Technical Analysis of the Financial Markets — John J. Murphy
Macro Overview & Strategic Value
Chapter 10 supplies the direct counterpart to Chapter 9’s moving averages: oscillators, designed specifically to extract signal from the third-to-half of market time when trend-following tools underperform in sideways, range-bound conditions. Murphy’s core thesis is that oscillators measure velocity (momentum) rather than direction, making them uniquely capable of flagging overbought/oversold extremes and trend exhaustion before that exhaustion becomes visible in price action itself.
The practitioner value is significant: oscillators are explicitly framed as a secondary, subordinate tool that must never override basic trend analysis — a discipline Murphy repeats throughout the chapter to guard against the single most common oscillator misuse (fading a strong trend because it’s “overbought”). Getting this subordination right is what separates a properly-timed pullback entry within a trend from a costly counter-trend trade against strong momentum.
Structurally, the chapter also introduces the Principle of Contrary Opinion, adding an explicit third analytical dimension — market sentiment/positioning — alongside price and volume, directly extending the COT-based contrarian logic from Chapter 7 into a quantified, poll-based framework used across futures and stocks.
Core Concepts & Mechanics
- Momentum construction and leading quality — Calculated as the latest close minus the close N days ago, oscillating around a zero line; because it measures the rate of change rather than price level, it turns before price itself, giving traders an explicit early-warning function ahead of visible trend reversals.
- Rate of Change (ROC) — A ratio-based variant (latest close divided by close N days ago, ×100) oscillating around 100 instead of zero; functionally identical to momentum in interpretation but scaled differently.
- Dual moving-average histogram — Plotting the difference between two moving averages as a histogram reveals divergences and pinpoints crossover timing earlier than the raw crossover signal itself, giving traders advance warning of a coming trend-following signal.
- Commodity Channel Index (CCI) normalization — Uses mean deviation to constrain oscillator values to a fixed ±100 range regardless of the underlying instrument, making the same overbought/oversold thresholds comparable across different markets.
- RSI construction and failure swings — Bound to a fixed 0-100 scale via smoothed average gains/losses (default 14-period); overbought >70, oversold <30, with the midpoint (50) acting as trend support/resistance; a “failure swing” (a second RSI peak/trough failing to confirm price’s extension) is Wilder’s single most important signal.
- First-move vs. second-move overbought/oversold discipline — The first oscillator move into extreme territory during a strong trend is only a warning, not an exit signal; the actionable signal is the second move into the danger zone failing to confirm a new price extreme — an explicit rule protecting traders from exiting a strong trend prematurely.
- Stochastics (%K/%D) mechanics — Measures where the close sits within the recent high-low range (0-100 scale); %D (a smoothed average of %K) generates the primary divergence signal, with the actual buy/sell trigger occurring when %K crosses %D within an overbought (>80)/oversold (<20) zone.
- MACD (Moving Average Convergence/Divergence) — Combines dual exponential moving average crossover logic (typically 12/26-period) with oscillator-style zero-line and overbought/oversold behavior via a signal line (9-period EMA of the MACD line); the MACD histogram (the spread between the two lines) turns before the actual crossover, giving early exit warnings for existing positions.
- Principle of Contrary Opinion — Adds a third, sentiment-based analytical dimension via polled bullish/bearish consensus (e.g., Bullish Consensus, Investors Intelligence); when an overwhelming majority (>75-90%) already holds one directional view, they’re assumed fully positioned already, leaving insufficient fresh capital to extend the trend — making extreme one-sided sentiment itself a contrarian signal.
Technical Terminology & Reference Table
| Term | Operational Definition |
|---|---|
| Momentum | M = V − Vx; latest close minus close x days ago, oscillates around zero |
| Rate of Change (ROC) | 100 × (V / Vx); ratio-based momentum variant, oscillates around 100 |
| Divergence | Oscillator fails to confirm a new price high/low; primary reversal warning |
| Failure swing | RSI peak/trough (in extreme territory) that fails to exceed the prior one |
| RSI (Relative Strength Index) | Smoothed 0-100 oscillator; >70 overbought, <30 oversold, 50 = midpoint |
| Stochastics (%K, %D) | %K = 100×[(C−L14)/(H14−L14)]; %D = 3-period MA of %K; >80/<20 = extremes |
| Larry Williams %R | Inverse-scaled variant of stochastics measuring close’s position in the range |
| MACD | Difference between 12- and 26-period EMAs; signal line = 9-period EMA of that difference |
| MACD histogram | Bar plot of the spread between MACD line and signal line |
| Contrary Opinion | Theory that overwhelming consensus positioning signals trend exhaustion |
| Bullish Consensus | Polled sentiment index; >75% overbought, <25% oversold, 55% = equilibrium |
The Author’s Market Philosophy
Murphy’s model treats oscillators as measuring the psychological over-extension of crowd behavior — velocity, not direction, reveals when a trend has moved “too far too fast” relative to what fundamentals or sustainable participation can support. He is emphatic that edge from oscillators comes from disciplined subordination to trend, not from independent signal generation; his repeated warnings against fading strong trends on overbought readings reflect an assumption that most oscillator losses come from misapplied contrarian impulses rather than the tool itself being flawed. The Contrary Opinion framework extends this same logic to aggregate trader psychology directly: when positioning becomes lopsided, the “buying/selling power” runs out mechanically (all bulls already committed), giving sentiment extremes genuine predictive content rather than being merely descriptive.
