1st Workshop on Approaches to Modelling Heterogeneous Interacting Systems

In Association with Financial Cryptography '22
May 6, 2022 (note date change!)

Radisson Grenada Beach Resort


The program of talks and discussions is available here.

The workshop will be hybrid; if you'd like to join remotely, please send an email.

About the workshop

Modelling plays an important role in how individuals, organizations, and governments understand, reason about, and make decisions about systems and their security. Economic, game-theoretic, logical, graphical, statistical, and systems models --- and more --- are all used to understand the behaviour of systems, estimate risk, evaluate system designs, and plan policies. Modern systems are complex in themselves and integrated into systems of systems. Systems no longer exist in isolation, but interact with each other, forming ecosystems --- and new approaches are needed to model such systems.

Being able to model and reason about ecosystems is becoming increasingly important. Consider financial systems. The Basel III proposals have placed a strong emphasis on providing operational risk factors for financial institutions. Furthermore, the MiFID II/MiFIR places a strong emphasis on tracking intention in financial instruments. The majority of mechanisms and tracking approaches rely on secure information ecosystems formed by overlapping architectures of varying vintages that map sets of systems together. The security of such systems requires quite sophisticated modelling to combine both the security and economic components. Modelling requires that information is securely transmitted and then to conduct valid computations and transformations, often as a function of a number of states of nature.

Movement towards combinations of centralized and decentralized financial information networks substantially complicates the modelling requirements for the security of such systems. The integrity of both the underpinning infrastructure and the economic integrity (for instance margin monitoring) are now correlated, particularly under `tail events' or edge cases. In some cases these events are by happenstance, but often there is a strategic component that has been designed to obtain beneficial positions, often at the expense of the sustainability of the markets facilitated by the information systems. Examples of this kind of issues range from Ponzi schemes in Ethereum DAOs to sudden shocks to auction rate securities and sudden liquidity dry-ups in CDOs and swap markets.

The objective of this session will be to explore how models can be used to understand the stability and security of these complex ecosystems. A particular interest will be focused on the complexity of modelling the interaction of decentralized and centralized information systems (such as brokerage networks trading in bitcoin). At present there is almost no significant research focused on providing guidance to financial regulators/policy makers and large institutions on how to understand risks from integrating these systems or how to record and report anomalies.

The workshop program will consist of talks and discussions led by the program committee members.

We welcome additional contributions in the form of talks, work in progress, or discussion sessions around the following areas:

  • modelling decentralized systems, particularly financial systems,
  • decision-making and edge-case testing financial infrastructure that combines DeFi and centralized platforms (notably brokerage, margining validation, order flows, double auctions, and associated order books),
  • challenges and approaches to ecosystem modelling,
  • the role of interfaces and composition in modelling,
  • modelling the integration of new components into existing systems,
  • using models to understand risk, resilience, and sustainability of interacting systems.
  • model checking for ecosystem models.

Program Committee

Tristan Caulfield University College London
Patrik Eklund
Maribel Fernandez
Xian Gu
Christos Ioannidis
Timo Lang
David Pym
Jonathan Spring
Will Venters
Julian Williams
UmeƄ University
King's College London
Durham University
Aston University
University College London
University College London
CERT / Carnegie Mellon University
London School of Economics
Durham University