Technical Analysis Handbook · Part II of II

Advanced Playbooks.

Confluence, Elliott Wave, Wyckoff, smart money & order flow, multi-timeframe analysis, risk & execution, backtesting, psychology, and a complete workflow — every concept with exact entry/stop/target rules. Sections 11–21 of the complete handbook.

Confluence Elliott Wave Wyckoff Smart money & order flow Multi-timeframe Risk & execution Backtesting Psychology Glossary

11Confluence — Putting It Together

The single most important concept in this handbook. No tool has an edge alone. The edge is the stack: several independent signals pointing to the same price, at the same time.

Confluence means multiple, ideally uncorrelated, factors agreeing on one zone. Each factor is a weak prior; together they compound into a high-probability decision.

horizontal support rising trendline EMA 61.8% fib 4-factor confluence → high-probability long
One zone, four reasons to act: horizontal support + rising trendline + 61.8% Fib + EMA, capped by a bullish reversal candle. That is a trade. A single factor alone usually is not.

Counting honestly — independence matters

Factors only count if they are independent. EMA20 + SMA21 is one factor. RSI + Stochastic is one factor. A genuine four-factor stack draws from different families: structure (level), geometry (trendline/Fib), a dynamic line (MA/VWAP), momentum/volume, plus the higher-TF trend.

Grading and sizing

GradeConfluenceHTF alignmentTrigger candleRisk size
A3+ independent factorsWith trendYesFull (1R)
B2 factorsNeutral / withYesHalf — or skip
C0–1 factor / "looks good"AnyNo trade
The pre-entry confluence checklist
  1. Higher-TF trend direction — am I with it? (Counter-trend instantly caps the grade at B.)
  2. List the independent factors at this zone, by family. Count honestly.
  3. Is there a confirming trigger candle at the zone, closed?
  4. Is R:R to the next opposing level ≥ 1:2?
  5. Grade it. A = full risk. B = half or pass. C = walk away.

12Elliott Wave Theory

A model of crowd psychology: markets move in repeating, fractal wave patterns — five waves with the trend, three against it — at every degree of scale.

The 5–3 structure

12 345 ABC 5-wave impulse (motive)A-B-C correction
Five waves up, three down. Knowing which wave you're in tells you whether to expect continuation (waves 2/4) or a counter-move (after wave 5).

The three unbreakable rules

  • Wave 2 never retraces more than 100% of wave 1.
  • Wave 3 is never the shortest of 1, 3, 5 — and is usually the longest (often a 161.8% extension of wave 1).
  • Wave 4 never overlaps wave 1's price territory (in a standard impulse).

Fib governs the proportions: wave 2 typically retraces 50–61.8% of 1; wave 4 about 38.2% of 3 (alternating in character with wave 2: one sharp, one sideways).

How to trade Elliott (only the A-grade moments)
  • The wave-3 entry — the franchise trade. After a clear impulse (candidate wave 1) breaks structure, buy the wave-2 pullback at the 50–61.8% retracement with a trigger candle. Stop: below the start of wave 1. Targets: wave-1 high (partial), then the 161.8% extension. R:R is routinely 1:3+.
  • The wave-5 fade — exit tool first. Wave 5 on momentum divergence at a 161.8% cluster = take profits on longs. Fading it outright is counter-trend: only with a confirmed CHoCH, half size.
  • The C-completion entry: in a larger uptrend, buy where C ≈ A in length and lands on the golden pocket of the prior impulse.
  • Skip the middle. Inside complex corrections, counts multiply and edge vanishes.
Use it as a map, not a GPS Elliott's real value is context — impulse (trade with trend) vs correction (expect chop). Keep alternate counts as live hypotheses, and let the three rules — not hope — decide when a count dies.

13The Wyckoff Method

A century-old framework for reading the footprints of large operators as they accumulate and distribute inside ranges — the deep logic behind springs, false breakouts, and stop-runs.

The three laws

  • Supply & Demand — price rises when demand exceeds supply.
  • Cause & Effect — time spent building a range (the cause) is proportional to the move that follows. Big bases, big moves.
  • Effort vs Result — volume (effort) should agree with progress (result). High volume with no progress = absorption; a warning at extremes.

