Measuring Similarity Between Candlestick Price Series

Summary

The article presents a quantitative way to compare two candlestick price series, motivated by the technical-analysis idea that historical price patterns may recur. It defines each candle by its open, high, low, and close, and notes that comparisons depend on both the lookback length and the candle interval.

Its similarity score is the average of four correlations: one each for the open, close, low, and high series. A score of 1 represents matching movement, -1 represents opposite movement, and values near zero indicate little relationship. The example compares two Chinese stocks over 60 daily candles and reports a similarity score of 0.57, described as not especially high. The article says this approach can support price forecasting, market analysis, and pattern-based stock selection, but the excerpt offers no predictive validation or evidence that similarity produces profitable signals.

Key ideas

• The proposed score averages correlations for the open, close, low, and high price series.
• A score ranges from -1 for opposite movement to 1 for identical movement, with values near zero indicating weak relation.
• The comparison depends on the selected lookback length and candle interval.
• An example uses 60 daily candles for two stocks and reports a similarity score of 0.57.
• The excerpt does not validate forecasting accuracy or trading profitability.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.