ABSTRACT
In this chapter we discuss the multivariate analysis of historical time series that typically are subject to messy features such as missing observations and outlying observations. Our primary motivation is the multiple analysis of monthly commodity prices in Babylonia between 385–61 BC. We consider monthly price series for barley, dates, mustard, cress, sesame and wool. As can be expected, the data set is far from complete. The monthly time series spans over 300 years and hence we could have 3, 888 observations for each price series. However for most series we only have around 530 monthly observations available. This means that the vast majority of data entries are missing, and we need to treat 3, 358 missing observations in the analysis.
