A1272
Title: A Bayesian GARCH EWMA control chart for monitoring Bitcoin
Authors: Majika Jean-Claude Malela - University of Pretoria (South Africa) [presenting]
Abstract: Statistical process control (SPC) techniques have been applied across various fields and industries, including medicine, education, and finance. SPC is useful for monitoring the stability of a process by identifying abnormalities and investigating the sources of variability. However, traditional control charts struggle to effectively capture the dynamic nature of financial data, as such data is often characterised by volatility and autocorrelation. This situation violates the assumption of independent and identically distributed observations. Therefore, it is essential to identify an appropriate time-series model for statistical control, often explored through the Box-Jenkins approach. A new Bayesian exponentially weighted moving average (EWMA) control chart based on a generalized autoregressive conditional heteroscedasticity (GARCH) model is introduced. The newly developed control chart along with the three other GARCH EWMA-type control charts are used in a case study to monitor Bitcoin prices. The results show that the properties of the new charts are attractive and more effective in detecting abrupt shifts in the financial time-series data.
