From collateral to liquidity management: the importance of a Golden Repository

Heykel Jelassi, chief back office architect, Murex

A fragmented data architecture can threaten a financial institution’s ability to keep track of its assets. All too often, front office systems are separated out into silos of individual asset classes and lines of business, making the integration of the transaction flows from these systems very difficult, writes Heykel Jelassi.

For example, the way that transaction flows are designed and defined can be quite different; the point at which a flow is considered in ‘the past’ changes from one system to another. One platform may identify the past flows based on their value day-to-day; another may look at all of the committed flows that are already in transit and did not yet reach their value date. Constantly reconciling such a fundamental issue requires a significant level of technology resource.

An impetus to resolve this challenge has been created by the push to centrally clear over-the-counter (OTC) derivatives trades. The process of central clearing requires trading firms to post collateral as a marginal buffer with central counterparties (CCPs). This buffer protects counterparties against the bankruptcy of other traders, by allowing the CCP to fulfil the bankrupt firm’s trading obligations, while expending the collateral to cover its costs.

Although this process strengthens the financial system as a whole, it poses a few problems for individual firms. Firstly, collateral must be both highly liquid and of a strong credit rating. Such assets – government bonds and cash – are not always held by investment firms and therefore they must be exchanged with other securities or loaned by a third party.

Secondly the frequency of margin calls has increased by a factor of five, with calls made intraday, making the dynamic tracking of assets absolutely imperative. This means that firms must have an enterprise-wide view of their assets. The concept is not new, but historically this requirement was typically only found in treasury departments, operating at a much slower pace. Banks would try to have the view of the data over a period of several days.

Struggling to cope

The faster pace of movement increases risk and cost for the affected firms while regulatory requirements such as Basel 3 revised liquidity coverage ratios. Tracking which securities have been assigned to be swapped for collateral-grade securities, tracking which collateral has been posted as margin, and understanding the firm’s liquidity position are made more difficult by a lack of integration of transactional data.

As the level of activity has increased fivefold, if an error occurs it can also be multiplied, and the number frequency of errors is likely to grow in proportion with the increase in activity.

To manage this and to limit the cost, firms try to increase visibility by generating reports that detail which securities are being used for collateral from their back offices. While it may once have been feasible to manage collateral by taking an end-of-day extract of the firm’s securities position, today firms will need that information in real time. They need to take intraday decisions and to manage their counterparty exposure so that the situation that firms experienced with Lehman Brothers, where it took months to find the positions they had with a single counterparty, cannot be repeated.

Securities financing activity faces a similar challenge. In order for a firm to lend stock, either its own or that of clients, it must have a firm grip on where those securities are, and how they are being used. Again this is not necessarily easy. The trading system used in the front office may not be directly tied to the systems used to account for lending. To have the data correctly represented in the back office it must mirror the custody accounts for the securities. To know where the securities are there must be communication both ways from the front office and its depository to the back office and its depositary. For this you need uniform data in the repository so the trader can look at his position as it is impacted by other activities.

What was being managed a couple of years ago by enterprise data management (EDM) systems is being reviewed today in the light of a much more dynamic business. A centralised source of data is needed for use on a day-to-day basis by those trying to manage liquidity or collateral; if consolidation of data is done on a batch basis from across many systems, then it is impossible to meet the demands of regulators and the moderns business. The consequences can be very bad.

The golden repository

Murex sees the development of a middle layer that sits between the front and back office as the best solution to the challenge of fragmentation. This golden repository for data sits within the organisation and can facilitate working across all of the different business lines, by providing a single point of exchange in which data is normalised and held. The different front office functions can acquire trade data for risk allocation and valuation; the back office functions can use it for confirmation, settlement and accounting, custody and third party management (TPM).

Business lines such as collateral management and transformation and securities financing can be more easily supported. The technology used to underpin each individual activity can depend on the bank, however the concept of developing a central layer which sits between systems, is becoming increasing accepted.

Using this model, the bank has one position, but this is now accessible for functions such as collateral management, finance, investment and settlement, with the specific representation required by those areas. Each business line can look at the same data but see it differently.

The model reduces the costs of operating multiple data siloes, the risks involved in operating non-dynamic position-keeping and data aggregation, plus the complexity of managing different flows and types of data.

Transactional data is the source for decision-making across the organisation from the front to the back office. Pulling this information together with a single repository and a common approach can significantly improve firms’ operational efficiency.

Heykel Jelassi will be speaking at Sibos 2013 in Dubai