busy bringing NBT to market to satisfy an increased demand. Selling them from T4 or T6 to T1-3 to hold the peg was a tough mission, which was accomplished due to the help of so many.
Thank you for that!

Especially during the mid of calendar week 47 there has been serious demand, emptied sell walls on Poloniex even with danger for the peg.
The FLOT is being formed and hopefully will make some things easier and more reliable.
Fixed cost compensation for liquidity providing will increase the incentive for liquidity providers to buy NBT from Nu if the sell walls start to run dry and aid the liquidity situation.
Poloniex is at the moment the exchange with the highest liquidity and volume for trading NBT. I might dislike the fact that so much depends on Poloniex, but that doesn’t change the fact.
Measures to strengthen other exchanges are not subject of this post, though.

This post shall deal with a possible answer to the question: how to bring NBT to market efficiently?

Implementation

Last week I created an NBT entry gateway to allow injecting NBT into market without manual intervention from my side. I don’t want to repeat all of [Concept] NBT entry and exit gateways here, but highly recommend reading that post to develop an understanding of what I’m going to propose.

TL;DR
If Nu needs to bring NBT to market, it can either deal with individuals who buy NBT and pay with BTC, which can be a cumbersome and slow process.
This proposal makes use of custodians who provide Nu with NBT entry gateways. The custodians run NuBots on an exchange (e.g. Poloniex) in a mode that only sells available NBT, but doesn’t place BTC on the buy side. NBT get exchanged for BTC and the custodian transfers them on a regular basis to the T4 buy side reserve.

Evaluation
An obvious drawback is that funds that are owned by Nu are temporarily in the hands of custodians. They are prone to theft, loss and exchange default - just to name a few risks.
In difference to the liquidity operations of the custodians @KTm and @jmiller who started providing liquidity with funds owned by Nu, this custodianship poses only a temporary risk for Nu’s funds. And this risk only needs to be taken if other methods to sell NBT fail or are too slow to fill the demand for NBT.

Proposal RIPEMD160 hash: ceb2e883a7706d3ba0e8ff4a5caec088b50e8b82

=##=##=##=##=##=## Custodian Hash starts with this line ##=##=##=##=##=##=

Custodial Address: BETwD8nSjtj9ADSvej2na34xmsMYwPRymv
Amount Requested: 1 NBT

@masterOfDisaster - below called “the operator” - will operate a NuBot on the exchange mentioned below. The liquidity is being broadcast using the custodial address to allow tracking the liquidity situation of the bot. If the gateway function is required Nu (FSRT, FLOT) sends NBT to the exchange account deposit address. The NBT get sold and the earned BTC are not put into orders, but get transferred to a Nu address (T4 buy side) on a regular basis.
The process of selling NBT can be tracked with the getliquiditydetails RPC.
Disclaimer
The operator charges no fee for this entry gateway, but offers no guarantee for anything - neither the availability of NuBot nor malfunctions of NuBot nor outages of the internet access nor the funds on the exchange. The operator promises to send all (remaining) NBT and traded BTC to Nu (FSRT, FLOT) on a regular basis or burn the NBT. The operator won’t take any of the funds or trade them.

Operator: @masterOfDisaster
Type: NuBot
Exchange: poloniex.com
Pair: NBT/BTC
Mode: Sell side only
Spread After Fees: < 1%
Duration: 60 days
Operator Fee: 0 NBT
Total Grant: 1 NBT

=##=##=##=##=##=## Custodian Hash ends with this line ##=##=##=##=##=##=

Verify. Use everything between and including the <custodianhash></custodianhash> tags.

Forecast

This type of gateway can be used to remove NBT from the market, by turning the NuBot operation upside down and making it place only buy side orders, selling BTC for NBT. I’m going to draft a grant for that as well, because I feel the buy side might benefit from having such a tool. Loosing the peg in the other direction would be even worse.

If Nu is close to releasing other products, this type of gateway might be very helpful as the demand for new products can hardly be estimated before. This gateway is basically unlimited in its capabilities of bringing Nu products to market. Only withdrawing the proceeds might take some time depending on withdrawal limits.

