<img height="1" width="1" style="display:none;" alt="" src="https://px.ads.linkedin.com/collect/?pid=1132820&amp;fmt=gif">
Skip to content
All posts

Case Study | From 6 Hours to Under 1 Minute: Migrating Equity Attribution to Snowflake with FactSet

Return to Resources

Continuus, an Acuity Analytics company, moved equity attribution from flat files into Snowflake, cutting processing time from up to six hours to under one minute and giving business users self-service portfolio restatement.

The Client

A global investment management firm that relies on equity attribution, which explains what drove a portfolio's returns, such as stock selection versus sector weighting.

The Problem

The firm ran attribution through a slow, multi-hour flat-file process. Restating a portfolio, or rerunning attribution for a past date, meant coordinating business users, IT, and FactSet every time.

Business users couldn't trigger a restatement themselves, so each request became an IT ticket and a vendor support call.

The Solution

Continuus moved the attribution process out of flat files and into Snowflake.

  • Rebuilt the attribution workflow in Snowflake, removing the manual file handling between systems
  • Connected directly to FactSet through Snowflake data sharing, so data arrives without file transfers or vendor coordination
  • Built stored procedures that let business users restate any portfolio for any historical date on their own, without IT or a FactSet support call
  • Compared the new outputs with historical flat-file results to confirm they matched before going live

The Outcome

  • Processing time reduced from up to six hours to under one minute, so analysts can rerun and explore portfolio results during the day instead of waiting overnight
  • Business users restate portfolios themselves in Snowflake, with no data engineering request or new file extract
  • Removing the flat files took a fragile manual step out of the workflow, lowering the risk of data errors and the effort to manage recurring runs
  • The same setup can take on additional attribution models and datasets

Download the Case Study

View or download the full case study below.