nep-for New Economics Papers
on Forecasting
Issue of 2005‒11‒09
eight papers chosen by
Rob J Hyndman
Monash University

  1. Real-Time or Current Vintage: Does the Type of Data Matter for Forecasting and Model Selection? By Hui Feng
  2. Model Selection, Forecasting and Monthly Seasonality of Hotel Nights in Denmark By Nils Karl Sørensen
  3. Economic Forecasting for Large Russian Cities By Raymond Struyk; Douglas Wissoker; Ioulia Zaitseva
  4. Modelling cross-border transport: Three cases in Öresund By Tom Petersen
  5. Can fear beat hope? A story of GARCH-in-Mean-Level effects for Emerging Market Country Risks By Maurício Yoshinori Une; Marcelo Savino Portugal
  6. Artificial Neural Networks in Finance Modelling By Crescenzio GALLO
  7. Defining Scenarios through shift - share models. An Application to the regional employment By Matías Mayor Fernández; Ana Jesús López Menéndez; Rigoberto Pérez Suárez
  8. Spatial structure and prediction of Land use change considering development projects in urban promotion districts By Yoshitaka Kajita; Masaya Kawano; Tetsunobu Yoshitake; Hiroshi Tatsumi; Satoshi Toi

  1. By: Hui Feng (Department of Economics, University of Victoria)
    Abstract: In this paper we investigate the impact of data revisions on forecasting and model selection procedures. A linear ARMA model and nonlinear SETAR model are considered in this study. Two Canadian macroeconomic time series have been analyzed: the real-time monetary aggregate M3 (1977-2000), and residential mortgage credit (1975-1998). The forecasting method we use is multi-step-ahead non-adaptive forecasting.
    Keywords: Vintage Data, Real-time Data, Model Selection, SETAR Model, ARMA model, Forecasting
    JEL: C22 C53
    Date: 2005–08–24
  2. By: Nils Karl Sørensen
    Abstract: Foreigners’ demand for hotel nights in Denmark by nationality are examined using monthly time series covering 30years, and divided into 11 nationalities. Special attention is given to the role of seasonality. Three univariate seasonal presentations of non-stationary data with different characteristics are considered, a stochastic, deterministic, and an error correction mechanism (ECM) approach taking into account economic as well as climatic variables. Based on a presentation of different measures to evaluate the forecasting performance a model selection is under taken. It is found that the single variable presentations in most cases are superior to the ECM. On the other hand the ECM presentation provides a more detailed description of the evolution of inbound tourism. In many cases it is found that the climatic indicators have significant influence on tourism. With regard to the single variable models it is found that seasonality in general is of stochastic nature, but the deterministic presentation is in many cases superior in forecasting performance. Theme: Tourism, regional econometric modelling Key words: Tourism, seasonality, fore casting, climatic variables. JEL Classification: C32.
    Date: 2004–08
  3. By: Raymond Struyk; Douglas Wissoker; Ioulia Zaitseva
    Abstract: The Budget Code of the Russian Federation requires that local self-governments prepare their budgets for the next year taking into account the likely economic situation in that year. To date these governments have had little guidance to use in preparing their budgets. This paper reports the results of initial steps to develop a procedure for forecasting key economic parameters at the local level. “Local level” is defined as cities that are capitals of Subjects of the Federation (similar to U.S. states); generally these are cities of over 100,000 population. Econometric models are reported for employment, manufacturing production, retail sales, average wage rates, volume of newly constructed housing, and fixed capital formation. The choice of estimation procedures was significantly constrained by data availability. The current document is an interim report, prepared after the basic econometric work has been completed but before the model is tested in actual forecasting. The paper consists of six further sections. The first lists the economic variables to be projected. The second describes the economic logic underlying the models specified for each variable. The third section then outlines the econometric strategy. This is followed in the fourth section with an overview of the data employed in the estimates. The fifth section presents the final models. The paper closes with a short discussion of the plans for future work in this direction. In the next phase of the work the forecasting qualities of these models will be evaluated.
    Date: 2004–08
  4. By: Tom Petersen
    Abstract: Transport models, and transport/land-use interaction models, are important decision support tools for large-scale infrastructure investments, for example in the road network. A bothersome feature of these tools are their distant forecasting horizon of 10--30 years ahead, and the uncertainty following from this. Although widely used, these models have rarely been put to test after this period have passed. The predictive power of a transport model is dependent on its ability to reproduce reality, which is assessed by validation. Apart from modelling the specific transport demand, which is based partly on socio-economic (demand) factors and partly on the supply of transport facilities (infrastructure), a number of scenarios of the future socio-economic development must be set up, called the scenario assumptions. In this paper we will present three different transport models: FREDRIK/ SSV, COMVIN and SAMPERS/ Skåne, out of which the first two have been used to model the transport across the Öresund Strait. The model structure, forecast results and scenario assumptions are considered in order to identify the key sources of uncertainties, and to prepare for the estimation of the true model error versus the error caused by incorrect scenario assumptions. Key words: infrastructure planning, transport models, validation, Öresund, before-and-after data.
