Big data was among the hot topics at this year’s Sibos 2012 trade show, so Richard Hartung sat down with a trio of technologists from Microsoft for their take on how financial institutions can best leverage big data. Hear from Microsoft’s industry solution director Colin Kerr, industry technology strategist for banking, Victor Dossey, and SQL server senior project manager, Sherri Brouwer.
“If there’s one thing the banks have, it is data,” said Microsoft’s industry solution director, Colin Kerr. On the one hand, that data means they can understand their customers and manage their risks better, while hopefully connecting better with treasuries. On the other hand, it can be harder to manage because much of the data is unstructured data, such as email and pictures, which can create reputational issues. According to Kerr banks are now looking at how to harness big data and turn it into something that manages customer connections and ups their revenue per customer.
“Using that data can require new ways of thinking,” says Kerr, while speaking to Richard Hartung at the Sibos 2012 trade show organised by SWIFT in Osaka, Japan, on 29 Oct-1 Nov. For many years banks have become used to receiving monthly or quarterly reports and tweaking them, he said. The data explosion means that they need a new model, since opportunities can come and go in milliseconds. The new model can offer master data management, which enables the user to pull data in or access it better and aids the move towards real-time analytics.
Once a bank gets control of its data, explains Microsoft’s industry technology strategist for banking, Victor Dossey, then the next step is to get it out to the people to serve them. It’s often a self-service solution, since more business user tools are available.
The Microsoft Example
Kerr explains how Microsoft itself provides a good example of how a large corporation can use big data effectively. Microsoft has approximately US$70bn spread across 1,000 accounts at 100 banks, and wants constant reports about its finances.
“In the last year we have gone through a risk management exercise that includes cash reporting, the creditworthiness of our suppliers and our portfolio of investments. In the past we had a team of analysts who built spreadsheets and external reports. Now that’s being funnelled into Paraview [the open-source, multi-platform application] for visualisation.” Microsoft staff are using self-service tools to build graphs of how trading partners are changing, or producing bubble charts to show the exposure. Treasury staff do not have to be specialists to set up a query.
Another example comes from Premier Bankcards. In the past, Kerr said, understanding how a change in their scoring mechanism might affect default rates could take months and months to understand. Now, they can do it in hours, thanks to big data technology and analysis tools.
Analysing Data to Build Better Services and Relationships
Microsoft’s Sherri Brouwer, SQL server senior project manager at the technology giant, says that banks and corporates are also leveraging big data to interact with their customers in entirely new ways. Companies can take ideas and look at them from new contexts, including using social data. The car industry and car loan financiers, for example, can now respond directly to consumers in real time using the power of data and analytic tools. Banks can also use a cloud platform to segment where data is stored and managed, so that they can parse it for information and ensure they only use the parts that are legal for them to access. The big power is taking in new data types and using it in the context of new business scenarios to create new opportunities.
What makes this customer and business analysis possible is the three layers of big data and the big computing power to interrogate it. The three layers are:
1. Data management.
“Using parallel processing power,” says Dossey, “means that getting query results back from massive amounts of data is now far faster and easier than ever before.”
Behind the scenes, Apache Hadoop software, an open source framework that enables the distributed processing of large datasets, allows users to take in poorly-structured data, identify sentiment or impact, and then use connectors to structure unstructured data and move it into analytical tools, says Dossey.
One example from the social realm was the Occupy movement, which sent out a message asking supporters to move their money out of banks on a single day as an act of protest against the bank bailouts. Research by Javelin showed that 640,000 people moved their money on a single Saturday, and millions more subsequently followed their lead. Only by penetrating the data and finding the sentiment could a big financial institution find out what happened, he added.
In transaction banking, Kerr believes banks have a tremendous opportunity to offer new services to their customers and those that monetise their data best will be the most successful. If a bank is able to gather information using big data analytics, it can offer services that provide details such as sentiment analysis, profiles of ageing of accounts receivable (A/R) or investment portfolio data to commercial banking clients. The same skills sets and technology can be used on retail banking customers.
Marketing functions at banks can also take advantage of big data analytics and the successful interrogation of traditional structured and modern unstructured data, emanating from email and social media. Business intelligence (BI) solutions can be used as commercial offerings if the parameters are correctly set. Banks and corporates can provide new revenue streams by monetising their data, adds Dossey.
If a commercial bank wants to move a level up in supporting their customers, they can map in factors such as credit risk information, internal receivables, swings in currency rates or geopolitical risk. A corporate with a supply chain could map in geopolitical factors, weather patterns and shipments from China to the US, while also becoming predictive about contextual data – some banks may be able to provide such services themselves.
Better Risk Management with Big Data
Big data produces new tools for banks, retailers and other corporations. Banks can access search engines’ social networks within appropriate privacy policies by using key words, explains Dossey. Risk management can then get a far better pulse on their customers, just if the noise or chatter about the customer or company client goes up. If they dig down a little, they might start to create triggers that can identify risk even better and share fraud alerts. “We have seen banks cleansing the data,” he says, “taking credit card data by spending and demographics and providing it to the social networks. Back-end technologies allow customers to decouple the information to ensure that the bank meets compliance standards and geographic constraints.”
The availability of this data will also result in a change in how financial institutions think, since they will be able to connect disparate data points, make decisions based on the analysis and be proactive about how to manage the business.
What is Next?
While many of these services are available from Microsoft and other software vendors, most banks and other end users have yet to take full advantage of the possibilities of big data, says Kerr. Microsoft manages its risks and cash position in its internal treasury, for example, but not factors such as the impact of weather on its operations. For retail banks the opportunity is there to learn more about their customers to serve them better or to facilitate fraud checks more seamlessly.
“Our perspective has been to drive operational efficiency and manage risk,” explains Kerr. “Streamline operations and you get a better view of your business. You don’t have to embark on a big data project per se, as you can build up to this to find value-add services later on. You can slowly start to consume elements of big data; you start with cleaning operations up, look at how to frame it and move on from there.” The important element is to have an end vision in place.
• A version of this article originally appeared in gtnews.