Pre-matching: not the silver bullet as hoped
27 June 2019
Catherine Talks of UnaVista suggests that pre-matching is not the silver bullet the industry is hoping for and may even create additional complications
Image: Shutterstock
Matching the two sides of a trade is one of the most important and historically difficult parts of regulatory reporting. Recent data from the European Securities and Markets Authority (ESMA) showed that just 40 percent of swaps trades reported under the European Market Infrastructure Regulation (EMIR) are matched in the trade repository (TR). When matching rates are this low it makes it very difficult for the regulator to accurately monitor systemic risk. To increase the success rate the industry has created a number of solutions, aiming to improve the data quality and reduce the number of breaks that result in the reconciliation process.
There are two schools of thought on the best way to improve the process, pre-matching and post-matching.
Pre-matching is a process where two parties send their portfolio of messages into a platform for a reconciliation to take place prior to reporting to the TR. The idea is that if the data is matched at source then it will naturally match once inside the TR. There are many market vendors looking to implement pre-matching processes that could be used as platforms to disseminate the unique transaction identifiers (UTI) between parties, ensuring improved pairing within the TR. Although there are some benefits to attempting pre-matching, there are several complications which could cause pre-matching to have a reduced success rate.
Time to resolve validation failures
With such a short timeline between trade execution and the reporting obligation cut off time, there is very little time to perform and correct the matching and reconciliation prior to sending the report. If the report is not corrected by the reporting required time, then it must be sent to a TR regardless of the number of residual pre-matching breaks. In addition, time must be allowed for potential resolutions after a report is sent into a TR and there may be a number of validation failures to address before the message can be accepted. It is only once the message is accepted by the TR that the reporting obligation is complete.
Disparate vendors and data
Pre-matching requires both counterparties to use the same pre-matching vendor for the same report dataset to attempt a match. As pre-matching is not a regulated requirement there may be some firms who simply decide not to utilise a pre-matching vendor. If the counterparty is using a different vendor then a match will not be possible as there isn鈥檛 an inter-vendor matching process in place. Many of the pre-matching services also only offer the service for some of the four reportable asset classes, meaning a firm may need to use multiple pre-matching services to get the sought-after benefit, but this leads to the undesired outcome of your data in more disparate systems.
Data issues
Most pre-matching and reconciliation takes place on a subset of data. To ensure a good level of matching, the number of fields required to match is often restricted to core economics, depending on the pre-matching vendor model. In contrast, the number of fields required for the inter-TR matching process is vast (upwards of 90 fields over the matching phases). It is therefore possible that even with a pre-matching process that there may be breaks at the TR matching level.
An alternative method
UnaVista believes that the most effective route to better matching is using the intra- and inter-matching capabilities of the trade repositories. These will provide the best results with the fewest delays. This has also been recognised by the regulators, who have prescribed that TRs must perform the inter-TR matching process.
Intra-TR matching
Once a transaction is accepted by a TR it is then subject to the first level of trade repository reconciliation, called an intra-TR matching. This is where the TR uses the matching key (UTI, reporting party, other counterparty and master agreement) to seek the other side of the transaction within its own repository of transactions. If both parties report to UnaVista then the results of this will be displayed back to the clients along with any breaking fields. If there is no match within UnaVista鈥檚 Trade repository then UnaVista will give the transaction a reconciliation status of 鈥榰npaired鈥�.
Inter-TR matching
UnaVista will keep seeking an internal match until the inter-TR matching process is undertaken on T+2. The inter-TR process works by each TR providing a file in XML format containing the matching key, where two TRs have the same key then a trade is 鈥榩aired鈥� and a separate file containing the trade economics is exchanged. UnaVista then displays back to the client both sides of the reconciliation i.e. the client side and counterparty side along with any differences. If there are no differences, then the transaction moves to a state of 鈥榬econciled鈥�.
Improving the inter-TR matching rates
Although matching rates within EMIR still have a long way to go, there are a number of reasons that SFTR should see an improvement. Firstly, unlike EMIR, ESMA has introduced a standardised format for the reporting to a trade repository, this means no conversions will need to be made during the inter-TR matching process, this should really improve matching rates. ESMA has also prescribed some field level tolerances which allow a marginal difference in some values to count as a match, which means even if certain attributes are a decimal place off due to a rounding error, those reports can still match.
Industry bodies anticipate that more than half of the volume of SFTR transactions sent to TRs will actually be single sided. The total number of transactions that are eligible (and able) to be reconciled will, therefore, be relatively small in comparison to the number of submitted transactions. So, matching SFTR reports may not be as onerous as it has been in previous regulations.
