It has been over a month since RBS customers first reported problems with their accounts due to an IT failure with the bank’s systems, where an upgrade made to batch processing software CA 7 from CA Technologies is believed to be at the heart of the failure. However, debate has since moved from the cause of the problem to astonishment that it is yet to be resolved over four weeks later.
Computerworld UK reported earlier this week the one of RBS’ subsidiaries in Ireland, Ulster Bank, will not be fully operational until Monday next week, which has led to shock amongst the analyst community, branding the lengthy fix period as ‘unprecedented’ .
Alex Kwiatkowski, research manager for IDC Financial Insights, argues that RBS is now “stretching the fabric of excuses a bit thin”.
He says: “For customers to still be effected this long after the problem started is probably unprecedented. Regardless of whether it is a smaller subsidiary of RBS or not, it is not an excuse.”
However, he believes that even though the initial problem was identified, the banking group is likely to have exposed a set of vulnerabilities complex interdependencies in its systems that are hard to remediate
“It’s those unknown, unknowns. When you start diving deep into the heart of the banks’ core systems, it can be a real minefield – you can adjust one thing and it can have an impact on something else unexpected,” says Kwiatkowski.
“There has been a lot of customisation over the years, so what started looking like a circle now looks like a hexagon. So, sending in a team to deal with the problem quickly may actually be very difficult if you don’t actually know what you are dealing with in the first place.”
Jost Hoppermann, vice president of banking applications and architecture at Forrester Research, agrees with Kwiatkowski, citing a similar situation that occurred in Japan where three of the country’s banks merged their IT systems.
“Typically, this should be sorted by now, I do not know of any other failure that has had problems for a similar length. The incident in Japan did not take longer than a couple of weeks to fix,” says Hoppermann.
“The three banks experienced a major systems failure after some partial integration, which was a much more serious failure than the one at RBS, but definitely took less than a month to get back up and running,” he adds.
“RBS suffered due to operational issues, whereas in Japan it was because of a weakly planned execution. However, in Japan, they were certainly back to fully automated processing for clients.”
Research director at Gartner, John Morency, says that even though the initial problem may have created a ripple effect for the bank, where it has led to deeper issues, it should still only have taken eight to 10 days to rectify. He says: “A month is unusual”.
He suggests that RBS is likely to have assumed that the patch from CA would work without fault, but actually carried out little testing.
“Given the fact that the patch was applied after hours, which is concurrent with batch processing procedures, it was probably the case that the UK staff didn’t have a lot of planned downtime to test it. RBS probably made the assumption that because the patch was coming from CA, for a system that has been in place for a while, then it would have been pretested and prequalified,” he says.
“However, there is a question as to whether the problem with the patch would have impacted all CA7 customer environments, or if it just caused unanticipated side effects specific to the RBS environment. It isn’t clear. But given that the patch is likely to have been applied to a lot of organisations, it sounds like it might be specific to the RBS environment,” he adds.
“RBS probably made a good faith assumption that the patch was tested by CA and didn’t anticipate any side effects.”
Both Gartner’s Morency and IDC’s Kwiatkowski agree that RBS’ delay in getting the problem resolved within a reasonable amount of time is likely to lead to its customers leaving the bank altogether.
“I think this has pushed RBS customers – from what sentiments I am picking up – to thinking that this has gone from being an inconvenience to downright abysmal service. That is the sort of catalyst that will push people towards swapping banks and not waiting for it to happen again,” says Kwiatkowski.
Morency concurs: “There is definitely a possibility that RBS is going to lose customers because of this. It is assumed that financial services have the most mature, robust processes, so it is certainly going to have a negative effect – there are going to be real financial consequences.”
He adds: “It’s different for sectors like retail; their systems are far more distributed, so a central failure may not impact a branch. This is not the case for RBS – the exposure and impact is much greater for a financial institution.”
To read Computerworld UK's analysis on the cause of the RBS problems, click here.