|
on Forecasting |
| By: | Rafal Weron |
| Abstract: | A variety of methods and ideas have been tried for electricity price forecasting (EPF) over the last 15 years, with varying degrees of success. This review article aims at explaining the complexity of available solutions, their strengths and weaknesses, and the opportunities and treats that the forecasting tools offer or that may be encountered. The paper also looks ahead and speculates on the directions EPF will or should take in the next decade or so. In particular, it postulates the need for objective comparative EPF studies involving (i) the same datasets, (ii) the same robust error evaluation procedures and (iii) statistical testing of the significance of the outperformance of one model by another. |
| Keywords: | Electricity price forecasting, Day-ahead market, Seasonality, Autoregression, Neural network, Factor model, Forecasts combination, Probabilistic forecast |
| JEL: | C22 C24 C38 C53 Q47 |
| Date: | 2014–05–12 |
| URL: | http://d.repec.org/n?u=RePEc:wuu:wpaper:hsc1407&r=for |
| By: | Oscar Claveria (Faculty of Economics, University of Barcelona); Enric Monte (Department of Signal Theory and Communications, Polytechnic University of Catalunya); Salvador Torra (Faculty of Economics, University of Barcelona) |
| Abstract: | This study compares the performance of different Artificial Neural Networks models for tourist demand forecasting in a multiple-output framework. We test the forecasting accuracy of three different types of architectures: a multi-layer perceptron network, a radial basis function network and an Elman neural network. We use official statistical data of inbound international tourism demand to Catalonia (Spain) from 2001 to 2012. By means of cointegration analysis we find that growth rates of tourist arrivals from all different countries share a common stochastic trend, which leads us to apply a multivariate out-of-sample forecasting comparison. When comparing the forecasting accuracy of the different techniques for each visitor market and for different forecasting horizons, we find that radial basis function models outperform multi-layer perceptron and Elman networks. We repeat the experiment assuming different topologies regarding the number of lags used for concatenation so as to evaluate the effect of the memory on the forecasting results, and we find no significant differences when additional lags are incorporated. These results reveal the suitability of hybrid models such as radial basis functions that combine supervised and unsupervised learning for economic forecasting with seasonal data. |
| Keywords: | forecasting; tourism demand; cointegration; multiple-output; artificial neural networks. JEL classification: L83; C53; C45; R11 |
| Date: | 2014–05 |
| URL: | http://d.repec.org/n?u=RePEc:ira:wpaper:201417&r=for |
| By: | Barbara Rossi; Tatevik Sekhposyan |
| Abstract: | We propose new methods for evaluating predictive densities that focus on the models’ actual predictive ability in finite samples. The tests offer a simple way of evaluating the correct specification of predictive densities, either parametric or non-parametric. The results indicate that our tests are well sized and have good power in detecting mis-specification in predictive densities. An empirical application to the Survey of Professional Forecasters and a baseline Dynamic Stochastic General Equilibrium model shows the usefulness of our methodology. |
| Keywords: | predictive density, dynamic mis-specification, forecast evaluation |
| JEL: | C22 C52 C53 |
| Date: | 2014–01 |
| URL: | http://d.repec.org/n?u=RePEc:bge:wpaper:758&r=for |
| By: | Jared Bullen (Treasury, Government of Australia); Jacinta Greenwell (Treasury, Government of Australia); Michael Kouparitsas (Treasury, Government of Australia); David Muller (Treasury, Government of Australia); John O’Leary (Treasury, Government of Australia); Rhett Wilcox (Treasury, Government of Australia) |
| Abstract: | Treasury’s forecasting framework has evolved over the past 21 years from the outlook for a single financial year to the outlook for the Australian economy 40 years ahead for intergenerational analysis. A constant through this evolution has been the sharp distinction between the methodologies used for near and longer term forecasts. The economic estimates underlying Australian Government fiscal projections divide the forecast horizon into two distinct periods: the near term forecast period which covers the first two years beyond the current financial year; and the longer term projection period which includes the last two years of the forward estimates, and up to 36 more years for intergenerational analysis. The economic estimates over the forecast period are based on a range of short run forecasting methodologies, while those over the projection period are based on medium to long run rules. Treasury routinely assesses medium to long run projection rules in light of new data, improved modelling techniques and structural changes to the economy. The measured cyclical weakness of recent years calls for an enhancement to the existing trend growth rate rules, which recognises the need for an adjustment period over which the economy transitions from a cyclical high or low to its potential level of output. Working towards that end, this paper details changes to the projection methodology that overcome the cyclical limitations of the existing framework. Applying these methodological changes to the economic estimates in the 2014 15 Budget leads to a slight improvement in the Underlying Cash Balance of $0.9 billion (0.05 per cent of GDP) in 2017 18 and $3.4 billion (0.12 per cent of GDP) in 2024 25. |
