Understand Bancor — A Real Passive Income for Token Holders

korpi
14 min readJun 28, 2021

--

The immense rise in popularity of Automated Market Maker (AMM) technology shows that DeFi has definitely found its product-market fit. Permissionless nature of decentralized exchanges allows hundreds of young projects to easily bootstrap liquidity for their tokens. However, this would not be possible were it not for liquidity providers who lend their idle assets to the protocol in anticipation of passive returns from trading fees. The dirty secret of most AMMs is that many liquidity providers lose in comparison to simply holding the tokens on their own. This risk, somehow ironically named “impermanent loss”, is still widely underappreciated in the DeFi space. This paper seeks to explain the impact impermanent loss has on returns for liquidity providers and how this risk can be mitigated by a unique AMM design proposed by Bancor.

TL;DR:

  • Liquidity providers to traditional AMMs, like Uniswap v2 or Sushiswap, often underperform a simple buy-and-hold strategy due to the impact of impermanent loss, which can substantially outstrip returns from trading fees.
  • Advertised APR figures for traditional AMMs ignore the impact of impermanent loss and, therefore, are misleading indicators of future returns. Real net returns for liquidity providers are often much lower, even negative because of impermanent loss.
  • Impermanent loss protection is a killer feature for passive liquidity providers. This is the Bancor’s unique value proposition.
  • Comparative analysis between Uniswap and Bancor pools shows that liquidity providers on Uniswap often end up with less money than they would have by “hodling”, while liquidity providers on Bancor always outperform a buy-and-hold strategy when full impermanent loss protection is achieved.
  • Single-sided exposure combined with impermanent loss protection turns Bancor pools into interest-bearing accounts for token holders who can enjoy their guaranteed passive income.

Is liquidity provision to AMMs a source of passive income?

Liquidity provision involves investors lending their idle assets to the protocol which uses them in a more productive way by offering its services to the end users.

Money markets, like Aave or Compound, take capital provided by lenders and offer borrowers overcollaterlized loans. Borrowers pay interest which is the source of passive income for lenders, i.e. liquidity providers to money markets.

In terms of AMMs, assets deposited by liquidity providers are used to facilitate trading on a DEX in an automated and non-custodial manner. Traders pay fees on their buys and sells which accrue to the deposited capital. Although trading fees generate a revenue stream for liquidity providers to AMMs, it’s not always guaranteed passive income. It would be, were it not for the infamous risk, impermanent loss.

“Impermanent loss” is rarely impermanent

The untold truth is that many liquidity providers to traditional AMMs, like Uniswap v2 or Sushiswap, often underperform a simple buy-and-hold strategy. Trading fees they earn are not high enough to compensate for AMM-native decrease in capital value which crypto industry somehow ironically named “impermanent loss”.

Impermanent loss is the difference in value between holding tokens in an AMM liquidity pool and holding them in an investor’s wallet (buy-and-hold strategy). A liquidity pool contains a pair of tokens and their amounts fluctuate as trades go through the pool — each trade increases the amount of the sold token and decreases the amount of the bought token. Therefore, the initial amounts of both assets deposited to the pool may be different from their current amounts. The difference in value between the initial assets and current assets is defined as impermanent loss (IL).

Impermanent Loss (IL) = current assets at current prices — initial assets at current prices

The magnitude of impermanent loss is dependent on the price ratio between both tokens in the pool. The bigger the divergence between the current price ratio and the initial price ratio, the greater the impermanent loss. If the prices of both tokens double, the price ratio doesn’t change and there is no impermanent loss. However, if only one token appreciates in price and the other one is stable, impermanent loss negatively impacts the value of the tokens in the pool.

Why is it “impermanent”? Because if the price ratio between tokens in the liquidity pool returns to the initial state when a liquidity provider entered the pool, the loss disappears. It also indicates another characteristic of impermanent loss — it’s only a realised loss when a liquidity provider exits the pool. As long as tokens are inside the liquidity pool, impermanent loss is an unrealised loss and can theoretically still go away.

However, this is rarely the case in real life. Many liquidity providers have painfully experienced that more often than not, impermanent loss becomes permanent and the expected passive income from providing liquidity becomes a net loss.

Read more about impermanent loss here: Beginner’s Guide to (Getting Rekt by) Impermanent Loss

How to mitigate the impact of impermanent loss?