    Date: 2004–08
  5. By: Maurício Yoshinori Une (Banco Itaú S.A.); Marcelo Savino Portugal (PPGE/UFRGS)
    Abstract: Upon winning the 2002 presidential elections, event that considerably increased the Brazilian country risk levels and volatility, Lula celebrated by declaring: “hope has beaten fear”. Extending Une and Portugal (2004), the aim of this paper is twofold: to empirically test the interrelations between country risk conditional mean (“hope”) and conditional variance (“fear”) and cast light on the role of country risk stability in the conduction of macroeconomic policies in developing small open economies. We compare the forecasting performance of various alternative GARCH-in-Mean-Level models for n-step conditional volatility point forecasts of the Brazilian country risk estimated for the period May 1994 - February 2005. The results support the idea that both hope and fear play important roles in the Brazilian case and confirms that hope and fear act in the same direction.
    Keywords: nonlinear GARCH, GARCH-in-Mean-Level effect, country risk, fear of disruption, forecast performance
    JEL: C22 F47 G14
    Date: 2005–09–04
  6. By: Crescenzio GALLO (Università di Foggia-Dipartimento di Scienze Economiche, Matematiche e Statistiche)
    Abstract: The study of Artificial Neural Networks derives from first trials to translate in mathematical models the principles of biological “processing”. An Artificial Neural Network deals with generating, in the fastest times, an implicit and predictive model of the evolution of a system. In particular, it derives from experience its ability to be able to recognize some behaviours or situations and to “suggest” how to take them into account. This work illustrates an approach to the use of Artificial Neural Networks for Financial Modelling; we aim to explore the structural differences (and implications) between one- and multi- agent and population models. In one-population models, ANNs are involved as forecasting devices with wealth-maximizing agents (in which agents make decisions so as to achieve an utility maximization following non- linear models to do forecasting), while in multi-population models agents do not follow predetermined rules, but tend to create their own behavioural rules as market data are collected. In particular, it is important to analyze diversities between one-agent and one-population models; in fact, in building one-population model it is possible to illustrate the market equilibrium endogenously, which is not possible in one-agent model where all the environmental characteristics are taken as given and beyond the control of the single agent. A particular application we aim to study is the one regarding “customer profiling”, in which (based on personal and direct relationships) the “buying” behaviour of each customer can be defined, making use of behavioural inference models such as the ones offered by Artificial Neural Networks much better than traditional statistical methodologies.
    Keywords: Artificial Neural Network, Financial Modelling, Customer Profiling
    JEL: C9
    Date: 2005–09–07
  7. By: Matías Mayor Fernández; Ana Jesús López Menéndez; Rigoberto Pérez Suárez
    Abstract: The analysis of different economic situations and risk factors is necessary in order to properly define forecasting scenarios. In this paper we focus on the shift-share model as a useful tool in the definition of economic scenarios, based on the different components that contribute to the change of a given economic magnitude (the so called national, sectoral and competitive effects). Although the most commonly used methodology is based on the “constant shift” and the “constant share” hypotheses, additional options can be considered based on the expected behaviour of the competitive effect, thus leading to more realistic scenarios. Once these new options are developed, this approach is applied to the definition of scenarios for the future evolution of the regional employment.
    Date: 2004–08
  8. By: Yoshitaka Kajita; Masaya Kawano; Tetsunobu Yoshitake; Hiroshi Tatsumi; Satoshi Toi
    Abstract: n this paper, the actual conditions and the change structure of land use by using mesh data are studied in urban promotion area in a local hub city of Fukuoka, Japan. Firstly, all meshes are classified into 15 patterns based on distribution of land use. Then, transition probability models are made out based on the change of these 15 patterns. The Change structure of land use in an area depends on whether development projects are carried out or not. Therefore, all of the meshes are divided into two groups, and different transition probability models are proposed. Finally, a prediction method of land use is proposed under the consideration of the changing structure of meshes. Though our proposed approach is a macroscopic forecasting method of land use, it is useful to evaluate the effects of urban policies for development projects.
    Date: 2004–08

This nep-for issue is ©2005 by Rob J Hyndman. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.