On April 29, the Bancor (BNT) project revealed its plans for the V2 upgrade of its platform. The overhaul addresses some of the major usability issues that have plagued the project since its launch in 2018.

With Bancor V2, scheduled for launch in Q2 2020, the team believes to have solved several risks incurred by liquidity providers to its platform. Bancor operates through automated market makers. This eliminates the need to maintain an order book. Instead, it relies on pools of liquidity and a price slippage mechanism to emulate natural fluctuations in price.

Liquidity providers earn trading fees, but in many cases they will suffer an “impermanent loss” that diminishes the value of their staked liquidity.

Nate Hindman, head of growth at Bancor, told Cointelegraph that this happens when the relative prices of two tokens change. He explained through an example:

“When ETH’s price goes up relative to DAI, that essentially gives an opportunity for arbitrageurs to balance the pool. And this can cause impermanent loss.” 

Since the relative value of each side of the pool changes, it is possible for a user’s initial stake to be a different percentage of the total pool, especially if it was initially a stablecoin. The user would thus withdraw less money than they put in.

Another issue that limited Bancor’s adoption was the need for projects to purchase its network token. Many users were faced with a dilemma, as Hindman noted:

“A lot of liquidity providers don't want to lose their long position, or token projects that are very rich in their own token, don't necessarily want to convert some of those tokens to BNT.”

The Chainlink solution

The solution to both these problems was to use a price oracle provided by Chainlink (LINK). The oracle can be described as a crutch for Bancor to lean on when balancing the relative liquidity between different tokens. 

As Hindman explained, “it's allowing Bancor to build these pegged reserve pools where the relative reserve values are not changed.” In these pegged pools, each conversion will trigger an oracle call. These will “balance the liquidity pools” according to the relative contribution from each user.

The impermanent loss problem does not exist on “stable” pairs, such as conversions between different stablecoins, or wrapped and unwrapped versions of the same token. The Chainlink integration thus reduces risk for liquidity providers, who could otherwise lose money from staking.

Oracles are found in many decentralized finance, or DeFi, products. However, many of these projects created their own proprietary versions. Asaf Shachaf, Bancor’s head of product, explained why the team decided to use a third-party partner, Chainlink:

“We are experts in liquidity pools. This is where our focus is and what we do best. Chainlink are experts in oracles. They know how to make oracles that are […] more resilient to market changes.”

Facilitating the rise of automated exchanges

Hindman referred to the impermanent loss issue as “DeFi’s dirty little secret.” He claimed that Bancor’s competitors, like Uniswap, also suffer from the same problems. 

According to the team, three key features present in Bancor V2 will help make this type of exchange more popular. While the removal of impermanent loss and exposure to multiple tokens were mentioned, a third problem is the excessive slippage experienced by users. 

Bancor solved this issue by adding an amplification coefficient. This reduces the amount of slippage relative to the total value in the liquidity pool.

That approach comes with its own risks, however, as it can result in the liquidity pool being drained completely. This is why it was previously only used on stable pools, as Shachaf explained:

“This risk is eliminated when you take it into ‘stable-to-stable’ pools, because you know that the price of the token is always the same. It's always aspiring back to the same value.”

With Bancor V2, it will be usable on volatile pairs as well thanks to the oracle integration.

As Hindman revealed, market feedback was not very positive when dealing with users or institutions due to these “secret” issues, though he emphasized that they are not unique to Bancor. He concluded:

“We expect and we hope that this [V2 upgrade] will bring tons more liquidity to the protocol. And that we won't have to have these conversations about impermanent loss or providing liquidity and also holding another token in addition.”

In the future, Bancor will also integrate with lending protocols to provide liquidity and drive further profits. This would decrease the opportunity cost of staking on Bancor, Hindman said.

A previous version of this article incorrectly stated that the amplification factor for slippage would only be applied to stable token pools.