AMA with Bancor
Regain lost ground, the sword refers to impermanent loss
The first project in the world to experiment with the AMM model was created by Bancor. And due to being trapped in its own model, Bancor’s share of the #DEX market was gradually decreasing.
After a period of research, Bancor recently launched Bancor V2, which protected against impermanent loss and allowed a single asset to provide liquidity.
AMA Q&A Part
Host: Jack Yang from Token Damo
Guest: Nate Hindman from Bancor
Q1: According to LinkedIn, you joined the Bancor Team in January 2018. Prior to this, you had been engaged in marketing and sales related work for many years. What attracted you to join the Bancor team? How have the members of the Bancor team changed over the last 3 years?
I studied economics and started my career as a journalist in the U.S. covering new financial technologies. Afterwards, I decided to learn some programming, launched and worked in a couple mobile Internet startups.
In 2017 I discovered Ethereum and started researching decentralized exchanges and trading tokens. One of the DEXs I used to trade was Bancor and after poking around the exchange I noticed a peculiar feature: there were no order books. Instead a set of smart contract-based automated market-makers (AMMs) was used to perform the exchange’s core operations, like listing, order-matching and the circulation of trading fees.
I started reading about AMM liquidity pools (Bancor called them Relay or Smart Tokens at the time) and learning about how they could be programmed. I reached out to the Bancor team and started working on the project full-time in late 2017.
Having launched the first AMMs on Ethereum in 2017, the team has learned a lot about the primitive over the past 3 years and to watch the AMM explosion in DeFi has been nothing short of incredible. The past year in particular has made the team more confident than ever that this new form of asset exchange and liquidity provision can have a transformative impact on both crypto and the broader financial system.
Q2: Bancor was actually the pioneer of the AMM model, and it attracted a lot of followers. However, over the past year or so, the market share has been gradually overtaken by other project teams, such as Uniswap. As a team member who witnessed the change, what do you think led to the change? What opportunities did the Bancor team fail to seize during the competition?
As you probably know, Liquidity is the King in AMM-based DEX protocols. The more tokens locked in the protocol, the cheaper the swaps, and the more trade volume (and fees) generated.
While Bancor was the first-ever protocol to enable AMM liquidity provision, there was always a point of friction for users: In order to provide liquidity, you had to do so with an equal value of BNT (Bancor’s Network Token) and the base ERC20 token. Uniswap removed the BNT token and used ETH in its place, so Uniswap users could provide liquidity in ETH + the base ERC20 token. This simplified the process for users (as more people hold ETH in their wallet than BNT).
However, liquidity providers to AMMs have over time begun to realize that taking on involuntary token exposure to any asset (whether it’s ETH or BNT) is not ideal. Not only because it requires giving up your long position on your tokens, but also because the price movements of multiple tokens in a pool leads to Impermanent Loss.
This is why Bancor, in the latest v2.1 version, has introduced single-asset exposure. Users can retain 100% exposure to a single token in the pool. So, for instance, if I want to provide $100 worth of LINK to a pool, instead of having to split half in LINK + half in ETH (Uniswap), or half in LINK and half in BNT (Bancor v1), I can provide just LINK and retain 100% exposure to my LINK tokens while earning swap fees. If LINK doubles in price, I am protected against impermanent loss, so the protocol will pay me $200 when I withdraw my LINK, plus the swap fees accrued.
These two key features — single-asset exposure and impermanent loss protection — offer LPs a unique value proposition that will help Bancor attract more liquidity and regain market share. And it is Bancor’s use of its protocol token, BNT, as the counterpart asset in every pool which allows for these novel features.
Follow this step by step guide to single-sided AMM staking on Bancor v2.1.
Q3: Bancor raised a record $153 million in 2017, with half the BNT in circulation. In August, ParaFi Capital invested in Bancor through BNT tokens. What were the main USES of the $153 million raised? Why did it restart a new round of financing in August? What is the exact amount?
The funds raised by the Bancor Protocol Foundation in 2017 continue to be used as grants to the core developer team and to numerous teams around the globe to develop the protocol. If you are interested in building on Bancor Protocol, we encourage you to contact us. The easiest way to get in touch is through our Telegram or Bancor’s WeChat.
Regarding ParaFi’s involvement, we did not start a new round of financing. ParaFi purchased BNT on the open market, and we added the team as advisors and chose to announce it given ParaFi’s deep expertise in the field of decentralized finance. We continue to work with the ParaFi in this respect. We are not able to disclose the amount of tokens purchased by the ParaFi team.
Q4: In July, Bancor proposed a solution to AMM’s impermanent loss, but after three months of inaction, it suddenly came up with a new solution, one that had nothing to do with the previous version. Why did this change happen?
