The dark horse of DEX in the DeFi ecosystem: Bancor

This article was published on Chain News Blockchain.

The year was 1944. An economist's proposal by John Maynard Keynes ultimately failed to pass at the Bretton Woods Conference, laying the groundwork for what we now know as the U.S.-centric international monetary system. Bancor was conceived as a supranational currency, inspired by the French word "banque" or "bank gold," which would be devalued to retain its conceptual form.

That is to say, at least until 2017. To pay homage to Keynes's idea of a universal, neutral unit of account for settling all international trade and maintaining global trade balance. The Bancor protocol, along with its native token (BNT), was envisioned as a kind of spiritual successor. Here, BNT serves as the base pair for every token trading pair listed on decentralized exchange (DEX) platforms.

This article aims to delve deeper into the design of the Bancor protocol and the development trends over the years. The focus area is the innovations introduced by the latest Bancor v2.1 released in October 2020, aimed at addressing the existing pain points faced by DEX liquidity providers. Specifically, these are impermanent losses (IL) caused by price discrepancies between token pairs and mandatory two-sided liquidity collateralization, leading to involuntary token exposure. Bancor's solutions to these issues have generated a compelling unique selling proposition (USP), which market participants have begun to take notice of, driving a significant increase in adoption rates in recent months.

Automated Market Maker

Bancor is a pioneer of the Ethereum automated market maker (AMM) model. Before the emergence of AMMs, DEXs had to rely on traditional order book styles to match submitted bids and required execution of trades. However, Bancor chose to completely abandon the order book and instead used an on-chain liquidity pool network to pair token pools with each other. Users can then trade based on the trading pairs in these pools, with algorithmically set prices based on the scale of user trades and the depth of the corresponding token pools.

This new market-making approach has proven to be a paradigm shift in the DEX space. Now, investors can access on-chain liquidity through transparent pre-priced guarantees without needing a counterparty. Token holders can also convert their existing assets into productive assets by providing liquidity to these DEX pools and earn returns from the protocol trading fees generated by the platform.

Bancor Native Token

The eponymous Bancor token model is unique among existing DEXs. BNT is deeply integrated into the protocol, serving as the base asset paired with each ERC-20 token (TKN) asset listed on Bancor alongside ETH. Any trade between a given TKN A and B will occur through BNT as a "connector": TKN A > BNT > TKN B, and vice versa, facilitated by using trading pairs TKN A / BNT and TKN B / BNT.

However, this has proven to be a fatal weakness for Bancor, contributing to its initial adoption challenges. Token holders wishing to provide liquidity on the platform must acquire matching amounts of BNT to enter any token pool. This requirement prevents holders of two directly paired assets from making upfront capital expenditures or reallocating assets. With no option for unilateral liquidity provision, market participants had little incentive to commit to the Bancor ecosystem by purchasing BNT tokens.

In response to this issue, competitors in the form of Uniswap quickly emerged in November 2018. This new AMM DEX completely eliminated the need for a native token and allowed anyone to establish liquidity pools between any TKN and ETH. This reduced friction among users and quickly attracted an increasing number of liquidity providers (LPs) and traders, creating a positive feedback loop. Since then, Uniswap has become a leader in the DEX vertical market, consistently capturing the largest share of trading volume.

DEX Dark Horse

Despite Uniswap's tremendous success and the rise of similar competitors like SushiSwap, the Bancor team remained committed to iterating on their original product. With the release of Bancor v2.1 in October 2020, unilateral liquidity provision and impermanent loss insurance seemed to be the breakthrough needed for the USP to succeed in the competitive DEX landscape. This has been clearly reflected in parabolic growth metrics such as TVL and trading volume since the protocol update.

Breaking the Code

Bancor's new model is supported by a novel elastic BNT supply mechanism. Elastic supply refers to the protocol's ability to mint and burn BNT as needed. This operation is only triggered under specific circumstances and is made possible by the unique pairing of BNT across all liquidity pools.

Unilateral Liquidity Provision

The elastic supply model aids liquidity providers by offering unilateral supply for any ERC-20 token, something that other DEXs currently cannot achieve. When choosing to provide unilateral liquidity for a specific TKN, the protocol introduces a matching value of BNT into the TKN / BNT pool to provide liquidity for the other side. Newly minted BNT remains in the liquidity pool and is burned after the TKN liquidity provider withdraws their initial funds.

However, before the burn event occurs, it is now possible for other participants to unilaterally provide BNT. Any BNT balance specified by the protocol in the liquidity pool can be replaced by BNT from external sources to meet the demand for unilateral liquidity provision on both sides of TKN and BNT. This replacement is done by the protocol burning an equivalent amount of BNT upon receiving user-staked BNT.

