Dow Theory
Definition
The oldest framework in technical analysis (100+ years), built from Charles Dow’s writing and formalized after his death by William Peter Hamilton and Robert Rhea. It reads the market through the correlation between two indices Dow created — the Dow Jones Industrial Average (DJIA) and the Dow Jones Transportation Average (DJTA) — and distills six principles.
Everything downstream in technical analysis traces back here: Elliott Wave, Andrews’ Pitchfork, trend lines, chart patterns. Any technical-analyst certification tests it.
Dow’s own framing was tidal: markets move in trends like waves hitting a beach, leaving patterns in the sand that mark where high and low tide occurred.
Core Ideas
The six principles
① Price discounts everything. All knowable market information — including expectations about future events — is already in the price. Supply/demand factors, fundamentals, and even unforecastable shocks (earthquakes, disasters) end up reflected in the average. This principle is the justification for technical analysis itself: if the chart contains all factors, analyzing the chart suffices.
② Trends come in three degrees, nested inside each other and coexisting:
| Trend | Character | Duration |
|---|---|---|
| Primary | the bull or bear market — overall direction | months to years |
| Secondary | corrections against the primary: declines within a bull market, rallies within a bear | weeks to months |
| Minor | daily noise | hours to weeks |
③ A primary trend has three phases.
Bull market:
- Accumulation — begins at the tail of the prior decline
- Public participation — the trend becomes legible, attracting the crowd and technical analysts; participation accelerates the move into a buying frenzy
- Excess — near the top, smart institutional money starts selling while the crowd, unaware, keeps buying
Bear market:
- Distribution — earlier buyers unload to a still-willing bid, but demand < supply and price starts falling
- Public participation — retail sells in a rush, driving price lower
- Panic — indiscriminate selling with no regard for value; price can fall far below value, then volatility subsides and a new cycle begins
The operational reading: accumulate after the crash, distribute after the run — the smart-money side of each phase.
④ The averages must confirm each other. One index is not enough. DJIA and DJTA must point the same way for a trend to be credible; when they disagree it is divergence — e.g. the Industrials making a new high while the Transports fail to, which suggests the uptrend may be changing. The modern generalization: check correlation across related markets (S&P 500, Nasdaq 100, broader indices) rather than only the original two.
⑤ Volume must confirm the trend.
- In an uptrend: volume rises with price, falls on declines.
- In a downtrend: volume rises as price falls, falls on rallies.
Dow treated volume as a secondary indicator to price: within a healthy primary trend volume oscillates, and steadily contracting volume signals an approaching reversal.
⑥ A trend persists until a clear reversal signal. Analogous to inertia — once started, a move continues until an outside force acts. This is the theoretical basis for trend-following being the dominant strategy class. The reversal signal is price breaking the prior high/low that defined the trend. The warning attached: a genuine reversal is hard to distinguish from a secondary correction.
What it seeded
- Trend lines — connect lows for an up-trendline, highs for a down-trendline
- Chart patterns — double/triple tops and bottoms, head-and-shoulders — always read together with volume
- Advance/decline and divergence indicators — originally DJIA vs Transports, since evolved into many intraday variants
For intraday traders specifically, four Dow-derived fundamentals still matter: trend (without it you trade blind), the closing price (carries more information than open or high/low), volume (for supply/demand and support/resistance), and divergence indicators.
Honest limits
- Charles Dow never claimed it predicts the market. He described it as a barometer of general market direction — not a forecasting tool, not investment guidance.
- Rhea stressed the same: it is a device for improving a speculator’s or investor’s knowledge, not a complete technical system that works detached from economic fundamentals and current market conditions.
- Empirical record is mixed: Alfred Cowles’ 1934 test found generally low returns; a later tracking analysis argued Cowles’ work was insufficient and put Dow-signal returns at 11.4% vs 10.6% for buy-and-hold — an edge, but a thin one.
- The source’s own conclusion: the goal is not to win a speculative game or find an infallible method, but to obtain “a vague correctness” (模糊的正确).
The reason it endures: it is really a theory of market psychology — the phases describe crowd behavior, and the patterns are statistical summaries of that behavior. Learn the assumptions before the patterns, so the patterns don’t become blind faith.
Relationships
- Mean Reversion and Momentum — principle ⑥ is the origin of trend-following; the momentum half of that dichotomy
- Quantitative Trading — the systematic descendant, where Dow’s rules become testable signals
- CAN SLIM — combines this kind of trend/volume reading with fundamental screens
- Thinking Fast and Slow — the phase model (frenzy, then panic) is System 1 behavior at market scale
- Trading and Finance
References
- 一文读懂”道氏理论”精髓 — 交易员说, 雪球 (2024-04-12)