Weekday × hour heatmap Measured from our own data through 2026-07-21
Median hourly range by weekday over the last 12 months. Darker = wider typical range.
- Most active
- Tue 14:00 UTC · 20.4p
- Quietest
- Mon 22:00 UTC · 4.8p
The most liquid forex pair globally. Tight spreads, deep order books, and clear technical structure make it the default surface for systematic strategies — and the first market to hold any EA claim against real tick data.
Avg. spread (typical) Typical estimate (broker-published)
0.7 pips
Broker-published, standard account
Average daily range Measured from our own data through 2026-07-21
≈54 pips
20-day measured average
Best sessions (UTC) Measured from our own data through 2026-07-21
London · NY
12:00 – 16:00 UTC
Volatility level
moderate
Profile classification
Typical values from each broker's own published EUR/USD figures (2026-07) — confirm the live spread on your own account type.
Symbol profile
Pearson correlation of daily log returns over the last 20 common trading days.
Strategy & timing
Data
Session windows, broker spreads, and cost math for this symbol — measured and typical values, not marketing.
Measured from our own data through 2026-07-21 Measured from our own data Typical estimate (broker-published) Typical estimate (broker-published) Broker-dependent Broker-dependent
| Account type | Typical spread | Commission | Avg. daily range | Spread / range | Slippage sensitivity |
|---|---|---|---|---|---|
| Standard account | 0.7 – 1.6 pips Typical estimate (broker-published) | $0 Typical estimate (broker-published) | ≈54 pips Measured from our own data through 2026-07-21 | 1.3% – 3.0% Typical estimate (broker-published) | Low Broker-dependent |
| Raw / ECN account | 0.0 – 0.3 pips Typical estimate (broker-published) | $3 – $7 / lot Typical estimate (broker-published) | ≈54 pips Measured from our own data through 2026-07-21 | < 0.6% Typical estimate (broker-published) | Very low Broker-dependent |
Values are typical ranges across brokers and vary by broker, time and market conditions — confirm on your own account type.
| Broker | Standard spread (pips) | Raw-type commission |
|---|---|---|
| Exness | 0.7 | $3.5/lot on Raw |
| TitanFX | 1.0 | $7/lot on Blade |
| AXIORY | 1.1 | $3.5/lot on Nano |
| FXGT | 1.2 | $3/lot on PRO |
| HFM | 1.2 | $3/lot on Zero |
| XM | 1.6 | $3.5/lot on Zero |
Standard-account spreads are each broker's own published figures (checked 2026-07); raw-type accounts compress to sub-pip spreads plus the listed commission. Confirm live values on your own account type.
Live market
Live market data isn't available right now.
Indicative, delayed prices for context only — not a trading feed or an execution quote.
Volatility & timing
Three measured views of the same tick-derived bars: which weekday-hours run hottest, how monthly volatility is trending, and which hours lean directional.
Median hourly range by weekday over the last 12 months. Darker = wider typical range.
Average daily range per calendar month — is the pair speeding up or calming down?
Recent months are calmer than the prior quarter — ranges are compressing.
Share of hourly bars that closed up, by UTC hour, over the last 12 months. Above the midline leans bullish; below leans bearish.
Hourly bias is a mild historical tendency around 50% — not a standalone signal.
Analysis
The best EA for EUR/USD is the one whose full backtest you can read before you run it — and the only backtest you can read in full is one you ran yourself. Hold every EUR/USD result, bought or built, to that bar: a multi-year every-tick test at your own spread, across the regimes a sales page leaves out, with the worst losing streak read before the headline.
This page covers what EUR/USD gives an automated strategy that other pairs do not, which strategy shapes the pair’s character actually supports, and the sessions that decide results. It also covers the test bar a EUR/USD EA should clear — including the losing streaks and the regime dependence that the sales pages of other sites leave out. The session timeline, broker spread table, and cost calculators live in the Trading conditions section above; this analysis is the interpretation.
EUR/USD accounts for roughly 23% of global forex volume. The combination of European Central Bank policy and Federal Reserve policy gives it two distinct catalysts per cycle. That is why systematic models built on rate-differential mean-reversion or trend-momentum tend to back-test cleanly here.
For an EA, the pair’s character reduces to three measurable properties:
The trap inside that last property: a fixed hour means your broker’s server clock, not UTC. Move a clock-anchored EA between brokers with different server timezones and every range boundary silently shifts. Re-check the time inputs (and re-run the backtest) whenever the broker changes.