Ultimately the more a firm can do to ensure their data is correct at every stage of the reporting process the better. However, our belief is that pre-matching is not the silver bullet the industry is hoping it is and may even create additional complications for accurate reporting.
There are two schools of thought on the best way to improve the process, pre-matching and post-matching.
Pre-matching is a process where two parties send their portfolio of messages into a platform for a reconciliation to take place prior to reporting to the TR. The idea is that if the data is matched at source then it will naturally match once inside the TR. There are many market vendors looking to implement pre-matching processes that could be used as platforms to disseminate the unique transaction identifiers (UTI) between parties, ensuring improved pairing within the TR. Although there are some benefits to attempting pre-matching, there are several complications which could cause pre-matching to have a reduced success rate.
Time to resolve validation failures
With such a short timeline between trade execution and the reporting obligation cut off time, there is very little time to perform and correct the matching and reconciliation prior to sending the report. If the report is not corrected by the reporting required time, then it must be sent to a TR regardless of the number of residual pre-matching breaks. In addition, time must be allowed for potential resolutions after a report is sent into a TR and there may be a number of validation failures to address before the message can be accepted. It is only once the message is accepted by the TR that the reporting obligation is complete.
Disparate vendors and data
Pre-matching requires both counterparties to use the same pre-matching vendor for the same report dataset to attempt a match. As pre-matching is not a regulated requirement there may be some firms who simply decide not to utilise a pre-matching vendor. If the counterparty is using a different vendor then a match will not be possible as there isn鈥檛 an inter-vendor matching process in place. Many of the pre-matching services also only offer the service for some of the four reportable asset classes, meaning a firm may need to use multiple pre-matching services to get the sought-after benefit, but this leads to the undesired outcome of your data in more disparate systems.
Data issues
Most pre-matching and reconciliation takes place on a subset of data. To ensure a good level of matching, the number of fields required to match is often restricted to core economics, depending on the pre-matching vendor model. In contrast, the number of fields required for the inter-TR matching process is vast (upwards of 90 fields over the matching phases). It is therefore possible that even with a pre-matching process that there may be breaks at the TR matching level.
An alternative method
UnaVista believes that the most effective route to better matching is using the intra- and inter-matching capabilities of the trade repositories. These will provide the best results with the fewest delays. This has also been recognised by the regulators, who have prescribed that TRs must perform the inter-TR matching process.
Intra-TR matching
Once a transaction is accepted by a TR it is then subject to the first level of trade repository reconciliation, called an intra-TR matching. This is where the TR uses the matching key (UTI, reporting party, other counterparty and master agreement) to seek the other side of the transaction within its own repository of transactions. If both parties report to UnaVista then the results of this will be displayed back to the clients along with any breaking fields. If there is no match within UnaVista鈥檚 Trade repository then UnaVista will give the transaction a reconciliation status of 鈥榰npaired鈥�.
Inter-TR matching
UnaVista will keep seeking an internal match until the inter-TR matching process is undertaken on T+2. The inter-TR process works by each TR providing a file in XML format containing the matching key, where two TRs have the same key then a trade is 鈥榩aired鈥� and a separate file containing the trade economics is exchanged. UnaVista then displays back to the client both sides of the reconciliation i.e. the client side and counterparty side along with any differences. If there are no differences, then the transaction moves to a state of 鈥榬econciled鈥�.
Improving the inter-TR matching rates
Although matching rates within EMIR still have a long way to go, there are a number of reasons that SFTR should see an improvement. Firstly, unlike EMIR, ESMA has introduced a standardised format for the reporting to a trade repository, this means no conversions will need to be made during the inter-TR matching process, this should really improve matching rates. ESMA has also prescribed some field level tolerances which allow a marginal difference in some values to count as a match, which means even if certain attributes are a decimal place off due to a rounding error, those reports can still match.
Industry bodies anticipate that more than half of the volume of SFTR transactions sent to TRs will actually be single sided. The total number of transactions that are eligible (and able) to be reconciled will, therefore, be relatively small in comparison to the number of submitted transactions. So, matching SFTR reports may not be as onerous as it has been in previous regulations.
Ultimately the more a firm can do to ensure their data is correct at every stage of the reporting process the better. However, our belief is that pre-matching is not the silver bullet the industry is hoping it is and may even create additional complications for accurate reporting.
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 Securities 麻豆影视传媒 Times
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Securities 麻豆影视传媒 Times
