Part II Chapter 7 — Filtering with Western Indicators

{fullwidth}

The Power of Japanese Candlestick Charts by Fred K. H. Tam

Note: This is the chapter that formally launches Part II (“Advanced Candlestick Techniques”). It is long and indicator-dense (10 Western tools), but every single indicator is applied through the exact same three-step template — this guide highlights that repeating structure so the volume of material is easier to hold in memory.

1. Core Concept & Theoretical Foundation

Primary objective of this chapter: To deliver on the promise made at the end of Chapter 6 — to teach the mechanics of filtering, i.e., pairing candlestick pattern signals with Western technical indicators to define the primary trend and screen out false candlestick signals. The chapter systematically works through ten Western tools (Moving Averages, MACD, RSI, Stochastic, Momentum, Williams %R, DMI, CCI, Volume, Bollinger Bands) plus an introduction to Elliott Wave Theory, applying an identical analytical template to each.

Foundational theory: The chapter opens by quoting two authorities to justify the entire filtering approach:

Steve Nison (Japanese Candlestick Charting Techniques): candlestick methods are valuable alone, but become “even more powerfully significant” when they confirm a Western technical signal.
Arthur Sklarew (Techniques of a Professional Commodity Chart Analyst): the more technical indicators that align at the same price area, the greater the chance of an accurate forecast — the theoretical basis for combining multiple tools rather than relying on any single one.

Underlying market psychology / historical context: No new pattern psychology is introduced here (that’s Chapters 2–5’s territory); instead, the chapter treats each Western indicator itself as a different psychological lens — moving averages read crowd sentiment as a lagging trend-following signal, oscillators (RSI, Stochastic, Momentum, Williams %R, CCI) measure the speed/intensity of buying vs. selling pressure to flag overbought/oversold exhaustion, and volume measures the conviction behind a move. Elliott Wave theory (attributed to accountant Ralph Nelson Elliott, who discovered the Wave Principle at the turn of the 20th century) is introduced as a structural map of where in a bull/bear cycle the market currently sits, so candlestick reversal signals can be weighted differently depending on which wave they appear in.

The universal two-step filtering process (repeated as the header logic for the entire chapter):

Define the trend — using trendlines, moving averages, or oscillators. There are only three trend classifications: up, down, sideways.
Trade only in the direction of that trend — using candlestick patterns to time entries and exits.

Three trading scenarios that frame the whole chapter:

Scenario 1 (Bull market): Take bullish candlestick signals to enter/add; ignore bearish candlestick signals except to close (not reverse) longs, until the trend itself turns bearish.
Scenario 2 (Bear market): Mirror image — take bearish signals to enter/add shorts (noting that in Malaysia, short-selling stocks is prohibited, so this applies to futures for local traders); ignore bullish candlestick signals except to cover shorts.
Scenario 3 (Overbought/oversold): Look for bearish reversal patterns to exit when overbought; look for bullish reversal patterns to enter when oversold.

2. Key Terms & Definitions

Term	Definition
Filtering / Rule of Multiple Techniques	Combining candlestick signals with Western technical indicators to confirm trades and reduce false signals (formally introduced in Ch. 6, operationalized here).
Triple Screening	Using three time-frame charts (e.g., monthly/weekly to determine trend direction, daily to trigger signals, and optionally hourly for earlier signals in the direction of the longer-term trend) to screen trade signals. Credited conceptually to the idea of multi-timeframe confirmation; the author references Marcel Link’s High Probability Trading and his own second book for more detail.
Golden Cross	When price closes above its moving average (single MA method) or when a shorter MA crosses above a longer MA (dual MA method) — a bullish signal.
Dead Cross	The bearish mirror of the Golden Cross — price closes below its MA, or a shorter MA crosses below a longer MA.
MACD (Moving Average Convergence Divergence)	A trend indicator developed by Gerald Appel, built from the difference between two EMAs (commonly 12- and 26-day) plotted against a signal line (commonly a 9-day EMA of the MACD line itself).
RSI (Relative Strength Index)	An oscillator developed in 1978 by J. Welles Wilder Jr., scaled 0–100, measuring the internal strength of price advances vs. declines over a period (commonly 14 days).
Stochastic Oscillator	Developed by George Lane, compares the latest closing price to the total price range over a period (commonly 5 or 14 days depending on use case), scaled 0–100, using %K and %D lines.
Momentum	The most basic oscillator; compares the current closing price to the closing price n days ago (commonly 10 days), plotted against a zero line.
Williams’ %R	Developed by Larry Williams, closely resembles the Stochastic Oscillator; scaled 0–100 (or inverted depending on software), commonly using a 9-day period.
DMI (Directional Movement Index)	Developed by J. Welles Wilder Jr., plots +DMI and –DMI to assess trending quality; paired with the ADX (Average Directional Index) to judge trend strength.
CCI (Commodity Channel Index)	Developed by Donald Lambert; compares a security’s mean price to its average mean price over a period, normalized to oscillate around +/–100.
Bollinger Bands	Developed by John Bollinger; a moving average (commonly 20-period simple) plus/minus a multiple (commonly 2) of standard deviations, widening/narrowing with volatility.
Elliott Wave Principle	Discovered by Ralph Nelson Elliott; describes market cycles as 5 impulse/corrective waves in the trend direction (Waves 1–5) followed by 3 corrective waves against it (Waves a, b, c), forming an 8-wave complete cycle.
Impulse Waves	Waves 1, 3, and 5 in Elliott’s model — waves that move with the primary trend.
Corrective Waves	Waves 2 and 4 (which correct Waves 1 and 3, respectively) plus the a-b-c sequence that corrects the entire 1–5 impulse sequence.
Selling Climax	A sudden, unusually large increase in volume after an extended decline, suggesting the market may be bottoming.
Divergence	When price makes a new high/low but an oscillator (RSI, CCI, MACD) fails to confirm with a matching new high/low — a warning of weakening momentum, most meaningful when the oscillator is already in overbought/oversold territory.
Extreme Point Rule (DMI)	A whipsaw-reduction rule: on the day +DMI and –DMI cross, note the extreme price (low of that day if now long-biased, high of that day if now short-biased); wait for price to move beyond that extreme before actually entering the trade.