| Keywords: | potential output, NAIRU, fiscal budget |
| JEL: | E30 E66 H68 |
| Date: | 2014–05 |
| URL: | http://d.repec.org/n?u=RePEc:tsy:wpaper:wpaper_tsy_wp_2014_2&r=for |
| By: | Chen, Cathy W.S.; Gerlach, Richard |
| Abstract: | Intra-day sources of data have proven effective for dynamic volatility and tail risk estimation. Expected shortfall is a tail risk measure, that is now recommended by the Basel Committee, involving a conditional expectation that can be semi-parametrically estimated via an asymmetric sum of squares function. The conditional autoregressive expectile class of model, used to indirectly model expected shortfall, is generalised to incorporate information on the intra-day range. An asymmetric Gaussian density model error formulation allows a likelihood to be developed that leads to semiparametric estimation and forecasts of expectiles, and subsequently of expected shortfall. Adaptive Markov chain Monte Carlo sampling schemes are employed for estimation, while their performance is assessed via a simulation study. The proposed models compare favourably with a large range of competitors in an empirical study forecasting seven financial return series over a ten year perio d. |
| Keywords: | Semi-parametric; Markov chain Monte Carlo method; Expected; Asymmetric Gaussian distribution; Nonlinear; CARE model |
| Date: | 2014–04 |
| URL: | http://d.repec.org/n?u=RePEc:syb:wpbsba:2123/10457&r=for |
| By: | Jared Bullen (Treasury, Government of Australia); Michael Kouparitsas (Treasury, Government of Australia); Michal Krolikowski (Treasury, Government of Australia) |
| Abstract: | Australia’s terms of trade rose significantly over the eight years to 2011 12 following a period of relative constancy over the preceding 40 years. Australian Government fiscal projections from the 2010 11 Budget to the 2013 14 Budget, assumed that beyond the near term forecast period the terms of trade would fall by 20 per cent over the subsequent 15 years. This approach was silent on when the expected decline would end and the level at which the terms of trade would eventually settle. This paper details the projection methodology underlying the terms of trade projection assumption in the 2013 14 MYEFO. In contrast to the earlier approach it provides guidance on the timing of the end of the current expected decline and the associated long run level of the terms of trade. The centrepiece of the new approach is detailed price and volume forecasting modules for Australia’s major export categories, including global demand and supply models for the three major bulk commodities (iron ore, metallurgical coal and thermal coal). Based on this methodology, Australia’s terms of trade are expected to fall at a more rapid rate than previously predicted in the 2013 14 Budget from 2012 13 to 2017 18 and remain reasonably constant thereafter at a level roughly equal to that recorded in 2006 07. As noted in the 2013 14 MYEFO, there are a number of downside risks to this outlook including uncertainty around the global economy, the nominal exchange rate and non-bulk commodity price forecasts. Applying prudent judgement to the model’s outcome results in a long-run terms of trade that settles at the level observed in 2005 06 by 2019 20. |
| Keywords: | Confidence commodity prices, production cost, mining boom |
| JEL: | Q00 F17 E37 |
| Date: | 2014–05 |
| URL: | http://d.repec.org/n?u=RePEc:tsy:wpaper:wpaper_tsy_wp_2014_1&r=for |
| By: | Rickard Nyman; Paul Ormerod |
| Abstract: | We describe an exercise of using Big Data to predict the Michigan Consumer Sentiment Index, a widely used indicator of the state of confidence in the US economy. We carry out the exercise from a pure ex ante perspective. We use the methodology of algorithmic text analysis of an archive of brokers' reports over the period June 2010 through June 2013. The search is directed by the social-psychological theory of agent behaviour, namely conviction narrative theory. We compare one month ahead forecasts generated this way over a 15 month period with the forecasts reported for the consensus predictions of Wall Street economists. The former give much more accurate predictions, getting the direction of change correct on 12 of the 15 occasions compared to only 7 for the consensus predictions. We show that the approach retains significant predictive power even over a four month ahead horizon. |
| Date: | 2014–05 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1405.5695&r=for |