Trend & Pattern Analysis
Identifying and interpreting regularities, cycles and breaks in change over time.
At a glance
- Produces
- Descriptions of direction, rate, volatility and structural break in observed change.
- Depends on
- Series long enough to separate a change of level from a change of character.
- Fails when
- A pattern is reported without any candidate mechanism that would produce it.
Trend and pattern analysis examines how phenomena change over time and where regularities, cycles, structural breaks or recurring configurations appear. It provides much of the empirical grounding for foresight, and it is where quantitative evidence and interpretation meet.
How we approach it
A trend is a description, not an explanation. We treat the search for pattern as a step towards asking what would produce that pattern.
- We distinguish level, direction, volatility and structure. A series can be rising and still have changed character, and that distinction is usually the interesting one.
- Series are examined for breaks and reversals, not only for slope. Change points are treated as events requiring explanation.
- Trends are analysed in relation to one another, since the interaction between two trends is rarely the sum of their separate trajectories.
- Every quantitative pattern is paired with a candidate mechanism. Patterns without mechanisms are reported as descriptions and labelled as such.