01Foundations of Technical Analysis
What technical analysis assumes, the raw material it works on, the different lenses you can put on the same price — and how each foundation concept translates into an actual trading decision.
What it is — and the three premises
Technical analysis (TA) is the study of price and volume to estimate the probability of the next move. It does not value a business; it reads the behaviour of the people trading it. Three classic assumptions:
- Price discounts everything. All known fundamentals, expectations, and emotion are already expressed in price, so the chart alone is sufficient to study.
- Price moves in trends. A move in motion is more likely to continue than reverse — until structure says otherwise.
- History rhymes. Crowd behaviour repeats, so recurring patterns carry statistical (not guaranteed) information.
- Premise 1 — you don't need to out-research the news; you need to read the reaction to it. Trade what price does, not what you think it should do.
- Premise 2 — your default trade is continuation, not reversal. Statistically, joining a trend beats calling its top.
- Premise 3 — patterns are probabilities with failure rates. Every setup needs a stop, because "usually" is not "always."
Technical vs fundamental vs quantitative
| Dimension | Technical | Fundamental | Quantitative |
|---|---|---|---|
| Question | When to buy / sell? | What is it worth? | What has an edge, statistically? |
| Inputs | Price, volume, time | Earnings, rates, macro | Data + tested rules |
| Horizon | Minutes – months | Months – years | Any, if it backtests |
| Blind spot | Regime shifts, news shocks | Timing, sentiment | Overfitting, regime change |
Anatomy of a candlestick
- Long body = conviction. Small body = indecision.
- Long lower wick = sellers pushed down and failed — bullish information at support.
- Close location is the verdict. Always wait for the candle to close before acting — an intrabar signal can vanish.
Timeframes and trader styles
| Style | Holding period | Primary TF | Trades / week |
|---|---|---|---|
| Scalper | Seconds–minutes | M1–M5 | Dozens+ |
| Day trader | Intraday, flat by close | M5–M15 | Several / day |
| Swing trader | Days–weeks | H4–Daily | 1–5 |
| Position trader | Weeks–months | Daily–Weekly | < 1 |
- Choose the style your life allows: if you can't watch the screen, you are a swing trader, not a scalper.
- Rule of thumb: bias TF → 4–6× your setup TF, setup TF → 4–6× your trigger TF (e.g. Daily → H4 → M15).
- Trade where liquidity is: FX moves in London, New York, and their overlap. Avoid dead-session chop.
Gaps
- Common gap (range, normal volume): fade it — stop beyond the gap's far edge.
- Breakaway gap (out of a base, high volume): do not fade. Buy the first pullback; the gap top often acts as support.
- Runaway gap: trend is healthy — hold with-trend; it often marks the halfway point.
- Exhaustion gap (late in trend, climactic volume, quickly filled): exit longs; a close back through is a reversal trigger.
02Market Structure & Trend
Before any indicator, learn to read structure from price itself: the sequence of swing highs and lows that defines who is in control — and the specific entries each structural event offers.
The three states of a market
- Uptrend — long only. Buy pullbacks to higher lows / support. Invalidation: the last HL breaks.
- Downtrend — short only. Sell rallies into lower highs / resistance. Invalidation: the last LH breaks.
- Range — fade the edges. Buy support, sell resistance, stop just beyond the boundary. Stop range-trading the moment a boundary breaks with conviction.
- Unclear / transitioning — no trade. "I can't name the state" is itself a signal: stand aside.
Structure breaks: BOS & CHoCH
- BOS — continuation entry. Wait for the pullback to the broken swing level; enter on a confirming candle. Stop below the pullback low; target the next structural level.
- CHoCH — defence first, offence later. Tighten stops / take partials. Only trade the new direction after the market confirms it: a lower high forms, then breaks the new low (first BOS of the new trend).
- Failed break — fade it. If price breaks a swing and immediately closes back inside, the trapped traders are fuel — trade back toward the other side.
Liquidity, equal highs/lows & premium/discount
Stops cluster beyond obvious swing highs/lows. Price is often drawn to equal highs/lows, sweeps the resting liquidity, then reverses. Within any leg, the upper half is a premium (better to sell) and the lower half a discount (better to buy), split at the 50% equilibrium.
