ASTS market bars
Raw open, high, low, close, and volume for every completed ASTS session.
Freeze 1,334 completed ASTS sessions with SPY, QQQ, and IWM context into one auditable row per day. The clock matters as much as the values.
These readings update independently. They are not model inputs, deployment evidence, or a trading recommendation.
TECHNICALS THROUGH · TIMESTAMPED FALLBACK
At the end of session t, after the regular-session bar is finalized, the model will predict whether raw ASTS close at the fifth next trading session is above raw close at t. “Next” follows the exchange calendar, never calendar days.
Today's finalized close may be used to form a prediction, but the model cannot assume a trade executed at that same close.
News and fundamentals can wait. First build a synchronized price-and-volume baseline whose timing and provenance can be audited.
Raw open, high, low, close, and volume for every completed ASTS session.
Raw SPY, QQQ, and IWM bars joined on the same ASTS session date.
Expected sessions, holidays, and a five-session offset for the target.
Retrieval time, source URL, row count, and a SHA-256 snapshot fingerprint.
session_datedateOne joined ASTS session key
asts_open · high · lowdecimalRaw ASTS session prices
asts_close · volumenumberRaw target close and reported volume
spy_*_rawnumberRaw SPY market-context bar
qqq_*_rawnumberRaw QQQ growth-context bar
iwm_*_rawnumberRaw IWM small-cap-context bar
session_indexintegerAscending exchange-session sequence
snapshot_idtextOne SHA-256 fingerprint per asset
row_quality_statustextPASS, warning, or blocking failure
Holidays and weekends do not create rows. The label date is exactly the fifth valid session after t.
The target and market context use raw closes consistently. Cash-distribution price effects in the ETFs are documented and are not presented as total returns.
ASTS began trading under its current symbol on April 7, 2021. Do not silently splice predecessor SPAC history into the first model baseline.
A failed gate is a data issue to explain—not a row to quietly delete or fill forward.
1,334 joined rows cover April 7, 2021 through July 29, 2026.
ASTS session dates are unique, ascending, and create no join explosion.
All required ASTS and ETF close fields are present.
There are zero blocking data-quality failures.
Two non-model ETF field warnings are documented; all ETF closes remain valid.
The common 23-feature eligibility window contains 1,254 rows.
Exactly 1,249 rows have both complete features and a five-session label.
The final five completed sessions remain unlabeled until t + 5 exists.
All 740 later out-of-sample dates receive one unique prediction.
One missing SPY volume and one inconsistent IWM OHLC field remain visible as warnings. Neither enters the 23-feature matrix, and every required ETF close is valid.
One frozen 1,334-row table with common ASTS and market-context dates, snapshot fingerprints, and explicit warning status.
Session logic follows the Nasdaq market calendar. The downloadable workbook records separate source snapshots and SHA-256 fingerprints for ASTS, SPY, QQQ, and IWM. The NPA-to-ASTS identity change is confirmed by the Nasdaq corporate-action alert and the company's April 2021 SEC filing.
Available now in Model 1.