Multi-input Improvement of One-day Performance Indicator

Stock market forecast can be built using different technical indicators. The relative change between the closing prices of two consequent days can be called one-day performance indicator (ODP indicator). The interpretation of ODP indicator chart is very simple - an uptrend starts after a big positive value of ODP indicator (the chart has been presented in the recent post). It works by a simple scheme: input₁ → output.

However, a closer look at ODP indicator chart reveals also a typical pattern before starting uptrend - it fluctuates while moving down and then it has a big positive value. So that the more informative scheme would be in case of using n days of ODP indicator signals: (input₁, input₂, . . input_n) → output. It reminds candlesticks pattern chart. A candlestick figure consists of Real Body and Upper and Lower Shadows. The Real Body size is proportional to the difference between opening and closing prices of one day. Since it is no a big difference between closing prices of previous day and opening price of the next day, there is a similarity between Real Body and ODP indicator.



Considering this uptrend case, it is important to notice that there could be many other ODP indicator patterns. Investors could analyze these multi-day patterns on charts (in the same way as candlesticks pattern is used as a tool to predict future prices) or by using an automatic statistical method to map the correlation (input₁, input₂, . . input_n) → output. Evidently, one of the simple and powerful statistical methods that could be used for this purpose is Neural Network.

© Alex Shmatov. Published with permission of the copyright owner. Further reproduction strictly prohibited without permission.