When IT infrastructure collapses, companies quickly realise how business critical it is. RBS learnt this lesson the difficult way this week when its systems collapsed on the busiest online shopping day of the year, leaving customers unable to use cards or pay for goods online. At the highest levels, the bank has admitted that its IT systems have suffered from chronic underinvestment for decades. RBS still cannot identify the cause of the meltdown, and is now committing spend to improving infrastructure.

IT systems are often neglected in budgets, usurped by more high profile areas of spend and leaving companies saddled with overlapping and outdated systems ill-equipped to cope with current business needs – so-called legacy systems. The financial services industry is riddled with reliance on platforms developed in the eighties and nineties, and paid heavily for it during the crisis. Some companies took weeks to identify their exposure to failed US Investment bank Lehman Brothers as a result of using ancient technology platforms governed by slow and inaccurate manual processes. Indeed, the very collapse of MF Global in 2011 has been in part attributed to the inability of the company to monitor risk on its antiquated IT systems.

There is often reluctance to make sufficient investments into renewing IT structures, particularly in times of recession when budgets are already under pressure. This is one of the factors that has led to dependence on legacy technology instead of state-of-the-art investment platforms (and no doubt partly why recent research showed that up to one-in-four investment managers is reliant on legacy systems).

However, this can prove to be a false economy, as maintaining outdated technology can carry significant costs in ongoing maintenance compared to their newer counterparts. It can also put a straight-jacket on a company’s growth ambitions as legacy systems are notoriously difficult to adapt to new business requirements. And as the RBS example shows, there is a significant cost to compensating customers when systems fail to perform their basic function. What’s more, the reputational risk associated with this incident will be hard to put a price on.