A System Dynamics Model of Stock Price Movements

Abstract

Stock prices are known to exhibit strongly non-linear behaviours, e.g. apparent random volatility, bullish or bearish trends, crashes etc. Using finite difference equations it is easy to generate pseudo-random behaviour patterns. The hope is that richer patterns can be generated in the continuous case using System Dynamics (SD). In the paper we test this possibility. Three basic behavioural attitudes are described by introducing three corresponding families of investors. α-investors are rational fundamentalists striving to stabilise the stock price toward a goal value. Short-term βS-investors are traders destabilising the market as they follow immediate movements and fads. Long-term βL-investors use arbitrage by comparing returns on stock and on risk-free assets. Although they are partly rational they provoke important departures from the fundamentals in an expanding market. It is shown that the behaviour of βL-investors is an important explicative factor of instability on the stock market.