Current Liquidity

I never got this:
In nu client, liquidity in vote page shows
type —buy sell
unkown 0 0
T1 60k 35k
T2 27k 33k
T3 14k 29k

alix shows
ask bid
70k 59k

Why don’t they agree?

I have no explanation why the report from the Nu client for T1 is lower than the ALix report.

I could provide explanations for differences, but only, if it were the other way round:

  • funds on T1 with a spread above what ALix reports (e.g. my Poloniex gateway)
  • not properly reset liquidity broadcasts

But for ALix reporting bigger values than Nu client I have no idea.
Regarding liquidity situation:

Liquidity; ALix T1 + gateways (poloniex dual side, numerous sell side):

Gateway status at Poloniex (dual side), NuLagoon MLP at Poloniex and several sell side only exchanges

Sat Apr 23 21:00:01 UTC 2016
status of mOD dual side NuBot at Poloniex:
nud getliquidityinfo B | grep BFGMPykfKxXZ1otrCZcsbnTwJjKHPP9dsP -A 2
        "BFGMPykfKxXZ1otrCZcsbnTwJjKHPP9dsP" : {
            "buy" : 16623.67,
            "sell" : 15287.0549
status of zoro dual side NuBot at Poloniex:
nud getliquidityinfo B | grep BJs4YbtaqCmxeHLiR6zzjnZEotYVFAPfMo -A 2
        "BJs4YbtaqCmxeHLiR6zzjnZEotYVFAPfMo" : {
            "buy" : 9317.08,
            "sell" : 17591.3965
status of Cybnate dual side PyBot at Poloniex:
nud getliquiditydetails B | grep B954pkUEdkeT1G5Lq14Cisij5no3RVxHYe -A 20 | grep poloniex -A 2
        "1:NBTBTC:poloniex:LiquidBits" : {
            "buy" : 0.0,
            "sell" : 2000.0
status of NuLagoon dual side NuBot at Poloniex:
nud getliquiditydetails B | grep BTRnV9uLSPVJw4jn1JMV2Ki2cfFqPYip9o -A 1000 | grep poloniex -A 2 | tail -n 1  | awk -F':' ... and some more stuff
        "1:NBTBTC:poloniex:0.3.2a_1461382012298_fbecd4" : {
            "buy" : 11267.0875,
            "sell" : 3247.4015
--
        "2:NBTBTC:poloniex:0.3.2a_1461382012298_fbecd4" : {
            "buy" : 0.75,
            "sell" : 0.0
status of mOD single side NuBot at hitbtc:
nud getliquiditydetails B | grep BETwD8nSjtj9ADSvej2na34xmsMYwPRymv -A 200 | grep hitbtc | tail -n 1 | awk -F':' ...and some more stuff'
        "1:NBTBTC:hitbtc:0.4.1_1461097479835_167440" : {
            "buy" : 0.0,
            "sell" : 499.98
--
        "2:NBTBTC:hitbtc:0.4.1_1461097479835_167440" : {
            "buy" : 0.09,
            "sell" : 1499.936

Most obvious reason to me would be that some liquidity might be provided without using NuBot or PyBot. The client won’t register it, but Alix sees it in the exchange books.

If this is true then the difference is the real liquidity by the traders. We should use the ratio of the sell side of this difference and the buy side of the client data to calculate real sell pressure.

Here is the current result. It looks like the real buy/sell pressure (diff_buy/nud_sell and diff_sell/nud_buy) is rather small

Sun Apr 24 02:39:33 UTC 2016
Alix buy 62784.2554 Alix sell 43286.3168
 nud buy 63420.9684  nud sell 52499.5876
diff buy   -636.7130 Alix sell -9213.2708
diff_buy/nud_sell       -1.21% diff_sell/nud_buy     -14.53%
1 Like

The 1.5% cutoff doesnt matter for the reported section of the front panel though. The front panel should be a straight readout from the client. @willy

Upon more looks this isn’t that simple because often alix has less liquidity than the client, resulting a negative difference.

You might be right. We need to better understand what we are looking at on Alix. It fluctuates a lot which make it appear that it looks indeed at reported data from the client. That begs the question why the numbers are different from the T1 in the client.

Maybe when an ALP client reset the wall it pulls liquidity for a short moment and the exchange reports to alix that it is having less liquidity on the book. And at the same time the ALP server still reports the pulled liquidity, causing the nu client to still report it.

Reported liquidity comes straight from the client already.
My opinion on that is, that in a decentralized liquidity database, not everyone has the same data at the same time.

