Market-Time Report #1

What was measured

Window

2026-09-19 06:37 UTC to 2026-09-21 00:52 UTC

42.25 hours

Readings

35,130 across 15 pools

Largest hole in the record

03:16:22

Reported rather than interpolated across. Nothing in this report is filled in over it.

Every 60 seconds the collector reads slot0, liquidity, tickSpacing, fee and the initialised ticks spanning at least +/-60% around spot from each pool on X Layer, and appends the reading with its block, a hash of the payload and the payload itself. The figures below are those stored readings. `liquidity` is the in-range liquidity L at the observed tick, the quantity a swap consumes first; it is not a dollar depth, and it is not converted into one here.

Findings

In-range liquidity, pool by pool

The first and last reading of the window, and the range it moved through in between. A negative change means the pool ended the window with less in-range liquidity than it started with.

Scroll the table sideways for the range and the pool address.

Sources behind this report

What this report does not show

• The window observed so far is entirely outside the underlying markets' regular sessions. It therefore measures how liquidity behaves while they are shut, and cannot yet compare that with how it behaves while they are open. That comparison needs a session in the record and will be added once one is.
• In-range liquidity L is not executable depth in dollars. Kerb computes C(i) by walking ticks, which uses more of the stored state than L alone; L is used here because it is a single stored number per reading and needs no re-derivation.
• The record has a hole. It is reported in the window above rather than interpolated across.

Reproduce it

pnpm --filter @kerb/engine market-time-report regenerates this file from the observation store. Every row is a SELECT over append-only tables whose UPDATE, DELETE and TRUNCATE are rejected by database triggers.

The proof page carries the live row counts these figures come from →