Bancor: The History of DeFi’s Founding Father

On March 30, 2022, TheTie Labs released a research report on the history of Bancor and the latest Bancor V3. This article is the first part of the history of Bancor.


Bancor is one of the most overlooked and misunderstood DeFi-native projects, and its complexity can leave even the savviest DeFi-native project scratching their heads. That being said, it has maintained a TVL of over $1 billion for over a year and continues to have one of the strongest and most intellectually diverse teams in the field.

In this series, I had the pleasure of speaking with Bancor team members Mark Richardson (Head of Research) and Nate Hindman (Head of Growth). The interview will be covered in the sequel. For this in-depth analysis, I will simplify the core components of the Bancor architecture.

Before diving into where Bancor is headed, we must first understand their tangled past.

A brief history of Bancor


Before there was Defi, there was Bancor . Its name comes from the Keynesian concept of currency, called the International Clearing Union (ICU). The ICU used the “bancor” as a unit of account in the 1940s to encourage international trade. Although Bancor was founded in 2016, it was not until February 13, 2017 that Bancor released its first white paper. Fast forward four months to June 12, 2017, and finally, the prodigal son was born. They officially launched the largest ICO (Initial Coin Offering) at the time, raising over $153 million.

Bancor is one of the creators of DeFi. It all started when they noticed that the tokens pouring into the space needed a market to accommodate and facilitate their trading and pricing — a place that was safe and away from the constraints of a centralized monetary system. In 2016, this idea gave birth to the AMM (Automated Market Maker) model, the cryptographic primitive that sparked DeFi Summer and led to a glorious bull run in 2020.

Innovation is never an easy task. In 2018, Bancor suffered a security breach in which hackers stole $10 million worth of protocol-owned BNT, $12.5 million in ETH, and $1 million in NPXS tokens. The company responded by freezing the stolen BNT tokens and contacting multiple exchanges in an attempt to track down and retrieve the remaining stolen tokens. Unfortunately, in the end, they could only save BNT, losing a total of $13.5 million.

This sparked controversy within the industry. It’s not the exploit that people are upset about – it’s the team’s ability to freeze BNT at any time. This mechanism creates a single point of failure and brings protocol security into focus, which no investor wants to question. Bancor listened to the community and immediately started taking action.

That was their first and last attack. Since then, the team has not only changed their OPSEC system and removed the BNT freeze feature from the protocol, but also completely overhauled their codebase. Therefore, the protocol has been running smoothly since then, with no security issues.

Enter Bancor v1

Bancor invented liquidity pools and called them “relays” or “smart tokens” at the time. This is where the bonding curve mechanism we all know and love comes from, and it enables protocols like Uniswap or Sushiswap to exist the way they do now.

Before diving into Bancor, we must take a moment to highlight the real importance of the AMM model.

What is an Automated Market Maker (AMM)?

AMM is the underlying protocol that constitutes a decentralized exchange (DEX). On centralized exchanges, there are order books managed by professional market makers. AMMs eliminate the need for a custodial entity that creates a marketplace and the necessity of a counterparty. 

So what are the benefits of using AMM?

The AMM was created to decentralize the decision-making process for pricing different assets. With AMM, Bancor has developed an alternative to traditional order book-based models and facilitates on-chain liquidity by leveraging algorithmically managed token reserves. 

Gone are the days of having to wait for a centralized exchange to provide stable liquidity. With Liquidity Pools (LPs), anyone can provide trading depth with their favorite tokens, and they are incentivized to do so by having transaction fees as well as liquidity mining rewards. This creates a healthy environment where AMM liquidity pools are mutually beneficial for both token holders and token issuers. New tokens are able to generate liquidity without having to spend money to bootstrap initial liquidity themselves and hire market makers to help maintain liquidity.

How does AMM work?

AMM works by creating a joint curve consisting of the equation x*y=k. In this formula, x represents the value of ReserveTKN1 and y represents the value of ReserveTKN2. K remains the same. Let’s best illustrate how the AMM bonding curve works through a small case study.


Example AMM equation

In this example, Bob is a Link Marine who bought his first LINK in 2018 and has not sold a single cent to date – we love Bob. However, by early May 2021, Bob felt that the market would soon reverse. Because of this, he decided to break his oath and cash out some of his LINK into a benchmark asset he believes is safer, such as ETH. In this case, Bob removes ETH from the LINK/ETH pool while adding LINK to it. Our bonding curve reacts to this by raising the price of ETH and lowering the price of LINK to keep the pool balanced. 

It’s worth noting that AMM doesn’t change its price based on other markets around it. The price of the assets in the pool only changes with the reserve ratio. This creates differences between asset prices across different AMMs, creating arbitrage opportunities. These mispricing events are often collected by bots, but for ardent investors, there are still opportunities on lesser-known AMM-based DEXs. While AMMs have been incredibly successful over the past few years, they also have their drawbacks. 

The perks of being number one


After Bancor launched V1, they sat in a unique position to lie down and observe the world while interacting with DeFi. They have seen more and more DEXs adopt their formula and determined that the most prominent issues revolve around slippage, forced token positioning, high network fees and one big problem – impermanent losses. 