I checked my client values with those of ALix and those of ALix’ nud. All three were the same.
Maybe we all have the same set of nodes.

The liquidity reported is a 15 minute average value which, in my opinion, is necessary due to quite strong fluctuation of liquidity data.

1 Like

I wrote a script to sample nud liquidity every 10 sec and compare its 15min average with alix 15min average

Sun Apr 24 15:50:21 UTC 2016
Alix buy 57184.5561 Alix sell 68334.7440
nud  buy 61742.8000  nud sell 54697.8000
diff buy  -4558.2439 Alix sell 13636.9440

The difference is 10-20%. (in case you wonder, the buy and sell sides aren’t mistakenly swapped)

ALix script does it every minute.

How many connections does your client have?

The rising BTC rate continues the pressure on the buy side:

Liquidity; ALix T1 + gateways (poloniex dual side, numerous sell side):

Gateway status at Poloniex (dual side), NuLagoon MLP at Poloniex and several sell side only exchanges

Sun Apr 24 20:52:50 UTC 2016
status of mOD dual side NuBot at Poloniex:
nud getliquidityinfo B | grep BFGMPykfKxXZ1otrCZcsbnTwJjKHPP9dsP -A 2
        "BFGMPykfKxXZ1otrCZcsbnTwJjKHPP9dsP" : {
            "buy" : 16692.72,
            "sell" : 15287.0549
status of zoro dual side NuBot at Poloniex:
nud getliquidityinfo B | grep BJs4YbtaqCmxeHLiR6zzjnZEotYVFAPfMo -A 2
        "BJs4YbtaqCmxeHLiR6zzjnZEotYVFAPfMo" : {
            "buy" : 0.99,
            "sell" : 27146.1363
status of Cybnate dual side PyBot at Poloniex:
nud getliquiditydetails B | grep B954pkUEdkeT1G5Lq14Cisij5no3RVxHYe -A 20 | grep poloniex -A 2
        "1:NBTBTC:poloniex:LiquidBits" : {
            "buy" : 0.0,
            "sell" : 1600.0
status of Cybnate dual side PyBot at Bittrex:
nud getliquiditydetails B | grep B954pkUEdkeT1G5Lq14Cisij5no3RVxHYe -A 20 | grep bittrex -A 2
        "1:NBTBTC:bittrex:LiquidBits" : {
            "buy" : 1236.515,
            "sell" : 800.0
status of NuLagoon dual side NuBot at Poloniex:
nud getliquiditydetails B | grep BTRnV9uLSPVJw4jn1JMV2Ki2cfFqPYip9o -A 1000 | grep poloniex -A 2 | tail -n 1  | awk -F':' ... and some more stuff
        "1:NBTBTC:poloniex:0.3.2a_1461472134955_5dc4ed" : {
            "buy" : 10794.2259,
            "sell" : 4006.7679
--
        "2:NBTBTC:poloniex:0.3.2a_1461472134955_5dc4ed" : {
            "buy" : 15.71,
            "sell" : 0.0
status of huafei sell side only NuBot at bter
nud getliquidityinfo B | grep BRUuKfGur7CZSLy65gYUPJUQyQZeT1XQnD -A 2
        "BRUuKfGur7CZSLy65gYUPJUQyQZeT1XQnD" : {
            "buy" : 0.08,
            "sell" : 2025.0529
status of mOD single side NuBot at hitbtc:
nud getliquiditydetails B | grep BETwD8nSjtj9ADSvej2na34xmsMYwPRymv -A 1000 | grep hitbtc | tail -n 1 | awk -F':' ...and some more stuff'
        "1:NBTBTC:hitbtc:0.4.1_1461097479835_167440" : {
            "buy" : 0.0,
            "sell" : 499.98
--
        "2:NBTBTC:hitbtc:0.4.1_1461097479835_167440" : {
            "buy" : 0.09,
            "sell" : 1499.936

@zoro’s buy side is empty. Balancing necessary.

1 Like

8 connections. but i don’t think it matters. if it takes more than 15min to propagate liquidity, we have trouble with nu network.

My ALix nud has around 50 connections, so I suspect that the majority of liquidity providers are directly connected to it.
It might have a time plus. I’m not sure.

I had two nud’'s running. one in N America and one in Asia. They both had 8 connections and agreed with each other well.

I have three running in central Europe, also all agreeing with each other.

Can I get some advice on how to calculate, please? Do we have any neat way I can know how much is required where to balance everything?

