Here's a great thread on X by @mononautical:
So, tl;dr: Are pools losing money?So far they've collected an extra ~ 3.8 BTC in fees, at a cost of no more than a single stale block reward worth 3.15 BTC (which statistically was probably unrelated to Sub-1 Sat/vB Summer).But we can't know for sure until we see more data.
Mononaut starts by showing us the rise of sub 1 sat/vB transactions in blocks:
Then they go on to calculate how many fees such transactions have earned miners:
Since the start of Sub-1 Sat/vB Summer, pools have mined a total of 2,410,130 sub-sat transactions across 1776 blocks, earning an additional 3.69 BTC in fees for only 681 MvB of otherwise unused blockspace.
Allowing these transactions into their mempool also unlocked an extra 0.0956 BTC in CPFP revenue, where sub-1 sat/vb transactions were boosted up above 1 sat/vb by higher fee children.
However, there is a question of whether these sub 1 sat/vB transactions are leading to an increase in stale blocks (blocks that get mined, but then do not become part of the longest chain because other miners don't learn about them in time to start mining on them and end up mining on some other block, effectively leaving the original block as a stub, or a stale chain).
Mononaut found 11 stale blocks between 90,000 and 910,000, but 10 of them didn't include any sub 1 sat/vB transactions and so such transactions couldn't be responsible for causing those stale blocks.
One stale block at height 906343 mined by AntPool and reorged by Foundry (before they lowered their minimum relay fee) contained 1422 sub-1 sat/vb transactions, weighing ~274 kvB.
This reorg cost AntPool their entire 3.14505215 BTC block reward.
However, Mononaut points out that you can't really conclude that this stale block and loss of reward were caused by sub 1 sat/vB transactions because such transactions made up 18% of all transactions during this period, and so you'd expect at least 2 of the 11 stale blocks to include some sub 1 sat/vB transactions just because they were there.