SFTR testing updates, time is the challenge
03 July 2020
IHS Markit discusses current testing, the vexing issue of UTI generation and ISO 20022 schema challenges under SFTR, which many firms will be going down to the wire to solve
Image: Colorlife/shutterstock.com
With the start of the Íø±¬³Ô¹Ï Financing Transactions Regulation (SFTR) reporting delayed until 13 July, firms are all keen to use this extra time given by regulators as effectively as possible: testing has never been more important. Data quality was stated to be expected by regulators as of go-live date, so where do we stand now on SFTR testing?
The focus has moved from user acceptance to production like volumes and scenarios testing and we have observed an accelerated shift to production end-to-end testing until reporting to the trade repositories (TR), more complex agency lending flows, triparty collateral data and central clearinghouses data testing. However, the testing window is shrinking day-on-day while some areas still remain to be more tested such as some agency lending flows, collateral re-use, delegated reporting and parties have expressed concerns around unique transaction identifier (UTI) pairing.
In quest of the UTI
UTI design, its generation and exchange are certainly at the core of current tests with an intense investigation on ‘alleged’ and ‘unpaired’ trades, with some actors able to leverage on vendor pre-matching platforms to help find the other side, where relevant.
A major challenge relates to the dual-sided nature of the reporting regime with upfront UTI bilateral agreements with all counterparts and consistent UTI configuration has proved quite difficult, ending up in bespoke solutions requests. Clear communication to the counterparties is key to address UTI reporting issues as stressed recently by the European Íø±¬³Ô¹Ï and Markets Authority (ESMA) during the virtual event hosted by IHS Markit on June 10.
Exchanging minimum key ‘pairing’ fields proved quite challenging, especially as both sides of the trades may not always ‘synchronised’ in submitting their events in the appropriate order and in the complete data set whilst corrections on UTI wrongly assigned could prove operationally intensive if trades were already reported to the trade repository.
On a positive note, we observed parties could streamline their process to some extent and rectify key pairing elements, sometimes engaging in additional clean up exercises on counterparty and agency lending correct identification or adjusting their source systems and operational processes.
After getting the UTI, the next challenge lies with the individual fields reconciliation, where pre-matching platforms could help. In current testing, parties look at major ‘day one’ breaks: recurrent discrepancies are on execution timestamp, master agreement type, trading venue, value date, collateral basket identifier, quantity or nominal amount. It is worth mentioning exceptions in all the fields related to master agreements where some firms may not populate consistently when their trade contract names could differ across countries with a value not necessarily covered by the setlist allowed by ESMA in the relevant field.
While we expect some of those discrepancies to be resolved during the course of the testing such as around reference data, some others will take more time to remediate, and some may only be fixed after post-go-live along with operational processes and control changes. Addressing some of the gaps may take time and could occur at different timelines across participants, which could result in ‘persistent’ exceptions and realignment efforts post-go-live.
In quest of the ‘ACK’
Another major challenge is the making of the ISO 20022-compliant message and the successful report to the trade repository with an acknowledged (ACK) message.
Common issues relate to XML messages failing to be compliant with the ISO schema: missed mandatory fields, skipped cross-validation rules or incorrect format.
A significant part of the rejections in user acceptance testing and production were related to inconsistent dates sequencing and missed ‘action type choreography’ such as subsequent lifecycle events and collateral, valuation messages being reported without prior successful submission of the initial new trade report.
Some of the rejections observed were due to LEI status or country having an invalid format or being inconsistent with the Global Legal Entity Identifier Foundation database.
Security reference data, issuer LEI, country of jurisdiction of the issuer on the security loaned/borrowed or collateral still presented some gaps, in which case, correction/process should be put in place to address such missing data. But, how does one manage a whole collateral message rejection due to a single erroneous line within the whole set of collateral securities? One approach could be to retrieve the erroneous line out of the message and report the rest of the lines. Specific error in the Pre-NACK/exceptions screen could be pointed out to facilitate errors management.
In addition, analysing the feedbacks from TRs were insufficiently clear to accurately remediate issues on a timely fashion, hence pre-TR exceptions (pre-NACKs) monitoring could sometimes be useful. For example, ‘bulk’ rejection messages (NACK) on collateral messages could make it difficult to identify errors on specific collateral components when the whole ‘collateral block’ is rejected.
