5 Key Takeaways from the CPMI-IOSCO “Harmonisation of critical OTC derivatives data elements (other than UTI and UPI)” – second batch

by Alan McIntyre

CPMI IOSCO

In October 2016, CPMI-IOSCO published their Harmonisation of critical OTC derivative data elements (other than UTI and UPI) – second batch 

Here are our 5 Key Takeaways:

  1. Is ISO 20022 the standard?

Risk Focus are enthusiasts regarding the benefits that a clearly defined data schema such as ISO 20022 can bring to improve data quality. And as I’ve commented in various arenas, the adoption of ISO 20022 in European regulatory reporting has a clear momentum. Outside of Europe, though, we have yet to see this data schema embraced.

It’s therefore interesting to see CPMI & IOSCO citing ISO 20022 as the standard for 7 different fields including ‘Confirmed’, ‘Day Count Convention’ and ‘Strike Price’. I’m a fan of ISO 20022 but there is no denying that FpML is the lingua franca for derivatives globally, and yet isn’t referenced at all in the consultation. This is despite FpML reporting being supported by various regulators including The CFTC, The SEC, The JFSA and The HKMA.

I wasn’t therefore surprised to see the ISDA/FpML response to the consultation comment on this:
The coverage of OTC derivatives in ISO 20022 is currently under development and one of the observations is that there is an attempt to establish new codes in ISO 20022 to reflect national regulations, rather than to leverage established codes used by the industry. Where there are established codes used by the industry, we urge CPMI-IOSCO to recommend the use of these established codes rather than introducing new ones. ISO 20022 can work with pre-existing external codes.

  1. Can we harmonize fields that don’t exist?

As Risk Focus stated in the response we submitted to this consultation we commend CPMI-IOSCO for taking the lead on this area and are fully supportive of these efforts. We know from our day to day discussions with our client community how much effort has gone into and continues to go into improving the quality of the data reported to TRs, ARMs and regulators. This attempt to harmonize fields at a global level therefore fully deserves support.

That said it’s a slightly unusual situation whereby CPMI & IOSCO seem to be harmonizing some fields that don’t currently appear to be requested by any G20 regulators. I assume the intention is that by introducing new fields as distinct concepts, ambiguity will be remediated on the related fields.

‘Final Settlement Date’ would appear to be one such case. By attempting to define a new field hopefully there is a beneficial effect on the existing related fields such as ‘Settlement Date’ (in EMIR) and ‘Date of Settlement’ (in CFTC Dodd-Frank), both of which support multiple dates currently.

Less clear though are fields like ‘Booking location of counterparty 1’ and ‘Location of counterparty 1’s trading desk’. The exact usage of these proposed fields and clear definitions of the relationship to related concepts such as branch codes and the ongoing work of the LEI ROC would therefore be welcome if these fields are to be adopted.

  1. Reaching a final settlement

Like ‘Final Settlement Date’, the field ‘Settlement Currency’ is eminently sensible at first glance but proves more challenging upon further investigation. We commented in our response that the CFTC have a corresponding field in Dodd-Frank whereas ESMA do not under EMIR.

The GFMA commented in their response on their unease with the definition “the currency for the cash settlement of the transaction” as this is problematic for NDF’s where the actual currency is not delivered. They argue in favour of a revised definition: “the settlement currency of the transaction in the case of cash settlement”.

  1. How to capture the null and void

This one brings a smile to my face as I recall many discussions on N/A. At times it felt like trying to get consensus on a solution to peace in the Middle East might be a more pragmatic exercise. The argument starts something like this. Does N/A mean Not Applicable, Not Allowed or Not Available? If we assume (at our peril no doubt!) that it’s Not Applicable, then what is the required system behaviour if an N/A field is populated? It might seem fairly simple at first glance but I recall some fairly existential and circular discussion with clients on this one.

The CPMI IOSCO consultation has 27 references to ‘null’. These range from null for “in the case there is no interest to be paid”, to null as “not applicable”, and null when “is not known at the time of reporting”.

Next comes the tricky matter of “does null mean the field is left blank or is the intention to populate actual value ‘null’?” As someone involved in building systems I would always prefer to have an actual value populated for the simple reason it shows that the submitter has intentionally used null rather than simply neglected to populate the field.

This is less straightforward that it might seem and I am reminded of the original draft of the ESMA Level 1 Validations where ESMA mandated that the value ‘na’ was to be used for used for various fields when the value was “not yet available”. The problem was that a round peg into a square hole often doesn’t fit. Your database administrators will no doubt confirm that date and numeric formatted fields simply cannot store the value ‘na’. As a result we ended up having to compromise with some fairly inelegant workarounds such as ‘1900-01-01T00:00:00Z’ and ‘999999999999999.99999’ to indicate ‘not available’.

  1. The key timeline question

We have travelled a good distance down the transaction reporting road. EMIR will be 3 years old in February 2017, MAS was 3 in October and CFTC Dodd-Frank reporting will be the envied older sibling with a later bedtime when it turns 4 in January 2017. A lot of lessons have been learned in those 4 years and a lot of progress has been made. For those of us on the front-line working on the various projects, the pace of change has seemed at times relentless.

EMIR has a new RTS with revised fields and field definitions due to go live late next year. MAS has new asset classes going live in the middle of next year. And The CFTC issued a consultation in December 2015 which it seems likely will lead to changes sometime in 2018. And all of the above is before we consider any eventual changes that may arise from this CPMI IOSCO consultation.

Further harmonization across the G20 derivative reporting regimes is a laudable goal and one that in time will not only further the ability for regulators to aggregate the data globally but will also make it simpler for firms to meet their various reporting requirements. But the banks and reporting firms who have been dealing with this ‘tsunami’ of regulatory reporting demands (and let’s not lose sight of MiFID II and SFTR) know that achieving all these reporting projects takes time, money and expertise.

Many will be wondering what the uptake of these CPMI-IOSCO recommendations will be across the G20 reporting regimes and more to the point what the likely timelines will be. We may be four years down the transaction reporting road but it looks like this road stretches out a long distance towards the horizon. I guess there’s no choice other than to fasten our seatbelts, enjoy the ride and see what perspectives come into view as the journey continues.

Click here to contact us