Proposal: Improved Value Capture and Fast Lane POC

Expected on Snapshot Feb 5, 2023

TLDR:

  • Since the events of June 2022 the Bancor DAO has been working to reduce the deficit
  • V2.1 uses all of its revenue (network fees) to power the vortex
  • V3 uses 90% of revenue to collect BNT here: Bancor: Bancor Network V3 | Address 0xeEF417e1D5CC832e619ae18D2F140De2999dD4fB | Etherscan
  • DeFi involves many different ecosystem players that each perform some function and collect some payment
  • This proposes that if ANYONE in the community:
    • Can identify value leaving the protocol
    • Can identify a way for the protocol to recapture that value
    • Can do so in a way where all benefits are directly returned to the protocol/LPs
    • That community member should propose to the DAO their mechanism and anything required to allow it to run.
      • For example, this proposal will suggest a proof of concept (POC) for Bancor Fast Lane.
      • Bancor Fast Lane is a contract whose goal is to take the profit typically earned by external arbs and use those gains to reduce the deficit
      • To work, Bancor Fast Lane requires a new smart contract or upgrade to V3 contracts. Details below.

Introduction:

Before the events of June 2022, the Bancor DAO was working to make the protocol more efficient by fine tuning the trading liquidity and pool fees.

After the events of June 2022, there have been an abundance of innovative ideas from the community on how to improve protocol efficiency. For example:

  • Using Almanak for parameter suggestions
  • Dynamic fees
  • Protocol level arbs
  • Changes to the behavior of the vortex (who can trigger, what to burn, reward levels)
  • Integrations
  • Programmatic ways to define V3 settings
  • Dynamic pool weights
  • And more…

However, while many of these ideas show potential, there are some cases where small upgrades to the smart contracts can magnify the impact of the suggestion.

In such cases, it is proposed that the proposer outlines precisely how they can make the protocol more efficient and what is needed in order to do so.

Proof of Concept - Bancor Fast Lane

Any member of the community can make a proposal with the following recommended elements for consideration, as demonstrated here by the Bancor Fast Lane POC:

Identify Value Leaving the Protocol

  • Arbs in AMMs keep prices balanced with external markets
  • Arbs perform this function to profit when prices diverge across protocols
  • This proposal suggests that instead of allowing arbs to extract value from Bancor protocol with no competition, the protocol essentially also arbs itself and uses gains to reduce the deficit

Identify a way for the protocol to recapture that value

  • The Bancor Fast Lane (outlined in more detail below) will arb the protocol
  • All profits generated by the Fast Lane will be converted to BNT and then burnt (excluding a fee to whoever identifies the arb and sends it to the contract).
  • In addition to making the protocol more efficient, this is likely to reduce the deficit

Do so in a way where all benefits are directly returned to the protocol/LPs

  • The Bancor Fast Lane will not take any profits for itself. All gains are converted to BNT and burned (excluding a fee to whoever identifies the arb and sends it to the contract)
  • The finders fee will be 10% of the profit capped at 100 BNT (These are DAO controlled parameters)
  • This will be verifiable on-chain

Propose to the DAO a mechanism and anything required to allow it to run.

  • This proposal requires a new contract (or upgrade to V3) to allow for the flow below:
  • User finds an arb
  • User puts relevant inputs in the contract
  • Contract does a BNT flashloan
  • Contract does the trades exempt from pool fees
    (Note: an experiment will be done to test first without an exemption from pool fees to see if this is required. If it is found that it performs well without the exemption, an update will be shared on this post.)
  • Contract gives the wallet who called the function a finders fee (10% of the profit capped at 100 BNT - These are DAO controlled parameters)
  • Sends the rest of the BNT to the token contract where it is burned: Bancor: BNT Token | Address 0x1F573D6Fb3F13d689FF844B4cE37794d79a7FF1C | Etherscan
  • This proposal suggests making all V3 pool fees at least .5% (not including ETH, DAI, LINK, WBTC) as their pool fee changes are coming from Almanak (link)
  • Raising the pool fees allows the Fast Lane a significant advantage against generalized arb bots
  • It is expected to result in a decline in trading fees as LPs will not directly benefit from trades performed with 0 fees.
  • However, the arb-ing of the pools and burning of any BNT gained, is likely to result in better economic outcomes for LPs by reducing the deficit in the pools.

