Radar Embedding Example

The IZA/Fable Data Consumption Indicator: Germany is based on consumer and small business transactions from Fable Data, a leading provider of anonymized, pan-European spending data was developed collaboratively by IDSC and Fable Data. Fable Data sources bank and credit/debit card operators and compiles a dataset of expenditures of distinct consumers. 

The index presented here is constructed using high frequency daily data with a lag of 3-4 days. The method used to construct it is the so called “one year look back rolling panel” which is found to correlate best with official data from Eurostat and others. This method provides monthly percentage difference YoY and solves the problem of a dynamic composition of the panel structure.

Since consumers are joining and leaving the panel we proceed as follows: for each month in each year we aggregate the expenditures in that month in that year as well as those in the same month of the previous year using the consumers that are active in both months. We then take percentage difference. This way the index gets better as the panel gets richer and more representative and at each point in time we have the best information available. Since we look at percentage differences the numbers remain comparable over the entire time coverage.