[Passed] Grant to provide liquidity using NuBot on hitBTC - the begin of modPuddle

Liquidity status status of modPuddle

Wed Oct 14 17:36:09 UTC 2015
nud -unit=B getinfo | grep version -A 8
    "version" : "v2.0.1-unk-beta",
    "protocolversion" : 2000000,
    "walletversion" : 1,
    "walletunit" : "B",
    "balance" : 1.0,
    "newmint" : 0.0,
    "stake" : 0.0,
    "parked" : 0.0,
    "blocks" : 570389,
    "moneysupply" : 4625112.9069,
    "totalparked" : 2006.33,
nud getliquidityinfo B | grep BBZ4h88BwYCyE9q268LoArps6eodq9PDGH -A 2
        "BBZ4h88BwYCyE9q268LoArps6eodq9PDGH" : {
            "buy" : 0.11,
            "sell" : 1511.937

Balancing the funds initiated##

As the buy side is below 45% of total funds, balancing happens


17:49:55.528 INFO  - Price Updated. blockchain:1 BTC = 254.5 USD - [c.n.n.t.PriceMonitorTriggerTask:353]

750 NBT equivalent in BTC will be put into buy side.
Based on 254.5 NBT/BTC that equals to 2.947 BTC.

As the hitBTC sell side is much bigger than the hitBTC buy side, inter exchange balancing will be done (instead of trading the NBT into NuLagoon’s buy side).
Poloniex has at the time of balancing a significantly higher ask volume and will be used for the balancing.

Ask order target  35737.96 NBT filled of maximum 20000 NBT ( 178.69% )
Bid order target  14659.23 NBT filled of maximum 20000 NBT ( 73.3% )

Liquidity status status of modPuddle after balancing##

Wed Oct 14 18:10:09 UTC 2015
nud -unit=B getinfo | grep version -A 8
    "version" : "v2.0.1-unk-beta",
    "protocolversion" : 2000000,
    "walletversion" : 1,
    "walletunit" : "B",
    "balance" : 1.0,
    "newmint" : 0.0,
    "stake" : 0.0,
    "parked" : 0.0,
    "blocks" : 570433,
    "moneysupply" : 4625112.8869,
    "totalparked" : 2006.33,
nud getliquidityinfo B | grep BBZ4h88BwYCyE9q268LoArps6eodq9PDGH -A 2
        "BBZ4h88BwYCyE9q268LoArps6eodq9PDGH" : {
            "buy" : 751.69,
            "sell" : 750.0

You made money :smile:

1 Like

Between exchanging and transferring the BTC the price went up :wink:

But in fact we lost money.
The 3.044 BTC from the start are now only 2.947 BTC. The 3.044 BTC would be worth 774.698 NBT at the rate of 254.5 NBT/BTC.
23 NBT in the red on the second day.
The transfer fees are tiny, but the 0.2% fee at poloniex was paid from my own pocket, which is another 1.5 NBT.

I’m not too excited about the current balance, but about the NuBot operation in general :smile:

But fortunately I get a share of the (future) reward for the liquidity operation and the two other guys bear the risks :stuck_out_tongue:

Minus 1.5 NBT for me.
Minus 23 NBT for the fund owners.
Operating a liquidity pool is obviously the more reliable business^^

Shareholders don’t account in BTC, they account in NBT. If we account for the 1.5 NBT fee loss, you broke even. When the price starts going down or remains stable we’ll see how much you really lose. Who’s to say the fund owners wouldn’t have sold in this time and lost out on that 23 NBT? In my opinion, funds can only be lost or gained in the moment of rebalance, and only from the perspective of NBT worth. The result will be that when BTC goes down you’ll have to account for some hard losses.

1 Like

It’s day two, BTC is in the last days more volatile than before, there are some more days to go.
The final balance will be after day 30 and not earlier.

Only 8 days left of the current operation.

Time for a summary.

Technical part

The internet access worked well as far as I can tell.
The RaspberryPi was running fine.
The CPU load was only high during startup of NuBot, RAM consumption was moderate and nud and NuBot were running on the same RaPi.
The temperature of the RaPi remained below 40°C (in an environment of 20°C).

I had some issues with broadcasting liquidity once, and ran into another issue recently (which isn’t fixed yet).
The NuBot documentation might need some tweaks and quality management, but overall it was a great experience to work with NuBot and @desrever!

hitBTC was running reliably, the API was performing well.

That much about technical summary.
I dare say that running nud (2.0.1) and NuBot (0.3.2) on a RaspberryPi 2 works great with little effort and little energy costs (RaPi running at below 3 Watts)

Now the (preliminary) economical summary.

