Here only includes atomic arbitrage and doesn't include stat arb, cross-chain arb, etc
Here only includes liquidation on Solend. The other protocols have a much smaller liquidation volume, except Mango and Larix, which had a bug.
Some liquidators also got rekt decent amount because they liquidated UST, sold UST in different transactions, and got rekt from Slippage, so the profit might be a little off.
May 15 • 20 tweets • 12 min read
$43m in total MEV on Solana 👇🧵
The collapse of LUNA / UST edition
I calculate the total arbitrage in the past 7 days following UST/LUNA collapse.
Here are what I found:
- total arbitrage profit: $2,743,624.75
- total liquidation volume $40,514,596.20
- total liquidation profit $2,025,729.81
May 2 • 15 tweets • 4 min read
2 of the most used blockchains Solana and Ethereum became *unusable* for a while yesterday.
Both actually share 2 sides of the same problem:
The Race for BlockSpace 🧵👇 2) For Context:
🦧ETH became *unusable* because it costs too much gas (1.5 ETH) to do anything as APEs bid for high gas to mint Otherdeed land
◎ Solana on the other hand got spam >100Gbps so bad by NFT bots for block space that the network had to shut down and restart
I wrote a custom script to ingest the data and calculate the total arbitrage in the past 3 months.
Here are what I found:
There were $20m of arbitrage profit made in the past 90 days. Here is the time vs. profit chart
Mar 14 • 4 tweets • 1 min read
1/ A Thought Experiment:
The Ship of Theseus of Blockchain and L2.
Blockchain has 3 comps: Execution, Consensus, and Data Availability (DA).
At first, L2 starts out as replacing just the execution layer and rollup on the Eth consensus layer.
There is also work being done like Celestia and Avail to replace DA but still relying on ETH as a consensus layer.
Finally, there is a solution like Arbitrum's Anytrust to replace ETH consensus with a committee and post block hash on ETH.
would love @aeyakovenko and @buffalu__ thoughts on this to ensure accurate information. #MEV#Solana
Solana currently only has an internal mempool of pending TXs and not external/public.
This means MEV opportunities only exist in terms of
- running your own validator
- inspecting TXs that come your way
- running your own MEV code on top (not easy but feasible)