About us



ACEFINMOD stands for Agent-based Computational Economics and Financial Modelling.

The original website was set up in 2007 by the Essex Group (Professor Sheri Markose, Simone Giansante, Matuesz Gatkowski and Ali Rais Shaghaghi), during the course of the COMISEF EC Marie Curie Research Training Network (2007-2013). We gratefully acknowledge funding from this EC project that started this research program initially at the Centre for Computational Finance and Economic Agents (CCFEA).  Sheri was the founder director of CCFEA from 2002-2009 and Giansante et. al. were PhD students there till

This marks a 2018 update of the website that aims to showcase developments we have made in non-traditional computational methods for economics and financial modelling.  The main add on is that this website is interactive and we hope it will stimulate some interesting interactions.

Why Do We Need New Models of the Economy ?


Many of the macro and micro economic models are too generic, fail to include relevant institutional details and often display a flawed scientific basis.

Four mega-trends are driving economic systems that have little purchase in the standard models

  • Globalization and off shoring of supply chains have made indigenous production networks in many advanced economies, especially the US and UK, fragile. This loss of connectivity has threatened the functionality of the production networks.
  • Financialization with the growth of the financial sector arrogating a larger and larger share of gross operating profits has reduced real investment in many OECD countries.  UK banks, for instance, lend less than 15% of their total lending to private non-financial corporations, though textbooks have you believe that the main function of banks is to lend to businesses.
  • Cashlessness starting with Electronic Fund Transfers at Point of Sale with card payments has reduced transactions demand for high powered money (M0: notes and coins of the state) has not been integrated into macro-economic models; This may have accelerated the fall of inflation in advanced cashless economies and the fact that the Philips curve is horizontal.
  • The 4th industrial revolution and AI; This is in addition to the digitisation of trading systems from stock markets to retail. The ambition of the 4th industrial revolution and super-AI exceeds machine learning associated with pattern recognition, voice recognition. The important paradigm shift with Information Communication Technology (ICT) based digital systems relates to the role of real time data feeds to assist in decision making in economic models for the 21st century. ICT based digital products and the blockchain technology, in particular, involves computational modelling of real time remote validation by direct data access of key distributed variables as a means of vitiating ambiguity, risk, asymmetric information and above all reliance on econometric/statistical data mining for determining many key decision support variables. Personally, my first experience of this came with my 2003 research (Markose and Loke, 2003) on electronic fund transfer at point of sale (EFTPOS) technologies. This allows deposit balances to be remotely verified and in due course for debit/credit cards to break the monopoly of state supplied cash. The latter can be regarded to be the analog variant of verifying quid pro in payments for goods and services. In the pre www world, we had to rely on bank guarantee cards and other reputational devices to replace cash in payments at physical points of sale. As with Google maps that can zoom in at many levels of data granularity (see, Markose, 2013), it is my opinion that many agent characteristics in ACE models can be provided by direct data feeds, minimizing the need for modelling assumptions. These are a far cry from equation and econometrics based models. This new aspect of digital markets and products that relates to direct access to data should feature more in ACE models.

Of these agent-based computational economics (ACE) is at the forefront of addressing major problems of mainstream economics and finance. Multi-agent models include agents that are both inanimate (eg. repositories of data bases) as well as behavioural agents capable of varying degrees of computational intelligence from fixed rules to fully adaptive agents representing real world entities (such as banks or consumers). These agents operate in artificial environments that depict real time orientation and also complex interactions. By abandoning homogenous and rational agents paradigm, the ABM research can finally build a bridge between real world complexity and its simulated depiction in a manner in which it is fruitful both for the advance of science and for practitoners.

Increasing computational power and quality of data sources can be herald a new era for finance and economics.



Our aim is to provide community with our contributions in the field of Agent-based modelling with special focus on large scale data based driven models of the financial sector from a network perspective, design and backtesting usage of algo trades in fully rebuilt order book and other complex economics system platforms. We will provide in depth tutorial of the methods and algorithms of the methods that we deploy.



If you have any news and important information from the Agent-Based world, please inform us!

You can be not only a witness of new era, but also become a part of its rise.

If you know any interesting, worth sharing publication, conference, book or paper in the field. Let us know.



For any further information, please contact scher@essex.ac.uk