- Sweep-and-reclaim entry: when price pokes beyond an obvious high/low and closes back inside, enter in the reclaim direction. Stop just beyond the sweep wick; target the liquidity at the opposite side.
- Don't park stops at the obvious level. Place them beyond the zone plus an ATR buffer.
- Premium/discount filter: in an uptrend, only execute longs in the lower half of the current leg (discount).
03Support & Resistance
The levels where supply and demand have flipped before — and are likely to react again. Everything else (Fib, pivots, round numbers) is a way of locating these.
Role reversal
Once broken, support becomes resistance and resistance becomes support — participants trapped at the level defend it on the retest.
- Qualify the level: at least two prior rejections, visible on your setup TF or higher.
- Qualify the break: a full candle close beyond the level, ideally on above-average volume.
- Entry: (a) limit order at the level on retest, or (b) wait for a bullish rejection candle at the retest, enter on its break.
- Stop: beyond the zone's far edge plus a buffer (~0.5×ATR). If the flip is real, price should not trade back through.
- Target: the next opposing level; require ≥ 1:2 R:R.
Pivot points — formula
R1 = (2 × P) − Low S1 = (2 × P) − High
R2 = P + (High − Low) S2 = P − (High − Low)
R3 = High + 2×(P − Low) S3 = Low − 2×(High − P)
- Bias: opening above P = long bias; below P = short bias. Trade pullbacks to P in the bias direction.
- Range day: fade R1/S1 with a reversal candle, target P.
- Trend day: do not fade. Buy pullbacks; expect the day to reach R2–R3.
- Confluence rule: a pivot alone is a B-grade reason. Pivot + horizontal level + VWAP = tradeable.
Supply & demand zones
Mark the origin of a sharp move — the small base from which price exploded. Drop-base-rally (demand) and rally-base-drop (supply) are the key reversal patterns. Fresh, un-retested zones are strongest.
- Mark the zone from the base's open-to-extreme, not the whole move.
- First touch only: limit at the zone's near edge, stop beyond its far edge. By the third touch, treat the zone as spent.
- Demand the impulse: only trade zones that produced a fast, one-directional departure with range expansion.
04Trendlines, Channels & Geometry
Diagonal support and resistance, and the geometric tools that frame a move — with three entry models: the bounce, the break-retest, and the channel fade.
Trendlines & channels
A valid trendline needs at least two touches and is confirmed by a third. In an uptrend, connect higher lows (rising support). A channel adds a parallel line on the opposite side.
- The bounce (3rd touch+): wait for a bullish rejection candle at the line — enter on its break. Stop below the candle's low. Target the channel's upper boundary or prior high.
- The break-and-retest: after a confirmed close through a mature trendline, trade the retest of the line from the other side. "Steeper is weaker."
- The channel fade: trade from boundary to boundary — but only with the slope in a trending channel.
Andrews' Pitchfork
Drawn from three pivots (a major swing and the next two). Price gravitates to the median line and reacts at the outer tines. In an up-fork: buy reactions at the lower tine, target the median. A confirmed close through the lower tine ends the fork.
05Price Action: Candlestick Patterns
The market's most granular language. Single, double, and triple-candle formations that signal exhaustion, rejection, or a shift in control.
Single-candle patterns
The pin bar — a small body with a dominant wick of 2× the body — is the workhorse rejection signal. The hanging man is a hammer shape after a rally (bearish); the inverted hammer is a shooting-star shape after a decline (bullish).
- Context first: the pin must form at a level with its wick rejecting the level, and ideally with the higher-TF trend.
- Entry A — break of the nose: stop-order 1 tick beyond the body side. Stop: beyond the wick's tip.
- Entry B — the 50% retrace: limit at the pin's midpoint. Better price, lower fill rate. Stop: beyond the wick.
- Quality checks: wick ≥ 2× body; close back inside the level; bar range above average.
Two-candle patterns
- Engulfing entry: at a level + with trend, enter on the break of the engulfing candle's extreme. Stop beyond the pattern's opposite extreme.
