Regression Analysis is the way that we compare two securities (usually an equity and a benchmark) and describe the relationship between them. Portfolio Managers will often use the individual components of regression analysis to assist in their portfolio selections. The regression chart plots changes in price / value for a dependent code (y axis) and independent code (x axis). By default weekly values are used.
The initial step is to examine the two securities and measure the weekly returns over the past year (52 observations).
For each security, the weekly return is measured and that value is plotted on a XY scatter plot.
Optuma then uses a Least Squares algorithm to determine the line that would best fit all the observations on the chart.
Note: Regression Analysis on securities is typically performed on the percentage changes in their respective prices. This is calculated using natural logarithms in line with standard financial practices. Our tools do have the option of using price–instead of percentage change–for analysis of datasets. This is so the functions can be used on derived datasets which have already been detrended.
From the regression chart, we are able to determine the individual components. These are also available as individual indicators in Optuma:
- Alpha - the measure of relative performance.
- Beta - the relative volatility.
- Correlation - how closely the two securities move together.
- R Squared (r²) - how significant is the relationship.
You can view a 30 minute video on Regression Charts by clicking the image below:
Actions & Properties
Clone Chart: Makes a carbon copy of the source chart alongside the original.
Print Chart: Creates a PDF file of the chart which can then be sent to a printer.
Restore Default Settings: Click this action if you have adjusted the default settings of the chart, and wish to return to the standard properties originally installed with Optuma.
Save Settings as Default: If you have adjusted any of the chart's properties (such as the Colour Scheme or bar thickness, for example) you can save the adjustments as your new default setting. Each time you open a new chart it will display using the new settings.
Send To: Allows you to send the chart to an image file, or if you have the Publishing module to MS Word, Powerpoint, etc.
Date Range: Set to Last Year by default, allows you to adjust the amount of historical data you can scroll back through on the chart (via the history slider).
Auto Scale: Allows you to set how the scale of the Regression Chart is calculated (Current Values, Historical Min / Max, or Manually).
Title: Allows you to enter a title for the chart which will display along the top center of the Regression Chart.
Dependent Script: Set to Change() by default, but can be adjusted to use any script which provides a value other than True / False.
Independent Code: Set the code the dependent symbol is being compared to. For US Equities SPX will be used by default, for AU Equities, XJO.
Independent Script: Set to Change() by default, but can be adjusted to use any script which provides a value other than True / False.
Point Colour: Allows you to set the colour of the data points displayed on the Regression Chart.
Line Colour: Allows you to set the colour of the line of best fit.
Bars on Screen: Allows you to adjust how much historical data is being used in the current view of the Regression Chart. This can also be adjusted via the history slider along the top of the chart.
Scales Font Size: This slider bar allows you to increase / decrease the font size of the x and y scales.
Show History Slider: Allows you to hide / display the history slider along the top of the Regression Chart.