Bancor piloted a dynamic weighting system earlier this year that utilized oracles in order to mitigate impermanent loss risk. The strategy was quickly adopted by several newer AMM protocols (including DODO and CoFix). However, we soon found (and have since seen other developers discover) that adjusting weights based on a price feed actually increases the risk of IL in certain market conditions, due to arbitrageurs moving faster than any price feed can update the pool.
So in Bancor v2.1, we decided to take a different approach. What if we could eliminate the risk impermanent loss for LPs by diversifying the risk across a wide array of pools?
And what if we allowed the protocol’s owners — BNT holders — to not only manage this risk, but also participate in the upside via swap fees collected by the protocol. We could potentially deploy a system where the revenue from swap fees exceeds the network-wide cost of impermanent loss.
Q5: What are the advantages of the current Bancor V2 protocol over other protocols in terms of solving the impermanent loss? How is it going? What are the potential risks for liquidity providers? Can you give me some example?
To our knowledge, Bancor v2.1 is the only true solution to impermanent loss. As mentioned above, other solutions we’ve seen can actually expose you to more impermanent loss in certain market conditions.
Here’s how Bancor v2.1 works: When an LP supplies liquidity to a given pool, the protocol Co-Invests BNT alongside liquidity providers, by minting an equal value of BNT to the other side of the pool(Keep in mind, every pool in Bancor uses BNT as its counterpart asset).
The protocol generates fees off its co-invested BNT, both of which are eventually burned whenever the LP withdraws their liquidity. The burned fees are used to compensate the LP for any impermanent loss on their stake when they withdraw, while any excess fees are used to reduce the overall supply of BNT (benefitting all BNT holders). Notably, a BNT holder can always provide their BNT to a pool, which effectively takes over the protocol’s position in the pool and burns the co-invested BNT along with its earned fees.
In terms of risks, there are a few important things to consider.
Protection against impermanent loss Vests over time — meaning, you are not protected from impermanent loss until you are staked in the pool for a certain number of days. The current default settings are as follows: 0% for the first 30 days, then 30% on day 30, then insurance increases by 1% per day (i.e., 40% on day 40), until day 100, when you get 100% protection thereafter. Keep in mind you can always pull liquidity out, but you would be subject to the same IL you might experience on a Uniswap or Balancer pool.
The second thing to keep in mind is in certain cases, if you experience IL, the insurance you receive when you withdraw might be paid out in part in BNT. Here’s such an example:
- stake $100 worth LINK
- LINK price doubles and LP stake now worth $200 (ignoring swap fees for simplicity sake)
- LP has experienced some imp loss, but has achieved 100% protection
- LP decides to withdraw and there are not enough LINK in the pool to compensate the LP fully in just LINK
- LP receives $170 in LINK and $30 in BNT
Learn more here.
Q6: An important aspect of the success of a protocol is the value capture of native token. Compared with the value capture of other DEX protocols, what are the advantages of Bancor V2 token in value capture?
V2.1 turns BNT into a Cash Flow Token, since the protocol now generates revenue off BNT it co-invests alongside LPs. Fees from these co-investments are effectively buybacks which benefit all BNT holders. In a way, we’ve figured out a way to generate value accrual to BNT without siphoning money from LPs. Something other AMM solutions are struggling to do with their own protocol tokens.
The key is to ensure protocol revenue exceeds the network-wide cost of IL protection. When protocol fees > cost of IL protection, the protocol profits.
Lastly, within every pool on Bancor, half the liquidity is always being provided in BNT, either by the protocol or by BNT holders. As liquidity on the network grows, and returns from trade volume increase, this drives demand for users to acquire BNT and stake it, in order to enjoy the resulting yield.
Q7: In June, the BancorNetwork V0.6 contract was found to be vulnerable. In 2018, Bancor’s contract breach led to substantial capital losses. What does the Bancor team think about smart contract security? What actions are currently being taken to address vulnerability related issues?
We’ve greatly expanded both our internal security team and our budget for regular audits. Bancor v2 has undergone 4 seperate audits, including audits by Consensys Due Diligence, Certik and Peckshield, as well as by a leading independent auditor. You can find those audits in the security section of our docs. In addition, we have supported third-party integrations with smart contract insurance providers like Nexus Mutual, as well as an ongoing public bug bounty to identify such issues.
DefiSafety finished evaluating us: https://defisafety.com/bancor-process-quality-review/
The score is 96% which only second to mStable’s 97%.
Q8: Community governance is very important to a decentralized protocol. What does Bancor plan look like on the road to community governance? How to expand the community?How to achieve deep user involvement in project governance?