The pair’s frontmatter lists trend, scalping, mean-reversion, and breakout as suitable — but suitability is not proof, and our own testing points the same way: defaults lose, validation pays. In our own stock-settings baseline of the Builder’s templates — a single USD/JPY M5 run over one year, default inputs, no optimisation — most finished below a 1.0 profit factor, and the only EUR/USD exposure in it came from the two basket templates, both of which also finished below 1.0. Almost none cleared the Builder’s own bar at stock settings, so treat any edge on EUR/USD as something your own settings and test have to earn. Look for it in mechanics like pending stop orders and session windows, not in the indicator on the box.
What the pair’s character supports on EUR/USD specifically:
The session timeline and the per-window notes in the Trading conditions section above show when EUR/USD is liquid. What decides an EA’s results is which of those windows its logic is allowed to trade.
EAs that restrict trading to London-and-NY hours typically out-perform 24-hour variants on this pair. A session filter on EUR/USD is not an optimisation flourish; it is part of the strategy definition. The Tokyo window is the structural reason. Its range moves often reverse at the London open, so a 24-hour EA spends a third of each day trading signals the next session is statistically likely to unwind.
The one caveat: scheduled ECB and Fed events sit inside those “good” windows, and spreads widen exactly then. A news filter (or a pause around rate decisions) protects fast strategies from paying triple spread at the worst moment.
EUR/USD is the lowest cost-of-trading surface among the pairs we cover, but the gap between brokers is wider than most users expect. The broker-by-broker table in the Trading conditions section above lists each broker’s own published standard-account spread, from 0.7 to 1.6 pips.
The arithmetic that matters: the 0.9-pip gap between the tightest (0.7) and widest (1.6) standard spread is about $9 per standard-lot round-turn. For a swing EA trading a few times a week, that is noise. For a high-frequency system trading many times a day, the same gap is a recurring tax on every entry. High-frequency EAs therefore belong on raw-type accounts where the spread compresses to sub-pip levels and the commission is a fixed, known number.
Two cost rules specific to this pair:
We have not published a EUR/USD backtest of our own, and a result from another pair — even another major — does not carry over to EUR/USD’s cost structure and session clock. The honest route to a EUR/USD EA is to build one and hold it to a bar like this before you fund it:
| Test bar | Bar |
|---|---|
| Test window | 2019–2025 (wider than the calm window marketing prefers) |
| Price model | Every-tick, at your account’s real spread |
| Profit factor | ≥ 1.2 |
| Max drawdown | ≤ 20% |
| Outcome | Recorded either way — a failed run tells you as much as a pass |
A build that fails this bar has saved you the deposit you would have lost finding out live. The site’s methodology sets out the same checks in full, and the Builder templates at the end of this section exist to build a EUR/USD candidate and run it through them.
An averaging ladder adds to a losing position at fixed distances and closes the whole basket on a small retrace. On a moderate-volatility pair in a calm window, that retrace nearly always arrives, so the equity curve looks smooth and the profit factor looks healthy. Run the same ladder across a window that contains a long one-way move and the basket keeps growing while the retrace never comes. Judge any ladder on its deepest basket and its equity drawdown, not on the smooth part of the curve.
In the Builder, EUR/USD is the anchor leg of the dual-pair RSI template (RSI period 14, SL 30 / TP 60 pips per leg). It is also the anchor leg of the risk-on/risk-off basket (one MA-cross signal firing EURUSD and AUDUSD together, SL 40 / TP 80 pips). Both generate a ready-to-run MT5 Expert Advisor (.ex5; the editable .mq5 source comes with Pro) with every parameter exposed, so the correlation exposure described above is something you configure deliberately rather than discover later.
No figure on this page is a promise: spreads are broker-published typical values, the test bar is a threshold rather than a result, and any backtest you run is sensitive to spread, server clock and regime. The correct reading of “best EA for EUR/USD” is the one whose worst losing streak you have already budgeted for.
Suitable strategiestrendscalpingmean-reversionbreakout
Open the AIStrategyMiner Builder with this symbol already selected, build the rules from blocks, then test them on your own price history before you decide anything.
Test any EA in the MetaTrader 5 Strategy Tester before you rely on it. Past results do not predict future ones.