- Inside bar (with trend): bracket the mother bar with stop-orders; enter on the break in the trend direction. Stop: opposite side of the mother bar. Works best on Daily/H4.
- Harami / doji after a strong run: not an entry — take partials, tighten stops, and wait for direction.
Three-candle patterns
Candlestick execution reference
| Pattern | Bias | Entry | Stop |
|---|---|---|---|
| Hammer / pin bar | Bullish | Break of high, or 50% limit | Below wick tip |
| Shooting star | Bearish | Break of low, or 50% limit | Above wick tip |
| Bullish engulfing | Bullish | Break of engulfing high / its close | Below pattern low |
| Bearish engulfing | Bearish | Break of engulfing low / its close | Above pattern high |
| Morning star | Bullish | Close / break of 3rd candle | Below star's low |
| Evening star | Bearish | Close / break of 3rd candle | Above star's high |
| Inside bar | With trend | Break of mother bar (trend side) | Other side of mother bar |
06Chart Patterns
Larger formations split into reversal (trend ending) and continuation (trend pausing) — each with its measured-move target and exact entry/stop placement.
Reversal — Head & Shoulders
- Entry A: confirmed close below the neckline. Entry B (preferred): retest of the broken neckline from below.
- Stop: above the right shoulder (aggressive) or above the head (conservative).
- Target: head-to-neckline distance projected from the break. Take partials at first support.
- Quality tells: declining volume across the three peaks, rising volume on the break.
Reversal — Double & Triple Tops/Bottoms
- Entry: break of the neckline (the middle peak), or its retest. Aggressive: long at the second low if it sweeps the first and reclaims with a reversal candle.
- Stop: below the lows. Target: pattern height projected above the neckline.
- Tell: second low on lower volume + momentum divergence is the strongest version.
Continuation — Triangles
- Entry: a close through the flat side on rising volume; or the retest of the broken line.
- Stop: inside the triangle, beyond the most recent swing within it.
- Target: the triangle's widest height projected from the breakout point.
- Timing rule: best breaks occur between half and three-quarters of the way to the apex.
Continuation — Flags & Pennants
- Qualify: impulsive pole with range expansion, then a shallow drift retracing < 50% of the pole on shrinking volume.
- Entry: break of the flag's upper line. Stop: below the flag's low.
- Target: pole length projected from the breakout. Partial at 1:2, trail the rest.
Pattern execution reference
| Pattern | Type | Entry | Stop | Target |
|---|---|---|---|---|
| Head & shoulders | Reversal | Neckline break / retest | Above right shoulder | Head–neckline projected |
| Double top / bottom | Reversal | Neckline break / sweep-reclaim | Beyond the twin extremes | Pattern height projected |
| Triangle | Continuation | Line break + volume / retest | Inside, beyond last swing | Widest height projected |
| Flag / pennant | Continuation | Break of consolidation | Beyond flag's far edge | Pole length projected |
| Rising / falling wedge | Reversal | Break against the slope | Beyond last swing inside | Wedge origin |
| Cup & handle | Continuation | Handle breakout | Below handle low | Cup depth projected |
07Indicators — Trend
For each indicator: the formula, how the number is actually computed, the long/short signals it generates, and how to trade it.
Moving Averages — SMA, EMA, WMA, HMA
k = 2 / (n + 1) // EMA smoothing factor; n=20 → k ≈ 0.095
EMA_t = (Close_t − EMA_prev) × k + EMA_prev // seed: first EMA = SMA(n)
WMA(n) = Σ(Cᵢ × wᵢ) / Σwᵢ, w = n, n−1, …, 1 // linear weights
HMA(n) = WMA( 2×WMA(n/2) − WMA(n), √n ) // Hull: near-zero lag
- Use them as filters, not triggers. Crossovers lag badly. Let the cross define the regime; enter on a pullback.
- The EMA20 pullback (trend entry): in an established trend above EMA20 & EMA50, wait for a retrace into EMA20; demand a rejection candle; enter on its break. Stop below the swing low / EMA50.
