ABSTRACT
This chapter discusses how to model, interpolate, and forecast a vector of time series sampled at different frequencies. To avoid cumbersome wording, it refers to the low-frequency variables as “annual” and to the high-frequency variables as “quarterly.” The results, however, are valid for any combination of sampling frequencies. The chapter studies how aggregation over time affects the dynamic structure of a vector of time series and its observability. It shows that the basic dynamic components remain unchanged but some of them, in particular those related to seasonality, become unobservable. Building on these results, the chapter presents a structured specification method based on the idea that the models relating the variables in high and low sampling frequencies should be mutually consistent.
