Steps to Better Data Quality
OSIS™ currently accepts loan level data in several formats e.g. Excel worksheet, XML and XML compressed where each data field is named as per the central bank template. Please contact us if your data are spread across several worksheets or are currently held in different formats and we will help you.
Here is a link to the Demo LLD Quality Portal. In 2016, we will extend our data quality services to support banks in compliance with the ECB requirements for the EU-wide credit register, Anacredit and the FED requirement for FRY-14Q.
Loan Level Data Quality Scoring
Each quarter OSIS™ executes 12 billion data quality checks on the entire SME and RMBS universe published at the European DataWarehouse (ED). Banks continue to improve the quality of their information (see below), however, big differences in the quality of individual transactions remain. In our tool, you can look at the quality (scores) of about 400 European RMBS deals and 120 SME deals organized by country of origination. We show the percentage of correct loans or cells based on mandatory and must-have data fields of the ECB data templates and the evolution over time. After consulting with banks, investors and calibrating our models we have identified 27 must-have data fields for mortgages and must-have 27 data fields for SME’s. Hover over the “About” button to see which fields are checked for data quality. For mortgages we also apply more detailed data quality scores but these are not published on our website. For more details on the latter please have a look at an example of the OSIS™ scorecard.
We score the quality of the data in two ways. First, we simply check for each field in each loan whether there are any mistakes and divide the good loan-field combinations by all loan-field combinations. Secondly, we score at loan level where we show all loans where every field is correct and divide that by all loans.
We have different levels of aggregation: at country level, originator level and transactions level. The lowest level is available on request.
At the same time you can narrow your view at country level and compare individual originators.
Furthermore it is possible to analyse the data quality evolution through time by selecting OSIS LoanQualifier™ Time Series. To export the diagram as a picture, data, cross tab, or PDF, click the export symbol on the bottom centre.
More detailed analysis at transaction level is available as part of our online loan analytical tool LoanCracker™.
The last update of the quality scores was on 1 April 2017 which is based on availability of reports in the ED database.
To contact us please email firstname.lastname@example.org or call +31 70 32 60 370.