The age of analytics: what banks have to gain

The familiar yet wholly unwelcome trio of a tough economic environment, increased regulatory pressure and changing customer behaviour and demands has been driving banks to change and innovate in many parts of their operations. In response, data, never the most glamorous of variables, is becoming more high profile than ever. Anyone with even half an eye on the world of business technology cannot have failed to notice an increased focus on the topic, particularly when it comes to buzz-worthy 'big data' touted by many a research firm or technology vendor.

The attention is very much warranted, however. In the banking world, data and its analysis are taking on more significance than ever, helping financial services firms comply with legal requirements, boost efficiency, identify and meet customer needs more effectively, and refine trading strategies and tools.

Edwin Van der Ouderaa, global managing director of financial services business at Accenture, says the use of analytics is growing tremendously year on year, with 20% to 30% growth of revenues from analytics projects. Analytics are fast becoming the 'core engine' for many banking activities.

Analytics pioneers

Capital markets are, in many ways, pioneers in the use of data analytics. After all, data and its analysis are essential to any successful trading strategy. This is truer now than ever, as markets are becoming increasingly inter-related, and the different data sets and variables required to trade successfully are multiplying.

Analytics are being used to give trading operations a clear-cut edge in areas such as cross-asset trading, says Rich Brown, global head of quantitative and event-driven trading solutions with Thomson Reuters. "If you look at how tightly markets are inter-related [with] each other in the quantitative area and beyond, there's a lot of the cross-asset analysis people need to do," he says.

Sources of data used to formulate strategies are also expanding far beyond mere market data to include news and other text sources. Thomson Reuters, for example, offers a product that measures the tone (how positive or negative) of news articles relating to a particular company, based on a dictionary of about 20,000 words and phrases. As standard, the product bases analysis on about 60 sources, typically news wires. However, it can be adapted to draw on social media sources. There is even a hedge fund, Derwent Capital Markets, which bases its trading strategies entirely on the analysis of overall sentiment on Twitter.

Less exotic usage is also on the increase. Already very common, analytics are ever more important to day-to-day operations as markets become more difficult and more expensive to trade in, liquidity gets harder to come by, and awareness of risk parameters grows. The changes are dramatic. The latest in analytics technology allows banks to make calculations far faster and far more accurately. Where value-at-risk volatility calculations used to be performed for the next day because they took about 20 hours, now they can be done in seconds or minutes, allowing banks to use risk analytics in their daily operations, rather than as a management activity performed afterwards to sum up exposures.

Now pricing and exposure can be adapted in the day, according to Mr Van der Ouderaa. "In the future it will become an activity of a matter of seconds, so you can dynamically change pricing and follow portfolios," he says. "If you can overlay 'what if' scenarios over current operations, you can start managing your business in a different way with a different perspective on what the possibilities and range of risks are. That for me is the most important change in capital markets."

Swift analysis of an ever-increasing influx of data is a competitive necessity, adds Mr Van der Ouderaa, particularly with the technological 'arms race' that is so integral to the investment banking world. He predicts that in three or four years, any firm not working in an at least quasi-real-time manner and integrating analytics into the day-to-day management of their trading desk will not be able to compete.

Customer focus

On the retail side of operations, analytical efforts tend to be more customer-focused. "I've had a number of requests, including from CEOs of large banks, asking how they can get better customer insight. How they can understand what customers want, which perhaps the banks haven't thought of, and their habits," says Dave Stuckey, a technology leader with PricewaterhouseCoopers. "They want to know how to leverage technology to get that knowledge, that understanding of those habits in order to create the right offering for the customer. That's becoming an absolute push right now."

Indeed, banks are moving from a broad categorisation of general tiers of customers to microsegmentation; an understanding of individual groups of customers with the same beliefs, values and demographics who are likely to react in the same way to branding, pricing and marketing. Banks can then tailor the products they offer those groups.

It must be an alluring thought for marketing executives. Gone will be the days of mass mailings with a miniscule uptake, of spamming customers aged 18 to 80 with offers of the same services regardless of income, goals, background, savings and myriad other considerations that might dictate how likely they are to take up a credit card, pension, mortgage, auto-finance or student loan offer.

Customers too may benefit; a world free from banking junk mail has a distinct appeal. And a happier customer is one who is far more likely to stick around. The cliché that customers are a bank's most important asset has never had bigger implications. The cost of acquiring a customer is increasing, so departure can have a big impact on profitability. At the same time, customers are a source of deposit funding. And if a bank does not have capital to lend, it has to turn to the wholesale markets and buy it in, which can be costly in current markets.

Such technology can be extended beyond predictive capabilities to profitability or financial analytics, says Laurence Trigwell, IBM's global financial services executive for business analytics. Targeted analytics can help banks to understand the profitability of a specific customer and forecast the success of the next product that will be marketed to them. This should, then, be win-win. For a customer, the bank will only offer very relevant products that anticipate their needs, while equally they should only receive offers that are profitable to the bank.

Customer turn-off

It may not be quite that simple, however. It is certainly true that customers dislike mass marketing. But too easily they might feel that every aspect of their life is being monitored to help boost a bank's bottom line; a scenario not likely to provoke outpourings of gratitude.

"Customer data analysis is fraught with ethical dilemmas, and the way banks handle and apply analytics will affect their reputation massively," says Harvey Lewis, research director with Deloitte Analytics. "Banks have unprecedented access, but getting things wrong in terms of individual customers is a big risk."

According to a recent survey by Deloitte on the collection and use of data, only 17% of customers were happy to receive obviously tailored communications from retailers. Of course, that does not stop well-known retailers such as Amazon doing so on a regular basis. "There's a massive gap between what organisations want to do with customer data and what customers want them to do," says Mr Lewis. "If you get something wrong with an offer, it is counterproductive, to say the least."