Using protocol-based BNT for initial liquidity matching is a powerful feature that resolves the "chicken or egg" dilemma without diluting the circulating supply of BNT. The eventual BNT burn, whether from TKN withdrawals or replacing BNT, will include a share of the trading fees accumulated by the protocol from the liquidity pool. As long as a trade occurs between the minting and burning events, this will lead to greater BNT scarcity, as the amount of BNT burned will exceed the amount initially minted. Thus, the protocol automatically co-invests in itself (with the upper limit determined by community governance) and returns the profits to the accumulated value of BNT.

Impermanent Loss Insurance

Crucially, the elastic BNT supply creates the possibility for the protocol's impermanent loss insurance. When liquidity is withdrawn, it can cover any impermanent losses experienced by liquidity providers in the pool by compensating them with BNT. Unlike unilateral liquidity providers, competitors like Uniswap or SushiSwap cannot offer this feature. The BNT used for this purpose first comes from the trading fees accrued by the protocol (from co-invested BNT) and is only minted as a measure when the fee reserves are insufficient.

This is particularly important for the long-term viability of liquidity providers, as returns on investment can sometimes exceed profits from trading fees. For example, the Chainlink token saw its value increase by over 700% between April 2019 and 2020, resulting in liquidity providers on the LINK/ETH trading pair losing over 50% of their value relative to holding both assets.

Moreover, Bancor's implementation is carefully designed to avoid excessive inflation due to impermanent loss payouts. As mentioned, impermanent loss insurance can only be "realized" when liquidity is withdrawn, with no claimable expenses in between. Since minting BNT can support insurance payouts, theoretically, the cost of providing this insurance is borne by all BNT holders through sharing. However, in practice, this cost has so far been fully covered by the trading fee income generated by the protocol's invested BNT.

From the protocol activity between November 2020 and January 2021, it can be seen that for every dollar earned by the protocol, only $0.07 was used for impermanent loss compensation. Note that this aggregates all impermanent loss insurance costs and trading fee income in the protocol, effectively spreading the impermanent loss risk across all liquidity pools. This will provide self-sustaining impermanent loss insurance for all liquidity providers participating in the Bancor ecosystem.

The Bancor team has simulated the worst-case scenario for IL insurance payouts, assuming all liquidity providers withdraw all liquidity simultaneously (100% coverage). The chart below clearly shows that the inflation caused by covering all impermanent losses provided by the protocol only accounts for a small fraction of the total BNT supply. It is estimated that if all liquidity were withdrawn today, Bancor would only need around 4.2% of the current BNT supply to achieve complete impermanent loss coverage. Any withdrawn BNT will be locked for 24 hours to prevent liquidity providers from making rash decisions.

By granting impermanent loss coverage incrementally, the system also encourages liquidity providers to make long-term commitments. Increased liquidity is only secured at a rate of 1% per day, meaning liquidity providers must provide liquidity for 100 days to receive 100% impermanent loss coverage upon withdrawal. Impermanent loss insurance takes at least 30 days to become effective, so withdrawing liquidity early will not provide any coverage.

Currently, no other DEX on the market offers a similar solution to mitigate IL. The native token distribution to liquidity providers from SushiSwap can serve as an indirect method to offset IL, but it is not a targeted solution and the distribution rate is unaffected by IL. THORChain has chosen to adopt the Bancor design, but the IL coverage is paid from limited protocol reserves. Purchasing token options for providing liquidity is another viable strategy, but it is not beginner-friendly and depends on the availability of liquidity options for the tokens involved.

Given the increased price volatility of various crypto assets, IL insurance will help alleviate the primary concern that hinders more users from becoming LPs. With the protocol-backed IL coverage in place, this depends on the liquidity options market available for the relevant tokens as trading fees are accrued.

Liquidity Black Hole

Notably, impermanent loss insurance and unilateral liquidity also apply to the liquidity supply of BNT. This synergizes well with the liquidity mining incentive mechanism Bancor is implementing, as BNT rewards can be directly funneled into the liquidity pools. The incentives for BNT liquidity providers to earn yield farming rewards help strengthen the depth of the protocol's liquidity pools while doubling the supply, minimizing the "farm and dump" phenomenon. As of March 2021, 78% of BNT liquidity provider rewards have been re-earned.

Due to the co-investment limits of the BNT protocol, this also helps to ultimately increase the overall size of the liquidity pools. As new unilateral BNT provided to the pool replaces the existing BNT specified in the protocol, it opens up more space for TKN liquidity. The symbiotic relationship here is clear, as token holders are rewarded for contributing to the overall functionality of the platform.

Additionally, Bancor's liquidity is highly concentrated, as each TKN only has one liquidity pool on the platform—TKN/BNT. The TKN/TKN pairs introduced by Uniswap V2 tend to lead to the dispersion of available TKN liquidity across shallower pools (such as TKN/ETH, TKN/USDT, TKN/USDC, etc.). Higher liquidity generates higher trading volume from lower slippage, charging higher fees to investors holding BNT, making BNT more attractive for purchasing and rejoining the protocol.