There are a few ways liquidity providers can mitigate the negative impact of impermanent loss by taking a more active approach to liquidity provision:

  • Provide liquidity to pools with highly correlated assets. If prices of tokens in a pool are highly correlated, they should move in the same direction and the price ratio between them should be relatively stable. Historical correlation doesn’t imply future correlation though and even if prices move in the same direction they can still diverge from each other (e.g. one price grows faster than the other) leading to impermanent loss.
  • Provide liquidity to pools with high Volume to Liquidity ratio. The higher Volume to Liquidity ratio, the more fees liquidity providers earn and the higher a chance that trading fees will cover impermanent loss. Volume to Liquidity rarely stays high for long (more liquidity providers enter the pool and the ratio decreases) so it’s hard to predict whether it will be sufficient to compensate for IL.
  • Average entry to the pool. Impermanent loss disappears when the price ratio returns to the initial state when a liquidity provider entered the pool. Each time liquidity is added, the initial ratio is averaged between all the entries and becomes closer to the current ratio. The more entries at different price ratios, the lower the risk and impact of impermanent loss. It requires a lot of effort and transaction costs though.
  • Wait for a good time to exit the pool. Impermanent loss becomes a realised loss only when a liquidity provider exits the pool. It’s always possible that the price ratio will return to the initial state, however, it may require a lot of patience and never happen regardless.

None of the above methods guarantee that liquidity providers will not suffer from impermanent loss and eventually incur a net loss from liquidity provision. They are only mitigation strategies which help in favorable conditions but are dependent on future factors which cannot be predicted.

Impermanent loss protection is the killer feature for passive liquidity providers

If liquidity providers are expected to constantly monitor their positions and use time-consuming (and often unreliable) mitigation strategies to avoid significant losses, liquidity provision becomes a game that is reserved for only the most advanced players. This hardly resembles the concept of passive income.

The only way liquidity providers to AMMs can enjoy their passive earnings from trading fees without the risk that they will underperform a buy-and-hold strategy is by completely eliminating impermanent loss. While AMM design makes it impossible to eliminate impermanent loss, it is possible to transfer and distribute this risk. This is Bancor’s unique value proposition.

Bancor protects liquidity providers from impermanent loss

Impermanent loss protection on Bancor means a liquidity provider can get back the same value of tokens originally deposited to the pool plus trading fees. In other words, liquidity providers on Bancor always outperform a buy-and-hold strategy (when full impermanent loss protection is achieved) which is far from guaranteed in other AMMs.

It is possible because Bancor eliminates impermanent loss for liquidity providers by shifting the risk to the protocol itself. Using an elastic BNT supply, the protocol is able to co-invest BNT in pools alongside liquidity providers and pay for the cost of impermanent loss with trading fees earned from its co-investments. If fees earned by the protocol are greater than impermanent loss compensation, the protocol is able to offset its costs without emitting new BNT.

Protection against impermanent loss for liquidity providers accrues over time. It grows 1% every day so full protection is effective after 100 days, covering any impermanent loss experienced up until then or any time after. Withdrawals prior to the 100-day maturity are only eligible for partial compensation and there is no compensation for withdrawals within the first 30 days.

Read more: How does Impermanent Loss Protection work?

Bancor Protocol has invented and implemented a novel mechanism to distribute the risk of impermanent loss across a wide array of pools in order to completely eliminate this risk for liquidity providers. IL-protected pools on Bancor are, therefore, an attractive and risk-free source of passive income for liquidity providers.

Bancor vs Uniswap — a quest of passive income for liquidity providers

To better illustrate the impact of impermanent loss on the returns from liquidity provision, I used historical data to compare the performance of two strategies: providing liquidity to IL-protected pools on Bancor and providing liquidity to IL-exposed pools on Uniswap. The purpose of the comparison is to check if liquidity provision on these two AMMs is a viable passive income strategy.

IL-protected vs IL-exposed liquidity provision — a comparison of LP returns

No analysis is possible without some arbitrary choices (e.g. time period of the analysis, tokens included in the comparison, etc.). To maintain maximum objectivity, the comparison between Bancor and Uniswap was conducted with respect to the following guidelines:

  • Long term preference. The impact of impermanent loss is dependent on the start and end date of the liquidity provision period. The start date for this analysis is 2020/11/01 — first IL-protected pools on Bancor went live in mid-October 2020, therefore, all the compared pools must have been already whitelisted for the protection by this day. The end date is 2021/06/26 — the last day available when this analysis was made. This is almost 8 months of data — as long a period as possible.
  • Only tokens with IL protection. Not all the pools on Bancor are whitelisted for impermanent loss protection. To comply with a long term preference described above, only pools which had IL protection activated in October 2020 were used in the analysis. These include: LINK, LRC, MKR, MTA, OMG, REN, renBTC, RSR.
  • Liquidity provider perspective. While the performance of the pool (pool ROI) and the net outcome for a liquidity provider (LP ROI) are one and the same number on Uniswap, they are different on Bancor due to the existence of IL protection. As the purpose of this analysis is to find a better strategy for liquidity providers, LP ROI is used as a deciding factor.

