Software errors that destroy millions every second
While computer problems on the trading floor have often lead to serious breakdowns, these days a software error paralyzed the New York Stock Exchange for almost four hours and disclosed what is already longer visible: Software has become immensely complex and companies that save on maintenance face high risks. In his latest contribution of his Manager Magazin column Dr. Johannes Bohnet therefore commented the circumstances of the failure and showed possible solutions.
A software error paralyzes the New York Stock Exchange
With a duration of three hours and 38 minutes, the New York Stock Exchange (NYSE) these days registered the longest outage in its history. A software error was to blame for the breakdown that caused “major technical difficulties” and appeared due to “a problem with the configuration” of the computer systems. The momentous software error caused an incorrect system configuration which occured after a system update and lead to investors being unable to buy or sell shares for more than half of the trading day.
Only the fact that the US share trading has undergone an increasing deregulation and fragmentation over time prevented a momentous economic debacle. This ensured that the business could continue as usual at other major trading centers. Over the long run, more and more trading centers had been established and made it possible to trade NYSE stocks on other exchanges and alternative trading venues.
And yet the failure of the NYSE is by far not the only example of a blatant software error in the context of the stock market. For instance a software error on the 22nd of August 2013 shut down the US technology exchange Nasdaq for approximately three hours. Examples of such failures occur worryingly often – in his column Dr. Bohnet names some more of them – and quickly the question for causes and treatment options for these failures caused by software errors arises. At least in the context of the NYSE the maintenance of software seems to play an important role.
On the relevance of software maintenance and modernization
Finally, the nearly four-hour outage of NYSE especially makes its operator Intercontinental Exchange (ICE) look bad. The company had bought the tradition exchange in 2012 for $8.2 billion and has probably speculated on particular cost synergies. Although ICE neglected accusations that the software error might even be provoked by cost cuts at the wrong place immediately after the failure, there remains the impression that the software error in the New York Stock Exchange may be related to its low software maintenance.
However, the balance between the from a technical perspective important software quality and the fast time to market results intended by the management is also not easy to accomplish. Appropriate budgets and a focus on software quality in the internal incentive and reward systems are required, which in particular requires that management and development departments at any time are able to see if solutions are programmed properly or threaten to fall victim to a software error.
This is possible by the intuitive visualization of technical risks as carried out by automated data-driven software management. Such software analysis methods make it possible to recognize and deal with software vulnerabilities in a semi-automated manner. Thus, software errors and vulnerabilities can already be eliminated in advance and the desired time to market can be met.