Resolution boundaries

Measured, or not printed

The terminal operates on a single engineering principle: a figure is either derived from data the system actually holds, or it does not appear. There is no interpolation layer, no synthetic fill, no heuristic standing in for a measurement. This is a deliberate constraint, and it is what separates a research instrument from a plausible-looking one.

The consequence is that the envelope is sharp rather than fuzzy. Inside it, figures are exact and reproducible. Outside it, the terminal says so — and that boundary is documented here so you can position around it rather than discover it mid-trade.

Execution microstructure is out of scope

The data layer resolves price, session, volatility, fundamentals and reference statistics. It does not resolve order-book microstructure: there is no bid, ask, resting depth, spread, slippage surface or fill model for either instrument.

Order sizing therefore does exactly one thing, and states it: it converts a notional into units at the last price. That is an arithmetic identity, valid to the cent. A market-impact estimate would be a model of a book the system does not observe — a different class of claim entirely, and one this terminal does not make.

Instrument-side versus venue-side

Volume, liquidity and turnover in this terminal are resolved for the underlying share on its US exchange. Pool-side activity — traded volume, depth, trade count on the token itself — is not published by any source in the stack.

Reading this correctly is a genuine analytical edge rather than a caveat: a listing can show heavy institutional turnover on the reference side while its on-chain pool has been static for hours. Those are two different liquidity pictures, and the terminal keeps them labelled.

Comparative scope

The market scan resolves the registry along session performance — gainers, losers, range width. Cross-listing activity ranking is not part of that surface, so a question about which listing is “most traded” has no measured answer, and the resolver returns the boundary instead of an inference.

Correlation, not causation

The system quantifies how unusual a move is and whether the broad market moved with it — a listing at two sigma against flat benchmarks is measurably idiosyncratic. It does not assign cause. Published headlines are available and are labelled as commentary, never promoted to explanation.

Series alignment, verified

One boundary is worth showing rather than asserting, because it determines what historical questions the system can answer. The listing feed’s daily series and the reference source’s daily bars were aligned by trading day and compared:

symbol   sessions matched   identical   max divergence
NVDA     23                 23          0.00%
TSLA     23                 23          0.00%
AAPL     23                 23          0.00%

Daily history is therefore a single authoritative series, not two that might disagree — which is exactly what you want underneath a chart. It also means historical premium is not a derivable quantity: the premium lives in the live pool quote, and the historical record does not carry a second series to measure it against. The terminal resolves the present gap precisely and does not extrapolate a past one.

The envelope, at a glance

• The boundary is enforced in the resolver layer, not merely described here — the agent cannot reach beyond it, and is built to name the edge rather than approximate past it.