Abstract not found

In this paper we compare the relative efficiency of different methods of forecasting the aggregate of spatially correlated variables. Small sample simulations confirm the asymptotic result that improved forecasting performance can be obtained by imposing aprioriconstraints on the amount of spatial correlation in the system. One way to do so is to aggregate forecasts from a Space-Time Autoregressive model (Cliff et al., 1975), which offers a solution to the ‘curse of dimensionality ’ that arises when forecasting with VARs. We also show that ignoring spatial correlation, even when it is weak, leads to highly inaccurate forecasts. Finally, if the system satisfies a ‘poolability ’ condition, there is a benefit in forecasting the aggregate variable directly.

In this paper we carry a disaggregated study of the monthly US Consumer Price Index (CPI). We consider a breakdown of US CPI in four subindexes, corresponding to four groups of markets: energy, food, rest of commodities and rest of services. This is seen as a relevant way to increase information in forecasting US CPI because the supplies and demands in those markets have very different characteristics. Consumer prices in the last three components show I(2) behavior, while the energy subindex shows a lower order of integration, but with segmentation in the growth rate. Even restricting the analysis to the series that show the same order of integration, the trending behavior of prices in these markets can be very different. An I(2) cointegration analysis on the mentioned last three components shows that there are several sources of nonstationarity in the US CPI components. A common trend analysis based on dynamic factor models confirms these results. The different trending behavior in the market prices suggests that theories for price determinations could differ through markets. In this context, disaggregation could help to improve forecasting accuracy. To show that this conjecture is valid for the non-energy US CPI, we have performed a forecasting exercise of each component, computed afterwards the aggregated value of the non energy US CPI and compared it with the forecasts obtained directly from a model for the aggregate. The improvement in one year ahead forecasts with the disaggregated approach is more than 20%, where the root mean squared error is employed as a measure of forecasting performance.

disaggregate forecasts or combining disaggregate information to forecast an aggregate

JEL classification: C13,C22,C53Hierarchical forecasts for Australian domestic tourism Abstract: In this paper we explore the hierarchical nature of tourism demand time series and produce short-term forecasts for Australian domestic tourism. The data and forecasts are organized in a hierarchy based on disaggregating the data for different geographical regions and for different purposes of travel. We consider five approaches to hierarchical forecasting: two variations of the top-down approach, the bottom-up method, a newly proposed top-down approach where top-level forecasts are disaggregated according to forecasted proportions of lower level series, and a recently proposed optimal combination approach. Our forecast performance evaluation shows that the top-down approach based on forecast proportions and the optimal combination method perform best for the tourism hierarchies we consider. By applying these methods, we produce detailed forecasts for the Australian domestic tourism market.

publications feature a motif taken from the €100 banknote. NOTE: This Working Paper should not be reported as representing the views of the European Central Bank (ECB). The views expressed are those of the authors and do not necessarily reflect those of the ECB. This paper can be downloaded without charge from