Turning 28 Currency Pairs and 5 Timeframes Into 8 Verdicts

Short answer: you cannot chart the euro. You can only chart EUR/USD, EUR/JPY, EUR/GBP and so on — every one of which is contaminated by whatever the other currency is doing. To get a technical read on a currency itself, you have to analyse every pair it appears in, flip the sign wherever it is the quote currency, and combine the results across timeframes.

Doing that produces two numbers, and the second one is more useful than the first. The score says which way the currency is leaning. The agreement says whether the score means anything. Right now the New Zealand dollar scores 65/100 bullish with agreement of 1.00 — every pair it trades in tells the same story. Sterling scores 54 with agreement of 0.41 — its pairs are contradicting each other, and the score is close to noise.

This article walks through the whole aggregation, with live numbers. The technical currency outlook runs it every 30 minutes across 28 pairs and 5 timeframes, free and without an account.

Problem 1: A currency has no price

EUR/USD falling tells you one of three things: the euro weakened, the dollar strengthened, or both moved and one moved more. The chart cannot distinguish them.

The standard fix is to look at several EUR pairs. If EUR/USD, EUR/GBP, EUR/JPY and EUR/AUD are all falling, the euro is weak. If only EUR/USD is falling, that is a dollar story.

Formalising that requires sign adjustment. For each pair a currency appears in:

• If the currency is the base (EUR in EUR/USD), a bullish reading on the pair is bullish for the currency: multiply by +1.
• If the currency is the quote (USD in EUR/USD), a bullish reading on the pair is bearish for the currency: multiply by −1.

Every one of the eight majors appears in seven of the 28 pairs. Score all 28 pairs, then for each currency take its seven sign-adjusted readings and aggregate.

This is the step that goes wrong most often in homemade versions, and the failure is silent — a sign error in one leg produces a plausible-looking number that is simply upside down for that currency.

Problem 2: Which timeframe is right?

None of them, individually. A currency can be in an uptrend on the weekly and a downtrend on the 15-minute, and both facts are true and relevant to different traders.

The approach here is to score five timeframes independently and weight them:

The daily carries the most weight because it is where most trend structure is legible without being so slow that it cannot respond. The weekly anchors the longer picture. M15 is included at a deliberately small weight so intraday noise cannot flip a verdict.

The per-timeframe breakdown is at least as useful as the blended score. Here is the New Zealand dollar as of 15 August 2026:

Every timeframe bullish, with the strongest reading on H1 and the weakest on the weekly. That is a currency in an established uptrend that has accelerated recently — a different setup from one that is strong on the weekly and fading intraday, even though both might blend to a similar headline number.

Problem 3: Which indicators, and how much does each count?

Twenty weighted signals are computed per pair per timeframe, grouped into five families:

Two things about this weighting are worth defending.

Trend gets the largest single share because trend persistence is the most robustly documented behaviour in price series. It is not the most exciting family; it is the most reliable one.

Volatility gets the smallest because it is a regime indicator, not a directional one. An ATR percentile of 15 tells you the environment is quiet, which changes how you size and where you place stops — but it does not tell you which way to lean. Giving it directional weight would be a category error.

The volatility family still surfaces separately in the output for exactly that reason. The kiwi currently shows an ATR percentile of 15.7 with 17.1% of its pairs in a Bollinger squeeze — a low-volatility, compressed regime. A bullish score in a compressed regime is a different trade from a bullish score in an expanding one.

The output: score, confidence, and why confidence matters more

The current board

As of 15 August 2026, all 28 pairs analysed:

Confidence is a herding measure, not a certainty measure

This is the part that distinguishes a useful aggregate from a misleading one.

Confidence measures how much the seven constituent pairs agree with each other, after sign adjustment. It says nothing about whether the analysis is correct — only whether it is coherent.

Compare two rows:

• CHF: score 23, confidence 1.00. Every single franc pair says the same thing. This is not seven weak signals averaging to a strong one; it is seven strong signals pointing the same way. The franc is broadly, unambiguously heavy.
• GBP: score 54, confidence 0.41. Sterling's pairs disagree substantially. The headline "Neutral" is not a considered verdict of balance — it is the arithmetic residue of contradictory readings cancelling out. Sterling is doing different things against different currencies, and a currency-level read is the wrong tool for it entirely.

Practical rule: below about 0.6 confidence, discard the currency-level score and go back to individual pairs. A low-agreement currency is telling you that "the pound is strong" is not a well-formed statement this week.

Note also that GBP and EUR have identical scores of 54 but different confidence — 0.41 versus 0.56 — and different trend readings (Up/29 versus Sideways/23). Two currencies that look interchangeable on a strength meter are meaningfully different once you look at the dispersion behind the number.

Risk level

Derived from volatility regime, cross-timeframe conflict and structural proximity. GBP and EUR both score Medium risk while everything else is Low — again reflecting internal disagreement rather than large moves.

The best-expression problem

Knowing the kiwi is strong is not a trade. You still have to pick a pair.

The naive answer — buy the strongest against the weakest, so NZD/CHF here — is often wrong in practice. That pair may be illiquid, may have already run, may have a wide spread, or may be sitting in the middle of a range with no clean structure.

The engine picks a primary pair per currency based on which one expresses the view most cleanly given current structure, and generates a concrete plan on the H4 timeframe. The current NZD plan:

Note it chose NZD/USD, not NZD/CHF, despite the franc being the weakest currency on the board. The stop is anchored to a structural level with a documented number of touches rather than to a round number of pips — which is what makes it a plan rather than a suggestion.

