Experiment of Trading Volume at Different Spreads Using a Gateway

mhps 2016-05-29 08:55:44 UTC #1

Now that poloniex has 0 buy side within 1.5% spread, I think this is a very good opportunity to test Jordan’s theory that there is foundamentally no way to predict market so there will always speculators to buy and sell.

Now there is no buying possiblly because the sell side of Nubot has too big spread for the small flectuation of btc prices that traders could profit from by scalping. So shall we lower the spread of a pool to super low to attract these traders?

There is no risk as the buyside is already gone.

If these traders appear, @JordanLee’s theory is validated.

@henry @woolly_sammoth

Current Liquidity

masterOfDisaster 2016-05-29 09:15:25 UTC #2

We will need to have a look at both the price and the liquidity.
I bet there’s always some liquidity. And I bet the spread will be within certain limits.
Hopefully the NuBot gateway parameters are just at the border of that limit.
Oh, and I bet that this limit depends on the market situation.
The result will not exactly be what we imagine as “deep liquidity”.

mhps 2016-05-29 09:31:13 UTC #3

Are you replying regarding the experiment? Of course to validate the theory there has to be very noticeable difference in trading volume when spread is lowered.

We can start with $10k sell at 0.5% SAF. The proceed goes also to 0.5% SAF.

ps. I just realize there can be both obligatory low spread liquidity (what Jordan wanted but with a cap) and high ones at the same time. Was your motion allowing that?

cryptog 2016-05-29 09:57:37 UTC #4

Very good suggestion. I am in favor.

Cybnate 2016-05-29 09:59:49 UTC #5

Yes it is, same for my motion but mine is more specific. Both spreads exist at the same time.

zoro 2016-05-29 10:43:55 UTC #6

Sell side in polo is already below 0.005 offset.
Both my nubot gateway and NuPool, but with a very little trading.
NuPool has already buy side offset at 0.005. Whenever i put some btc there, are traded in minutes!
The thing is, where we can find BTC to put in this tight buy offset!
In theory sure, there will always traders but with what are you going to “feed” them?

masterOfDisaster 2016-05-29 11:14:11 UTC #7

Your motion has a bit too tight limits for the max spread.
We would have been out of BTC since yesterday with your motion, even faster with Jordan’s, but we still have BTC at a high offset, which is only allowed by my motion (and Nagalim’s, but that one is still in draft version)

mhps 2016-05-29 15:05:54 UTC #8

If the buyside has big spread trader will be hesitant to buy nubits. Traders could test buy side spread by buying a specific amount and observe where the proceeds btc are placed. But I agree that we cannot show where the buy wall is adds friction.

mhps 2016-06-01 09:05:25 UTC #9

@woolly_sammoth and @Cybnate you are the only ALP pool managers on polo, right? Is there any problem in lowering the spread to 0.75% SAF for 24hr for the above experiment that could help to validate @JordanLee’s theory?

Cybnate 2016-06-01 09:08:33 UTC #10

I’m only running a gateway on Poloniex, no pool.

However I do have a working ALPv2 server sitting idle and waiting for some experiments and testing. But I do not have a custodial grant or other funding for it.

woolly_sammoth 2016-06-01 09:21:34 UTC #11

It’s technically possible to lower the spread on the running NuPool instance. I don;t know where that leaves the pool in terms of operation though as it;s running under the parameters that were voted for to get the funding.

mhps 2016-06-01 09:23:43 UTC #12

Is this [Decommissioned] PyBot passive gateway on Poloniex and on Bittrex its current status? So if you have nubits on the sell side with say 0.6% SAF on both sides, then traders could buy the sell side with 0.3%+fee and sell it back from the gateway’s buy side with 0.3%+fee?

Is it unfunded? All you need is FLOT to fund it?

mhps 2016-06-01 11:09:08 UTC #13

I don’t know if the pool motion approved prohibits such change of spread. As all LPs can freely set his/her own minimum offset, I suppose lowering pool spread doesn’t force anyone to trade at a lower spread. Only those who don’'t care will get lowered.

But the uncertain amount of liquidity at different spread makes the experiment environment messy and difficult to interpret the result. It’s better to have a dedicated pool or gateway. @Cybnate’s gateway seems to be ideal.

Cybnate 2016-06-02 01:44:17 UTC #14

Had about 2000 NBT volume in last 24h at current spread (1.4%) on my Pybot gateway on Poloniex. That is encouraging. I’m prepared to seek the margins of the motion the gateways are running under and reduce the spread with the transaction fees for the experiment. With average fee of 0.2% on bot sides, the effective margin will be 1%. The Pybot can’t work with uneven buy and sell side.

I recommend just start with the 1.4% to test the waters. When BTC buy side liquidity is gone within hours, than I’m not sure whether it is worth to reduce the spreads even further at this stage. I believe Jordan’s comments re tight peg reflect a normal situation where decent reserves are available, not really what we are in now.

What about the equivalent of about 2000 NBT to start with? If ok, I will withdraw 2000 NBT to FLOT to keep the balance of funds the same on Poloniex in order to keep the exchange risk the same.

Regardsless I will raise a motion which request more flexibility for the spreads of the PyBots.

mhps 2016-06-02 03:50:36 UTC #15

Of course it’s worth it. I suggest start with all sell side and no buy side. Just observe. Take advantage of this singular opportunity.

Cybnate 2016-06-02 04:04:20 UTC #16

Not sure what you are suggesting. There are 2000NBT on sell side at 0.7% offset (1.4% spread on pair). As said I’ve seen some volume (trading backward and forward in last 24h), but I doubt that will continue with BTC having upward pressures again. Reducing the sell side spread is unlikely to make a difference in BTC bull market.

Cybnate 2016-06-02 04:09:31 UTC #17

So what holds FLOT back to use my Poloniex Pybot gateway to test the uptake at 1.4% spread levels?

mhps 2016-06-02 04:15:03 UTC #18

Jordan asks for 1% spread and that is 0.5% offset on each side.

That is just opinions. We all have ours. What we have is Jordan thinks there will always be tradings going on and reducing spread to 1% attracts traders significantly. moD and you think the effect will be not or not very significant. We want data to tell whose view is closer to reality. So I am suggesting this experiment that has low spread. Your gateway can be a clean test bed although its volume isn’t up to big arbitrageur’s apetite. But still it’s worth a try.

Can you find out your trade volume with 1.4% spread in the last 24hr?

To look at it from another way , it’s now either that or do nothing.

Cybnate 2016-06-02 04:17:01 UTC #19

1500 NBT sell and 1500 NBT buy in 5 transactions on Poloniex in last 36h.

mhps 2016-06-02 04:18:46 UTC #20

Great. So that was from 2000 liquidity provided.