What the publication is
QAXUS Intelligence is the research publication of QAXUS Corp. It publishes dated perspectives on Bitcoin, markets and macroeconomics. Published material combines provider data, automated analysis and language-model synthesis. It may contain errors and should be checked against its dated sources before it informs a decision.
From observation to publication
The research workflow gathers market and economic observations and keeps generated specialist analysis, opposing interpretations and synthesis in private research records. Full public editions pass an automated evidence review after specialist analysis, opposing interpretations and synthesis. A separate model checks every displayed text block against dated source records; code checks coverage, quoted evidence, numeric consistency and source timing. Failed claims go through bounded correction and a fresh review before publication. An unresolved review keeps the new draft private and leaves the existing archive edition online. A report’s own method disclosure records what is known about that report; the intended workflow does not establish that every historical stage completed.
Historical audit evidence identified a period in which analyst and debate stages were absent after an upstream authentication failure while briefs continued to publish. The original evidence remains preserved. Historical corrections preserve the public archive and original publication dates. Dated market records and primary documents support corrections; later historical reconstructions are identified, and unavailable evidence is stated explicitly. The archive does not represent this period as uninterrupted completion of the full research process.
Sources and observation times
Sources consulted lists URLs retrieved through the named data or search provider, with available observation and retrieval times. It records the inputs consulted and does not establish independent verification of every claim. Model-supplied source URLs must match a successfully gathered public input before a new automated report can be published.
The time a page is rendered is not the time the underlying market observation was measured. Missing observations remain unavailable rather than being displayed as zero. Historical reports without a recorded public source list disclose that limitation beside their text.
Outcome calculations
Outcomes compare daily closes at 1, 7 and 30 calendar days. The starting price is the daily close on or before the brief’s publication date, looking back at most five calendar days. The horizon price is the daily close on or after the target date, looking forward at most five calendar days. A horizon matures at 23:59:59 UTC on its target date. This convention is not an executable entry or exit price.
A range hit means the measured price or percentage change falls within the numeric base-case interval, inclusive of its bounds. A missing or unparseable range is unmeasured, never a miss. MARKETS percentage forecasts are compared with the 1-day percentage change; their bounds are displayed as percentages. The 7- and 30-day range scores remain unmeasured.
The range-hit rate is hits divided by measured ranges. Each table shows that denominator, along with evaluated and matured counts. Price change is (horizon close ÷ starting close − 1) × 100. It excludes fees, slippage, position sizing and any trading strategy, and is not a portfolio return.
Corrections and the historical record
Corrections preserve the original publication time and update the report revision. The track record includes only reports whose forecast narrative has explicit editorial approval. Unreviewed historical forecasts, withdrawn reports and factual snapshots are excluded; their stored evidence is retained. An earlier implementation capped the fetched outcomes at 1,000 rows and omitted MARKETS; the current tables remove that truncation and show MARKETS with its scoring limitation. Historical figures from the two implementations may differ for those reasons.
Send a correction through the contact form, including the report URL, the disputed claim and a supporting source. QAXUS does not publish a guaranteed response time.
Limits and responsibility
Forecast probabilities and scenario ranges express the analysis at publication. They are not guarantees, and range hits alone do not establish predictive skill against a baseline. Coverage, market conditions and the number of measured observations change over time. QAXUS may have interests in assets it discusses; the public archive is not a disclosure of private holdings.
This research is informational and does not consider your financial circumstances. The publication does not identify a verified individual author or promise human review of every brief.