Bancor Vortex

Bancor's latest product, Bancor Vortex, elevates the utility and capital efficiency of BNT to a new level by unlocking entirely new realms of potential use cases for vBNT tokens. To recap, vBNT is Bancor's governance token and is minted based on unilateral BNT liquidity provision, doubling as a liquidity provider token representing a percentage of ownership in the liquidity pool. Through Vortex, a BNT/vBNT liquidity pool is created, allowing BNT participants to exchange their vBNT for more BNT. This effectively serves as an interest-free and liquidation-leveraged means with a cap of 1.0x. The additional BNT obtained can then be exchanged for any supported TKN or deposited into the protocol in the form of leveraged yield farming.

Ultimately, by simply repurchasing the initial amount of vBNT through the same Vortex pool, the underlying BNT position can be unwound (plus accumulated fees). It is important to note that any debt incurred from using Vortex as incremental leverage is denominated in BNT. If the value of BNT rises, the value of the outstanding debt will also increase accordingly (in USD terms).

In the near future, the protocol will implement additional swap fees and direct them towards purchasing and burning vBNT. This provides ongoing upward price pressure to offset the impact of vBNT holders exchanging their tokens for BNT.

Now take a moment to consider the impact of this design on the circulating supply of BNT. In its current form, the additional fees granted to BNT liquidity providers help encourage more participants to hold BNT. Once the protocol's vBNT repurchase is implemented, it effectively also serves as a one-way bridge, depositing BNT into Vortex and gradually absorbing more BNT over time.

When predicting the long-term effects of this ongoing repurchase activity, things become even more interesting. Compared to the BNT that created vBNT, the total supply of existing vBNT decreases, leading to the hypothetical situation of 1 vBNT > 1 BNT. This creates an arbitrage opportunity where anyone can buy BNT, deposit it into any liquidity pool at a 1:1 ratio to mint vBNT, and sell it to Vortex for more BNT than initially purchased. With principal and profit, arbitrageurs have no incentive to return to the protocol to redeem the deposited BNT, effectively keeping it permanently. This is another way Vortex can bring externally circulating BNT into the protocol and enhance its liquidity reserves.

In the short term, this effect may not materialize, as vBNT holders will be incentivized to exchange their vBNT for BNT when the Vortex price peg tends toward 1:1. The closer the exchange rate is to parity, the less risk users take when creating leveraged positions in Vortex. In any case, 1 vBNT > and 1 BNT will also lead the protocol to effectively "pay" existing vBNT holders to take on leverage. Either way, opening more positions with BNT helps create "sticky" liquidity and attract more BNT into the Bancor system.

Potential Drawbacks

It is prudent to discuss the potential limitations of the Bancor model. The core features in v2.1 ultimately must follow a governance whitelisting process, where only approved tokens can benefit from unilateral liquidity provision and IL insurance, thus limiting potential growth. The liquidity mining rewards for any liquidity pool must also receive the same governance approval and must be renewed after 12 weeks.

Newly established projects that have not yet been whitelisted are unlikely to gain approval. Therefore, considering the additional inconvenience of acquiring BNT, these projects have little incentive to add liquidity to Bancor.

Stablecoin pools have also been found to contribute significantly to Bancor's IL insurance costs. The stronger the market trend in either direction, the greater the price deviation of any stablecoin/BNT.

The Road Ahead

With a comprehensive AMM foundation established, Bancor can now look to the future with optimism. In the face of fierce competition for market dominance, the Bancor team continues to boldly venture into uncharted territories within the DEX design space to differentiate its product portfolio and seek ways to add value to the industry. Market leaders Uniswap and SushiSwap stand to gain significantly from Bancor's upcoming series of products:

• Original Pools: Bancor plans to incentivize the listing of newer projects on Bancor with unilateral IL protection to replace SushiSwap's Onsen program.

• Shadow Token Stablecoin Pools: Bancor's stablecoin swap pools aim to attract TVL and challenge Curve as the primary stablecoin swap DEX.

• Layer 2 Expansion: Bancor may pioneer DEX on the Arbitrum mainnet, planned for release in Q2 2021.

• Cross-Chain Expansion: Using the Polkadot bridge as the first step towards a multi-chain future for cross-chain swaps using BNT.

• Further improvements to on-chain governance (gasless voting) and UI reforms since Bancor v2 are the final touches for a comprehensive overhaul.

Just as Keynes envisioned Bancor as a means to revitalize traditional finance and rebuild the world economy, we may see Bancor completely reform decentralized finance, establishing itself as a central exchange in the ever-evolving DeFi ecosystem.