Can you enter — and still exit?

Local observations since this page opened. Missing periods remain gaps. Green shading marks the best observed window; red points mark a 30% depth deterioration and the last point marks the latest observation.

Best window uses approximately 15 continuous minutes. Its aggregate is retained until reload; raw observations retain only the last 4 hours. All subscribed live markets are sampled in memory while the page is visible. Accelerated demo observations are separate and discarded on exit.

“Usual this session” means the median of observations collected in this browser during the selected window. Nearby orders means same-side visible depth within 5 bps of mid. These comparisons do not predict prices or profits.

Reference size is 25% of the first usable same-side 5 bps depth, fixed until reload. Each market and side has its own reference. Scores compare this session, not markets or profitability. Until usable depth appears, quality is zero; collecting a reference restarts the sample.

Score weights: spread 20%, depth 25%, reference-order impact 25%, fillability 20%, stability 10%. Components are clamped to 0–1: spread = 1 − spread/10 bps; depth = current / initial 5 bps depth; impact = 1 − reference slippage/20 bps; fillability = filled/requested; stability = current / previous 30-second median depth. Missing stability scores zero.

Comparisons need 30 observations over 3 minutes with 80% coverage. Partial windows are labeled. Ties use midpoint percentile ranks. Reference cost comparisons exclude partial fills and apply the current fee to all samples. TRADE NOW requires quality ≥70, percentile ≥60, full fillability and additional impact ≤2 bps. Quality below 45 or 30% depth deterioration means WAIT. These are transparent heuristics, not validated predictions.