Tutorial

How to match I/B/E/S vintage without the landmines?

A practical step-by-step Stata recipe for cross-vintage matching.

The problem

Working with I/B/E/S data across different vintages can be tricky. Identifier changes, data updates, and formatting inconsistencies create landmines for researchers. This tutorial provides a practical guide to navigating these challenges.

Key steps

  1. Download the Stata codes and linking table from the link above
  2. Understand the structure of I/B/E/S detail history files
  3. Use the provided linking table to match broker IDs across files
  4. Apply the reverse engineering codes for missing analyst IDs
  5. Validate your matches using the diagnostic checks in the code

Common pitfalls

Pitfall 1: Assuming analyst IDs are consistent across time. They are not always. Use our codes to identify and handle inconsistencies.

Pitfall 2: Ignoring the difference between ESTIMATOR (detail history) and ESTIMID (recommendation history) broker IDs. Our linking table maps between these.

Pitfall 3: Not checking for duplicate forecasts after merging. Always deduplicate carefully.

Resources

For the full methodology and validation, see our paper: Good-bye I/B/E/S (or not?)

We update this analysis annually. Check the paper page for the latest updates.

← Back to Good-bye I/B/E/S