Generating a Summary Statistics Report

The summary stats report contains statistics for your entire company, as well as statistical data sorted by surveillance model per trader, account, or instrument.

To generate a Summary Stats report:

1. On the Trade Surveillance dashboard, select the Reports tab and click Generate Report.

   

   The reports data selection panel is displayed on the left side of the screen.

2. Select Summary Stats in the Select a Report field in the reports data selection panel.

   

3. Enter data for the following Inputs fields in the panel:
   • Start Date - End Date — Sets the time period for reporting the statistics.
   • Group By — Allows you to filter the report by "trader", "account", or "instrument".
   • Display By — Sorts and displays statistics by surveillance model.
4. Click Generate Report.

   The report is displayed on the screen.

   

   Each section of the report contains statistics in the following columns:

   • Model — Name of the surveillance model. This column is displayed for company level statistics, which can be viewed only by Trade Surveillance administrators.
   • Trader/Account/Instrument — The name of the trader/account/instrument. This column is displayed per surveillance model based on your "Group By" data selection.
   • % High Scoring — The percentage of clusters with a score of 75 - 100.
   • Change % High Scoring — Shows the percentage change in high scores since the last time period (Start Date - End Date) the report was generated. Periods must be of the same duration in order to calculate this change.
   • Clusters (0-100) — Total number of clusters scored by the model.
   • Clusters (1-100) — Total number of clusters receiving a score.
   • Clusters (75-100) — Total number of clusters receiving a high score.
   • Max Score — The highest cluster score.
   • Mean Score (0-100) — The mean of all clusters scored by the model.
   • Mean Score (1-100) — The mean of all clusters with a score greater than "0".
   • Mean Score (75-100) — The mean of all clusters with a high score.
   • Min Score (0-100) — The lowest score of the clusters.
   • Min Score (1-100) — The lowest score of the clusters with a score greater than "0".
   • Min Score (75-100) — The lowest score of the clusters that received high scores.
   • Prev Period % High Scoring — The percentage change of clusters with a score of 75 - 100 from the previous period.
   • Prev Period Clusters (0-100) — The total number of clusters scored by the model in the previous period.
   • Prev Period Clusters (1-100) — Total number of clusters receiving a score in the previous period.
   • Prev Period Clusters (75-100) — Total number of clusters receiving a high score in the previous period.
   • Prev Period Max Score — The highest cluster score in the previous period.
   • Prev Period Mean Score (0-100) — The mean of all clusters scored by the model in the previous period.
   • Prev Period Mean Score (1-100) — The mean of all clusters with a score greater than "0" in the previous period.
   • Prev Period Mean Score (75-100) — The mean of all clusters with a high score in the previous period.
   • Prev Period Min Score (0-100) — The lowest score of the clusters in the previous period.
   • Prev Period Min Score (1-100) — The lowest score of the clusters with a score greater than "0" in the previous period.
   • Prev Period Min Score (75-100) — The lowest score of the clusters that received high scores in the previous period.