[PRE-GIP 2/3] START Gearbox Allowed Tokens and Protocols

The purpose of this proposal is to determine what protocols and tokens should be included in the initial GearBox Launch.

I propose the following protocols are included at launch. This list aligns with testnet. I believe we should be rather conservative with which protocols are added at first, as protocols are far more complex than single tokens. Note all of these except yearn only cover swapping (not LPing):

  1. Uniswap v2
  2. Uniswap V3
  3. Sushiswap same adapter as uni v2
  4. Yearn
  5. CRV 3Pool [NEW]
    edit - added adapter links

I propose the following tokens are included at launch (to be clear, these are tokens you could swap to - not tokens you can open a credit account with). This list aligns with testnet:

  1. WETH
  2. BTC
  3. USDC
  4. DAI
  5. LINK
  6. SNX
  7. yvDAI
  8. yvUSDC
    edit - removed ZRX, REPv2

I propose also adding the following tokens, which are in the top non-stable trading pair tokens by volume on Uniswap and Sushiswap.

  1. 1INCH
  2. AAVE
  3. COMP
  4. DPI
  5. FEI
  6. LINK
  7. UNI
  8. MATIC
    edit - revamped list per discussions below
21 Likes

Awesome! Ser please include contract links! (most can be found in the proposal template)

2 Likes
  • I would also add Curve 3Pool to be able to swap stables with minimal slippage (yeah, Uni V3 is quite competitive now, but have a choice here is a good, especially taking into account that later other DEXes can be integrated - so it’s good to test how pathfinder works etc)
  • re Tokens: I’d remove REPv2 - don’t see big volumes on Uniswap (Dune Analytics). So I would add some big DeFi tokens and top by volume which satisfy criteria mentioned in article
1 Like

I propose a little bit different list. It is created in accordance with principles and ideas specified in this article. Besides DEX liquidity, one of the main metrics was the presence at AAVE/Compound lending markets with substantial market size. SUSHI and CRV were excluded due to high volatility and the fact that protocol [probably] won’t be able to offer substantial leverage for these assets.

For each asset, I collected market data for 5 min volatility for 1 year and 3 months time periods and calculated Liquidation Threshold value.

Asset Max 5min volatility for the last 1year Max 5min volatility for last 3m Suggested LT
1INCH 11.50% 9.31% 77.5
AAVE 13.75% 8.74% 75
COMP 8.65% 4.41% 80
DAI 3.59% 2.74% 85
DPI 11.11% 11.11% 77.5
FEI 3.12% 3.12% 85
LINK 14.27% 6.69% 75
SNX 10.92% 9.91% 77.5
UNI 13.53% 8.65% 75
USDC 3.22% 3.22% 85
WBTC 5.47% 3.53% 85
WETH 9.34% 9.34% 85

Contracts and oracle price feeds source addresses for a selected list of assets is presented below:

Asset Address Oracle
1INCH 0x111111111117dc0aa78b770fa6a738034120c302 0x72AFAECF99C9d9C8215fF44C77B94B99C28741e8
AAVE 0x7Fc66500c84A76Ad7e9c93437bFc5Ac33E2DDaE9 0x6Df09E975c830ECae5bd4eD9d90f3A95a4f88012
COMP 0xc00e94cb662c3520282e6f5717214004a7f26888 0x1B39Ee86Ec5979ba5C322b826B3ECb8C79991699
DAI 0x6B175474E89094C44Da98b954EedeAC495271d0F 0x773616E4d11A78F511299002da57A0a94577F1f4
DPI 0x1494ca1f11d487c2bbe4543e90080aeba4ba3c2b 0x029849bbc0b1d93b85a8b6190e979fd38F5760E2
FEI 0x956F47F50A910163D8BF957Cf5846D573E7f87CA 0x7f0d2c2838c6ac24443d13e23d99490017bde370
LINK 0x514910771AF9Ca656af840dff83E8264EcF986CA 0xDC530D9457755926550b59e8ECcdaE7624181557
SNX 0xc011a73ee8576fb46f5e1c5751ca3b9fe0af2a6f 0x79291A9d692Df95334B1a0B3B4AE6bC606782f8c
UNI 0x1f9840a85d5af5bf1d1762f925bdaddc4201f984 0xD6aA3D25116d8dA79Ea0246c4826EB951872e02e
USDC 0xA0b86991c6218b36c1d19D4a2e9Eb0cE3606eB48 0x986b5E1e1755e3C2440e960477f25201B0a8bbD4
WBTC 0x2260fac5e5542a773aa44fbcfedf7c193bc2c599 0xdeb288F737066589598e9214E782fa5A8eD689e8
WETH 0xc02aaa39b223fe8d0a0e5c4f27ead9083c756cc2 0x5f4eC3Df9cbd43714FE2740f5E3616155c5b8419
YFI 0x0bc529c00c6401aef6d220be8c6ea1667f6ad93e 0x7c5d4F8345e66f68099581Db340cd65B078C41f4

Regarding the original list presented by @ov3rkoalafied.

