Search results

1

Unobserved Component Model with Observed Cycle Use of BTS Data for Short-Term Forecasting of Industrial Production

100%

Dudek S., Pachucki D.

Prace i Materiały Instytutu Rozwoju Gospodarczego SGH

|

2011

|

vol. 86:Business Surveys, Business Cycles. Polish Contribution to the 30th CIRET Conference

83-100

EN

In the paper we are checking the explanatory power of business tendency survey data (BTS) in short-term forecasts of industrial production within the framework of the unobserved component model (UCM). It is assumed that the "unobserved cyclical component" is common for reference quantitative variable and qualitative variable. In that sense the cyclical fluctuation of industrial production can be approximated by the fluctuations of BTS indicators. We call such a specification of structural time series model the “Unobserved component model with observed cycle" (UCM-OC). To estimate the system we are using the Kalman filter technique. Then we compare the model recursive one-period ahead forecasts to the historical path of the reference series to check its out-of-sample data fit. The forecasting properties are also evaluated against alternative models, i.e. "pure" UCM and ARIMA model. The analysis was performed for Poland and selected European Union countries.

2

Econometric modelling of compound cyclicality of using telecommunication services

100%

Kaczmarczyk P.

Econometrics. Ekonometria. Advances in Applied Data Analytics

|

2021

|

vol. 25

|

issue 2

27-45

XX

The aim of this study was to test the effectiveness of an econometric model with dichotomous (binary) explanatory variables in the approximation and prediction of different kinds of cyclicality (annual, monthly, weekly, and daily) of demand for telephone services. The analyses were conducted with the use of data provided by the selected telecommunication network operator. The data included hourly combined demand for specified telephone services (in seconds) of outgoing calls within the framework of the particular subscriber group (business or individual), the given day, the particular month, and the specific category of connection. All kinds of the cyclicality were confirmed in models with 70 explanatory variables (i.e. in models without holidays). The inclusion of the variables set denoting specific holidays improved goodness of the models fit to the data. The econometric modelling of cyclical components and the forecasting of it with the use of dichotomous variables was effective.

3

Short-term forecasting and composite indicators construction with help of dynamic factor models handling mixed frequencies data with ragged edges

63%

Łupiński M.

Przegląd Statystyczny

|

2012

|

vol. 59

|

issue 1

48-73

EN

This paper aims at presenting practical applications of latent variable extraction method based on second generation dynamic factor models, which use modified Kalman Filter and Maximum Likehood Method and can be applied for time series with mixed frequencies (mainly monthly and quarterly) and unbalanced beginning and the end of the data sample (ragged edges). These applications embrace short-term forecasting of Polish GDP and construction of composite coincident indicator of economic activity in Poland. Presented approach adopts the idea of short-term forecasting used by Camacho and Perez-Quirioz in Banco de Espana and concept of Arouba, Diebold and Scotti index compiled in the FRB of Philadelphia. According to the author's knowledge, it is the first such adaptation for Central and Eastern Europe country. Quality of the forecast obtained with these models is compared with standard methods used for short-term forecasting with series of statistical tests in the pseudo real-time forecasting exercise. Moreover described method is applied for construction of composite coincident indicator of economic activity in Polish economy. This newly-created coincident indicator is compared with first generation coincident indicator, based on standard dynamic factor model (Stock and Watson) approach, which has been computed by the author for Polish economy since 2006.

PL

Celem niniejszego artykułu jest prezentacja praktycznych aspektów estymacji nieobserwowalnych zmiennych za pomocą modeli czynnikowych drugiej generacji opartych o zmodyfikowany filtr Kalmana i metodę największej wiarygodności (MNW). Opisane w pracy modele wykorzystywane są do krótkookresowego prognozowania polskiego PKB i konstrukcji równoczesnego wskaźnika aktywności ekonomicznej w Polsce na podstawie szeregów czasowych o mieszanych i częstotliwościach, posiadających niezbilansowany koniec próby. Przedstawione podejście stanowi adaptację koncepcji modeli używanych do krótkookresowego prognozowania przez Camacho oraz Pereza-Quirioza w Narodowym Banku Hiszpanii oraz struktur analitycznych stosowanych przez Aruobę, Diebolda i Scotti’ego przy kompilacji indeksu aktywności ekonomicznej na potrzeby Banku Rezerwy Federalnej w Filadelfii. Zgodnie z wiedzą posiadaną przez autora jest to pierwsza tego rodzaju adaptacja w Europie Środkowo-Wschodniej. W ramach wykonanych badań za pomocą grupy testów statystycznych porównania została jakość prognoz uzyskanych za pośrednictwem zaproponowanych modeli z wynikami uzyskanymi metodami standardowymi. Ponadto ocenione zostały własności statystyczne obliczonego równoczesnego wskaźnika aktywności ekonomicznej w Polsce z wersją tego wskaźnika estymowaną regularnie z wykorzystaniem modeli czynnikowych pierwszej generacji (podejście Stocka-Watsona) począwszy od 2006 r.

Refine search results

1 Econometrics. Ekonometria. Advances in Applied Data Analytics

1 Prace i Materiały Instytutu Rozwoju Gospodarczego SGH

1 Przegląd Statystyczny

1 Dudek S.

1 Kaczmarczyk P.

1 Pachucki D.

1 Łupiński M.

1 2021

1 2012

1 2011

Unobserved Component Model with Observed Cycle Use of BTS Data for Short-Term Forecasting of Industrial Production

Econometric modelling of compound cyclicality of using telecommunication services

Short-term forecasting and composite indicators construction with help of dynamic factor models handling mixed frequencies data with ragged edges