05 · Forex Technical Analysis and Practical Patterns
Forex is the world's deepest market and the most "pure" arena for technical analysis: no price limits, no T+1, no insider earnings reports — price forms through a global auction. Technical tools are far more usable here than in A-shares — but precisely because everyone uses them, the same pattern can behave completely differently across pairs and sessions.
This chapter is the technical deep-dive of chapter 01 "Forex Trading Practice": support/resistance, trends and channels, Fibonacci retracement, candlestick patterns, moving averages and Bollinger Bands in forex — each covered properly, ending with an executable multi-timeframe workflow. The goal isn't adding another indicator; it's knowing which parts of forex technicals are "structure" and which are "noise".
1. What Makes Forex Technical Analysis Different
Establish three forex-specific facts first, or every tool below gets misused:
1. Forex Is a 24-Hour Market with No Unified Candle Close
- Forex has no single exchange; daily closes follow session convention: most brokers use 5:00 PM ET (near NY close / Sydney open) as the daily close point.
- The same pair can differ by one candle between brokers because their close times differ. Confirm your platform's timezone before reading patterns — extra care when comparing across platforms.
- From Friday 17:00 to Monday open there are weekend gaps, but they're usually far smaller than stocks/crypto.
2. Spreads and Overnight Interest "Eat" Small-Timeframe Profits
- Major pairs carry spreads of 0.1–1 pip; cross pairs 1–3 pips. On a 5-minute chart, one spread cost can equal 20%–50% of your target profit.
- Overnight positions pay/earn swap (overnight interest): shorting high-yield currencies may cost interest daily; longing them may earn it. Small-timeframe traders must total "spread + swap + slippage" — many patterns don't make enough on small timeframes to cover costs.
3. Technical Consensus Is Stronger, So Self-Fulfillment Is More Visible
- Traders worldwide watch the same charts and same key levels (round numbers, prior highs/lows, daily MAs). Key levels get respected more easily — everyone's orders sit there.
- The side effect: false breakouts multiply too. Stop-loss clusters (above prior highs, above round numbers) get "swept then reversed" constantly. Hence in forex, the retest after a breakout matters more than the breakout itself.
2. Support/Resistance and Round Numbers
Psychological Levels
In forex, 00/50 levels are natural magnets: EUR/USD's 1.1000, 1.1050; USD/JPY's 150.00, 150.50. The reason is simple — huge amounts of retail limit orders and option barriers cluster near round numbers.
| Level Type | Behavior | Trading Implication |
|---|---|---|
| 00 levels | Strong magnets, many false breakouts | After breaking 1.1000, expect a retest first — don't chase; wait for confirmation |
| 50 levels | Secondary levels, intraday support/resistance | Good for small-timeframe scalping between levels |
| Prior highs/lows | True stop clusters | After breaks, expect "stop-sweep → pullback → go" scripts |
Support/Resistance Are "Zones", Not Lines
- Effective S/R in forex is a zone (prior high/low ± 10–20 pips), never an exact line. Reasons: spreads, inter-broker quote differences, dispersed order placement.
- Three factors validate a zone: touch count (≥3), candle reaction at touches (long wicks / engulfing), timeframe of the zone (daily zone > hourly zone).
- After support breaks, support flipping to resistance is one of forex's most reliable rules — trapped longs at the old support sell into any rebound.
Worked Example: EUR/USD Round Number Script
Suppose EUR/USD consolidates between 1.0950–1.1050 on the daily, with 1.1000 mid-range:
- Price rebounds from 1.0950 toward 1.1000; the first reaction is not chasing long but observing: any 15-minute stalling signal (doji, upper wick)?
- If 1.1000 breaks on volume, wait for the retest — a hold at 1.1000–1.1008 with a support signal allows a light trend-following long.
- If price breaks away without looking back (strong momentum), skip the chase entirely — in forex, "missing" is always cheaper than "chasing wrong".
3. Trendlines and Channels
Trendline Drawing Points in Forex
- Draw trendlines on candle body closes, not wicks — more reliable (wicks are often stop-hunting spikes).
- ~45° trendlines mean the most; too steep (>60°) or too flat (<20°) lines fail with high probability — steep slopes eventually get digested sideways.
