Trend and Market Structure — Reading a Trend Objectively

Where this starts: Dow, and what is still true

The strategy is built on Dow theory, updated with a different way of marking the chart and with volume added to it. Charles Dow never published a trading method — the theory was assembled afterwards from his editorials — and it does not predict anything. What it gives you is a set of observations about how markets behave, and four of them still carry the whole of what follows:

The market has three movements — a primary trend, secondary reactions against it, and minor noise.
A primary trend has three phases: accumulation, public participation, distribution.
Price already contains the news. Whatever you just read, the market read it first.
Averages must confirm each other. One index making a high alone is not a market making a high.

A rising market with the primary trend drawn through it, two secondary reactions against that trend marked in red, and minor noise running through the whole path — All three at once. The primary trend is the one worth trading; the other two are what makes sitting in it difficult. Click to enlarge

Three daily charts side by side labelled down trend, up trend and sideways — The three shapes a market can be in, as he labels them: bearish, bullish, and range-bound.

Why build on trends at all

Before the mechanics, the case for the style — because it explains which of the rules later on are load-bearing and which are convenience.

The first reason is arithmetic. Most of what a market does in a year happens in a minority of its sessions, and a method that keeps entering and leaving misses those by construction. Following a movement instead of trading round it is how you stay in for the part that pays.

The second is that it reduces the number of judgement calls. Once the direction is settled, the remaining decisions are rule-shaped: where the stop goes, what counts as the trend still being intact, when it is finished. That is also what makes risk manageable — in an uptrend there is an obvious place for a stop, and it is not a matter of opinion where it is.

The third is that it travels. The same reading works on minutes and on months, and on shares, currencies, commodities and indices, because it is built on structure rather than on any property of a particular market. That is worth knowing now, because everything in this lesson is demonstrated on currencies and none of it is about currencies.

What this course adds to that is volume — and adding volume is what forces the awkward detour in the next section.

One advance where most of the gain comes from three legs, shown once as a position held throughout and once as three short spells in the market — Most of what an advance pays comes from a few legs. Trading round them is how you end up in the market for the rest of it instead. Click to enlarge

The S&P 500 across several years with the bull and bear boundary recomputed at three separate peaks — Years of one index on a single screen. The moves worth having are long, which is the whole argument for following one.

The three phases, and why they do not feel the same

Accumulation. Informed money is building a position while the story is still bad. Price goes almost nowhere, the news gives you no reason to be there, and it is the least comfortable place to be right. Dow's own description is that this phase passes unnoticed by the public, and that is not incidental — it is unnoticed because nothing on the price chart is worth noticing yet.

Public participation. The trend becomes obvious, the news catches up, and volume arrives. This is the longest of the three and the one most trend following is actually done in. It is also the easiest to be in, which is why most methods look good when tested on it and stop looking good either side.

Distribution. The same informed money sells into the enthusiasm. This is the phase that costs people money, because price is still making highs while the buying that produced them is being sold into. Nothing about the price chart says so.

Naming the phase matters because the same setup has a different expectancy in each one. A pullback bought in accumulation and the identical pullback bought in distribution are not the same trade, and the candles do not distinguish them. Most of the tooling later in this course exists to tell the first from the third — which look alike on a price chart and do not look alike on a volume one.

One market cycle divided into accumulation, public participation and distribution, with a volume histogram beneath showing low volume in the first phase and high volume in the last — The distinction the rest of the course is built to make. On price alone the first and third phases are the same picture; underneath them they are not. Click to enlarge

A daily chart with movements and corrections shaded in green and red and a volume histogram beneath, annotated with how volume behaves in each — The same move with volume beneath it and his notes on what volume is doing at each stage of it.

Why volume analysis sends you to the futures market

This is the part most volume material skips, and skipping it makes the rest meaningless.

Start with the problem. Forex has no volume figure — not a hidden one, not a hard-to-get one. It is an over-the-counter market with no central exchange, so no record of total trading exists to be published. Six trillion dollars a day changes hands and nobody counts it, because there is no single place where the counting could happen.

What you trade on a retail platform is usually a CFD — a contract for difference. It is a derivative: you never own the currency, and the volume your platform shows is your broker's own flow. Two brokers will report two different volumes for the same hour, and both are honest, because each is counting only itself. Neither is the market.

The alternative would be direct market access, trading straight into the interbank market rather than through a broker's book. In practice that is closed to retail: it needs the capital and the infrastructure to deal with major banks and liquidity providers directly, which is why the CFD exists in the first place.

The futures market solves the counting problem by being centralised. Buyers and sellers from everywhere meet in one venue and prices come from the balance of competing orders there, so the volume is total, standardised, and the same number for everyone looking at it. Currencies, indices, metals, energy and agricultural products all trade this way.

