The Most Important Problem to Solve in 2023: Exclusive Orderflow

This commit is contained in:
dr. mia von steinkirch, phd 2023-02-09 13:02:44 -08:00 committed by GitHub
parent 172df3fb2c
commit 5afbd3533c
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -2,27 +2,36 @@
<br>
<img width="802" src="https://user-images.githubusercontent.com/1130416/217937850-9b7e9434-ee72-4c2a-9e5e-ae9fe795310e.png">
### tl; dr
<br>
<img width="759" src="https://user-images.githubusercontent.com/1130416/217937850-9b7e9434-ee72-4c2a-9e5e-ae9fe795310e.png">
<br>
<img width="759" src="https://user-images.githubusercontent.com/1130416/217938181-3ea5947c-b847-4f4c-a41c-a40addceed8b.png">
<br>
### rpc-based "payment for order flow protocols" (POFPs)
---
### types of private order flows
<br>
#### rpc-based "payment for order flow protocols" (POFPs)
* this type of private order flows can be realized by modifying either the user's rpc or by modifying the protocol's frontend
* the rpc endpoint re-broadcasts (forwarding) the tx to nodes.
* in respect to every pending tx they receive, they can: 1) censor, 2) forward to some mempool, or 3) forward it directly to some block-builders or validators.
<br>
---
#### validator transaction reordering protocols (VTRPs)
### validator transaction reordering protocols (VTRPs)
<br>
* private tx reordering
* two goals: 1) identify valuable user txs, 2) replace the POFP's value-extracted tx with its own value-extractive tx
@ -32,25 +41,7 @@
---
### stat arbs
<br>
* if CEX is greater than the price on chain (DEX), stat arbs can buy on the DEX and hedge on the CEX -> confident that the DEX price will drift towards the CEX price, but not confident that will happen to the CEX price
* problem: settlement mismatch -> CEX trades are fast but blockchains aren't -> start arbs are unsure if the trade will land or competitors will beat them
* from the time between trade submission and receipt of the next block, the trade (their exposure to the assets) is uncertain
* the greater the uncertainty, the more the profit estimate needs to be marked down
* while volatility of the token is important, the longer the time period between blocks, the longer tx remains, and the longer it can go without being fully hedged
* as block time increases, stat arb MEV profit estimates get marked down -> payments made by stat arbers to miners/validators or to users/traders decrease
* conclusion: the battle between PFOF-style MEV could lead to increase block times for stat arbers, decreasing combined realized MEV
<br>
---
#### resources
### resources
<br>