Bancor: Smartcoins With Infinity Liquidity

[saburgeon]

The concept behind Bancor is so simple. Couldn't something like this easily be replicated by having bitshares as the reserves and building a token on top of it?

Yeah, but I think most BTS shareholders are apprehensive of diluting BTS tokens for new features based on past experiences.

Me personally... I say we keep development moving and implementing new features, otherwise we will be left behind... as far as blockchain tech goes.

I agree we should move forward, because if this is successful, then exchanges will become obsolete. I think Bitshares is the better platform to implement something like Bancor since Ethereum has scaling issues and increasing gas costs. The great thing about Bitshares is that a protocol like Bancor is easily within their capacity, so adjustments can be made if needed, but I want to see how Bancor turns out first.

[yvv]

The concept behind Bancor is so simple. Couldn't something like this easily be replicated by having bitshares as the reserves and building a token on top of it?

This would be a very good PR for bitshares, if it could show to all those "investors" that you don't actually need $150M ICO to implement something like this.

[CoinHoarder]

The concept behind Bancor is so simple. Couldn't something like this easily be replicated by having bitshares as the reserves and building a token on top of it?

Yeah, but I think most BTS shareholders are apprehensive of diluting BTS tokens for new features based on past experiences.

Me personally... I say we keep development moving and implementing new features, otherwise we will be left behind... as far as blockchain tech goes.

[saburgeon]

The concept behind Bancor is so simple. Couldn't something like this easily be replicated by having bitshares as the reserves and building a token on top of it?

[david johnson]

Bancor Network is a protocol that is based on built-in price discovery and mechanism of liquidity for cryptocurrencies on smart blockchains. In a more basic way explaining, it is primarily a platform for smart token creation and exchange of those tokens. Tokens created in the network are backed by Bancor’s own token called BNT. The BNT reserve supplies a strong liquidity for trading the tokens. Next to smart tokens, it is also possible to trade other Ethereum tokens within the platform using smart contracts.
(http://blog.bancorprice.org/bancor-network-an-introduction/)

[pc]

Bancor will be going live in about 2 hours: https://blog.bancor.network/bancor-network-token-bnt-activation-e1766b62a73d

[yvv]

Bit20 does pretty much the same job as shapeshift's prism. May be we need a bit easier way to create such tokens?
 

[CoinHoarder]

The main showstopper I see with Prism is that you need a counterparty. The blockchain itself should not act as a counterparty to speculators, IMO. This means that creating a Prism is not sufficient, you must also find someone as your counterparty. The latter is what makes Prism much more complicated than Bancor, IMO.

I had imagined that Bitshares users would be the counterparty on Prisms, similar to the way they are the counterparty for Smartcoins.

Prisms would be traded on a separate tab/page on the DEX than Smartcoins.

I do agree that the necessity for a counterparty is a negative though.

[pc]

Bancor:
High liquidity
More complicated economics
More complicated to implement into Bitshares

Hm, maybe it's because the Bancor whitepaper is more detailed than what I could find about Prism, but I think Bancor is much simpler than Prism.

The main showstopper I see with Prism is that you need a counterparty. The blockchain itself should not act as a counterparty to speculators, IMO. This means that creating a Prism is not sufficient, you must also find someone as your counterparty. The latter is what makes Prism much more complicated than Bancor, IMO.

[CoinHoarder]

Perhaps we should implement it after all, just to prove our superiority... https://twitter.com/joeykrug/status/874450730182000641

I had forgot about Bancor Smart Tokens the other day when I was discussing adding Prisms to Bitshares. If Bitshares were to add a feature, then I think this or something more along the line of Prisms (https://bitsharestalk.org/index.php/topic,24271.0.html) would be a good idea.

After learning how Bancor works, I have realized that Smartcoins still have their place, as you cannot create a Prism/Bancor Smart Token with Gold, Silver, USD, GOOG, APPL, (etc...) in it without using something like a Bitshares smartcoin to go in the Prism/Bancor Smart Token portfolio.

Copying either Bancor Smart Tokens or Prisms on the Bitshares blockchain would each have their own separate pros and cons:

Bancor:
High liquidity
More complicated economics
More complicated to implement into Bitshares

Prisms:
Low liquidity
Less complicated economics
Less complicated to implement into Bitshares

IMO, I think the liquidity provided by the Bancor Smart Token model be worth the higher development costs, because I think that liquidity is still the main thing holding Bitshares back. Implementing Prisms on Bitshares will probably take another 3 to 5 to 10 years before we see them liquid enough to be very useful. We would have to wait for the Prism's market liquidity to bootstrap itself slowly over time, similar to how we have had to wait a long time for decent Smartcoin liquidity.

