It can be difficult at times to separate the hype that surrounds big data from the reality. Nevertheless, there is no doubt that big data will continue to have a profound impact on our lives. It is also clear that big data is transforming how companies operate. But what specifically will it mean for investment banks?

The Emergence of Big Data
Big data will always be difficult to define precisely and will change with time as the scale of data captured and used in the digital age continues to grow. In time it might disappear as a concept when working with vast amounts of data becomes the norm. However, today Wikipedia quotes an article that gives a succinct definition, “Big Data represents the Information assets characterized by such a High Volume, Velocity and Variety to require specific Technology and Analytical Methods for its transformation into Value”.

As the idea of big data has gained attention in recent years, there has been enthusiasm and concern in equal measure about the scale of change it will bring to business and life in general. It now seems that big data has emerged as a discipline and firms are beginning to approach it as the next frontier on the competitive battlefield. Evidence of this came last year when Gartner chose to remove big data from its hype cycle. The important question this raises is, “What now?”

What does Big Data Means for Financial Institutions?
The data problems that many financial institutions face are constantly growing. These are being driven by an increasingly electronic world, regulatory reporting needs, and the search for deeper insight to deliver the best customer service. This is in addition to other data problems that firms have been dealing with for much longer, such as pricing and risk calculation.

Both the understanding of what is possible and the technologies that have been developed to support massive data processing and analytics are developing at a rapid pace. As a result, a bank’s infrastructure needs to adapt to solve these new problems.

Whereas data was often the by-product of an operational process, it is now the primary asset that drives business decisions. This will require a significant shift towards information systems that inform the strategic decision makers rather than primarily managing and processing transactions.

This means that banks and other financial institutions no longer need to think about building solutions to specific data problems. Instead they now need to think about a unified approach to data. This will allow businesses to be run analytically and decisions to be taken with supporting evidence from the data. The more enlightened firms will realise that the biggest opportunity will be to use the data to deliver the best service to their customers.

However, customer analytics is only one area that will be of interest. At a time when revenue growth seems like a distant memory, data and analytics will help to optimise where a firms resources are directed.

The Barriers to Successfully Harnessing Big Data in Financial Institutions
Today, the biggest barrier that buy and sell-side institutions must overcome is the need to adapt structurally from what worked in the past to what is needed now. Banks for instance, are typically organised along product lines with their own technology infrastructure. Entire systems that begin with the customer, include the banking applications and then right out to market utilities, are all organised along a line of business.

In the near term, it is unrealistic to believe that banks will be able to redesign and upgrade these applications to unify their data. The danger in trying to work around this constraint is that they will add new layers to their systems, further increasing the complexity in their architecture. And yet again, as with many areas of emerging technologies, the consistent theme is that banks are short of the key skills needed to harness the potential of big data.

Another barrier is delivering the insight gleaned from big data to the end users. By definition, big data is a problem for machines rather than humans. However, making the data useful will often necessitate that it is presented in a form that can be understood by humans. For this reason, data visualisation will be a significant area of research and a beneficiary of the needs of users.

Furthermore, the value created by bringing together the vast amounts of data generated results in new problems. This is due to the increasing threats posed by cyber security and also the need to manage appropriate access to the data by employees. To solve this data entitlements will be a necessary part of the solution. As the problem of collecting and using data is solved, it will be important to secure it and manage how it is used and by whom. Ensuring that employees have appropriate data entitlements will become a core function.

What is obvious is that to overcome these barriers and seize the opportunities described earlier, firms will need a clear vision of what they want and a strategy to get there.

The Pathway to Data Analytics
For example, whilst structural barriers to enterprise wide adoption of data analytics by sell-side banks continue to exist, the potential for the technology to deliver these institutions with a competitive advantage across the breadth of their FICC sales and trading operations is vast and fast becoming a reality. As the volume and complexity of electronic trade data increases, the ability of financial institutions to derive actionable business intelligence by applying effective aggregation, standardisation and real-time data analytics could determine winners from losers in an increasingly challenging trading environment.

By working with specialist vendors and focusing on cost effective integration at a modular level, banks can begin to derive the benefits of analysis of client trading activity, venue performance, market share and profitability analytics on FICC instruments in real-time.

While this might sound like a daunting task, by applying a modular approach to integration, technology vendors can assist financial institutions to take a step-by-step approach, starting with benchmarking the health of their current data landscape, before applying effective normalisation, integration, analysis and insight to multiple sources of electronic transaction data.

The next and most advanced stage is breaking into the field of predictive analytics and machine learning, where the ability to predict future client trading behaviour based on historical patterns sets institutions streets ahead of their peers.

In what has become a challenging trading environment for all, the real winners in the race to harness and utilise big data will be those institutions that partner with the technology specialists that deliver expertise and innovation on a cost effective, modular basis.

By Diane Castelino Ph.D,

Data Science And Research Lead, Mosaic Smart Data