How we score currency strength
Every score on BearPaws is derived from primary data and computed the same way every time — no discretionary nudging, no black box you have to take on faith. This page explains exactly how the numbers are built, where the data comes from, how often it refreshes, and — just as importantly — what the scores can't do.
The framework runs in four layers, moving from a single currency's fundamentals out to how a pair lines up and whether independent signals confirm it.
The four-layer framework
- 1Fundamental valuation. Each currency's macro releases — interest rates, inflation, growth, employment, and trade — are turned into a single intrinsic-strength score. The model is intrinsic-first: it weighs the pressure a currency is under and anticipates how its central bank is likely to respond, because those second-order policy reactions (cuts and hikes) drive the larger, more durable moves than the first-order data surprise itself.
- 2Cross-pair alignment. Selecting a pair weights each leg's intrinsic strength by relative market-force intensity to produce a cross-pair score. A wide, aligned gap means the two sides genuinely agree on direction and confidence is higher; a narrow gap means the pair is closer to a coin flip, whatever either leg looks like alone.
- 3Sentiment & bias. Risk-on versus risk-off is read from real futures performance — one-month and three-month windows, with the longer window weighted more heavily so a single-day spike doesn't swing the read. News sentiment supplements this for shifts that haven't reached the futures data yet.
- 4Confirmation. The fundamental read is cross-checked against where institutional money actually sits in the weekly CFTC Commitments of Traders report (non-commercial positioning), and against breaking news scored per currency. Confirmation raises confidence; divergence is flagged rather than hidden.
Gold (XAU/USD) runs through the same framework as the eight majors — its own strength score, positioning, and risk read — not the chart-only treatment most tools give it.
How the fundamental score is built
The intrinsic score blends two components. A macro component reads the directional consensus across a currency's available indicators — each release is normalized so that a stronger reading pushes the score up, with the direction carrying the weight and the relative magnitude modulating it within a capped band so no single print dominates. A momentum component captures recent price behaviour over rolling multi-week lookbacks.
Negative scores indicate deflationary pressure, positive scores inflationary pressure, and near-zero a balanced backdrop. The split is auditable: each currency's momentum and macro contributions are stored and exposed alongside the headline score, so a reading can always be traced back to its inputs.
Where the data comes from
We build on primary, mostly official sources and normalize everything ourselves. No scraped forums, no anonymous “leaks.”
- U.S. Federal Reserve (FRED) — Interest rates, inflation, employment, and other U.S. macro series.
- Eurostat · OECD · national statistics (via DBnomics) — Non-USD macro indicators for the euro, sterling, yen, and the commodity dollars.
- CFTC Commitments of Traders — Weekly institutional (non-commercial) positioning across the majors and gold.
- Real-time market-data feeds — Spot prices, daily candles, and futures performance for the eight majors and XAU.
- ForexFactory economic calendar — Scheduled releases, forecasts, and prior readings.
- Major financial newswires — Reuters, FT, FXStreet, and others — filtered, scored, and tagged per currency.
How often it updates
Currency strength is a weeks-to-months signal, so the refresh cadence is matched to each source rather than chased tick-by-tick. Prices and pair snapshots refresh through the trading day; the economic calendar and news refresh on a short cycle; CFTC positioning updates weekly with the official Friday release; and macro indicators refresh as each statistics agency publishes. Pages are served from cache and rebuilt on these intervals, so the data you see is current to its underlying source — not artificially “live” on numbers that only move weekly.
How we use AI
AI helps us work faster — scoring the impact of breaking news per currency and drafting recaps and outlooks from the day's data. It is a tool, not the author. It never invents prices or figures: every number traces back to the sources above. When a piece is wrong, that's on us, not the model — and a person reviews what we publish.
What the scores can't do
We'd rather you know the limits than over-trust a number:
- Macro coverage is uneven by currency. The U.S. dollar has the deepest official-data coverage; the euro, sterling, yen, and commodity dollars are covered through public European and OECD statistics. Where a currency lacks fresh macro data, its score leans more on price momentum until that coverage improves.
- Some indicators are proxies. A few survey-style inputs (such as PMI) use the closest freely available public series because the licensed originals are proprietary.
- Scores describe a weeks-to-months horizon. They are not timing tools, entry/exit signals, or price forecasts — markets can move against a fundamental read for a long time.
- Stale data degrades gracefully. A freshness guard drops macro series once they age past a set window, so a frozen upstream source quietly reduces its own weight rather than poisoning the score with old numbers.
BearPaws is market context, not advice. For the full risk picture, read our risk disclaimer.
See the live strength scores