The accumulation schematic

resistance (creek) support (ice) SC AR ST Spring SOS markup →
SC selling climax → AR automatic rally → ST secondary test → Spring (false breakdown, red dot) → SOS sign of strength → markup. Distribution mirrors this at tops with a UTAD instead of a spring.
How to trade Wyckoff
  • The spring entry (best R:R): price breaks range support, fails, and closes back inside. Enter long on the reclaim close. Stop: below the spring wick. Tell: spring's break on low volume, reclaim on rising volume.
  • The SOS/LPS entry (conservative): after price rallies through resistance on expanding volume, buy the pullback that holds above the broken creek. Stop below the LPS low.
  • The UTAD short: mirror at a distribution top — false break above resistance that closes back inside on weak follow-through volume.
  • Phase discipline: no positions during phase B chop. The trade is the test (C) and the breakout-retest (D), nothing else.
Why it still works Strip the jargon and Wyckoff is supply/demand + volume confirmation + liquidity. It explains why double bottoms, false breakouts, and stop-runs happen — and maps directly onto the Smart-Money vocabulary next.

14Smart Money Concepts & Order Flow

The modern price-action vocabulary that reframes S/R around institutional order flow — plus the literal order-flow tools — assembled into one executable entry model.

Divergence — the bridge from momentum

Price higher high RSI lower high price up, momentum down → bearish divergence
Regular divergence warns of reversal. Hidden divergence is the continuation version: price makes a higher low while the oscillator makes a lower low — the trend is resting, not dying.
How to trade divergence
  • Regular (reversal): only at a meaningful level, only with a trigger candle. Divergence into thin air = early and wrong.
  • Hidden (continuation): in an uptrend, price makes a higher low while RSI makes a lower low — a with-trend entry at pullbacks.
  • Stacked divergence (visible on two timeframes at once) is the strongest version.

Order blocks, fair value gaps & liquidity

order block fair value gap (imbalance) prior low (liquidity) sweep → reversal
Order block — the last opposing candle before an impulsive move. FVG — a price imbalance between candle wicks that tends to be filled. Liquidity sweep — a poke beyond an obvious low to trigger stops before reversing.
The SMC entry model (one repeatable sequence)
  1. HTF bias: Daily/H4 trend and the liquidity pools above/below (equal highs/lows, old swings).
  2. The sweep: price runs one of those pools — the trap springs.
  3. Displacement: an impulsive move away from the sweep that breaks structure (CHoCH on the trading TF) and leaves an FVG.
  4. Entry: limit order in the FVG (50% of the gap) or at the order block that launched the displacement.
  5. Stop: beyond the sweep wick. Target: the opposing liquidity pool. Geometry routinely gives 1:3 to 1:5.

Literal order flow tools

DOM / order book

Resting bids/offers. Large resting size hints where price may stall — but can be spoofed. Use: context near your level, never standalone.

Footprint chart

Bid-vs-ask volume inside each candle. Use: spot absorption — heavy selling into a level that refuses to drop = passive buyer.

Delta & cumulative delta

Buy minus sell market orders. Use: price makes a new low but delta doesn't (sellers exhausted) — reversal tell at support.

Volume Profile / POC

Volume by price. POC = most-traded price (magnet). Value Area = ~70% of activity. Low-volume nodes = price crosses fast. Full playbook below.

How to trade the Volume Profile
  • Open inside value — expect rotation: fade VAH/VAL toward the POC with a rejection candle; stop beyond the edge.
  • Open outside value and accepted — trend day: go with it; do not fade. The 80% rule: re-entry into the value area that holds has ~80% odds of crossing to the other side.
  • Naked/untested POCs from prior days are magnets — natural targets.

15Multi-Timeframe & Intermarket Analysis

No chart trades in isolation. Two habits separate professionals: aligning timeframes top-down, and reading the other markets that drive your instrument.

Top-down, three-screen alignment

ScreenTimeframe (swing example)Job
BiasDaily / WeeklyTrend direction + major S/R. Only trade this way.
SetupH4 / H1Find the pattern / level / confluence.
TriggerM15 / M5Precise entry candle + tight stop.
The alignment rules
  • Conflict rule: when timeframes disagree, the higher one wins — and the honest answer is usually no trade.
  • Stops and targets live on the setup TF, not the trigger TF.
  • One drill-down only: bias → setup → trigger. Chasing a 4th TF is how traders talk themselves into C-grades.