Fast Lane Functionality

  • Continuously search all pools on Bancor
  • Find potential arbitrage opportunities on:
    • Uniswap V2
    • Sushi
    • Uniswap V3
    • Curve
    • Dodo
    • Balancer
      • Note: Not all of these will be live on day 1 and this list is expected to expand with time.
  • Calculate optimal trades
    • This will start as a simple calculation that improves over time.
    • This must account for expected gas cost and trade fees to ensure the trade is profitable.
  • Inform the contract of the trade

Note: Arb performance should be expected to improve over time.

Implementation Components include:

  • New contract or upgrade to V3
  • Fee Changes
    • Please see the table in the appendix

For

Create a contract allowing trades exempt from pool fees as part of the Fast Lane and move all pool fees below .5% to .5% on V3 (excluding ETH, WBTC, LINK, DAI).

Against

Do nothing.

Appendix

Token Current Fee Proposed Fee
ETH Almanak Almanak
DAI Almanak Almanak
LINK Almanak Almanak
WBTC Almanak Almanak
1INCH 0.20% 0.50%
1ONE 0.20% 0.50%
AAVE 0.20% 0.50%
ALCX 0.20% 0.50%
ALEPH 0.20% 0.50%
ALPHA 0.20% 0.50%
AMP 1% No Change
ANKR 0.20% 0.50%
ANT 0.50% No Change
APW 0.20% 0.50%
ARCONA 5% No Change
ARMOR 0.20% 0.50%
AST 0.20% 0.50%
AUC 0.20% 0.50%
BAL 0.20% 0.50%
BAT 0.50% No Change
bBADGER 0.30% 0.50%
BBS 0.20% 0.50%
bDIGG 0.25% 0.50%
BMI 1% No Change
BNB 0.20% 0.50%
BOBA 0.20% 0.50%
BOND 0.20% 0.50%
BOR 0.50% No Change
BORING 0.30% 0.50%
BZRX 0.20% 0.50%
CHZ 0.20% 0.50%
COMP 0.20% 0.50%
COT 3% No Change
CRO 0.20% 0.50%
CRV 0.20% 0.50%
CTSI 0.20% 0.50%
DAO 0.20% 0.50%
DAPP 0.20% 0.50%
DATA 0.20% 0.50%
DDX 1% No Change
DEXE 0.20% 0.50%
DIP 0.20% 0.50%
DRC 0.20% 0.50%
DUSK 0.20% 0.50%
DXD 1% No Change
DYDX 0.20% 0.50%
EDEN 0.20% 0.50%
ELF 0.20% 0.50%
ENJ 0.50% No Change
ENS 0.20% 0.50%
eRSDL 0.20% 0.50%
EWTB 0.30% 0.50%
FARM 0.20% 0.50%
FODL 0.30% 0.50%
FOX 0.20% 0.50%
FRM 0.20% 0.50%
FTT 0.50% No Change
FXS 0.20% 0.50%
GNO 0.50% 0.50%
GRT 0.20% 0.50%
GTC 1% No Change
GUSD 0.20% 0.50%
HEGIC 0.20% 0.50%
HOT 0.20% 0.50%
HY 0.20% 0.50%
ICHI 1% No Change
IDLE 0.20% 0.50%
INDEX 0.50% 0.50%
INST 0.20% 0.50%
KTN 0.20% 0.50%
LPL 5% No Change
LQTY 0.20% 0.50%
LRC 0.30% 0.50%
MANA 0.20% 0.50%
MASK 0.20% 0.50%
MATIC 0.20% 0.50%
MFG 0.30% 0.50%
MFI 0.20% 0.50%
MKR 1% No Change
MLN 0.30% 0.50%
MONA 1% No Change
MPH 0.20% 0.50%
MTA 1% No Change
NDX 0.20% 0.50%
NEXO 0.20% 0.50%
NMR 0.50% No Change
NOIA 0.20% 0.50%
OMG 0.50% No Change
OPIUM 0.20% 0.50%
PATH 0.20% 0.50%
PERP 0.20% 0.50%
PHTR 1% No Change
PLR 0.20% 0.50%
POOLZ 0.20% 0.50%
POWR 0.20% 0.50%
PSP 0.50% No Change
QNT 0.20% 0.50%
QUICK 0.20% 0.50%
RAIL 0.20% 0.50%
RARI 0.20% 0.50%
REN 0.20% 0.50%
renBTC 0.20% 0.50%
renZEC 0.20% 0.50%
REQ 1% No Change
rETH 0.20% 0.50%
RLC 0.20% 0.50%
RNB 0.20% 0.50%
ROOK 0.50% No Change
RPL 0.20% 0.50%
SATA 0.20% 0.50%
SFI 0.20% 0.50%
SHEESHA 0.20% 0.50%
SMARTCREDIT 0.20% 0.50%
SNX 1% No Change
STAKE 1% No Change
SUSHI 0.20% 0.50%
TEMP 0.20% 0.50%
TOMOE 0.20% 0.50%
TRAC 2% No Change
TRU 0.20% 0.50%
UMA 0.20% 0.50%
UNI 0.20% 0.50%
UOS 0.20% 0.50%
USDC 1% No Change
vBNT 1.50% No Change
VISION 0.20% 0.50%
VITA 0.20% 0.50%
VLX 0.20% 0.50%
wNXM 1% No Change
WOO 0.20% 0.50%
wstETH 0.20% 0.50%
WXT 0.20% 0.50%
xSUSHI 0.20% 0.50%
YFI 0.20% 0.50%
ZCN 0.50% No Change
ZRX 0.50% No Change
LYRA 0.20% 0.50%
POOL 0.20% 0.50%
PDI 0.20% 0.50%
THALES 0.20% 0.50%
SHIBGF 0.20% 0.50%
CROWN 0.20% 0.50%
USDT 1% No Change
OCEAN 0.20% 0.50%
ACRE 0.20% 0.50%
6 Likes