Systemic & Portfolio Integration
Oscillators are explicitly positioned as the necessary complement to Chapter 9’s trend-following tools — this pairing (trend tool + oscillator, filtered by the ADX regime indicator introduced later) forms the backbone of most systematic trading frameworks that must operate across both trending and non-trending regimes. The weekly-for-direction, daily-for-timing combination approach recurs as the book’s standard multi-timeframe risk-management structure, while Contrary Opinion sentiment extremes function as an independent confirming/disqualifying filter layered on top of standard trend and oscillator signals.
Important Formulas, Data, or Initial Examples
- Momentum formula: M = V − Vx (V = latest close, Vx = close x days ago); 10-day period most common.
- ROC formula: 100 × (V/Vx), oscillating around a 100 midpoint.
- RSI construction: 14-period average up-move divided by average down-move (RS), inserted into the RSI formula to yield a fixed 0-100 scale.
- Stochastics %K formula: %K = 100 × [(C − L14)/(H14 − L14)]; %D = 3-period MA of %K (slow stochastics = further 3-period smoothing).
- MACD defaults: 12- and 26-period EMAs for the MACD line; 9-period EMA of that line as the signal line.
- Bullish Consensus thresholds: equilibrium ~55%; overbought/oversold warning zones at 75%/25%; extreme contrarian zones at 90%/20%.
- Investors Intelligence thresholds: bullish readings >55% = excess optimism (negative); <35% = excess pessimism (positive); breadth readings >70%/<30% (of stocks above 10/30-week averages) signal overbought/oversold.
Active Recall Evaluation
- Why does Murphy insist that the first move of an oscillator into overbought/oversold territory during a strong trend should not be treated as an exit signal, and what does the second move need to show to become actionable?
- Explain mechanically why the MACD histogram consistently turns before the actual MACD/signal-line crossover, and what practical use Murphy assigns to that early turn.
- How does the Contrary Opinion framework’s “strong hands vs. weak hands” argument work when 80% of traders are positioned on one side of a zero-sum futures market?
- Why does Murphy caution against taking a contrarian position while open interest is still rising, even when sentiment polls show an extreme reading?
- What specific problem in constructing a simple momentum line does Wilder’s RSI formula solve, beyond just bounding the oscillator to a fixed 0-100 range?
Answer Key (spoiler)
- A strong trend typically pushes an oscillator into extreme territory quickly, and if a trader exits (or worse, fades) on that first extreme reading, they risk abandoning a still-powerful trend prematurely; the second move into the danger zone only becomes a genuine warning if it fails to confirm a corresponding new price extreme (forming a double top/bottom on the oscillator itself), which constitutes an actual divergence rather than just a normal extension of trend strength.
- The histogram plots the spread between the faster MACD line and the slower signal line; as that spread narrows (even while the faster line is still above the slower one), the histogram bars shrink toward zero before the lines actually cross — Murphy uses this early narrowing as an advance warning to tighten protective stops or take defensive action on existing positions, though not as a standalone signal to initiate new counter-trend trades.
- Since futures is a zero-sum game, every long position is matched by an equal-sized short position; if 80% of traders (by headcount) hold longs, the remaining 20% of traders must collectively hold short positions equal in total size to the majority’s longs, meaning each of those minority traders is, on average, carrying a much larger position and must be better capitalized — making them “strong hands” relative to the more numerous but thinly-capitalized “weak hands” majority who would be forced to liquidate on any adverse price move.
- Rising open interest indicates fresh capital is still actively entering the market in the direction of the current trend, which increases the odds the trend continues regardless of how lopsided sentiment appears; a genuine contrarian setup requires that fresh capital commitment to be exhausted, which only becomes visible once open interest flattens or begins declining — taking a contrarian position while open interest still climbs risks fighting a trend that still has genuine new participation behind it.
- Beyond providing a fixed 0-100 range, Wilder’s RSI formula solves the problem of erratic, jumpy momentum-line movement caused by a single old data point suddenly dropping out of the calculation window (e.g., a sharp price move from 10 days ago disappearing abruptly); by using smoothed average gains and average losses rather than a raw point-in-time price difference, RSI removes that artificial jumpiness while still preserving genuine trend and extreme-reading information.