Types of Dynamic Models Financial Market Simulation

30TH ANNIVERSARY ISSUE 2004 Large and detailed asynchronous (event-based) simulation models are widely used in the planning and analysis of systems such as manufacturing, logistics, and warfare. They have been used relatively little in financial analysis, especially as compared with continuous-time models. We believe the situation will change, starting early in the 21st century. Here we describe asynchronous simulation generally, and illustrate its capabilities in terms of a particular asynchronous and discrete-event stock market simulator, the JLM Market Simulator (JLM Sim). The JLM Sim is not a model of a market per se. Rather, it is a tool investors can use to create a model of the market using their own inputs. JLM Sim users can vary the number of securities, statisticians, portfolio analysts, investors, and traders in the simulated market and the decision rules they use. Prices are set endogenously, as new orders encounter already placed limit orders. At present, all JLM Sim investors are mean-variance investors; more advanced versions of JLM Sim later may accommodate additional investor types. Dynamic models may be described as synchronous, asynchronous, or continuous. Asynchronous simulation has some advantages over synchronous or continuous-time financial models.We summarize the salient features of the current JLM Sim and some desirable features that may be added to more advanced versions.