Indicator library · Definitions
Downtrend, Three Definitions in Use
The only page here about a word rather than a formula. Three definitions of a downtrend are in common use, they disagree on ordinary data, and almost every argument about whether one is under way is really an argument about which rule is being applied.
Three rules, one word
Ask three analysts whether an instrument is in a downtrend and they may all be right and disagree, because they are answering with different rules. Each is defensible; none is standard.
| Definition | The rule | Hidden parameter |
|---|---|---|
| Moving average | Price below an average that is itself falling. | The period, and every artefact the average carries. |
| Structural | Lower highs and lower lows in sequence. | What counts as a swing high or low. Two readers, two answers. |
| Retracement | A fixed percentage below the peak, 10 or 20 per cent. | Nothing about time or shape enters it at all. |
Why the disagreement is not academic
The interval in the middle of that figure is where money is made and lost. It is precisely the kind of rally that ends a mechanical downtrend and does not end the decline, and someone who entered on one rule and exited on another has no way of knowing whether their result came from the market or from the switch.
Each rule also fails in its own characteristic way, which is worth knowing before adopting one. The average definition changes state too readily: an instrument can flip in and out of a downtrend several times inside one decline, and each flip is a signal to a system built on it. The structural definition is the closest to how a decline looks and the hardest to automate. Its swing window is a judgement, so two people coding the same rule get different results. The retracement definition is unambiguous and blind to shape: a 20 per cent fall over a week and over two years satisfy it identically.
Timeframe is a fourth, silent parameter
Every rule on this page is applied to a series of bars, and nothing in any of them says which bars. The same instrument can be in a downtrend on a daily chart and an uptrend on a weekly one, with no contradiction whatsoever; the two statements are about different series.
This is where most cross-purposes conversations about a market come from, and it is settled by the same discipline as the rest: say the timeframe alongside the rule. "Below a falling 50-day average on daily bars" is a claim someone else can check. "In a downtrend" is not, and the person disagreeing with you may be looking at a chart that agrees with you completely.
What none of them measures
All three are statements about price alone, and two things they leave out change how a decline resolves often enough to be worth naming.
The first is participation. A decline in which most of the list is falling is a different market from an index dragged down by a handful of heavy constituents while breadth holds, and no definition on this page can tell them apart. That is what the breadth measures are for.
The second is volume. A decline on expanding volume is meeting supply; the same decline on progressively thinner volume is running out of sellers. Both are downtrends under every rule here, and the second is the condition that most often precedes a turn. Neither is part of the definition, and both belong in the description.
Where it misleads
| Situation | What goes wrong |
|---|---|
| Definitions switched mid-argument | Entering on one rule and holding on another means the result is a property of the switch, not of the market. |
| A published claim without its definition | "Downtrends persist" is untestable until the rule is stated. Most such claims never state one. |
| Structural rule automated | The swing window is a judgement. Two implementations of the same rule identify different structures. |
| Average rule in a choppy market | The state flips repeatedly inside one decline, and every flip is a signal to whatever is built on it. |
| Retracement rule across timeframes | Blind to time and shape: a week and two years satisfy it identically. |
| Treated as a forecast | Every definition here describes what has already happened. None contains a claim about the next bar. |
The habit this page is really about
Almost every page in this library ends by saying that a reading is a description rather than a signal. This one starts a step earlier: before a description can be right or wrong, the words in it have to mean something fixed.
The practical version costs nothing. Decide which definition you are using, write it down beside the position, and use the same one to decide when the condition has ended. Most disagreements about whether a market is in a downtrend dissolve the moment both parties state their rule, and the ones that survive are then about the market, which is the only kind worth having.
Frequently asked questions
Is there a standard definition of a downtrend?
No, and that is the practical point of this page. Three definitions are in common use — price below a falling average, a sequence of lower highs and lower lows, and a fixed percentage decline from a peak — and they disagree with each other on ordinary data. Any statement about what downtrends do is a statement about whichever definition was used, and a claim quoted without one cannot be checked.
What is the moving-average definition?
Price below a moving average that is itself falling. It is the easiest to compute and the easiest to automate, it is unambiguous once the period is fixed, and it inherits every property of the average underneath, including the lag, and the window-exit artefact that can turn the average on a quiet session. Because it is a rule about today rather than about structure, it flips in and out of a downtrend more often than the alternatives.
What is the structural definition?
A sequence of lower highs and lower lows, the classical description, and the one most chart readers mean by the word. It is closer to how a decline actually looks and it contains a hidden parameter: what counts as a high or a low. Two analysts using different swing windows identify different structures on the same chart, which is why the definition resists mechanical testing.
What is the retracement definition?
A fixed percentage decline from a peak, 10 per cent for a correction, 20 for a bear market, in the conventional usage. Its virtue is that it is completely unambiguous and requires no parameters beyond the threshold. Its weakness is that it is a statement about magnitude with no reference to time or shape: an instrument that fell 20 per cent in a week and one that drifted down 20 per cent over two years are the same thing under this rule.
Which definition should I use?
Whichever matches the decision, stated in advance. A trend-following system needs a mechanical rule, which points at the average. A discretionary reader describing market structure is using the structural definition whether they say so or not. A risk report quoting drawdowns is using the percentage rule. The error is not choosing wrongly, it is switching between them mid-argument, which is easy to do because all three share one English word.
Why do the definitions disagree most in the middle of a decline?
Because a substantial rally inside a longer decline satisfies some rules and not others. Price can climb back above a falling average, ending the downtrend by the first definition, while still making lower highs than the previous swing and remaining well below the peak. That is exactly the interval where someone using one definition sees a trend change and someone using another sees a pullback, and both are correct on their own terms.
Does a downtrend have to be confirmed by volume?
No definition in common use requires it, and the question is worth asking anyway. A decline on expanding volume and one on progressively thinner volume are both downtrends by every rule here, and they have historically resolved differently often enough to be worth distinguishing. Volume is not part of the definition; it is part of the description, and the page on volume spread analysis is where that reading lives.
When does a downtrend end?
Under whichever definition began it, and this is where the switching does most damage. A trader who entered on the average definition and then argues the downtrend continues because the structure still shows lower highs has changed the rules mid-position. Deciding the exit condition at the same moment as the entry condition, and writing both down, is the whole of the discipline.