spb/anomaly-atlas Public License

1---2project: anomaly-atlas3document: expA_data_reality/hypothesis4author: Simon-Pierre Boucher5contact: contact@spboucher.ai6data_source: hfmarketdata.io7created: 2026-08-128modified: 2026-08-129status: final10---1112# Hypothesis — expA_data_reality1314```text15Hypothesis16  hfmarketdata.io advertises 1min→1day bars across stock/etf/futures/index/17  fx/crypto plus daily options chains. We believe the advertised granularity18  is real, but that session semantics, timestamp conventions, adjustment19  behavior, missing-data patterns and response limits are NOT documented20  precisely enough to design honest experiments — they must be measured.2122Falsification criterion23  Not a statistical hypothesis: this experiment FAILS if any downstream-24  critical property (granularity, timestamp basis, adjustment arithmetic,25  row caps, completeness) cannot be pinned down empirically, or if the26  advertised 1-minute granularity turns out to be resampled/absent.2728Artifact null(s)29  None (no anomaly is claimed). This experiment EXISTS to seed the artifact30  taxonomy that later experiments must beat.3132Method33  Eight probe families (A1–A8) through the single cached client:34  inventory; history bounds per class (representative tickers AAPL, SPY, ES,35  SPX, EURUSD, BTC); intraday session structure incl. a sparse ticker (AIZN);36  daily-vs-1min aggregate comparison; AAPL 2020 4:1 split across all three37  adjustment series; row-cap + pagination check on one year of SPY 1min;38  latency profile; options coverage. All responses cached; manifest indexed.3940Result41  See analysis.md and results/expA_data_reality/20260812T054515Z/results.json.42  54 network requests, 274 571 rows, 0 retries; re-run = 55 cache hits,43  0 network requests.4445Interpretation46  Confidence n/a (no anomaly claim). Dataset is fit for Q1–Q3 research with47  five load-bearing caveats (see data_source_profile.md §9).4849Next experiment50  expB_artifact_baselines: build the measured null distributions for51  bid-ask bounce, stale prices, and non-synchronous lead-lag on THIS data.52```53