Search results

First and higher order digits in data sets of natural and socio-economic processes often follow a distribution called Benford's law. This phenomenon has been used in many business and scientific applications, especially in fraud detection for financial data. In this paper, we analyse whether Benford's law holds in economic research and forecasting. First, we examine the distribution of leading digits of regression coefficients and standard errors in research papers, published in Empirica and Applied Economics Letters. Second, we analyse forecasts of GDP growth and CPI inflation in Germany, published in Consensus Forecasts. There are two main findings: The relative frequencies of the first and second digits in economic research are broadly consistent with Benford's law. In sharp contrast, the second digits of Consensus Forecasts exhibit a massive excess of zeros and fives, raising doubts on their information content.

Main description: Wer die richtigen Lehren aus der Vergangenheit gezogen hat, dem wird vor der Zukunft nicht bange sein. Dieser Erkenntnis folgt die in diesem Buch entwickelte vorhersageorientierte Modellierungstechnik. Zu Beginn werden wesentliche Aspekte der Vergangenheitsentwicklung mit Hilfe von Zeitreihenmodellen transparent gemacht, um damit anschließend Prognose-Experimente am aktuellen Rand der Zeitreihe durchzuführen. Ausgewählte wird am Ende ein adäquates Modell optimaler Kompliziertheit mit überlegenen Prognoseeigenschaften. Der Leser kann den mehrstufeigen Entstehungsprozess von Punkt- und Intervallprognosen schrittweise nachvollziehen und gegebenenfalls korrigierend eingreifen. Im praktisch wichtigen Bereich der Kurzfristprognosen wird großer Wert auf einfach strukturierte Modelle mit niedrigem Korrekturaufwand und hoher Flexibilität gelegt. Der Autor arbeitet seit mehr als 30 Jahren forschend und lehrend auf dem Gebiet der angewandten Zeitreihenanalyse und Prognoserechnung und hat an der Fachhochschule Stralsund verschiedene Studienprogramme zur Wirtschaftsinformatik entwickelt und betreut. Zusatzmaterial für Studierende bietet der Autor auf seiner Website http://goetze.fh-stralsund.de an.

This paper considers factor forecasting with national versus factor forecasting with international data. We forecast German GDP based on a large set of about 500 time series, consisting of German data as well as data from Euro-area and G7 countries. For factor estimation, we consider standard principal components as well as variable preselection prior to factor estimation using targeted predictors following Bai and Ng [Forecasting economic time series using targeted predictors, Journal of Econometrics 146 (2008), 304-317]. The results are as follows: Forecasting without data preselection favours the use of German data only, and no additional information content can be extracted from international data. However, when using targeted predictors for variable selection, international data generally improves the forecastability of German GDP. -- forecasting ; factor models ; international data ; variable selection

We look at how large international datasets can improve forecasts of national activity. We use the case of New Zealand, an archetypal small open economy. We apply data-rich factor and shrinkage methods to tackle the problem of efficiently handling hundreds of predictor data series from many countries. The methods covered are principal components, targeted predictors, weighted principal components, partial least squares, elastic net and ridge regression. Using these methods, we assess the marginal predictive content of international data for New Zealand GDP growth. We find that exploiting a large number of international predictors can improve forecasts of our target variable, compared to more traditional models based on small datasets. This is in spite of New Zealand survey data capturing a substantial proportion of the predictive information in the international data. The largest forecasting accuracy gains from including international predictors are at longer forecast horizons. The forecasting performance achievable with the data-rich methods differs widely, with shrinkage methods and partial least squares performing best. We also assess the type of international data that contains the most predictive information for New Zealand growth over our sample. -- Forecasting ; factor models ; shrinkage methods ; principal components ; targeted predictors ; weighted principal components ; partial least squares ; ridge regression ; elastic net ; international business cycles

In this paper, we put DSGE forecasts in competition with factor forecasts. We focus on these two models since they represent nicely the two opposing forecasting philosophies. The DSGE model on the one hand has a strong theoretical economic background; the factor model on the other hand is mainly data-driven. We show that by incooperating large information set using factor analysis can indeed improve the short horizon predictive ability, as claimed by manyresearchers. The micro founded DSGE model can provide reasonable forecasts for inflation, especially with growing forecast horizons. To a certain extent, our results are consistent with the prevailling view that simple time series models should be used in short-horizon forecasting and structural models should be used in long-horizon forecasting. Our paper compareds both state-of-the art data-driven and theory-based modelling in a rigorous manner.