|
on Forecasting |
| By: | Jose Ramon Cancelo; Antoni Espasa; Rosemarie Grafe |
| Abstract: | This paper discusses the building process and models used by Red Eléctrica de España (REE), the Spanish system operator, in short-term electricity load forecasting. REE's forecasting system consists of one daily model and 24 hourly models with a common structure. There are two types of forecasts of special interest to REE, several days ahead predictions for daily data and one day ahead hourly forecasts. Accordingly, forecast accuracy is assessed in terms of their errors. For doing so we analyze historical, real time forecasting errors for daily and hourly data for the year 2006, and report forecasting performance by day of the week, time of the year and type of day. Other aspects of the prediction problem, like the influence of the errors in predicting temperature on forecasting the load several days ahead, or the need for an adequate treatment of special days, are also investigated. |
| Keywords: | Energy forecasting, Hourly and daily models, Time series, Forecasting practice |
| Date: | 2007–12 |
| URL: | http://d.repec.org/n?u=RePEc:cte:wsrepe:ws078418&r=for |
| By: | Bruno Eklund |
| Abstract: | This paper considers the modelling and forecasting of the Icelandic business cycle. The method of selecting monthly variables, coincident and leading, that mimic the cyclical behavior of the quarterly GDP is described. The general business cycle is then modelled by a vector autoregressive, VAR, model. The cyclical behavior of the business cycle is summarized by a composite coincident index, which is based on the root mean squared forecast error over a pseudo out of sample. By applying a bootstrap forecasting procedure, using the estimated VAR model, point and interval forecasts of the composite coincident index are estimated. |
| Date: | 2007–09 |
| URL: | http://d.repec.org/n?u=RePEc:ice:wpaper:wp36&r=for |
| By: | Jana Eklund; Sune Karlsson |
| Abstract: | The problem of having to select a small subset of predictors from a large number of useful variables can be circumvented nowadays in forecasting. One possibility is to efficiently and systematically evaluate all predictors and almost all possible models that these predictors in combination can give rise to. The idea of combining forecasts from various indicator models by using Bayesian model averaging is explored, and compared to diffusion indexes, another method using large number of predictors to forecast. In addition forecasts based on the median model are considered. |
| Date: | 2007–05 |
| URL: | http://d.repec.org/n?u=RePEc:ice:wpaper:wp34&r=for |
| By: | Jonas Dovern; Christina Ziegler |
| Abstract: | In this paper we analyze the power of various indicators to predict growth rates of aggregate production using real-time data. In addition, we assess their ability to predict turning points of the economy. We consider four groups of indicators: survey data, composite indicators, real economic indicators, and financial data. Almost all indicators are found to improve short-run growth forecasts whereas the results for four-quarter-ahead growth forecasts and the prediction of recession probabilities in general are mixed. We can confirm the result that an indicator suited to improve growth forecasts does not necessarily help to produce more accurate recession forecasts. Only composite leading indicators perform generally well in both forecasting exercises. |
| Keywords: | leading indicators, forecasting, recessions |
| JEL: | C25 C32 E32 E37 |
| Date: | 2008–01 |
| URL: | http://d.repec.org/n?u=RePEc:kie:kieliw:1397&r=for |
| By: | Shiyi Chen; Kiho Jeong; Wolfgang Härdle |
| Abstract: | In recent years, support vector regression (SVR), a novel neural network (NN) technique, has been successfully used for financial forecasting. This paper deals with the application of SVR in volatility forecasting. Based on a recurrent SVR, a GARCH method is proposed and is compared with a moving average (MA), a recurrent NN and a parametric GACH in terms of their ability to forecast financial markets volatility. The real data in this study uses British Pound-US Dollar (GBP) daily exchange rates from July 2, 2003 to June 30, 2005 and New York Stock Exchange (NYSE) daily composite index from July 3, 2003 to June 30, 2005. The experiment shows that, under both varying and fixed forecasting schemes, the SVR-based GARCH outperforms the MA, the recurrent NN and the parametric GARCH based on the criteria of mean absolute error (MAE) and directional accuracy (DA). No structured way being available to choose the free parameters of SVR, the sensitivity of performance is also examined to the free parameters. |
| Keywords: | recurrent support vector regression, GARCH model, volatility forecasting |
| JEL: | C45 C53 G32 |
| Date: | 2008–01 |
| URL: | http://d.repec.org/n?u=RePEc:hum:wpaper:sfb649dp2008-014&r=for |
| By: | Philip Hans Franses; Henk Kranendonk; Debby Lanser |
| Abstract: | Official forecasts of international institutions are never purely model-based. Preliminary results of models are adjusted with expert opinions. What is the impact of these adjustments for the forecasts? Are they necessary to get ‘optimal’ forecasts? When model-based forecasts are adjusted by experts, the loss function of these forecasts is not a mean squared error loss function. In fact, the overall loss function is unknown. To examine the quality of these forecasts, one can rely on the tests for forecast optimality under unknown loss function as developed in Patton and Timmermann (2007). We apply one of these tests to ten variables for which we have model-based forecasts and expert-adjusted forecasts, all generated by the Netherlands Bureau for Economic Policy Analysis (CPB). For almost all variables the added expertise yields better forecasts in terms of fit. In terms of optimality, the effect of adjustments for the forecasts is limited, because for most variables the assumption that the forecast are not optimal can be rejected for both the model-based and the expert-adjusted forecasts. |
