Share the article
twitter-iconcopy-link-iconprint-icon
share-icon
CommentOctober 1 2016

How data can be used internally to prevent staff going 'rogue'

Banks can avoid 'rogue' behaviour by their employees by repurposing the technology already used for customer data to monitor staff exposures. Chris Skinner explains.
Share the article
twitter-iconcopy-link-iconprint-icon
share-icon

It is interesting to see the heat meted out to Wells Fargo over firing 5300 employees for faking account openings. Bearing in mind there are more than 265,000 people working at the US bank, that’s a relatively puny number, but the bank’s leadership team is being held to account over the issue. Should it be accountable, and can technology help? I think so.

I remember being asked by the head of audit at one bank for some form of dashboard to check that staff were following company policy. Another head of compliance requested the same, as did a leader of fraud and cybersecurity. What they all were seeking is some way to monitor frontline activity and whether it follows company policy or is breaching the rules.

This is a challenge in all aspects of banking. For example, in the case of Wells Fargo, the issue is actually culture, where the leadership of the bank has created a management that incentivises staff to be product pushers rather than customer advisers. A product pusher aims purely to attain sales targets and, in the case of Wells Fargo, the fear of missing those targets and losing your job became a key aspect of the cultural poison.

Profit-focused culture

Similarly, payment protection insurance (PPI) in the UK was a cultural issue. Here, Lloyds Bank in particular was noted as having staff adding acceptance of PPI to customer loans after the customers had left the branch.

The same issue arose at UBS with Kweku Adoboli, who lost the bank $2bn in 2011, and Jerome Kerviel, who lost Société Générale $5bn in 2008. Both ‘rogue traders’ blamed the corporate culture for their actions, and so the bottom line has to be how to instil a profit-focused culture while maintaining behavioural controls.

This is what the leaders of audit, compliance, fraud and cybersecurity were asking me, and it is the core challenge of every financial institution: how to increase market share and profitability while minimising risk and potential exposure. 

Data to the rescue

There is a way to do this with today’s technologies. We talk a lot about ‘big data’, data analytics, artificial intelligence and machine learning. These are the technologies that are being applied to customer data to look for sales opportunities, robo advice and fraudulent transactions.

What these technologies have shown is that, in real time, they can see who, what, where and how activities are taking place on a micro level, account by account. For example, several big banks are using IBM’s technology platform Watson to analyse ultra-high-net-worth individuals' accounts to provide predictive and proactive advisory services. Every morning, these clients receive personalised analytics of their investment portfolios based on their risk appetite, which show them dynamically updated information about where they have weaknesses and opportunities.

These same technologies could and should be applied internally to employee data. If a bank wishes seriously to avoid the wanton behaviours of frontline staff in retail, commercial or investment banking, then monitoring these using staff’s digital footprint of data access and update is the way to do it. This can be achieved using the same controls we are using for customer data but, unlike chasing sales for profits, this would be chasing staff for exposures. 

The real question is whether a bank has the motivation and the capability to use such technologies for these areas. After all, increasing profit is far more a focus than avoiding fines. Or is it?

Chris Skinner is an independent financial commentator and chairman of the London-based Financial Services Club.

Was this article helpful?

Thank you for your feedback!

Read more about:  Analysis & opinion , Comment