Ellipsis Finance BSC: Volume Spike

[bye43]

Two spikes in the billions

• Looks like the first on 5/29/2021 is due to the pricing of 0x160caed03795365f3a589f10c379ffa7d75d4e76
• Second one on 07/15/2022 is due to 0x6dbebd8fa691124e48829de1805be681920fbadf

[bye43]

@dhruv-chauhan Can you take this on too? Should be high priority (above multi and thena) as it is a prod subgraph

[dhruv-chauhan]

For the second pool, 0x6dbebd8fa691124e48829de1805be681920fbadf there's a mispricing of DUET token, will fix that.

For the first pool, 0x160caed03795365f3a589f10c379ffa7d75d4e76, found a bunch of strange transactions which swap $190,000,000 worth of BUSD to USDT and back multiple times within each transaction (possibly for rewards?). And the sum of inputVolume + outputVolume of these swaps combined is >35bil, as the subgraph reflects.

0x50b0c05dd326022cae774623e5db17d8edbc41b4f064a3bcae105f69492ceadc 0xc4d4156aab1fca85c99e85352b836274d3c53bafe98a2c9867b68950e1eafde9 0xb57acfeab13d52664416aa2ada9b490e340292731fced049fc8c4a730b7af700 0xcca1ebf01e694bb4c447f6018eebb34a3b829cff9ea1ec5fce236eb3cc2ef99c 0x7719e1bae25dbe80539edea37c962e941ec4141145e6eabe63540b7178ffd0d0 0xd790026feda9a16646647e9df0779dc4a7b173053369847691b8f3f678da1f66 0xf598e092ab82ce08798f9dab7ea6ade64f152aa91db897f3449b23ab591baa1d 0x7b3b727a56d1649ee325c42416a1199f4a9b4f4eb024a60b5848a7b1485953b1

@bye43

[bye43]

@dhruv-chauhan Ahh I remember now this is from the venus hack, so that is legit https://quillhashteam.medium.com/200-m-venus-protocol-hack-analysis-b044af76a1ae. We can leave that 35b spike on the first pool, my b. Thanks!

[bye43]

@dhruv-chauhan Just checking if there a deployment link for the misprice of 0x6dbebd8fa691124e48829de1805be681920fbadf. Thanks!

[bye43]

Context on fix for this issue: For the ellipsis volume spike due to a pool: 0x6dbebd8fa691124e48829de1805be681920fbadf Issue

token: dUSD

• no price obtained from existing pricelib
• hardcoded as stable here
• all liquidity drained on July 15, 2022 (attached screenshot), now trading at $0.000000003298

Now,

1. Easier fix is to remove it from harcoded stables list, so all it's USD values will be 0. Should not make a lot of difference, but will affect the protocol TVL by a little.
2. I think I can fix things like these by another pricelib upgrade, similar to what I'm thinking for optimism. But that will take some time. Image To give you a summary, basically while using Uniswap forks for pricing a token (say, tokenA), we currently use the default swap path as [tokenA -> WETH -> USDC].

Now what's happening for optimism, Uniswap fork there is Velodrome Velodrome does not have a lot of WETH pools, so the default path does not exist in most cases. Hence, price of tokenA is returned as 0.

What I intend to change there is remove this compulsory WETH pool condition, so ideally a path [tokenA -> tokenB -> USDC] should work too. Lastly, how this relates to dUSD on BSC for ellipsis,

Uniswap fork there is pancake swap There is no dUSD -> WETH pool, but there is a dUSD -> BUSD pool. And by the above logic, this can work to give dUSD price.