I highly encourage others to do the same - provide NuBots as entry and exit gateways. The more NuBots, the better the availability and the less money in the hands of a single custodian. Please don’t get shied away from my operator fee of 0 NBT. I think this can be a service of certain value for Nu and NSR holders should seriously consider voting for grants from other people who charge a fee.
After all it impacts the usability of the exchange account, requires some efforts, but most of all it’s valuable for Nu!
I just though it’s important to get the ball rolling, to have this means of peg defence. I’m not interested in earning money with it, but want to help keeping the peg intact.

Famous last words

For emergencies two NuBots stand ready on Poloniex - one sell side only, one buy side only, each operating on the same exchange account, but with individual API credentials.
They can be used to defend the peg in either side if all else fails. @JordanLee and @FSRT have been informed how to use them.

I crated this draft to authorize using custodians and their bots as gateway formally.

I’m interested in reading your comments.
And once again the recommendation to read the whole thread [Concept] NBT entry and exit gateways - some questions you might already be answered there

edit:

NBT address to fund account (this was the NBT entry gateway / sell side, but is no longer; it’s now dual side!):
BNCaRW9ZjSLeGkwCfMb48izHy2tXLAVR2i

BTC address to fund account:
1L4DcRj9qu7EjUSnuyceEyUunThx41fmhY

Remark:
as the FLOT has started its operation, addresses controlled by JordanLee or FSRT will no longer be used to withdraw funds from the gateway.
Any funds that are sent to this address will be converted to BTC by NuBot and the proceeds will be withdrawn to a FLOT T4 BTC address. NBT funds that are not yet converted might be withdrawn to a FLOT T4 NBT address.

Update:

Current Liquidity

Outlook: the gateways will very soon not be “always on”, but can be activated with short lead-time.

This will reduce the utility of the gateways as they can’t be used without manual activation (from me) after being funded by e.g. @JordanLee or @FSRT.As only the FLOT did fund them the last times, Nu might be able to live with that.If not, additional gateways need to be created.

[zoro]

The buy side nubot will need BTC to be operational, correct?
Since all the T4 BTC are used for NSR buybacks, what BTC reserve exists to be send to buy side if needed?

[Nagalim]

all the T4 BTC

We currently keep $80k. If that T4 Circulating NBT Threshold motion passes, we’ll have at least a 15% reserve that doesn’t get used for buybacks.

[masterOfDisaster]

The buy side nubot will need BTC to be operational, correct?

Correct!

If that T4 Circulating NBT Threshold motion passes, we’ll have at least a 15% reserve

And this is why the T4 Circulating NBT Threshold motion is so important - the more NBT are circulating, the bigger the buffer to deal with the effect of people dumping NBT on the buy side is required!

[Cybnate]

I think this is a sensible emergency strategy and I hope we don’t need it often.
The only issue I have with it is its availability, an emergency strategy should be available at times we need it. We have seen what happened with FRST, we shouldn’t make the same mistake again. Is adding more custodians the only way to strengthen this emergency strategy?

[masterOfDisaster]

The only issue I have with it is its availability, an emergency strategy should be available at times we need it.

As I’m (meanwhile) an elected member of the FLOT it’s likely that I will be available to fix issues with the gateway in times it’s needed! The FLOT is not operational, but I think it will soon start to ramp up its operation.

That’s why I hope additional members of the FLOT are the next to join this gateway idea, given that the shareholders approve of it - they are required for the liquidity operations and can have an eye on their NuBot.
Plus the shareholders already gave them a vote of confidence: they trust them with funds in property of Nu!

It would be good for decreasing the centralization to have several of these gateways and I’d welcome if community members outside FLOT provided Nu with such gateways.
The setup is quite simple and a single RaspberryPi 2 can run both nud and 2 instances of NuBot. This requires a little experience with Linux, but not that much.

If there’s no opposition against this idea in general or demand for important adjustments, I’m going to hash the grant as soon as the assistant is awake and see what happens in terms of voting.

