Modelling Financial Derivatives Markets in a Firm Based Evolutionary Macro Model (MOSES)-On the market integration of computing, communications, and financial services

Abstract

The evolutionary firm based model MOSES of an Experimentally Organized Economy (EOE) is introduced and then outfitted with increasingly more sophisticated financial markets that take us from a stylized industrial banking system (standard in the model), to a (capital) market oriented monetary system, and then on to a securitized financing system where options and market derivatives play a role in allocating financial resources. We study the consequences for long term economic growth of introducing different financial market regimes.

The markets of the financial system and those of the real economy are integrated within the business plans of model firms. The immense micro to macro complexity thereby created reduces the transparency of the economic environment of all actors in the model, including Government as a policy maker, to a fraction of what is needed for informed decisions. The policy maker thereby loses its presumed information advantage of economy wide overview and becomes one bounded rational actor among all other ignorant actors. Because of the strong leverage on the entire economy of its actions the competence and information demands on Government, when conducting policy, however, becomes far more exacting than what is required of any other actor.

When Government engages in ambitious policy making its interventions in markets affect the behavior of all other actors. The entire market regime is therefore affected to the extent that the Central Government often cannot possibly be in control of what it is doing, especially when it comes to the long run consequences of its actions, and therefore often would do better by taking it easy (Eliasson and Taymaz, 1992). We find that financial markets matter for long term economic development and that seemingly insignificant respecifications of the financial market regime can cause considerable differences in long term GNP growth outcomes, amounting to as much as a difference of 1.5 percentage points per annum and a per capita GNP, that is up to 2.5 times larger on the 60 year horizon than in the worst scenario.

The capital market regime is found to outperform the industrial banking system. When introducing a crudely modeled securitized system (call options), the only one our computer system could handle, financial market transactions were, however, speeded up to the extent that resource allocations worsened. Mixing the capital market regime with the simple securitized system generally lowered long run economic growth. Adding options not only increased the speed of arbitrage, destabilizing fundamentals, but also made expectations unreliable and raised the exit rate, notably by making manufacturing firms lose money on their financial investment portfolios. The results on options cannot however be generalized beyond the very specific experimental designs implemented, but they indicate that speeded up market arbitrage, which is what financial derivatives are all about, has mixed economy wide consequences, that may be long term negative in an economy that is not otherwise diversified and robust. While the simulation analysis of the markets for derivates in the MOSES model is still too primitive to allow any empirical generalizations, the results so far should be taken seriously as indicative of possible surprise dynamics that makes it even more important to conduct more inquiries of this kind.

A final methodological note. The excessive demands on computer capacity of the options experiments have prevented us from testing the results for robustness (as for instance in Eliasson and Taymaz, 2000) by running a whole spectrum of experiments with small changes each time in the four different blocs where stochastic specifications are used. The non-linear specifications of the model often generate large differences in very long run outcomes when the stochastic seeds are changed marginally. Hence, the results to be reported should be regarded as explorative, suggesting new credible hypotheses, rather than empirically calibrated results, and also telling what kind of future analyses the new expanded model is capable of.