[Nubot Gateway] Poloniex @zoro (sell offset:0.25% , buy offset:5% ) - stand by

[zoro]

[BTC] Bitcoin 19.28535532
On orders: 3.48952500
BTC value: 22.77488032

[NBT] NuBits 8355.76893059
On orders: 1500.00000000
BTC value: 23.10724448

[zoro]

[BTC] Bitcoin 19.34640532
On orders: 3.42847500
BTC value: 22.77488032

[NBT] NuBits 8355.76893059
On orders: 1500.00000000
BTC value: 22.74819882

[Nagalim]

I’d like to point out that this last week you haven’t made any trades but you’ve held $20k on exchange and at risk. In my opinion we should think about maybe withdrawing half of each side.

[zoro]

It seems ALPs are enough ATM. If FLOT has no objection i will transfer 4800 NBT and 11 BTC to their addresses.

[masterOfDisaster]

Depositing funds, especially BTC can take several hours.
In case the BTC would start another rollercoaster, $5,000 on each side wouldn’t last long - I know what I speak of.

There were no trades, because the ALP worked. No big BTC volality means easy money for the liquidity providers. It was rational to put funds in ALP.
Let’s see what happens, if the BTC price fluctuates.
I’m not comfortable with having only $5,000 standing between losing the peg, once ALP failed.
And this is what I expect if BTC is suddenly highly volatile again: liquidity providers in ALP make the rational choice and pull funds.

Do we really want to put that pressure on FLOT and zoro, because of the USD equivalent that gets paid each month (more or less) for liquidity provision in total?

[zoro]

What if i these funds are to be sent to another Nubot in Polo? “More nubots with less funds each” is my moto

[masterOfDisaster]

That limits operator related risks, but not exchange related ones…
Only if NuBot operators withdraw part of the funds that would reduce the exchange related risks.
But where to?
The FLOT might not be fast enough in case a deposit to the NuBot account is needed.
Singlesig is faster, but less secure.

Would 2-of-3 multisig with only a few thousand USD that could be activated much faster be a solution?
A subset of FLOT could do that.

[Nagalim]

Even just single sig. Let @zoro withdraw to an address he holds the key for. The only cost is withdrawal and network fees, and the risk is significantly lower.

[masterOfDisaster]

This way you replace exchange risk with operator risk.
While I have no idea to assess the operator risk of an individual operator, I need to think it’s smaller than exchange risk - otherwise Nu shouldn’t send any funds to that individual in the first place.

I thought of that as well, but shied away from posting it.
Afer thinking again, I need to say it’s just the most simple and reliable way if NuBot operators are allowed to do that: withdraw some funds from the exchange to addresses for which they own the private key.

I haven’t looked at the current terms of NuBot (and PyBot) operations, but I’m almost sure that operators are only allowed to withdraw to FLOT addresses.
That could be changed for all operations in a single motion.

Having the operators handle the funds is a good solution, because operators tend to keep an eye on liquidity.
It would be nice if they weren’t only allowed to fund their own operation, but other official Nu funded operations as well, although that would make accounting much more complex.

Thanks to Cointoolkit you could do all that stuff from anywhere you have the private key, access to the internet and a web browser!

[Nagalim]

Id prefer specific motions listing the addresses explicitly and specifying that the operator can’t hold more in T3 than T2+1. That’s technicality though, there are a lot of ways to contractualize this so it doesnt get abused. Gateways have plenty of unconsidered attack vectors anyway, this would be a step in the right direction.

[zoro]

I like your resolution guys!
Then we can make the nubot operators not only keeping an eye in the liquidity but also behave as a private T3 operators. And perhaps we could extent this T3 operations in a more general way
Although this could raise the nubot’s cost, it eliminates the double risk of operator+exchange to only operator risk.
I will definitely try this scheme in my next nubot proposal.

[huafei]

private T3 operators

Is that something like nulagoon tube?

[zoro]

Yes, it would be a mixture of MLP+nubot with funds from FLOT. A crazy combination!

[Nagalim]

Basically no, the T3 we’re talking about here are specifically for @zoro on poloniex and would not be allowed to trade outside of the exchange. My T3 contract is much closer to NuLagoon than this will be because i have a mechanism for picking prices off-exchange whereas zoro’s only allowed mechanism for trading is to make an order on poloniex.

[zoro]

[BTC] Bitcoin 19.36924935
On orders: 3.40563000
BTC value: 22.77487935

[NBT] NuBits 8355.76934982
On orders: 1500.00000000
BTC value: 22.56586806

[zoro]

[BTC] Bitcoin 12.54537865
On orders: 3.38854500
BTC value: 15.93392365

[NBT] NuBits 11350.83780235
On orders: 1500.00000000
BTC value: 29.31031936

[Nagalim]

How about putting 2500 on the close walls instead of 1500? So that would be ‘bookbuywall’ and ‘booksellwall’ at 2500.

[zoro]

why to change it?

[Nagalim]

I just think we should consider the upsides and downsides. I mean, how much volatility risk are we really avoiding by keeping most of the funds on exchange but off the books? A 2500 baseline on poloniex just seems more appealing to me than 1500.

[zoro]

The idea was to limit the exposure of T1 funds in case Nubot goes down during a high volatility period.
An increase of 2000 or 2500 NBT is acceptable?