Visualizing large financial networks

DNB is the Dutch central bank and prudential supervisor. It is responsible for setting interest rates and safeguarding financial stability. It supervises banks, pension funds and insurance firms, amongst others. Data comes to DNB is many shapes. We receive written reports, regular structured financial reporting, and ad-hoc unstructured data. The challenge is to use these data optimally to aid DNB in guarding financial stability. In facing this challenge, a Data Science Hub has been set up to experiment with in a cutting edge analytical environment.

The information we receive is on a firm level, but we generally get the information for all firms. This means that DNB is in a unique position as we can see the financial network as a whole. It also means that the datasets are growing in volume. Many are now moving into billions of transactions. One way to analyze this data is to isolate a single market (or ‘layer’). Examples of statistical and visualization work are, respectively, Squartini et al (2010) and Heijmans et al (2016). Alternatively, the data can be seen as a multi-layer network where an institution’s activity in one market – interbank lending – is affected by its positions in another – for example derivatives market. Visualizing such large multilayer
networks is not trivial. We are especially interested in researching how visualization can help to understand econometric model outcomes (see, e.g. dieter.wang/contagionchain/).

We are looking for students interested in an analytical challenge in a policy relevant environment. You will be working in a multi-disciplinary team of analysts from Statistics and various business units. We offer internship expenses with travel. Working hours are flexible and remote access is possible. Since you will be working with sensitive data you will have to be screened by the Justice department

References

Heijmans, Ronald, Heuver, Richard, Levallois, Clement, & van Lelyveld, Iman, 2016. Dynamic visualization of large financial networks, Journal of Network Theory in Finance 2(2), 57–79
Squartini, Tiziano, Iman van Lelyveld, and Diego Garlaschelli, 2013, “Early-Warning Signals of Topological Collapse in Interbank Networks”, Scientific Reports, 3 (3357).