Intermarket relationships

RelationshipTypical directionTrading use
DXY ↔ commodities, EUR/GBPInverseCheck DXY at major S/R before any USD-pair trade
Yields ↔ growth stocksInverseRising yields = headwind for index longs
VIX ↔ equitiesInverseRegime filter: calm (<20) = normal size; panic (>30) = no new longs
Risk-on (stocks up) ↔ JPY / CHFInverseEquity rallies favour short-yen carry pairs
Credit spreads ↔ equitiesInverseWidening spreads warn before stocks roll over

16From Signal to Trade: Risk & Execution

Analysis finds the setup; risk management decides whether you survive to compound it. Every formula here comes with a worked example.

Think in R, not dollars

Define one unit of risk — 1R — as the distance from entry to stop. Every outcome is measured in R-multiples: +2R, −1R. At 1:2, a 40% win rate is profitable.

Expectancy — the only number that matters

Expectancy (R per trade) = (Win% × avgWin_R) − (Loss% × avgLoss_R)
// 0.40 × 2.5R − 0.60 × 1.0R = 1.00 − 0.60 = +0.40R per trade
Required win rate to break even at R:R = 1/(1+R) // at 1:2 you need just 33.4%

Position sizing

Risk$ = Equity × risk% // $10,000 × 1% = $100
Size = Risk$ ÷ (stop distance × value per unit)
// FX: $100 ÷ (30 pips × $10/pip per lot) = 0.33 lots

Kelly f* = W − (1 − W)/R // growth-optimal; trade quarter-Kelly or less in practice

Drawdown asymmetry

DrawdownGain needed to recover
−10%+11.1%
−20%+25%
−30%+42.9%
−50%+100%

Recovery gain = L/(1−L). At 1% risk, even 10 straight losses = a −9.6% drawdown, fully recoverable. At 5% risk the same streak = −40%.

Trade management

  • Scaling out — bank a partial at +1R or the first level, trail the rest.
  • Breakeven — move the stop to entry only after a structure event (a new swing forms in your favour), never on a timer.
  • Trailing — behind swing points, the chandelier (3×ATR), or a MA/Kijun. Mechanical trailing beats discretionary exits.
  • Correlation cap — two open trades in the same risk direction are one trade in disguise: halve each, and cap total open risk (e.g. ≤ 3%).
The non-negotiables
  • Fixed risk per trade: 0.5–1% (2% absolute ceiling for proven edges).
  • Daily loss stop: −3R or −3% — flat, done for the day. Consecutive-loss stop: 3 losses → walk away.
  • Never widen a stop. Never average down a loser. Never take a revenge/recovery trade.
  • Bracket at entry (OCO) so exits never depend on in-the-moment courage.
  • Journal every trade: setup grade, R outcome, reason for entry and exit.

17Backtesting & System Development

How to find out whether any of the preceding actually has an edge — without fooling yourself.

The validation pipeline

  • In-sample vs out-of-sample. Build and tune on one slice; test untouched on another. A strategy that only works in-sample is curve-fit.
  • Walk-forward analysis. Roll the in/out windows forward, re-optimising each step — the gold standard for robustness.
  • Monte Carlo. Shuffle trade order thousands of times to see the distribution of worst-case drawdowns.

Worked example — reading a 100-trade sample

42 wins avg +2.1R · 58 losses avg −1.0R
Gross profit = 42 × 2.1 = 88.2R   Gross loss = 58 × 1.0 = 58R
Profit factor = 88.2 / 58 = 1.52 // > 1.3 → viable
Expectancy = (88.2 − 58) / 100 = +0.30R / trade

The biases that fake an edge

Overfitting

Too many parameters fit to noise. Prefer few rules that work across a plateau of parameter values.

Look-ahead bias

Using information not available at decision time (repainting indicators). Silent and lethal.

Survivorship bias

Testing only instruments that still exist ignores the delisted losers.

Data-snooping

Test enough variants and one looks great by luck. Reserve a final hold-out set you touch exactly once.

Performance metrics

MetricFormula / definitionHealthy zone
Profit factorGross profit ÷ gross loss> 1.3 (≥ 1.5 good)
ExpectancyAverage R per tradePositive, stable across regimes
Sharpe ratio(Return − risk-free) ÷ σ of returns> 1 decent, > 2 strong
Max drawdownLargest peak-to-trough drop< 15–20%
Sample sizeTrades tested≥ 100 before any conclusion
How to apply this to your own trading
  • Your journal is a backtest. Tag every live trade with its setup name; after 30+ trades per setup, compute PF and expectancy per setup — then cut the losers.
  • Never tune on the last 10 trades. A true 45%-win system loses 5 in a row about once every 20 sequences. Judge over ≥ 100.
  • Try to break it: shift entries one bar, add realistic spread/slippage, halve the sample.
  • Promote slowly: backtest → paper/sim → small live size → full size.