Let’s talk about what slippage is and why it matters. 

Slippage occurs when there is not enough liquidity to fill an order at the exact price at which it will be executed. This can cause significant changes in the price of an asset from the time an order is placed to when it is fully filled. This is especially true when the user is trying to place a relatively large market order. This forces larger investors to default to execution strategies such as TWAP (Time Weighted Average Price).

Slippage can also occur when trading activity is high. This is especially true for ecosystems like Ethereum, which currently use PoW (Proof of Work) consensus, limiting their ability to process transactions. When a transaction is executed on a DEX, the user’s transaction is placed in a queue. The longer the queue and the longer the transaction takes to process on the blockchain, the more likely it is that there will be a discrepancy between the price you think you are getting and the price you actually get.

Impermanent loss IL

Impermanent loss is the difference in value between the tokens held in the user’s wallet and the liquidity pool. Earlier in this article, we discussed how AMMs constantly rebalance themselves based on token withdrawals/deposits. This means that as the price of the reserve token starts to fluctuate, the party that is rising is sold by the AMM and invested in its counterparty assets to keep the liquidity pool at a predetermined ratio.  

Why would anyone suffer from IL?


Believe it or not, it took people a long time to realize the idea of ​​impermanent loss. When the idea of ​​being a liquidity provider started to explode in mid-2020, it was easy to put on rose-colored glasses and repeat adages like “It’s not a realized loss until I sell!” As staking grows, what everyone cares about is “passive income” from yield-generating assets. The idea is that if the transaction fees you generate from LP are higher than your IL, then the strategy can be profitable with very little effort – or so you think. 

Due to the lack of transparency of DEXs on IL, multiple external studies have since revealed the truth behind LP profitability. Bancor and Topaz Blue published a study stating that 49.5% of LPs lost money. Defiant found that 52% of LPs using Uniswap V3 were also unprofitable. In hindsight, these numbers are not surprising at all, given the volatility of cryptocurrencies over the past 2 years.

ttUJ5NrDId8wgwr4o6Ss8F1eqH9qo9oU5JLqIObj.pngWhile IL is a natural phenomenon stemming from the current state of cryptography, the real innovation comes from the protocol, which finds ingenious ways to solve problems and improve the user experience.

Better Models – Enter Bancor v2.1

Over time, the novelty of Bancor V1 faded away, and different protocols took the standard CPMM formulation and devised their own. Notable unique approaches include Curve using hybrid CFMMs. That said, most AMMs still have specified counterparty assets that LPs need to contribute, think ETH (underlying asset)/USDC (counterparty asset). This is extremely inconvenient for LPs as it means they are forced to split asset exposures to meet the requirement to provide liquidity within the pool. 

Bancor v2.1 directly addresses this issue, allowing LPs to provide liquidity to the pool while maintaining 100% exposure to a single asset (one-sided staking). In this way, Bancor greatly reduces slippage and facilitates extremely deep liquidity. Now, the liquidity of a token is spread across multiple pools (eg TKN/ETH, TKN/WETH, TKN/LINK), and all the liquidity of a single token is concentrated in one pool.

These pools then continuously generate fees for LPs and protocols. Fees earned by the protocol are used to reimburse LPs for impermanent losses once users withdraw their liquidity positions. In cases where the impermanent loss of the pool is greater than the fees it generates, the protocol mints BNT from the DAO co-investment to repay the remaining IL. At first glance, this sounds like a good way to dilute the token supply, but small changes can have big consequences. In the next section, we will examine BNT token economics to clarify our understanding of these implications. 

Token Economics

All of the features brought by the v2.1 upgrade are achieved by leveraging the elastic supply of BNT, which allows the protocol to mint and burn BNT as it deems necessary. 

Whenever a protocol wants to whitelist their tokens on Bancor, they need to formally submit a proposal within the Bancor DAO for BNT holders to vote on. If the proposal passes, the DAO will vote to co-invest (in BNT) in the liquidity pool. In this case, the DAO is essentially given the ability to mint BNT for matching user deposits to the pool and guarantee impermanent loss insurance. 


Example of how it works

Joe is looking for a place to earn money on his LINK. He noticed that there is a LINK pool on Bancor that allows him to one-sided staking his LINK (LINK/BNT). Joe then proceeds to deposit $100,000 worth of LINK tokens into the pool. The protocol simultaneously minted an equivalent amount of BNT dollars to match Joe’s position in the pool. Now, both the protocol (Chainlink) and the user (Joe) are incurring fees. 

In another scenario, if Joe withdraws his liquidity, the protocol will burn the corresponding BNT. On paper, this sounds like a zero-sum game, but it actually constitutes a key deflationary component of Bancor v2s token economics. When Joe stakes his LINK in Bancor, the LINK/BNT pool is generating fees paid in BNT. Assuming that the fees incurred are greater than the impermanent losses, the network will end up with more BNT than Joe contributed initially when he deposits – causing the deflationary system to keep burning BNT as LPs withdraw their positions. 

