Why Are Some Polymarket Markets Illiquid?

• Structural thinness — honest low interest: few traders, wide spreads, shallow ladders from open.
• Terminal thinness — one-sided by design: the final-minute book of a settling Up/Down market where the winning side holds 42,000+ shares at 0.99 and the losing side has nearly nothing.
• Event-driven thinness — depth that leaves when a headline lands: the 26-second repricing where five markets hit lower bounds as depth departed.
• Meted-out thinness — sports books before lineups or state news; deep on new info, empty in between.

Across 17,880 one-second snapshots of resolved BTC five-minute contracts, both sides were never simultaneously empty — the book does not vanish. But the ask is empty in most final-minute frames: 90%+ of books end one-sided, 98.6% of winning-side quotes sit at 0.99 or better.

So what feels like illiquidity near settlement is actually the market having answered. Trying to trade it as liquidity — buying the 0.99 bid at 42,000 shares of size — is an execution error the recorded book exposes on sight.

Measure the two states separately: compute usable depth on your side only, and treat one-sided-terminal books as settled, not liquid. Recorded 250ms rows separate the states numerically (bidDepth vs askDepth per timestamp).

For structural thinness, spread analysis over history tells you when to stand aside; for terminal thinness, the answer is to already be holding the winning side. Both are research questions the full book answers and the mid cannot.

• Measure one-sided share for a week on a single Up/Down market (expect the high-90s texture) and for a sports series; the contrast is the "illiquidity" number in practice.
• Compare top-of-book sizes to the median trade (42,254 shares); the size x share product says how many trades a book absorbs.
• Count the final-minute ask-absent frames (the measured extreme is 3 of 18); the drain toward resolution is the venue's structural last-minute illiquidity.
• Replay a reprice and count the seconds the floor was effectively the whole book; illiquidity is worst exactly when price discovery is loudest.

The venue is illiquid the way a narrow ladder is illiquid: tick-constrained prices, mostly one resting side, and trade sizes that repeatedly exceed top-of-book depth — a combination that makes "illiquid" a structural description, not a complaint.

The causes are measured across the catalog: one-sidedness (90%+ of frames on the families), round-grid bunched pricing (98.6% at 0.99/0.995), trade sizes above top depth (median 42,254), and a final-minute drain that leaves ask presence at 3 of 18 frames on the extreme window.

The four-pattern answer generalizes beautifully across coins and sports; measure the one-sided share, the top-size-versus-trade-size ratio, the bunching rate, and the final-minute drain on any two series and the structural story reproduces without a single model.

Illiquidity on Polymarket is structural: thin, one-sided, ladder priced, and drained at the end, all measurable from the archive.

The answer to "why" is the four measured patterns, not a feeling — one-sided share, size-vs-depth, on-grid bunching, and final-minute drain.

Every liquidity fix (staged exits, maker quotes, size-to-resting-side) is downstream of those four measurements.

This page is the "why" appendix of the whole catalog: the four measured patterns (one-sided, bunched, thin, drained) it names reappear everywhere.

Read it after the liquidity and depth explainers and before the arbitrage or market-making pages, which assume the structural thinness.

Its four-pattern answer is the reference for every subsequent discussion of venue liquidity.

The answer is four measured patterns: one-sidedness, bunched pricing, thin tops versus trade size, and the final-minute drain.

Every liquidity remedy in the catalog is downstream of those four measurements.

• Order book data
• Price history
• Crypto data
• Provider matrix