The analysis was conducted using graphs from https://amm.vav.me/ — an insightful tool which charts returns from Bancor, Uniswap and Sushiswap over time. Below are a few guidelines on how to read the graphs:

  • Each pair of graphs represents two pools for the same token and the same period of time:
  • Left graph — IL-exposed pool on Uniswap
  • Right graph — IL-protected pool on Bancor
  • Vertical axis measures percentage return vs buy-and-hold strategy. Positive values mean liquidity providers earn profits from providing liquidity. Negative values indicate they are at loss — they would do better if they simply held their tokens outside of the liquidity pools.
  • Collected Fees (blue) — percentage return vs buy-and-hold strategy from the trading fees collected in the pool.
  • Impermanent Loss (red) — percentage loss vs buy-and-hold strategy due to the impact of impermanent loss.
  • Pool Return (green) — net percentage return vs buy-and-hold strategy when both collected fees and impermanent loss are accounted for. Pool Return = (1+Collected Fees) * (1+Impermanent Loss). On Uniswap this is the equivalent to the net return for a liquidity provider (LP ROI). On Bancor, this would be the LP ROI if there was not IL protection.
  • Bancor Protected (yellow) — net percentage return vs buy-and-hold strategy for liquidity providers on Bancor when IL protection is accounted for. This is LP ROI on Bancor.
Figure 1: Percentage returns vs a buy-and-hold strategy for LINK liquidity providers on Uniswap and Bancor
Figure 2: Percentage returns vs a buy-and-hold strategy for LRC liquidity providers on Uniswap and Bancor
Figure 3: Percentage returns vs a buy-and-hold strategy for MKR liquidity providers on Uniswap and Bancor
Figure 4: Percentage returns vs a buy-and-hold strategy for MTA liquidity providers on Uniswap and Bancor
Figure 5: Percentage returns vs a buy-and-hold strategy for OMG liquidity providers on Uniswap and Bancor
Figure 6: Percentage returns vs a buy-and-hold strategy for REN liquidity providers on Uniswap and Bancor
Figure 7: Percentage returns vs a buy-and-hold strategy for renBTC liquidity providers on Uniswap and Bancor
Figure 8: Percentage returns vs a buy-and-hold strategy for RSR liquidity providers on Uniswap and Bancor

Net returns for liquidity providers realised on the end date of the analysis period (2020/06/26) are presented on the below graph:

Figure 9: Net returns for liquidity providers (LP ROI) on Uniswap and Bancor vs a buy-and-hold strategy in the analysed time period (2020/11/01–2021/06/26)

There are a few interesting conclusions which come from the analysis of the above graphs:

  • Impermanent loss is often permanent. Even though a relatively long period of time (8 months) was used in the analysis, only MKR liquidity providers were lucky to experience an impermanent nature of impermanent loss — its negative impact cancelled out when MKR/BNT in the Bancor pool and MKR/ETH in the Uniswap pool went back to their initial ratios. In the other pools impermanent loss would become a realised loss if liquidity providers wanted to withdraw their liquidity on the end date of the analysis period (2021/06/26).
  • Impermanent loss is a substantial risk for liquidity providers. While returns from collected fees grow steadily over time, in a stable slowish way dependent on a trading activity in the pool, the impact of impermanent loss can be sudden and often of a much bigger magnitude. Although the pool return can stay positive for some time when impermanent loss is close to 0, it quickly plunges into negative territory when asset prices in the pool diverge and impermanent loss outstrips the return from fees. Because the prices of crypto assets are highly volatile, the risk of impermanent loss is significant.
  • Impermanent loss makes it hard for liquidity providers to exit the pool at their preferred time. Impermanent loss disappears if asset prices in the pool revert to their initial ratio, therefore, staying in the pool and waiting while earning fees may be a good exit strategy. Nevertheless, it forces liquidity providers to keep a position they may no longer be comfortable with and with no guarantee that the loss will ever be cancelled out. Protected liquidity on Bancor fixes this issue by allowing liquidity providers to exit at any time with guaranteed compensation for impermanent loss.
  • Net returns for Uniswap liquidity providers are highly volatile and often negative. LP ROIs (Pool Returns) for the analysed tokens and the chosen time period ranged from -23% to 18%. It suggests that Uniswap liquidity providers have to accept a high level of uncertainty when it comes to the expected returns. Moreover, a positive outcome is not guaranteed — only 3 out of 8 analysed Uniswap pools (RSR-ETH, MKR-ETH and renBTC-ETH) earned profits for liquidity providers (see: Figure 9). The other ones ended at a loss.
  • Returns from fees alone on Uniswap are higher than on Bancor but it doesn’t imply higher net returns for liquidity providers. Blue curves (Collected Fees) for the Uniswap pools are steeper than for the Bancor pools which indicates Uniswap generated more volume (and fees) per liquidity in the analysed period than Bancor. Fees APRs, which are often oversimplified as an approximation of the returns from liquidity provision, would be higher on Uniswap than on Bancor. It could attract more liquidity providers to Uniswap, in anticipation of a higher yield, however, most pools would underperform Bancor because of impermanent loss. In the conducted comparison liquidity providers of 6 out of 8 analysed assets earned more on Bancor than on Uniswap (see: Figure 9). Only 2 Uniswap pools (RSR-ETH and MKR-ETH) were more profitable for liquidity providers than their corresponding pools on Bancor.
  • Bancor liquidity providers have guaranteed profits when full impermanent loss protection is activated. Yellow curve (Bancor Protected) goes into a positive territory after 100 days from the pool entry and stays there until the end of the analysis period regardless of the impermanent loss impact. 100 days is when full IL protection is accrued. After this time liquidity providers don’t have to worry about the return vs a buy-and-hold strategy — it will always be positive because the protocol will fully compensate for any impermanent loss.