I’d be happy to read more efficient and powerful ways to bring a lot of coins at market. So far this is the fastest and most efficient approach I could think of.

[zoro]

These gateways could work along with LPs but only in case of emergency?
Moreover, the gateways will not get fees from ALPs’ budget during operation?
In general i think it could be a great tool since we are going to introduce more NU products

[masterOfDisaster]

The case of emergency was what make me think of this solution.
It can be used for much more, if NSR holders approve of that.
Introduction of new products is only one use case, which I already outlined:

If Nu is close to releasing other products, this type of gateway might be very helpful as the demand for new products can hardly be estimated before.

I don’t know how it can be verified that these funds will not receive compensation from ALP.
I promise I won’t operate an ALP bot on an exchange I provide an entry or exit gateway.
But even if somebody did it - this would in case of emergencies be the least to worry about…

This is different if you think about introducing new products.

edit: grant promoted to [Voting]

[cryptog]

To me this could be a crucial in case of an emergency.
A last resort.
I will add it to my feeds very soon.

ceb2e883a7706d3ba0e8ff4a5caec088b50e8b82 verfied.

[Nagalim]

I don’t think we need to worry about the operators of gateways double dipping other than telling them that we would like them not to. They are already being entrusted with credit, it’s kinda the name of the game that the gateway is trusted not to do shady stuff with it.

I would like you to consider running the bot at a 0% SAF.

[masterOfDisaster]

To me this could be a crucial in case of an emergency.

Thank you for your support. The complementary grant (NBT exit / BTC entry gateway) is under construction and will be posted soon.
I’m going to post it as [Voting] immediately, because I see no reason to be against that variant if nobody opposed this one.
I hope that others, especially form the FLOT join this idea and operate this kind of gateway.

I would like you to consider running the bot at a 0% SAF.

I was just thinking of a little bit above 0% SAF for the benefit of Nu.
Wouldn’t 0% SAF lead to endless shift of the walls and in case of the walls being shifted too late a loss for Nu?
An example of the current results of NuBot parameters is this:

12:27:00.107 [pool-1-thread-1] INFO - Computing USD prices with sellOffset:0.005$ and buyOffset:0.005 : buy @ 1.0062999999999998 buy @ 0.994 [com.nubits.nubot.streamclient.Subscriber:426] 12:27:00.119 [pool-1-thread-1] INFO - Actual prices from bitfinex (using 1 BTC = 322.15 USD) : sell @ 0.0031237 BTC; buy @ 0.00308552 BTC [com.nubits.nubot.streamclient.Subscriber:436]

Is selling for $1.0063 (before fees) really to much? I doubt it, but willing to adjust it. “<1%” leaves some room.

[Nagalim]

So this is something I thought about a good bit, actually, when I was running the bot as a gateway. I had up a 5% buy side, which seemed alright, but it really shoulda been more like 10%. The sell side I ended up at 0.1% total offset, even with 0.25% deviation. At the worst that’s selling at $0.9985. Anyway, is the price even known within 0.25%?

Something that ends up being interesting is that if you sell at $0.9985 you may end up catching a buy side custodian. Now, here’s my question to you: is that even a problem? Ultimately, my answer was no.

[masterOfDisaster]

So this is something I thought about a good bit, actually, when I was running the bot as a gateway.

I’m glad that I’m not alone thinking about it

I had up a 5% buy side

This NuBot’s buy side is effectively not existing.

The sell side I ended up at 0.1% total offset, even with 0.25% deviation.

To be honest I don’t know how NuBot calculates that. The parameters are different. There’s no deviation, just

  "bookSellOffset": 0.005,

if you sell at $0.9985

You make a loss if the fee is 0.2%, but only little.

Now, here’s my question to you: is that even a problem? Ultimately, my answer was no.

Except for wasting a little bit of money to fees (see above) I can’t see any more drawbacks. My ultimate answer would be no as well.
I wouldn’t try to tell custodians and non-custodians apart in this case.