- A trendline is only confirmed by its "second touch": first touch is just a candidate; second touch plus bounce makes it real; third touch is the low-risk entry.
Ascending/Descending Channels
- Forex channels are tidier than stocks' (continuous volatility, no limit interruptions).
- The channel midline isn't a signal; only upper/lower boundary + reversal candles form trade references; channel ends (price touching the lower bound a 3rd–4th time) often coincide with false breakouts.
- Post-breakout target: channel height (vertical distance from breakout point to the opposite boundary).
Golden Rule of Trend Filtering
Trade only trend-direction "pullback entries"; never catch falling knives against the trend. With MAs bullishly stacked (short > mid > long), look only for dips to buy; bearishly stacked, only rallies to sell. Directional filtering removes 70% of invalid signals.
💀 Iron Rule: Trade Only Trend-Direction Pullbacks — Never Catch Falling Knives
Trade only trend-direction pullback entries; never catch knives against the trend. Bullish MA stacks → buy dips only; bearish stacks → sell rallies only — directional filtering removes 70% of invalid signals. So forex technical analysis's first principle isn't "how many indicators you know" but "fix the direction first" — until direction is set, every pattern is noise.
4. Fibonacci Retracement: Forex's Most-Used Tool
Fibonacci retracement (38.2% / 50% / 61.8%) is forex technical analysis' most frequently used and most abused tool. Using it right comes down to: retracement levels must be combined with "structure", not read as standalone numbers.
Correct Drawing
- Draw only on a clean trending leg (swing low to high, or reverse), never across choppy ranges.
- Wicks or bodies for endpoints? Mainstream forex practice: daily level uses extreme wicks; small timeframes use bodies. Inconsistent anchoring shifts levels by dozens of points — pick one method per chart and stick to it.
- Watch for clusters of three levels: where 38.2% overlaps a prior high, 50% a round number, 61.8% a moving average — those confluence zones are far more reliable than any single level.
Behavioral Stats per Level (teaching figures)
| Level | Common Script | Trading Reference |
|---|---|---|
| 38.2% | Shallow retracement marking strength | In strong trends, a hold above 38.2% supports continuation |
| 50% | The most-touched "neutral level" | Weak alone, but effective when confluent with round numbers |
| 61.8% | The "golden ratio" key level | Limit of deep pullbacks; losing 61.8% usually signals trend reversal |
The Counterintuitive Point
- A retracement level isn't "arrived = bounce" — the signal requires arrival plus a reversal candle. Fib level + reversal pattern = signal; fib level alone = reference line.
⚠️ Counterintuitive: Retracement Levels Don't Mean "Arrived = Bounce"
A retracement level doesn't trigger on arrival alone — arrival plus a reversal candle is the signal. Fib level + reversal pattern = signal; fib level alone = reference line. So don't treat 38.2%/50%/61.8% as automatic buy/sell buttons — they only mark where price may pause; the signal still needs reversal-candle confirmation.
- Losing 61.8% doesn't guarantee instant reversal, but it puts the original trend in "questionable" status — reduce and observe rather than averaging down against the move.
- Fibonacci extensions (161.8%, etc.) measure targets: after breaking a prior high, 1.272 / 1.618 serve as trailing-profit references.
5. Candlestick Patterns, Forex Edition
Candlestick patterns (engulfing, hammer, doji, three crows, etc.) got their basics in the technical-analysis chapter; here we cover only forex-specific usage:
1. Patterns Need "Location" to Mean Anything
- The same hammer at a daily-level support zone is a buy signal; mid-trend it's noise; at the end of a rally + higher-timeframe overbought it may be a reversal warning.
- Order of judgment: location first (support/resistance/mid-trend) → then the pattern → finally confirmation (next candle).
2. The "Timezone Trap" of Forex Candlestick Patterns
- Asian-session (Sydney–Tokyo) pattern signals are weak — thin liquidity, wide spreads, big money absent.
- Patterns formed at the London open (~15:00 Beijing) and NY open (~20:00–21:00 Beijing, DST) are far more credible — that's when real money prices the market.
- Discount Friday-late patterns: weekend flattening and profit-taking distort them.
