Proposal : putting a cap on the maximum size of orders

[desrever]

I was reading @KTm 's proposal (ID goes here) and thinking : does it make sense to have a sell wall 1800000 high, even when daily volume is very low? Maybe part of the grant can be taken off-the-exchange for added safety.

Shouldn’t we put a CAP on the size of the order instead of putting “the whole balance” ?
There is also another situation to consider : a new exchange opens while the 1800000 wall is up on exchange1. We now need to move at least some NBTs to the other exchange, but they are all ‘on order’. So it will require some manual work to turn down the bot, take down on order, and then turn the bot back on (or it will always try to put all the balance on order).

If you think that having a ‘max-order-size’ parameter will benefit us, please say it. If you see possible drawbacks to this, also, speak up.

[Ben]

In talks with the exchange that I’ve been negotiating with, if all goes as planned, we’ll likely have more than a single market at launch.

• NBT/USD
• BTC/NBT
• PPC/NBT

I presume that if this is true, we’ll need to split the initial liquidity up, so unless something changes, or if @KTm has a different opinion of why your concern isn’t valid that she will be able to explain, I doubt we’ll run into the situation you described.

For your suggestion, what type of guide would the NuBot use to determine what “max-order-size” is? Is it dynamically generated, or is it something that a user would manually set?

[desrever]

I imagine that is a parameter that can be set by the custodian from options.json.

I understand that we can have 3 pairs on day one. But what about the security risk of keeping all the balance on a single exchange? Let really say that we have an initial 1000USD daily dollar volume ? It will take ~ five year for a 2MIL wall to deplete and in the meantime the whole sum is subject to risk. Isn’t it wise for a custodian to increase wall size dynamically?

[Ben]

I suppose that exchange risk is always going to be a (potential) problem. If the custodian is moving funds on and off the exchange, I see a possible problem with transparency. The protocol’s broadcast of liquidity information works because it pulls in information from real exchanges* so if funds were moved off of the exchange, where would they go? For crypto (like NBT), it could go back into either the grant address, or another one that is identified by the custodian, but for fiat … like the profits that are recycled in @KTm’s proposal … I worry that an even larger logistical problem may be introduced.

Are traders going to be expecting that there will be available, visible liquidity in on the exchanges? Does the preception of the size of the walls make a difference for how “secure” nubits appear to be as a currency?

* Though I suppose it may be suscepitble to being spoofed, but I am not sure. I was not involved in that part of the development so I do not want to speak out of turn or introduce a confidence problem.

[desrever]

Does the preception of the size of the walls make a difference for how “secure” nubits appear to be as a currency?

I don’t know, only time will tell us. It shouldn’t , and if it does its irrational. Custodian grants are public anyway, so knowing that they are in the custodian wallet’s ready to be sold, or already on sale, shouldn’t make a difference , confidence-wise

[Cybnate]

Higher walls do translate in higher confidence I believe. Not everyone tracks the custodian grants, but the walls on the exchange will be very visible for everyone. It also shows confidence in the exchange although others may say that you shouldn’t have all eggs in one basket. Therefore I think we should ideally have a second exchange up and running in the short term, either to transfer the wall to in case of availability issues or as a second ongoing exchange with walls.

The risk of moving the funds off the exchange is that we need to rely on the 24x7 alertness of the custodian to replenish the walls in case of sudden changes in volumes caused by unexpected events.

My conclusion is that we should have them all up initially given the uncertainty on the volumes. After a couple of weeks demands and trends will be more clear and walls can be adjusted when difference between volumes and walls are unbalanced.

[JordanLee]

I agree with David that if we have an NBT/USD pair available to us then we should not be putting LPC liquidity on the BTC/NBT or PPC/NBT pairs. This will result in the potential for loss due to minor latency in repricing orders. As David indicated, these pairs will be arbitrage opportunities that will naturally attract volume.

[desrever]

pair available to us then we should not be putting LPC liquidity on the BTC/NBT or PPC/NBT pairs. Th

@JordanLee I have a related question .
Next week the danish exchange will support us with three pairs : NBT/USD NBT/BTC and NBT/PPC . We will then likely have the whole @KTm custodial grant as NBT balance on her’s account.

Would you then use the bot only with NBT/USD pair for the whole amount and let the market figure out with arbitrage for the other two pairs?

[JordanLee]

Yes. Exactly this.

[Chronos]

I agree with Jordan’s approach to put all LPC liquidity on the USD currency pair. It allows the walls to be extremely accurate and dependable, even when the market value of BTC or PPC is moving quickly.