| Keywords: | Expert-Adjusted Forecasts; Optimality |
| JEL: | C53 E17 |
| Date: | 2007–12 |
| URL: | http://d.repec.org/n?u=RePEc:cpb:discus:92&r=for |
| By: | Michael Artis; José G. Clavel; Mathias Hoffmann; Dilip Nachane |
| Abstract: | Strongly periodic series occur frequently in many disciplines. This paper reviews one specific approach to analyzing such series viz. the harmonic regression approach. In this paper the five major methods suggested under this approach are critically reviewed and compared, and their empirical potential highlighted via two applications. The out-of-sample forecast comparisons are made using the Superior Predictive Ability test, which specifically guards against the perils of data snooping. Certain tentative conclusions are drawn regarding the relative forecasting ability of the different methods. |
| Keywords: | Mixed spectrum, autoregressive methods, eigenvalue methods, dynamic harmonic regression, data snooping: multiple forecast comparisons |
| JEL: | C22 C52 C53 |
| Date: | 2007–09 |
| URL: | http://d.repec.org/n?u=RePEc:zur:iewwpx:333&r=for |
| By: | Silvestro Di Sanzo (Department of Economics, University Of Alicante) |
| Abstract: | Recent studies have showed that it is troublesome, in practice, to distinguish between long memory and nonlinear processes. Therefore, it is of obvious interest to try to capture both features of long memory and non-linearity into a single time series model to be able to assess their relative importance. In this paper we put forward such a model, where we combine the features of long memory and Markov nonlinearity. A Markov Chain Monte Carlo algorithm is proposed to estimate the model and evaluate its forecasting performance using Bayesian predictive densities. The resulting forecasts are a significant improvement over those obtained by the linear long memory and Markov switching models. |
| Keywords: | Markov-Switching models, Bootstrap, Gibbs Sampling |
| JEL: | C11 C15 C22 |
| Date: | 2007 |
| URL: | http://d.repec.org/n?u=RePEc:ven:wpaper:2007_03&r=for |
| By: | Dominique Guegan (CES - Centre d'économie de la Sorbonne - CNRS : UMR8174 - Université Panthéon-Sorbonne - Paris I, Ecole d'économie de Paris - Paris School of Economics - Université Panthéon-Sorbonne - Paris I) |
| Abstract: | The detection of chaotic behaviors in commodities, stock markets and weather data is usually complicated by large noise perturbation inherent to the underlying system. It is well known, that predictions, from pure deterministic chaotic systems can be accurate mainly in the short term. Thus, it will be important to be able to reconstruct in a robust way the attractor in which evolves the data, if this attractor exists. In chaotic theory, the deconvolution methods have been largely studied and there exist different approaches which are competitive and complementary. In this work, we apply two methods : the singular value method and the wavelet approach. This last one has not been investigated a lot of filtering chaotic systems. Using very large Monte Carlo simulations, we show the ability of this last deconvolution method. Then, we use the de-noised data set to do forecast, and we discuss deeply the possibility to do long term forecasts with chaotic systems. |
| Keywords: | Deconvolution, chaos, SVD, state space method, wavelets method. |
| Date: | 2008–01 |
| URL: | http://d.repec.org/n?u=RePEc:hal:papers:halshs-00235448_v1&r=for |
| By: | Nwaobi, Godwin |
| Abstract: | Indeed, the specification of equilibrium in the world economy depends on the exchange rate regime and thus, the early contributions to the postwar literature on exchange rate economics are to a large extent concerened with the role of speculation in foreign exchange markets. However, the world has known several exchange rate systems beginning with the fixed-gold standard, the adjustable-peg system, adjustable-parity system and the flexible exchange rate system. Yet, in 1997, when foreign exchange was deregulated, independent traders finally had access to the biggest trading market of the world; and these forex traders attempt to make money from the simultaneous buying and selling of foreign currencies. And within the forex market, many types of instruments can be used:futures market,spot market, and forward market.However, the degree of volatility tends to increase with the frequency with which observations are sampled and this can be seen clearly as one moves from monthly to daily observations on exchange rates. Thus the basic thrust of the paper is to analyse the forecasting accuracy of the full vector autoregressive(FVAR), mixed vector autoregressive(MVAR) and Bayesian vector autoregressive(BVAR) models of the selected currency pairs(based on the monetary/asset model of exchange rate determination). |
| Keywords: | exchange rate; foreign exchange; forex; forecasting; vector autoregression; regimes; volatility; world;future markets; spotmarket;futures; options; assets; portfolio balance; brettonwood; IMF; Fixed rate; Floating rate; adjustable peg; purchasing power parity(PPP); Uncovered interest rate parity(UIP); internal balance; external balance; devaluation; overvaluation; pips; currency pairs; trading platform; forex allocation; parallel(black) market; banks; brokers; misalignment |
| JEL: | F00 F37 G1 C53 E42 G15 E44 F31 |
| Date: | 2008–02–01 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:6958&r=for |