First of all, what metrics it was based on?
Also, some assets look like not suitable for Gearbox services due to liquidity/farming opportunities:

  1. REPv2. It has very limited DEX liquidity (around $10k for -2% metric): Uniswap V2: REPv2 2 | 0x8979a3ef9d540480342ac0f56e9d4c88807b1cba. Almost 100% of liquidity and trading volumes are on CEXes: Augur (REP) price today, chart, market cap & news | CoinGecko
  2. ZRX. The same is applied to ZRX (limited DEX liquidity) - it is around 150k$ for -2% based on 1inch testing.
  3. SUSHI and CRV - too big volatility.
  4. yvDAI and yvUSDC - technical assets without developed secondary markets + I don’t understand leveraged farming opportunities for them.
  5. MATIC - it looks like all farming opportunities for these assets are on Polygon, while Gearbox operation will start on Ethereum.

In my opinion, we need to exclude these assets from the Gearbox whitelist.

I will be happy to discuss my list and other suggestions here.

10 Likes

wow, fantastic work. Absolutely agree that token volatility is a key metric to understand whether to add a token or not. I would probably increase LT for stablecoins a little. It seems that it is safe enough to take more leverage for DAI/USDC pair than for ETH/DAI pool. @vasilysumanov wdyt?

yvDAI and yvUSDC are technical assets, yeah. but they can be in the allowed token list if and only if Yearn yDAI and yUSDC vaults are in Allowed Contracts list. So, if Yearn get approved, I haven’t see any problems to add these tokens.

3 Likes

Of course, DAI/USDC is safer than ETH/DAI!

Regarding yvDAI and yvUSDC - yeah, it could work this way.

1 Like

Hmmm… I think Abracadabra has proven how beneficial adding some of these more volatile tokens is though. We can’t offer substantial leverage, but we can still offer leverage, correct? Defi users are degens, gearbox was built around degen score. Give them a volatile asset to leverage with expensive parameters and they will still likely use it. If anything, the high volatility is more potential for revenue going to gearbox. If we truly feel we can’t protect liquidators with the amount of volatility there is, I’m ok leaving those tokens out.

Re: the original list - the first 4 were all on the gearbox testnet. I am in favor of removing REPv2 and ZRX. Not convinced yet on SUSHI and CRV.

RE: matic farming - most of the assets we are adding won’t be able to be farmed initially (ie, only yearn strategies and uni/ sushi pools could really be farmed). Still, it doesn’t actively harm us to have in included and offers at least additional options for non-farm related strategies.Ie, I’m not actually seeing a drawback to including matic, compared to the rest of your list.

1 Like

To clarify - do we need to add contracts for every pool that we want to use on UNI/SUSHI? Ie, if a user wallet connects their credit account to Uniswap, and Uniswap is “approved”, can they use all Uniswap functionality? Or do we have to split it up and approve the DAI/WETH pool speficially, etc, etc?

I thought it was the former, but based on you suggesting an individual CRV pool (and including the specific Yearn vaults) it seems it’s the latter.

And - if we approve a specific pool contract, users would be able to both swap and LP with that contract thru gearbox? Or, would we need to also approve the LP token for that to be possible?

Excuse my lack of full technical understanding here! Thank you both for stepping in on this one!

1 Like

as far as I remember you approve UniswapRouter, so no need to add each trading pool.

The list looks good. I’d also suggest we add FRAX to the proposal. FRAX fits the preferences / requirements in the path to proposal medium post - 1.48bn circulating supply, 1.5bn of total liquidity in the FRAX-3CRV pool, #6 by liquidity on Univ3 and #16 on Univ2. FRAX also has a chainlink oracle and has been very stable since launch.

