many people are calling for a hyperliquid type fee-on-volume. here is why it is not needed and why The Burn SIMD is sufficient imo. i believe average swap size per txn will probably go up a bit but not much over the course of the next few years if the onchain market thesis plays out. we may get huge blocks from institutions, but we will also get a lot more retail flow and the mean order size will be roughly the same order of magnitude. as such, taking 1 bp per swap now is going to be roughly the same as taking 1 bp per swap in a few years (on average). due to some optimizations, i believe compute resources per swap will go down a bit but not much over the course of the next few years. as such, resource fees per swap now is going to be roughly the same as in a few years. if you also believe these two statements, then you believe that resource fees per swap and bps per swap will be fairly correlated (and both go up with increased volume). given this correlation, it's best to just do resource fee for two reasons: 1) it reduces protocol and toll complexity. resource fee is already destined to be implemented and deployed, and taking fee on volume is INCREDIBLY messy and invasive on apps. 2) it naturally (and softly) incentivizes larger orders, which is a form of "vip" discount AND a form of IBRL