Perhaps unsurprisingly, when it comes to access to, and analysis of, data, the survey also found that customers are particularly sensitive about their financial information, making the issue even more pressing for banks.

Indeed, banks that have been as proactive as the information allowed them to be – using what in marketing parlance is described as a 'push' strategy – have not always been net beneficiaries, says Accenture's Mr Van der Ouderaa. "[Push strategies] had, to their surprise in the beginning, an enormous backlash. Banks were trying to do the right thing, because they thought people wanted personal, rather than blanket, impersonal marketing. But customers said it was too much like Big Brother," he says.

Instead, the most advanced marketing strategies now 'pull', so that when a customer walks into a branch, picks up a phone or logs in online, the bank is ready to have a personalised discussion. They will know if your child is reaching university age, for example, but will not bring it up unprompted. Instead, it will be ready to have a conversation about student or housing loans, and already be in a position to offer or decline various services without lengthy approvals. "Pull strategies are very important and are the future," says Mr Van der Ouderaa. "They mean banks are ready to personalise, but are not so aggressive."

This too will become essential in the near future, says Barrie Neill, a banking consultant at SAS UK and Ireland. "By implementing big data analytics, banks can detect much earlier signs of disengagement from their customers, strengthen customer interactions in marketing campaigns or make smarter product pricing predictions by revealing additional insights into the sales process. It will move from value-add to a necessity when regulation creates an environment where account switching is simplified and barriers to churn fall."  

Analytics also offer the potential to improve security, ensure customer identity, and have anti-fraud and anti-money laundering capabilities. "Analytics allow for unprecedented techniques for detecting fraud in real time. And so that's a vast area of development which has huge implications in terms of making transactions more reliable," says Mr Van der Ouderaa.

Analytics for business

Uses of analytics are prevalent in the business banking world too, although they are being used in a somewhat different manner. In common with retail operations, though, customer analytics are a big focus, says Michael Knorr, head of integration and data services for Citi Transaction Services. However, given the nature of its operations, it also possesses a detailed knowledge of specific customers.

"Our target market in Citi [Transaction Services] is very well defined. We know our core customers," says Mr Knorr. "Data analytics is more about understanding what the right opportunities for different product solutions that we might want to approach them with."

To do so, the bank has developed internal business management tools, allowing client and product managers a better view of client activity across Citi's operations, helping target the right products and services to its clients, and better managing revenue attrition.

These data sources do not just include payment volumes; Mr Knorr also expects to draw on more unstructured sources of data, such as call transcripts or deal memos.

Analytics are also offered as a service to business customers. For example, card transactions have been analysed for some time, with types of product sales and payments by banks collated, and in turn provided to their customers. This, says Mr Knorr, was then enriched by data obtained externally, linking transactions with underlying commercial data. It is now moving, he adds, to include cross-product analytics, supplying customers with wholesale information across multiple product types, on what a business's customers want.

Regulatory demands

A tough competitive landscape has certainly driven the development of analytics in each business segment. But some drivers are more universal. In the post-Lehman world, regulatory demands have skyrocketed: greater oversight and the need to get a better grip on all aspects of risk is being mandated, in the hope of anticipating when a major or systemic problem might occur.

IBM's Mr Trigwell says that analytics will prove crucial in meeting the demands of regulators and central banks as they exert greater control over the financial services industry. IBM is working with both extensively. Central to addressing regulatory demands is greater transparency and a much more granular understanding of an organisation's risk position, as well as ensuring that such information is calculated appropriately and used intelligently.

In fact, Mr Van der Ouderaa estimates that about 40% of analytics investment is currently being directed towards dealing with various forms of risk, whether it be in underwriting the right sort of credit or improving the market risk management system to better handle trading book concentration, tail risk, stress risk and other related exposures.

However, meeting regulatory requirements necessitates a broader view of risk, across the business. It is not as simple as merely adding values from investment, retail and business banking arms. When it comes to risk, two plus two equals four does not apply; different correlations and concentrations of risk makes things rather more complicated. Enter data analytics. "What we're seeing is the application of risk analytics to create an aggregated view and perspective on risk which is distributed across the organisation," says Mr Trigwell. "It's about intelligently aggregating risk with all the business rules."

Harmonisation drive

There may now be a pressing need to perform all risk calculations across silos, but unfortunately this has not been the case in the past. Instead, data gathering, processing and definitions grew organically and, as a result, are likely to be rather different depending on the business unit, leaving banks with a sizeable task in harmonising standards across all product areas and silos.

"A lot of the work banks are doing now is to clean that up and turn it into enterprise-wide architecture so that everyone talks the same language and true enterprise-wide risk management is possible," says Accenture's Mr Van der Ouderaa. "It's one of the big practical investments and it is easier said than done."

Analytics also have the potential to find cost excess in particular processes, geographies, product lines or channels. "We have seen banks try and slash costs, which gets you a reduction, but it is not intelligently extracting cost from the business, and that is what has to happen," says Mr Trigwell. "We need to create a much more aligned and granular understanding of the cost excesses that sit in the business functions, from finance and IT to human resources, and to do that we need to understand the services those things are providing and correlate those things together." He adds that some movement in this direction is being pioneered in internal shared service centres, using financial analytic process understanding to drive operational efficiency improvements.

The advantages of data analytics for banks are huge, but the challenges can be equally big. Making the most of analytics will require deciding how best to analyse data and, in an age of unprecedented volumes of information, which data is the most beneficial. There is doubtless still much to be done in terms of trial and error. Those that get it right, however, will reap benefits across business segments, improving their competitive profile, operating far more efficiently and avoiding being crushed by an oppressive array of regulatory requirements.