Bolstering DEX Liquidity (previously my rant/opinion on Bitshares' DEX)

[Xypher]

An alternative, extremely ambitious approach
https://bitsharestalk.org/index.php/topic,20855.0.html

[Nagalim]

Not sure where you got the 7% from, fixed cost does not specify a rate but rather a cost ($/day, not %/day).
Edit:  this was disingenuous of me.  The 7% is correct, but only up to a specified limit at which competition drives down rates which makes the economics fundamentally different from saying a blanket 7%.  Fixed cost is the new model and will be adopted with the NuBot ALP upgrades.

We announce each week how many nbt are in circulation.  It's ~700k.  Coinmarketcap has a hard time accurately giving stats to Nu because we're doing things that haven't been done before.
https://discuss.nubits.com/t/passed-motion-to-begin-nsr-buyback-immediately/2654/120

You are entitled to your opinion.  A lot of people think USD is unsustainable, yet here we are talking about pegging to it as a standard.  All Nu is democratic contracts enforced by a decentralized central banking system.  I'm sorry the success of literally 4000 years of economics to develop the concept of a banking system offends you so.

[chryspano]

In my biased opinion nubits are unsustainable and we should stay away from them, it is easy for them to "print" out of thin air the 7% that is required to pay their market makers but who is actually paying this fee? to me it's like a blockchain loan they can never pay back.

One question...How many nubits exist? coinmarketcap says 840,000 where did they got this number? the explorer lists 27,700,000 https://blockexplorer.nu/status  and 100 wealthiest addresses show something in between https://blockexplorer.nu/topNBTaddresses/1

[CoinHoarder]

I think you're still a little confused.  The bitusd will be held in a multisig reserve as a tier 4 fund, it will not be spent unless the nubit peg is in serious imminent danger and we've spent all our btc (you know, depending on how we actually end up specifying the terms of the bitusd reserve.  It'll be a zoology reserve, again for anyone keeping track).

You are right, my bad I still do not fully grasp how Nushares works and the terminology.

The providers will be paid in nbt.  For every nbt we give providers, this proposal would imply that the bitasset community owes Nushareholders 0.5 bitusd per nbt used.  These can be created using bts y'all print or whatever, it's not really my concern where you get your money from.  I'd suggest you print some fresh bts and lock it to make bitusd then give us the bitusd to Nu under the assumption that they probably won't even spend it.  Then you basically get free liquidity.  But again, not my concern.

I'd suggest we start with something like $1/day until we get a real handle on the price feed.  Of course, there's going to be plenty of beaurocratic hurdles to overcome to get a proper contract written and passed on the Nu block chain, then there's the matter of the ALP bot upgrade not being ready yet.  So this whole concept may take time (months) to come to fruition, but I truly believe it would be a lucrative endeavor.

I agree that it sounds like it would be a win/win agreement. I wonder if there is enough support in the Bitshares and Nushares community to make it happen. I guess I will let others here voice their support or concern so we can judge the Bitshares community's opinion..

[Nagalim]

I think you're still a little confused.  The bitusd will be held in a multisig reserve as a tier 4 fund, it will not be spent unless the nubit peg is in serious imminent danger and we've spent all our btc (you know, depending on how we actually end up specifying the terms of the bitusd reserve.  It'll be a zoology reserve, again for anyone keeping track).

The providers will be paid in nbt.  For every nbt we give providers, this proposal would imply that the bitasset community owes Nushareholders 0.5 bitusd per nbt used.  These can be created using bts y'all print or whatever, it's not really my concern where you get your money from.  I'd suggest you print some fresh bts and lock it to make bitusd then give us the bitusd to Nu under the assumption that they probably won't even spend it.  Then you basically get free liquidity.  But again, not my concern.

I'd suggest we start with something like $1/day until we get a real handle on the price feed.  Of course, there's going to be plenty of beaurocratic hurdles to overcome to get a proper contract written and passed on the Nu block chain, then there's the matter of the ALP bot upgrade not being ready yet.  So this whole concept may take time (months) to come to fruition, but I truly believe it would be a lucrative endeavor.

[CoinHoarder]

"Isn't that how a "liquidity pool" works in Nushares... the people fronting the Nubits make interest on their deposits, right? I understand liquidity providers take a risk, but so is the person fronting the bitUSD..."

I'm not sure I fully understand the question here, so I'll go ahead and clarify by talking about Nubits only and leave bitusd out of it.

So there are two parties with nbt: the operator and the provider.  The operator is granted funds by shareholders, and so must be trusted and contracted properly to give out the funds fairly to providers.  The providers then put nbt (and btc) up as market orders on their own account.  They are always in control of these funds, but as long as they prove the market orders are theirs by providing API info, the operator credits the provider and gives out some of the nbt granted by shareholders.

