Trend bias
Which way the last four swing labels lean — a small, fast read that this product never allows to be the final word.
- Assumes:
- The basics of this family
- Markets:
- Crypto · Equities · Indices
- Timeframes:
- 1h · 4h · 1d
What it means
Trend bias is the direction implied by the most recent structure labels. It answers a narrow question: of the last few swings, how many pointed up and how many pointed down.
It is not the same as saying a market is trending. A market can lean while going nowhere, and the distance between those two statements is the reason this product computes both.
How we detect it
The engine takes the last four structure labels. Three or more upward labels — higher highs and higher lows — return up; three or more downward labels return down; anything else returns range. Fewer than two labels also returns range, because there is nothing to read.
That is the whole rule. It is deliberately crude, and its output feeds the regime classifier rather than the reader.
The word range here means only that the labels did not agree. It does NOT mean the engine has found a range, which is a separate and much stricter test.
When the read carries information
- The swings behind the labels are genuine turns on a timeframe that suits the question.
- The market has been doing one thing for long enough that four labels describe one episode rather than three.
- It is used as an input. Read alone it is the weakest statement in the taxonomy.
When it misleads
- A choppy market routinely produces three labels leaning one way while price has finished exactly where it started. This is the failure the regime classifier exists to correct.
- Four labels is a short memory. A trend that paused and resumed can read as range in the middle of itself.
- Because the labels ignore magnitude, four tiny swings outweigh one enormous move in the opposite direction.
How it reaches you
On our surfaces the trend bias is rarely quoted on its own. What a reader sees is the market state, which requires the price to have travelled as well as the labels to agree.
It is an observation about recent swings, never a forecast and never a suggestion.
A worked example
Constructed example — not market data
Six advances, each with a pullback that stops above the previous one. Run over this series the engine labels the swing sequence higher highs and higher lows, and reads the trend from those labels rather than from the direction of the last bar.
Worked example
A worked example for this concept has cleared review and is shown with it.