A vast quantity of structured and unstructured data is now generated and mined in the financial industry. It is analysed by firms as well as regulators to provide insight. For those that do it well, the business value they can derive can confer a competitive advantage. Yet, big data in the regulated world of financial services is not without its dangers. To negotiate the fine line between risk and opportunity, we can expect to see growing demand for data analytics skills and solutions in financial services.
Financial institutions – both those that are highly regulated and those that are less so – are having to compete in rapidly changing markets. They need to make best use of all the tools at their disposal to secure a competitive advantage. That includes using big data analytics to provide valuable insights from data they generate all the time on customers and products. This has resulted in the creation of a market in analytics that is being met by big data vendors.
Regulatory reporting and structured data
Since the events of 2007/08 huge demands have been placed on the regulated financial sector. Globally, regulators now demand more data. Firms can’t just send an unstructured data download either; the demand – and expectation – now is for structured data.
Reporting institutions have to formulate their data to match defined common standards with strict validations for everything from equity transactions, retail financial instrument sales data, Over The Counter derivative trades, line-by-line reports of cash flows, to assets and their capital requirements. The structured nature of the billions of records collected by the Federal Reserve, or the European Banking Authority and other regulators is in place precisely to enable SQL, as opposed to NoSQL, to find the hot spots and the outliers they need, from within that structure.
Commercial use of behavioural data
Inevitably there are business uses within financial institutions for the behavioural data of clients. This can be analysed and turned into information to be embedded for business intelligence and marketing purposes, especially when it is combined with social media data and its business equivalent.
Commercial players may now own in-house a large enough pool of behavioural data, or may be able to buy or licence access to such data, as the commercialisation of financial big data analytics starts to take root. Valuable datasets are related to the financial behaviours of customers (and not just of those in the retail space); there are enough behavioural trends of everyone involved in the financial sector and they are all driven by emerging disintermediation in finance at large. Big data will inform this.
The data overhead of the ‘unregulated’
The inroads peer-to-peer banking are making into traditional banking represents an example of market disruption occurring as a result of market forces, and not just because of the availability of big data analytical processing vendor solutions.
This sector is unregulated in the traditional sense but still has extensive obligations around personal data protection. In Europe, firms need to be sure that data is being used only for the purpose it was originally collected. The difference between ticking or not ticking a box in the small print can mean a lot to the survivability of an individual’s data further down the line.
While tapping into the abundance of data available may bring benefits of insight to firms and transparency to regulators, it is not without its risks.
For example, it is possible that the relevance of some data may only become clear after it has been analysed and a trend or outlier has been discovered. Thus, the very sensitivity of the data - and therefore its potential misuse - may only become apparent after the event.
Also, for regulators and marketers to make predictions they have to collect more data than might appear necessary at the time, as some data sets may only become valuable way into the future. By this time, issues around Personally Identifiable Data or other concerns may have already been caused. For this reason, firms collecting data should employ segregation or siloing.
This becomes especially relevant in the instance of a merger or acquisition. As company data pools are merged they may reveal information which is contrary to the initially acceptable use of the collected data.
Data lakes owned by firms, held by regulators or quite deliberately built by data aggregators may give rise to ’open data’ where massively sourced information may become public domain material. But data quality and reliability would be questionable in that context, as no provider of ‘open data’ would accept liability for the quality nor the potential non-infringing nature of it.
Firms have to stick with best practice. Chiefs of Risk, Compliance and Data are aware that specific compliance rules for the use of big data are as much an inevitability as has been the technological capability to find the information within it in the first place. To ensure compliance, firms need to mark the compliance-sensitive elements of the data they collect; Big data vendors have filters and probes which can highlight sensitivities in home-grown, acquired or licenced data sets.
These issues may result in new regulatory disciplines around the use of financial behavioural data. There may be pressures to adjust consent notification so that old and non-consented data goes stale and expires; opportunities open up as new and better consented data become available.
Big data is a part of the paradigm shift happening in financial services overall. Lombard Risk is now seeing a trend in firms making data processing investments as a result of regulatory costs. This will be another way for regulated firms to get business value from collecting vast volumes of regulated data.
By James Phillips, Regulatory Strategy Director, Lombard Risk.