The end result is that we can get large amounts of funds (thousands of $$) by only rewarding a small, continual payout (single digit $/day).  So the shareholders take the risk that the liquidity provision will make the network more valuable than the cost for liquidity while the providers take on all default and volatility risks and get rewarded for it.  Everybody's happy, we make a contract, and the whole system becomes reliable and dependable for customers.

Thank you, I think I understand it better now. I understand now that we are paying for liquidity (at most $9 a day?) and not investing in the market making services themselves. Eventually the bitUSD funds will all be spent on Liquidity costs because there is a daily fee, so the bitUSD that will be provided upfront is pretty much a donation to the network to afford us more liquidity, correct?

The Bitshares shareholders would need to decide if having more bitUSD liquidity is worth the expense, and how would we fund the initiative? I can think of several different ways. As JonnyBitcoin mentioned we can use the reserve pool, we could create a worker proposal, or simply run the operations on donations (althoguh I'm not sure we would be able to raise enough via this method.)

[Nagalim]

"Isn't that how a "liquidity pool" works in Nushares... the people fronting the Nubits make interest on their deposits, right? I understand liquidity providers take a risk, but so is the person fronting the bitUSD..."

I'm not sure I fully understand the question here, so I'll go ahead and clarify by talking about Nubits only and leave bitusd out of it.

So there are two parties with nbt: the operator and the provider.  The operator is granted funds by shareholders, and so must be trusted and contracted properly to give out the funds fairly to providers.  The providers then put nbt (and btc) up as market orders on their own account.  They are always in control of these funds, but as long as they prove the market orders are theirs by providing API info, the operator credits the provider and gives out some of the nbt granted by shareholders.

The end result is that we can get large amounts of funds (thousands of $$) by only rewarding a small, continual payout (single digit $/day).  So the shareholders take the risk that the liquidity provision will make the network more valuable than the cost for liquidity while the providers take on all default and volatility risks and get rewarded for it.  Everybody's happy, we make a contract, and the whole system becomes reliable and dependable for customers.

[Empirical1.2]

So let's do a thought experiment:
BitAssets pays $5,000 to one liquidity provider.  For some reason we trust that provider with our funds (like a T3 trusted custodial grant for anyone keeping track) but let's ignore that for now.  So the provider puts up the funds as a sell order and they get bought.  Now the provider needs to do the arbitrage shuffle to get the funds back on exchange.  In the meantime, there is no liquidity.  The provider gets tired of doing this all the time and starts charging a premium.  This generates competition, but we're still just picking the best offer out of the ones on the table and almost invariably have to pay premium.  Then the exchange defaults and we lose all our money.  Welcome to the first 6 months of Nu.

We then developed automatic liquidity provision such that we can decentralize this process and lay the exchange default risks on the providers shoulders instead of the network.  Anyway, long story short, no matter what the costs for liquidity provision exist.  You say providers profit, but they take on risk and opportunity cost and their profit is well earned.  The network pays a little to get functionality, which brings in new money.  Economies are not a zero sum game, sometimes you gotta spend a little to grow the network.

I still don't get why there is no obvious incentive for the person(s) fronting the bitUSD (other than the value to the BTS token from the added liquidity.) Isn't that how a "liquidity pool" works in Nushares... the people fronting the Nubits make interest on their deposits, right? I understand liquidity providers take a risk, but so is the person fronting the bitUSD...

I just realize you run a liquidity pool for Nubits.. http://nupond.net/ ... right? I guess you know a lot more about this than I do, so maybe I am still misunderstanding something. Is the only benefit of people depositing into liquidity pools the added liquidity, and the pool pockets all interest? I am interested in continuing this conversation because liquidity is a major problem with Smartcoins and the Bitshares DEX.

I don't know but as a guess, I would say I think liquidity increases demand for NuBits, so  NuShares holders are incentivised to create NuBits to pay for these liquidity services. As NuBits can just be created without backing, shareholders only have to balance the costs with the risks that at some point more NuBits will be sold in a short space of time than there is demand and reserves for and thus break their peg.

[Empirical1.2]

I was also not a fan of NuBits and think it will at some time come unstuck but the volume is pretty good and I was impressed that they paid out over $400 000 in dividends to shareholders over the last year representing a 25% yield to NuShares shareholders.
[...]
Do any people who follow NuBits more know how much the market makers need to be subsidized?

So far the track record of Nu is ok for the first year. February 2015 and the months after that were bad because Nu lost a lot of money due to exchange defaults. Back then the liquidity providing was done with funds that Nu owned and handed over to custodians to provide the market with liquidity.

