Staying Compliant: A new type of reconciliation
Having the processes and controls in place to reconcile reported data with internal data is an obligation, but efficiency is hard to realise.
The sheer volume of data required by regulators is a burden for reporting firms. It is an operational challenge to gather and process the data. Under-reporting, misreporting and over-reporting are offences under many regulations giving additional motivation, or pressure, to get the process right.
Under the current raft of capital market regulations, reporting has become more complex and demanding than ever before. For example, a firm may have some disclosures processed by third parties while others are issued by various in-house teams and a greater frequency of reporting is required. Consequently it can be challenging for firms to establish that what they have reported corresponds with their internal information.
Historically reconciliation is a function that is tied to the checking of bank balances, primarily a back office process. Most of the technology platforms that exist in the reconciliation space have been built to service that function. Yet as regulatory reports increasingly look like forms of transaction, the need for regulatory reconciliation is growing.
The UK’s Financial Conduct Authority (FCA) and its predecessor have levied substantial fines upon firms that have failed to report in line with regulatory standards; £33 million in fines have been set since 2009, with £13.2 million of those set in 2015 alone. Fines have a massive impact for the firms affected, not only in terms of the money spent on the fines but also in the impact on share price, and the less tangible but very real reputational damage.
Resolution and clarity
When a firm is found to have transgressed the rules, its operations team will often review the model of data submission, comparing what has been submitted against the data held in the firm’s own systems. However, if this is only checked when a problem has been identified then the damage will have already been done. A more frequent system of checks is needed to ensure that the reports are up to scratch according to the FCA’s guidance on ‘Systems and controls’.
The weakness of that model under the existing regime has proven challenging, with nine major sell-side institutions being fined by the FCA for reporting failures since 2009. When the Markets in Financial Instruments Directive (MIFID) was first introduced in 2007, 28 fields covering the economics of a trade were required. The FCA advised firms to conduct reconciliation by requesting reported data back from it. That required market participants to access the regulator’s website and download a file, which would then be reconciled with the source system. Engaging in the process demonstrated that a firm had systems and controls in place.
Now regulators are changing up several gears; the revision of MiFID (MiFID II and MiFIR) that will come onto effect anywhere up to a year after January 2017 will require some 65 fields to be completed, containing a lot more granular level detail on individuals and on counterparties’ client data, capturing personal data elements as well as more traditional transactional data, and far more detail in the process of order generation and execution. And more recently, the latest ESMA consultation paper on MiFIR reporting clearly places a greater emphasis on this reconciliation control framework being in place for organisations.
The European Market Infrastructure Regulation (EMIR), the Dodd-Frank Act in the US and various similar regimes in G20 and non-G20 countries – such as Singapore – all demand reports for derivatives trades that in some – but not all – cases cover listed instruments, standardised contracts as well as over-the-counter non-standard contracts. Reporting is burdensome, checking that reporting even more so. Simply put, increasing the number of data elements increases the potential for error.
The question is how to approach the task in a smart manner. Firstly the business needs to consider the complexity of the task, based upon the rules it is exposed to and the inherent complexity of its own organisational structure.
Secondly the level at which reconciliation is mandated and managed should be decided. Different models for different regulations would be unworkably expensive and unmanageable for all but the simplest of firms. Manual processing of reconciliation can only really take a ‘spot-check’ approach due to the scale of the task. That will only provide limited effectiveness. An alternative would be to adapt a firm’s existing reconciliation technology in order to support this process. To take this approach will require a flexible system given the complexity of the reports. Ensuring that ownership of the project is not kept at the level of the unit or division but is managed as an enterprise-wide project will decrease system entropy now and into the future.
Development or adoption of a reconciliation platform for regulatory reports must take several functional elements into consideration:
- Firstly, whatever happens to a trade must be accurately represented in the company’s own data. If a trade’s details are altered post-trade the regulator will want to see that a firm has made the amendment and has visibility over the change.
- Secondly, and specifically within MiFID reporting now, when information is sent to the FCA, it changes data elements because it has its own reference data. Consequently the file a firm can request back from the FCA will not look exactly the same as the version sent by the ARM. Having a platform supported by a dedicated regulatory reconciliation team, offering expert or peer support for interacting with the national competent authority, allows for a more confident approach.
- Thirdly, a platform should be as automated as possible. Comparing one piece of data with another is straightforward but even if just 1% of trades are exceptions, a major firm processing five million trades a week will need to sort through 50,000 records.
A platform must offer effective workflow, auditing and data. Many generic reconciliation platforms will come up short. Many instruments, for example bonds, are largely traded over-the-counter (OTC), consequently there is not the electronic paper trail that often exists with platform-traded instruments. Nevertheless under MiFID II they will be subject to reporting requirements. Functionality must be wedded to expertise and the capacity to data across from one format to another so that is can be reconciled on a fully matched basis. To understand this, the provider of the reconciliation platform has to have knowledge of the rules, the instruments, the regulators and the processing that the data will be subject to. Anything less will not match up.