18Trading Psychology

A profitable system run by an undisciplined operator loses money. This part you can't outsource to an indicator — but you can engineer around it.

The biases that cost you money

BiasHow it shows upCounter
Loss aversionCutting winners early, holding losersMechanical exits; bracket orders at entry
Recency biasOverweighting the last few trades; revenge after a lossThink in 100-trade samples
Confirmation biasSeeing only signals that fit your positionWrite the opposite case before entry
Gambler's fallacy"I'm due for a win" — oversizingEach trade is independent; size stays fixed
Sunk costAdding to a loser to justify itNever average down a losing trade
FOMOChasing a move that already left"Missed" is a 0R outcome — better than a bad entry
OverconfidenceOversizing after a winning streakFixed-fractional sizing, always

Process over outcome

Any single trade is mostly noise. Judge yourself on decision quality, not results. A good process / bad outcome trade is a success — repeat it. A bad process / good outcome trade is the most dangerous event in trading — it pays you to build the habit that eventually ruins you.

The pre-trade routine (60 seconds, every trade)
  1. State check: calm, rested, not chasing a loss? If no — no trade.
  2. Daily limits: am I inside my −3R day and 3-consecutive-loss caps?
  3. The setup: grade (A/B/C), confluence list, HTF alignment.
  4. The numbers: entry, stop, target, size before clicking; R:R ≥ 1:2.
  5. Bracket in (OCO). Then hands off.
The hard truth Most traders don't fail because their analysis is wrong — they fail because they can't execute correct analysis under pressure. Fix the operator and a mediocre system becomes profitable. The reverse is never true.

19Common Pitfalls

A concentrated list of the mistakes that recur across every market and every experience level.

PitfallWhy it hurtsFix
Indicator stackingFive momentum tools = one signal counted five timesOne tool per job (trend, momentum, volatility, volume)
Counter-trend tradingFighting the higher timeframe is low-probabilityTrade with HTF bias; treat fades as rare exceptions
Ignoring regimeTrend tools whipsaw in ranges and vice-versaName trending vs ranging before choosing tools
Pattern without locationA signal mid-range is noiseDemand a tested level + confluence
No confirmationAnticipating breaks gets you trappedWait for the close / retest before acting
Moving the stopWidening turns a small loss into a large oneStop is fixed at entry; only trail in profit
Over-tradingForcing B/C-grades bleeds the accountWait for A-grades; flat is a position
Revenge tradingWinning it back = oversized, unplanned riskHard consecutive-loss stop; walk away
Risking too muchLarge per-trade risk — one streak ends youFixed 0.5–1% per trade
Curve-fittingA backtest tuned to noise fails liveOut-of-sample + walk-forward; fewer parameters
The meta-pitfall Constantly switching systems after a few losses ("strategy hopping") guarantees you never accumulate the sample size to know if anything works. Pick a tested edge, execute it for 100+ trades, and change it based on data — never on the feeling from the last red day.

20A Complete Workflow

Everything in this handbook assembled into one repeatable top-down routine — bias before setup, setup before trigger, trigger before risk.

The seven-step routine

  1. Regime & bias (HTF). Name the state (trend/range) and direction from structure (§02); confirm with the 200 MA and slope (§07); check volatility (ATR). This decides which direction you're allowed to trade.
  2. Map the levels. Prior swing highs/lows, flip levels, supply/demand zones, the active Fib of the last impulse, pivots, naked POCs.
  3. Find the setup (MTF). Is price approaching a mapped level with a recognisable structure (flag, double bottom, order block, golden pocket, spring)?
  4. Score the confluence. Count independent factors by family (§11). Grade A/B/C. Not an A — aligned, 3+ factors, with trend? Pass.
  5. Wait for the trigger (LTF). A confirming candle at the zone (rejection/engulfing/pin, §05) or a clean break-and-retest. No trigger, no trade.
  6. Define risk first. Stop where the idea is invalidated. Target the next opposing level. R:R ≥ 1:2 or skip. Size from the stop: 0.5–1%.
  7. Execute & manage mechanically. Bracket in (OCO); breakeven on a structure event; trail behind swings / chandelier; bank partials per plan. Journal it. Then leave it alone.

