Smoothing and forecasting mixed-frequency time series with vector exponential smoothing models

Abstract

The analysis of mixed-frequency (MF) time series has been limited mainly to the vector autoregressive integrated moving average (ARIMA) framework, even though the exponential smoothing (ETS) method—a competing model to ARIMA—has made considerable progress in recent years. The ETS method provides a useful multivariate time series specification for estimating missing observations of low-frequency variable(s) and constructing forecasts of future values. Hence, this study proposes the vector ETS (VETS) method as a suitable alternative to ARIMA for smoothing and forecasting MF time series. To illustrate the superiority of the VETS method, we obtain high-frequency smoothed estimates of low-frequency variables and forecasts of MF vector time series using US data on four monthly coincident indicators and quarterly real gross domestic product. Furthermore, the method's forecast accuracy is investigated through a Monte Carlo simulation. The results show that the proposed method is suitable for short and medium-term forecasting.