Stock markets as evolving complex systems : Simulations and statistical inferences
For a long time Finance Theory hold on to the idea of efficient markets that convert every piece of new information into the price of an asset, so as to reflect the true value. The upcoming of empirical facts that contradict this view gave rise to new theoretical models. One of the most important contributions was the so called Behavioural Finance Theory that describes people and traders in particular on a more psychological basis. Based on this approach, other concurring ideas to the efficient market theory developed in the following years. A relatively new branch in finance theory that borrows its main ideas from natural siences is known under the name of econophysics. This approach both comprises the statistical features of financial time series as well as simulations that try to picture real asset markets as complex systems. These systems are characterised by many different traders that interact and influence each other and so create time series that are encountered in many other fields like earthquakes, mass extinction or solar flares. This work applies the theory of complex systems in order to understand the mechanics of real financial markets in particular the stock markets. First it comprises the existing literature on econophysics. Then it provides new statistical work that confirms the former result where the main characteristics of financial time series turned out to be the non-normality of the distribution of price changes, the multifractality and the volatility clustering. This is followed by two new simulation-models. The first is an Ising-model where neighbour influence plays the crucial part. The second is a more economically based simulation where the traders have explicit strategies after which they decide how to act. As it turns out, both models are able to produce time series that possess all the characteristics of real time series.