How Macrotome sources and processes its data — no black box.
Central bank data — rate decisions, statements, press conferences, and projections — comes directly from official publications: the Federal Reserve, European Central Bank, Bank of England, Bank of Japan, Reserve Bank of Australia, Reserve Bank of New Zealand, Bank of Canada, and Swiss National Bank.
Positioning data (COT) comes from the CFTC's weekly Commitments of Traders report, the same public dataset used by institutional desks.
News is aggregated from a curated set of forex and financial news feeds, refreshed automatically every two minutes.
Every headline is classified into one of three tiers based on keyword matching against known market-moving terms:
HIGH — rate decisions, inflation prints, employment data, and named central bank officials speaking publicly.
MEDIUM (CB) — central bank-sourced headlines and secondary macro data (PMI, retail sales, trade balance).
LOW — everything else relevant to FX markets that doesn't meet the above criteria.
Each central bank meeting is scored on a five-point scale from very dovish to very hawkish, based on the actual decision (hike, hold, cut), the vote split, and the tone of the statement and press conference relative to market expectations going into the meeting.
This score determines the bias shown for that currency until the next scheduled meeting.
Retail positioning is sourced from Myfxbook's community outlook data, covering long/short percentages across 43 pairs including FX majors and crosses, commodities, crypto, and indices.
When positioning on one side exceeds 75%, we flag it as a contrarian signal — historically, retail traders are more often wrong than right at positioning extremes. Between 60-75%, the signal is shown as a milder lean rather than a full contrarian flag.
News feed — every 2 minutes
Retail sentiment — hourly
COT positioning — weekly, following CFTC's Friday release
Daily report — every evening at 22:00 UTC
Weekly report — every Sunday at 21:00 UTC
We're working toward a single confluence score per currency, combining central bank stance, positioning, and sentiment into one number. We won't publish a score until the underlying methodology is solid enough to stand behind — and when we do, the calculation will be documented here in full.
If anything here is unclear, or you'd like more detail on a specific data point, reach out at macrotomefx@gmail.com.