Price structure

Trend: a definition you can check, not an opinion you hold

Most arguments about whether something is "in an uptrend" are arguments about nothing, because neither side has said what the word means. Define it in terms of swing points and the argument becomes arithmetic.

About 10 minutes to read. Free, no account.

The definition

An uptrend is a sequence of swing highs that each exceed the previous swing high, and swing lows that each exceed the previous swing low. Higher highs and higher lows. A downtrend is lower highs and lower lows. Anything else is neither, and "neither" is a perfectly good answer that most people refuse to give.

This is worth having because it is checkable. You do not need to feel anything about the chart. You list the swing points - using the fixed rule from the highs and lows lesson - and you compare numbers:

Six swing points, compared in order

low   1   100.0
high  1   112.0
low   2   104.0    higher than low 1    (104.0 > 100.0)   yes
high  2   120.0    higher than high 1   (120.0 > 112.0)   yes
low   3   110.0    higher than low 2    (110.0 > 104.0)   yes
high  3   118.0    higher than high 2   (118.0 > 120.0)   NO

every test passes until high 3. That is an uptrend, by definition,
up to and including low 3.

Nothing in that table is a judgement. Either 104.0 is larger than 100.0 or it is not. This is the single cheapest upgrade available to most people reading a chart: replace the word "trend" with a list of comparisons.

What breaks it, and what merely interrupts it

High 3 at 118.0 failed to exceed high 2 at 120.0. That is a failure to extend, and on its own it is not a broken trend - the sequence of lows is still intact, and plenty of advances pause without ending. Measure the failure and move on:

The size of the failure

high 2                 =  120.00
high 3                 =  118.00
short by               =  2.00
as a percentage        =  2.00 / 120.00  =  1.67%

The trend breaks when the sequence of lows breaks, because that is the half of the definition that says buyers kept stepping in higher. Say the next swing low prints at 106.0:

The break, and the thing people miss about it

low 3                  =  110.00
low 4                  =  106.00     lower than low 3  ->  broken
the break, as a %      =  4.00 / 110.00  =  3.64%

and yet, from the start of all this:
low 1                  =  100.00
low 4                  =  106.00
price is up            =  6.00 / 100.00  =  6.00%

Read that last line carefully, because it is the honest and slightly uncomfortable part. Price is 6% higher than where this started and the uptrend is over by definition. The definition is not telling you what price will do. It is telling you that one specific pattern in the swing points has stopped holding. Those are different claims, and conflating them is how people short things that keep rising.

The timeframe is not a detail, it is the answer

A swing point is defined in bars, and a bar is defined by the timeframe you selected. So the swing points change when you change timeframe, and therefore the trend does too. This is not a flaw to be resolved. There is no single true trend for a coin, and asking which timeframe is correct is like asking which magnification of a map is correct.

Put numbers on how much room the lower chart has to disagree:

One higher-timeframe pullback, seen from below

one 4-hour bar          =  240 minutes
in 15-minute bars       =  240 / 15      =  16 bars

a 3-bar pullback on the 4-hour chart
                        =  3 x 16        =  48 fifteen-minute bars

with a 3-bar swing rule, 48 bars can hold roughly
48 / 8                  =  6 pivots of their own

Six pivots is more than enough for a complete sequence of lower highs and lower lows. So the 15-minute chart can show a textbook downtrend for two days inside what the 4-hour chart records as one pullback bar in an uptrend. Both descriptions are accurate. They are descriptions of different things.

How to use two timeframes without lying to yourself

Looking at more than one timeframe is reasonable and extremely easy to abuse. The abuse has a shape: you have a position or an opinion, the timeframe you chose disagrees with it, and you go looking for one that agrees. That search always succeeds. On any liquid coin at any moment there exists a timeframe showing an uptrend and another showing a downtrend, because the definition is applied to different bars.

Name the timeframe you are trading on before you look, and write it down with the trade. That is the one whose swing points govern the decision.
A higher timeframe is for context only - it tells you whether you are trading with or against the larger sequence, which is useful to know and is not a trigger.
Two timeframes is a frame of reference. Five is a search for agreement, and you will find it.
If a decision flips depending on which chart you happen to be looking at, the decision was never resting on structure - the same test the RSI lesson applies to indicator periods.

What the definition cannot do

It cannot be known in the present. Every term in it is a swing point, and a swing point is confirmed late by construction. A trend is always declared a few bars after it is already true.
It says nothing about how far or how fast. An uptrend made of 1% steps and one made of 20% steps satisfy the definition identically. Size and speed are separate measurements.
It carries no probability. "This is an uptrend" is a statement about bars that have printed. It does not say the next swing high will be higher, and the whole of trading is in that gap.
It is sensitive to the swing rule. Widen the rule and small pullbacks stop counting as swing lows, so a trend that looked broken stays intact. Pick the rule first, or you are choosing the conclusion.

The terms, defined

Uptrend: A sequence in which each swing high exceeds the previous swing high and each swing low exceeds the previous swing low.
Downtrend: The mirror image: each swing high below the previous one and each swing low below the previous one.
Higher high / higher low: The two comparisons an uptrend is made of. Both are needed; one of them alone is not a trend.
Failure to extend: A swing high that does not exceed the previous swing high while the sequence of lows is still intact. A pause, not a break.
Structure break: The first swing low below the previous swing low in an uptrend, or high above high in a downtrend. It ends the definition, and nothing more.
No trend: The correct description when the comparisons do not all pass in one direction. The most commonly skipped reading on any chart.

What this means in practice

Replace the word "trend" with the list of comparisons. If you cannot write the list, you do not have a trend, you have a feeling.
Separate a failure to extend from a broken sequence of lows. They are different events with different meanings.
When a trend breaks, say "no trend" rather than "downtrend". One lower low is half of a downtrend, not all of it.
Pick the timeframe you are trading before you look, and let the higher one be context and nothing else.
Remember that a trend can break while price is still above where you first noticed it. The definition is about swing points, not about profit.
Keep the swing rule constant across coins and across timeframes, so that "trend" means the same thing everywhere you use the word.

Nothing in this lesson is financial advice or an instruction to trade. See the risk disclaimer. We never recommend a leverage setting - that decision is yours.