Definition and purpose
A worked example of trend identification is a fully spelled-out scenario where you apply a consistent set of rules to historical price observations, label the direction (trend vs range), and show the intermediate steps. The goal is not to predict the future; it is to help a reader understand the mechanics of deciding what the current behavior most closely resembles, and to be able to repeat the decision process.
Trend identification typically relies on comparing “higher highs” and “higher lows” for an uptrend, or “lower highs” and “lower lows” for a downtrend, with a third possibility: a range, where those relationships do not hold over the chosen lookback window. What you call a “high” or “low,” and how large a move must be to count, are rule choices and therefore assumptions.
How the example works (mechanism)
Below is one simple, rule-based approach that you can verify by hand. It separates stable mechanics (the logic) from variable conditions (the data window and thresholds).
Working rules (assumptions stated):
- You observe prices for 10 time points labeled T1–T10.
- You choose a “pivot” rule: a pivot high is a local maximum among the last 3 points; a pivot low is a local minimum among the last 3 points.
- You label each pivot and then apply trend logic:
- Uptrend: at least two successive pivot lows are higher than the prior pivot lows, and pivot highs are also higher.
- Downtrend: at least two successive pivot highs are lower than prior pivot highs, and pivot lows are also lower.
- Otherwise: classify as range/unclear.
- You use closing prices as the only input.
Input data (assumption stated): Assume the following closing prices:
- T1: 100, T2: 103, T3: 101, T4: 106, T5: 104, T6: 110, T7: 107, T8: 112, T9: 109, T10: 111.
Step-by-step pivot identification:
- Pivot highs (local maxima): T2 (103) vs T1/T3, T4 (106) vs T3/T5, T6 (110) vs T5/T7, T8 (112) vs T7/T9.
- Pivot lows (local minima): T3 (101) vs T2/T4, T5 (104) vs T4/T6, T7 (107) vs T6/T8, T9 (109) vs T8/T10.
Apply trend logic:
- Pivot highs: 103 → 106 → 110 → 112 (each is higher than the previous).
- Pivot lows: 101 → 104 → 107 → 109 (each is higher than the previous).
- Result under the stated rules: uptrend.
Notice the mechanics are deterministic given the rules and the input series. If you change the pivot definition (for example, require a higher/low to exceed a minimum distance), the label can change even with the same underlying prices.
Evidence or example comparison (two outcomes from the same idea)
A worked example should also show how alternative assumptions change the output.
Alternative pivot-low threshold (assumption change): Suppose you add rule (new assumption): a pivot low must be at least 3 units higher than the previous pivot low to “count” as a higher low; otherwise, you treat it as a minor fluctuation.
Using the same data, pivot lows are 101, 104, 107, 109. Under the new rule:
- 104 vs 101 is +3 → counts.
- 107 vs 104 is +3 → counts.
- 109 vs 107 is +2 → does not count.
With only two counted successive higher lows (101→104→107) and depending on whether you also apply a similar threshold to pivot highs, your classification could remain “uptrend” or become “unclear” if your exact decision requires multiple counted pivots. This illustrates why independent verification must include the exact rules.
Limitations and failure modes
Several limitations are common and are largely independent of any specific charting tool.
- Rule sensitivity (assumptions change the label): The window size for “local” maxima/minima and thresholds for what counts as a meaningful move can flip the outcome. 2. Whipsaw during ranges: In sideways or choppy conditions, local maxima/minima occur frequently. You may observe alternating higher lows and higher highs that look like a trend, but the sequence may not persist. 3. Inconsistent classification across people: Two analysts using different pivot rules can produce different trend labels even when both are “reasonable. ”
- No forward guarantee: Historical relationships do not establish future results.