Bancor on the other hand has liquidity baked in from the get go. This would bolster liquidity on the DEX. You could create one asset Bancor Smart Tokens that only contain one Smartcoin in its portfolio, such as bitUSD, and let the free market decide which market pegged asset design they like more. Then there's the added benefit of easily creating multi-asset  Bancor Smart Token portfolios too.

It seems like those knowledgeable with Ethereum think Bancor will not work on Ethereum due to scaling issues, so I think it would be really good marketing if the same model were made to work on Bitshares. Along with all of the added benefits of adding the feature. I am normally against dilution for development/advertising, but I think such a feature would be a big boost to the value of the Bitshares token and its DEX.

[yellowecho]

Perhaps we should implement it after all, just to prove our superiority... https://twitter.com/joeykrug/status/874450730182000641

lol!

[pc]

Perhaps we should implement it after all, just to prove our superiority... https://twitter.com/joeykrug/status/874450730182000641

[yvv]

This is crazy. They just started their crowd sale and they already raised +$80M worth of ETH

https://etherscan.io/address/0x5894110995b8c8401bd38262ba0c8ee41d4e4658

Bitshares community should learn how to promote our product. The secret of successful marketing is: very low fractional reserve, no interest, no dividends, not even scheduled buy back. The basic message should be: give us your money and fuck off. Look people, this works

[Permie]

Opposite is also true. It should not be difficult to make a better bitUSD in ethereum, it is surprising that nobody did it yet.

Disagree, oracles are still very shitty on ethereum.

Would you mind expanding on why BitShares is superior to ethereum in this case?
The more information like this that is public on the forum the better, imo
Cheers

[fav]

Opposite is also true. It should not be difficult to make a better bitUSD in ethereum, it is surprising that nobody did it yet.

Disagree, oracles are still very shitty on ethereum.

[yvv]

Interesting... but I don't quite get it.

The first example in the whitepaper creates a BANCOR asset with an initial supply of 300k BANCOR backed by 60k ETHER, with a CRR of 20%. So the initial BANCOR price is 1 BANCOR = 1 ETHER. Anyone can buy or sell BANCOR at any time for that price - which moves the price in the right direction. Ok, so the BANCOR/ETHER market is always liquid, which is nice.

But what is the use case for BANCOR? Why should anyone buy 1 BANCOR for 1 ETH if BANCOR is only backed by 0.2 ETH? Where is the value proposition?

BANCOR is a crowd funding token, something like OBITS. In contrast to OBITS, you'll be able to sell back and burn BANCOR any time you want, and you don't need exchange for this.

Apart from this it looks like it should be quite simple to implement that mechanism in BitShares - any takers? I bet abit already has it in his test network. ;-)

Opposite is also true. It should not be difficult to make a better bitUSD in ethereum, it is surprising that nobody did it yet.

[Permie]

Apart from this it looks like it should be quite simple to implement that mechanism in BitShares - any takers? I bet abit already has it in his test network. ;-)

Does abit do lots of coding for BitShares?
I'm unaware

[pc]

Interesting... but I don't quite get it.

The first example in the whitepaper creates a BANCOR asset with an initial supply of 300k BANCOR backed by 60k ETHER, with a CRR of 20%. So the initial BANCOR price is 1 BANCOR = 1 ETHER. Anyone can buy or sell BANCOR at any time for that price - which moves the price in the right direction. Ok, so the BANCOR/ETHER market is always liquid, which is nice.

But what is the use case for BANCOR? Why should anyone buy 1 BANCOR for 1 ETH if BANCOR is only backed by 0.2 ETH? Where is the value proposition?

Apart from this it looks like it should be quite simple to implement that mechanism in BitShares - any takers? I bet abit already has it in his test network. ;-)

[yvv]

The idea behind Bancor is pretty simple: a smart token issuer holds a reserve in other ERC20 token(s) which is used to buy back their smart token.  This allows immediate exchange of smart token for reserve token(s) without the need to be listed at any exchange. The price is set by a smart contract such that the reserve ratio stays constant. This does not negate the need for normal exchange with bids and asks. Ethereum has at least one decentralized exchange for its assets running live by the way, check etherdelta.

Is there a distilled rundown of how the algorithm works?

It is in their white paper, including math derivation of price discovery algorithm.

[CoinHoarder]

In my own words (someone correct me if I'm wrong, as I'm still trying to understand it):

Bancor's Smart Tokens are backed by the real assets they contain, which are autonomously held in reserve funds.
Smart Tokens can consist of a single asset or a basket of assets.
Smart Tokens have an infinity amount of liquidity, liquidity of which is not dependent on finding someone to match your buy/sell order.

