|
on Forecasting |
| By: | Gerhard Rünstler (WIFO) |
| Abstract: | Forecasts from dynamic factor models potentially benefit from refining the data set by eliminating uninformative series. The paper proposes to use forecast weights as provided by the factor model itself for this purpose. Monte Carlo simulations and an empirical application to forecasting euro area, German, and French GDP growth from unbalanced monthly data suggest that both forecast weights and least angle regressions result in improved forecasts. Overall, forecast weights provide yet more robust results. |
| Date: | 2010–07–13 |
| URL: | http://d.repec.org/n?u=RePEc:wfo:wpaper:y:2010:i:376&r=for |
| By: | Michael McAleer (Erasmus University Rotterdam, Tinbergen Institute, The Netherlands, and Institute of Economic Research, Kyoto University); Les Oxley (Department of Economics and Finance, University of Canterbury) |
| Abstract: | Time series data affect many aspects of our lives. This paper highlights ten things we should all know about time series, namely: a good working knowledge of econometrics and statistics, an awareness of measurement errors, testing for zero frequency, seasonal and periodic unit roots, analysing fractionally integrated and long memory processes, estimating VARFIMA models, using and interpreting cointegrating models carefully, choosing sensibly among univariate conditional, stochastic and realized volatility models, not confusing thresholds, asymmetry and leverage, not underestimating the complexity of multivariate volatility models, and thinking carefully about forecasting models and expertise. |
| Keywords: | Unit roots, fractional integration, long memory, VARFIMA, cointegration, volatility, thresholds, asymmetry, leverage, forecasting models and expertise. |
| JEL: | C22 C32 |
| Date: | 2010–08 |
| URL: | http://d.repec.org/n?u=RePEc:kyo:wpaper:710&r=for |
| By: | Marcus Scheiblecker (WIFO) |
| Abstract: | For nowcasting and short-term forecasting of industrial production and GDP, business surveys are a vital source of information. They cover information of the recent past as well as developments in the near future. Whereas variations in industrial production indices potentially cover weather conditions as well as variations due to the different number of work days, it is unclear to which extent business surveys mirror them as well. Ignoring such information can lead to model misspecifications if used for nowcasting or forecasting. This paper sheds light on the effects of temperature changes as well as the varying number of work days on business survey results and on the production index of the Austrian construction industry. We find that survey data do not contain sufficiently the effects of the different number of work days necessary for explaining variations in industrial production of the construction sector. No statistical evidence was found that changing temperatures beyond their typical seasonal pattern influence the survey results and production. |
| Date: | 2010–07–12 |
| URL: | http://d.repec.org/n?u=RePEc:wfo:wpaper:y:2010:i:374&r=for |
| By: | Jesus Felipe; Utsav Kumar; Arnelyn Abdon |
| Abstract: | We forecast average annual GDP growth for 147 countries for 2010-30. We use a cross-country regression model where the long-run fundamentals are determined by countries’ accumulated capabilities and the capacity to undergo structural transformation. |
| Keywords: | Capabilities; Forecast; Growth |
| JEL: | C53 |
| Date: | 2010–08 |
| URL: | http://d.repec.org/n?u=RePEc:lev:wrkpap:wp_609&r=for |
| By: | Lundtofte, Frederik (Department of Economics, Lund University); Leoni, Patrick (EUROMED Management) |
| Abstract: | We analyze how a benevolent government agency would optimally release information about the growth rate of the stochastic dividend process of the financial market. We investigate the effects of the agency's signal on the agents' optimal strategies and equilibrium asset prices. In the case where all investors are rational Bayesian updaters, we show that the agency's optimal choice is to release a manipulative signal (lie) with probability one. However, if there are some nonupdating (inattentive) agents, we find cases where it is optimal for the government agency to send a revealing signal with probability one. |
| Keywords: | Social welfare; information; forecasting; asset pricing; inattention |
| JEL: | D80 G11 G12 |
| Date: | 2010–07–31 |
| URL: | http://d.repec.org/n?u=RePEc:hhs:lunewp:2010_008&r=for |