[Nagalim]

We need a standard for what price gateways should sell at. I guess for now we just let the gateway operator decide, but I was thinking we could use an incentive, like you get paid x% of each T4 NBT you sell at $1 or above on-exchange.

[masterOfDisaster]

This is rather meant for emergencies. I don’t think it’s in general a good idea to put NBT or BTC from T4 directly on T1.
Trading it off exchange (T4 -> T3) should be the preferred way to go.

In the end it doesn’t make a big difference whether the funds hit T3, because they are intended to be put on T1 soon or they are directly put there.
It shouldn’t mess much with the liquidity mechanism.
I hope that I’ve written enough about benefits and drawbacks of the custodian’s role.
Maybe this sort of gateway is useful not only in emergencies. We’ll see…

With B&C Exchange things will be a lot easier and this sort of gateway is no longer required

I was thinking we could use an incentive, like you get paid x% of each T4 NBT you sell at $1 or above on-exchange.

This would be a nice idea, but it would require evaluating the logs to determine the reward. If NuBot is not configured to put all funds in a single order, it may sell some NBT for even more.

I think a compensation (e.g. grant above 1 NBT) could be a better way to go. I wouldn’t want to create conflicts of interest with a custodian who wants to have a high x% to increase the compensation while Nu wants a low x% to have a sound peg.
As a “gateway custodian” normally shouldn’t have much work (if any), a small compensation for effectively disabling the exchange account and the need to run NuBot(s) somewhere might be more fair than a share of the traded funds - normally they won’t be traded via the gateway.

If you want to have a predictable fee, you need a “bookSellType” linear with “bookSellSteepness” flat, I think.
@desrever knows for sure

If you are interested in NuBot parameters, have a look here:
https://bitbucket.org/JordanLeePeershares/nubottrading/src/7dea551e0794f102b8bc904268e4c42e6e48b064/docs/SETUP.md?at=master&fileviewer=file-view-default

[Nagalim]

well, so x can be small if it’s not much work. But what I’m looking at is that when funds are put in a gateway, it is intended that they will leave the gateway and integrate with buyers and custodians within a reasonable period of time. The T4 signers should not put more funds in a gateway than they want to sell right now. Therefore, it is a good idea that we stimulate gateways to sell as many as possible, as long as they are above a certain price according to a well defined price feed. The issues here are technical, but we can mostly hand-wave them by saying that gateways are trusted with credit anyway and, as long as we are clear about the rules, the operators will follow them.

[masterOfDisaster]

The T4 signers should not put more funds in a gateway than they want to sell right now.

Very important observation!

Therefore, it is a good idea that we stimulate gateways to sell as many as possible, as long as they are above a certain price according to a well defined price feed.

People who sell BTC for NBT to get rid of BTC volatility (by putting that burden on Nu) should willingly pay $1.0063 for it - a small premium for such a fine service.
And taking the volatility into consideration, they might very well be able to buy the NBT at close to $1.
The closer the price moves to $1, the bigger the risk to sell below $1 if the price moves faster than the NuBot can adjust the orders.
As I’m only providing the gateway and don’t intend to make money operating it, I’m fine with anything between 0% and 1% SAF

But seriously, I think we should postpone a part of that discussion or move it to a thread that deals with using this sort of gateway on a regular basis.
This gateway is intended to bring NBT to market at a price of $1.0063 in times where demand for NBT can’t be satisfied by other means. Is it in this case really important whether it’s $1.0063 or $1.004 or $1.002?

[Nagalim]

Is it in this case really important

No, it’s nowhere near important, it’s just something I’ve been thinking about. This is an excellent grant.

[masterOfDisaster]

No, it’s nowhere near important, it’s just something I’ve been thinking about.

I find it important to think about such things! These thoughts are important in general for liquidity operations, but maybe not exactly for such a last resort (that offers NBT at below $1.01) - at least in my opinion.

[Nagalim]

yah, I definitely don’t want to see us offering $0.97 again in times of a peg break over $1. It just seems so counter-intuitive.