Tore Opsahl

Tore Opsahl

tore@opsahl.co.uk   LinkedIn  
Head of the Quantitative Solutions Advisory group within Nomura Securities’ Global Markets division. The group delivers quantitative strategies and bespoke advisory to institutional clients and internal trading desks across portfolio construction, execution, risk management, and systematic investing. Additionally, I serve as an adjunct professor at NYU, where I have taught Advanced Machine Learning in Finance since 2017.

Previously, I was a Senior Trading Strategist within Global Markets at Bank of America (BofA Securities), where I led the Systematic Investment Group that helped institutional clients build and execute systematically-aware portfolios. Before BofA, I founded a fintech company that combined real-time data from the Internet’s DNS backbone with statistical models to forecast website traffic and retail sales. Earlier roles include Chief Scientist at DeepMile Networks, a government contractor in the DC area focused on big data analytics, and Research Associate in the Innovation and Entrepreneurship group at Imperial College London’s business school.

My published work centers on methodological advances in network science: the study of relational data such as collaboration patterns, neural connections, and online communication. This research traces back to 2003, when I founded ANTfriender, an early online community at the University of California, Irvine. While the venture itself didn’t succeed, it generated a rich dataset that informed my Ph.D. thesis, “The Structure and Evolution of Weighted Networks” and a series of peer-reviewed articles. Much of this work explores richer network structures (e.g., weighted, two-mode, and time-stamped) and their potential for uncovering new organizing principles. Further details are available on the Network Science pages.

I have also applied this research to problems in financial industry. One example is quantifying systemic risk in the derivatives market, where post-crisis regulation mandating central clearing of standardizable contracts has fundamentally reshaped the market’s structure. In fact, it shifted it from a core-periphery network to a star network. Related work has examined concentration and contagion effects in stress-testing and loss forecasting, spill-over effects across portfolios, and the impact of supplier and buyer concentration on default risk and credit default swap (CDS) spreads.

To make my methodological contributions accessible, I developed tnet, an open-source package implemented in R that provides functions for computing the network measures I have proposed alongside related measures from the literature.

One of my first academic presentations at Applications of Social Network Analysis 2006 in Zurich. I presented the Valued Clustering Coefficient, which was later published in Social Networks as Clustering in Weighted Networks.