How Deja-view computes chart similarity
Deja-view compares normalized adjusted-close shapes. The output is a retrospective similarity table and outcome distribution, not a prediction model.
Method
- Use adjusted close to reduce split and dividend artifacts.
- Normalize each window so its first point starts at 100.
- Resample windows to the configured canonical length.
- Rank historical windows by shape distance.
- Summarize what happened after 1, 7, and 30 trading days.
Limitations
Similar shape does not imply causality. The corpus can still contain survivorship bias, source-data delays, and missing historical cases. See sources.
Page Information
Frequently Asked Questions
Is deja-view investment advice?
No. Deja-view is a retrospective comparison tool for historical price patterns. It does not provide buy/sell signals, return guarantees, or investment advice.
How often are pages updated?
Static explanation pages are updated when their content changes. Data-backed stock pages and hubs show the generation timestamp from the data artifact.
Who writes and maintains this site?
Deja-view is created and maintained by Baek Kyoungjung (BK). The About page links to public Medium and LinkedIn profiles for author verification.