Centre for Digital Finance (CDF)

Digital Finance

The financial markets are the largest and most important electronic marketplaces in the world today, by far. Digital Finance is the application of cutting-edge computer technologies to financial problems and products. These technologies include: Blockchain, Cryptocurrencies, Artificial Intelligence / Machine Learning, Agent Based Simulation Models, Natural Language Processing, Big Data and Post-time-series Statistical Analysis. Digital Finance draws together researchers from different academic backgrounds for collaborative research and innovation for the purpose of gaining deeper insights into financial system dynamics. We also work with industry partners to develop solutions for contemporary business problems.

Transformative changes in financial technology range from the emergence of an alternative virtual money (cryptocurrencies) and the direct, internet-mediated access to investors' capital (crowdfunding), to the explosion of new data types (e.g., social media, text) and new analytical techniques (Artificial Intelligence, Machine Learning, Natural Language Processing). These changes trigger shifts in industry practice and they can both empower and subjugate consumers. Historically, the global investment banks and fund managers were siloed organisations who jealously guarded the intellectual property in their systems and models, as they considered these their competitive edge. In recent years, they have all come to embrace the Financial Open Source (FinOS) movement and now openly share their cherished in-house data management systems and pricing algorithms with each other. The reason for this shift in behaviour is that they understand they cannot out-compete the emerging wave of FinTech technology and so they are adapting to it in order to position themselves to prosper in a different financial world. The philosophy underlying cryptocurrencies aims to engage and empower communities at the expense of disintermediating corporate entities. However, scammers, programming errors and poor operational practices have exposes neophyte cryptocurrency investors to losses. At the same time, corporations gain greater influence over their customers via richer, personalised tracking data and sophisticated classification and forecasting methods. Regulators are wrestling with this new landscape with a view to protecting consumers.

Bitcoin cryptocurrency is one type of financial instrument that is built on the Blockchain. Another fast-growing Blockchain technology is Decentralised Finance (DeFi). This automates financial functions (distributed applications or "dApps") that were historically the preserve of banks and other financial institutions. These new financial instruments pose a new educational need: the engineering and programming skills, combined with the understanding of finance, required to build the financial products and services of the future. This in turn presents a new research need: the capacity to test the efficacy of new dApps, to examine their impact on financial system dynamics and to provide a conceptual framework to help policymakers to monitor and regulate these new systems.

As an empirical approach, Digital Finance has the same objective as traditional time series econometric analysis but use different tools. Namely, it aims to test theory from finance and economics. Digital Finance's empirical tools have roots in computer science. In some settings, Digital Finance complements the econometric approach. In other settings, it can provide an alternative view and can often offer new insights, e.g. via simulation, where the historical data required by the econometric approach is lacking. Digital Finance draws in finance researchers from a broad set of background disciplines and perspectives, most notably computer science but also from engineering, physics and mathematics. This rich diversity of backgrounds and philosophical outlooks is reflected in the wider array of academic journals in which Digital Finance research is published. While much Digital Finance research finds a home in finance and economics journals, a good deal of it is published in Information Systems or Operations Research journals, which are receptive to research on the effects of technology on business and finance.

A relatively simple technological advance, hardware improvements that speed up order routing, led to some high frequency traders gaining an advantage over rivals because they could place orders milliseconds earlier. This spawned the industry of high frequency trading. To study the latter requires working with data of such high resolution (tick data) that it is not evenly spaced in time and the time in milliseconds at which assets trade do not line up with each other. Traditional mainstream financial time series econometrics tools become useless in this setting. Digital Finance provides a set of new, post-time-series, statistical tools for analysing irregular high frequency data.

New technology also offers finance researchers an alternative tool for testing theory: a simulation sandbox called agent-based modelling. On the one hand, this is an alternative tool to empirical analysis of history for validating finance theory and which can yield new insights into financial dynamics. On the other hand, this tool has the additional power of being able to simulate the likely impact of new policies and regulations. Agent based modelling involves coding up automated trading behaviour (the "agents"), which will be triggered when selected by a simulation programme. Large scale simulations are run in which different agents (programmes) are randomly selected, paired and made to trade. The resulting prices are recorded and used as input for the subsequent trade. And this step is repeated. These simulations produce price paths, along with volume and volatility patterns. Tens of thousands of trades can be easily simulated. Different sets of simulations can be run with different prevailing conditions and the variety of emergent behaviours can be observed for each. An important feature of this approach is that it enables endogenous behaviour to emerge within a complex adaptive financial system. This contrasts with traditional financial theory and empirical approach which is inherently deterministic and does not allow for endogenous outcomes.

Centre for Digital Finance (CDF)

The CDF was established in 2014 and has built up an enviable track record of publications in top international journals and research grants on topics linking finance with technology. In addition, our members regularly disseminate their research via seminars, workshops and international conferences. We actively cultivate collaborations with other leading international research centres and institutions, as well as with industry partners.

Our Mission

• Research Leadership: To be a global leader in curating knowledge and creating new knowledge and understanding about the co-evolution of finance and the emergent technologies which are reshaping finance.
• Research-led Education and Dissemination: To equip the financier of tomorrow with the skills they will need to navigate and master a more technological world, to inform and guide policymakers, and to contribute to academic debates in order to advance communal understanding of Digital Finance.
• Research-led Enterprise: To collaborate with industry to solve real-world business problems via sponsored research and consulting, and to contribute to a better society by participation in the formation of policies, regulations and standards by serving on boards and writing policy/industry reports.
• Research Funding: To generate revenue to fund Digital Finance research via academic research grants and commercially sponsored research projects.
• Profile Promotion: To generally promote and enhance the profile and reputation of the Centre for Digital Finance by engaging with external parties, serving on external boards and committees, presenting at conferences (industry and academic), participating in news media, writing industry reports, promoting our research on social media, etc.