Analyze liquidity
How to Analyze Polymarket Liquidity
Liquidity analysis on Polymarket is the study of the executable side: how much rests at the bound, how often the other side is empty, and how long the spread stays wide. This page is the analysis recipe built on the recorded 250ms grid.
Figures measured as of 2026-09-24 on the published PolyOrderbooks archive.
Metrics
The five numbers to compute
Method
Compute from the right layer
All five metrics read from one source: the full-L2 rows at 250ms with prices and metrics on the same grid. Aggregation has a price — a per-second series hides the one-sided sub-second states, so the honest analysis reports the raw-rate along with any summary.
The metrics endpoint and the raw books agree by construction (both derive from the same frames); a study should say which layer it computed from.
Worked example
A liquidity profile for one market
Take a resolved 5-minute market and compute per-minute executable depth and one-sided rate from its ~1,200 frames. The profile shows the market as traders met it: two-sided early, one-sided in the final minute (76.2% baseline), deep at the bound at settlement.
That profile is the honest input to any capacity and slippage estimate on the venue.
Honest
The honest limits
Liquidity shown is liquidity resting, not liquidity trading: the honest analysis pairs depth frames with the tape that consumed them, because a deep book that nobody trades and a thin book that turns over are different liquidity pictures with the same summary.
FAQ
How do you measure Polymarket liquidity?
Depth at the executable side, one-sided rate, spread persistence, cross rate, and bound rest — all from the recorded 250ms grid.
What is normal one-sidedness?
The venue baseline is 16.9% of 5-minute snapshots, rising to 76.2% of final-minute books.
What separates real from fake liquidity?
Consumption: pair depth frames with the tape that traded; resting size that never prints is a different asset than flowing size.