If we are also including the Vortex pool (vBNT/BNT) then shouldn’t we also include a buy vBNT buy/burn in order to drive demand for BNT as well as lock vBNT into the Vortex now that ILP is disabled for stakers?

Who holds the keys? And what is the incentive to take on the regulatory/tax risk in return for no profit?

2 Likes

Chart should be under “FOR”?

2 Likes

I didn’t want to make people scroll dont he whole list.

It is now labelled as appendix.

Thanks!

2 Likes

I like it. We should start the creation of an Arb-Bot for Bancor that is given explicit advantages over the market. And we should use it to reduce the deficit.

I will think more on it, but it seems like a solid, ultra-low-cost medium-return type of decision.

6 Likes

Agree with @ZenoBNT . Overall sounds very positive for the protocol.

7 Likes

I’m a fan of all of this!

4 Likes

Very excited about this proposal and I fully support it.

5 Likes

Love this and support 100%!

2 Likes

Hi Everyone.

I made 2 major changes.

  1. There is no longer an exempt wallet. Now there is a smart contract that anyone who finds an arb can use. This contract will be exempt from pool fees.

  2. Whoever does this, can get a finders fee - 10% of the profit capped at 100 BNT

7 Likes

Shall we move forward with this? Seems promising!

3 Likes

Looks like it’s expected on Snapshot December 10

3 Likes

just a random suggestion here:
what if this arb just arb the protocol like any other arb and instead of collecting gains, it would split them in to:

  1. pay for the gas costs
  2. burn any access BNT gains

simpler to implement, faster to the finish line, make us all happy.

3 Likes

@teamAsaf if I understand correctly you are suggesting skipping the fee removal in the proposal?

The main problem I see is that the transactions would be much more likely to be front-run by generalized arbitrage bots, who would inevitably sell BNT profits & reduce/eliminate any benefit gained by this. Also the bot would be continually losing ETH in gas fees that it wouldn’t recover due to transactions being consistently reverted.