Bancor formally launched its on-chain governance portal in September and since then we have seen a large amount of engagement by the community with numerous Bancor Improvement Proposals being discussed and voted on via https://gov.bancor.network/
Some of the best ideas have come from the community and we are thrilled to be building in the open with such brilliant minds.While anyone can join the discussion on Discourse, Discord or in our Telegram channels, in order to vote in Bancor governance, you must be staking BNT In a Bancor pool, which generates vBNT, Bancor’s new governance token. This design ensures that BNT liquidity providers and governance participants are one in the same. You can find more details here.
Q9: Some time ago, the YFI community proposed to invest some of the value captured in tokens into the future growth of the community, rather than simply destroy or buy back. How does Bancor think about investing in the community?
Yes, we are actively exploring a community treasury funded via protocol revenue, through which governance participants can vote on funding grants and other activities.
Q10: In the next two years, what goals do Bancor want to achieve in terms of team building, project development, community governance, algorithm research, market size, etc.?
Bancor has a thrilling year ahead of it, but the most immediate item on Bancor’s roadmap is to:
1) expand the current caps on pools to allow for far more liquidity to enter the system; and
2) launch Bancor’s liquidity mining, which will distribute BNT staking rewards to qualifying liquidity pools on the network.
These rewards will greatly enhance APY for LPs and in order to incentivize greater liquidity on the network. In addition, there will be unique features in BNT liquidity mining that will encourage LPs to re-invest their BNT rewards in pools. And unlike most liquidity mining rewards programs, users will not be exposed to impermanent loss, magnifying the ROI of collected rewards. We expect LPs to come for the rewards, and stay for the protection!
As the original creators of AMMs on the blockchain, the Bancor team will continue to develop bleeding-edge AMM technology, optimized for both liquidity providers and traders — including implementations of Bancor on additional chains and layer-2 solutions.
Questions from Chinese Community:
Q1：Will you corporate with Nexus Mutual, do some marketing events like Shiled Mining?
Nate：We currently have an integration with Nexus Mutual — you can take out smart contract insurance on your current stakes. You also think about our existing solution as “mining” protection — the longer you stay in the pool, the more you are protected against impermanent, until 100 days when you are fully protected thereafter.
Q2：What if current coverage could not cover loss?
Nate: it could, but this is offset by the fact that fees are always being burned for BNT. When fees collected from protocol BNT are greater than the cost of impermanent loss protection, the protocol profits.
We have modeled some scenarios here.
Here you can see protocol revenue exceeding liquidity protection cost protection after 3 months.
Q3: Bancor V2 is now behind its competitors in user numbers and transaction volume. How can Bancor attract more users in the future, like some new incentive mechanism?
Nate：Protocol incentives will help, and we will do this with BNT liquidity mining. We also believe the value of our impermanent loss protection & single-asset exposure is something very unique and will generate a lot of liquidity on the network.
For instance, we announced our first whitelisted pool yesterday (OCEAN) and within a few hours it had attracted $1.5M and was yielding 50–100% APR
Now it has at $2.7M and almost the size of OCEAN Uniswap pool.OCEAN holders don’t want to experience impermanent loss and split their OCEAN into ETH, so they are moving over to Bancor
Q4: AMM has been criticized for requiring a lot of liquidity to reach the same slide point as order book based exchanges. Can AMM be combined with order book based exchanges? What does Bancor V2 do?
Nate: AMMs may require more liquidity than order books but the key innovation is ANYONE can provide that liquidity. Order books can’t compete with that and it’s why AMM-based decentralized exchanges are beating order book-based DEXs by a lot.
This is the power of liquidity from the crowd, instead of a system that is controlled by professional market makers. Now anyone can store their idle liquidity in a pool and earn interet.
That said, there is some interesting research that combines order books with AMMs, and it is something we are heavily researching.
Q5：Since Bancor v2.1 don’t use amplification (v2.0) to lower slippage, what is the different approach to reach capital efficiency? Do you have any schedule?
Nate: We are working on ways to add liquidity amplification to certain pairs on Bancor. It is easier to stablecoin pairs.
Even without amplification, we think we can reduce slippage through large liquidity. If everyone stakes their idle tokens in Bancor pools (why wouldn’t they if there is no impermanent loss?), than the slippage will be very low.
Q6: Could Bancor add pair permissionless like uniswap in the future? When?
Nate: Bancor is already permissionless. It has been permissionless since 2017. Anyone can create a pair. Just go to bancor.network -> swap -> “create a pool”.
Here is AMA in Chinese version from Token Damo.
AEX: Your Best Asset Guardian
AEX is a global digital assets integrated service platform with information, community, financial management and digital currency trading since 2013.