- Slope beats position. A flat MA means no trend regardless of which side price sits.
- The 200 on the Daily is the regime line: above = buy dips; below = sell rallies.
MACD
Signal = EMA(9) of MACD line
Histogram = MACD line − Signal // histogram peaks = acceleration; leads the MACD cross
- Filter by the higher TF: take only crossovers that agree with the Daily/H4 trend.
- The strongest setup: price pulls back to a level in an uptrend → histogram bottoms and ticks up → MACD crosses up. Stacks location + momentum turn.
- Histogram as exit timer: first histogram peak-and-shrink = take partials (leads the MACD cross by several bars).
- In ranges: MACD whipsaws when EMAs flatten. Require ADX > 20–25 before trusting crossovers.
ADX / DMI — trend strength
+DI = 100 × Wilder14(+DM) / ATR(14) −DI = 100 × Wilder14(−DM) / ATR(14)
DX = 100 × |+DI − −DI| / (+DI + −DI)
ADX = Wilder14(DX) // Wilder smooth: Avg_t = (Avg_prev×13 + x_t)/14
- It's a filter, not a trigger. Only take MA/MACD/breakout signals when ADX > 20–25 and rising. This single filter removes most whipsaw losses.
- High ADX is not "overbought." ADX 45 says the trend is strong — never fade strength because the number is big.
- The turn that matters: ADX rolling over from a high while price makes a marginal new extreme = last legs. Tighten to breakeven+.
Ichimoku Kinko Hyo
Kijun (26) = (HH26 + LL26) / 2
Senkou Span A = (Tenkan + Kijun) / 2 → plotted 26 bars ahead
Senkou Span B = (HH52 + LL52) / 2 → plotted 26 bars ahead
Chikou (lagging) = Close, plotted 26 bars back
// Kumo (cloud) = area between Span A and Span B
- Trade only full alignment: price above cloud + Tenkan > Kijun + Chikou clear of price.
- Kijun is your trailing stop in an Ichimoku trend trade.
- Thin cloud = cheap to cross: expect breakouts through thin Kumo; respect thick Kumo as real S/R.
Parabolic SAR
// AF starts at 0.02, +0.02 each new EP, capped at 0.20
// EP = extreme point of the current trend (highest high in an uptrend)
- Best use: the exit. Enter on your own setup; trail the position at the SAR dot.
- As an entry: only take flips when ADX > 25 and in the higher-TF direction; ignore flips inside ranges.
08Indicators — Momentum
Oscillators that measure the speed of price. They shine at exhaustion and divergence — but in strong trends they stay pinned, so the rules below are trend-adjusted, not naive overbought/oversold.
RSI — Relative Strength Index
AvgGain = Wilder14(gain) AvgLoss = Wilder14(loss)
RS = AvgGain / AvgLoss
RSI = 100 − 100 / (1 + RS)
- Re-band by regime: in an uptrend use 40/80 (buy dips to 40–50, expect tops near 80). The naive 30/70 fade only works inside ranges.
- Divergence is the A-signal: price makes a new extreme, RSI doesn't, at a level, with a reversal candle. Stop beyond the extreme; target the prior swing.
- The 50 line is a trend health check: pullbacks in healthy uptrends hold above ~40–45. A deep cut through 40 warns the trend is changing character.
Stochastic oscillator
%D = SMA3(%K)
- One-directional use: in an uptrend take only the oversold crosses (buy signals) and ignore every overbought cross.
- Use slow stochastic (14,3,3); best in ranges and rhythmic pullback trends.
The wider family — formulas at a glance
Williams %R = −100 × (HH14 − C) / (HH14 − LL14)
ROC = 100 × (C − C_n) / C_n Momentum = C − C_n
StochRSI = (RSI − minRSI_n) / (maxRSI_n − minRSI_n)
09Indicators — Volatility & Volume
Two families that answer "how much is price moving?" and "who is behind the move?" — the context that decides whether a breakout is real, where the stop belongs, and how large the position should be.
