By Praveen Thakur
Financial service firms are among the most data driven of all businesses. The regulatory environment that commercial banks and insurance companies operate in requires them to store and process many years of transaction data. The pervasiveness of electronic trading today has also meant that the capital market firms both generate and act upon hundreds of millions of market related messages every day.
So far, these firms have relied on relational technologies, coupled with business intelligence tools to handle their ever increasing data and analytics needs. It is however increasingly clear that while such technologies will continue to play an important role; it is actually a clutch of new technologies that will have a transformative role in data management and analytics within this industry. It is interesting to note that many of these technologies have been developed in response to the data analytics challenges first faced in e-commerce and internet search industries.
Data-driven customer insight and product innovation
Banks for example are looking at ways to offer new and targeted services to their customers in order to increase revenues, customer engagement and reduce churn. To offer targeted services, they must integrate their transactional data (customer call records, customer emails and claims data etc.) with external information from Web logs, social media feed etc. to derive a 360 degree view of customer behavior and preferences. Integration of these different sets of data requires deployment of Big Data technologies.
Banks can then use Real-Time Decision (RTD) making software to generate offers like a new savings scheme that matches customer need at that point in time. The same can be sent to customers in real-time through smartphones, Facebook, website and call-center channels.
Sentiment analysis and brand reputation
Whether looking for broad economic indicators, market indicators, or sentiments concerning a specific organization or its stocks, there is a mass of data on the Web that can be harvested, from traditional as well as social media sources. While keyword analysis and entity extraction have been with us for a while, and are available from several data vendors, the availability of social media sources is relatively new, and has certainly captured the attention of many institutions looking to gauge public sentiment.
For example, sentiment analysis by studying social media is finding increasing acceptance. A hedge fund like Derwent Capita ? there are others too ? is basing its strategies on trading signals generated by twitter feed analytics. Even outside the rarified world of hedge funds, most firms at this point are using sentiment analysis from social network to gauge public opinion about specific companies, markets or the economy as a whole.
On-demand risk analytics
?On-Demand? risk analytics is now the desired goal of global banks, especially at a trading desk level. The objective is not just a faster measurement and reporting of risk, but also measurement across asset classes. Aggregation of global positions, pricing calculations, and Value at Risk (VaR), all fall within the realm of Big Data. This is due to the mounting pressure to speed these calculations up well beyond the capacity of current systems and the need to deal with ever growing volumes of data.
While firms have adopted the use of compute grid technologies to enable faster risk computations, the feeding of data into these grids has become a bottleneck. Technologies like Java based in-memory data grids help overcome these bottlenecks. They enable faster calculations by allowing real time access to in-memory positions data and Map Reduce style parallelization.
For example, such architecture helped a global bank reduce VaR calculation time from 15 hours on a 200 node grid to 20 minutes on a 60 node grid, while at the same time the number of simulations was increased almost 25-fold.
Rogue trading detection
Rogue trading continues to be a huge threat for the financial institutions. Deep analytics that correlate accounting data with position tracking and order management systems can provide valuable insights to detect signs of rogue trading, which are not available using traditional data management tools.
For example, in couple of well-known cases (UBS and Soci?t? G?n?rale), inconsistencies between data managed by different systems could have raised red flags if found early on. This might have prevented at least a part of the huge losses incurred by the affected firms. Here too, a lot of data needs to be crunched from multiple, inconsistent sources in a very dynamic way, requiring a new technical approach to the analytics platform.
Fraud detection involving cards, debit, and wholesale payments is also quickly becoming a big data problem, in as much as correlating data from multiple, unrelated sources has the potential to catch more fraudulent activities early on, than what is possible with the current methods. Consider for instance the potential of correlating Point of Sale data (available to any credit card issuer) with Web behavior analysis (either on the bank’s site or externally), and potentially with other financial institutions or service providers such as First Data or SWIFT, to detect suspect activities.
Payment providers have developed fraud detection tools that take advantage of massive data assets containing not only financial details for transactions, but IP addresses, browser information, and other technical data. These tools help financial institutions refine their models to predict, identify, and prevent fraudulent activity. These tools enhance the traditional approaches to fraud prevention, which are mostly based on sanctions lists and pre-defined rules.
Investigation and compliance (e-Discovery)
A few recent cases where financial institutions were found guilty of misappropriating their customers? funds or misguiding them about fund allocation, has caused litigators and regulators to push for far greater degrees of scrutiny of information than ever before.
To detect such malpractices, investigators have to go through records of all the interactions associated with a financial transaction such as a trade order ? including emails, phone transcripts, text messages, contracts, etc. All this data is unstructured in nature. Retention of these records has been mandated for years, but the difficulty in associating them with the corresponding transactions has caused regulators to look for a far greater degree of correlation between structured transaction records and unstructured interaction records. Big Data solutions help create this relation by bringing disparate levels of structure into the holistic data management platform, which can maintain their relationship and correlations.
To conclude, data driven decision making tools are helping financial institutions grow their business by improving customer experience, reducing risk, and meeting regulatory needs. In fact, financial firms are slated to be one of the fastest adopters of Big Data and analytics solutions.
The author is the vice president for technology at Oracle Asean