How to Regulate TechFins and Data-Based Finance

In a new research paper, we consider the impact of a group of new entrants into financial services and regulation. These new entrants include technology, e-commerce, social media, and telecommunications companies with often large pre-existing bases of non-financial services customers. These firms (loosely termed “TechFins”) may be characterized by their capacity to leverage data gathered in their primary businesses into financial services by the use of Big Data analytics, machine learning, and artificial intelligence. Initially they often act as conduits linking their customers to regulated financial firms.

China’s Alibaba with its subsidiary Ant Financial is the frontrunner, and its founder, Jack Ma, is often said to have coined the term TechFin. However, many tech firms have started to offer TechFin services in recent years, among them Amazon, Apple, Facebook, Google, Microsoft, and Uber (all in the U.S.), Samsung (in Korea), Tencent (in China), and Vodafone (in the UK, India, and Africa). All offer some form of payment, lending, or other financial service.

The advent of TechFins signals a shift from financial intermediary (FinTech) to data intermediary, and gives rise to important questions such as, how do these firms fit within the framework of financial regulation, and to what extent do their activities signal arbitrage opportunities and deficiencies in the current regulatory system?

Our paper outlines the challenges the TechFin business model poses to financial regulation. Given that serving a mere conduit function (such as e-commerce, search, or social media) does not meet the criteria for regulated financial activities, TechFins can assemble a large scale customer base and leverage the trust and control generated in this non-financial business in moving into financial services. Laws applying to specific financial services often do not apply to TechFins serving as conduits, mere data aggregators, data analysts, or other roles. Both these factors give tech giants a head start when competing for customers, as they are not burdened by financial regulatory requirements such as capital rules. Using Big Data and extraordinarily rich datasets, TechFins have the potential to disrupt and dominate significant portions of financial markets. Alibaba’s financial products are good examples. Its money market fund became the world’s fourth largest within only nine months and the world’s largest in less than four years. This rapid pace of change is a major challenge for financial regulation and regulators.

We argue that the rise of TechFin may be the single most important development in financial services since the 2008 global financial crisis for five reasons.

First, TechFins are a new type of market participant. They have their origin in tech or e-commerce environments with typically a multitude of clients (both consumers and small businesses) and a very deep pool of data. As TechFins reach a significant size, they also often build significant international networks and datasets. These data resources give them a significant advantage in the provision of financial services. TechFins may first enter the world of finance by providing their data, either raw or processed, to established financial services firms or FinTech startups, and then move to serving as conduits between their customers and regulated financial services firms. However, over time, many will probably begin providing financial services directly to their customers.

Second, TechFins may be able to provide far more efficient financial services. In particular, they may reduce transaction costs and improve decision-making by using or providing a more comprehensive dataset than that to which established financial intermediaries have access. We are already seeing such data – particularly payment data combined with social data – used as the basis for financial services decisions, particularly lending. This sort of credit provision based on cash flow and relational analysis is likely to result in far greater access to finance, particularly for small and medium-sized enterprises.

Third, established thresholds for the imposition of financial regulation, such as the solicitation of customers, deposit-taking, pooling of assets, or discretion over client assets, may not apply to TechFins. Accordingly, regulators may be unable to enforce customer protection measures and may face challenges in monitoring and mitigating systemic risk. Moreover, – use of data analytics combined with automation by TechFins may result in financial or other decisions being made on the basis of race or other discriminatory factors which correlate with income, ability to repay etc. Such discrimination may not be immediately obvious, particularly in the machine learning or artificial intelligence context.

Fourth, if financial regulation matters in furthering market efficiency and customer protection, TechFins should be subjected to it when offering financial services. Moreover, TechFins will have a competitive advantage over established licensed intermediaries if they are not constrained in their early stages by risk or compliance rules, and they do not bear minimum compliance and capital costs as regulated entities do. Such competitiveness concerns will likely result in regulatory arbitrage and new sources of shadow banking risks.

Fifth, in the world of TechFin, most customers give their data away for free, looking for some side service, so following the money (as traditional financial law does) is likely to fail. Following the data may provide an alternative, however. This alternative is not a mere policy choice; it is a necessity when the value of data exceeds the value of traditional production if measured by market valuation. In a world where data is the new currency, and where special legislation regulates intermediaries that manage financial assets owed to and owned by others (as banks and asset managers do), it is necessary to regulate data intermediaries as well as financial intermediaries, given that they pose similar risks to individuals and society.

We conclude that financial data gathering and analytics should be regulated if they exceed a certain size. One threshold could be coverage of a minimum percentage of people in a given market. Above this threshold, regulators should gather information and have access to data-based business models to ensure that analytical methods are sound and that financial decisions are not, unintentionally or intentionally, based on discriminatory factors. Additional restrictions should apply if there is a danger of systemic risk because, for example, TechFin data is essential for one significant financial institution, or the TechFin provides infrastructure (such as cloud computing services) essential to financial market functioning.

This post comes to us from Professor Dirk Zetzsche of the Faculty of Law, Economics, and Finance at the University of Luxembourg, Professor Ross Buckley of the Faculty of Law at UNSW Sydney, Professor Douglas W. Arner of the Faculty of Law at the University of Hong Kong, and Janos Barberis at SuperCharger FinTech Accelerator. The post is based on their recent paper, “From Fintech to Techfin: The Regulatory Challenges of Data-Driven Finance,” available here.