Metric Transformations

Metric transformations apply a mathematical transformation to the metric to gain further insight into it's statistical properties.

Display Metric Pct Rank

Display a third pane with a **n-period** percentage rank of the metric. The default rank shown is 100 days and the pane current value will be the last metric value ranked from 1 (lowest value) to 100 (highest) in the past 100 periods.

To test the percentage rank of the metric, change the **Backtest Settings** from **Metric** to **Percent Rank Metric**.

Display Metric Rate of Change

Display a third pane with a **n-period** Rate of Change of the metric. The default Rate of Change shown is 100 periods and the pane current value will be the last metric value **subtracted** from the metric value 100 periods ago. When looking at volatile metrics it could be more meaningful to look at smaller period Rate of Changes (than 100) such as a 2 period or 5 period Rate of Change.

To test the Rate of Change of the metric, change the **Backtest Settings** from **Metric** to **ROC Metric**.

Display Metric Realized Volatility

Display a third pane with a **n-period** Realized Volatility of the metric. The default Realized Volatility is 100 periods. Realized Volatility of the metric is calculated as the n-period Standard Deviation of outright metric values.

To test the Realized Volatility of the metric change the **Backtest Settings** from **Metric** to **Realized Volatility Metric**.

Display Metric Z-Score

Display a third pane with a **n-period** Z-Score of the metric. The default Z-Score shown is 100 rolling periods and the Z-Score is calculated as the ((current metric value - the mean of the metric values over the rolling n-periods) / the standard deviation of the metric values over the rolling n-periods).

To test the Z-Score of the metric change the **Backtest Settings** from **Metric** to **Z-Score Metric**.

Last modified 6mo ago