13th January 2014 by Mick McAteer
The phenomenon that is ‘Big Data’ is much discussed these days so I thought it would be worth giving an overview of what this might mean for consumers.
Big Data is not new. Quantum physicists, meteorologists, astronomers, geneticists and, of course, internet providers have relied on a combination of unimaginably large volumes of data, technological innovation and massive processing power for some time now. Financial markets – particularly capital markets, stock exchanges and insurers – also rely on big data to run operations. But a more recent development is the application of big data, coupled with the huge growth in social media, in retail financial services.
It all seems very exciting and there are plenty of people around evangelising about the benefits of big data and its ‘transformative potential’. But what are the implications for consumers? Can we say with confidence what the potential benefits and risks are for consumers? Are we genuinely on the cusp of something transformative? Or should we be more sceptical and treat it as just another one of those ‘innovations’ which promised much but delivered little?
Civil society and consumer groups are fully engaged in the debate about big data and technology with regards to surveillance and human rights (for example governments and internet providers snooping on citizens). However, we have not really engaged in the debate about the potential impact of big data and technology in financial services.
So, how do we approach this issue? As usual, the best way to assess the impact of any market innovation on consumers is to use the established consumer principles to identify the potential benefits and risks and detriments for consumers. These consumer principles (or outcomes) are: access and choice; safety and security fairness and integrity; efficiency and genuine innovation and what are termed externalities.
In theory, there are a number of clear potential benefits for consumers if the financial sector exploits the advances in information science and technology in an efficient, fair and responsible manner.
But the same features of big data that create potential benefits can also create risks and detriment for consumers. It is not possible to cover the entirety of big data in a single article so it may be helpful to focus on several important aspects: customisation and personalisation; the effects on distribution; risk management; and investment behaviours.
Customisation and personalisation: the capacity to track and analyse huge volumes of data on trends, activities and behaviours means firms should be in a position to manufacture customised and personalised financial products and services better suited to the needs and preferences of individual consumers and groups of consumers. In theory, big data should lead to more effective financial innovation.
Conversely, customisation and personalisation could result in restricted access and choice for growing numbers of consumers through the extension of price differentiation and red-lining (where financial firms are able to define more precisely who they don’t want to serve). There is a very real risk that big data will lead to more, not less, financial exclusion.
Moreover, consumer advocates argue that there has been very little genuine financial innovation in financial services over the years except for things such as the ATM (a technological innovation) and UCITS funds (a legal/ regulatory innovation). Many innovations have been variations on a theme providing the same core products but with unnecessary added features or innovations such as securitisation – of mortgage and other debt – innovations which were originally a good idea but were distorted. The increased use of big data could simply result in more confusion marketing and more socially useless or what might be called hyper financial innovation.
There is a strong argument for saying that too much choice is just as bad as too little choice. Will yet more choice really add anything to the sum of human welfare?
In addition, civil society advocates have more fundamental concerns about the use of big data.
Consumers are often not even aware that they are giving away behavioural data to financial providers. This means that we have to rely on firms holding that data to act with integrity. But this could be incredibly difficult to oversee and regulate given the sheer volume of data involved and complexity of the algorithms employed by firms.
Distributional effects involve the ability to target more precisely the customers for a product or service and could result in less waste in the supply chain and make the distribution and marketing of products in turn more efficient. This could allow the industry to extend access and choice to more consumers.
But, again, if big data simply leads to more fragmentation and hyper innovation, this could simply push up product manufacturing costs, search costs and, ultimately, distribution costs. Moreover, more efficient, precise targeting of profitable customers could lead to embedded cross subsidies being exposed. This could result in consumers who are costly to serve paying disproportionately more in distribution costs.
Risk management: the essence of risk management is having enough relevant data and robust risk models to produce information that allow markets to make good choices and decisions. Big data by definition seems to fit the bill. The sophistication of big data systems allows new forms of financial intermediation such as peer-to-peer (P2P) lending to bypass traditional models of lending which can be resource intensive and unresponsive.
But, despite the advances in sophistication of data analysis techniques, the jury is still out on whether more complicated risk models involving millions of data points means financial models have any better predictive-ability or leads to better financial decision making – for example, when lending to consumers, stockmarket trading and asset management. There is a danger that quantitative models even more become a substitute for judgment and experience. The financial crisis is a prime example of how delegating too much decision making to financial models can lead to serious market failure.
Similarly, delegating too much authority for trading decisions to algorithmic/ big data based trading strategies runs the risk of market dislocation and, in a more highly connected world, major systemic risks.
Investment behaviours: big data, if used well by insightful experts, has the potential to allow investment analysts and fund managers to spot emerging consumer trends and behaviours in advance of those being monetised into actual revenue and profits. This could provide first mover advantage, allow investment managers to spot undervalued companies, new or emerging product lines and create value for investors. This takes active fund management to a whole new level.
But, the evidence of history doesn’t give us much confidence that providing access to more complex data and information will make active fund managers any more efficient at asset allocation and stock selection – especially hedge funds who seem to be more gullible than most when jumping on the innovation bandwagon and being taken in by the next big idea. Instead, it is just as likely to lead to investment managers losing sight of the big picture, overtrading and, ultimately, higher costs for investors.
To conclude, I think it is safe to assume that big data will have a major impact on financial services and consumers. But, as we can see from this brief overview, it is not yet clear whether big data has the power to transform financial services for the better or is just another one of those ‘innovations’ which promised much but delivered little just making life more complex, difficult, and expensive along the way. If big data is used properly, in the right hands, it has much potential for much good. But, equally, there is a very real risk the outcomes could be greater exclusion, less efficiency, more unfair, unsafe, and less trustworthy financial services.
Who knows which way it will turn out? It is early days. But we do have an opportunity to shape this innovation now so it produces the right outcomes for consumers. This requires a robust public policy framework supported by a strengthening of the law and regulatory ground rules relating to the application of big data. Let’s not repeat the mistakes of the past by waiting until we see the emergence of major consumer detriment which will then require more costly regulatory interventions to clean up the market and protect consumers.