Unique Aurora | UA-RN-2026-01 v2.2
First publication: 2026-10-01. Editorial revision: 2026-10-12.
Observation date: 2026-09-28.
Research question: how do quoted spreads and displayed first-level quantities correspond in seven selected historical quotation records?

Sample conventions
clock: retained observation label; timezone and exchange-event timing are not established. IDs preserve record order, not equal time spacing.
last, mark, bid, ask, mid, spread: USDT per QNT.
bid_qty, ask_qty: QNT under the recorded base-asset convention, without contract-specification reconciliation.
S07 quantities: missing; an empty value never means zero.
display_buy_share: recorded percentage indicator; aggregation scope unspecified. It is excluded from first-level imbalance.

Calculations
mid=(bid+ask)/2; spread=ask-bid; spread_bps=10000*spread/mid.
imbalance=(bid_qty-ask_qty)/(bid_qty+ask_qty), for a positive quantity sum; dimensionless.
last_mark_bps=10000*(last/mark-1).
An aggressive buy is compared with displayed ask quantity; an aggressive sell with displayed bid quantity. The common 1 QNT reference measures first-level displayed capacity, not simulated execution.

Interpretation
Static quantity imbalance is distinct from event-based order-flow imbalance and is not a price-movement probability.
Last-to-mark difference compares two price measures; it is not a perpetual-to-spot basis.
Cross-field measurement times may differ. Records are not interpolated or time weighted and are not a representative market sample.
The analysis does not reconstruct full depth, hidden liquidity, trade events, transaction costs or realized execution.
Inputs, formulas, calculations and research scope are documented in the bilingual HTML study.