Disruptive Innovation by Foundational Research at the Interfaces of Economics, Computer Science and Mathematics

The problem which the technology of OiCOS is now able to approach originated in Viktor Winschel's PhD in economics started in 2000. The question was: "Which group from about 30, western, middle and eastern European countries is suited to form a monetary union?" The models were formulated in hundreds of nonlinear stochastic forward looking difference equations. To check the models and to derive a recommendation tens of thousand of macro- and microeconomic time series had to be processed by code written in mixtures of Fortran, C++ and Matlab and run on large beasts of parallel computers.

But the real problems arose in the aftermaths of the meetings with the  PhD advisor: He asked for seemingly simple adjustments of the questions to ask the data which repeatedly resulted in months of tedious, error prone and manual reorganisations of models, data and code.

Think about the complex models industry has to build when designing, running and reorganising virtual power plants or complex banking products.

How to automate, simplify and make the modelling and programming process cheaper, more reliably structured and organised?

The problem, now understood and at the core of OiCOS know-how, was the non-compositionality of economic models. To be clear, our industrial revolution in machines is the result of compositional methods in physics and mathematics and more and more understanding of compositionality for software architectures. But economics, business models and human organisations can not be designed so far by parts in teams of specialised experts and composed into an overall economic or business model, organigram, process flow diagram or economic model in a software on top of some physical simulations and control structures.

Modelling is another name for problem solving in teams but for that models have to be compositional, i.e. they have to be carefully formulated with their interfaces to their context - they need to be open. These possibilities will change the way your staff will engage in the development of your organisation.

Compositionality is the best method we know to cope with complexity. This is also the case for economic models.

And this is especially the case in an economy, where money helps being organised by the division of labour. But it is not the case in thinking about the economy, i.e. the economic modelling process itself is not compositional and organisable by a devision of labour since traditional economic models are not compositional. By that management and operations are desynchronised and the root of many organisational problems.

This was the research question of OiCOS in the last 15 years: "How to make economic modelling compositional and by that organisable by a division of labour?" We have solved it in about 2018 by compositional open games. Since then software tools are under way to also make it applicable in industry. There, since about 2018, when OiCOS became operational, we also had to understand how to communicate what we have discovered to those who may need it. It took a while and of course, if we knew what we were doing it would not be called research.

What is the gain from compositionality and division of labour in engineering? 

Think about the process of designing a car. The engineers of the gearbox, motor and axle deliver their models and in some holo cave (as with tractors at John Deer) the car is virtually assembled and simulated. Then for example a repair team can check whether for repairing the axle the motor needs to be removed - bad - or the drivers can check whether all knobs are reachable during a simulated drive. These interactions of the parts in the whole is mission critical during the design of the car and interaction analysis - aka synthesis or composition - is a major technology to manage the risk of an unsuited model of parts. But synthesis and simulation is also critical for the toolmakers who, in parallel to the design of the car, design and build the toolchain to build the car. Synchronisation of these two processes - building the car and building the tools to build the car - was a major innovation step in the manufacturing industry. Now we aim to synchronise similarly the object and the meta level - operations and management where economic models are at stake.

Now imagine similar power tools for business and economic modelling or management and compare it to the current technology of Power Point slides. How about taking the organigram to the John Deer holo cave? Live - on top of your ERP system's data? You get a glimpse of what this may mean for your products and the organisation of your company, value chain, sector or public administration.

What is the gain from compositionality and division of labour in business, economic and organisational modelling?

Modelling and organisation can be done in specialised teams, preferably by those who are organised, sub-models compose to wholes and are testable and replaceable. Libraries of building blocks of organisations emerge and can be reused. Mass customisation at the level of business and organisational models is possible. Scalability, maintainability, adaptivity and quality of your organisation and product improve. You may simulate different business models and check how they perform in different contexts like macroeconomic scenarios in the areas of your international activities. You can simulate before you actually reorganise.

The organisation by a division of labour is synchronised with the division of labour when thinking about and modelling the organisation.

Synchronised operational and management processes. This is the promise of our technology. In biological living beings the synchronisation of different levels of the body like tissues and cells is a prerequisite of life and consciousness and a lack of synchronisation brings about ageing and ultimately death. In this sense our technology is far from being the nightmare of a mechanistic approach to life but quite the opposite. It is the dream of learning the wonders of nature and life and make them available in the way we organise our common lives.

Research Partners

Prof. Dr. Florian Heiss

Chair for Statistics and Econometrics

Prof. Dr. Philipp Zahn

  • Mechanism design (good rules for games) and game theory

  • Behavioural economics (beyond mechanistic rationality)

  • Experimental design (to understand human behaviour)

Philipp's personal Website

Dr. Evguenia Sprits

Prof. Dr. Neil Ghani

  • Mathematics

  • Programming language design

  • Open games

Neil's personal Website

MSP Group (Mathematically Structured Programming)

Dr. Jules Hedges

  • Open Games

  • Mathematics

  • Computer science

Jules' personal web blog

Dr. Renée Menéndez

Renee's personal web blog


Uncertainty Quantification and Global Sensitivity Analysis for Economic Models

(2019), Quantitative Economics, (preprint), Bruno Sudret, Stefano Marelli, Daniel Harenberg, Viktor Winschel

Compositional Game Theory (2018), Logic in Computer Science, (preprint) Neil Ghani, Jules Hedges, Viktor Winschel, Philipp Zahn

Higher-Order Decision Theory (2017) Algorithmic Decision Theory, (preprint), Evguenia Winschel, Philipp Zahn, Jules Hedges, Paulo Oliva

Selection Equilibria in Higher Order Games (2016), Practical Aspects of Declarative Languages, (preprint),  Philipp Zahn, Jules Hedges, Paulo Oliva, Viktor Winschel, Evguenia Sprit

Coalgebraic Analysis of Subgame-perfect Equilibria in Infinite Games without Discounting (2015), Mathematical Structures in Computer Science, Samson Abramsky, Viktor Winschel

Solving, Estimating and Selecting Nonlinear Dynamic Models without the Curse of Dimensionality (2010) Econometrica, Markus Krätzig, Viktor Winschel

Likelihood Approximation by Numerical Integration on Sparse Grids (2008) Journal of Econometrics, Florian Heiss, Viktor Winschel

Public Deficits and Borrowing Costs: The Missing Half of the Market Discipline

(2001), Journal of Public Finance and Public Choice, Friedrich Heinemann, Viktor Winschel

A Coalgebraic Semantics of Compositional Games in Economics

(2013), arXiV, Achim Blumensath, Viktor Winschel

The Empirical Analysis of Exchange Rate Regimes and Nonlinear Structural Econometrics (2005), PhD Thesis University of Mannheim, Viktor Winschel


Dusko Pavlovic, Brendan Fong, Bob Coecke, Samson Abramsky, David Spivak, Alexander Kurz, Jocelyn Ireson-Paine, John Baez and many more at Oxford, MIT, Strathclyde Glasgow, Amsterdam, Nijmegen, Hawaii and other universities and research institutions.


renée-menéndez-foto.256x256 copy.jpg

Academic Network