Within the context of Analyze > Trends, segments are constructed and processed on-the-fly, displaying trends and their underlying contributors for performance, risk, exposures, and your custom indicators.
Segment construction is performed by filtering out portfolio positions based off any combination of the following characteristics:
- average daily volume
- market cap
- asset class
- long / short
Once the segment is created, all trend graphs and their respective contributors panel can be used as usual, clicking on any date of interest to view the contributors that are part of the segment.
Segments should feel similar; just like Securities Screener under the Market tab, filters are employed to search your portfolio for any positions that match your criteria.
Compare Long / Short Segments of a Portfolio
Evaluate the effectiveness of each side of a long / short portfolio using Analyze > Compare. This tool can be split a portfolio into its corresponding long and short books for instant side-by-side comparison.
Once the splits are loaded into the compare screen, each side of the book automatically populates all available metrics — performance, risk, exposure, composition, custom signals -- where each column can be viewed as a standalone entity by selecting the “normalize as standalone” option from the column header, or reverted back to the contributor view.
Currently, Analyze > Compare only supports long/short segmentation.
Research Topics: using long/short splits to view baskets
When creating smart trade experiments in a research topic, the experiment can now be split into its long/short segments.
This allows you to hone in on the basket as a standalone segment and view its positions based on the selected normalization.
Once split, a basket’s positions can be explored by selecting the menu option “view segment,” which opens the positions panel with basket summary data and a full list of positions.
From this segmented positions panel, the download icon can be used to retrieve the basket positions — with the selected normalization — into a CSV file.