Ranges breakouts and failed breaks

EN source master · P05-L03 · 30 minutes estimated · needs_review

Offline formative study. No wallet connection, real funds, private keys, signatures, live trade or personal portfolio inputs. Visuals are specifications. This source master remains needs_review; completing the formative exercise does not issue certification.

A historical chart can teach the difference between crossing a boundary and continuing outside it, provided the data and chosen rule remain visible.

Identify the record

The frozen package retains seven dated FXUSDCAD values visible in the retrieved official CSV search rendering [BOC]. The direct CSV fetch was unsupported; this is not a byte-identical publisher archive. The Bank describes its rates as indicative daily averages [BOCM]. They are not closes from a tradable venue. The teaching case therefore tests reference-rate observations, not executions or crypto candles.

Freeze a deliberately small range

Use only December 30–31 to define [1.4379,1.4389] CAD per USD. In this original rule, an upper crossing is any later published rate strictly above 1.4389. A failed upper persistence hypothesis means a later rate is at or below that upper boundary. This includes a move below the entire initial range; it does not require returning inside both bounds.

Walk forward in observation order

January 2 and 3 exceed the upper boundary. January 6 is below both original bounds, so upper persistence fails. At January 2, January 6 must stay hidden. Replacing the formation window after seeing January 6 would be hindsight. Holidays and weekends without records are missing publication dates, not zero prices.

Bound the conclusion

This retrospectively chosen tiny window demonstrates a descriptive test, not statistical validation. Different formation dates or tolerance would change crossings. Publication averages hide intraday paths and cannot establish a tradable stop or gap. No false-break frequency or predictive edge is estimated. The original sample's parameter dependence, delayed publication and possible later publisher revisions must accompany any annotation.

• Upper crossing: a sampled value exceeds a frozen upper bound.
• Persistence hypothesis: later observations stay above that bound.
• Failed persistence: a later observation violates the defined condition.
• Reference rate: publisher's indicative average.

HISTORICAL source-rendering excerpt, CAD per USD: 2024-12-30 1.4379; 2024-12-31 1.4389; 2025-01-02 1.4418; 2025-01-03 1.4442; 2025-01-06 1.4348; 2025-01-07 1.4340; 2025-01-08 1.4390. See docs/academy-2/datasets/P05-L03-BOC-USD-CAD-20250108-v1.json.

Historical reference-point line plus frozen first-two-point range; label Jan2/3 upper crossings and Jan6 failed upper persistence below lower bound. Catalog's annotated-candlestick family is adapted to points: no fabricated OHLC or volume. Hide later points in the Jan2 information panel.

Dated values in the retained search rendering and the original boundary classifications computed from them.

Full-file authenticity, intraday candle path, executable fills, a return inside the full range or a predictive strategy.

• OBSERVED: copied source rendering contains Jan2 rate 1.4418; scope is rendering, not a captured full CSV.
• INFERRED: Jan6 violates upper persistence under the frozen author rule.
• UNKNOWN: Jan2 intraday high/low and executable spread.
• INSUFFICIENT EVIDENCE: upper crossings provide a profitable directional edge.

• Calling daily averages candles.
• Calling below-range failure an inside-range return.
• Using later outcome at Jan2.

Freeze the two-point range. Classify Jan2, Jan3, Jan6 and Jan8 against it. Calculate Jan2 excess above the upper bound. Write what could be known with Jan2 as the cutoff and one missing source requirement for execution research.

Deliver calculations or annotations, claim/source table and limitations. Suggested allocation: study 12 minutes, exercise 8, correction/quiz 10; estimate subject to calibration.

• Identify series and unit.
• Preserve source-rendering limitation.
• Separate formation/cutoff/outcome.
• Distinguish upper failure from inside-range return.

The historical record supports a bounded sampled crossing-and-failure exercise. It does not supply candles or predictive performance.

• Keep retrospective formation separate from later observations.
• Report failure without inventing intraday events.

P05-L04 after correction review.

NONE in the authoritative catalog. The supplied offline fixture/package is sufficient; no paid feature or unverified Production capability is required. Lab/certification metadata denotes downstream associations, not access gates or live awards.

• [BOC] Bank of Canada — selected official CSV search rendering — Seven selected dated USD/CAD observations visible in retrieved search rendering; full CSV fetch unsupported. Checked 2026-10-01; locator turn8search16.
• [BOCM] Bank of Canada — Exchange rates — Published daily rates are indicative averages, not executable OHLC. Checked 2026-10-01; locator turn11view7.