As we reach the first anniversary of implementation of the Securities Financing Transactions Regulation (SFTR) for the first cohort of firms, it seems an appropriate time to ask ourselves what we have learnt. Jonathan Lee, senior regulatory reporting expert at Kaizen Reporting, shares his assessment
In spite of the pandemic, and some would say against the odds, the industry delivered and should be given credit for doing so in such difficult circumstances. This can be seen in the trade repository acceptance rates, which have been far higher than many expected. However, before we get too carried away, we must remember that producing valid SFTR reports is the easy part.
Data proves to be ‘valid but wrong’
One of the biggest issues highlighted during the first year is that the majority of reported activities, while passing validation, are in fact inaccurate. This is what at Kaizen we call the ‘valid but wrong’ problem. A staggering 93 per cent of SFTR reports tested by Kaizen for the first time exhibit errors and questionable population, typically reported with valid (they pass trade repository validation rules) but wrong data.
Data in the public domain is low grade but informative
The public data published weekly by the trade repositories appears to indicate under-reporting and double counting. In spite of aspirations, this data also holds very little value to the industry in terms of making markets less opaque or countering asymmetries of market data. Under-reporting, especially of collateral, is a problem and one likely to come under greater regulatory scrutiny.
To report completely and accurately, reporting counterparties need the specific details.
We also identify a lack of clarity in areas such as bilateral margin reporting, collateral re-use and cash reinvestment reporting — and the reporting of commodity financing transactions has led to an apparent lack of reports covering these areas. In spite of repeated industry presentations prior to go-live about the significance of bilateral margining for repo, this has proven something of an afterthought in the reporting taxonomy.
Trade Repository reconciliations are far from perfect
The principal mechanism regulators put in place for promoting data quality under SFTR is the trade repository reconciliation. The overarching concept is that if both firms submit reports directly from their books and records, then trades and post-trade lifecycle events should ultimately match. If both sides agree, this should provide assurance that the data is correct. However, this has proved to be rather a naive premise for two key reasons.
First, the vast majority of SFTR reports are single-sided and not subject to reconciliation. According to DTCC, the Depository Trust and Clearing Corporation, these figures currently stand at 76 per cent in the EU and 86 per cent in the UK. In a recent Kaizen webinar, only 32 per cent of participants polled appreciated the ineffectiveness of the trade repository reconciliations (selecting the level of TR reconciliation at 0-25 per cent).
Second, SFTR was so ambitious in its data reporting requirements that a large proportion of reporting counterparties did not, and still do not, have enough reference data or static data to complete a report. Therefore, much of the data is sourced from third parties, not the reporting firm’s books and records, and is of mediocre to poor quality. This data is far below the standard required to ensure systemic risks are controlled and that market manipulation is detectable.
So, despite a start that should be applauded, there is still work to do. As we head into the second year, firms should focus on the completeness and accuracy of their own reporting. But these efforts should not be at the expense of resources devoted to addressing the challenges of trade repository reconciliation.
Regular, periodic, full universe quality assurance through regulatory testing will provide a significant advantage in meeting the SFTR compliance challenge.