That assessment is correct - from a NSR holders’ persective or from the perspective of people who prefer a stable crypto currency over volatile ones.
If it’s non-shareholders providing liquidity they might account in BTC - especially if you want to attract liquidity on the buy side!

The current liquidity situation of modPuddle is
750 NBT, 2.4 BTC

The value in NBT is still above 1,500 NBT (BTC rate at the time of writing 390 NBT):750 NBT + 2.4 BTC * 390 NBT/BTC = 1,686 NBT

If you calculate that in BTC, the funds would be worth roughly 4.32 BTC (1,686 NBT / 390 NBT/BTC = 4.32 BTC) at the moment.
Even if the complete compensation of 225 NBT would be distributed to the fund providers (which isn’t the case as I get some of the compensation for my efforts), they would end up with less BTC than before (225 NBT / 390 NBT/BTC = 0.58 BTC -> 4.32 BTC + 0.58 BTC = 4.9 BTC).

The fund providers already indicated that they are not going to provide this operation with funds for another round.
And I understand their reasoning.
They started with close to 6 BTC, half of which were converted to NBT.
From their perspective they have lost more than 1.1 BTC or above 400 USD in this month.

Considering that this operation was at a relatively high compensation rate (due to the uncertain exchange risks), I need to come to the conclusion that this operation was a total failure for the liquidity providers.
The compensation was not sufficient to provide the liquidity providers with an incentive to contine.
The only perspon who earned money from it is me, 25 NBT.
The lesson I learned is more valuable than the 25 NBT and the money was not the reason to do this - obviously.

I consider pulling funds from ALP to continue the operation as I have long ago started to calculate in NBT. I have not yet come to a conclusion with how many funds I can continue the operation and what terms I’m going to offer.

Is there demand for modPuddle on hitBTC or is Nu comfortable with having NuLagoon provide liquidity there?

If they sold half the BTC immediately, then they really came in with <3 BTC and 750 NBT. They have lost ~0.5 BTC. In return they are given 225 NBT, or ~0.5 BTC at current prices. From my perspective, they came out about even and are just sore that they did that initial sell of half their funds to get to a balanced pool. That sell technically occurred before operations began, and should not be included in the summary.

Unfortunately it’s not up to me or you to tell them how they need to account that.

I’ve heard complaints about the funds being worth less (BTC) than when the operation started. They knew about that risk and they took it. They didn’t ask me to terminate the operation (of course they kew about the terms), but indicated they don’t want to provide liquidity after the 30 days.

If they had kept the ~6 BTC, they still had 6 BTC worth 2,340 USD.
I can’t deny that this is a valid perspective.

If the terms of the operation hadn’t stated very strictly that the operation continues no matter what happens, they might have pulled the funds days ago.

Imagine a hypothetical currency that does nothing but increase in price at a fast rate. Imagine Nu started pegging it. Nu would have to pay buy side custodians a higher rate than the increasing price to make it worth their while if they account from the perspective of that currency.

Sounds to me like they are accounting in nbt as btc falls and btc as btc rises. This will always look like a loss from all perspectives. Putting all your eggs in one basket gives more profit in hindsight, sure, but also vastly more risk in practice. No offense to your friends, but good riddence to bad economics.

1 Like

You don’t have to convince me.
As I indicated before I consider continuing the operation with funds currently being in ALPs.
I know about the risks and don’t account in BTC.

I’m happy with getting a compensation for my NBT that would otherwise sit idle in my wallet.

But as it happens those who need to be attracted to stuff BTC in the buy sides might account in BTC - why else should they have them instead of NBT?

Nu shouldn’t ignore the fact that a lot of people are accounting differently and maybe even irrational, for it will not prevent them from making strange choices.
Nu needs to be prepared for that or do you expect that in times of stress the buy side is provided by NSR holders? With what funds?
I see a potential danger for the liquidity of tiers 1 to 3, because if the shit hits the fan, only tier 4 and 6 are available to provide the lower tiers with BTC (in one way or another).

I very much hope that we have a bunch of fixed cost ALPs soon as they are well-suited to address that by offering a variable compensation - the more compensation, the less funds are on tier 1.

I guess it’s the way that the btc was stuffed into the pool. As we’ve discussed, it was that initial sell to balance the pool that caused them so much sorrow. If instead they had provided 100% buy side and Nu had matched them with 1,500 nbt and simply forced you to burn 1,500 at the end of the month, I think the numbers would very much have worked out the way I stated.

The issue with actually doing that, of course, is that Nu has no way of forcing you to burn nbt. If btc drops in value precipitously, it is possible that the total funds in the pool could drop below 50% of the original value, meaning the participants would have to fork over more money from pocket. Without any way to make that happen, a default is probable.