Vendor platforms can bridge that gap providing more details within the error messages prior to TR reporting, categorising exceptions by field category and criticality with the possibility to assign errors resolution to a specific member of operational teams: this exception monitoring covers files ingestion errors, fields cross-business rules and points out critical issues preventing the XML message creation or preventing from getting a UTI.
Categorising errors and having appropriate operational process around exceptions remediation is key. Consequently, we observed queries on user interface and tools/reports to facilitate investigations were increasingly numerous, which relates back to ‘data quality’ requirements, and on ‘how can data be verified, how are the data given back to reporting entities?’ As was pointed out by ESMA recently.
We also observe that interfacing to several TRs represents a challenge: whilst the core of the ISO 20022 XML schema is the same for all TRs, there could be some technical divergences in some peripheral elements, which sometimes requires bespoke developments and re-testing such as responses files handling, entity permissioning, opening hours of the channel, business days, headers and technical records.
In addition, on the field reporting level, the validation rules themselves, the ISO schema and the ESMA guidelines sometimes gave room to interpretation, for example, around fields marked as conditional in the rules but optional in the ISO schema or vice versa (currency, collateral basket identifier, collateral component), action type corrections on trades with net exposure, the way to report collateral and reconcile with late guidance clarification presented some implementation challenges among participants.
Industry bodies such as the International Capital Market Association, the International Íø±¬³Ô¹Ï Lending Association and the Association for Financial Markets in Europe have certainly helped establish best practices to work around some of those ambiguities and testing issues were actively conveyed to each trade association. In some countries, other trade associations have actively supported ‘broadcasting’ members concerns, such as AMAFI in France and BVI in Germany.
Epilogue
At the eve of go-live, a lot of challenges remain as shown in current testing. At this stage, the focus is still on UTI pairing on trades, collateral and successful reporting to the TR on production-like scenarios.
Some regulatory guidance and subsequent clarifications and TR implementation amendments left insufficient time for participants to change their build.
Multiple actors engaged ‘in the reporting chain’ also included external parties which were out of the scope of SFTR reporting obligations.
On the whole, current testing windows still seem quite narrow while firms are still facing severe operational challenges related to COVID-19 measures, and reduced testing resources. Time is the main challenge.
On the positive side, we have seen a lot of engagement from all market participants and industry associations along with TR and market infrastructures and vendors to build reporting solutions, get reports ‘out-of-the-door’ and move to production without losing track of more strategic implementations in line with best practices and regulatory requirements.
The focus has moved from user acceptance to production like volumes and scenarios testing and we have observed an accelerated shift to production end-to-end testing until reporting to the trade repositories (TR), more complex agency lending flows, triparty collateral data and central clearinghouses data testing. However, the testing window is shrinking day-on-day while some areas still remain to be more tested such as some agency lending flows, collateral re-use, delegated reporting and parties have expressed concerns around unique transaction identifier (UTI) pairing.
In quest of the UTI
UTI design, its generation and exchange are certainly at the core of current tests with an intense investigation on ‘alleged’ and ‘unpaired’ trades, with some actors able to leverage on vendor pre-matching platforms to help find the other side, where relevant.
A major challenge relates to the dual-sided nature of the reporting regime with upfront UTI bilateral agreements with all counterparts and consistent UTI configuration has proved quite difficult, ending up in bespoke solutions requests. Clear communication to the counterparties is key to address UTI reporting issues as stressed recently by the European Íø±¬³Ô¹Ï and Markets Authority (ESMA) during the virtual event hosted by IHS Markit on June 10.
Exchanging minimum key ‘pairing’ fields proved quite challenging, especially as both sides of the trades may not always ‘synchronised’ in submitting their events in the appropriate order and in the complete data set whilst corrections on UTI wrongly assigned could prove operationally intensive if trades were already reported to the trade repository.
On a positive note, we observed parties could streamline their process to some extent and rectify key pairing elements, sometimes engaging in additional clean up exercises on counterparty and agency lending correct identification or adjusting their source systems and operational processes.
After getting the UTI, the next challenge lies with the individual fields reconciliation, where pre-matching platforms could help. In current testing, parties look at major ‘day one’ breaks: recurrent discrepancies are on execution timestamp, master agreement type, trading venue, value date, collateral basket identifier, quantity or nominal amount. It is worth mentioning exceptions in all the fields related to master agreements where some firms may not populate consistently when their trade contract names could differ across countries with a value not necessarily covered by the setlist allowed by ESMA in the relevant field.
