FX TERMINAL
Top Setups Analysis turns the terminal's currency research into a ranked, two-sided board. Every major pair is scored across five factor families — policy and rate expectations, economic momentum, multi-timeframe trend, institutional positioning and crowd positioning — and the ones that agree rise to the top. Each setup opens into the full case: which factors voted, which disagreed, what the two economies actually printed, the scheduled releases that could reprice it, and a volatility-scaled level plan.
Most research tells you how a single currency looks. You still have to hold eight of those views in your head, pair them off, and decide which of the twenty-eight resulting pairs has the strongest case. Top Setups Analysis does that step: it scores every major pair across five independent factor families and ranks them by how strongly those families agree, so the board opens on the pair with the most support behind it rather than on an alphabetical list.
The comparison is made family by family, not on the headline. Two currencies can both look bullish for completely unrelated reasons — one on a repricing of its rate path, the other on a run of data beats — and differencing their headline scores would credit that as agreement when the two are not even arguing about the same thing. Scoring the gap within each family means a pair only earns conviction where the same kind of evidence leans the same way on both legs.
Policy and rate path measures the change in what the market expects a central bank to do — the last decision against consensus, how futures and front-end yields have repriced, and how close the next meeting is. Economic momentum is the weighted surprise across each economy's release calendar, where a CPI beat counts for far more than a second-tier survey. Institutional positioning reads CFTC Commitments of Traders against its own multi-year range. Those three are compared leg against leg.
The other two are measured on the pair itself, because that is strictly better than differencing two currency averages. Multi-timeframe trend is a technical read of the actual instrument across five timeframes from fifteen minutes to weekly. Crowd positioning is the retail crowd's own book on that exact symbol, read contrarian — a heavily one-sided crowd is a source of fuel for the other side, not a reason to join them.
Each factor family casts one vote and spends its whole weight on it; how far past the threshold its reading sits changes nothing. That sounds blunt and it is deliberate — every family is itself an average of components, so a weighted mean of averages collapses towards zero and everything reads neutral. Families that have data but no directional view keep their weight in the denominator, which is why a quiet market cannot produce a hundred-out-of-a-hundred setup.
The board is as explicit about doubt as it is about conviction. Every setup shows how much of the factor book reported at all, how much of it actually voted, and what share of the voting weight pointed the same way. A pair whose heaviest family disagrees with the other four is shown with that disagreement named, not buried, and one where too little of the book reported is published as no-trade rather than as a confident call on thin evidence.
A setup can be perfectly reasoned and still be unplayable. Every board row lists the high and medium impact releases scheduled for either currency over the next four days with a live countdown, and a release inside the next twenty-four hours visibly reduces that setup's conviction — the reduction is stated as a percentage next to the release that caused it, never applied silently.
Each directional setup also carries a level plan built from the pair's own recent volatility: a stop at a fixed multiple of its average daily range, and targets at set multiples of that stop. Sizing this way rather than off the nearest swing high is what makes the risk on every card mean the same thing, so a hundred-and-sixty-pip stop on a yen cross and a fifty-pip stop on a euro pair are comparable positions rather than comparable-looking numbers.
A ranked list quietly invites you to act on the top five, which on a board built from eight currency views is very often one view repeated five times. A bullish euro-dollar call and a bullish sterling-dollar call are not two ideas; they are one bearish-dollar view expressed twice, and they are wrong together. The exposure view answers that directly: it groups the actionable setups into distinct ideas, marks the best expression of each, and shows the net currency bias acting on all of them would amount to.
That grouping is structural rather than statistical. Two pairs sharing a currency leg in the same direction are correlated by construction, whatever last month's returns happened to do, so the grouping is defined for every pair on the board, is always current, and cannot quietly go stale the way a scraped correlation matrix can.
All data, readings, analysis and numbers on FX Terminal are for informational purposes only and do not constitute financial advice. FX Terminal is not responsible for any investment decisions or outcomes. Please read the full disclaimer at fxterminal.app/disclaimer.