You are right - it might have worked better starting with 100% on buy side…

i don’t know about you but my sell side as an ALP custodian is decreasing
although BTC’s value is increasing!
not only i am not loosing money but i have a high profit! a very strange behavior i would say!
edit: it seems that the reason is the high BTC volatility :slight_smile:

1 Like

You don’t have to convince me.

Nor do I think we have to convince those who account in and only in BTC.

Liquidity provision for Nubits is a special kind of investment that is part BTC and part NBT; you earn NBT interests but at the same time you “risk” holding a BTC position of less than 100% which causes you to earn less during a BTC surge. Plenty of people would be content of this kind of investment; it’s worse than being able to catch a falling knife but much better than catching it on the blade. We just have to make it clearer that all our interests do are to insure against losses of NBT value.

By the way, if the BTC/NBT spreads do follow 0ec0be7f113a0bf6ff603545a974cd6410458e00 then liquidity providers have earned several thousand USD over the past week in spreads, more than 1000 USD just in the past 24 hours (>220k volume). Enough to cover a large part of LP interests.

1 Like

While I totally agree, they also probably lost more than that in the volatility. The real gain is something like (spread - volatility)*volume, but the ‘volatility’ term is super hard to quantify.

1 Like

Tomorrow at 16:00 UTC the 30 days are over.

The fund providers consider that a lesson learned and a big loss (because the calculate in BTC)
They started with roughly 6 BTC worth 1,500 USD.
Current status is 3.1 BTC (at currently 340 USD/BTC worth 1,054 USD) and 292 NBT.
That makes a total of $1,346. Together with the 200 NBT (their share from the grant for which I’m going to propose a draft soon) they are at $1,546.
They consider it a loss, because they are down from 6 BTC to 4.54 BTC.

I consider that a gain, because I calculate in NBT. If I add up the 25 NBT I asked from them for my service it’s a gain of 71 NBT for 30 days or 4.5% per 30 days.

If there’s interest for such a small liquidity provision (1,500 NBT total) in the light of NuLagoon being at hitBTC as well, I would continue the operation and just pay off my buddies with funds from my ALP bots.

From a technical perspective it went very well on my RaspberryPi2 and my private internet access.
The hitBTC worked very well as far as I can tell
The only outage I recognized (and only for some hours) was due to trouble with the streaming server.

The fees I proposed for this grant were quite high and I don’t intend to lower them for the next term.
I offer less liquidity than NuLagoon and likely at a higher compensation rate.
But I can promise to keep up the walls no matter what (except for technical issues, but not for economical issues).
If customers of NuLagoon do a “bank run” at it might be different.
And with 2 liquidity providers and 2 independent NuBots at the same exchange it feels more decentralized :wink:
I’m not sure whether that’s worth it for Nu, though. But I wouldn’t want to do it for less compensation with my own money.

1 Like

I d support the continuation of mod puddle under the same conditions.

1 Like

I hope I had paid more attention to the wall movement on hitBTC. There’s a lot to learn here.

On NuLagoon where the actual interest rate is somewhere like 7% for Pool A, we observed about 20% gain in NAV over 6 months (early April - early October) where the BTC price at the end points were similar. About a 3-percent gain per month. The actual cost of volatility (and possibly exchange fees) was then something like 4% per month.

Similarly, 4 months from the start of Pool D (late May) the BTC prices were about the same at the end points, and the NAV went from 1.087 to 1.276. About a 4% increase per month. The actual interest rate of NuLagoon was between 11% to 13% depending on the interest rate of Pool C, so the volatility costs were roughly 7-9% per month.

Over your 30 days, the cost of volatility was apparently over 10%. Perhaps a lot of that was due to the recent wild swings and a large exposure to BTC.

Dear NSR holders,
the 30 days of modPuddle term 1 are over.

The funds at the end are
3.1 BTC * 312 USD/BTC = 967,2 USD

  • 292 USD
    = 1259 USD

The liquidity pool is 241 USD in the red. 225 NBT from the grant which will be drafted soon (for the back pay of the 30 days) will compensate most of the loss.

I pay off the participants who don’t want to continue, let NuBot run and propose a new draft for the next term soon.
I will move 241 NBT to hitBTC to fill the liquidity volume to 1,500 USD again.


modPuddle is currently out of operation.

The device on which it was operated on (RaPi2), was urgently required to operate the NBT exit and the NBT entry gateway simultaneously. The current BTC price swings drained first the sell side, then the buy side. This might not be over.

RAM issues with nud 2.0.3 prevent operating two NuBots on the same RaPi2.
As I have no 2.1.0 blockchain synced yet, I can’t tackle the RAM issues.

I decided that supporting the peg on Poloniex is more important than on hitBTC at the moment.