kStable Peg Integrity · Live Market Data

kStable Tracking-Error Explorer

Deviation of every kStable FX pool's price from its on-chain oracle target and from the independent market FX rate, in basis points (1 bp = 0.01%), across Base and Arbitrum. Pool vs oracle is our own mark; pool vs market is the independent external check. A pool being actively recentered by the keeper is the system self-healing — healthy, not broken.

Loading live data from api.bankcoin.capital…

Chain

Pool (currency pair)

Window

Histogram TE definition

Auto-refresh 60s

Three-series overlay

Pool price vs on-chain oracle vs independent market FX

Tracking error over time bps

Signed deviation of each series pair; dashed line = 0 bps (perfect peg). 10-min bin means.

RMS tracking error headline metric

RMS = √(mean(TE²)) over the window — penalises magnitude regardless of sign (the number a risk desk looks for). Sample-weighted, pooled exactly from the 10-minute bins.

Tracking-error distribution the headline

Binned mean tracking error, with mean / median / RMS markers

Slippage vs trade size

On-chain QuoterV2 execution cost across order sizes (latest snapshot, both directions)

Institutional tracking-error statistics

All pools — current tracking error

Latest 10-min-bin RMS(pool vs oracle) per pool on the selected chain. Click a row to inspect it above. Colour bands: on peg ≤ 10 bp, drifting 10–100 bp, above band > 100 bp (often actively recentering).

On peg ≤ 10 bp Drifting 10–100 bp Above band > 100 bp

Pivot & explore — 10-minute TE bins (selected pool)

Drag fields to pivot by definition / bin; switch renderer to Plotly line, bar or heatmap.

What you are looking at. Raw ~60-second observations of each pool are pre-bucketed server-side into fixed 10-minute bins (mean / RMS / min / max per bin); the browser downloads bins, never raw rows. Three tracking-error definitions, all signed, in bps: pool vs oracle = (pool − oracle)/oracle ×1e4 — the AMM price vs our on-chain oracle target (keeper-lag / LVR signal, our own mark); pool vs market = (pool − market)/market ×1e4 — the all-in dislocation a taker experiences vs the independent market FX rate (the honest external check); oracle vs market = (oracle − market)/market ×1e4 — the peg integrity of our oracle itself vs the real market. RMS = √(mean(TE²)) is the more generally accepted institutional measure — unlike the mean (where positive and negative deviations cancel and can flatter a wandering peg), RMS penalises magnitude regardless of sign. Cells with fewer than 30 samples are shown as insufficient data, never fabricated. High tracking error on a pool the keeper is actively recentering is the system working as designed — a mean-reverting correction, not a broken peg. This is public market-data transparency, not a financial-product offering, and carries no offer, solicitation, or advice.