While we expect some of those discrepancies to be resolved during the course of the testing such as around reference data, some others will take more time to remediate, and some may only be fixed after post-go-live along with operational processes and control changes. Addressing some of the gaps may take time and could occur at different timelines across participants, which could result in ‘persistent’ exceptions and realignment efforts post-go-live.
In quest of the ‘ACK’
Another major challenge is the making of the ISO 20022-compliant message and the successful report to the trade repository with an acknowledged (ACK) message.
Common issues relate to XML messages failing to be compliant with the ISO schema: missed mandatory fields, skipped cross-validation rules or incorrect format.
A significant part of the rejections in user acceptance testing and production were related to inconsistent dates sequencing and missed ‘action type choreography’ such as subsequent lifecycle events and collateral, valuation messages being reported without prior successful submission of the initial new trade report.
Some of the rejections observed were due to LEI status or country having an invalid format or being inconsistent with the Global Legal Entity Identifier Foundation database.
Security reference data, issuer LEI, country of jurisdiction of the issuer on the security loaned/borrowed or collateral still presented some gaps, in which case, correction/process should be put in place to address such missing data. But, how does one manage a whole collateral message rejection due to a single erroneous line within the whole set of collateral securities? One approach could be to retrieve the erroneous line out of the message and report the rest of the lines. Specific error in the Pre-NACK/exceptions screen could be pointed out to facilitate errors management.
In addition, analysing the feedbacks from TRs were insufficiently clear to accurately remediate issues on a timely fashion, hence pre-TR exceptions (pre-NACKs) monitoring could sometimes be useful. For example, ‘bulk’ rejection messages (NACK) on collateral messages could make it difficult to identify errors on specific collateral components when the whole ‘collateral block’ is rejected.
Vendor platforms can bridge that gap providing more details within the error messages prior to TR reporting, categorising exceptions by field category and criticality with the possibility to assign errors resolution to a specific member of operational teams: this exception monitoring covers files ingestion errors, fields cross-business rules and points out critical issues preventing the XML message creation or preventing from getting a UTI.
Categorising errors and having appropriate operational process around exceptions remediation is key. Consequently, we observed queries on user interface and tools/reports to facilitate investigations were increasingly numerous, which relates back to ‘data quality’ requirements, and on ‘how can data be verified, how are the data given back to reporting entities?’ As was pointed out by ESMA recently.
We also observe that interfacing to several TRs represents a challenge: whilst the core of the ISO 20022 XML schema is the same for all TRs, there could be some technical divergences in some peripheral elements, which sometimes requires bespoke developments and re-testing such as responses files handling, entity permissioning, opening hours of the channel, business days, headers and technical records.
In addition, on the field reporting level, the validation rules themselves, the ISO schema and the ESMA guidelines sometimes gave room to interpretation, for example, around fields marked as conditional in the rules but optional in the ISO schema or vice versa (currency, collateral basket identifier, collateral component), action type corrections on trades with net exposure, the way to report collateral and reconcile with late guidance clarification presented some implementation challenges among participants.
Industry bodies such as the International Capital Market Association, the International Íø±¬³Ô¹Ï Lending Association and the Association for Financial Markets in Europe have certainly helped establish best practices to work around some of those ambiguities and testing issues were actively conveyed to each trade association. In some countries, other trade associations have actively supported ‘broadcasting’ members concerns, such as AMAFI in France and BVI in Germany.
Epilogue
At the eve of go-live, a lot of challenges remain as shown in current testing. At this stage, the focus is still on UTI pairing on trades, collateral and successful reporting to the TR on production-like scenarios.
Some regulatory guidance and subsequent clarifications and TR implementation amendments left insufficient time for participants to change their build.
Multiple actors engaged ‘in the reporting chain’ also included external parties which were out of the scope of SFTR reporting obligations.
On the whole, current testing windows still seem quite narrow while firms are still facing severe operational challenges related to COVID-19 measures, and reduced testing resources. Time is the main challenge.
On the positive side, we have seen a lot of engagement from all market participants and industry associations along with TR and market infrastructures and vendors to build reporting solutions, get reports ‘out-of-the-door’ and move to production without losing track of more strategic implementations in line with best practices and regulatory requirements.
NO FEE, NO RISK
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Íø±¬³Ô¹Ï Finance Times
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Íø±¬³Ô¹Ï Finance Times
