Article Contents
This page explains how to map indices across mixed granularities (weekly, mid-month, monthly, quarterly) so they can be compared or combined. For definitions and background, see Data Granularity.
Index Structure & Scaling
- Normalisation: The first month = 1.0. Subsequent period values evolve from that base.
- Cumulative value: A running total through each
PERIODENDDATE.
- Weekly availability: Where available, weekly values are scaled relative to the first month.
- Example: a weekly value of 0.1 indicates 10% of the first month’s volume.
Recommended Approach
To maximize comparability across tickers and periods, we recommend merging all available granularities into a single daily dataset. This is not true daily measurement, but rather a daily-frequency approximation derived from cumulative monthly/weekly values. It gives you a consistent canvas to:
- Align tickers with different granularities, and
- Leverage higher-frequency data where available, without losing broader monthly trends.
How to derive daily from cumulative indices:
We provide a SQL template that merges the available durations (week, month, quarter, etc.) and creates a daily time series. Below is a worked-example if you prefer to build from first principles:
Made with Bullet