This model has been reworked and now Nu pays for market makers who provide liquidity with their own funds.
The average monthly revenue for market makers is approximately 7%. That seems big, but apart from the exchange default risk it needs to compensate BTC volatility (the main trading pair is still NBT/BTC, although USD/NBT and CNY/NBT are available as well).
These 7% are paid by Nu up to a maximum volume (some tens of thousands USD value each side all exchanges combined, Poloniex has the biggest volume: https://alix.coinerella.com/walls/?).
A dutch auction model kicks in if the liquidity volume is above that threshold and reduces the effective interest.

This is speaking of the so-called ALP (automated liquidity pool) where a custodian operates a server software and liquidity providers provide funds with ALP clients. They stay in full control over the funds as the ALP client puts orders via exchange API and reports them to the ALP server, which credits them each minute. The money never leaves the exchange account of the liquidity provider (unless stolen, etc.; it might just get converted of the orders get filled)

A second way to provide liquidity is via MLP (managed liquidity pool). In this version a liquidity provider uses NuBot to place orders. The funds are under direct control of the NuBot operator.

For more information on liquidity providing have a look here: http://docs.nubits.com/liquidity-pools/

There are a lot of changes on the road map.
The ALP and MLP software are currently being merged and will in the future be based on NuBot (https://bitbucket.org/JordanLeePeershares/nubottrading/src/master/docs/SETUP.md).
The reward scheme will be changed to a fixed compensation scheme where x NBT are paid per side and liquidity providers fight over the compensation. That is expected to have some advantages over the dutch auction model as it makes providing liquidity especially attractive (in terms of interest) if the order sizes on the book are small for whatever reason.
This is expected to help the money flow between the different tiers (of Nu's tiered liquidity model), different exchanges, because it incentivizes tracking wall sizes to put orders preferred at exchanges with low order volume.
The motion regarding the tiered liquidity model: https://discuss.nubits.com/t/finalized-evolution-of-liquidity-operations/618
An interpretation of it: https://discuss.nubits.com/t/interpretation-of-the-liquidity-tiers-a-waterfall-model-triggers-metrics-and-actions/2914
The begining of the ALP: https://discuss.nubits.com/t/trust-less-liquidity-pool/1686 (back then called "trustless liquidity pool")

This might sound confusing, has a lot of links and I bet I used some words that are not really self-explanatory. Sorry for that. But the liquidity providing is a quite complex area if you are interested in the inner workings and one of Nu's important functions.
If you just want to make some money providing liquidity, it can be as easy as sending funds to the major MLP "Nulagoon" (http://nulagoon.com/lqpools.html) or downloading and configuring an ALP client.
If you have questions that are not answered here (because I think not too many from the Nu community are frequently here), feel free to ask at https://discuss.nubits.com/!

  Thanks very much for taking the time to give such a detailed response.

As you say it is a bit confusing to me at first glance but I'll try wrap my head around it, I'm sure guys like BM & other forum members will understand it better & find the info very useful.

What do you think of Market Maker Incentivization Worker Proposal? Could that provide useful incentives for liquidity?

I don't believe bitusd will ever have liquidity on real exchanges if all market making happens on the virtual exchange.  What I propose is fundamentally different from all the market maker approaches this community has put up because it will create real liquidity instead of just more inbred trading on the virtual exchange amongst bitasset holders.  Bitasset holders are not your target group, merchants and general adoption is.  You want people using bitusd that don't even really know what bitassets are.

Yes, make a bunch of smartcoins.  Then give them to Nu and let us provide liquidity for you.

I think I probably agree that unless we can incentivise market making on popular exchanges, BitAssets will struggle to gain traction.

[CoinHoarder]

So let's do a thought experiment:
BitAssets pays $5,000 to one liquidity provider.  For some reason we trust that provider with our funds (like a T3 trusted custodial grant for anyone keeping track) but let's ignore that for now.  So the provider puts up the funds as a sell order and they get bought.  Now the provider needs to do the arbitrage shuffle to get the funds back on exchange.  In the meantime, there is no liquidity.  The provider gets tired of doing this all the time and starts charging a premium.  This generates competition, but we're still just picking the best offer out of the ones on the table and almost invariably have to pay premium.  Then the exchange defaults and we lose all our money.  Welcome to the first 6 months of Nu.

We then developed automatic liquidity provision such that we can decentralize this process and lay the exchange default risks on the providers shoulders instead of the network.  Anyway, long story short, no matter what the costs for liquidity provision exist.  You say providers profit, but they take on risk and opportunity cost and their profit is well earned.  The network pays a little to get functionality, which brings in new money.  Economies are not a zero sum game, sometimes you gotta spend a little to grow the network.

I still don't get why there is no obvious incentive for the person(s) fronting the bitUSD (other than the value to the BTS token from the added liquidity.) Isn't that how a "liquidity pool" works in Nushares... the people fronting the Nubits make interest on their deposits, right? I understand liquidity providers take a risk, but so is the person fronting the bitUSD...

