Essays on macroeconometrics and short-term forecasting

The thesis, entitled "Essays on macroeconometrics and short-term forecasting",is composed of three chapters. The first two chapters are on nowcasting,a topic that has received an increasing attention both among practitioners andthe academics especially in conjunction and in the aftermath of the 2008-2009economic crisis. At the heart of the two chapters is the idea of exploiting theinformation from data published at a higher frequency for obtaining early estimatesof the macroeconomic variable of interest. The models used to computethe nowcasts are dynamic models conceived for handling in an efficient waythe characteristics of the data used in a real-time context, like the fact that due to the different frequencies and the non-synchronicity of the releasesthe time series have in general missing data at the end of the sample. Whilethe first chapter uses a small model like a VAR for nowcasting Italian GDP,the second one makes use of a dynamic factor model, more suitable to handlemedium-large data sets, for providing early estimates of the employment inthe euro area. The third chapter develops a topic only marginally touchedby the second chapter, i.e. the estimation of dynamic factor models on data characterized by block-structures.The firrst chapter assesses the accuracy of the Italian GDP nowcasts basedon a small information set consisting of GDP itself, the industrial productionindex and the Economic Sentiment Indicator. The task is carried out by usingreal-time vintages of data in an out-of-sample exercise over rolling windowsof data. Beside using real-time data, the real-time setting of the exercise isalso guaranteed by updating the nowcasts according to the historical release calendar. The model used to compute the nowcasts is a mixed-frequency VectorAutoregressive (VAR) model, cast in state-space form and estimated bymaximum likelihood. The results show that the model can provide quite accurateearly estimates of the Italian GDP growth rates not only with respectto a naive benchmark but also with respect to a bridge model based on thesame information set and a mixed-frequency VAR with only GDP and the industrial production index.The chapter also analyzes with some attention the role of the Economic SentimentIndicator, and of soft information in general. The comparison of ourmixed-frequency VAR with one with only GDP and the industrial productionindex clearly shows that using soft information helps obtaining more accurateearly estimates. Evidence is also found that the advantage from using softinformation goes beyond its timeliness.In the second chapter we focus on nowcasting the quarterly national accountemployment of the euro area making use of both country-specific andarea wide information. The relevance of anticipating Eurostat estimates ofemployment rests on the fact that, despite it represents an important macroeconomicvariable, euro area employment is measured at a relatively low frequency(quarterly) and published with a considerable delay (approximatelytwo months and a half). Obtaining an early estimate of this variable is possiblethanks to the fact that several Member States publish employment data andemployment-related statistics in advance with respect to the Eurostat releaseof the euro area employment. Data availability represents, nevertheless, amajor limit as country-level time series are in general non homogeneous, havedifferent starting periods and, in some cases, are very short. We construct adata set of monthly and quarterly time series consisting of both aggregate andcountry-level data on Quarterly National Account employment, employmentexpectations from business surveys and Labour Force Survey employment andunemployment. In order to perform a real time out-of-sample exercise simulatingthe (pseudo) real-time availability of the data, we construct an artificialcalendar of data releases based on the effective calendar observed during the first quarter of 2012. The model used to compute the nowcasts is a dynamicfactor model allowing for mixed-frequency data, missing data at the beginningof the sample and ragged edges typical of non synchronous data releases. Ourresults show that using country-specific information as soon as it is availableallows to obtain reasonably accurate estimates of the employment of the euroarea about fifteen days before the end of the quarter.We also look at the nowcasts of employment of the four largest MemberStates. We find that (with the exception of France) augmenting the dynamicfactor model with country-specific factors provides better results than thoseobtained with the model without country-specific factors.The third chapter of the thesis deals with dynamic factor models on datacharacterized by local cross-correlation due to the presence of block-structures.The latter is modeled by introducing block-specific factors, i.e. factors thatare specific to blocks of time series. We propose an algorithm to estimate the model by (quasi) maximum likelihood and use it to run Monte Carlosimulations to evaluate the effects of modeling or not the block-structure onthe estimates of common factors. We find two main results: first, that in finite samples modeling the block-structure, beside being interesting per se, can helpreducing the model miss-specification and getting more accurate estimatesof the common factors; second, that imposing a wrong block-structure orimposing a block-structure when it is not present does not have negativeeffects on the estimates of the common factors. These two results allow usto conclude that it is always recommendable to model the block-structureespecially if the characteristics of the data suggest that there is one.Doctorat en Sciences économiques et de gestion