Skip to content
English
Sign in
Go to theorem.io
  • There are no suggestions because the search field is empty.

Incremental Matching

How the matching engine updates results as new trade data arrives.

Matching Only What Needs to be Matched

Theorem’s matching engine runs as new trade data arrives. It does not re-match the full population of trades from the beginning of the day at every update.

Instead, new data is evaluated against relevant unmatched activity. When new matches are found, the matching engine creates or updates Match Groups based on the matching rules and assignment priority.

This allows match results to update throughout the day without continually reshaping activity that has already been matched.

Understanding the Data Pipeline

Incremental matching occurs just after new data arrives into the Theorem data repository.

  1. A file is detected on a remote server
  2. The file is processed and imported into the Data Viewer
  3. Incremental Matching is triggered
  4. Data Viewer updates with revised match scores and results
  5. Automated outbound reports run

Tip:  Use the Match Score Bubble to tell if Matching is Currently Running

match_score_bubble_circle

The circle in the match score is green when idle and pulsates red when incremental matching is running.  After a few moments, the light should return to green. Press the refresh button to update your data. 

A report or Data Viewer search may show different unmatched results after new data arrives. An item that was previously straight out may later match, become part of a quantity break, or be tagged with a more specific discrepancy.

Matched results should remain stable, while unresolved exceptions may change as the matching engine receives more information.

What Happens to Matched Activity

Once trades are matched and assigned to a Match Group, later matching runs do not continually reprocess those trades.

This keeps matched results stable. Later updates may add new matched activity or resolve unmatched activity, but existing matched Match Groups are not rebuilt from scratch at every update.

For more information, see Match Groups and Match Group Assignment.

What Happens to Unmatched Activity

Unmatched activity remains eligible for re-evaluation.

As new data arrives, the matching engine compares it against the remaining unmatched population. New opposing activity may create a match, change a discrepancy label, or change how unmatched activity is grouped.

This is expected behavior. Unmatched Match Groups represent the matching engine’s current understanding of unresolved activity, not a permanent assignment.

Related Articles