Statistical process control in asset management

Abstract

Statistical process control (SPC) suggests tools for the on-line detection of changes in the parameters of the process of interest. For the purpose of the data analysis the observations are divided into two parts, historical and online observations. The historical observations are used to make statements about the distributional properties of the process. Such results are necessary for the calculation of the design of control charts, which are the most important monitoring instruments in SPC. Every newly incoming information is immediately exploited. The new observations are analyzed on-line. It is examined at each time point whether the process parameters identified from the historical observations remain unchanged. The control chart gives a signal if the process parameters have changed in a statistically significant way. A decision maker should carefully analyze possible causes and consequences of any signal. The rest of the chapter is organized as follows. Section 20.2 provides a brief review of the terminology, instruments, and procedures of SPC. Section 20.3 discusses the application of SPC in asset management for passive as well as for active portfolio investors. In particular, Section 20.3.1 deals with monitoring a fund manager's performance, while in Section 20.3.2 the monitoring the GMVP weights is discussed. Section 20.4 concludes. © 2008 Springer-Verlag Berlin Heidelberg.