A worked example — EUR/USD swing long

Step 1 — bias: Daily prints HH/HL above a rising 200 SMA; ADX(14) = 28 and rising with +DI on top — a confirmed, strengthening uptrend. Longs only.

Step 2 — levels: the shelf at 1.0850 is a prior breakout (flip level) that also lands in the 50–61.8% golden pocket of the last Daily impulse, with the rising trendline passing through this week.

Step 3–4 — setup & grade: on H4, price is pulling back in a tidy bull flag toward that zone on shrinking volume; H4 RSI has cooled from 68 to 44 — the uptrend pullback zone. Factor count: flip level + golden pocket + trendline + EMA20 + with-trend = A-grade.

Step 5 — trigger: on M15, price sweeps the flag's low into 1.0848, then prints a bullish engulfing that closes back above the level at 1.0858. Enter 1.0860.

Step 6 — risk: stop below the sweep wick at 1.0822 (−38 pips = 1R ≈ 1.4×ATR). Target the prior high 1.0985 (+125 pips = 3.3R). Risk 1% of $10,000 = $100 → 0.26 lots.

Step 7 — manage: OCO bracket in. Partial at +1R; stop to breakeven after the first HL forms above entry. Trail the remainder behind H4 swing lows. Win or lose, it was an A-grade trade executed to plan — the only thing you control.

The whole game in one line Higher-TF bias → level → pattern → confluence grade → trigger candle → risk-defined entry → mechanical management. Skip any step and you're guessing. Run every step and you've converted "reading charts" into a process with an edge you can measure and improve.

21Quick-Reference Glossary

The vocabulary of this handbook in one place.

TermMeaning
OHLCOpen, High, Low, Close — the four prices defining a candle/bar.
HH / HL / LH / LLHigher high / higher low / lower high / lower low — building blocks of trend structure.
BOSBreak of Structure — price breaks a swing in the trend direction (continuation).
CHoCHChange of Character — price breaks a swing against the trend (early reversal signal).
ConfluenceMultiple independent factors agreeing on one zone — the core of an edge.
Pin barA candle with a small body and a dominant rejection wick (≥ 2× body).
Golden pocketThe 50–61.8% Fibonacci retracement zone; the prime trend-entry band.
PRZPotential Reversal Zone — the D-point completion of a harmonic pattern.
DivergencePrice and a momentum oscillator disagree (regular = reversal, hidden = continuation).
Golden / death cross50 MA crossing above / below the 200 MA — regime signals.
ATRAverage True Range — a normal bar's movement; the basis for stops and sizing.
Chandelier exitTrailing stop at highest high − 3×ATR (longs).
VWAPVolume-Weighted Average Price — intraday fair-value line and institutional benchmark.
POC / Value AreaVolume Profile's most-traded price / the ~70% range of fair value.
80% ruleRe-entry into the value area that holds tends to traverse to its other side.
Order blockThe last opposing candle before an impulsive move; an institutional footprint zone.
FVG / imbalanceFair Value Gap — a price inefficiency between wicks that often gets filled.
Liquidity sweepA brief push beyond an obvious high/low to trigger stops before reversing.
Premium / discountUpper/lower half of a range vs its 50% equilibrium — where to sell/buy.
Spring / UTADWyckoff's false breakdown (bullish) / upthrust after distribution (bearish).
SOS / LPSSign of strength / last point of support — Wyckoff's breakout and its retest.
Impulse / correctionElliott's 5-wave trend move / 3-wave counter-move.
R / R-multipleOne unit of risk (entry-to-stop); all outcomes measured as multiples of it.
ExpectancyAverage profit per trade in R: (Win%×avgWin) − (Loss%×avgLoss).
Profit factorGross profit ÷ gross loss; > 1.3 is the rough viability floor.
DrawdownPeak-to-trough equity decline; recovery needs L/(1−L) — the asymmetry that justifies small risk.
Walk-forwardRolling out-of-sample testing that simulates live re-optimisation.
RegimeThe prevailing market condition — trending vs ranging, calm vs volatile.
Kelly criterionGrowth-optimal bet fraction f* = W − (1−W)/R; trade a fraction of it.

That is the full arc — from a single candlestick to a complete, risk-managed, testable process, with every tool's formula and playbook on one shelf. Master structure first, add tools sparingly, demand confluence, and let risk management keep you in the game long enough for your edge to compound. Reign your edge.