The price of Smart Tokens is calculated as: Price = valueHeldInTheReserveFund / ( totalSupplyOfTheSmartToken * anArbitraryPercentageWhichIsDefinedWhenASmartTokenIsCreated)

This price moves with the amount that is bought/sold, because it is continuously recalculated for every fraction of a coin bought or sold. For instance, when selling one Smart Token you would get a different price for every 0.000000001 of that one token.

The higher the arbitrary percentage, the less volatility there will be in the Smart Token's price due to there being more reserve funds for every 1 Smart Token coin.

There is no spread since both a buy and sell price are calculated with the same formula.

When the price of a Smart Token changes due to a buy or a sell, it then creates arbitrage opportunities for market makers who have incentive to correct the price closer to the value of the reserveFundValue / totalSmartTokenSupply. If the Smart Token is under or overvalued then an arbitrageur can instantly liquidate or purchase the Smart Token and profit by selling it on an exchange. This opportunity is amplified because there is immediate and unlimited buy/sell liquidity without the need of someone to match an order with, which increases market depth.

Optionally, the Smart Token issuer can set a fee for each exchange which can be reinvested back into the reserve fund. This increases the Smart Token price with every time an exchange takes place, because the value of the reserve fund increases compared to the value that was transferred. This works similar to interest bearing Smartcoins in Bitshares.

[CoinHoarder]

Is there a distilled rundown of how the algorithm works?

The explanation in the whitepaper is fairly easy to follow.

Advantages of Smart Tokens

Smart tokens introduce multiple advantages over the traditional exchange model:

1. Continuous Liquidity​ - Since purchasing and liquidating is done through the smart
contract, smart tokens are always liquid, irrespective of their trading volume.

2. No Extra Fees​ - The only mandatory fees applied by a smart token are the blockchain
platform fees (gas) which are relatively low.

3. No Spread​ - Since the price calculation is done algorithmically by the smart token, the
same price applies for purchasing and liquidating the smart tokens.

4. Predictable Price Slippage ​- Smart tokens allow pre-calculation of the precise price
slippage, based on the transaction size, before it is executed.

5. Lower Volatility - ​A smart token with a 10% CRR (for example) is comparable to an
exchange with 10% of the entire supply of a token in its order-book at all times, forming
substantial market depth. In a typical crypto-exchange, the share of the supply in the
market depth at any given moment is well below 1%. The higher the CRR, the lower the
smart token’s price volatility. The lower the CRR, the more “new credit” is created relative
to the original reserve amount.

A New Method for Price Discovery

A smart token utilizes a novel method for price-discovery which is based on a “Constant Reserve
Ratio” (CRR). The CRR is set by the smart token creator, for each reserve token, and used in price
calculation, along with the smart token’s current supply and reserve balance, in the following way:

Price = Balance / (Supply * CRR)

This calculation ensures that a constant ratio is kept between the reserve token balance and the
smart token’s market cap, which is its supply times its price. Dividing the market cap by the
supply produces the price according to which the smart token can be purchased and liquidated
through the smart contract. The smart token’s price is denominated in the reserve token and
readjusted by the smart contract per each purchase or liquidation, which increases or decreases
the reserve balance and the smart token supply (and thus the price) as detailed below.
When smart tokens are purchased (in any of their reserve currencies) the payment for the
purchase is added to the reserve balance, and based on the calculated price, new smart tokens
are issued to the buyer. Due to the calculation above, a purchase of a smart token with a less
than 100% CRR will cause its price to increase, since both the reserve balance and the supply are
increasing, while the latter is multiplied by a fraction.

Similarly, when smart tokens are liquidated, they are removed from the supply (destroyed), and
based on the current price, reserve tokens are transferred to the liquidator. In this case, for a
smart token with a CRR less than 100%, any liquidation will trigger a price decrease.
This asynchronous price-discovery model works by constantly readjusting the current price
toward an equilibrium between the purchase and liquidation volumes. While in the classic
exchange model price is determined by two matched orders in real-time, smart token prices are
calculated over-time, following every order.

The above formula calculates the current price, however, when a purchase or liquidation is
executed, the effective price is calculated as a function of the transaction size. The calculation
can be described as if every transaction is broken up into infinitely small increments, where each
increment is changing the smart token’s supply, reserve balance, and thus its price. This ensures
that purchasing the same amount of smart tokens in a single or multiple transactions would yield
the same total price. Additionally, this method ensures that the CRR will be kept constant and the
reserve can never be drained. Essentially, the effect of the transaction size on the price (due to its
changing the smart token’s supply and reserve balance) is incorporated into the effective price
for any transaction. The mathematical functions for calculating price per transaction size are
presented further in this document.