Since BNT is the denominated asset for mining pool rewards, some investors are concerned that the inflationary nature of the “farm token” will create continued selling pressure. Combined with the other inflationary factors we mentioned, there is every reason to doubt the effectiveness of a deflationary system. While inflationary feedback loops are sometimes preferred over deflationary feedback loops, the purpose is to spread risk across multiple pools, resulting in an average net positive outcome.

Bancor Vortex and vBNT

Vortex launches on February 17, 2021, a few months after Bancor v2.1 was released. With its launch, it introduces new ways for users to interact with BNT, as well as some additional deflationary features to complement BNT’s token economics. This is achieved by using the derivative product vBNT introduced with this update.

When staking BNT on the platform, the protocol mints vBNT at a 1:1 ratio. The token represents the depositor’s share in the pool, while also providing some additional use cases such as:

  • Liquidity – vBNT is liquid BNT, and every token on Bancor is liquid. Stakers can now choose to utilize their vBNT by exchanging it for any platform asset.
  • Governance – vBNT is the token used to vote for protocol changes and whitelist proposals in the Bancor DAO. This allows staked users to vote.
  • Earnings – stakers earn a share of transaction fees by staking vBNT in the vBNT/BNT pool.

Let’s bring Joe back to demonstrate the Vortex use case and highlight some important nuances.

  • Joe deposits 500BNT in the LINK/BNT pool. The protocol mints 500 vBNT to represent pool shares. 
  • Joe sees AAVE falling to a key support level and thinks now is a good time to buy. So, he swapped his vBNT for AAVE
  • Joe then decides that he wants to hold his newly purchased AAVE for a while and decides to use it to provide liquidity to the BNT/AAVE pool. 
  • Fast forward 12 months and now, due to all the fees Joe has accrued, his BNT share has tripled and is now worth 1500 BNT. He thinks now is a good time to cash in and buy the home he’s been looking for. 
  • In order for Joe to withdraw his LP, he needs to have the same amount of vBNT in his wallet as when he first deposited it – in this case 500 vBNT.
  • Joe buys 500 vBNT with some USDC in his wallet. He can now withdraw his staked BNT from the LINK/BNT pool. This burns 500 vBNT and returns 1500 BNT. 

Note that while Joe has tripled the amount of BNT staked in the pool, the amount of vBNT he needs to withdraw BNT in his wallet remains the same.


Dune Chart – If you sold your vBNT in the summer of 2021, it would cost twice as much to buy it back at the current vBNT price.

While implementing leverage on Bancor is a great addition to the protocol, it is still leverage and there is always risk. Borrowing against pledged BNT is effectively a vBNT/BNT pegged bet. The cost is the difference between the vBNT price when you redeem and the price when you buy it back.


Bancor foresees the difficulties posed by the complexity of the system. Despite the best efforts to highlight the risks, there is sure to be a group of people with irresponsible influence. With these users in mind, Bancor is introducing some different deflationary token economics to vBNT to help offset some of the risks involved with leverage and help the protocol as a whole. 

feel the burn

On April 4, 2021, Bancor announced the official deployment of Bancor Vortex Burner, which means that vBNT burning has gone live. Enabling vBNT burning means: 

  • Increase liquidity depth by locking a portion of each exchange into the protocol.
  • Offsets the continued upward pressure on vBNT by continuing to burn, reducing borrowing risk.
  • Permanently reduce the circulating supply of BNT by continuing to buy and permanently lock BNT.  
  • Increasing the amount of liquidity the protocol owns through the Treasury allows them to provide more space for liquidity pools and pay for insurance. 

The release of vortex introduces an important element in Bancor’s monetary policy. Vortex enables the protocol to adjust the burn rate of vBNT by collecting up to 15% of transaction revenue. Remember, for every vBNT that exists, there is 1 BNT equivalent. Therefore, continuously burning vBNT is equivalent to locking BNT in the protocol forever.


Less than a year since the burn began, a total of $4,314,188.26 of vBNT has been burned in the protocol so far, and $7,325,296.08 of BNT has been locked. 

Limitations of V2.1

There is no doubt that Bancor v2.1 successfully provides a solution to the limitations imposed by the first generation of AMMs. While this may be true, that’s not to say that this version of the protocol doesn’t have its own limitations. We analyzed the following key factors:


Enter Bancor V3

Bancor V3 proposes a major upgrade to the protocol architecture, aiming to address all major issues present in V2 and to improve the user experience by introducing several new features, including:

  • Immediate Impermanence Loss Protection
  • Redesigned single-sided pool token framework
  • Introducing “Ultra Liquidity” and Unlimited Pools, Unlimited Deposit Caps
  • flash loan
  • Improve liquidity and reduce transaction costs by using Bancor’s new Omnipool
  • Double-sided pool rewards are issued.
  • Contract-level auto-compounding rewards.
  • External impermanent loss protection
  • New integrations, multi-chain and L2 support, improved UI, and more. 

Keep an eye out for the lower part of this series, where we break down each of these components in more detail.

Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2022-03-30 11:35
Next 2022-03-30 11:37

Related articles