3. Anatomy, Rules, & Construction Mechanics

Every indicator below follows the same “Rule” format: a simple inequality that defines bullish vs. bearish, paired with candlestick patterns to use for timing.

Moving Averages

Simple Moving Average (SMA): equal weight to all prices in the period. Formula: sum of closing prices over N days ÷ N.
Weighted Moving Average (WMA): more weight assigned to recent prices (weighting scheme is user preference).
Exponential Moving Average (EMA): includes all historical prices, but the most recent period carries the most weight, decaying smoothly for older data.
Rule (Single MA Crossover): Close > MA = Bullish (Golden Cross) = Buy; Close < MA = Bearish (Dead Cross) = Sell. Author’s parameter: 50-day SMA.
Rule (Dual MA Crossover): Shorter MA > Longer MA = Bullish (Golden Cross) = Buy; Shorter MA < Longer MA = Bearish (Dead Cross) = Sell.

MACD

Formula: MACD = EMA1 − EMA2 (typically 12-day and 26-day EMAs); signal line = 9-day EMA of the MACD line.
Rule: MACD line crosses above signal line = buy; crosses below = sell.
Author’s preferred parameters: 5, 34, 5 (popularized by Dr. Bill Williams) for more timely crossings; also used to distinguish Elliott Wave 3 (most extreme histogram reading) from Wave 5 (less extreme, producing divergence).
Can also be read as a histogram (overbought/oversold extremes) or via zero-line crossings.

RSI

Formula: RSI = 100 − (100 / (1 + (Avg. up price change / Avg. down price change))).
Rule: RSI > 70 = Overbought; RSI < 30 = Oversold; RSI > 50 = Bullish = Buy; RSI < 50 = Bearish = Sell.
Also used for divergence (price makes new high/low that RSI fails to confirm) — most meaningful in overbought/oversold zones.

Stochastic Oscillator

%K and %D lines, scaled 0–100. Formula for %D (slow): average of three %K(slow) readings ÷ 3.
Author’s parameter: 14 days for short-term trend identification.
Rule: Overbought > 80, Oversold < 20; %K > %D = Bullish = Buy; %K < %D = Bearish = Sell.

Momentum

Formula: Momentum = (Closing price today) − (Closing price n days ago). Author’s parameter: n = 10 days.
Rule: Momentum > 0 = Bullish = Buy; Momentum < 0 = Bearish = Sell. Signals given at zero-line crossings.

Williams’ %R

Formula: %R = (Highest high of X period − Current close) / (Highest high of X period − Lowest low of X period). Author’s parameter: X = 9 days.
Rule: Readings of 80–100% = market is oversold (standard Williams convention) or overbought (inverted convention used in some software) — the chapter explicitly flags this inconsistency across platforms. Wait for price to actually turn before trading an overbought/oversold reading. %R notably tends to peak/trough a few days before price does.

DMI (Directional Movement Index)

Plots 14-period +DMI and –DMI together, typically alongside ADX.
Rule: +DMI > −DMI = Bullish = Buy; +DMI < −DMI = Bearish = Sell.
Works best when ADX > 25; avoid trend-following signals when ADX < 20. For a valid buy/sell signal, ADX should be rising at the time of the DMI crossover (or begin rising within a few bars).
Extreme Point Rule: on a DMI crossover day, mark the extreme price (low if now bullish-biased, high if now bearish-biased); only enter once price moves beyond that extreme.

CCI (Commodity Channel Index)

Normalized to oscillate roughly within +/–100.
Rule: CCI > 0 = Bullish = Buy; CCI < 0 = Bearish = Sell; readings beyond +/–100 imply overbought/oversold. Also used for divergence analysis.

Volume

No formula — a qualitative confirming filter: greater volume = greater force behind a move; used to confirm breakouts.
An unusually large volume spike after an extended decline may signal a selling climax (potential bottom).

