Correlations

Español·Deutsch·Русский·Italiano·Français·Türkçe·العربية

Historical analogue search compares a selected price segment with similar past segments. Shape is compared first; subsequent changes are studied separately. Resemblance alone does not determine continuation.

Sample

SampleSimilar shape [100, 96, 98, 94, 102]94102Time →
  • Similar shape

Historical analogue

Historical analogueOne shape at different scales [200, 192, 196, 188, 204]188204Time →
  • One shape at different scales
Teaching diagrams with hypothetical numbers. They explain concepts, not the next price move.

What changes the sample?

Sample length sets scale; normalization determines the role of absolute level and amplitude. Similarity measure and exclusion of overlapping episodes affect membership. Choose these conditions before viewing outcomes to limit outcome-driven tuning.

The section name does not mean every search method is Pearson correlation. Correlation and time-series distance measure different properties; see DTW and Pearson.

Reading results

A positive-outcome fraction describes the selected history, not a ready-made future growth probability. Magnitude, dispersion, horizon and dependence matter. Predictive value needs independent evaluation and a predefined baseline.

Begin with historical analogues, sample length and reading outcomes.

Different outcomes

  • A
  • B
Different outcomesContinuation paths represent hypothetical completed past cases.. Similarity measures shape, not the future. [[100, 103, 99, 106], [100, 98, 103, 94]]94106Time →
  • Continuation paths represent hypothetical completed past cases.
  • Similarity measures shape, not the future.
Teaching diagrams with hypothetical numbers. They explain concepts, not the next price move.

Testing conclusions

Continue with look-ahead bias, hypothesis testing, expectancy, costs, confidence intervals and survivorship bias. The limits of candles and chart scales clarify measurement. These are educational tools, not trade selection advice.