Currency futures track spot closely — forex settles instantly and the front-month contract follows it — so the two are correlated enough to read one and trade the other. That gives the working method: analyse the future, place the trade on the CFD. The chart you make decisions from and the instrument you execute on do not have to be the same thing, and for this style they usually are not.

Seeing that volume needs a platform that reads exchange data — Level 2 rather than the Level 1 price feed a standard retail platform gives you. Level 1 is bid, ask, last and a volume number; Level 2 carries depth, per-trade size, and buying separated from selling. Every technique in the later lessons is built on that separation.

The pictures below include his own pairing table: which futures contract to read for each instrument you actually trade.

Three brokers reporting three different volume figures for the same instrument in the same hour, beside a single centralised futures venue reporting one figure — Same instrument, same hour, three answers — and none of them is the market. The exchange has one number because every buyer and seller meets there. Click to enlarge

One minute of futures beside the same minute on three different CFD brokers — The same minute on four feeds. The prices disagree; the shape of the candles does not — which is what makes reading one and trading the other possible.

The same chart minutes later with volume added, and a cluster view showing per-price volume — The same chart with volume attached. He marks the high-volume points as where the reversal came from.

A table pairing futures symbols with their CFD equivalents, beside explanatory text — His pairing table: which futures contract to read for each instrument you actually trade.

Marking the structure: P1, P2 and P3

The important area

Here is where this departs from the usual description. Rather than "higher highs and higher lows", every turn gets a name, and the names have rules.

The reason for bothering is that "higher highs and higher lows" is a description, not a test. It tells you what an uptrend looks like once you already agree it is one. It does not tell you whether the thing in front of you right now qualifies, and it gives two people looking at the same chart no way to settle a disagreement.

In an uptrend:

P1 — where the structure starts. The low the first movement leaves behind.
P2 — a movement point. The high the market reaches before it pauses.
P3 — the correction point. The higher low the pullback ends at.

An uptrend labelled P1, P2 and P3 with the important area shaded at each movement point, and the candle that closes above it circled — Three legs of an uptrend. The shaded band at each movement point is the important area; the circled candle is the close that makes the next point definitive. Click to enlarge

A daily uptrend fully labelled with P(1), P(2) and P(3), the important area at each point, and the caption definitive uptrend structure — A whole uptrend named point by point, with the important area marked at each one. This is the picture the section describes.

Certain and uncertain — the rule that removes the argument

A point is not a point because it looks like one. It becomes definitive when a candle closes beyond the important area of the previous movement — above it in an uptrend, below it in a downtrend.

Until that close happens the point is uncertain, and an uncertain point is not something to trade from. A wick through the area is not a close beyond it, and this is the single most common way a structure is misread.

The value of the rule is not that it is clever. It is that two people looking at the same chart get the same answer, and that you can be wrong in a way you can check afterwards. A method you cannot be objectively wrong about is a method you cannot improve.

It also decides what you are allowed to do next. Every entry technique in lesson two is written against points that have become definitive, and several of them are separated from each other by nothing more than whether they wait for that close. The difference between the safest setup in the next lesson and one he labels high risk is exactly this rule being applied or skipped.

Practically: mark the area, then wait for the candle to finish. A price that trades through the level intrabar and comes back has told you something about pressure, but it has not made the point definitive, and the structure is unchanged.

The same important area twice: on the left a candle whose wick pierces it but closes back below, on the right a candle that closes above it — A wick through the area is not a close beyond it. Left proves nothing; right makes the point definitive. Click to enlarge

A daily downtrend labelled with P(1), P(2) and P(3) and an annotation reading closing below the important area of the previous movement — The same structure downward, with the annotation that defines the rule: closing beyond the important area of the previous movement.

Movement and correction

A movement is a directional leg. A correction is the temporary move against it.

Corrections usually retrace somewhere between 30% and 80% of the movement before them. A correction can go as far as 100% and the structure still holds — provided it does not close below the last low of the previous movement. That close is the difference between a deep correction and a broken trend.

Shallow corrections under 20% are not free entries. They tend to mean the market has not finished, and entering into one carries more risk than the neat retracement it resembles.

The purpose of a correction, in his framing, is that price has moved too far too fast and is being brought back to somewhere defensible. That is why depth on its own is not the signal. A shallow correction and a deep one can both be healthy, and the thing that separates them is not the percentage — it is what volume did while it happened, which is the next section.

One line from the end of the section is worth keeping: the potential of the next movement is related to how deep the correction went and how far volume fell during it. A deep correction on collapsing volume is a different proposition from a shallow one on rising volume, and the pictures below are the two cases side by side on real charts.