Using this method, the Bancor protocol can enable liquidity and asynchronous price discovery for
existing standard tokens -- through smart tokens holding them in reserve, enabling backward
compatibility.

A Solution to the Coincidence of Wants Problem

The coincidence of wants problem , in the current asset exchange model, creates a situation
where assets are required to be traded at a certain minimal volume or else face liquidity risk.
The cause for this limitation is that the chance of finding a second party with opposite wants to
exchange with, correlates to the asset’s trading activity level. Smart tokens solve this problem
through the use of reserve tokens which embed market depth directly into the smart token’s
smart contract.

Smart tokens are a technological solution​ to the coincidence of wants problem for asset
exchange, rather than a labor-based solution as used in traditional (or decentralized) exchanges.
The current laborers in asset exchange are the professional market-makers who provide liquidity
and facilitate collaborative price discovery. In the domains of information exchange and trade, the
technologies of writing and currency replaced labor-intensive solutions (speaking and barter) with
technological ones, creating mass efficiencies for societies and unlocking collaboration on a
global and intergenerational level. The Bancor protocol proposes to similarly advance the domain
of asset exchange by replacing the need for labor with a technological solution to the existing
coincidence of wants problem.

The above describes is effectively describing Bitshares' Smartcoins when it refers to "labor-based solutions".

[fluxer555]

Is there a distilled rundown of how the algorithm works?

[CoinHoarder]

What do you guys think of the economics behind Bancor? I would ask on Bitcointalk, but I think this is the best place to have this discussion. I am still digesting the economics of it, but essentially the Bancor project is purporting that they have conceptualized a way to create Smartcoins with infinity liquidity.

https://www.bancor.network/
https://www.bancor.network/whitepaper/en
https://bitcointalk.org/index.php?topic=1789222.0

So far the best proposals I have seen for highly liquid Bitshares Smartcoins are:
- Using inflation to provide autonomous market depth:
https://www.docdroid.net/2OcoImM/autonomous-smartcoin-liquidity-funded-by-dilution.pdf.html
- Tonyk's idea to make BTS unspendable, thus all BTS trading would take place on the BTS chain (unless exchanges program a special IOU trading module specifically for Bitshares trading... which is unlikely IMO due to the risk/reward):
https://bitsharestalk.org/index.php/topic,21409.0.html
- Using a worker proposal to use reserve pool funds to create smartcoins and sell them into the market at feed price plus 10%:
Discussed many places
- Using inflation to pay people to provide liquidity, similar to how Nubits does (this was not the reason Nubits failed btw):
Again, discussed many places.

But... the main question I am proposing in this thread... has Bancor come up with a better idea than all of the above (and better than Bitshares' historical idea of a "if you build it, they will come" mentality)?

[off topic]
If the economics of Bancor are sound, then this could be a Bitshares killer considering the DEX space is getting quite competitive recently (some implementations of which are better IMO, as far as decentralization goes). Again, IMO, Smartcoins is one of the last few things Bitshares has going for it.

In the DEX space... you now have:

Bitshares (derivative and IOU based)
cons: derivative and IOU based
pros: convenient and more established market

Waves (IOU based)
cons: IOU based
pros: highly funded project, big userbase and hype, clean and easy to use GUI

Supernet is apparently nearing a release of a DEX (autonomous multisig)
cons: multisig
pros: convenient and hold/trade real assets, not derivatives

Blocknet (atomic swap based)
cons: Have to download the entire blockchain, but this inconvenience can later be reduced to simply running SPV nodes of coins you want to trade. This inconvenience could technically be further reduced by building a multi coin SPV wallet which is integrated in the Blocknet wallet.
pros: fully decentralized, no IOUs and no derivatives or IOUs

Honorable mention to: B&C Exchange ... although the Nubits dev abandoned the project, they did conceptualize a great idea for how to implement a DEX that would be superior to IOU and derivative DEXs.

There are too many autonomous multisig and IOU-based DEX projects to list really...

Considering Bitshares is derivative based and/or IOU based, I value it around the same level as other autonomous multi sig, derivative, and IOU exchanges. I think Blocknet has the best design for a DEX thus far (as long as it is made to be convenient by implementing multicoin SPV wallets directly in the Blocknet client).

So, that leaves Smartcoins as Bitshares main "killer feature", which is why I am interested when new smartcoin-like proposals come out.
[/off topic]