Dow Theory for CMT Candidates: Six Principles, One Coherent Trend Test

Where Dow Theory fits

The current 2026 CMT Program Guide lists Dow Theory under Level I history and theory. Its public objectives refer to the two averages, basic tenets, closing price, confirmation, and price movements. This article explains those ideas as an interpretive framework. It does not imply that every historical rule has a precise universal numerical threshold or that future exam items will use this example. For a hands-on observation worksheet, see Dow Theory Fundamentals.

Charles Dow's market writing was later organized by successors. A common six-principle summary is useful, but it is a later synthesis. Read it as a set of hypotheses to evaluate, not six independent indicators that can each issue a trade.

1. Averages absorb available information

The phrase “the averages discount everything” means that transactions aggregate participants' responses to known information. It does not mean prices are always correct, that private information is reflected instantly, or that technical analysis can ignore data quality. New information, changing expectations, and liquidity alter prices. The analyst studies the recorded outcome, then asks whether a proposed trend interpretation survives new data.

This principle is about the market average, not the certainty of an individual stock call. It also differs from claiming the strongest form of the Efficient Markets Hypothesis. An analyst can use market prices as primary evidence while remaining skeptical of predictability.

2. Markets move on several horizons

A primary trend is the broad movement of interest; secondary movements correct it; minor fluctuations occur within both. In a stylized uptrend, a rise from 1,000 to 1,200 followed by a decline to 1,140 is still 140 points above the initial 1,000, though the decline is meaningful to a shorter-horizon trader. The correction is 60/200 = 30% of the previous advance. This percentage describes the pullback and does not, by itself, prove that the primary trend continues.

The time horizon must be stated. A daily chart may show a minor downtrend while a weekly chart still shows higher primary lows. A risk exit may trigger during a correction even when the broad classification remains up. Trend naming and trade management should be written separately.

3. Primary trends may pass through phases

An uptrend is often described as accumulation, wider public participation, then distribution; a downtrend has analogous development. The value of phase language is its hypothesis about participation and expectations. Its weakness is that it is easiest to see in hindsight. “Institutions are accumulating” cannot be read directly from a green candle. Support a phase label with observable price, volume, leadership, and breadth evidence, then name what would refute it.

Imagine price makes new highs while fewer components participate and volume contracts. That can motivate a distribution hypothesis, but it is not proof. A fresh broad advance on stronger participation would weaken that hypothesis. See Market Breadth Indicators for methods to count participation.

4. Related averages should confirm

Dow's original reasoning used industrial and rail averages to test whether production and transportation moved together; modern summaries often refer to industrial and transportation averages. The central idea is that a broad economic trend has more credibility when related groups make corresponding significant highs or lows. Their compositions have changed, so confirmation is an analytical test rather than a timeless mechanical verdict.

Consider two hypothetical averages. Industry previously closed at a significant high of 800 and now closes at 840. Transport previously closed at 500 and now closes at 490. Industry is 5% above its old high: (840/800 − 1) × 100 = 5%. Transport is 2% below: (490/500 − 1) × 100 = −2%. The highs are unconfirmed. If Transport later closes at 510, both averages have exceeded their reference highs. The analyst should document the reference points and avoid moving them after seeing the outcome.

A nonconfirmation is a warning, not a complete reversal. It says the evidence is weaker than a joint breakout. It does not specify the direction or date of the next move.

5. Volume offers supporting evidence

In the classic interpretation, volume tends to expand with movements in the primary trend and contract on countertrend moves. Suppose an advance of five sessions averages 2.0 million shares per day while a pullback averages 1.2 million. The ratio is 2.0/1.2 ≈ 1.67. That observation is consistent with greater participation during the advance. It cannot tell who bought, whether the move was informed, or whether price will rise again. Compare equivalent periods and comparable trading venues.

Volume can conflict with the two-average signal. If averages confirm but volume fades, record both facts instead of forcing consensus. A method that can acknowledge conflicting evidence is more useful than one that labels every new high “strong.”

6. Keep the trend until contrary evidence is sufficient

This principle protects against changing a primary trend designation after every correction. It also demands a definition of “sufficient.” One illustrative rule is: require both averages to close below significant secondary lows after failing to establish confirmed new highs. Another may use swing structure plus volume. No threshold in this example is presented as an official rule. Predeclare the evidence that would move your assessment from uptrend to warning to reversal.

Suppose Industry's significant secondary low is 780 and Transport's is 470. A close of 775 and 465 respectively would break both reference lows. This is materially different from Transport merely failing to confirm a high. Even then, a technician should examine whether data adjustments, gaps, or unusual index events distorted the readings.

A decision sequence

First, specify interval and closing-price series. Second, mark significant primary and secondary highs and lows before viewing later bars. Third, classify each average's trend. Fourth, record confirmation or nonconfirmation. Fifth, review volume and breadth as supporting or opposing evidence. Finally, write an invalidation level and a position-risk plan. The distinction between the last two matters: an investor may reduce exposure to protect capital before a primary Dow Theory reversal is established.

Common mistakes include using intraday highs against a closing-price criterion, declaring a reversal from a lone nonconfirmation, treating phases as directly observable, and switching between a broad index and one stock mid-analysis. The theory structures evidence; it does not eliminate uncertainty.

Self-check and next action

Industry has an old high of 1,250 and now closes at 1,275. Transport has an old high of 900 and now closes at 918. Do they confirm, and by how much did each exceed its reference? Answer: Yes, under a closing-high rule. Industry is 25/1,250 = 2% above; Transport is 18/900 = 2% above. This says nothing about how far the trend will continue. Mark each average's most recent secondary low to state what later evidence would challenge the conclusion.

Next, make a two-column chart note for the actual averages you study: one column for observations, one for conditional interpretations. Compare the wording with the official Level I objectives.