I just realize you run a liquidity pool for Nubits.. http://nupond.net/ ... right? I guess you know a lot more about this than I do, so maybe I am still misunderstanding something. Is the only benefit of people depositing into liquidity pools the added liquidity and the pool pockets all interest? Perhaps that is my misunderstanding.. I am interested in continuing this conversation because liquidity is a major problem with Smartcoins and the Bitshares DEX.

[CoinHoarder]

I'm going to change the name of the thread to DEX liquidity discussions as that is the way the conversation has shifted.

What do you think of Market Maker Incentivization Worker Proposal? Could that provide useful incentives for liquidity?

Yes, I think it would certainly help. However, I think we need to pursue all possible avenues of increasing liquidity because that is the main issue with the DEX right now.

we need to use the reserve pool as collateral to create a bunch of smartcoin to help liquidity.

I like this idea, we would just need to garner shareholder support. What do you think the pros and cons would be of using this money in the way you proposed in your thread versus using the money in the way Nagalim is proposing? I am thinking Nagalim's proposal provides more liquidity, but there is more risk involved with his proposal than with your proposal.

(reference: Jonny's thread: https://bitsharestalk.org/index.php/topic,20836.0.html)

[Nagalim]

So let's do a thought experiment:
BitAssets pays $5,000 to one liquidity provider.  For some reason we trust that provider with our funds (like a T3 trusted custodial grant for anyone keeping track) but let's ignore that for now.  So the provider puts up the funds as a sell order and they get bought.  Now the provider needs to do the arbitrage shuffle to get the funds back on exchange.  In the meantime, there is no liquidity.  The provider gets tired of doing this all the time and starts charging a premium.  This generates competition, but we're still just picking the best offer out of the ones on the table and almost invariably have to pay premium.  Then the exchange defaults and we lose all our money.  Welcome to the first 6 months of Nu.

We then developed automatic liquidity provision such that we can decentralize this process and lay the exchange default risks on the providers shoulders instead of the network.  Anyway, long story short, no matter what the costs for liquidity provision exist.  You say providers profit, but they take on risk and opportunity cost and their profit is well earned.  The network pays a little to get functionality, which brings in new money.  Economies are not a zero sum game, sometimes you gotta spend a little to grow the network.

[CoinHoarder]

You say it's risky for the bts holders fronting the bitUSD and say nothing about the shareholders fronting nbt.

True, my bad... whomever is fronting the Nubits also has risk.

You need to understand that liquidity costs money.  You try to 'solve' the cost of liquidity by widening the peg and you end up right back where you started.  If your network can't put up a couple dollars a day for a working product, what good is you whatever a million marketcap?

Ya'll will learn sooner or later that liquidity provision is about opportunity cost and you need to reward liquidity providers with a portion of your marketcap if you want a tight peg.

So.. in your proposed deal only the liquidity provider makes a profit, is that correct? In that case I'm having trouble understanding why we couldn't fund the purchase of the NBT and the bitUSD and be the liquidity provider ourselves. It would help negate the risks of fronting the money and provide incentive for people to do so. That may be the only way that this would be feasible... unless the money is raised by a worker proposal, or by the fees that are already locked on the chain (those options would of course require everyone to be on board with it).

I feel like I must still be missing something in the deal you are proposing?

[Nagalim]

I don't think shareholders want to store TUSD in T4 reserves because of the centralization.  A service like NuVault is much better than something like tether or tusd because when things do hit the fan and the centralized third party fails, at least with NuVault we have the BKS collateral to sell.
Without shareholder support, this is dead in the water.  It basically would devolve into privately funded liquidity operations, getting all the costs without any of the flair that Nu has.  The software is open source, anyone can write a wrapper and run it.  However, at that point using a US-NBT/TUSD market is unnecessary and you might as well just do TUSD/BTC and reinvent tusd as a centralized nbt knockoff.

[Samupaha]

What do you think of Market Maker Incentivization Worker Proposal? Could that provide useful incentives for liquidity?

I don't believe bitusd will ever have liquidity on real exchanges if all market making happens on the virtual exchange.  What I propose is fundamentally different from all the market maker approaches this community has put up because it will create real liquidity instead of just more inbred trading on the virtual exchange amongst bitasset holders.  Bitasset holders are not your target group, merchants and general adoption is.  You want people using bitusd that don't even really know what bitassets are.

Yes, make a bunch of smartcoins.  Then give them to Nu and let us provide liquidity for you.

Yeah, I get what you mean. I was mostly interested if the incentive program will be any good or just useless. Seems to me a good way to get things started and then provide a reasonable incentive to grow the markets bigger to get full benefit of potential profits.

Don't know if you already noticed but we have now also the TUSD, which is a privatized market pegged asset. Maybe @bitcrab is interested in cooperation to provide TUSD/NuUSD markets?