[Next] [Previous] [Top] [Contents]

Modelling the Process of Market Emergence

4 The simulation set-up - general


In order to capture the main, essential elements of the foregoing account of the position of Russian enterprises, we assume the decision variables of the enterprises to be planned output, output price, the wage rate, the offer of employment, orders placed with suppliers and payments to creditors. In addition, the enterprises choose the suppliers with whom they place their orders and the customers whose orders they will fill in whole or in part. Each enterprise notes at each date whether its suppliers have filled the orders placed with them and it customers have paid for the goods previously supplied to them. These notes take the form of endorsements attached to the enterprises records of its customers and suppliers. Orders are allocated among known suppliers in proportion to their records of reliability. Sales are allocated first to orders from known customers with the best records of payment. In effect, each enterprise builds up models of the enterprises with which it trades.

These endorsements are also used by enterprises for formulate views about which other firms are the most successful. It is natural to assume that those suppliers which are best at supplying orders and those customers which pay most quickly are also the strongest enterprises. On this basis, enterprises take into account any observable information they have about these trading partners and assume that their behaviour is highly functional. In the model reported here, the only information which one agent can observe about an enterprise is the price it sets for its output and information arising from the transactions in which they engage (supplies, orders and payments). Thus, if one enterprise observes its best trading partners lowering (or raising) prices, it will assume that to lower (or raise) prices increases the values of goal variables and, so, will conjecture a model to that effect.

The goals of the enterprises are sales and cash. There is no attempt at optimisation of these values but, rather, the agents seek strategies which will increase the value of one or the other of these goals. In the present set-up, neither is given pride of place. In the event that changing the value of one decision variable is expected to increase the value of one goal value and diminish the other, then the action which is considered most likely to have the anticipated outcome will dominate the decision. If the agent has more confidence that the goal value diminution will occur then he will change the decision variable value to reduce or prevent the diminution. If he has more confidence that the other goal value will be increased, then he will change the value of the decision variable in the appropriate direction.

The pre-defined intermediate variables observed by the enterprises are their own purchases, their own sales, their current stocks of real goods (inputs and unsold outputs), their current financial asset holdings (only cash, so far), and the wage bill (the product of the wage rate paid and employment by the enterprise).

The wage bill is the only intermediate variable to be calculated from other variables observed by the enterprise. It gets this special treatment because we assume that it is paid in the same period as the employment it covers. This assumption itself seemed appropriate because of the relatively insignificant incidence of delayed wage payments and also because, in inflationary conditions, the impact of delayed wage payments is much the same as offering a lower wage rate. A rather more elaborate setup would be required directly to capture the effects of wage arrears. If such a model were thought likely to be useful, it would be a straightforward extension of the model reported here.

In general terms, the set-up reported here was devised only to capture a coarse-grained account of the development of Russian enterprises and to demonstrate how our modelling techniques perform on problems relating to the emergence of new markets and market institutions.


Modelling the Process of Market Emergence - 17 MAY 96
[Next] [Previous] [Top] [Contents]

Generated with CERN WebMaker