This is a starting point, not a signal service. The reasoning is exposed precisely so you can disagree with it.

The institutional layer

Alongside the score, each currency gets a breadth read on where price sits relative to volume-weighted references. For the kiwi:

• 57% of NZD pairs trade on the NZD-bullish side of their daily VWAP. VWAP breadth is a participation measure — it says how broadly the move is supported rather than how far it has gone.
• NZD/USD daily: price above the volume point of control at 0.58639. The POC is where most volume transacted; trading above it is constructive.
• NZD/USD daily bearish MACD histogram divergence — price made a higher high while the MACD histogram made a lower high.

That last item contradicts the bullish headline, and it is reported rather than suppressed. An aggregate that only surfaces confirming evidence is a marketing document. The divergence is a legitimate reason to be more careful with entry timing on a currency the engine is otherwise positive on.

Cross-checking against everything else

A technical score is one input. It becomes considerably more useful when it agrees or disagrees with independent measures.

Take the US dollar as of this week:

Price is falling, the data is missing expectations, and speculators are heavily positioned the other way. Those three facts fit together into a specific, actionable observation: a crowded long-dollar position is being squeezed by deteriorating data. That is a much stronger read than any one lens alone, and it is only visible because the three measures are independent.

The franc is the mirror image: technically Strong Bearish with perfect agreement, macro score −73 (weakest of the eight), and speculators net short 29.5% of open interest. Here everything agrees, which means the move is well-established and the marginal information is low — a good trend to follow, a poor one to initiate.

Where this approach falls down

Aggregation destroys pair-specific information. A currency-level score cannot tell you that EUR/GBP is in a tight range while EUR/JPY is trending hard. That is exactly what the confidence number is for — it flags when aggregation has thrown away too much.

Indicator weights are reasoned, not fitted. The 30/20/20/20/10 split reflects a defensible view about which families of signal are most reliable. It has not been optimised against forward returns, and optimising it on a limited sample would produce a curve fit rather than an improvement. Treat the weights as a transparent assumption you can inspect, not a discovered truth.

Everything here is a description of the past. Technical analysis characterises current state. The score says the kiwi has been trending up across seven pairs and five timeframes; it does not say it will continue.

Thirty-minute refresh, not real-time. Good enough for swing and position decisions, too slow for scalping.

Frequently asked questions

How do I tell which currency is strongest technically? Analyse every pair the currency appears in, flip the sign where it is the quote currency, and aggregate across timeframes. As of 15 August 2026 the New Zealand dollar scores highest at 65/100 with perfect cross-pair agreement, and the Swiss franc lowest at 23/100, also with perfect agreement. The technical currency outlook does this across 28 pairs and five timeframes automatically.

What is the difference between a currency strength meter and a technical score? A strength meter typically measures recent price change across a currency's pairs — a momentum snapshot. A technical score combines trend structure, momentum, volatility regime, market structure and volume-weighted references across multiple timeframes into a directional read. The strength meter answers "what just happened"; the technical score answers "what state is this currency in". Both are useful; they are not substitutes.

What does cross-pair agreement mean and why does it matter? It measures how consistently a currency's seven pairs point the same direction after sign adjustment. High agreement means the currency itself is driving; low agreement means the pairs are being moved by their other legs. Sterling's 0.41 agreement this week means its "Neutral" score is contradictory readings cancelling out, not a considered balance — and the correct response is to analyse GBP pairs individually rather than trust the aggregate.

Which timeframe should I weight most heavily? The daily, for most traders. Here it carries 30%, with the 4-hour at 25% and the weekly at 20%. The daily is where trend structure is legible without being unresponsive. The 15-minute is included at just 10% specifically so intraday noise cannot flip a longer-term verdict.

How do you combine RSI, MACD and ADX into a single signal? Group them by what they measure and weight the groups, rather than averaging indicators directly. RSI, MACD and stochastic RSI are all momentum measures and are highly correlated — averaging them triple-counts the same information. Here they contribute 20% collectively as a momentum family, while trend indicators (EMA stack, Supertrend, Ichimoku, ADX) carry 30% and volume-based measures 20%.

What is the best pair to trade a currency view? Not automatically the strongest against the weakest. The best expression depends on current structure, liquidity, spread and how far each pair has already moved. This week the strongest currency is NZD and the weakest CHF, but the generated plan uses NZD/USD rather than NZD/CHF, because NZD/USD is sitting in a clean H4 pullback zone with a structural stop level behind it.

How often does the analysis update? Every 30 minutes, covering all 28 major and cross pairs across M15, H1, H4, D1 and W1. Historical scores are retained so you can see how a currency's technical read has evolved rather than only where it currently stands.

Does the engine use AI to generate its scores? No. Every score is deterministic arithmetic over indicator values — the same inputs always produce the same outputs. Narrative text is generated separately to describe the numbers; the numbers themselves never depend on it, and are identical whether narration runs or not.

Conclusion

Most currency-strength tools give you one number and no way to know when to distrust it. That is the failure mode this design targets.

The score is the easy part. The confidence measure — how much a currency's own pairs agree with each other — is what turns an aggregate into something you can act on, because it tells you when the aggregate has thrown away the information that mattered. A currency at 65 with perfect agreement and a currency at 54 with 0.41 agreement are not two points on the same scale. One is a coherent trend and the other is a statistical artefact of contradiction.

Both are on the free board at fxterminal.app/technical-outlook, along with the per-timeframe breakdown, the indicator families behind each score, the volatility regime, and the specific pair and structural levels that best express each view.