The Frax protocol also has substantial protocol controlled value that provides liquidity to borrowing markets on Aave, Rari and other platforms. If added to Gearbox, the Frax protocol is capable of providing substantial liquidity directly to Gearbox, which should be help things take off.

I also agree on dropping some of the low liquidity tokens from the list - adding them is probably more effort than it’s worth.

1 Like

I propose also adding the following tokens, which are in the top non-stable trading pair tokens by volume on Uniswap and Sushiswap.
OHM ,TOKEMAK, SPELL ,Convex CRV

5 Likes

I originally had OHM, SPELL, and CRV included. Per above, it seems like they are very volatile, which requires more conservative parameters. While I would like to push for all of these assets (and TOKE) in the future, I think it would be better for more of the DAO to get a hang of how everything works with the more conservative tokens that have been proposed for now.

Also, OHM is a rebasing token, which per the docs may not play nice with GearBox, so that should also be a separate proposal / investigation IMO. Also, OHM doesn’t make a lot of sense unless you can stake it from your gearbox credit account, OR use gOHM instead. Given gOHM is still in the transition period, I think it’s best to give them more time to fully move over to gOHM then asses if adding gOHM is possible in the future.

1 Like

Hi, very happy to be here! I wanted to propose adding the MIM stablecoin to the token list. It meets the criteria in the Medium post — MIM is a decentralized, over-collateralized stablecoin with high liquidity on Ethereum and other chains. It’s listed on Bitmart. It also has one of the highest-volume Curve pools.

对于项目组的决策我十分钦佩 做的滴水不漏 审时度势 真的很难得 良心团队 加油
在这个过程中我收货的不仅是经验 更多的是对良团队的专业性的认可 在我以后的工作学习当中一定会有很大的帮助

I agree ohm without staking does not make sense did not immediately think
for the protocol, the best is probably trading with leverage for multivat EVRO GB multi-currencies from https://www.angle.money/?
https://fixedforex.live/. ?keep3v ?

1 Like

My two cents for allowed tokens list:

  1. It’s easier to analyze short list that long one. DAO can add new allowed tokens at any time.
  2. All of these tokens have proved ERC20 contracts and looks safe. USDT, for example, has fee options which is not an issue, however, better to start without it.
  3. Appreciate for ov3rkoalafied work - this selection looks nice and make sense. However, during kovan tests users claimed they want to use many tokens for trading - so vasilysumanov list looks better suits this purpose.
  4. In vasilysumanov lists I cant understand LT options, they should be computes between two tokens. For example two stable coins could have LT ~90%, when DAI/WBTC ~ 60-80, cause high volatility between than.
  5. My advise is to start with lower LT and increase them over time. It could be perceived better in user’s eyes.
  6. Also, please check that UNISWAP V2 pool has enough liquidity. When account is closed, it automatically converts all asset on UniV2 compartible router, so small pools on UniV2 could cause big slippage.
  7. yvDAI and yvUSDC SHOULD BE ALLOWED if we decide to use YEARN DAI/USDC Vaults, and NOT as opposite. No economical sense to use them without allowing YEARN.
  8. DAO later could provide individual allowed token setup for each pool. If someone beleive we need to do that now and have separate lists - please, tell your args.
  9. Another significant thing to take into account: Chainling accuracy. To convert asset A into asset B, we compute the following: A-> ETH, than ETH → B, so double check final accuracy to compute Liquidation thresholds more carefelly. (Spoiler: ETH pool has better accuracy that others!)
2 Likes

For the protocol selection, all protocols from this list is audited to work with Gearbox properly. So, the major thing we should take into account - the real demoand from commuinty

Why no crv, cvx, mim, and spell in the list?

Probably in that order. Gotta get the flywheel going.

for crv, cvx, spell - high volatility, harder to set parameters. Can evaluate them individually after protocol launches and we verify everything works with more conservative assets>

mim - Hmm it does have a yearn vault, so as long as it meets all the other requirements maybe it could be included? Got a real nice APR on yearn through the crv vaults. However see mikael’s point #1 above, we can add it later.

1 Like

Why don’t we whitelist stable vaults of Convex?
I believe many choose Convex over Yearn.

Regarding vaults, I think that the yields for yvDAI and yvUSDC are too low compared to the borrowing interest rates and they can’t provoke enough demand. I believe vaults for FRAX, 3pool, alUSD, etc. will do.

Also, while it is not stable vaults, there must be a strong demand for Convex/Yearn vaults for stETH.

1 Like