week 38 in numbers: a 6.3-gwei spike, the same two pegs off, and $566m that went in circles

week two of the habit: the numbers from my own scripts, with the commands next to them, so you can rerun the week and tell me where i am wrong. week 38 ends for me on a tuesday morning again, with the same three tools and one public node.

gas, by the hour

gasweek pulled 50,400 blocks, a week to the block (25,932,130 to 25,982,529, tuesday 10:44 utc to tuesday 11:21 utc), in fourteen seconds this time. the base fee of the next block, in gwei:

the median came down from 0.071 to 0.061 gwei and the shape stayed: 04:00 to 17:00 berlin is a factor of 4.5, last week it was 4.2. the max changed: one block hit 6.3 gwei, three times last week's spike, and the half-hour medians touched 3 gwei once, 3.0 gwei on friday around 16:00 berlin. by weekday, saturday and sunday sit at 0.054 and 0.051 gwei and monday at 0.100, the priciest of the seven. a transfer on the busiest monday hour still cost less than a coffee.

nine pegs

stablepeg at block 25,982,530 (11:21 utc): the uniswap v3 spot price against usdc, a ten-minute twap from the pool's own observations, and coingecko next to it.

the same picture as last week, to the third decimal. seven of nine sit within fifteen basis points of a dollar; crvusd and gho are about ten basis points under, where they tend to sit. frax at 0.991 and lusd at 1.009 read the same on the pool, on the twap and on coingecko, and have for weeks. the curve 3pool quote for 100,000 usdt into usdc was 99,970.67, a spread of 2.9 basis points, and 99,984.74 for dai; the pool's virtual price stands at 1.0398.

one hour of large transfers

bigmoves over the 300 blocks before 11:21 utc (25,982,236 to 25,982,519), everything above $5m in usdt, usdc, dai, weth and wbtc: 46 lines, 26 of them with a name on at least one side, up from 16 of 76 last week, because morpho blue has a label now. the biggest:

• $98.4m usdc out of morpho blue and back within one transaction: a flash loan, folded into one line.
• 1,009 wbtc ($77.8m) borrowed from morpho blue in a single transaction, walked through three unlabeled wallets (0x6cbe…, 0xa0f1…, 0xb40d…) and returned, four legs in one block with nothing kept.
• the same 9,810.81 weth ($24.4m), borrowed from morpho blue and returned 16 times inside 180 blocks, to two addresses in turn (0x06cf… and 0xd226…): one bot running one loop, $390m of gross volume that never existed as a position.
• $41.0m usdt from binance 14 to binance 15: the exchange moving its own money between its own wallets.
• three fresh mints of $10m usdc each into 0xe69f…, forwarded to 0xee7a… within ten blocks, and $10.0m usdc into circle from an unlabeled wallet: issuance and redemption in the same hour.
• coinbase 10: $7.2m out and $7.1m in, ten blocks apart.

last week's pattern holds at a larger size. the 46 lines add up to $1.05bn of transfers, and $566m of that is 18 round trips, borrowed and returned inside one block, which a naive scanner reports as the largest moves of the hour. the transfers where money changed hands are the exchange shuffle, the mints and the coinbase hops, and together they come to a sixth of the loops.

rerun it

gasweek --tz +2 --json --svg week.svg
stablepeg --json
bigmoves --blocks 300 --min-usd 5000000 --json

python 3.10+, standard library, public rpcs. numbers from a public node drift a little between runs, a block later or a different endpoint; a different picture would be a bug, and i want to hear about it.

none of it is advice. the scripts are gasweek, stablepeg and bigmoves. last week's numbers are in week 37.