Bancor provides liquidity providers with guaranteed attractive returns by protecting them from impermanent loss

There are two main conclusions which come from the conducted analysis:

  • Uniswap returns are not guaranteed. Because of impermanent loss, liquidity providers often end up with less money than they would have by “hodling”.
  • Due to impermanent loss protection, Bancor guarantees long term liquidity providers positive returns vs a buy-and-hold strategy.

Bancor should be the obvious choice for passive liquidity providers who want to book guaranteed profits. Uniswap is an option for investors with a bigger risk appetite — in favorable conditions (low IL) returns can be higher than on Bancor but they are often negative when impermanent loss substantially outstrips return from fees. APR from fees alone is not a good approximation of future returns.

Advertised fees APRs are often misleading indicators of future returns

When liquidity providers decide where to allocate their capital, they often don’t take impermanent loss into account as it’s impossible to predict its future impact. The only easily available information is APR from fees, however, it’s not a reliable indicator of future returns.

Liquidity providers on Uniswap and other similar AMMs should be aware of the following caveats:

  • Advertised fees APRs on many platforms are often based on the performance in the last 24 hours only. Trading volume and liquidity in the pool can substantially change in future.
  • Fees APRs for IL-exposed pools, like on Uniswap, ignore the impact of impermanent loss and, therefore, don’t represent net returns for liquidity providers which are often much lower, even negative. Net returns must take into account both fees and impermanent loss and the latter is often higher than the former.

Bancor is a go-to platform for token holders who want to generate passive income on their holdings

So far this article has discussed Bancor and Uniswap in terms of their attractiveness for liquidity providers. Such a comparison, however, doesn’t do justice to Bancor which can tap into a much wider target audience than “liquidity providers” only. In contrast to Uniswap and other similar AMMs, Bancor allows any token holders with long positions on their preferred assets to benefit from the passive income generated on the network.

Uniswap and other similar AMMs require liquidity providers to provide liquidity in both assets in the pool. Even if external tools, like Zapper, are used to enter the pool with a single asset only, in fact half the value of the provided asset is exchanged for the other one before liquidity is added to the pool. Therefore, liquidity providers in other AMMs are always forced to maintain exposure to both assets in the pool.

Bancor fixes this inconvenience by allowing users to provide liquidity in a single asset only. This is a real single-sided exposure, not an interface overlay to improve user experience. Depositor’s stake in the pool is expressed in the provided asset only which means any investor who wants to hold a long position on the preferred asset (with no forced exposure to other assets) can become a liquidity provider on Bancor.

Single-sided exposure combined with impermanent loss protection turns Bancor pools into interest-bearing accounts for token holders. Users can deposit the assets they hold and have guaranteed positive returns after full IL protection is accrued (100 days). By transferring the risk of impermanent loss from the liquidity providers to the BNT token holders, depositors on Bancor always have guaranteed profits vs buy-and-hold strategy and this is what other competitive AMMs don’t offer.

Action Items:

--

--

korpi
korpi

Written by korpi

Full time in DeFi. Fond of all the financial experiments conducted on the blockchain.