3. Engulfing-Pattern Filters
- A bullish engulfing needs a prior bearish candle fully engulfed by the next bullish body, ideally with expansion (in forex, watch volatility expansion over volume — rising ATR).
- Wait for the retest after the signal: if price retests the engulfing body's midpoint and holds, that's the low-risk entry.
6. MAs, Bollinger Bands, and ATR in Practice
Moving Averages: Forex's "Axiomatic" Tools
| MA | Mainstream Forex Usage |
|---|---|
| EMA20 | Short-term trend filter: long only above EMA20 |
| EMA50 | Swing divider: above = bullish bias, below = bearish |
| SMA200 | Bull/bear line: the single most important daily-level MA, price oscillating around it |
| Golden/death cross | Lagging and noisy alone; must stack directional filters (e.g., EMA50 slope) |
- In forex, EMAs beat SMAs (faster reaction), though fast MAs also get faked by spikes more easily. Use the distance between fast and slow MAs to gauge trend strength: widening = accelerating; narrowing = transition brewing.
Bollinger Bands: The Volatility Envelope
- Default forex parameters 20/2; bandwidth matters more than the bands themselves: extremely narrow bandwidth (squeeze) often precedes directional breakouts — direction from fundamental catalysts or higher-timeframe trend.
- Touching the upper band ≠ sell signal: in strong trends price walks the band. Combine RSI divergence and candlestick patterns to judge whether "upper-band top" is real.
ATR: Forex's Only "Position Ruler"
- ATR (average true range) on the 1-hour chart: EUR/USD typically 8–15 pips, GBP/USD 15–25, GBP/JPY 30–50. Set stop distances with ATR (e.g., 1.5×ATR) instead of guessing "30 pips".
- ATR doubles as a volatility filter: when ATR jumps 50%+ (data events / central bank decisions), either stand aside or halve size — spreads widen then too.
7. Multi-Timeframe Analysis Workflow
Professional forex traders almost universally run three-timeframe analysis. A workflow you can copy directly:
① Daily (direction): trend up/down? Where are key S/R levels?
↓
② 4H/1H (trade frame): find entry zones (retracements / round numbers / patterns)
↓
③ 15M/5M (execution): wait for confirmation (reversal candle / held retest)Worked Example: Long EUR/USD
| Step | Timeframe | Observation | Decision |
|---|---|---|---|
| 1 | Daily | Price above SMA200, EMA50 rising | Bias long; look for long setups only |
| 2 | 4H | Pullback into 38.2%–50% retracement + prior-high support zone | Mark entry zone 1.0980–1.1000 |
| 3 | 1H | Bullish engulfing forms, ATR not abnormally elevated | Wait for retest confirmation |
| 4 | 15M | Retest holds engulfing midpoint, small bull candle appears | Enter, stop below 1.5×ATR |
| 5 | Holding | Trail along the 4H trendline | Trail stop; target prior high |
Three Disciplines
- Higher timeframes decide direction; lower timeframes decide timing — when a lower-timeframe signal opposes the higher-timeframe direction, skip it (counter-trend filtering).
- Every signal must answer "on which timeframe does this hold" — a signal that can't name its frame is treated as noise.
- Add size only when timeframes align: daily, 4H, and 1H pointing the same way with confluence-zone overlap is what justifies larger positions.
✅ Conclusion: Higher Timeframes Decide Direction, Lower Timeframes Decide Timing
Higher timeframes decide direction; lower timeframes decide timing — when a lower-timeframe signal opposes the higher timeframe, skip it. That's why professionals run three-timeframe analysis: daily for direction, 4H/1H for entry zones, 15M/5M for confirmation. Every signal must answer "which timeframe does this hold on" — anything that can't is noise.
⚠️ Risk Warning
- Technical analysis is a probability tool, not prophecy: every pattern, retracement level, and MA fails sometimes — stops are mandatory.
- High leverage magnifies technical errors: the same signal costs a trial at 1:10 leverage and a blow-up at 1:100.
- Spreads and slippage balloon during data releases and central bank decisions; technical signals within ±30 minutes of major events are unreliable — stand aside or cut size.
- All levels, parameters, and behavioral statistics here are teaching references — defer to the latest markets and broker terms.