A movement with its thirty to eighty per cent retracement band shaded, and below it the last low of the previous movement drawn in red as the level that ends the structure — The trend survives a correction of any depth — until a candle closes below that red line. That close is the difference between a deep pullback and a broken trend. Click to enlarge

Two charts labelled downtrend uncertain and uptrend uncertain with movement and correction marked on each, and a note that these are not yet trends — Movement and correction marked on both directions — with his own warning underneath that these are not yet trends.

What volume should be doing

The same price, two different meanings

Structure says where the market is. Volume says whether anyone is behind it.

The healthy pattern is simple: volume rises through a movement and falls through a correction. People are participating in the direction of the trend and losing interest when it pauses.

There are four combinations, and each one says something different:

Movement on rising volume — the move is being paid for. Many participants are involved and the momentum is confirmed.
Movement on falling volume — the move continues while fewer people support it. Not a reversal signal, but the trend is losing the thing that was driving it.
Correction on falling volume — healthy. The pullback is a lack of interest rather than a change of mind, and this is the normal case inside a working trend.
Correction on rising volume — the warning. Especially if the volume in the correction exceeds what the movement before it managed, which says the pullback is attracting more business than the trend.

Two identical price sequences, one with volume rising into the movement and falling in the correction, the other with volume doing the opposite — Identical candles. On the left volume confirms the trend; on the right it contradicts it. Price alone cannot tell you which one you are looking at. Click to enlarge

A movement and correction with a shaded retracement band and a volume histogram, beside the note that potential relates to correction depth and volume reduction — Both halves of the judgement in one picture: how deep the correction went, and how far volume fell while it did.

When a trend breaks

A break is only a break at a definitive point. A candle closing through an uncertain one has broken nothing.

The weaker case is a close inside the important area. It counts, but it is the version that gets retested and often fails.

The clean case is a close below the higher low in an uptrend, or above the lower high in a downtrend. That is a definitive break, and it is the one worth acting on.

On the left a candle closing inside the important area, on the right a candle closing below the higher low with the close circled — Both are breaks. The left one counts and is the version that gets retested; the right one is the one worth acting on. Click to enlarge

A chart headed Broken Trend with an uncertain P(2), a certain P(3), the important area shaded and the breaking candle circled — His broken-trend slide, with the candle that did it circled and the note that a break only counts at a definitive point.

When a trend reverses

A break is not a reversal. A reversal has three stages, in order:

A definitive break at the last P3.
A definitive P2 and P3 form in the new direction after that break.
The last uncertain P2 is renamed P1 — the old structure is finished and a new one has started.

An uptrend breaking below its last P3, then forming a new movement point and correction point downward, with the old uncertain P2 relabelled as P1 — All three stages, in order. Stop after the first and you have a break, not a reversal — which is a different trade. Click to enlarge

A chart headed Reversed trend listing the three stages of a reversal, with the broken trend marked and the old P(2) relabelled P(1) — The three stages listed beside the chart they happen on, ending with the old uncertain P(2) renamed P(1).

Bull market, bear market, and the trend in front of you

These are two separate questions and confusing them causes real damage.

The deck uses a simple line for the first: take the instrument's highest recorded price and multiply it by 0.8. Above that level is a bull market, below it a bear market.

For the S&P 500 with a record of 4820.3, that line sits at 3856.2.

The point of the calculation is what it allows you to say: the market can be in a bull market while the daily trend is down. Both statements are true at once, they answer different questions, and a position taken on one while thinking about the other is where a lot of avoidable losses come from.

A market trading above eighty per cent of its record high, so inside the bull zone, while the most recent leg of the price line points down — Both true at the same time: above the line it is a bull market, and the trend in front of you is down. They answer different questions. Click to enlarge

The formula for the bull and bear boundary with the highest price multiplied by nought point eight, and the two outcomes colour coded — The calculation itself, with the worked example: 4820.3 times 0.8 gives 3856.2 on the S&P 500.

The S&P 500 daily chart with the boundary drawn across it and a note reading the market is bullish but trend is down — The note he left on the chart. Above the line it is a bull market; the trend in front of you is still down.

What to take from this one

Three things, in order of how much they matter:

A point is definitive only on a close beyond the important area. Everything else here rests on that.
A correction can retrace fully and the trend survives — until it closes below the last low.
Volume rising into a movement and falling into a correction is the trend agreeing with itself. The opposite is a warning, not a signal.

The three rules stacked as three panels: a candle closing beyond the important area, a correction reaching the last low, and volume rising into a movement — The three in the order they matter. Everything else in this lesson rests on the first one. Click to enlarge

Next: where to actually get in

Lesson two takes the structure from this page and turns it into entries — the pullback setup, the outside setup, and where the stop belongs so the risk is fixed before the trade is.

Back to the twelve lessons