Bollinger Bands
Upper / Lower = Middle ± 2 × σ₂₀ // σ = standard deviation of the last 20 closes
%B = (C − Lower) / (Upper − Lower)
Bandwidth = (Upper − Lower) / Middle // the squeeze metric
- Squeeze breakout: after a multi-week bandwidth low, enter on the first close outside a band with above-average volume. Stop at the middle band. False-break tell: immediate close back inside — exit, consider reversing.
- Range reversion: only when the middle band is flat — fade band touches that print a rejection candle, target the middle.
- Regime test: flat middle + regular touches both sides = range rules; sloped middle + one-sided band rides = trend rules.
ATR — the trader's ruler
ATR = Wilder14(TR)
- Stops: structure level ± 1–1.5×ATR buffer. Stops tighter than ~1×ATR are inside the noise.
- Position sizing: Size = Risk$ ÷ (ATR-multiple stop × value-per-point). When ATR doubles, the same dollar risk buys half the size — automatically volatility-normalised.
- Trailing — the chandelier exit: trail at highest high − 3×ATR for longs.
- Regime filter: ATR spiking to 2× its average = news/panic — widen stops and halve size.
Volume — the conviction behind the move
MFM = ((C−L) − (H−C)) / (H−L) // money-flow multiplier, −1…+1
A/D = Σ(MFM × V) CMF = Σ20(MFM×V) / Σ20(V)
- Breakout rule: require volume ≥ 1.5–2× its 20-bar average on the breakout bar. Below that, treat the break as suspect.
- Climax: extreme volume spike after an extended trend with a huge bar and a weak close = exhaustion.
- OBV divergence: price grinds to new highs while OBV flattens — participation is leaving; tighten stops.
- Pullback quality: healthy trend pullbacks come on shrinking volume; expanding volume on a pullback is distribution in disguise.
VWAP — the institutional benchmark
- Bias line: price holding above a rising VWAP = long bias. Trade pullbacks to VWAP in the bias direction with a PA trigger.
- First pullback after an open drive into VWAP is the classic day-trade entry.
- Reversion fade: on rotational days, price stretched to the 2σ band tends to revert to VWAP.
10Fibonacci & Harmonic Patterns
A ratio toolkit for locating where a pullback ends and an extension reaches — and the geometric patterns built entirely from those ratios.
Retracements & the golden pocket
Extensions: 127.2% · 161.8% · 261.8%
// ratios from the Fibonacci sequence: each ÷ next ≈ 0.618; φ ≈ 1.618
- Filter: higher-TF uptrend, fresh impulse leg up.
- Wait for the retrace into 50–61.8% and demand at least one other factor (flip level, trendline, EMA, pivot).
- Trigger: a bullish rejection/engulfing candle inside the pocket.
- Stop: below the 78.6% level or the swing low. A retrace beyond 78.6% says the "pullback" is a reversal.
- Targets: TP1 = impulse high (partial); TP2 = 127.2%; TP3 = 161.8%. Move stop to breakeven at TP1.
Harmonic patterns
Precise five-point (XABCD) structures whose legs must hit specific Fib ratios. When the ratios align, the D point is a high-probability reversal zone (the PRZ). They demand exact measurement.
| Pattern | B retrace (of XA) | D completion |
|---|---|---|
| Gartley | 61.8% | 78.6% of XA |
| Bat | 38.2–50% | 88.6% of XA |
| Butterfly | 78.6% | 127–161.8% of XA |
| Crab | 38.2–61.8% | 161.8% of XA |
| Cypher | 38.2–61.8% | 78.6% of XC |
- Don't front-run. Let price reach the PRZ and print a reversal candle.
- Stop: beyond X (retracement patterns) or beyond 113% of XA (extension patterns).
- Targets: TP1 = 38.2% retrace of the AD leg; TP2 = 61.8% of AD. Move to breakeven at TP1.
- Best version: a PRZ that lands on independent confluence (horizontal level, prior POC).
Continue reading
Part II: Advanced Playbooks covers confluence, Elliott Wave, Wyckoff, smart money & order flow, multi-timeframe analysis, risk & execution, backtesting, psychology, and a complete workflow.
Read Part II: Advanced Playbooks →