Management of liquidity risk in financial institutions has been a major thrust of regulation enacted in the wake of the Great Recession. Many American and European banks recently went through stress tests to certify the quality of assets in their books. The Basel III framework from the Bank of International Settlements – often called the Central Bank for Central Banks – has raised the percentage of senior capital required to be maintained by banks in liquid and near-liquid assets. Since liquidity risk, in simple terms, is faced when a bank runs out of money, these measures are not difficult to understand.
However, there’s another angle to liquidity risk that has nothing to do with asset or capital and lies outside the purview of regulatory oversight. In plain language, this pertains to a situation when a bank has enough money to meet its commitments but is unable to push its high value payments messages out the door before the expiry of stringent daily deadlines – called cutoffs – imposed by payment schemes. In other words, if a bank has to make a payment to a corporate or a financial institution on a particular day, its payment instruction must reach the clearing agency (say, TARGET2, which is the RTGS scheme operated by European Central Bank for Euro cross-border payments) before the cutoff time (5PM CET in the case of TARGET2). When it comes to multi billion dollar payments – large banks do many of them every day, by the way – missing the cutoff is almost tantamount to default since the intended receiver of the payment might (wrongly) conclude that the sending bank has failed to make the payment because it doesn’t have enough money to do so.
From my experience at a Top 5 UK bank that processes around EUR 250 billion of payments every day, I learned how liquidity risk of this sort can be caused by technical instability, inaccurate reference data, and other mundane but critical problems that plague high value payments landscapes in many banks.
In one instance at this bank, a faulty EUR 100 payment message corrupted the payments message queue, which in turn prevented the processing of a 10 billion Euro payment by the cutoff time. In another instance, the software’s insistence upon 11-character BIC codes led to thousands of payments bearing 8-character BIC codes getting diverted from the STP route, with barely enough time for them to be manually repaired and resubmitted before cutoff. Both ‘incidents’ caused grave threat to the bank’s liquidity position and called for extensive damage control measures from its top management.
These examples clearly underscore the importance of “stress tests” on a bank’s high-value payments processing technologies and system landscapes. Covering throughput, end-to-end latency, ability to withstand burst volumes and a plethora of other technical stuff, these tests could be carried out either by a bank’s operations team or outsourced to an external technology vendor who has a good grasp of the payments and liquidity domain.