Sumit Sourabh

   Mailing address

Universiteit van Amsterdam
P.O. Box 94323
1090 GH Amsterdam

Email:   s[dot]sourabh[at]uva[dot]nl

Visiting address (Fridays)

Room C3.136b
Science Park 904
1098 XG Amsterdam

About me

I am currently working as a front office Quantitative Analyst within Financial Markets at ING Bank in Amsterdam. At ING, I focus on pricing and risk management of financial derivatives using mathematical modelling and data-driven techniques.

I have a joint position as a Research Scientist at Computational Science Lab within Informatics Institute, University of Amsterdam. At UvA, the primary focus of my research is network-based techniques in finance for trading and credit risk management. I am associated with H2020 EU BigData Finance project: Machine Learning for Trading and Risk Management.

I did my PhD in Mathematics at Institute for Logic, Language and Computation, University of Amsterdam, and Integrated Masters in Mathematics and Scientific Computing from Indian Institute of Technology (IIT) Kanpur, India.

Research Themes

Systemic Risk and Financial Complexity

Derivatives markets play a crucial role in the interconnectedness of financial systems, and currently a large volume (with outstanding notional amounts in trillions of USD) of derivatives are actively being traded between financial institutions, corporates and individual investors. Post the financial crisis of 2007-08, regulators have introduced a number of valuation adjustments (xVA) to cover credit, funding and capital cost which financial institutions face in OTC derivatives transactions. In context of xVA, wrong-way risk (WWR), which is caused by an adverse correlation between exposure and credit, can have significant consequences for risk management. Our objective is to use data-driven network-based methodologies for xVA modelling, focusing particularly on WWR.

PyData presentation on Learning Credit Networks

Data-driven methods for the audit

I am co-supervising Marcel Boersma who is working on his PhD project funded by KPMG on this theme. The main objective of this research is to develop a framework that uses and combines various advanced analytical methods for the purpose of the audit. This framework can be used to understand the structures of the entity and confirm our understanding of the entity being audited. Due to the increased operational complexity of large entities we propose to use advanced methods from the complex systems to build a mathematical representation of the entity. The most important contribution is to develop novel techniques using state-of-the-art computational methods to analyse available data of the entity to obtain audit evidence.


    Quantitative Finance, Network Theory and Machine Learning

  • Reducing the Complexity of Financial Networks using Network Embeddings
    M. Boersma, A. Maliutin, S. Sourabh, L.A. Hoogduin and D. Kandhai,
    Submitted, 2020.

  • Wrong-way risk in Credit using Bayesian Networks,
    S. Sourabh, M. Hofer and D. Kandhai,
    Risk, March, 2020.

  • Contagious defaults in a credit portfolio: A Bayesian network approach,
    I. Anagnostou, J. Rivero, S. Sourabh and D. Kandhai,
    Journal of Credit Risk, Vol. 16, No. 1, March, 2020.

  • Financial statement networks: an application of network theory in the audit,
    M. Boersma, S.Sourabh, L. Hoogduin and D. Kandhai,
    Journal of Network Theory in Finance, Vol. 4, No. 4, 2018. [ PDF ]

  • Incorporating Contagion in Portfolio Credit Risk Models Using Network Theory.
    I. Anagnostou, S. Sourabh and D. Kandhai,
    Complexity , Article ID 6076173, 15 pages, 2018. [ PDF ]

  • Liquidity risk in derivatives valuation: an improved credit proxy method,
    S. Sourabh, M. Hofer and D. Kandhai,
    Quantitative Finance 18 (3), 467-481, 2017. [ PDF ]

    Mathematical Logic

  • Bisimulations for coalgebras on Stone spaces
    (with Sebastian Enqvist)
    Journal of Logic and Computation, doi:10.1093/logcom/exy001 2018. [PDF]

  • Algorithmic correspondence for Intuitionistic modal mu-caculus
    (with Willem Conradie, Yves Fomatati and Alessandra Palmigiano)
    Theoretical Computer Science, Volume 564, pp. 30-62, 2015. [ PDF ]

  • Sahlqvist preservation for flat topological fixed-point logic
    (with Nick Bezhanishvili)
    Journal of Logic and Computation, doi:10.1093/logcom/exv010 [ PDF ]

  • Jònsson-style canonicity for ALBA-inequalities
    (with Alessandra Palmigiano and Zhiguang Zhao)
    Journal of Logic and Computation, doi:10.1093/logcom/exv041 [ PDF ]

  • Generalized Sahlqvist theory for impossible worlds
    (with Alessandra Palmigiano and Zhiguang Zhao)
    Journal of Logic and Computation, March 2016, doi:10.1093/logcom/exw014. [ PDF ]

  • Algebraic modal correspondence: Sahlqvist and beyond
    (with Willem Conradie and Alessandra Palmigiano)
    Journal of Logical and Algebraic Methods in Programming, doi:10.1016/j.jlamp.2016.10.006 [ PDF ]

  • Subordinations, closed relations and compact Hausdorff spaces
    (with Guram Bezhanishvili, Nick Bezhanishvili and Yde Venema)
    Applied Categorical Structures, doi:10.1007/s10485-016-9434-2 [ PDF ]

  • An application of ALBA: relativised canonicity via pseudo-correspondence
    (with Willem Conradie, Alessandra Palmigiano and Zhiguang Zhao)

PhD Thesis

      Correspondence and canonicity in non-classical logic

In this thesis we study correspondence and canonicity for non-classical logic using algebraic and order-topological methods. Correspondence theory is aimed at answering the question of how precisely modal, first-order, second-order languages interact and overlap in their shared semantic environment. The line of research in correspondence theory which concerns the present thesis is Sahlqvist correspondence theory --- which was originally developed for classical modal logic, and provides a systematic translation between classical modal logic and first-order logic.


Together with Prof. Drona Kandhai, I am partially responsible for the lectures, lab assignments and supervision of individual projects in the following courses in Graduate School of Informatics at University of Amsterdam.

  • Computational Finance
    Computational finance, generally referring to the application of computational techniques to finance, has become an integral part of modelling, analysis, and decision-making in the financial industry. In this course an introduction will be given to the theory of derivative pricing. Several computational approaches such as Monte Carlo methods, lattice methods, numerical PDE (Partial Differential Equation) and network theory based techniques will be covered. The application of these algorithms on distributed computing architectures will be outlined.
  • Advanced topics in Computational Finance
    In this project based course students will further develop on the knowledge gained in the computational finance course by implementing and analysing the behaviour (such as accuracy, stability or performance) of an advanced computational method.

Funding Sources

We gratefully acknowledge the funding from the sources below which support our research activities.