This year is vital for the
European Central Bank (ECB) AnaCredit project as it goes live for Stage 1 after
7 years in the making. In previous commentary pieces, we have focused on the
differences in the way EU member states have implemented AnaCredit in terms of
scope and timelines. This commentary will give some guidance and ideas on how
firms can benefit more broadly from their ‘AnaCredit investment’.
At its
heart, AnaCredit requires Monetary Financial Institutions (MFIs) to submit 88
contract level attributes on their banking book loans, and this is without
taking identifiers into account. 72 need to be reported monthly and 16 on a
quarterly basis. To ensure the quality of the data the ECB has defined over
2,000 validation rules for consistency, completeness and referential integrity
as per table 1. Many National Central Banks (NCBs) have also added to the
number of attributes and validation checks.
Table 1: Summary of the
validation rules defined by the ECB

This level of data validation is unheard-of for this scope of
financial products at this level of detail and frequency of reporting. Which
means the data firms need to submit to their NCBs is a great source for
internal risk management and business intelligence insight. MFIs would be
making a serious mistake if they failed to capitalize on this.
However,
MFIs do need to be cautious. Data might have passed a large number of
validation checks that have been defined, but is it really correct? What if
nominal amounts or fair values being reported are of the wrong magnitude? What
if interest rates being specified are not correct? What if deal dates are also
being reported inaccurately? What if the accounting classification of assets
hasn’t been mapped correctly? Data that has been inaccurately specified can
still pass validation checks.
Whilst the ECB has defined many validation
rules, there are still several credibility checks that an MFI should do prior
to submission. These are needed so the MFI can attest that once aggregated,
the contract level data does reflect their balance sheet (assets).
Following on the point made above, it becomes obvious that there is a need
for AnaCredit solutions to comply with BCBS 239 – Principles for Effective
Risk Data Aggregation. Data analysts and regulatory reporting professionals are
used to dealing with aggregated data, not loan-by-loan details. They would
deal with information at the line of business, country, region, portfolio and
balance sheet level. They wouldn’t need to try to assess if each one of their
MFIs’ 100,000+ loans are accurately reported, attribute-by-attribute. It’s
just not humanly possible.
For years, MFIs have invested large sums in
building out their reporting framework, but they have never had to define their
data quality checks in such a level of detail. We have seen this through many
AnaCredit projects where completeness of data has been the number 1 challenge
to date before even considering quality. However, for AnaCredit projects both
are paramount.
As such in 2018, firms will have at their disposal a huge
pool of the highest quality data that is also submitted to their regulator.
This pool will also grow over time as more submissions are prepared and made.
In addition, the pool will also expand as Stages 2 and 3 of AnaCredit are
defined over the next few years. So what can firms do with this massive pool
of high quality, consistent data? The answer is “rather a lot”.
For
data points that are similar to those used in current risk and reporting
systems, the AnaCredit attributes could be retrofitted. Conceptually speaking,
this should work, however unpicking existing solutions attribute by attribute
in order to cherry pick better quality data from AnaCredit projects can also
be rather difficult. In addition, MFIs would have to deal with the impact of a
definitional change on time-series analysis – which can be tricky.
Given
that Stages 2 and 3 of AnaCredit will soon be defined, it’s worth considering
building analytic capability directly on top of a MFIs’ AnaCredit solution.
Looking solely at the attributes and measures being reported, a vast amount of
analysis and analytic capability can be delivered, without making this a
complex project. For example, a business intelligence reporting suite can sit
on top of customers’ AnaCredit data, providing MFIs with analytic insight not
only with respect to AnaCredit, but also with a level 2 analysis which
supports data quality checks across their statistical and prudential
submissions in general.
While firms are focused on compliance, they
should also take the time to look at the big picture. Having spent a
reasonable amount of time, money and resources on satisfying the highly
granular and specific requirements of AnaCredit, there are simple steps MFIs
can take to turn this cost into a benefit.
After all, it is the same
data that the ECB will use to support decision-making in monetary policy and
macroprudential supervision, so it’s key that individual firms take advantage
of this data, too.