Ross McEwan, the chief executive of Royal Bank of Scotland, says he is “angry” with a “small group of people” following the bank’s shaming in the recent foreign exchange trading scandal. You can understand why.

But the most infuriating thing of all about this sorry saga is that RBS and the four other banks collectively fined £2.6bn by UK and US regulators didn’t need to end up where they did.

In reality, they had all the information they needed to prevent their forex teams landing them in hot water, but the problem was that they couldn’t see, let alone make sense of it.

Evidence aplenty

As a scandal that went back years — the whacking fines relate to activities going back as far as 2008 — there was a mass of evidence.

To top it all off, what the traders were doing wasn’t some sophisticated exercise that couldn’t have been detected. The fact that compliance got there in the end confirms as much.

This was really just standard ‘manual lag’ (it took ages for people to find what a machine could have found in a fraction of the time).

Afternoon ‘fix’

So what happened? The banks’ traders were able to rig the forex markets at crucial moments in order to rip off their clients by planning their strategies in online chatrooms.

They then placed a handful of very large trades at just the right time – usually shortly before the afternoon ‘fix’, when benchmark exchange rates are agreed across the currency markets.

Both those activities left an electronic footprint that could have been used to raise the alarm at the traders’ employers, but the alarm never sounded, and the all-important data remained in the dark.

Legacy systems

The forex-rigging scandal is certainly a reminder that, for all the excitement about the potential of ‘Big Data’, many firms, especially financial services ones, have not yet learned how to harness the information they have.

These firms are immersed in vast data ecosystems, but their reliance on legacy IT infrastructures means they have little idea what these ecosystems really consist of and even less of an idea how to connect one bit of data to the next — especially when companies are heavily siloed.

This poses an obvious problem: it’s only when you join the dots, and add context, that you can see the bigger picture — in the example above, highly suspect trades.

Shadowy data

The forex scandal is a good example of ‘Dark Data’ at work — or perhaps that should be ‘not at work’.

The evidence of the traders’ wrongdoing was always there, but was hidden in the shadowy unstructured data of their chatroom logs and the curious trading patterns recorded at specific times.

The banks, of course, aren’t alone in their Dark Data failings: most financial services companies are unable to exploit the intelligence lying in Dark Data, whether that’s a missed chance to increase revenues, or a lost opportunity to spot wrongdoing.

Dark Data is an issue all financial services firms must now get to grips with, especially with regulators breathing down their necks more than ever.

After all, a business that can’t shine a light on all its data risks drawing misleading conclusions from the data that it can see – like, for example, making the mistake of assuming its forex traders are all little angels.