Does Order Simultaneity Affect the Data Mining Task in Financial Markets? Effect Analysis of Order Simultaneity Using Artificial Market

This study analyzed the effect of order simultaneity in financial markets on data mining tasks, using multi-agent simulations. In financial markets, multiple orders are submitted almost simultaneously or within very quick succession; such orders are thought of as independent of one another. We call this phenomenon order simultaneity. If order simultaneity increases, tick-time-level data mining methods are assumed to worsen because the randomness of the order sequences increases. The present study analyzed this effect using artificial market simulations, which enable experimentation in a fully-controlled environment. As a data mining task, we employed a Generative Adversarial Network (GAN) for the financial market to perform next order generation (prediction). We analyzed the impact of order simultaneity by applying the GAN to simulated data in artificial market simulations with various environmental parameters. We found that the effect of order simultaneity is limited for the next order generation task, which can be said to be the ultimate prediction task in financial markets. This analysis also supports the validity of the current approach of utilizing GANs to model order time series in financial markets. Moreover, our study demonstrates the utility of combining artificial market simulations and data mining.