The Exchange commissioned an independent review from Ernst & Young to establish the original cause of the problems and to recommend steps to be taken to reduce the risks of a recurrence. The Exchange fully accepts Ernst & Young's conclusions and recommendations, and has discussed with the Financial Services Authority the findings of the review and the steps that have been taken and are under way to prevent a similar occurrence in future.
The review found that a combination of circumstances affected the processing of data that occurs at the end of each trading day. This process is an overnight procedure which normally would take 40 minutes to complete. On this occasion a software fault in handling the data caused the process to take over four hours. This fault was exacerbated by the mechanical break-down of a back-up tape, which had caused an earlier delay of over 40 minutes.
As a consequence of this over-run, elements of the overnight process were still running when the start-of-day procedures commenced automatically at 4 a.m. This created a conflict in the system which resulted in the incorrect processing and subsequent communication of some of the previous day's data to customers.
Although on the basis of the initial assessment it was thought that a solution might be available quickly, after full investigation it was found necessary to close the system and re-start it - a process that takes some three hours.
Commenting on the outcome of the review, Don Cruickshank, Chairman-designate of the London Stock Exchange, said: "We very much regret the disruption and inconvenience caused on 5 April. The investigation has identified the series of events that led to the late opening of the market. As a result, additional systems control and managerial measures have been implemented in order to ensure that such a problem should not occur again."
In order to ensure reliability for the future, the following steps have been taken:
Software coding fault
Additional controls to software management have been implemented, including increased monitoring and assessment of system performance.
Management of IT systems
Procedures and controls in respect of the identification and escalation of faults have been revised and steps have been taken to minimise the risks of further systems malfunctions.
Back-up system
Although the Exchange has back-up systems to cover hardware faults, their use was not appropriate on this occasion. Our contingency arrangements are therefore being reviewed in order to cover data corruption issues and software faults.
Contingency planning and communication
Contingency arrangements have been broadened to ensure comprehensive coverage of decision-making in the event of problems, and measures have been taken to improve communication with customers and ensure consistent messages to the market.