I want to refill or balance zoro’s and Cybnate’s dual-side gateways at Poloniex. Only balancing should be my concern unless I want to increase the amount, right? How much do gateway operations cost/profit Nu in trades (excluding operator fees)?

Is there any limits or thresholds set as rules or guidelines for gateways about how much should be kept on them respectively?

Because my math skills are fleeting, I’ll publish my calculations for verification.

Balance zoro’s Poloniex gateway

(rounded)
Buy side: 0
Sell side: 27146

27146 / 2 = 13573
13573 USD = 29.31768 BTC

Send: 29.31 BTC
Receive: 13573 US-NBT

That was easy. What about when there is still some on both sides …

Practise

Buy side: 7533
Sell side: 17529

17529 - 7533 = 9996
9996 / 2 = 4998
4998 USD = 10.79568 BTC

Send: 10.79 BTC
Receive: 4998 US-NBT

Correct?


We only have ≈13.20 BTC in FLOT BTC Multisig #2.

I was being very pragmatic about it and heading to a 50/50 balancing of @zoro’s gateway account, because the bid side was lower than the ask side network wide.
I initiated a deposit of 29 BTC; 13.2 BTC from FLOT, 15.8 BTC from JordanLee:

@zoro already withdrew 13,500 NBT

I didn’t consider @Cybnate’s gateway. Shame on me.
Balancing is more straight-forward than increasing or lowering the amount, right.

That will require a bit of diligence to find out.
Basically you have to sum up all deposits/withdrawals and their respective USD value and compare it with the current account balance.
As @zoro’s and my gateway operate at a big offset, you can safely assume that they rather make Nu money when funds get traded there - unless the NuBot is stuck and the orders can’t be moved. In that case the gain will be reduced and worst case Nu makes s slight loss.
The only vector for ongoing loss in gateway operation/trading is the volatility risk combined with the hedging - as long as Poloniex doesn’t default or the funds get lost in another way.

As the gateways are an important line of defence that kicks in after ALP and NuLagoon failed, I recommend to keep at least $20k on each side. I can’t give you an algorithm how to calculate that number. It’s what my gut feeling tells me.
Do you remember what a pain the permanent balancing was when NuPool failed (LP pulled funds), NuLagoon didn’t have sufficient funds there and all that kept the peg was a pair of NuBots trading at a too small spread?
I speak of early this year.
The NuBots I operated had way more than $20k each side and it was still hard to prevent sides from running dry during the BTC swings.

I guess with an offset of above 1.5% and a funding of $20k each side, a bot operation on Poloniex is well suited to provide a viable line of defence.

Looks reasonable.

1 Like

I won’t beg for it :slight_smile: It is there anyway. But FLOT may consider adding some buy side to my gateway. Even though it is on 0.7% offset on either side I still made 200 NBT in the last 6 weeks or so. Bot is stable now. My Nu wallet issues have been resolved. Only the odd known PyBot glitch which resolves itself automatically within 15 minutes.

1 Like

What about $2,000 in BTC to your Poloniex account in exchange of 2,000 NBT?
The current liquidity situation network wide does rather have an overweight on buy side, but the gateways on Poloniex look poorly funded on the buy side:

Mon Apr 25 20:13:08 UTC 2016
###
status of the gateways in total:
bid total 1230 USD, bid side ratio 6%
ask total 17670 USD, ask side ratio 93%

This output includes the PyBot funds from Bittrex!

Bot status Poloniex only:

Mon Apr 25 20:23:32 UTC 2016
status of mOD dual side NuBot at Poloniex:
nud getliquidityinfo B | grep BFGMPykfKxXZ1otrCZcsbnTwJjKHPP9dsP -A 2
        "BFGMPykfKxXZ1otrCZcsbnTwJjKHPP9dsP" : {
            "buy" : 0.0,
            "sell" : 0.0
status of zoro dual side NuBot at Poloniex:
nud getliquidityinfo B | grep BJs4YbtaqCmxeHLiR6zzjnZEotYVFAPfMo -A 2
        "BJs4YbtaqCmxeHLiR6zzjnZEotYVFAPfMo" : {
            "buy" : 1.02,
            "sell" : 13646.1263
status of Cybnate dual side PyBot at Poloniex:
nud getliquiditydetails B | grep B954pkUEdkeT1G5Lq14Cisij5no3RVxHYe -A 20 | grep poloniex -A 2
        "1:NBTBTC:poloniex:LiquidBits" : {
            "buy" : 0.0,
            "sell" : 2000.0