Bollinger Bands

Constructed from a moving average (Bollinger’s default: 20-period SMA) plus/minus a multiple of standard deviation (Bollinger’s default: 2 deviations). Periods under 10 are noted as not working well.
Key behaviors: bands narrow in quiet markets, expand in volatile ones; a price move outside the upper band implies uptrend strength, outside the lower band implies downtrend strength; a sharp move outside the bands followed by immediate retracement signals exhaustion; a top/bottom made outside the bands followed by one made inside the bands calls for a trend reversal; price tends to travel from one band to the other (useful for price targets).

Elliott Wave Principle (introductory treatment)

A complete market cycle = 8 waves: 5 impulse waves (1, 3, 5, moving with the trend) + 3 corrective waves (2 and 4 within the impulse sequence, plus a-b-c correcting the whole sequence).
Each impulse wave subdivides fractally into smaller 5-wave sequences, each corrective wave into smaller 3-wave (a-b-c) sequences — across nine degrees of magnitude, from the multi-century “Grand Super-Cycle” down to a “subminute” degree.

4. Practical Trading Application

The repeating three-part “Proper Action” template applied to every single indicator in this chapter:

Directional signal → matching candlestick patterns. For each indicator’s bullish state (e.g., Golden Cross, MACD > signal, RSI > 50, %K > %D, Momentum > 0, +DMI > −DMI, CCI > 0), the author lists specific bullish candlestick confirmations to trade off of: Hammer, Inverted Hammer, Bullish Engulfing, Piercing Line, Morning Star, Doji-Star (bullish), Fred Tam’s White Inside Out Up. For each bearish state, the bearish mirror list: Shooting Star, Hanging Man, Bearish Engulfing, Dark Cloud Cover, Evening Star, Doji-Star (bearish), Fred Tam’s Black Inside Out Down.
Overbought/oversold extremes → reversal-specific patterns. At RSI > 70% or CCI > 100 or %R overbought zones, the author narrows the pattern list further to the strongest top-reversal formations (Tweezers Top, Bearish Meeting Line, Bearish Harami, Three-River Evening Star); the mirror applies at oversold extremes (Tweezers Bottom, Bullish Harami, Three-River Morning Star).
Rule of Multiple Techniques (the trend-following discipline layer), applied identically for every indicator:
- Bullish trend intact → look for bullish candlesticks to buy/add; ignore bearish candlestick signals, or at most use them only to close longs (never to initiate a short).
- Bearish trend intact → look for bearish candlesticks to sell/add shorts; ignore bullish candlestick signals, or at most use them only to cover shorts (never to initiate a long).

Multi-timeframe execution guidance (Triple Screening):

Futures traders: use daily and weekly charts to establish trend direction; use 15-/30-minute charts to trigger entries, but only when aligned with the higher timeframe trend.
Malaysian stock traders: use weekly/monthly charts for trend, daily charts to trigger signals; hourly charts may be used for even earlier signals, but only in the direction of the longer-term trend.

Indicator-specific risk-management notes:

DMI/ADX: don’t trade the DMI crossover system when ADX < 20 (non-trending market); use the Extreme Point Rule to avoid whipsaws — wait for price to clear the crossover-day extreme before entering.
Overbought/oversold oscillators generally (RSI, Stochastic, %R, CCI): the chapter explicitly warns that a security can remain overbought/oversold for a long time while price keeps climbing/falling — don’t sell purely because a reading looks “overbought”; wait for price itself to turn before acting.
Volume: treat a volume spike with a price rise as a possible top warning (pair with bearish single candles: Shooting Star, Hanging Man, Doji, Doji-Star); treat a volume spike with a price decline as a possible bottom warning (pair with bullish single candles: Hammer, Inverted Hammer, Doji, Doji-Star).
Bollinger Bands: the specific reversal cue is price moving outside a band and then re-entering it — that re-entry is the trigger to look for a reversal candlestick pattern (bearish patterns after re-entering from above the upper band; bullish patterns after re-entering from below the lower band).
Elliott Wave: the practical, worked chart examples (Figures 7.16–7.17) show using MACD to spot the top of Wave 3 (historically the most extreme move) and riding it, then watching for bearish divergence and a moving-average cross to exit — a direct illustration of stacking multiple indicators (Elliott Wave structure + MACD + moving average) on top of candlestick confirmation signals (Fred Tam’s White Inside Out Up, Hammer confirmation) in a single trade plan.

5. Active Recall Quiz

What two authorities does the author quote to justify the filtering approach, and what is the core claim each one makes about combining candlestick analysis with other technical tools?
State the two-step “define the trend, then trade with the trend” process that frames this entire chapter, and explain why — under the Rule of Multiple Techniques — a trader should still avoid initiating a short position even when a strong bearish candlestick pattern appears during a confirmed bullish trend.
For four of the following indicators — MACD, RSI, Stochastic, Momentum, DMI, CCI — state the simple bullish/bearish rule (the inequality) the author uses to define trend direction.
What is the “Extreme Point Rule” in DMI trading, and what specific market condition (measured by ADX) does Wilder say should make a trader avoid using a trend-following DMI system altogether?
According to the chapter’s guidance on overbought/oversold oscillators (RSI, Stochastic, %R, CCI), why is it risky to sell a security purely because an indicator shows “overbought,” and what should a trader wait for before acting on that signal instead?