3 Likes

Hey All,

Made a video about just this issue: https://youtu.be/hPKBzJOaG_E

That is my worry as well.

4 Likes

Excellent video! B3 does support flashloans :sunglasses:

2 Likes

Is this still getting fleshed out or was this set to hit Snapshot today?

2 Likes

I moved the snapshot date to next week and might push it back further.

This proposal is fairly detailed, so I want to make sure that nothing here will need to be changed during the implementation process that would require another vote.

2 Likes

:sparkles:Hi Bancorians! :sparkles:

As ALMANAK is currently involved in one part of the innovative ideas @foxsteven proposed (Almanak ongoing recommendations), we would like to highlight some important considerations in terms of risks and opportunities:

  1. Technicity of arbitrage could be a limiting factor to community usage

Even though Bancor should incentivize users to interact with the protocol and are keen to reduce the deficit, relying on the userbase to find profitable transactions can be cumbersome and does require additional knowledge on how to make the most profitable call with multiple external sources.

  1. Potential negative externalities to Bancor users
  • Arbitrageurs exploit of the Fast Lane bot

An additional source of risk of less benefits would be the integration of auto-calling this function by arbitrageurs, seeking to extract value, as the risk-reward is high enough. In fact, since any user can interact with the contract once they identified an arbitrage, professional arbitrageurs can exploit it really easily, given they already have all the tech stack in place and their game scales with volume and number of arb transactions… It is not bad in itself, but to be considered.

  • Profit cap too high to prevent from arbitrageurs exploit

Little to no arbitrage is more profitable than the maximum 100 BNT proposed as reward (source). We see a risk of too high an incentive for professional arbitrageurs leaving this contract as a source of more profits for those already in the game (market makers) rather than the interested and active Bancor user. By lowering the cap, this would help discourage those players.

  1. Moating strategy

What is the moating strategy for the new smart contract? A potential downside would be the open source of this contract and therefore an expected behaviour to be exploited by arbitrageurs

  1. Flash loan cost coverage

An additional problem could be a potential loss happening through the flashtrade: who will cover for potential costs of missed arbitrage & gas usage and is there a fail-safe to avoid it happening in the first place? Defining a detailed specification to ensure a robust usage is already part of this proposal and we suggest considering this outcome for any further development.

  1. Opportunity for inter-pool rebalancing

Lastly, an aspect to be covered when considering protocol-owned arbitrage is the opportunity of rebalancing pools with triangular trades. Triangular trades include passing through (at least) three different pools to gain a profit over all transactions. The Fast Lane bot could be a lever for Bancor to rebalance its pools and adjust for better trade execution. This does need further consideration how these strategies could be incorporated but would open up a path for more user-friendly trades along the way.

Almanak Team.
@Michael_Almanak @Nano @almanak-philipp 55-Stephan

3 Likes

Stephan, thank you for your thoughts. I wanted to address some of your points:

  1. You’re right that not many community members will be able to run arbitrage strategies through this. If even a couple do, however, it is a substantial win for Bancor.
  2. Can you explain what you mean by arbitrageurs exploiting the Fast Lane? The way I see it is that if arbitrageurs (or anyone for that matter) use the contract, it is a significant win for Bancor. I don’t see why the 100 BNT maximum reward is problematic, or how lowering it would discourage anyone.
  3. The contract is open for anyone to use. Using it is beneficial for Bancor - regardless of who uses it. If you are alluding to smart contract risk, I see the worst-case exploit from this contract is that someone is able to make a few trades without paying fees.
  4. What do you mean? Given this is open for anyone to use, the contract caller will pay for their own gas usage like any Ethereum transaction - regardless of the outcome of the transaction. The nature of flash loans prevents a loss (beyond gas costs) as the transaction would fail.
  5. Could you explain further what you mean? The nature of triangular trades will rebalance the pools, aligning prices on Bancor more closely with other AMMs, but without bleeding value to arbitrageurs to do so - are you suggesting something beyond this?
3 Likes